1996 TMS ANNUAL MEETING AND EXHIBITION

TECHNICAL PROGRAM THE 1996 TMS ANNUAL MEETING AND EXHIBITION

February 4-8, 1996 Anaheim Convention Center Anaheim, California

MONDAY AM ADVANCES IN THE SCIENCE AND TECHNOLOGY OF TITANIUM ALLOY PROCESSING·I: Hot Working I

Sponsored by: SMD Titanium Committee and MDMD Shaping and Forming Committee Program Organizers: Prof. Isaac Weiss and Prof. Ragbavan Srinivasan, Mechanical and Materials Engineering Dept., Wright State University, Dayton, OH 45435; Dr. Paul Bania, TImet Corporation, TImet-Henderson Technical Laboratory, P.O. Box 2128, Henderson, NV 89009; Prof. Daniel Eylon, Graduate Materials Engineering, University of Dayton, Dayton, OH 45409 Monday, AM February 5,1996

Room:B5-6 Location: Anaheim Convention Center

Session Chairpersons: G. W. Kuhlman, Alcoa Forged Products, 1600 Harvard Ave., Cleveland, OH 44105, and R. Srinivasan, Mechanical and Materials Engineering Dept., Wright State University, Dayton, OH 45435 8:30 am Invited HOT-WORKING OF TITANIUM ALLOYS - AN OVERVIEW: S.L. Semlat1n, Wright Laboratory, Materials Directorate, WUMLLN, Wright-Patterson Air Force Base, OH 45433-7817 1be thermomechanical processing of ingot metallurgy titanium alloys will be discussed with special emphasis on microstructure evolution and workability considerations for a1pba-beta, alpha-two titanium a1uminide, and gamma titanium aluminide alloys. The conversion of ingot structure to fine equiaxed wrought structure will be addressed. In this regard, the breakdown of lamellar microstructures, the occurrence of cavitation/wedge cracking ('strain induced porosity'), and the development of crystallographic texture will be described. Special methods to breakdown the difficult-to-work titanium aluminide alloys will also be summarized. 1bese methods include canned hot extrusion and canned conventional forging. Secondary processes such as bare and pack sheet rolling, superpiastic forming of sheet, and closed-die forging will be reviewed. The emergence of non-standard methods for microstructure control, e.g., forging of metastable microstructures, will also be summarized.

9:00am MICROSTRUCTURE DEVELOPMENT IN A TITANIUM ALLOY DURING HOT DEFORMATION PROCESSING: Gangshu Shen, David Furrer, Ladish Co., Inc., P.O. Box 8902, Cudahy WI 53110-8902 Methods for control and manipulation of titanium alloy microstructures bave been developed for many years by iterative trials. Variations in titanium heating, deformation and cooling practice result in greatly varied microstructural morphology and volume fraction for the various alloy pbases. Efforts are being undertaken to develop an analytical tool to predict the development of titanium alloy microstructures under heating, bot working, and cooling conditions. Results of this tool bave been successfully correlated to the actual forge shop applications. Furtberutilization of this technology will allow more rapid process development for forged titanium components with greater robustoess with respect to resultant microstructure. 9:20am ADVANCING THE STATE-OF-TllE-ART IN TITANIUM ALLOY CLOSEDDIE FORGINGS: FABRICATION OF TIlE F-18E/F SINGLE PIECE TI-6Al-4V BULKHEADS: G. W. Kuhlman, K.A. Rohde, Alcoa Forged Products, 1600 Harvard Ave., Cleveland, OH 44105 Very large closed-die titanium alloy forgings bave presented a significant challenge to forgers, due to the material's very high unit pressures when sulrtransus forged, difficulty in deformation processing for required TMP and pressure constraints on available beavy presses,even at 50,000 tons. Until recently, the Boeing 747 Ti-6-4 Main Landing Gear Beam (4,000 sq.in. PVA) and the McDonnell-DouglasF-15 TI-6-4 Lower Bulkheads (3,200 sq.in. PV A) represented the state-of-the-art in closed-die forging size producible, particularly for the a1pba-beta alloy TI-6-4. Recently boweverdriven by the development of the new McAirlNortbrop F-18EIF fighter, the state-of-the-art in Plan View Area oflarge closed-die Ti forgings bas been extended by nearly 25%. The F-18Fl F aircraft included a series oftbree very large (PV A of 4,800 to 5,000 sq.in.) single piece Ti-6-4 bulkheads manufactured with incremental forging techniques. Fabrication of these parts as single forgings eliminated joints and reduced part fabrication costs and flow time. Presented is a summary of the fabrication approaches and mecbanical properties achieved.

1996 TMS ANNUAL MEETING AND EXHIBITION

FOUNDED 1871

TIRt:1IS

Minerals' Metals • Materials

9:40am RESEARCH ISSUES IN TIlE HOT EXTRUSION OF TITANIUM ALLOYS: Rajiv Shivpuri, Dinesh Damodaran, The Ohio State University, Columbus, OH 43210; Tony Esposito, Plymouth Tube Co. Hot extrusion is a basic metal working process, but for materials like titanium alloys where experience with extrusion is insufficient, many tryouts are required to determine the operating conditions due to the complex relationships among the process variables. 1be flow stresses of titanium alloys are higher than those of steel and hence these alloys are difficult to extrude. In comparison with steel, the high temperature flow stress of titanium alloys increases very rapidly with increasing strain rate. Besides this, the bebavior of the glass lubricant used in hot extrusion is very sensitive to the surface temperature and the contact time between the hot billet and the tooling. Information about the interactions between the process variables is necessary to optimize the process, but it is difficult to obtain by experiment only. In this study the entire hot extrusion process is mathematically modeled including the induction heating, billet transfer, glass lubrication and metal flow. The objective of the study is to investigate the interactions between the major process variables involved in bot extrusion of the titanium alloy TI6AI-4V. 1be effects of ram speed, and billet initial temperature on theextrusion pressure and lubricant behavior are presented. 1be theoretical results are compared with experimental observations of titanium alloy extrusions done at an extrusion plant. 10:00 am INFLUENCE OF STARTING MICROSTRUCTURE ON THE HIGH TEMPERATURE PROCESSING OF CP TITANIUM ALLOY: S. Guillard, M. Thlrukkonda,P.K. Cbaudbury,ConcurrentTechnologies Corporation, 1450 Scalp Ave., Johnstown, PA 15904 Commercially pure (CP) titanium grades 2 and 3 have been tested in compression at temperatures and strain rates corresponding to those encountered in industrial hot working operations, i.e., temperatures from 500" to 725°C, and strain rates from 0.001 to 20 S·l. Most deformation conditions led to either inhomogeneous flow, shear banding, or cracking of the grade 2 material, whereas the grade 3 material deformed homogeneously under most conditions. In addition, the flow stresses of the grade 2 material were higher than those of the grade 3 material by approximately 200 to 700%, although grades 2 and 3 materials had very similar chemical compositions. 1bese differences were traced to the difference in starting microstructures, grade 2 exhibiting a Widmanstlitten structure in contrast to grade 3 exhibiting slightly elongated grains. Samples of the grade 3 material were heat treated to duplicate the microstructure of the grade 2 material, and mechanically tested under the same processing conditions. The results, including the localization characteristics and flow stresses, are compared to study the influence of starting microstructure on hot forming of CP titanium. It is suggested that to insure successful manufacturing of CP titanium alloys, formation of the Widmanstlitten structure should be prevented during prior processing. 10:20 am BRINGING TITANIUM INTO A COMMERCIALLY ACCEPTABLE REALM FOR GENERAL USE: M. Miller, Stealth Engineering and Technologies, Inc., 1489 Cedar St, Holt, MI 48842 Issues relating to bringing titanium alloys into a commercially acceptable realm will be discussed. The topiCS will include (1) Bringing raw material costs to a representative comparison with other currently used materials; (2) Maintaining proper structural morphology in forging; (3) Product opportunities in areas other than aerospace, sucb as medical, automotive high performance, and original equipment automotive equipment; and (4) General description of those operations involved in near net-sbape forging. 10:40 am BREAK 10:50 am ELEVATED TEMPERATURE DEFORMATION BEHAVIOR OF TI-6.8Mo4.SFe-l.SAl: I. Philippart, H. J. Rack, Materials Science and Engineering Program, Clemson University, Qemson, SC 29634-0921 1be high temperature deformation bebavior ofTIMETLCB(Ti-6.8Mo-4.5Fe-I.5Al) bas been investigated as a function of temperature, strain rate and strain. Stable and unstable flow regimes were establisbed by Dynamic Materials Modeling and correlated with microstructural restoration processes. Unstable flow was observed at low strain rates «10.2 S·l) independent of temperature, and was found to be related to grain boundary failure phenomena. Stable flow was observed at high strain rates being associated with dynamic recovery at low temperature and dynamic recystallization at higher temperatures. The observed flow behavior will be discussed considering the effect of deformation temperature, strain rate and strain on grain boundary sliding and dislocation processes. 11:10 am SUBMICRONSTRUCTUREDTlTANIUMALLOYSPREPAREDUSINGMULTlSTEP ROLLING + ANNEALING TECHNIQUE: P.E. Marlcovslcy, Institute for Metal Physics, National Academy of Sciences of the Ukraine, 36 Vemadsky str., 252142, Kiev, Ukraine The method of obtaining ultra-fine structural (UPS) state (with grain size less than

B1

0.5J1111) in two-phase titanium alloys from massive starting material (without stage of powder preparation) is discussed. This method is based on the results ofinvestigation of "undeveloped" recrystallization. The tensile testing of these UFS materials was carried out in the temperature range form 106O·C to 550·C. It was found that UFS materials have a specific deformation behavior and significantly improved mechanical properties (at low and room temperatures) and a superplastic-like behavior at elevated temperatures. 11:30 am MICROSTRUCTURAL EVOLUTION IN TI-6AI-4V DURING HOT DEFORMATION: M.B. Gartside. IRC in Materials for High Performance Applications. The University of Binningham. Edgbaston. Birmingham. BI5 2TT. United Kingdom Evolution of microstructure and its effect on mechanical behavior has been investigated in Ti-6AI-4V for different initial a and 8 morphologies. Hot tensile tests were performed in the temperature range 720-920·C and at strain rates of t04 to to-2S-I. The stress-strain behavior exhibited a maximum stress followed by a decrease in flow stress which was most pronounced at low temperatures. high strain rates and for the transformed 8 microstructures. Optical microscopy of the transformed 8 microstructures showed shearing and fragmentation of the a needles. particularly at high temperatures. leading to a softer microstructure of spherical a particles. In the as-received microstructure some rearrangement of the a and 8 grains was observed. Tests with large rapid changes in strain rate showed that strain rate has an effect on the evolution of microstructure. These results indicate that rearrangement of the microstructure is responsible. to some extent. for the development in flow stress seen at this alloy in high temperatures. 11:50 am SOLUTE SOFTENING OFa TITANIUM - HYDROGEN ALLOYS: O.N. Senkov. J J. Jonas. Department of Metallurgical Engineering. McGill University. 3450 University Street. Montreal, Quebec. Canada H3A 2A7 Compression tests were carried out on a series of titanium - hydrogen alloys within the alpha phase field. The dependence of the flow stress. rate of work hardening and mechanical anisotropy on temperature. strain rate and hydrogen content were determined. Anomalies in the rate and temperature dependencies of the flow stress and work hardening rate were detected in the unalloyed specimens. which indicated that dynamic strain aging was occurring. Hydrogen in solid solution shifts the ranges of the anomalies towards lower temperatures; it also decreases the work hardening rate. produces marked flow softening, and decreases the mechanical anisotropy of alpha titanium. The relative values of the hydrogen - induced softening are shown to depend on strain rate and temperature.

ALUMINUM REDUCTION TECHNOLOGY I: Fundamental Information Sponsored by: LMD Aluminum Committee Program Organizer: Ms Fiona J Stevens. Comalco Research and Technology. Comalco Research Centre. PO Box 316. Thomastown. Yictoria 3074. Australia

Monday. AM February 5.1996

Room:A9 Location: Anaheim Convention Center

Session Chairperson: Jomar Thonstad. Department of Electrochemistry, Norwegian Institute of Technology. University of Trondheim. N-7034 Trondheim. Norway 8:30am THE TRANSPORTED ENTROPIES OF IONS IN SOLID STATE FLUORIDES AND B"·ALUMlNA: V.S. Sharivker. Institute of Structural Macrokinetics. Russian Acadamy of Sciences. Chemogolovka. Moscow Region. Russia 142432; S.K. Ratkje. Department of Physical Chemistry, Norwegian Institute of Technology. University of Trondheim, N-7034. Trondheim. Norway Transported entropies of Na+ in solid state mixtures of NaP and Na3AIF6 are determined from thermocell experiments. The experiments were favourably described by the electric work method. The variation observed in the thermocell emf with composition is explained by the probable path of charge transfer in the electrolyte. The transported entropy ofsodium ions in cryolite is Sa<'YNa+= 140 ±7 J K·I mol-I between 380·C and 500·C. The Thomson coefficient in this temperature range is aery N.+ = 33 ± 10 JK·I mol· l . It is predicted that the transported entropy for Na+ in the liquid electrolyte probably is larger than 140 J KI mol-I. This makes us predict that the transported entropy for Na+ in the molten electrolyte mixture for aluminium production is substantial. and that the reversible heat effects in the aluminium electrolysis cell are the same. It implies that the reversible heat balances calculated for the electrodes in the aluminium electrolysis need to be revised. 9:00am PELTIER HEATS IN CRYOLITE MELTS WITH ALUMINA: Belinda E. Flem. Signe Kjelstrup Ratkje. Division of PhYSical Chemistry. Norwegian Institute of Technology. University of Trondheim. N-7034 Trondheim, Norway; Asmund Sterten. Division of Industrial Electrochemistry. Norwegian Institute of Technology, University of Trondheim. N-7034 Trondheim. Norway The Seebeck coefficient was measured for cells with molten mixtures of NaP and AIF3 saturated with AI20). The electrodes were either a pair of oxygen electrodes or a pair of aluminium electrodes. For the molar ratio NaF/AIF3 equal to 1.8. 1.2 and 1.0. we obtained the Seebeck coefficients -1.80 mY K·I at 971'C, -1.63 mY K· I at 8I3.6'C and -0.583 mY KI at 758'C. respectively. for the oxygen electrodes. For the aluminium electrodes, we obtained the Seebeck coefficient - 1.23 mY KI at 962"C. for NaP/AIF3 of 1.8. These suggest that there is a substantial reversible heat consumption at the anode during aluminium electrolysis and a large reversible heat production at the cathode. The

B2

highest temperature in the Hall-Heroult cell is then closer to the cathode than the anode. The transported entropies of AI) and 02- were calculated to 77 J mole-I KI and 1OJ moleI K- I• respectively. when NaP/AIF) was 1.0. 9:30am THE DISSIPATED ENERGY OF THE ELECTRODE SURFACES IN THE ALUMINIUM ELECTROLYSIS CELL: Signe KjelstrupRatkje.E/lenMarie Hansen. Belinda FIem, Department of Physical Chemistry. The Norwegian Institute of Techno1ogy, The University of Trondheim, 7034 Trondheim, Norway; Dick Bedeaux. Department of Physical and Macromolecular Chemistry, University of Leiden, 2300 Leiden. The Netherlands The theoretical expression for the dissipated energy, 1"0', of the electrode surfaces in the aluminium electrolysis cell is derived from nonequilibrium thermodynamics, considering polarizable fluids at an interface. Here a" is the excess entropy production rate at the surface and 1" is the surface temperature. It is shown that the energy dissipated at the electrode surfaces is a significant part of the total energy dissipated in the cell. Both vectorial and scalar contributions to the dissipated energy are presented. The vectorial contributions describe processes parallel to the surface. while the scalar contributions describe transport processes normal to the surface. Coupled transport of heat. mass. and charge is discussed in particular. The premises used in the derivation of the theoretical expressions are local thermodynamic equilibrium and electroneutraiity at the surface. Discontinuities in the intensive variables at the surfaces (e.g. temperature and chemical potential). are predicted and estimated. 10:00 am DEFINING BATH RATIO IN THE PRESENCE OF L1F AND MgFl: Warren Haupin, 2820 Seventh Street Road. Lower Burrell. PA 15068 When Lewis acids other than AIF) and Lewis bases other than NaP are present. the ratio of NaP to AIF) is no longer a good measure of Lewis acidity. The formula Req = 1/ 2{[(NaP)/42 + 0.32 (LiF)J25.4]/ [AIF) /84 + 0.41 (MgF2)/62.3J) produces nearly constant acidity of AIF4- and nearly constant vapor pressures. However. at constant Req. LiP lowers: liquidus temperatures, metal solubility, alumina solubility and bath density while raising electrical conduc:tivity. At constant Req. MgF2 lowers: liquidus temperature, alumina solubility and elr.ctrical conductivity but raises metal solubility, and bath density. 10:25 am BREAK 10:35 am ELECTRICAL CONDUCTIVITY MEASUREMENTS IN CRYOLITE ALU· MlNA MELTS IN THE PRESENCE OF ALUMINUM: G.M. Haarberg. J. Thonstad, Department of Electrochemistry, The Norwegian Institute of Technology. N-7034 Trondheim, Norway; JJ. Egan. Brookhaven National Laboratory. Upton. NY 11973; R. Oblakowski. S. Pietrzyk, Academy of Mining and Metallurgy. Krakow. Poland Ac resistance measurements were performed in cryolite alumina melts to study the effect of dissolved metal on electrical conductivity. The specific electrical conductivity was found to increase substantially in the presence of AI. A certain activity of sodium is established due to the interaction between AI and the melt. Dissolution ofN a gives rise to the formation of mobile electrons, which are responsible for the increase of the conductivity. The electronic conductivity was taken as the difference between the conductivity in melts equilibrated with AI and the ionic conductivity. In the pure cryolite alumina melt saturated with AI at IOOOC the electronic conductivity contributes to more than 20% of the total conductivity. Results are in good agreement with previously reported data using a polarization technique. Effects of temperature CaP2 and xs AIF) were studied. Model calculations showing the effect of electronic conduction on the current efficiency for producing AI were made. 11:05 am DENSITY,ELECTRICALCONDUCTIVITY AND VISCOSITY OFLOW MELT· ING BATHS FOR ALUMINIUM ELECTROLYSIS: M. Chrenkova. Y. Danek. A. Silny. Institute of Inorganic Chemistry. Slovak Academy of Sciences. SK-842 36 Bratislava. Slovakia. Department. of Metallurgy and Materials Science, T.A. Utigard, University of Toronto. MSS IA4, Toronto, Canada Density. electrical conductivity and viscosity of melts of the Na)AIF6 - AIF) - LiP AI~3 system have been measured. Empirical equations for the concentration and temperature dependencies of the~e parameters have been developed. The reliability of these equations is compared with those published by other authors. An equation for density. was derived; which is valid from 850-105OC. AIF) up to 30% andAh~up to the solubility limit. An equation for electrical conductivity was obtained; which is valid in the same ranges. An equation for viscosity (within the standards deviation of 1.83.10 -2 mPa.s) is given which is valid from 850 - 105OC, AlF) from 9-26% and LiP up to 7%. 11:35 am METAL SOLUBILITY IN LOW RATIO BATH: Qiu Zhuxian. Tie Jun. Yu Yaxin, Northeast University. Shenyang IlOOO6. China For the molten salt system of cryolite - excess AIF) - CaPrAI20), the metal solubility was measured by gravimetric method. The metal solubility in the molten salt was obtained by plotting the relationship curve of "Metal wt loss versus time" in a time duration of 6 hours. An increase in the excess AIF3 - content from 10% to 28% (mass). will cause the metal solubility to decrease significantly, for two super-heat temperatures of 15'C and 50'C above the corresponding liquidus temperatures. In a graphite crucible with BN-Iining the metal solubility may be decreased to 0.1% (mass). A comparison between the gravimetric method and volumetric method is discussed.

1996 TMS ANNUAL MEETING AND EXHIBITION

10:10 am BREAK

CAST SHOP TECHNOLOGY I: Equipment and Operations

Sponsored by: LMD Aluminum Committee Program Organizer: Julian V. Copenhaver, Technical/Quality Manager, NSA A Division of Southwire, PO Box 500, Hawesville, KY 42348 Monday, AM February 5,1996

Room: AIO Location: Anaheim Convention Center

Session Chairperson: Mr. Edmund J. Elder, Alumax of South Carolina, P. O. Box 1000, Goose Creek, SC 29445 8:30am COMPARATIVE SAMPLING MOLDS EVALUATION: Laura Pierrard, Research Engineer, Phillippe Jarry, Jean Charbonnier, Pechiney Centre de Recherches de Voreppe, BP 27,38340 Voreppe, France; Colette Rigaut, Aluminium Pechiney Division Alliages de Moulage, BP 7,38340 Voreppe, France The metallurgical industry needs to cast alloys with narrow tolerances in their chemical composition in order to reduce variability of their use properties. Therefore, appropriate sampling practices and analytical methods are required. Both accuracy and precision of the analytical results are limited by the non-homogeneity of as cast-disk or cylinder samples, which results from macrosegregation phenomenon. This paper presents a comparison hetween Pechiney and other commonly used molds (ASTM designation: center-pour molds type B and vacuum mold). Two complementary approaches are exhibited for the different molds designs: Solidification modelling in order to predict macrosegregation localization using the Simulor software; Experimental characterizations. Radial and axial segregation profiles are determined by Analytical Scanning Electron Microscopy as well as analytical precision evaluation by spark emission and xray fluorescence spectrometries for a given cutting depth. 8:50am DEVELOPMENT OF A SINGLE WHEEL, HIGH SPEED ALUMINUM STRIP CASTER: Lawrence D. Ray, Reynolds Metals Company, PO Box 27003, Richmond, VA 23261; George A. Sloan,Reynolds Metals Company, PO Box 1155, Hot Spring, AK 71901 The aluminum industry has been in search of a viable continuous casting process for

many years. Such a process would provide an increase in production as well as lower production costs. The melt drag process is capable of producing thin aluminum sheet at extremely high casting rates. Over the past ten years, many milestones have been reached in making this process acceptable for use in production. Some ofthese accomplishments are reviewed in this paper along with a general overview of some of the technical aspects of melt drag casting. In addition, a comparison is made with other processes available for continuously casting aluminum slab and strip. 9:10am PROCESS AUTOMATION FOR HlGH-SPEED-THIN-GAUGE-TWIN ROLL CASTING TECHNOLOGY: Klaus Peter Maiwald, Bruno Mariethoz, Lauener Engineering, Stationsstrasse 21B, CH-3645 Gwalt, Switzerland The successful Joint Venture development between Lauener-Engineering and Hydro-Aluminium in respect to thin gauge - high-speed - casting demanded a highly sophisticated Process-Control System. The targets of 100% (and above) Productivity increases and gauges of 2 mm and below within tolerances of ±I % and below where reached on time. The paper will explore details of the Process Control system applied from fumace to Caster and Coiler. 9:30am MOLTEN ALUMINIUM: RECENT ADVANCES IN WEIGHING AND TRANSPORTATION: Palmi Stefansson, Technical Manager, Hydro EquipmentAS, P. O. Box 80, N-1321 Stabekk, Norway; Ole Ingar Vee, General Manager, Procon als, Kronvegen 3B, N-1265 Oslo, Norway; Dr. Thorsteinn I. Sigfusson, Science Institute, Austorbrun 17, IS-I04 Reykjavik, Iceland Transportation of liquid metal is an important aspect of the efficiency of any aluminium smelter operation. The paper discusses a tapping method developed by Hydro Aluminium which has proven superior to the conventional crane or fotklift tapping of potroorn metal. The overall manning can be halved by the extensive rationalization of this method. The remote computer operated control of metal mass transfer by electronic scales mounted on the tapping vehicle is explained. As the system uses no skimming, the dross generation and sodium content of metal is shown with collected data. The impact and advantages of this novel engineering on the casthouse are finally explained and discussed in detail.

10:20 am BENEFITS OF MASTER ALLOY MELT TREATMENTS IN THE ALUMINIUM FOUNDRY INDUSTRY: P. S. Cooper, R. Cook, Dr. M. A. Keams, London & Scandinavian Metallurgical Co. limited, Fullerton Road, Rotherham, South Yorkshire, S60 IDL, England The general benefits of the master alloy approach to foundry melt treatment in terms of environmental and technical advantages are reported. New data on the environmental impact of master alloy treatments compared with other addition methods are provided. New application areas for melt treatment are described with reference to Sr in high pressure die castings. The results of a controlled trial on A380 alloy are presented and it is shown that Sr offers benefits in terms of the consistency of casting properties. Sr modifies the microstructure, particularly of thicker section castings and virtually removes gross porosity. The paper concludes with a field study showing the technical benefits ofTiBAi grain refinement compared with other methods of refining AI-Si sand castings. 10:40 am LOW-NO.OXY-FUEL FLAMES FOR UNIFORM HEAT TRANSFER: Dr. Lao Tjay Yap, The BOC Group Technical Center, 100 Mountain Avenue, Murray Hill, NJ 07974; Mohammed Pourkashanian, University of Leeds, Department of Fuel and Energy, Leeds LS2 9JT, UK Aame characteristics such as controllable heat-transfer rates and NO. formation rates have been optimized for a novel self-cooling oxy-fuel burner using CFD techniques. Contrary to traditional axisymmetric flames, here, luminous large aspect-ratio flames are established resulting in uniform heat transfer rates. Extreme flame stability allows a wide operating range with low noise emissions. Luminosity provides effective flame cooling as well as dampening of turbulent fluctuations resulting in reduced NO. formation rates. Computational modeling has been verified with physical probing as well as in-situ Laser Induced Auorescence (LIF) measurements of OH. Increased thermal efficiencies, improved productivities, unaffected metal yields, lower costs, and decreased NO. emissions are demonstrated in commercial aluminum furnaces. 11:00 am REFRACTORY SOLUTIONS DESIGNED TO OVERCOME CORUNDUM GROWTH IN ALUMINIUM FURNACES: Duncan Jones, Development & Technical Applications Manager, Morganite Thermal Ceramics Ltd., Liverpool Road, Neston, South Wirral L64 3RE, England Corundum is an extremely hard, high temperature form of aluminium oxide. Its formation in aluminium melting furnaces and metal treatment units can lead to a reduction in operating efficiency and premature failure of the furnace lining. The formation of corundum is known to be dependent on several factors, including furnace atmosphere, operating temperature, alloy composition, and refractory type. The paper describes solutions to this phenomenon, identifying its cause and effect whilst highlighting refractory design criteria. Results from an extensive testing programme, clasifying resistance to corundum growth against various qualities of refractory product, are discussed. 11:20 am

CORROSION OF REFRACTORIES AT THE BELLYBAND OF ALUMINUM HOLDING AND MELTING FURNACES: Sylvio Quesnel, S. Afshar, C. Allaire, CIREP-CRNF, Ecole Polytechnique, Metallurgy and Materials Engineering Department, 8475 Christophe-Colomb Road, Montreal, Quebec, Canada H2M 2N9 Different phenomena leading to the degradation of the refractory lining at the bellyband of aluminum holding and melting furnaces are described. The effect of temperature, alloy composition, fluxing agent, and surrounding atmosphere, on the corrosion of aluminosilicate refractories in such applications has been studied. A laboratory test that simulates the conditions at the bellyband in the above furnaces was developed. Results suggest that sidewall build-up originates from corundum formation at the surface of the molten metal, which then grows inside the refractory above the metal line. 11:40 am PRIMITIVE PRACTICES IN TODA Y'S CAST SHOPS (IS YOURS ONE OF THEM?): George J. Binczewski, S.C. Systems, PO Box 6154, Moraga, CA 94570 The continuing evolution of the aluminum industry cast shop has involved the introduction of many production operations which were quite innovative and considered important for their time. Too often, the original reasons for them have become obscured, and the result is the perpetuation of daily practices which have become inefficient, cost, and open to technological scrutiny and challenge. Current practices are identified and appropriate remedies are suggested concerning grain refining, filtration, furnace practices, skimming, casting, alloying, sampling, and quality control.

9:50am MELT FLOW IN SPOUTISTOPPERSYSTEMS AND ITS IMPACT ON METAL QUALITY IN ALUMINIUM CASTHOUSES: C. J. MorilZ,Dr. Wolfgang Schneider, H. P. Krug, V.A.W. AluminiumAG,ResearchandDevelopment, Georg-von-BoeselagerStrasse 25, 0-53117 Bonn, Germany In aluminium casthouses spout/stopper systems of different designs are in operation for regulation of the melt flow during transport of the melt for example, from a launder into a D.C. casting mOUld, from a launder to another launder, or from a launder into a casting ladle of an ingot casting carousel. Due to the different designs, variable and high melt flow rates in the system occur. This leads to the formation of Significant underpressures which cause turbulences, and as a result of this, a pollution of the melt with inclusions. The under-pressure courses in different spout/stopper systems will be described and discussed and their impact on inclusion content reported, investigated by LiMCA measurements.

1996 TMS ANNUAL MEETING AND EXHIBITION

B3

10:35 am

COMPOSITES FOR HIGH PERFORMANCE ELECTRONIC PACKAGING AND THERMAL MANAGEMENT APPLICATIONS

Sponsored by: MDMD Powder Metallurgy Committee and Jt. SMDIMSD Composites Materials Committee

Monday,AM February 5,1996

Room: Grand H Location: Anaheim Marriott Hotel

Session Chairperson: Michael J. Koczak, Drexel University, Philadelphia, PA 19104 8:30am

DESIGN AND FABRICATION OF COOLED GRiCU STRUCTURES: Daniell. Vrab/e, Fred L. Beavers, SPARTA, Inc., 9455 Towne Centre Dr., San Diego, CA 921211964 This presentation summarizes and discusses some of SPARTA's efforts in the design, development and fabrication of GrlCu structures for highly loaded thennallstructural applications. The applications include; leading edges and engine components for the National Aerospace Plane, radiators for space power, flexible high conductance thermal straps for space power and electronic SEM-E modules. The components for these applications have common design requirements of managing the transfer of high heat fluxes or minimizing component weight, tailoring the coefficient of thermal expansion to minimize thermal strains, and enhancing other structural attributes such as stiffness orstrength. Discussed is the methodology that is used in developing the Gr/Cu composite components. This methodology includes; the composite design, thennallstructural analysis, materials characterization and property determination, composite part producibility, proto-type development and testing, and full scale component fabrication and tesL Several examples are provided to explain and help illustrate the successful insertion of advanced GrlCu composite materials into high technology components. The discussion addresses the selection of the fiber type, fiber architecture, fiber volume fraction, enhancement of the short transverse thermal conductivity and various processing methods. 8:55am

COPPERIGRAPHITE BASED METAL MATRIX COMPOSITES FOR IMPROVED THERMAL MANAGEMENT OF ELECTRONIC DEVICES: Joseph K. Weeks, Jared L. Sommer, Technical Research Associates, Inc., 2257 S. 1100 E. Salt Lake City, UT 84106 As electronic components continue to shrink while they increase in capability, the need to remove waste heat is becoming significantly more difficult. TRA has developed a series of copper/graphite products with better properties than can be obtained in conventional materials. High conductivity graphite fibers are incorporated in a pure copper matrix througb a simple fiber coating and casting process. The resulting composites offer Significant improvements in thermal conductivity, lower weight, and lower thermal expansion than alternative materials. Since molten copper wets the coated fiber surface, superior bonding between the fibers and the copper is produced. Theefforts to produce economical composite materials through the utilization oflower cost graphite fibers will be discussed. Properties of the composites will be presented. New products and applications only made possible by the fiber coating will be discussed.

9:20am METAL MATRIX COMPOSITE RADIATOR PANELS FOR SPACECRAFT THERMAL MANAGEMENT: Uday KashaliluJr, Foster-Miller Inc., 350 Second Ave., Waltam, MA 02154 During this reaearch program, Foster-Miller demonstrated an innovative, effective and potentialJy low cost metal matrix composite (MMC) radiator panel for spacecraft thermal management. The component used ultrahigh thermal conductivity graphite fiber in an aluminum 6063 matrix. The light weight was achieved by using corrugated radiator panel design that provides flexural stiffness equivalent to a conventional "sandwich" panel without the weight penalty of a honeycomb core. Additionally, a housing for the heat pipe was cast in place, thereby eliminating the high cost joining step. An MMC radiator component was fabricated to net shape via pressure casting. The P 120 graphite fiber reinforced Al-6063 component showed a 16 to 30 percent lower thermal impedance than the baseline aluminum sandwicb component. Use of the KIlOO grapbite fiber is expected to produce over SO percent reduction in the radiator thermal impedance along with substantial reduction in panel weight. 9:55am IMPROVED THROUGH-THICKNESS THERMAL CONDUCTIVITY IN ADVANCED POLYMER COMPOSITES: Thomtls G. Campbell, Composites Technology Division, Foster-Miller, Inc., 350 Second Ave., Waltham, MA 02154 Advanced composite materials are used extensively in aircraft but have bad limited use in areas of thermal applications. Inplane conductivities can be greatly enhanced by using pitcb based carbon fibers. However, most areas of beat transfer require heat flow through the thickness of the laminate. This property is controlled by the epoxy wbicb has low conductivity. A special application of the Z-Fiber process has been developed that permits introduction of up to 10% of the surface area of higb conductivity pitcb fibers in the through-thickness direction. This enables carbon epoxy laminates to generate beat transfer capabilities approaching aluminum. A 3-D laminate is constructed from 20 materials generating improved properties without the loss of inplane properties generally found in 3D materials. The progress to date will be presented, including test data verifying the through-thickness conductivities of greater than 8 W/mK, compared to the I W/mK found in standard laminates, at minimal loss of inplane strength. 10:20 am BREAK

B4

CERMETS FOR MICROELECTRONICS PACKAGING: M.K. Premkumar, Alcoa Technical Center, 100 Technical Drive, Alcoa Center, PA 15069 The availability of suitable new materials for packaging is critical to the continuing advancement of microelectronics technology. As circuit densities and switching speeds continue to increase, and electronic devices are used in more bostile environments, there is an increasing need for packaging materials with superior properties. Traditional packaging materials such as Kovar that have been successfully used for decades are no longer capable of meeting dlC thermal and reliability requirements of the new designs. New classes of ceramics and composite materials are being developed to meet these cballenges and AI-SiC cermet is one sucb material system. These cermets are composites with near optimum combinations of thermal expansion coefficient, thermal conductivity, density and elastic modulus. This paper describes the attributes of this material system as well as Alcoa's manufacturing technology for production of cermet microelectronics packaging products. The unique capabilities of this technology will be illustrated via application examples. 11:00 am

CAST SILICON CARBIDE REINFORCED ALUMINUM MATRIX COMPOSITE FOR REMOTE POWER CONTROLLER (RPC) PACKAGES: Suraj RQWal, Tom Green, Bruce Lanning, Bruce Benson, Martin Marietta Tecbnologies, Inc., R & T, Mecbanical, P. O. Box 179, MS F 3085, Denver, CO Cast silicon carbide particulate reinforced aluminum matrix composites (SiCplAl) provide desired properties and characteristics for lightweight remote power controller packages. These composites with high volume percent (2 55 v/o) particulates offer excellent combination of low density, high strength, bigh conductivity, properties to provide performance improvements and weigbt savings compared to baseline kovar and copperltungsten packages. Using pressure assist infiltration process, both the baseplates and integrated electronic enclosures (i.e. baseplate with sidewalls) were successfully cast. In addition, a few SiCplAI baseplates were also cast by higb pressure infiltration process. Eacb of these components were electroplated with nickeUcopper and gold. The plated baseplates were used to assemble electronic enclosures by soldering copper or kovar sidewalls. These basepla.tes and electronic enclosures were extensively evaluated using optical and scanning electron microscopy, thermal-mecbanical tests and salt-fog tests, and thermal cycling tests. Subsequently, the beryllia substrate (with electronic components) was bonded to the composite baseplate, and performance tests were conducted to verify that each electronic component remained within the desired operational temperature range. This paper presents the results of materials characterization, and thermal performance tests and discusses the technology issues to be resolved to ensure metal matrix composite technology insertion into RPC packages. 11:25 am APPLICATIONS FOR ULTRA-mGH THERMAL CONDUCTIVITY GRAPH· ITE FIBERS: Thomas F. Fleming,C.D. Levan, Amoco Performance Products Inc., William C. Riley, Research Opportunities Inc. Thermal management has become a limiting problem for both military and commercial systems. Higb thermal conductivity graphite fibers provide a breakthrougb in the efficiency of conductive cooling. The fibers have exceedingly high specific modulus and a negative coefficient of thermal expansion; these properties can be used to tailor the thermal expansion of composite material to virtually any required level. This paper discusses the nature of fibers, the translation of fiber properties into composites including metal, organic, and carbon matrix. Emphasis is on applications for the resulting bigh thermal conducti\'ity composites. These applications include electronic packaging, aircraft and spacecraft. 11:50 am FABRICATION TECHNIQUES FOR ADVANCED CARBON-CARBON COMPOSITE HEAT EXCHANGERS: William H. Altergott, Patricia J. Scbwartz FosterMiller Inc., 350 Second Ave., Walbam, MA 02154 This paper describes various i1movative brazing techniques that address the thermal expansion mismatcb problems associated with the joining of carbon-carbon (ClC) composites to refractory metals and high thermal conductivity metals for advanced beat excbanger applications. Upon cooling from braze temperature, thermal expansion mismatcb between the ClC and the metal results in formation of deleterious residual stresses throughout the joint. Residual intedaminar stresses may lead to cracking of brittle refractory metals. Even barring material failure, the residual stresses may compromise the ability of the joint to withstand the mechanical and thermal stresses imposed during the service. Specific brazing innovations discussed in this paper include: 10 use of multiple interlayers in tbe joint where a compliant material accommodates thermal expansion mismatch stresses via plastic flow, wbile a stiff intedayer reacts shear stresses to off-load the ClC, 2) hybrid brazeldiffusion bonding process in whicb the resulting joint may withstand higher service temperatures than the braze temperatures, and 3) coatings in faying surfaces that prevent formation of brittle and unstable compounds at the braze interface, and promote chemicallthermal stability of the Joint during prolonged exposure to elevated service temperatures. We will describe potential applications for advanced ClC beat excbangers including bypersonic vehicle structures, advanced propulsion systems, spacecraft therlllal management and higb density electronics packaging. 12:15 pm POWDER INJECTION MOLDING OF THERMAL MANAGMENT MATERIALS FOR MICROELECTRONIC APPLICATIONS: R. M. Gennan, 118 Researcb West, Penn State University, University Park, PA 16802-6809 Thermal management materials create cballenges both with material selection and component fabrication. Powder injection molding provides an important fabrication route for near net shaping high quality composite systems with tailored thermal properties. Analysis bas identified several candidate materials, including W-Cu, Mo-Cu, and SiC-AI. These have been developed to a demonstration state using actual geometries and tested for thermal properties. W-Cu is the most viable material for beat sink and packaging applications, delivering thermal conductivities in the 230 WImf'C range. But,

1996 TMS ANNUAL MEETING AND EXHIBITION

major opportunities are apparent in systems based on AlN . This presentation outlines the powder injection molding of thermal management structures and isolates processing options and applications for these liquid phase sintered composites. Demonstration heat spreaders. boilers. and microwave packages are used to illustrate the state of technology.

ENGINEERING OF MECHANICALLY ALLOYED MICROSTRUCTURES

Sponsored by: MDMD Powder Materials Committee Program Program Organiurs: M.G. McKimpson. Institute of Materials Processing. Michigan Technological University. 1400 Townsend Drive. Houghton. MI 49931; S. AxteU. Dept. of Mechanical Engineering. University of Nebraska-Lincoln. 225 WSEC. Lincoln. NE 68588 Monday. AM February 5.1996

Room: Grand G Location: Anaheim Marriott Hotel

Session Co-Chairpersons: M.G.McKimpson. S. Axtell. Department of Mechanical Engineering and Center for Materials Researcb and Analysis. University of Nebraska. 255 WSEC. Lincoln. NE 685880656 8:30am GRAIN GROWTH IN NANOCRYSTALLINE MATERIALS PREPARED BY MECHANICAL ALLOYING: S.c. Axtell. Department of Mechanical Engineering and Center for Materials Researcb and AnalysiS. University of Nebraska. 255 WSEC. Lincoln. NE 685880656 Althougb commercial quantities of nanocrystalline powders can be produced by mechanical alloying. a consolidation step is required to produce usable parts and devices from these materials. Mecbanically alloyed powders are generally consolidated at temperatures where grain growth occurs; therefore an understanding of the grain growth behavior of nanocrystalline materials produced by this method is desirable. This talk will present an overview of studies of grain growth in mecbanically alloyed materials. The presentation will include. in part, a discussion of the wOlk being performed at the University of Nebraska on Cu and Cu-based alloys using in-situ and non in-situ x-ray diffraction techniques. Support for this wolk was provided by NSG EPSCoR grant OSR255225. 9:00am PHASE TRANSFORMATION & STRUCTURAL EVOLUTION OFMECHANICALLY ALLOYED Fe-Zo-AI PHASES: O. Uwalcweh, Z. Liu. R. Edwards. Department of Materials Science & Engineering. University of Cincinnati. Qncinnati. OH 45221-0021 Fixed composition ratios of Fe and Zn corresponding to Fe3ZnIO' Fe~21' FeZn7•and FeZn13• with the addition of 5% Al (wt) were mechanically alloyed. Nonisothermal kinetic analyses based on DSC measurements of the milled alloys revealed two diffusion controlled processes for the FeZn7 + 5% Al with activation energies of227 +/- 12kJ/mole and 159 +/-IIkJ/mole respectively. The formation of FeAI2 phase at 44O"C during the evolution ofFeZnI3 + 5% AI suggests the need for are-evaluation of the existing Fe-ZnAI equilibrium diagrams. The Fe3ZnlO + 5% Al and Fe~1 + 5- Al compositions evolved similarly except at 400"C where the former decomposed to Fe + FeZn7 + Fe~IO' and the latter to Fe + FeZn7. 9:20am INVESTIGATION OF THE MICROSTRUCTURE AND MAGNETIC PROPERTIES OF MECHANICALLY MILLED SmCo,IFe: R. L. Schalek. S.C. Axtell. Department of Mechanical Engineering. University of Nebraska. 225 WSEC. Lincoln. NE68S88-0656; Alice M. Milton and DiandraL. Leslie-Pelecky. DepartmentofPbysics and Astronomy. University of Nebraska. Lincoln. NE 68588-0656 Theoretical calculations indicate that hard magnets (SmCoS> can excbange couple across soft magnets (Fe) producing large coercivity materials exhibiting enhanced remanence and saturation magnetization. We have produced (SmCo,).:FeulO.•. with x 90. 70 and 50 vol.% via mechanical milling of smCo, and iron powders. X-ray diffraction. scanning electron microscopy. and magnetic bysteresis measurements are used to characterize the materials. To investigate the effects of the lamellar microstructure 011 the magnetic properties. as-received iron powder or iron premilled for 6 bours was milled with the smCo,. Powders containing the premilled iron have significantly different magnetic properties than powders conSisting of as-received iron. Furthermore. increasing the iron concentration decreases the coercivity and remanence. but increases saturation magnetization. Results of powders annealed at temperatures between the Curie temperatures of Fe and smCo, will also be presented. Support for this wOlk was provided by NSG EPSCoR grant OSR-9255225.

=

9:40am CHARACTERIZATION AND CONSOLIDATION OF CuAlNIMnTI SHAPE MEMORY ALLOY POWDER PREPARED BY MECHANICAL ALLOYING: W.G. Uu*. C.Y. Cbung. Dept ofPbysics & Materials Science. City University of Hong Rong. Tat Chee Avenue. Kowloon. Hong Kong. * also Dept of Materials Engineering. Dalian University of Technology. Dalian 116023. China The elemental powder mixtures ofCu-12Al-5Ni-2Mn-ITi(wt%)have been processed by mechanical IUoying (MA). cold pressing and sintering with the aim of fabricating a Cu-based high temperature shape memory alloy. The mechanically alloyed powders as well as consolidated products in compacted. sintered and quenched states have been characterized by X-ray diffraction. SEM. DSC and hardness testing. The MA process. whicb was carried out in a blgb energy ball mill for up to 4Ob. was found to be a typical one like that wbicb generally occurs in the MA of ductile/ductile systems. The final product of MA was a single pbase of fcc structure with lattice parameters close to that of Cu. The green densities of cold pressed compacts strongly depended on the MA time

1996 TMS ANNUAL MEETING AND EXHIBITION

with a maximum value after 10 bours milling. As the MA time increases. the size of pores in compacts formed during sintering reduced and the expansion wbicb occurred in sintering decreased. All the sintered compacts gave the same dual phase structure conSisting of and Y2phase regardless of the milling time. The dominant structure of the quenched compacts from powder milled for less than Ib was martensite with the same crystal structure as observed in conventional cast CuAINiMnTI alloy. With longer MA processing. the amount of martensite in quenched compacts reduced and little martensite was found when MA time exceeded 2Ob. The martensite demonstrated no reversible transformation during beating from 50-590'C 10:00 am BREAK 10:20 am CONSOLIDATION OF MECHANICALLY ALLOYED MATERIALS-AN OVERVIEW: M. G. McKimpson. Institute of Materials Processing. Michigan Technological University. 1400 Townsend Drive. Hougbton. Ml49931 Mechanical alloying is widely used to produce powders exhibiting a variety of novel microstructures and pbysical properties. including nanoscale structures. extended solid solutions and dispersion-strengthened alloys. Frequently there is a need to process these powders to nearly full density while retaining as mucb of the as-milled structure as possible. Achieving this goal often requires extended-and expensive-iterative process development efforts. In this overview presentation. phenomenological changes likely to be observed during consolidation processing of mechanically alloyed powders will be reviewed. and possible guidelines for minimizing consolidation process development efforts will be suggested. 10:40 am COMPACTION MICROSTRUCTURE OF MECHANICALLY PROCESSED IRON POWDER: I. Rawers. G. Slavens. Albany Researcb Center. U.S. Bureau of Mines. 1450 Queen Ave SW. Albany. OR 97321; J. Groza. University of Califomia Davis. Davis. CA Although mucb bas been written about high-energy milling of powders and the properties of these powders. there bas been limited information about consolidation of these materials. This study describes the microstructure of mecbanically processed iron and mechanically alloyed iron and 2 WI% carbon powder that have been compacted using two different techniques: conventional bot-pressing and a new technique plasma activated slntering. Compactcbaracterization will include: denSity. hardness. grain size. and microstructure as a functlon of processing time. temperature. and pressure. This study will sbow that fully dense materials can be produced wbile retaining the mechanically milled nanostructure. 11:00 am THE EFFECT OF PROCESSING HISTORY ON SECONDARY RECRYSTALLlZATlON IN MA754: Thomas R. Bieler. R. Saminathan. J. McDougall. Department of Materials Science and Mechanics. Michigan State University. East Lansing. MI; S.K. Mannan. INCO Alloys Intemational. Huntington. WV 25720 MA754 is a Ni-Cr based superalloy with relatively few additional alloying elements. Past wOlk bas sbown that a directional anneal is effective for obtaining elongated large grains in a secondary recrystaUization anneal at temperatures above O.9Tm. However. directional anneals are not convenient for sbeet metal structures that could be fabricated out of sheet material; a static anneal is desirable. Sheet forms of MA754 must be bot rolled. and the bistory of bot rolling conditions affects the potential for a large grain secondary recrystalization. The microstructures and textures were investigated for MA754 with various bot rolling histories. When it is bot rolled near 2OOO'F. partial secondary recrystallization occurs during rolling. resulting in cube texture. Subsequent secondary recrystaUization anneals do not result in large grain sizes. Wben it is bot rolled at temperatures near 1700'F. secondary growth is mucb less. and a diffuse FCC rolling texture is formed. This condition will provide large elongated grains in a static anneal above 0.9Tm, with a strong (110)<001> texture. Precipitation of carbides during bot rolling. and subsequent cold work before annealing strongly affect the kinetics of microstructural evolution. 11:20 am HOT ISOSTATIC PRESS CONSOLIDATION OF NANO-SCALE TITANIUM TRIALUMINIDE POWDERS: EA. Laitila. D.E. Mikkola. Department of Metallurgical and MaterialsEngineering. Michigan Technological University; M. G. McKimpson. Institute of Materials Processing. 1400 Townsend Drive. Hougbton. MI 49931 Reactive mechanical alloying bas been used to produce nanoscale (Al.CrhTI-TIC composites containing nominally 10 WI. % TIC. Powders milled in air or argon were consolidated in a hot isostatic press (HIP) at pressures of 414 MPa. temperatures from 6()()OC to l000'C. and times of either 15 or 120 minutes in an attempt to preserve the nanoscale microstructure of the powders and optimize the mechanical properties. Near full density was achieved by HIPing the powders at 800'C for 15 minutes. but all materials consolidated at 800'C sbowed limited resistance to cracking during hardness testing. Powders milled in argon and consolidated at 414 MPall000'C sbowed significant improvements in cracking resistance. and increasing consolidation time from 15 to 120 minutes increased the load required for cracking from 2kg up to 5kg. This increase occurred with only a sligbt decrease in hardness. The results of compression testing the consolidated material will also be reported. 11:40 am MECHANICAL ALLOYING FOR PREPARATION OF MAGNESIUM MATRIX DEFORMATION PROCESSED COMPOSITES: I.A. lensen. A.M. Russell. L.S. Chumbley. Ames Laboratory nd Iowa State University. T. W. Ellis. Ames Laboratory. U.S. Department of Energy. Ames. Iowa 50011 Althougb the conventional precipitation hardened alloys of Mg dominate today's engineering use of Mg. advantages in mechanical properties can be achieved by use of immiscible refractory metal phases as strenghtening particles in an Mg matrix. Since the low melting and boiling points of Mg preclude use of crdlnary melt processing of Mg with refractory metals. the authors have preparead Mg-20 vol. % Fe. Mg-20 vol. % Nb.

B5

and Mg-20 vol.% 11 composites by mechanical alloying of the pure metal powders in liquid nitrogen followed by deformation processing to increase the strength of the composites. The resulting composites possess ultimate tensile strengths of 300 MPa (44 ksi) at room temperature and show no change in microstructure and no loss of room temperature strength after exposure to 400°C (750°F) for 500 hours.

A SYMPOSIUM DEDICATED TO PROFESSOR JOHANNES WEERTMAN'S SEVENTIETH BIRTHDAY: SESSION I: High Temperature Deformations Sponsored by: JL SMDIMSD Nuclear Materials and MSD Flow and Fracture Committees and FEMS (Federation of European Materials SocIeties) Program Organizers: RJ. Arsenault, Deptartment of Materials Science and Nuclear Engineering, University of Maryland, College Park, MD 20742-2115; David Cole, CRREL, 72 Lyme Rd., Hanover, NH 03755; Todd Gross, Department of Mechanical Engineering, University of New Hampshire, Durham, NH 03824; Gernot Kostorz, Institut fUr Angewandte Physik, ETH Hllnggerberg, CH-8093 ZUrich, Switzerland; Peter Liaw, Department of Materials Science and Engineering, University of Tennessee, Knoxville, TN 37996-2200; Sivan Parameswaran, NRC-Institute for Aerospace Research, Ottawa, Canada KIA OR6; Howard Sizek, Inco Alloys International Inc., Huntington, WV 25705-1771 Monday, AM February 5,1996

Room: Orange County 3 Location: Anaheim Marriott Hotel

Session Chairpersons: H. Sizek, Mechanical Testing, Inco Alloys International Inc., Huntington, WV 25720-1958; Julia Weertman, Department of Materials Science and Engineering, Northwestern University 8:30am OPENING REMARKS: J.B. COHEN, NORTHWESTERN UNIVERSITY 8:35 am Invited mGH TEMPERATURE DEFORMATION IN ULTRAHIGH CARBON STEELS: O.D. Sherby, Department of Materials Science and Engineering, Stanford University, Stanford, CA 94305; O.A. Ruano, M. Carsi, CENIM, Avda. Gregorio del Amo, 8, Madrid, Spain; T. Oyama, WESGO Co., 477 Harbor Blvd., Belmont, CA 94002 Ultrahigh carbon steels (UHCSs) contain from I to 2.1 % carbon and have been shown to be superplastic at elevated temperature and strong, tough and tensile ductility at low temperature. The basis for achieving such properties is the ability to prepare these steel with fine ferrite grain sizes (Ito 5 11m) and fine spheroldized carbides. The elevated temperature mechanical properties of these steel will be described with special emphasis on their behavior at large strains utilizing torsion and tension tests. The influence of carbon content, grain size and subgrain size on the deformation characteristics are considered. The results are discussed on the basis of deformation mechanisms involving grain boundary sliding and diffusion-controlled dislocation creep, with consideration for the important influence of grain growth during plastic deformation. 9:00 am Invited CREEP BEHAVIOR OF INTERMETALLICS WITH SPECIAL REFERENCE TO NI~I: T.S. Rong, I.P. Jones, R.E. Smallman, School of Metallurgy and Materials, The University of Birmingham, Edgbaston, Birmingham, B15 2TT, United Kingdom Intermetallic compounds are currently showing promise as high temperature structural materials, particularly for airspace applications. Two areas of considerable concern, however, for intermetallics in general, are low temperature brittleness and high temperature creep resistance. In this paper, it is this latter aspect which is reviewed. Conventional strain-rate deformation of intermetallics has been extensively studied. By comparison, investigation of creep deformation has been limited to only a few intermetallics, and particularly Ni3AI, which is considered to be a basic compound for understanding intermetallics in general. The aim of the present paper is to outline the current understanding of creep of Ni3Al. Its creep properties will be summarized with respe~tto stress and temperature dependence, and the microstructures corresponding to the different stages and temperatures of creep outlined. The mechanisms which are thought to control creep deformation, especially at intermediate temperatures below the peak in the flow stress, will be discussed. General comments of creep behavior in intermetallic will then be made. 9:30am STRAIN AND STRAIN RATE BEHAVIOR OF A LOW CARBON 18Cr-12 NI (304L) STAINLESS STEEL UNDER CONDITIONS OF CREEP-LOW CYCLE FATIGUE INTERACTION: J. Cadek, Institute of Physics of Materials, Academy of Sciences of the Czech Republic, 616 62 Brno, Czech Republic Results of an investigation of the application of a low frequency "rectangular" cyclic stress, during both the initial and advance stages of primary creep, on the time dependent strain and strain rate behavior of a low carbon 18Cr-12Ni(304L) stainless steel are presented. The strain rates immediately before, as well as immediately after, any stress decrement and/or increment in any loading cycle are measured accurately. It is shown tha~atany given relative applied stress arnplitudeAo/o and superimposed applied stress pennd At.:, these strain rates depend linearly on time. The relations between these strain rates and the relative amplitude of stress cycling, Ao/o are also linear at any constant cyclic loading period At.:. This finding makes it possible in principle, to estimate the contribution of the net strain due to superimposed cyclic loading. Further, it is shown that the ~reep strain. rate as .measured after the cyclic loading superposition lasting}At. =36 kS,ls only relal1vely slightly affected by this superimposed loading. An interpretation of

B6

strain and strain rate behavior under conditions of cyclic loading superposition is based on such processes as creep strain strengthening and creep recovery, though the backward straining after any stress reduction must be taken into account as well. All the above mentioned processes depend not only on temperature and the "mean" applied stress, but especially on cyclic loading superposition history. 9:50am DYNAMICAL ASPECTS OF THE ANOMALOUS TEMPERATURE DEPENDENCE OF THE FLOW STRESS OFNI3AISINGLE CRYSTALS: G. Saada,LEM CNRS ONERA-BP72-F-9222, Chatillon, Cedex, France Ni3AI single crystals exhibit a mechanical behavior which is atypical in many respects: The flow stress increases by a factor larger than 5 between 300K and the peak temperature Tp aboutl000K; The schmid law is violated; The strain hardening rate is larger than 10.2 times the shear modulus 11; The strain rate sensitivity is small, comparable to that of fcc metals; Creep below the peak temperature consists of an exhaustion ~tage followed by inverted creep. The first stage is atypical (the creep rate rn:creases With temperature) while the inverted creep is typical (the creep rate increases With temperature). Below the peak temperature, glide occurs mostly in octahedral planes during constant strain rate tests and during the primary creep while glide occurs be cube glide during inverted creep. After reviewing the main physical aspects, we shall show that taking into account the multiplication and exhaustion processes explains semiquantitatively the experimental observations. 10:10 am THE EFFECTS OF MICROSTRUCTURAL CONTROL ON THE MECHANICAL BEHAVIOR OF Cr2Nb-BASED INTERMETALLIC ALLOYS: J.A. Cook, P.K. Liaw, Department of Materials Science and Engineering, The University of Tennessee, Knoxville, TN 37996-2200; C.T. Liu, Metals and Ceramics Division Oak Ridge National Laboratory, Oak Ridge, TN 37831-6155 ' Microstructural evaluations and mechanical testing of Laves-phases alloys based on Cr2Nb were examined in order to optimize microstructural and mechanical properties through secondary processing techniques and thenno-single-phase Cr2Nb alloys are very hard and brittle due to the complicated crystal structure(C-15). The following results were revealed through examination of the Cr-Cr2Nb two-phase region:(a) with increasing amounts of the soft chromium-rich phase, the compression strength and hardness decrease; (b) the annealing treatments studied provided the best break-up of the coarselbrillleeutectic structure in the 94 al.% Cr-6 al.% Nb(CN-7) alloy; (c) 2 days (d) at 1~~oC, lead to a substan~~1 improvement in the room-temperature compressive ductility over the as-cast condition and previous annealing treatments (3d at 1100°C). In addition, Hot Isostatic Pressing (illpping) + annealing led to a substantial refinement of the b~ttle Cr2Nb phase in ~eutectic structure in theCN-7 (Cr-6al.% Nb) composition. Hlppmg Ic:ct to only a n,wgmal refmement of the brittle, interconnected Cr2Nb phase in the eutectic structure 10 the CN-4(Cr-12 al.%) composition. A combination of hot forging and annealing is promising in refming the brittle Cr2Nb phase in the eutectic structure of the CN-7 composition. 10:30 am BREAK 10:40 am PHENOMENA AND MECHANISM OF HIGH TEMPERATURE CREEP OF SICIlI24 Al METAL MATRIX COMPOSITES: S.H. Hong, Kyung H. Chung, Department of Materials Science and Engineering, Korea Advanced Institute of Science and Technical, 373-1 Kusung-dong, Yusung-gu, Taejon, 305-701, Korea The high temperature creep behavior of SiC12124A1 metal matrix composites was investigated at 300°C. The steady state creep rates of SiCl2124AI composites were sensitively dependent on the aspect ratio and alignment of SiC reinforcements. Based on the concept of load transfer between reinforcement and matrix, a new parameter of effective aspect ratio was proposed considering the combined effects of aspect ratio and alignment on load transfer efficiency. The steady state creep rates and sub-grain sizes were measured similar when the effective stress for creep deformation of2124AI matrix is identical. It is suggested that the SiC reinforcements enhance, the creep resistance of SiCl2124AI composites by reducing the effective stress for creep deformation of2124AI matrix. A modified power law creep equation was proposed to describe the high temperature creep behavior of SiC12124AI composites. 11:00 am TENSILE AND COMPRESSIVE CREEP PROPERTIES OF ALUMINUM CONTAINING 2S VOL% OF ALUMINA DISPERSOIDS: A.M. Redsten, D.C. Dunand, Massachusetts Institute of Technology, 77 Massachusetts Avenue Room 8-328 Cambridge, MA 02139 ' , ~ creep properties in ~~ tension and compression of dispersion-strengthened cast a1ummum (DSC-AI) contammg 25 '101.% ofO.3llIDalumina particles are presented for two v~ry d!fferent gra~n sizes (~cron-and millimeter-range). Continuum models predict that dl~uslOnal creep IS the dommant deformation mechanism for fine-grained DSC-A1. Expenmental data for this grain size, however, indicate that power-law creep controls the deformation and that diffusional creep is negligible. This result is discussed in tenos of reduced mobility of grain-boundary dislocations due to the presence of particles. The stress exponent and activation energy of both fine-and coarse-grained DSC-A1 are significantly higher than those for pure aluminum, indicating a threshold stress inn the pow~r-~aw creep region. The measured power-law threshold stresses are compared to predlcl10ns from models based on dislocation-particle interactions. Finally, deformation is significantly faster in tension than in compression, a result discussed in tenos of cavitation. 11:20 am INTERPRETATION OF CREEP BEHAVIOR IN TERMS OF A MODIFIED POWER LAW: FA. Mohamed, Department of Chemical Engineering and Materials Science, University of California, Irvine, California 92717 Recent experimental data have suggested that the low-stress creep behavior of some materials may be best described by a modified creep power law which incorporates a

1996 TMS ANNUAL MEETING AND EXHIBITION

threshold stress. The presence of a threshold stress signifies thatan effective stress,rather than the applied stress, is responsible for the observed creep rate. Micrograin superplastic alloys and powder metallurgy Al alloys used as matrices in the development of discontinuous SiC-AI composites provide typical examples for materials whose lowstress creep characteristics are interpreted in terms of a modified creep power law. 11:40 am CONSTANT STRUCTURE CREEP STUDIES OF THE MECHANISMS OF DEFORMATION IN DISPERSION STRENGTHENED COPPER: S.E. Broyles, J.C. Gibeling, Division of Materials Science and Engineering, Department of Chemical Engineering and Materials Science, University of California, Davis, CA 95616 A new series of constant structure, stress reduction creep experiments on oxide dispersion strengthened copper alloys is described. The results of these experiments reveal that the constant structure creep rate is found to be significantly higher than the steady state rate corresponding to the reduced stress, reflecting the important role of dislocation density in the creep process. By considering the effects of a high dislocation density, the steady state and constant structure creep behavior can be described by a new model based on the thermodynamics and kinetics of dislocation detachment from hard obstacles. Unlike previous models, the current method does not require the introduction of threshold stresses or comparisons with arbitrary values of the activation energy for creep. A relaxation parameter, k, is shown to provide a measure of the degree to which a dislocation's line energy is reduced at the particle matrix interface. However, the behavior of dispersion strengthened materials can be adequately described by a combination of an athermal flow stress, (J which represents the strength of the microstructure, and a Helntholtz free energy of activation, AF. The magnitudes of AF and (J are determined from the constant structure creep data and are shown to be in the range expected for deformation controlled by relatively strong obstacles.

HIGH TEMPERATURE COATINGS II: Ceramic Coatings I

Sponsored by: MDMD Surface Modification and Coatings Committee Program Organizers: Narendra B. Dahotre, Center for Laser Applications, University of Tennessee Space Institute, Tullahoma, TN 37388; Janet M. Hampikian, School of Materials Science & Engineering, Georgia Institute of Technology, Atlanta, GA 30332; Jacob J. Stiglich, PO Box 206, Sierra Madre, CA 91025

Monday, AM February 5, 1996

Room:BI Location: Anaheim Convention Center

Session Chairpersons: Sun K. Kim, Department of Metallurgical Engineering, University of Ulsan, Kyung Nam, Korea 680-749; Narendra B. Dahotre, Center for Laser Applications, University of Tennessee Space Institute, Tullahoma TN 37388 8:30 am Opening Remarks Narendra B. Dahotre, Center for Laser Applications, University of Tennessee Space Institute, Tullahoma TN 37388

8:40 am Invited FORMATION OFWEAR·RESISTANTVANADIUMCARBIDEAND NITRIDE LAYERS ON TOOL STEElS: Sun K. Kim, Department of Metallurgical Engineering, University of Ulsan, Kyung Nam, Korea 680-749 When steels are heated at 900 - lOOO·C in a fused borax bath containing Fe-V powder, vanadium carbide layers are formed on the steel surfaces. Effect of steel composition and the treating condition on the growth rate of the vanadium carbide layers were investigated. Hardness and wear resistance of such layers were determined. Also nitriding and vanadium carbide forming was combined to obtain better properties of treated specimen. EPMA analysis was done for this V(C,N) type layer. Hardness and wear resistant properties of such treated specimen were investigated. 9:05am DEVELOPMENT OF A CONVERSION COATING FOR SILICON CARBIDE: S.c. Kung, Babcock & Wilcox, Alliance Research Center, 1562 Beeson Street, Alliance, OH44601 The formation of a chromium-carbide conversion coating on SiC bas been demonstrated using the pack cementation technique. The conversion coating would improve the high-temperature corrosion resistance of SiC and its derivatives, such as SiC-based composites and continuous fibers, by forming a protective Cr203 scale upon exposure to corrosive environments. The coating consists of a multi-layered structure with each of the sub-layers containing a high Cr concentration. In addition, the conversion coating is relatively dense compared to the underlying porous SiC substrate material. Such a coating morphology is highly desirable for many applications when impermeability is required. 9:25am PLASMA SPRAYED MoSl2ISlC COATINGS: J. Wolfenstine, XLiang, E.J. Lavemia, Department of Chemical and Biochemical Engineering, University of Califomia, Irvine, CA 92717-2575; A. Sickinger, Sulzer Metco (Irvine) Inc., 17426 Daimler Street, Irvine, CA92714 Various processing techniques for the synthesis of MoSi2 composite coatings reinforced with SiC particles using low pressure plasma spraying will be presented. The microstructure and thermal behavior of the coatings were characterized using x-ray diffraction, scanning and transmission electron microscopy, x-ray photoelectron spectroscopy and dilatometry. In addition, the room temperature fracture toughness and high temperature creep behavior were investigated and compared to MoSi2-SiC particulate composites prepared by powder metallurgy methods.

1996 TMS ANNUAL MEETING AND EXHIBITION

9:45am A MODEL FOR THE MICROSTRUCTURAL DEVELOPMENT OF PLASMA SPRAYED CERAMIC COATINGS: Seiji Kuroda, National Research Institute for Metals, Advanced Materials Processing Division, 1-2-1, Sengen, Tsukuba-shi, Ibarald, 305 Japan Plasma sprayed ceramic coatings are widely in use for high temperature applications such as thermal barrier coatings in jet engines and power generators. However, the relationship between a number of process parameters and the microstructure of these coatings is not well established. Sprayed coatings consist of layers of rapidly solidified splats. The paper deals with the most important microstructural features of ceramic coatings, i.e., microcracks within splats and interlamellar gaps. A theoretical model to correlate the spray conditions and the density of microcracks is presented first. Then, experimental evidences on the pore structure within coatings are compared with the theory. Importance of the substrate temperature during spraying on the interlamel1ar bonding is also demonstrated. 10:05 am CHEMICAL PROPERTIES OF SODIUM BERATE COATINGS: R.G. Reddy, J.Y. Yen, Department of Chemical and Metallurgical Engineering, University of Nevada, Reno, NV 89557 A thermodynamic study of ionic species distribution of high temperature sodium borate glass coating melts was undertaken to understand the boron anomoly. Mole fraction of BO), B)06' and B409 structural units in the depolymerized portion of borate melts were calculated at 1123K and as a function odcomposition. The addition of Na20 cause the boroxol ring breakdown into basic triangular structure unit BO). 10:25am BREAK 10:35 am EVOLUTION OF MICROSTRUCTURE DURING LASER INDUCED REAC· TION COATING OF ALUMINA ON SiClAI10 3 COMPOSITE: Narendra B. Dahatre, C. Xiao, W. Boss, M.H. McCay, T.D. McCay, Center for Laser Applications, University of Tennessee Space Institute, Tullahoma, TN 37388 The protection of SiCiAl20) composite at high temperature from deleterious reactions occurring within and with the surrounding environment is required for several high temperature applications. The development of continuous Al20) coating on SiC/Alp) ceramic composites for such protection was achieved using the laser assisted in-situ reaction technique. As deposited alumina coatings were analyzed by scanning electron microscopy, energy and wavelength dispersive x-ray spectroscopy for their morphology and composition. The interface between the substrate and the laser deposited alumina coating were characterized for chemical composition using above techniques. The performance and microstructural and chemical compositions were studied after exposing the coated samples to high temperatures (>800°C) for long time durations. 10:55 am OPTIMIZATION OF THE THERMAL SPRAY GUNS AND COATING PRO· CESSES USING NUMERICAL SIMULATIONIS: Shmuel Eidelman, Xiaolong Yang, Isaac Lottati, Science Applications International Corporation, McLean, VA 22102 Thermal Spray (TS) systems are low cost versatile methods of significantly improving the material properties of the base materials by coating them with the high performance materials. We show that the current TS systems are very inefficient in terms of energy and material utilization as well as in quality control. In addition, barrel erosion and coating lead to additional equipment maintenance costs and problems with coating consistency. We use a recently developed and validated three dimensional simulation capability for modeling TS systems gas and coating powder flow for the TS process analysis, and illustrate the roots of TS systems inefficiencies. The same capability can be used to design optimized TS gun systems, and for coating process optimization and control. Examples are given for TS process designs with improved performance and system efficiency. A reduced version of the developed numerical capability can be used for effective selection of the plating parameters and process control. 11:15Bm APPLICATION TO CERAMIC COATING AND SYNTHESIS OF SPHERICAL CERAMIC PARTICLES USING LASERSPRAYTECHNIQUE:Mamoru Okutomi, Koich Tsukamoto, Photun-process section, Electro-technical Laboratory, 1-1-4 Umezono Tsukuba iharaki Japan A laser spray method, which is melting and quenching of random shaped powder dropped perpendicularly to a focused CO2 laser beam in the air, has been developed. First, by observing using high speed camera, we investigate whether the ceramic particles in the Zr02-9 % mole Al 20), y-AI20), Ti02and AI20)-Cr20) system are spheroid while passing through laser beam ornot. The melting kinetics, residence time of particles inner laser beam and melting efficiency of particle with changing of laser power has estimated. Second, this rapid melting-solidification technique was applied to ceramic coating for improving several problems due to poor efficiency, production rates, controllability of the deposited structure. The properties of the coated layer are discussed with the melted particles passing through the laser beam. 11:35 am STRUCTURAL CHANGES ON FLAME·SPRA YED ALUMINA AND CHROMlA: R.S. lima, J.A.C. Martins, C.P. Bergmann, Department of Materials, School of Engineering, Federal University of Rio Grande do SuI. Av. Oswaldo Aranba 99n05, 90035-190 Porto Alegre RS, Brazil flame-sprayed substrate-free ceramic coatings of alumina and chromia were produced by applying the over preheated aluminum substrates around 200·C. During the deposition temperatures above 500·C were not allowed to avoid the substrate warpage. The coatings were easily detached from the substrate when the system coating-substrate reached the ambient temperature, due to the difference between the thermal expansion coefficient of the ceramic layer and the aluminum substrates. The coatings of alumina were reheated in air for periods of 2 hours at 1300 to 1700·C. The coatings of chromia

B7

were reheated at 1500'C for 2 hours. 1be structural changes occurred due to the heating on phase transitions. microstructure. true density. specific surface area for as-sprayed and sintered coatings were examined by X-ray diffraction. seaMing electron microscopy. helium pycnometer and gas sorption analyzer respectively. 11:55 am RESIDUAL STRESS, ADHESION AND THERMAL SPRAYED COATINGS: C. Richard. G. Beranger. Universtie de Technologie de Compiegne. Department de Genie Mecanique. URA 1505 du CNRS. 60206 Compiegne, France; J. Lu. Universtie de Technologie de Troyes, Department de Genie des Systems. Mecaniques, 10000 Troyes. France Most coatings are applied with a specific aim in mind, such as improving the base material resistance to corrosion and wear, or providing a thermal barrier against high temperatures. These aims can obviously be achieved if the coating is properly bonded to the substrate. One important factor which contributes to failure of plasma sprayed coatings is the residual stress produced during its munufacture. This paper gives a review of the results on residual stresses obtained in different coating systems (NiCrAIY metallic bonding layer. a Cr203 ceramic layer. and a WC-CO cermet layer). To study the effect of the residual stresses on the properties of the coatings (particularly on the adhesion of coatings to the substrate). two techniques are used: a Vichers indention test is pedormed at the substrate/coating intedace (in order to determine. by a simplified model based on mechanics concepts, "an intedacial toughness") and fout point bending test assisted by an acoustic emission system.

HIGH TEMPERATURE SUPERCONDUCTORS· SYNTHESIS, PROCESSING, AND LARGE SCALE APPLICATIONS I: Conductor Development

Sponsored by: JL EMPMDISMD Superconductlng Materials Committee, MSD Electronic, Magnetic and Optk:al Phenomena Committee, FEMS (Federation of European Materials Societies) Program Organizers: U. Balachandran. Energy Technology Division. Argonne National Laboratory. Argonne. IL 60439; Paul J. McGinn. University of Notre Dame. Notre Dame. IN 46556; Stuart Abell. University of Birmingham. Edghaston. BirminghamBl52TT. UK Monday. AM February 5. 1996

Room:B7 Location: Anaheim Convention Center

Session Chairpersons: R. Wesche. EPFL. Centre de Recherches en Physique des Plasmas. Switzerland; J. SchWartz. Florida State University. Tallahassee. FL 8:30 am Invited DEVELOPMENT OF HTS CONDUCTOR AND COILS FOR ELECTRIC POWER APPLICATIONS: P. Haldar. G. Hoehn, Jr.• Intermagnetics General Corp., Latham, NY 12110; U. Balachandran. Argonne National Laboratory, Argonne. 1L 60439 Long lengths ofbismuth-2223 based silver sheathed superconductors by the powderin-tube approach have been produced. to fabricate high temperature superconducting (HTS) components for electric power applications such as generators. motors. transmission cables. transformers. SMES. etc. Prototype lengths of multi filament tapes are presently being manufactured with critical current densities useful for most electric power applications. We bave recently demonstrated the fabrication oflengths approaching I km of Bi-2223 multi filament conductor that carried> 10.000 Ncm2at 77 K. self field. A shorter piece length (20 m) exhibited a critical current of 42 A with a correspondingJ. of21.ooo Ncm2. Work at ANL and part of the work at IGC is supported by the U.S. Department of Energy (DOE). Energy Efficiency and Renewable Energy. as partofa DOE program to develop electric power technology. under Contract W-31-109Eng-38. 8:50 am InvIted DEVELOPMENT OF BlzSrlCalCulO,HTS CONDUCTORS AND COILS FOR ruGH FIELD GENERATION: K. Togano. H. Kurnakura. H. Kitaguchi. National Research Institute for Metals, TSukuba 305. Japan; M. Okada, Hitachi Research Laboratory. Hitachi 319-2, Japan; I. Kato. Hitachi Cable Ltd.• Tsuchiura 300. Japan; T. Hasegawa, Showa Cable & Wire Company. Ltd .• Kawasaki 210. Japan We have concentrated on developing long lengths of Bi2Sr2CalCu20J tape conductors and their coils aiming at the high magnetic field generation in Liq. He or by refrigerator operation.1be tapes were fabricated by the powder-in-tube method (PIT) or coating process on Ag substrate tape. Multifilamentary tape processed by PIT showed good strain tolerance and uniformity of Jc along the tape. Using this multifilamentary tape. we have fabricated a compact magnet with 13mm and 45mm in inner and outer diameters. respectively. which was designed as a booster of the 20T class superconducting magnet. 1be coil generated 3.3T in zero hias field at 4.2K. In order to improve the strain tolerance of the coating-processed tape. a laminated tape was fabricated by sandwiching several coated tapes between Ag foil, which was heat treated continuously by being passed through a pipe furnace with precise temperature gradient. 1be laminated tape bas good strain tolerance which is enough forcoil malting by "react & wind" method. Coil tests in strong bias field in Liq. He or refrigerator operation are in progress. 9:10 am InvIted CONTROLLED PROCESS OF Ag-CLAD BI-BASED SUPERCONDUCTING TAPES: S. X. Dou. M. lonescu. W. G. Wang. Y. C. Guo. H. K. Liu. Centre for Superconducting and Electronic Materials, University of Wollongong. NSW 2522. Australia; E. Babic and I. Kusevic. Dept. of Physics. Univ. of Zagreb. Zagreb. Croatia;

B8

and E. W. Collings. Dept. of Materials Science and Engr.• 1be Ohio State University. Columbus. OH 43210-1179 A continuous forming process was successfully tested as an alternative process of the powder-in-tube process for fabrication of long length superconducting tapes. A "sandwich rolling" process has been developed to prevent the formation of sausaging and cracks in longitudinal direction of the Ag-clad Bihased superconducting tapes. Higb Ie for AglBi-2212 tapes achieved through MTG in the alternating oxygen and nitrogen atmosphere. While N2 annealing on cooling ensures high Tc' O2treatment during melting and slow cooling period enhances the stability of 2212 allowing for grain growth with ample liquid formation. Direct observation of the interface achieved by removing Ag sheath with Hg alloy show:s highly dense. textured and thin-film like structure. Grain growth in 2223 tape must be coupled with the 221212223 phase transformation in order tooptimiseJ.- No initial drop in J.-B for 2212 MTG tape as typical feature of2223 tape. In highT,J. forboth2212 and 2223 tapes is limited byintragrainJ•. 1be U.for2223 tape is larger than that for the best 2212 tape and epitaxial thin films. 9:30 am InvIted BlzSr1CaCu10. CONDUCTOR PREPARATION FOR INSERT COILS IN 20 T MAGNETS: J. Schwartz. J. Kessler. W. Wei. H. Weijers. Florida State University. National High Magnetic Field Laboratory. Tallahassee. FL 1be superconducting properties of many high temperature superconductors are excellent for applications in high magnetic fields. One of these materials is Bi2Sr2CaCu20.. Li-doping of this superconductor greatly improves the superconducting properties. Powder-in-tube wires and tapes were prepared using Ag and AgMg tubes. Li-doped and undoped powders. Additionally. hot rOlling of the tapes was pedormed to improve the conductor pedorrnance. Short samples were characterized by critical current measurements in high magnetic fields up to 20 T. Investigations of the ceramic microstructure were performed by SEM to correlate the critical current with the grain structure and impurities inside the core. Wind-and-react layer-wound and pancake coils were prepared to analyze the critical current .in longer conductors lengths at 4.2 K and fields up to 20 T. This work was supported by the National Science Foundation through the National High Magnetic Field Laboratory. Grant No. 90 16241. 9:50 am Invited RECENT ADVANCES IN HTS COMPOSITE SUPERCONDUCTORS: J. M. Seunliens. American Superconductor Corp.• Two Technology Drive. Westborough. MA 01581 Continued progress in the pedorrnance of HTS composite superconductors has been made. Using powder metallurgical techniques that are based on scaleable processes, filament critical current densities in excess of 40.000 Ncm2 (77 K. self field) have been achieved for multifilamentary conductors. Composites have been fabricated which demonstrate the ability to twist and bend the composite without excessive degradation in critical current. Astrong interest in HTS composites exists for alternating current (AC) applications. AC losses of composite conductors have been characterized to quantify eddy current. hysteresiS, and coupling current losses. Microstructural analysis of the HTS composite superconductors provides insights into mechanisms that limit supercurrent, strain tolerance. and AC loss. Results of recent efforts in fabrication of multistrand conductors and coils will be presented. 10:10 am BREAK 10:20 am OPTIMIZATION OF mE FABRICATION PROCESS OFBI(2223)TAPES: G. Grasso. F. Marti. A. Jeremie.B. Hensel. R. FlUkiger, UniversiU! de Gem!ve. Departement de Physique de la Matiere Condensee. 24. quai ErnestAnsermet. 1211 Geno!ve 4. Switzerland In this work. we describe in detail the technique we have developed for the preparation of rolled. high ic Bi(2223) tapes with Ag and Ag-Mg sheaths. New results about the correlations between the critical current density and the powder heat treatment temperature and time. the cold deformation steps. as well as the tape heat treatment temperature and time are presented. Moreover. we have found that the local pressure exerted on the tapes during the rolling deformation between the heat treatments has a strong influence on the transport properties: the highest ic (77 K. 01") value (30 kNcm2) was found for a local rolling pressure of 0.6 GPa. The addition of Mg to the Ag sheath has been used to improve the mechanical resistance of the tapes. We have found that the Mg addition has also remarkable effects on the principal physical properties of the tapes. 1bermaI transport measurements have been pedormed on both types of tapes. The Ag sheathed tapes have shown a thermal conductivity 2 and 7 times higher than the Ag-Mg sheathed ones at 77 K and 4.2 K. respectively. Preliminary results have shown that at 77 K the transport AC losses are about 20% lower in Ag-Mg sheathed tapes. 10:40 am ALTERNATIVE SHEATH MATERIALS FOR BI-2223: J. D. Col/on. Los Alamos National Laboratory, Los Alamos. NM 87545; G. N. Riley. Jr.• American Superconductor Corp.• Westborough. MA 01581 1be oxide-power-in-tube (OPIT) process for producing Bi-2223 superconducting tapes has been used with considerable success. This is largely due to the good chemical compatibility of the tube material. pure silver. with the Bi2223 phase. Additionally, silver is formable. has a suffiCiently high melting point. is oxygen-permeable. and has a low electrical resistivity. However. (:ertain device designs would benefit from a sheath material wbich has a higher electriccal resistivity than pure silver. This presentation will report the results of our investigations into alternative ssheath alloys for processing Bi2223. Alloys investigated include Ag-Mn. Ag-Pd and Ag-Pt.1be electrical resistivity at 75 and 300 K is determined, as well as hardness, oxidation response. chemical compatibility and superconducting transition temperature.

1996 TMS ANNUAL MEETING AND EXHIBITION

11:00 am Invited FABRICATION AND PROPERTIES OF SUPERCONDUCTING AgNlMg AND AgBl-llll WIRES: R. WescM. A. M. Fuchs. B. Jakob. G. Pasztor. EPFL. Centre de Recherches en Pbysique des Plasmas. CH-5232 Villigen PSI. Switzerland Superconducting AgNiMg and AglBi-2212 monocore wires of I and 1.5 mm diameter have been fabricated by the powder-in-tube technique from the same batch of powder. The effects of the different sheath materials on the optimum heat treatment conditions and the superconducting properties of the Bi-2212 wires have been studied. At4.2 K and B =0 short sample critical current densities up to 70000 Alcrrr- and 100000 Alcm2 have been achieved for AgNiMg and AglBi2212 wires of I mm diameter. respectively. For both wire diameters the critical temperature is lower in the case of the AgNiMg. The optimum heat treatment conditions have been found to depend on the sheath material. This may be a consequence of rapid oxygen diffusion through the AgNiMg sheath via grain boundaries. Finally. rust results for multifilamentary AgNiMg and AglBi-2212 wires will be presented.

9:20 am InvIted UNIVERSAL ATOMIC SIZE CRITERION FOR GLASS FORMATION: T. Egami. Department of Materials Science and Engineering, University of Pennsylvania. Philadelphia. PA 19104 The importance of the atomic size difference in the stability of metallic alloy phases has long been known ever since the workofHume-Rothery. However. the rules remained largely semi-empirical. We show that the size factor can be quantitatively understood from the view point of the local topological instahility, and a single principle explains melting. glass transition. solid state amorphization by irradiation as well as range of composition for glass formation by liquid quenching. The local topological instability occurs when the relation between the local volume and the local coordination number deviates by a cenain amount from the ideal relation which can be deduced from purely geometrical considerations. Glasses become relatively stable when crystals of the same composition are destabilized by the size effects. These effects are demonstrated by computer simulation.

11:20 am CRITICAL CURRENT ANISOTROPY AND EFFECTIVE TEXTURE IN BSCCO-PHASE CONDUCTORS: J. Everen. M. DhalI~. M. Johnston. A. D. Caplin. Centre for High Temperature Superconductivity. Blackett Lsboratory. Imperial College. London SW7 2BZ, UK The powder-in-tube (PIT) fabrication of BSCCO-phase conductors aims to achieve a high degree of texturing. with the abplane of the crystallites parallel to the plane of the conductor. However. it is now well-known that the current distribution in these tapes is far from uniform, and that texture may play an important role in determining which are the dominant current paths. Therefore. the anisotropy of the critical current (at some voltage level) with respect to the direction of an applied magnetic field reflects the texture of the significant current pahts. which may be very different from the texture averaged through the core of the conductor. as obtained from structural studies. We repon anisotropy studies of such conductors. analyze the nature of their texture. and draw inferences above the form of the current paths.

9:50am TIGHT-BINDING APPROACH TO ORDER ANDMETASTABILlTY IN AMORPHOUS ALLOYS:P.EA. Turchi. LLNL (L-268) P.O. Box 80S. Uvermore. CA 94551. Dicker Mayou. LEPES-CNRS. 25 Avenue des Martyrs. BP 166. 3S042 Grenoble Cedex 9. France We present a real-space approach based on the tight-binding approximation to study order and metastability in amorphous alloys. The extended recursion coupled with the coherent potential approximation and the generalized penurbation method allows a realspace treatment of ordering processes in this class of materials. The interactions obtained from these electronic structure calculations which build up the ordering pan of the total energy can then be used with a statistical model to study ordering trends. Finally. in combination with molecular dynamics simulations and the full tight-binding expression for the energetics, the most probable atomic configuration of an alloy can be predicted and analyzed on the basis of fundamental electronic parameters. The advantages and the future of this new approach together with some preliminary results will be discussed. Wark performed under the auspices of the U.S. Department of Energy by the Lawrence Uvermore National Laboratory under contract No. W-7405-ENG-48. Partial suppon from NATO under contract No. CRG-941OZ8 is gratefully acknowledged.

11:40 am POWDER-IN-TUBE TAPES OFTI-I223 MATERlALPREPARED BY D1FFERENTMETHODS: R. LOw. T. Riepl. G. ROOI. B. Steinberger. W. Huber. R. Stierstorfer. G. OsWald. K. F. Renk. Institut fur Angewandte Pbysik. UniversitlitRegensburg. 93040 Regensburg. Germany We have synthesized (TI.Bi)lSr16B80.4~CU3Ox (TI-I223) with a Tc up to 117 K by the usual ceramic route and by a single step sol-gel process using citric acid and ethylene glycol. These methods lead to T1-1223 material with different grain sizes (ceramic route about 10 m - sol-gel process about 50 m). Our powder-in-tube tapes from material prepared by the ceramic route showed.k5.5 x 103 Alcrrr-(I,,13 A) at 77 KandOTwhen rolled longitudinally and transversally.

HUME·ROTHERY AWARD SYMPOSIUM FOR WILLIAM L • .JOHNSON: Thermodynamic And Kinetic Issues In The Synthesis Of Metastable Materials I

Sponsored by: Jt. EMPMDlSMD Alloy Phases Committee Program Organizer: Professor Carl C. Koch. Materials Science and Engineering Depanment, North Carolina State University. Box 7907. Raleigh. NC 27695; Dr. Roben D. Shull. NIST. Bldg. 223 B152. Gaithersburg. MD 20899 Monday. AM February 5. 1996

Room: Orange County I Location: Anaheim Marriott Hotel

Session Chairperson: Carl C. Koch. Materials Science and Engineering Department,

Nonh Carolina State University. Box 7907. Raleigh. NC 27695

8:30am OPENING REMARKS: Carl C. Koch 8:40 am Keynote THERMODYNAMIC AND KINETIC ASPECTS OF GLASSY PHASE FORMATION IN NON-EQUILIBRIUM METALLIC SYSTEMS: William L. Johnson. California Institute of Technology. I3S-7S. Pasadena, CA 91125 A glass or amorphous solid is a condensed phase. which like the related liquid phase. lacks long range periodicity in the arrangement of atoms. Amorphous solids can be formed wben the liquid phase is deeply undercooled and configurationally freezes without crystallizing. when a vapor condenses under restricted kineteics. when a crystalline solid is driven from equilibrium by a disordering processes so that long range order collapses. or during chemical reactions in the solid state where kinetic contstraints prevent the attainment of full equilibrium. The latter two cases can be usefully described as "solid state melting" or "solid state amorpbization". The fundamental thermodynamic and kinetic factors which govern these various types of glass formation will be discussed and compared. Experimental results relating to glass formation. melting. and arnorphization in metallic systems will be surveyed. reviewed. and analyzed to illustrate how a fundamental understanding of these phenomena has developed.

1996 TMS ANNUAL MEETING AND EXHIBITION

10:10 am BREAK 10:30 am Invited THERMODYNAMICS AND METASTABILITY IN GLASS FORMING SYSTEMS: H.J. Fecht. TV Berlin. Hardenbergstr. 36 PN 2-3. 10623 Berlin. Germany Glasses are generally produced from the highly undercooled liquid state by rapid quenching methods or quasi-statically at slow cooling by the effective control of potent heterogeneous nucleation sites. For metallic systems the latter method recently has led to the development of bulk metallic glass with a complex multicomponent chemistry and advanced engineering properties. As long as crystallization can be avoided the relevant thermodynamic properties of the metastable glassy and undercooled liquid phases can be measured below and above the glass transition temperature. respectively. The obtained data give new insight into the nature of the g1a6S transition suggesting that it is not a phase transition in the classical sense but kinetic freezing triggered by an underlying entropic instability. Furthermore. glass formation can be achieved by solid-stateprocessing without passing through the liquid state. Conceptually, the formation of the glassy state from the liquid and the solid can be described within a thermodynamic framework under appropriate kinetic constraints resulting in a universal phase diagram with a pseudo-critical point. 11:00 am Invited DEFECTS, DEFECT MEDIATED TOPOLOGIAL ORDER-TO-DISORDER TRANSITIONS AND SOLID STATE AMORPmZATION: Mo U. W.M. Keck Laboratory. 13S-7S. California Institute of Tecnology. Pasadena. CA 91125 Crystal-ta-g1ass transition can be considered as one of the defect induced topological order-to-disorder transitions. Stability of a crystalline phase is closely related to the presence of defects that are either introduced during experiments or pre-exist in the host crystalline phases. It was shown. through computer simulations. that it is the topological defects that are ultimately responsible for the crystalline instability. or amorphization. The strain fluctuations caused by the defects give rise to softening of elastic constants, leading to the so-called "mechanical instability" at amorphization. The arnorphization can occur in different modes depending on the state of the defects. We show that the normally observed nucleation and growth in amorphization can change if a mechanical spinodal precipitates preemptively. We will show that the (static) displacement fields of atoms from their stable lattice positions can be used as a general order parameter to describe the topological order-to-disorder transition. Connections of this order parameter to experimental measurements will be discussed. 11:30 am CALCULATION OFPHASESELECTION IN NON-EQUILIBRIUM PROCESSING: G. Shaw andP. Tsakiropoulos. Department of Materials Science and Engineering. University of Surrey. Guildford. Surrey GU 25XH. United Kingdom Predominant nucleation maps in rapidly solidified alloys were calculated by combining phase equilibria data, derived using the CALPHAD method and extended to undercooled melts, with time dependent nucleation and steady state nucleation theories. The study includes AI.1M, compounds in undercooled gas atomised droplets of AI-TM alloys. stoichiometric compounds and a solid solution phase in melt spun AI-Ti-V alloy ribbonsandthefcc.hcpandorderedLl.(y)phasesinTiAl-Vaiuminides. Phase selection under rapid solidification conditions will be discussed in terms of transient nucleation and steady state nucleation at different melt cooling rates or melt undercoolings.

B9

HYDROMETALLURGY: RECENT DEVELOPMENTS I: Progress in Electrometallurgy

Sponsored by: EPD Aqueous Processing Committee Program Organizer: Professor David Dreisinger, The University of British Columbia, Department of Metals and Materials Engineering, 309-6350 Stores Road, Vancouver, B.C., C V6T lZ4 Monday, AM February 5,1996

Room: A14-l5 Location: Anaheim Convention Center

Session Chairperson: Dr. Arash Kasaaian, Research Electrochemist, Elkem Metals Co., PO Box 299, Marietta, OH 45750 8:30am SPOUTED BED ELECTROWINNING OF COPPER AND ZINC: S. Siu, J. W. Evans, University of California, Department of Materials Science and Mineral Engineering, Berkeley, CA 94720; J. Salas, CODELCO Division Chuquicamata, Chuquicamata, Chile; M. Newman, Pasminco Research Center, Boolaroo, New South Wales, Australia Prior investigations into fluidized bed electrowinning (FBE) have shown good results for copper but rather poor current efficiencies and high energy consumption rates for zinc (electrowon from conventional acid electrolytes). This paper describes an alternative electrode, the spouted bed electrode (SBE) which, in laboratory-scale investigations, appears superior to both FBE and conventional electrowinning. Current efficiencies and cell voltages comparable to those of conventional cells were achieved at current densities greatly in excess of those used in conventional cells for both copper and zinc. The paper describes the results of measurement of SBE cell performance as a function of electrolyte composition, particle size, current density and other experimental parameters. Preliminary work oriented towards cell scale-up is also described. 9:00am NEURAL NET MODELLING OF ZINC ELECTROWINNING PROCESS: D.H. Rubisov, V.G. Papangelakis, University of Toronto, Department of Chemical Engineering and Applied Cbemistry, Toronto, Ontario, Canada M5S lA4; G. Houlachi, C. Bowen, Noranda Technology Centre, 240 Hymas Boulevard, Pointe-Claire, Quebec, Canada H9R lG5 The neural networks modelling approach is one of the most elegant black-box type methods to model industrial processes. In the paper, this approach is applied to model the performance of the zinc electrowinning process. The model that was built was found to exhibit an excellent predictive ability for voltage and current density. The model was also used to investigate the sensitivity of the process to various key variables such as temperature, current density, zinc and sulphuric acid concentration, etc. Optimal regimes for zinc production were identified from the power consumption standpoint. 9:30am RECENT OPERATION IN TANKHOUSE OF TAMANO SMELTER: Y. Oda, M. Kusano, K. Noda, Hibi Kyodo Smelting Co., Ltd., Copper Refinery, Tarnano Smelter, No. 6-1-1 Hibi, Tamano City, Okayama Pref. 706, Japan The tankhouse of Tarnano Smelter has adopted a high current density electrolysis operation applying PRC (Periodic Reverse Current) method and before 1991, it has reconciled both high production capacity of 161,000 tonslyear and high labor productivity. However, recent rapid rise of Japanese currency in foreign exchange market has urged us to further exert our efforts in reduction oflabor cost. In order to achieve this goal of labor cost reduction, we have completed, in 1994, our phase four production expansion project which incorporated significant re-structuring of the organization and accomplished annual production of 191,000 tons/year. This project can be outlined as increase of number of cells for electrode exchange per day in the commercial electrolysis section to meet the increased number of cells and increase of the capacity of impurity treatment section. These improvements were done with minimum installation of new equipment and facilities and with achievements in efficiency and manpower saving of the existing facilities. Together with these improvements, rationalization in the start sheet section was carried out by changing the material for mother blanks as well as changing their maintenance procedures. At the same time, the crane in this section was modified to enable automatic operation by a fuzzy control. As a result, we were able to save manpower equivalent to 5 operators in this section. With these improvements, the production capacity has been increased by 18% and labor productivity by 25%. 10:00 am IMPROVEMENT OF THE PERFORMANCE OF ANODES FOR ZINC ELEC· TROWINNING: J.M.S. Rodrigues, Mintek, Hydrometallurgy Division, Private Bag X3Ol5, Randburg, South Africa 2125 The use of sandblasting for preconditioning anodes made from two different leadbased alloys for zinc electrowinning was investigated and compared with conventional fluoride preconditioning. Laboratory results showed that sandblasting and preconditioning both accelerated the formation of stable Pb02 layers on the anode and reduced oxygen overpotentials. The oxygen.overpotential was lower on the lead alloy containing silver and calcium than on that containing silver only. It was found that sandblasting improved the physical stability of the lead dioxide layer compared with that obtained by fluoride treatment. This procedure was tested on an industrial scale in the cell house of a local zinc producer and in addition to conftrming the laboratory results, corrosion rates of the lead anodes were established. An unexpected finding of the large scale tests was that the sandblasting appeared to work harden the lead anodes reulting in reduced warping during handling and operation. Sandblasting is easier to implement on a plant scale as no electrolysis or chemical treatment is required and the operation can be completed in a much shorter time.

B10

10:30 am THE REGENERATION OF HYDROCHLORIC ACID IN A TWO MEMBRANE NICKEL CHLORIDE ELECTROWINNING CELL: L. Liao, A. van Sandwijk, G. Van Ween, Delft University of Technology, Department of Raw Materials Technology, Faculty of Mining and Petroleum Engineering, Mijnbouwstraat 120, 2628 RX Delft, The Netherlands During electrowinning of nickel chloride, hydrochloric acid instead of chlorine was regenerated in a two membrane cell, by segregating the chloride catholyte with an anion exchange membrane, and tho~ sulphate anolyte with a cation exchange membrane form the middle electrolyte (ampholyte) compartment. Successful operation of such a cell depends on controlling chloride leakage through the cation exchange membrane to the anolyte. Small amounts of chlorine gas could be detected in the oxygen evolved at the anode. In this paper, the chloride leakage as function of various types of cation exchange membrane and different composition of the anolyte is reported and discussed. Ag+ was added to the anolyte to eliminate chloride leakage to the anode. The cell voltage, electrode potential and voltage drop over the membrane as function of current density has also been investigated. 11:00 am OPERATION OF THE MCLE (MATTE CHLORINE LEACH ELECTROWIN· NING) PLANT FOR NICKEL REFINING AT SUMITOMO MOTAL MINING CO., LTD.: S. Makino, M. Sugimoto, F. Yano, N. Matsumoto, Niihama Nickel Refinery, Sumitomo Metal Mining Co., Ltd., 3-5-3, Nishibara-cho, Niihama, Ehimo, Japan Sumitomo Metal Mining developed a new nickel refining process (MCLE process) based on chlorine leaching and electrowinning technologies and started the operation of the pilot plant with product capacity of 4,000 tons per year of cathode nickel in 1989. Then the commercial plant with design product capacity of 24,000 tons per year was constructed and has been successfully operated since August, 1992. Today the refinery product capacity is increased to 30,000 tons per year. Through these operations our MCLE process has been demonstrated to be one of the mot superior processes in the world with productivity, cathode quality and working environment. This paper describes the equipments and operation of the Sumitomo's MCLE process. 11:30 am GOLD UPTAKE BY PYRITE IN BROMIDE SOLUTION: M. Fofana, Dept of Mineral Engrg, K. Osseo-Asare, Dept of Matis Sci & Engrg, Pennsylvania State University, University Park, PA 16802 The uptake of dissolved gold by pyrite in aqueous gold (III) bromide solution was investigated as a function of pH (I to 12), ionic strength (0.0 to 0.5M) and time, using various surface and solution chemical techniques, including: (a) zeta potential measurements, (b) atomic absorption spec:tropbotometry (AA), (c) ion chromatography (IC).and (d) ultraviolet-visible spectroscopy (UV-Vis.). Gold uptake was constant (99% gold removal) up to pH-9, and then decreased with further increase in pH. The trend is rationalized in terms of the combined effects of gold speciation, specific chemical interactions, and repulsive electrostatic interactions. The relatively high gold removal at low pH indicates that the pyrite sutface can be a major source of gold loss during bromine leaching of sulfidic gold ores. Also, the high extent of gold removal even for extremely low pyrite quantities (e.g., I mg of Au removedlmg of pyrite) can adversely affect the efficiency of industrial operations. The substantial gold removal from solutions of high pH values and high ionic strengrh also demonstrates that the pyrite surface can be efficient at scavenging gold eve al high pH.

INTERMETALLICS AND THEIR COMPOSITES: PROCESSING AND APPLICATION TECHNOLOGIES I: Titanium Aluminides

Sponsored by: SMD High Temperature Alloys Committee Program Organizer. P.W. Keefe, Special Metals Corporation, New Hartford, NY 13413; M. Sohi, Allied Signal Engines, Phoenix, AZ 85072-2118

Monday, AM February 5, 1996

Room:B4 Location: Anaheim Convention Center

Session Chairperson: P.W. Keefe, Special Metals Corporation, New Hartford, NY 13413 8:30am mGH TEMPERATURE FLOW BEHAVIOR OF TITANIUM ALUMINIDES REINFORCED WITH mGH VOLUME PERCENTAGES OF TlB1: J. S. Marte, R. J. Martin, and S. L. Kampe, Virginia Polytechnic Institute & State University, Blacksburg, VA 24061-0237 Near- '"(fiAl and AI)Ti-based interrnetallic matrices reinforced with 30, 40 and 50 v% TiB2 have been produced by XD® processing and densified into bulk form by powder metallurgy techniques. The compressive flow behavior has been evaluated at temperatures ranging from 800·C to 1200·C and strain rates from 0.0001 Hz to om Hz. The results will be analyzed by a correlation to the resulting microstructure, especially with regards to the intermetallic matrix composition, TiB2 size, and interparticle spacing. This research is motivated by the potential application of these materials in capacities currently dominated by structural ceramics.

1996 TMS ANNUAL MEETING AND EXHIBITION

8:50am PROPERTIES OF TITANIUM INTERMETALLICS-BASED PARTICULATE COMPOSITES: Satoshi Emura, Masuo Hagiwara, Yoshikuni Kawabe, National Research Institute for Metals, 1-2-1 Sengen Tsukuba Ibaraki 305 Japan Ti-AI-Nb intermetalJics, such as Ti3AI and Ti 2AINb, are considered to be very promising as a new turbine engine material. In order to improve their high temperature mechanical properties, TiB particulate-reinforced composites were developed by blended elemental (BE) powder metallurgy method. A new technique was introduced, whereby matrix forming powders were blended with TiB2 ceramic powder to form TiB in-situ in the matrix. Processing conditions such as the sintering temperature/time for obtaining homogeneous composites were discussed in detail. Microstructural observations and high temperature mechanical tests, particularly tensile and creep tests, were performed. The results were also compared with unreinforced matrix material. 9:10am SYNTHESIS OF TITANIUM ALUMINIDES BY MECHANICAL ALLOYING: F. H. Froes, C. Suryanarayana, Institute for Materials and Advanced Processes, University of Idaho, Moscow,lD 83844-3026; C. M. Ward-Close, P. Goodwin, ORA Famborough, RSO Building, Hampshire GUI4 6TD, United Kingdom; F. Biancaniello, F. Gayle, National Institute of Standards and TechDology, Gaithersburg, MD 20899 Synthesis of the titanium a1uminides (TixAI, where x =I or 3) using a mechanical alloying (MA) approach allows production of novel materials with potentially enhanced behavior. This results from the ability to achieve novel constitutional and microstructural effects by MA, including solubility extension, modification of the crystal structure, and refinement of the grain size. Recent results will be discussed for both Ti3AI and TiAItype compositions. Consideration will also be given to the influence of the milIing atmosphere on phase selection; particularly the effect of nitrogen and oxygen contamination. 9:30am REACTIVE SINTERING AND REACTIVE POWDER FORGING OF A TIAIBASED ALLOY: David E. Alman, Materials Science Division, U. S. Bureau of Mines, Albany Research Center, Albany, OR 97321 High melting temperature, low density, and oxidation resistant intermetalJic compounds, such as TiAl, are desirable for use as structural materials at elevated temperatures and in aggressive environments. Recently its has been recognized that these compounds may have applications in industries, such as the automotive industry, where cost is frequently a major factor in material selection. For intermetallic compounds to be widely used for these types of applications, low cost fabrication methods will be a requirement. One potential economical method for component product is coupling Reactive Syntheses with traditional powder metallurgy techniques (e.g. press and sinter of powder forging). Reactive Synthesis involves forming compounds insi/u from elemental powders. Discussed will be the effect of alloy composition and processing parameters on the microstructure and properties of a TiAI-based alloy fabricated by reactive sintering and reactive powder forging. 9:50 am BREAK 10:10 am EFFECTS OF W, Mo AND SI ON THE CREEP BEHAVIOR OF A CAST GAMMA TITANIUM ALUMINIDE: P. R. Bhowal, H. F. Merrick, AlIied Signal Engines, Phoenix, AZ 85034 Gamma titaniumaluminides exhibit attractive high temperature specific strength and stiffness characteristics, and oxidation resistance. However, improvements in the creep resistance of gamma aluminides are needed if the promise for more general application in gas turbines is to be realized. A new investment cast gamma titanium aluminide, nominally Ti-47A1-2Nh-IMn-0.5W-O.5Mo-O.2Si(At.%), has been identified which exhibit significantly higher creep resistance over first generation cast a1uminides such as Ti-47AI-2Nb-2Cr and Ti-47AI-2Nb-2Mn. Creep strength improvement over these latter a1uminides was achieved by the addition of small amounts ofW, Mo and Si. In this study, the effects of these alloying elements on tensile and creep hehavior have been evaluated in more detail in order to optimize the new gamma a1uminide chemistry. Tensile tests were performed on material in the HIP and heat treated conditions at temperatures from RT to 815"C and creep tests were conducted in the temperature range 650-8 15C. The influence of the W, Mo and Si alloying additions on microstructure and mechanical properties of the new gamma a1uminide will be described. 10:30 am MICROSTRUCTURAL REFINEMENT OF THE TI46A12WO.5SICAST ALLOY BY STATIC HEAT TREATMENT: E. Evangelista, W. J. Zhang, L. Francesconi, E. Cetri, S. SpigarelIi, Department of Mechanics, University of Ancona, Via Brecce Bianche, 1-60131 Ancona, Italy Investment casting may be the most feasible and cost effective approach for the manufacture of near-net-shape components such as blades and vanes, compared to wrought or powder metallurgical route. But the key shortcoming of cast TiAl alloy is the remain of coarse ingot microstructure in the final product, which deteriorates the property performance. In this paper, static heat treatment; quenching/tempering process, was performed in an attempt to refine the microstructure of a cast Ti46AI2WO.5Si alloy. It was found that a fine near-fully lamellar structure with the colony size of about 30 11m, or a duplex structure comprising of equaixed wains and fine a2 plates or spheroids, can be obtained by tempering the quenched massive-transformed structure at1280-117OC. Preliminary compression tests and hardness measurements conftrmed that the ductility and strength of the studied alloy are greatly improved with the fine tempered microstructure. The microstructure change during heat treatment were examined by TEM, and the existence of Pand TisSi3 phase was confirmed due to the addition of W and Si.

1996 TMS ANNUAL MEETING AND EXHIBITION

10:50 am HALL PETCH BEHAVIOR IN GAMMA-BASED TITANIUM ALUMINIDES: IMPLICATIONS FOR PROCESSING AND APPLICATIONS: W.O. Soboyejo,C. Mercer, Department of Materials Science and Engineering, The Ohio State University, 2041 College Road, Columbus, OH 43210-1179 The inverse square root dependence of yield/fracture properties on the average equiaxedllamellar packet size in gamma-based titanium aluminides is demonstrated in this paper. Room- and elevated-temperature yield/ultimate tensile strength and ductility , and room-temperature fracture toughness are shown to exhibit such grain size dependence in binary (Ti-48AI) and ternary (Ti-48AI-1.4Mn, Ti-48AI-2Mn and Ti-48AI1.5Cr) gamma alloys. The strong effects of the average equiaxed equiaxedllamellar packet size in duplex alloys with a wide range of lamellar volume fractions is attributed to the controlling influence of soft deformation modes. The observed trends are also rationalized by considering the critical conditions associated with dislocation pile-ups in lamellar and equiaxed grains. The implications of the observed grain size dependence are discussed for future processing and alloy development that may impact the near/ midterm applications of gamma alloys. 11:10 am DEFORMATION AND FRACTURE UNDER COMPRESSIVE LOADING IN LAMELLAR TIAI MICROSTRUCTURES: Ming Dao, Bimal K. Kad, Robert J. Asaro, Department of Applied Mechanics and Engineering Sciences, University of California-San Diego, LaJolla, CA 92093-0411 A rate dependent elasto-viscoplastic, fmite strain, crystallographic Slip model is applied to study the mechanical behavior oflamellar and nearly lamellar TiAI polycrystals. Orientation dependent mechanical behavior of single PST is modeled using softmode and hard-mode slip systems at the single crystal level. At the polycrystallevel, finite element method is used to catch the non-uniform deformation modes developed. Stress-strain curve, texture development and especially the failure modes under the compressive loading are examined. Intergranular fracture and internal buckling are found to be the most important failure modes under compression. The computational results are shown to be in very good agreement with our proof of the concept' experiments. Results under compressive loading are also compared to the results under tensile loading.

INTERNATIONAL MOBILITY OF ENGINEERS AND ENGINEERING ACCREDITATION I

Sponsored by: TMS International Activities Committee Program Organizers: Dr. Axel Meisen, Dean of Applied Science & Professor of Chemical Engineering, Office of the Dean, Faculty of Applied Science, The University of British Columbia, 2006-2324 Main Mati., Vancouver, B.C. Canada V6T IZ4; Dr. J. Keith Brimacombe, A1can Chair, Matis. Process Engr., Univ. of British Columbia, Centre for Met. Process Engr., 309-6350 Stores Road, Vancouver, BC V6T 1Z4, Canada; Dr. William P. Imrie, Principal Engineer, Bechtel Group Inc., 50 Beale Street, San Francisco, CA 94105; Prof. Carl McHargue, Univ of Tennessee, Dept of Matis Sci & Engrg, 102 Estabrook Hall, Knoxville, TN 37996-2350; Dr. Fiona Doyle, Univ of California, 551 Evans Hall #1760, Berkeley, CA 94720-1760; Peter P. Schepp, Managing Director DGM, Adenauerallee 21, Oberursel, 0-6370, Germany Monday, AM February 5, 1996

Room: All-12 Location: Anaheim Convention Center

Session Chairs: TBA The increasing globalization of the world's economies demands that engineers practice internationally and that barriers to their mobility across national boundaries are elimi-

nated or, at least, minimized. Many countries impose licensing requirements on engineers to ensure that they have the appropriate education and experience for practice. Two sessions are planned, consisting of papers and panel discussions, to examine means of increasing the mobility of engineers (especially for countries with stringent licensing requirements such as Canada and the USA) and of widening the recognition of highquality engineering programs offered in countries with and without accreditation systems. The sessions will conclude with identifying special initiatives for members of TMS and related organizations regarding international mobility and accreditation.

B11

MATERIALS AND TECHNOLOGIES FOR MICROELECTRONICS I Sponsored by: EMPMD Electronic Packaging and Interconnection Materials Committee Program Organizers: M. McCormack, AT&T Bell Laboratories, Rm IA-123, 600 Mountain Ave., Murray Hill, NJ 07974; S. K. Kang IBM, T. J. Watson Research Center, P. O. Box 218, Yorktown Heights, NY 10598; M. R. Notis, Department of Materials Science and Engineering, Lehigh University, 5 East Packer Ave., Whitaker Laboratory, Bethlehem, PA 18015-3195; K. N. Tn, Department of Materials Science and Engineering, UCLA, Los Angeles, CA 90095-1595 Monday, AM February 5,1996

Room: Grand K Location: Anaheim Marriott Hotel

Session Chairperson: M. R. Notis, Department of Materials Science and Engineering, Lehigh University, 5 East Packer Ave., Whitaker Laboratory, Bethlebem, PA 18015-3195 8:30 am Invited ELECTRICALLY CONDUCTING ADHESIVES FOR MICROELECTRONIC PACKAGINGAPPLICATIONS:S.K. Kang, IBM,T. J. Watson Research Center, PO Box 218, Yorktown Heights, NY 10598 Solder interconnection technology is in need of alternatives to address not only environmental issues associated with lead (Pb) abatement and elimination of fluxes and flux cleaning solvents, but also the technical challenges related to fine pitch assembly and low temperatore processing. In the electronic industry, two different groups of materials are being investigated currently for the possibility of substituting the Pb-containing solder materials; Pb-free solder alloys and electrically conducting adhesives. The present paper discusses the development and applications of electrically conducting adhesive materials. An electrically conducting adhesive is made of metallic filler particles dispersed in the matrix of a polymer resin. Depending on the mode of electroconduction, electrically conducting adbesivescan becategorized into two groups; isotropic and anisotropic materials. Silver-filled epoxy material used in die attach applications is the most common example of an isotropic material. Anisotropic conducting film (ACF) used in liquid crystal display (LCD) applications belongs to the category of anisotropic conducting material. In this talk, both materials are discussed in terms of their current status, trends, and cballenges for further development. 8:55 am Invited ANISOTROPICALLY CONDUCTIVE INTERCONNECTION MATERIALS FOR MICROELECTRONIC PACKAGING: S. Jin, M. McCormack, AT&T Bell Laboratories, Murray Hill, NJ 07974 The enormous imbalance between the micron resolution at the IC device level and the hundreds-of-micron features at the interconnectionlpackaging level leads to ineffecient device integration. In order to make the electronic devices and products smaller, faster, and cheaper, a substantial increase in interconnection density is necessary. In this paper, some recent progress in the development of new, anisotropic conductive materials for high density interconnections will be described. The possibility of interconnection schemes using micro-solder arrangement, as well as metalpolymer composites, will also be discussed. 9:20 am Invited DEVELOPMENT OF NEW SOLDER BONDING MATERIALS FOR ELECTRONIC PACKAGING APPLICATIONS: R. Emigh, J. Li, M. Pinter, Johnson Matthey Electronics, Spokane, WA 99216 New solder based bonding materials (Pb containing and Pb-free) are being developed for advanced electronic applications. These are targeted for use where reduced temperature and low stress bonding is required. Wark is focused on developing materials that can tolerate subsequent reflow temperatores that are above the initial bonding temperature. The materials are also designed with enhanced thermal conductivity and fatigue resistance. 9:45 am Invited MICROSTRUCTURAL MECHANISMS OF ELECTROMIGRATION FAIL· URE IN AL-CU THIN·FILM CONDUCTING LINES: J. W. Morris, Jr., C. Kim, S.H. Kang, Department of Materials Science, University of California Berkeley, and Center for Advanced Materials, Lawrence Berkeley Laboratory, CA 94720 This paper discusses microstructural control of intemal electromigration failure in narrow AI-Cu thin film conduction lines on Si. The lines of interest have near-bamboo structures; their microstructores consists of polygranular segments that are terminated by "bamboo" grains that span the line. The common mode of electromigration failure mode is a transgranular separation of the bamboo grain that terminates its longest polygranular segment. The polygranular segment and terminating grain are swept free of Cu, which is gathered into a Cu precipitate at the opposite end of the segment. The time-to-failure increases exponentially with the inverse of the segment length. These observations suggest three ways for increasing the lifetime of narrow AI-Cu-Si lines; low-temperatore pre-aging to stabilize the distribution ofCu, high-temperature pre-aging to narrow the distribution of polygranular segment lengths, and periodic current reversal to reverse Cu flow. All three methods have been tried with some success. This work was supported by the Director, Office of Energy Research, Office of Basic Energy Sciences, U.S. Department of Energy, under Contract No. DE-AC03-76SF00098.

10:25 am Invited EFFECT OF THERMAL AGING ON THE MICROSTRUCTURE AND RELI· ABILITY OF BALL GRnD ARRAY (BGA) SOLDER JOINTS: E. E. Bradley, P. Lall, K. Banerji, Motorola Advanced Manufacturing Technology Center, Land Mobile Products Sector, Ft. Lauderdale, FL 33322 A major cause of solder joint failures in electronic assemblies is due to the high strain rate effects of bending. Sin,ce solder is very strain-rate sensitive, solder joint failures caused during high strain rate deformation of the printed circuit board are often located at the interfaces where the so.lder bas wetted to the PCB metallization. The formation and growth of intermetallics at these interfaces can substantially impact package-level reliability. This study investigated the solder joint reliability of Ball Grid Array (BGA) packages under three point bending as a function of aging. Parts were aged between 75°C and 125°C for up to 1000 hours and then reflowed onto printed circuit boards using a conventional surface mount profile. The relationship of the failure modes in three point bending to interfacial microstructural development will be discussed. 10:50 am Invited MATERIALS·BASED MODEL TO PREDICT SOLDER JOINT LIFETIME IN ADVANCED ELECTRONIC PACKAGES: D. R. Frear, S. N. Burchett, M. K. Neilsen, Sandia National Laboratories, Albuquerque, NM 87185 The long term reliability of solder joints in electronic packaging is a critical competitiveness issue. The reliability of an electronic system is frequently limited by the failure of soldered electrical interconnections due to thermomechanical fatigue. As an added complication, thermollll:chanical fatigue tests at Sandia have clearly sbown that the microstructure of the solder undergoes a heterogeneous coarsening process followed by failure through the coarsened regions. A microstructurally based constitutive model bas been developed at Sandia to simulate thermomechanical response including microstructural evolution leading to prediction of the reliability and lifetime of solder joints. The metallurgical inputs for this model are: the constitutive relations of the solder in mathematical form and a quantitative relation of the solder microstructure as it evolves during thermomechanical fatigue. The microstructural state variables are then incorporated into the temperature-dependent, viscoplastic constitutive framework in a finite-element code. A time integration algorithm for the complete constitutive model is implemented into JAC3D, a three-dimensional finite-element code. The output of the simulation of three dimensional stress and strain and the solder microstructural evolution will be presented for a variety of solder jOint designs. We present the metallurgical results that are used in the fmite element model and simulation results of the model. Computational and experimental results will be presented that examine the effects of a the variety ofjoint structures used in advanced packages ranging from leadless surface mount, to flip chip, to ball grid array. This work was performed at Sandia National Laboratories, whicb is supported by the U.S. Dept. of Energy undercontractnumberDE-AC04-94AL85000. 11:15 am ADHESION OF TWN TANTALUM NITRIDE FILMS ON Alz03 AND ALN SUBSTRATES: N.R. Moody, D. Medlin, D. Norwood, Sandia National Laboratories, Livermore, CA 94551-0969; S. Venkataramant, Applied Materials, Santa Clara, CA 95054 Thin tantalum nitride films are often used in microelectronic applications because of their long term stability and low thermal coefficients of resistance. however, the lack of microstructure and adhesion data of these films on A~03 and ALN substrates limits our ability to assess reliability. We therefore combined nanoindentation, continuous microscratch testing, and higb resolution transmission electron microscopy to determine structure-property relationships in thin tantalum nitride films on single crystal sapphire and polished aluminum nitride substrates. The films were sputter-deposited to a thickness of 100 nm on eacb type of substrate and then tested at room temperature. The results revealed that there was a difference in the structure, morphology, and properties of the films. In this presentation, the results will be discussed and used to show how interface structure and film morphology affects the resistance to fracture of thin tantalum nitride films. This work supported by U.S. DOE Contract DE-AC04-94AL85000. 11:40 am INFLUENCE OF LASER ANNEALING ON ELECTROPHYSICAL CHARAC· TERISTICS METAL·SILICON CONTACT UNDER IRRADIATION IN CARBON·CONTAINING LIQUIDS: A. M. Chaplanov, A. N. Sbibko, Institute of Electronics BAS, 220090, 22, Logoi:lkii Teakt, Minsk, Belarus Mucb attention bas been given to physico-chemical processes which occur in thin films under irradiation with highpowered laser beams owing to a wide use of such films in microelectronics. Peculiarities of pbase composition and electrophysical properties of titanium-silicon and vanadium-silicon compositions under laser treatment in hydrocarbon-containing environment have been studied. It is found tbat the use of certain aromatic hydrocarbons (toluene, styrene, etc.) favors synthesis of metal carbides. The use of benzene does not result in carbide synthesis, even though it causes structural changes in the films. Benzene serves as a protective medium against laser damage and formation of carbides in the films. Laser annealing of T-Si, V-Si systems in aromatic bydrocarbon medium results in changes of electrophysical parameters of the contacts due to changes in the phase composition and interface.

10:10 am BREAK

B12

1996 TMS ANNUAL MEETING AND EXHIBITION

MATERIALS PROCESSING FUNDAMENTALS SYMPOSIUM I

Sponsored by: Jt. EPDIMDMD Synthesis, Control, and Analysis In Materials Processing Committee, EPD Process Fundamentals, Aqueous Processing, Copper, Nlckel·Cobalt, Pyrometallurgy, Lead, Zinc, Tin Committees, MSD

Thermodynamic & Phase EquUlbrla Committee Program Organizers: R.G. Reddy, Department of Chemical and Metallurgical Engineering, University of Nevada, Reno NV 89557; S. Viswanathan, Oak Ridge National Lab., Oak Ridge, TN 37831·6083; J.e. Malas, Wright-Patterson AFB, OH 45433·6533 Monday, AM February 5, 1996

Room: A16-17 Location: Anaheim Convention Center

Session Chairpersons: R. G. Reddy, Department of Chemical and Metallurgical Engineering, University of Nevada, Reno NV 89557; E. J. Grimsey, Department of Mineral Engineering and Extractive Metallurgy, Western Australian School of Mines, Kalgoorlie, Western Australia 6430 8:30am SULFIDE CAPACITIES OF CaO·AlOl,! AND CaO·FeO-S102 SLAGS: R. G. Reddy, W. Zhao, Department of Chemical and Metallurgical Engineering, University of Nevada, Reno, NV 89557 The sulfide capacities (Cs) of CaO-AIOI.' melts were experimentally de~ermined using encapsulation method at 1773 K. The Cs ofCaO·AlOI.' and CaO-FeO-Sl~ melts were calculated a priori using the Reddy· Blander model. The calculated Cs results are in very good agreement with the experimental data. The sulfur distribution ratio (sulfur in slag/sulfur in metal) in CaO·FeO·Si~ slags as a function of slag composition was calculated. The results were used to explain the sulfur distribution in several industrial slag systems. 8:!!am MODELING OF SULPHIDE CAPACITIES OF SOME SILICATE MELTS: R. Nilsson, Du Sichen, S. Seetharaman, Division of Theoretical Metallurgy, Royal Institute of Technology, S·IOO 44 Stockholm, Sweden In the present work, an assessment of the sulphide capacities of a numher of binary and ternary slag systems have been carried out in the light of this model. The model is based on the expression: C, =exp (·AG'!R1) oa.;l·lfs2· where AG' is the standard Gibbs energy change for the above reaction, a.;l- is the activity of oxygen ions and fs~' is the activity coefficient of the sulphide ion in the slag. In the present model, the ratio (a.;lIfs2-) is optimized for the experimental data available in literature. The systems assessed are the binaries MnO-Si02, CaO·Si02, Al20:l-CaO, as well as the ternaries CaO-Al20 3• Si~ and CaO-MnO-Si~. In the case of all these systems, the experimental data available in literature are complemented by measurements carried out at the Division of Theoretical Metallurgy. The sulphide capacities of some FeO- containing binary and ternary slags are also examined. The potentiality of the model to estimate the sulphide capacities of complex slag systems from binary and ternary data is demonstrated. 9:20am REDUCTION OF MnO·FeO-SIOzSLAGS BY CARBON-SATURATEDLIQUID IRON: Jo~ R. de Oliveira, Jorge A. S. Tenorio, Department of Metallurgical and Materials Engineering, University of Sio Paulo 05508, Slio Paulo, Brazil 05508 The aim of the present work was to investigate the reduction process of MnO-FeOSi02slags by carbon-saturated liquid iron. The effect of the initial Mn content in the bath and also the effect of the initial Si02content were explored. The initial Mn content was in the range of 0 t04Owt.% and the Si~content was in the range of 14 to 26 wt.%. The MoO content in the slag was kept constant at 40 wt%. The tests were performed in laboratory apparatus. Samples of slags and bath were taken. In the fICSt five minutes a very strong reduction of the FeO was observed. This first step promoted a MnO·Si~ enrichment of the slag and formation of CO bubbles, which increased the slag level. The increase of Mn and Si02 decreased the MnO reduction rate. 9:!!am REDUCTION OF MnO PELLETS BY CARBON·SATURATED LIQUID IRON: Jo~ R. de Oliveira, Jorge A. S. Tenorio, Department of Metallurgical and Materials Engineering, University of Sio Paulo, Slio Paulo. Brazil 05508 The goal of this work was to investigate the reduction process of MoO pellets by carbon-saturated liquid iron. The experiments were performed in a laboratory apparatus. The tests were performed at 1500 and 1600'C. The initial Mo content in the bath was in the range of 0 to 4Owt%. Two sizes of pellets were tested. The reaction time was observed by the evolution of CO. 10:20 am BREAK 10:30 am REMOVAL OF As AND Sb FROM MOLTEN COPPER BY NazC03 FLUXING: Baozhong Zhao, Nickolas J. Themelis, Deparment of Chemical Engineering, Materials Science and Mining Engineering, Columbia University, New York, NY 10027 Research to-date has shown that fluxing with soda ash (Na2C03) can remove As and Sb from molten copper at the rate of As>Sb, but is relatively ineffective in removing Bi. Oxygen in the copper is essential for soda ash reactions. If the oxygen content in the copper is reduced after soda ash fluxing. the impurity in the soda ash slag can return to the copper. In this paper, the influences of various factors, such as the initial oxygen content in the copper, the flow rate of carrier gas, the soda ash feed rate, and the effects of soda ash fluxing by surface addition and injection, on the rate ofimpurity removal are

1996 TMS ANNUAL MEETING AND EXHIBITION

discussed. In the light of experimental results, a mathematical model was developed to describe the impurity removal phenomena. 10:!!am DISTRIBUTION OF Se AND Te BETWEEN MATTE AND SLAG IN NICKEL SULFIDE SMELTING: N. Choi, W. D. Cho, Department of Metallurgical Engineer· ing, University of Utah, Salt Lake City, UT 84112 The distribution behavior of selenium and tellurium in nickel sulfide smelting was investigated in the temperature range of 1250-1300'C under an inert gas atmosphere and various oxygen partial pressures. The distribution coefficients of selenium and tellurium between nickel-copper-iron matte and silica-saturated iron silicate slag were determined in terms of matte and slag compositions, temperatue, and oxygen partial pressure. The effect ofslag additives such as CaO, MgO, and Ah03 on the distribution coefficients of these elements was also investigated. It was observed that the distribution coefficients of selenium and tellurium increased with increasing matte grade and temperature. The solubility of selenium and tellurium in the slag decreased with increasing partial pressure of oxygen. The distribution coefficients of selenium and tellurium increased with the amount of CaO and MgO. Tellurium distribution coefficients also increased by the addition of A120 3. Meanwhile, the addition of AI20 3decreased the distribution coeffi· cients of selenium at high-matte grade. 11:20 am KINETICS OF REACTIONS BETWEEN IRON OXIDES AND LIQUID IRON· CARBON ALLOYS: Marcelo B. Mouriio, Cassio A. Rodriguez, Ricardo S. Maltez, Department of Metallurgical and Materials Engineering, Politechnic School, University of Sio Paulo, 05508-900 Sio Paulo, SP, Brazil An experimental investigation was carried out to study the rates of reduction of pellets containing iron oxides by carbon dissolved in liquid iron. at temperatures ranging from 1720 K to 1973 K. Several different pellet types were employed; the materials used in the pellet's manufacture were pure hematite, two kinds of commercial iron ores, and mixtures hematite/silica and hematiteJlime. Reaction times were determined employing the constant volume pressure increase technique (CVPI). It has been observed that the reaction rate increases as the temperature and the pellet's specific area increases, and as the pellet's size decreases. The reaction rate is also affected by the ore type and by the kind of compound formed between hematite and other oxide; pellets containing calcium oxide react faster than the ones containing silicon oxide.

MELT.SPINNING, STRIP CASTING, AND SLAB CASTING I: Emerging Technologies and Industrial Applications Sponsored by: MDMD SoIldlftcation and SCAMP Committees Program Organizers: E.F. Matthys, Mechanical Engineering Department, University of California, Santa Barbara, CA 93106; W.G. Truckner, Technical Director--Product Development, Alcoa Technical Center, Alcoa Center, PA 15069 Monday, AM February 5,1996

Room:B3 Location: Anaheim Convention Center

Session Chairperson: A. Cramb, Carnegie Mellon University, Dept of Matis Sci, Pittsburgh, PA 15213; E.F. Matthys, University of California, Dept of Mech Engrg. Santa Barbara, CA 93106 8:30 am Invited CONTINUOUS CASTING OF ALUMINUM· AN ALCOA PERSPECTIVE: STATUS AND POTENTIAL FOR PROGRESS: P.N. Anyalebechi. R. Bachowski, S.F. Baumann, O. Richmond, AJ. Sartschev, W.G. Truckner, Alcoa Technical Center. Alcoa Center, PA 15069 Over the past 25 years, continuous casting is the AI industry has grown steadily, although this growth has been mostly confined to "common a1loylproducts" such as foil, conductor rod, building products, selected food packaging allOYS, etc. This growth has been driven essentially from process economics (compared to ingot based metal). Development of products requiring higher levels of performance such as beverage can stock, automotive sheet, surface sensitive products, etc. has been difficult. This difficulty can he traced to the lack of understanding of the complex processes occurring during continuous casting. Experience over the years has provided some valuable information; however, much more understanding is required to answer metallurgical and surface concerns. This paper will address, from Alcoa's perspective, these problems and will be discussed and presented in terms of research opportunities that need to be solved before continuous cast products hegin to compete from both a performance and economic standpoint in the higher performance markets now served by ingot based processes. 9:10am AN UPDATE ON THIN·GAUGFJHIGH-SPEED ROLL CASTING OF FIN AND FOIL PRODUCTS: Bill Carey, Norandal USA,Inc., Brentwood, TN. Bruno Taraglio. Fata Hunter, Inc., Torino, Italy; Chris Romanowski, Fata Hunter, Inc., Riverside, CA 92507 The cast house at Norandal's Huntingdon facility has been expanded to accommodate a fourth FATA Hunter Casting line. The new line features a 2184 mm (86") wide SpeedCasterTII. scheduled for start-up in late 1995. This installation will be the world's first production thin-gauge/high-speed twin roll caster for aluminum. The specifications for each of the major pieces of equipment in this installation are reviewed in the written paper and the presentation will include a video showing the operation of the casting line. An overview of the start·up program and initial performance of the caster in terms of productivity, product quality and up-time will also be presented. The written paper outlines and compares the investment cost for a new foil plant based on this thin-gauge! high-speed casting technology with the cost of a conventional SuperCastere plant.

B13

9:35am FASTCAST - THE FIRST OF A NEW GENERATION OF ROLL CASTERS: P.M. Thomas. Davy International. Poole. BHI2 5AG. United Kingdom Over the past few years. high speed thin strip casting has gained credibility as an emerging technology. It has been established that the process diffe~ from con~entio~a1 roll casting in a number of key areas and offers a number of potential benefits mcludmg reduced capital and operating costs. increased produ~tivity and freedom from the a1~oy limitations associated with conventional roll casting. As a result of an extensive development program on both laboratory and pilot scale casters. the importance of separating force as a process variable has been established. Based on the knowledge gained during the experimental program, it has been possible to engine~r a cas';Cr for commercial operation. This machine differs fundamentally from conventional twm roll casters in a number of ways. To accommodate the high separating forces necessary to product good quality material at thi? gauges tIM; casti.ng roll~ are sup,,?rted.by back .up rolls in a 4 - High construction. This construction Will prOVide matenal With supenor gauge and profile tolerances and with a microstructure: co~atible with ~owns~am processing requirements. The frrst FASTCAST ~c~e IS currently bel~g .bul.lt f~r Granges Eurofoil for the production of 1800mm Wide foil stock and cODl1DlsslOmng IS scheduled for November. 1995. 10:00 am BREAK 10:20 am PROCESS AUTOMATION FOR HIGH SPEED - TmN-GAUGE - TWIN ROLL CASTING TECHNOLOGY: K.P. Maiwald. B. Mariethoz. Lauener Engineering LTD. CH-3645 GwatllIbun. Switzerland The successful joint venture development between Lauener-Engineering and HydroAluminum in respect to thin gauge-high speed casting demanded a higbly sophisticated process-control system. The targets of 100% (and above) productivity increases. and gauges of2mm and below within tolerances of +/- I % and below w~ reached on time. The paper will explore details of the process control system appbed from furnace to caster and coiler.

10:45 am DEVELOPMENT AND COMMERCIALIZATION OF A 3000 SERIES BODY STOCK ALLOY USING THE LAUENER BLOCK CASTER: B. Newton. Golden Aluminum. Golden. CO 80401 The paper describes the start up experiences at the San Antonio Mill - Texas and follows alloy and product development which lead to the commercialization ofGolden's Body Stock. The paper examines the transition from the initial 5349 alloy to the current 3000 series alloy. 11:10 am ELECTRO-SLAG REFINING AS A LIQUID METAL SOURCE FOR MELTSPINNING, STRIP CASTING AND SLAB CASTING: M.G. Benz. W.T. Carter. RJ. Zabala. B.A. Knudsen. P.L. Dupree. GE Corporate Research and Development. Schenectady. NY 12301; R.M. Forbes-Jones. Teledyne Allvac. Monroe. NC 28110 Electro-slag refining (ESR) is a process used to melt Fe-. Ni-. or Ti- base alloys by passing an electric current through an electrode of the alloy partially immersed in a liquid slag. The liquid slag is maintained at a temperature high enough to melt the end of the electrode by this current. As the electrode melts. a refining action takes place - oxide inclusions are exposed to the slag and are dissolved. Droplets of molten metal fall through the slag and are collected in a liquid metal pool contained in a water-cooledcopper crucible below. By addition of an induction-heated. segmented. water-cooled copper guide tube to the bottom of the crucible. a liquid metal stream can be extracted from the liquid metal pool. rather than allowing it to sol~dify as an ingot.. ~s ~tream makes an ideal liquid metal source for: powder atomization. spray deposttlon. mve~t­ ment casting. melt-spinning. strip casting. and slab casting. Results from a 15 kg/mID. pilot-scale system are discussed. 11:35 am MELTSPINNING OF SIC REINFORCED ALUMINIUM COMPOSITES: L. Katgerman. N. Fei. W.H. Kool. Laboratory of Materials. Delft University of Technology. Delft. The Netherlands . . AI MMCs have been prepared by compocasting and subsequent meltspmnmg. The effect of rapid solidification process parameters on the microstructure and particle distribution in the ribbons have been investigated. SiC particles tend to segregate to the air-side of the ribbons and the segregation effect is influenced by particle size and volume fraction. Particle pushing models have been extended with a boundary layer model for fluid flow and heat transfer in the puddle to give valuable insight into the nature of the SiC segregation.

PHYSICAL CHEMISTRY OF HIGH TEMPERATURE COMPOSITES AND INTERMETALLICS I: Interfaces in Composites Sponsored by: EPD Process Fundamentals Committee, MSD Thermodynamics & Phase EquDlbria Committee, Japan Institute of Metals Program Organizers: Prof. R.Y. Lin. University of Cincinnati; Prof. Y. Austin Chang. University of Wisconsin-Madison; Prof. R. Reddy. University of NavadaReno and Dr. C.T. Liu. Oak IUdge NL Monday. AM February 5.1996

Room: B2 Location: Anaheim Convention Center

Session Chairperson: Ray Y. Lin. Dept. Materials Sci. & Eng .• University of Cincinnati. M.L. #12. Cincinnati. OH 45221-0012; Li Lu. Dept. Mech. & Production Engineering. The National University of Singapore. 10 Kent Ridge Crescent. Singapore OSIII 8:30 am Keynote Address IN-SITU COMPOSITE SYNTHESIS OF METAL SILICIDE REINFORCED WITH SILICON CARBIDE VIA SOLID STATE DISPLACEMENT REACTIONS: C. R. Kao. Y. A. Chang. Department of Material Science and Engineering. University of Wisconsin. Madison. WI 53706 The principle of thermodynamics and kinetics which governs solid-state displacement reactions to synthesis structural composites in-situ will be briefly described. Application of these principles will be illustrated using experiments being carried ?ut a! Madison to synthesize NbSiiSiC composite by reacting NbC powder compact WIth SI single crystal. Diffusion couple experiments consisting NbClSi are being carried out at 1300=BOC for tirnes=up to 100b. The resulting microstructure is a layer of continuos NbSi2 matrix dispersed with SiC particle with an average size of Imm. The rationale .for occurrence of this type of microstructure is discussed as well as our proposed reaction mechanism. Growth of the reaction layer involves long-range diffusion ofSi through this layer and short-range diffusion ofNb and C around the reaction front. The particles sizes of SiC appear to be controlled by the voids in the NbC powder compact. 9:00 am Invited CRITERIA FOR MATRIX SELECTION IN CONTINUOUS FIBER ALUMINUM MATRIX COMPOSITES: Colin McCullough. 3M Company. MMC Program. S!. Paul. MN 55144 Traditionally Continuous Fiber Aluminum Matrix Composites have been fabricated using high strength fibers and high strength matrices. often con;unercial alloys. wi~ the goal of producing high strength composites. However the speCIfic role of the matrix on all areas of composite performance have not been well established. One way to control composite properties. which will be: shown. is to use coatings to introduce a low strength interface between fiber and matrix which tends to reduce the importance of the matrix selection. but unfortunately renders the transverse strength unacceptably low. Thus using a system with strong bonding between fiber and matrix is desirable. such as an alumina fiber/aluminum alloy mabix system. In this case the transverse strength is a strong function of the matrix strength. However the longitudinal strength the~ de~nds largely on the characteristics of the matrix. The important role of the m.atrix wdl be discussed using examples of a1uminll fihers in various aluminum alloy matrices. The key attributes are matrix constituent phllses. alloy chemistry.a1loy reactivity with the fiber and matrix strength. From the understanding developed in these systems a set of design rules may be established for selecting an optimum matrix. 9:25 am Invited SYNTHESIS AND CHARACTERIZATION OF LOW SILICA MoSI2 COMPOSITES USING IN-SITU DISPLACEMENT REACTIONS: M. J. Kaufman. A. Costa e Silva. S. Jayashanar. M.J. Kaufman. Department of Materials Science and Engineering. University of Florida. Gainesville. FL 32611 . . Previously. we have shown that MoSirmatrix composites can be synthestzed usmg in-situ displacement reactions involving carbon (to form SiC) and aluminum (to form A1203) additions. These reactions lead to the simultaneous reduction and elimination of silica from the micro-structures. In the present study. we have extended these efforts to produce MoSirboride (MoB. M02Bs and TiB2) composites using a combination of the lower silicides and boron. For the boride-reinforced materials. the resulting microstructures are fine scale and consist of interpenetrating silicide and boride phases. In this presentation the microstructural evolution in these composites will be. Our characterization of the mechanical properties of the low-silica material indicate that both silica and grain size are extremely important in determining the mechanical properties of these composites and that careful control of the processing strategies allow what appears to be superplastic deformation at small grain sizes and relatively high creep strengths at larger grain sizes. This work was supported by the Office of Naval Research. Grant No. NOOOI4-91-J4132. under the direction of Dr. Steve Fishman. 9:50 Invited IDENTIFICATION OF THE INTERFACE FAILURE MECHANISM DURING TRANSVERSE LOADING OF SCS-6ll'I-6AI-4V COMPOSITES: Sunil G. Warrier. Systran Corporation. Dayton. OH; Bhaskar S. Majumdar.UES. Inc .• D~yton. OH; Dani~l B. Miracle. Materials Directorate. Air Force Wright Laboratory. Wnght-Patterson Atr Force Base. Dayton. OH 45433 During transverse testing of continuous fiber-reinforced composites. along with the increase in the normal stress parallel to the loading axis. large tangential shear stresses develop at the interface at about 45° to the loading axis. Depending on the properties of the interface. these shear stresses may initiate interface failure. Once frictional sliding

B14

1996 TMS ANNUAL MEETING AND EXHIBITION

has initiated, either the interface can lose a part of its load carrying capacity or lower the nonnal strength of the interface. It is therefore necessary to establish a fundamental understanding of the governing interface failure mechanism in these composites. In order to clearly quantify the tangential shear strength of the interface, shear stresses have to be isolated from mechanically induced normal stresses. In the present study, the interface has been subjected to pure shearstresses through torsion testing. Result indicate that in SCS6ro-6A\-4V composites shear failure is likely to have preceded nonna! failure of the interface during transverse loading. Detailed evaluation of the micromechanical aspects of the interface response to torsion testing will be presented. 10:15 am BREAK 10:15 am Invited PERFORMANCE OF BARRIER COATINGS FOR ALz0 3 FIBERS IN TITANIUM ALLOYS: Rob Kieschke, J.R. Heffelfinger, C.B. Carter, 3M Company, MMC Program, St.Paul, MN 55144 Abstract not available. 10:50am . KINETICSOFINTERFACIAL REACTION IN COMPOSITES: AIlInconelInconel 601. A. V. Ruzette, F. Boland, A. Colin, F. Delannay, Universite catholiquede Louvain, Departement des sciences des materiaux et des procedes, PCIM Place Sainte Barbe, PCIM Place Sainte Barbe; 2, B-1 348 Louvain-Ia-Neuve, Belgium Reinforcement of AI-base alloys with continuos Inconel601 fibers is studied as a means for enhancing high temperature properties. The volume fraction of the fibers was varied between twenty and sixty volume percent. The composites were processed by squeeze casting. Adequate ductility requires that the extent of interfacial reaction be minimized. A low thermal inertia squeeze casting set up was designed in order to achieve rapid solidification of the metal after inftltration of the preform. The severity of reaction was controlled by varying the preheating temperature of the mold and of the fiber preform and the mass and temperature of the liquid metal cast in the mold. The influence of these parameters was monitored by measuring the temperature evolution from the moment of infiltration. The nature and amount of reaction compounds was characterized by EPMA and image analysis. The reaction kinetics was modeled assuming a parabolic law. The influence of fiber heat treatment, fiber volume fraction, and silicon content of the matrix alloy on reactivity was also elucidated. 11:15 am Invited INTERFACIAL CHARACTERlSTICS OF SQUEEZE-CAST SIC-REINFORCED AZ91D MAGNESIUM-BASED COMPOSITES: Jason La, Graham Carpenter, MTIlCANMET, Department of Natural Resources Canada, 568 Booth St., Ottawa, Ontario KIAOOI, Canada The method of producing magnesium-based composites with silicon carbide (SiC) particulates embedded in the matrix of magnesium AZ9ID alloy by the squeeze casting technique are described. This method involves the fabrication of a SiC particulate preform and the subsequent squeezing of molten magnesium into the preform. In this study, alumina, phosphate and aproprietary binder were used for the fabrication of SiC particulate preforms. The procedures employed in the fabrication of SiCfAZ91D composites are presented, together with the results of characterization of the composites using scanning and transmission electron microscopy. Special attention is given to the reaction products in the bulk alloy and at the interfaces between SiC particles and the AZ9ID matrix alloy especially when different binders were employed. The significance of the presence of some reaction products in relation to their effects on mechanical properties are discussed.

11 :40 am Invited THERMAL CYCLING OF ALz03FIBER.REINFORCEDSUPERALLOY COM· POSITES: Gopal Das, Robert J. Miller, D. Wheeler, Pratt & Whitney, P. O. Box 109600, W. Palm Beach, FL 33410-9600 The coefficient of thermal expansion (erE) mismatch between the fiber and matrix is known to cause damage in a composite following thermal cycling, the severity of which intensifies with increasing erE mismatch. Two superalloy matrix, HA 214 and MA 956 with significantly different erE were unidirectionally reinforced with single crystal AI 20] fibers (Saphicon) and subjected to thermal cycling between room temperature and 1800 F in air. The SaphiconfHA 214 composite bar started to exhibit bending after 50 cycles while the SaphiconfMA956 composite showed negligible bending after 400 cycles. Thermal cycling in air was accompanied by a substantial weight gain (1.77%) in SaphiconfHA 214 composite and a negligible weight gain (0.197%) in Saphiconl MA956 composite after 400 cycles. Microstructural characterization of thennally cycled SaphiconfHA 214 composite bar revealed twinning in fibers and with increasing cyclic exposure all these fibers appeared cracked. In contrast, fibers in Saphicon!MA 956 composite did not reveal any twinning or cracking. The presence of both oxi.des and nitrides were detected in the fiberfmatrix interface of thermally cycled composItes by a combination of x-ray, SEM and electron microprobe studies. The observation of twinning and cracks in fibers will be discussed in terms of residual stresses developed during cooling and heating of the composite.

1996 TMS ANNUAL MEETING AND EXHIBITION

PHYSICAL PROPERTIES OF COMPOSITES I:

Sponsored by: Jr. SMD·MSD Composite Materials Committee Program Organizer: P.K. Liaw, Materials Science and Engineering Department, The University of Tennessee, Knoxville, TN, 37996-2200; R. Pitchumani, Mechanical Engineering Department, University of Connecticut, Storrs, cr 06269-3139; S.G. Fishman, Office of Naval Research, 800 N. Quincy Street, Arlington, VA 22217 Monday, AM February 5,1996

Room: Marquis 1&2 Location: Anaheim Marriott Hotel

Session Chairperson: R. Pitchumani, Mechanical Engineering Department, University of Connecticut, Storrs, cr 06269-3139; P.K. Liaw, Materials Science and Engineering, The University of Tennessee, Knoxville, TN, 37996-2200 8:30am PROPERTY·MICROSTRUCTURE·PROCESSRELATIONSHlPSOFTEXTILE STRUCTURAL COMPOSITES: Tsu-Wei Chou, Jerzy L. Nowinski, Professor of Mechanical Engineering and Materials Science, Center for Composite Materials and Department of Mechanical Engineering, University of Delaware, Newark, DE 19716 The term "Textile structural composites" is used to identify a class of advanced composites utilizing fiber performs produced by textile forming techniques for structural and functional applications. The recent interest in textile structural composites stems from the need for improvements in intra-and inter-laminar strength and damage tolerance. Textile composites offer the potential of providing adequate structural integrity as well as shapeability fomear-net-shape manufacturing. This presentation will review recent advances in the processingffabrication, microstructural design/analysis and performance characterization of textile composites. The relationship among processing, microstructure and property is demonstrated by (a) a multi-step braiding process for fabricating three-dimenstional preforms with controlled microstructure, (b) thermal and electrical property modeling and analysis of two-dimensional woven fabric composites, and (c) non-linear behavior and damage evolution in two- and three- dimensional brittle matrix composites. Textile composites based upon polymer and ceramic matrix materials with fiber preforms produced by weaving, braiding and knitting will be examined. 9:00am INVESTIGATIONS ON EDDY CURRENT EVALUATION OF METAL MA· TRIX COMPOSITES: Ranga Pirchumani, Mechanical Engineering Department University of Connecticut, Storrs, cr 06269-3139; Peter K. Liaw, Materials Science and Engineering Department, University of Tennessee, Knoxville, TN 37996-2200 Particulate-reinforced metal matrix composites are being considered as strong candidates for use in many engineering applications. Widespread deployment of these materials, however, requires reliable material qualification techniques. Nondestructive evaluation using ultrasonics or eddy currents provides a viable means of characterizing the composites and their properties. In this regard, considerable theoretical and experimental investigations have been carried out by the authors in recent years on the application of eddy current technqiues to evaluate metal matrix composites. Studies include development of predictive models for the anisotropic composite condustivities. for binary and multiphase composites; development of a technique for nondestructive inspection of product quality in terms of the constituent phase concentrations; and eddy current measurements on a wide array of A1fSiCp composites extrusions. A summary of the research accomplishments will be presented and discussed. 9:30am THERMAL AND ELECTRICAL CONDUCTIVITY OF COMPOSITES, CAL· CULATED USING SIMPLE UNIT CELL MODELS: K.S. Ravichandran, Department of Metallurgical Engineering 412 WBB, The University of Utah, Salt Lake City, UT84112 Prediction of thermophysical properties including thennal and electrical conductivity of composites is of interest in the contexts of their use in high temperature structural, electrical and electronic applications. In the present research, simple microstructurebased unit cell models will be used to predict the thermal and electrical conductivities of composites having discrete second phases in continuous matrix materials. This method is based on the divisions of the unit cell into parallel and series configurations of two phases. Comparisons of predicted trends with experimental data on metal-metal, metal-ceramic and ceramic-ceramic composites will be made. The experimental thermal conductivity data of epoxy-graphite, Zr02-Ni, Zr02-Mo as well as electrical conductivity data of several eutectic alloys including Bi-Bi2Pb, Pb-Mg2Pb and pseudoalloys such as Cu-Pb, Cu-Fe will be used for comparison. The accuracy of the present method will also be evaluated in the light of other methods. 10:00 am MODULI OF CONTINUOUS FIBER REINFORCED CERAMIC COMPOS· ITES: P.K. Liaw, N. Miriyala, Department of Materials Science and Engineering, The University of Tennessee, Knoxville, TN 37996-2200; N. Yu, Department of Mechanical and Aerospace Enginering and Engineering Science, The University of Tennessee, Knoxville, TN 37996-2030; X. Mao, Department of Mechanical Engineering, University of Calgary, Calgary, ABT2NIN4, Canada; D.K. Hsu, Center for NDE, Iowa State University, Ames, IA 50011 The moduli of woven Nicalon fiber fabric reinforced Alp] matrix composites have been investigated. Both through-thickness and in-plane (fiber fabric plane) moduli were measured using ultrasonic techniques. The through-thickness moduli were found to be much less than the in-plane moduli. Increased porosity significantly decreased both inplane and through-thickness moduli. An analytical model using a homogenization method was formulated to predict the effect of porosity on the moduli of woven fabric composites. Moreover, numerical modeling work was formulated to predict the influence of porosity on the moduli. The predicted results using both analytical and numerical

B15

techniques were found to be in good agreement with the experimental data. The present wad is supported by the Oak Ridge National Laboratory. the Department of Energy under contract Nos. Martin Marietta IlX-SNI9IV and SL261V. 10:30 am MODELING OF YOUNG'S MODULUS OF COMPOSITES REINFORCED WITH 3-D ORIENTED SHORT FIBERS: Y.T. Zhu. W.R. Blumenthal. M.G. Stout. T.C. Lowe. Materials Science and Technology Division. Los Alamos National Laboratory. Los Alamos. NM 87545 Young's modulus is one of the most important design parameters in applications of composite materials. This paper develops a theory for calculating the Young's modulus of composites reinforced with three-dimensionally (3-D) oriented short fibers. Assuming a composite strain in the loading direction. the fiber stress and its ineffective length with different orientations are first calculated as a function of E. The total load carried by a composite sample is calculated by integrating the load contributions from these short fibers and the matrix. The composite stress and Young's modulus are subsequently derived. Comparisons with other models show the advantages of our new approach. 11:00 am YOUNG'S MODULUS MEASUREMENT OF 10800. FILM USING DYNAMIC AND STATIC METHODS: Youngman Kim. R&D Center. Korea Gas Corp. Ansan. Korea; Min-Tae Kim. KIA Motors. Seoul. Korea ITO(lnSnO.) thin film was produced on plate glass substrates by reactive sputtering under various conditions for automobile applications. Young's modulus of the ITO films was measured using the sonic resonance method and strain gages. For the dynamic Young's modulus measurement (sonic resonance method) a simple beam vibration theory was used to obtain film modulus with the knowledge of the substrate modulus values. For the static method (strain gage method) strain gages were attached to the specimen surfaces. parallel to the beam axis. The specimens were then bent in a fourpoint bend fixture with a dead load. The elastic moduli were determined from the strains recorded. The perfect bonding. linear elastiCity and no stress relaxation were assumed throughout this study.

RECYCLING· GENERAL

Sponsored By: Jt. LMDIEPD Rec:ydlog Committee Program Organizer: Ms. PragnaN.H. Bhakta. U.S. Bureau of Mines. 810 7th St.. MS 6204. Washington. D.C. 20241 Monday. AM February 5. 1996

Room:A6 Location: Anaheim Convention Center

Session Chairperson: Pragna Bhakta. U.S. Bureau of Mines. 810 7th St.. MS 6204. Washington. D.C. 20241 8:30am SOLID ELECTROLYTE BASED SENSOR FOR MONITORING THE MAGNESlUM LEVEL DURING RECLAMATION OF ALUMINUM SCRAP: Jeffrey W. Fergus. Shiqiang Hui. Materials Research and Education Center. 20 I Ross Hall. Auburn University. AL 36849 Aluminum alloy scrap often contains excess magnesium which must be removed during recycling by a process referred to as "demagging". The efficiency of this process could be improved with an in-situ magnesium sensor. which could be used to optimize the process parameters to the changing magnesium content. The sensor developed in this work consists of galvanic cell with a magnesium fluoride (MgFz) solid electrolyte and a molten magnesium reference electrode. The voltage output of the sensor changes by about 100 mV for the change in magnesium contents which occur during the demagging process (5 wt% to 0.1 wt%) and is in excellent agreement with thermodynamic measurements using molten chloride electrolytes. Although the best response is provided using a molten magnesium reference electrode. a solid reference electrode could be used to produce a completely solid-state sensor. 8:55am RECOVERY AND SEPARATION OF HIGH VALUE PLASTICS FROM DISCARDED HOUSEHOLD APPLIANCES: D. E. Karvelas. BJ. Jody. B. Arman. J. Pomykala and EJ. Daniels. Argonne National Laboratory. 9700 South Cass Avenue. Argonne.IL 60439 This paper presents the results of research. being conducted at Argonne National Laboratory. to develop a cost-effective and environmentally acceptable process for the separation of high value plastics from discarded household appliances. The process under development has separated high-purity (greater than 99.5%) acrylonitrile-butadiene-styrene from commingled plastics generated by appliance-shredding and metalrecovery operations. The process consists of size-reduction steps for the commingled plastics. followed by a series of gravity-separation techniques to separate materials of different densities. Individual plastics of similar densities are further separated by USing a chemical solution. By controlling the surface tension of the chemical solution. we are able to selectively float plastic materials that have different surface energies. This separation technique has proven to be highly effective in producing high-purity plastic materials from household appliances. A conceptual design of a continuous process for separating high-value plastics from discarded household appliances will be presented. In addition to plastics separation research. Argonne National Laboratory has also been conducting research to develop cost-effective techniques for improving the mechanical properties of the recovered plastics. Results from this research will also be presented.

B16

9:20am PROGRESS IN RECYCLING OF AUTOMOBILE SHREDDER RESIDUE: B.J. Jody and E. J. Daniels•.r\rgonne National Laboratory. 9700 South Cass Avenue. Argonne. IL. 60439 Argonne National Laboratory has been developing a process to recycle automobile shredder residue (ASR). We built and tested a field demonstration plant for recycling polyurethane foam. and produced several thousand pounds of-recyclable foam. Several 300 pound samples were sent for evaluation. and were found to be of marketable quality. We are also in the process ctf preparing for a large scale test in which about 200 tons of fines (<0.25 inch) derived flrom ASR. will be used as a raw material in cement making. Small samples of fines prepared in the laboratory were evaluated by a major cement company and were found to meet their requirements. as a substitute for iron ore or mill scale. This paper presents a brief discussion of the process as a whole. and summarizes the results obtained from tlK~ field work on foam and fines recycling. 9:45am EUREKA-Project: PACK·EE: G. Nijhof, Y. Bertaud. R. Guillermet, Hoogovens Groep. P.O. Box 10000. 1970 CA Umuiden. The Netherlands A large research project with this title has been started in June 1993. The project carries out research and innovation into collection. soding and valorisation of types of packaging. with improvements to sorting procedures capabie of meeting the needs of the recovery and recycling industries. The project includes studies of consumer behaviour concerning wastes. which will influence the choice of sorting techniques. Materials involved are: aluminium. glass. paper. and plastic. In the Aluminium Group the participants are: Hoogovens Groep. Pechiney and VAW. The following subjects of research will be discussed; I) improvements of the collection of waste packagings. 2) improvements of the Eddy Current equipment 3) separation optimization and efficiency. 4) economic balance.

10:10 am BREAK 10:30 am EVALUATION OF RECYC.LABILITY OF USED AUTOMOTIVE OIL FILTERS AS FERROUS SCRAl': Darrell E. Roberts. Kent D. Peaslee. University of Missouri-Rolla. Department of Metallurgical Engineering. 218 McNutt. Rolla. MO 65401 Over 425 million used automotive oil filters are discarded in the U.S. each year. Only about 10% are currently recycled as ferrous scrap and the remainder are typically landfliled. Oil filter scrap is not lx~ing used by many operations because of concerns with scrap consistency and quality. Consistent scrap quality depends both on the content of the used oil filters and the processing technique used to produce the scrap. This paper investigates the content ofused oil filters and compares the recyclability of different oil filter designs and manufacturing methods. Included is a model that predicts the scrap yield. chemistry. and energy effects of various oil filter deSigns in combination with different recycling methods. 10:55 am INVESTIGATIONS ABOUT THERMAL SEPARATION OF ALUMINUM PACKAGING: B. Filion. Y. Bertaud. Y. Caratini. Pechiney CRY. BP 27. 38340 Voreppe-France. J.C. Templier. H.. Rossei. and D. Mackenstedt. YAW AluminumAG. Postfach 2468. 530 14 Bonn. Germany Aluminium packaging scrap are combined with a lot of other materials. like plastics. paper. varnish. glue. colors and ink. In addition the scrap can be contaminated with residues of the packaged goods (cosmetics. food etc.). Most of the non aluminium components would be removed by incineration during the melting procedure. The local increase in temperature resulting from the combustion process and mixing with carbonaceous residues decreasss the metal yield and produces non aceptable emissions of hydrocarbons. Low temperature pymlysis. i.e. endothermic cracking of organics in short gaseous chains recovered and/or burnt (energy recovery). is the only solution. The investigation has been divided in design experiments and model calculations and pilot scale experiments. The research is largely focused on controlling the thermal cracking profile to avoid the formation of soot while degrading a maximum number of compounds at low temperatures. Special attention is being paid to all technological requirements. mechanical pretreatment, monitoring. automation of the thermal steps. collection and treatments of the gases given off by the tbermal degradation of the organics. U:2Oam PURIFICATION OF REMELTED ALUMINUM PACKAGING: G. Nijhof, H.M van der Dook. Y. Bertaud. S. Bouvet. Hoogovens Groep. P.O. Box 10000. 1970 CA Umuiden. The Netherlands Remelting of aluminium requires only 5% of the energy required for primary smelting. Therefore it is important to keep aluminium in the materials cycle as long as possible. Making new packaging from production scrap or waste packaging is thus very important. Remelted scrap may contain an amount of undesired alloying elements. this amount must be reduced. State of the art is to mix in molten scrap with purest material to reach the required composition. Future aim is to use less primary metal. thus technologies for the removal of these undesired elements. particularly iron. are necessary. Investigations are focused of intermetallic compounds. by adding other elements to the melt it is possible to form intermetallic compounds. which can be removed from the melt by: gravity segregation. partial solidification or filtering. Having accurate knowledge of the thermodynamic reactions and the phase diagrams. it is possible to predict the amount ofmaterial to be added to achieve a final composition. The results will be discussed. 11:40 am LEAD REMOVAL FROM BRASS AND BRONZE SCRAP: Frederick J. Dudek. Dwight R. Diercks. Edward J. Daniels. Argonne National Laboratory. 9700 S. Cass Avenue. Argonne.lL 60439 The brass and bronze ingot manufacturing industry melts lead-bearing copper-base scrap and casts specification alloy ingots for sale and conversion into a wide range of

1996 TMS ANNUAL MEETING AND EXHIBITION

products, including plumbing fixtures. Under proposed regulations limiting the lead content of plumbing fixtures, the industry could not use c~rrent technology to rec~c1e a large fraction of the scrap stream. To provide for the contmued and most econormc use oflead bearing scrap, a process is needed that selectively removes lead from copper-base melts andlor scrap. Several alternatives for selective lead removal from copper based scrap are liquation, slagging, precipitation of intermetallics, vacuum distillation, and chemical leaching of selected portions of the scrap stream. This paper reviews an effort to identify and assess the technical and economic feasibility of options for selective lead removal from brass and bronze melts andlor scrap. Results of experimental screening are presented.

models have produced results in the early stages of growth in good agreement with experiments for a number of different quantities, including the island density and island·size distribution. For multilayer growth, however, there exists a competition between microscopic effects such as the Ehrlich-Schwoebel step barrier and the crystalline microstructure. Therefore, the crystal structure and geometry are important in determining the evolution of epitaxial morphology. We present the results of large-scale realistic kinetic Monte Carlo simulations of multilayer epitaxial on fcc(loo) and bcc( 100) surfaces. The influence of crystal structure on the formation and coarsening of mounds and facets will be discussed. We also discuss and compare our results with recent experiments. 10:20 am BREAK

STRUCTURE AND MORPHOLOGY OF EPITAXIAL THIN FILMS SESSION I: Homoepitaxy·From the Submonolayer to Multilayer

Sponsored by: EMPMD Thin FUms & Interfa~s Com~ttee Program Organizer: Dr. David E. Jesson, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831-6030 Monday, AM Room: Orange County 4 February 5, 1996 Location: Anaheim Marriott Hotel

Session Chairperson: Andrew Zangwill, School of Physics, Georgia Tech., Atlanta, GA30332 8:30 am Invited DYNAMICS OF ISLAND GROWTH DURING SUBMONOLAYER EPITAXY: G. S. Bales, Computational Materials Science Department, Sandia National laboratories, Uvermore, CA 94550 Under ideal conditions, the size distribution of islands which grow during the early stages of epitaxial growth obeys a simple scaling form. The distribution is specified completely by the average size and a ·universal· scaling functio~. Moreover, t.IJe ~tal number of islands has a simple power law dependence on the ratio of surface dIffUSIOn rate D and deposition rate F. These simple scaling rules apply to infinitely flat surfaces with no defects or impurities, fixed critical island size, and values of D/F greater than about 1()6.ln this worle, a formalism is presented for understanding the scaling behavior when the above conditions are not satisfied. The shape of the distribution function and the value of the number density become dependent on a limited set of coverage dependent "dynamical state variables· which reflect the various competing processes which occur during growth. A systematic study using Kinetic Monte Carlo simulations and self-consistent rate equations identifies the important variables. The results are compared to experimental data. Worlc supported by the Office of Basic Energy Sciences of the Department of Energy. 9:00am MORPHOLOGY OF THIN FILMS AT INITIAL STAGES OF GROWTH: MONTE CARLO SIMULATIONS: Ashok Chol/a, Center for Solid State Science, Arizona State University, Tempe, AZ 85287-1704; Timothy S. Cale, Center for Solid State Electronics Research, Arizona State University, Tempe, AZ 85287-6206 The growth mechanism and resulting atomic morphology at initial stages of epitaxial growth of thin films by vapor phase depositic.m. ~s investigated with Mo~te Carlo simulations using a solid-on-solid model. The Imtial stage of film growth IS clearly divided into two stages: a nucleation stage and a growth stage. Thin film growth can be characterizedas2Dlayer-by-layergrowthor3Dislandgrowth.Wehavedevelopeda3D island growth model to study the growth morphology of the thin films at sub-monolayer coverages. The model incorporates the potential barriers near edges of the island that suppress the diffusion of adatoms from upper layer to lower layer and vice versa. The role of these potential barriers on the island morphology is investigated. We also examine the influence of reversible and irreversible attacbrnent and restructuring of adatoms at the island edges on the 3D island growth morphology. The island morphologies obtained from the simulations are compared to the experimentally observed morphologies. 9:20 am Invited ROUGHENING ANDSMOOTmNG KINETICS OF Ge(OOI): David G. Cahill, S. Jay Chey, Joseph E. Van Nostrand, Department of Materials Science, University of Illinois, 1101 W.Springfield Ave., Urbana, IL 6180 I We use in-situ scanning tunneling microscopy to characterize the rough surface morphologies of Ge(ool) produced by MBE growth and low-energy ion etching at low temperatures. For crystal growth, a regular pattern of multi-layer growth mounds develops over a wide range of temperatures, 6O-200°C. A similar pattern of etch pits is formed during ion-sputtering by 240 eV Xe ions at 250 and 275°C. In both cases, crystal growth and etching, the lateral length scale of the roughn~ss evolves .directly f~om ~ submonolayer island separation. We attribute the roughemng mechanIsm to a diffUSIOn bias for surface defects- adatoms during growth, and surface vacancies during sputtering. We have also studied the relaxation of multilayer roughness by thermal annealing at 225,250 and 275°C. Despite the relatively low annealing temperatures and short lateral length scales studied in our experiments, the data are consistent with the Herring-Mullins (continuum) deSCription of surface smoothing. 9:50 am Invited EFFECTS OF CRYSTAL MICROSTRUCTURE ON EPITAXIAL GROWTH: Jacques G. Amar, Fereydoon Family, Physics Department, Emory University, Atlanta, GA30322 In the past, simulations of epitaxial growth have used solid-on-solid (SOS) models to simulate the crystalline structure of both the substrate and the growing crystal. These

1996 TMS ANNUAL MEETING AND EXHIBITION

ISSUES IN METAL EPITAXY Session Chairperson: R. F. C. Farrow, IBM Research Division, Almaden Research Center, 650 Harry Road, San Jose, CA 95120-6099 10:40 am Invited THE ROLE OF PLACE EXCHANGE AND ADATOM INCORPORATION IN METAL EPITAXY: R. J. Behm, J. A. Meyer, Abt. Oberflachenchemie und Katalyse, Universitat Ulm, D-89069 Ulm, Germany In simple models, epitaxial growth is limited to processes on top of the respective surface layer, and the substrate serves as a passive support. In a rapidly growing number of cases, however, it turned out that the situation is less ideal. Incorporation of diffusing metal adatoms into the surface layer and place exchange with metal atoms in that layer can play an important role in metal epitaxy. These processes can strongly affect both nucleation and growth behavior of adislands and hence roughness and morphology ofthe resulting film (i) by modifying the density of diffusing adatoms, (ii) by providing nucleation sites for subsequent heterogeneous nucleation and (iii) by creating an adlayer of mixed composition. These processes are identified from recent scanning tunneling microscopy data, which also allow an estimate of the kinetic barriers for the respective processes. 11:10 am InvIted THE STRUCTURE OF CLOSE·PACKED SURFACES: C. B. Carter, Department of Chemical Engineering and Materials Science, Amundson Hall, University of Minnesota, Minneapolis, MN 55455; R. Q. Hwang, J. C. Hamilton, Sandia National laboratories, Livermore, CA 94551 Several different structures have been observed for the close-packed III-fcc surfaces of pure materials and when layers ofa second element are grown on aOOOI-hcp substrate. We will review both the experimental observations of the phenomenon and computer modeling studies. The strain may be relaxed locally in different ways that involving threading dislocations, misfit Shockley partial dislocations, and both rotation and buckling of the surface layer. We will then describe and use the cubic-hexagonal (CH) model for the structured III surface to show the relationship between the different observed structures which include trigon arrays, node networks and the herringbone pattern. Real surfaces tend to show mixtures and distortions of these different structures. We will relate the different terminologies which have been used to describe the same phenomenon and relate these concepts to the structure of III heterojunctions in thin films. 11:40 am Invited STM STUDIES OF INTERMIXING AND ALLOY FORMATION IN ULTRATmNMETALFILMS:D.D.Chambliss,IBMResearchDivision,650Harry Road, San Jose, CA 95120-6099 Intermixing can play an important role in determining the morphology of epitaxial thin films. Conversely, the structure of the substrate can control the stability of alloys formed when deposited materials react with each other or with the substrate. Using the scanning tunneling microscope (STM) we have found that ultrathin films of Au on Cu(loo), Fe on Cu(IOO), and Ag and Co on Mo(llO) demonstrate many aspects of this interplay. Au/Cu(loo) forms a surface alloy analogous to bulk Cu3Au, modi~ed to relieve misfit strain. Fe/Cu(loo) tends to intermix at the monolayer level, deSPIte the miscibility gap for bulk Fe and Cu. This inJermixing leads to substantial roughness of the Fe/Cu interface. Ag and Co, though immiscible in bulk, form an intermixed "quasialloy· phase when codeposited on Mo( 110), to establish an epitaxial fit with the substrate. This metastable phase is replaced at high temperatures by a phase separation between pure Ag and a Co-Mo surface alloy. 12:10 pm STM STUDIES OF THE STRUCTURE AND GROWTH DYNAMICS 2·D METAL ALLOYS: A. K. Schmid, J. L. Stevens, R. O. Hwang, Sandia National Laboratories,Livermore, CA 94551-0969 Strain has been recently shown to playa key role in determining not only the structure of thin metal films, but also their composition. This is especially true in mUlti-component films and superlattices. To investigate the effects of strain on such films, we have chosen to study monolayer binary mixtures of AglCu, AglCo, and Co/Cu on a Ru(OOO 1) substrate. These metals all exhibit a substantial miscibility gap in their bulk binary phase diagrams. However on the Ru(OOOI) substrate, strain induced by the lattice mismatch gives rise to mixing behavior substantially different than the bulk. In this talk we will describe our STM experiments on the structure and the dynamics of formation of these alloy combinations. New 2-D phase diagrams are found for the various combinations of metals, which are strongly driven by strain relief. Furthermore, growth morphology of these alloys differ significantly from those of the single component films. These differences point to a variety of new atomistic processes not found in the growth of the single component case.This work is supported by the Office of Basic Energy Sciences, Divsion of Materials Science of the US DOE under Contract DE-AC04-94AL85000.

B17

TANTALUM: MINING, EXTRACTION AND REFINING: SESSION I

Sponsored by: SMD Refractory Metals and Materials Committee and Jt. MDMDIEPD Synthesis, Control and Analysis In Materials Processing Program Organizers: Andrew Crowson, U.S. Army Research Office, Research Triangle Park, NC; Edward S. Chen, U.S. Army Research Office, Research Triangle Park, NC; Prabhat Kumar, Cabot Corp, Boyertown, PA; William Ebihara, Picatinny Arsenal, Picatinny, NJ; Enrique J. Lavernia, UC Irvine, Irvine, CA Monday, AM February 5,1996

Room: A4-5 Location: Anaheim Convention Center

Session Chairpersons: Prabhat Kumar, Cabot Corp, Boyertown, PA; Robert H. Tritran, NASA Lewis Research Center, Cleveland, OH 9:00am TANTALUM-AN OVERVIEW: George J. Korinek. Tantalum-Niobium International Study Center, Rue Washington 40,1050 Brussels, Belgium Tantalum in nature is closely related to niobium and tin. The raw material basis is provided by tantalite and tantalum containing tin slags. Structural changes in the tin industry affected very seriously the production of tin slags, so that tantalite became by far the most important resource. These developments will be discussed in detail. The tantalum industry is of relative modest size. The total world consumption in 1994 was 2.2 million Ibs. Ta. Over 60% ofTa was used in electronics in the form of powder, wire and furnace hardware for the production of solid electrolyte capacitors. Cemented carbides, equipment for the chemical industry and other uses in mill product form and addition to super alloys are the other main applications. A detailed discussion of the development of the individual applications will be given. 9:30am BENEFICIATION OF TANTALUM ORE-HOW IS IT ACHIEVED AND COULD IT BE BETTER?: Richard Burl, General Manager, Tantalum Mining Corporation of Canada Limited Typically a tantalum ore contains less than I kg of tantalum in every tone ore; beneficiation of the ore is therefore an essential step in the production of tantalum 'This paper briefly reviews the various tantalum minerals, and studies the mining and beneficiation routes used to concentrate the ore, citing examples from mines and mills throughout the world. The paper also looks at some newer techniques that have been tried, attempting to extend the envelope of recovery, and thereby improve production. The failures are noted; the successes highlighted. 9:50am THE PRODUCTION OF TANTALUM BY THE SODIUM REDUCTION PROCESS: Dr. Terrance B. Tripp, H.C. Starck Inc., 45 Industrial Place, Newton, MA021611951 Tantalum metal was first produced in Germany on a conunercial scale in the early 1900's by the reduction of potassium heptaf!uorotantalate with sodium. Today, virtually all of the tantalum metal produced in the world is obtained by variations of a stirred reactor version of this process. The powder is isolated in the form of a powder usually with properties designed to meet the requirements of the solid tantalum capacitor industry or capacitor grade wire production. 'This paper will briefly trace the history of tantalum production and describe how potassium heptafluorotantalate is reduced to tantalum metal with sodium in a molten salt system using a stirred reactor. The procedures for converting the primary powder to a form suitable for use as a capacitor grade powder or wire and furnace parts feedstock will be described. Examples of the performance of typical tantalum powders available today will be presented. 10:10 am BREAK 10:30 am

SPREADSHEET SIMULATION OF A TANTALUMlNIOBIUM SEPARATION CIRCUIT: Donald VOil, Process Consultant, 4444 Parkview Dr., Ogden UT 84403 A spreadsheet has been developed that simulates the behavior of a tantalum/niobium solvent extraction circuit. The circuit consists of a niobium extractor, a solvent wash section, a tantalum extractor, a niobium scrubber, and a tantalum stripper. The simulation requires selection of the number of stages in each section and specification of feed composition, flows, and distribution coefficients. The simulation calculates compositions at each node in the circuit, chemical consumptions, yields and product quality. Reasonable choices for input variables and their impact on the system are discussed. Literature distribution behavior for the 2-methyl-4-pentanone (mibk) -H2S04-HF aqueous syslem are reviewed. Regularities in the data perrnitdistribution coefficient behavior prediction over a wide range of chemical environments. The logarithm of the niobium distribution coefficient increases linearly with the acidity. The tantalum/niobium separation factors decreases with niobium distribution coefficient. 10:50 am

.. STUDIES ON TANTALUM EXTRACTION: A.K. Suri, T.K. Mukherjee, C.K. Gupta; Materials Group, Bhabha Atomic Research Centre; Trombay, Bombay 400 085, India 'This paper provides a brief account of research and development program on tantalum extraction at Bhabha Atomic Research Centre. In India, sources of tantalum are rather limited. It is known to occur as columbite-tantalite in association with mica pegmatites, with deposits of tin as columbite-tantalite and also in thc crystal lattice of cassiterits. Besides these, some quantity of niobium and tantalum is generated as sludge dnring the processing of cemented carbide tools. Processes have been developed to treat all these sources to prepare pure K2TaF7, and Ta20s. Extensive studies have been carried out and

B18

number of processes have been developed to prepare pure tantalum using each of these compounds. While fused salt electrorowinning and sodium reduction techniques have been developed for the reduction of K2TaF7, reduction of oxide has been accomplished with calcium, aluminum, carbon and carbon-nitrogen. Final purification of metal, if necessary, was carried out generally by electron beam melt refining. An alternate purification route based 011 fused salt electrorefining has also been developed. 11:10 am Nb-Ta ALLOYS BY TIlE ALUMINOTHERMIC REDUCTION OF (NblOsl TalO!) MIXTURES AND ELECTRON BEAM MELTING: Carlos A. Nunes, University of Wisconsin-Madison11500 Eng. Drive, 53106 Madison WI, on leave from FAENQUIL-DEMARI Polo Urbo-Ind. s/n-126OO-000, Lorena, SP, Brazil; Daltro G. Pinatti, Alain L.M. Robin, F AENQUIL-DEMARI Polo Urbo-Industrial s/n-126()()..000, Lorena, SP, Brazil The a1uminotherrnic reduction (ATR) of Nb20 s followed by the electron beam melting (EBM) of the ATR -'llectrodes has been successfully applied to the conunercial production of pure niobium. In this work we will discuss the technical viability for obtention of Nb-Ta alloys by the same route. We will also present results on the application of metal-gas equilibrium (N)/steady-state (0) equations to estimate the limits rorrefining Nb, Ta and Nb-Taalloys in terms of oxygen and nitrogen, during EBM under practical vacuum conditions. Finally, it will be shown results of chemical! microstructural analysis and metallic yield on the obtainment of an ingot (- 20 kg) of the Nb-20 wt% Ta alloy by the previously mentioned route.

11:30 am THE INDUSTRIAL APPLICATION OF PYROMETALLURGY, Cin.ORlNATION AND HYDROMETALLURGY FOR PRODUCING TANTALUM COMPOUNDS: Dr. Joachim Eckert, H.C. Starck GmbH & Co. KG, 1m Schleeke 18-91, 038642 Goslar, Germany The modern industrial production of tantalum compounds, like potassium heptafluorotantalate (K2TaF7), is not restricted to the wet chemical processing of tantalum bearing ores, but includes the pyrometallurgical upgrading of tantalum containing slags and residues as well as the chlorination of ferroalloys and ores. Considerable progress in producing tantalum compounds by hydrometallurgy was made in the early seventies. Since then, this technique has been continuously modified and improved. Today, tantalum compounds of high purity are manufactured in large quantities for use as starting materials for the production of tantalum metal. Recent process improvements now make it possible to utilize raw materials with extremely unfavorable TalNb ratios, thus assuring a stable supply of tantalum for the foreseeable future. Using actual flow charts, the state-of-the-art of the different techniques for producing tantalum compounds are described and discussed in dtltail.

TREATMENT AND MINIMIZATION OF ALUMINUM AND LIGHT METALS INDUSTRY WASTES I

Sponsored by: Jt. LMDIEPD Recycling Committee, EPD Waste Treatment & Minimization Committee Program Organizer. Donald L. Stewart, Jr., Alcoa Technical Center, 100 Technical Drive, Alcoa Center, PA 15069; James A. Clum, Binghamton University, Watson School of Engineering, PO Box 6()()(), Binghamton, NY 13902-6000 Monday, AM Room: A1 February 5, 1996 Location: Anaheim Convention Center Session Chairperson: Donald L. Slewart, Jr., Alcoa Technical Center, 100 Technical Drive, Alcoa Center, PA 15069; Jatnes A. Clum, Binghamton University, Watson School of Engineering, PO Box 6000, Binghamton, NY 13902-6000 8:30am INTERFACIAL REACTION BE1WEEN AL ALLOYS AND SALT FLUX DURING MELTING: M. Mehdi Taghiei, University of Science & Technology, Materials Science Department, Narrnak, TehClm, Iran A ternary mixture of KQ NaCI and KF may be used as a molten flux cover during melting of AI scrap. 'This practice mini.mizes oxidation of the liquid metal and allows the existing oxide layer to be stripped from the scrap more readily. The coalescence of metal droplets dispersed in molten salt was found to be influenced by the formation of a highmelting layer that forms at the interface between the molten alloys and the salt. This layer was identified as KMgF3 in some cases and as the mineral Elpasolite, K2NaAIF6in other instances. Either forms appear to depend on conditions that exist in the melt at the time of formation. There was no indication of any changes in the structure or chemical composition of the layer that occur as a function of time of contact between metal and salt. The presence of metallic potassium and sodium in the solidified AI was detected by SEM microanalysis of the sample. 8:55am EFFECT OF IMPURITIES IN INDlISTRlAL SALTS ON ALUMINUM SCRAP MELTING: Jian Ye, Yogeshwar Saha;, Department of Materials Science and Engineering, The Ohio State University, Columbus, OH 43210; Annette Revet, Kalium Canada, Ltd., Regina, Sask., Canada S4P 4B5 Aluminum scrap such as Used Beverage Containers (UBC) are melted under a protective molten salt cover. An appropriate salt protects metal from oxidation, promotes coalescence of molten droplets, and separates clean metal from the oxide contamination. Generally, the salt compositions for aluminum scrap recycling are based on equi-molar mixtures of NaCI and KCI. A small amount of fluoride is also added in the salt. In the past, laboratory research at universities and industrial laboratories has been limited to pure salts. However, the industrial salts have impurities such as sulfates and other

1996 TMS ANNUAL MEETING AND EXHIBITION

insoluble materials. These impurities, particularly tbe presence of sulfates in salt, have a pronounced effect on the efficiency of the scrap remelting process. In tbis paper, tbe role of impurities in industrial salts in terms of their chemical interactions witb the metal are summarized. The efficiency of different industrial grade salts containing varying amounts of sulfates and other insoluble impurities for scrap recycling is compared. 9:20am ALUMINUM RECYCLING USING OXYGEN COMBUSTION - FUNDAMENTAL AND PRACTICAL ASPECTS FROM CONVERTING AN AIR-FUEL BURNER TO AN AIR-COOLED OXYGEN-FUEL BURNER IN A ROTARY FURNACE: Torbjorn Dahl, Canadian Liquid Air Ltd., 6880 Loius-H. Lafontaine, Anjou (Montreal), Qul!bec, Canada HIM 2T2 This paper will present practical aspects and results from a major Canadian recycler of aluminum dross and scrap (A1can Recycling, Guelph, Ontario, Canada), which successfully converted (1994) one of its two rotary furnaces to oxygen combustion using an air-cooled oxygen-fuel (natural gas) burner and related combustion-technology (Canadian LiqUid Air). Recent production improvements (over 30% productivityincrease), together witb data on melt-time, fuel consumption, firing rates, recovery, and the positive impact this conversion had on the off-gases and omitted particulates (baghouse loading), will be presented. Some of the furnace modifications in relation to tbe oxygen-conversion will also be mentioned. The air-cooled oxygen-fuel burner and tbe control system and principles of operation will be presented. In order to explain the improvements, the theory behind oxygen combustion, especially in a rotary (dross) furnace, will be presented (graphs, formulas). This will include how heat is transferred to tbe furnace (conduction, convection, radiation) and tbe thermal balance for tbe furnace when the amount of inert nitrogen is decreased as the oxygen content is increased. References will also be made to other similar operations, including recycling of otber nonferrous metals; e.g., lead, copper. 9:45am THE PERFORMANCE OF CURRENT OXY-FUEL COMBUSTION TECHNOLOGY FOR NONFERROUS MELTING: Randy Abernathy, Jim McElroy, Loo Tjay Yap, BOC Gases, 575 Mountain Ave., Murray Hill, NJ 07974 Higher tbermal efficiencies, lower exhaust-gas flows, exhaust particulate loading, superstructure temperatures, and NO. emissions have contributed to more widespread cost-effective use of oxygen-based firing. Operating characteristics of oxy-fuel fired furnaces are compared witb air-fired furnaces for the secondary melting of nonferrous metals. Typical performance gains resulting from oxy-fuel firing are established for reverberatory and rotary aluminum furnaces, reverberatory and rotary lead furnaces, as well as rotary brass furnaces. Representative examples are presented in detail. The large improvements in thermal performance of a novel self-cooling oxy-fuel burner providing uniform heat-transfer to large melt areas are demonstrated for aluminum furnaces. 10:10 am BREAK 10:20 am

THE HOT ALUMINUM DROSS RECYCLING (HDR) SYSTEM: Han Spoel. Spalco Metals inc., 210 MacPherson Ave .• Toronto, Ontario, Canada MSR IW8; William A. Zebedee, MICA Metals Inc., 500 North Woodward Ave., Suite 350, Bloomfield Hills, MI 48304 The difficulties previously encountered by aluminum dross recyclers in directly processing hot dross from a nearby generating facility have been overcome. Dross is skimmed as usual into pots from the melting, batching, or holding furnaces. These pots are then placed in hot dross holders, and enough hot dross is accumulated for a full charge of tbe rotary processing furnace. The hot dross is then dumped into an accumulator and charged, together witb such flux as is needed, as rapidly as possible into the rotary furnace, where it is processed in the usual way. The first HDR system became fully operational at MICA Metals, Bedford, Indiana, in March 1995. There has been a substantial improvement in metal recovery. Usage of salt flux, energy consumption, batch times, and fugitive emissions have all been reduced. 10:45 am SAFETY ASPECTS OF THE DESIGN AND OPERATION OF ROTARY ALUMINUM DROSS COOLERS: Han Spoel. Spalco Metals Inc., 210 MacPherson Ave., Toronto, Ontario, Canada M5R IW8 Rotary aluminum dross coolers are used to cool hot dross skimmed from aluminum melting, batching and holding furnaces. The dross is introduced in batch mode into one end of an unlined steel drum which rotates about the longitudinal axis. Cooled dross is discharged from the otberend of the drum. The outside of the drum is water-cooled, and tbe cooled shell in tum extracts heat from the dross. The avoid dangerous conditions, the equipment must be designed and operated in such a way tbat hot aluminum dross and water do not come into contact witb each other. 11:10 am

RECOVERY OF ALUMINUM FROM ROTARY FURNACE SALT CAKE BY LOW IMPACT ROTARY TUMBLING: David J. Roth, ALTEK International Inc., 585 Exton Commons, Exton, PA 19341 The goal of secondary dross recyclers is to maximize the aluminum units recovered from the dross they process and minimize the amount of material tbey have to send to a landfill. Salt Cake is a by-product of the most rotary furnace processors. Altbough there are systems available to totally recycle this material, the financial aspects of the process can be debated depending on the volume of salt cake generated by the facility. Normally from 5-10% aluminum is contained in tbis salt cake by weight. For a generator of only a million pounds a montb of salt cake, even tbis amount of aluminum is a significant number. Before now the equipment to extract the aluminum from tbe salt cake has been very expensive, difficult to maintain, and taken up a lot of real estate. Most processors have not gone witb tbis option and all of tbe material is sent to tbe landfill. The Tumbler as manufactured by Didion and ALTEK separates the aluminum from the salt cake in an efficient manner. The equipment is simple and easy to run in difficult environments. We

1996 TMS ANNUAL MEETING AND EXHIBITION

will review the operation of tbis type of system on typical salt cake. White, black dross, and batb are also materials tbat are processed by tbis system and will be touched on in tbe paper. 11:35 am THE ECONOMICS OF SALT CAKE RECYCLING: D. Graziano,]. N. Hryn, E. J. Daniels, Argonne National Laboratory, Energy Systems Division, 9700 Soutb Cass Avenue, Building 362, Argonne, IL 60439-4815 The Process Evaluation Section at Argonne National Laboratory has a major program aimed at developing cost-effective technologies for salt cake recycling. This paper addresses tbe economic feasibility of technologies for tbe recovery of aluminum, salt, and residue-oxide fractions from salt cake. Three processes were assessed for salt recovery from salt cake: (i) base case: leaching in water at 25·C, with evaporation to crystallize salts; (ii) high Tlhigh P case: leaching in water at 250·C, witb flash crystallization to precipitate salts; and (iii) solvent/anti-solvent case: leaching in water at 25·C, concentrating by evaporation and reaction witb acetone to precipitate salts. Base case economics were evaluated at two leaching effluent brine concentrations, 12 and 22 wt%. All test cases for salt recovery had a negative present value, given current priCing structure and 20% return on investment. Altbough manufacturing costs (variables plus fixed) could reasonably be recovered in tbe sales price of the salt product, capital costs cannot. The economics fortberecycling processes are improved,however,ifthe residueoxide can be sold instead of landfilled. For example, the base case process would be profitable at a wet oxide value of 10 cents/lb. The economics of alternative scenarios were also considered, including aluminum recovery witb landfilling of salts and oxides, and aluminum and residue-oxide recovery witb discharging of concentrated salt solutions. The economics associated witb additional processing of residue-oxide were also evaluated.

MONDAY PM ADVANCES IN THE SCIENCE AND TECHNOLOGY OF TITANIUM ALLOY PROCESSING II: Hot Working II

Sponsored by: SMD Titanium Committee, MDMD Shaping and Forming

Committee

Program Organizers: Prof. Isaac Weiss and Prof. Raghavan Srinivasan, Mechanical and Materials Engineering Dept., Wright State University, Dayton, OH 45435; Dr. Paul Bania, Timet Corporation, Timet-Henderson Technical Laboratory, PO Box 2128, Henderson, NV 89009; Prof. Daniel Eylon, Graduate Materials Engineering, University of Dayton, Dayton, OH 45409 Monday, PM February 5, 1996

Room: B5-6 Location: Anaheim Convention Center

Session Chairpersons: S.L. Semiatin, Wright Laboratory, Materials Directorate, WU MLLN, Wright-Patterson Air Force Base, OH 45433-7817; I. Weiss, Mechanical and Materials Engineering Dept., Wright State University, Dayton, OH 45435 2:00 pm Invited MEETING THE CHALLENGE OF HOT WORKING TITANIUM ALLOYS INTO COST EFFICIENT FINISHED COMPONENTS: G.W. Kuhlman, Alcoa Forged Products, 1600 Harvard Ave., Cleveland, OH 44105 Titanium alloys, especially advanced a1uminides, are among the most difficult to fabricate engineering materials. Management of titanium tbermomechanical processing (TMP) in hot working processes (forging, rolling. extrusion) has become a skillfully executed science, utilizing sophisticated process engineering. modeling and computer design and control technologies, rather than relying solely upon art or the experience of an operator. At the heart of overcoming tbe challenges of cost-efficient titanium component fabrication is tbe growing body of science and knowledge available on the interactions between required Ti alloy TMP for critical properties and tbe hot working processes. Using sophisticated deformation simulators, artificial intelligence and expert systems, FEM and otber modeling technologies, and state-of-the-art computer controlled equipment, tbe titanium forger, roller, or extruder has unparalleled power at his disposal in terms of deptb of deformation understanding, accurate process models and equipment capabilities. Witb these tools, fabrication of titanium alloy components is being efficiently accomplished. Reviewed in tbis paper is tbe state-of-the-art in key TMP, materials performance, deformation and commercial scale hot working processes for Ti Alloys. 2:30pm THE DESIGN, PRODUCTION, AND METALLURGY OF ADVANCED, VERY LARGE, TITANIUM AEROSPACE FORGINGS: T.E. Howson. R. G. Broadwell Wyman Gordon Forgings, P.O. Box 8001, North Grafton, MA 01536-8001 This paper discusses the development of methods and dies for production of large aidrame and engine forgings in several titanium alloys. Applicable modelling, metallurgical considerations, and mechanical properties are also presented.

B19

2:50pm A MECHANISTIC STUDY OF HOT WORKING OF CAST MULTI-PHASE NEAR-"{I'I-AIALLOYS: DA. Hardwick. P.L. Martin, Rockwell Science Center, 1049 Camino Dos Rios, Thousand Oaks, CA 91360 Microstructural structural refinement and chemical homogeneity are the twin goals of primary ingot breakdown, and both are a necessary prerequisites for further deformation processing. Compression workability testing, coupled with extensive metallographic characterization (SEM backscatter and TEM), has been used to study the evolution of these characteristics in three near- TiAl alloys: Ti-46A1-O.2B, Ti-48AI-0.2B, and Ti48A1-2Mo-O.2B. Details of the response to deformation of each phase in these multiphase alloys, including the partitioning of deformation between the various phases, will be discussed. 3:10pm MICROMECHANISMSOFSUPERPLASTlCDEFORMATION OF A GAMMA TITANIUM ALUMlNIDE ALLOY: C.M. Lombard, S.L. Semiatin, WIlMLLN, Wright-Patterson AFB, OH 45433-7817; A.K. Ghosh, Dept. of Materials Science and Engineering, University of Michigan, Ann Arbor, MI48109-2316 A near-gamma TIAI material (Ti-45.5AI-2Cr-2Nb) was hot rolled to produce thin sheets. Samples of the sheet were tension tested at strain rates of 1O.3-1O·:!s.1 and temperatores between 900 and 1200·C to establish the flow behavior and microstructural changes. Results showed that the optimum superplastic forming conditions were at 1200·C and an initial rate of 1O.3s·1 Under these conditions, the maximum elongation (980%) and the minimum amount ofcavitation were achieved. Various micromechanisms, such as dynamic/static grain growth, solute redistribution between constituent phases, and changes in the morphologies of the phases, will be discussed in relation to the observed flow and fracture behavior. 3:30pm A CONSTITUTIVE EQUATION FOR THE mGH TEMPERATURE DEFORMATION BEHAVIOR OF TI-(40-54)AI ALLOYS: Catherine M. Sabinash, McDonnell Douglas Corporation, St. Louis, MO 63166; Shankar M. L Sastry, Department of Mechanical Engineering, Washington University, St. Louis, MO 63130-4899 The elevated temperature deformation behavior of Ti-(40-54)AI was studied by compression testing at 1200·C to 1300·C and strain rates of O. I to 0.00 lis. Based on the operating mechanisms, a constitutive equation for the dependence of flow stress on temperatore, strain rate, strain, and volume fraction of constituent phases was developed for single phase alpha and gamma alloys, two phase gamma + alpha2 titanium aluminides and Ti-AI-Cr and Ti-AI-Mn alloys. The equation, which utilizes readily available parameters eliminates the need for expensive high temperature testing. The equation predicts the values of peak and steady state flow stresses of alloys across the compositional range within a factor of two. 3:50 pm BREAK 4:00pm A PM METHOD FOR MAKING GAMMA TITANIUM ALUMlNIDE WIRE: C. F. Yolton,I. P. Beckman, I. H. Moil, Crucible Materials Corporation, Crucible Research, 6003 Campbells Run Road, Pittsburgh, PA 15205-1022 Gamma titanium aluminide alloys have attractive elevated temperature properties which make them a desirable material for advanced gas turbine engines and advanced airframe applications. These alloys are of interest for aircraft applications because they generally have a low density, good elevated temperature tensile and creep strength, and good oxidation resistance. Considerable progress has been made in the development of titanium aluminide alloys for aircraft applications. One fabrication method which has been very successful in making gamma titanium a1uminide components is casting. Gamma titanium aluminide alloys such as TI-48AI-2Nb-2Cr have good castability and the castings have good mechanical properties. As with all titanium castings, gamma titanium aluminide castings must be hot isostatically pressed to close internal cavities and weld repaired to fill surface dimples. Weld repair requires filler metal to fill the dimples, typically in the form of wire. Although gamma titanium aluminide casts well it is very difficultto hot work and process by other conventional methods. Production of welding wire is therefore difficult and costly. One means of minimizing the difficulties associated with the fabrication of gamma titanium aluminide welding wire is through the use of powder metallurgy. This paper will describe a powder metallurgy method for making gamma titanium aluminide wire in which wire is extruded directly to finish size from prealloyed powder. 4:20pm LAMELLAR TO EQUIAXED GRAIN TRANSFORMATION BY TORSIONAL DEFORMATION PROCESSING OF ALPHAIBETA TI ALLOY: G. Welsch, Case Western Reserve University, Cleveland, OH 44106; I. Weiss, Wright State University, Dayton, OH 45435; D. Eylon, University of Dayton, Dayton, OH 45469; F.H. Froes, University ofIdaho, Moscow, ID 83843 The lamellar alpha + beta microstructure of cylindrical Ti-6AI-2Sn-4Zr-2Mo specimens was transformed to the equiaxed grain structure by torsionally induced shear deformation at elevated temperature. The effects of shear strain and time at elevated temperature on the microstructural development were analyzed by optical and electron metallography. By taking advantage of the strain dependence of the transformation and of the strain gradient in torsion, a duplex microstructure can be created of crackinitiation-resistantequiaxed grains in the surface region and crack-propagation-resistant lamellar structure in the interior. 4:40pm INFLUENCE OF THERMOMECHANICALPROCESSING CONDITIONS ON THE MATERIAL PROPERTIES OF TI-6%A1-4%VtrlC COMPOSITES: P. Wanjara, R.A.L. Drew, S. Yue, McGill University, 3450 University Street, Department of Mining and Metallurgical Engineering, Montreal, Quebec, Canada, H3A 2A7 Titanium alloy composites containing up to 20 volume percent TiC particulates were processed by powder metallurgy processing. Investigation of the as-consolidated

B20

composites indicated that, although high temperature consolidation is beneficial in achieving near complete densification, degradation in the composite properties may result from extensive <:arbon loss from the TiC particulate reinforcements. Thermomechanical procetlsing can reduce consolidation temperature and time while simultaneously improving the properties of the composite through elimination of the residual porosity and modification of the matrix microstructure. High temperature deformation processing was performed by compression testing at strain rates in the range of 3 x 10-4 to 2 S·I at various temperatures above and below the beta transus (around 988·C forTi-6%AI-4% V). Through systematic variations in these deformation processing parameters, the conditions for obtaining a fully dense material having a range of microstructores and mech81Dical properties have been investigated. 5:00pm SOLUTE STRENGTHENING IN 8 TITANIUM - HYDROGEN ALLOYS: O.N. Senkov and 11. lonas, Department of Metallurgical Engineering, McGill University, 3450 University Street, Montreal, Quebec, Canada H3A 2A7 Compression and torsion rests were carried out on aseries of titanium-hydrogen alloys containing up to 37 atomic % of dissolved hydrogen. The tests were performed within the beta phase field. The dependence of the flow stress on temperatore, strain rate and hydrogen content was studied. The addition of hydrogen to beta titanium leads to noticeable hardening, the amount of which increases with hydrogen concentration according to a quadratic law ..I\nalysis of the activation parameters of plastic flow shows that they depend weakly on hydrogen concentration. The flow behavior is described in terms of thermally activated glide and takes into account internal stresses that depend on the hydrogen concentration ru:cording to a power law relationship. 5:20pm THE EFFECT OF GRAIN SIZE ON AMBIENT TEMPERATURE CREEP BEHAVIOR OF a-TITANIUM ALLOYS: S. Chandu, S.Ankem, Department of Materials and Nuclear Engineering, University of Maryland, College Park, MD 20742-

2115

Recently it has been shown that grain size has a significant effect on ambient temperature creep behavior of
ALUMINUM REDUCTION TECHNOLOGY II: Inert Anodes And Other New Developments

Sponsored by: LMD AlumlnuDI Committee Program Organizer: Ms Fiona 1 Stevens, Comalco Research and Technology, Comalco Research Centre, PO Box 316, Thomastown, Victoria 3074, Australia Monday,PM February 5, 1996

Room: A9 Location: Anaheim Convention Center

Session Chairperson: Alton Tabereaux, Reynolds Metal Company, Corporate Research and Development, 3326 East Second St, Muscle Shoals, AL 35661-1258 2:00pm AN IMPROVED METHOD FOR CURRENT EFFICIENCY DETERMINATION IN A LABORATORY AI.UMlNUM CELL: M.R. Domen, M. Hyland and BJ. Welch, Department of Chemical and Materials Engineering, The University of Auckland, Auckland New Zealand; I.M. Purdie, Comalco Research and Technology, PO Box 316, Thomastown, Victoria, Australia Obtaining rapid, continuous current efficiency measurement is important when using a laboratory cell to investigate the disputed relationship between current efficiency and alumina concentration. An improved method of measuring continuous current efficiency in a laboratory cell has been developed. The method uses mass spectrometry to performatotal mass balance of oxygen in the cell. The cell is flushed with argon, carrying the anode gases through solid state mass flow meters to a previously calibrated quadrupole mass spectrometer. An algorithm corrects the CO concentration by subtracting the "shadow" CO resulting from fragmentation of C~ in the mass spectrometer. The calibration algorithm also accounts for mass spectrometer drift during the course of the experiment. Cell data and gas composition are recorded, so that with a known starting hath composition the current efficiency can be calculated with respect to cell temperature and aluminum concentration. 2:30pm PROCESSINGSILlCON,SILUMINANDALUMINIUMFROMFELDSPARSA METHOD TO REGULATE THIS SI, AISI ALLOYS AND AL AMOUNTS BY A CONTINUOUS WAY: Jan R. Stubergh, Oslo College, School of Engineering, Section of Chemistry, Cort Adelers gate 30, N-0254 Oslo, Norway Feldspars are mixed with cryolite and electrolyzed at about lOOO·C.ln the first bath silicon "metal" is deposited in a high purity state and in a desired chosen amount. A carbon cathode is placed at the top of the bath and the carbon anode in the bottom of the bath. By electrolysis COz is formed at the anode and bubbles through the bath in good contact with the silicon crystals.Si4+ diu + 4e ~ S~.). In the second bath the rest of deposited silicon "rnetal"andofSi(IV) I'rom the first bath is reduced by aluminium metal at the same temperature. Silicon is deposited in aluminium metal as a silumin alloy and removed from the bottom of the bath. 3Si4+ diu + 4Al(1) ~ 3Si(l) +4AIl+ diu S~.) + AIoJ ~AlSi(l).InthethirdbaththeSi(IV)poorelectrolytefromthesecondbathiselectrolyzed

by using aluminium metal as a cathode. The Al metal is deposited in the bottom of the bath. All+ diu + 3e ~ AI(l).

1996 TMS ANNUAL MEETING AND EXHIBITION

3:00 pm BREAK 3:10pm INERT ANODES FOR THE PRIMARY ALUMINIUM INDUSTRY: AN up. DATE: R.P. Pawlelt, Technical Info Services and Consulting, Avenue du Rothom 14, CH-3960 Sierre, Switzerland The recent development of inert anodes for the primary aluminium industry is reviewed. The types of higb temperature conducting oxide and metal electrodes are considered as well as the developments and researcb under various problems. Of special interest are the stability and corrosion resistance, the metal contamination, mechanical integrity and practical and economic fabrication. 3:40pm THE BEHAVIOUR OF NICKEL FERRITE CERMET MATERIALS AS INERT ANODES: Espen Olsen and Jomar Tbonstad, Department of Electrochemistry, The Norwegian Institute of Tecbnology, N-7034, Trondbeim, Norway Various compositions of nickel ferrite cermets were tested in a laboratory cell. The materials were based on the copper-containing cermet material (NiFC204+NiO+Cu) originally developed by Alcoa. Electrical and chemical properties were studied as function of the content of excess nickel oxide. A new method of powder preparation was used, giving smaller grain size and somewhat higher electrical conductivity than hefore. Unusual electrical properties were found for a material containing no excess nickel oxide. The materials were tested as anodes in a conventional electrolyte for up to 50 boura at a current density of 0.8 Alc~. The contamination of anode components in the deposited metal varied for the different materials, but typically 2200 ppm Fe, 400 ppm Ni and 4S0 ppm Cu were found. The corrosion rates were very low, corresponding toO.12 cm/year. Mass transfer of imparities from the bath into the metal was slow with msss transfer coefficients of the order of 10.7 mls. 4:10pm TIN DlOXIDE·BASED CERAMICS AS INERT ANODES FOR ALUMINIUM SMELTING: A LABORATORY STUDY: A.M. Vecchio-Sadus, D.C. Constable, R. Dorin and E.J. Frazer,CSIRO Division of Minerals, PO Box 124, Port Melbourne, Victoria 3207, Australia; M.B. Trigg, G.S. Neal, S. Lathabal and I. Fernandez, CSIRO Division of Materials Science and Technology, Locked Bag 33, Clayton, Victoria 3168, Australia The behaviour of tin dioxide-based ceramics as inert anodes was examined in a laboratory· scale aluminium smelting cell over a range of electrolyte compositions with operating temperatures between 830-97S'C. Anodes of a nominal composition Sn
CAST SHOP TECHNOLOGY II: Grain Refining and Pha.e Modification Sponsored by: LMD Aluminum Committee Program Organizer: Julian V. Copenhaver, Technical/Quality Manager, NSA-A Division of Southwire, P. O. Box 500, Hawesville, KY 42348 Monday, PM February S, 1996

Room: AIO Location: Anabeim Convention Center

Session Chairperson: Dr. Ray Peterson, Reynolds Metals Company, 3326 East Second Street, Muscle Sboals, AL 3S633-12S8 2:00pm THE RELATIVE IMPORTANCE OF NUCLEATION ANDGROWfHMECHA· NISMS TO CONTROL GRAIN SIZE IN VARIOUS ALUMINUM ALLOYS: unnan Bilc1terud, Dr. Mats Johnsson, Department of Inorganic and Structural Chemistry, Arrhenius Laboratory, Stockholm University, S-I06 91 Stockholm, Sweden Control of grain size involves control of nucleation as well as growth. A critical review involving criteria for heterogeneous nucleation like particle size, abundance, stability, epitaxy and so forth will be given. Some potential candidates for heterogeneous nucleants are discussed in detail. Being able to supply a sufficient number of potent nuclei constitutes the basis for grain refinement. But the final grain size will be determined by the subsequent growth rate of -AI crystals. This bas been found to be controlled by a growth restriction factor [(kt-I)1I1jCJ whicb, at least for low alloying levels, seems to be additive in nature. Experimental evidence sbows a range of composition where optimum grain size can be achieved. This can be related to a change over from cellular to dendritic growtb in the equlaxed zone. 2:20pm HIGH PERFORMANCE PHOSPHORUS ADDITIVES FOR MODmCATION OF SIUCON IN AI-SI ALLOYS: Dr. Ben Heshmatpour, Shieldalloy Metallurgical Corporation, 12 West Boulevard, P. O. Box 768, Newfield, NJ 08344 A series ofFeP,CuP,NiP,FeP+CuP,FeP+NiP,CuP+NiP,andFeP+CuP+NiPbased additives were developed as high performance silicon modifiers for application in bypereutectic AI-Si alloys. These additives are manufactured by cold compaction of powder constituents with no subsequent higb temperature annealing as required in prnduction of

1996 TMS ANNUAL MEETING AND EXHIBITION

competitive products. Primary silicon grain size of below 40 microns is achieved with low phosphorus addition rate (40 ppm) with five minutes contact time. These additives produce higb performance at both higb (800C) and low (73OC) alloy temperatures. The particular sizing of the constituents, as developed and defined in this work, is the key factor in performance of these additives. The broad range of chemistry, form, performance, and low cost offers significant flexibility in application of these additives. 2:40pm THERMAL ANALYSIS OF ALUMINIUM ALLOYS AS A TOOL TO EVALUATE THE GRAIN REFINERS EFFICIENCY: Mohamed M. Hariedy, Aluminium Company of Egypt, Nag Hammadi,Egypt; Prof. A. H. Niazy, AssiutUniversity,Faculty of Engineering, Assiut, Egypt; Prof. A. A. Nofal, Central Metallurgical Researcb & Development Institute (CMRDI), Al Tabin, Cairo, Egypt AI-S'lI>Ti-I'lI>B and AI-S'lI>TI-I'lI>C master alloys were used to grain refine some aluminium alloys. The factors that affect the efficiency of the grain refiners (addition rate, bolding time, pouring temperature and metal composition) were studied. TbermaI analysis was used to evaluate the efficiency of the two grain refiners in parallel with a conventional method (KBI test). The characteristic parameters of the cooling curves have been correlated to the grain size of the castings. It was found that thermal analysis could be used as a tool to evaluate the efficiency of the grain refiners of aluminium alloys prior to casting. Also, it was found that AI-S'lI>TI-I'lI>C master alloy is more efficient grain refiner than AI-S'lI>TI-I'lI>B. 3:00pm METALLURGICAL EFFECTS OF STRONTIUM ON WROUGHT 6061 AL· LOYS: F. Paray, J. E. Gruzleski, Department of Mining and Metallurgical Engineering, McGill University, 3450 University Street, Montreal, Qu~bec, Canada, H3A 2A7; Bahadir Kulunlt, D. J. Zuliani, Researcb and Development Center, Tmuninco Metals, A Division ofTimminco Limited, Haley, Ontario, Canada, KOJ IYO Aluminum extrusion alloys contain a variety of intermetaUics that form during ingot solidification. In the 6061 series of alloys, iron combines with aluminum and silicon to form two types ofintermetallics, pAlFeSi and aAlFeSi. The type of the intermetaUic that is present in these alloys will bave an important bearing on the bomogenization time, Workability and quality of the surface finisb of the parts produced from these alloys. It bas been demonstrated that the addition of about ISO- 300 ppm ofstronium to 6061 alloy will make the preferred a the dominant intermetaUic phase at room temperature whicb forms during billet solidification. This enables the T4 billet bomogenization time to be significantly shortened by eliminating the lengthy time, in excess of 6 boura, needed to transform the as-cast Pphase to the more desirable a form. It bas also been demonstrated that 6061 extrusions produced from strontium-free billets require approximately 6 boors of T4 billet bomogenization at S7S'C to achieve the same surface finisb obtained in strontium containing extrusions with billet T4 times of less than 2 bours. 3:20pm RECENT ADVANCES IN UNDERSTANDING THE MECHANISM OF ALU· MINIUM GRAIN REFINEMENT BY TIBAI MASTER ALLOYS: Dr. M. A. Keams, S. R. Thistlethwaite, P. S. Cooper, London & Scandinavian Metallurgical Co. Limited, Fu\1erton Road, Rotherbam, South Yorkshire, S60 lDL. England Experimental data and observations from a number of complementary researcb programmes on the microstructure and behaviourofTiBAI grain refiners is summarized. These include analysis of the microstructure of nucleant particles within the master alloy and the final product and studies on particle agglomeration phenomena. The studies provide new insight into the nature of potent nuclei and bow microstructural features influence grain refiner efficiency. A model is proposed to describe current understanding of bow TiBAI grain refiners work. This is in the form of a series of stages from grain refiner addition through to solidification. The model predictions are compared with casthouse experience and lessons are drawn for manufacturers and users of grain refiners. 3:40 pm BREAK 3:SOpm THE EFFECT OF Sr-MODIFICATION ON THE MELT HYDROGEN CONTENT AND THE HYDROGEN SOLUBIUTY IN THE SOLID AND UQUID AI· SI ALLOYS: Daryoush Emadi, Department of Mining and Meta\1urgical Engineering, McGill University, Mon~a1, QlKlbec, Canada, H3A 2A7, and D6pl. des Sciences Appliqu~s, Universitt! du ~bec l Cbicoutimi, Chicoutimi, QlKlbec, Canada, G7H 2B 1; J. E. Gruzleski, Department of Mining and Meta\1urgical Engineering, McGill University, Montreal, Qu~bec, Canada, H3A 2A7; H. Liu, M. Bouchard, D6pt. des Sciences Appliqu6es, Universitt! du Qu~bec l Chicoutimi, Chicoutimi, Qu~bec, Canada, G7H2BI Sr-modification of AI-Si alloys is nonna1ly accompanied by an increase in porosity in the casting. In an attempt to understand the nature of this problem, a study was initiated to investigate the effect of Sr on the melt bydrogen content, and the bydrogen solubility in solid and liquid AI-Si alloys. The melt bydrogen content before and after Sr addition remains essentially constant for all ranges of Si content, and Sr treatment does not introduce bydrogen into \be melt. In addition, strontium does not affect the rate of bydrogen pick-up by molten AI-Si alloys. Some problems associated with bydrogen measurement in Sr-modified melts are also discussed. Hydrogen measurements in the solid state indicate that Sr does not change the equilibrium bydrogen content in solid A3S6 alloy. Sr addition decreases the bydrogen solubility in liquid AI-Si alloys. This decrease seems to be more significant at lower temperatures and at higher silicon contents. At lower bydrogen solubilities, the bydrogen level in the liquid reaches the solubility limit at an earlier stage of solidification. As a result, pores can nucleate earlier and grow over a longer period of time resulting in more overall porosity in the solidified alloy.

B21

4:10pm DEVELOPMENT AND EVALUATION OF AI-TI-C MASTER ALLOYS AS GRAIN REFINERS FOR ALUMINIUM: Eng. M. A. Hadia. Aluminium Company of Egypt, Nag Hammadi. Egypt; Dr. A. A. Ghaneya. Assiut University. Faculty of Engineering. Assiut. Egypt; Prof. A. H. Niazy. Assiut University. Faculty of Engineering. Assiut. Egypt A series of AI-11-C master alloys with different amounts of titanium and carbon (AI3.5%11-0.5%C. AI-3.5%11-0.7%C. and AI-5%Ti-I%C) has been developed by adding activated graphite to molten AI-11 binary alloys. The microstructures of the developed alloys were investigated. The efficiency of the produced alloys as aluminium grain refiner was evaluated. The effects of the following factors were investigated: addition rate of each master alloy. the pouring temperature of the refmed metal and the holding time before pouring. Also. the contact time for each refiner was studied. The results were significant and the refiner AI-5%11-1 %C was the most effective. 4:30pm MICROSTRUCTURES AND PROPERTIES OF STRONTIUM TREATEDALUMINIUM ELECTRICAL CONDUCTOR ALLOYS: Dr. B. Closset. Timminco S.A .• 44. ch. Petite-Boissi~re. 1208 Geneva. Switzerland; F. Paray. J. Gruzleski. McGill University. 3450 University Street, Montreal. Qu~bec. Canada H3A 2A7; Dr. H. Mulazimoglu. American Racing. 19200 South Reyes Avenue. Rancho Dominguez. CA 90221 Aluminium electrical conductor wires must meet specified strength and electrical resistivity. In this study the composition ofa typical 6201 electrical conductor alloy was altered by varying respectively the Mg. Si. and Fe levels. After a standard T4 treatment and quenching. various T6 artificial aging cycles were applied to Sr modified and unmodified conductor wires. Electrical resistivity and machanical properties were investigated. For a given chemical composition and T6 heat treatment cycle. it was shown that the tensile strength of strontium-treated alloys can be improved by 10%. The microstructure of the as-cast samples were evaluated by optical microscopy and the morphology of intermetallic compounds related to the properties. 4:50pm HIGH-RESOLUTION TRANSMISSION ELECTRON MICROSCOPY OF GRAIN-REFINING PARTICLES IN AMORPHOUS ALUMINIUM ALLOYS: Peter Schumacher. University of Oxford. Department of Materials. Parks Road. Oxford. OXI 3PH. United Kingdom; A.'L. Greer. University of Cambridge. Department of Materials Science and Metallurgy. Pembroke Street. Cambridge CB2 3QZ. United Kingdom Crystallization from aluminium melts and crystallization from amorphous aluminium alloys occur by a similar nucleation mechanism on heterogeneous nucleation sites. The low atomic mobility in the amorphous aluminium alloy permits the observation of nucleation events in high resolution electron microscopy. 11B zparticles exhibit on their basal facets a thin layer of A1311 which nucleates a-AI. All three phases reveal an epitaxial orientation relationship between their close-packed directions and planes. The underlying 11B z substrate stretches the a1uminide layer on the basal facets of the boride closer to the lattice of the aluminium. The A1311 thereby becomes a better nucleation substrate by better lattice matching. However. the thinly stretched layeris notonlyagood nucleation substrate for aluminium. but also can act as an adhesive between borides once these make physical contact. 5:10pm THEFLEXIBLEAISrI0115:P. C. van Wiggen.KBMMaster AlloysB.V .• K100sterlaan 2. 9936 TE Delfzijl. The Netherlands In this paper. three different perspectives are presented to outline the success of Aluminium Strontium 10/15% rod (AISrlo/I5). This rod is commercially applied for the modification of aluminium alloys. An illustration is given of the improved ductility of this super-fine microstructured alloy. thus offering the required flexibility and reliability for in-line additions made from coils. It is explained how the in-line addition method itself proves to yield multiple improvements in today's aluminium casting operations. Last minute (in-line) addition is synonymous to flexibility in production. Finally. the paper will describe the successful use of AISrl0/15 rod in the production of aluminium wires with improved flexibility (springs, for example).

ENVIRONMENTAL EFFECTS AND WEAR TMS General Abstract Session Monday,PM February 5,1996

Room:A2-3 Location: Anaheim Convention Center

Session Chairperson: TBA 2:00pm TENSILE PROPERTIES AND STRUCTURE OF SEVERAL SUPERALLOYS AFTER 5000 HOUR EXPOSURE TO LlF AND VACUUM AT 1173 K: J. Daniel Whittenberger. NASA Lewis Research Center, Cleveland, OH 44135 The solid-to-Iiquid phase transformation of LiF is under consideration for thermal energy storage in space-based solar dynamic system. Unfortunately, the melting point of this salt (1121 K) is beyond the commonly accepted upper use temperature of 1100 K for Cr-bearing superalloys in vacuum. The commercially available Ni-base superalloy HastelloyR B-2 is Cr-free. but its high Mocontent alloy can form phases which cause high temperature embrittlement. HastelloyR B-2 and the Cr-containing Co-based Haynes AlIoyR 188 and Ni-based Haynes Alloy 230 were exposed to LiF and vacuum at 1173 K for 5000 h. Neither exposure had any significant effect on HastelloyR B-2 in terms of microstructural sudace damage or weight change. Nevertheless. measurement of the post exposure tensile properties of HastelloyRB-2 revealed low tensile ductility at 1050 K. Little evidence of attack by LiF was seen in either of the Cr-containing superalloys;

B22

however considerable weight loss and near sudace microstructural damage occurred in both alloys exposed to vacuum. 2:20pm PROTECTIVE ALUMINA SCALE OVER SUPERALLOYS: Vivek Saxena. Department of Metallurgical Engineering. Malaviya Regional Engineering College. Jaipur. India The growing use of superalloys for high temperature applications needs oxidation resistance which can be effectively attained by forming continuous. thin A I z03 scale (4-6 m) using selective oxidation technique. Spontaneous oxidation of AI and various other alloying elements takes place as the Gibbs free energy of oxide formation is large and negative. Preferential a1mnina scale development depends upon AI interdiffusion coefficient. solubility and diffusivity of oxygen in the alloy having critical weight of aluminium as depicted by oxide mapping. This paper is an attempt to review the principles of alloy oxidation along with factors affecting scale formation and its adherence to superalloy surface. The role of reactive elements. dispersoids and effect of sulphur has also been critically analysed. 2:40pm EVALUATION OF CORROSIVENESS OF FIRE-RETARDANTS ON ALUMINUMALLOYSUSEDINAIRCRAFfINTERIORS:J.E.Talia.J.Chaudhuri.S.N. Cheema. A. Arif. Dept. of Mech. Eng .• Wichita State University. Wichita. KS 67208 This research evaluates the corrosive fire retardant products present in general aviation small aircraft interiors and eliminates those chemical compounds which are suspected of inducing corrosion. To this end. various existing fire retardant products were evaluated in conjunction with various aluminum alloys making up the aircraft interior skin. In order to supplement the literature search of the Service Difficulty Reports data:base. extensive experimentation was carried out with varying parameters. such as temperature and humidity in the sandwich test technique. This conclusively proved that the presence of halogen based salt particles may be the source of corrosion. 3:00pm EFFECTS OFHe+ IRRADIATION ON THE MECHANICAL PROPERTIES OF POLYCRYSTALLINE NIAi: S. T. Eck, A. J. Ardell. Department of Materials Science and Engineering. University of California. Los Angeles. CA 90095-1595 Polycrystalline NiAI was irradiated with 2 MeV He+ ions at temperatures of 150 and 225 K to fluences ranging from 2.9 x lOIS to 5.8 X 10 16 He+/cmz•The peak doses under these conditions range from 0.1 to 2.0 dpa. The yield stress was found to increase with increasing fluence at a fixed temperature. but decreased with increasing temperature at a particularfluence. The yield Sb'ess increased significantly from 148 MPa forunirradiated specimens to 177 MPa for specimens irradiated to a fluence of2.9 x lOIS He+ Icm2 at 225 K. but only increased gradually 'with further increases in the fluence. A similar effect was found at 150 K. with the exception that the yield stresses were slightly higher than those at 225 K at each particular fluence. We attribute the increase in yield stress to at least partial irradiation-induced disordering of the NiAI. but this awaits confirmation by transmission electron microscopy investigations. which are in progress. This work is supported by the Department of Energy. 3:20 pm BREAK 3:30pm MICROMECHANISM OF HYDROGEN-ASSISTED FATIGUE CRACK GROWTH IN TI-6AI-4V TITANIUM ALLOY: P. S. Pao. C. R. Feng. R. A. Bayles. Naval Research Laboratory. Washington. D.C. 20375; G. R. Yoder. Office of Naval Research. Arlington. VA 22217 The micromechanism of hydrogen-assisted fatigue crack growth under high stress ratio in ~annealed 11-6AI-4V titanium alloy has been investigated. At stress ratio of 0.90. the threshold stress intensity of Ti-6AI-4V containing 1000 wppm hydrogen is significantly lower than that for alloy that contains only 60 ppm hydrogen. The fatigue crack growth rates of hydrogen charged material are also substantially higher than those for the as-received material. In the intermediate K region. an extensive secondary cracking along ~ intedaces and the concomitant reduction in the power-law exponent are observed in the hydrogen charged Ti-6A1-4V. Subsequent TEM analyses of tensile deformed materials indicates the precipitation of titanium hydride along the ~ intedaces and the presence of deformation twins in the hydrogen charged alloy. These deformation twins are terminated at the a/bydride intedaces and do not traverse across either hydrides or p phase into the neighboring a grains. The observed secondary cracking along ~ intedaces can be rationalized by the accumulation of stresses and the fracture at the a/bydride intedaces. 3:50pm OXIDATION KINETICS, MICROSTRUCTURE AND PROPERTIES OF A VSCr-STI ALLOY EXPOSED TO 300-600°C IN AIR: K. Natesan. W.K. Soppet. EnergyTechnology Division. Argonne National Laboratory. Argonne.1L 60439;M. Uz. Chemical Engineering Department. Lafayette College. Easton. PA 18042 A study was carried out to assess the oxidation kinetics and the effects of air exposure on the microstructure and properties of a V-5Cr-5Ti alloy sheet. Samples were evaluated in the as-rolled and as-annealed (I 050°c/1 h) conditions. and after subsequent exposure to air in the temperature range of 300-6OO"C for times ranging from 24 h to 2060 h. Tensile properties were determined following exposure at 500°C for up to 2060 h. Oxidation behavior at 300. 400. 500. ~i75. and 650°C was studied by a thermogravimetric testing method. Microstructural attributes (including phase. grain size. oxide thickness and intergranular fracture zone) were determined by various metallographic and X-ray diffraction techniques. The Vickers hardness profile along the thickness of each sample was measured to determine the depth of the hardened layer and to estimate oxygen diffusivity in the alloy. The results will be discussed. with the emphasis on oxidation kinetics and the possible correlation between processing. microstructure. and properties. Work supported by the U.S. Department of Energy. Office of Fusion Energy Research. under Contract W-31-109-Eng-38.

1996 TMS ANNUAL MEETING AND EXHIBITION

4:10pm ADSORPTION OF H10 AND 0 1 ON NI3(AI,T1) AND NI~e SURFACES: CORRELATION OF SURFACE REACTIVITY WITH DUCTIUTY. W. J. Chiao Y. W. Chung. Dept. of Mat. Sci. and Eng .• Northwestern University. Evanston. IL 60208 Water adsorption on clean and oxygen-covered NiJ(AI.11) (100) and NiJFe has been investigated by x-ray photoelectron spectroscopy and thermal desorption. It will be shown that water adsorbs molecularly on Ni J(AI.11) (100) at 140K and dissociates at higher temperature. On the oxygen-precovered sudace hydroxyls are produced upon water adsorption at 14OK. Compared with water adsorption on the clean sudace. more hydroxyls are produced. Some of the sudace hydroxyls are produced by direct water dissociation. and some by adsorbed oxygen abstracting hydrogen atoms from H20. Similarly. the reactivity of polycrystalline Ni~e towards water will be reported. The mechanical properties of NiJ(AI.Ti) and Ni3Fe such as ductility will be correlated with sudace chemical properties. 4:30pm SURFACE HYDRIDES IN LOW AND HIGH PRESSURE HYDROGEN CHARGED T\48AI-2CR-2NB: K. Li. U. Habel. T. M. Pollock. M. DeGraef. A. W. Thompson. MSE Dept. Carnegie Mellon University. Pittsburgh. PA 15213 The development of hydrides at the sudaces of duplex Ti-48AI-2Cr-2Nb samples charged in gaseous hydrogen at pressures from 0.1 MPa to 13.8MPa have been studied. Under conditions where trace amounts of oxygen were available during charging experiments up to 60 hrs in duration. the sudace microstructures were dramatically different from those of the bulk material. At higher pressures the bulk microstructure consisted of approximately 25 vol% hydrides including FCC hydrides and a recently identified 0 hydride. The 0 phase is body centered tetragonal with lattice parameters of a =0.528nm and c =2.845nm. The hydrides were embedded in a TiAI matrix with no traces of the ~ phase. The sudace microstructure was composed of an outer mixed layer of1102 and AI 20 3and an inner layer consisting of a continuous matrix of ehydride with a small volume fraction of discontinuous alumina particles. At O.lMPa no changes in bulk microstructure were observed. However. near sudace regions contained some titania and alumina and very small amounts of the e phase within the ~ phase. 4:50pm EFFECTS OF SUBSTRATE CURVATURE ON RESIDUAL STRESSES IN OXIDE FILMS: A COMPARISON OF FEM SOLUTIONS AND OPTICAL FLORESCENCE EXPERIMENTS: R. Williamson. J. Wright. Idaho National Engineering Laboratory. P.O. Box 1625. Idaho Falls. 10 83415-2218; P. Hou. R. Cannon. Lawrence Berkeley Laboratory. Berkeley. CA. 94720; B. Veal. Argonne National Laboratory. 9700 S. Cass Ave .• Argonne.IL 60439 Finite element analyses are utilized to investigate the effects of substrate curvature on the thermal residual stresses within an alumina film on an iron-a1uminide substrate. For films that are normally compressive on flat substrates. large tensile stresses are observed as the substrate radius of curvature is decreased. Comparisons are made between FEM predictions and residual stress measurements using optical florescence techniques. The results are useful in designing components which utilize oxide coatings for sudace protection. 5:10pm SUDING WEAR AND FRICTION CHARACTERISTICS OF ACuZlnc ALLOYS: M. David HanlUJ. Jon T. Carterand • M. S. Rashid. Metallurgy Department. General Motors Corporation. Research and Development Center. 30500 Mound Road. Box 9055 Warren. MI48090-9055 ACuZinc. a GM-patented. high-pedormance zinc-copper-aluminum alloy. which is suitable for manufacturing net shape die castings. has demonstrated superior wear and friction properties compared to traditional zinc alloys. The wear and friction characteristics of ACuZinc. Zamak 3. ZA 27. and pure zinc were determined by pressing blocks of these materials against warm, oil-lubricated. rotating rings of AISIII44 steel. The mass wear rate of the alloys generally correlated directly with coefficient of friction and inversely with hardness.

EXPERIMENTAL METHODS FOR MICROGRAVITY MATERIALS SCIENCE SESSION 1 Sponsored by: MSD Thermodynamic and Phase Equllbrla Committee NASA Mlcrogravlty Sciences Program Organizers: Dr.Robert A.Schiffman. R. S. Research Inc .• Barton. VT 05822. Professor Carlo Patuelli. Universita di Bologna. 1-40126 Bologna. Italy

Monday. PM February 5. 1996

Room: Orange County 2 Location: Anaheim Marriott Hotel

Session Chairperson: Edward A. Sioot. Canadian Space Agency. 6767 rue de I'Aeroport. St-Hubert. Quebec. Canada J3Y 8Y9

2:00pm INTRODUCTION AND WELCOME BY THE ORGANIZERS 2:15pm THE INFLUENCE OF CONVECTION AND THERMAL BOUNDARY LAYER THICKNESS ON NUCLEATIONOFSOUDPHASEININFILTRATEDMETAL MATRIX COMPOSITES: C. Patuelli. S. Sprio. R. Tognato. Dipartimento di Fisica. Universita di Bologna. Via Imerio 46. 40126 Bologna. Italy In a recent work. infiltrated Al metal matrix composites (MMCs) with different volume fractions of ceramic preforms (10.19% Al203 and 23% Sic) were melted and

1996 TMS ANNUAL MEETING AND EXHIBITION

solidified up to 720'C and the crystallites size distribution were measured by X-ray diffraction analysis. It was found. in particular. that convection and thermal boundary layer (t.b.l.) thickness influence the nucleation rate of the solid phase. The phenomenon was more evident in the sample with the ceramic preform of 10% Al20 3 which has the highest aspect ratio. ie. much higher convection. The aim of this study is to investigate more deeply such phenomenon by melting and solidifying the same MMCs at higher temperatures (greater than 720·C.). The results are discussed taking into account the effects of gravity or reduced gravity on t.b.1. thickness and viscosity. ie. on nucleation and growth of the AI matrix solid phase. 2:35pm X-RAY TRANSMISSION MICROSCOPY STUDY OF THE DYNAMICS OF SOLIDILIQUID INTERFACIAL BREAKDOWN DURING METAL ALLOY SOLIDIFICATION: Peter A. Curreri. Marshall Space Right Center. National Aeronautics and Space Administration. MSFC. AL 35812; William F. Kaukler. Center for Microgravity and Materials Research. The University of Alabama in Huntsville. Huntsville. AL 35899 In order to study the detailed dynamics of the solidification process. we are developing an apparatus to pedorm real-time x-ray microscopy of solidifying metallic systems. Only recently has technology become available to achieve this with the necessary resolution. speed. and contrast. We have applied the technique to AI-Cu and AI-Ag single phase alloys to study the dynamics of morphological instability. Due to the x-ray a1Jsorbing solute layer. AI- Ag alloys provide better contrast. With a thermal gradient of 45 Clcm. AI- IOAg alloy retains a planar intedace up to a growth rate of I micron per second. The initial stages of cellular breakdown occur at a rate between 1.5 and 2 microns per second and steady state cellular growth is attained at 2 microns per second. Curvature of the cellular interface was recorded as the intedace approached and engnlfed voids. For both alloys we could clearly image the intedacial morphology in real time and determine the growth rate dependence on feature size and spacing. 2:55pm CONVECTION IN HVPERMONOTECTIC ALLOYS AND THE INFLUENCE OF RESULTING COMPOSITIONAL VARIATIONS ON MICROSTRUCTURE: Ll. Hayes. J .B. Andrews. Department of Materials Science and Engineering. Uni versity of Alabama at Birmingham. Birmingham, AL 35294 This investigation focuses on the role of convection during vertically upwards directional solidification of hypermonotectic alloys. It should be possible under intedacially stable processing conditions (sufficiently high thermal gradient to growth rate ratio) to form a composite structure in these alloys consisting of aligned rods of the minority phase in a majority phase matrix if convection is minimized. Convective instability. however. is inherent in hypermonotectic alloys and results from adverse density gradients due to solute concentration variations in the liquid adjacent to the solidification front. Results from this work indicate that convective flow at low flow velocities (one-g processing) causes a composition variation along the length of the sample but does not impede the coupled growth process. Vigorous convective flow such as that experienced during portions of parabolic flight appears to completely disrupt the formation of the composite structure. Composition profiles and morphologies of both ground processed and KC-135 processed samples will be discussed. 3:15pm CRYSTALLIZATION KINETICS OF THE UNDERCOOLED Zr41.2TIl3.8Cu12.5NIlO.0Be22.5 ALLOY DURING CONTAINERLESS ELECTROSTATICLEVITATIONPROCESSING: Y.l. Kim.R.Bush.W.L.Johnson.MS 138-78. California Institute of Technology. Pasadena. CA 91125; S.K. Chung. W.K. Rhim. MS 183-401. Jet Propulsion Laboratory. Pasadena. CA 91109; AJ. Rulison. Microgravity Research, Space SystemsJLorai. MS 003. Palo Alto. CA 94303 A glass forming alloy Zr41.2Ti13.8CuI2.5NilO.OBe22.5 was used to investigate the crystallization kinetics at different undercooling levels by constructing the timetemperature-transformation (TIT) diagram experimentally. The experiments were pedormed using the high temperature high vacuum electrostatic levitator at JPL. The measurements indicate that there are two distinct crystallization events at different undercooling levels. At low undercooling. there is a primary crystallization from the homogeneous liquid. while. at large undercooling. a secondary crystallization appears due to the phase separation in the supercooled liquid state. Based on the measured TIl diagram. the critical cooling rate required to bypass the crystallization was obtained to be 1.8 Kls. 3:35 pm BREAK 3:55pm X-RAY TRANSMISSION MICROSCOPY OF AI-Pb MONOTECTIC ALWYS DURING SOLIDIFICATION: W.F. Kaukler. Center for Microgravity and Materials Research. The University of Alabama in Huntsville. Huntsville. AL 35899-0001; P.A. Curreri. Marshall Space Flight Center. National Aeronautics and Space Administration. MSFC. AL 35812 In-situ observations of AI-Pb alloys bothduring and after unidirectional solidification will be discussed. X-ray microscopy permits viewing the features developing on either side of the sll intedacial isotherm in real time. Our X-ray Transmission Microscope results reveal a diversity of phenomena not previously observed. Recent improvements to our techniques applied to this alloy have produced post-solidification images of fiber and aligned particulates with diameters of order of 10 micrometers. Previously reported features like solute layers in the melt and droplet formation can be observed with higher definition during solidification. Delicate. periodic striations parallel to the intedace. presently only observed via x-ray microscopy. have been detected in these alloys with growth rates down to I micrometer/sec. It was previously thought there was a lower critical velocity for striation formation. but this new evidence suggests the possibility that this morphological feature simply becomes impractical to detect as growth rate is reduced.

B23

4:1Spm MICROGRAVITY SOLIDIFICATION MICROSTRUCTURES AS ILLUS· TRATED BY NICKEL·IRON AND STONY·IRON METEORITES: P.z. Budka, 2135 Morrow Avenue, Schenectady, NY 12309; J .R.M. Viertl, 1403 Oifton Park Road, Schenectady, NY 12309; S.V. Thamboo,2444 Brookshire Dr., Schenectady, NY 12309 The macroImicrostructurai features of nickel-iron and stony-iron meteorites provide visual, logical evidence that these materials solidified as castings in a microgravity environment. These features are remarkable because they exhibit regular patterns of phase formation over a wide range of masses and compositions. Such patterns, visible at low magnification, provide information on general conditions experienced during solidification: microgravity, undercooling and a shallow thermal gradient. This paper will present microstructural evidence to support this scenario. It will also illustrate classic solidification reactions such as eutectic, peritectic and isolllOlJlhous series using typical meteoritic microstructures. Understandings derived from these meteoritic structures can provide insights into fundamental segregation behavior and the solidification of large iron-nickel superalloy ingots. 4:3Spm MICRO· AND MACRO-SEGREGATION IN ALLOYS SOLIDIFYING WITH EQUIAXED MORPHOLOGY: D.M. Stefanescu, B.K. Dhindaw,J. Leon, Solidification Laboratory, The University of Alabama, Tuscaloosa, AL 35487; S. Sen, USRA, MSFC, Huntsville, AL 35812; P.A. Curreri, NASA, MSFC, Huntsville, AL 35812 A model has been developed to describe micro-segregation based on the 'closed system' assumption, i.e. no net mass transfer enters of leaves the domain during solidification. Further, this model has also been extended to include the impact of the 'open system' assumption (mass transport in and out of the system) and of the 'expanding system' assumption (coarsening and coalescence) on micro-segregation. Under purely diffusive conditions the 'closed system' assumptions can be considered to be valid. Solidification experiments during aircraft parabolic maneuvers and solute concentration measurements in AI-2%Cu and AI-5%Cu samples are being performed to test the 'closed system' model predictions. High and low g experimental data will be used to test the validity of the 'open system' model. Initial results and data analysis for these experiments will be discussed. 4:SSpm THE INFLUENCE OF CONVECTION ON BANDING IN PERlTECTIC SYS· TEMS: A COMPARISON OFSn·CdANDPb-BI: K.L Zeisler-Mash!, T.A. Lograsso, Inst. for Physical Research and Tech., Iowa State University, Ames, IA 50011; R.K. Trivedi, Dept. of Matis. Sci. and Eng. and Ames Laboratory, US-DOE, Iowa State University, Ames,lA 50011 During directional solidification in peritectic systems, the primary phase and peritectic phase can form at the solid-liquid interface as alternating layers perpendicular to the growth direction for certain combinations of alloy composition, growth velocity and temperature gradient in the liquid. This banded microstructure is related to repeated solute accumulation and depletion in the liquid ahead of the interface. For a system such as Sn-Cd, in which convective mixing in the liquid can be minimized during growth in a terrestrial environment, banding repeats over significant solidification distances. For a system such as J>b.Bi, convective mixing in the liquid is significant during growth in a terrestrial environment at the slow rates required for band formation. As a consequence, the liquid composition progresses through the regime in which banding occurs. The resulting microstructure exhibits a narrow region containing a small number of bands that cbange in length, with the length of the primary bands decreasing and the length of the peritectic bands increasing as the fraction solidified increases. Current ground-based directional solidification experiments using Sn-Cd alloys will be compared with corresponding results from the J>b.Bi system.

FERROUS METALLURGY TMS General Abstract Session Monday,PM February 5, 1996

Room: Grand D Location: Anaheim Marriott Hotel

Session Chairperson: TBA 2:00pm CRYSTALLOGRAPY ANDTHREE·DIMENSIONALRECONSTRUCTIONOF PEARLITE COLONIES IN FE·C·X SYSTEMS: M. A. Mangan, G. J. Shiflet, DepartmentofMateriais Science and Engineering, University ofVirginia, Charlottesville, VA 22903 Previously, electron backscatter diffraction (EBSD) in a scanning electron microscope (SEM) has been applied to the pearlite transformation in Fe-.8C-12Mn to analyze the orientation relationship between ferrite and cementite lamellae in a single COlony. This study has been extended to other Fe-C-X systems where X is Ni, Cr, Mn and Si. In addition, any ambiguity associated with the three-dimensional (3-D) development of a pearlite colony can be removed through repeated sectioning and subsequent reconstruction. The interlamellar spacing and colony size can now be measured unambiguously from the resulting volume, improving data for studies of growth kinetics. Combining the 3-D information with the orientation relationships between ferrite and cementite lamellae obtained through EBSD, the factors resulting in different ferrite/cementite relationships can be determined. 2:20pm PHASE TRANSFORMATION KINETICS AND MICROSTRUCTURAL EVO· LUTION IN HYPOEUTECTOID STEELS: R. Pandi, M. Militzer, E. B. Hawbolt, T. R. Meadowcroft, The Centre for Metallurgical Process Engineering, The University of British Columbia Vancouver, B.C. V6T 1Z4 Cansda The austenite-to-ferrite/pearlite phase transformation is characterized by nucleation

B24

and growth phenomena. Usually, early site saturation is obtained in the austenite decomposition and further progress of the phase transformation is followed by ferrite growth. In ferritic steels, tile final microstructure can be evaluated by ferrite grain size and morphology. The ferrite grain size depends on the number of grains produced at the early stage of phase transformation. In other words, the ferrite nucleation rate determines the grain size of the final products. Chemical composition, austenite grain size, cooling rate and retained strain have major effects on the phase transformation kinetics and final microstructure. This work deals with modelling of the austenite decomposition kinetics and final microstructure of hypoeutectoid steels (low carbon, plain carbon and HSLA steels) by quantifying the above parameters. To follow the transformation kinetics experimentally, dilatometrit: continuous cooling transformation (CCT) tests have been performed on the Gleeble 1500 thermomecbanical simulator. Both semi-empirical and fundamental approaches are proposed todescribe the phase transformation behavior and good agreement between experimental data and model predictions are obtained. 2:40pm MODELING OF HEAT AFFECTED ZONE FORMATION IN FLAME CUT· TING OF MILD STEEL: Peter Robinson, BOC Fabrication Technology Center, 24 Deer Park Road, London, UK SWI93UF A numerical model has been developed, using a finite-volume based approach, to simulate the conductive heat transfer leading to the formation of a heat-affected zone in oxy-acetylene cutting of steels. The model considers both the heat supplied by the iron-oxygen reaction, and thai from the preheating flame. Phase transformations in the steel are considered, by the inclusion of non-Iinearities in the governing partial differential equation. Experimental validation has shown the model to be accurate in the prediction of both the size and shape of the heat affected zone. Correlations for dependence of heat affected zone size on processing parameters (cutting speed, f10wrate of gas, steel alloy content and nozzle size) are discussed in relation to the model, and their effects are described. 3:00pm HIGH STRENGTH HIGH DUCTILITY COMPACTED GRAPHITE CAST IRON: MECHANICAL PROPERTIES: Cheng-Hsun Hsu, Shen-Chih Lee, Dept. of Materials Engineering, Tatung Inst. of Technology, Taipei, Taiwan 10451 This research applied alloy addition and heat treatment methods to obtain austenitic, upperllower bainitic, and tempered martensitic compacted graphite (CG) cast irons. The effects of vermicularity (percentage of compacted graphite in total free carbon), and matrics on strength and ductility were evaluated. It was found that high nickel base austenitic CG iron resulted in lowest hardness and tensile strength, followed by as-cast pearlitic-ferritic iron. Bainitic CG iron exhibited both superior strength and ductility of all CG irons studied. As to tempered marten-sitic CG iron, the high strength grade would be accompanied by relatively low ductility as compared to bainitic one. Optical microscopy, SEM and X-ray diffractometty were carried out in order to correlate the microstructural features to that of the properties obtained. 3:20 pm BREAK 3:30pm HETEROGENEOUS NUCLEATION BEHAVIOR IN CAST IRON: Toshiaki Mizoguchi, John H. Perepezko, Deptartment of Materials Science and Engineering, University of Wisconsin-Madison, 1509 University Ave., Madison, WI 53706 During the solidification of cast iron a modest melt undercooling is a common observation, but the identity of the nucleation catalyst limiting the arnountofundercooling has been elusive. In order to examine the undercooling behavior in a systematic study, a droplet method has been developed based upon an AI,03SiO,-CaO slag containing a dispersion of 30-300 /1Itl diameter droplets. For hypoeutectie alloys, primary y-Fe is a poor nueleant for both graphite and Fe3C based on a nucleation undercooling of 375·C below the eutectic. For hypereutectic alloys, primary graphite is catalytic fory-FeandFe3C at8·C and 31·C undercoolingrespectively. The solidification microstructures have been examined to identify the reaction path. The support of the Nippon Steel Curp. and ARO(DAAH-04-93-G-0296) is acknowledged. 3:50pm MORPHOLOGICAL DEVELOPMENT OF CELLULAR (DISCONTINUOUS) COLONIES IN A 19Cr·SNI AUSTENITE STEEL: S.Matsuoka, M. A. Mangan, G. J. Shiflet, Department of Materials Science and Engineering, University of Virginia, Charlottesville, VA 22903 The dependence of grain boundary misorientation on the occurrence and morphological development of cellular precipitation was studied using a combination of electron backscattered diffraction (EBSD) with an SEM and three-dimensional re-construction from serial sectioning of cellular colonies. In this study, cellular colony growth rates are shown to be dependent on the grain boundary misorientation forming the transformation growth front. The growth rate increases from about 20 degs. misorientation and reaches a maximum at 40 degs. By applying the repeated sectioning method and our computer graphic analysis program, a three-dimensional reconstruction of a cellular colony is obtained resulting in an improved perspective of the colony. Once the cellular colony is reconstructed, precipitate plate lengths and spacings can be accurately measured without the usual errors associated with single sectioning. 4:10pm THE EFFECT OF GRAIN SIZE AND LOAD ON THE TRANSIENT CURRENT RESPONSE AND MICROSTRUCTURES DURING THE IMPACT AND CON· TINUOUS SCRATCHES IN Fe·18%Cr·S%NI ALLOY AND 304 STAINLESS STEELS: Richard Raul Romero, S. K. Varma, Department of Metallurgical and Materials Engineering, The University of Texas at EI Paso, EI Paso, TX 79968-0520 The scratch technique has been modified to include the continuous scratches besides the conventional impact scratches to study the corrosive wear behavior of Fe-18%Cr5%Ni alloy and 304 stainless steel. The transient current generated during the two different scratching processes have been compared to study the depassivation kinetics. The near surface microstructures evolved as a result of the scratches have been

1996 TMS ANNUAL MEETING AND EXHIBITION

determined by the transmission electron microscopy (TEM). The microstructural details at the TEM level have been examined to study the effect of initial grain size and load on the transient current generated due to depassivation. The load on the stylus and the grain size not only influence the microstructures, they also affect the depassivation and repassivation kinetics during the corrosive wear process. This research has been supported by the National Science Foundation through the grant number HRD-9353547.

GENERAL TOPICS IN MATERIALS TMS General Abstract Session Monday, PM February 5,1996

Room: Grand H Location: Anabeim Marriott Hotel

St!ssion Chairpt!rson: TBA 2:00pm CALIBRATION METHOD OF THE INDENTER TIP SHARPNESS FOR MICROINDENTATION TESTS: Youngman Kim, R&D Center, Korea Gas Corp. Ansan, Korea; Min-Tae Kim, KIA Motors, Seoul, Korea A calibration method of the indenter tip sharpness is suggested for microindentation tests. Microindentation tests were carried out on quartz glass plates and silicon wafers using a microhardness tester with continuous depth recording. The worn-out height of the indenter tip was obtained from the load-depth data based on a derived linear relation between the load and the square of the penetration depth. The as-obtained values of load and penetration depth were calibrated in terms of 'worn-out' height considering near-surface effects of specimens. The elastic modulus values of quartz glass obtained from tbe calibrated data showed a good agreement with those from references. 2:20pm MATERIALS, EVOLUTION OF THE CHOICE CONSIDERATIONS: Tawfik Bt!nabdallah, Mohammed B. Benabdallab, University of Science and Technology of ORAN-ALGERIA Up to the last few years, choosing any material was obeying considerations strongly related to the conditions of use. The coming years and the next century will definitely add new requirements that should be taken into account in order to fulfill conditions of satisfaction. The meaning of that is, neither the environment effect nor the energy consumption of the process are to be neglected. The aim of this contribution is showing that, the nature of the chosen material is no more firstly considered, as the Quality/price ratio is strongly related now to energy, recycling, and environment All these considerations will make engineers facing new priorities for decision. A personal point of view will be given closed to Algerians engineers considerations as a sample of a third world country. 2:40pm CONDUCTING POLYMERS: Ajay Pratap Singh, Rabul Asthana, Department of Metallurgical Engineering, Malviya Regional Engineering College, Jaipur India Conducting polymers are emerging as new class of electronic material. These polymers have a highly anisotropiC quasi-one-dimensional structure, which makes them fundamentally different from conventional inorganic semiconductors. The conducting polymers area has experienced phenomenal growth over past few years with the discovery that the polyacetylene (CH).,could be doped with electron acceptors and electron donors to conductivity levels (10000. 1em· 1) approaching those of some metals. This paper attempts to review the novel concepts associated with this new class of electronic material by describing some important properties of polyacetylene (CH).-a well studied sample. 3:00pm TOUGHENING OF CERAMICS: Rajeev Agrawal, Rajendra Kumar Goya, Department of Metallurgical Engineering, Malaviya Regional Engineering College, Jaipur, India Advanced C.eramics have drawn much attention for engineering parts. They are finding increasing applications in electrical, electronics, defence, thermal, structural, chemical, biological, nuclear plant etc. They possess high hot strength, high compressive strength, good wear and corrosion resistance, low density, but they are brittle. In this paper recent development in toughening of ceramics have been critically summarized and revieWed. Paper deals with special case of alumina toughening. Addition of about 10 to 20 % by weight of partially stabilized zirconia (PSZ) raises the toughness and transverse rupture strength of alumina by transformation toughening. In which metastable zirconia is transformed to stable monoclinic zirconia with an increase of about 3 to 5% volume and induction of shear strain (about 7%) during the crack propagation or during machining the matrix. This enables increase in toughness and transverse rupture strength of the alumina. 3:20pm LIFETIME ASSESSMENT OF INSULATION FOAMS USING TIDN FILM BENDING: Sung-Sik Shin, Youngman Kim, R&D Center, Korea Gas Corp., AnsanCity, Kyunggi-do, Korea The PU(polyurethane) foams for LNG (Liquified Natural Gas) pipe insulation are applied using RIM (Reactive Injection Molding) and covering processes. For the RIM process the adhesion force at the interface between PU foams and substrate is associated with the various properties of foams and substrate. The cracks formed at the interface by the CTE(coefficient of thermal expansion) difference have influence on the properties and lifetime of the insulation foams since two different materials with different CTE's are bonded. In this study thermal stress variations were measured as a function of temperature to assess the lifetime of the insulation foams. The foams were applied on thin SUS304 substrates under controlled atmosphere to measure the bending caused by the

1996 TMS ANNUAL MEETING AND EXHIBITION

CTE difference between the two materials. The amount of bending was converted to the therrnal stress, in tum, the stress was used to predict the lifetime of the insulation foams.

A SYMPOSIUM DEDICATED TO PROFESSOR .JOHANNES WEERTMAN'S SEVENTIETH BIRTHDAY: SESSION II: Fatigue & Fracture

Spansored by: Jt. SMDIMSD Nudear Materials and MSD Flow and Fracture Committees and FEMS (Federation of European Materials SocIetIes) Program Organizers: R.J. Arsenault, Department of Materials Science and Nuclear Engineering, University of Maryland, College Park, MD 20742-2115; David Cole, CRREL, 72 Lyme Rd., Hanover, NH 03755; Todd Gross, Department of Mechanical Engineering, University of New Hampshire, Durham, NH 03824; Gernot Kostorz, Institut fiIr Angewandte Physik, ETH Htlnggerberg, CH-8093 ZUrich, Switzerland. Peter Liaw, Department of Materials Science and Engineering, University of Tennessee, Knoxville, TN 37996-2200; Sivan Parameswaran, NRC-Institute for Aerospace Research, Ottawa, Canada KIA OR6; Howard Sizek, Inco Alloys International Inc., Huntington, WV 25705-1771 Monday, PM February 5, 1996

Room: Orange County 3 Location: Anaheim Marriott Hotel

Session Chairpersons: T. Gross, Department of Mechanical Engineering, University of New Hampshire, Durham, NH 03824; A.K. Vasudevan, Office of Naval Research, Arlington, VA 2:00 pm Invited CRACKS ON INTERFACES: R. Thomson, Emeritus, NIST, Gaithersburg, MD 20899 Cracks on interfaces can cleave on the interface, cleave or branch to a cleavage plane off the interface, emit a (nonblunting) dislocations on the interface, or emit a blunting dislocation on a slip plane inclined to the interface. Criteria for all these events will be presented in terms of the balance between a driving force for the event and a lattice resistance to the event The driving force can be written as an extension of the standard crack extension force for a crack. The lattice resistance for cleavage is related to the intrinsic interface surface energy, but is not identical to it for the branching crack. The lattice resistance for nonblunting emission is the unstable stacking fault, but involves both the surface and unstable stacking fault energies in the case of blunting emission. The role of chemical segregants on the interface will be discussed. 2:30 pm InvIted EXPERIMENTAL AND NUMERICAL STUDIES OF THE BDT TRANSmON IN NlAl: P. Neumann, P. Ochmann, H. Vehoff, Max-Planck-Institutfiir Eisenforschung GmbH, Max-Planck-Strasse I, D-402337 DUsseldorf, Germany In pre-cracked mono- and bi-crystalline samples of NiAI the fracture toughness was measured as a function of temperature and the distance between crack tip and grain boundary of the bicrystals. The results are compared to numerical simulations of dislocation emission from a crack tip in a vicinity of a grain boundary. The kinetics of dislocation motion and the image forces on the dislocations are fully taken into account. The grain boundary is either modelled as an impenetrable obstacle 10 the dislocations, a source of dislocations on other slip systems or a weak interface with a critical cohesive strength. The qualitative features oflhe experimental findings could be reproduced in the numerical simulations. 3:00 pm InvIted DISLOCATION STRUCTURES IN DEFORMED METALS: N. Hansen, Materials Dept., RiSil National Laboratory, DK-4000 Roskilde, Denmark The deformation of single crystals and polycrystals produces dislocation structures which evolve with continued deformation. These structures have been observed to be common to a broad range of materials, alloys, deformation modes and deformation temperatures. Recent work within this field is analyzed with emphasis on the characteristics of various dislocation boundaries and the causes of their formation taking into account the misorientation across the boundary, the boundary plane and the slip pattern in the adjoining crystallites. It is found that both the dislocations in the boundaries and the characteristic microstructural and crystallographic features of the boundaries relate to the actual slip pattern and to the tendency of dislocations to accumulate in low energy cOnfigurations. A basis has therefore been created for relating dislocation structures to the plastic deformation behavior over a large range of stress and strain. 3:30pm SLIT-LIKE INHOMOGENEITIES UNDER UNIFORM STRESS: CRACKS, ANTICRACKS AND QUASICRACKS: J.A. Hurtado, Department of Materials Science and Engineering. J. Dundurs, T. Mura, Department of Civil Engineering, Northwestern University, Evanston, IT.. 60208 Solutions for slit-like inhomogeneities in an infinitely extended isotropic solid subjected to a uniform stress at infinity are obtained by using Eshelby's equivalent inclusion method. First, two limiting cases are studied: Cracks (i.e. inhomogeneities with elastic moduli identically zero), and Anticracks (i.e. inhomogeneities with infinitely large elastic moduli). Solutions are obtained for different modes of uniform loading in plane strain (mode I and mode II), and antiplane strain (mode III). It is observed that the stress field of a mode I anticrack is the sum of the stress field of a mode I crack (which arises from a crack plane dislocation dipole distribution. Finally the most general case ofinhomogeneities with finite non-zero elastic moduli, which here are calledquasicracks, is considered and new solutions are provided. An interesting example of quasicracks in engineering practice is that of welded cracks where the elastic moduli of the welding material differ from those of the matrix.

B25

3:50pm FATIGUE DAMAGE IN A20 VOL% SIC..-6061 AI ALLOY MATRIX COMPOSITE: S. Homdee, J.G. Byrne, Department of Metallurgical Engineering, 412 W.C. Browning Bldg., University ofUtah, Salt Lake City, UT 84112-1183 Both 20 vol % Sic... reinforced and unreinforced 6061-T6 AI alloy R. R. Moore specimens were fatigue tested under constant maximum alternating stress to establish SN curves. The addition of SiC whiskers increased the fatigue strength and fatigue life of the composite relative to the unreinforced AI alloy. In the composite, the positron annihilation Doppler broadening technique was able to detect increasing fatigue damage up to failure, whereas in the unreinforced alloy fatigue damage could be detected only for the initial 8 to 17% of total fatigue life due to saturation of the Doppler line shape parameter used in the measurements. At low stress levels, the fatigue fracture process in the composite is localized near the fracture surface relative to the unreinforced alloy. As the stress level increases, the depth of the fatigue damage region beneath the fracture surface increases. Fatigue crack initiation in the composite occurs by the formation of microcracks near the interface between SiC whiskers and the aluminum alloy matrix beneath the free surface. Pull-outofSiCwhiskers coated with metal matrix material was noted. The connection of these microcracks results in fatigue crack propagation. In the final fracture region, the fracture process consists of the formation and coalescence of voids in the matrix and at whisker ends in dense SiC cluster areas. 4:10pm PARTICLE SHAPE EFFECTS ON THE FRACTURE MORPHOLOGY AND DUCTILITY OF DISCONTINUOUSLY-REINFORCED COMPOSITES: N. Shi, S.G. Song, M.A.M. Bourke, J.A. Goldstone, Los Alamos National Laboratory, Los Alamos, NM 87545; G. Liu, Department of Metallurgical and Materials Engineering, Illinois Institute of Technology, Chicago, IL 60616 Particle shape effects on the fracture and ductility of a spherical and an angular particulate-reinforced 6061-AI matrix composite containing 20 vol. pet AI20 l were studied using SEM fractography. The results were correlated tu finite element method (FEM) simulations. SEM fractographic examinations show that during tensile deformation the spherical composite fails through void nucleation and linking in the matrix near the reinforcement/matrix interface. In contrast the angular composite fails through particle fracture and matrix ligarnentrupture. FEMresults indicate that the failure modes for these two composites can be attributed to the differences in the development of internal stresses and strains. While the maximum strain in the matrix is similar under the same load, particle stress in the angular composite is considerably higher, inducing premature particle fracture which constitutes the composite failure mode. 4:30pm EFFECT OF ymLD STRENGTH AND ELASTIC MODULUS ON FRACTURE SURFACE INTERFERENCE OF REMOTELY LOADED MODE II CRACKS: T.S. Gross, R.U. Goulet, Department of Mechanical Engineering, University of New Hampshire, Durham, NH 03824-3591; D.A. Mendelsohn, Department of Aeronautical Engineering, Applied Mechanics and Aviation, The Ohio State University, Columbus, OH43210 Fracture surface interference bas been shown to shield the crack tip from a remote Mode II load as well as induce a Modell stress intensity factor. Specimens of 7075 AI in the T6 and 0 condition and specimens of MSO tool steel in the annealed and quenched and tempered condition were subjected to pure remote Mode II loads on a modified four point bend fixture. These four specimens were chosen to determine the extremes of the effect of elastic modulus and yield strength on the magnitude of Mode II shielding and induced Mode I stress intensity factor. The effect of differing boundary conditions on the loading pins will be discussed and a compliant modified four point bend fixture will be described. 4:50pm FATIGUE AND FRACTURE IN DUCTILE-PHASE REINFORCED, BRITTLE INTERMETALLICLAMINATED COMPOSITES: D. B/oyer,K.T. Venkateswara Rao, R.O. Ritchie, Department of Materials Science and Mineral Engineering, University of California, Berkeley, CA 94720-1760 With their superior creep properties, several intermetallic alloys may become lightweight alternatives to currently-used superalloys for many engine applications, particularly associated with aerospace. Unfortunately these materials invariably suffer from low intrinsic toughness and ductility at ambient temperatures. Indeed, improvements in their damage tolerance bas necessitated the use of extrinsic toughening methods, i.e., by promoting sources of crack-tip shielding, to improve their fracture behavior. Although ductile-phase reinforcements are commonly used to this end, their effect on fatigue behavior is less certain. Accordingly, in the present study, the effect of metallic Nb reinforcements on the fatigue-crack propagation and fracture toughness hehavior of NblAI is examined in a medium-scale laminated composite consisting of 125-l1m thick layers of Nb bonded to 500-l1m thick layers of NblAI. It is shown that whereas the toughening response of such materials is far superior to particulate-reinforced or microlaminates of NblNblAl, the fatigue properties are much less affected. The microstructural mechanisms underlying such behavior, and the corresponding fatigue and fracture properties at elevated temperatures (> IOOO°C), will be discussed. Work supported by the Air Force Office of Scientific Research under the AASERT Program.

HIGH TEMPERATURE COATINGS II: Ceramic Coatings II Sponsored by: MDMD Surface Modiftcatlon and Coatings Committee Program Organizers: Narendra B. Dahotre, Center for Laser Applications, University of Tennessee Space Institute, Tullahoma, TN 37388; Janet M. Harnpikian, School of Materials Science & Engineering, Georgia Institute of Technology, Atlanta, GA 30332. JacobJ. Stiglich, PO Box 206, Sierra Madre, CA 91025 Monday, PM February 5, 1996

Room:BI Location: Anaheim Convention Center

Session Chairpersons: Ashok Kumar, Department of Electrical Engineering, University of South Alabama, Mobile, AL 36688; Jacob J. Stiglich, P.O. Box 206, Sierra Madre, CA 91025 2:00pm GRADED NI-AIIOl COATINGS VIA ELECTRODEPOSITION: S.W. Banovic, C.M. Petronis, Barmak, K., Marder, A.R. Department of Materials Science and Engineering, Lehigh University, Bethlehem, PA 18015 Functionally Graded Materials (FGMs) are heing investigated for use as thermal barrier coatings (TBCs) in advanced gas turbine systems. The principle failure mechanism in conventional TBCs is the spallation of the ceramic top coat from the metallic bond coat due to dissimilar thermal expansion coefficients. In order to eliminate this problematic interface, FGMs are being evaluated. A model system of codeposited metallic nickel matrix with second phase alpha-alumina particulates (Ni-AI20 l ) was studied in an initial effort to produce a graded structure. Electrodeposition process variables were investigated and optimized to produce Ni-Al20 l composite coatings with varying AI20 l volume percents. Through manipulation of these process variables, a discretely graded structure of Ni-Al20 l ranging from 0 to 35 volume percent alumina particulate in the deposit has been produced. Light Optical Microscopy (LOM), Scanning Electron Microscopy (SEM), and Quantitative Image Analysis (QIA) techniques were used to characterize the electrodeposited microstructure in detail. Microhardness tests were JX,rforrned to evaluate the mechanical properties of the electroplated coatings. The relationships between the microstructure and microhardness were analyzed. 2:20pm EROSION OFTHERMALSPRAY MCrAIY-Cr3C2 CERMETCOATINGS: KJ. Stein, B.S. Schorr,A.R. Marder', Energy Research Center, Lehigh University, Bethlehem PA 18015 Solid particle erosion (SPE) is a serious problem for the electric power industry, costing an estimated $150 million a year in lost efficiency, forced outages, and repair costs. In an attempt to reduce lhe damage caused by SPE many industrial users have turned to high velocity oxy fuel (HVOF) cermet coatings. The ceramic content of these cermet coatings needed for optimum erosion resistance is still unclear. The corrent work investigated the erosion resislaJ1lce of FeCrAlY-Crl C2 and NiCr-Crl C2 cermet coatings with carbide levels ranging from 50-100% (in the pre-sprayed powder). The as sprayed coating microstructures were fully analyzed using light optical microscopy (LOM), scanning electron microscopy (SEM), x-ray diffraction, and microprobe analysis. This line of testing revealed the carbide levels in the starting powder are much higher than in the final as sprayed coating. It appears that the carbides react with the HVOF jet resulting in the formation of various oxides. Erosion testing was carried out using a particle accelerator apparatus set at 40 mis, 80 gimin, and 4OO'C with 30011m A120l particles for both 30' and 90' impact angles. Upon erosion testing of the sprayed coatings it was discovered that the erosion rate decreased as the carbide content, and overall hard phase content (oxides and carbides), was reduced for 90!l impact. In addition the erosion rate remained fairly constant regardless of carbide content, or hard phase content, for 30' impact. 2:40pm CUBIC CALCIA STABALIZED ZIRCONIA COATINGS WITH SPUTTER DEPOSITION: Mahendra Pakal, Ray Y. Lin, Department of Materials Science, University of Cincinnati, M.S. 12, Cincinnati, OH 45221; Heather Walls, Crucible Magnetics, Magnet Drive, Elizabethtown, KY 42701 Calcia stabalized zirconia films were deposited on the surface of < III> silicon wafers using RF magnetron sputter deposition. Deposition was carried out at substrate temperature varying between 80°C and 900°C. X-ray diffraction results show that all films consist of cubic zirconia. The fracture surface morphology of the films were studied using scanning electron microscope (SEM). Depending on the deposition temperature, the structure of the film is either columnar or equiaxed. The transition temperature from columnar structure toequiaxed structure was found to lie between the Trrmratios of 0.39 and 0.43. The microhardness of the films is found to increase deposition temperature. This is due to probably that the hardness of ceramic materials decreases with the increase in the defect concentration. 3:00pm PHASE COMPATIBILITY OF Zr0z-ZrSI.-SI01 AT 2073 K: Charles A. Odegard, Mario Cuen, Arturo Bronson, Materials Research Center of Excellence (MRCE), 321 Burges Hall, The University of Texas at El Paso, EI Paso, TX 79968 Multiphase compatibility has been investigated for the reaction couple zirconium oxide/zirconium silicide/silica at 2073 K in air. The multiphase system was assembled by encapsulating zirconium disilicide in a quartz capsule. In a previous study of this system at 2273 K, annealing resulted in an inner core consisting of liquid zirconium silicide and an interdiffusion zone. The interdiffusion zone developed between the liquid silica layer and liquid silicide was composed ofzirconia precipitates with liquid silicide globules dispersed in a liquid silica matrix. The microstructural features and their

B26

1996 TMS ANNUAL MEETING AND EXHIBITION

consequences on the oxygen flux of the zirconium silicide/oxide coating will be discussed. 3:20 pm BREAK 3:30 pm Invited EXPERIENCE IN SEEKING HOT CORROSION-RESISTANT STABILIZERS FOR ZIRCONIA: Robert L. Jones, Chemistry Division, Code 6170, Naval Research Laboratory, Washington, DC 20375-5342 This paper summarizes the results of a research effort aimed at identifying hot corrosion-resistant stabilizers for use in zirconia thermal barrier coatings (TBCs) for engines. Among tbe oxides investigated were TI02, Ce02, Y203, MgO, Sc203, In203 and Sn02. A brief overview of the hot corrosion of zirconia TBes is given to illustrate the rationale used in evaluating the hot corrosion performance of the candidate oxides. 3:55pm STRUCTURE AND PROPERTIES OF SPUTTER-DEPOSITED BN THIN FILMS: R. Capelletti, Istituto Nazionale di Fisica della Materia; Dipartimento di Fisica, Universita' di Parma, Parma, Italy; M. Elena, CMBM, 38050 Povo (TN), Italy; D. Ielmini Dipartimento di Ingegneria Nucleare, Politecnico di Milano, 20133 Milano, Italy; Miotello, Istituto Nazionale di Fisica della Materia; Dipartimento di Fisica, Universita di Trento, 3 8050 Povo (TN), Italy; P.M. Ossi, Istituto Nazionale di Fisica della Materia; Dipartimento di Ingegneria Nucleare, Politecnico di Milano, 20133 Milano, Italy This paper presents results obtained on BN thin films deposited with an RFmagnetron sputtering machine, adopting RF bias of the samples and process atmospheres compo~ of argon, or of an argon-nitrogen mixture. Comparisons with ~Ims produced by ~pplymg DC bias are also made. The techniques used to charactenze the films, which were deposited on silicon substrates, were IR spectroscopy, AES, SEM and nanoindentation, providing information on structure, composition microstructure and mechanical properties (hardness and Young's modules). By a suitable choice of experimental parameters, both hexagonal and cubic phases were produced. The type of bias (DC or RF) has a major influence on the film structure of the samples. Film composition, which is significantly influenced by process atmosphere is also important for detennining the relative abundance of the hexagonal and cubic phases.

A.

4:15pm PROCESSING AND CHARACTRIZATION: PLASMA ASSISTED-CVD DIAMOND FILMS: Mohit Sisodia, T.Y. Rajan, Dept. of Metallurgical Engr., Malaviya Regional Engr. College, Iaipur-17, India Recent advancement in surface modification by high temperature coating technology that allows the chemical vapour deposition (especially the plasma assisted) of diamond films on many different materials. Present paper critically reviews the morphologies! physical properties of Diamond and Diamond like carbon (DLC) films pr~uced ~y various techniques like Magnetron sputtering, microwave plasma, laser aSSisted With much emphasis on Plasma assisted - CVD along with its characterization an~ process variables. Besides it, Iapans latest technique of Stable Plasma process are also discussed. Due to the high temperature properties such as melting, solid-solid equillibria and thermal expansivity by laser heated cell, these films have been used extensively for fabrication of high temperature, high power and high frequency semiconductor/electronics devices. 1bermal imaging windows and mechanical components of industrial practices, are also studied at length. 4:35pm MICROSTRUCTURE OF LASER CLAD ZrOz-(NI ALLOY) COMPOSITE COATING: Y. T. Pei, T.C. Zoo, Institute of Applied Lasers, Department of PhYSics, Beijing Polytechnic University, Beijing 100022, China 1be microstructure of laser clad 6Ovol.% Zr02 (partially stabalized with 2mol% Y203) plus 4Ovol. % Ni alloy composite coating on stain!ess steel ~rl3 was in~estig~ted by scanning electron microscopy, electron probe IDIcroanalYSIS, X-ray diffraction, energy-dispersive X-ray analysis, and microhardness tests. A stratification is observed in tbe clad coating which consists of a pure ZrQ clad layer in the outer region and a bonding zone of Ni alloy adjacent to the substrate. 1be pure ceramic layer exhibits columnar Zr02 dendrites growing from the ceramic layer-bonding zone interface. This ceramic layer is composed of metastable xl -Zr02 phase and a very small amount of m-Zr02phase distributed along the boundary of tl grains. 1be high ~ati?g and cooli~g rate caused by laser cladding restrains the t->m phase transformation In the ceranuc layer. The bonding zone ofJ.1-Ni alloy shows the multi-phase structure of primary ll-Ni dendrites and interdendritic eutectics, and have good metallurgical characteristics between tbe clad ceramic layer and the substrate. 4:55pm ZnO COATINGS GROWN BY PULSED LASER DEPOSITION: M. Baleva, A. Chalkov, E. Mateena, Faculty of Physics, Sofia Unviersity, Sofia 1126, Bulgaria Abstract not available.

HIGH TEMPERATURE SUPERCONDUCTORS. SYNTHESIS, PROCESSING, AND LARGE SCALE APPLICATIONS II: Kinetics, Phase Equilibrium, and Phase Development

Sponsored by: Jt. EMPMDISMD Superconductlng Materials Committee, MSD Electronic, Magnetic and Optical Phenomena Committee, and FEMS (Federation of European Materials SocIeties) Program Organizers: U. Balachandran, Energy Technology Division, Argonne National Laboratory, Argonne, IL 60439; Paul I. McGinn, University of Notre Dame, Notre Dame, IN 46556; Stuart Abell, University of Birmingham, Edgbaston, BinninghamBI52TT, UK Monday,PM February 5,1996

Room: B7 Location: Anaheim Convention Center

Session Chairpersons: K. C. Goretta, Argonne National Laboratory, Argonne, IL 60439; W. Wong-Ng, National Institute of Standards and Technology, Gaithersburg, MD20899 2:00 pm Invited KINETICS OF FORMATION, c-AXIS ALIGNMENT, AND CRITICAL CURRENTS OF BI(2:2:2:3)1Ag TAPES: Y.L. Wang, W. M. Bian, Y. Zhu, Y. Fukumoto, T. Thurston, M. Suenaga, Dept. of Applied Science, Brookhaven National Laboratory, Upton, NY 11973 Employing a precursor powder, which was produced by an aerosol spray pyrolysis technique, highly textured Bi(2:2:2:3)/Ag tapes were synthesized. Using the tapes, a comprehensive study was carried out for kinetics of Bi(2:2:2: 3) formation, the platelets alignment, and their critical currents utilizing transmission electron microscopy, transmission x-ray diffraction, and high sensitivity Y-I characteristic measurements. 1be results of the study will be described in detail. This work was performed under the auspices of the U.S. Department of Energy Division of Materials Sciences, Office of Basic Sciences under Contract No. DE-AC02-76CHOOOI6. 2:20 pm Invited FORMATION OF NUCLEATION SITES FOR GROWTH OF BI(2223) ON PbDOPED BI(2212): D. K. Finnemore, M. Xu, Ames Laboratory, Iowa State University, Ames, IA 500 11; U. Balachandran, Argonne National Laboratory, Argonne, IL 60439 Scanning electron microscopy studies of mixed powders of Pb-doped Bi(2212), calciumcuprate, and CuO show that nucleation sites of 100 nm in size grow on Bi(2212~ grains during the ramp up to temperature from 25 to 800·C. Growth of these nuclei begins at 650·C and seems to require the presence of Ag; it occurs at both 3 and 150TO!f of partial pressure of oxygen. The nuclei grow in the same temperature range as the rapid changes in X-ray peak heights while pellets of mixed powders are ramped from 25 to 800·C. Work at Ames Laboratory was supported by the U.S. Department of Energy, under Contract W-7405-Eng-82; work at ANL was supported by the U.S. Department of Energy (DOE), Energy Efficiency and Renewable Energy, as part of a DOE program to develop electric power technology, under Contract W-31-109-Eng-38. 2:40 pm Invited MELTING EQUILIBRIA OF THE HIGH T.PHASES IN THE SYSTEM BI-SrCa-Cu-O: W. Wong-Ng, L. P. Cook, Ceramics Division, National Institute of Standards and Technology, Gaithersburg, MD 20899 Knowledge of melting equilibria in the system Ba-Sr-Ca-Cu-O (BSCCO) is important for melt processing of bulk superconductors and for melt growth of single crystals. A general procedure developed at NIST to detennine melt composition has been applied to study the melting equilibria of the superconducting phases in the BSCCO syste~.1be primary crystallization fields of the Raveau phase, 2212 and 2223 phases Will be discussed. Effects of Ag addition on the melting equilibria will also be discussed. 3:00 pm Invited THE INFLUENCE OF LEAD ON PHASE FORMATION AND MICROSTRUCTURE DEVELOPMENT IN AglBl-2223 COMPOSITE CONDUCTORS: V. A. Maroni, N. Merchant,I. S. Luo, I. Leboy, A. K. Fischer, K. Wu, W. Zhong, S. E. Dorris, Argonne National Laboratory, Argonne, IL 60439 Investigations performed during the processing of AglBi-2223 composite conduct~rs have revealed that lead has numerous significant influences on Bi2223 phase formation and microstructure development. 1be mode of lead incorporation in the precursor powder is the dominant factor controlling the length of the induction period preceding the onset of Bi-2223 formation. Back reaction of Bi-2223 to Bi-2212 during cool down appears to be accelerated by the lead rich "liquid phase." Changes in the cbemical state oflead during ramp up, heat treatment, and cool down have been monitored using Raman microscopy and x-ray diffraction. The influence of stoichiometry variations involving Bi, Pb, and Sr on Bi-2223 phase development and core microstructure has been investigated. 3:30 pm Invited NOVEL PROCESSING ROUTES FOR BSCCO PHASES THROUGH CONTROL OF PHASE CHEMISTRY: J. L. MacManus-Driscoll, L. F. Cohen, A. D. Caplin, Centre for High Temperature Superconductivity, Imperial College, London SW72BZ,UK 1be main limitations to the formation oflong length of BSCCO conductors with good Ic performance are phase purity, microstructural inhomogeneities and large scale imperfections such as cracking and voiding. Understanding and control of the phase chemistry is essential in order to obtain reprodUCible properties. We have detennined phase equilibria in the Bi-2212 system and phase stability of(Bi,Pb)-2212and (Bi,Pb)2223 as a function of temperature and reduced oxygen partial pressures. This informa-

1996 TMS ANNUAL MEETING AND EXHIBITION

B27

tion bas allowed alternative precursor pbase combinations and processing routes to be explored. In particular, reduced temperature and oxygen partial pressure processing schemes for (Bi,Pb)-2212 and (Bi,Pb)-2223 sball be discussed. For (Bi,Pb)-2212, processing in the supersolidus pbaseregion at-830°Cand p02= 3xlO·3atm.leads to well aligned, near pbase pure conductors with J.'s of up to lOS A cm·2at 4.2 K. For (Bi,Pb)2223, rapid formation of the pbase occurs when processing from reduced oxide precursor phases. 3:50 pm Invited

PHASE DEVELOPMENT IN PARTIAL-MELT PROCESSING OF SILVFR· CLAD BI11,srl.l3C....1CU10. TAPES: Ming Jiang, M. T. Lanagan, W. E. Delaney, K C. Gort!na, Energy Technology Division, Argonne National Laboratory Argonne IL ~9

'

,

Ag-clad Bi2.15SrmCal 02CU20. (Bi-2212) tapes were fabricated by the powder-intube method. Using the partial-melt processing approach, we heat treated the tapes in air and in various 01Ar mixtures. The tapes were then characterized by differential thermal analysis, scanning electron microscopy energy dispersive X-ray analysis, and X-ray diffraction. Particular eropbasis was placed on conditions above the peritectic melt temperature. Relationships between processing and microstructure will be discussed. Work supported by the U.S. Department of Energy (DOE), Energy Efficiency and Renewable Energy, as part of a DOE program to develop electric power technology, and Basic Energy Sciences (Materials Sciences), under Contract W-31-109-Eng-38. 4:10 pm Invited PHASERELAT10NS ASA FUNCTION OF TEMPERATURE AND POl IN THE BI~rlCalCuO.·Ag SYSTEM: R. W. McCallum, L. Margulies, M. J. Kramer, Ames Laboratory, Iowa State University, Ames, IA 50011 The promise of usable forms of high temperature superconducting wires and tapes rests on developing techniques for processing highly textured materials. Currently, meltpr~ssing seems most suited to this goal, as the formation of liquid enhances grain allgnmenl Although much work bas been done in determining the effect of Ag and P02 on the first solidus temperature, there bas been no detailed study of the Bi2212-Ag pseudo binary pbase diagram. Understanding the effect of Ag and P02on the chain of peritectic reactions initiated during melting is necessary in order to optimize the meltproc:essing ~ters of Bi22121Ag composites. Previously, work on the solubility of Ag m the BI2212 melt revealed a large liquid immiscibility gap and a eutectic on each side~fthis gap. The sub-liquidus p~reactions in this system were examined in greater detail forO.OI, 0.2, and 1.0 bar P02. This work suggests that the initial peritectic reactions may be different than those previously reported. Supported by the United States Department of Energy under contract No. W-7405-ENG-82. 4:30pm EFFECT OF SINTERING PERIODS ON THE PINNING FORCE, ACTIVA· T10N ENERGY, AND MICROSTRUCTURE OFIDGH·TC SUPERCONDUCT· ING BI-(pb)-8r·Ca·Cu·O TAPES: N. V. Vo, H. K. Liu, S. X. Dou, Centre for Superconducting and Electronic Materials, Univ. of Wollongong, NortbfIelds Ave., Wollongong, NSW 2522, Australia Short and long periods of sintering of short single core Bi(Pb)-2223 tapes bave been studied. Coroparisons are made between three batches of these tapes with the same thermo-mechanical deformation and sintering conditions but with the starling precursor powders being different in their processing. It is found that the volume density of the ~ro~~ic pinning force in fields up to IT is greater for a particular sintering period. Reslsuvlty measurements further show the activation energy also reaches a higher value for the same sintering period cbosen. SEM images of the surfaces indicated better n,'orpt,wlogy and density can be achieved for this particular sintering period. Prolonged smtermg leads to greater porosity while short duration sintering inhibits grain growth and thus is of no assislance to the healing of cracks induced by the intermediate deformation belWeen sintering periods. 4:50 pm Invited

COMMON MECHANISM OF TEXTURE FORMATION IN BI-BASED HTSC PRODUCED BY DIFFERENT METHODS: T. D. Aksenova, S. V. Sbavkin, ISC&SSF, RRC Kurchatov Institute, 123181, Moscow, Russia; P. V. Bratukhin. Kashirskoe sb. 31, Bochvar's Inorganic Materials Institute, 123060 Moscow. Russia In spite of different therrnomechanical processes for fabrication of the best Bi-2212based and 2223-based Ag-sbeatbed composite conductors, our model of texture formation is supposed to be valid for bothcases. Model treats the anisotropic crystalline growth process in quasi-2D volume of crystallization. Coroparative analysis of texture in 2212 and 2223 composite tapes have been performed. We bave also concluded the absence of influence of Ag sheath texture on growing oxide layer. 5:10pm EVOLUTION OF PHASES IN POSTANNEALED BlSrCaCuO THIN FILMS: B. Vengalis, A. Deksnys, L. Dapkus, V. Jasutis, A. Jukna, Semiconductor Physics Institute, A. Gostauto II, 2600 Vilnius, Lithuania .Evolution of highly oriented2201 and 2212 pbasesin various composition pos1annealed BISrCaCuO films bas been investigated in-situ using high temperature XRD. Magnetron sputtered amorphous Bi-Sr-Ca-Cu-O precursor films and sequentially evaporated Bi(SrF2, CaF2)-Cu mullilayers both prepared onto MgO(IOO) bave been used for the investigations. The superconducting compounds were found to be formed via the intermediate layered Bi- and Sr-based oxide (3-x.x.O,O) following the sequence: (3x,x,O,O) -> 2201 -> 2212. Kinetics of the (3-x,x,O,O) -> 2201 -> 2212 structural transformations bave been investigated. Small amount of a fugitive liquid depending strongly on the annealing schedule bas been indicated at 1'>7500C. At certain conditions the liquid stimulated oriented growth of the intermediate pbase and modified formation con~i~ons of the pbases towards liquid pbase epitaxy. As a result, we succeeded in obtainmg superconducting 2212-phase films with smooth surface being unusual for most postannealed BiSrCaCuO.

B28

HUME·ROTHERY AWARD SYMPOSIUM FOR WILLIAM L• .JOHNSON: THERMODYNAMIC AND KINETIC ISSUES IN THE SYNTHESIS OF METASTABLE MATERIALS II Sponsored by: JL EMPMDISMD AHoy Phases Committee Program Organizers: Professor Carl C. Koch, Materials Science and Engineering Department, North Carolina State University, Box 7907, Raleigh. NC 27695; Dr. Robert D. Shull, NIST. Bldg. 223 B152. Gaithersburg, MD 20899 Monday, PM February 5, 1996

Room: Orange County I Location: Anaheim Marriott Hotel

Session Chairperson: Dr. Robert D. Sbull, NIST, Bldg. 223 BI52 Gaithersburg MD 20899 ' , 2:00 pm Invited RADIATlON·INDUCED CRYSTAL·TO·GLASS TRANSFORMATION IN IN· TERMETALLIC COMPOUNDS: P.R. Okamoto. N.Q. Lam, Materials Science Division, Argonne National Laboratory, Argonne, IL 60439 Molecular dynamic simulations of defect-induced amorphization of intermetallic compounds suggests that static atomic displacements can be used as a general order parameter for both bomogeneous and inhomogeneous defect structures. The results indicate that a unified approach to defect-induced amorphization can be based on a generalized version of the Lindemann melting criterion, which in its simplest form assumes that melting ofa defective crystal occurs when the sum of the static anddynarnic mean square displacement exceeds a critical value identical to that for melting of the perfect crystal. A direct consequence of this hypothesis is that the effective Debye temperature, average shear modulus and melting temperature of the defective crystal, all must decrease with increasing static atomic disorder. Experimental observations and measurements in support of the generalized Lindemann melting criterion will be discussed. 2:30 pm Invited NON·EQUILIBRIUM PROCESSING OF MATERIALS BY ENERGETIC ION IRRADIATION AT IDGH AND LOW TEMPERATURE: R.S. Averback, P. Partyka, Mai Ghaly, Department of Materials Science and Engineering, University of Illinios at Urbana-Champaign, Urbana, IL; Y. Lee, C.P. Flynn, Department of Physics, University of Illinois at Urbana-Champaign, Urbana, 11.; P. Bellon, SRMP, Saclay, France Irradiation with energetic particles is a highly controlled and convenient means to synthesi~ ~~table ~rials. At 10':" temperatures, irradiation drives systems away from eqwhbrium by nnxmg atoms m energetic collision cascades, while at high temperatures, radiation-induced defects restore systems to their equilibrium state. Although ~ concepts are clear, the details of these processes are complex and bave been controvefSlal for many years. Recently, molecular dynamics coroputer simulations bave begun to provide a detailed understanding of the mixing process in cascades, and the results ~f these simulations wil.l be reviewed here. The high temperature phenomena are more difficult to understand because of a competition between disordering processes in cascades, and ordering processes from mobile defects. This competition will be illus!rB~ ~y ~i~tion-induced disordering in order-disorder alloys and interfacial roughen109 IR Imrmsclble systems. Results from both experiment and computer simulation will be presented. 3:00pm SYNTHESIS, STRUCTURAL STABILITY AND PHASE SEPARATION OF AMORPHOUS MoS11Nx: H. Kung, T.R. Jervis, T.E. Mitchell, M. Nastasi, Los Alamos National Laboratory, Los Alamos, NM; J-P. Hirvonen, Tecbnical Researcb Centre of Finland, Espoo. Finland . The structural stability and phase separation of amorphous MoSi2N. films, with D1troge~ conte~t x varyi?g between 0 and 4. bave been studied. Tbe MoSi2N. films were syntheSized usmg reacuve magnetron sputtering technique. Cross-sectional transmission electron microscopy was used to examine the structural evolution of as-deposited an.d. annea1~ (uP. to lOOO'C in "acuum) films. Electron energy loss spectroscopy was utlhzed to mvesugate the local bonding structure. Nanoindentation was employed to measure the corresponding mechanical properties. An amorphous structure is observed in the as-sputtered MoSi2N•. For large nitrogen content (02.5) films, the amorphous s~ture is preserv~ ev~n after lOOO·C annealing but undergoes a pbase separation (spmodal dec~mpos1Uon) mto two pbases. At lower nitrogen concentration, the MoSi2N. films crystalhze to form the C40 structure, which is a metastable structore ofMoSi . The hardness and modulus is observed to vary with the nitrogen content and ~aling conditions. The role of nitrogen in influencing the structural stability, pbase separation and mechanical properties will be discussed and presented. 3:20 pm BREAK 3:40 pm Invited SOFT AND HARD MAGNETIC PROPERTIES OF Fe·RICH NANOCRYSTALLINE ALLOYS CONTAINING INTERGRANULAR AMORPHOUS PHASE: Akihisa Inoue,lnstitute for Materials Research Tohoku University Sendai 980 77, Japan " .The partial crystallizati~n of amorphous Fe ricb Fe-M-B (M=:a. Hf, Nb or Nd) alloys With 90 at'll> Fe concentration was found to cause finely mixed structures of nanoscale bec-F~ particles surrounded by the remaining amorphous pbase in the Zr-, Hf- and Nbconta.t~ng alloys and of nanoscale bec-Fe and bet-FeI4N~B particles embedded in the remalRlDg amorpbous phase. The particle size is about 10 to IS om for the bec-Fe pbase

1996 TMS ANNUAL MEETING AND EXHIBITION

and 20 to 30 nm for the bec-Fe and Fe14N~B pbases. Furthermore, the phase transition of bee-Fe + amorphous to bec-Fe + Fe14N~B + amorphous in a Few(Zr-Nd},B) system occurs around 4 at% Nd. The mixed pbase alloys exbibit good magnetic properties combined with bigh magnetization, i.e., permeability above l.5xl04 at 1 kHz and saturation magnetization above 1.5 T for the former phase alloys, and coercivity of 250 kAlm, remanence of 1.3 T and maximum energy product of 145 kJ/m) forthe latter phase alloys. The appearance of the soft and bard magnetic properties in coexistence with the remaining amorpbous phase for the Fe-ricb alloys is important for future development of new nanocrystalline magnetic alloys.

4:10pm STRUCTURAL EVOLUTION OF NdlFel~ DURING MECHANICAL MILLING: W.F. Miao, J. Ding, R. Street, P.G. McCormick, Researcb Centre for Advanced Mineral and Materials Processing, University of Western Austra1ia, Nedlands, W.A. 6900 The development of metastal3le structures during the mechanical milling ofN~Fe14B has been investigated. Changes in structure and composition accompanying milling were followed with x-ray diffraction, Mosshauer spectroscopy and differential scanning calorimetry measurements. Milling was found to initially result in disordering and amorpbization of the Nd2Fe14B phase. Continued milling caused disproportionation of the amorphous pbase to a two phase a-Fe I amorpbous mixture at longer milling times. Structural and compositional changes ocurring during heat treatment of the as-milled material will also be reported. 4:30pm THE EFFECT OF NITROGEN AND OXYGEN ON THE THERMAL STABIL· ITYOFNANOCRYSTALLINEFeSYNTlIESIZEDBYCRYOMILLING:Benlih Huang, Robert J. Perez, Eurique J. Laverina, Department of Chemical Engineering and Materials Science, University of California, Irvine, CA 92717-2575 Recently, cryomilling of aluminum and NiAl bas been found to result in the in-situ formation of nanoscale g-AI2OJ and AlN, whicb impede the grain growth of the nanocrystaIline structures during heat treatment. The present study bas been conducted to synthesize Fel g-AI20), AlN nanocomposites by cryomilling of Fe-IO wI. %AI in liquid nitrogen. The as-milled powders following 25 bours of cryomilling were consolidated at 550·C at 350 MPa for 30 minutes, and subsequently annealed at elevated temperatures for 60 minutes. 1EM dalt field imaging indicated that nanocrystalline structure with crystallite size of 10-20 om was maintained for the consolidated powders annealed at 950°C. Selected area diffraction (SAD) indicated the formation of g-AI20) and AlN. In contrast, cryomilling of Fe-IO wl.%Al in liquid argon, and cryomilling of pure Fe in liquid nitrogen have not been al3le to produce nanocrystalline structures with thermal stability at 650°C. Therefore, incorporation of nitrogen, oxygen, and aluminum to Fe during mechanical alloying is sbown to be imperative to the achievement ofthermaI stability of nanocrystalline structures. The authors would like to acknowledge the financial support provided by the Office of Naval Researcb under grants NOOOI4-931072 and NOOOI4-94-OO17. 4:50pm VIBRATIONAL ENTROPY OF NI)V: Laura Nagel, Brent Fultz, Department of Materials Science, California Institute of Tecbnology, Pasadena, CA 91125; Lee Robertson, Steve Spooner, Oak Ridge National Laboratory, P.O. Box 2008, OakRidge, TN 37831 We have measured the difference in vibrational entropy for two different ordered states of Ni)V: as a cbemically ordered cubic crystal and as the equilibrium tetragonal ~2 structure. From these measurements, we have determined the Debye temperatures for NiJV in two states of order. Data were obtained from low-temperature calorimetry (80K to 360K). From these data we estimate differences in vibrational entropy between the two states to be on the order of 0.1 kB per atom. About balf of this result can be explained by harmonic contributions, and the other half by anbarmonic effects. In order to interpret these results, we have performed thermal expansion measurements and inelastic neutron scattering experiments. This work was supported by the U.S. Department of Energy under Contract No. DE-FG03-86ER45270. 5:10pm MECHANICAL CRYSTALLIZATION BY CRYOGENIC BALL MILLING: Benlih Huang, Robert J. Perez, Paula J. Crawford, Enrique J. Lavernia, Department of Chemical Engineering and Materials Science, University of califomia, Irvine, CA 92717-2575; Steven R. Null, Department of Materials Science and Engineering, University of Southern California, Los Angeles, CA 90089-0241 Cryogenic ball milling of Metglas F~.BllSi9 in an attritor in liquid nitrogen bas resulted in the synthesis of nanocrystals of a-Fe(Si) and Fe2B with crystallite size of26 nm in diameter. The mechanical crystallization during cryogenic attritor milling consists of two stages: the initial stage involves the crystallization through the formation of shear bands, and the latter stage involves the crystallization through a wear-like mecbanism. To further provide insigbt into the mechanisms of mechanical crystallization, thermal analysis indicated that the formation of a-Fe(Si) consisted of both thermal and athermal mechanisms, wbile that of the F~B consisted of only thermal mecbanism. To understand the effect of composition on the stability of the glassy pbase, elements of Co and Ni were incorporated to the Metglas F~.BllS~ using both cryogenic attritor milling and Spex milling. TEM study and thermal analysis bave been used to investigate the mecbanisms whicb control the stability of Metglas Fe,.B l)S~ during mechanical crystallization. The authors would like to acknowledge the financial support provided by the Office of Naval Researcb under grants NOOOI4-93-1072 and NOOOI4-94-0017.

1996 TMS ANNUAL MEETING AND EXHIBITION

HYDROMETALLURGY:RECENT DEVELOPMENTS II: Progress In Leaching

Sponsored by: EPD Aqueous Processing Committee Program Organizer. Professor David Dreisinger, The University of Britisb Columbia, Department of Metals and Materials Engineering, 309-6350 Stores Road, Vancouver, B.C., Canada V6f lZA Monday, PM February 5,1996

Room: A14-l5 Location: Anaheim Convention Center

Session Chairperson: Professor David Dreisinger, The University of Britisb Columbia, Department of Metals and Materials Engineering, 309-6350 Stores Road, Vancouver, B.C., Canada V6T lZA 2:00pm NEW EVALUATION OF NON·CYANIDE LIXIVIANTS FOR GOLD EXTRACTION FROM COMPLEX AND REFRACTORY GOLD ORES: L.M. Abrantes and Y.Z. Zhang, University of Lisbon, ICAT, Faculty of Science, 1700 Lisbon, Portugal Cyanide bas generated environmental concerns because of its toxicity, and it may not respond to the treatment of complex gold ores and refractory gold ores. Non-cyanide lixiviants thiourea and thiosulpbate are the most promising substitutes of cyanide. However, the commercial adoption of the non-cyanide process has been hindered by three factors: more expensive and higher consumption than cyanide, and the gold recovery step still requires more development. The consumption of gold lixiviants include the stoichiometric usage for gold dissolution, the degradation loss, and the loss due to the impurity interference (fOfDling complexes, precipitates and adsorption). The degradation processes of cyanide, thiourea and thiosulphate systems are discussed on thermodynamics, kinetics and electrocbemistry basis. The influences of various factors on the degradation reactions of gold lixiviants are analysed and illustrated for gold leacbing from different refractory gold ores. New progress on decreasing the consumption of non-cyanide and avoiding gold passivation are summarized. Finally, progresses on gold recovery from the non-cyanide pregnant solutions by various techniques sucb as cementation, CIP, RIP, solvent extraction, bydrogen reduction and electrowinning are reviewed. The potential of non-cyanide leaching as a replacement for cyanidation is discussed. 2:30pm THE LEACIDNG BEHAVIOR OF PLATINUM POWDER IN IODIDE SOLUTIONS WITH OXYGEN: X. MengandK.N. Han, South Dakota Scbool of Mines and Technology, Department of Metallurgical Engineering, 501 East Saint Josepb Street, Rapid City, SD 7701-3995 The leaching behavior of platinum powder in iodide solutions under oxygen pressure in an autoclave was investigated. Ammonium-balogen salts were used at neutral pH as the key lixiviants for the dissolution study of platinum. The effect of temperature, partial pressure of oxygen, concentration of iodide and ammonium salts on the dissolution of platinum was examined. Under experimental conditions, the leacbing behavior of platinum powder exhibited a complex reaction mecbanism. A precipitate was found during the leaching and identified to be platinum ammine iodide, Pt(NH)hI4' The dissolution rate was found to be controlled by a cbemical reaction and limited by the equilibrium concentration of the platinum ammine iodide. A series of cbemical reactions are believed to take place. 3:00pm FERRIC CHLORIDE LEACHING AND METALS RECOVERY OF POLYMETALLIC GOLD ORE: X,I. Guo, Beijing General Researcb Institute for Nonferrous Metals, Beijing, 100088, China; J.L. Hendrix, University of Nevada, Reno, Mackay Scbool of Mines, Reno, NV In this paper, an alternative process to recover the value metals from polymetallic gold ore by FeCI) leacbing was proposed. In the process, gold was recovered by direct cyanidation from the leaching residue, the isolative electrolyte in chloride solution was selected to recover lead, the cementation with copper powder was used to recover silver and the solvent extracting was applied to recover copper and zinc. The total recovery of gold, lead, copper, silver and zinc in the process is 90.8%, 96.05% , 84.3%, 81.98% and 54% to the bullion of gold, lead and silver and sulfate of copper and zinc, respectively. The process conditions selected for FeCI) leacbing are the temperature of 104°C, leacbing time of 6 bours, Fel+ content of 140 gil and stoichiometry of FeCI). 3:30pm PRETREATMENT USING ALKALINE SULFIDE LEACHING AND NITRGOEN SPECIES CATALYZED PRESSURE OXIDATION ON A RE· FRACTORY GOLD CONCENTRATE: Corby G.Anderson, SuzzannM. Nordwick, Sunshine Mining & Refining Company, PO Box 1080, Kellogg, ID 83837 Hydrometallurgical tecbnology for the pretreatment of refractory gold concentrates has been developed at Sunshine. It utilizes the patented nitrous-sulfuric acid pressure leacbing process whicb bas been industrially proven at the Sunshine refinery. Additionally, antimony and some other impurities can be removed by also employing alkaline sulfide leaching. This paper will discuss the successful application of these two technologies for pretreatment prior to CIL gold recovery. A complex gold concentrate composed of pyrite, arsenopyrite and stibnite was tested. Results of antimony leaching in alkaline sulfide media will be discussed including some details of a novel and selective recovery metbod for gold leacbined by polysulfide or thiosulfate. Also, the factors important in and the results of the application of nitrogen species catalyzed pressure leacbing to refractory gold concentrates will be detailed.

B29

4:00pm AN OXIDATIVE CHLORIDE LEACHING OF PALLADIUM POWDER: H. Zhang, X. Meng and K.N. Han, South Dakota School of Mines and Technology, Department of Metallurgical Engineering, Rapid City, SD 57701-3995 The leaching behavior of palladium powder in aqueous solutions containing NaCI or NI4CI was investigated. Cupric iron was used as an oxidant for the dissolution reaction. The effect of a number of leaching parameters such as temperature, concentration, stirring speed and surface area on the leaching rate was investigated. The leaching system containing chloride and cupric ion was found to be very effective for the dissolution of metallic palladium. Based on the experimental results, the overall leaching rate was believed to be controlled by a chemical reaction step. The dissolution rate of palladium was found to be a first order with respect to the concentration of cupric ion and a second order with respect to the concentration of chloride. 4:30pm APPLICATION OF FACTORIAL OPTIMIZATION TECHNIQUE FOR GENERATING KINETIC DATA FOR HYDROMETALLURGICAL PROCESSES: K. Sanjay, S. Anand, R.P. Das, Regional Research Laboratory, Bhubaneswar, 751 013, Orissa, India The statistical optimization technique using full factorial design of experiments is generally applied to determine the boundary conditions for maximum yields of desired products. The response surfaces are generated to provide flexibility of operating parameters without affecting the yield. The regression equations developed from these set of experiments will show the dependence of yield on individual parameters as well on interactions for simultaneous variations of parameters. These experiments do not give an insight into the kinetics or mechanism of various reactions taking place in the process. In the present work, kinetic data is generated from the factorial optimization results for dissolution of manganese from manganese oxide using Sulphuric acid as a leachant and carbon as a reductant. The boundary conditions for a 23 full factorial design were fixed to achieve variations in the yield in the wider range (30 to 90%). The results obtained are discussed both for process optimization and kinetics of leaching reactions. 5:00pm THE QUANTITATIVE DETERMINATION OF SUB-PPM QUANTITIES OF Au AND Pt IN SULFIDE MINERALS: LJ. Cabri, G. McMahon, CANMEr, 555 Booth Street, KIA OGI Canada The ion microprobe (SIMS technique) is a very sensitive instrument for the quantitative determination of trace quantities of precious metals (often at the ppb level) occurring in common sulfide minerals, either chemically-bound or as discrete particles too small for resolution by SEM. The technique typically involves sputtering the surface of sulfide grains to a depth of one j1III and analysis of an area of 62.5 I1Jll. The combination of quantitative mineral-specifc sub-ppm analyses together with depth-profiling and the ability to produce secondary ion maps of trace preciOUS metals makes this a very powerful technique in quantifying the precious metal distribution in refractory and other sulfide ores and process products. Applications of the technique to gold and platinum ores and process products will be discussed.

INTERMETALLICS AND THEIR COMPOSITES: PROCESSING AND APPLICATION TECHNOLOGIES I: Nickel Alumlnldes

Sponsored by: SMD High Temperature Alloys CommIttee Program Organizer. P.W. Keefe, Special Metals Corporation, New Hartford, NY 13413; M. Sohi, Allied Signal Engines, Phoenix, AZ 85072-2118 Monday, PM February 5, 1996

Room:B4 Location: Anaheim Convention Center

Session Chairperson: M. Sohi, Allied Signal Engines, Phoenix, AZ 85072-2118 1:30pm PROCESSING, STRUCTURE AND PROPERTIES OF REACTION-INFILTRATED NIAI AND NIAI COMPOSITES: T.A. Venkatesh, C. San Marchi, D.C. Dunand, A. Mortensen, Massachusetts Institute of Technology, 77 Massachusetts Avenue, Room 4-117 Cambridge, MA 02139 Bulk NiAl is synthesized by infiltration of nickel preforms with molten aluminum and chemical reaction to form the interrnetallic. Continuous fiber reinforced NiAI composites are fabricated with the same process by adding tungsten fibers to the nickel preform. Success of the process is found to be highly sensitive to the relative proportions of aluminum and nickel used. This issue is analyzed, and critical parameters are identified. Microstructures of the materials are characterized, particular attention being paid to macrosegregation, porosity, grain size and fiber-matrix interface structure and reaction. Mechanical properties of the material produced by this process are studied by compression creep testing and discussed for both bulk and reinforced NiAl. 1:50pm ENGINEERING OF TOUGHENED NICKEL ALUMINIDE INTERMETALLICS: R. Rarnasundaram, R. Bowman, Y. He, J. J. Lannutti, W. O. Soboyejo, Department of Materials Science and Engineering, the Ohio State University, 2041 College Road, Columbus, OH 43210-1179 A review of efforts to design and process toughen nickel a1uminide composites is presented in this paper. This includes the use of micromechanics-based design methodologies to engineer the development of composites with improved fracture toughness and fatigue crack growth resistance. The paper will discuss the use of: ductile phase toughening with Mo particles and fibers; transformation toughening with partially

B30

stabilized zirconia (I'SZ); hybrid/synergistic toughening with PSZ and Mo fibers! particles, and functionally graded architectures in the toughening of NiAI. The paper illustrates how the processing and applications technologies have guided the development of composites W1ith improved damage tolerance. The damage tolerance issues that may hurt the potential aerospace applications of toughened nickel a1uminide composites are also discussed. 2:10pm FABRICATION OF STAINLESS STEEL TOUGHENED NIAI INTERMETALLIC COMPOUND MATRIX COMPOSITE: Shou-Yi Chang, Su-Jien Lin, Department of Materials Science and Engineering, National Tsing Hua University, Hsinchu, Taiwan, China The stainless steel fibers with adiameterof 150 11m were wound and then nickel plated to a diameter of about 225 11JII. Alternate layers of the properly spaced Ni-coated stainless steel fibers and aluminum foils were stacked and hot pressed at 500'C, 100 MPa in vacuum for 10 minutes to produce a S.S.c/NilAl composite. A following attractive process, hot pressing at 650'C ISO MPa for a long period of time, 20 hours, allowed the nickel and aluminum to react to form the NiAI interrnetallic compound matrix. Fewer pores were left if the hoI pressing at 9OO"C was followed. The variance of microstructures, which resulted from the different conditions of heat treatment and hot pressing, was analyzed by optical microscopy, scanning electron microscopy and X-ray diffrllCtion. And the analyses revealed thatsome Ni-AI reaction phases existed in the composite. 2:30pm MECHANICALBEHAVIOROFIN-8ITUDIRECTIONALSOLIDIFlEDCr(Mo)l NlAI COMPOSITE: S. M. Jeng, J.-M Yang, Department of Materials Science and Engineering, University of California, Los Angeles, CA 90024-1595; K. Bain, R. A.Amato, GE Aircraft Engines, Cincinnati, OH 45235 The NiAl-based composites reinforced with refractory metals in situ by directional solidification (DS) technique composites have been recognized as one of the most promising candidate matelials for high temperature structural applications. The ductile reinforcing phase in the OS eutectics can assume a fibrous or lamellar morphology. The well-aligned structure can provide significant improvement in strength and toughness over the monolithic NiAI. The Cr(Mo)/NiAl eutectic composite has been successfully fabricated using an Edge-deflned Film-fed Growth (EFG) technique. The mechanical behavior of the composites including tensile properties at room and elevated temperatures, fracture resistance ami creep behavior were studied. The damage evolution and fracture characteristics at room and elevated temperatures were also investigated. 2:50 pm BREAK 3:10pm NiAl SINGLE CRYSTALS WITH HIGH TENSILE DUCTILITY: Igor A. Bul, Vladimir I. Levit, Michael J. Kaufman, Department of Materials Science and Engineering, University of Florida, Gainesville, FL 32611 NiAl single crystals were grown by the Bridgman method from arc-melted rods. Some crystals were prestrained by compression at a temperature of 1273K. The dislocations b=< 100> in the prestrained crystals were primarily located at low angle boundaries. Very low dislocation density was observed inside the subgrain volumes. Tensile tests were performed on both prestrained and undeformed NiAl single crystals oriented in (123) direction. The test temperatures were from 300K to 1273K, with a strain rate of about 10' 4 sec-I. All the samples tested at RT obtained high tensile elongation (up to 13.6% ).The prestrained samples showed slightly lower yield points and slightly higher elongation than undeformed ones at all temperatures. The factors related to ductility of NiAI single crystals are discussed: toichiornetry, point defects and interstitials, dislocation density and mobility, sample design, method of machinery and quality of the sample surface preparation. This work is sponsored by the Air Force Office of Scientific Research (URI Grant F49620-93-0309) under the direction of Dr. Charles Ward. 3:30pm AN ANALYSIS OF TENSILE DUCTILITY OF NIAI [110) ORIENTED SINGLE CRYSTALS: Jeffery S. Winton, Vladimir I. Levit, MichaelJ. Kaufman, Department of Materials Science and Engineering, University of Florida, Gainesville, FL 32611 The tensile properties of (110) oriented NiAI single crystals were investigated between RT and 700·C. The tensile ductility exhibited three temperature regimes. At high temperatures of >500'C, tile ductility increases with temperature because of an increase in strain rate sensitivity, which results in less localized necking. At intermediate temperatures of 200T-500'C, the ductility shows a plateau owing to a loss of strain hardening rate, which results in a reduction of maximum uniform strain. At low temperatures of RT-200'C, fracture occurs after yielding, but before necking. The change of ductility of these three temperature regimes will be discussed in terms of strain hardening behavior, slip localization and necking. This work is supported by the Air Force Office of Scientific Researc:h (URI Grant F49620-93-0309) under the direction of Dr. Charles Ward. 3:50pm THE EFFECT OF ORffiNTATION, TEMPERATURE, AND STRAIN RATE ON THE MECHANICAL PROPERTIES OF SINGLE CRYSTAL NIAI: Jeffery S. Winton, Vladimir I. Levit, Michael J. Kaufman, Department of Materials Science and Engineering, University of Florida, Gainesville, FL 32611 While the mechanical behavior of NiAl has been studied extensively, questions concerning the fundamental deformation mechanisms and mechanical properties of single crystal NiAl still remain. Little tensile testing has been performed, and a review of the literature points out that there is relatively poor agreement for such properties as yield strength and ductility, derived from tensile testing. The tensile properties of single crystal NiAl were investigated between RT and 1273K. The four crystallographic orientations tested were (110), [Ill), (135), and (100). The single crystals tested exhibited higher plasticity and a lower yield point at RT then previously published data for all orientations tested. The dislocation mechanism at RT for these tests will be discussed. Kinking was observed from RT 10 773K in samples slightly deflected from

1996 TMS ANNUAL MEETING AND EXHIBITION

[100]. Strain rate sensitivity was measured using strain rate changes between lx1O-5, lx 10-4 and lxlO-3 and the results for all orientations will be presented. The mechanical behavior corresponds to TEM observations of the dislocation density and distribution. This work is supported by the Air Force Office of Scientific Research (URI Grant F49620-93-0309) under the direction of Dr. Charles Ward.

INTERNATIONAL MOBILITY OF ENGINEERS AND ENGINEERING ACCREDITATION II

Sponsored by: TMS International Activities Committee Program Organizers: Dr. Axel Meisen, Dean of Applied Science & Professor of Chemical Engineering, Office of the Dean, Faculty of Applied Science, The University of British Columbia, 2006-2324 Main MatI., Vancouver, B.C. Canada V6T lZ4; Dr. J. Keith Brimacombe, A1can Chair, Matls. Process Engr., Univ. of British Columbia, Centre for Met. Process Engr., 309-6350 Stores Road, Vancouver, BC V6T lZ4, Canada; Dr. William P. Imrie, Principal Engineer, Bechtel Group Inc., 50 Beale Street, San Francisco, CA 94105; Prof. Carl McHargue, Univ of Tennessee, Dept of Matls Sci & Engrg, 102 Estabrook Hall, Knoxville, 1N 37996-2350; Dr. Fiona Doyle, Univ of California, 551 Evans Hall #1760, Berkeley, CA 94720-1760; Peter P. Schepp, Managing Director DGM, Adenauerallee 21, Oberursel, D-6370, Germany Monday,PM February 5, 1996

Room: All-12 Location: Anaheim Convention Center

Session Chairs: TBA The increasing globalization of the world's economies demands that engineers practice

internationally and that barriers to their mobility across national boundaries are eliminated or, at least, minimized. Many countries impose licensing requirements on engineers to ensure that they have the appropriate education and experience for practice. Two sessions are planned, consisting of papers and panel discussions, to examine means of increasing the mobility of engineers (especially for countries with stringent licensing requirements such as Canada and the USA) and of widening the recognition of highquality engineering programs offered in countries with and without accreditation systems. The sessions will conclude with identifying special initiatives for ~~rs of TMS and related organizations regarding international mobility and accreditation.

MATERIALS AND TECHNOLOGIES FOR MICROELECTRONICS II

Sponsored by: EMPMD Electronic Packaging and Interconnection Materials Committee Program Organi'[J!rs: M. McCormack, AT&T Bell Laboratories, Rm IA-123, 600 Mountain Ave., Murray Hill, NJ 07974; S. K. Kang, ffiM, T. J. Watson Research Center, P. O. Box 218, Yorktown Heights, NY 10598; M. R. Notis, Department of Materials Science and Engineering, Lehigh University, 5 East Packer Ave., Whitaker Laboratory, Bethlehem, PA 18015-3195; K. N. Tu, Department of Materials Science and Engineering, UCLA, Los Angeles, CA 90095-1595 Monday, PM February 5, 1996

Room: Grand K Location: Anaheim Marriott Hotel

Session Chairperson: S. K. Kang, ffiM, T. J. Watson Research Center, P. O. Box 218, Yorktown Heights, NY 10598 2:00 pm Invited MICROSTRUCTURES AND PROPERTIES OF NEW, ALTERNATIVE SOLDERALLOYS WITH LOWER MELTING POINTS: M. McConnack, S. Jin,G.W. Kammlott, AT&T Bell Laboratories, Murray Hill, NJ 07974 Lower temperature processing in manufacturing allows less expensive materials with greater temperature sensitivities to be used and lower overall production costs. Several new PI>-free and Pb-bearing solder alloys have been developed with the goal of meeting various material property requirements. The metallurgical design of these new solders will be discussed in terms of their strength, creep-resistance, strain-rate sensitivity, ductility, and fatigue properties. In addition, the wetting behavior of these new, alternative alloys will be discussed in terms of wetting balance data and area of spread tests. The effects of different surface fmnishes, atmospheres, and fluxes on the wetting behavior of the new alloys will also be discussed. 2:25 pm Invited SOLDER INTERACTIONS WITH MULTICOMPONENT SUBSTRATE CHEMISTRIES: P. T. Vianco, R. Grant, Sandia National Laboratories, Albuquerque, NM 87185 Traditional solder joints are comprised of a tin-based solder in contact with copper, or in some cases, with the nickel finish of a low expansion alloy I/O. The simple tin! copper metal couple, and the interface reactions which characterize this binary couple, have been found to adequately describe such cases. On the other hand, complex, multicomponent substrates are found in thick film conductors used on hybrid microcircuits. These systems present an opportunity to examine the growth kinetics and chemistry hierarchy ofinterfacial phases which develop between the solder and substrate materials or complex chemistries. A study was performed to investigate the system of 63Sn37Pb solder on 76Au-2IPt-3Pd (wt%) alloy. TIme dependent solid state growth kinetics were evaluated for aging temperatores of 55°C - 170°C and time periods of up

1996 TMS ANNUAL MEETING AND EXHIBITION

to 5000 hours. The interface phase compositions were determined by electton microprobe analysis. A parallel study was performed on the 63Sn-37Pb/lOOAu system to serve as a baseline to the 63Sn-37PblAu-Pt-Pd work. The experimental results will be used to assess the engineering of solder/substate systems which minimize such interactions. This work was performed at Sandia National Laboratories and supported by the U.S. Dept. of Energy under contract DE-AC04-94AL85000. 2:50 pm Invited NOBLE METAL CIRCUIT BOARD SURFACE FINISHES FOR NEW ASSEMBLY TECHNOLOGIES: U. Ray, G. M. Wenger, I. Artaki, D. A. Machusak, AT&T Bell Labs, Princeton, NJ 08542 The suitability of noble metal printed wiring board surface finishes was assessed for new technology applications thai incorporate both direct chip attachment (wire bonding) and peripheral leaded device attachment (soldering) on the same product vehicle. The robustness of the surface finishes to temperature and humidity exposures was evaluated first for screening purposes. The performance of potential candidates was then assessed relative to wire bondability and solder joint integrity. Gold and aluminum balUwedge bonding characteristcs were monitored as a function of assembly processes. Solder joint integrity was determined as a function of temperature cycling and thermal shock. 3:15 pm Invited COMPARISON OF SN-PB AND SN-AG SOLDER JOINTS SUBJECTED TO VARIOUS THERMAL CYCLING PROFILES: Y.-H. Pao, S. Badgley, E. Jih, R. Govila Research Laboratory, Ford Motor Company Dearborn, MI 48121-2053 The thermal cyclic and fatigue behavior of 63Sn-37Pb, 90Pb-IOSn, 63Sn-O.7SI>-IAg-lIn-Pb (SS-IOO) and 96.5Sn3.5Ag solder alloys are studied using a modified biomaterial specimen. The specimen is subjected to temperature cycling with two different profiles, one is between -40'C and 125°C with a two hour and 15 minute period, and the other has a 4O-minute period between 4O'C and 140'C with a 10 minute ramp rate. The shear stress and strain in the joint during thermal cycling are measured with the aid of four high temperature strain gages. The steady state creep parameters of these solders are determined based on the stress relaxation data. Viscoplastic constitutive equations are assumed for the solder alloys and are implemented in a fmite element programABAQUS to model the test for each solder alloy. The fracture surface as a result of thermal cycling is examined with scanning electron microscopy to identify the dominant failure mode and associated failure mechanisms. In addition, a comparison of the evolution of hysterics loops is also made for the above solder alloys, and the results are used to develop a damage parameter in the constitutive equations and to establish an accelerated thermal cycling factor or the solderjoint. The results serve as part of the effort at Ford to assess Pb-free solders versus some conventional PI>- containing solders. 3:40 pm BREAK 3:50pm MICROSTRUCTURAL EVOLUTION DURING NATURAL AGEING OF A Sn-Ag SOLDER: A. Gibson, T R. Bieler. K.N. Subramanian, Department of Materials Science and Mechanics, Michigan State University, East Lansing, MI 48824-1226 The need for lead-free solders in automotive electronics in the next few years requires an investigation of how mechanical properties are affected by the microsttocture and microsttoctural changes that occur in an automotive environment. For automotive applications, solders must withstand thermal fatigue resulting from termperature chages from -40°C to 150°Cas well as mechanical fatigue from vibration. Lead-tin solders have shown improved fatigue resistance when the size of microstructural features can be kept smail by pinning strategies that prevent coarsening of the microstructure. Consequently the natural ageing behavior of Sn-Ag solders is investigated. In the soldering process used in production, the heating and cooling history, along with the size of the joint, will affect the microstructural scale within the solder joint. Several specimen configurations having different geometry and size that are convenient for laboratory testing of creep and fatigue properties will be presented, as well as comparisons of the microsttocture with realistic solder joints. In each of these, the evolution of microsttocture and the hardness is monitored with time. These observations will be compared with feature sizes and cooling rates following solidification. 4:10pm KINETIC ANALYSIS OF THE SOLDERING REACTION BETWEEN EUTEC· TIC Sn-Pb AND Cu: H.K Kim. K.N. Tu, Department of Materials Science and Engineering, UCLA, Los Angeles, CA 90095-1595 Owing to the advance in VLSI technology, there are increasing concerns about the functionality and reliability ofinput-outputconnections using solder joints. The reliability of such connections is, however, degraded by intermetallic compounds formation between the solder and metal substrate. The morphology and kinetics of Cu-Sn intermetallic compound formation in the eutectic Sn-Pb/Cu soldering reaction have been studied. At the interface, Cu-Sn compounds do not form layered structures, rather the CU6Sn5 phase grows as scallop-like grains and ripening occurs between the grains. It is a complicated growth process. We have analyzed tbe growth by a combination of two fluxes: the flux of interfacial reaction and the flux of ripening. The mathematical model will be presented. Also we have measured the rate of consumption ofCu substrate in the soldering reaction. 4:30pm ASSEMBLY YIELD PREDICTION USING SURFACE EVOLVER WITH SOLDER PROCESS VARIABLES AS INPUTS: D. Lewis, M. R. Notis, Department of Materials Science and Engineering, Lehigh University, Bethlehem, PA 18015-3195; G. C. Munie, AT&T Network Systems, Naperville, IL 60566; D. Noctor, AT&T Bell Laboratories, Allentown, PA 18103 Surface Evolver,a public domain program, was used to predict solder joint geometry using component and assembly parameters that effect solderability. To verify modeling results, as assembly trial was completed using different combinations of solder volume, component mechanical tolerance, component aging condition, and solder alloy. This

B31

study was developed to correlate manufacturing yields to theoretical modeling, and the final application is intended to provide a simple method for surface mount assemblen to predict solderability and yield. Application examples fromSulj'ace Evolvermodeling will be provided to highlight results from the assembly trial. 4:!Opm COMPARISON OF Sn-Ag AND Sn-Ag-BI LEAD-FREE SOLDERS ON ENTEK COATED PCBs: A. Grusd. Heraeus Inc., Cermalloy Division, West Conshohocken, PA 19428 In addition to eliminating lead, many manufacturers desire to develop and use higher melting point solders as compared to eutectic tin-lead in order to meet demands of higher operating temperatures. Solder alloys of eutectic Sn-Ag (Tm=22 I 0c) and Sn-Ag-Bi (l88°C to 218C) were made into psstes, using both no-dean and RMA fluxes, and printed onto Entek coated PCBs, and then reflowed. The Sn-Ag-Bi exhibited improved wettsbility over Sn-Ag in air reflow on Entek. The wetting characteristics of the solder pastes on chip capacitors with various end termination metallurgies including SnIPb, Ag, AglPd, and AgIPdIPt will also be discussed. 5:10pm THE DESIGN OF PB-SN BASED COMPOSITES FOR INTERCONNECT APPLICATION IN MICROELECTRONICS: R.K Mahidhara, S. Hariprasad, High Performance Materials Inc. (HPM), Campus Box 1087; S. Baybnam, S.M.L. Sastry, K.L. Jerina, Mechanical Engineering Department, Campus Box 1185, Washington Univenity, St. Louis, MO 63130 Solden are critical materials for preserving the integrity of components in microelectronic psckaging. Current tinlead solders lack sbear strength, resistance to creep and resistance to thermal-mechanical fatigue damage. Pb-Sn based in-situ composites were developed by rapid solidification processing (RSP) following additions of suitable ternary alloying elements to Pb-Sn eutectic. The fme microstructure due to RSP has culminated into the development of two kinds of in-situ composite solders for application in microelectronics as solder paste. One kind of solder exhibits higber tensile strength, creep and high fatigue resistance compared to commercially available Pb-Sn eutectic solders. The other kind exhibits higberductility and low cycle fatigue resistance. Tensile, creep and fatigue tests were conducted on single lap sbear joints made with each solder composition to assess the reliability of the composite solders. The performance of HPM's composite solders exceed those for commercially available lead-tin solders.

MATERIALS PROCESSING FUNDAMENTALS SYMPOSIUM II

Sponsored by: JI. EPDIMDMD Synthesis, Control, and Analysis In Materials Processing Committee, EPD Process Fundamentals, Aqueous Processing, Copper, Nickel-Cobalt, Pyrometallurgy, Lead, Zinc, Tin Committees, MSD Thermodynamic &: Phase Equilibria Committee Program Organizers: R. G. Reddy, Department of Cbemical and Metallurgical Engineering, University of Nevada, Reno NV 89557; S. Viswanathan, Oak Ridge National Lab., Oak Ridge, TN 37831-6083; J. C. Malas, Wright-Patterson AFB, OH 45433-6533

Monday, PM February 5, 1996

Room: A16-17 Location: Anabeim Convention Center

Session Chairpersons: D.G.C. Robertson, Department of Metallurgical Engineering, Univenity of Missouri-Rolla, Rolla, MO 65401; S. Seetharaman, Division of Theoretical Metallurgy, Royal Institute of Technology, Stockholm, Sweden

1:00pm THE BOUNDARY LAYER AND SOLID-GAS REACTIONS: Farhi Habashi, DepartmentofMiningandMetaliurgy,LavaiUniversity,QuebecCity,CanadaGIK7P4 Solid-gas reactions of metallurgical importance can be classified into three groups: I) physical processes, e.g., drying and sublimation, 2) cbemical processes, which may be an addition or a thermal decomposition reaction, and 3) electrocbemical processes, which may be addition, oxidation-reduction, or displacement reactions. Examples are given and tbe role of the boundary layer is emphasized. 1:15pm CONVERSION OF ILMENITE TO TIC ANDFe USING A NON-TRANSFERRED ARC THERMAL PLASMA REACTOR: Patrick R. Taylor, Ilaria Accorsi-Lamb, Department of Metallurgical and Mining Engineering, College of Mines, Univenity of Idaho, Moscow ,10 83843. Milton Manrique, Department of Materials Science, Simon Bolivar University, Caracas, Venezuela Ilmenite concentrate and methane were fed into a non-transferred thermal plasma reactor to produce ultrafine powders of TIC and Fe. Free energy minimization plots were used to predict the feasibility of the reaction. An estimate of the optimum molar ratio between the ilmenite concentrate and methane was made and used in the experiments. Argon-hydrogen mixtures were used as plasma gases. Power input, carrier gas flow rate, and feeding rate were the variables studied on the behavoir of the system. A mathematical model to describe the distribution psttern in the hot zone of the reactor was formulated. The product has been characterized by x-ray diffraction and atomic absorption. It was found that minor amounts of TI2~, Fe2C, and free C are also present in the product. Finally, the possibility of directly uniaxial pressing and sintering to produce a Fe-TiC based cermet is discussed. 1:!Opm A THERMODYNAMIC ANDKINETIC ANALYSIS OF THE CARBOTHERMIC REDUCTION OF KALGOORLIE PENTLANDITE CONCENTRATE IN THE

832

PRESENCE OF LIME: Jillian Spindura, Animesh Jha, Depsrtment of Materials Technology, Brunei University, Kingston Lane, Uxbridge UB8 3PH, England This work investigates the carbothermic reduction of the complex sulphide mineral, pentlandite, in the presence of lime. The effects of change in temperature and the use of different carbon sources are studied and the role of these variables on the reducibility of pentlandite mineral is discussed. The microstructures of the beat-treated pellets are examined using X-ray diffraction and scanning electron microscopy techniques. The latter analytical method also allows any variation in the morphology and composition of the metallic phases formed, with temperature, to be considered. The kinetics of the reduction process is studied and the magnitude of the overall activation energy is determined from the plots of percentage reduction versus time. 3:15 pm BREAK 3:15pm THERMALDECOMPOSmON OF ZIRCON IN A NON-TRANSFERRED ARC THERMAL PLASMA FLOW REACTOR: Milton Manrique, Judith C. Gomez, Department of Materials Science, Simon Bolivar Univenity, Caracas, Venezuela; Patrick R. Taylor, Department of Metallurgical and Mining Engineering, College of Mines, University of Idaho, Moscow, 10 83843 The thermal decomposition of zircon concentrates in a non-transferred arc thermal plasma flow reactor was investigated. 1be plasma torch was operated using argonbelium and argon-nitrogen mixtures as plasma gases. 1be effect of power, feed rate and zircon particle size was studied. Products were characterized by XRD, SEM, TEM and chemical analytical techniques. The average size of the ceramic product is in the submicron range. Free energy minimization plota were used to predict the behavior of the system under different operating conditions. A kinetic model todescribe the different stages of the thermal decomposition process is proposed. The results of this investigation showed a significant segregation process wbere zirconia-rich product is collected in tbe reactor while silica-rich product is collected in the filter. lEM analysis indicated that silica vapor is condensed on the surface of the zirconia particles. 3:!Opm REDUCTION PROCESS OF IRON ORES IN A TWO-STAGE FLUIDIZED BED SYSTEM: Yoon-Bong Hahn, Yeon-Ho 1m, Jac-Sung Lee, Department of Chemical Engineeringand Technology, Chonbuk National University, Duckjin-Dong IGa, Chonju 561-756, Korea A mathematical model has been developed to analyze and predict the prereduction and final reduction process of iron ore particles in a two-stage fluidized bed system. This model includes the degradation pbenomena of iron ore particles sized 1-5 mm to elucidate its effect on reduction degree. Although the degradation of particles is dependent on the kind of iron ores, it was found that 80 to 90% of overall degradation occurs in the prereduction stage mainly due to thermal stress at reaction temperature. The degradation of particles in the final stage is not much, but is mainly due to interparticle collision and abrasion. In spite of different kinds of ore, the reduction degree showed similar results. It was found that to obtain a targeted reduction degree, it is important to control the residence time of particles in reactor RI in which particles larger than 0.5 mm are fluidized. Optimurnrange of residence time was 15-20 minutes in the prereduction stage; 30-35 minutes in the final reduction stage. 4:15pm LIME ENHANCED REDUCTION OF MOLYBDENITE WITH CARBON: R. Padilla, M.C. Ruiz, Department of Metallurgical Engineering, Univenity ofConcepci6n, Concepci6n, Chile; H.Y. Sobn, Department of Metallurgical Engineering. Univenity of Utah, Salt Lake City, UT 84112-1183 A thermodynamic analysis of the MoS2-C-CaO system has been carried out in order to undentand the reduction of molybdenite with carbon in the presence of lime. Reduction experiments have also been carried out with various mixtures ofM~:C:CaO in the temperature range of 900-1200-C. It was found that the reduction of Mo~ in this system proceeds through the formation of intermediate oxidized molybdenum species M~ and CaO·MoO], which subsequently undergo reduction by CO. In this respect the experimental findings are in agreement with the thermodynamic analysis. For mixtures of MoS 2:C:Cao 1:2:2 at a temperature of 1200"C, complete conversion ofMo~ can be obtained in less than 20 min.; the corresponding product species were metallic Mo, MIIlC, and CaS.

=

MATERIALS SCIENCE OF CONTACTS AND METALLIZATION AND INTERCONNECTS I Sponsored by: EMPMD ThIn FBm &: Interfaces CommIttee Program Organizers: Jagdish Narayan, Dept of Matls Sci & Engrg, Box 7916, NC State Univenity, Raleigh, NC 27695; John Sanchez, Advanced Micro Devices, M3 160, PO Box 3453, Sunnyvale, CA 94088

Monday, PM February 5, 1996

Room: Orange County 5 Location: Anaheim Marriott Hotel

Session Co-chairs: D. Crawford. National Science Foundation, Arlington, VA 22230; J. Narayan, North Carolina State University, Raleigh, NC 27695-7916

1:55 pm Opening Remarks J. Narayan, North Carolina State University, Raleigh, NC 27695-7916 1:00 pm invited THE SCIENCE AND APPLICA.TION OF LOW TEMPERATURE GROWN GaAs FOR NONALLOYED OHMIC CONTACTS: J.M. Woodall, T.P. Chin, M.R. Melloch. Department of Electrical and Computer Engineering, Purdue Univenity, West

1996 TMS ANNUAL MEETING AND EXHIBITION

Lafayette, IN 47907-1285; M.P. Patkar, Villesse Semiconductor Co., Camarillo, CA 93010 In spite of voluminous research on metal contacts to compound semiconductor materials and devices, very little is known about the basic charge transport physics at the metal-semiconductor interface of widely used contacts. For example, there is still no consensus concerning the Schottky barrier physics for metal/GaAs interfaces used for gate electrodes and diodes. Also, the exact transport physics and its relationship to interface structure and chemistry for the Au-Ge-Ni ohmic contact to GaAs is not well understood. Here we discuss the transport physics and ohmic contact applications of well defmed MBE grown and non-alloyed contact structures using low temperature grown (LTO) GaAs. For n-type contacts we use a tandem structure of2 nm of LTG on a 10-100 om thick layer of GaAs doped with Ix I ()20 Si atoms-cm· 3. Ohmic electron transport from the metal to the bulk GaAs occurs by a tandem processes, first via a mid-gap band of deep donor states in the LTO GaAs and then via tunneling though the high density space charge layer caused by the highly Si doped layer, with a resulting non-alloy contact resistance of mid JO.7 ohm-cor. For non-alloyed ohmic contacts to p-type GaAs we grow at 350'C and dope with 1x1()20 Be atoms-cm·3. The key here is that growth at 350'C produces As rich GaAs which in turn greatly reduces the rapidly diffusing interstitial Be concentration, hence very high surface space charge densities can be obtained leading to a non-alloyed contract resistance of <10'7 ohm-cor, without degrading device performance by uncontrolled Be diffusion. 2:30 pm Invited SOLID·PHASEREGROWTHOFCOMPOUNDSEMICONDUCTORSATLOW TEMPERATURES: APPLICATION TO SHALLOW, LOW-RESISTANCE OHMIC CONTACTS: T. Sands, Department of Materials Science and Mineral Engineering, University of California, Berkeley, CA 94720-1760; L.C.Wang, DepartmentofElectrical Engineering, Texas A&M University, College Station, TX 77843-3128 Bilayer metallizations consisting of a first layer ofPd or Ni and an overlayer of a group III, IV or V element react with III-V semiconductor substrates in two stages. In the first stage, reactions occur independently at both interfaces as the systems move toward local equilibrium. In the second stage, the reaction between the two metallization layers dominates, and the Pd-III-V (or Ni-llI-V) reaction product decomposes to release Pd (or Ni) and regrow the III-V compound. Since the proposal and demonstration of this solid-phase regrowth (SPR) process [Sands, MarshaJJ and Wang, J. Mater. Res. 3 (1988) 9141, SPR has provided a mechanism for the controlled modification of III-V surfaces at low temperatures «350·C) with principal application in the fabrication of shallow, low-resistance ohmic contacts. In this talk we will review the SPR mechanism and highlight recent advances in the understanding of doping by SPR, the application of SPR contacts to p-InP and the development of SPR contacts to n-GaAs using processing temperatures as low as 150·C.

3:00 pm Invited ELECTRICAL AND MICROSTRUCTURAL CHARACTERISTICS OF Cu/Ge OHMIC CONTACTS TO N·TYPE GaAs: M.O. Aboelfotoh, S. Oktyabrsky, J. Narayan, Department of Materials Science and Engineering, North Carolina State University, Raleigh, NC 27695-7916 We show that Cu-Ge'alloys prepared by depositing sequentially Cu and Ge layers onto GaAs substrates at room temperature followed by annealing at 4OO·C form a low-resistance ohmic contact to n-type GaAs over a wide range of Ge concentration that extends from 15 to 40 at %. A contact resistivity of(4-6)10·7 Ocm2 is obtained on n-type GaAs with doping concentrations of _lx)(Jl7 cm-3. The contact resistivity is little affected by varying the Ge concentration in the range studied and is not influenced by the deposition sequence of the Cu and Ge layers. 10 addition, the contacts are electrically stable during annealing at 450·C after contact formation. Cross-sectional high-resolution TEM results show that Cu reacts only with Ge to form the ~ and £I,-CU3Ge phases, and that the interface between the contact metal and the GaAs substrate is planar and structurally abrupt The results suggest that the formation of the ~,and £ I-Cu3Ge phases leads to the formation of a highly doped n'-GaAs surface layer which results in the low contact resistivity. 3:30 pm BREAK

4:00 pm Invited RECENT DEVELOPMENT OF OHMIC CONTACTS FOR COMPOUND SEMI· CONDUCTORS: M. Murakami, T. Oku, Y. Koide, and K. Tsukui .. Department of Materials Science and Engineering, Division of Engineering Science, Kyoto University, Sakyo-ku, Kyoto 606-01 Japan Recent several breakthroughs in wide energy gap compound semiconductor materials have inspired many researchers to develop new electronic devices capable of high temperature and/or high power operation, blue light-emitting-diodes,and ultra-short wave length lazer-diodes. One of the restrictions for mass production of these devices is lack of low resistance Ohmic contact materials which satisfy the device requirements. In this paper our current status of development of Ohmic contacts for n-GaAs, p-GaN, and p-ZnSe are: briefly reviewed. Special emphasis will be made on recent progress of the intertnetallic compound contacts to n-GaAs. 4:30 pm Invited

CRYSTAL STRUCTURE AND CHEMISTRY OF NOVEL Cu-GeOHMIC CON· TACT LAYERS: S. Oktyabrs/cy, M.O. Aboelfotoh, J. Narayan, Department of Materials Science and Engineering, North Carolina State University, Raleigh, NC 27695-7916 We report systematic studies of chemistry, structure and electrical characteristics of novel Cu-Ge alloyed ohmic contacts to n-GaAs with very low specific contact resistivity (4xlO-7 cm2 for n-lxI()I7cm·3). Results of TEM investigations on chemistry of phase formation, crystal structure and defect microstructure in Cu-Ge alloyed layers are presented. Structure and properties of Cu-Ge contact layers having lower and higher Ge concentrations from the stoichiometric Cu3Ge composition are compared. Unique properties of the contact layers, namely low specific contact resistiVity, high thermal stability, interface sharpness and high contact layer uniformity are related to the

1996 TMS ANNUAL MEETING AND EXHIBITION

formation of an ordered orthorhombic £1-Cu3Ge phase with average lattice parameters: 5.30,4.20 and 4.56 A. High-resolution TEM has been used to study the ordering in the £ICu)Ge which is responsible for orthorhombic distortion of the parent hexagonal ~­ phase. No compounds due to the chemical reactions with GaAs were found in the alloyed layer as well as in the interface region, demonstrating chemical inertness of the £1-Cu3Ge ordered phase with respect to GaAs. 5:00 pm Invited METAL CONTACTS TO III-V COMPOUND SEMICONDUCTORS: W.O. Barnard, G. Myburg, F. D. Auret, W. E. Meyer, Department of Physiscs, University of Pretoria, South Africa In this talk a summary will be given of Schottky barrier height data on (1 (0) OMVPE grown GaAs, as well as on methods to improve Schottky and ohmic contacts to loP. The Schottky barrier height data of 41 metals processed in the same laboratory, following the same cleaning procedure and using the same high quality OMVPE grown (100) n-type GaAs, will be presenteL From I-V and C-V measurements it was found that there exist no obvious linear relationship between Schottky barrier height and metal work function as is suggested by the Schottky-Mott theory, if all the above-mentioned metals are taken into account. Similar results were obtained if the metal work function was replaced by the Pauling or Meidema electronegativities. In contrast with this, if only metals are chosen with high melting points which were deposited by means of an electron gun, a linear relationship does exist belween Schottksy barrier height and metal work function. The Schottky barrier heights ofRu contacts to InP showed a dramatic increase from 0.49 eV befare passivation to about 0.88 eV afIer wet chemical passivation in an acid containg Ru ions.

MELT.SPINNING, STRIP CASTING, AND SLAB CASTING II: Emerging Technologies And Modelling

Sponsored by: MDMD SoIldlftcation and SCAMP Committees Program Organizers: E.F. Matthys, Mechanical Engineering Department, University of CaJifornia, Santa Barbara, CA 93106; W.G. Truckoer, Technical Director--Product Development, Alcoa Technical Center, Alcoa Center, PA 15069 Monday, PM February 5,1996

Room:B3 Location: Anaheim Convention Center

Session Chairperson: D. Thoma, Los Alamos National Laboratory, MST-6, Mail Stop G770, Los Alamos, NM 87545; W. Truckner, Alcoa Technical Center, Alcoa, PA 15069 2:00 pm Invited PROGRESS IN STEEL CONTINUOUS CASTING: Alan W. Cramb, Department of Materials Science, Carnegie Mellon University, Pittsburgh, PA 15213 In the last ten years steel casting has gone through a revolution that has allowed new steel mills to be constructed for flat products for the first lime since the early sixties. These new mills will contribute 18 million tons of extra capacity by 1997 and mark the largest expansion of the US Steel industry since the second world war. The casting of thin slabs has caused this expansion and yet strip casting may not be far from a commercial reality. This paper will focus on reviewing thin slab and strip casting in North America and around the world to critically assess the technologies and discuss potential future trends.

2:40pm THERMAL ISSUES IN SHAPE DEPOSmON MANUFACTURING: Cristina H. Amon, Kevin S. Schmaltz, Mechanical Engineering, Carnegie Mellon University, Pittsburgh, PA 15213 Shape Deposition Manufacturing (SDM) is a layered manufacturing process that combines the benefits of Solid Freeform Fabrication (SFF), thermal spray and other manufacturing processing operations for fabricating multi-material objects of arbitrary three-dimensional geometric complexity with controlled microstructures and for embedding electronic components and sensors in conformal shape structures. SDM involves molten metal droplet deposition, remelting and solidification as well as the use of sacrificial support materials. Therefore, a thorough understanding of the thermal aspects of the process is required for selecting SDM process parameters and for enabling the integration with SDM CAD-hased design systems. Important thermal issues in the production of high-quality SDM objects are the creation of inter-layer metallurgical bonding through substrate remelting, the control ofcooling rates of both the substrate and the deposition material, and the minimization of residual thermal stress build-up which may induce warping and deboding between deposited layers. This paper presents brief descriptions of the thermal modeling approach, numerical predictions of the cooling rates and substrate remelting depths of steel deposited on steel and on copper, and experimental verifications by temperature measurements and optical metallography. 3:05pm FACTORS LEADING TO THE MICROSTRUCTURAL REFINEMENT AND ELIMINATION OF MICROPOROSITY IN DROPLET·BASED FREEFORM FABRICATION OF STRUCTURAL MATERIALS: M. Ortne, J. Courter, C. Huang, K. Willis, Department of Mechanical and Aerospace Engineering, University of California, Irvine, CA 92717-3975 The development of a droplet-based solid freeform fabrication technique for the fast manufacture of structural components directly from CAD information is presented. The technique relies on the use of precisely controlled molten nano-Iiter droplets as the deposition element. Encouraging experimental results indicate that typical material porosities as low as 0.03% and microstructures with grain sizes on the order of a micron can be achieved with molten droplet throughout rates as high as .18 mils with a single

B33

droplet stream. Extremely high control of the droplet trajectories (measured angular dispersions of the order of I micro-radian), and droplet speeds (measured speed dispersions of the order of 3 x 10-7 times the average stream speed) provides the ability for precise control over the solidification characteristics and hence microstructure. Experimental results in combination with numerical analyses on the solidification characteristics of the droplets determine a control scheme for achieving a component with minimum porosity and microstructure. 3:30 pm BREAK 3:50pm ON THE HEAT TRANSFER AT THE INTERFACE BETWEEN A SOLIDIFYING METAL AND A SOLID SUBSTRATE: G.-X.Wang,E.F. Matthys, Department of Mechanical Engineering, University of Califomia, Santa Barhara, CA 93106 The interfacial thermal conductance between a solidifying molten metal layer and a cold substrate is a critical parameter for many applications in material processing and manufacturing such as continuous slab casting, strip casting, melt-spinning, spray deposition and others. Relatively few quantitative data on the interfacial thermal conductance for such applications are available, however, and -accordingly- we have initiated an experimental study to investigate further this issue. In these experiments, a thin molten metal layer was put into contact with a thicker solid metal substrate. The interfacial heat transfer coefficient between the melt and substrate can then be estimated by matching calculations from a phase change model to the measured temperature of the top surface of the layer. Estimates of the variation of this coefficient with time can then be generated. The results indicate that the interfacial heat transfer coefficient varies significantly as the liquid/solid interface becomes a solid/solid interface. The results also showed a strong effect of the layer/substrate material combination, of the melt superheat, and of the substrate surface roughness on the interfacial conductance. Some of the results from these parametric studies will be presented and discussed. The effect of the surface roughness on the thermal conductance will also be discussed in terms of the contact surface characteristics. 4:15pm FLUID AND HEAT FLOW IN PLANAR-FLOW MELT-SPINNING: EXPERIMENTS AND MODELING IN THE LINEAR GROWTH REGIME: Paul H. Steen, Cornell University, Department of Chemical Engineering, Ithaca, NY 14853; Tetsuhara Ibaraki, Nippon Steel Corporation, Kimitsu,Japan The success of steady planar-flow casting requires that the rates of delivery of mass and removal of latent heat from the solidification front be compatible. This requirement couples the fluid mechanics and heat transfer. On the other hand, in the regime of linear (on average) front growth, the fluid flow is only weakly coupled to the heat flow. We study this regime by casting 5 cm wide aluminum ribbons on a copper-beryllium wheel in air. High speed cinematography of the menisci (yielding the puddle length) and temperature probes within the substrate wheel (yielding the mold temperature) give the key measurements. These data are input into ID fluid and heat flow models. The heat flow model delivers the heat-transfer coefficient (puddlelwheel) to within 5%. Both models (independently) deliver a growth rate. In addition to these results, we shall present evidence that the 'sink-of-masss' effect of the solidification front exerts a considerable influence on the fluid flow. 4:40pm CONTINUOUS CASTING AND RAPID SOLIDIFICATION OF WIRES PRODUCED BY SHAPE FLOW CASTING - SFC-TECHNIQUE -- PROCESS DEVELOPMENT, PARAMETER SIMULATION, MICROSTRUCTURES: G. Frommeyer, W. Frech, Max-Planck-Institut fiIr Eisenforschung GmbH, Diisseldorf, Germany A new production technique for wires by continuous shape flow casting has been developed in order to reduce the steps required to fabricate wires up to 3 mm in diameter by conventional processing routes. An instrumented and computerized melt spin facility allows quantitative parameter studies with subsequent process modelling. The achieved microstructures are correlated with the governing process parameters, such as superheat of the melt, velocity of the melt jet, nozzle geometry and adjustment, etc. The paper describes the principles and the technology of the SFC-process which enables a flexible production of wires out of aluminum, nickel hase alloys, and stainless steels. 5:05pm SIMULATION OF MOLTEN STEEL FLOWFIELD IN INTERMEDIATE TIDCKNESS SLAB CASTER OF THE TSJ'TM PROCESS: Sung-Woo Lee, Changwon Research Institute, Sarnsung Heavy Industries, Changwon Co., Korea The TSP is a new method of thin slab casting and cost-effective getting to high quality hot strip production designed to satisfy the mini-mill company. Here the casting width can be from 1000 to 2500 mm or up to 3000 mm wide, and the facility consists of: casting thickness of 100 to 150 mm, a straight mould of conventional design, width and taper with adjustable during casting, a casting radius of approximately 5.5 m, casting speeds of 1.0 to 3.5 mlmin depending upon grade and slab dimensions, a roller apron using split by closely spaced rollers arranged in segments, computer controlled air-mist cooling for different grades and casting speeds to improve internal soundness and low maintenance deign of components, etc. Basically the TSP caster combines the benefits of thin slab as well as the traditional thick slab casters by using conventional mould design. The main features of this research paper are a flowfield control of molten steel of intermediate thickness slab in the TSP. The simulation method is suggested using a computer modelling in order to describe the turbulent properties of molten steel. The formulation of molten steel behaviour is based on 3-dimensional turbulent Navier-Stokes equation in which allowance is made for spatially distributed turbulent properties. Namely, the basic governing equations are those of 3-dimensional, steady state, incompressible transport of mass, momentum and enthalpy. The simulation model using two-equation k-e turbulent model, in which it is assumed that the influence of turbulence on the transport of momentum can be modelled by the addition of a 'turbulent viscosity' to the existing laminar viscosity, has been applied. In addition to cold modelling, a considerable help in the research for the flow visualization comes from the use of 0.5 scale models

B34

of tundish and mould in which water is used as operation fluid. The measurement data were obtained by using 3-dimensional Laser Dopper Velocimeter.

PHYSICAL CHEMISTRY OF HIGH TEMPERATURE COMPOSITES AND INTERMETALLICS II: Joining

Sponsored by: EPD Process Fundamentals Committee, MSD Thermodynamics & Phase EquUlbrla Committee, Japan Institute of Metals Program Organizers: Prof. R.Y. Lin, University of Cincinnati; Prof. Y. Austin Chang, University of Wisconsin-Madison; Prof. R. Reddy, University of NavadaReno and Dr. C.T. Liu, Oak Ridge NL Monday, PM February 5,1996

Room:B2 Location: Anaheim Convention Center

Session Chairperson: A. Mikula, Institut fur Anorganische Chemie, Universitat Wien, Wahringerstrasse 42, A-I09O Vienna, Austria; S.K. WU,lnstitute of Material Science & Engineering, National Taiwan University, Taipei, Taiwan 2:00 pm invited FUNDAMENTALS OF'BONDING BY ISOTHERMAL SOLIDIFICATION FOR HIGH TEMPERATURE SEMICONDUCTOR APPLICATIONS: RainerSchmidFetzer, Techn. Universitat Clausthal, AG Elektronische Materialien, Robert-Koch-Str. 42,0-38678 Clausthal-Zellerfield, Germany Isothermal solidification can be used to form strong bonds at low temperatures which remain solid at much higher temperatures. The exploitation of this principally known technique for new high temperature processing steps and applications in electronic assemblies is the aim of this study. Three basically different variants of this bonding process are given. A cOllljlrehensive metallurgical screening is performed and the feasibility of binary material systems is given on a quantitative scale. Very attractive systems are spotted that have never been studied before. Detailed examples for the realization of bonding with complex ternary and multilayer systems are given. A highly innovative example is the chip stack or wafer bonding of unmetallized Si or SiC at 250·C. The proposed bond remains stable above the melting point of Si and enables entirely new semiconductors manufacturing steps. An improved die bonding process was proposed and verified experimentally. 2:25pm INFRARED JOINING OF SILICON CARBIDE FORMATION AT THE JOINT DURING INFRARED BONDING OF SILICON CARBIDE: Randall A. Blue, Ray Y. Lin. Department of Materials Science and Engineering, M.L. #12, University of Cincinnati, Cincinnati, OH 45221-0012 Joining of Silicon Carbide with infrared usingamixture ofSi - 30wt.%C as the brazing material has been investigated.J oining was performed at a temperature of 1500=BOC in a flowing argon atmosphere for proceSSing times ranging from 5 to 60 seconds. Cross section examinations of the joined parts have shown that all joints are nearly void free with excellent wetting between the base SiC and the brazing material. Silicon carbide has been identified as the reaction product in the joining woe with a transmission electron microscope (TEM) and selectc>.,d area diffraction (SAD) analysis. At 1500=BOC with a holding time of 60 seconds, a nearly homogenous SiC jOint forms, producing strengths comparable to other brazing t,~hniques. A kinetic model of forming SiC from liquid silicon and carbon will be presented. 2:50pm ON THE METALLURGY OF ACTIVE BRAZING OF SILICON NITRIDE: J.K. Kivilahti, M. Paulasto, F.U. van Loo, Helsinki University of Technology, Dept. of Materials Science and Engineeling, FI-02150 Espoo, Finland Advanced ceramics like silicon nitride are increasingly used as structural components in demanding high temperature applications as well as in electronics industry. Complex, multicomponent structures for engineering applications generally. The interfacial microstructures formed when Si3N4 is brazed with Ag-Cu-Ti braze alloys have been discussed in several recent studies. In order to resolve which chemical reactions are possible under various joining conditions and thereby to be able to control the brazing of Si3N4 the reactions in brazing has to be studied using both microstructural and thermodynamics. In this work tile earlier determined ternary Ti-Ag-Cu and TI-Is-N systems have been applied to the brazing of silicon nitride with AgCuTi alloys. Using direct brazing experiments at various temperatures and thermodynamic analyses performed by combining the most recent available thermodynamic data and experimental information the chemical reactions controlling the brazing process are determined and a deSCription of the am is presented. It is shown that the miscibility gap found in the ternary Ti-Ag-Cu system exists during brazing and divides the liquid braze into Ag-rich (LJ) and Ti- and Cu-rich (U) liquid. The liquid U segregates at the ceramic interface and reacts with the silicoa quateffiil1)' Ti-Si-Cu-N compound. 3:15 pm Invited BONDING OF TmN COPPER mscs TO SAPPHIRE BY A CUPROUS OXIDE INTERLAYER: G.P. Martins, Wen-Ming Chen, Carlos A. Silva, Colorado School of Mines, Golden, CO 80401 Joining of metaVmetal, ceramic/ceramic systems can be effected by a transient liquid phase (TLP) scheme or an adaptation of it. The bonding of copper to AI20 3( alumina or sapphire) by this technique has been of interest during the last twenty years in the electroniCS industry. In general, the process can be conveniently analyzed by identifying three distinctstages which lead to the formation of ajoint-member whereby an interlayer is completely consumed. The research conducted was devoted to a primary effort on the development of mathematical model!. describing the behavior of three stages associaled with the process. A secondary effort was focused on bonding of copper disk (screens or

1996 TMS ANNUAL MEETING AND EXHIBITION

wasbers) to sappbire lenses. Several bonded parts were successfully manufactured in the laboratory scale equipment wbicb incorporated an isothermal cell as a critical system component. However, only a cursory analysis was conducted on one assembly in regard to the interfacial bond region between the copper and sapphire. The predictions obtained are the first known quantification of the time scales related to the mecbanisms associated with 1LP bonding of copper to sappbire by use of a cuprous-oxide interlayer. 3:40 pm BREAK

filmfluoromica(AI20fKMg3[AlSi31011f~Al203)laminatesundertheconditionsT=25-

4:00 pm Invited INFRARED JOINING OF TIAI TO TIAI WITH ALUMINUM FOILS: Sunil K. Annaji, Ray Y. Un, Department of Material Science & Engineering, University of Cincinnati, M. L. 12, Cincinnati, OH 45221-0012; S. K. Wu, Institute of Material Science & Engineering, National Taiwan University, Taipei, Taiwan It has been sbown that wben liquid Al is in contact with Ti there is the formation of an intermetallic TIAI3 phase. This is the basis for the brazing ofTIAl using aluminum foils. In this study, infrared joining ofTIAl was investigated at temperatures between 1050·C and 13500C in a flowing argon environment for times varying between IS seconds to 60 seconds. The growth ofTIAl3 in the joint appears to be diffusion controlled with aluminun being the dominant moving species. Cross sectional studies on the joined specimen bave sbown different degrees of reaction depending on the temperature and time of bolding. Microhardness studies across the jOint sbow a gradual decrease in the hardness as the indentations are made towards to the center of the joint. However, it is observed that at higher temperatures of processing and longer periods of time, the hardness values towards the center increase considerably as compared to the increase at the peripbery of the joint. TEM studies have been done to identify the TIAI3 pbase. 4:25pm THERMODYNAMICS AND PHASE DIAGRAM OF NEW SOLDER MATERIALS: S. Karlbuber, A. Mikula, Institut fur Anorganische Chemie, Universitat Wien, Wabringerstrasse 42, A-I09O Vienna, Austria The unique physical and mechanical properties have made the Pb-Sn solders attractive for low temperaturejoining of electronic parts. But since lead in the solder became a great environmental problem and a bealth concern, it is safer and more economical instead of cleaning up the waste dumps to replace lead in the solder materials. This idea has caugbt on and a lot of researcb and development efforts are going on to replace lead in the soft solder materials. Some of these materials are ternary tin alloys and not mucb is known about the thermodynamics and the phase diagram of these alloys. We investigated (Ag, Au, Cu)-Sn-Zn systems and determined the thermodynamic properties over the wbole composition range and did some work on the phase diagrams in the Sn-ricb comer. 4:50pm PERIODICLAYERFORMATIONDURINGSOLIDSTATEREACTION:MR. Rijnders, A.A. Kodentsov, F.J.J. van Loo Lab, Solid State Chem. & Mat. Sci., Eindboven Univ., of Tecb., P.O. Box 513, 5600 MB Eindhoven, the Netherlands A periodic layered morpbology of the reaction zone has been observed in several annealed solid/solid diffusion couples. This morpbology is cbaracterized by a regular array of hands of particles embedded in an intermetallic matrix phase. An understanding of the mecbanism underlying periodiC layer formation is important in predicting the long-term high temperature stability of joints. From experimental observations we will show that various factors like widely different mObility's of components, bigh vacancy fluxes, particle coarsening, dislocation movement, sbifting pbase interfaces and development of stresses sbould be included in a comprebensive theory concerning this type of reaction.

PHYSICAL PROPERTIES OF COMPOSITES II Sponsored by: JL SMD-MSD Composite Materials Committee Program Organizer P.K. Uaw, Materials Science and Engineering Department, The University of Tennessee, Knoxville, TN, 37996-2200; R. Pitcbumani, Mechanical Engineering Department, University of Connecticut, Storrs, CT 06269-3139; S.O. Fishman, Office of Naval Researcb, SOO N. Quincy Street, Arlington, VA 22217 Monday, PM February 5, 1996

2:30pm ELEVATED TEMPERATURE-VARIABLE-RATE MECHANICAL RESPONSE OF ALUMINAIFLUOROMICA LAMINATES: Todd T. King, Reid F. Cooper, Department of Materials Science and Engineering, University of Wisconsin, Madison, WI 53706 Elevated-temperature flexural delamination experiments are used to evaluate the mecbanical performance of tbermocbemically and structurally stable a-alumina/thin-

Room: Marquis 1&2 Location: Anaheim Marriott Hotel

Session Chairperson: R. Pitcbumani, Mechanical Engineering Department, University of Connecticut, Storrs, CT06269-3139; R.F. Cooper, Depts. MS&E and NEEP, University of Wisconsin, Madison, WI 53706 2:00pm HIGH-TEMPERATURE FLOW OF SIC CONTINUOUS FlBER-GLASS-CERAMIC MATRIX COMPOSITES: THE EFFECT OF INTERFACEIINTERPHASE DUCTILITY: B.G. Nair, J.N. Almquist, R.F. Cooper, M.E. Plesba, Depts. MS&E and NEEP, University of Wisconsin, Madison, WI 53706 The bigh-Trheology of continuous SiC-calcium alurninosilicate glass-cerarnic matrix composites is evaluated in uniaxial compression creep experiments (-aI = 20-to-4O MPa; T = 1300-1320·C). The steady-state strain rate is higbly sensitive to the orientation of the reinforcement relative to the maximum compressional stress, with highest bulk specimen strain rates noted for conditions in whicb the sliding between the fiber and the matrix is optimized as a kinetic flow response, i.e., a fiher orientation of approximately 40-50· from 0"1' (These conditions produce far higber strain rates than when the fibers are perpendicular to 0"1)' One further discovers that the temperature sensitivity of flow increases as the amount of interface flow/sliding increases. The experimental results suggest that the higb-temperature,low-stress interface response in this composite system is related to the ductile flow of the "planar" Si02reaction-layer interpbase that exists (in addition to the well-recoguized planar carbon interpbase) in these materials.

1996 TMS ANNUAL MEETING AND EXHIBITION

800°C: E=IxIO's - 5xlO·3s· l ; air environment. At low T, the laminates display a purely brittle delamination response, one based on the intrinsic cleavage behavior of the mica. For T>4OO°C, a mixed, brittle-ductile interfacial response is noted, wbicb is a clear function of strain rate: consistent with the pbysics of the brittle-ductile transition, more rapid rates allow for a brittle interfacial response (usually desired in a ceramic composite). The onset of thermally activated ductile deformation increases the energy for delamination. Microstructural analysis indicates both (partially anelastic) kinking and slip mechanisms are active in the mica interphase at high T. The results suggest that the mica can be engineered cbemically so as to produce mechanically functional interfaces for alumina-fiber composites over a broad range of (T,E) conditions. 3:00pm INFLUENCE OF SUBSTRATE ON THE ELASTIC REACTION OF FILMS FOR THE MICROINDENTATION TESTS: Youngman Kim, R&D Center, Korea Oas Corp. Ansan, Korea; Min-Tae Kim, KIA Motors, Seoul, Korea The influence of substrate on the elastic reaction of thin film was investigated using a depth-sensing microbardness tester. Silicon oxide films were produced in a radio frequency (rf) plasma witbbexamethyldisilane and oxygen, on silicon wafers and glass ribbons simultaneously for the first set of samples and on silicon wafers but with various film thickness for the second set Young's moduli of the silcon oxide films were measured to be different for the different substrates for the same thickness of film. Young's moduli of the silicon oxide films decreased for the same substrate with increasing film thickness. The elastic deformation was considered for the normal point load to a flat quasi-infinite surface using an elementary theory of elasticity under the assumption of perfect adbesion between film and substrate. 3:30pm PROPERTY -BASED OPTIMAL TAILORING OF COMPOSITE MATERIALS: D. Sadagopan, R. Pitcbumani, Department of Mechanical Engineering, University of Connecticut, Storrs, CT 06269-3139 Composite materials offer designers the advantage of tailoring structures and materials to meet a variety of property and performance requirements in changing and demanding environments. However, the wide variety of materials combinations, reinforcement geometries and architectures to cboose from, poses a bewildering problem of selections. Thus, an appropriate, and furthermore optimal, tailoring of materials for composites applications is a challenging design problem and forms the focus of the paper. Specifically, the present work addresses the problem of selecting optimal combinations of matrix and reinforcement materials, reinforcement morpbology, architecture, and volume fraction so as to meet specified property and performance requirements. Matrix materials considered in the study span the material classes of polymers, metals and ceramics, while reinforcement geometries considered include continuous fibers, particulates and short fibers, and two-dimensional textile fabrics. Analytical models available in the literature are used to relate the composite properties to their microstructures. The property model base, in conjunction with the combinatorial optimization technique of simulated annealing, is used to solve the optimal materials tailoring problem. The overall approacb and key results of the study will be presented and discussed. 4:00pm EFFECTS OF NEUTRON IRRADIATION ON THERMAL EXPANSION OF SICISIC COMPOSITES: D.J. Senor, Pacific Northwest Laboratory, Ricbland, WA 99352; DJ. Trimble, Westinghouse Handord Company, 1.1. Woods, Lockheed-Martin, Schenectady, NY 12301 The effect of irradiation on thermal expansion of several SiC/SiC composite types was studied. Thermal expansion was measured using an alumina pusbrod dilatometer between room temperature and lOOO°C. The irradiated specimens were subjected to a neutron dose of either 33 or 43 dpa-SiC in the Experimental Breeder Reactor-II at approximately lOOO·C. The composites possessed a variety of constituent materials including two matrix types (CVI SiC and liquid infiltrated polymer-precursor SiC), three fiber types (Nicalon, Tyranno, and HPZ), and two interface types (pyrolytic carbon and boron nitride). In the unirradiated condition, thermal expansion was primarily dependent on tbe matrix material. The presence of the fibers resulted in a contribution to the unirradiated thermal expansion only for bighly anisotropic fiber architectures. Mter irradiation, the fibers were observed to bave completely debonded from the matrix, resulting in thermal expansion bebavior that was solely dependent on the matrix type. 4:30pm INFLUENCE OF FIBRE NETWORK INTERCONNECTION ON THE THERMAL EXPANSION ANISOTROPY IN METALLIC COMPOSITES: L. Tao, F. Dolannay, Universite catholique de Lonvain, oepartement des sciences des materiaux ei des procedes, PCIM, Place Sainte Barbe 2, B-1 348 Louvain-Ia-Neuve, Belgium Transverse isotropic fibre networks were prepared by pressing mats of steel fibers and sintering under protective atmospbere. The number and strength of the bonds at fibre interconnections varied as a function of sintering conditions, fibre volume fractions, and fibre cross-sections. These networks were used as preforms for creep strengthening ZnAl foundry alloys. The composites were processed by squeeze casting. They presented widely different thermal expansion coefficents in the directions parallel and perpendicular to the plane of the preform. The magnitude of the anisotropy was found to depend on the degree of fibre interconnection in the network. Reversibility of the strains during thermal cycling up to 180°C suggested no plastic yielding of the matrix. A model is developed for accounting for this bebavior.

B35

5:00pm FABRICATION OF SILVER-A1 z0 Jp ELECI'RlCAL CONTACf BY ELECTROLESS PLATING AND HOT PRESSING: Jiunn-Homg Un, Su-lien Lin, Dept of Matis Sci & Engrg, National Tsing Hua University, Hsinchu, Taiwan, China In this research, A1103p reinforced silver matrix composites with different amounts of reinforcements, 10, 20 and 40 vol. % and different partical size, 3 IllD and I lun, were fabricated by electroless silver plating followed by cold compacting and hot pressing at SOO-C or 6OO-C, 300 MPa in air. The distributions of the AI10lp were rather uniform. Densification was obtained by comparing the true densities of the composites measured by Archimedes'law with the theoretical densities calculated from ICP-AES analysis. Density measurement coniumed that the composites can be fully densified. Hardness and bending properties of the composites were also tested and found to be much better than pure silver. Electrical conductivity were measured by four-point probe method and revealed that the silver-AI10lp composites have a higher electrical conductivity than the traditional silver CdO electrical contacL

STRUCTURE AND MORPHOLOGY OF EPITAXIAL THIN FILMS SESSION II: Surfactant. And Surface Reaction.

Sponsored by: EMPMD ThIn Fllms & Interfaces Committee Program Organizer: Dr. David E. Jesson, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831-6030 Monday, PM February 5, 1996

Room: Orange County 4 Location: Anaheim Manion Hotel

Session Chairperson: C. J. Paimstrom, Department of Chemical Engineering and Materials Science, University of Minnesota, MS 2:00 pm Invited MANIPULATING EPITAXIAL GROWTH AT THE ATOMIC SCALE: FROM SURFACfANTS TO QUANTUM WIRES: Juan Jos6 de Miguel,Juan de laFiguera, and Rodolfo Miranda, Laboratorio de Fisica de Superficies, Departamento de Ffsica de la Materia Condensada,lnstituto de Ciencia de Materiales "N. Cabrera", Universidad Aut6noma de Madrid, Spain Heteroepitaxial growth is usually controlled by kinetic harriers which determine the island shape and density, the film morphology or the stacking of layers. In order to achieve thin films or supedattices with the desired properties it is often necessary to find ways to overcome these kinetic limitations. The purposeful use of surface active agents has been employed to modify crystal growth. We will review the use of surfactants to control the mode of growth of metallic films, the defect structure of vertical superlattices and the fabrication of lateral superlattices of magnetic wires by a multitechnique approach including TEAS, STM, LEED, surface-sensitive XRD and Molecular Dynamics. 2:30pm INFLUENCE OF SURFACfANTS ON SURFACE MORPHOLOGY OF GelSI HETERO EPITAXIAL FILMS STUDIED BY mGH RESOLUTION ELECTRON DIFFRACTION: M. Hom von HOI!gen, Universitat Hannover, Institut fur Festkorperphysik, Appelstr. 2, 0-30167 Hannover, Germany Surfactants have been very effectively used for growth mode engineering of hetero epitaxial systems. For the GeiSi system Sb,As, Bi orH has been tested to act as surfactant by forcing layer growth instead of island formation as observed without surfactant Two growth regimes have to be considered: Elastic relaxation Prior to the generation of dislocations or defects Ge growth pseudomorphic with the Si lattice constant and is strongly strained. The continuous Ge film can relieve strain by forming a rough and open surface (microrougbness on a om scale) which allows a partial elastic relaxation towards the bulk lattice constants. The variations of the parallel and vertical lattice constantcould be observed with LEBO. At very high temperatures on Si(OOI) the Ge forms regular arranged cones with a uniform size of 30 nm covering the whole surface. Reversible changes of the shape of the cones could be induced by variations of the surfactant coverage (i.e., by changes of the surface free energy). Plastic relaxation. After a particular "critical" thickness finally dislocations are generated. On Si(lll) those dislocations are arranged in a threefold regular array at the interface and matches exactly the lattice constants. For the first time the formation and evolution of such a network was observed in situ during deposition with high resolution LEBO. 2:50pm SILICIDE FORMATION BY Mo INCORPORATION IN SI(I00): Peter J. Bedrossian, L-3SO, Lawrence Livermore National Laboratory, Livermore CA 94551 The binary MoISi system has a particularly rich bulk phase diagram with a variety of stoichiometries. The structure of Mo-silicide thin films is complicated even further by the influence of interfacial energies, the importance of which is underscored by the ready observation of thin-film phases which do not even appear in the bulk binary phase diagram. Previous investigations of Mo adsorption on clean Si( I00) have reported the formation of an amorphous interface and transitions among various disilicide phases, but the transition temperatures were found to be highly process-dependent. Using STM and glancing-incidence x-ray diffraction, we have identified the temperature- and coveragedependence of the formation and interconversion of a variety of surface and thin-film phases raulting from adsorption of up to five monolayers of Mo on clean Si(IOO). Submonolayer Mo adsorption below 6SO·C leads to metastable, globally-homogeneous surface phases which exhibit short-range order in one-dimension, but were previously thought to be amorphous. Above 6SO"C, both the disilicide precipitates and precursors to grain boundaries can be identified even after initial formation of disilicide microcrystals. Microcrystals of two different phases of MoSi1, which form at different temperatures, can be identified with STM and x-ray diffraction. Recent modeling of the

B36

electronic structure of Mo-silicides offers insights which account qUalitatively for many of these experimental observations. Despite the tremendous complexity of this prototypical transition metal-silicide system, it is now possible to identify in detail the microscopic consequences of various processing parameters on evolving thin-film morphology. 3:10pm CoSlzFORMATION IN GelSI HETEROSTRUCfURES: G. R. Carlow, M. ZiokeAllmang, Department of Physics, University of Westem Ontario, London, Ontatrio, Canada The evolution of Co/Ge films on Si substrates has been investigated. Room temperature deposition of Ge films followed by Co films was done by Molecular Beam Epitaxy (MBE) and post-deposit annealing was done at 700·C. While planar CoSi1 films are formed at this temperature when no intermediate Ge layer is present, the effect of the Ge creates a clustered morphology and results in the formation of non-planar CoSi1regions. Using Rutherford baclcscattering spectroscopy and transmission electron miscroscopy in combination with electron energy loss spectroscopy, we detail the evolution of the cluster morphology and the spatial distributions of the Co and Ge. 3:30 pm BREAK COMPOSmONAL ORDERING Session Chairperson: L. J. Schowalter, Dept. Physics and Ctr.lntegrated Electronics and Electronic Manufacturing, Rensselaer Polytechnic Institute, Troy, NY 12180 3:50 pm invited SURFACE STRUCTURE AND ORDERING IN GalnP: G.B. Stringfellow, L. C. Su, Department of Materials Science and Engineering, University of Utah, Salt Lake City, UT84112 The surface structure of GainP layers grown by organometallic vapor phase epitaxy lattice-matched to GaAs,substrates has been studied using atomic force microscopy. For exactly (001) oriented su bstrates, islands 30-60 Ain height are observed that are formed of bilayer (5.9 A) steps. '.lbe results allow an understanding of the microstructure of the Cu-Pt type ordering observed in these layers, particularly the presence of order twin boundaries. For layers gIown on substrates intentionally misoriented by angles of from 3 to 9· to produce [110] steps on the surface, the steps are found to bunch, forming (001) facets, (lln) facets, and vicinal regions. Both the step height and the value ofn are found to depend on growth conditions. The lateral dimensions of the structures formed are found to scale as approximately (growth rate)·-, indicating a diffusion limited size. The supersteps are found to correlate with the anti-phase boundaries observed in the ordered structure. 4:20 pm invited SURFACE RECONSTRUCfION AS A DRIVING FORCE IN ATOMIC WNG RANGE ORDERING: Alex Zunger, NREL, Golden, CO 80401 Vapor-Phase (MOCVO, MBE, ALE) growth of many A•.•B.C semiconductor alloys results in spontaneous long-range order, most often in theformofmonolayer-altemation (ACMBC). superlattices along the (III) direction (the "CuPt" Structure). At the same time, it is known theoretically that: (i) the lowest energy state of bulk alloys is phase-separation into AC + BC, and that (ii) the lowest energy state of epitaxial alloys is the AB~ chalcopyrite structure. A combination of first-prinCiples total energy calculations and lattice-gas thermodynamic models clarifies that: (i) phase-separation is inhibited by the epitaxial coherence with the substrate, (ii) the chalcopyrite structure is suiface-unstable relative to the CuPt structure, and (iii) dimerization, buckling and titling of surface actions stabilizes ev~n at T-900K a special variant ("CuPt-B") of the CuPt structure. Detailed calculations on surface reconstruction in the Ga,.ln•.•P/GaAs (001) epitaxial films will be presented. They show that (I) the (2X2) cation terminated reconstruction leads to CuPlg ordering and In surface segregation, (2) the 2(4X2) leads to CuPtA ordering and Ga &egregation, (3) the 2(2X4) leads to CuPts ordering and In segregation, and (4) the C(4X4) leads toCuPtA ordering and In segregation. Spontaneous ordering changes profoundly the band structure of the alloy, leading to (a) handgap reduction, (b) splitting of the degeneracy of the valence band maximum, and (c) new polarization selection rules. In collaboration with S. B. Zhang and S. Froyen. 4:50pm X-RAY DIFFRACTION STlJDIES OF NONEQUILlBRIUM ORDER AND 00MAINS IN SlxGel-x FILMS ON MlSCUT SI(ool): J. D. Budai, J. Z. TIschler, O. E. Jesson, P. Zschack, Oak Ridge National Laboratory, Oak Ridge TN 37831-6030; J.-M. Baribeau, O. C. Houghton, National Research Council of Canada, Onawa, Canada Si.Ge •.•films grown on Si(OOI) continue to provide ideal model systems for investigating the rich interplay between surface morphology during growth and the resulting nonequilibrium compositional order. We have utilized synchrotron x-ray scattering measurements to characterize the atomic structure, the order parameter, the domain populations, and the domain sizes for Sio.sGeo.s films. The films were grown by MBE on substrates with a large range of miscut angles away from Si(OO I), including the Si( I05) surface. Measurements of integral and half-integral reflections reveal that compositional modulations along both (III) and (100) type directions are present in all samples. The substrate miscut controls the domain orientation, consistent with a growth mode govemed by ordering at surface ledges. The coherence lengths for the composition modulations in the two different directions are approximately equivalent over a widely varying range of values, suggesting that the structures cnexist rather than compete. Research sponsored by Division of Materials Sciences, U.S. DOE. under contract 0E-AC05-840R21400 with Lockheed Martin Energy Systems, Inc. 5:10 pm Invited CONTROL OF THE GROWTH AXIS AND CHEMICAL ORDERING IN INTERMETALLIC MAGNETIC ALLOY FILMS: R. F. C. Farrow, O. Weller, R. F. Marks, M. F. Toney, IBM Research Division, Almaden Research Center, 650 Harry Road, San Jose, CA 95120-6099 The magnetic and magneto-optical properties of intermetallic Co-Pt and Fe-Pt alloy

1996 TMS ANNUAL MEETING AND EXHIBITION

films are of considerable current interest. For example, the magnetic anisotropy of equiatomic FePt correlates with the degree of cbemical ordering and is predicted to reach the highest value of all transition metal alloys for the fully-ordered Ll. phase. Films containing this phase of FePt or CoPt are being studied as high-anisotropy longitudinal magnetic recording media. The challenge in growth of intermetallic alloy films is in controlling both the extent of chemical ordering and its orientation. We show how this control can be achieved in the case of FePt by using seeding techniques to select the orientation of the tetragonal c-axis. Films were grown by co-evaporation (MBE) of Fe and Pt onto a seed film of Pt grown on MgO substrates. Perpendicular or in-plane ordering was controlled by selecting the (00 I) or (110) plane, respectively, of substrate and seed film. The long-range order parameter (S) increased smoothly with substrate temperature and near-complete chemical ordering (S=O.95O.05) was achieved for growth at soooC.This work raises several interesting questions. Why does one unique orientation of the c-axis (c-axis in plane and along Pt[OO I]) dominate for growth ofFePt on Pt(llO)? Why do minority orientations occur more readily for compositions which are Pt-rich? These issues will be discussed and examples given of the growth of other intermetallic compounds including the newly-discovered hcp ordered phase of CIl] Pt.

SYNTHESIS AND PROCESSING OF NANOCRYSTALLINE POWDER I: Overview and Novel Synthesis Routes

Sponsored by: MDMD Powder Metallurgy Committee and FEMS (Federation of European Materials Societies) Program Organizers: Dr. David L. Bourell, The University of Texas at Austin, Materials Science & Engineering, MC C220l, Austin TX 78712; Dr. Uisa KuhnSpearing, Laboratory for the Study of Skeletal Disorders and Rehabilitation, Harvard Medical School, Children's Hospital, 300 Longwood Avenue, Boston MA 02115; Professor Dr. Herbert Gleiter, Karlsruhe Research Center, P.O. Box 3640, 0-76021 Karlsruhe, Federal Republic of Germany Monday,PM February 5,1996

Room: Grand G Location: Anaheim Marriott Hotel

Session Chairperson: Dr. David L. Bourell, The University of Texas at Austin, Materials Science & Engineering, MC C220I, Austin TX 78712 2:00pm NANOSTRUCTURED MATERIALS - AN OVERVIEW: JiJrg Weissmiiller, Nationaiinstitute of Standards and Technology, Bldg 223 Rm AI53, Gaithersburg MD 20899 Nanocrystalline materials are polycrystals with a grain size of the order of 10 nm. The earliest studies of these materials were motivated by the strive to explore the limits of metastability of solid matter and to create materials with novel properties. Dimensionality effecta and the large fraction of atoms in the core of topological defects (the grain boundaries) with non-lattice atomic short-range order contribute to the unique properties of the materials. The recent past has brought considerable progress in the characterization of the grain boundaries and of the microstructure of nanocrystalline solids. Physical properties are determined not only by the grain size (or rather the distribution of sizes), but also by additional parameters reflecting the thermal and mechanical history of the material. A large number of methods is now available for producing powdera with nanometer grain size, but the number of fundamental studies of their consolidation behavior is limited, and producing dense volume materials while avoiding excessive grain growth remains a challenge. The need for controlling the sintering process, together with the observation of grain growth at room temperature in some materials, indicates the fundamental importance ofexploring scbemes for stabilizing the grain size, andlorforproducing nanocrystalline dense volume materials directly ,circumventing the powder stage. 2:35pm NANOSTRUCTUREDPOWDERSYNTHESISVIASELF·ASSEMBLEDMEM· BRANES: G.M. Chow, M. Markowitz, A. Singh, Laboratory for Molecular Interfacial Interactions, Center for BiolMolecular Science and Engineering, Code 6930 Naval Research Lab, Washington DC 20375 Nanocrystalline powder can be synthesized using self-assembled molecular membrane structures as reaction vials. This chemical approach has the advantages of producing size-controlled, unaggJomerated nanoscale particles. An overview of this approach and our current work at NRL will be presented. In specific, gold and Col Co(OHh nanoparticles have been synthesized inside polymerized vesicles using surface bound palladium ions as catalysts for electroless metallization. HRTEM studies have demonstrated that particle nucleation and growth occurred within the vesicles and that particle growth could be initiated by either single or multiple nucleation sites. Results of formation of nanoscale metal particles using non-polymerized vesicles will also be discussed. 3:00pm CARBONATED APATITE NANOCRYSTALS FROM BONE: L. Kuhn-Spearing, MJ. Glimcher, Laboratory for the Study of Skeletal Disorders and Rehabilitation, Harvard Medical School, Children's Hospital, Boston MA 02115; H-M. Kim, College of Dentistry, Seoul National University, Seoul, Korea; C. Rey. Laboratoire de Physicochimie des Solides,I.N.P.T., Toulouse, France At the microscopic level, bones, antlers and teeth are all composed of numerous,

1996 TMS ANNUAL MEETING AND EXHIBITION

nanometer-sized mineral crystals embedded in an organic matrix. In bone, mineralization is initiated within the organic micro-compartments (40 nm gaps or holes baving a 67 nm periodicity) formed when the organic collagen molecules self-assemble into microfibrils. The calcium phosphate crystals (Ca-P) formed in bone are nearly monodisperse, nm-sized plates of carbonated apatite (a metastable phase of Ca-P), with a unique composition, crystal structure and surface character reflecting the complex biological environment they were nucleated in. A process has been recently developed to isolate and disperse the mineral crystals essentially free of organic matrix -- without any significant change in their composition. overall structure, or internal short range order. Tbe processing, characterization and potential medical applications of the bone mineral crystals will be presented. 3:25pm NANOCRYSTALLINE PROCESSING FOR STRUCTURAL AND BIO·CERAMICS: Atsushi Nakahira, DepartrnentofCbemical Engineering, Darren T. Castro, Department of Materials Science and Engineering, Jackie f. fing, Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge MA 02139 Nanocryslalline processing has demonstrated exciting potential in deriving oxide ceramics with low-temperature sinterahility and net-shape formability. However, sucb approach has been little explored with more complex ceramic compositions for hightemperature structural or bio-ceramic applications. This talk will focus OP two different techniques for generating high-quality nanocrystalline powdera. The first involves a modified forced-flow gas-condensation tubular reactor for flexibly producing large quantities of nanocrystalline non-oxide powders, notably SiJN4-based systems. The second exploits wet-cbemical precipitation to achieve proper phase control and dopant uniformity in hydroxyapatite-based nanocomposite. Tbe thermal stability and sinterability of the nanostructured ceramics are related to the processing. microstructure and interfacial chemistry of these systems. 3:50 pm BREAK 4:15pm MECHANOCHEMICAL SYNTHESIS OFNANO POWDER: J. Ding, W.F. Miao, T. Tsuzuki, R. Street, P.G. McCormick, Research Centre for Mineral and Materials Processing, University of Western Australia, Nedlands, WA 6097 Australia The synthesis of nanocrystalline powders by mechanochemical solid-state reduction reactions has been studied. X-ray and transmission electron microscopy measurements showed that Fe, Ni and Cu powders with particle sizes in the range 5-50 om were formed during the reduction process. Synthesized Fe powders exhibited coercivities of -500 Oe, indicative of separated nano-size particles. This study demonstrates that mechanocbemical processing has significant potential forthe synthesis of a wide range of nanocrystalline powders in an economic and efficient manner. 4:40pm A NEW APPROACH TO SYNTHESIS OF CERAMIC NANOPARTICLES: Ganesh Skandan, Structured Materials Industries. Inc., 120 Centennial Ave., Piscataway NJ 08855-3908; Yijia Cben, Nick Glumac, Bernard Kear,Rutgers -The State University of New Jersey, P.O. Box 909, Piscataway NJ 08854-0909 Bulk nanostructured materials, and nanostructured coatings are becoming increasingly importantin structural and functional applications. In order to achieve the requisite structure in the end product, a non-agglomerated powder with < I0 nm particle size needs to be used as starting material for most of these applications. Furthermore, it is important to prevent coarsening of the structure during processing. In recent research carried out at Rutgers University, we have developed a scaleable process for the synthesis of non-agglomerated nanoparticles of single phase (alumina, zirconia, silicon oxynitride), multiphasic(alurninalzirconiacomposites) and multicomponentoxides (gallates, ferrites), carbides and nitrides. The process, called Cbemical Vapor Condensation, is a modification of the Inert Gas Condensation process, and involves pyrolysis of chemical precursors in a hot zone with short residence times. Efficient pyrolYSis in a low pressure «30 mbar) atmospbere, either inert or reactive, has been found to be tbe key to high rate production of powders with desirable characteristics. Studies have been carried out on consolidation of these powders into dense materials while still preserving the nanostructure. For example, near theoretical densities with an average grain size of - 50 nm, have been obtained by pressure assisted sintering of nanoparticles of zirconia in air. Similar results have been obtained by air-sintering powder mixtures of alumina and zirconia. Correlation between powder characteristics, processing and structure will also be discussed.

5:05pm THE TRANSITIONS "SOL·GEL·STONE·LlKE CONDITION" OF FOUNDRY WASTES: VA. Mymrin, Scientific Centre for Engineering Geology and the Environment, Russian Academy of Sciences, P.O. Box 145, Ulanski per. 13, Moscow 101000 Russia It was established by a wide complex of modem investigation methods that the crystal bodies in burnt foundry sand (foundry slag) may be completely dissolved in a porous alkaline solution of the same industrial waste, forming a sol. By increasing the concentration, the sol transforms into a gel. This gel transforms through several stages to a ·stone-like" state. After 28 days, the strength of the samples was 1.1-2.2 MPa, but after 90 days the strength rose to 5-6.2 MPa. This is the top of the upper level of strength requircments for Russian standards for reinforced soils at 90 days. The strength almost doubled at 360 days. These materials possess very high freeze-thaw resistance. These new materials can be used as environmentally safe, cost effective, low-grade concrete in potential applications including road bases, industrial/municipal waste dumps, etc.

B37

3:00 pm BREAK

TANTALUM: Fabrication And Processing: Session II

Sponsored by: SMD Refractory Metals and Materials Committee and JL MDMDIEPD Synthesis, Control and Analysis In Matertals Processing Program Organizers: Andrew Crowson, U.S. Army Research Office, Research Triangle Park, NC; Edward S. Chen, U.S. Army Research Office, Research Triangle Park, NC; Prabhat Kumar, Cabot Corp, Boyertown, PA; Willam Ebihara, Picatinny Arsenal, Picatinny, NJ; Enrique J. Lavernia, UC Irvine, Irvine, CA

Monday, PM February 5, 1996

Room:A4-5 Location: Anaheim Convention Center

Session Chairpersons: Andrew Crowson, U.S. Army Research Office, Research Triangle Park, NC; Enrique Lavernia, University of California at Irvine, Irvine, CA

1:30pm SPRAY ATOMIZATION AND DEPOSITION OF TANTALUM ALLOYS: Weidong Cai, Huimin Liu, Enrique J. Lavernia, Department of Chemical Engineering and Materials Science, University of California, Irvine, CA 92717. Roger H. Rangel. Department of Mechanical & Aerospace Engineering, University of California, Irvine, CA92717 In previous numerical analysis, the droplet-gas interactions !hat are present during spray atomization and deposition of a Ta-2.5W alloy using N2 gas were investigated. The numerical results demonstrate that at any axial distance, the droplet velocity, temperature, cooling rate and solidification rate all exhibit a maximum at the spray axis, and decrease to a minimum at the periphery of the spray cone, except for the locations where solidification occurs. The droplets in the periphery region solidify wiibin a shorter flight distance relative to those at the spray axis due to longer night time in the periphery. Hence, the microstructure of spray deposited materials is predicted to be fme in the edges of the deposits as a result of high cooling rates associated with small droplets. Accordingly, in the present paper, a comparison with the experimental results is performed. To accomplish ibis, the spray atomization and deposition of tantalum alloys is conduced on a specially designed elevated temperature spray atomization and deposition facility. The alloys are melted in an induction skull melting crucible and then poured into a tundish in which the melt is delivered through a nozzle to an atomizer where the melt is atomized. Following the atomization, the droplets impinge onto a watercooled substrate and eventually form a deposit. The microstructure of as-sprayed deposits is then analyzed by using optical and electron scanning microscopy. 2:00pm TANTALUM PRODUCTION BY VACUUM MELTING PROCESSES: Trung Q. Luong, H.C. Starck, 45 Industrial Place, Newton, MA 02161-1951 Tantalum and tantalum alloy wrought products are extenSively used in the chemical process industry and in high temperature vacuum fumaces. Tantalum parts typically possess high ductility, excellent corrosion resistance, and very high melting temperatures. These properties can be enhanced or reduced by varying the amount of contained interstitial and other alloying elements. Careful control of operating parameters is critical in optimizing or negating the effects of these elements. This paper will discuss the consolidation of recycled tantalum by electron beam melting. It will also discuss the vacuum arc remelting of tantalum to refine grain size and to make alloy additions. The effects of material preparation, melting parameters, and alloy additions on finished products will be described. 2:20pm THERMOMECHANICAL PROCESSING OF Ta-IOW: R. W. Buckman Jr, Refractory Metals Technology, Pittsburgh, PA 15236; C. Bagnall, Concurrent Technologies Corporation, Johnstown, PA 15904 The effect of the amount of prior cold work on properties of the solid solution strengthened alloy Ta-IOW has been investigated. Recrystallized Ta-IOW was reduced by both compressive and tensile operations at room temperature to four different levels of cold work ranging from 20% to 90% reduction in area. Tensile an compressive properties were measured and the recovery and recrystallization behavior was monitored. From the data obtained, a thermomechanical processing schedule is identified for producing Ta-IOW tubing with optimum high temperature properties in the recrystallized condition. This work was conducted by the National Center for Excellence in Metalworking Technology, operated by Concurrent Technologies Corporation on under contract to the U.S. Navy as part of the U.S. Navy Manufacturing Science and Technology Program. 2:40pm POWDERED TANTALUM METAL EXPLOSIVELY FORMED PROJECTILE LINERS: Ernest C. Faccini, Textron Defense Systems (TOS), 201 Lowell Street, Wilimington, MA 01887 TOS bas successfully demonstrated aerostable EFPs produced from powdered Ta. A discussion of the result which TOS bas obtained using a powdered metallurgy technique for the formation of EFP liner starting stock is presented. The data discussed shall include: grain size of the material, texture, pseudo-static engineering properties, dynamic stress-strain curves, flash x-rays of the EFPs produced from the starting stock, photographs of the recovered EFPs, cost ramifications of the process, comparison to wrought material which has been orbitally forged and comparison to wrought material which has been formed into plate. A brief discussion of the future of the process is also presented.

B38

3:15pm PIM PROCESSING, CHARACfERIZATION, AND APPLICATION OF TalOW: H. Clemens, R. Grill, P. Rodbammer, A. Voiticek, Plansee AG, A-6600 Reuttel Austria The tantalum alloy Ta-IOW (comp. in wt.%) was investigated with regard to its suitability as substrate material for coated fasteners to be used in the 'hot' structure of reusable space vehicles. The paper describes the PIM processing route of Ta-IOW rods which are the base for falStener production. Static recrystallization behavior as well as the grain growth dependence at elevated temperatures were investigated. The mechanical characterization was conducted for different material conditions, e.g., as-worked, recrystallized, and recrystallized + aged. The brittle-to-ductile transition temperature was determined by Charpy impact testing. Tensile tests and double-shear tests were performed at room temperature and elevated temperatures up to IQOOoC and 1300°C, respectively. The crack propagation behavior was measured at room temperature. The obtained mechanical properties will be compared with those published for meltmetallurgically manufactured Ta-IOW. In addition, the feasibility of modified silicide coatings for oxidation protection of Ta-l0W has been investigated. The results of oxidation tests and the <:onsequence of local defects will be outlined. 3:45pm TANTALUM POWDER CONSOLIDATION,MODELING AND PROPERTIES: Sherri R. Bingen, Victor D. Vargas, Haskell Shein berg, Materials Science and Technology; Metallurgy, P.O. Box 1663, MS G770, Los Alamos National Laboratory, Los Alamos, NM 87545 A systematic approach has been taken to investigate the consolidation of tantalum powders. The effects cf hot isostatic pressing (HIP) temperature and time; sinter temperature, time and ramp rate; oxygen content; and alloying an consolidation kinetics, microstructure, crystallographic texture, and mechanical properties have been evaluated. A micromechanics model, the Ashby HIP model bas been employed to predict the mechanisms active during the consolidation (HIP and sinter) processes. This model also predicts the density of the end product and whether grain growth should expected under the applied processing conditions. It, however does not predict the overall shrinkage and part shape during consolidation. A finite element model, such as MARC, is an appropriate solution to these predictions, and is being applied (in concert with Ashby Model) to tantalum powdler processing in order to optimize near-net shape consolidation conditions. Experimental and modeling results to date will be presented. 4:05pm MECHANICAL ALLOYING IN THE IMMISCIBLE Cu-Ta SYSTEM: Toshihara Fukunaga, Department of Crystalline Materials Science, School of Engineering , Nagoya University, Furo-cho, CMkusa-ku, Nagoya 464-01, Japan The mechanical alloying process bas been studied on the Cu-Ta system which is characterized by a positive heat of mixing. The neutron diffraction and EXAFS measurements have been employed as a main tool to analyze the structural changes taking place during milling and DSC, XPS and the low temperature specific heat measurements as supplementary tools. The formation of an amorphous phase was confirmed not only from the structural studies but also from a change in the thermodynamical and electronic properties. The local atomic structure in the amorphous Cu-Ta is compared with that in the amorphous Ni-Ta with a large negative heat of mixing. The amorphization process can be understood as the preferential penetration of smaller atoms Cu into the bee Ta crystalllites. Studies of the ambient temperature effect in the Cu-Ta system thermodynamically suggest that an increase in the interfacial energy is large enough to allow the formation of an amorphous phase. 4:25pm SHOCK COMPRESSION SYNTHESIS OF BI-TYPE TANTALUM NITRIDE: K.S. Vandersall, N.N. Tbadbani, MSE, Georgia Institute of Technology, Atlanta, GA 30332-0245 Shock compression wall used to synthesize tantalum nitride with the B I-type (cubic) crystal structure which has been theoretically predicted to have high hardness and high super-conducting critical temperature. Other processing methods generally yield a nonstoichiometric low recovery product !hat must be further processed before use in the bulk polycrystalline state. In the present work, hexagonal phase (Co-Sn Structure) tantalum nitride powder was packed with densities of -35% and -60% T.M.D into steel capsules, and shock loaded at 1.0 kmls impact velocity, corresponding to a 40-60 GPa calculated peak pressure. X-ray diffraction and optical and scanning electron microscopy were used to characterize the recovered compacts and starting powder. The conversion of the hexagonal to the BI phaso, was observed to depend on the initial porosity and shock condition (pressure and temperature) wiibin the sample. The highest yield was obtained from the capsule with the lower packing density in the regions toward the non-impact face which correspond to the highest pressure and temperature. The lattice parameter of the shock synthesized B1 phase was calculated to be 0.433 nm which coincides with a stoichiometry of approximately TaNo.96•

1996 TMiS ANNUAL MEETING AND EXHIBITION

TRANSIENT THERMAL PROCESSING OF MATERIALS: SESSION I: Silicon Related Processes

Sponsored by: EMPMD Thin FUms & Interfaces Committee Program Organizers: N. M. Ravindra, New Jersey Institute of Technology, Newark:, NJ; R K. Singh, University of Florida, Gainesville, FL Monday, PM February 5,1996

Room: Grand J Location: Anaheim Marriott Hotel

Session Chairmen: Arun Nanda, Sematech, Austin, TX; Ahmad Kermani, CVC Products, Inc., Fremont, CA 2:00 pm Invited MODELING,MEASUREMENT AND CONTROL OF RAPID THERMAL PROCESSING: Krishna C. Saraswat, B. T. Khuri-Yakub, Department of Electrical Engineering, Stanford University, Stanford, CA 94305 A highly flexible Rapid Thermal Multiprocessing (RTM) reactor is described. This flexibility is the result of several new innovations; a lamp system, an acoustic thermometer and a real-time control system. The new lamp has been optimally designed through the use of a "virtual reactor" methodology to obtain the best possible wafer temperature uniformity. It consists of multiple concentric rings composed of light bulbs with horizontal filaments. Each ring is independently and dynamically controlled providing better control over the spatial and temporal optical flux profile resulting in excellent temperature uniformity over a wide range of process conditions. An acoustic thermometer non-invasively allows complete wafer temperature tomography under all process conditions - a critically important measurement never obtained before. For real-time equipment and process control, a model based multi variable control system has been developed. 2:30 pm Invited RAPID THERMAL CHEMICAL VAPOR DEPOSITION OF SILICON-BASED HETEROSTRUCTURES: James C. Sturm, Department of Electrical Engineering, Princeton University, Princeton, NJ 08544 This talk willdescribe the use ofRTCVDforthe growthofsilicon-basedheterostructures such as Si/Sil-x Gex and SiSil-x-y Gex Cy. Such heterostructures can be used to improve the performance of conventional silicon devices such as MOSFETs and bipolar junction transistors (through heterojunction MOSFETs and heterojuntion bipolar transistors), and also to implement new opto-electronic functions onto silicon such as infrared detectors and emitters. Because of the strain involved in such pseudomorphic structures, the tendency towards three dimensional growth, and the desire for abrupt interfaces on a scale of nm, these structures are typically grown at temperatures below 650-c. These low temperatures bring both advantages and disadvantages for RTCVD compared to traditional silicon epitaxy at higher temperature. For example, due to lower growth rates, thermal transients need not be as fast as at higher temperatures, and gas switching of processes can be used instead of temperature switching. However, increased sticking coefficients ofimpurities cause an increased sensitivity to gas purity. We have found that high quality RTCVD growth of these heterostructues is only possible due to the fortitous effects of hydrogen passivation present during the growth, which can reduce oxygen sticking coefficients several orders of magnitude. There is also a need for more precise temperature measurement because of the stronger dependence of growth rate on temperature, and a non-invasive non-pyrometric method used daily in our lab for accurately measuring the substrate temperature in the reactor in this range will be shown. Finally, the relative technical merits based on fundamental considerations ofRTCVD vs MBE and UHVCVD for the growth of these structures will be discussed. 3:00 pm Invited USING PHOTONIC EFFECTS IN OPTICAL PROCESSING FOR SILICON DEVICE FABRICATION: Bhushan L. Sopori, National Renewable Energy Laboratory, 1617 Cole Boulevard, Golden, CO 80401 Using optical energy for semiconductor processing has many advantages over conventional furnace processing. Some of these advantages, generally made use of in RTA processing, include ability to rapidly increase the wafer temperature, system cleanliness, possibility of excitation from outside the furnace, and low thermal budget. In an RTA process, the light is used to generate heat in the semiconductor wafer, and reactions are thermally activated. Recently, we found that RTA processing can also produce photonic effects that activate several reactions at much lower temperatures as compared to the thermal activation alone. We call this mode of RTA as Optical Processing (OP). We have developed several processes for silicon solar-cell fabrication that can also be used for other semiconductor devices. In general, the photonic phenomenon invoked in OP can produce reactions at the interface(s) of a multilayer device or in the bulk of the semiconductor. Suitable choices of process parameters such as the incident spectrum, light intensity, and the process time can be applied to favor the desired process. The interface- related processes are used to: (I) form very lowreSistivity Si-AI ohmic contacts with controlled optical properties at the interface, graded alloy compositions, and textured interfaces for light trapping in solar cells; (2) grow high-quality, thin Si02 at low temperatures (<.500 degrees C), with average growth rates reaching 5 nm/minute. The bulk effects are used for: (I) low-temperature, very rapid gettering of metallic impurities like Cu, Ni, and Fe; (2) impurity/defect passivation of solar cells by hydrogen. OP is a very low thermal budget, low-temperature process that typically uses only 4-6 W/cm2 of broadband incoherent light Lower temperatures are needed because photonic effects diminish at higher temperatures. Typical process times for the above applications are 6O-9Os. These features make OP well suited for a variety of commercial applications that require high throughput and continuous or batch processing. This paper will review basic concepts ofOP and the operational differences

1996 TMS ANNUAL MEETING AND EXHIBITION

between conventional RTA and OP. We will also describe applications in solar-cell and detector fabrication. 3:30 pm BREAK 3:45pm IN SITU ANALYSIS DURING RAPID THERMAL ANNEALING OF THIN FILM REACTIONS IN SUBMICRON CMOS STRUCTURES: L.A. Clevenger, C. Cabral, Jr., RA. Roy, C. Lavoie, K.L. Saenger, K. Rodbell & J. Jordan-Sweet, IBM, T J. Watson Research Center, Yorktown Heights, NY 10598; G. Morales, K.L. Ludwig, Jr., Boston University, Boston, Massachusetts 02215; G.B. Stephenson, Argonne National Laboratory, Argonne, IL 60439 The formation of thin film conductors is a critical processing step in CMOS (complementary-metal-oxide-semiconductor) logic and memory device manufacturing. In this work, we have studied the formation of low resistance titanium silicide and Al(Cu,Si)/Ti interconnections used in submicron device structures (0.1 to 1.0 micron in width, 50 to 500 nm in thickness) in situ, during rapid thermal annealing using synchrotron x-ray diffraction analysis on a millisecond time scale. In addition to the collection of x-ray diffraction spectrum, resistance and optical scattering (lambda = 632.8nm) measurements were simultaneously done to gather data on thin film resistance and surface roughness. Our results demonstrate that the formation of the low resistance C54-TiSi2 phase is limited by the submicron Iinewidth of the silicide structure. For patterned Al(Cu,Si)ffi conductors, the formation of the conductive AlTi) intermetallic phase is restricted by the amount of Si in the Al(Cu,Si) thin film 4:05pm SELECTIVE EPI OF SILICON USING RTCVD TECHNIQUES: M.C. Ozturk et.al., Department of Electrical & Computer Engineering, North Carolina State University, Raleigh, NC 27695-7911 Abstract not available. 4:25pm EFFECT OF TITANIUM FILM THICKNESS ON SILICIDE FORMATION & KINETIC: S. Sharan, Micron Technology Inc., MS 306, 5683 S Alyssum PI., Boise, ID 83705; RK. Singh, R Nagabhusnam, Department of Materials Science & Engineering, University of Rorida, Gainesville, FL 32611 Abstract not available. 4:45pm RAPID ELECTRICAL CHARACTERIZATION OF OXIDES AND NITRIDED OXIDES GROWN IN A SINGLE-WAFER REACTOR: 1. Sagnes, D. Laviale, France Telecom, CNET-CNS, BP 98, F-38243 Meylan-cedex, France; F. Martin, CEALETIlDMEL, BP 85X,F-38041 Grenoble-cedex, France; F. Glowacki, L. Deutschman, ASTElektronik, DaimlerstraBe 10, D-89160 Dornstadt, Germany Rapid Thermal Oxidation (RTO) and Nitridation (RTN) are performed in a module of a single-wafer cluster tool. The heating in the module makes use of a dedicated lamp array system for uniformity control with processes which can be achieved at atmospheric and low (H)i'-6 torr) working pressures. Electrical results obtained via a SCA (Surface Charge Analyzer) equipment on thin oxides and NO nitrided oxides «=6nm total thickness) will be presented. It will be shown that oxides obtained by RTO present a charge density equivalent to that obtained in batch oxidation furnaces as soon as some process conditions are fulfilled. The NO nitridation of the Si02lSi interface presents different charge states according to the process conditions (temperature, time and working pressure). We will show that after an anneal at low temperature, the initial level of charges is recovered. This work has been carried out within the GRESSI consortium between CEA-LETI and France Telecom-CNET. 5:05pm SELECTIVE RTCVD OF T1SI2: M.C. Ozturk et.al., Department of Electrical & Computer Engineering, North Carolina State University, Raleigh,NC 27695-7911 Abstract not available.

TREATMENT AND MINIMIZATION OF ALUMINUM AND LIGHT METALS INDUSTRY WASTES II

Sponsored by: Jt, LMDIEPD Recycling Committee, EPD Waste Treatment & Minimization Committee Program Organizer. Donald L. Stewart, Jr., Alcoa Technical Center, 100 Technical Drive, Alcoa Center, PA 15069; James A. Clum, Binghamton University, Watson School of Engineering, PO Box 6000, Binghamton, NY 13902-6000 Monday,PM February 5, 1996

Room: A7 Location: Anaheim Convention Center

Session Chairperson: Donald L. Stewart, Jr., Alcoa Technical Center, 100 Technical Drive, Alcoa Center, PA 15069; James A. Clum, Binghampton University, Watson School of Engineering, PO Box 6000, Binghamton, NY 13902-6000 2:00pm MINERALOGICAL AND PHYSICAL CONSIDERATIONS RELATED TO THE PHYSICAL SEPARATION AND RECOVERY OF ALUMINUM SMELTER WASTE PRODUCTS: Anhur Plumpton, lean-Francois Wilhelmy, Denis Blackburn, Centre de recherches min6rales (CRM), 2700 rue Einstein, Sainte-Foy, Qu6bec, Canada, G1P3W8 Several secondary and waste products of aluminum smelting were characterized mineralogically and physically in terms of their constituent mineral phases so as to

B39

evaluate the potential application of various mineral processing technologies for their partial decontamination or recovery of valuable fractions. The investigated products include red mud, bath-alumina mixtures, prebaked anode recycle residues, spent pot linings, drosses, sal leake, and dusts. Several of the materials possessed chemical and mineral compositions which varied with particle size and contained highly liberated mineral phases. The favorable mineralogy led to an evaluation of physical separation possibilities using the differing physical properties of mineral and metallic constituents. Wet processing methods such as gravity and magnetic separation and flotation are potentially attractive in several cases. Particular attention was given to the potential of recent fine particle processing techniques, as well as to dry separation technologies, such as electrostatic separation, pneumatic separation and gravity tabling, which best comply with the aluminum smelting environment. A literature review was made in regard to recovery or decontamination of aluminum smelter wastes using mineral processing technologies. 2:25pm PHYSICAL RECOVERY OF CARBON AND BATH PRODUCTS FROM GROUND SPENT ANODE W ASTES,Anhur J. Plumpton, Denis Cotnoir, Centre de recherchesmin~les (CRM), 2700 rue Einstein, Sainte-Foy, Qu~bec, Canada, G IP3W8 The constituent mineral phases contained within several aluminum smelter waste products are sufficiently liberated from each other to permit the application of mineral processing technologies for their separation and recovery. CRM studied both dry and wet physical processing methods for the recovery of carbon and bath products from wastes obtained during shot-peening of the spent anode assembly. The principal test material for the experiments was a composite sample of fine and coarse wastes, analyzing 30% C and 70% electrolysis bath material and having a 143~m average particle size and a 16 wt% - 45 ~ fme fraction. Summary results are presented and discussed for two dry processing approaches, electrostatic separation and pneumatic tabling (on des1imed feed), as well as for two wet methods, gravity tabling and froth flotation. Using a limited number of scavenging and cleaning steps, each wet method yielded two clean products, one being of 95 to 98% carbon purity, the other 97 to 99% bath grade. Similar high grade products were obtained using dry separation methods, but at the expense of having to deslime and/or classify the feed, and of accepting lower product yield and a middlings fraction equal to 40% or more of the feed weight. The approximate cost of a 1000 tpy recovery facility is discussed, as are the perspectives for improved physical separation of spent anode residues. 2:50pm A PROCESS FOR TREATMENT AND RECOVERY OF SPENT POTLINER (SPL): R. J. Adri~n, J. Besida, T. K. Pong, T. A. O'Donnell, D. G. Wood, G. H. Covey, Department of Chemical Engineering, The University of Melbourne, Parkville, Victoria 3052, Australia; J. J. Giansiracusa, D. E. Price, DEPCO-lRH Ply Ltd., Melbourne, Victoria 3000, Australia Spentpotliner(SPL) is classified as a hazardous waste as it contains cyanides and high levels offluorides. This paper describes a novel process for the conversion ofSPL at near ambient temperatures for the recovery of useful products and a benign residue. The initial treatment of crushed SPL removes over 90% of the cyanide and a large percentage of the fluoride. By appropriate selection of the chemical species and wash conditions, aluminium fluoride and other fluorides are effectively recovered in subsequent stages. After this treatment by a series of chemical washes, the residue produced is a carbonaceous material containing only carbon and brick fragments; i.e., refractory aluminium compounds. 3:15pm PRODUCTION OF VALUE-ADDED PRODUCTS THROUGH PHOSPHATE STABILIZATION OF ALUMINUM INDUSTRY WASTE: Arun S. Wagh, Energy Technology Division, Argonne National Laboratory, 9700 S. Cass Avenue, Argonne, IL 60439; W. W. Phil Robinson, Environmental Solutions Inc., 5711 Staples Mill Road, Richmond, VA 23228 Aluminum smelter plants produce potliner waste material during the electrolytic production of aluminum. For every ton of aluminum produced, =25-35 kg of spent potliner is generated. Annual U.S. generation of this waste is .. 120,000 metric tons. There is a large accumulation of this materia1 from earlier operations in the U.S. In addition, U.S. companies produce a large amount of this material abroad, primarily in Australia. A land ban on this waste is scheduled to be enforced in the U.S. in 1996, and similar bans are already enforced in several other countries. Spent potliner contains up to 50 wt% carbon (the basic liner material), contaminated by 15% each of fluorides of several alkaline metals, and up to 0.5% free or complexed cyanide. Many companies have developed processes to destroy the cyanides, stabilize the fluorides, and convert the potliner into harmless landfilled material. However, landfilling of this material is expensive, costing ..$150 per ton in the U.S. Pacific Northwest and even more in the Northeast For these reasons, we have conducted research to find ways to recycle this treated waste, and the results are presented in this paper. Representative samples of the waste were treated by a novel chemically bonded phosphate ceramic process developed at Argonne National Laboratory. This process is aroom-temperature treatment in which an acid-base reaction route is used to develop a binder using MgO and a phosphoric acid solution. The waste, which is mixed with MgO before the reaction, also participates in the reaction; the reacted slurry sets into hard and dense ceramics within .. I hour. Physical and mechanical properties and leaching characteristics of the stabilized product will be presented. Results of microstructure studies and mineralogy will be used to demonstrate the superior quality of the fmal waste form. The process can be used to produce valueadded structural products. Economics of the process will be discussed. Such product development may help to delist the treated spent potliner as an EPA hazardous waste. 3:50pm COMMONWEALTH ALUMINUM IMPROVES WASTEWATER TREATMENTPLANTPERFORMANCE AND REDUCES COSTS WITH NEW POLYMER TECHNOLOGY: Dr. Neil Isles, Finishing Mill-MRU Manager, Commonwealth Aluminum Company, P. O. Box 480, Lewisport, KY 42351; Scott Stedeford, Account Specialist, Betz Water Management Group, 8977 Technology Drive, Suite A2,

B40

Fishers, IN 46038 Commonwealth Aluminum, a manufacturer of aluminum coil products, located in Lewisport, Kentucky, was experienc::ng problems with their wastewater treatment process. The wastewater facility is designed to remove heavy metals from their paint line effluent stream. At the end of 1993, Commonwealth Aluminum shut down their old paint line and expanded the output of their modernized paint line. The result was a reduction in effluent water flow and solids loading, creating a problem with suspended solids carryover in the clarifier. To address this problem, Betz proposed the use of their Novus® polymer technology to replace ferrous sulfate and an anionic flocculant. The polymer trial began in December, 1994, and the process improvements were evident immediately. In January, 1995, Commonwealth Aluminum modified their treatment process to fully incorporate the Novus® technology. 1fhe results from the incorporation of this polymer include: improvement of clarifier effluent clarity; reduction of wastewater treatment costs; reduction in manpower; and ,;onsistent compliance of environmental permit limitations.

TUESDAY AM ADVANCES IN THE SCIENCE AND TECHNOLOGY OF TITANIUM ALLOY PROCESSING III: Cold Working I

Sponsored by: SMD TItanium Conunlttee, MDMD Shaping and Forming Committee Program Organizers: Prof. Isaac W,~iss, Prof. Raghavan Srinivasan, Mechanical and Materials Engineering Dept., Wright State University, Dayton, OH 45435; Dr. Paul Bania, Tunet Corporation, TImet-Hc:nderson Technical Laboratory, P.O. Box 2128, Henderson, NV 89009; Prof. Daniell Eylon, Graduate Materials Engineering, University of Dayton, Dayton, OH 45409

Tuesday, AM February 6, 1996

Room:B5-6 Location: Anaheim Convention Center

Session Chairpersons: P. Allen, TImet Corporation, TImet-Henderson Technical Laboratory, P.O. Box 2128, Henderson, NV 89009; D. Musi, NEFM Intemational Inc. Monaca, PA 15601

8:30 am Invited COLD FORMING OF BETA TITANIUM ROUNDS AND FLATS: Charles Pepka, Renton Coil Spring Co., 325 Burnett N., PO BOX 880, Renton WA 98057 The manufacture of springs and sheet metal components at Renton coil spring has changed throughout the years. The exotic materials of the late 70's is common place in today's manufacturing. Just in tillll: production has brought additional demands requiring faster setup times and more accurate work to be done. Cold formability of titanium has allowed the material to become fmancially competitive to stainless and steel material. The considerations of weight and corrosion resistance have allowed titanium materials to dominate in many ven~les. The paper will cover forming of beta titanium rounds and flats in many sizes. Cold forming of spring wire and stamp and form parts from sheet materials will be discussed. 9:00am TI-6AI-ZSn-4Zr-2Mo ALLOY F'OILS COLD ROLLED FROM STRIP CAST BY THE PLASMA MELT OVERFLOW PROCESS: I. M. Sukonnik, Texas Instruments, MS 10-28, Attleboro, MA 02703; T. A. Gaspar, Ribbon Technology Corporation, P.O. Box 30758, Columbus, OH 43230; W. D. Brewer, R. K. Bird, NASA Langley Research Center, MS 188A, Hampton, VA 23665 The combination of direct strip casting and cold rolling to produce titanium alloy foils offers many economic and metallurgical advantages over conventional ingot metallurgy practices. A TI-6AI-2Sn-4Zr-2Mo alloy strip, 0.5-mm-thick and lOO-mm-wide, cast by the plasma melt overflow process was cold rolled to a O.07-mm-thick, fully dense foil. The microstructure of the direct cast + cold rolled foil was more isotropic than ingot metallurgy foil. The chemical composition, gas content, microhardness and tensile properties of the direct cast + cold rolled foil was measured and found to meet or exceed MIL specifications. 9:20am THE RELATIONSIDP BETWEEN TEXTURE AND ANISOTROPY OF MECHANICAL PROPERTmS IN TITANIUM ALLOYS: Susan M. Kazanjian, Hinrich Hargarter and E. A. Starke, Jr., Department of Materials Science and Engineering, University ofVtrginia, Charlottesville, VA 22901 The texture of a material is the primary basis of its anisotropic behavior. Consequently, texture development during various processing steps is Important in order to understand and optimize processing parameters to develop the desired isotropy of mechanical properties and to obtain an insight into the mechanisms associated with deformation and fracture. A well bown example of the importance of texture is in deep drawing where a high r-value and low planar anisotropy is necessary to prevent tearing. During the past twenty years significant advances have heen made in the characterization of textures in a very quanti.tative way by the introduction of the orientation distribution function (ODF) which Clm be calculated from several pole figures. This talk will describe texture measurements of two titanium alloys and the correlation of parameters

1996 TMS ANNUAL MEETING AND EXHIBITION

determined from the ODF's with mechanical properties. 9:40am ADIABATIC SHEAR BANDS FORMED BY PUNCHING IN TI-6A1-4V ALLOY: Y. KosaJca, J. Kosin, Oregon Metallurgical Corporation, 530 34th Ave., Alhany OR 91321 Itis generally understood that the formation of adiabatic shear bands (ASB) is closely correlated to shock or dynamic loading. In this study, the formation of ASB along fracture surface was investigated by punching test with moderate cross-head speeds in TI6A1-4V alloys. Microstructure and strength of specimens were changed by heat treatment which includes mill anneal, recrystallize anneal, beta anneal, and solution treatment and aging. ASB was observed along fracture surface in any specimens. 1be width of ASB was in a range of 10 to 20 micrometers. Shear strength of samples heat-treated at beta appeared to be lower than that of alphaJbeta processed samples. 1bese values will be discussed in conjunction with their microstructures and other metallurgical factors. 10:00 am

THE EFFECTS OF RESIDUAL WORK ON THE AGEING RESPONSE OF BETA ALLOYS: P. J. Bania, P. G. Allen, TImet Corporation, Tunet-Henderson

Technical Laboratory, P.O. Box 2128, Henderson, NV 89009 When metastable beta alloys undergo deformation processing, particularly at elevated temperatures, the stored residual work is usually utilized to drive subsequent recrystallization. However, due to other considerations such as, for example, grain size, the recrystallization process may not always be driven to near completion. Also, at times an intermediate step of subgrain formation or polygonization is developed. Whether it is simple incomplete recrystallization or sub-grain formationipolygonization, the lack of complete recrystallization leaves added driving force that can significantly influence the precipitation process during ageing. This paper will review some experiences with metastable beta alloys relating to the effects of incomplete recrystallization on ageing behavior. 10:20 am COLD WORKING OF LCB TITANIUM ALLOY: I. Weiss, R. Srinivasan, M. Saqib, N. Stefansson, S. leClair, Mechanical and Materials Engineering Dept. Wright State University, Dayton. OH 45435; A. Jackson, Manufacturing Technology Branch. Wright LaboratorieslMaterials Directorate. Wright Patterson AFB, OH 45433 A study was undertaken to evaluate the cold forming characteristics of the low cost beta titanium alloy (LCB-TItanium) containing iron and molybdenum. Room temperature compression tests were conducted at strain rates in the range of 0.0 Is·, to 5s·'. 1be material work hardens at slow strain rates, while deformation heating and localization causes the material to flow soften at fast strain rates. 1be deformation behavior of the alloy will be discussed in terms of the microstructure of the alloy and changes in the mode of deformation observed by optical and transmission electron microscopy. 10:40 am BREAK 10:50 am DEFORMATION BEHAVIOR OF BETA21S SHEET: J. Reshad. I. Weiss. Mechanical and Materials Engineering Dept.. Wright State University, Dayton, OH 45435; T. F. Broderick, S.L.Semiatin, Processing Science Group. Wright Laboratory. Materials Directorate. Wright Patterson AFB, OH 45433 1be general bending and stretching characteristics of .060" beta2IS sheet will be presented. In particular, the sensitivity of the material to fail by localized necking will be discussed relative to strain hardening and rate sensitivity behavior. As the alloy is heat treatable, the influence of pre-conditioning heat treatments on deformation and failure will also be presented. 11:10 am FLOWFORMING AND SPINNING OF TITANIUM AND ITS ALLOYS TO PRODUCE WW COST INDUSTRIAL COMPONENTS: D.M. Zimmennann, Spin Forge International. EI Segundo, CA 90245 1be use of titanium and its alloys is usually justified for aerospace applications on the basis of using a cost premium to save weight compared to steel and other materials. Most industrial applications do not offer similar cost vs. weight trade off calculations in order to decide bill of materials. F1owforming and spinning technology bas recently been applied to produce a number of parts in the industrial sector with reasonable success. 1be advantage of limited machine cycle (spinning) times. reduced scrap generation. and higher input to application yield weight bas reduced costs to attractive levels. In some components. titanium can even be less expensive than steel. Several different applications are reviewed: temperature probe casing. micro-tunneling bands. small pressure tanks. and nuclear waste cannisters. Alloys reviewed include commercially pure titanium. beta titanium alloys, and TI-6AI-4V. 11:30 am

HYDROGEN TECHNOLOGY OF SEMIPRODUCTS AND PRODUCTS PRODUCTION FROM mGH STRENGTH TITANIUM ALLOYS: A.A. Ilyin. V.K. Nosov, A.M. Mamonov, V.N. Uvarov, A.V. Alexandrova. Moscow State University of Aviation Technology after K. Tsilkovsky. Metals Science Dep., Petrovka St. 21. K-31, Moscow. Russia. 103161 High strength titanium alloys such as TI-IQ-2-3 and VT22 are processed by hot forging. New treatment technology of titanium alloys base on reversible hydrogen alloying

1996 TMS ANNUAL MEETING AND EXHIBITION

give the possibility the use of deformation process at the room temperature (cold deformation). In the paper, the scientific basis and technological schemes for the production of sheets, foils and wire from TI-IQ-23 and VT221 high strength titanium alloys by means of cold rolling are described. Products, such as, hemispheres and "cups" produced by cold sheet forming are presented in material. Alloying with hydrogen not only gives the opportunity to produce shapes and semiproducts at the room temperature, but also. allows the development of special texture. such as. {1I1 J <1I0> and {Ill J<2II>. in the 8 phase titanium alloys. This type of texture is favorable for further cold deformation and for decreasing anisotropy in the mechanical properties of the deformed products. 1be application of hydrogen technology in this manner permits high strength to be developed in the products without quenching and ageing. Obtaining the high strength level of this alloys under vacuum annealing is important for sheet and foil products. 11:50 pm MANUFACTURE OF TITANIUM ALLOY FASTENERS: Viacheslav A. 'lblodin. Igor A. Vorobiov, Normal Corporation, 14 Utvinova St., Nizhny Novgorod. 603600, Russia TItanium and its alloys. which are very attractive for commerical applications, are used in aerospace structural and engine components where weight is important, and in chemical, oil and gas. nautical and medical applications, where corrosion resistance is required. TItanium bas also been used in consumer products, such as sporting goods and jewelry. Russian industry produces a wide range of titanium alloys for use as fasteners. This paper discusses the application of cold headed fasteners in several applications which meet DIN and ISO standards. One of the best Russian alloys is BT-16. containing a smaIl amount of AI and relatively large amounts of Mo and V, which can be heat treated to strengths up to 1410 MPa. This alloy is compared with TI-6AI-4V for various applications. 1be application of titanium alloys for fastening fragile composite and glass-fiber plastic materials will also be discussed.

ALUMINA & BAUXITE TECHNOLOGY I: Fundamentals I

Sponsored by: LMD Aluminum Committee Program Organizer: C.G. Hendriks. Kaiser Aluminum and Chemical Corp .•9131 Interline Drive. Baton Rouge, LA 10809 Tuesday, AM February 6. 1996

Room:A6 Location: Anaheim Convention Center

Session Chainnan: Fred S. Williams, Alcoa, Point Comfort. TX 11918

8:30 am Invited MASS TRANSFER CONSIDERATIONS DURING THE PRECIPITATION OF

DEPOSITS ON ALUMINA REFINERY HEAT EXCHANGER SURFACES: Dr. Victor Rayvnan. 1329-D Alicante Rd .• Carlsbad, CA 92009 A theoretical mass transfer equation was established for the analysis and calculation of the heat exchanger fouling during alumina manufacture as a function of flow velocity. heating pipe diameter, and heated medium temperature. In order to determine the interaction of all hydrodynamic quantities with respect to scale deposit formation. a pilot heat exchanger of the type "tube-in-tube" was designed. built and tested for heating the aluminate-silicate solution. 1be suitability and reliability of the theoretical and practical results were corroborated by a comparison with known data for the heating various bauxites and nepheline slurries. 1be data from one of the tests were applied to the design and operation of the aluminum plant. 1be results showed that lowering the concentration of the deposit forming substances by prior heating of the slurry or aluminate-silicate solution is the only alternative for mitigating the precipitation of deposits on heat exchanger surfaces. 9:00am RECENT DEVELOPMENTS IN STANDARD PROCEDURES FOR THE ANALYSIS AND TESTING OF SMELTER GRADE ALUMINA: Grant Foster. P.O. Box I. Alcan Australia Ply. Ltd., Kurri Kurri Works. Kurri Kurri N.S.W. 2321 Standards Australia committee MN9 - (Alumina and Materials Used in the Production of Aluminum) was formed in 1983 in response to requests from the alumina and aluminum industries in Australia for standardized analytical procedures and test methods for measurement of physical and chemical properties of alumina. From 1983 to 1995 the committee has published the following seven standard methods. AS 28191986 Part I) Determination of Loss of Mass at 300000C and lOOOooC. AS 2819-1991 Part 2) Determination of particles passing a 20um sieve. AS 2819-1991 Part 3) Determination of alpha alumina content by X-ray diffraction. AS 2819-1991. Part 4) Determination of specific surface area by nitrogen adsorption AS 2819-1994. Part 5) Determination of angle of flow AS 2819-1995. Part 6) Determination of the mass distribution of particles sizes using electroformed sieves AS 2819-1995. Part 1) Determination of trace elements - wavelength dispersive x-ray fluurescence spectrometric method. This paper reviews some of the principles involved in the development of these analytical procedures and test methods. 1be current work program. which includes methods for determination of dustiness and attrition index is also discussed. 1bese methods and the ongoing development and review program provided by MN9 are meeting the industries' need for up-to-date, relevant standard methods.

B41

9:30am IMPURITIES BUILD UP AND BEHAVIOR OF LOW ORGANIC BAUXITE OF PANCHPATMALIIN THE BAYER PROCESS: R.C. Mohanty. B. K Satpathy. National Aluminum Co. Ltd.• IDCO Tower. Janpath. Bhubaneswar- 151 001. Orissa. India The major portions of the impurity enter the Bayer process through bauxite. Beyond the critical levels the impurities alter the kinetics of the process. Among the various impurities. organic carbon is the most prominent one that affect the precipitation operation and the product quality. depending on the level of organics with respect to the precipitation process adopted. A lower level of less than 0.2% organic carbon in the Panchpatmaii bauxite of Nalco although has allowed smooth operation of the plant for the last seven years. the same has been a matter of concern today. The present paper has analyzed the situation at Nalco plant with respect to various impurities build up over the years. The occurrence of organic in bauxite and its hehavior in the Bayer process has been analyzed. Even such low level organic has been a point to watch for achieving productivity and quality product. The paper also highlights how the problems of sodium oxalate. emerging from the organics has been resolved at Nalco with the help of Aluminum Pecheney. France. Results are supported with experiments. plant data and characterization of bauxites and hydrates through instrumental techniques. 10:00 am LAYERED DOUBLE HYDROXIDE FORMATION IN BAYER LIQUOR AND ITS PROMOTIONAL EFFECT ON OXALATE PRECIPITATION: A. J. Perrotta, Alcoa Technical Center, Alcoa Center, PA 15069-0001; F. S. Williams. Alcoa Operations. Pt. Comfort. TX 11918-0101 Enhancing the precipitation of sodium oxalate from Bayer process liquor to improve the quality of alumina product remains an important objective for Bayer refining. The formation of layered double hydroxides by the reaction of alkaline earth oxides. such as lime and magnesia. with Bayer liquor gives a crystal structure which is capable of intercalating anions, both inorganic and organic. within its structure. Both lime and magnesia. with long contact times in Bayer liquor, show layered double hydroxide formation. This layered double hydroxide formation is accompanied with a decrease in the sodium oxalate content in the liquor from about 3 gil to below 1 gil. Short contact times lead to a destabilization of the liquor which facilitates sodium oxalate precipitation. Additional work on magnesium hydroxide shows. in comparison to lime and magnesia. much less layered double hydroxide formation with equivalent residence time in the liquor. Destabilization of the liquor also occurs giving enhanced oxalate precipitation with less alumina being. consumed in agreement with lower layered double hydroxide formation. Thermal regeneration of these structures followed by in situ recrystallization in Bayer liquor also gives enhanced oxalate precipitation. suggesting that there is an opportunity for a regenerable oxalate reduction system. 10:30 am BREAK 10:50 am DETERMINATION AND EVALUATION OF ORGANIC BALANCES OF ALUMINA REFINERIES: Dr. Karoly Solymar. Mrs. Marta Gimpl- Kazar. Mrs. Elconora Molnar. Hungalu Eng. and Dev. Center. ALUTERV-FKI Ltd. H-1502 Budapest. P.O. Box 308. Hungary The adequate measures to control the organic level and especially that of humates and oxalate need the determination and evaluation of the organic balance of the given . alumina plant. The organic materials of the different alumina refineries have special compositions. therefore it is very difficult and limited to draw general and common conclusions for all alumina refineries concerning the role and control of organics. The methodology of the determination of the organic balance of an alumina plant will be discussed. The qualitative and quantitative determination of the organic materials has been performed by means of gas-chromatography of the esterized organic materials. The organic balance of three different alumina refineries will be compared and evaluated relating to the total organic. humate and oxalate contents. 11:20 am SODIUM HUMATES, HOW BAD ARE THEY FOR THE BAYER PROCESS?: Kwat The. Principal R&D Consultant. Sharon Rowe. Technology Department. Queensland Alumina Umited (QAL). PO Box 1. Gladstone Queensland 4680. Australia Numerous process problems have been attributed to sodium humates. A study was therefore undertaken which examines the effect of "Bayer humates" on the properties of the liquor. the behavior of sodium oxalate and the precipitation process of the hydrate. 11:50 am BOUND SODA INCORPORATION DURING HYDRATE PRECIPITATION: EFFECTS OF TEMPERATURE, CAUSTIC, AND ORGANICS: K. McCormick. J. Hunter. H. Warren. L Armstrong. Senior R&D Chemist. Technology Department. Queensland Alumina Umited (QAL), P.O. Box I. Gladstone, Queensland 4680. Australia Soda is incorporated into aluminum hydroxide ("Hydrate") during the precipitation stage of the Bayer Process. A review of general literature shows the predominant effect is alumina supersaturation. Our research extends the literature by quantifying the secondary effects of temperature. caustic and organics on soda incorporation beyond their effect through alumina supersaturation. This work advances industry knowledge towards better control of soda incorporation in the refinery in the pursuit of higher and more consistent product quality for smelter grade alumina.

B42

ALUMINUM ALLOYS FOR PACKAGING II: Session I

Sponsored by: SMD Non-Ferrous Metals Committee, Light Metals Division Program Organizer. Professor James G. Morris. Chemical & Materials Engineering. Director. Ught Metals Research Labs .• University of Kentucky. Lexington. KY 40506 Tuesday. AM February 6. 1996

Room:A2-3 Location: Anaheim Convention Center

Session Chairperson: James G. Morris. 115 Anderson Hall, University of Kentucky, Lexington, KY 40506

8:30 am Invited METALLURGICAL ASPECTS OF ALUMINIUM FOR CANMAKING: P. D. C. Rogers, CarnaudMetalbox pIc Research and Development. Downsview Road. Wantage. Oxon. OXI2 9BP. UK The properties of aluminum alloys used for canmaking have been the object of much research, as the demands of the Canmakers for zero defects and trouble-free running have grown. The main requirement is consistency. but the puzzle still remains why coils will not run one day. but will nm the next. Work has shown the importance of controlling earing. mechanical properties. coil shape and thickness. both from coil to coil and within the coil. This mean:1 that traditional testing is inadequate, and more emphasis has to be placed on process control. It is not difficult to write a specification which describe the main metallurgical and dimensional features of body. end and tabstock. but it will not describe why material from one mill will run better. whilst the reverse is true elsewhere. While we do not have clear answers to all problems, our understanding is growing. 9:00 am Invited MICROSTRUCTURE AND TEXTURE EVOLUTION DURING THERMO MECHANICAL PROCESSING: Niels Hansen. Materials Department. Riso National Laboratory. DK-4000 Roskilde. Denmark The microstructure and texture evolution during thermomechanical processing of materials and alloys is at present studied extensively by modem automated techniques such as EBSP (Electron Back Scattering Pattern) in a SEM and by the Kikuchi pattern technique in a TEM. These experimental studies are followed by the development of physical based models of deformation and recrystallization microstructures and textures. On this basis recent studies of aluminium and aluminium alloys deformed by rolling or plain strain compression 3Jre summarized with emphasis on grain subdivision on different size scales into crystallites separated by dislocation boundaries. The scale of this subdivision refines with increasing stress and strain which also significantly effects the crystallographic orientation of the crystallites. These observations are analyzed on basis of the actual slip pat~~rn and the tendency of dislocations to accumulate in low energy configurations. Finally the observations and the results of the subsequent analysis are related to the thermomechanical behavior of aluminium and aluminium alloys. 9:30am THE EVOLUTION OF CRYSTALLOGRAPmC TEXTURE DURING THE RECRYSTALLIZATION OF HOT ROLLED 5182: Phillip A. Hollinshead, Alloy Technology Division. Alcoa Labomtories. Aluminum Company of America. Alcoa Center. PA 15069 The progression of recrystallizati.on in initially unrecrystallized hot rolled 5182 was studied by various techniques. Recrystallization nuclei could plainly be observed in an optical microscope after heating 22 seconds at the hot mill exit temperature of 615 OF and recrystallization was complete :It 10 seconds. Using the backscattered Kikuchi diffraction (BKD) technique in the scanning electron microscope (SEM) the cube texture was found to be the only preferred orientation in the early stages of recrystallization. Pole figures from nearly fully recrystallized material were very similar to those from the initial recrystallized grains. imlicating that the final texture was determined at a very early stage of the recrystallization process. The copper (Cu) component. {112}<111>, showed signs of decl1~asing more rapidly in the early stages of recrystallization relative to other components of the rolling texture. The last portion of the microstructure to recrystallize was found to be bands of the brass (Bs) component. {110 }<112>. In 8 cube oriented grains at the start of recrystallization. 6 were associated with deformed regions of the cube orientation. 9:50am EVOLUTION OF ROLLING AND RECRYSTALLIZATION TEXTURES IN A HOT ROLLED CAN BODY STOCK ALLOY: Boolute Ren. Manufacturing Technology Laboratory, Corporate Research & Development, Reynolds Metals Company, 3326 East Second Street. Muscle Shoals. AL 35661-1258 Hot rolling is a very important step in fabrication processes of aluminum beverage can body stock. II has been indicated that the development of microstructure and texture of AA 3104 alloy during hot rolling depends significantly on hot rolling procedures.ln industrial thermomechanlical processing. a hot band material can be produced with a deformation structure and a dominant deformation texture, or with a recrystal-

1996 TMS ANNUAL MEETING AND EXHIBITION

lized structure and a dominant recrystallization texture, even though the hot band is not subjected to a hot line anneal. In addition, a partially rec~s~lized hot ban~ ma~rial can be obtained, which has both deformation and recrystalbzatlOn texture onentatlons. In the present work, the texture evolution of AA 3104 alloy during hot rolling was studied. It was found that the Copper orientation was predominant for the deformation texture on the surface of the hot band while the Brass orientation was predominant at the mid-thickness. In tbe subsequent annealing or self-annealing process, this difference in the texture orientation influenced the recrystallization behavior for the structure from the surface to the mid-thickness of the hot band material.

ALUMINUM REDUCTION TECHNOLOGY III: Cell Materials

10:10 am BREAK

Tuesday,AM February 6, 1996

10:30 am ANISOTROPY IN ROLLED ALUMINUM SHEET AS IDENTIFIED BY ULTRASONIC AND METALLURGICAL ANALYSIS: Wei-yang Lu, Shermann Min, Darcy Hughes Sandia Nat. Lab., Uvermore CA 94550; Baolute Ren Reynolds Metals Company, Mussel Shoals, AL 35661-1258 The use of ultrasonic techniques to measure the material anisotropy of rolled metal sheets has been demonstrated in many studies. Recently, electromagnetic acoustic transducer (EMAT) and laser ultrasonic (LU) methods have been favored since both are noncontact in nature and therefore potential candidates for in-situ process monitoring and control in rolling mills. In general wave propagation velocities are measured and then used to calculate the texture coefficients W400' W'1Jl and W.... In addition to W4XX values, more information can be obtained from an ultrasonic interrogation. This paper describes the identification of crystallographic textures in aluminum plates using the above noncontact ultrasonic approaches. The textures of samples which include a variety of crystallographic components are evaluated using ultrasonic, x-ray methods, and transmission electron microscopy and micro-texture analysis. The microstructure and texture analysis results are compared tu the results of anisotropy measured by mechanical testing. This work supported by U.S. DOE under contract No. DE-AC04-94AL8500. 10:50 am THE ASYMMETRY OF EARING IN ROLLED ALUMINIUM: H. E. Ekstrom, Granges Technology, Finspang, Sweden, A. Oscarsson, Swedish Institute for Metals Research, Stockholm, P. Bate, IRC, Univ. of Birmingham, UK In practice earing may cause two different types of problem. The lowest valley of the cup rim determines the level of cutting and the ~un! of was.tage in tri.mn.ung. ~ highest peak may cause failure in cupping due to pmchmg: Earin~ hehavlO~ !S assocIated both with mechanical anisotropy in the sbeet and With toobng conditions. The latier often cause a skewness of the complete 0-360' profile and friction between tools and sheet decreases the overall earing. The contribution from the material depends mainly on the average through thickness c~stallographic texture of the s~.t in non-~at treatable aluminium alloys. DependlDg on sheet alloy composlllOn and Its therrnomechanical history the texture can vary significantly giving rise to very different earing profiles. Textures and experimental earing profiles have been examined .for different alloys in different conditions. Calculated earing profiles show good qualitative agreement with experimentally measured ones. The present ISO standard for an earing test does assume a greater symmetry of earing than actually occurs. A new me~od is proposed to replace the current one in which the difference between the mean height of tbe highest pair of peaks and the lowest pair of valleys is related to the mean cup height. 11:10 am THE EFFECT OF INITIAL MICROSTRUCTURE ON THE EVOLUTION OF CRYSTALLOGRAPHIC TEXTURE IN AA 3004 DURING SIMULATION OF HOT ROLLING: G.J. Marshall, M.W. Merodith, RK Bolingbroke, Alcan International, Banbury Laboratory, Southam Road, Banbury, OX16 7SP, UK Controlling property anisotropy, particularly earing, in can body sheet is dependent upon the annealed microstructure after hot rolling which is used to ~a1ance the textures generated during cold rolling. Essential during the hot deformation and subsequent annealing is the formation of sufficiently strong cube texture to ensure low final gauge earing. Low earing has been the industry accepted standard for many years and current research aims are addressing the growing concern of earing consistency. Significant factors in obtaining consistent hot rolled texture are the deformation conditi~ns and ~e initial starting microstructure. This study has addressed the role of Mn sobd solution content and AIMnSi dispersoid size and distribution on the evolution of texture during laboratory simulation of hot rolling. It will be shown that both process conditions and the size and spacing of dispersoids are critical factors in controlling texture. In contrast, significant changes to the Mn solid solution content wen: considerably less influen~al. The results will be discussed in relation to current theones for cube texture evolution during hot rolling of aluminium alloys. 11:30 am STATIC RECRYSTALLIZATION OF AN AI·IMn ALLOY DURING THERMOMECHANICAL PROCESSING: P.l. Orselli Rossi, C.V.G. Industria Venezolana de Aluminio, C.A., Zona Industrial Matanzas, PO BOX 302, Puerto Ordaz, Venezuela; C.M. Sellars, The University of Sheffield, Department of Engineering Materials, Mappin Street, Sheffield, S1 3JD, UK The kinetics of static recrystallisation of an AI-IMn alloy have been investigated by rolling tapered slabs and by plane strain compression (PSC) tests at high temperatures. A significant through-thickness recrystallisation gradient (macro-heterogeneity) has been measured on samples from rolled and annealed tapered slabs. The kinetics follow reasonably well an Avrarni equation as long as significant precipitation does not take place. The Avrarni exponent increases systematically with strain, probably a result of an increasing micro-homogeneity in deformation. Since in this alloy, precipitation occurs concurrent with recrystallisation, irregular recrystallised structures are obtained.

1996 TMS ANNUAL MEETING AND EXHIBITION

Sponsored by: LMD Aluminum Committee Program Organizer: Ms Fiona J Stevens, Comalco Research and Technology, Comalco Research Centre, PO Box 316, Thomastown, Victoria 3074, Australia Room:A9 Location: Anaheim Convention Center

Session Chairperson: Jarg Mittag, SGL CARBON AG, Werk Griesheim, Stroofstrasse 27, D-65933 Frankfurt, Germany

8:30am USE OF OLIVINE AS A DRY MIX BARRIER FOR THE PRODUCTION OF ALUMINIUM REDUCTION CELL POT LINING REFRACTORIES AND METHOD FOR THE MINERALOGICAL IDENTIFYING OF CREATED PHASES: Ove Thornhlad, Svenska Hoganas AB, Bruksgatan, S-263 83 Hoganas, Sweden; Egil Grolnes, Hoganas Eldfast AS, Fjellhamarveign 46, Postboks 54, N1472, Fjellhamar, Norway In recent years, flfeclay bricks and low conductivity insulations have been used in the bottom of alumina reduction cells to save power energy. In order to protect the thermal insulation from attach by bath between the carbon cathode and bottom insulation, an olivine based refractory barrier is developed. Bath resistant tests have shown that the barrier has good resistance to bath attach, bath penetration and thermal conductivity. The material layer, instead of the fireclay bricks layer, was used in a Soderberg alumina reduction cell in Norway. After 16 months operation the bath was not penetrated through the barrier material layer and there was normal heat loss through the bottom. A petrological method, by microscopical examination of thin sections, has been used for the identifying of created phases.

9:00am A CRJTICALANALYSIS OF SODIUM MEMBRANES TO PROTECT SODIUM RELATED CARBON CATHODE DAMAGE IN THE HALLHEROULT CELL: I.A. Sekhar, J. Uu, J.J. Duruz,lnternational Center for Micropyretics, University of Cincinnati, Cincinnati, OH, 45221-0012; V. de Nora, MOLTECH S.A., 9, Route de Troinex, 1227 Carouge, Geneva, Switzerland During the past few years, cathodic membranes have become available which are designed for improved wettahility and reduced sodium attack of the cathode. The microstructure and sodium trapping mechanisms of such membranes are discussed. The efficacy of the different microstructures in relation with their electrical and chemical properties towards the prevention of sodium attack is illustrated for various start-up conditions and current densities. Results from using such membranes in commercial cells will also be shown. 9:30am INFLUENCE OF TEMPERATURE AND CONTACT PRESSURE BETWEEN CAST IRON AND CATHODE CARBON ON CONTACT RESISTANCE: Frank Hiltmann, Jorg Mittag, SGL CARBON AG, Werk Griesbeim, Stroofstrasse 27, D65933 Frankfurt, Germany. Anne Store, Harald A. Oye, Institute of Inorganic Chemistry, The Norwegian Institute of Technology, N-7034 Trondbeim, Norway The contact resistance between cathode block and cast iron as collector bar casting agent is said to give a Significant contribution to the cathode voltage drop (CVD) measured in aluminum electrolysis cells. Five typical cathode carbon qualities were examined in this respect from room temperature to 95QooC. At cell operating temperature the contact resistance is found to be usually negligible in comparison to other resistance effects. However, inadequate casting parameters and cell ageing processes leading to an improper contact or contact pressure may impede the current flow between cast iron and cathode block considerably and hence result in an increasing CVD with cell age. 10:00 am BREAK 10:15 am CHARACTERJZATION OF ANODE STUB CORROSION IN HALL REDUCTION CELLS: Xiangwen Wang and Ray D. Peterson, Manufacturing Tecbnology Laboratory, Reynolds Metals Company, 3326 East 2nd Street, Muscle Shoals, AL 35661-1258 Mild steel is widely used as a structural material in tbe aluminum smelting industry. In prebaked-anode reduction cells, the stability of the steel used as an anode "stub" against high temperature oxidation and corrosion is very important with regard to its full service life and maintaining aluminum cleanliness. This paper deals with the accelerated corrosion of the steel material used as anode stubs in the presence of the sulfurcontaining anode gases. Oxidized scale and tbe interface region of the oxidation reaction zone in a stub from a reduction cell were fully examined using scanning electron microscopy and Z-ray diffraction. The sulfur from the bath and the anode carbon, released as SO" play an important role in accelerating the anode stub corrosion process. A sulfidation-oxidation corrosion mechanism is proposed to support the corrosIon phenomena observed on tbe steel anode pieces.

B43

10:45 am CRITICAL ISSUES FOR THE DEVEWPMENT OF AN OXIDATION TREATMENT FOR CARBON ANODES: I.A. Sekhar, J. Uu, International Center for Micropyretics, University of Cincinnati, Cincinnati, OH, 45221-0012; V. de Nora, MOLTECH S.A., 9, Route de Troinex, 1227 Carouge, Geneva, Switzerland The oxidation of the carbon anodes in the Hall-Heroult Cell, leads to an increase of the cost of aluminum on account of a consumption of more than 0.43kg of carbon per kg of aluminum produced instead of the theoretical amount which is 0.333kg. Several oxidation prevention treatments have been proposed which have worked in the laboratory, but have fallen short of the expected performance when the same treatments have been applied to a commercial anode. An investigation of this effect was carried out and a new theory is proposed to explain the anode degradation. It is found that a brittleness limit may be encountered in the cell prior to oxidation related loss. This limit is examined and a new treatment is proposed to improve the anode consumption by reducing the anode brittleness and further improving the oxidation resistance. U:15 am THE UTILISATION OF COMPOSITE CARBON-SILICON CARBIDE SIDEWALL BLOCKS IN CATHODES: Edward I. Curtis, Paul D. Mascieri, Alton Tabenlaux, Reynolds Metals Company, Corporate Research and Development, 3326 East Second St, Muscle Shoals, AL 35661-1258 A recently developed carbon-silicon carbide composite sidewall block has performed perfectly in the cathode lining of 180kA prebake reduction cells for nearly two years. The composite block provides the opportunity to extend the potlife of cells operating with conventional carbon sidewalls, or can offer excellent cost saving while maintaining the desired operational results in cells using full size silicon carbide bricks. The composite blocF consists of a calcined anthracite carbon block, nitride-bonded silicon carbide and a glue or carbonaceous cement. The exact thickness or dimensions of the carbon and silicon carbide layers are optimised for each reduction cell sidewall design. Considerable time and research was dedicated to the bonding of the composite materials in order to develop the required strength to withstand the extremely harsh environment in the reduction cell. The advantage of the silicon carbide layer is to provide excellent resistance to erosion and oxidation of the sidewall lining while the carbon provides the added flexibility in design specification due to the case of machining. This paper discusses the research work involved in developing the required composite bond strength, thermal conductivity calculations for application of the composite block in cells, and measured plant results of side ledge freeze profiles and cathode shell temperatures to date.

CARBON TECHNOLOGY I: Binder Pltch/Petroleum Coke

Sponsored by: LMD Aluminum Committee Program Organizer: Use Castonguay, Alcan International Ltd., Arvida Research and Development Centre, PO Box 1250, 1955 Mellon Blvd, Jonqui~re, Qu~bec, Canada, G7S4K8 Tuesday,AM February 6, 1996

Room:A7 Location: Anaheim Convention Center

Session Chairperson: Mr. William R. Roder, Reilly Industries, Inc., 1500 South TIbbs Avenue, PO Box 42912, Indianapolis, IN 46242-0912

8:30am

LABORATORY PITCH PAR AND POM STUDY: E. R. McHenry, W. E. Saver, Koppers Industries, Inc., 1005 William Pitt Way, Pittsburgh, PA 15238 The polynuclear aromatic hydrocarbon (PAH) measurements for a series of pitches are reported as a function of feedstocks and product softening point. The relationship of PAH to the polycyclic organic matter (POM) which evolves during the pitch carbonization (baking) cycle, is also studied. By combining selected feedstocks and processes, a reduction in toxic PAH and POM content in pitches can be achieved.

9:00am EVAPORATION AND VAPOUR CHARACTERISATION OF LOW-PAR BINDERS FOR SODERBERG CELLS: Marl Eie, Harald A. Oye, Institute of Inorganic Chemistry, The Norwegian Institute of Technology, University of Trondheim, N-7034 Trondheim, Norway; Morten Sorlie, Elkem als Research, PO Box 40 Vagsbygd, N-4602 Kristiansand, Norway Polyaromatic hydrocarbon (PAH) vapour emissions from coal-tar pitches used as binders in the anode paste production can be an environmental concem for aluminium smelters. Pitch producers have responded to this by offering PAH-reduced anode binders, but with quite variable PAH distributions. Some producers have reduced both light and heavy PAH compounds while others have reduced the content of heavy PAH but reintroduced some light PAH compounds by using PAH-containing cut-back oils. As each country has different regulations and legislation, similar PAH emission profiles may be classified as satisfactory in one country and unacceptable in another. The paper gives emission characterisations of such PAH-reduced pitches. 9:20am SHOT COKE: P. I. Ellis, Great Lakes Carbon Corporation, PO Box "C», Port Arthur, TX 77640; J. D. Bacha, Chevron Research and Technology Company, 100 Chevron Way, Richmond, CA 94802-0627

B44

Shot coke is not desired in aluminum anodes, but 55% of all delayed coke in USA contains shot coke. Shot coke formation is due to refinery crude oil economics and improved operations which increased the asphaltene content in the coker feedstock. Diagrams showing how shot coke fornls in the coke drum, along with a review of current technology on shot coke will be covered.

9:40am A STUDY ON COKE DUST GENERATION IN A ROTARY CALCINATION KILN: Y. S. Kocaefe, F. Dahmane, R. T Bui, Process and Systems Engineering Research Group (GRIPS), Department of Applied Sciences, University of Qu~bec at Chicoutimi, 555, boul. de l'UniversiW, Chicoutimi, Qu~bec, Canada, G7H 2BI; Andre L.Proulx, Alcan International LtIIl., 1955 Mellon Blvd, PO Box 1250, Jonqui~re, Qu~bec, Canada G7S 4K8 One of the steps in the production of carbon electrodes is the green coke calcination. This allows the removal of moisture and volatile matter, and also the restructuring of the coke. Coke dust is a by-product of tbis process and has an important impact on the process. It causes environmental problems and loss of production. Also, it affects the process through its combustion and its participation in the heat transfer. A study has been carried out to investigate (both experimentally and through mathematical modelling) this dusting phenomenon. This paper will present the results of the experimental study which show the influence of various parameters on the coke dust generation. 10:00 am BREAK 10:20 am

MATllEMATICAL MODELLING OF A ROTARY HEARTH CALCINER: R. Fernandez, STATOIL Technical Servicl: Centre, PO Box 3, N-5154 Mongstad, Norway; H. C. Meisingset, and J. G. Balchen, The Norwegian Institute of Technology - NTH, Department of Engineering Cybernetics, N-7034 Trondheim, Norway Calcination of petroleum coke is a tllermal process where green petroleum coke is heat-treated to a pre-determined telllpl,rature. During heat treatment, the associated moisture is removed and volatile combl~sti1e material (VCM) released. This is bumed in the gas phase giving the energy to sustain the process. In addition, structural changes take place. The combination of the final calcination temperature and the residence time determine the final density of the calcin,ed coke. Depending on its further use, different real density requirements may arise. It is important to control the dynamics of the calcination process so that the specified final quality is achieved. A dynamic mathematical model of a Rotary Hearth Calciner is presented. The model is based on physicochemical laws involving the most important phenomena taking place and the relevant calcination parameters. The temperature profile in the coke bed is predicted which in terms is related to the real density of the coke. 10:40 am STEPS FOR GREEN COKE CALCINATION - MATHEMATICAL MODEL AND PRACTICAL TESTS AND EXPERIENCES: Dr H. Pnldel, ESSO AG, Karlsruhe Refinery, ESSO-Str. 1,76187 Karlsruhe, Germany The different steps for green coke calcioation like: water evaporation, drying, VCM evaporation, VCM burning, heating-up ramps, soaking period are calculated with a mathematical model. The results are compared with practical experiences for regular calcined coke production and with sampling programs during calcination process. The results are important for adjusting calcination conditions to achieve best calcined coke qUality.

CAST SHOP TECHNOLOGY III: Semi-Solid Metal Processing

Sponsored by: LMD Aluminum Committee Program Organizer: Julian V. Copenhaver, Technical/Quality Manager, NSA A Division of Southwire, PO Box 500, Hawesville, KY 42348 Tuesday, AM February 6, 1996

Room:AIO Loc:ltion: Anaheim Convention Center

Session Chairperson: Dr. Andreas N. Alexandrou, Worcester Polytechnic Institute, 100 Institute Road, Worcester, Massachusetts 01609-2280

8:30am SEMI-SOLID PROCESSING: NEW ADVANCES IN NET SHAPE FORMING: Dr. Stuan B. Brown, Failure Analysis Associates, Three Speen Street, Framingham, MAOl701 Net shape and near net shape manufacturing using semi-solid metals has been undergoing a remarkable surge in activity. A number of new companies and initiatives have begun in semi-solid metal and compolnent production. All exploit the well-known advantages of semi-solids, e.g., lower temperatures, reduced porosity, and improved mechanical properties than comparable (:ast metals. The flow behavior of semi-solids is still incompletely characterized, however, and must be understood to fully exploit these advantages. This presentation sUllUll8lizes the last decade of research into the flow of semi-solid metals and discusses the implications of this research for net shape manu-

1996 TMS ANNUAL MEETING AND EXHIBITION

facturing using semi-solids. This research includes the effect of temperature, liquid! solid fraction, particle morphology, flow history, and methods for producing the semisolid state. The presentation closes with a discussion of some particularly exciting new semi-solid processes. 8:50am SEMI-SOLID FORMING OF LIGHT METALS: NEW CHANCES FOR LIGHT WEIGHT STRUCTURES IN TRANSPORTATION APPLICATIONS: Dr. G. Hin, R. Cramer, A. Wmkelmann, T. Witulski, M. Ziligen, EFU Gesellschaft fUr Ur-lUmformtechnik mbH, Jligemeusstrasse 22, 0-52152 Simmerath, Germany Near net shape Semi-solid Forming of Ught Metals offers the chance to achieve high strength, complex shaped components at economic cost levels. Possible applications include unsprung masses (i.e., steering knuckle) as well as thin-walled body structures or pressure-tigbt components in the brake system. The paper will present the actual status of R&D work carried out by EFU in the field of raw material production, reheating and semi-solid forming: Pilot scale equipment has been used to produce thixotropic billets from the following alloys: AISi7Mg (356); AlMgSii (6082); AlCuMg2 (2024); AISil8CuMgNi; and AI-Metal Matrix Composites. The typical billetdiarneters are 3 incb and 4.5 incb (length up to 3.5 m); For inductive reheating of these billets a computer controlled system bas been developed, which allows closed loop control of the beating cycle by using a special inductive sensor to monitor the billet condition (liquid fraction, softness, melt loss). Recent experiments demonstrate that this control system allows a flexible response, i.e., in case of sbort term interruptions caused by the forming system. Various demonstrator parts have been produced from all alloys mentioned above and Magnesium alloy AZ 91. The demonstrator parts include a steering knuckle and a thin-walled rib-type structure (average wall thickness 2.5 mm, outer dimensions 500 mm x 150 mm). In the case of alloy AlMgSil, mechanical properties whicb have been determined from specimens taken from semi-solid formed parts are close to the requirements according to DIN 1749 standani for forged parts. This implies an advantage of 6xxx alloys, compared to other thixoformed alloys (356, AZ 91). In SSM-Forming of MMC's, the mechanical properties are according to the specification after conventional processing. In this case the major advantages are material saving and reduced macbining due to near net shape semi-solid forming.

9:10am SEMI-SOLID METAL (SSM) CASTING: APPLICATIONS AND CASE STORIES IN ALUMINUM AND MAGNESWM: Dr. Kenneth P. Young, Buhler Incorporated, CH-924O Uzwil, Switzerland A semi-solid (SSM) casting process based around real time controlled borizontal cold chamber die casting machines is described. SSM casting has been successfully applied to both aluminum and magnesium alloys with the largest sbot weigbt to date being 11.5 kilogram (25 pounds). SSM casting offers a unique combination of near net shape and premium quality mechanical performance. Aluminum components generally require either partial (T5) or full (T6) heat treatment, providing attractive combinations of strength and ductility. In the case of magnesium alloys, the ability to completely eliminate liquid metal handling from the casting floor allows SSM casting to compete with magnesium cold cbamber die casting. In these instances, less massive gating than for aluminum is required and normal trimming operations can be considered. A series of case studies are presented together with projections for future developments. As raw material costs decrease and secondary recycling becomes viable, it is expected that SSM casting will be considered for a mucb wider range of applications. 9:30am RHEOMOLDING A ONE-STEP PROCESS FOR PRODUCING SEMI-SOLID METAL CASTINGS WITH WWEST POROSITY: Dr. K. K. Wang, Nan Wang, Hsuan Peng, Sibley Scbool of Mechanical and Aerospace Engineering, Cornell University, Upson Hall, Ithaca, NY 14853 Rheomolding is a new idea of combining rheocasting in an injection-molding-like apparatus to achieve net-shape manufacturing of semi-solid metal parts with lowest porosity. Instead of solid pellets or chips in conventional injection molding or thixomolding, as starting material, rheomolding begins with molten metal in a crucible. As the fluid enters the extruder, it is always subject to cooling along the barrel and begins to crystallize on the cold wall. The screw rotation and the moving fluids will break the dendritic crystals on the wall and form solid particles suspended in the remaining fluids. By adjusting the temperature profile along the barrel and screw rotating speed, one could achieve various degrees of solid fraction and perhaps manipulate particle sizes. A prototype Rheomolding machine has been built, and extensive experiments have been carried out with Sn-15% Pb alloy. In this paper, some details on temperatore control, metallographical examinations, and resulting product quality and repeatability are discussed.

9:50am SEMI-SOLID FORMING OF HYPEREUTECTIC AI-SI ALLOYS: [ttipon Diewwanit, Dr. Merton C. Flemings, Massacbusetts Institute of Technology, Department of Materials Science and Engineering, Building 8-407, 77 Massachusetts Avenue, Cambridge, MA 02139-4307 Hypereutectic aluminum-silicon alloys, containing up to about 18% Si, are commercially cast but present difficulties due to their high liquidus temperature and wide freezing range. This paper describes research on fundamentals of semi-solid forming of these high silicon alloys. Viscosities of semi-solid alloys containing up to 50% Si are measured in a Searle type viscometer. Structure and viscosity are related to processing conditions. Behavior of the alloys with respect to structore and formability is similar to that of hypoeutectic alloys. The research demonstrates the feasibility of part forming of alloys containing up to 50% silicon, at temperatores as low as the Al-Si eutectic, 577 celsius.

10:10 am ORIENTATIONAL RELATIONSIDPS IN RHEOCAST AI-6.5wt%SI: S. Sannes, SINTEF Materials Technology, N-7034 Trondheim. Norway; Dr. Lars Amberg, Department of Metallurgy, Norwegian Institute of Technology, Alfred Getz vei 2b, N-7034 Trondheim. Norway; Merton C. Flemings, Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139-4037 The orientational relationships between particles in agglomerates of semisolid AI6.5wt%Si bave been investigated. The alloy has been sheared at a rate of 900 5'\ for two hours at a fraction solid of 0.2 and then quencbed. The orientational relationships between particles in the agglomerates have been investigated on polished samples from electron beam scattering patterns in a scanning electron microscope. The results show that all particles belonging to the same agglomerate are related either by low angle grain boundaries with misorientations of less than 10 or by misorientations of about 60 around the <111> axes which corresponds to coherent twins. All particles in the microstructure can be identified as belonging to an agglomerate. No preferred orientational relationship between different agglomerates has been observed. 10:40 am SEMISOLID PROCESSING OF SPRAY CAST ALLOYS: EXPERIMENTS AND MODELING: E. Tzimas, Dr. Antonios Zavaliangos, Assistant Professor, A. Lawley, Department of Materials Engineering, Drexel University, 32nd and Chestnut Streets, Philadelphia, PA 19104 Metallic systems which possess a fine, equiaxed, non-dendritic microstructore behave thixotropically and can be formed to net shape in the semisolid state. Semisolid processing provides several advantages over conventional casting, forging, and PIM techniques, but has not been accepted widely until recently. A combined theoreticalnumerical-experimental program is under way at Drexel University to assess quantitatively the potential of thixoforming, as applied to spray cast materials. Here we present experimental results on the flow/deformation behavior of semisolid spray cast binary Al-4.5w/oCu and 2014 aluminum alloys at different temperatures (590-64OC) and strain rates (O.I-IO/sec) under simple unconstrained compression. The resistance of both alloys to flow increases initially with strain. During deformation the solid skeleton that is formed initially by the solid grains appears to be destroyed. This results in a significant decrease of the flow resistance which finally reaches steady state. The steady state stress in the binary alloy is significantly lower than in 2014. This difference is attributed to the presence of unmelted secondary phases in 2014. Similarities with the behavior of low cohesion fully saturated soils are noted and utilized establishing a modeling framework for semisolid processing. 11:00 am THREE DIMENSIONAL FLOW OF THIXOTROPIC MATERIALS IN A CHANNEL WITH SUDDEN EXPANSION USING AN IMPLICIT SEGREGATED FINITE ELEMENT SOLUTION: Dr. Andreas N. Aluandrou, Associate Professor, Director, Aerospace Program. Department of Mechanical Engineering, Worcester Polytechnic Institute, 100 Institute Road, Worcester, MA 01609 The importance of processing of metal and metal matrix composites in semi-solid state is increasing quite rapidly. This manufactoring approach offers distinct advantages over similar methods of near-net-shape making, such as better control of the microstructore and thus mechanical properties and better control of the processing conditions due to the "fluid"-like flow of the material. The effective use of this process presumes good knowledge of the material behavior. Unfortunately, the constitutive behavior is quite complex and depends not only on the volume solid fraction but also on the processing conditions. Currently, the only mathematically sound theory attempting to describe their rheology is based on an internal variable approach. In this paper we are taking an alternative phenomenological approach where the behavior of semisolids is expressed through a continuous shear thinning-thickening Bingham fluid constitutive relation whose construction relies on the generalization of available experimental data. This constitutive relation is used to study the flow in a three-dimensional channel with a sudden expansion. Results will be presented for different geometries and flow parameters. 11:20 am PANEL DISCUSSION: SEMI-SOLID PROCESSING TECHNOLOGIES, CHALLENGES AND FUTURE PROSPECTS: Moderator: Prof. A. Alexandrou The contributing authors and audience will be invited to join in a panel discussion on the future of Semi-solid Processing in light of recent developments in the field.

EXPERIMENTAL METHODS FOR MICROGRAVITY MATERIALS SCIENCE: SESSION II

Sponsored by: MSD Thermodynamic and Pbase EquIlbrla Committee, NASA Microgravlty ScIences

Program Organizers: Dr.Robert A.Schiffman, R.S. Research Inc., Barton, VT 05822;

Professor Carlo Patuelli, Universita di Bologna, 1-40126 Bologna, Italy Tuesday,AM February 6, 1996

Room: Orange County 2 Location: Anaheim Marriott Hotel

Session Chairperson: Dr. R.W. Smith, Department of Materials and Metallurgical Engineering, Queen's University, Kingston, Ontario, Canada, K7L 3N6

10:10 am BREAK

1996 TMS ANNUAL MEETING AND EXHIBITION

B45

8:30am COMPOSITION DEPENDENCE OF THE SURFACE TENSION OF VANADIUM-TITANIUM ALLOYS NEAR LIQUIDUS TEMPERATURE: S. Bannerjee, A.B. Hmelo, T.G. Wang, Department of Applied and Engineering Sciences, Box 1593-B, Vanderbilt University, Nashville, TN 37235 We report the surface tension of V-TI alloys as a function of temperature and composition. We've gathered high resolution data using the OSCillating drop technique by simultaneously viewing the drop from three orthogonal directions, over a temperature range of 150-200 C. above the liquidus down to the undercooled regime. The composition dependence of the surface tension of V-TI deviates from ideality. The d gamma!dT values of the alloys imply some surface adsorption effects at and below the melting temperature, as corroborated by surface composition calculations based on the d gamma! dC values we've obtained. There is clear evidence of clustering of like-atoms in the region of the congruent melting composition. Both surface tension and undercooling data support this conclusion. As expected, the clustering effect is dominant at temperatures below the liquidUS. The surface composition becomes approximately equal to the bulk composition at superheated temperatures. The congruent melting composition appears to be V-60 wt.%TI rather than at V-68 wt.%TI as has been reported earlier. 8:50am DYNAMIC REDOX REACTIONS AND THE FORMATION OF FINEGRAINED, POLY CRYSTALLINE OXiDElSILICATE CERAMICS FROM INVISCID MELTS: Reid F. Cooper, Department of Materials Science and Engineering, University of Wisconsin, Madison, WI 53706 The dynamics of oxidation or reduction reactions in multicomponent, transitionmetal-cation-bearing oxides involves the coupled diffusion of electron holes (polarons) and component cations. As a consequence, intema1 reactions, resulting in the nucleation/crystallization at a reaction front of higher-order oxides or the more-noble metal component, respectively, dominate dynamic behavior. In the amorphous state, such reactions result in homogeneous nucleation of this product phase, suggesting an approach to the preparation of fine-, uniform-grained ceramics directly from inviscid melts: the redox-reaction-front metal or oxide phase can act as a dispersed heterogeneity for nucleation of other, "majority" phases. This approach, for which microgravity containerless processing is ultimately required, will be illustrated with results from Fe2+ bearing aluminosilicate melts and glasses. For example, oxidation of the melt at 1400 C. results in "isothermal undercooling": the liquidus temperature and primary phase are a function of Activity of 02 and thus less than or equal to I micron magnetite (Fe304) forms at the reaction front. 9:10am PARTICLE REDISTRffiUTION DURING REMELTING OF AIISICp COMPOSITES AN EXPERIMENTAL AND COMPUTATIONAL STUDY: BX. Dhintiaw, P. Ganguly, F. Juretzko, D.M. Stefanescu, Department of Metallurgical and Materials Engineering, The University of Alabama, Box 870202, Tuscaloosa, AL 35487-0202 Low gravity processing of particulate metal matrix composite samples includes two distinct phenomena: (a) the remelting of the originally prepared sample when the fluid flow within the melt plays a dominant role in the particle redistribution and (b) the solidification stage when the pushing of the particles by the solidification front is the determining factor. The effects of these two sub-processes on the ceramic phase distribution in the final composite microstructure are distinctly and relatively independent of each other. The present work aims at the analysis of the effects of the melt fluid flow occurring during the melting of the cylindrical composites specimens and comparison of predictions with experiments. The gravity driven buoyancy and the surface tension driven Marangoni flows have been assumed to be the dominant flows. The model analyzes the flow patterns at three gravity levels viz., Ig, O.lg and O.Olg. The model has been validated using Al-2I3% SiCp composites specimens. 9:30am GRAVITATIONAL EFFECTS IN LIQUID PHASE SINTERING: R.M. Gennan, 118 Research West, Penn State University, University Park, PA 16802-6809 Liquid phase sintering is applied to high solid contents, since rigidity is needed to retain compact shape during sintering. Experiments have been performed under microgravity to determine the difference in microstructures due to gravitational compression of contacting grains. The conditions present during microgravity processing allow liquid phase sintering over a wider range of liquid-solid ratios. A widely held assumption is that solid grains will remain dispersed in the absence of gravitational forces. This paper shows the action of a weak agglomeration force couples with Brownian motion causes solid grain agglomeration, but that the rate of grain coarsening is actually slower under microgravity conditions for all solid contents. One consequence of this work is a clear demonstration of a coalescence contribution to grain coarsening during liquid phase sintering. Other unique aspects of the microgravity experiments will be described, including shape distortion and pore growth. 9:50 am BREAK 10:10 am SELF-PROPAGATING WGH-TEMPERATURE SYNTHESIS OF POROUS CERAMICS IN MICROGRAVITY: T. Manor, J. Moore, Colorado Center for Advanced Ceramics, Colorado School of Mines, Golden, CO 80401 Porosity formation and distribution in boron carbide - alumina foamed ceramics is studied in a microgravity environment. Porous boron carbide - alumina is created through a self-propagating high-temperature synthesis (SHS) process in an inert environment. The variables under investigation are: reaction chamber pressure, additions of excess B203-Si02 mixtures, additions of excess aluminum, form of carbon (graphite or lampblack), green pellet diameter, acoustic levitation level, and gravitation level. Each variable is studied independently under a variety of reaction chamber pressures. Poros-

B46

ity and distribution is established as a fun,;tion of the above variables, and the underlying mechanism is explained. 10:30 am CONVECTIVE FORCE EFFECTS ON THE CRITICAL VELOCITY OF ENGULFMENT OF PARTICLES BY A PLANAR SOLIDILIQUID INTERFACE: S. Sen, USRA, MSFC, Huntsville, AL 35812. H. Pang, D.M. Stefanescu, B.K. Dhindaw, Solidification Laboratory., The University of Alabama, Tuscaloosa, AL 35487. P.A. Curreri, NASA, MSFC, Huntsville, AL 35812 Directional solidification experiments were performed on mixture of transparent succinonitrile and spherical polystyrene particles. Ground based experiments performed using a thermal gradient stage indicated a decrease in critical velocity of engnlfment with increase in particle size between 1-10 microns. Further, for a given particle size there was a variation in critical velocity with variations in experimental cell thickness. Results from low-gravity experiments performed during aircraft parabolic maneuvers will be discussed. The effect of convective forces on the critical velocity for particle engulfment will be quantified with respect to several theoretical models describing particle pushing and engulfment. 10:50 am CRYSTALLIZATION STUDIES ON HEAVY METAL FLUORIDE GLASSES USING PARABOLIC FLIGHT AIRCRAFfS AND SOUNDING ROCKETS: S. Varma, S.E. Prasad, Sensor Technology Limited, B.M. Hi-Tech Div., P.O. Box 97, Collingwood, Ontario, Canada L9Y 3Z4; A. Ahmad, T.A. Wheat, MTI)CANMET, Natural Resources Canada, 405 Rochester Street, Ottawa, Ontario, Canada KIA 001 The predicted low losses in heavy lII!~tal fluoride glass (HMP) fibres have not yet been realized in spite of over 20 years of world wide research. The main reason is believed to be excess scattering due to homogeneous and heterogeneous nucleation during glass synthesis and fibre drawing. The primary reason for heterogeneous nucleation is container contamination, while gravity driven density segregation could trigger homogeneous nucleation. Although containerless glass synthesis on the ground could minimize heterogeneous nucleation, homogeneous nucleation could still cause excess scattering. Heat treatment of these glasses for fibre drawing can cause further nucleation and optical degradation due to density segregation. A microgravity environment could, therefore, reduce homogeneous nucleation and optical degradation during fibre drawing. This was examined with the help of ground based experiments, and microgravity experiments aboard the T-33 aircraft and the CSAR-I and CSAR-U sounding rockets. The results of these experiDII:nts, which indicate that microgravity helps in reducing crystallization and optical degradation in these glasses during their heat treatment for fibre drawing, are presented in this paper. 11:10 am EFFECTS OF COMPOSITION ON THE PORE BEHAVIORS IN MICROGRAVITY PROCESSED FE..CU COMPACTS: Z Xue, A.K. Kuruvilla, J.E. Smith, Jr., The Department of Chemical and Materials Engineering, The University of Alabama in Huntsville, Huntsville, AL 35899 Several liquid phase sintered samples from the Fe-Cu, Co-Cu and W-Ni-Cu systems have been processed aboard STS-57, STS-60 and STS-63. ProceSSing times range from 75 seconds to over 66 minutes for samples with aspect ratios of 0.5 to 1.0. Samples processed under 12 psia of argon did not exhibit slumping but did show shape distortion as a function of the amount of liquid present during the melt. Extensive pore formation was observed for the microgra'lity processed samples that underwent either pore metamorphosis to smaller pores or coalescence to large pores as a result of instabilities within the compacts. The particular type pore behavior was found to depend on the amount of liquid present during the melt. Microgravity results from Fe-Cu system with the Cu concentration of 33, 43 and ~;3 weight percent will be presented to show the various behaviors. The results demonstrate the use of microgravity to study pore metamorphosis within systems that undergo large scale sedimentation effects during unit gravity processing.

GAS INTERACTIONS IN NONFERROUS METALS PROCESSING I

Sponsored by: SMD Non-Ferrous Metals Committe, EPD Pyrometallurgy and Process Fundamentals Committees Program Organizers: Debabrata Saha, Metals Industries Group, Air Products and Chemicals, Inc., Allentown, PA 18195; Dr. William E. Frazier, Naval Warfare Center, Warminster, PA 18974; William P. Imrie, Bechtel Group, San Francisco, CA 94105; Prof. David G. Robertson, Department of Metallurgical Engineering University of Missouri, Rolla, MO 65401 Tuesday, AM February 6, 1996

Room: B8 Location: Anaheim Convention Center

Session Chairpersons: Debabrata Saba, Metals Industries Group, Air Products and Chemicals, Inc., Allentown, PA 18195; David G. Robertson, Department of Metallurgical Engineering, University of Missouri, Rolla, MO 6540 I

8:30am GAS-METAL REACTIONS DURING DEBINDING AND SINTERING OF INJECTION MOLDED NONFERROUS METALS: Randall M. Gennan, R.

1996 TMS ANNUAL MEETING AND EXHIBITION

Tandon, A. Griffo, Penn State University, University Park, PA 16802 Injection molding of mixed elemental powders is the basis for forming sintered cop' per, steel, titanium, cobalt alloys, tungsten heavy alloys and various heat sink materials, including W-Cu and Mo-Cu. An organic binder (wax-polymer) is used to form a powder slurry that allows injection molding. However, during heating to the sintering temperature it is necessary to extract all of the binder. This involves careful control of the process pathway, including temperature, time, heating rate, and process atmosphere. For various powders it is possible to employ highly reducing atmospheres, but for other systems the processing is difficult because of undesirable thermochemical reactions, example hydride formation. Research results are presented on several material systems and the thermal processing options deemed necessary for full binder removal prior to reaction, pore closure, or system densification. Several examples of late stage pore generation via impurity reactions are used as illustrations of the difficulties in effective debinding. 8:50am EFFECT OF ATMOSPHERE COMPOSITION ON SINTERING OF BRONZE COMPONENTS: D.Garg, K.R. Berger, OJ. Bowe, J.G. Marsden, Air Products and Chemicals, Inc., Allentown, PA 18195 Premixed bronze powders are widely used to produce a wide variety of PIM components including bearings and bushings by the powder metal industry. The physical and mechanical properties of sintered bronze components such as dimensions, strength, microstructural characteristics, and sintered density are well known to change with sintering furnace design, particle size distribution in powders, green density, preheating and sintering time and temperature, and protective atrilosphere flow rate. However, not much is known about the effect of protective atmosphere composition on physical and mechanical properties of sintered bronze components. This paper describes effect of the protective atmosphere composition on physical and mechanical properties of sintered bronze components. 9:10am INTERACTIONS BETWEEN CHEMICAL REACTION, PORE DIFFUSION AND HEAT GENERATION DURING THE ZINC SULFIDE EXOTHERMIC OXIDATION: F. Palisson, M. Galant Francois, D. Ablitzer, LSG2M Ecole des Mines, Parc de Saurupt,54042 Nancy Cedex, France During gas-solid exothermic reactions, heat generation makes temperature rise and modifies the rate of the various mass transfer processes, namely the chemical reaction itself, the pore diffusion and the external transfer. We have investigated this thermal effect during the oxidation of zinc sulfide porous pellets in a thermobalance furnace, between 550 and 900°C and under O,-N, atmosphere. The results have heen analyzed by using, flfstIy, the law of additive reaction times derived from the isothermal Grainy Pellet Model, and, secondly, a new numerical model taking into account non-isothermal conditions. Both models can satisfactorily simulate conversion versus time, but the numerical model also depicts the temperature evolution within the pellet. Finally, it is shown that owing to the heat generation the zinc sulfide oxidation takes place under diffusion control (3/4 internal, 114 external) in all the cases studied. During the course of reaction, the pellet undergoes strong temperature variations, with thermal gradients confined to a shell oxide layer. 9:30am MECHANISMS OF FORMATION OF HYDROGEN POROSITY IN 7XSO AND 2X24 ALUMINUM ALLOYS; EFFECTS ON MECHANICAL BEHAVIOR: Femand Marquis, Dept. of Metallurgical Engineering, South Dakota School of Mines and Technology, Rapid City, SO 57701-3369 In this investigation, we consider the mechanisms of formation of two types of p0rosity: primary and secondary, their characterization and their effects on fracture, fatigue and failure behavior of these alloys. The mechanisms of formation of three forms of primary or cast porosity: (a) macroscopic shrinkage porosity (MSP); (b) microscopic interdendritic porosity (MIP); and (c) hydrogen gas porosity (HGP) or primary hydrogen porosity ( PUP) are discussed and each of these forms of porosity is characterized. The mechanisms of formation of four forms of secondary or thermallthermomechanical porosity: (a) high temperature oxidation (HTO), internal hydrogen precipitation (IUP), or secondary hydrogen precipitation (SUP); (b) preheating blistering porosity (PBP); (c) eutectic melting porosity (EMP); and, (d) dissolution microvoid formation (DMF) are discussed and each of these forms of porosity is characterized. The evolution of the primary and secondary types of porosity (with special emphasis on PUP and IHP) during cotumercial and laboratory processing and their effects on the fracture and failure behavior of both types of alloys, with special focus on fatigue crack initiation, was investigated. Methodologies for the decrease of the total porosity: volume fraction and size, its redistribution and the minimization of its effects on the mechanical behavior of both types of alloys are discussed. 9:50 am BREAK 10:00 am THE FLAMMABILITY OF ALLUMINUM IN OXYGEN: D. Bruce Wilson, Quantos Consulting, Mesilla Park, NM 88047; Joel M. Stoltzfus, NASAIWhite Sands Test Facility, Las Cruces, NM 88004-9544 The reaction of oxygen and solid aluminum produces the well-known surface coating of aluminum oxide, A I ,01 Thermodynamic properties and kinetic mechanism of this reaction is extensively documented. Less well-defined are the corresponding liquid phase reaction and reactions which would be characterized as "burning" of the aluminum. This paper reviews the solid aluminum-oxygen reaction as a foundation for discussing the flammability of aluminum in oxygen enriched atmospheres. Appropriate

1996 TMS ANNUAL MEETING AND EXHIBITION

thermodynamic analysis and kinetic reaction mechanisms are developed. Available models of aluminum burning are reviewed and evaluated. The appropriate information for design of engineering applications of aluminum in oxygen atmospheres is summarized. 10:20 am EFFECT OF ATMOSPHERE COMPOSITION ON HOMOGENIZING Al-L1 AND Ll-Mg ALLOYS: Z. Zurecki, Air Products and Chemicals, Inc., Allentown, PA 18195-1501 The selection of atmosphere composition is important for controlling surface defects of ingots while homogenizing reactive aluminum alloys. An improper selection of annealing atmosphere can result in surface blistering in aluminum-magnesium and subsurface depletion oflithium in aluminum-lithium alloys. To determine atmosphere compositions required for homogenizing aluminum-magnesium and aluminum-lithium alloys with acceptable surface quality, a number of exploratory experiments were carried out in nitrogen and carbon dioxide based atmospheres including nitrogen or carbon dioxide and mixed with reactive impurities or additives such as oxygen, moisture, and sulfur hexafluoride. This paper will describe the influence of these atmospheres on the surface quality of aluminum-magnesium and aluminum-lithium alloy ingots. 10:40 am THE EFFECTS OF OXYGEN CONCENTRATION ON THE SURFACE QUALITY AND OTHER PROCESSING VARIABLES FOR ANNEALING ALUMINUM AND ITS ALLOYS: G.R. White, T. Philips, H.S. Nayer, BOC Gases, Murray Hill, NJ 07974 Traditional atmospheres used for aluminum annealing have heen air, exothermic gas, and cryogenic purity nitrogen. Air and exothermic gas have several problems associated with them such as surface discoloration, toxic emissions, and explosion hazards. High oxygen levels can cause magnesium streaking and oil staining, and therefore, pure nitrogen has heen the atmosphere of choice for high quality aluminum coils. A comparative study of the surface quality of key aluminum alloys annealed in nitrogen with different amounts of oxygen content was made. The atmosphere composition inside the furnace was analyzed during the purge and heat cycle to determine its effects on purge times and exhaust emissions. The aluminum coils after annealing were visually inspected for surface staining, followed by lab analysis of the surface oxides. The study revealed that a small concentration of oxygen is acceptable, and did not adversely effect the surface quality or any of the other processing variables. 11:00 am CHEMICAL REACTION AND HYDRODYNAMIC SCIENCE ISSUES FOR GAS-NONFERROUS LIQUID METAL REACTIONS: R. MUlharasan, MJ. Koczak, S. Kalidindi, Drexel University, Dept. of Materials Engineering, Philadelphia, PA 19104 The phenomena of gas-molten metal reactions for the production of metal matrix composites consist typically of a combination of one or more gas transport steps and a liquid gas reaction step. Reaction products include such strength enhancers, grain refiners, or high thermal conductivity phases as TIC, AIN, SiC, Si,N., TIN in aluminum alloys. Nitrides, carbides and oxides can be produced at various volume fractions and in fairly fine sizes (500 nm to 5 um) in laboratory melts by Reactive Gas Injection method (RGI). Gas injection methodology to produce the reinforcements is attractive because it can potentially be scaled to production scales in excess of 10,000 kglh. For continuous production, the reactant molten metal can be continuously fed into the crucible where the reactive gas is injected. The molten metal phase reactant may also be involved in a controlling transport step, depending on its concentration and Viscosity. It is well known that the reaction product reinforcement size and shape of these phases have significant impact on the physical, mechanical and thermal properties. The current state of knowledge is unable to predict the kinetics and growth of solid reaction products obtained. The primary goal is tu develop a fundamental framework for the understanding and prediction of the particle size in such reaction systems. In addition, the role of gas transport, liquid hydrodynamics and decomposition are important issues. A reaction-transport model is being developed using a reaction, transport, and thermodynamic framework, both at macro and at atomistic level to assess gas - liquid interactions. 11:20 am VALIDITY OF THE KINETIC LANGMUIR'S LAWS FOR THE VOLATIZATION OF METALLIC ELEMENTS IN VACUUM METALLURGY: A NUMERICAL APPROACH: J. P. Bellol, H. Duval, D. Ablitzer, Ecole des Mines, Parc de Saurupt, 54042 Nancy Cedex, France During the non-ferrous metal processing, the melting step under vacuum or under rarefied gas leads to volatile metal losses. It is the case of the metallurgical processing of the nickel-based, titanium-based or zirconium-based alloys where the evaporation losses don't allow a sufficient control of the alloy chemical composition. A theoretical approach has been carried out involving both a continuum description in the liquid phase and a molecular approach in the rarefied gas phase. The finite volume method is used in order to solve the coupled transport equations in the liquid, whereas the Boltzmann equation is modelled by using a Monte Carlo method. At the liquid metal and metallic vapor interface, the volatilization and recondensation fluxes link up the two calculations. Simulation results are given for the melting and refining of reactive metals such as titanium alloys where the control of the chemical composition remains a major difficulty. The validity of the kinetic Langmuir's law is discussed for different operating conditions.

B47

A SYMPOSIUM DEDICATED TO PROFESSOR .JOHANNES WEERTMAN'S SEVENTIETH BIRTHDAY: SESSION III: Dynamic Plasticity

Sponsorrd by: JL SMDlMSD Nuclear Materials and MSD Flow and Fracture Committees and FEMS (Federation 01 European Materials SocIeties) Program Organizers: RJ. Arsenault, Department of Materials Science and Nuclear Engineering. University of Maryland. College Park. MD 20742-2115; David Cole. CRREL. 72 Lyme Rd.• Hanover. NH 03755; Todd Gross. Department of Mecbanical Engineering. University of New Hampshire. Durham, NH 03824; Gemot Kostorz. Institut fUr Angewandte Physik. ETH Hilnggerberg. CH-8093 Zllrich. Switzerland; Peter Liaw. Department of Materials Science and Engineering. University of Tennessee. Knoxville. TN 37996-2200; Sivan Parameswaran. NRC-Institute for Aerospace Research. Ottawa. Canada KIA OR6; Howard Siuk, Inco Alloys International Inc .• Huntington. WV 25705-1771 Tuesday. AM February 6. 1996

Room: Orange County 3 Location: Anaheim Marriott Hotel

Session Chairpersons: S. Parameswaran. NRC-Institute for Aerospace Research. Ottawa. Canada KIA OR6; H. Ishii. Department of Mechanical Engineering. Shizuoka University. Hamamatsu City. Japan

8:30 am Invited DIPOLE ORIENTATION JUNCTIONS AND BORDONI RELAXATION: J.J. Gilman. Department of Materials Science and Engineering. University of California at Los Angeles. Los Angeles. CA 90024 In 1949. Bonloni reported internal friction peaks in cold-worked Ag. AI. Cu. and Pb at temperatures equal to roughly one-third of their Debye temperatures; and frequencies of 1040 kHz. These peaks have since been observed in several other metals. and have been Interpreted in terms of the nucleation of dislocation kink-pairs lying in a Peierls-Nabarro crystal potential well. An altemative interpretation is based on the recognition that dislocation dipoles have two orientations relative to a plane that is pelpCndicular to the glide plane and which contains the axis of the dipole. Furthermore. along a given dipole the orientation can change creating a dipole orientation junction (DOJ) between the two differently oriented pieces of dipole. The DOJs can hop from one atomic plane to another. In the presence of an applied shear stress this causes anelastic relaxation. The activation barrier that resists this is small; hence the low temperature of the relaxation peak. There are two kinds of dipole lines (a vacancy type. and an interstitial type); hence the two peaks. The height of the activation barrier for DOl motion is determined approximately by the stacking fault energy in the specimen metal. Measured parameters confirm this. 9:00 am Invited

DEFECT GENERATION UNDER SHOCK-WAVE DEFORMATION: M.A. Meyers. Department of AMES. University of California. San Diego. San Diego. CA 92093; G. Ravichandran. California Institute of Technology. Pasadena, CA; Y.F. Nesterenko. Lavrentiev Institute of Hydrodynamics. Novosibirsk, Russia Different models for the generation of point, line. and interfacial defects onder shockwave deformation are reviewed with emphasis on the pioneering work ofWeertman on high-velocity dislocations and plastic-wave and shock-wave dislocation-based mechanisms. Recent results of predictions of the decay of the deviatoric component of stress hehind the shock front are presented. A constitutive description of the slip-twioning transition is proposed. The possible Influence of non-linear. non-equilibrium temperature fluctuations on the process of defect generation is analyzed. Research supported by the U.S. Army Research Office ContractARO DAAL03-92-G-OI08 and NSF Grant DMR 9396132. 9:30 am Invited MlCROMECHANISMS OF THE DISLOCATION MOTION IN CRYSTALS WITH DEEP PEIERLS RELIEF: V.I. Nilcitenko.lnstitute of Solid State Physics. Russian Academy of Sciences. 142432 Chemogolovka, Russia Problems of the influence of point defects on dislocation mobility are analyzed. The two-level intermittent loading technique is applied to study kink motion along a dislocation line in Oe. Si. and SiOe crystals that are characterized by different ratio of the influence of point defect barriers and Peieds potential on the dislocation motion. The experimental data are analyzed in the framework of a model. considering the joint interaction of point defects with a dislocation. It is shown that Cottrell atmospbere not only determines the barriers for the kink motion but stimulates the kink pair return to formation centers as well as the specific mode of kink drift along the dislocation line due to the motion in the field of random forces. It is determined by step-like changes of a dislocation energy with attachment of a point defect to the dislocation core oc its detachment caused by kink motion and can lead to the nonlinearity of the kink drift or even to localization of kinks. 10:00 am Invited ON THE ROLE OF DISLOCATIONS IN HEAVILY STRAINED MATERI· ALS: J.Th.M.De Hosson. Department of Applied Physics. Materials Science Center. University of Groningen. Nijenborgh 4. 9747 AG Groningen. The Netherlands. Otrnar Kanert. Department of Physics. University of Dortmund. 0-44221 Dortmund.

B48

Germany As a tribute to professor Hans Weertman upon the occasion of his 70th birthday this paper deals with the role of dislocations in materials deformed at high strain rates. In contrast to classical crystal deformation experiments which occur typically at strain rates of to"s" to 100's·'. ballistic deformation experiments involve strain rates of 100s" to IO"s". These high strain rates yield completely different deformation mechanisms. This paper describes TEM and HREM study of the dynamic compacted samples of high 11: superconducting ceramic msterials such as YBa,Co, at various ElM values. i.e. the ratio between the mass density of tlH~ explosives and the mass density of the material to be compacted. The ElM values for these samples increases from ElM - 0.7 to FJ M=2.1. Higher ElM values correspond to longer pulse duration and slightly higher peak pressures. Apart from the well estai.lished <100>( lOO} glide system. the role of a novel (110)(10) and (010)(100) glide systems is studied. All glide systems are found to interact with the ferroelastic domains of the material. each in a different way. Possible advantageous effects of the novel dislo<:ations on the superconducting properties are also considered. A comparison is made with in-situ NMR experiments on the deformation-induced generation of point defects in ceramic materials and metals. 10:30 am BREAK 10:40 am ON THE INCREASED STRAIN RATE SENSmVlTY OF THE FLOW STRESS OF COPPER AT mGH STRAIN RATES: H.D. Chandler. University of The Witwatersrand. Johannesburg. Private Bag 3. WITS 2050. South Africa Experimental work on the mechanical behavior of fcc metals indicates that when the strain rate increases to about 100sec" there is an apparent sharp increase inn the flow stress (measured at constant strain) with strain rate. This has been interpreted in terms of a change in deformation mechanism. e.g. as a change from dislocation glide control to a dislocation drsg mechanism. However. much of the recent work on the subject has explained this change in tenns of dislocation glide accompanied by an increase in the rate of dislocation production. A similar approach using a glide rate equation together with a structure evolution description is adopted for the present work. However. instead of the dislocation increase heing modelled by the evolution equation. it is regarded as initially heing forced by the clictates of the glide equation. This leads to a simpler description of the effect requiring fewer empirical values. There is good correlation between modelled behavior and available experimental results on copper. 11:00 am mGH STRAIN RATE DEFORMATION MECHANISMS IN ALPHA TITANIUM: D.R. Chichili. K.T. Ramesh. K.J. Heroker. Department of Mechanical Engineering. The John Hopkins University. Baltimore. MD 21218 The present work addresses the deformation mechanisms that control the high strainrate mechanical behavior of alpha titanium. The mechanical properties have been measured using a combination of quasistatic and dynamic Kolsky bar compression tests. and data for strain-rates ranging from(U),' to 100s") has been acquired. At the macr0scopic level. the material shows a strong strain hardening behavior for fixed strainrates and rate sensitivity of the flow stress for fixed strains. Kolsky bar specimens have heen recovered after being subjected to a single known stress pulse. and the deformed specimens have been investigated using both optical and transmission electron microscopy (TEM). This recovery technique allows us to relate the observed microstructural mechanisms with the measured mechanical properties. Macrotwins are evidenced in the optical micrographs. TEM is employed to charscterize these deformation twins and to identify the nature and density of dislocations that are present after deformation at different strain-rates. To date. the density of dislocations and the number of active twin planes has been found to increase with the strain-rate. These studies are ongoing. and a more detailed picture of the relationship between the macroscopic behavior and the microstructural mechanisms will also he discussed. 11:20 am

CONCERNING THE EVALUATION OF THE INTERNAL STRESS IN DISLOCATION DYNAMICS: C.Y. Iswaran. R.E. Reed-Hill. MJ. Kaufman. Department of Material Sciences and EIllgineering. University of Florida. Gainesville. FL32611 The measured flow stress during pla!:tic deformation. in the absence of any strain aging phenomena. consists of two parts: a thermally activated effective stress. (J'. and an athermal internal stress. (Js' On the CIne hand. as far as theories of dislocation dynamics are concerned. (J' is by far mo~e interesting. and a vast amount of literature exists on the subject of the thermally activated movement of dislocation line segments through barriers posed by obstacles such as point defects to such motion. On the other hand. the experiments measure the total tlow stress. (Jpo which includes (Js' In addition. we have also shown. that for a number of polycrystalline materials. data in the literature conform to the power law. providing the effective stress is evaluated by subtracting the correct value of the internal stress from !:he flow stress. 11:40 am DISLOCATION MECHANISM OF SHOCK-INDUCED FRACTURE AND MARTENSITE FORMATION IN AUSTENITIC STEEL: E. Zarrtsky. Pearlston Center for Aeronautical Engineering Studies. Department of Mechanical Engineering. Ben-Gurion University of the Nege',. P.O.B. 653. Beer-Sheva. 84105. Israel An interferometric (VISAR) technique was used to measure Hugoniot Elastic Limit, dynamic viscosity and spall strength of austenitic Fe-Cr-Ni steel at temperatures of 173 - 293 K. The samples were impacted by 3-mm steel projectiles with velocities of 100350 mlsec. The recovered samples were studied metallographlcally. The material behavior during shock-induced martensitic transformation at temperatures of 173 -240 K was found to be sirni1ar to the dynamic response of the material near its dynamic fracture threshold at 293 K. The suggests the same governing mechanism in both phenom-

1996 TMS ANNUAL MEETING AND EXHIBITION

ena. A model capable of explaining the phenomena is suggested. It is based on the motion and multiplication of Shockley-type partial dislocations in presence of shockinduced shear stress, resulting in the development of a lens-like colony of stacking faults. At room temperature the perfect dislocations interact with the faults of the colony, transforming it into the colony of Frank loops. Under a tensile stress this transformation results in the nucleation of vacancy discs which coalesce into voids. At lower temperatures the tensile stress favors the transformation of the stacking fault lens into a lenticular martensite crystal.

HIGH TEMPERATURE COATINGS II: Intermetallic/Ceramic Coatings I

Sponsored by: MDMD Surface Modlncatlon and Coatings Committee Program Organizers: Narendra B. Dahotre, Center for Laser Applications, University of Tennessee Space Institute, Tullahoma, TN 37388; Janet M. Harnpikian, School of Materials Science & Engineering, Georgia Institute of Technology, Atlanta, GA 30332; Jacob J. Stiglich, PO Box 206, Sierra Madre, CA 91025 Tuesday, AM February 6, 1996

Room: BI Location: Anaheim Convention Center

Session Chairpersons: Peter F. Tortorelli, Metals and Ceramics Division, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831-6156; Janet M. Hampikian, School of Materials Science & Engineering, Georgia Institute of Technology, Atlanta, GA30332

8:30 am Keynote Presentation THE FORMATION OF PROTECTIVE SCALES DURING IUGH-TEMPERATURE OXIDATION: F.H. Slott, Corrosion and Protection Centre, UMIST, PO Box 88, Sackville Street, Manchester MOO IQD, UK Many high-temperature nickel-or iron-base alloys rely on the formation of slowgrowing alumina-rich scales for protection against environmental degradation. Although A~O, scales can be developed on binary Ni-A1 or Fe-AI alloys, the presence of a third element, usually chromium, promotes their establishment on alloys oflower aluminium concentration. Following establishment of the scale, the subsequent growth kinetics are controlled mainly by transport of reactants along short-circuit paths in the scale. Despite their slow growth rates, mecbanical failure of such scales can cause enhanced rates of metal loss. This failure occurs when the strain arising from residual growth stresses and thermal-cycling stresses exceeds a critical value. However, the addition of small amounts of more reactive elements, such as yttrium, to the alloy can give a considerable improvement in mechanical integrity of the scale, resulting in a much improved overall oxidation performance. 8:55am EVALUATION OF TBC-COATED boN1A1 SUBSTRATES WITHOUT A BOND COAT: B.A. Pint, Oak Ridge National Laboratory, PO Box 2008, MS 6156, Oak Ridge, TN 37831-6156; B. Nagaraj, General Electric Aircraft Engines, One Neumann Way, MD H85, Cincinnati, OH 45215-6301 bo NiAI substrates with various alloy additions and oxide dispersions were coated with Y,O,-stabilized Zi'O, using electron beam-physical vapor deposition (EB-PVD). Cyclic oxidation experiments (I 000-1200-C) were conducted to study the effect of substrate dopants such as Y and 'h on the coating lifetime without the intermediate metallic bond coat layer used in conventional thermal barrier coatings. Oxidation kinetics and A~O, scale microstructures were compared to those for uncoated substrates. Degradation of the substrate-alumina interface by void formation was the primary mechanism leading to coating spallation. Coating lifetime was correlated to the alumina scale adhesion on the uncoated substrates. 9:15am HIGH-TEMPERATURE OXIDATIONISULFIDATION RESISTANCE OF IRON-ALUMINIDE COATINGS: P.F. Tortorelli, I.G. Wright, G.M. Goodwin, M. Howell, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831 Iron a1uminides show excellent corrosion resistance in oxidizing and sulfidizing environments at temperatures well above those at which these alloys have adequate mechanical strength. Therefore, it is anticipated that appropriate compositions of iron aluminides may find application as coatings or c1addings on materials which are less corrosion-resistant at high temperatures but have higher strengths. Toward this end, the behavior of weld-overlay iron-aluminide coatings on steel substrates and of co-extruded Fe,AUstainless steel tubing in elevated temperature air and H,S-H,-H,O mixed gases is being studied. Results based on weight cbanges and corrosion product chemistry and morphology have shown that the corrosion resistance of these layers can be essentially equivalent to that of wrought alloys with similar aluminum and cbromium concentrations when coating defectss are minimized by appropriat material selection and process control. The cbanges in corrosion behavior as a function of compositional modifications resulting from the coating process or from the choice of starting materials can be explained on the basis of what is known from previous high-temperature oxidation! sulfidation studies of bulk iron aluminides. Research sponsored by the U.S. Department of Energy, Fossil Energy AR&TD Materials Program under contract DE-AC0584OR21400 with Lockheed Martin Energy Systems.

1996 TMS ANNUAL MEETING AND EXHIBITION

9:35am THE EFFECT OF PLASMA SYNTHESIZED ALUMINA COATINGS ON THE OXIDATION BEHAVIOR OF IRON ALUMINIDES: P. Y. Hou, Lawrence Berkeley National Laboratory, I Cyclotron Rd., Berkeley, CA 94720; K.B. Alexander, PO Box 2008, Oak Ridge National Laboratory, Oak Ridge, TN 378316376; Zhi Wang, I. G. Brown, Lawrence Berkeley National Laboratory, I Cyclotron Rd., Berkeley, CA 94720 Plasma synthesized alumina films were deposited on two types of Fe,AI substrates with or without a small amount of 'h addition. After subsequent oxidation treatments at lOOO-C in air, the coating on the 'h-doped Fe,AI remained strongly adherent and suppressed the growth of the alumina scales from the alloy. The same coating on the undoped Fe3AI however, spalled off completely upon cooling to room temperature. These results suggested a strong substrate effect on coating adherence. Interface chemistry, studied by Auger Electron Spectroscopy, and coating microstructures, studied by TEM, are presented to explain the observed behavior. Results are discussed in terms of oxide adhesion properties in general. 9:55am PLATINUM ALUMINIDE DIFFUSION BARRIER FOR THE OXIDATION PROTECTION OF TITANIUM AND TITANIUM INTERMETALLIC ALLOYS: J.R. Nicholls, M.I. Deakin, Cranfield University, Cranfield, Bedford, UK; M. R. Wins/one, S. Kerry, Structural Materials Centre, DRA Pyestock, Famborough, Hampshire UK GUI4 OLS TItanium alloys find wide applications in gas turbine engines, with creep resistant alloys operating to over 6OO'C but the drive to reduce engine weight will lead to future engines using titanium aluminide alloys at temperatures between 700 and 9OO-C. This will only be possible with oxidation resistant coatings. This paper presents results of the development of a PtAI, intermetallic barrier coating for titanium and titanium aluminide alloys. A 3mm coating is manufactured by multilayer sputter deposition, coupled with high temperature reaction synthesis. Oxidation studies have shown that the PtA~ layer successfully protects titanium alloys against the ingress of oxygen for exposures up to 1000h at 700-C. Longterm exposure tests on aluminide alloys are in progress at 800"C. Under creep conditions of 350MPa at 650'C the coating extends the life of alpha-2 aluminide alloys by a factor of 2 by reducing environmentally assisted cracking. 10:15 am BREAK 10:15 am A STUDY OF THE MECHANISM OF ''PEST'' IN NIOBIUM ALUMINIDE: R.I. Hanrahan Jr., Los Alamos National Laboratory, TA-3, MS-G770, Los Alamos, NM 87545; M. Puga-Lambers, E. S. Lambers, University of Florida, MAIC, Materials Science and Engineering Department, 121 Rhines Hall, Gainesville, FL 32611; S.P. Withrow, Oak Ridge National Laboratory, PO Box 2000, Bldg. 3003, MS-6048 Oak Ridge, TN 37831 The "pest" phenomenon of accelerated oxidation and disintegration is exhibited by numerous intermetal\ic compounds exposed at intermediate temperatures. We have studied this phenomenon in NbAl, using a combination of ion implantation, SIMS, and AES. A continuous oxide layer is formed in the surface of one side of the specimen at room temperature using ion implantation. Using 018 as the implanted species a built in marker is provided for subsequent. SIMS analysiS. The implanted specimens were exposed at 7OO'C under Ar-5Oppm 0, to induce pest. The implanted surfaces survived largely intact while the rest of the specimen disintegrated. Comparison of SIMS profiles from implanted and non-implanted regions of the same specimen show that oxygen transport through the implanted layer was dramatically reduced. This result shows that the controlling mecbanism in pest of NbAI3 involves oxygen dissolution in the substrate. 10:45 am THE DIFFUSIONAL AND MICROSTRUCTURAL PHENOMENA IN TWO DIFFERENT SUPERALLOYS WITH A NICoCrAIYTa COATING UNDER CREEP AFTER DIFFERENT PRE-DAMAGE TREATMENTS: Alejandro Sanz, P; Bemadou, Laboratoire de Metallugrgie, 10 Avenue E. Belin, 31055 Toulouse cedex, France. L. Uanes, M. Anglada, M. Gomez, UPC ETSn, Dep de Ciencia de loc Materiales y Metalurgia, Av; Diagonal 647, 08028 Barcelona, Spain; M. B. Rapaccini, Polytecnico di Milano, Dipartimento de Ingenieria Aeroespaziale, Piazza Leonardo da Vinci, Milano, Italy NiCoCrAlYTh coatings are frequently used to protect the components in the hot path of the gas turbines engines. A careful selection of the coating which offers sufficient chemical compatibility with the substrate and a low interdiffusivity is necessary for these high temperature systems. The effect of different pre-damage treatments on the creep behavior and in the diffusional phenomena of single crystals, [(01) oriented, of the AM-3 and MC-2 superalloys coated with a Low Pressure Plasma Spray (LPPS) NiCoCrAlYTa coating is described. The pre-damage treatments comprise oxidation treatments at 1373 K (II OOC) during eight hours as well as creep treatments during the same period of time and at the same temperature under stress of 100 and 140 Mpa. After the pre-damage treatments, all samples were submitted to high temperature creep tests at 1323 K (105OC) and 140 Mpa until fracture. The phenomenalogical description of multicomponent diffusion by extending the Fick's law to a n-components system requires the use on n-I independent diffusion couples for the determination of (n-I') concentration-dependent interdiffusion coefficients at one composition. An altemative approach in the study of isothermal diffusion in multicomponent systems is to analyze the concentration profiles for interdiffusion fluxes of all components without invoking Pick's law. This approach allows to by-pass the need for (n'-I) interdiffusion coefficients and provides a direct insight into the direction as welI as magnitudes of tbe fluxes at any section within the diffusion zone. Using the data obtained from Energy

B49

Dispersive microbrobe Spectroscopy (EDS) the diffusional phenomena between the coating and the substrate is characterized and correlated with the chemical nature of the substrate and to the different pre-damage treatments. 11:05 am SILICIDE COATINGS FOR NIOBIUM ALLOYS: M. ViIasi H. Brequel, R. Podor, J. Steinmetz, Laboratoire de Chimie du Solide Mineral, URA CNRS 158, Faculte des Sciences. BP 239-54506 Vandoeuvre Les Nancy Cedex - France A pack cementation process for the codeposition of Si, Cr, Fe and TI has been adapted to pure niobium and niobium alloys. Through choice of a masteralloy rather than pure elements, and an adequate activator, it is possible to inexpensively produce complex silicide coatings in a single step which offer significant promise to protect niobium base materials in oxidizing environments to the highest temperatures of their application. AcIear knowledge of the Nb-T-Si (T= Cr; Fe; TI) ternary phase diagrams, previously studied, and of the Nb-Fe-Cr-Si quaternary phase diagram, partly determined in this study, allows the choice of a "diffusion path" leading to the formation of coatings with phases selected with regard to their oxidation resistance. Two different chromia former coatings have been particularly studied between 973 to 15,73 K in air. Finally, the results of a complete crystallographic characterization of the different silicides, showing possible Nbm, Nb/Cr, CrlFe and SilFe exchanges, are useful to determine the additional elements which could improve both the physical and mechanical properties of these coatings. 11:25 am Invited ALUMINIDES: YESTERDAY OR TOMORROW COATINGS TO PROTECT AGAINST HIGH TEMPERATURE CORROSION? Roland Streiff, Universite de Provence, 13331 Marseille Cedex 3, France Aluminides were the ftrst coatings used to protect turbine hot components against high temperature oxidation. However, in aggressive environment such as marine environments or burning lower grade fuel, they suffer hot corrosion, the result of the fluxing of the protective oxide scale by molten salts. Despite the development of the MCrYalY class of overlay coatings which have been especially designed to withstand hot corrosion attack, aluminide coatings are still widely and the most used high temperature protective coatings. This paper will review the status of current research on the aluminides used as coatings either on superalloys or on titanium alloys and titanium aluminides. Aluminide coatings modified by "reactive elements", such as hafnium, yttrium, cerium, tantalum or silicon have been developed. Very recent results lead to the conclusion that substrate hafnium addition might be tbe avenue of economical oxidation resistance improvement for tomorrow's alurninide coatings. 11:50 am DEVELOPMENT OF OXIDATION RESISTANT COATINGS: B.D. Chattoraj, A. Garg, M.S. Mukhopadhyay, Research and Development Center for Iron and Steel, Steel Authority of India limited, Ranchi - 834002, India Surfaces of steel and graphite electrodes are corroded by oxidation at high temperatures. These can be protected by using suitable oxidation resistant coatings. For steel, the coat should be easily removable after heat treatment, whereas for the electrode, it should adhere even during cooling and be capable of withstanding a higher temperature. Accordingly, different materials were chosen for different substrates, but bonding components were similar. Coatings for steel were based on magnesia-alumina and on alumina-silicate for the electrode. Adherence and other properties may be controlled by using an appropriate sodium oxiae-potassium oxide-silica bond. The coefficient of thermal expansion (CIE) of the coating is an important criterion and may be controlled by adjusting the cbemistry. The CI'E of coating mass should be very close to that of graphite electrode and should be widely different from that of steel slab. Repeated experiments in the laboratory followed by evaluation through plant trials have given encouraging results. The scale formation and electrode consumption were reduced by 50-70% and 20%, respectively.

HIGH TEMPERATURE SUPERCONDUCTORS· SYNTHESIS, PROCESSING, AND LARGE SCALE APPLICATIONS III: Thallium Conductor Development

Sponsored by: Jt. EMPMD/sMD Superconductlng Materials Committee, MSD Electronic, Magnetic and Optical Phenomena Committee, and FEMS (Federation or European Materials SocIeties) Program Organizers: U. Balachandran, Energy Technology Division, Argonne National Laboratory, Argonne, n.. 60439; Paul J. McGinn, University of Notre Dame, Notre Dame, IN 46556; and Stuart Abell, University of Birmingham, Edgbaston, Birmingham B 152TT, UK Tuesday, AM February 6, 1996

Room:B7 Location: Anaheim Convention Center

Session Chairpersons: D. J. Miller, Argonne National Laboratory, Argonne, n.. 60439; V" Selvamanickam, Intermagnetics General Corporation, Latham, NY 12110

B50

8:30am PROCESSING OFTI-1223 TAPES Jl:Y PARTIAL MELTING: J. C. Moore, S. Fox, D. Hyland, D. Dew-Hughes, C. R. M. Grovenor, Dept. of Materials, Oxford Univ., Parks Road, Oxford, OXI 3PH, United Kingdom; Oxford Univ., Oxford, OXI 3PJ, UK In common with several other grOUP!I, we have shown that it is quite simple to prepare 11-1223 tapes carrying critical cunent densities of 1-2.10' Alcm', but that in only very modest magnetic fields (0.1 T) at 77 K the J,drops by a factor of about 20-50. We have also shown that by using a carefully controlled partial melting process it is possible to increase the grain size in the 122:3 core by a factor of 100, and to achieve much improved field tolerance. This paper will describe the chemical reactions that occur inside the core during the melting and re·annealing processes in Bi- and Pb- substituted 1223, and report on the best current cBnying properties. 8:50am PREPARATION AND PROPERTIES OF Ag(Au)·SHEATHED T1(1223) TAPES: R. E. Gladyshevskii, A. Perin, B. Hensel, and R. AUkiger, Departement de Physique de la Matiere Condensee, 24, quai Emest-Ansermet, 1211 Geneve 4, Switzerland Superconducting (TI,Pb,Bi)(Sr,Ba).Ca,Cu,o.- tapes were produced by the powderin-tube technique from either a pre-reacted 11(1223) powder or a mixture of precursors in Ag-based tubes. High-purity powdeIS (>95 wt% 11(1223), Tc -120 K) were prepared by reaction under high pressure (He) which avoids 11 losses, allowing a better control of the composition. A critical current density of 14 kAlcm' (77 K, °2 T) was reached for Ag-sheathed tapes with pre-reacted powder. Values of 9 kAlcm2 were obtained for Ag(Au)-sheathed tapes with in-tape-reacted powder ic at 0.5 T being decreased by a factor of 16 with respect to the value at 0 T (23 for Ag-sheathed tapes). A small anisotropy of ic was introduced during the tape preparation, the ratio of the critical current densities parallel and perpendicular to the tape surface exceeding 1.7. Increasing the reaction time had a positivl~ effect on the grain growth, but a progressive decomposition of the 11(1223) phase WIIS observed. Locally textured samples with up to 50 m-Iarge plate-like grains were obtained by in-tape-reaction involving partial melting. 9:10 am Invited OPTIMIZATION OF SUPERCONDUCTING TRANSITION TEMPERATURE IN THE T1·2223 AND 11·1223 PHASES VIA ADJUSTING CHEMICAL COMPOSITION AND PROCESSING: R. S. Liu, Department of Chemistry, National Taiwan University, Taipei, Taiwan, China An efficient and highly reproducible method for the preparation of monophasiC 112223 and TI-1223 materials by fin,: tuning the chemical compositions TI.,Ba,Ca,+,Cu,O.., and (11..,Pb..,.)sr•.,Ca"" Cu,o.. respectively, has been developed. The as-synthesized samples of x = 0.4 in both TI,.,Ba,Ca,..CU,)O.o. and (11•.,Pb0.5)Sr2.Ca,.,Cu,o.. compositions have T:s around 120 K and 115 K, respectively. The T:s can be further increased up to 128 K and 124 K in both 112-,Ba,Ca,..Cu,o.o. and (11.~Pb~,)SrhCa,.,Cu,O.. samples, respectively, after the appropriate post-annealing process. The corresponding phase transformation and microstructure studies will be reported. 9:30 am Invited DEVELOPMENT OF LONG LENGTH T1·1223 CONDUCTORS: V. Selvamanickam, P. Haldar, Intermagnetics General Corp., Latham, NY 12110; E. J. Peterson, K. V. Salazaar, Los Alamos National Laboratory, Los Alamos, NM 87545 11-1223 superconductor is an attractive candidate for conductor applications at liquid nitrogen temperatures because of its strong flux pinning properties. 11-1223 conductors by powder-in-tube (PIT) and thick film approaches. Thermomechanical treatments have been developed in both approaches to achieve current densities in the order of 10' Alcw at 77 K in short tapes (4 coo long). While densification of the core is the key factor in the achievement of high current densities in PIT tapes, development of a high friction of c-axis textured regions is one of the main factors in attaining high current densities in thick film deposits. We are investigating the scale-up of 11-1223 conductors to long lengths (10+ m) with current densities in the range of those of short tapes. Uniform current densities in the range of 10' have been achieved in PIT tapes in lengths of several meters. Novel proces!ling techniques have also been developed for the fabrication of thick film 11-1223 conductors in lengths of several meters. 9:50 am Invited PHASE ASSEMBLAGES AND TRANSIENT PHASES FOR GROWTH OF THE (11,BI)-1223 SUPERCONDUCTOR: D. J. Miller, Y. S. Sung, K. C. Richter, M. T. Lanagan, Energy Technology Division; N. J. Zaluzec, Materials Science Division, Argonne National Laboratory, Argonne, n.. 60439; Z. F. Ren, C. A. Wang, J. H. Wang, Dept. of Chemistry and Einstein Chair Dept., State University of New York, Buffalo, NY Various phase assemblages and transient phases involved during the growth of (TI,Bi)1223 have been explored by quenching powder-in-tube as well as film samples. In particular, we have looked for evidence of fluxing liquids and explored the role of Ag in samples in which it has been added intentionally as well as samples in which Ag is introduced only during growth through the external container. Selected intermediate phases have been synthesized for additional characterization. The evolution of superconducting phases and influence of the initial and intermediate phase assemblages will be discussed. Work supported by the U.S. Department of Energy (DOE), Energy Efficiency and Renewable Energy, as part c,f a DOE program to develop electric power technology, and Basic Energy Sciences (Materials Sciences), under Contract W31-109Eng-38.

1996 TMS ANNUAL MEETING AND EXHIBITION

10:10 am BREAK 10:20 am Invited FABRICATION OF BIAXIALLY ALIGNED POLY CRYSTALLINE HTSC CONDUCTORS: A. Goyal, D. P. Norton, E. D. Specht, D. M. Kroeger, M. Paranthaman, Q. He, J.-D. Budai, D. K. Christen, Metals and Ceramics Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831 Measurements of grain boundary misorientations between hundreds of contiguous grains in well-linked, high-Jc Bi-2223 powder-in-tube, Bi-2212 thick films and 111223 thick films using Electron Backscatter Diffraction indicate the presence of a connected network of low energy (low angle and low sigma) boundaries. The observed grain boundary texture in the samples was used in limiting path calculations to determine the effective critical current density. Good agreement with the observed J, was observed. These results suggest that to forther increase the J" macroscopic biaxial texture in the superconductor is required. Several research groups have now demonstrated that ion-beam-assisted deposition (mAD) can be used to deposit biaxially aligned ceramic buffer layers on metal substrates. Subsequent deposition ofYBa,Cu,O, results in biaxially aligned films with high-J:s. We have developed a new method "RABIT' for fabricating biaxially textured substrates. In-plane aligned lITSC films have been grown on these substrates using several methods. Microstructure and properties of these films will be discussed. Research sponsored by U.S. Depanment of Energy under contract DE AC05840R21400. 10:40 am TBCCO HTS TIDN FILMS AND PASSIVE MICROWAVE DEVICES: K. Kale, A. P. Jenkins, J. D. O'Connor, D. Dew-Hughes, D. J. Edwards, Dept. of Engineering Science, Univ. of Oxford, Oxford OXI3PJ,UK; A. P. Bramley, C. R. M. Grovenor, S. M. Morley, Dept. of Materials, Univ. of Oxford, Oxford, OXI 3PJ,UK Thin films of 11-based HTS have been deposited on various substrates, with and without buffer layers. A variety of microwave devices have been fabricated from such films (both single sided and double sided). These include delay lines, planar filters and low phase noise oscillators. Experimental results for such devices will be presented that demonstrate performance enhancements over similar copper implementations of the devices. In addition preliminary details will be given for the integration of these individual components into a complete microwave sub-system. 11:00 am Invited ON THE DEVELOPMENT OF ADVANCED SENSORS TO MONITOR HIGH TEMPERATURE SUPERCONDUCTOR (HTS) MANUFACTURING PROCESSES: S. Parhasarathi, M. D. Aesoph, J. Kajuch, Concurrent Technologies Corporation Johnstown, PA 15904 Silver sheathed, ceramic lITS wire and tape manufacturing processes are currently being scaled up for long length applications. In this context there is an urgent need for improved quality management and control of these processes. The development of advanced sensing techniques for process and quality control through variuos stages of HTS manufacturing promises to reduce manufacturing costs and enhance lITS quality. Uniformity of conductor dimensions and ceramic core density are of critical importance to the inservice performance of lITS wires and tapes. This paper reports on the development of advanced sensors based on nondestructive techniques to monitor conductor dimensions and core density in HTS wires and tapes. Preliminary results from sensor testing for on-line, real-time monitoring of lITS manufacturing processes are also presented. This work was conducted by the National Center for Excellence in Metalworking Technology, operated by Concurrent Technologies Corporation, under contract to the U.S. Navy as part of the U. S. Navy Manufacturing Science and Technology Program. 11:20 am Invited FREEZE DRYING SYNTHESIS OF HIGH 'Ie SUPERCONDUCTORS: O. A. Shlyakhtin, Yu. D. Tretyakov, Dept. of Chemistry, Moscow State University, 119899 Moscow, Russia Freeze drying (FD) method has been successfully applied in the synthesis of the main groups of high Tc phases. Melt processing ofFD powders to obtain dense ceramics needs special procedures to avoid encapsulation of air in micropores. The application of FD in the synthesis of melt grown composites of Y-123 and Bi-2212 with nonsuperconductive dopants allows to reduce the size of dopant particles up to submicron level. Due to easily contrOlled grain size, phase composition and reactivity of precursor powders FD has a good perspective in the Ag-BSCCO tapes preparation. The most effective way ofFC precursors application is 11 and, especially, Hg-HTSC's preparation due to extensive porous structure of particles promoting vapour transport of volatile components. 11:40 am MICROSTRUCTURE OF 11-1223 TAPES PREPARED BY PIT TECHNIC: P. Dileo, S. Fox, Institute of Experimental Physics, SAS, Watsonova 47, 04353 Kosice, Slovakia; J. C. Moore, C. Grovenor, M. J. Gorringe, Depanrnent of Materials, Oxford University, Parks Road, Oxford OXI 3PH, UK The microstructure of Ag and Ag alloy sheathed 11-1223 tapes prepared by PIT technic was studied by polarized light microscopy. The porosity, 1223 platelet size and alignment, secondary phase content and defects were given into the relationship with the annealing time at 840°C and cold rolling thickness reduction.

1996 TMS ANNUAL MEETING AND EXHIBITION

HUME·ROTHERY AWARD SYMPOSIUM FOR WILLIAM L. JOHNSON: THERMODYNAMIC AND KINETIC ISSUES IN THE SYNTHESIS OF METASTABLE MATERIALS III

Sponsored by: Jl. EMPMDISMD Alloy Phases Committee Program Organizer: Professor Carl C. Koch, Materials Science and Engineering Depanrnent, North Carolina State University, Box 7907, Raleigh, NC 27695; Dr. Robert D. Shull, NIST, Bldg. 223 BI52, Gaithersburg, MD 20899 Tuesday, AM February 6, 1996

Room: Orange County I Location: Anaheim Marriott Hotel

Session Chairperson: Professor Brent Fultz, Depanment of Materials Science, California Institute of Technology, Pasadena, CA 91125

8:30 am Invited ATOMIC DIFFUSION IN ALLOYS: BEYOND MEAN·FIELD THEORIES: M. Atvnon, Depanrnents of Nuclear Engineering and Materials Science and Engineering, University of Michigan, Ann Arbor, MI 48109-2104 The atomic flux in alloys is commonly assurued to be proportional to the gradient of the chemical potential. Such an approach is implicitly based on a mean field approximation - the migration rate of an atom is calculated based on its average environment. The effect of the average environment is expressed through the composition-dependent chemical potential, where the composition is a mean field, obtained by averaging the occupations of many lattice sites. Since the atomic jump frequencies are exponential functions of the activation enthalpy for migration, it is obvious that calculation of the jump frequencies based on average site occupation is a poor approximation. In this paper, a calculation of chemical diffusion coefficients will be presented for a dynamic Ising model. In contrast to the simple mean-field theory, the distribution oflocal atomic environments is taken into account. For alloys with large enthalpies of mixing, it will shown that the atomic flux is not proportional to the gradient of the chemical potential. The effect of the assumed jump dynamics on the interdiffusion coefficient will be discussed. The model will also be applied to ordering kinetics. Predictions will be cornpared with experimental results, where available, and the need for molecular dynamics simulations will be discussed. Work funded by NSF grant DMR-9200132.

9:00am SOLID-STATE ALLOYING OF IMMISCIBLE ELEMENTS: E. Ma, J.-H. He, Depanrnent of Mechanical Engineering, Louisiana State University, Baton Rouge, LA 70803; PJ. Schilling, Center for Advanced Microstructures and Devices (CAMD), Louisiana State University, Baton Rouge, LA 70803. F. Pan and B.-. Liu, Department of Materials Science and Engineering, Tsinghua University, Beijing, China Solid-state alloying and amorphization reactions in binary systems immiscible in equilibrium, unlike those in systems exhibiting a negative heat of mixing, are not yet well understood. In this work, we have employed the Extended X-ray Absorption Fine Structure (EXAFS) technique, in addition to XRD, DSC, and TEM, to monitor the formation of thermodynamically unstable alloys induced by mechanical alloying and thin-film interdiffusion reactions in immiscible systems Cu-Fe, Cu-Ta, and Ag-Fe. The combined use of these analytical techniques yielded information regarding local microstructural evolution as well as energy (enthalpy) states of various metastable phases in systems with a positive heat of mixing. A simple model is used to explain the presence of thermodynamic driving force for alloying in certain positive-heat-of-mixing systems when the elemental domain sizes are reduced down to a few nanometers. Mechanically alloyed metastable CU'QO.,Fe, powders were used as the starting material to produce in situ formed two-phase nanocomposites during phase separation upon sinter forging consolidation into full density bulk samples. 9:20 am STRUCTURAL TRANSFORMATIONS OF Fe-AI AND TI·A1 SYSTEMS INDUCED BY MECHANICAL ALLOYING: E. Bonetti, G. Va/dre, INFMUniversity of Bologna, Italy The structural evolution of the Fe-AI and n-Al systems induced by mechanical alloying has been followed by modem techniques of analytical electron microscopy (EDS and PEELS) on cross-sectioned samples and by mechanical spectrOscopy measurements (Internal Friction and Dynamic Elasticity Modulus) on cold consolidated samples. The formation of an extended solid solution was observed for both systems after the first times of milling. The the n,AI system collapses to a quasi-homogeneous amorphous phase through a preferential path in the destroying the host n lattice. Whereas, the FeAl system tranforms by a moderate thermal treatment at about 800 K in a partially long range ordered intermetallic phase. Details of both the transformations will be discussed in details. 9:40am SYNTHESIS OF METASTABLE CARBON·SILICON·NITROGEN COMPOUNDS: B. Park, C. Uslu, D.H. Lee, Y. Betta, School of Materials Science and Engineering, Georgia Institute of Technology, 778 Atlantic Drive, Atlanta, GA 30332-0245; D.B. Poker, Solid State Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831 Preliminary studies have been performed on the feasibility of metastable carbon-

B51

silicon nitride formation (b-Si,..cI.,N" the homologue of equilibrium b-Si,N, or hypothetical b-C,N,). Thin films were formed using high dose N-implantation into SiC with varying ion doses up to 2.7 X 10" N/cm', and target temperatures between -196C and 980·C. X-ray diffraction with a position-sensitive detector and cross-sectional transmission electron microscopy revealed that the as-implanted surfaces (up to 860'C) contained 0.1 m thick buried amorphous layers. Rutherford backscattering spectroscopy showed that the peak concentration of nitrogen saturated up to approximately 54 at. % with increasing doses, suggesting formation of a new phase. Implantation to doses of 1.1 x 10" and 2.3 x 10"N/cm' at 980'Ccaused enhanced surface hardness. The responsible mechanisms for the observed phase formation, nucleation and phase stability will be discussed. 10:00 am BREAK 10:20 am Invited THE KINETICS AND ENERGETICS OF SOLID STATE AMORPHIZATION: E.J. Cons, Physics Department, Binghamton University, State University of New York, Binghamton, NY 13902-6016 Rates of solid state reactions in thin film diffusion couples are directly monitored with measurements of the heat flow during heating. Differential scanning calorimetry and x-ray diffraction analysis are utilized to investigate the kinetics and energetics of solid state amorphization reactions. This calorimetric technique also provides thermodynamic information on the amorphization of alloys during hydrogenation, such as the Zr-Rb alloys of Johnson et ai's pioneering work on solid state amorphization. Further examples include studies in the Ni-Zr system of the heat of formation of amorphous alloys from Ni and Zr, and of the averaged interdiffusion coefficient. Effects of atomic mobility on solid state amorphization reactions are examined in a comparison of reactions in NiIZr 11, composites as the Zr concentration, x, is varied. Studies in meta1lSi systems, such'as lVSi, are also discussed. 10:50 am ON THE THERMODYNAMICS AND KINETICS OF BALL MILLING INDUCED COMBUSTIVE REACTIONS: L Takacs, University of Maryland, Baltimore County, Department of Physics, Baltimore, MD 21228 A variety of highly exothermic reactions can tum intu self propagating combustion during high energy ball milling. Examples include combination reactions tu form metal sulfides, borides, carbides, aluminides, and silicides and displacement reactions between an oxide or halide and a more reactive metal. Thermodynamics, in particular the adiabatic temperature, is a useful guide to predict whether a reaction can sustain itself during a combustive process. However, estimating the activation time hefore the ignition of combustion is a much more difficult task. It requires consideration of all the relevant processes during the activation period. Our current understanding of these processes will be discussed based on the currently available experimental data. Particular attention will be paid to the formation of lattice defects and its influence on the reaction thermodynamics and kinetics. 11:10 am PHASE SEPARATION OF AMORPHOUS PHASE IN MECHANICALLY ALLOYED Fe5OZr5O POWDERS: S.E. Lee, H. Y. Ra, Department of Metallurgical Engineering, Seoul National University, W.T. Kim, Department of Physics, Chongju University, Chongju, Korea; T.H. Yim, Korea Academy of Industrial Technology, Incheon, Korea Amorphous Fe..Zr.. alloys have been successfuly manufactured by mechanical alloying of elemental Fe and Zr powders in a conventional ball mill under an Ar atmosphere. Phase formation and microstructure were studied by using X-ray diffractometery and transmission electron microscopies. Also temperature dependence of magnetization of the alloy powders was determined by using SQUID. Selected area diffraction patterns taken from the mechanically alloyed powders showed two hallow rings, indicating phase separation into Fe rich and Zr rich amorphous phases. Radial distribution funtion obtained by analizing hallow X-ray diffraction patterns also showed two separate nearest atomic distance, supporting the phase separation obserbed by TEM. From temperature dependence of magnetization of Fe-Zr alloy powders, Curie temperature of the Fe-Zr amorphous phase was determined accurately by using Arrot plot. Curie temperature of Fe-Zr amorphous phase indicated that Fe content in the ferro-magnetic amorphous phase is higher than the mean Fe composition of the alloy powders. The result also suggested that phase separation of amorphous phase into Fe rich and Zr rich phases occurred, which is in agreement with the previous results obtained from X-ray diffraction study and transmission electron miroscopy. 11:30 am ON METASTABLE PHASES FORMATION MECHANISM IN ELECTRODEPOSITED FILMS: v.G. Shadrow, T.A. Tochitskii, A.V. Boltushkin, Institute Solid State Physics and Semiconductors, Acad. Sci. of Belarus, P.Brovki, 17, 220072 Minsk, Belarus The mechanism of the polymorphic metal and alloy structure formation in electrodeposited Co based films, taking into consideration the influence of the film content, deposition conditions and included hydrogen, is proposed. In particular, hydrogen localized in the tetrehedral pore promotes pronounced increasing lattice parameter and probabilities of the atom successions ABAB ... and ABCABC ... become equilprobable, which leads to the formation of the fcc phase at small pH values. The addition of P or W deforms Co lattice and promotes a heterogeneous structure formation. The succession of the atum layout becomes random with a short-range local order. The microcrystallites with such successon evidently are the regions of a higher mass density, where the first crystallites appear after isothermal annealing.

B52

HYDROMETALLURGY:RECENT DEVELOPMENTS III: IEnvironment, Equipment And Metal Recovery

Sponsored by: EPD Aqueous Processing Committee Program Organizer. Professor David Dreisinger, The University of British Columbia, Department of Metals and Materials Engineering, 309-6350 Stores Road, Vancouver, B.C., C V6T IZ4 Tuesday, AM February 6, 1996

Room:AI4-15 Location: Anaheim Convention Center

Session Chairperson: Professor Manorlmjan Misra, University of Nevada-Reno, Department of Metallurgical Engineering, MacKay School of Mines, Reno, NV 895570136

8:30am ENERGY AND VOLUME REQUIRE:MENTS FOR OXYGEN TRANSFER TO SAND SLURRIES IN THE DIP (DEI.FT INCLINED PLATE) AERATOR: G. Van Weert and A. Borleffs, Delft University of Technology, Faculty of Mining and Petroleum Engineering, Department of Raw Materials Technology, Mijnbouwstraat 120, 2628 RX Delft, The Netherlands Hydrornetallurgical oxidation of sulphides is capital and energy intensive, with autoclave and bio-technology competing for lowest overall cost status. The use of surface aerators instead of stirred tank reactors could provide lower costs for bio-oxidation of concentrates. With this in mind, the Delft Inclined Plate reactor was designed with inclined plates above a settler shaped reac;tor to return slurry to the surface of the liquid. The multiplate superstructure overcomes the scale up restriction of the standard circular aeration patterns. DIP aerator energy and volume requirements for oxygen transfer from air will be presented in terms of kg O,JkWh and kg O/m'lh transferred as a function of WI % sand in the slurry and compared to stirred tank performance. Scale up from the 4 m' pilot unit will be discussed. 9:00am BACTERIA OXIDATION OF SULFIDES DURING ACID MINE DRAINAGE FORMATION: A MECHANISTIC STUDY: K. Nyavor and N.O. Egiebor, University of Alberta Department of Mining, Metallurgical & Petroleum Engineering, Edmonton, Alberta, Canada T6G 206; P. Fedorak, University of Alberta, Department of Biological Sciences, Edmontun, Alberta, Canada TOO 2G6 The indirect, and the direct contact mechanisms of bacteria oxidation of sulfides were studied using Thiobocillus(T) jerrooxidans and pyrite. The aim was to determine which of the two generally accepted mechanisms played a dominant role, if any, during acid mine drainage (AMD) formation from the oxidation of sulfide minerals. The results show thalat a pH of 4.5 and E., of 0.1 volts where T. jerrooxidans are quite active, but where initial chemical oxidation of pyrite to Fe" is limited, the bacteria were unable to directly oxidize pyrite to any Significant extent. Whereas, in the presence of Fe", T. jerrooxidans assisted oxidation of pyrite by the indirect mechanism was observed and sustained over the pH and E., nmge where the bacteria is active. The results lead to the conclusion that the direct contact mechanism of pyrite oxidation by T. jerrooxidans is insignificant. The predominant, and possibly the only, pathway involves the indirect bacteria oxidation of availabJle Fe" to Fel+, which in tum oxidizes FeS, to generate additonal Fc" and acidity in a cyclic process. 9:30am AMBffiNT PRESSURE PRODUCTION OF CRYSTALLINE SCORODITE FROM ARSENIC-RICH METALLURGICAL EFFLUENT SOLUTIONS: D.J. Droppert and G.P. Demopoulos, McGill University, Department of Metallurgical Engineering, Montreal, Quebec, Canada, H3A 2A7. B. Harris, 3670 St-Famille, Montreal, Quebec, Canada This paper describes the precipitation-production of crystalline scorodite (FeAsO•. 2H,O) from arsenic-rich sulphllte-based metallurgical solutions at ambient pressure (95°C). Arsenic is precipitated by means of a controlled precipitation technique in which the supersaturation of the: solution is carefully controlled. By controlling the supersaturation in the presence of seed material the formation of amorphous ferric arsenate is prevented. Scorodite is Imown to be the most stable As(V) compound and is therefore preferred as host mineral for arsenic immobilization. The effect of the following paramters on the precipitation behavior and kinetics of scorodite will be discussed: type of base, seed concentration, initial [AS), Fe/As ratio, sulphate concentration, neutralization rate, seed recycling and method of Fe(lII) addition to Fe-deficient arsenic solutions. 10:00 am METAL RECOVERY FROM A COMPLEX SULPmDE CONCENTRATE USING A FERRIC CHLORIDE LEACH PROCESS: C.J. Ferron, R.G. Williamson and A.D. Zunkel, Lakefield Research, Lakefield, Ontario, Canada A modified Cuprex chloride leach process was used to extract Cu, Ni and Co from a complex sulphide concentrate. The standard leach was modified to a two-stage countercurrent leach process which provided emf control of the copper SX feed and minimized iron dissolution. A f10wsheet was developed to recover the base metals and silver from the leach solution and gold and platinum group metals from the leach residue,

1996 TMS ANNUAL MEETING AND EXHIBITION

which was predominantly elemental sulphur. Thchniques required to maintain an iron balance in the system including pressure oxidation and solvent extraction are described. 10:30II1II ACID RECOVERY FROM SPENT ACIDS &: ELECTROLYTES VIA CONTINUOUS ION EXCHANGE: R.S. DeMis, Tetra Thchnologies, Inc., 2502S I45 North, The Woodlands, Texas TI380 Acid recovery employs ion exchange resin as separation media for inorganic acids from dissolved metallic salts by physical adsorption and without the use of chemicals. This technology has been used for yeara on spent etching and plating liquors to recover acid for reuse in the bath and to reduce waste acid neutralization costs. It is also used for electrolyte treatment to facilitate valuable metal recovery. Tetra's CCIX'" continuous I-X contactor has been used successfully in a simple, continuous operating mode for acid separation from its metallic salt counterpart. Efficient separation is enhanced by acid reconcentration to its original strength or as high as 14-20%. Continuous treatment also recovers heat to the tankhouse or bath and improves metal qualiry due to consistent acid qUality. Conventional fixed bed ion exchange (I-X) systems are impractical for acid recovery because of complex control design and large resin inventories needed to process very low Iiquid-to-resin treatment ratios associated with strong acid concentrations. 11:00 am THE DESTRUCTION OF FREE AND METAL-COMPLEXED CYANIDE BY PHOTOLYSIS: T. Valentine, S.P. Cashin, C.A. Young, Montana Tech., Dept of Metallurgical Engrg., Butte, MT 59701-8997 Cyanide is a strong lixivant for metals making it favorable for finishing and treatment, flotation control, and leaching. As a result, industrial effluents vary considerably in volume and conentration of free and metal-complexed cyanide. In this study, photolytic techniques are described and experimentally compared for the destruction of free and metal-complexed cyanides. Several weak-acid dissociables (WADs) and strongacid dissociables (SADs) are examined. Various reaction mechanisms are determined by ion chromatography (lC) and inductively-coupled plasma (ICP) emission spectrography. 11:30 am THE REMOVAL OF METAL CATIONS FROM ACID-MINE DRAINAGE USING RHIZOPUS JAVANICUS: J.J. Ballard, D. Dysinger, Ctr for Advanced Minerai Processing, Montana Tech, Butte, MT 59701; P.P' Wood, C.A. Bradley, Mycotech Corp., Butte, MT 59701; C.A. Young, Montana Tech Dept of Met Engrg, Butte, MT 59701-8997 Biosorption is a proven technology for the extraction of metal cations from industrial waste waters; however, most research activity has been conducted with algae, moss, and bacteria. By comparison, little research has involved filamentous fungi. In this regard, an inexpensive and easily cultivated fungus, Rhizopus favanicus, was investigated using acid-mine water from the Berkeley Pit, Butte, Montana. Various operating parameters such as temperature, pH, fungus concentration, fungus size, fungus age, and fungus pretreatment method were studied. Results indicate that extraordinary loading capacities can he achieved.

INTERACTIONS BETWEEN CREEP, FATIGUE AND THE ENVIRONMENT: SESSION I

Sponsored by: SMD Mechanical Metallurgy Committee Program Organizers: W.W. Milligan, Department of Metallurgical and Materials Engineering, Michigan Technological University, Houghton, MI 49931; D.L. Davidson, Southwest Research Institute, PO Box 28510, San Antonio, TX 78228OSlO; M.F. Henry, General Elecbic, CRD, KI-MB229, Schenectady, NY 12345; H.W. Sizek, INCO Alloys International, PO Box 1958, Huntington, WV 25720 Tuesday, AM February 6, 1996

Room: Grand D Location: Anaheim Marriott Hotel

Session Chairpersons: W.W. Milligan, Department of Metallurgical and Materials Engineering, Michigan Technological University, Houghton, MI 49931; D.L. Davidson, Southwest Research Institute, PO Box 28510, San Antonio, TX 782280510

8:30am WELCOME AND INTRODUCTORY REMARKS: w.w. Milligan, Department of Metallurgical and Materials Engineering, Michigan Technological University, Houghton, MI 49931 8:35am A PHENOMENOLOGICAL EXAMINATION OF TIME DEPENDENT CRACK PROPAGATION IN THE PARIS AND NEAR THRESHOLD REGIMES IN NICKEL-BASE SUPERALLOYS: M.F. Henry, General Electric, CRD, KI-MB229, Schenectady, NY 12345 lime dependent fatigue crack propagation has been observed in nickel-base supera!loys in ambient environment at temperatures of 4OO'C and above. This JllIPCr discusses generic techniques for the phenomenological treatment of these data. The hehavior can he treated as a thermally activated process. Fully time dependent data is extrapolated to estimate cyclic periods required to achieve this full time dependence. In addition, the hehavior of these materials in the near threshold regime is explored. Compared to the

1996 TMS ANNUAL MEETING AND EXHIBITION

hehavior in the Paris Law regime, crack propagation rates in the near threshold regime are seen to have an inverse relationship with both temperature and cyclic period. The similitude of the cyclic stress intensity factor, crack closure, creep interaction and envi· ronmental interaction are examined as potential cause of this "cross-over" behavior. 9:00am ENVIRONMENTALLY ASSISTED CRACK GROWTH IN NICKEL-BASE SUPERALLOYS AT ELEVATED TEMPERATURES: R.P. Wei, S.F. Chen, M. Gao, Department of Mechanical Engineering and Mechanics, Lehigh University, Bethlehem, PA 18015 Environmental enhancement of crack growth in nickel-base alloys is re-examined in the light of recent understanding of the role of niobium. In this presentation, results from a coordinated study of crack growth kinetics (in oxygen and water vapor), surface chemistry, and microstructure in a commercial IN718 are reviewed. Crack growth is suggested to result from the formation and rupture of niobium oxides at the grain boundary surfaces, and to he controlled by the rate of oxidation and decomposition of niobium carbides at these boundaries. Additional data on a niobium-free, Nil8Crl8Fe ternary alloy and on data from the literature are presented to corroborate this interpretation. The technological Significance of these rmdings is considered. Research supported by the National Science Foundation under Grant DMR-9102093. 9:25am FATlGUElTlME·DEPENDENT CRACK GROWTH IN ALLOY 718: H. Ghanem, Mechanics of Materials Laboratory, Department of Mechanical Engineering, University of Rhode Island, Kingston, RI 02881 This paper summarizes studies focusing on the crack growth hehavior of Alloy 718 due to a combined influence of environment and loading frequency at 6S0'C. The parameters that have been examined include the crack growth rate, fracture surface morphology and slip line density at and below the fracture surface. Results of these studies suggest that the governing mechanism of the crack tip response is the degree of homogeneity of plastic deformation and associated slip density. For loading conditions promoting homogeneous plastic deformation with a high degree of slip density, the creep-environmental damage conbibution is shown to he limited thus permitting the dominance of the cyclic damage effects which is characterized by a transgranular crack growth mode and lower crack growth rate. Under conditions leading to inhomogeneous plastic deformation and lower slip density the crack tip damage is described in terms of grain boundary oxidation and subsequent fracture. On the basis that the variation in crack growth damage mechanism in Alloy 718 from cyclic-dependent to environmentdependent is governed by the slip character in the crack tip region, experimental observations under different fatigue-creep-environment conditions are examined. Approaches to modify the alloy resistance to environmental effects through microstructure control as well as sensitizing by mechanical conditioning are discussed. 9:50am THE EFFECT OF OXYGEN PARTIAL PRESSURE ON TIME DEPENDENT CRACK PROPAGATION IN N1-BASE SUPERALLOYS: P.F. Browning, M.F. Henry, General Electric, Research and Development, PO Box 8, Schenectady, NY 12301; K. Rajan, Rensselaer Polytechnic Institute, Department of Materials Science and Engineering, Troy, NY 12180 Tune dependent crack propagation in elevated temperature, oxygen hearing environments is thought to occur as a result of oxygen penetration along grain boundaries in front of the advancing crack tip. However, direct evidence to support this argument is lacking, and there is little understanding of either the mechanism(s) by which oxygen acts to reduce intergranular cohesion or the rate limiting step in oxygen transport. In this study, H,IH,O environments are used to control oxygen partial pressure and investigate its effect on crack tip oxide phase stability and crack propagation rate in two Nibase superalloys used for structural applications in the aircraft engine industry. Experimental results are compared with those obtained by other workers in high vacuum, controlled oxygen pressure studies. 10:15II1II BREAK 10:20II1II ENVIRONMENTAL EMBRlTI'LEMENT OF IRON ALUMINIDES UNDER MONOTONIC AND CYCLIC LOADING: D.A. Alven, N.S. Sloloff, Materials Science and Engineering Department, Rensselaer Polytechnic Institute, Troy, NY 12180-3590 The tensile properties and fatigue crack growth behavior at room temperature in air, in hydrogen, and in oxygen for two Fe,AI alloys: Fe-AI-Cr and Fe-A1-Cr-Zr, are reported. In the case of the temary alloy hydrogen and air were embrittling while in the Fe-AI-Cr-Zr alloy hydrogen was the only embrittling environment. Crack growth behavior was adversely affected by moisture, while crack growth rates increased as frequency was decreased for both alloys. Oxygen provided the lowest crack growth rates in each case. Fractographic features have been determined as a function of environment. Fatigue crack growth results have been compared with those of other iron a1uminides. Zirconium appears to improve the crack growth characteristics of Fe,A1type alloys. 10:45II1II EFFECT OF ENVIRONMENTAL DAMAGE ON THE CREEP BEHAVIOR OF A TI·14A1-11Nb ALLOY: R.S. Mishra, A.K. Mukherjee, Department of Chemical Engineering and Materials Science, University of California, Davis, CA 95616 The influence of environmental damage on the creep behavior of a li-24 AI-II Nb alloy has been studied by pre-oxidizing the specimens at 923 K for 30-495 h. All the specimens show extensive surface cracking. The quantitative results of surface cracking agrees well with Riedel model. The technological implication of surface cracking is presented. The pre-oxidized specimens show change in creep strength and life. The dependence of creep life on stress of pre-oxidized specimens is different at lower and

B53

higher stresses. The origin of this difference is explained in terms of concurrent embrittlement kinetics. 11:10 am CREEP-FATIGUE INTERACTIONS DURING HIGH-TEMPERATURE FATIGUE-CRACK GROWTH IN INTERMETALLICALLOYS: K.T. Venkateswara Rao, R.O. Ritchie, Department of Materials Science and Mineral Engineering, University of California, Berkeley, CA 94720-1760 High temperature fatigue-crack propagation behavior in aluminide and silicide intermetallic alloys is reviewed with emphasis on the role of creep deformation, creepfatigue interactions and environment. Specifically, crack-growth data in various twophase (""fI"a,) TIAI alloys and TINbIy-TIAI composites at800'C, and MoSi" MoSi, composites and Nb,AI alloys at 1200"C are presented. It is found that creep is the more dominant damage mechanism in MoSi,-based materials at 1200'C than mechanical fatigue. Crack growth at high temperatures is governed by mutual competition between creep cavitation damage at the grain boundaries and crack bridging by the viscous SiO, glassy phase. However, the crack-growth results are well characterized by the linearelastic fracture mechanics parameter, OK. The influence of creep is less prominent in TIAI alloys for temperatures up to 800'C; fatigue crack growth rates are fastest in the 600-650'C range just below the ductile-to-brittle transition temperature for r-TIAI. Potential implications of such crack growth response on the damage-tolerant approach to component design and their potential structural use are also discussed. Work supported by the U.S. Air Force Office of Scientific Research. 11:35 am ELEVATED-TEMPERATURE FATIGUE-CRACK GROWTH IN GAMMATITANIUM ALUMINIDES: A. McKelvey. K.T. Venkateswara Rao, R.O. Ritchie, Department of Materials Science and Mineral Engineering, University of Califoroia, Berkeley, CA 94720-1760 Intermetallic alloys based on the y-TIAI system are promising materials for many elevated-temperature applications, in particular to replace titanium alloys and nickelbased superalloys in gas-turbine engines. In light of these applications, ambient and high-temperature fatigue-crack growth behavior has been measured in several commercial TIAI-base alloys, including the TI-48AI-2Cr-2Nb ("GE") and TI-46.5AI-2Cr3Nb-0.2W ("KS") alloys, in both the equiaxed duplex and lamellar (a, + 1J microstructures, at ambient, 600 and 800c (above and below the ductile-brittle transition temperature for TIAI). To delineate contribotions from intrinsic mechanisms (which promote fatigue damage) and extrinsic mechanisms (which impede crack growth), "longcrack" tests have been performed under both constantIoad ratio (K..,IK_ =0.1 to 0.7) and constant-K_/increasing K... loading conditions. Crack-profile analysis and Auger spectroscopy, in combination with extensive fractography, have been used to evaluate the roles of microstructural damage and crack-tip shielding in order to delineate the salient microstructure features controlling crack growth. Work supported by the U.S. Air Force Office of Scientific Research.

INTERMETALLICS AND THEIR COMPOSITES: PROCESSING AND APPLICATION TECHNOLO· GIES I

Sponsored by: SMD High Temperature Alloys Committee Program Organizer: P.W. Keefe, Special Metals Corporation, New Hartford, NY 13413, M. Sohi, Allied Signal Engines, Phoenix, AZ 85072-2118 Tuesday, AM February 6, 1996

Room:B4 Location: Anaheim Convention Center

Session Chairperson: P.W. Keefe, Special Metals Corporation, New Hartford, NY 13413

8:30am RECENT ADVANCES IN FIBER COATING TECHNOLOGY FOR USE IN SICtn-MMCs: D. Upodhyaya, Institute for Materials and Advanced Processes, University of Idaho, Moscow, ID 83844-3026; M. J. Wood, C. M. Ward-Close, Structoral Materials Centre, ORA Famborough, Hants, GUI4-6TD, UK; P. Tsakiropoulos, Department of Materials Science and Engineering, University of Surrey, Guildford, Surrey, GU2 5XH,UK; F. H. Froesa, Institute for Materials and Advanced Processes, University ofidaho, Moscow, ID 83844-3026 Continuous silicon carbide fiber-reinforced titanium-based metal matrix composites offer many advantages over conventional monolithic materials including high strengthto-density ratio. However, there is a need to design an improved protective coating for the silicon carbide fiber to combat interfacial chemical reactions as well as to accommodate large mismatch in the coefficient of thermal expansion (CTE) between the SiC fiber and titanium matrix. This paper describe the strategy for designing coatings for continuous silicon carbide fiber and reports on a new coating system based on a compliant layer/reaction barrier concept developed for titanium metal matrix composites including those based on g-TIAI. The newly developed coating system, designated FUWT, consists of GdlGdB, where the Gd layer (adjacent to SiC) works as a compliant layer while the GdB, layer works as a reaction barrier. The results of this work indicate that the FUWT coating perform well both as a reaction barrier (protect the SiC fiber from reactions with titanium up to IIOOC) and a compliant layer (absence of thermal cracking in TIAI based composite). The interfacial shear strength of an as-fabricated SiCIFUWTITI-6AI-4V was found to be higher than the presently available SCS6/TI6AI-4V and Sigma (SMI240)/TI-6AI-4V composites.

B54

8:50am SIC FIBER STRENGTH BEFORE AND AFTER CONSOLIDATION IN A TIALLOY MATRIX COMPOSITE: M. L Gambone, Materials Directorate, Wright Laboratory, Wright-Patterson AFB, OH 45433; V. Fry, Amercom, ARC Sequa Corporation, Chatsworth, CA 91311 The strength distribution of three spools of SiC fiber was measured before fabrication in TI-22AI-23Nb (atomic percent) luatriX composite panels. Fibers were then extracted from the composite and tensile te:lted to measure the post-consolidation strength distribution. The tensile properties of each fiber spool tested before and after composite processing were compared. Fibers from the spool which showed the best pre-consolidation strength demonstrated little loss of strength after extraction from the matrix; however, indications of low strength fiber in the pre-consolidation tensile. data were mirrored in the extracted strength distributions. Fractography of extracted fibers indicated tensile failure initiated from the ,core, mid-radius, and surface of fibers. The relative strength of failures initiating from each location will be discussed as well as their impact on the overall fiber strength distribution. 9:10am THE INDENTATION SHEAR STRENGTH OF SCS-817075 AL COMPOSITES: Liang-Guang Chen, Su-Jien Un, Department of Materials Science and Engineering, National Tsing Hua University, Hsinchu, Taiwan, China Push-out tests were conducted to detennine the interfacial shear strengths of SCS-81 7075 AI composites prepared using diffusion bonding process. The effects of strengths of matrix after aging and annealing treatlllCnt on the debonding behavior were studied. The interfacial shear strength of the peak-aged composite was found to be higher than those of the under-aged and over-aged c,amposites. Due to the low strength of the matrix after annealing treatment relative to the interfacial strength, the premature fracture of the SCS-8 fiber occurred and resulted in a lower interfacial shear strength during the test. The interfacial shear strengths were also variant for the different SCS-8 fibers of a specimen because the extent of matrix plastic flow during hot-pressing was not the same. 9:30am THE DEVELOPMENT OF MONOLITHIC AND GRADED CERAMIC! INTERMETALLIC COMPOSITES IIY TAPE CASTING AND REACTIVE INFILTRATION: S. F. Corbin, P.S. Apt\!, Sherrill Inc. Fort Saskatchewan, Alberta, Canada TSL 2P2 Intermetallic and ceramic materials are candidates for a number of applications including high temperature oxidation-resistant structural materials for aerospace and turbine service. The fabrication of composites from these two class of material allows one to tailor the properties of a part to suit the desired application. This is particularly true if a manufacturing process can be developed where a functionally graded component can be produced with controlled composition and microstructural architecture. This paper will describe the development of a process whereby monolithic and functionally graded ceramic/intermetallic components are formed using a tape casting/reactive infiltration route. Development of tape casting formulations and debindering/sintering techniques will be described which resullt in the formation of ceramic/metal preforms with both graded composition and porosity. In addition, the reactive infiltration of these preforms to form intermetallic/ceramic composites will be described. The dependence of the microstructural characteristics of these composites on the preform microstructure (i.e. porosity and ceramic/metal composition) and infiltrating conditions (i.e. lime and temperature) will also be descri bed. 9:50 am BREAK 10:10 am ON TIlE PERFORMANCE OF TITANIUM BASED HYDROGEN STORAGE ALLOY ELECTRODE WITH ALUMINIUM ADDITION: B. Luan, N. Cui, H. J. Zhao, S. Zhong, H. K. Uu, S. X. Dou, Centre for Superconducting & Electronic Materials, University of Wollongong, Northfields Avenue, Wollongong, NSW 2522 Australia Effects of aluminium addition on the performance to TI,Ni hydrogen storage alloy electrode have been studied in the present work. The cycle life of the electrode was found greatly increased with aluminium addition. XRD (X-Ray Diffraction) analysis indicated the co-existence of a new phase, TI,AI, with TI,Ni. Electrochemical studies revealed that the electrode with aluminium additions are characterized of having active-to-passive transition. The reason fOI" this was attributed to the passivation of the new TI,AI phase. The passive film was confirmed to be composed of titanium and aluminium by AES (Auger Electron Spe:ctroscopy) analysis. Passivation of the electrode reduce the oxidation of alloy powd,~r during charge-discharge process and therefore increase the cycle life of the titanium based hydrogen storage alloy electrode. 10:30 am ELECTRODE PERFORMANCE OF Mg,NI-IOwt. %TI1NI COMPOSITE ALLOY: N. Cui, B. Luan, H. J. Zhao, S. Zhong, H. K. Liu, S. X. Dou, Centre for Superconducting and Electronic Materials, University of Wollongong, Northfields Avenue, Wollongong, NSW 2522 Australia Mg,Ni-1 Owl %TI,Ni sintered composi~e alloy has been successfully syntheSized. The performance of the electrode fabricated from this alloy has been investigated for the first time. X-ray diffraction results showed that Mg,Ni-IOwt.%TJ,Ni sintered composite alloy composed of Mg,Ni, TI,Ni and TINi. Composite of Mg,Ni with TI,Ni greatly increased the discharge capacity of the Mg,Ni-type alloy electrode. The discharge capacity of the electrode was increased from 8 mAh/g for Mg,Ni electrode to 126 mAhlg at ambient temperature. This may be atbibuted to the co-existence of TI,Ni and TINi with Mg,Ni alloy which enhanced the hydrogen reaction electrocatalytic activity of Mg,Ni alloy, resulting in the improvement of the hydriding-dehydring kinetics.

1996 TMS ANNUAL. MEETING AND EXHIBITION

INTERNATIONAL SYMPOSIUM ON REACTIVE METALS: Processing & Applications I: Pure Metals and Alloys Sponsored by: LMD Reactive Metals Committee Co-sponsored by: Canadian Institute of Mining, Metallurgy & Petroleum, Montreal, Canada; The Japan Institute of Metals, Sendai, Japan; Mining & Materials Processing Institute of Japan, Tokyo, Japan; Society for Mining, Metallurgy & Exploration, Littleton, CO Program Organizers: B. Mishra, Department of Metallurgical & Materials Engineering., Colorado School of Mines, Golden, CO 80401; GJ. Kipouros, Department of Mining and Metallurgical Engineering, Technical University of Nova Scotia, Halifax, Nova Scotia, Canada B3J 2X4; R.G. Reddy, Department of Chemical and Metallurgical Engineering, University of Nevada, Reno, NY 89557 Co- Organizers: W.A. Averill, Rocky Fiats, Inc., Golden; R.G. Bautista, University of Nevada - Reno, Reno; M.C. Bronson, Lawrence Livermore Natl. Lab., Livermore; J.A. Sommers, Teledyne Wah Chang Albany, Albany; C.B. Wilson, The Dow Chemical Company, Freeport Tuesday,AM February 6, 1996

Room:B9 Location: Anaheim Convention Center

Session Chairperson: Dr. B. Misbra, Department of Metallurgical & Materials Engineering, Colorado School of Mines, Golden, CO 80401; Dr. GJ. Kipouros, Department of Mining & Metallurgical Engineering, Technical University of Nova Scotia, Halifax, Nova Scotia, Canada B3J 2X4

8:45am MODERN MAGNESIUM PRODUCTION PROCESSES: C.B. Wilson, Magnesium Operations, 2301 N Brazosport Blvd., A-1230, Room 114, Freeport, TX 77541-0286 There are currently two major routes practiced industrially for the production of primary magnesium metal. The flfSt of these is the electrolysis of magnesium chloride and the second is the thermal reduction of magnesium oxide. Within these two major categories, there are many variations of the two technologies which are practiced industrially. The dominate production method is the electrolysis of magnesium chloride and this is practiced by the Dow Chemical Company, Norsk Hydro, Mag Corp., and the plants in the Former Soviet Union. Each of the processes practiced at these facilities uses a different means of feed preparation and this in tum causes the electrolytic cell end of the processes to he different also. The second technology is the thermal reduction of magnesium oxide and there are at least three different variations which are practiced industrially. The flfSt of these is the Pidgeon process which reacts dolime (MgO.CaO) with ferro-silicon (FeSi) to form a di-calcium silicate and magnesium metal at an elevated (about 8ooT) temperature and reduced pressure in an externally heated retort. This is practiced by 11mmenco and produces a very high purity magnesium. The second process reacts the same starting raw materials at a high (about lloo'C) temperature in an internally heated reactor under reduced pressure. This is practiced by Brasmag in Brazil. The final variation of the ferro-silicon process is the Magnetherm process developed by Pechiney and practiced by Pechiney and Alcoa which uses slag resistance heating and very high (about 1550'C) temperatures. This paper will examine each of these production methods and discuss both their merits and their weaknesses. N

9:10am ULTRA HIGH PURITY MAGNESIUM REFINING: R.K.F. Lam, D.R. Marx, Materials Research Corporation, 542 Route 303, Orangeburg, NY 10962 Ultra high purity magnesium was purified by a proprietary vacuum method and apparatus to increase purity by approximately five hundred times in a single step. Purity increased from 99.95% to 99.999%, exclusive of zinc impurity. Properties of starting material and refined material are discussed. A novel purification apparatus was designed and developed for the magnesium refining. The purification process was carried out automatically by a computerized control system, and the system was monitored continuously by a computerized data acquisition system. 9:35am PRODUCTION OF ruGH PURITY TITANIUM BY ELECTROREFINING: M. Kanda, K. Sato, E. Kimura, Mitsubishi Materials Corporation, 1-297 Kitabukurocho, Omiya, Saitama 330, Japan The hehavior of impurities in electrorefining of 11 by NaQ-KCI-11CI, molten salts and the production process of its molten salts have been studied. As a result, purification of salts, materials for the apparatus and quality of anode 11 were important to produce the high quality 11. By using purified molten salts and the apparatus constructed of nickel, impurities of the deposited 11, such as Fe, Ni, could be lowered to about 0.01 ppm. The content of AI, Cr of this 11 were about 0.5 ppm. The volatile impurities were removed from the obtained 11 in this process by the electron beam melting and 6N up grade 11 ingot (except for gas elements) were obtained. 10:00 am BREAK

1996 TMS ANNUAL MEETING AND EXHIBITION

10:15 am HYDROGEN STORAGE WITH NANOCRYSTALLINE MAGNESIUMNICKEL ALLOYS: M.A. Imam, D.A. Meyn, Materials Science & Technology Division, Code 6320, US Naval Research Laboratory, Washington, DC 20375; R.L. Holtz, H. Aaronson, Geo-centers, Inc. 10903 Indian Head Highway, Fort Washington, MD 20744 Of potential practical hydrogen storage materials, pure Mg would be best from the standpoint of hydrogen capacity per unit mass. However, pure Mg has poor hydriding reactivity due to surface oxide poisoning and low catalytic activity. Mg,Ni is an alternative which is much more stable against oxidation and the Ni aids catalysis of the II, dissociation, but at the expense of a 50% decrease in storage capacity. Empirically, it would seem that an ideal hydrogen storage material would be pure Mg particles coated with a thin layer of Ni. Theoretical considerations suggest that this optimum hydrogen storage material may be realized with nanocrystalline Mg-Ni alloys with Ni concentrations comparable to the maximum solid solubility of Ni in Mg. We have prepared nanocrystalline Mg-Ni alloys in powder and thin film form with various concentrations of Ni \>y ball-milling and physical vapor deposit, and report on studies of the surface segregation of Ni and on the hydrogen storage characteristics of these pseudoalloys. 10:40 am EFFECT OF CASTING PARAMETERS ON THE MICROSTRUCTURE OF ASTM F-75 IMPLANT ALLOY: H. Mancha, M. Gomez, J.L. Rodriguez, J. Escobedo, M. Castro, M. Mendez, Centro de Investigaciones y Estudios AvanzadolPN, Unidad Saltillo, Carretera Saltillo-Monterrey Km. 13, Apdo. Postal No. 663, Saltillo, Coahuila, Mexico Hip replacement implants fabricated using the ASTM-F75 alloy sometimes fail in a sudden catastrophic way. In general, the fractures start at microstructural defects exposed to chemical attack by body liquids subjected to stress-corrosion. In this paper, the results of a study on the effect of casting parameters on the microstructure are presented. The mold and the liquid temperatures were varied between 900 and lOOO'C and 1410 and 1470'C, respectively. The results show that the volume fraction of eutectic carbides that are present in the microstructure strongly decreases as the mold and/or pouring temperature increase. The volume fraction of "blocky" crabides remains at a high level and in all instances were found located mainly in interdeuritic regions. It wa also found that the grain size increases rapidly with increasing mold and pouring temperatures, developing long dendrites that grow from the surface to the center of the bars. 11:05 am SYNTHESIS OF Sm-Co ALLOYS BY MECHANOCHEMICAL REDUCTION: W. Richmond, W.F. Miao, P.G. McCormick, Research Center for Advanced Mineral and Materials Processing, University of Western Australia, Nedlands, WA 6097, Australia The synthesis of samarium cobalt alloys from SmCI, and Co precursor powders by mechanical milling with Ca or Na has been investigated using x-ray diffraction, transmission electron microscopy, differential scanning calorimetry and magnetic measurements. With the synthesis of SmCo" using Ca as the reductant, the as milled structure was entirely amorphous. Heat treatment above 8OO'C was required to form the magnetically hard SmCo, phase. Procedures for removal of the CaCI, by-product phase following milling and heat treatment will be discussed. Attempts to synthesis SmCo, from SmCI, and Co using Na as the reductant resulted in a nanocrystalline mixture of Sm, Co and NaCI with no evidence for alloying of Sm and Co. In this case, the treatment at temperatures above 400'C resulted in reversion of Sm to SmCI,.

MATERIALS AND TECHNOLOGIES FOR MICROELECTRONICS III

Sponsored by: EMPMD Electronic Packaging and Interconnection Materials Committee Program Organizers: M. McCormack, AT&T Bell Laboratories, Rm IA-123, 600 Mountain Ave., Murray Hill, NJ 07974; S. K. Kang, IBM, T. J. Watson Research Center, P. O. Box 218, Yorktown Heights, NY 10598; M. R. Notis, Department of Materials Science and Engineering, Lehigh University, 5 East Packer Ave., Whitaker Laboratory, Bethlehem, PA 18015-3195; K. N. Tu, Department of Materials Science and Engineering, UCLA, Los Angeles, CA 90095-1595 Tuesday, AM February 6, 1996

Room: Grand K Location: Anaheim Marriott Hotel

Session Chairperson: M. McCormack, AT & T Bell Laboratories, Rm lA-123, 600 Mountain Ave., Murray Hill, NJ 07974

8:30 am Invited INTERFACIAL REACTIONS DURING SOLDERING WITH LEAD-TIN EUTECTIC AND LEAD (Pb)-FREE, TIN-RICH SOLDERS: S.K. Kang, R. Rai, S. Purushothaman, IBM T.J. Watson Research Center, P.O. Box 218, Yorktown Heights, NY 10598 Lead (Pb)-containing solders used for the interconnections of microelectronic suhsystem assembly and packaging have been a subject of the environmental issues recently. Extensive research and development activities for replacing Pbcontaining solders with Ph-free solders are in progress in electronic industries, universities, and nationallaboratories. One key technical issue recognized with the Pb-free, Sn-rich solders is a need to develop a good barrier metallurgy to control the interfacial reactions,

B55

namely, dissolution of the base metal(s) and concurrent formation of intermetallics at the soldering interfaces. In this study, the interfacial reaction of Cu and Ni metalization with several Pb-free and Pbcontaining solders are investigated. The dissolution kinetics of the base metal(s) as well as the growth kinetics of the intermetallic are discussed. 8:55 am Invited PHASE EQUILIBRIA, COMPOUND FORMATION AND INTERFACIAL INSTABILITY IN THE Au-Cu-Sn SYSTEM: M. R. No/is, Lehigh University, Bethlehem, PA 18015-3195 Phase equilibria in the Au-Cu-Sn system is important in determining which phases and compounds form in solder connection in several electronic device applications such as TAB and BGAjoining. Ternary Au-Cu-Sn compounds were formed by reaction diffusion in Cu,Au, CuAu, or CuAu, vs. Sn solid-state diffusion couples annealed at l7OC. These compounds are identified as (at 25Au-55Cu-20Sn, atomic %) and X (at 33.3Au-33.3Cu-33.3Sn). Significant interfacial breakdown occurred in the CuAu, vs. Sn diffusion couple. Diffusion paths for the various CuAu vs. Sn couples were determined by EPMA and were plotted onto the 170·C isothermal section of the phase diagram. The location of the ternary phases in relation to other known phases is displayed, and compared with the results of other studies in the literature. The present study also included measurement of the liquidus tempemture of a number of ternary alloy compositions. These are also displayed together with other literature data and used to discuss the applicability of this ternary system to joining technologies. 9:20 am InvIted MICROSTRUCTURAL EVOLUTION OF Ag.Pd·pt TWCK FILM CONDUC· TOR DURING STEP SOLDERING: O. Millin, G. M. Crosbie, T.-Y. Pan, H. O. Blair, R. P. Cooper, J. M. Nicholson, Ford Research Labomtory, Ford Motor Company, Dearborn, MI 48121-2053 Thin film packages with ceramic substmte have been widely used in automotive and military applications. The material usually involved in thick film are the AlP, substmte, a Ag-based conductor pad, one layer of the low-Sn, high-Pb solder (solder I), and one layer of the eutectic or near-eutectic Sn-Pb solder (solder 2). Both solder layers are reflowed sequentially, with the process referred to as "step-soldering." In this study, microstructuml changes in a AgPd-Pt thick film conductor bonded to an AlP, substmte were observed during both stages of the step soldering. Samples were reflowed using a solder paste (solder 1) having a composition of 88Pb-IOSn-2Ag. Initial microstructural observations exhibited minimal reaction between solder I and the conductor. Samples were then dipped into a bath of eutectic Sn-Pb solder (solder 2) at varying times and tempemtures. Solder 1 was found to mpidly dissolve at all tempemtures tested. At lower tempemtures, the mtes of dissolution were found to follow a classic diffusion based dissolution equation. This equation was used in a processing model that provided an optimum processing window. It was used to predict the soldering lifetime of the conductor based upon alloy composition, time, and tempemture. The model does not account for convection effects, that are likely to play an important role at higher tempemtures and during the dynamic soldering processes. Catastrophic damage to the conductor, in the form of intermetallic formation, occurred once the conductor contacted molten solder 2. 9:45am A COMPARISON STUDY OF THE WETTING REACTIONS OF EUTECTIC SnPb ON Cu, Pd, AND NI FOILS: Jessica P. AIIfUlIYJZ, Carmen Fernandez, Shanay R. Trunnell, Y. Yang, H.K. Kim, and K.N. Tu, Department of Materials Science and Engineering, UCLA, Los Angeles, CA 90095-1595 We performed a comparison study of the morphology and reaction mte of molten eutectic Sn-Pb solder on Cu, Pd, and Ni foils. Overall, the reactions on Cu and Ni are quite similar except that the wetting angle on Cu is smaller and reaction mte is faster. However, no stable wetting angle can be observed on Pd and the solder spreads out until the reaction is completed. The reaction mte of intermetallic compound formation on Pd is much faster than those on Cu and Ni. Only Sn forms compounds with them and no Pb compounds can be detected. The morphology of the Sn compounds has been examined by selectively etching the Pb away from the cross-section of the samples. We observed a strong dependence of the reaction mte on the morphology of the compounds.

Engineering, UCLA, Los Angeles, CA 90095-1595 The soldering reaction on Ni is of interest because of the use of Ni as a cushion layer on Cu to reduce the consumption of Cu by the solder. We have conducted a systematic study of the soldering reactions of eutectic Sn-Pb, eutectic Sn-Bi and pure Sn with Ni. The wetting angles and wetting interfaces have been studied by using SEM and EOX, The wetting angles do not change much with time, yet a reaction band has been observed to grow in front of the wetting tip Or the Sn-Pb solder. The cross-sectional views showed the formation of Ni-Sn intermetallic compounds. The morphology of the compound interfaces is mther smooth. 11:15 am WETTING BEHAVIOR OF EUTECTIC Sn.Ag ALLOY ON Cu SURFACE: H.K Kim, K.N. Tu, Department of Materials Science and Engineering, UCLA, Los Angeles, CA 90095-1595 The Pb-bearing solder used in microeltlClronic industry is becoming an environmental issue. Since the eutecffc Sn-Ag alloy is one of the Pb-free solden which have received much attention, we studied its wetting and aging behavior on Cu. The Cu-Sn andAg-Sn intermetallic compound formation has been studied. At the wetting tip, Cu-Sn compounds form and grow into the solder side, and a scaIlopJike Cu.Sn, compound structure forms at the interface between the solder and Cu substrate. Large Ag,Sn grains have been observed to grow in the matrix of the solder and the amount increases with aging time. Upon solidification, the solder surface showed crack formation. 11:40 am INTERFACIAL REACTIONS OF MOLTEN Sn AND In WITH THE Cu SUBSTRATE: L. Su, S.Chen, Department of Chemical Engineering, National Tsing-Hua Univenity, Hsin-Chu, Taiwan 33043, China The interfacial reactions of molten Sn Imd In with the Cu substrate were determined by studying the reacffon diffusion couplles. Cu foil was encapsulated with Sn or In shots in a quartz tube. The sample capsule was then annealed at a predetermined tempemture. The annealing tempemtures were 3OO·C and 450"C. These tempemtures were higher than the melting points of Sn and In, whilst than lower than that of Cu. At these annealing tempemtures, the Sn and In shots became molten and reacted with the Cu substrate. The phases formed at the interface were examined by using OM, SEM, and EPMA. The thickness of the reaction layers was measured by using an image analyzer. Only Culllln.phase was observed at the interface of the Culln couples annealed at both temperatures. For CulSn couples, two phases, and were found. In comparison with the results available in the litemture for couples of solid CU/solid Sn or In, the phase formation sequences were similar, however, the interfacial morphology and the reaction rates were quite different. For the solidlliquid couples, the reaction mte was much faster and interface was nonplanar.

MATERIALS PROCESSING FUNDAMENTALS SYMPOSIUM· III

Sponsored by: Jt. EPDIMDMD Synthesis, Control, and Analysis In Materials Processing Committee, EPD Process Fundamentals, Aqueous Proc:esslng, Copper, Nlckel·Cobalt, PyrometaIlurgy, Lead, Zinc, TIn Committees, MSD Thermodynamic & Phase EquDlbria Committee Program Organizers: R. G. Reddy, Department of Chemical and Metallurgical Engineering, Univenity of Nevada, Reno NV 89557; S. Viswanathan, Oak Ridge National Lab., Oak Ridge, TN 37831-6083: J.C. Malas, Wright-Patterson AFB, OH 45433-6533 Tuesday, AM February 6, 1995

Room: AI6-l7 Locaticln: Anaheim Convention Center

Session Chairpersons: G.W. Warren, Department of Metallurgical and Materials Engineering, Univenity of Alabama, Tuscaloosa, AL 35487; Y.B. Hahn, Department of Chemical Engineering & Technology, Chonbuk National University, Ouckjin-Dong 1Ga, Chonju, Korea

10:10 am BREAK 10:15 am XPS STUDY OF THE OXIDATION BEHAVIOR OF Cu-Sn INTERMETALLIC COMPOUNDS AT LOW TEMPERATURE: O. Wang, A. C. Miller, M. R. NoIis, Lehigh Univenity, Bethlehem, PA 18015-3195 The oxidation of Cu-Sn intermetallic compounds is of significant technological interest in relation to the wetting behavior of Sn-base solder alloys in contact with Cu and Cu alloys. In this present study, oxide films formed in-situ in pure O2 on a number of different Cu-Sn intermetallic compounds held at temperatures <350·C have been analyzed using XPS. Single-phase Cu-Sn intermetallic compounds used for this study were made by hot isostatic pressing of mpidly solidified powden and were provided by NIST. In order to characterize the nature of the oxide layers formed on the intermetallic compound surfaces, spectm obtained from the Cu-Sn intermetallics are compared with speetm for pure copper and pure tin oxidized at the same conditions. The role of copper and tin on oxidation behavior in Cu-Sn intermetallics will be discussed in terms of available phase equilibria diagrams.

8:30am KINETICS OF MILLERITE DISSOLllJTION IN CUPRIC CHLORIDE SOLUTIONS: R.C. Hubli, T. K. MukheIjee, C. K. Ghupta, Bhabha Atomic Research Centre, Trombay, Bombay, India; S. Venkatachalam, Department of Metallurgical Engineering, and Materials Science, Indian Institute of TechnologyBombay, Bombay, India. R.G. Bautista, Department of Chemical and Metallurgical Engineering, Univenity of Nevada, Reno" NV 89557 A kinetic study of the dissolution of millerite in cupric chloride media has been conducted to determine the effect of different variables in the mte expression. Experimental data were obtained on naturally available millerite in a labomtory scale stirred pressure reactor under inert atmosphere. The variables evaluated include acid strength, initital copper (II) concentration, total chloride concentration, and temperature. The mte equation is expressed in amount of mineral dissolved per unit area of mineral surface per unit time.

10:50 am MORPHOLOGY OF THE WETTING REACTIONS OF EUTECTIC SnPb, EUTECTIC SnBl AND PURE Sn ON NI: Shanay R. Trunnell, Carmen Fernandez, Jessica P. AJmarnz, H.K. Kim, K.N. Tu, Department of Materials Science and

8:55am NEW EVALUATION OF NON·CYANIDE LIXIVIANTS FOR GOLD EXTRACTION FROM COMPLEX AND REFRACTORY GOLD ORES: 1.. M. Abrantes, Y.z. Zhang, ICAT, Faculdade dI: Ci~ncias, Universidade de Lisboa, 1700 Lisboa, Portugal

B56

1996 TMS ANNUAL MEETING AND EXHIBITION

Cyanide has generated environmental concerns because of its toxicity, and it may not respond to the treatment of complex gold ores and refractory gold ores. Noncyanide lixiviants thiourea and thiosulphate are the most promising substitutes of cyanide. However, the commercial adoption of the non-cyanide process has been hindered by three factors: more expensive and higher consumption than cyanide, and the gold recovery step still requires more development. The consumption of gold Iixiviants include the stoichiometric usage for gold dissolution, the degradation loss, and the loss due to the impurity interference (forming complexes, precipitates and absorption). The degradation processes of cyanide, thiorea and thiosulphate systems are discussed on thermodynamics, kinetics and electrochemistry basis. The influences of various factors on the degradation reactions of gold Iixiviants are analyzed and illustrated for gold leaching from different refractory gold ores. New progress on decreasing the consumption of non-cyanide and avoiding gold gold passivation are summarized. Finally, progresses on gold recovery from the non-cyanide pregnant solutions by various techniques such as cementation, CIP, RIP, solvent extraction, hydrogen reduction and electrowinning are reviewed. The potential of non-cyanide leaching as a replacement for cyanidation is discussed. 9:20am A MODEL TO DESCRmE THE CHEMICAL SPECIES CONCENTRATION CHANGES IN THE DISSOLUTION OF MILLERITE IN CUPRIC Cm..ORIDE SOLUTIONS: R.C. Hubli, T.K. MukheIjee, C.K. Gupta, Bhabha Atomic Research Centre, Trombay, Bombay, India; S. Venkatachalarn, Department of Metallurgical Engineering and Materials Science, Indian Institute of TechnologyBomhay, Bombay, India; R. G. Bautista, Department of Chemical and Metallurgical Engineering, University of Nevada, Reno, NV 89557 The model computes the initial concentrations of the chemical species present in the Iixiviant and leach solutions and the subsequent changes in their concentrations with small increments of dissolution of the mineral for the specified amount or period of dissolution. The dissolution step is simulated under initial cupric and acid concentrations. The initial analytical concentrations of the components are specified. The initial concentrations of all the species are obtained by solving a set of simultaneous nonlinear equations for the independent mass action expressions, mass and charge balance equations and the dissolution rate equation. 9:55 am BREAK 10:05 am THE GAS COMPOSITIONS CALCULATIONS ON THE SYSTEM OF FeSH 0 AT ELEVATED TEMPERATURE: GUO Xian ]ian, Beijing General R:search Institute for Nonferrous Metals, Beijing, 100088, China; James L. Hendrix, Mackay School of Mines, University of Nevada, Reno, NV In this paper, the gas compositions calculations on the FeS-H,o system at elevated temperature were performed. The results show that the gas constituents are ~O, SO" H,S, S., ~, and 0 , and the order of the compositions partial pressure is PHlO > Pill> P > P.> P > POl when FeS is oxidized into Fe,O. and FeO by stearn in the ternpe":itureSJrang:;"'of 700 to 960"C, the Qaus process occurs with the decrease of gas temperature in the off-gas and the SO, and ~S in the gas are transformed into Sj' the critical temperature for the transformation is 6OO"C, and conversion percent of Sj is 97% at 330"C. 10:30 am OXIDATION MECHANISM OF ARSENOPYRITE IN ACIDIC FERRIC SULPHATE MEDIA IN THE ABSENCE AND PRESENCE OF BACTERIA T. FERROOXlDNAS: Y. Z. Zhang, Y.Y. Lu, T.C. Zhao, Department of Metallurgy, Central South University of Technology, Changsha 410083, Hunan, China The mechanism of arsenopyrite oxidation by ferric ion and bacteria has been investigated. Rest potential, polarization resistance measurements, polarization curve, cyclic voltammetry experiments are conducted in a ferric sulphate medium of 10gIL Fe"', temperature 25-3S"C, pH 1.5-2.5 in the absence and presence of bacteria Tferrooxidans, respectively. The anodic oxcidation process of arsenopyrite is expressed as: at low p0tential: FeAsS + 3~0 Fe" + H,AsO, +S· +3H+ + 4e; and at high potential: FeAsS + 8H,o =Fe'" + H,AsO. +S01' + 148+ + 14e. The cathodic process is the reduction of oxygen. The oxidation behavior of arsenopyrite in sterile and inoculated media are compared. The preferential attachment of bacteria onto the surface of arsenopyrite decreases the rest potential of arsenopyrite from 455 (sterile) to 402 (inoculated) mV(SCE) and the charge transfer resistance from 206 (sterile) to 80 (inoculated) nlcm . The ' absorption of oxygen by bacteria increases the transfer of electrons from anode to cathode pots, and accelerates the dissolution rate of arsenopyrite.

=

10:55 am MATHEMATICAL MODELING OF THE SEPARATION OF LIQUIDLIQUID DISPERSIONS IN A DEEP-LAYER GRAVITY SETTLER: M.e. Ruiz, R. Padilla, Department of Metallurgical Engineering, University of Concepci6n, Casilla 53-C, Concepci6n, Chile An experimental study has been carried out on the separation of phases in a laboratory scale mixer-settler, for a system consisting of 10% Acorga MS640 in Escaid 1030.25 M Na,SO. solution. The thickness of the dispersion band, the local dispersed phase holdup and the size distribution of the dispersed phase droplets were determined for various operational conditions. Based on the experimental fmdings, a mathematical model for the steady state operation of a deep-layer gravity settler has been developed by using a population balance approach. This model takes into account the size distribution of drops within the dispersion band and it uses rate expressions for the description of drop-drop and drop-interface coalescence phenomena The model predicts very well the thickness of the dispersion band as well as the growth of the dispersed phase drops occurring by drop-drop coalescence within the dispersion band.

1996 TMS ANNUAL MEETING AND EXHIBITION

MATERIALS SCIENCE OF CONTACTS AND METALLIZATION AND INTERCONNECTS II

Sponsored by:EMPMD ThIn FUm & Interfaces Committee Program Organizers: Jagdish Narayan, Dept of Matis Sci & Engrg, Box 7916, NC State University, Raleigh, NC 27695; John Sanchez, Advanced Micro Devices, M3 160, PO Box 3453, Sunnyvale, CA 94088 Tuesday, AM February 6, 1996

Room: Orange County 5 Location: Anaheim Marriott Hotel

Session Co-chairs: J.M. Woodall, Purdue University, West Lafayette, IN 47907; M.O. Aboelfotoh, North Carolina State University, Raleigh, NC 27695

8:30 am Invited PROPERTIES OF OHMIC CONTACTS AND SCHOTTKY BARRffiRS ON GaN: H. Morkoc, University of Illinois at Urbana-Champaign, IL 61810 GaN and its alloys have gained considerable interest recently owing to the introduction of blue LEDs and potential applications to coherent sources and high power devices. Having experienced a tremendous improvement in the quality of the material paves the way for an examination of metal contacts in order to harness the device potential. Thus, research into ohmic and Schottky metals have already begun and rapid progress ensued. With TilAlINiIAu annealed at 900°C for 30 sec, specific resistivities in the high 8.9 xlo-' cm' for a doping level of 4 x 10" cm"have already been obtained. On the Schottky barrier front, PtIAu has been found to lead to low ideality factors and low leakage currents approaching nearly ideal behavior. The electrical and structural properties of metaVGaN structures will be discussed. 9:00 am Invited OHMIC CONTACTS TO WIDE BANDGAP SEMICONDUCTORS HgSe AND Au-Pt-Zn'hlZnSeMQW'S ON p-ZnSe AND Au ON GaN: P.H. Holloway, J. Fijol, J.Trexler, S.Miller, Department of Materials SCience and Engineering, University of Rorida, Gainesville, FL 32611-6400 Achieving ohmic electrical contacts to p-type semiconductors is non-trivial as the bandgap becomes large, because the top of the valence band, E" normally is far below the vacuum reference energy level. Even for semiconductors such as ZnSe and GaN with unpinned Fermi energy levels, metals with large work functions are required since E, is so negative. As a result, we have studied an ex situ method of forming the large electron affinity semi-metal HgSe and its use for ohmic contacts to p-ZnSe. This approach still results in Schottky barriers ofO.55eV at RT. Ohmic contacts may be achieved using MQW's of p-ZnTelZnSe, but these degrade with time while transporting currents at densities of 30Alcm'. The characteristics and mechanisms of this degradation will be illustrated and discussed. Finally, the lessons learned from ZnSe will be discussed relative to the search for ohmic contacts to wide bandgap GaN. Specifically, the ability to form ohmic contacts using Au-based metallizations will be illustrated and discussed. This work supported by ONR Grant NOO 14-92-J-189S. 9:30am INFLUENCE OF PLATING PARAMETERS ON GROWTH MORPHOLOGY AND GOLD BUMP SHAPE DEPOSITED ON GaAs DEVICES: L-P. Lai, D. Ramirez, P. Goolsby, Motorola, Inc., MD EL609, 2100 E.E1liot Rd., Tempe, AZ 85284 The effect of pulsed-plating parameters such as voltage, frequency and duty cycle on the surface morphology and shape of plated gold bumps was investigated. The study was performed on the electroplating process to produce rectangular sbaped micro-gold bumps deposited on GaAs devices to form interconnections and flip chip bond sites. An internally developed plating system was used which provides precise control of the charge and discharge characteristics of the electrochemical diffusion layer while maintaining a constant current density independent of varying plating area. A set of plating parameters was selected with high frequency (>3000 Hz) and low duty cycle (10%) while maintaining a large difference of "on" and "off' voltages. This resulted in uniform deposition across a wafer which produces very flat bumps (less than 1.5 mm planarity) with smooth grain structure and soft gold to maintain adequate assembly yields. The process diminishes the effect of changing chemistry during a plating solution's effective lifetime and flow dynamics upon deposition morphology and rate, and greatly enlarges the process window. 10:00 am BREAK 10:30 am Invited SPREADING OF CONTACTS: ORIGIN AND RAMIFICATIONS: S. Mahajan, Department of Materials Science and Engineering, Carnegie Mellon University, Pittsburgh, PA 15213 State-of-the-art semiconducting devices impose stringent demands on contacts in terms of their dimensional and chemical stabilities. We have evaluated the dimensional stability of circular gold contacts deposited on InP surfaces of different orientations during annealing in the temperature range of 45()'6OO"C. Results indicate that the contacts spread during annealing and assume different shapes. The observed shapes reflect the crystallography of the underlying InP surface. Similar results have been obtained on the spreading of aluminum contacts on a (III) silicon surface. Arguments will be developed to rationalize the preceding observations. Ramifications of the contact spread-

B57

ing to the device tecbnology will also be discussed. 11:00 am Invited AN INVESTIGATION OF SnPbAg SOLDER FOR DIE A'ITACH OF GRAs DEVICES: J.M. Parsey, Jr., W. Cronin. L-P. Lai, P. Mobley. B. Scrivner, S. Valocchi. L. Sue.. Motorola. Tempe. AZ 85284; J. Mohr, K. Monarch, B. Keller. Motorola MCL. Mesa. AZ 85202 The interactions of Sn.,Pb36Au, solder with NiV-Au and TtPtAu backmetal systems were investigated. GaAs die were soldered to copper slugs (plated with Ni (barrier) and Au). Die attach and reflow cycles were carried out in the range of 190-25OC. Optical microscopy. SEM cross-sections, EDS and X-ray transmission were used for assessing the evolution of the microstructure through die attach and subsequent reflow operations. Five alloy phases were identified in the solder matrix. The Au-Sn phases produces a "girder like" needle microstructure, and strongly influenced the void formation and coalescence. There was no discernible difference in the microstructures for NiVAu or TtPIAu backmetals. Fine structure and dispersed voids found at die attach were replaced by coarse structure and large voids after reflow. The void volumes were in agreement with liquid-solid phase transitions. Ni was found to provide robust barrier to diffusion and interactions at the die and slug interfaces. We also identified the critical role of the thermal conductivities of the slug and die on the solidification of the solder and void formation. and deduced that the reflow process results in only a partial remelting of the solder due to the high-melting point Au-Sn microstructure.

MELT.SPINNING, STRIP CASTING, AND SLAB CASTING III: MODELLING AND MICROSTRUC· TURES

Sponsored by: MDMD Solldlftcatlon and SCAMP Committees Program Organizers: E.F. Matthys. Mechanical Engineering Department. University of California. Santa Barbara. CA 93106; w.G. Truckner. Technical DirectorProduct Development. Alcoa Technical Center. Alcoa Center, PA 15069 lUesday,AM February 6. 1996

Room: B3 Location: Anaheim Convention Center

Session Chairperson: C. Levi. University of California. Dept of Mech Engrg, Santa Barbara. CA 93106; and W. Truckner. Alcoa Technical Center, Alcoa. PA 15069

8:45am MECHANICAL DESIGN OF TWlN·ROLL IN TIDN STRIP CASTING OF METALS: Ampere A. 1Seng. Department of Mechanical Engineering and Mechanics. Drexel University, Philadelphia. PA 19104 In recent years. the technology of twin-roll caster has been widely investigated for the application of thin strip casting of both ferrous and non-ferrous metals and alloys. Most of the investigations focused on either microstructural evaluations or transport phenomena behavior. Only limited information on the mechanical aspect of twin-roll casting is available. However. in order to design and put twin-roll mill to practical use, the mechanical information including separating forces. torques. and thermal and mechanical stresses induced in the rolls have to be understood. In this paper. the development of a semi-analytical model to study the interactions between operating and design parameters is presented. While the operating parameters includes casting speed. strip thickness. height of molten pool, and pouring temperature, the design parameter consists of roll geometry. roll material. separating force. and torque. The model predictions agree very well with the on-line measurements. 9:10am A SOLIDIFICATION AND COOLING ROLL DEFORMATION ANALYSIS CONSIDERING THERMAL FWW IN TWIN ROLL STRIP CONTINUOUS CASTING PROCESS: C.G. Kang. Research Institute of Mechanical Technology, Pusan National University. Korea; Y.D. Kim, Engineering Research Center for Near Net Shape and Die Manufacturing. Pusan National University. Korea The twin-roll type strip continuous casting process of steel materials is characterized by two rotating water-cooled rolls receiving a steady supply of molten metal which solidifies onto the rolls. Mathematical models of the process to determine the role of various process parameter have been proposed to estimation of casting rolls life. The roll deformation leads to short roller life because thermal fatigue cracking of the casting roll. Solidification analysis of molten metal considering phase transformation and thermofluid is performed using finite difference method with curvilinear coordinate to reduce computing time and molten region analysis with arbitrary shape. An enthalpyspecific heat method is used to handle the latent heat effect during the phase change. The computed velocity field is used to determine the temperatures in the roll and the steel. The temperature distribution of cooling roll is calculated using two dimensional finite element method. because of complex roll shape due to cooling hole in the rolls and improvement accuracy of calculation result. The energy equation of cooling roll is solved simultaneously with the conservation equation of molten metal in order to consider heat transfer through the cooling roll. The calculated roll temperature is compared to experimental results and the heat transfer coefficient between cooling roll surface and rolling material (steel) is also determined from comparison of measured roll temperature and calculated temperature. Rolling force in strip continuous casting process has been computed using rigid viscoplastic finite element method. The relation between stress and strain rate is used as a function of temperature. The three dimension deformation analysis of cooling roll has also been performed by using of commercial package ANSYS. We focused our aoention on elastic-plastic stress analysis to the esti-

B58

mation of roll life caused by cracking on the hot roll surface. To more accurately model the complete thermal cycle of the roll. creep strain is included in an elastic-plastic analysis. In this study. roll life is predicted in terms of ultimate roll revolution without failure on the roll surface. The thermo-mechanical properties of roll material with copper alloy are obtained by uniaxial tensile test at controlled environmental temperature. The proposed analysis techniques are also used to developed an improved roller design. In particular. we will seek to det.ermine the effect of the various process and cooling roll design parameters on roll deJformation behavior and roll life prediction. 9:35am 3·D BULGING ANALYSIS OF CONl1NUOUSLY CAST STEEL SLABS: Taejung Yeo. Kyung-hyun Kim. Kyu Hwan Oh. Dong Nyung Lee, Department of Metallurgical Engineering and Center for Advanced Materials Research, Seoul National University. Seoul 151-742. Korl~a The deformation behavior of continuously cast steel slabs in the mold exit region has been investigated. To analyze bulging of the solidifying steel shell, it is necessary to obtain the profile of the shell in the mold region. The profile of the shell can be given from the heat transfer analysis which has liaken air gap into account and 3-D FEM mesh has been made. Static bulging due to ferrostatic pressure in the mold exit region has been calculated by FEM using thermo-elastic-plastic material model. The effect of support roll pitch and casting speed has been analyzed. Large strain was developed at the right beneath of support roll due to bulging. and its magnitude increased with increasing support roll pitch. Dynamic bulging which takes ferrostatic pressure and the movement of shell into account has been analyzed. This analysis predicted reported deformation geometry of solidifying shell mOle precisely than static analysis. 10:00 am BREAK 10:20 am DENSIFICATION KINETICS OF CAST AI-Mn-Mg SLABS AT SOLID STATE: P.T. Wang. Alcoa Technical Center. Alcoa Center. PA 15069 Densification kinetics of castAI-Mn-Mg slabs containing porosity is investigated at solid state through subscale experimental evaluation and constitutive modeling. Slab samples with two initial porosity levds. 0.4 % and 2% were deformed under axisymmetric compression tests. Experimental data generated from various deformation histories at elevated temperatures revealed that the densification rate, insensitive to strain rate. is higher at the beginning of deformation. The material point at this stage may correspond to the casting location after the solidification front. The densification rate slows down at the later stage of deformation. Optical microscopy revealed that silicon and ironconstituents are usually .;Iustered around the voids that may present weak links if they are brittle. A constitut:ive model based on the interaction of matrix and voids resulted from local plasticity flow. is tuned using a material function to fit experimental data and a reasonable agreement obtained. 10:45 am HEAT FLOW, DEFORMATION BEHAVIOR AND CRACK FORMATION DURING CONTINUOUS CASTING OF BEAM BLANK: Kyung-hyun Kim, Heung Nam Han. Tae-jung Yeo. Yong-gi Lee. Kyu Hwan Oh. Dong Nyung Lee. Department of Metallurgical Engineering and Center for Advanced Materials Research. Seoul National University. 56-1 Shinrim-dong. Kwanak-ku. Seoul 151· 742. Korea A two-dimensional transient finite element model has been developed to compute the heat transfer and deformation behavior of tbe solidifying shell in continuously cast beam blank. The mathematical model developed for this study is capable of treating the heat flow and deformation as coupled phenomena. and of taking into account steel composition through the calculation of the non-equilibrium pseudo binary Fe-C phase diagram. Also. the model incorporates the effects of microsegregation of solute elements on hot tears using the proposed mec:hanical property model of mushy zone which take d-g transformation into account. The maximum tensile stress develops on the web surface because mold inhibits the thermal contraction of beam blank near the fillet. The surface cracks originate on the web surface at the initial stage and ropagates into the solidified shell during casting. The intemal cracks form in the flange tip region at the intermediate stage. These predictions are in agreement with the experimental observation. 11:10 am MICROSEGREGATION DURING MELT·SPINNING OF DILUTE Pd ALLOYS: OJ. Thoma. E.M. Schwartz. S.R. Bingert. D.R. Korzekwa. R.D.Field, LA. Jacobson. Los Alamos National Laboratory. Los Alamos. NM 87545 Free jet melt-spinning has been used to produce dilute Pd-X alloys (X=Ni. Co. Cr) with reduced segregation. The microsegregation in the alloys was experimentally examined and compared with solidification theory. Increased microsegregation occurred in alloy systems with smaller partition co~fficients. steeper liquidus slopes. and higher solute concentrations. In order to further reduce segregation in alloys with higher solute concentrations (-10at. %), statistically designed computer simulations of the meltspinning process were performed. The modeling incorporated fluid flow. heat transfer. and solidification theory. and three process variables were examined: wheel speed. ejection pressure. and crucible orifice diameter. Based upon the optimized simulations. experiments were performed to yield thinner ribbon with greater compositional homogeneity. 11:35 am ANALYSIS OF THE RIBBON FORMATION BY PLANAR FLOW CASTING: L. Kubicar. S. Adamisova. Institute of Physics SAS. 84228 Bratislava. Slovak Republik A model of a basin with a cold bottom is used for ribbon formation by Planar Row Casting. The solidification starts by contlcting of the melt with the cold bottom. The

1996 TMS ANNUAL MEETING AND EXHIBITION

solidification front spreads from the cold bottom into the melt and the solidified ribbon is extruded by the moving bottom. The conditions of the ribbon formation due to the spreading of the solidification front are studied in great detail in the framework. of the model presented. The thickness of the ribbon depends on the period during which the solidification front spreads into the melt. This period is given by the shape of the puddle and by the surface velocity of the cooling wheel. The temperature gradient inside the melt influences the structure of the ribbon. While large temperature gradient leads to the freezing of the melt and thus to the glass formation small temperature gradient causes redistribution of the melt component, nucleation even crystallization. As a result a ribbon can be obtained which is not homogenous across its thickness. Any latent heat corresponding to crystallization retards the spreading of the solidification front. The theoretical predictions are compared with experimental facts.

PHYSICAL CHEMISTRY OF HIGH TEMPERATURE COMPOSITES AND INTERMETALLICS III: Intermetallics And Environments

Sponsored by: EPD Process Fundamentals Committee, MSD Thermodynamics & Phase EquUlbria Committee, Japan Institute of Metals Program Organizers: Prof. R. Y. Lin, University of Cincinnati; Prof. Y. Austin Chang, University of Wisconsin-Madison; Prof. R. Reddy, University of Navada-Reno, Dr.

C.T. Liu, Oak Ridge NL Tuesday, AM February 6, 1996

Room:B2 Location: Anaheim Convention Center

Session Chairperson: C. T. Liu, Metals and Ceramics Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6093

8:30 am Invited ENVIRONMENTAL EFFECTS AND INTRINSIC MECHANICAL PROPERTIES OF NI3Al: E. P. George, C. T. Liu, Metals and Ceramics Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6093 The effects of environment, strain rates, temperature, and boron doping on the mechanical behavior ofNi3A1 are reviewed. Environmental embrittlement now appears to be the main reason for the grain boundary brittlement ofB-free Ni,AI. The embrittlement mechanism involves the generation of atomic hydrogen at the grain boundaries, by either the reductioll of water vapor by AI, or the direct dissociation of molecular H,. Consistent with this ductility increases with increasing strain rate, decreasing temperature, and decreasing amounts ofH,OIH, in the environment. When the embrittlement is suppressed-by testing in H OIH, free environments (approximately 10·'0 torr) the tensile ductility ofB-free Ni ~I increases dramatically (to more than 40% ), and the fracture mode changes to piedominantly transgranular. Boron alleviates environmental embrittlement in Ni AI. Possible mechanisms include, slowing down H diffusion, and altering the surface ~hemistry to interfere with the generation and retention of atomic H. Research sponsored by the Division of Material Sciences, U.S. Department of Energy under contract DE-AC05-84OR214oo with Lockheed Martin Energy Systems, Inc. 8:55 am Invited INTERACTIONS OF WATER AND OXYGEN ON NI,(AI,Ti) AND Nl,Fe SURFACES-EVIDENCE FOR THE CRITICAL ROLE OF COMPETITIVE REACTIONS: Yip-Wah Chung. Department of Materials Sci. & Eng., Robert R McCormick, School of Engineering and Applied Science. Evanston. IL 60208 Recent studies indicate that much of the observed room temperature embrittlement occurring in intermetallics is due to environmental effects. Our research has focused on the specific interactions hetween water vapor and oxygen (the two key active components in air) and Ni (AI, TI) and Ni,Fe surfaces, using standard tools of surface science. Our studies show that when absorbed on to Ni,(AI, TI) surfaces, water indeed can dissociate to produce atomic hydrogen. a key proposed step in the embrittlement process. but only on specific crystallographic planes and above a certain critieal temperature. This result may explain the observation that Ni,AI does not suffer moisture-induced embrittlement at room temperature when solidified in certain orientation. Water also absorbs on Ni,Fe surfaces, dissociating to produce hydrogen above 2ooK. However, the water absorption process is approximately 1000 times slower than those of oxygen. In a normal room temperature ambient (the partial pressure ratio of oxygen to water being 15 to I), oxygen will out-compete water vapor when exposed to fresh Ni,Fe surfaces, thus quenching the hydrogen production process. This may explain its high ductility in regular air ambiance. 9:20 am Invited AN OVERVIEW OF ENVIRONMENTAL EFFECTS IN IRON ALUMINIDES: I. Baker, Thayer School of Engineering, Dartmouth College, Hanover, NH 03755 Iron a1uminides hold great promise for intermediate temperature applications (to approximately 6OO"C) because of their excellent oxidation and sulphidation resistance and reasonable strength. However, they have poor low temperature ductility which is exacerbated by water-containing environment. This paper will present an overview of the environmental effect on iron a1uminides. It will cover the effects of iron aluminum ratio; ternary alloying additions, particularly boron; strain rate and temperature. The mechanisms involved in the water-induced embrittlement will be discussed.

1996 TMS ANNUAL MEETING AND EXHIBITION

9:45 am Invited PREPARATION OF A TI-NIINTERMETALLIC COMPOUND BY USING A DROP SHAFf: Yoshikazu Suzuki, Katsuyoshi Shimokawa, Yoshinobu Ueda and Jiro Nagao, 2- 17 Tsukisamu-Higashi, Toyohira-ku, Sapporo Hokkaido 062, Japan A TI-Ni intermetallic compound could be formed rapidly from TI and Ni powder mixed compact with Self-propagating High-temperature Synthesis. Microgravity experiments using a drop shaft are carried out for the very short available time. This report shows the effect of microgravity or high gravity on formability of an intermetallic compound by using a drop shaft. It is concluded that the melted speeimen forms a porous macrostructure and a nearly amorphous alloy by rapid cooling under microgravity. Otherwise, when the capsule of the drop shaft stops after falling and the melted speeimen solidifies with rapid cooling under high gravity, the specimen forms a uniform and dense structure with better crystallinity. 10:10 am BREAK 10:10 am Invited HIGH TEMPERATURE DEFORMATION AND TERNARY ALLOYING OF NbCr, LAVES INTERMETALLlCS: T. Takasugi, Institute for Materials Reserarch, Tohoku University, Katahira 2-1-1, Aoba-ku, Sendai 980 Japan A study has been conducted to examine the effect of ternary alloying elements on high temperature compressive deformation property of the CIS NbCr, based intermetallics, which have attractive properties as high temperature structural materials because of their high strength, high melting point, low density and good chemical stability. The high temperature deformation of the CI5 NbCr, intermetallics can be improved by two principal alloying methods. One method is based on consideration of site occupation and size factor of the additive elements; in the CI5 lattice, addition ofMo, V and W can promote the synchro shear and resultant dislocation movement. Other method is based on modification of microstructure through alloying; moderate addition of these elements results in a variety of duplex microstructure between the CI5 phase and bee solid solution without forming any intermediate phases, and can enhance the high temperature deformahility. Being concurrent with mechanical observation, lattice property of the additive atoms in the CIS structure was investigated by XRD, TEM HREM and ALCHEMI observations and then discussed in association with the observed mechanical behavior. Also, phase relation between the CIS structure and the bec solid solution in Nb-Cr-X systems was investigated by OM, XRD and TEM observations. 10:45 am Invited EFFECT OF INTERFACIAL MODIFICATION ON MECHANICAL BEHAVIOUR OF INTERMETALLIC MATRIX COMPOSITES: J. M. Yang, Department of Material Science and Engineering, University of California, Los Angeles, CA 90095-1595 Intermetallic matrix composites are currently being developed for variety of high temperature structural applications. One of the most critical issue is the chemical, physical and mechanical incompatibilities between the fiber and matrix. Fiber coating appears 10 be one of the most promising approaches for designing a functional interface to improve the fiber/matrix incompatibility. In this paper the effect of fiber coating on the mechanical behaviour of several intermetallic matrix composites including TI,AI, TI,AINb and NiAI would be discussed. 11:10 am HIGH TEMPERATURE OXIDATION BEHAVIOR OF TIAI-BASED INTERMETALLlCS: M.F. Stroosnijder, J.D. Sunderkotter, V.A.C. Haanappel, Institute for Advanced Materials-Ispra Site, Joint Research Center-European Commission, 21020 Ispra(VA), Italy In spite of their high aluminum content, TIAI-based intermetallics do generally not form long lasting protective alumina scales. After longer exposure times, the scales, which are initially rich in alumina, deteriorate and scales consisting of mixed alumina and titania predominate, with similar high growth rates as pure titania. A great deal of research work. has been undertaken to improve the oxidation behavior of these alloys by adding ternary and quartenary elements. Among other candidates, especially niobium was found to improve the oxidation behavior of titanium aluminides. Nitrogen from the air appears to change the oxidation resistance dramatically. Despite this, in the literature the oxidation mechanism in air is often treated on the basis ofTI-A1-0 thermodynamiCS and kinetics only. In the present conrribution the influence of alloying elements, in particular niobium, and the effect of nitrogen in the oxidizing environment on the high temperature oxidation behavior ofTIAI-based intermetallics will be considered. 11:35 am MECHANISM OF DISINTEGRATION DURING METAL DUSTING: L. Levin, A.Katsman, L. Klinger, T. Werber, Department of Materials Engineering, Technion-Israel Institute of Technology, Technion, Haifa 32000, Israel The process of catastrophic carborization, i.e. "metal dusting", which converts a compact metallic material into fine particles, was analyzed. It was shown that disintegration of the compact metal matrix to a powder is the result of plastic deformation and subsequent fracture accompanying diffusional phase formation in the near-surface layer. The process is controlled by internal stresses arising during phase transformation. Competition between stress generation and relaxation may result in attaining the ultimate strength in the near-surface layer and its fracture. Such a mechanism of metal disintegration can occur in a certain temperature interval dependent on the kinetic and geometrical parameters of the system diffusivities of the alloy's components, the ratio between specific volume of the new and the old phases, and the ultimate plastic strain. The temperature interval for carbon steel disintegration is the carburizing atmosphere, the rate of disintegration, and the period of the disintegration cycle were evaluated and compared with experimental data.

B59

PHYSICAL PROPERTIES OF COMPOSITES III:

Sponsored by: JL SMD-MSD Composite Materials Committee Program Organizer: P.K. Uaw, Materials Science and Engineering Department, The University of Tennessee, Knoxville, TN, 37996-2200; R. Pitchumani, Mechanical Engineering Department, University of Connecticut, Storrs, CT06269-3139; S.O. Fishman, Office of Naval Research, 800 N. Quincy Street, Arlington, VA 22217 Tuesday, AM February 6, 1996

Room: Marquis 1&2 Location: Anaheim Marriott Hotel

Session Chairpersons: S.O. Fishman, Office of Naval Research, 800 N. Quincy Street, Arlington, VA 22217; R. Pitchumani, Mechanical Engineering Department, University of Connecticut, Storrs, CT 06269-3139

8:30am THERMAL BOUNDARY CONDITIONS FOR HETEROGENEOUS MEDIA: Mincheng iin, S.S. Sadhal, Department of Mechanical Engineering, University of Southern California, Los Angeles, CA 90089-1453 For composite materials, the process of homogenization of the physical properties leads to an average representation of the behavior of the material. In many of the analytical developments, such homogenization processes are based on a composite medium of infinite extent. The validity of such an average breaks down in the neighborhood of the boundaries. lbis is a well known phenomenon for the elastic properties of composite materials. lbis paper focuses on the thermophysical properties of such solids, and models have been developed for the representation of thermal boundary conditions in terms of the average properties. In particular, the new results consist of the derivation of a 'thermal constriction resistance' in the boundary region through some exact solutions for the heat conduction equation for a layered composite. An asymptotic analysis of the results gives physical insight into the thermal behavior near the boundary for the Neumann type boundary conditions. In addition, expressions for the equivalent conductance of the boundary regions have been developed. 9:00am

RESISTIVITY OF RETICULATED VITREOUS CARBON (RVC): H. Zahiri, S. Zhong, H.K. Uu, S.X. Dou, Centre for Superconducting and Electronic Materials, University Of Wollongong NSW 2522, Australia Electrical resistivity influences electrical conductivity of RVC materials. The work described in this paper compares two kinds of RVCs. One is made from Polyurethane (PU) foam and the other infused PU foam with Phenolic resin. In addition, a new method is proposed to measure resistivity of porous materials. Results indicated th~t the increase in density could lead to decrease in resistivity of the RVCs. Furthermore, It was found that RVCs which are made from PU have lower reSistivity than those made from infused PU when the density of both RVCs was the same. 9:30am COMPOSITE INTERFACE STUDY BY X-RAY MICRODIFFRATTOMETRY AFTER DIFFERENT QUENCHING TREATMENTS: C. Patuelli. S. Sprio, Dipartimento di Fisica, Universilt di Bologna, VIa Irenrio 46, 40126 Bologna, Italy The residual lattice strain and the crystallite size obtained by X-ray line broadening Fourier analysis using a Rigaku microdiffrattometer were correlated for a better knowledge of the dislocation generation mechanisms at the AI metal-rerarnic reinforce interface after different quenching treatments. Different volume fractions (10,20,23,27%) and morphologies (particles and whiskers) of the SiC reinforcing phase are considered. The results are discussed taking into account the role of the reinforcing phase and thermal treatment in the generation of lattice distortions at the composite interface. 10:00 am

STRUCTURE AND CONTENT MODELLING OF COMPOSITE MATERIALS UNDER THE ELECTRIC DISCHARGE INFLUENCE: R. V. Minakova, ML. Grekova, E. B. Homento, Institute for Problems of Material Science, Ukraine Electrical discharge machining was found to be suitable for materials which can not be treated by usual mechanical cutting. The process is discribed here by a model based on a heterogenious powder material surface exposed to a constant plane heating source of a constant flux over the pulse duration. lbis model has allowed to calculate the volume proporlons and grain size of components in the composite materials according to working conditions of machining. For Cu-W composite material the optimal mass content is W-70%, Cu-30% and grain size changes from 3.5 mkm for finishing machining to 30 mkm for rough one. Theoretical predictions were verified by machining of hard alloy by Cu-W tools with different components content Sirniliar calculations were also carried out for Cu-Cr composite material for electric contacts, working in the power vacuum switchboards. The optimal content is 25% Cr, 75% Cu. These results completely agree with data about loss of heterogeneity of Cu-Cr material in metastable condition when copper content decreasing to 71 %. It happens because of the phase BCC-FCC change bY consecutive substitution chromium atoms for Cu. 10:30 am STRUCTURE DEGRADATION IN THE WORKING LAYER AND THE WORKING CAPACITY OF CONTACTS IN SWITCHBOARD APPARATUS: R. V. Minakova, M.L. Grekova, E.B. Homenko, Institute for Problems of Material Science, Ukraine

B60

The different kind contacts working capacity is reached by using composite materials with heterogenious structure and connected with capillary effect. To clear up the influence of composite materials natum on electric erosion stability in this work it was carried out the comparative investigation of degradation in composite materials on the base of refractory metals, copper and silver degradation in contacts working layers. Irreversible structure changes occuring in these layers under the electric discharge thermal flux influence and secondary structure forming has been shown. The secondary structure is a variable composition lay.~r with heterogeneities and defects forming under influence of whole complex of prOl:esses in the inter electrodes space and working layer. The peculiarities of this structure has been illustrated for contacts from composite materials on the base of W, Mo, Cr, Cu and Ag. Forming of secondary structure in the working layer decreases the electric erosion under single pulse influence. When the operations of switching on - switching off and arc extinguishing repeats many times the electric erosion process consists in secondary structure forming and destruction of working layer in vapour, liquid and solid phases. The relative decreasing of electric erosion and other parameters happen if secondary layer thin and uncontinious. It may be achieved by changing of main elements and special additions content, choice of optimal conditions for using of contact materials.

11:00 am MECHANICAL PROPERTIES OF SQUEEZE CAST EXTRUDED AIISICp COMPOSITES: N. Ramakrishnan, Indian Institute of Technology, Powai, Bombay400 076, India Squeeze cast extruded AVSiCp composites provides an economic material with many potential applications in defence, automobile, etc. The mechanical properties of squeeze cast extruded AVSiClI OpI2Op13Op c0ffil:0sites in a quenched condition, in tension, compression and shear, and hardness were e'.valuated. Fixtures necessary for the evaluation of properties were designed and fabricated as per the ASlM standardS. Strength, modulus, strain hardening index were analYlled as function of volume of reinforcement in composites. The highest improvement in the properties was evident in the tensile mode, in which, elastic modulus and yield strength were improved by 75% and 25% respectively. A large increase in the strain hardening index depicts the extent of strain hardening taking place in the composites. Maximum increase in the brinell hardness of composites was about 35%. Experiments to calculate strain rate sensitivity and the dependence of flow stress of composites on te:mperature are being done, the results of which will also be presented in this paper. Understanding mechanical properties of the composites is focal to their further application and proceSSing. 11:30 am HOT ISOSTATIC PRESSING OF DllAMOND-CONTAINING INORGANIC COMPOSITES: Shaich; Kume, Kazutaka Suzuki, National Industrial Research Institute of Nagoya (NIRIN), I-I Hirak:-cho, Kua-ku, Nagoya, Japan. Haruo Yoshida, National Institute for Advanced Interdisciplinary Research (NAIR), 1-1-4 Higashi, Tsukuba, Ibaraki, Japan Diamond/partially stabilized zirconia (PSZ) composites were fabricated by hot isostatic pressing (HIPing). The fabrication of the composites was successful even under conditions where the diamond stayed thermodynamically metastable. The transformation of the diamond into a graphite in the composites HIPed under 150MPa at II23K for 3h was successfully suppressed. The: monoclinic and tetragonal phases which consist of PSZ remained unchanged in the composites even after HIPing. The diamond/ PSZ composites containing diamond up to 30 vol % were successfully densified. Their densities were lighter than 96% of the dJeorticai value obtained by ROM. Hardness of the composite containing 10 vol% diamond was markedly higher than that of PSZ sintered without diamond.

RADIATION MATERU~LS SCIENCE IN TECHNOLOGY APPLICATIONS I: Radiation Effects in Structural Alloys

Sponsored by: JL SMDlMSD Nuclear Materials Committee Program Organizers: L. K. Mansur, Metals and Ceramics Division, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831-6376; C. L. Snead, Jr., Applied Technologies Division, Brookhaven National Laboratory, PO Box 5000, Upton, NY 11973-5000 Tuesday,AM February 6, 1996

Room: Grand H Location: Anaheim Marriott Hotel

Session Chairperson: L. K. Mansur, Metals and Ceramics Division, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831-6376

8:30 am Invited MECHANISM EVALUATION OF RADIATION-INDUCED SEGREGATION OF MAJOR AND MINOR ELEMENTS IN AUSTENITIC STAINLESS ALLOYS: E. P. Simonen, S. M. Bruemmer, Pacific Northwest Laboratory, PO Box 999 MS P8-15, Richland, WA 99352 Radiation-induced segregation (RIS ) occurs in austenitic stainless alloys by enriching Ni and Si and by depleting Cr and F
1996 TMS ANNUAL MEETING AND EXHIBITION

RIS are found to be much less than calculated using the interstitial RIS assumption. In the present paper, an integrated understanding of both major and minor elemental RIS in austenitic stainless alloys is developed. Unique, minor element interaction with defects are identified to rationalize the observed lack of Si and P RIS at grain boundaries. This work was supported by the Materials Sciences Branch, BES, U.s. Department of Energy, under Contract DE-ACQ6. 76RLO 1830. 9:00am MODELING OF RADIATION-INDUCED SEGREGATION IN AUSTENITIC Fe-NI-Cr ALLOYS USING LATTICE RATE THEORY: T. R. Allen, G. S. Was, The University of Michigan, Department of Nuclear Engineering, Ann Arbor, MI 48109 A significant set of radiation-induced segregation (RIS) measurements from proton irradiated and Ni ion irradiated austenitic iron-base and nickel-base alloys now exists. The compositions studied span the iron-rich (304 stainless steel) to nickel-rich (Inconel) regions of the Fe-Cr-Ni phase diagram. Modeling using the Perks vacancy-driven continuum RIS model has shown that to accurately predict segregation trends, alloy specific diffusion parameters must be used. This requirement for alloy-specific diffusion parameters limits the ability of the Perks model to predict segregation in new alloy systems. Additionally, some features of the segregation measurements cannot be explained using simple Inverse Kirkendall behavior. In order to improve on the ability to predict RIS in new alloy systems, a new model for calculating RIS is presented, using lattice rate theory with diffusion parameters based on local atomic configuration. The advantages to a lattice rate model calculation are that migration energies for multiple alloy systems are not needed and that the effects of local order can easily be incorporated into the calculations. 9:20am NEUTRON IRRADIATION AND INTERGRANULAR FRACTURE IN V-10 wt. % TI ALLOYS UNDOPED AND DOPED WITH P OR S: D. Y. Lvu, Chonju Technical College, Korea; T. E. Bloomer, O. Unal, J. Kameda, Ames Laboratory, Iowa State University, Ames, IA 50011 The effect of neutron irradiation (10'" nlm' at 400°C) on the deformation and fracture behavior in V-20 wI. % 11 alloys undoped and doped with P or S has been studied by means of a small punch (SP) testing method in conjunction with scanning Auger microprobe (SAM) and transmission electron microscopy (TEM) analyses. SP tests on unirradiated specimens indicated that the addition of S or P slightly increased the ductile-brittle transition temperature (DBTI). Impurity doped alloys showed a little stronger hardening during the irradiation. The neutron irradiation led to a large increase in the OBIT of a undoped alloy while it did not affect the OBIT in the impurity doped alloys. All the unirradiated and irradiated alloys exhibited smooth or rough intergranular fracture surfaces at the lower shelf regime. The formation of sulfides or phosphides, and segregated S or P at grain boundaries were examined by SAM and TEM. The irradiation effect on the intergranular fracture is discussed in light of defect trapping at incoherent precipitate interfaces in the grain interior and the migration of defect-solute complexes to grain boundaries. The beneficiaUdetrimentai effects of impurities are also discussed in developing fusion reactor materials. 9:40am EFFECTS OF RADIATION AND STRAIN ON INTERGRANULAR FRACTURE OF AUSTENITIC STAINLESS STEELS: E. P. Simon en, S. M. Bruemmer, Pacific Northwest Laboratory, P.O. Box 999 MS PS-15, Richland, WA 99352 Intergranular (IG) fracture of irradiated austenitic stainless steels can be promoted by either slow straining at temperatures above 400°C in inert environments or by slow straining at lower temperatures in water environments. The synergistic roles of strain and environment on IG fracture are evaluated in the present paper using an Ashby map approach for deformation and fracture kinetics. During post-irradiation straining, radiation retards kinetics for transgranular (TG) processes and provides an opportunity for dominance of IG processes. During in situ straining, radiation creep accelerates transgranular deformation. However, flow of radiation-produced defects to grain boundaries is expected to accelerate IG sliding and promote IG fracture. Radiation effects on IG and TG deformation and fracture are calculated in the present analysis to establish criteria for IG fracture relevant to observed component cracking for both boiling water reactors and pressurized water reactors. This work was supported by the Materials Sciences Branch, Department of Energy, under Contract DE-AC06-76RLO-1830. 10:00 am BREAK 10:10 am Invited RADIATION EFFECTS IN MATERIALS FOR ACCELERATOR-DRIVEN NEUTRON TECHNOLOGIES: M. S. Wechsler. C. Lin, North Carolina State University, Raleigh, NC 276957909 Accelerator-driven neutron technologies include facilities for neutron scattering research, accelerator transmutation of waste, accelerator production of tritium (APT), and accelerator-based conversion of Pu. In these systems, high-energy protons (E > 100 MeV) strike a heavy metal target, producing spallation neutrons with energies extending up to the incident proton energy. Radiation damage (displacements, He, and transmutation products) due to the incident protons and the spallation neutrons have heen calculated using the LAHET Code System. For W, a frequently used target material, the displacement cross section is calculated to reach 7300 barns at 800 MeV for both protons and neutrons. For a projected 800 MeV I MW neutron source, a peak displacement rate of 93 dpa/yr is calculated in W due to the incident protons and 8 dpa/ yr due to spallation neutrons, with 50% of the displacements due to neutrons of E > 20 MeV. The limited experimental information on spallation-radiation-damaged materials will be presented. Comparison with reactor irradiated materials will be discussed.

1996 TMS ANNUAL MEETING AND EXHIBITION

10:SOam MODELING FRACTURE TOUGHNESS INDEXED TEMPERATURE SHIFTS IN LOW ACTIVATION MARTENSITIC STAINLESS STEELS: G. R. Odelle, Departments of Mechanical Engineering and Materials, University of California, Santa Barbara, CA 93106 Low activation martensitic steels are attractive candidates for fusion applications, but may experience significant embrittlement as a consequence of hardening by fine scale precipitate and defect cluster features that form under irradiation. Fracture toughness (K) as a function of test temperature (T) curves reflect the changing nature of the fracture event with increasing T. At very low T (lower shelf and knee) stressed volume controlled cleavage fracture is initiated in the local plastic zone in front of the crack tip while the specimen is grossly elastic. At higher T (lower transition regime) cleavage occurs following general yielding of the specimen; at still higherT limited stable crack growth takes place prior to cleavage initiation (upper transition regime) by a mixture of stress controlled cleavage and strain controlled microvoid coalescence; and, finally, at high T, fracture takes place by a ductile tearing/microvoid coalescence mechanism. Detailed micromechanical models are needed to relate the crack tip stress strain fields, mediated by the constitutive law, to the local stress-strain-volume conditions leading to fracture. However, for the cleavage regime this complexity might be short circuited by using an empirically based hypotheses that the local conditions for fracture are independent of irradiation and test temperature and that there is a one-ta-one correlation between strain hardening and yield stress. In this case cleavage initiation in a given material is simply controlled by the yield stress: thus K .. (T,,> = K.- (Tunirr) when 0,.,=0,...., where aTK=T.. - T_ (To. > T......) represents the temperature shift atsome index of toughness (e.g., K = 100 MPam) The micromechanical basis for this hypothesis is described and it is shown that predictions of the Simple model are in agreement with limited data from the literature. The results also suggest that T decreases rapidly with T..... and increases non-linearly with y. Various limitations of the model are examined. Supported by the Office of Fusion Energy, U. S. Department of Energy Grant DEFG03-94ER54275. 11:10 am REFUTATION OF RADIATION SOFTENING IN PRECIPITATION· HARDENED ALUMINUM ALLOYS: K. Farrell, S. T. Mahmood, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831-6376 (Present address: General Electric Vallecitos Nuclear Center, Pleasanton, CA 94566) The 6000-series aluminum alloys contain Mg and Si and are hardened by a fine precipitate of Mg,Si, produced by solution treatment, quenching and tempering. Normally, the precipitation-hardened material is stable under displacive irradiation. However, there are two claims of radiation-induced softening, one during neutron irradiation, the other during high energy proton bombardment, in which tensile strengths are reduced to the fully annealed levels and the Mg,Si phase disappears. This paper descrihes new neutron and proton irradiations that disprove the claims and indicate that the reported softening is caused by thermal overaging due to inadequate control of irradiation heating. Research sponsored by the Division of Materials Sciences, U.S. Department of Energy under contract DE-AC05-840R21400 with Lockheed Mattin Energy Systems. One of the authors (S. T. Mahmood) was supported by an appointment to the Oak Ridge National Laboratory Postdoctoral Research Program administered by the Oak Ridge Institute for Science and Education. 11:30 am FORMATION OF Kr BUBBLES IN ZIRCONIUM ALLOYS: L. Pagano, A. T. Motta. Department of Nuclear Engineering, The Pennsylvania State University, University Park, PA 16802 Zirconium alloys are remarkably resistant to void swelling and bubble formation, but bubbles can be formed during irradiation when stabilized by insoluble gases. Several zirconium alloys, including Zircaloys and pure Zr, were irradiated with 100 keV Kr ions in the HVEMffandem facility at Argonne National Laboratory at temperatures ranging from room temperature to 800°C. The objective was to conduct a systematic study of the conditions for bubble development under irradiation, and study the influence of alloy microstructure and composition. Results indicate that a gas-driven regime, (where a fine dispersion of 30-60 A. bubbles is seen), exists from room temperature up to 500°C. At higher temperatures (above 700°C) larger faceted bubbles (about 500°) were observed, as a result of bubble mobility and coalescence. The results are discussed in terms of existing bubble growth models.

STRUCTURE AND MORPHOLOGY OF EPITAXIAL THIN FILMS SESSION III: Strain Sponsored by: EMPMD Thin FUms & Interfaces Committee Program Organizer: Dr. David E. Jesson, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831-6030 Tuesday,AM February 6, 1996

Room: Orange County 4 Location: Anaheim Marriott Hotel

Session Chairperson: D. E. Jesson, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831-6030

8:30 am Invited STRESS·DRIVEN SELF-ORGANIZATION IN THREE-DIMENSIONAL ''QUANTUM DOT' STRUCTURES: J. Tersoff, IBM T. J. Watson Center, PO Box

B61

218, Yorktown Heights, NY 10598 In some systems, e.g. InAs on GaAs(OOI), pseudomorphic growth of strained material leads to formation of islands with a relatively narrow size distribution. Such islands are considered possible candidates for "quantum dot" devices. Moreover, several groups have observed that when such islands are grown on successive layers of a multi-layer structure, there may be strong vertical correlations between islands in different layers. I will discuss how the distribution of islands evolves with the growth of successive layers. Not only do the islands become vertically correlated, but in addition their lateral arrangement and size distribution is predicted to become more uniform. Comparison will be made with experimental data from C. Teichert et al. 9:00 am Invited A MECHANISM OF COHERENT ISLAND FORMATION DURING HETEROEPITAXY: Andrew Zangwill, School of Physics, Georgia Tech, Atlanta, GA30332 1be phenomenon of coherent (undislocated) island formation during heteroepitaxy is discussed using the results of recent Monte Carlo simulations as a frarneworlc. Through an approximate treatment of the elasticity problem, the growth model permits dynamic stress relaxation at the free surface of both 2D and 3D islands although misfit dislocations are disallowed by construction. It is assumed that misfit strain reduces the energy harrier to atom detachment from island edges. Coherent islands form above a critical value of misfit that depends on both growth conditions and the material parameters. Both the lateral size and width of the size distribution become smaller as misfit increases. 1be latter occurs despite the complete absence of substrate-mediated elastic repulsion between islands. Comparison is made with other theories of island formation. 9:30am COHERENT STRANSKI·KRASTANOV GROWTH: STUDY OF THE FORMATION OF 3D ISLANDS VIA THE POWER SPECTRAL DENSITY: G. C. Hsueh, Department of Chemical Engineering, C.M. Reaves, S. P. DenBaars, Materials Department, W. H. Weinberg, Center for Quantized Electronic Structures (QUEST), University of California, Santa Barbara, CA 93106 1be evolution of the power spectral density (PSD) corresponding to the development of coherent 3D InP islands on GalnP/GaAs( 1(0) is investigated. Analysis of the PSD of the pseudomorphic wetting layer follows the correspondence of wavelength scaling laws with associated atomic proccesses on unstrained and strained surfaces. Upon the formation of coherent 3D InP islands, the qualitative nature of the PSD changes profoundly, with the appearance of short-wavelength oscillations, indicative of a narrow distribution of island sizes. 1be PSD of the surface displaying 3D InP islands is interpreted in terms of particle scattering and interference. By assuming an analytic approximation to the surface morphology, the island size distribution and spatial correlation are extracted and deconvoluted from the PSD. Specifically, the island size distribution is modeled by an ensemble of cylindrical functions, while the spatial correlation is accounted through the use of a "hard-spheres" radial distribution function. 9:50am THE STRAINED STATE OF HETEROEPITAXIAL ISLANDS ON MISFITIING SUBSTRATES-FINITE ELEMENT CALCULATIONS AND EXPERIMENTS: S. Christiansen, M. Albrecht, H. Michler, H. P. Strunk, A. Borbely, Institut fur Werkstoffwissenschaften VII, Universitat Erlangen-Nurnberg, Cauerstr. 6, 91058 Erlangen, Germany; T. Ungar, Eotvos University Budapest, Institute for General PhYSics, POB 323, Budapest VIII, Hungary; B. Dietrich, Institut fur Halbleiterphysik, Walter Korsing Ring 2 15230, FrankfurtlOder, Germany Several attempts exist to calculate the strained state in nanoscaled laterally limited structures by analytical, semi-empirical, and numerical methods. However, the experimental confirmation of these data encounters several problems: high resolution methods, such as convergent beam electron diffraction and quantitative high resolution transmission electron microscopy have to deal with inhomogeneous thin foil relaxation of single strain components, while laterally averaging methods such as high resolution x-ray diffraction and Raman spectroscopy are limited in local resolution. In this paper, we compare both classes of methods for the case of facetted pyramidal islands obtained by solution growth ofGe on Si(OOI). The Raman measurements yield the average strain distribution within the layer, and the x-ray measurements yield the average strain distribution within the substrate. We compare these data to the strain distribution calculated with the three-dimensional fmite element method for the substrate and for the islanded Ge layer. Steep strain gradients are present in certain regions of the structure and are corroborated by experiments using locally resolving convergent beam electron diffraction. 10:10 am BREAK 10:30 am InvIted NEW INSIGHTS INTO THE KINETICS OF THE STRESS-DRIVEN lD·TO·3D TRANSITION: K. M. Chen, D. E. Jesson, S. J. Pennycook, T. Thundat, R. J. Warmack, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6031 1be stress-driven 2D-to-3Dtransition associated with the Stranski-Krastanow growth mode has received considerable attention. In this worlc, the influence of temperature, film thickness, misfit strain, atomic steps, and supersaturation on the kinetics of the transition have been systematically investigated with RHEED and AFM. Our studies have revealed a number of surprising features, reflecting the interplay between kinetics and energetics during the initial and late stages of island growth. We establish that an energy harrier exists to the formation of 3D islands, and map the kinetic pathway to {SOl} faceted islands. It is demonstrated that die energy barrier exists before {50 I } faceting. We interpret the transition in terms of positive step free energy, which explains the observed sensitivity of the results to strain and temperature. In die late stages of growth, we observe quite different behavior including the fissioning of elongated

B62

islands (inverse Ostwald ripening), and we present a model for die remarkably narrow distribution of observed island sizes. This research was sponsored by the Division of Materials Sciences, U.S. Department of Energy, under contract No. DE-AC05-840R21400 with Lockheed Martin Energy Systems. 11:00 am IMPACT OF DEPOSITION SURFACE ON THE FORMATION OF SELF·ASSEMBLED InP EPITAXIAL ISLANDS: C. M. Reaves, G. C. Hsueh, W. H. Weinberg, S. P. DenBaars, Center for Quantized Electronic Structures, University of California, Santa Barbara, CA 93106 The formation of self-assembled, defect-free epitaxial islands in the StranskiKrastanov growth mode has been obselved for a wide range of semiconductor systems. Although various studies have focused on understanding the impact of growth conditions on these islands, one growth parameter that has not received much attention is the nature of the deposition surface. We have studied die formation of InP islands on a range of surfaces with the GaAs lattice constant (a lattice mismatch of 3.7 %). Comparisons between GaAs and GaInP surfaces, between rough and smooth GaInP surfaces, and between (IOO) and (311) surfaces will be presented. For example, InP islands grown on a GaInP(311)A surface exhibit a single distribution of islands of height 21 + 7 A and base diameter 850 + 120 A, whereas islands grown on a GaInP (100) surface exhibit a bimodal distribution, one mode of height 31 + 12 A and base diameter of 1320 + 200 A and anodler of height 280 + IS A and base diameter of 1430 + 120 A. The impact of the surface on transport, >nucleation, and growth processes will be discussed. 11:20 am ON THE INSTABILITY OF EPITAXIALLY-STRAINED ANISOTROPIC THIN FILMS: Jong K. Lee, Stephen A. Hackney, Department of Metallurgical and Materials Engineering, Michigan Technological University, Houghton, MI 49931 The morphological instability of an epitaxially-strained thin film is studied by means of a discrete atom method, which is predicated upon Hookean atomic interaction and Monte Carlo diffusion. Cubic systems of anisotropic ratios, A (= 2C../(C II - C,,)} = 0.5, 1, and 2.33 are analyzed in dislocation-free, two dimensional crystals. If the surface normal is parallel to an elastically soft direction (> =100 with A =2.33), a planar thin film is found to be relatively stable at a low temperature, but prone to create islands at a high temperature apparently with the aid of thermal fluctuations. If the surface normal orientation is elastically hard or isotropic, bowever, a planar film tends to create a wavy surface with characteristic wavelengths. Stress analysis shows that within a hilltop, the strain energy density decrease!> as die free surface is approached from inside, but increases within a valley. The instability of the film-substrate interface and its elastic interaction with die free surface are also discussed. 1be research was supported by the U.S. Dept. of Energy under Grant DE-FG0287ER45315. 11:40 am MOLECULAR DYNAMICS SIMULATION OF STRESS RELAXATION IN HETEROEPITAXIAL FILMS: Richard W. Smith, David J. Srolovitz, Department of Materials Science and Engineering, The University of Michigan, 2300 Hayward, Ann Arbor, MI 48109 Molecular dynamics simulations were performed to investigate die development of voids and stress during the growth of heteroepitaxial films. During low temperature deposition of atoms which possess a larger lattice constant than the substrate material, a compressive stress develops in the first few layers of the film. Void formation and dislocation formation were found to each playa significant role in relaxing the misfit stress. Misfit stresses can be relieved by the flexure of die free surfaces of the voids throughout die film. As the lattice constant of the deposited atoms is increased, die critical thickness for dislocation formation decreases and for sufficiently large mismatches, dislocations form immediately at the interface between film and substrate. While dislocations normally form at the film-substrate interface when no voids are present, when voids are present dislocations were found to be preferentially located within the voids. However, if the misfit strain were very large, dislocations were found to lie preferentially at the interface. Dislocations at the interface move rapidly along the interface in the absence of voids. A video tape showing stress and microstructure evolution during film growth will be presented.

SYNTHESIS AND PROCESSING OF NANOCRYSTALLINE POWDER II: Condensation Synthesis Techniques Sponsored by: MDMD Powder Metallurgy Committee and FEMS (Federation of European Materials Societies) Program Organizers: Dr. David L. Bourell, The University of Texas at Austin, Materials Science & Engineering, MC C2201, Austin TX 78712; Dr. Liisa KuhnSpearing, Laboratory for the Study of Skeletal Disorders and Rehabilitation, Harvard Medical School, Children's Hospital, 300 Longwood Avenue, Boston MA 02115; Professor Dr. Herbert Gleiter, Karlsruhe Research Center, P.O. Box 3640, D-76021 Karlsruhe, Federal Republic of Germany Tuesday,AM February 6, 1996

Room: Grand G Location: Anaheim Marriott Hotel

Session Chairperson: Dr. Liisa Kuhn-Spearing, Laboratory for the Study of Skeletal Disorders and Rehabilitation, Harvard Medical School, Children's Hospital, 300 Longwood Avenue, Boston MA 02115

1996 TMS ANNUAL MEETING AND EXHIBITION

8:30am NANOSCALE MAGNETIC MATERIALS - AN OVERVIEW OF PROCESSING AND PROPERTIES: Robert D. Shull, Magnetic Materials Group, Materials Science and Engineering Laboratory, National Institute of Standards and Technology, Gaithersburg MD 20899 Nanocrystalline materials represent one of the most active laboratories in recent times for the atomic tailoring of materials with specific properties and properry combinations. For magnetic materials, new magnetic behavior, enhanced magnetic properties, and unusual property combinations are observed wben some critical length scale (e.g., grain size or separation distance) is reduced to dimensions comparable to the magnetic exchange length. In addition, the properties may depend upon whether the reduction in size of this critical length was in I, 2, or 3 directions. The preparation of such materials also presents special problems, largely because of the magnetic dipolar attraction between the ferromagnetic species. For composite materials, a special subset of nanostructured materials, the preparation route is particularly important. In this presentation a review will be presented of many of the techniques which have been found successful for the preparation of nanometer-scale magnetic materials and how the nanometer dimensionality effects their magnetic character. 9:05am SYNTHESIS AND HIGH MECHANICAL STRENGTH OF AL-BASED ALLOYS CONSISTING MAINLY OF NANOSCALE CRYSTALLINE, QUASICRYSTALLINE AND AMORPHOUS PARTICLES: Akihisa Inoue, Institute for Materials Research, Tohoku University, Sendai 980-77 Japan The application of rapid solidification to AI-based alloys was found to cause the formation of nanoscale mixed structure consisting of amorphous plus nanogranular fcc Al phases, nanogranular icosahedral plus Al phases, and coexistent nanogranular amorphous plus Al phases. The particle size is about 3 to 5 nm for the nanogranular Al phase, 30 to 50 nm for the icosahedral particle and about \0 nm for the nanogranular amorphous phase. These mixed phase alloys exhibit high mechanical strength of 1300 to 1560 MPa combined with good bending ductility. It is to be noticed that the strength level exceeds largely that for AI-based amorphous single phase alloys. The nanoscale Al particles precipitate via a growth mechanism from the amorphous matrix, but the icosahedral and amorphous particles precipitate as a primary phase from supercooled liquid during rapid solidification. The detailed mechanism for the formation of the new nanophase structures as well as their nanostructural feature will be descrihed. 9:30am SYNTHESIS OF NANOCRYSTALLINE ALLOY POWDERS: S.A. Pirzada, T. Yadav, Nanomaterials Research Corporation, 10960 N. Stallard Place, Tucson AZ 85737 Synthesis of nanocrystalline alloy powders was investigated using a novel thermal process. This thermal process continuously vaporizes the metallic precursors and uses Joule-Thompson expansion to control the nucleation process. The design has enabled the synthesis of nanosize powders of several commercially important alloys, such as NiAI, FeNi, and Fen. The powders produced were characterized by various techniques for phases, size, morphology, size distribution, surface area, etc. The powder size typically ranged from 20-50 nm. This is a scaleable process to produce nanosized materials and is economically attractive. The design and the experimental results will be discussed.

clusters embedded in a co-evaporated matrix can be prepared by this technique. After a brief review of cluster production, analysis, and deposition techniques, the specific nucleation and growth processes which govern the formation of cluster assembled films will be presented. Then, some characteristic examples (covalent: c, Si, and magnetic metals: Fe, Co, Ni) of cluster assembled films and cluster embedded in matrices, with original structures and properties will be reported. 11:10 am SYNTHESIS TECHNIQUES, MICROSTRUCTURE AND MECHANICAL PROPERTIES OF NANOSTRUCTURED CERAMICS: Horst Hahn, Technical University Darmstadt, Department of Materials Science, Thin Films Division, Hilpertstra8e 31ID, 64295 Darmstadt Germany The synthesis of nanostructured materials using gas condensation processing is well established. The current knowledge for the synthesis of nanostructured ceramics (oxides, nitrides, carbides) including a novel technique, called chemical vapor condensation (CVC), using pyrolysis of chemical precursors, will be reviewed. Due to the ultrafine grain size and the narrow grain size distribution, sintering and creep deformation are enhanced and are observed at temperatures below II2 TM' Various models of creep deformation and superplasticity and their extension to the nanometer regime will be discussed with an emphasis on a new model which is based on grain boundary sliding as the rate controlling process. 11:35 am OXIDATION AND GAS ADSORPTION IN AIR BY NANOCRYSTALLINE NICKEL AND COPPER SYNTHESIZED BY SPUTTERING: R.L. Holtz, GeoCenters, Inc., 10903 Indian Head Hwy., Fort Washington MD 20744; V. Provenzano, Materials Science and Technology Division, Code 6323, U.S. Naval Research Laboratory, Washington, DC 20375 A key processing issue for nanocrystalline metals is their reactivity due to the high specific surface area, and how this reactivity effects the handling of nanocrystalline powder. Exposure of these materials to air results in Significant adsorption of gases, rapid initial rates of oxidation, and in many cases spontaneous ignition. To quantify the reactivity of nanocrystalline metals, and to better understand these phenomena and the implications for processing, we have synthesized nanocrystalline Ni and Cu and measured the weight gain as a function of time while exposed to room air. The n-Cu and n-Ni with particle sizes of around 5 nm were produced by the sputtering method of evaporation-condensation, which yields the cleanest possible starting material. We analyze the air weight gain data in terms of gas adsorption and oxidation processes, and contrast the results for the nanocrystalline metals with micron-scale Ni and Cu powders.

TANTALUM: High Strain Rate Deformation: Session III Sponsored by: SMD Refractory Metals and Materials Committee and Jt. MDMDIEPD Synthesis, Control and Analysis In Materials Processing Program Organizers: Andrew Crowson, U.S. Army Research Office, Research Triangle Park, NC; Edward S. Chen, U.S. Army Research Office, Research Triangle Park, NC; Prabhat Kumar, Cabot Corp, Boyertown, PA; Willam Ebihara, Picatinny Arsenal, Picatinny, NJ; Enrique J. Lavernia, UC Irvine, Irvine, CA

9:55 am BREAK 10:20 am PRODUCTION AND CONSOLIDATION OF NANOPHASE METALLIC AND CERAMIC POWDERS: L. He, J.-S Yin, E. Ma, Department of Mechanical Engineering, Louisiana State University, Baton Rouge LA 70803 Mechanical milling/alloying has been employed to produce nanophase metallic materials such as metals (e.g., Fe), intermetallics (e.g., Fe,AI), and composites (e.g., Fe-Cu). An aerosol synthesis route is being developed for nanophase ceramics (e.g., nO,). These inexpensive powder synthesis routes yield relatively large quantities ofnanophase powder materials for consolidation studies. A sinter forging process is developed to consolidate the powders into dense bulk samples while maintaining nanoscale microstructures. The process is effective for full-density processing not only for ceramics, but also for nanophase metallic materials prepared by mechanical alloying. In addition, nanophase ceramic-based composites (e.g., AI,O,tAI) and their sintering densification procedures have also been investigated. Preliminary results of the modeling of the densification processes and mechanical properry measurements of consolidated bulk samples will also be reported. 10:45 am FORMATION AND PROPERTIES OF NANOSTRUCTURED MATERIALS BY THE LOW ENERGY CLUSTER BEAM DEPOSITION TECHNIQUE: A. Perez, P. Melinon, V. Dupuis, J. Tuaillon, B. Prevel, JP. Perez, V. Paillard, Department de Physique des Materiaux, Universite Claude Bernard LYON I, F69622 Villeurbanne, France Intense and stable beams of clusters of various materials in the size range from few atoms to few thousands of atoms can be produced using the inert gas condensation sources. In the gas phase, depending on the source parameters, these clusters generally exhibit specific atomic and electronic structures interesting to stabilize to form cluster assembled materials with original properties. For this purpose the low energy cluster beam deposition technique has been recently developed. It consists in depositing selected mass distributions of neutral clusters having the very low energy gained in the supersonic expansion at the exit of the source. In this case, clusters do not fragment at the impact upon the substrate leading to the formation of controlled nanostructures. Both nanocrystalline films grown by random stacking of incident clusters or individual

1996 TMS ANNUAL MEETING AND EXHIBITION

Tuesday, AM February 6,1996

Room:A4-5 Location: Anaheim Convention Center

Session Chairpersons: Rodney Clifton, Brown University, Providence, RI; Mark Meyers, University of California at San Diego, La Jolla, CA

8:30am HIGH STRAIN-RATE, mGH-TEMPERATURE RESPONSE OFTANTALUM: S. Nemat-Nasser, CEAM, Department of AMES, UCSD, La Jolla, CA 920930416 Recent advances in high strain-rate, high-temperature experimental techniques have allowed the measurement of the isothermal (quasi-isothermal) flow stress of many ductile materials, over a broad range of strain rates and temperatures. With this technique, we have studied the response of tantalum at strain rates from a few hundred to 40,0001 s, and at temperatures of -200 to 750°C. The flow stress under adiabatic conditions is compared with the flow stress under isothermal conditions, at prescribed strain rates and initial temperatures. The experiments have revealed a dramatic change in the workhardening of tantalum, when the temperature exceeds a critical value (T,), estimated to be 150-175°C. Above this T" the workhardening rate increases dramatically, such that it compensates for the thermal softening accompanying the high strain-rate plastic flow. Hence, the material continues to deform adiabatically at high strain rates, at non-diminishing flow stress. Experiments with the hat-shaped specimens verify this fact. Preliminary TEM observations suggest that a large number of dislocation loops are created when screw dislocations jog. It is speculated that high temperature activates this mechanism, increaSing the number of loops which can then serve as barriers to dislocations, leading to greater workhardening. 9:00am HIGH-STRAIN, HIGH-STRAIN-RATE BEHAVIOR OF TANTALUM: V.F. Nesterenko, Department of AMES, UCSD, La Jolla, CA, 92093; M.A. Meyers, J.C. LaSaivia, Department of AMES, UCSD, La Jolla, CA, 92093; Institute for

B63

Mechanics and Materials. M.P. Bondar, Lavrentyev Institute of Hydrodynamics, Russian Academy of Science, Novosibirsk, 630090 Russia; Y.I. Chen, Department of AMES, UCSD, La olla, CA 92093; Y.L. Lukyanov, Lavrentyev Institute of Hydrodynamics, Russian Academy of Science, Novosibirsk, 630090 Russia Ta was subjected to large strains (10- 10) at high strain rates (E 100s") using the radial collapse of a thick-walled cylinder accelerated by a co-axially placed explosive charge. The resulting strains and temperatures achieved produced significant microstructural Changes: (i) elongated dislocation cells (r<2, T <600K);(ii) subgrains (2 10,1'>1000K). Grain-scale localization produced by anisotropic plastic flow and localized recovery and recrystallization was observed at strains y>4. Plasticity analysis predicts significant residual tensile "hoop" stresses near the central hole region upon unloading; profuse ductile fracturing of the Ta was observed within this region. Supported by the U.S. Army Research Office. 9:20am TANTALUM MICROSTRUCTURES FOR HIGH STRAIN RATE DEFORMATION: SHOCK LOADING, SHAPED CHARGES AND EXPLOSIVELY FORMED PENETRATORS: LE. Murr, S. Pappu, CoS. Niou, C. Kennedy, Department of Metallurgical and Materials Engineering, The University of Texas, EI Paso, TX 79968; M.A. Meyers, Department of AMES, UC San Diego, Lalolla, CA 92093 In applications of tantalum involving shaped charges and explosively formed penetrators (EFP), a strong shock is frrst imposed on the liner. In the case of shaped charge slug formation, the residual microstructures as observed in transmission electron microscopy (TEM) are considerably different from those in an EFP. Furthermore, the EFP microstructures are distributed around the EFP slug, while those in the shapedcharge slug are uniform throughout the slug. This uniformity, however, includes dynamic recovery and recrystallization near the slug center with elongated dislocation cells and more regular cells radiating outward from the slug center. These same microstructures are distributed around the EFP slug, with a much lower incidence of dynamic recrystallization. There is no evidence for deformation twins in the shaped charge or EFP microstructures, but very noticeable twinning has been observed in plane-wave shock loaded 'Ill around 45 GPa peak pressure. Systomatic study of twinning suggests that twins do not form below about 30 GPa peak pressure, and this may account for their absence in the EFP. In the shaped charge slug, the actual post-shock deformation completely obliterates any twins as a consequence of dynamic recovery and recrystallization. Research supported in part by U.S. Army Contract DAAA2I-94-C-0059 through Picatinny Arsenal (ARDEC). 9:40am mGH STRAIN RATE DEFORMATION BEHAVIOR OF TANTALUM: R. W. Armstrong, X.I. Zhang, University of Maryland, College Park, MD 20742; C. Feng, U.S. Army Research and Development Center, Dover, NJ 07801; J.D. Williams, University of Illinois, Urbana-Champaign, IL 61801; F.J. Zerilli, Naval Surface Warfare Center Indian Head Division, Silver Spring, MD 20903 Split Hopkinson Pressure bar measurements have been made at ambient temperature on different tantalum materials and are related to predictions made with a dislocation mechanics based constitutive equation analysis (J.D. Williams et aI. "Dynamically Deformed Thntalum: Stress-Strain Measurements, Dislocation Microstructures, and Constitutive Equation Consequences" ,4th Intern. Symp. on Plasticity Preprin!, 1993, Balto. MD). Items of interest are evidence of added dislocation generation at SHPB strain rates and, also, the role of interstitial solutes on the strain hardening behavior and ductility (F.I. Zerilli, R.W. Armstrong, "Description of ThntaIum Deformation Behavior", J. Appl. Phys. 68,1580,1990). 10:00 am BREAK 10:15 am VERY mGH STRAIN RATE BEHAVIOR OF TANTALUM: Kevin Duprey, Rodney Clifton, Division of Engineering, Brown University, Providence, RI 02912 Pressure-shear plate impact experiments are performed to determine the flow stress of tantalum at very high strain rates on the order of 10'8". These results are used to determine if there is a change in the rate controlling mechanism of the flow stress of tantalum at high strain rates, as is observed in other metals. The tantalum used in these experiments is examined using transmission electron microscopy to determine if there is any texturing of the material, and also to examine the dislocation structure both before and after the deformation. 10:45 am MECHANICAL PROPERTIES AND CONSTITUTIVE RELATIONS FOR TANTALUM AND TANTALUM ALLOYS UNDER HIGH-STRAIN·RATE DEFORMATION: Shoo Rong Chen, G.T. Gray III, Mail Stop G755, Los Alamos National Laboratory, Los Alamos, NM 87545 Tantalum and its alloys has received increased interest for ballistic applications. The stress-strain behavior of several tantalums of various compositions and manufacturing sources were investigated as a function of temperature from 17K to 1273K, and strain rate from Io-'s" to 8000s". The yield stress was found to be sensitive to the test temperature, the impurity and solute contents; however, the strain hardening remained very similar for "pure" 'Ill but increased with alloying. Powder-metallurgy tantalum with various levels of oxygen content produced via different processing paths were also investigated. Similar mechanical properties compared to conventionally processed tantalums were achieved. Constitutive relations based upon the Johnson-Cook, the ZerillArmstrong and the Mechanical Threshold Stress (MTS) models were evaluated. Parameters for fitting these models to various tantalurns and its alloys were derived. The

B64

capabilities and limitations of each mo:!el for large-strain applications will be examined. Work performed under the auspices of the U.S. Department of Energy. 11:05 am MECHANICAL BEHAVIOR AND MICRO-TEXTURE ANALYSES OF ANNEALED TANTALUM AND TANTALUM-TUNGSTEN ALLOY PLATE: David H. Lassila, Erik Randish, Gilbert F. Gallegos, University of California, Lawrence Livermore National Laboratory, P.O. Box 808, L-35, Livermore, CA 94550 At LLNL a wide range of mechanical testing was performed on numerous annealed tantalum and tantalum-tungsten alloy plate materials. These tests include, dynamic tensile testing, uniaxial testing in compression at strain rates of Io-'s" and 3000s". In general we found the stress-train response of all of the materials to be reasonably well behaved., i.e. test results were reproducible for a given plate and general trends in behavior such as work hardening as a function of loading direction were consistent. However, on a smaller scale the tantalum test materials exhibited inhomogeneous mechanical behavior evidenced by non-uniform plastic strains in the off loading axes. For example, all compression samples were found to exhibit non-uniform radial expansion along the compression samples were found to exhibit non-uniform radial expansion along the compression axis. Metallographic examinations and micro-texture analysis of the microstructure of tbe test materials indicate that the inhomogeneous mechanical behavior is related to microstructural banding in the plate materials. 11:25 am THE MICROSTRUCTURE OF 'Ill and Ta-W ALLOYS DEFORMED AT HIGH STRAIN RATES: C.L Briant, R.H. Batcheler, Division of Engineering, Brown University, Providence, RI 0291:2. D.H. Lassila and W. Gourdin, Lawrence Livermore National Labs, Livermore, CA 94551 This paper will report on a study of the microstructure of pure Ta, Ta-2.5W, and TalOW alloys that have been subjected to deformation over a range of strain rates between approximately 10-' scc:' and 10' scc:'. Prior to testing the samples had either been annealed or subjected to shock loading. The high strain rates were applied by testing the samples in compression in a split-Hopkinson bar. The samples were examined by both optical metallography and transmission electron microscopy; the analysis of the dislocation structure from the latter technique will be the focus of this talk. The results show that in the annealed samples the Ta and Ta-2.5W alloys contain parallel screw dislocations with Burgers vectors. In the Ta-IOW alloy, regions were observed with long, straight parallel dislo<:ations overlapping in specific directions. Deformation greatly increased the density of dislocations and showed the development of a cell structure. The formation and role of twins both during shock loading and deformation will be discussed. 11:45 am WORK HARDENING AND THERMAL SOFTENING PROPERTIES OF TANTALUM AND TANTALUM ALLOYS: Craig L Winman, Jon P. Swensen, Dave A Dehmer, Alliant Techsystems Inc., 600 2nd Street N.E., Hopkins, MN 55343 Tantalum and tantalum Alloys of tanlalum-2.5% tungsten and tantalum-10% tungsten have been tested in the annealed and cold worlced condition at strain rates ranging from 10' to 100s" and temperatures from 78-1000K. The work hardening and thermal softening response as the tungsten content is increased will be shown Additionally the work hardening and thermal softening responses due to various oxygen levels (30300ppm) in pure tantulum and tantalum·2 5% tungsten will be presented.

TECHNICAL, ECONOMIC COMMERCIAL & ENVIRONMENTAL CONSEQUENCES OF SULFUR RECOVERY IN COPPER PRODUCTION SYSTEMS I: S02-Emlsslon Abatement's Effect On Recent Smelter Operating Practices

Sponsored by: EPD Copper, Nickel, Cobalt Committee Program Organizers: Norbert L. Pire!, Piret & Stolberg Partners, 1m Licht 12,047279 Duisburg, Germany; Ivan M. Santos Moraes, Caraiba Metais, Dias D' Avila, Bahia, Brazil Tuesday,AM February 6, 1996

Room: AII-12 Location: Anaheim Convention Center

Session Chairpersons: G.A. Eltringhanl, Magma Copper, P. O. Box M., San Manuel, AZ 8563 I; H. J. Konig, Lurgi Metallurgie GmbH, Lurgi Allee5, 0-60295 Frankfurt am Main

8:30 am Invited RECENT OPERATION OF THE ACID PLANT AT HUELVA COPPER SMELTER: Patricio Barrios, (Vice President Copper Operations), J. L. Bonallo, (Smelter Superintendent), J. A. Suarez (IProcess Metallurgist), Rio nnto Metal, S.A., Apartado de Correos 110,21080 HuelvalSpain The Rio nnto Metal (RTMetal) copper smelter, located 5 km from Huelva, Spain, began operations in 1970. Since then a series of expansions and modifications have allowed a progressive increase in copper production. Environmental consideration having become increasingly important, environmental protection has come to represent an important aspect of the investments in production. As a result, modifications have been

1996 TMS ANNUAL MEETING AND EXHIBITION

carried out in the two existing acid plants: Modification of the RTM 1 gas cleaning section to closed circuit; Substitution of cascade filters and acid lines; Modification of the RTM 1 plant to double absorption. The Expansion and Environmental Improvement Project, completed end 1995, will allow 270,000 t Cu and 800,000 t H,SO. to be produced yearly, at the same time achieving a decrease in specific and overall eotissions. RTMetal's goal is to be an efficient smelter, withthe avowed intention of respecting the environment and adopting the best available technologies. 9:00am HERNAN VIDELA LIRA SMELTER'S PROMISING FUTLIRE OF THE SO,ABATEMENT PROGRAM: German Richter, Jorge Del Castillo, Enami, Chile, HernWl Videla Lira Smelter, Casilla 483, Copiap6, Chile To comply to the Chilean EPA regulations and ambient air quality requirements, the HernWl Viidela Lira Custom Smelter, having to cope with a wide variety of concentrates, has developed a two-step modernization plan; a second acid plant of 80,000 Nm'lh to be implemented in 1996 will increase the sulfur capture from 40% to 70%, in asaociation with the replacement of the Hoboken Converter by a second PS-Converter, thereby lowering operating problems due to treating high grade matte. Oxygen enrichment, to be incorporated in 1998 in the Teniente and P. S. Converters, in association with tuyere injection of dry concentrates, will result in capturing 9().93% of the sulfur input. By this program, sulfur eotissions will be decreased from today 40,000 tpy to less than 8,000 tpy and the environmental performance monitored. 9:25am SULFUR DIOXIDE EMISSION REDUCTION BY THE POLISH COPPER INDUSTRY: Jan Garbaczewski, KGHM Polska Miedz S.A., 59-301 Lubin, Poland, ul, M.C. Sklodowskiej 48 Polish Copper Industry, located in the Legnica-Glogow region and operated since the 1970's, includes four otines and two smelters, "Legnica" and "Glogow", the latter consisting of Glogow I and II both having different technologies. The "Legnica" and the "Glogow I" smelters using shaft furnaces, sulfur eliotination largely takes place during batchwise converting, resulting in fluctuating SO,-concentrations between less than 1 % up to 10 % in the converter gases to the sulfnric acid plants. The ensuing difficulties of the past have now largely been overcome. Introduction of flash smelting at "Glogow II", with subsequent acid making by double SO,conversion from the stable process gas, contributed significantly to further reduction of SO,-eotission. In 1994 a new double conversion sulfnric acid plant was installed at "GJogow I" as well as a "Solinox" desulfurisation plant at "Legnica", further reducing SO, eotissions considerably. The next step towards this goal under consideration is a seoti-dry desulfurisation system for the "Glogow I" power station. SO,-abatement measures by KGHM and the results thereby achieved are presented. 9:50am ECONOMICS OF GAS AND DUST COOLING IN COPPER PRODUCTION FROM SULFIDES: Kurt A. Westerlund, Hannu Holopainen, Ahlstrom Corporation, Ahlstlrom Pyropower, POB 5, FIN 00441 HelsinkilFinland All known applied gas cooling methods are outlined. The criteria that should be considered when choosing the best gas cooling methods for each case are discussed. The bottlenecks on production caused by undersizing and undermaintaining the gas cooling equipments are econootically compared to the incremental cost and earning from increased production. Dust sulfating energy and its effect on cooling equipment and ESP-availability are examined. The operating window for the fast cooling Fluxflow fluid bed cooler to avoid sulfating is thermodynamically analyzed and kinetically explained. 10:15 am BREAK 10:35 am SULFUR OXIDE CONTROL AT KENNECOTT'S UTAH COPPER SMELTER: David B. George, Kennecott SmelteriRefinery Modernization Project, PO Box 10, Magna, UT 84044; Ray J Gottling, Chris J. Newman, Technical Services, Kennecott Utah Copper, PO Box 329, Magna, UT 84044 The Modernized Kennecott Utah Smelter utilizes Outokumpu flash smelting and Kennecott-Outokumpu Flash Converting in conjunction with advanced acid plant and gas cleaning technology to achieve over 99.9 % capture of input sulfur. Sulfur oxide eotissions are less than 3 kgffonne of copper. 11:05 am ARSENIC CONTROL IN THE SULFURIC ACID PLANT BY THE PURIFICATION SYSTEM OF CARAIBA METAlS: Esther Carrera Maninez, Process Engineer, Caraiba Metais S.A, Via do Cobre no 3700, A 1.0. COPEC, Dias D-Avila, Bahia, Brasil. Carafba Metais is the sole Brasilian Copper Custom-Smelter. The sulfur present in the various concentrates processed is recovered as sulfuric acid and oleum which are sold in the local market. For quality assurance purposes, a study was carried out in order to deterntine how to control the arsenic content in the acid, allowing the continuous processing of concentrates with high levels of this impurity. The study was based on the process parameters of the Washing Towers and Mist Precipitators. Other process variables were investigated and their effects were used in support to confrrm the results. After many analyses performed, some interventions to improve the quality of the acid were implemented in the plant, which proved to be successful even by using a conventional purification system. 11:30 am Invited COPPER SMELTER ACID PRODUCTION: Ian C. Knight, Head of Metallurgy

1996 TMS ANNUAL MEETING AND EXHIBITION

Group, CRU International, 31 Mount Pleasant, London WCIX OAD Great Britain The paper will review the current levels of by-product acid production at copper smelters on a regional basis, and look at the possible changes in the future. There will be increasing pressure to close or modify copper smelters that currently either have no sulfur capture or have sulfur capture levels well below the otinimum level of around 95 %, a level that is considered the otinimum acceptable in developed countries. The plans for technology changes at smelters achieving low capture will be considered, and the effect of these smelter changes on future levels of acid production will be identified.

TRANSIENT THERMAL PROCESSING OF MATERIALS: SESSION II

Sponsored by: EMPMD Thin Films & Interfaces Committee Program Organizers: N. M. Ravindra, New Jersey Institute of Technology, Newark, NJ; R. K. Singh, University of Florida, Gainesville, FL

Thesday,AM February 6, 19%

Room: Grand J Location: Anaheim Marriott Hotel

Session Chairmen: Bhushan Sopori, Naitonal Renewable Energy Lab, Golden CO; Anthony Fiory, AT & T Bell Laboratories, Murray Hill, NJ

8:30 am Invited

RAPID THERMAL ANNEALING ISSUES IN SILICON PROCESSING: Richard B. Fair, Dept. of Electrical and Computer-Engineering, Duke University, Durham, NC 27708 Rapid ThermaI Annealing (RTA) in silicon technology provides a controlled environment for thermally activated processes that is increasingly difficult to achieve in batch furnace systems. However, the vast majority of RTA tools use optical radiant sources which create a complex heating system for silicon wafers with multiple absorbing and reflecting layers which are usually patterned in device structures. Optical interference at overlayer interfaces modulates the wafer eotissivity, and thus, the amount of energy absorbed causing large differences in temperature across the wafer. As a result, the introduction of RTA, RTO and RTCVD into manufacturing has had Iiotited success, since these are all activated processes that are sensitive to changes in temperature. This talk will review issues in applying RTA to manufacturing silicon ICs with particular emphasis on optical radiation effects on temperature uniforntity and measurement, dopant activation, and the particular challenge of Iiotiting dopant diffusion in shallow junctions during the simultaneous removal or establishment of defects in crystalline silicon. 9:00 am Invited TRANSIENT ENHANCED DIFFUSION IN BORON IMPLANTED SILICON: Kevin S. Jones, Department of Materials Science & Engineering, University of Florida, Gainesville, FL 32611-6400 An overview of recent results correlating transient enhanced diffusion (TED) and extended defects will be presented for boron implanted silicon. There appear to be several different types of TED depending on the extended defects that mediate the release of the excess interstitials. Results will be presented on how the implant conditions affect the transition between these different types of defects and therefore TED. Measurements using doped superlattices and SIMS of the point defect flux during different kinds of TED will be presented. Finally it will be shown that there are distinctly different activation energies for saturation of TED depending on the type of extended defect or cluster mediating the process. 9:30 am Invited DOPANT DIFFUSION IN RT PROCESSED MATERIALS: M.E. Law, Department of Electrical Engineering, University of Florida, Gainesville, FL 32611 Abstract not available.

10:00 am BREAK 10:15 am CURRENT STATUS OF RTCVD IN SILICON PROCESSING: O.H. Gokce & N.M. Ravindra, Department of Physics, New Jersey Insitute of Technology, Newark, NJ 07102 Abstract not available.

TUESDAY PM ADVANCES IN THE SCIENCE AND TECHNOLOGY OF TITANIUM ALLOY PROCESSING· IV: Cold Working II Sponsored by: SMD TItanium Committee, MDMD Shaping and Forming Committee

B65

Program Organizers: Prof. Isaac Weiss, Prof. Raghavan Srinivasan, Mechanical and Materials Engineering Dept., Wright State University, Dayton, OH 45435; Dr. Paul Bania, TImet Corporation, Timet-Henderson Technical Laboratory, P.O. Box 2128, Henderson, NV 89009; Prof. Daniel Eylon, Graduate Materials Engineering, University of Dayton, Dayton, OH 45409 Tuesday, PM February 6, 1996

Room: B5-6 Location: Anaheim Convention Center

Session Chairpersons: P.L. Martin, Rockwel1 Science Center, 1049 Camino Dos Rios, Thousand Oaks, CA 91360; J.C. Fanning, Henderson Technical Laboratory, TImet, P.O. Box 2128, Henderson, NV 89015 2:00 pm Invited COLD EXTRUSION OF TITANIUM ALLOYS AND TITANIUM MMCs: Hans W. Wagener, Joachim Wolf, Metal FOrming Laboratory, University of Kassel, 34109 Kassel, Germany The fol1owing metals were tested by the basic cold extrusion processes, forward bar extrusion, forward tube extrusion, and backward cup extrusion: pure titanium, TI-6AI4V, TI- 13AI-IICr-3Al, TI-3Al-8V-6Cr-4M04:lJ: (Betal-C) and two MMCs, TI - 99.9 + 10% TIC and TI-6AI-4V + 10% TIC. The manufacturing techniques, tooling, and the optimum coating and lubrication systems for cold extrusion of these metals are described. Deformation force vs. punch travel digrams are plotted. The values of the coefficient of friction acting between material and tool surface are determined experimentally and by computer simulation. In the case ofTI-6AI-4V the high strength combined with the small ductility makes cold forming problematic. The same problem exists for the MMCs. The fibrous and particle strengthening act like material seperation in the matrix. The MMCs are considered being brittle materials. To increase the workability the state of stress in the forming zone is shiften more to the compressive side by applying a counter pressure. 2:30pm COLD WORKING OF TI13Nb-13Zr: C.M. Bugle, E. W. Robare, Dynamet Inc., 195 Museum Road, Washington, PA 15301; K.P. Daigle, J .A. Davidson, Smith and Nephew TransTech, 1450 Brooks Road, Memphis, TN 38116 TI- 13Nlr 13Zr (TI- 13-13"") is a new titanium alloy which has a unique combination of properties that may make it an attractive choice for a variety of applications. Originally developed for use in biomedical implants, this alloy combines a low elastic modulus, high strength, excellent hot and cold workability, superior corrosion resistance, and the ability to be surface hardened to improve wear properties. Research on this agehardenable alloy has shown that the mechanical properties can be controlled over a significant range through hot working, heat treatment and cold working. The elastic modulus ofTI- 13-13 can be varied between approximately 6 and 12 Msi, and strengths as high as 190 ksi have been achieved. Some typical combinations of properties for various material forms and conditions will be presented. Further, the material does not seem to exhibit some of the typical processing difficulties encountered in some agehardenable beta-TI alloys. For example, in wire drawing trials, TI- 13-13 yielded exceptionally good surface finish in the "as-drawn" condition with over 85% cold work. Some examples of processing behavior will be discussed. 2:50pm TITANIUMALUMINIDE FOIL PROCESSING: C. C. 'MJjcik, Teledyne WahChaog, P. O. Box 460, Albany, Or 97321; R. Roessler, R. Zordan, Allison Engine Co., P. O. Box 420, S. C. W05, Indianapolis, In 46206-0420 Orthorhombic titanium aluminides based on the TI2AINb composition are ofinterest for high temperature composite applications because of their excellent strength and compatibility with SiC fibers. The performance improvements attainable with this new type of titanium alloy will only be realized if orthorhombic alloy foil and SiC fibers can be produced economically. Fabrication of these alloys into thin (i.e., 0.10 mm thick) foil presents some difficult problems. We will report on our experience fabricating TI22Al-26Nb into foil. The effects of hot working and annealing temperature have profound influences on microstructure and properties. Additionally, this alloy is sensitive to cooling rate which may be used to quench in metastable structures that are more workable than ordered orthorhombic or ordered alpha phases. The effects of heat treatment on microstructure, hardness and cold workability will be discussed. 3:10pm ABRASIVE WATER JET CUTTING OF TITANIUM VENT SCREENS FOR THE F-22ADVANCED TACTICAL FIGHTER: H.R. Phelps, Lockheed Aeronautical Systems Co., Marietta, Georgia The F-22 Advanced Tactical Fighter (ATF) has several titanium vent screens. These screens will be fabricated using abrasive water jet (AWl) cutting to produce up to 27,000 densely packed, rhombus shaped holes. This paper will compare the potential methods of producing these screens, why AWl was selected as the primary method, and issues associated with aerospace applications. 3:30pm CRYOGENIC MACmNING OF TITANIUM 64 ALLOY: Shane Y. Hong, Dept. of Mechanical & Materials Engineering, Wright State University, Dayton, OH 45435 Due to its abrasiveness,low thermal conductivity, and chemical reactivity, titanium has been long classified as a difficult-to-machine material. Cutting tool wears rapidly and tool life is extremely short. For this the cutting speed of 200 to 300 feet per minute is used for machining titanium in the industry, compared to a speed 10 times faster used in machining steels. The slow cutting process and the rapid tool wear makes the productivity very low and the production cost very high. This paper presents a new cutting process that uses liquid nitrogen to cool the high temperature usually generated in conventional machining. It enables higher speed cutting and better the productivity. On the other side, the production cost has been much reduced. In addition, this is an environ-

B66

mental safe manufacturing process. 3:50 pm BREAK 4:00pm CRYOGENIC MILLING OF TITANIUM 64 ALLOY: Shane Y. Hong, Dept. of Mechanical & Materials Engineering, Wright State University, Dayton, OH 45435 As part of effort of environmental conscious manufacturing, milling of titanium is one of the needed areas for the researcher to focus on. Due to the extremely short tool life in dry cutting of titanium alloys, cutting fluid is necessary in the milling process to remove the heat generated in the cutting process. But conventional coolant is not environment friendly, and costs at least double the purchasing price to dump it. This paper introduces a cryogenic milling process which is essentially a dry cutting process. Liquid nitrogen is injected to the milling tool cutting tips through a rotary distribution ring. It evaporates at the cutting tool tips and ;Ibsorbs the high heat generated in the cutting process. No residue is left in the work piece and the chips. The effects of this neat and clean process include longer tool life and better work quality. This new cryogenic milling system features a low nitrogen consumption so that the production cost is competitive compared to conventional milling process. 4:20pm HYDROGEN INFLUENCE ON MACmNING OF TITANIUM ALLOYS: B. A. Kolachev, Y. B. Egorova, Moscow SUIte Aviation Technological University after Tsiolkovsky, Petrovka 27,103767 Moscow K-31, Russia; V. D. ThJaev, Scientific-Production Association "AVITOM", U1anskyi pereulok 16, 101000 Moscow, Russia TItanium alloys are difficult to machine. The working hours of machining of titanium alloys is 7-8 times more than for aluminum alloys and 4-5 times more than for carbon steels. The machinability of tiuUlium alloys may be improved essentially by reversible hydrogen alloying. In this paper detailed investigations of hydrogen influence on machining of titanium alloys have been carried out. It has been shown that hydrogen alloying of titanium alloys leads to significant increasing of tool life (2-10 times). The specific tool life has a maximum value at hydrogen concentration 0.1-0.5 wt% depending on composition of alloys. The favourable influence of hydrogen on machining of titanium alloys is diminished with the growth of cutting rates. Hydrogen alloying causes changes of chip-formin~:: chip become fragile. The improvement of machinability of titanium alloys by hydrogen alloying is due to microstructure refinement, decrease of metal flow stresses and impact ductility, increase of thermal conductivity and changes in tool wear characteJistics. When machining titanium alloys with iniatial hydrogen content (0.003-0.005%wt) wear of cutter occurs on the front face as well as on the back one. The wear on tliIe back face of cutter takes place mainly at machining of hydrogenated metal. The machined hydrogenated details should be vacuum annealed to reduce hydrogen content in a, metal to safe level which is not followed by hydroden embrittIement during operation. 4:40pm TEMPERATURE DISTRIBUTION IN TITANIUM MACHINING BY FINITE ELEMENT ANALYSIS: Shane Y. Hong, Y. C. Ding, Dept. of Mechanical & Materials Engineering, Wright State University, Dayton, OH 45435 Temperature distribution in the cutting tool and chips will be analyzed using the finite element method, and will be presenlied in terms of various machining parameters, such as feed speed, depth of cut for both dry and cryogenic machining conditions.

ALUMINA & BAUXITE~ TECHNOLOGY I: Fundamentals II:

Sponsored by: LMD Aluminum Committee Program Organizer: C.G. Hendriks, Kaiser Aluminum and Chemical Corp.,9131 Interline Drive, Baton Rouge, LA 70809 Tuesday,PM February 6,1996

Room:A6 Location: Anaheim Convention Center

Session Chairman: Dr. Milind Chaubal, Reynolds Metals Co., PO Box 9911, Corpus Christi, TX 78469

2:00pm HYDRARGILLITE AND BAYERITE: TWO POLYMORPHS OF ALUMINUM TRIHYDROXIDE: Mauryay M. Pyzalski, Wieslawa Mocun - Wczelik, Technical University of Mining and Metallurgy, Faculty of Ceramics and Material Engineering, Ave. Mickiewicz 30, A 3,30-059, Cracow, Poland The two following forms of the aluminum trihydroxide have been produced as a result extensive studies:the first one, obtained by Bayer and Grzymek method,the second one in a new Polish technology on a semi-technical scale. Some new modification were synthesized at different technical parameters of the process and they were subsequently investigated by use of several methods. The results thus obtained allow to characterize the properties of hydroxides. They also put a new light on their formation and polymorphism. 2:30pm THE STUDY ON REACTION OF RUTILE IN THE BAYER LIQUOR: Bi Shiwen, Yang Yihong, Li Dianfeng, Fu Gaofeng, Zhao Xin, Li Yan, Department of Nonferrous Metallurgy of Northeastem University, Shengyang 110006, China TItania has a hindrance effect in digestion of diaspuric bauxite, the forms of titania in diaspore are rutile and anatase. In the papt:,r, the reaction of rutile in the Bayer digestion

1996 TMS ANNUAL MEETING AND EXHIBITION

liquor is studied, and the effects of CaO, Na20CAUST concentration and temperature on the reaction are probed into. Tbe substance phase of the reaction product in difference condition are analyzed by using X-ray. 3:00pm EFFECT OF THE GRAINS MORPHOLOGY OFSPECIALALUMINA ON THE GREEN DENSITY OF MOULDERS: M. A. Wojcik, J. Sidor & 0. Paiyz, Technical University of Mining and Metallurgy, Faculty of Materials Science and Ceramics, al. Mickiewicza 30, 30-059 Cracow, Poland Tbe successful fabrication of high alumina ceramics is a thorough preparation of the powders which depends on the various requirements of the different forming procedures. All of the aluminas used in these ceramics have defined ultimate crystal sizes, controlled grindability and green densities to ensure the very tight shrinkage tolerances which are critical for the spark plug production process. Production practices shows that alumina milling belong to very important unit operation. So, the effect of grains morphology of special alumina obtained after milling in rotary-vibration mill on the green density of moulders are presented in this paper. Tbe granulation parameters (d50) of monodispersed product was selected according to its application in the production of sparking plugs. 3:30pm NEW TECHNOLOGY FOR ALUMINATE LIQUOR DECOMPOSITION: Dr. V. Abramov, Scientific and Technical Center "Reactor", Russia. G. Stelmakova, AO "Aluminiy", Russia Tbe hasis for new technology for decomposition of aluminate liquor is the use of high seed ratio and countercurrent movement of liquid and solid phase with intensified mass transfer. The letter excludes the cycle of the main seed mass. To perform the process a new type of reactor without special mixing devices was created. It was named ATK "Sloy". Intensification of mass transfer makes it possible to shorten the period and to increase the degree of aluminate liquor decomposition. Tbe new technology and the reactor are the result of the theory of nonpermanent mass transfer. Tbe suggested technology was put to commercial tests on basis of a special experimental 100 m3 reactor. First the new technology was applied to aluminum hydroxide washing. Tbe experiment showed that it was possible to use countercurrent mode of aluminum hydrate washing with water consumption on more than 1 cubic meter per ton. Later tests were conducted on decompoSition of aluminate liquor with the initial Alumina/Caustic ratio of 3.2-3.3. Tbe tests showed that using the new technology the decomposition period of aluminate liquor can be shortened significantly. Thus, it took 20 hours to increase the AlC ratio from 3.2 to 4.1, while usually the highest ratio reached within this period of time was 3.7. At present we started installation of an industrial 1000 m3 reactor "Sloy" to use the new technology of liquor decomposition. 4:00 pm BREAK 4:20pm THE EFFECT OF VARIOUS ADDITIVES ON BAYER DIGESTION OF DlASPORIC BAUXITE: Zhang Lun-he, Xu Qi-li, Wang Xiang-Ii, Zhengzhou Light Metal Research Institute, Henan, Shangjie District, Zhengzhou, China In this paper, the effect of about ten kinds of additives on the Bayer digestion of diasporic bauxite has been investigated. Tbe experiment results showed that mineral additives such as dolomite, chlorite, medicinal stone, magnesite have a great influence upon the alumina extraction ratio of diasporite and the scaling property of slurry. Nevertheless a great quantity of additives must be used, the digestive results of bauxite is not satisfactory (except dolime). So it is difficult to put into practice. In addition, we treated lime and red mud or slurry with sodium-aluminate liquor to achieve synthesis catalyzators on the definite conditions. Compared to the results when lime being added, the alumina extraction ratio of diasporite increased 2-4% when this kind of synthesis catalyzator being added under otherwise equal conditions, and their molar ratio of digestion liquor (Na20/AL203) decreased 0.03-.011. Tbe effect is obvious and the technology is simple. Tbe experiment results approved that the synthesis catalyzator have industrial applied value.

ALUMINUM ALLOYS FOR PACKAGING II Session II

Sponsored by: SMD Non-ferrous Metals Committee and Light Metals DIvision Program Organizer: Professor James G. Morris, Chemical & Materials Engineering, Director, Light Metals Research Labs., University of Kentucky, Lexington, KY 40506

Tuesday, PM February 6, 1996

Room:A2-3 Location: Anaheim Convention Center

Session Chairperson: H. Scott Goodrich, Ravenswood Aluminum Corp., P.O. Box 98, Ravenswood, WV 26164 -

2:00 pm Invited CAN AND END DESIGN TRANSFORMATION: Ms. Selwyn M. Knap, Reynolds Metals Company, Can Division, 7900 Reycan Road, Richmond, VA 23237 This paper will walk tbrough a selection of major changes in can and end designs, as they relate to base stock, beginning with the early 1960's through 1995, with a glimpse of what the future may bring. Presented from a can maker's point of view, the ideas of the "perfect can stock" will be discussed as well as challenges surrounding conversion to reduced neck diameters, bottom design changes and downgauging.

TER THE PARTIAL ANNEALING: Shin Tsuchida, Hiroshi Yokoi, Technical Research Laboratories, Sumitomo Light Metal Ind., Ltd., Chitose 3-1-12, Minato-ku, Nagoya, 455 Japan The strain hardened sheets of AI-I %Mn-I %Mg alloy which were heat treated for the solutionizing during the rolling process showed the increases of tensile strength and elongation after the partial annealing below 200·C. In these partially annealed sheets, however, the decrease of formability was obviously observed after the additional cold working. The decline of the formability confrrmed by the hole expansion test was discussed in the relation with the elongations obtained by uni axial tensile tests. It is postulated that the very fine precipitates on the dislocation structures which were born by the low temperature annealing are the cause of the development of deformed micro structures which intend to decrease the non-uniform elongation. 3:00pm CRYSTALLOGRAPHIC TEXTURE IN COLD ROLLED 5182 AND ITS CHANGE IN THE PANEL WALL DUE TO SHELL FORMATION: Phillip A. Hollinshead, Alloy Technology Division, Alcoa Laboratories, Aluminum Company of America, Alcoa Center, PA 15069 Local orientation measurements by the electron backscattered Kikuchi (BKD) technique were made at various depths through the thickness of the panel wall of an extensively thermally recovered 5182 end shell. Measurements were done at the traosverse location relative to the rolling direction (RD). For comparative purposes, similar BKD measurements were made on the panel itself, and for further comparative purposes the crystallographic texture of the panel was measured by standard X-ray pole figure techniques. Orientation distribution functions (ODFs) were calculated from the X-ray and BKD data. Tbe crystallographic texture of the panel was typical of one commonly found in cold rolled 5182. By contrast, however, the BKD technique indicated a very low intensity of the Cu component, (112}<111>, which was believed to be due to a higher dislocation storage rate in the Cu component which interfered with Kikuchi pattern production. This was in spite of the extensive thermal exposure the shell had been subjected to. Tbe texture varied along the height of the panel wall but at alllocations the Cu component had very low or zero intensity. Other rolling texture components were also reduced compared to the panel, but appear to have been replaced by a new component close to (120)[212J. 3:20pm MICROSTRUCTURE-PROPERTY RELATIONSIDPS IN ALUMINUM FOIL ALLOYS: GJ Mahon & GJ Marshall,Alcan International, Banbury Laboratory, Southam Road, Banbury, OX16 7SP, UK In production, manufacturing and service, an important requirement of aluminum foil alloys is the ability to resist tearing. In lxxx and 8xxx foil alloys the microstructural features that influence general properties are known to be fine dispersoids, coarse constituent particles and solid solution levels, particularly Fe. To study the effect of these features, advanced testing and characterization techniques are needed, including a quantification of tearing performance. A thicker gauge tear test adapted for foil alloys, 5-200~m, will be discussed in the context of understanding tear resistance of commercial foil alloys. Secondly, to understand the importance of individual microstructural features, outlined above, alloys with 'engineered' microstructures have been tested. Preliminary results will be presented that further the understanding of this complex structure-property phenomenon. 3:40 pm BREAK 4:00pm INGOT HOMOGENIZATION STUDY OF 3104 CAN BODYSTOCK ALLOY: Rajeev G. Kamal, Corporate Research and Development, Department of Metallurgy, Reynolds Metals Company, P.O. Box 2700, Richmond, VA 23261 A study was conducted to investigate the effect of heating rate and the homogenization temperature on the microstructural evolution during homogenization of can body stock ingots. Tbe through-thickness as-cast ingot compositions and SDAS (secondary dendrite arm spacing) were measured and significant edge-center differences were observed. Industrial type homogenization practices were simulated and (a) Mn content in solid solution (Mn ss )' (b) IX-dispersoid size distribution and (c) IX-constituent fractions were evaluated. The Mnss increased with decreasing homogenization temperature but the IX-dispersoid size distribution was affected more by the prior thermal history. Tbe ingot edges showed a higher drop in the degree of Mn supersaturation than the center of the ingot due to lower SDAS of the edges. Transformation of AI6(FeMn) to IX-AI12(FeMn>JSi was enhanced by the higher homogenization temperature. Tbe influence of homogenization practice on the final sheet properties such as strength, earing and microstructure are also discussed. 4:20pm THE DEVELOPMENT OF HIGH ANGLE DEFORMATION BOUNDARIES AND LOCAL ORIENTATIONS IN ALUMINUM: D. A. Hughes, Center for Materials and Appl. Mechanics, Sandia Nat. Lab., Livermore, CA 94550 High angle boundaries and their local surroundings are important for all aspects of recrystallization including nucleation, growth and texture formation. Tbey are thus important for processing of aluminum alloys. Tbese boundaries have been observed and quantitatively characterized using transmission electron microscopy in aluminum following rolling from medium to large reductions. Tbe distribution of local orientations between individual dislocation boundaries and their angle/axis pairs have been measured using convergent beam kikuchi analysis and are compared to the macroscopic texture. The sequence of near neighbor orientations shows that individual grains subdivide into tbree to four different texture components separated by sharp boundaries. The local orientations surrounding high angle boundaries are much more diverse than suggested by simple models and single crystal studies. Five origins for these high angle boundaries are suggested and discussed. This work supported by U.S. DOE under contract No. DE-AC04-94AL8500.

2:30 pm Invited FORMABILITY OF THE COLD ROLLED SHEETS OF AA3004 ALLOY AF-

1996 TMS ANNUAL MEETING AND EXHIBITION

B67

4:40pm DISPERSOID MODELLING IN 3XXX ALLOYS: J.P. Suni, R.T. Shuey, Alcoa Technical Center, Alcoa Center, PA 15069; R. Doherty, Drexel University, Philadelphia, PA 19104 A physically based model was developed to predict dispersoid evolution during preheating of 3XXX alloys. The model combines the parallel processes of nucleation, growth and coarsening, resulting in evolution equations for particle size, number density and volume fraction. Model parameters relate to diffusivity, interfacial energy, nucleation kinetics and equilibrium solvus for the AII2Mn3Si phase. Model fitting consists of using conductivity data, employing simple relationships between conductivity change and volume fraction AI12Mn3Si dispersoid precipitated. Data used in fitting this model includes internal Alcoa data, as well as published, external results. The combined data set contains variations in solidification rate, alloy composition, namely Fe, Mn, Si and Mg (i.e. either 3003 or 3004), as well as time and temperature of thermal treatments. All of the trends in these variations are qualitatively consistent arnong the various data sets. A model incorporating these effects was developed for the combined data set and the rms error of fitting was .04 wt% Mn in solution. The resulting model enables predictions of the effects of process excursions, or the searching of new combinations of alloy composition and thermal practice, on the basis of physical principles and prior data. 5:00pm NECKING TWO PIECE BEVERAGE CANS IN THE \IO's: Dean Johnson, Bell Corporation, Metals Container Operations, 9343 West IOSth Circle, Broomfield, CO 80021-3682 This subject is fresh in the minds of canmakers because the past few years have been a challenge for the Two Piece canmakers. The introduction of 204 and 202 ends and cans created great opportunities for machine manufactures. This paper will outline the advantages and disadvantages and disadvantages of Standard Die Necking, Smooth Die Necking, Spin Nicking, and Spin Flow Necking. Each necking process will be looked at from the stand point of can line-front end quality requirements, can line spoilage, can defects, coatings required. Also included will be Ball's option on, "What will be the necking process of the future?"

ALUMINIUM REDUCTION TECHNOLOGY IV: Operational Performance Improvements Sponsored by: LMD Aluminum Committee Program Organiser: Ms FionaJ Stevens, Comalco Research and Technology, Comalco Research Centre, PO Box 316, Thomastown, Victoria 3074, Australia Tuesday, PM February 6, 1996

Room:A9 Location: Anaheim Convention Center

Session Chairperson: Peter ERtner, AIusuisse Lonza Services Ltd, Research and Process Development, 3965 Chippis, Switzerland

2:00pm THE ECONOMICS OF SHUTTING AND RESTARTING PRIMARY ALUMINIUM SMELTING CAPACITY: K.J. Driscoll, CRU International Ltd., 31 Mount Pleasant, London, WC1X OAD, England In recent years the aluminium industry in the Western world has been operating well below capacity, with cutbacks in production due largely to the depressed aluminium market conditions of 1992 and 1993. Since mid-I995, however aluminium producers have begun restarting idled capacity. Extensive efforts and preparation are required both to close capacity in an orderly manner and to restart idled capacity, thus CRU has carried out a comprehensive analysis of the costs of shutting capacity, maintaining idled capacity, and restarting potlines. Costs have been calculated for a smelter which may be considered representative of the industry as a whole. Technical aspects and commercial data are outlined for the representative smelter, with costs presented under a variety of shutdown and restart conditions. Additionally, the time required to bring capacity back on-line is examined for several scenarios and the economic impact of idled capacity is discuased. 2:30pm UNTAPPED OPPORTUNITIES FOR IMPROVING PERFORMANCE AND INCREASING PRODUCTION OF OLDER AND NEWER SMELTERS: Helge O. Forberg HOF Incorporated, 8A Quail Ridge Court, Owensboro, KY 42303 For older smelters to remain competitive and extend economic life, it is important to exploit opportunities to increase their production. These efforts have been going on for many years, and the most successful producers have been able to stay viable. Future opportunities are still promising for many older facilities as further increases of incremental production are waiting to be harvested. Newer smelters are interested in decreasing cost by increasing production. Even if the operating cost of newer plants is lower than for older ones, it is important that newer smelters find improvements which could help generate an acceptable rate of return on invested capital. During the last 1015 years the aluminum producers have learned the hard way the importance of being competitive. The industry has also learned it is just as important to have an acceptable return on the invesunent in each smelter. This paper will discuss both technical and operating opportunities that are open to older as well as newer smelters to stay viable. 3:00 pm BREAK 3:10pm THE ROLE OF SODIUM IN ALUMINUM ELECTROLYSIS: A POSSIBLE INDICATOR OF CELL PERFORMANCE: Alton T. Tabereaux Reynolds Metals Com-

868

pany, Corporate Research & Develepment, 3326 East Second St, Muscle Shoals, Alabama 35661-1258 The sodium concentration in aluminum in the metal pad of modern prebake cells, having superior magnetic compensation and stable operating conditions, is substantially higher compared with that measur'ld in older, less stable prebake and Soderberg cells. In some instances, the sodium cont,ent was found to increase in the metal after the cells were retrofitted with improved tecillnologies. The higher sodium level in aluminum in modem cells is due to the increase in the cathode polarization and build up of sodium containing species in the electrolyte near the bath/metal interface as a consequence of reduced stirring and cell hydJrodynamics forces. The relationship between the sodium content in aluminum measured in diverse types of aluminum cells and various operating circumstances are discusst:d. 3:40pm RETROFIT OF SODERBURG SMEI:I'ERAT ALUSAFBAYSIDE PLANT PART 1: CONCEPTUAL DESIGN AND ENGINEERING: D. Vogelsang, Ch. Droste, I. Eick, M. Segatz, YAW Aluminium-Tecbnologie GmbH, Friedrich - Woehler - Str. 2, 53117 Bonn, Germany A main objective of the Bayside smelter modernization at Alusaf Ply, is the retrofit of the Soderberg pots operating at 120 kA to prebake cells with a current load of 165 kA. The complete design and engineering of this modernization project was done by YAW. Based on computer simulations a tlilor-made concept for the retrofit project was developed. A pilot plant was taken succl:5sfully in operation in 1994. This paper presents the retrofit concept with details of Ithe magneto-hydrodynamic and thermo-electric calculations. Plant measurements demonstrate the good agreement with computer simulations. Some highlights of the detail engineering for the modernized pots elucidate concepts of integrating existing and new hardware. 4:10pm RETROFIT OF SODERBURG SMELTERATALUSAFBAYSIDE PLANT, PART 2: START-UP AND OPERATION OF PILOT PLANT: J.A. Femira, Alusaf Ply. Ltd., PO 284, Richards Bay 3900, Rep. of South Africa Construction, commiSSioning and the first year of operation of twelve 165 kA pointfed pilot cells with prebaked anodes at Alusaf works are described. Development of these pilot cells using YAW technology targette:d reduced emission levels and increased productivity of the 14 year old 120 kA VSS lines at Bayside smelter; while utilising as much of the existing hardware as possible and not exceeding the existing floor space. Extensi ve modelling showed that major parts of the existing busbar system could be kept and the amperage of the modernized pre bake cells increased to 165 kA. Twelve pilot pots that were constructed and taken in operation in June 1994. During the first year of operation excellent production results were achieved. The production of these cells are @ 50% higher than for the VSS cells. The current efficiency was in excess of 95%, the specific energy consumption is 13.6 kWhlkg Al and the net carbon consumption 410 kg/ton AI. 4:40pm A GOOD MORALE: THE KEY TO OUTSTANDING OPERATION RESULTS: Ono A. Knaisch. Aluminerie Alouette Inc:., PO Box 1650, Sept-lies, Quebec, Canada, G4R5C7 Three year experience in a greenfield smelter producing 215,000 tpy in Sept-lies, Quebec, Canada. Results are presented from start-up in mid-1992 up to steady-state operation in 1995. Achievements are a result of a combination of technical know-how and human dimension. Although technical expertise plays an important role in the management of an aluminum smelter, it must be complemented with the ability to build trust and promote a progressive attitude throughout all areas of the organization. AIouette is in a rather favourable situation by having a non-unionized workforce. Therefore, concepts such as multi-trade and multi-craft are used so that employees can optimize their potential and abilities.

CARBON TECHNOLOGY II: Anode Production! Performance

Sponsored by: LMD Aluminum Commilttee Program Organizer: Lise Castonguay, Alcan International Ltd., Arvida Research and Development Centre, PO Box 1250, 195:5 Mellon Blvd, Jonqui~re, Quebec, Canada, G7S4K8 Tuesday, PM February 6, 1996

Room:A7 Location: Anaheim Convention Center

Session Chairperson: Mr. Per Stokka, Norsk Hydro a.s., Research Centre Porsgrunn, N-390l, Porsgrunn, Norway

2:00pm MODELLING OF PACKING OF PETROLEUM COKE FRACTIONS: Harald Schreiner, Merten Sorlie, Elkem AlS Research, PO Box 40 Vagsbygd, N-4602 Kristiansand, Norway; Per Goltermann, VagnJohansen, G.M. Idorn Consult, Ramboll AlS, Bredevej 2, OK-283O VIRUN, Denmark The dry aggregate of anodes is compoS IOmm to sub-micron size. The amount of each fraction in the recipe, and hence the packing of these particles have a determining influence on properties such as dry aggregate density, pitch demand, green paste rheology, green and baked anode denSity, annde slumping behaviour, permeability and others. A semi-empirical mathematical model originally developed for calculating bulk porosity (bulk density) in concrete mixtures, has heen modified for use with paste plant petroleum coke fractions. The paper discusses the model and the necessary modifications to input

1996 TMS ANNUAL MEETING AND EXHIBITION

parameters that make it useful for aluminum smelter paste plants, whether they make prebaked anodes or Sllderberg anode paste. 2:30pm MASS LOSS, FUME AND POROSITY STUDIES OF PITCH + DUST BINDER MATRIX FOR ANODES: L. P. Lossius, M. Eie, S, Rorvik, H. A. Oye, Institute of Inorganic Chemistry, NTHlUnit, Sem Saelandsv, 12, N-7034 Trondheim, Norway A commercial coal-tar pitch binder for anodes was mixed with petroleum coke dust, carbon black or graphite to study the effect of binder level, type of filler and analytic parameters on the mass loss and the porosity that developed after controlled heat treatment in a thermobalance. The porosity was determined by a recently developed automatic method of image analysis. The evolving fume was collected during the heat treatment and analysed by gas chromatography to study the effect of these parameters on the quantity and composition of the fume. 2:50pm METHODS FOR DETERMINING THE DEGREE OF BAKING IN CARBON ANODES: Colin P. Hughes, Comalco Research Centre, PO Box 316 Thomastown, Victoria, Australia 3074 Anode baking temperature is recognised as a critical factor in determining anode quality and performance. It is difficult and costly to measure directly and an indirect method, the coke Lc technique, is often used. In this technique, baking temperature is estimated from the average crystallite size in the "c" direction (Lc) of a coke sample placed in the anode stubhole. The paper details the results of a large statistically designed experimental program in which coke Lc results are compared to anode properties routinely measured by smelters. Anode thermal conductivity and air and carboxy reactivity were found to correlate well with baking remperature. Real density and a direct anode Lc measurement technique were found to correlate best at high baking temperatures. Recommendations are given on the usefulness and simplicity of traditional anode property measurements for assessing baking temperatures as alternatives to the coke Lc method. 3:10pm A 3D MATHEMATICAL MODEL FOR THE HORIZONTAL ANODE BAKING FURNACE: Y. S. Kocaefe, E. Dernedde, D. Kocaefe, Process and Systems Engineering Research Group (GRIPS), Department of Applied Sciences, University of Quebec at Chicoutimi, 555 boul. de I'Universite, Chicoutimi, Quebec, Canada, G7H 2BI; R. Ouellet, Q. Jiao, Noranda Technology Centre, 240 boul. Hymus, Pointe Oaire, Quebec, Canada, H9R IG5; W. F. Crowell, Noranda Aluminum Inc., P.O. Box 70, New Madrid, M063869 In aluminum industry, carbon anodes are baked in large horizontal or vertical ring furnaces. The anode quality depends strongly on the baking conditions (heating rate, soaking time and final anode temperature). A three-dimensional mathematical model has been developed for a horizontal anode baking furnace to be able to assess the effects of different operational and design parameters on the baking process. The commercial CF.D code CF.DS-A..OW3D is used to solve the governing differential equations. The model gives the temperature, velocity and concentration distributions in the flue, and the variation of temperature distribution with time in the pit. In this paper, a description of the 3D model for the horizontal anode baking furnace will be given. Some of results from different case studies will also be presented. The results show clearly the importance of flue geometry on the gas flow distribution in the flue and the heat transfer to the anodes. 3:30 pm BREAK 3:50 pm

COKE CALCINATION LEVELS AND ALUMINUM ANODE QUALITY: ChristianDreyer, Bernard Samanos, Aluminium Pechiney, LRF-B.P. 114,73303 Saint-Jeande-Maurienne, Cedex, France; Franz Vogt, Calciner Industries, Inc., PO Box 1306, Chalmette, LA 70044 The calcination temperature of petroleum coke for aluminum anode applications has been generally increased by coke suppliers during the past 5 to 10 years. This paper discusses the effects of this trend for both the calciner and the user. High and low sulphur cokes were calcined industrially at different levels and used to manufacture laboratory scale anodes. Anodes were produced with these individual cokes and with blends of these cokes. The influence of anode baking level on the anode properties was investigated. The results are presented and discussed. 4:10pm BATH IMPREGNATION OF CARBON ANODES: R. Perruchoud, W. Fischer, R&D Carbon Ltd., PO Box 157, CH-3960 Sierre, Switzerland The mechanisms of the anode impregnation by bath in the electrolysis pots are reviewed. The rapidity of the penetration of bath in the anodes under different electrolysis conditions was measured in full size pots. In case of a cathodic contact of the anodes or of a cell current interruption, the butts (anode at the end of its service life) pick-up initially 0.2% of Na per minute. An exposition of the anodes to such conditions should therefore not exist otherwise a dramatic deterioration of the anode quality will occur when these butts are recycled. 4:30pm KINETIC STUDY OF THE CATALYTIC CARBONIZATION OF COAL TAR PITCH WITH PETROLEUM COKE: lilai Xue, Harald A. Oye,Institute of Inorganic Chemistry, The Norwegian Institute ofThchnology N-7034 Trondheim, Norway. Morten Sorlie, Elken als Research, PO Box 40 Vagsbygd, N-4602 Kristiansand, Norway The rate of carbonization has important impacts on the energy consumption and the unit productivity in baking process of prebaked anodes. In the present work the catalysts, such as S, AI03, ZnCI2, and FeCP2(CO)4, for pyrolysis or carbonization of pitches were seleted according to the defined criteria, and the kinetic behaviors of the

1996 TMS ANNUAL MEETING AND EXHIBITION

carbonization of coal tar pitch - petroleum coke mixtures with and without catalysts were mainly investigated by thermogravimetric analysis (TGA). It was found that the mass loss over the major range of temperature for releasing of non-coking volatiles was increased with the catalysts, while a high coke yield (to binder pilCh) was still possible. The observed kinetic effects of the catalysts will be discussed in connection with the anode baking process. 4:50pm ASTUDY OF MEANS FOR IMPROVING PREBAKEDANODESAND REDUCING THEIR CONSUMPTION DURING ELECTROLYSIS: S. F. Egorov, H. Post, V. A. Sverdlin, Russia, 198147, St-Petersburg, Malodetskoselsky av, 26-7, HACA Technocentr, St-Petersburg The modern tendency in the development of aluminum industry is the construction of high efficiency prebaked smelters. Improving cell effiCiency and increasing anode block size not only modify cell operation but also affect the anode performance due to higher thermal and mechanical stresses. The quality of the carbon anodes will affect the cell life and the technical and economical characteristics of the process. A higher prebaked anode consumption will affect the total net cost of aluminum, deteriorate the cell process and the environmental conditions in the potrooms. The objective of this paper is to present means of reducing prebaked anode consumption by improving anode quality. Results of laboratory and pilot plant experiments with various impregnating materials and an inhibiting substance are reported. The former are used to block anode pores and the latter reduces anode consumption during electrolysis.

CAST SHOP TECHNOLOGY IV: Metal Treatment Melting and Filtration

Sponsored by: LMD Aluminum Committee Progmm Organizer: Julian V. Copenhaver, Technical/Quality Manager, NSA A Division of Southwire, PO Box 500, Hawesville, KY 42348 Tuesday, PM February 6, 1996

Room:AIO Location: Anaheim Convention Center

Session Co-Chairpersons: Dr. Diran Apelian, Worcester Polytechnic Institute, 100 institute Road, Worcester, MA 01609-2280; Mr. Jerry Dassel, Commonwealth Aluminum Company, PO Box 480, Lewisport, KY 42351

2:00pm SAMPLING OF ALUMINIUM MELTS AND DETECTION OF NON-METALLIC INCLUSIONS USING LAIS: D. Sampath, P. G. J. Flick, J. Pool, W. Boender and W. van Rijswijk, Corporate Research Laboratory, Hoogovens Groep BV, PO Box 10,000, 1970 CA IJmuiden, The Netherlands The compelling needs to qualitatively/quantitatively monitor the inclusion content of aluminium melts led to the development and use of several inclusion assessment methods. The requirement to quantify the inclusion content arises from the adverse effects of non-metallic inclusions on the product quality of aluminium alloys. Simplicity coupled with ease of transport caused the LAiS method to be one of the preferred techniques to measure the inclusion contents. Several sampling trials were conducted with the LAIS method. An optimized sampling procedure was developed. The sampling time is primarily influenced by a combination of melt inclusion content and temperature. By examining the LAiS samples, different types of non-metallic inclusions present in production melts have been identified and will be presented. 2:20pm A STUDY OF THE ROLE OF PARTICLE CONCENTRATION AND FLUID VELOCITY ON DEEP BED FILTRATION BEHAVIOUR: F. Andres AcostaGonuilez, E. RuizA.,A. H. Castillejos E., M Mendez N., J. C. Escobedo B., ONVESTAV Unidad Saltillo, Apdo. Postal 663, 25000 Saltillo, Coahuila, Mexico In this work, a physical model of interconnected cells of ceramic foam filters is used to visualize the role of particle concentration and fluid velocity on deep bed filtration behaviour. The design of this model is based on the ReynOlds number, the Gravity number, and the ratio of the particle diameter to the cell diameter. It was found that the values for these dimensionless numbers for molten aluminum flowing at typical superficial velocities between 1-3 cmfs through ceramic foam filters, having a pore diameter between 1-3 mm, are similar to those obtained in a physical model consisting of a viscous glycerol-water mixture carrying bakelite or pliolite particles through several interconnected void glass spheres. These spheres are 2.54 em inner diameter, and the whole set of spheres is submerged in a transparent container filled with the liquid. In this way, refraction of light and lateral fluid flow are minimized. The observations were registered and analyzed using a high speed camera interfaced to a motion analysis system. It was found that at a high enough fluid velocity, flow recirculation occurs in the pores. Under this situation, it could be expected that the probability of particle attachment on the filter walls would increase. Also, a high particle concentration induces collisions among the particles, which results in changes of their original trajectories. The observations indicate that the perturbation enhances the probability of particle retention inside the filter cells. 2:40pm EVALUATION OF A BONDED PARTICLE CARTRIDGE FILTRATION SYSTEM: Dr. Wolfgang Schneider, Dr. H. P. Krug, YAW. Vereinigte Aluminium Werke AG, Georg-Von-Boeselager-Strasse, 25D-5300 Bonn, Germany; V. Dopp, Foseco G.m.b.H., Postfach 1220, 0-4280 Borken, Germany; Dr. N. I. Keegan, M. D. Bryant, Foseco Internaional Umited, 285 Long Acre, Nechells, Birmingham, England B75JR Metal cleanliness is a major issue in today's aluminium casthouse, especially in the production of critical products such as canstock, litho sheet and foil. Bonded particle

B69

cartridge filters are widely regarded as the most effective means available for inclusion removal from critical production items. V.A.W. and Foseco have carried out a joint programme of evaluation of a Metaullics 14 tube cartridge filter system in conjunction with ceramic foam filters and an in-line degassing unit in various configurations. The ceramic foam filters ranged from standard, coarse pore types to new generation allceramic bonded, fine pore types. Metal cleanliness was assessed using LiMCA, PoDFA, and LAIS sampling techniques, as well as metallographic and scanning electron microscope examinations. This paper outlines the findings of this work which was carried out a V.A.W. 's full scale experimental D.C. slab casting unit as Neuss in Germany. 3:00pm THE FILTRATION OF MOLTEN lXXX SERIES ALUMlNUM ALLOYS WITH RIGID MEDIA TUBE FILTER: K. Hoshino, T. Nishizaka, Mitsui Mining & Smelting Company Ltd., Corporate R&D Center, 1333-2 Haraichi, Ageo-shi Saitama 362, Japan; T. Yoshida, K. Kakimoto, Mitsui Mining & Smelting Company Ltd., TKR Division, 2-26-6 Higashinihonbashi, Chuo-ku Tokyo 103, Japan Recently it has been increasing that molten IXXX series aluminum alloys are filtered by Rigid Media tube filters (RMF). In that case, it occasionally happens that the pressure drop across the RMF rapidly increases in relatively early stage. We have investigated inclusions captured in the RMF and conducted some model tests to clear up the cause. Consequently, the mechanism of the rapid increase of pressure drop is given. 3:20 pm BREAK 3:30pm LIQUID METAL FILTRATION AND DISTRIBUTION USING FIBERGLASS FABRIC: C. Brochu, A1ean International Ltd., PO Box 1250, Jonquiere, Quebec, Canada G7X 4K8; R. Dault, J. Dery, Pyrotek Incorporated, 2400 Boulevard Lemire, Drummondville, Quebec, Canada J2B 6X9; Sylvain P. Tremblay, Pyrotek Incorporated, 1623 Manic, Chicoutimi, Quebec, Canada G7K 1G8 In the last decade, the aluminum industry has mainly focused on improving metal quality by working on furnace and casting practices and especially on in-line treatment units. fiberglass fabric is widely used throughout the industry, being often the last tool hefore the mold; and in this respect, few efforts have been reported to better understand or use fiberglass fabric in molten aluminum filtration and distribution applications. This paper will review the fiberglass fabric options. The different fiberglass fabric coatings and finishes will be described, as well as their properties and characteristics versus molten aluminum. Fiberglass fabrics use in filtration applications will be defined in function of selected parameters such as opening, finish, throughput, etc. Finally, the metal distribution will be discussed. A better understanding of the fiberglass fabric finishes and bag configuration improve the metal distribution and, ultimately, the fmal ingot quality. 3:50pm IMPROVED EXTRUSION BILLET PRODUCTIVITY AND QUALITY USING A DUAL STAGE CERAMIC FOAM FILTRATION SYSTEM: Leonard S. Aubrey, Michael A. Cummings, Clifton L. Oliver, SELEE Corporation, 700 Shepherd Street, Hendersonville, North Carolina 28792, Mitch M. Chow, EI Campo Aluminum Company, Route 3, EI Campo, TX 77437 The presence of non-metallic inclusions in aluminum extrusion billets can significantly affect both plastic deformation during the extrusion process and final surface quality. Traditionally, secondary extrusion billet casters have relied on a combination of furnace fluxing and a ceramic foam filter to obtain the reqnired melt cleanliness. With increasing quality demands on extrusion billet quality to obtain better surface quality and more complex shapes, as well as reduced press cycle times, there has been a trend towards the use of finer pore ceramic foam filters. The use of finer pore size ceramic foam filters (> 3Oppi) generally requires the use of an expensive in-line spinning nozzle degasser system. Secondary billet casters often do not have the floor space for an in-line system. In addition, the efficient utilization of an in-line system requires the use of chlorine, which is coming under increasing restrictions. A novel two-stage filter bowl and preheat system has been developed to allow the use of 50 ppi ceramic foam filters without the use of an in-line spinning nozzle degasser system. Improved metal cleanliness was quantified by taking LAIS inclusion samples prior to and after the installation of a dual stage filtration system utilizing a combination of 30 and 50 ppi filters compared to an existing 30 ppi filter. Statistical analysis of press cycle times indicated an overall reduction of 11.4%. 4:10pm EXPERffiNCE OF ALCOA-KOFEM WITH MHD INDUCTION STIRRER: Sandor Petho, Ingot Business Unit Engineering Manager, ALCOAKOFEM KFT., H8200 Szekesfehervar, Verseci u. 1-15, PO Box 102, Hungary I) Operating principle. 2) Description of the equipment. 3) Working experience (7 furnaces). 3.1) Application ofMHD. 3.2) Effect on heal melting process: Melting rate, productivity. Fuel consumption, Dross generation, Dross skimming, Temperature homogeneity, Chemical homogeneity (Forced metal tapping as a chance.) 3.3) Experience on maintenance. 4) Conclusion. 4:30pm SIDE MOUNTED EMS FOR ALUMINIUM SCRAP MELTERS: Magnus Eidem, Gtlte Tallback, ABB Industrial Systems AB, Dept: HS, S-721 67 Viisteras, Sweden; Patrick J. Hanley, ABB Industrial Systems, Incorporated, Metals Division, 100 Madison Corporate Park, Route 6, Brewster, NY 10509 Normally the electromagnetic stirrer (EMS) is placed below the furnace. However, it has recently been found that the EMS can also be placed at the side of the fumace, still giving good stirring. This makes it possible to install EMS on most existing furnaces. The side-mounted EMS is compared with the standard bottom-mounted stirrer with respect to installation, melting time, and flow pattern in the melt. The major conclusion is that a side-mounted EMS is practical and will give about as good a performance as the bottom-mounted. Melting time estimates are based upon 3-D fluid flow and hat transfer predictions in combination with a simplified scrap melting theory. Pre-

B70

dicted melting times are in fair agreement with operational data for non-stirred and electromagnetically bottom-stirred furnaces. 4:50pm ANEW ELECTROMAGNETIC CIRCULATION PUMP FOR ALUMINUM REVERBERATORY FURNACES: Dr. Richard S. Henderson, Director of Technology, Richard C. Chandler, President, Metaullics Systems Company L.P., 31935 Aurora Road, Solon, Ohio 44139; Warwick Brown, Director, EMP Technologies, Repton Road, Willington, Derby, DE65 6BX, England! The benefits of circulating molten mlltal in an aluminum reverberatory furnace are well documented, and include higher productivity, reduced fuel consumption, and excellent metallurgical and temperature homogeneity. Current methods to achieve circulation or metal movement include mechanical pumps, electromagnetic stirrers, and porous plugs. Each of these methods has limitations or drawbacks which can be overcome by the use of an electromagnetic pump of novel design. This new device was developed by EMPT, a UK company, and is marketed in North America by Metaullics Systems Company L.P. It combines high pumping rates with long-term reliability and realistic cost of purchase, installation and operation. In addition, it can be readily retrofitted to existing open-well or dlirect ch8Ql,e aluminum reverberatory fumaces. The EMPT pump has been installed in 6 furnaces in the UK and France over a two-year period. The design and operation of the electromagnetic pumping system will be presented with performance data including melt rates, metal homogeneity and temperature profiles. In addition, a video will be shown of the system processing a variety of scrap feeds. Finally, procedures and costs for accomplishing a retrofit to an existing furnace will be outlined.

DEFORMATION SCIENCE

TMS General Abstract Session Tuesday,PM February 6, 1996

Room: Orange County I Location: Anaheim Marriott Hotel

Session Chairperson: TBA 2:00pm AN OVERVIEW OF DISLOCATION MOBILITY IN CRYSTALS-FOUR DECADES IN RETROSPECT: V. R. Parameswaran, Institute for Aerospace Research National Research Council Canada, Ottawa, Ontario KIA OR6; R. J. Arsenault, Metallurgical Materials Laboratory, Dept. of Materials Nuclear Engineering, University of Maryland, College Park, MD 20742-2115 Studies carried out in the past four decades on dislocation motion in crystals under different velocity regimes are briefly reviewed. The evolution of the techniques starting from slow deformation under quasi-static loading leading to high-strain-rate techniques, such as torsion and impact streS!; pulses and ultrasonic methods, are recapitulated. High speed dislocations subjectee! to lattice dirag effects and their anomalous behaviour at relativistic velocities predil;ted by Weertrnan, are also briefly reviewed, with their implications in high-strain-rabl deformation of materials. 2:20pm DETERMINATION OF THE MAGNITUDES OF PEIERLS STRESS FOR DIFFERENT CRYSTALS: lian N. Wang, Chemistry & Materials Sciences, L-370, Lawrence Livermore National Laboratory, P. O. Box 808, Livermore, CA 94551-9900 The classical derivation of Peierls stress is critically examined. It is found that in Peierls-Nabarro treatments there may exist an error in representing the atom positions across the slip plane. With the correction of this error, it is shown that I) the misfit energy of a dislocation and the lattice friction to dislocation motion have an expected periodicity of the Burgers vector b rather than the unexpected periodicity of bl2 predicted by the original model, and 2) the Peierls stress formula for a dislocation in a lattice in which atoms just above and below the slip plane face each other may not be different from that for a dislocation in a lattice in which atoms alternate across the slip plane, as suggested before. Comparison with experimental data and atomistic calculations suggests that the delived new equation can be used to predict the magnitude of the Peierls stress for predominant slip systems in different crystals, and the accuracy of the prediction may degenerate greatly in the I:ases that the spreading of the dislocation core is non-planar and the application of an external stress changes the core structure significantly. 2:40pm THE BEHAVIOR OF THE BORDONI PEAK IN ULTRA mGH PURITY CU AND ITS IMPLICATION FOR DISLOCATION MECHANISM: S. Okuds Faculty of Engineering, Ibaraki University, Hitacbi, Ibaraki 316, Japan The size of the Bordoni peak is usually believed to decrease after high temperature annealing above recrystallization, becau se of a decrease in dislocation density. Recently, however, in ultra high purity Cu (higher than nominal7-nine) the Bordoni peak was found to grow even after annealing up to 800· C. This was true for specimens heavily defonned by drawing and also for specimens deformed by fatigue at low amplitude. The size of the peak can well be explained as caused by kink pair formation on dislocations which form networks after high temperature annealing. The decay of the peak in ordinary pure Cu after annealing should be attributed to impUrity pinning. Implication of the results for dislocation lIIe<:hanism and arrangement will be discussed in detail. 3:00pm ON THE DEFORMATION MECHANISM IN HARPER-OORN CREEP: lian N. Wang, Chemistry & Materials Sciences, L-370, Lawrence Livermore National Laboratory, P. O. Box 808, Livermore, CA 94551-9900 Harper-Dom creep, a Newtonian dislocation creep, has been observed in many me-

1996 TMS ANNUAL MEETING AND EXHIBITION

tallic, ceramic and silicate materials. Based on an early Weertman's model for power law creep, a micro physical model for Harper-Dorn creep is developed. Two assumptions are made. One is that a steady dislocation density is established not only by the applied stress but by the Peierls stress as well. The other is that the flow process is dislocation glide plus climb with the climb being rate-controlling. It is shown that the predicted dependencies of dislocation density and strain rate on the Peierls stress are in very good agreements with experimental data on a wide range of materials. 3:20 pm BREAK 3:30pm DYNAMIC DEFORMATION RESEARCH AND BIOSPHERIC PROTECTION: THE INTERMESIDNG OF MATERIALS SCIENCE AND ENVIRONMENTAL MANAGEMENT: Arthur H. Pun:ell, Preventive Environmental Management, 1745 Selby Ave., Los Angeles, CA 90024 Materials research and environmental protection are inextricably linked. Materials research leads to advances in material performance; and the better a material performs, the greater the opportunity for applying it to advanced resource conservation and environmental protection systems. Similarly, superior materials performance means reduced long-term resource and environmental impacts of material creation, use, and ultimate discard. Dynamic deformation research has particularly relevant implications for environmental management, since it directly supports the critical pollution prevention strategy of source reduction. Increasing durability, strength, and mechanical diversity (including recyclability) of materials leads directly to reduced energy, materials, and environmental intensiveness, and also facilitates utilization of alternative materials that can entail both enhanced environment/materials compatibility and reduced resource exploitation stress on the biosphere. 3:50pm KINETICS OF RECRYSTALLIZATION IN ADIABATIC SHEAR BANDS, Joy A. Hines, Kenneth S. Vecchio, University of CA-San Diego, Materials Science Group, Dept. of AMES, MC-04I1, LaJolla, CA 92093 Small crystallized grains have been observed to form within the adiabatic shear bands of a variety of metals and alloys. Adiabatic shear bands form under conditions of high strain rate and high shear strain; however, recrystallization behavior under these circumstances is somewhat unclear. In this study, the kinetics of two existing models for recrystallization, strain-induced boundary migration and subgrain coalescence, are compared with the shear band deformation and cooling times and the resulting recrystallized grin sizes in adiabatic shear bands of several materials. This comparison shows that the resulting recrystallized microstructure could not have occurred statically during cooling down of the shear band and must have occurred during deformation. The existing diffusion-dominated models for recrystallization will be shown to be inadequate to describe dynamic recrystallization under the present conditions of high strain, high strain rate adiabatic shear banding. It is hypothesized that an additional mechanical effect associated with subgrain rotations contributes to dynamic recrystallization under these conditions. Preliminary calculations of subgrain rotations during deformation will be presented.

EXPERIMENTAL METHODS FOR MICROGRAVITY MATERIALS SCIENCE: SESSION III

Sponsored by: MSD Thermodynamic and Phase EquJlbrla Committee, NASA Microgravlty ScIences Program Organizers: Dr.Robert A.Schiffman, R.S. Research Inc., Barton, VT 05822; Professor Carlo Patuelli, Universita di Bologna, 1-40126 Bologna, Italy

Tuesday,PM February 6, 1996

Room: Orange County 2 Location: Anaheim Marriott Hotel

Session Chairperson: Professor Carlo Patuelli, Universita di Bologna, 1-40126 Bologna,ltaly

2:00pm OVERVIEW AND PRELIMINARY MODELING AND GROUND-BASED RESULTS FOR THE OLIPSE EXPERIMENT: E. Atchley, T. Hamilton, A. Hedayat, J.E. Smith, Jr., The Department of Chemical and Materials Engineering, The University of Alabama in Huntsville, Huntsville, AL 35899 The Optizone Liquid Phase Sintering Experiment (OLiPSE) will begin on STS-67 studying samples of Fe, Co, Ni, Ni-W cemented with either Ag or Cu under vacuum. Eight ampoules, four from the Ag and Cu systems, each containing 10 samples, will be transported aboard STS-67 and processed as a function of time in the Russian Optizone furnace aboard space station MlR. Video tape recordings of the melt and solidification process will be available during the LPS process using this unique piece of Russian flight hardware. After processing, the samples will be returned on the next available space shuttle for metallurgical analysis. Thermal modeling of the multi-sample ampoule that installs into the Russian Optizone furnace have been accomplished. The results of this modeling will be compared to measurements of the tbermal profiles and power consumption of the ground based Optizone furnace. Specific details about the mission, its scientific objectives, numerical modeling results and an overview of the Optizone furnace and its ground performance will be presented. 2:20pm NUMERICAL SIMULATION OF DIRECTIONAL SOUDIFICATION OF HYPERMONOTECTICALLOYSWITHRESIDUALGRAVITY:Y.Arikawa,J.B.

1996 TMS ANNUAL MEETING AND EXHIBITION

Andrews, Department of Materials Science and Engineering, University of Alabama at Birmingham, Birmingham, AL 35294; S.R. Coriell, w.F. Mitchell, B.T. Murray, National Institute of Standards and Technology, Gaithersburg, MD 20899 In this study, a numerical simulation scheme is under development for simplified hypermonotectic alloy solidification. The simulation is based on a mathematical model following an analysis for eutectic systems. However, supplemental effects including high diffusivity in one of the product phases and Marangoni effects at the interface are considered as well. The interface shape has been calculated. Also, the level of residual gravity and g-jitter which cause the disruption of the interface solute layer have been investigated utilizing a popular fluid dynamics package. Results agree with analytical estimations. 2:40pm STATISTICAL DESIGN OF EXPERIMENTS· A POWERFUL AND ESSEN· TIAL TOOL FOR MICROGRAVITY MATERIALS RESEARCH: S. Varma, S.E. Prasad, Sensor Technology Limited, B.M. Hi-Tech Div., P.O. Box 97, Collingwood, Ontario, Canada L9Y 3Z4. A. Ahmad, T.A. Wheat, M1LICANMET, Natural Resources Canada, 405 Rochester Street, Ottawa, Ontario, Canada KIA OGI The progress in microgravity materials processing research has generally been slow and limited due to the lack of flight opportunities, the number of parameters that can influence the processes, and high associated costs for microgravity and extensive ground based experiments. Selection of optimum conditions of microgravity experiments, which could clearly show the effect of gravity on the studied response, are necessary but not easy to obtain in multi-component materials whose properties may be simultaneously affected by a number of process parameters. Statistical design of experiments is a powerful, and in our opinion an essential, tool for microgravity materials research as it has the capability of systematically screening a number of process parameters simultaneously, and then mapping the effect of most important process parameters to select the optimum conditions for microgravity materials processing experiments. Examples of a number of statistical deSigns used for the ground-based and microgravity experiments, including those to be used for heavy metal fluoride glass processing experiments on the Russian MIR space station in 1996, are presented to demonstrate the effectiveness of this approach. 3:00 pm BREAK 3:20pm DEVELOPMENT OF A UNIVERSITY BASED LOCAL TELEOPERATIONS SITE FOR THE PERFORMANCE OF EXPERIMENTS IN MICROGRAVITY: M. B. Koss, M.E. Glicksman, L.T. Bushnell, J.C. LaCombe, Materials Science and Engineering Department, Rensselaer Polytechnic Institute, Troy, NY 12180-3590; E.A. Winsa, NASA Lewis Research Center, Cleveland, OH 44135 The Isothermal Dendritic Growth Experiment (lOGE), a NASA space flight experiment, was designed to provide microgravity and ground-based data on dendritic growth velocities and radii of curvature for a critical test of theory. A key element in the operation of this experiment during its frrst flight aboard the United States Microgravity Payload (USMP-2) on STS-62 in March 1994 was its use of telescience for the nearreal-time monitoring and control of the experiment. For the lOGE's second flight (USMP-3), scheduled for February 1996 on STS-75, a remote operations and control center will be established at Rensselaer Polytechnic Institute. This local site, using a subset of the data and communication systems available at the Payload Operations and Control Center (POCC) at the Marshall Space Flight Center (MSFC), will provide a proof of concept for local teleoperational facilities. In this paper, we describe the telescience capabilities of the Rensselaer Operations and Control Center (ROCC), how we plan to use these capabilities, and how this approach to space flight experimentation might be relevant to the development of remote operations sites for Space Station operations and experiments. 3:40pm QUELD II • A MULTIPURPOSE EXPERIMENTAL FACILITY FOR THE SPACE STATION ALPHA: Reginald Smith, Mark Gallerneault, Kevin Best, Queen's University, Kingston, Canada K7L 3N6; Timothy Smith, Sydney Pugh, Millenium Biologix, Kingston, Canada K7M 7G3 In October 1992, the Space Shuttle Columbia took into orbit Canadian Astronaut Steve Mclean and the CANEX 2 mid-deck payload. One of the facilities was QUELD (Queen's University Experiment in Liquid Diffusion). Throughout the flight, the crew repeatedly processed liquid metal diffusion couples and subsequently brought back to Earth some very interesting diffusion data. In view of the success obtained with the manually operated QUELD unit, the Canadian Space Agency is sponsoring the design and construction of an automated version which is to be placed in the PRIRODA facility soon to be attached to MIR as part of the build-Up to ALPHA. Whilst on ALPHA, QUELD II will process a variety of specimens including metallic and non-metallic diffusion couples, infra-red glasses, ternary compound semiconductors and a number of phase-ripening systems. The experimental facility and the associated research activity will be described. 4:00pm SHUTTLE AND MIR MICROGRAVITY ENVIRONMENT: K. Hrovat, MJ.B. Rogers, Tal-Cut Company, 3355 Richmond Road, Suite 251A, Beachwood, OH 44122; R. Hakirnzadeh, R. DeLombard, NASA Lewis Research Center, 21000 Brookpark Road, Cleveland, OH 44135 The Microgravity Measurement and Analysis Project at NASA Lewis Research Center is tasked with characterizing the microgravity environment of the Shuttle missions on which science experiments are conducted. The on-going work accomplished by this group includes measurement of the microgravity acceleration environment on-board the NASA Orbiters, the Russian MIR space station and, eventually, the International Space Station. This paper will describe the overall microgravity environment ranging in frequency from 100 Hertz down to quasi-steady. The microgravity environment is generated by multiple sources and is dependent on the location within the vehicle. As such, the microgravity environment at one location is highly complex but unique. This

B71

paper will also describe specific examples of major disturbances along with samples of the acceleration environment for those disturbances. Examples of quiet and "noisy" times on tbe Orbiter and MIR will also be illustrated. Knowledge of the expected microgravity environment will assist scientists in properly preparing their experiment andlor experiment sample for the expected accelerations to be encountered during their operation. Information necessary to obtain microgravity environmental data will also he presented.

GAS INTERACTIONS IN NONFERROUS METALS PROCESSING II

Spollsored by: SMD Non-Ferrous Metals Committe, EPD PyrometaDurgy and Pro-

cess Fundamentals

Program Orgallizers: Debabrata Saba, Metals Industties Group,Air Products and Chemicals, Inc., Allentown, PA 18195; WilliamE. Frazier, Naval Warfare Center, Warminster, PA 18974; William P. Ime, Bechtel Group, San Francisco, CA 94105; Prof. David G. Robertson, Department of Metallurgical Engineering, University of Missouri, Rolla, M065401 Thesday,PM February 6, 1996

Room: B8 Location: Anaheim Convention Center

Sessioll ChairpersOllS: WilliamE. Frazier, Naval Warfare Center, Warminster, PA 18974; William P. Ime, Bechtel Group, Inc., San Francisco, CA 94105 2:00pm SOME GENERAL CHARACTERISTICS OF CONTINUOUS GAS-STIRRED REACTORS: L.Li, G. Trapaga, J. Szekely, Department of Materials Science and Engineering, Massachusetts Institute of Technology, Cambridge, MA02139 Abstract not available. 2:20 pm KINETIC STUDY OF As, Sb, AND BI REMOVAL FROM COPPER BY SF6INJECTION: B. Zhao, NJ. Themelis, Columbia University, Dept. of Chemical Engineering, Materials and Mining Engineering, New York, NY 10027 . Laboratory experiments have been conducted on the removal of As, Sb and BI from molten copper by SF6 injection. The results showed that tbe impurities can be removed by SF6 injection in the sequence of Sb>As>Bi and the rate of Bi removal is affected significantly by tbe concentration levels of As and Sb in tbe copper melt. Analysi~ of the experimental data indicates that tbe impurity removal process is under CheffilCal reaction control ratber than by mass transfer of impurity through the molten copper phase. On tbe basis of experimental results, a kinetic model was developed and the order of chemical reaction and tbe volumettic chemical reaction rate coefficients of impurity removal were determined.

2:40pm A STUDY OF THERMODYNAMICS D KINETICS OF NITROGEN ABSORPTION AND ITS REMOVAL FROM LIQUID NI-CR ALLOYS USING THE SIVERTS' TECHNIQUE: L. Petitnicolas, A. Jardy, D. Ablitzer, Ecole des Mines, Pan: de Saurupt, 54042 Nancy Cedex, France The experimental study of nitrogen absorption and its removal from a liquid Ni-20% Cr alloy has been undertaken. We use an experimental apparatus based on tbe Sieverts' technique in order to determine the solubility of nitrogen in tbe alloy, as well ~ tbe kinetics of nitrogen transfer across tbe gas/melt intedace. Experiments show that mtrogen solute equilibrium content can be calculated using th~ cl~sical.Sieverts' ru.le. ~ nitrogen absorption/removal behaves as an overall reaction IOvolvlOg convectivel~lf­ fusive transport of nitrogen solute within tbe bulk liquid phase, as well as tbe CheffilCal reaction at the gaslliquid intedace, which is based on an adsorption-desorption mechanism. Measurements show the nitrogen reaction rate to be well described by first order kinetics, whereas experiments on nitrogen removal reveal that tbe rate is better expressed using a second order equation. 3:00pm NITROGEN GAS REACTIONS WITH MOLTEN ALUMINUM AND ALUMINUMALLOYS: J.F. Flumerfelt, I.E. Anderson, Ames Laboratory, Iowa State University, Ames, IA 500 II -3200 Molten AI and AI alloys containing Si, Mn, n, and Y have been lanced with nitrogen-bearing gases. These lancing experiments provided the means for observing what phases form during exposure to nitrogen-bearing gases for flOite periods of time, <2 minutes, at high temperatures, > 1200·C. The experiments were conducted i? an induction heated drop tube furnace apparatus. The overall premise of these expenments was to select those Al alloys and gas compositions that show promise of forming secondary dispersoid phases, particularly AIN, in the AI mattix. The alloys selected will be ato~­ ized using gas atomization reaction synthesis (GARS) using tbe most favorable conditions. In this case, the chosen AI alloy will be atomized with a nitrogen-bearing gas, such that the AI and gaseous nitrogen species react to form a dispersoid phase within each powder particle. The powder particles that result from this "aggressive" GARS process would be essentially AI metal-mattix composites within a uniform disttibution of dispersoids. To date, only high-purity Al has been atomized using a mild GARS reaction. This work is funded by the Materials Science Division of DOE-BES under contract number W-7405-Eng-82.

3:30pm PHYSICAL MODEL STUDY ON SPLASHING BY TOP BLOWING GAS INJECTION ONTO LIQUID BATH: S. Okabe. H. Sato, F. Tanaka, E. Kimure, Mitsubishi Materials Corporation, Pyrometallurgical Process Laboratory, Omiya, Saitama, Japan Top blowing lance technique is employed by the Mitsubishi Continuous Copper Converter for the continuous supply of reaction gases onto molten bath. In order to minimize the refractory wear caused by the interaction between tbe fumace lining and the melt splash, optimum furnace design and blowing condition are required. An empirical equation obtained by the physical model study enables the estimation of the splattering rate on a vertical plane at a ct~rtain distance from the lance. The effect of the ceiling position was considered as well. The application of the equation to the fumace design of the Mitsubisbi Converter is discussed. 3:50pm INTERACTIONS OF LARGE GAS BUBBLES WITH FLUID INTERFACES: LJollg-Lellg, A. Guerra, N. Gray, Department of Chemical Engineering, University of Melbourne, Parkville, Victoria, Australia 3052 A survey of the interactions of large gas bubbles formed from top and bottom injection of gases into melts was carried out. Spherical cap bubbles were found not to produce much splash when they break through the bath free sudace except for that fonned by plunging of tbe spout its edges. In contrast, the bursting of balloons produces cavities which give rise to liquid jets being ejected from the bath free sudace. The position of the balloons relative to the bath free !:udace is an important parameter in the formation of the liquid jet. This is compared with bubbles formed from top and bottom gas injection to give plausible mechanism for the formation of splash in intensive bath smelting processes. It is found that the interaction of the free sodace movement with the gas plume is important in predicting the splash behavior of gas injected systems. 4:10pm THE CHLORINE-ALUMINUM INTERFACE: G. Saevarsdottir, T. Gudmundsson, T.I. Sigfusson, University of Iceland, Science Institute, Dunhaga 3, 107 Reykjavik, Iceland For decades the aluminum industry ~~s been practicing gas treatment of aluminum melts. This treatment is of great impommce because of its influence on gas and inclusion removal from the melt. The gas - melt interaction is greatly influenced by the gas composition and melt temperature. This paper gives an overview of the interaction between argon/chlorine gas mixtures and the aluminum melt. It covers the thermodynamics of the reaction between the chlorine and number of different elements commonly found in liquid aluminum. The paper reports an experimental study of the behavior of the chlorine - aluminum interface which is the key to understanding the role of chlorine in the liquid metal. 4:30pm INTERACTIONS BETWEEN MELT AND CHLORINE IN CHLORINE FLUXING OF ALUMINUM: Q. Fu, J. W. Evtflls, Department of Materials Science and Minerai Engineering, University of California, Berkley, CA 94720-1760 Although chlorine fluxing to remove magnesium has been widely practiced in the aluminum industry for many years, ther,e appear to have been few fundamental studies of the kinetics and transport phenomenal involved. The paper describes such a study in wbich the rates of reaction of individuall argon/chlorine bubbles with aluminum/magnesium melts were measured. The detachment of bubbles from a submerged nozzle, at variable depth in the melt, was detected by a microphone and the arrival of the bubble at the top sudace by conductivity probes. Analysis of exit gases and mass balances yielded tbe overall progress of reaction. Experimental results are interpreted in terms of heterogeneous reaction kinetics and ma,ss transport to the bubble-melt intedace. 4:50pm IMPACT OF GASES ON METAL PROCESSING AT INDAL: C. Ramasamy, Indian Aluminum Company, Ltd., Taloja, Dt. Raigad, Maharashtra 4 I0208, India The winds of liberalization blowing across India have sent the tariff barriers, that have provided protection to industry, crllshing. Where as, imports are perceived to be a major threat, it is also seen as an opportunity to penetrate global market with quality upgrades. India will also have to tighten environmental regulations in line with G7 countties over time. Our company mission on new equipment selection calls for design that is World Class in terms of quality, cost effectiveness and environment. In tbe above context, use of appropriate gases for processing aluminum assumes great importance. For example, aluminum processing involves chlorine - nitrogen during holding furnace operation, argon during treatment of dross and nitrogen/generator gas for annealing of metal. The paper discusses the currentlmd proposed quality standards in India and the implication of this on the uses of gases on one hand and the increasing sttingency of ecological regulation on the other. Essentially, therefore, it boils down to the question of balancing cost, quality and environment and the need, perhaps, to develop technology that addresses the needs. 5:10pm OXY-FUEL COPPER MELTING FOR INCREASED PRODUCTIVITY AND PROCESS ENHANCEMENT: c.T. Ebelillg, AGA Gas, Inc., Cleveland, OH 44131 The purpose of this paper is to discuss the implementation and results derived from installing oxy-fuel burners in a 180-ton copper reverbatory melting furnace. This installation has been in operation since the fall of 1993. Specific results will be discussed in the areas of melt chemistry, fum ace modifications, productivity, process enhancements and the combustion system capabilities. Similarly, the methodology of the installation and progress of more recent modifications will be addressed.

3:20 pm BREAK

B72

1996 TMS ANNUAL MEETING AND EXHIBITION

A SYMPOSIUM DEDICATED TO PROFESSOR JOHANNES WEERTMAN'S SEVENTIETH BIRTHDAY: SESSION IV: Plastic Deformation Of Ice

Sponsored by: JL SMDIMSD Nuclear Materials and MSD Flow and Fracture Committees and FEMS (Federation of European Materials Societies) Program Organizers: R.I. Arsenault, Deptartment of Materials Science and Nuclear Engineering, University of Maryland, College Park, MD 20742-2115; David Cole, CRREL, 72 Lyme Rd., Hanover, NH 03755; Todd Gross, Department of Mechanical Engineering, University of New Hampshire, Durham, NH 03824; Gernot Kostorz, Institut ftIr Angewandte Physik, E11I Hanggerberg, CH-8093 ZUrich, Switzerland; Peter Liaw, Department of Materials Science and Engineering, University of Tennessee, Knoxville, TN 37996-2200; Sivan Pararneswaran, NRC-Institute for Aerospace Research, Ottawa, Canada KIA OR6; Howard Sizek, Inco Alloys International Inc., Huntington, WV 25705-1771

Tuesday. PM February 6, 1996

Room:Orange County 3 Location: Anaheim Marriott

Session Chairpersons: David M. Cole, US Army Cold Regions Research and Engineering Laboratory, Hanover, NH 03755; Kathryn A. Forland, GMI Engineering and Management Institute, Flint, MI 48504-4898

2:00 pm Invited VALIDATION OF WEERTMAN'S THEORY OF BASAL MORAINE FORMATION BY BOTTOM FREEZING BENEATH POLAR ICE SHEETS: A.I. Gow, US Army Cold Regions Research and Engineering Laboratury, Hanover, NH 03755 Weertman postulated on theoretical grounds that wet-bottomed ice sheets could, by refreezing of basal meltwater generated by geothermal and frictional heating incorporate debris from the underlying bed. This "freeze-on" of basal moraine was offered as an alternative mechanism to that of "shearing-in" to explain the origin of so-called shear moraines at the margin of the Greenland Ice sheet. A viable test of Weertman's "freeze-on" hypothesis came several years later with the successful drilling to bedrock of the Antarctic Ice Sheet at Byrd Station. Penetration of the icelrock interface at this location was accompanied by significant up-welling of water into the 2164 meter-deep drill hole. The bottom 4.83 meters of ice core from Byrd Station contained abundant stratified debris testifying to episodic incorporation of bed sediment into the base of the Antarctic Ice Sheet. In this paper I will describe the nature and disposition of this basal moraine, which together with stable isotope and entrapped gas analyses of the enclosing ice are consistent only with a ''freeze-on'' mechanism as first postulated by Weertman. Formation of debris-laden ice in cores from the bottom of the Greenland Ice Sheet at Camp Century is also attributed to "freeze-on" of bed sediment. "Freeze-on" is the likely predominant mechanism of basal moraine formation in ice sheets at their pressure melting points. 2:30 pm Invited ON CRYSTAL ORIENTATIONS IN FLOATING ICE SHEETS: W. F. Weeks, Geophysical Institute, University of Alaska Fairbanks 99775-7320. J. Wettlaufer, Applied Physics Laboratory, University of Washington, Seattle, WA 98195 The facts that crystals of glacier ice show pronounced c-axis orientations associated with deformation and that ice crystalS forming in the atmosphere show striking morpho
ogical changes associated with changes in temperature and vapor pressure are well known. Less well known and less studied are the striking c-axis orientations that are commonly observed in lake and sea ice and presumably also when ice forms on the bottoms of ice shelves; a situation that occurs at a number of locations in the Antarctic. The general process at work here is commonly referred to as geometric selection; it is a process that also occurs during the solidification of a number of other materials; and that in ice can proceed to several different end points. In addition, if the salinity of the water is greater than -I 0/00, the coaxes are parallel to the current and, if the salinity is less than -1 0/00, the coaxes are perpendicular to the current. Finally these apparently contradictory experiments and observations are examined in view of recent developments in crystal growth theory and observations in the Russian literature in an attempt to develop a consistent, overall explanation. 3:00 pm Invited SCALE EFFECTS IN THE FRACTURE OF ICE: I.P. Dempsey, Department of Civil and Environmental Engineering, Clarkson University, Potsdam, NY 13699-5710 Do the fracture properties at large scale differ significantly form those at small scale? Can laboratory-scale testing be used to predict properties at large scale? What particular problems are faced with large scale testing? How useful are the size-effect theories developed to date? The answers to these questions are key to the applicability of fracture mechanics in the fields of ice engineering, concrete design and rock mechanics. The present focus is to determine the size-scale effect from nonlinear fracture models in terms of the ice microstructure, rate of loading and thermal regime. The required large-scale in-situ arctic experiments posed very challenging logistical and mechanical problems. A very important objective is to evaluate the scale effect over the range 10- 1 m (laboratory) to IOZ m (full thickness field experiments) employing a combination of on-site and laboratory experiments linked to theoretical fracture and constitutive models as well as theoretical ice property models. Nonlinear fracture models can then be predict the scale effect over the range Io-Im to approximately 103m. 3:30 pm Invited THE COMPRESSIVE FAILURE OF ICE: E.M. Schu/son, Thayer School of Engi-

1996 TMS ANNUAL MEETING AND EXHIBITION

neering, Dartmouth College, Hanover, NH 03755 Compressive failure limits the forces exerted by floating ice sheets and other ice features on engineered structures and is thus important in design. This paper considers the failure process in terms of its dependence on both environmental (temperature, loading rate, confinement) and microstructural factors (e.g., grain size, shape and texture), and discusses the deformation mechanisms. Particular attention is paid to brittle failure under multiaxialloading and to the transition from ductile to brittle behavior which occurs upon exceeding a critical strain rate. Evidence is presented for crack nucleation via grain boundary sliding, for the frictional crack sliding/wing crack mechanism of crack growth, and for the role of crack-tip dislocation creep in suppressing brittle failure. It is suggested that lessons learned from a stUdy of ice may be relevant to the failure of other materials. 4:00 pm Invited THE STRUCTURE OF A PLEISTOCENE GLACIATION CYCLE: T. Hughes, Department of Geological Sciences, Institute for Quaternary Studies, University of Maine, Orono, ME 04469 The structure of a typical Pleistocene glaciation cycle, as revealed by proxy data for ice volume, consists of a rather abrupt drop in sea level of about 50 m, followed by an irregular "sawtooth" drop of another 70 m or so over a period of about 80,000 years, and then terminating abruptly in less than 10,000 years. During the last glaciation cycle, the "teeth" of the sawtooth drop averaged about 10,000 years in duration, roughly corresponding to precession hemicycles of Earth's rotation axis. Each "tooth" displayed its own "sawtooth" pattern, with peaks every 1000 to 3000 years. In all cases, changes were abrupt, occurring in a few decades to a few centuries. This structure is interpreted as resulting from Northern Hemisphere ice sheets that formed rapidly when sea ice grounded on Arctic continental shelves, thereby damming rivers flowing into tbe Arctic. These marine ice sheets advanced onto land from relatively stable ice domes over Hudson Bay, the Gulf of Bothnia, and the Barents, Kara, and East Siberian Seas, lowering sea level some 50 m Unstable advances and retreats from these domes, caused by thawing and freezing of the bed beneath them, caused "sawtooth" sea-level drops of another 70 m, with overall lowering caused by continuing glacio-isostatic depression under the ice sheets. Termination of the glaciation cycle took place when this depression passed a critical threshold, causing the ice sheets to collapse. Computer ice-sheet reconstructions of a typical glaciation cycle are presented. 4:30pm CREEP OF A COMPOSITE AND FLOW OF DIRTY ICE IN GLACIER: T. Mori Ryobi Ltd., Toshima, Kita, Tokyo 114, Japan; K. Wakashima, Tokyo Inst. Tech. Nagatsuta, Midori, Yokohama 226, Japan Using a micromechanics approach, creep of a composite, a material consisting of a plastic matrix and reinforcements (inclusions), is examined. It is shown that if no relaxation process is operative, the creep of a composite eventually stops, contrary to the prediction of a model used by continuum plasticity. This is due to the accumulation of internal stresses. The combination of sliding and mass transport on the interfaces of inclusions plays a role, equal to or more effective than plastic deformation allowed in the inclusions, to relax stresses. If these two relaxation processes operate, the balanced state of stresses is achieved in the matrix and inclusions mass transport and sliding, leading to a stationary creep state. 4:50pm RHEOLOGY, MELT MIGRATION AND DILATATIONAL ATTENUATION IN EQUILIBRATED PARTIAL MELTS: R.F. Cooper, Department of Materials Science and Engineering and Geology and Geophysics, University of Wisconsin, Madison, WI 53706 The rheology of partial melts is first-order dependent on the (quasi) equilibrium distribution of the melt phase at the grain scale. The primary rheologic effect of partial melting is to enhance the rates of diffusional creep mechanisms far greater than those of dislocation mechanisms: for a given set of (a, T) conditions, change of dominant mechanism and strain rate increases from 2 to 104 are possible, depending on liquidsolid interfacial equilibrium The partially molten aggregate is thus considered as two interpenetrating fluids, and potentials for relative motion of the solid and liquid phases accompanies plastic deformation. The melt-migration response is distinctly anelastic; as a consequence, dilatational attenuation becomes a significant aspect of the internal friction signature of partial melts. lllustration of these phenomena will be made with data on silicate assemblages. Applications of these ideas range from the characterization of rheology of, and magma motion in, Earth's upper mantle and lower crust, to the forging of ceramics, to the high-temperature creep of ceramics composites.

HIGH TEMPERATURE COATINGS II: Intermetallic/Metallic Coatings II Sponsored by: MDMD Surface Modlftcatlon and Coatings Committee Program Organizers: Narendra B. Dahotre, Center for Laser Applications, University of Tennessee Space Institute, Tullahoma, TN 37388; Janet M. Harnpikian, School of Materials Science & Engineering, Georgia Institute of Technology, Atlanta, GA 30332; Jacob J. Stiglich, PO Box 206, Sierra Madre, CA 91025 Tuesday, PM February 6, 1996

Room:BI Location: Anaheim Convention Center

Session Chairpersons: Alejandro Sanz, Laboratoire de Metallurgie, Ecole Nationale Superieure, De L' Aeronautique EtDe L'espace, 10,Av. Edouard Belin, BP4032, 31055 Toulouse Cedex, France; Janet M. Hampikian, School of Materials Science of Engrg, Georgia Institute of Technology, Atlanta, GA 30332

B73

2:00 pm Invited MODELING THE INTERDIFFUSION OF IUGH TEMPERATURE COATINGS: J.E. Morral, Cheng Jin, W.D. Hopfe, Department of Metallurgy and Institute of Materials Science, University of Connecticut, 97 N. Eagleville Road, Storrs, CT 062693136 During processing and service, high temperature coatings can interdiffuse with their underlying substrate. The microstructure of the interdiffusion zone depends on the initial coating and substrate compositions in a way that can be predicted using newly developed principles of diffusion in complex systems. As an example, these principles will be applied to predicting the microstructure of MCrAlY coatings when they diffuse into Ni-base superalloys at 1200°C. It will be shown, both by computer modeling and experiments on diffusion couples, that a variety of microstructures are possible even when the initial coating is a "ttP alloy and the substrate is a "tty alloy.

Structural materials and coatings usedl for industrial gas turbine hot components are designed to resist high temperature corrosion attack through the development of protective scales of Cr203, AI203 or Si03. The effectiveness of these scales can be reduced by thermal cycling induced stresse!; which can lead to crack nucleation and growth and to consequent spallation. Synergistic interaction phenomena with HTC can play an important role in accelerating the process. A Dean Test was devised at the HTC Lab of CISE to simulate the gas turbine environment; a 48 hours thermal cycling from test (850 and 900'C) to room temperature was performed. The behavior of some Ni and Co-based superalloys, conventional coalings (both PtAI and NiCoCrAIY) and thermal barrier coatings was studied through gravimetric kinetics, loss-of-metal data and SEMI EDS analyses of corrosion products morphology and composition after exposure up to 3000 hrs. Based on the results, a model was applied to predict scale service life reduction due to this mechanism.

2:25pm CUBIC PHASE (Al,Cr)3T1 COATINGS FOR TITANIUM AND TITANIUM ALUMINIDEALLOYS: Dale K. Dewald, Waubik, Inc., P.O. BOX 457, Hancock, MI 49930; Donald E. Mikkola, Department Metallurgical & Materials Engineering, Michigan Technological University, 1400 Townsend Dr., Houghton, MI 49931 Chromium alloyed forms of cubic-modified Al3TI have been identified as potential high temperature protective coatings for titanium-based structural alloys. Coatings of the nominal composition Al67Cr8TI25 have been applied to both TI and TIAI alloy substrates by the Low Pressure Plasma Spray (LPPS) and Electro-Spark Deposition (ESD) processes. The coatings have been subjected to long term cyclic oxidation tests. Details of the coatings and methods will be presented along with the micro-structure and protective behavior of the coatings.

4:15pm IMPROVING CORROSION RESISTANCE OF CAST IRON BY SOLID-GAS CHROMOSILICONIZING: Xiaoxing Qiu, Dev Venugopalan, Materials Department, College of Engineering and Applied Sdence, University of Wisconsin-Milwaukee Siliconizing and chromosiliconizing are an inexpensive method of chemically treating for corrosion and high-temperature oxidation resistance in ferrous materials. Conventionally, siliconizing and chromosili(:onizing by the powder-pack method are used to improve the above properties in cas!. iron. However, limited success has been attained for gray cast iron. For example, corrosion resistance has been raised only two times in hydrochloride acid solution test. This poor performance is due to a high density of porosites and flake graphite in the case layer. The solid-gas method is a way of producing vapor from a solid treating agent directly within the retort. The present word was done on cast iron using solid-gas chromosiliconizing. A very dense layer was obtained on the surface of the cast iron with this processing. This results in a greatly improved resistance to corrosion in the test with hydrochloric acid and hot alkaline solutions.

2:45pm THE FORMATION AND RELEVANCE OF "MARTENSITIC" STRUCTURES WHICH DEVELOP IN ALUMINIDE COATINGS ON A RHENIUM CONTAINING NICKEL BASE SUPERALLOY: S.B. Newcomb, A.E. Dent, W.M. Stobbs, Department of Materials Science and Metallurgy, Cambridge University, Pembroke St., Cambridge CB2 3QZ, UK The compositional and structural changes that occur in aluminide and platinum aluminide coatings on a rhenium containing nickel base superalloy during high temperature thermal cycling have been assessed using "edge-on" TEM approaches. It is found that AI diffusion, up into the oxide and down into the substrate, happens differently for the two tvpes of coating with, in particular, differing relative effects on the development of the sub-coating substrate '"(Ii structure in relation to the relative segregation of the other elements present. Highly defective "martensites" based on Ni5Al3 form in the coatings and their role will be discussed. Their defective nature suggests that thev enhance the rates at which the coatings degrade but there is evidence that they are also important in relation to stress relief. 3:05pm COATING INDUCED LIFE REDUCTIONS OF SINGLE CRYSTAL SUPERALLOY GAS TURBINE BLADE MATERIALS: J. Bressers, Institute for Advanced Materials, Joint Research Centre, EC Petten, The Netherlands Single crystal nickel-based alloys are widely used as materials in aero gas turbine blades because of their excellant resistance to high temperature deformation. Coatings are used in order to provide the blades with adequate protection against environmental degradation. An issue of major concern is the effect of the presence of the coating on the operational life of the blades. This paper reports the results of a study evaluating the life of the single crystal nickel-based superalloy SRR99 under thermo-mechanical fatigue (TMF) loading, in the uncoated condition as well as protected with a nickel aluminide diffusion coating aod with an MCrAIY overlay coating. Different TMF cycle types with a minimum tempe:ature of 3OO-C and maximum temperatures of 850'C and 1050'C, and with different minimum-to-maximum strain ratios were used to simulate the thermal and mechanical loads at critical volume elements of a blade. At strain ranges relevant to aero engine blade service, the presence of both coatings causes the TMF life of SRR99 to decrease, the reduction by a factor of five being largest in the case of the overlay coating. The results are analyzed in terms of the effect of the coating on the number of cycles for microcrack initiation and on the crack growth rate. The faster crack growth in the coated material, which is found to be largely responsible for the observed reduction of the TMF life, is due to a combination of poor coating properties and composition/microstructure changes in the substrate layers just beneath the coating. 3:25pm THE HIGH TEMPERATURE OXIDATION PERFORMANCE OF ALVMINA COATINGS ON NI-BASE ALLOYS: M.R. Hendrick, J.M. Hampikian and W.B. Carter, School of Materials Science and Engineering, Georgia Institute of Technology, Atlanta,GA Alumina coatings were applied to nickel-base alloys by combustion chemical vapor deposition (CCVD). CCVD is an open air CVD technique performed in a flame produced from a nebulized liquid fuel solution. As- deposited alumina coatings on fused quartz were analyzed by transmission electron microscopy (TEM) and found to be crystalline. The oxidation kinetics of coated specimens were ascertained by oxidation tests carried out in pure flowing air at 800-1100'C and compared to the kinetics of uncoated specimens. The morphology and composition of the alumina coatings were determined by scanning electron microscopy, energy dispersive X-ray spectroscopy, Auger electron spectroscopy and X-ray photoelectron spectroscopy. 3:45 pm BREAK 3:55pm THE RESISTANCE OF SOME GAS TURBINE HOT COMPONENTS COATINGS TO IUGH TEMPERATURE CORROSION AND THERMAL CYCLING DAMAGE: G.P. DeGaudenzi, F. Uherti, CISE spa-P.O. Box 12081,1-20134 Milano, Italy; A. Colombo, G. Rocchini, ENEL Spa, DSRlCRAM, Milano, Italy

874

4:35pm HIGH TEMPERATURE CORROSION PROTECTION BY LASER·SURFACE MODIFICATION: Rahul Asthana, Anil Gupta, Ashish Gupta, Ajay Singh, E-277 'Nirupam', Kania Nagar, AGRA282005, India Laser treatment of metallic materials has gained wide interest due to great potentialities of the technique. It has several advantages i.e. localized heating, high temperature and high cooling rates, etc. Laser surface alloying or coating is a good way to prepare barrier layers on to metals. This paper deals with the method of elaboration of several alloys or coating on to iron, nickel, or nibomium by melting a predeposited solid powder and the substrate. The behaviour of the formed coating in Oz at high temperature have also been presented. 4:55pm CORROSION RESISTANTPd·SI COATINGS: LP. Efimenko, E.A. Antonova, L.P. Petrova, Institute of Silicate Chemistry of Russia Academy of Sciences, ul.Odoevskogo,24, korp.2, St. Petersburg, 199155, Russia The Pd-Si coating with palladium content more than 95 weight percent has been formed by powder-annealing technology for protecting superalloys and steels. The influence of temperature, time, and Si content in the composition on the coating structure were studied. The coating is formed at lOOO'C for ten minutes in the presence of liquid phase. After formation the base metal (:ontent in the coating is less than one weight percent. The solid solution formation in the transition zone provides the high adhesion of the coating to the protected metal. The oxidation resistance of the coating at 800'C for fifty hours is practically equal to that of pure palladium. The corrosion of Pd-Si and Pd-Ir-Si coatings in HN03 is studied. It is shown the iridium increses corrosion resistance by ten to twenty times. The aging ,at 700'C for twenty hours improves the corrosion resistance additionally by a factor of 1.5 to 2.

HIGH TEMPERATURE SUPERCONDUCTORS· SYNTHESIS, PROCESSING, AND LARGE SCALE APPLICATIONS IV: Flux Pinning And YBCO Conductor

Sponsored by: Jt. EMPMD/sMD Super
Room: B7 Location: Anaheim Convention Center

Session Chairpersons: P. J. McGinn, University of Notre Dame, Notre Dame, IN 46556; A. Roshko, National Institute of Standards and Technology, Boulder, CO 80803

2:00 pm Invited FLUX PINNING AND DISSIPATION IN IUGH Tc SUPERCONDUCTORS: C. S. Pande. D. U. Gubser, Naval Research Laboratory, Washington, DC 20375-5343 Many phenomena in high Tc supercoDlductors such as magnetic relaxation, thermally activated resistance, and flux noise can be interpreted quantitatively by applying the concept of flux motion under pinning energy barriers. We show that such vortex dynamics must assume flux motion under a well-distributed spectrum of potential wells. We also show analytically that under some plausible assumptions that such a distribu-

1996 TMS ANNUAL MEETING AND EXHIBITION

tion of potential wells must be log normal. Implication of this result in explaining the transport properties of these materials will be discussed. Review of NRL work in high Tc conductor development will also be presented. 2:20 pm Invited THE USE OF PHASE DIAGRAMS FOR THE ENGINEERING OF FLUX PINNING CENTRES IN BSCCO CERAMICS: P. Majewsi, Helsenbergstr. 5, D-70569 Stuttgart, Germany From considerations of the phase equilibrium diagram of the system Bi203-PbOSrO-CaQ-CuO, different annealing procedures were developed to precipitate various second phases in high-temperature superconducting Bi2Sf2CaCu203 and (Bi,Pbl2Sr2Ca2Cu301O ceramics with different chemical compositions. In each case, the precipitation results in an increase of the critical current density of the samples which is believed to express improved pinning properties of the superconducting ceramics, in particular an increased pinning energy, which reduces the probability for thermally activated depinning. 2:40 pm Invited FLUX PINNING ENHANCEMENT IN MELT PROCESSED YBa2Cu30,_ THROUGH RARE EARTH ION SUBSTITUTIONS: P. J. McGinn, C. Varanasi, Dept. of Chemical Engr., Univ. of Notre Dame, Notre Dame, IN 46556 Stoichiometric YBa2Cu3O,_(YI23) samples doped with excess Nd203 and La203 additions were melt processed in air. Because decomposition and reformation of 123 occurs during melt processing, light rare earth ion substitution in 123 takes place. As the light rare earth ion sizes (Nd, La) are comparable to that of Ba2+, partial substitution into both the Ba2+ sites and Y ion sites is anticipated. The light rare earth oxide doped, melt processed 123 shows significant improvement in magnetization compared to undoped samples, likely due to point defects created by substitutions of the light rare earth ions. 3:00 pm Invited MICROSTRUCTURE-PROPERTY RELATIONSHIPS IN EPITAXIAL YBa2Cu3o,_1) THIN FILMS: A. Roshko, F. J. B. Stork, D. A. Rudman, L. F. Goodrich, National Institute of Standards and Technology, Boulder, CO 80803 The excellent high temperature properties ofYBa2Cu3o,_1) thin films make them of interest for many active and passive devices. Understanding the role of microstructure in determining film properties will allow optimization of films for electronic applications, and may also aid in the development of bulk materials with specific properties. We have investigated the influence of microstructure on the superconducting properties ofYBa2Cu3o,_1) films. The zero resistance transition temperature, critical current density, and surface resistance have all been found to be relatively insensitive to the film morphology. However, small systematic trends have been observed which indicate the importance of surfaces and defects in determining the properties of these materials. 3:20 pm BREAK 3:30 pm Invited DISTINCTIVE MICROSTRUCTURE OF YBa2Cu3O,-jilPt OR Ce02 SUPERCONDUCTOR PRODUCED BY NEW POLY CRYSTALLINE SEEDING METHOD: G. Kozlowski, Wright State University, Physics Department, 3640 Col. Glenn Hwy., Dayton, OH 45436; C. E. Oberly, SLPOOX-2, 2645 Fifth St. Ste 13, WPAFB, OH 45433 A new polycrystalline seeding method for providing melt-processed YBa2Cu3O,_ 1)(123) large pellets (up to 2 inches in diameter) or bars (up to 5 inches long) placed on high density Y2BaCu05 (211) substrate is proposed and discussed. The 123 bars can carry a transport current of up to 1,500 A under criterion of I v/cm. Furthermore, a distinctive microstructure of the unusual defects or gaps, with the accumulated Ba-Culiquid, running parallel to (100) and (010) planes in the melt-textured 1231Pt or Ce02 superconductors was discovered. These defects may be interpreted as an instability of the planar front in the presence of fine 211 particles. A theoretical model of planar interface stability confirms a strong dependence of the stability region on the size of 211 particles in the melt.

°

3:50 pm Invited MICROSTRUCTURE DEVELOPMENT IN ISOTHERMALLY MELT-TEXTURED Y-Ba-Cu-O SUPERCONDUCTORS: V. Selvamanickam, Intermagnetics General Corp., Latham, NY 12110 Fabrication of melt-textured Y-Ba-Cu-O superconductors is useful for the development of trapped-flux devices that can operate at 77 K. Isothermal melt-texturing by seed-controlled solidification has been found to be the most effective method of fabricating single-grain materials that are capable of trapping high magnetic fields and exhibiting large levitative forces. In an effort to achieve a high yield of high quality melttextured Y-Ba-Cu-O, we have examined the microstructure development during the solidification process. Microstructure features such as the solidification interface, grain alignment, and Y2BaCu05 precipitates have been analyzed in samples fabricated at various conditions of superheat, undercooling, and cooling rates. The role of dopants that affect diffusivity and SOlid-liquid interfacial energy in the microstructure development has also been investigated. The quality of the materials fabricated under various conditions has been examined by trapped field measurements. This research was sponsored by the Air Force Office of Scientific Research. 4:10pm CRYSTAL GROWTH MODEL AND MORPHOLOGY OF 123 MELT-TEXTURED MATERIALS DOPED WITH 211 AND (RE)z03 PARTICLES: R. Cloots, N. Vandewalle, N. Mineur, M. Ausloos, SUPRAS, University of Liege, SartTilman, B4000 Liege, Belgium Kinetic growth mechanisms of melt-textured (RE)Ba2Cu3O, (with RE=Dy, Er) materials doped with (REl2BaCu05 and (REl203 are extensively discussed. We present optical observations and computer simulations of the kinetic processes. The model is

1996 TMS ANNUAL MEETING AND EXHIBITION

based on the well-known Eden model. Physicochemical interactions between the growth front aod the 21 I andlor (RE)203 particles are taken into account in a Uhlman-ChalmersJackson way. The surface morphology is studied through the TLK model. The 211 particles can be trapped at front instabilities leading to the formation of filamentary structures. Quantitative results are presented through Monte-Carlo simulations. Concerning the influence of the (REl203 influence on the crystal growth process, it is shown that modifications of the solidlliquid interfacial energy leads to patterning formation in the grains. 4:30pm REDUCED OXYGEN PRESSURE MELT PROCESSING OF YBa2Cu30xlAgPd WIRES: M. A. Black, P. J. McGinn, Dept. of Chemical Engr., Univ. of Notre Dame, Notre Dame, IN 46556; M. Tomsic, Plastronics, Inc., Troy, OH Powder-in tube processing has been utilized to fabricate YBa2Cu30x (YBCO)J90Ag10Pd composite wires. These wires have been melt textored at 960°C using reduced oxygen partial pressures (1000 Pa oxygen). Metal alloylliquid phase interaction results in small inclusions of BaCuO liquid within the alloy and dissolution of some Ag by the liquid. The use of reduced oxygen pressure and small Pd additions have been used to avoid melting of Ag in the YBCO system, so that composite wires can be fabricated which will retain their shape. Processing concerns, microstructural development during texturing in reduced ~ environment and the effects of metallliquid interactions will be discussed. 4:50pm TEM INVESTIGATION OF MELT PROCESSED YBa2Cu3O,_y DOPED WITH Y2BaCu05 AND Pt02: J. D. Riches, J. C. Barry, Physics Dept., UOlV. of Queensland, St. Lucia QLD 4072, Australia; A. Bhargava, Centre for Microscopy and Microanalysis, Univ. of Queensland, St. Lucia QLD 4072, Australia; P. McGinn, Dept. of Chemical Engr., Univ. of Notre Dame, Notre Dame, IN 46556 Melt texturing causes a significant increase in the Jc of bulk YBCO samples. Thus, it is of considerable interest to investigate the features of such textured microstructures and how they depend on processing parameters. To this end, a study was made of the microstructure of melt processed YI23 doped with Y211 and Pt~. The samples were processed via the SLMG route and their microstructures were examined via TEM. The grain boundaries, both between Y211 and Y123 grains and also between adjoining YI23 grains, were examined. Investigations into the structure of the grain boundaries, and the presence of any dislocations, defects or secondary phases, were undertaken. Previous studies of this material have shown the existence of Y211 precipitation of a size less than 100 nm. These small precipitates should be more beneficial to flux pinning than the micron sized Y211 precipitates that are usually found in melt processed samples. In this study the hold time above the peritectic temperature for the samples was varied and the effect on the size and distribution of the nm sized Y211 precipitates was examined. 5:10 pm Invited GROWTH OF LARGE-DOMAIN YBa2Cu30x AND THE STUDY OF INTERFACE COUPLING OF DOMAIN BOUNDARIES: Donglu Shi, Krishanu Lahiri, Sandhya, Department of Materials Science and Engineering, University of Cincinnati, Cincinnati, OH 45221-0012 Several seeded-melt texturing (SMT) methods were used to grow large-domain YBa2Cu30x for magnetic levitation applications. With various seed crystals, these SMT methods resulted in large single-domain pellets of high levitation forces. A domain growth model is proposed. We developed a novel method to determine the coupling strength of the domain boundaries. Instead of measuring the transport critical current density across a boundary where many technical problems could occur, this new approach employs the magnetization measurements on a sample that contains a single interface. The coupling strength of any interfaces can be determined by comparing with a standard obtained from a sample without the interface. It was found that the natural grown-in domain boundaries are strongly coupled. However, the coupling strength is less compared to that of the domain without any interface. It was also found that the interface coupling strength has a temperature dependence. At 30 K, it reached a maximum value of 0.85 assuming the standard coupling strength inside the domain is

F= 1.

INTERACTIONS BETWEEN CREEP, FATIGUE AND THE ENVIRONMENT: SESSION II

Sponsored by: SMD Mechanical Metallurgy Committee Program Organizers: W.w. Milligan, Department of Metallurgical and Materials Engineering, Michigan Technological University, Houghton, MI 49931; D.L. Davidson, Southwest Research Institute, PO Box 28510, San Antonio, TX 78228-0510; M.F. Henry, General Electric, CRD, KI-MB229, Schenectady, NY 12345; H.W. Sizek,lNCO Alloys International, PO Box 1958, Huntington, WV 25720 Tuesday, PM February 6,1996

Room: Grand D Location: Anaheim Marriott Hotel

Session Chairpersons: M.F. Henry, General Electric, CRD, KI-MB229, Schenectady, NY 12345; H.W. Sizek, INCO Alloys International, PO Box 1958, Huntington, WV 25720

2:00pm GAS PHASE EMBRITTLEMENT OF SUPERALLOYS: D.A. Wood/onl, Materials Performance Analysis, Inc., 1737 Union Street, Suite 543, Schenectady, NY 12309 This talk describes the phenomenology, mechanisms, and implications of high temperature gas phase embrittlement (GPE), associated with intergranular penetration of oxygen, and other aggressive species such as sulfur and chlorine. Examples of its oc-

B75

currence in nickel, cobalt, and iron-based superalloys are given, and it is shown to be manifested in a degradation in tensile ductility and impact resistance, creep fracture, and fatigue crack initiation and propagation. The detailed grain boundary processes associated with this embrittlement are explored relative to studies of various grades of nickel. These data lead to radical implications, which should change our general understanding of creep cavitation and high temperature fatigue cracking, and have broad implications relative to the performance of engineering alloys in aggressive environments. A new approach to materials development, design and life assessment, which bas evolved from this work will be outlined. Several methods of preventing or alleviating the damage resulting from oxygen penetration are described. In particular, the mechanical integrity of environmentally resistance coatings in preventing GPE becomes a crucial factor. 2:25pm THE DEVELOPMENT OF HYDROGEN RESISTANT SUPERALLOYS: D.P. DeLuca, Pratt & Whitney, MS 707-20, PO Box 109600, West Palm Beach, FL 334109600 Nickel base superalloys are used extensively in the hot sections of Pratt & Whitney's high pressure oxidizer and hydrogen fuel turbopumps now entering service in the NASA Space Shuttle's main engines. Mechanical property tests of cast y' strengthened alloys have shown substantial deleterious effects when conducted in high pressure gaseous hydrogen. Among the most pronounced (and life limiting) effects are orders of magnitude reduction in fatigue life and accelerated fatigue crack growth. Unexpectedly, the most extreme effects are observed at ambient temperature rather than elevated temperature. A summary is presented of Pratt & Whitney's efforts to develop cast y' hot section alloys which are resistant to the unique challenges posed by hydrogen embrittlement at low temperatures. Mechanisms of hydrogen degradation and the effects on fatigue and fracture will be discussed. 2:50pm A STUDY OF MICROSTRUCTURAL AND ENVIRONMENTAL EFFECTS ON FATIGUE CRACK PROPAGATION IN MODEL NI·BASE SUPERALLOYS: S.D. Antolovich, School of Mechanical and Materials Engineering, Washington State University, Pullman, WA 99163-2920; A de Bussac, SNECMA, BP 81,91003 Evry, France lbis study addresses some aspects of environmental damage in large grained model Ni·base alloys at intermediate temperature (427'C) and high frequency (10Hz) for which fatigue damage is expected to predominate. Four different compositions were used to evaluate the effects of APBE, mismatch, and volume fraction of y'. The alloys were heat treated to produce two y' sizes. Testing at both 25'C and 427'C showed that FCP resistance was slightly increased at elevated temperature, and was higher for compositions and heat treatments that promote reversible, planar slip and increased surface roughness. lbis suggests the importance of closure effects. Airlvacuumlair tests at elevated temperature allowed for separation of environmental and closure effects through an analysis of FCP data at the air/vacuum transitions. These tests demonstrated a strong negative environmental effect on the FCP resistance of the superalloys tested. However, while oxygen was damaging it was shown to promote crystallographic cracking and thus high roughness. 3:15pm A MODEL FOR OXIDATION·ASSISTED CREEP CRACK GROWTH IN APM SUPERALLOY: L Remy, P. BemMe, Centre des Materiaux de l'Ecole des Mines, URA CNRS 866, BP 87, 91003 Evry, France The purpose of this paper is to show the importance of oxidation on creep-fatigue and to outline how oxidation can be taken into account in life prediction, in the case of a PM superalloy, Astroloy. lbis alloy has been studied in a fine-grained condition (5 mm grain size after extruding and isothermal forging), and tested in creep-fatigue and pure creep in air at 65OC. The effects of air, vacuum, and frequency on the crack propa· gation rates and crack paths will be discussed. A life prediction model is proposed which considers the coupling between diffusion, oxidation, and inelastic strain. The inelastic strain is further partitioned into plastic and viscoplastic components. The model is in very good agreement with experimental results, particularly with respect to the effects of initial Kmax on the transient creep crack growth regime. lbis work was sponsored by the DRET (French Ministry of Defence) and was conducted in collaboration with SNECMA engineers. 3:40 pm BREAK 3:45pm GRAIN SIZE DEPENDENCE OF FATIGUE CRACK PROPAGATION IN A P/ M NI·BASE SUPERALLOY: c.P. Blankenship, Jr., M.F. Henry, General Electric, Research and Development, PO Box 8, Schenectady, NY 12301 Processing routes were developed to produce grain sizes of 7, 15 and 90 JUD, in PIM Ni·base superalloy Rene88DT. Heat treatment schedules were applied to ensure that y' sizes and distributions were identical, regardless of grain size. These microstructures are not indicative of production material, yet they provide a means to isolate the contri· bution of grain size to fatigue crack propagation resistance. Fatigue crack growth tests were run at temperatures between 1200F and 1400F with various tensile hold times. Significant improvements in time dependent fatigue crack propagation resistance were demonstrated with increasing grain size. 4:10pm STRESS ACCELERATED GRAIN BOUNDARY OXYGENEMBRITTLEMENT ON CREEP CRACK GROWTH BEHAVIOR OF IN'18 SUPERALLOY: B.S.·J. Kang, G. Zhang, P. Uu, M. ElIathur, Mechanical and Aerospace Engineering Depart· ment, West Virginia University, Morgantown, WV 26506 Oxidation environment can significantly reduce the tensile fracture ductility and notch stress rupture lives of several nickel-base superalloys at elevated temperatures. This brittle failure is often referred to as "stress accelerated grain boundary oxygen (SAGBO) embrittlement." In this research, the mechanism of SAGBO embrittlement on IN718

B76

superalloy was investigated. In situ, full Held crack tip displacement fields of IN718 SEN specimens under loads at 650'C were measured using high temperature moire interferometry and elastic crack growth W:ilS observed for specimens tested in air. The experimental stress intensity factors and crack-tip plastic yield zones were found to be in good agreement with the theoretical predictions based on linear elastic fracture mechanics. Post-mortem microstructural analysis showed the existence of Nb-rich oxide films at grain boundaries near the crael; tip region. This suggested that SAGBO embritllement was caused by oxygen diffusion to grain boundaries in high stress reo gions and reacting with Nb to form thin, Mtlle Nbz05 oxide layers on grain boundary surfaces. Chemical composition of the N~~05Iayers was verified using Auger electron spectroscopy. 4:35pm INTERACTIONS BETWEEN CREEl', FATIGUE AND ENVIRONMENT IN DAMAGE TOLERANT NIOBIUM ALUMINIDE INTERMETALLICS: J. DiPasquale, F. Ye, W.O. Soboyejo, Department of Materials Science and Engineering, The Ohio State University, 2041 College Road. Columbus, OH 43210-1179 The interactions between creep, fatigue and laboratory air/vacuum environments are discussed in this paper for damage tolerant niobium aluminide intermetallics. These inelude the effects of load/environmental interactions of crack initiation and propagation behavior onder monotonic and cyclic loading. The possible creep·fatigue·environment interactions are modeled using fracture mechanics- and chemistry-based models. Coating schemes are also proposed for thE: control of environmental interactions. 5:00pm STRESS RUPTURE OF A SiClSiC COMPOSITE: T.E. Steyer, F.W. Zok, Materials Department, University of California, Santa Barbara, CA 93106. D.P. Walls, United Technologies, Pratt and Whitney, West Palm Beach, FL 33410 The stress rupture response of a SiC/SiC [0/90) plain weave composite bas been studied at a temperature of 9OO'C. At this temperature, failure is governed by an oxida· tion embrittlement mechanism. Tests haVE: been performed at various stresses ranging from the matrix cracking stress (-70 MPll) to values approaching the composite fast· fracture ultimate tensile stress. The effects of pre-cracking at a high stress at ambient temperature on the stress rupture charactelristics at 900c have also been examined. The effects of oxidation on microstructure havl: been studied using both SEM and scanning Auger microscopy to examine the evolution of chemical species present in the matrix and fiber/matrix interphases. The implications of oxidation on durability and lifetime are explored.

INTERNATIONAL SYMPOSIUM ON REACTIVE METALS: Processing 1& Applications II: Molten Salts and Met,als Sponsored by: LMD Reactive Metals Committee Co·sponsored by: Canadian Institute of Mining, Metallurgy & Petroleum, Montreal, Canada; The Japan Institute of Metals, Sendai, Japan; Mining & Materials Processing Institute ofJapan, Tokyo, Japan; Society fOlt Mining, Metallurgy & Exploration, Uttleton, CO Program Organizers: B. Mishra, Department of Metallurgical & Materials, Engineer· ing., Colorado School of Mines, Golden, CO 80401; G.J. Kipouros, Department of Mining and Metallurgical Engineering, 1I:chnical University of Nova Scotia, Halifax, Nova Scotia, Canada B3J 2X4; R.G. Reddy, Department of Chemical and Metallurgical Engineering, University of Nevada, Reno, NV 89557 Co-organizers: W.A. Averill, Rocky Flat,;, Inc., Golden; R.G. Bautista, University of Nevada - Reno, Reno; M.C. Bronson, Lawrence Livermore Natl. Lab., Livermore; J.A. Sommers, Teledyne Wah Chang Albany, Albany. C.B. Wilson, The Dow Chemical Company, Freeport Tuesday, PM February 6,1996

Room: B9 Location: Anaheim Convention Center

Session Chairperson: Dr. R.G. Reddy, D€:partment of Chemical and Metallurgical Engineering, MS 170, University of Nevada, Reno, NV 89557; Dr. D.R. Sadoway, Department of Materials Science and Engineering, Massachusetts Institute of Technol· ogy,77 Massachusetts Avenue, Room 8-109, Cambridge, MA 02139

2:00pm A NEW ELECTROLYTIC MAGNESIUM PRODUCTION PROCESS: R.A. Sharma, Physical Chemistry Dept., General Motors NAO R&D Center, 30500 Mound Road, Box 9055, Warren, MI 48090-9055 Existing magnesium choride electrolysis and thermal magnesium oxide reduction processes for producing magnesium are described and their limitations are pointed out. There theoretical background of a patented, new process is outlined. In this process, magnesium oxide is dissolved ina rare earth chloride containing electrolyte and electrolyzed to produce magnesium and oxygen like that of alumina in the Hall-Heroult process. It is also shown that the efficiency of tbe existing magnesium chloride elec· trolysis process should be improved greatly by adding a rare earth chloride. In both cases, magnesium produced is expected to be free from the detrimental iron, nickel, copper and boron impurities. 2:25pm INVESTIGATION OF THE SYSTEM Mg.Nd.O·CI: J.E.Vindstad, H. Mediaas, T. Ostvold: Institute of Inorganic Chemistry, The Norwegian Institute of Technology, N· 7034, Trondheim, Norway; GJ. Kipouros, Department of Mining & Metallurgical Engineering, Technical University of NOVll Scotia, Halifax, Nova Scotia, Canada B3J 2X4; O. Tkatcheva, High Temp. Electrohemistry Institute, Russian Academy of Sci-

1996 TMS ANNUAL MEETING AND EXHIBITION

& Phase EquHlhrla Committee Program Organizers: R. G. Reddy, Department of Chemical and Metallurgical Engineering, University of Nevada, Reno NV 89557; S. Viswanathan, Oak Ridge National Lab., Oak Ridge, TN 37831-6083; J.C. Malas, Wright-Patterson AFB, OH 45433-6533

ences, 620219 Ekuterinburg, Russia The binary phase diagram NdCI3-NdOCI has been measured by thermal analysis (chloride liquidus) and oxide solubility measurements (oxychloride liquidus). the melting point of NdCI3 was detennined to be 759±I'C, in good agreement with the value of 759'C cited by Knacke, Kubaschewski, and Hesselmann. r The eutectic temperature was detennined as 739±I'C. In the pseudo-binary system MgCI2-NdOCl, liquidus temperatures have been measured at up to 15.7 mole % NdOCl. Melt samples have been withdrawn at 850, 800, and 730'C, and analyzed for oxide content (Leco TC-436) and MglNd ratio (ICP). The results indicate that a ternary oxygen-containing solid compound precipitates upon addition of NdOCI(s). A stoichiometric equilibrium constant for the equilibrium: MgCI2 (I) + NdOCI (l)->MgO (s) + NdCI3 (1) has been calculated from the melt compositions at the different temperatures.

Tuesday, PM February 6, 1995

2:50pm SOLUBILITY OF MgO IN MgCl2-NaCI-NaF MELTS: H. Mediaas, J.E. Vindstad, T. Ostvo1d, Institute of Inorganic Chemistry, The Norwegian Institute of Technology, N-7034, Trondheim, Norway The solubility of MgO in MgCI2-NaCI-NaF melts have been measured for melts with constant and varying concentration of Mg2+ with addition of NaF. The fluoridechloride melt samples have been analyzed by Leco TC-436 for total oxide content, and the different complexes have been quantitatively separated. The oxide content in the binary melt (MgCl2-NaCI) has also been analyzed by Iodoetric titration. The increase is smaller when NaF is added to a binary M~12-NaCI melt, compared to when MgCI2 is added together with NaF to keep the Mg2+ concentration constant. It seems to be possible to distinguish between two different oxide containing complexes in the ternary MgCI2-NaCI-NaF melt. The fluoride-containing complex oxidizes carbon at higher temperatures than the "MgOCI" complex. An introduction of 1.7 mole % NaF in industrial electrolytes does not seem to have any effect on the oxide solubility.

2:00pm A 3D NUMERICAL MODEL FOR REMOVAL OF INCLUSIONS TO GAS BUBBLES: B.i. Hoo, S.T. Johansen, SINTEF Materials Technology, N-7034 Trondheim-NTH, Norway, and B. Rasch, Hydro Aluminium, Metallurgical R&D Centre, N-6601 Sunndalsora, Norway The flow field set up by the standard Hydro gas purging unit has been predicted by a numerical model. In addition, turbulent trajectories of gas bubbles released from the rotor are predicted. The deposition rate of inclusions on the bubble surface is allowed to depend on local conditions. Bubbles passing regions with strong pressure gradients, such as close to the rotor, will experience large slip velocities and large deposition rates. This model can deal with any distribution of bubble sizes and inclusions. The results found from the multidimensional flow predictions are compared to the prediCtions due to a much simpler metallurgical engineering approach. The differences between these two approaches to predict the same process will be discussed. The model predictions are compared to experimental results for removal of SiC-particles from aluminium.

3:15 pm BREAK 3:30pm ANODIC OVERVOLTAGE ON GRAPHITE IN LI2C03-LlF-CaF2 MELTS: P.T. Velu, R.G. Reddy: Department of Chemical and Metallurgical Engineering, MS 170, University of Nevada, Reno, NV 89557 The anodic overvoltage on graphite anode during electrolysis of Li2C03 in LiFCaF3 melts at 1093 K was measured. A steady state current-voltage metod was used. The experimental results plotted with anodic overvoltage as a function of log current density, showed a straight line relationship given as as 11'" 4.17 + 1.38 log i. A two electron controlled charge transfer reaction between the complex oxyfluoride and graphite anode is proposed. The limiting current density in the melt is detennined to be 0.011 Alcm2. 3:55pm AN IN-BATH LIQUIDUS TEMPERATURE MEASUREMENT FOR MOLTEN SALTS AND SLAGS: EJ. Grimsey, N. Li, X.Y. Yan, Western Australian School of Mines, P.O. Box 597, Kalgoorlie WA, Australia 6430; A. Warczok, T.A. Utigard, University of Toronto, Toronto, Ontario, Canada MSSIA4 A technique is being developed to provide a rapid liquidus measurement for salts and especially metallurgical slags. A copper or nickel cylinder is immersed into a molten bath and a computer based data-logging system is used to record the heating profile of a type K thermocouple centered within the cylinder. The heat transfer is affected initially by the presence of a thin crust on the cylinder surface which freezes from the bath. When the crust melts, the rate of heating shows a subtle increase. Numerical differentiation is used to detect the on-set of this increase so as to provide an indication of the liquidus temperature of the melt. This paper outlines the technique and its application to a numher of molten salts and slags. Factors which affect the sensitivity of the measurement are discussed also. 4:20pm CaO INSULATOR COATINGS AND SELF-HEALING OF DEFECTS ON vALLOYS IN LIQUID LITHIUM AND LITHIUM-CALCIUM: J.-H. Park, T.P. Kassner: Energy Technology Division, Argonne National Laboratory, 9700 S Cass Avenue, IL 60439 Electrically insulating and corrosion-resistant coatings are required at the liquid meta1l structural interface in fusion flfSt-waiUblanket applications. The electrical resistance of CaOcoatings produced on V-5%Cr-5%TI by exposure of the alloy to liquid lithium that contained 0.5-85 wt. % dissolved calcium was measured as a function of time and temperatures between 250 and 6OO'C. The solute element, Ca in liquid Li, reacted with the alloy substrate at 400-420'C to produce a CaO coating. Resistance of the coating layer measured in-situ in liquid Li was '" 106 n at 400'C. Thermal cycling between 300 and 700'C changed the coating layer resistance, which followed insulator behavior. These results suggest that thin homogeneous coatings can be produced on variously shaped surfaces by controlling the exposure time, temperature and composition of the liquid metal. The technique can be applied to various shapes (e.g. inside/outside of tubes, complex geometrical shapes) because the coating is formed by liquid-phase reaction. Examination of the specimens after cooling to room temperature revealed no spallation, but homogeneous crazing cracks were present in the CaO coating. Additional tests to investigate the insitu self-healing behavior of the cracks indicated that rapid healing occurred at 360'C. This work has been supported by the US Department of Energy, Office of Fusion Energy Research, under contract W-31-109-Eng-38.

MATERIALS PROCESSING FUNDAMENTALS SYMPOSIUM· IV

Sponsored by: Jt. EPDIMDMD SyntheSiS, Control, and Analysis In Materials Processing Committee, EPD Process Fundamentals, Aqueous Processing, Copper, Nickel-Cobalt, PyrometalIurgy, Lead, Zinc:, TIn Committees, MSD Thermodynamic:

1996 TMS ANNUAL MEETING AND EXHIBITION

Room:AI6-17 Location: Anaheim Convention Center

Session Chairpersons: S. Viswanathan, Oak Ridge National Laboratory, Oak Ridge, TN 37831-6083; J.C. Malas, Wright-Parterson AFB, OH 45433-6533

2:25pm EQUILIBRIUM PARTITION RATIOS IN MULTI COMPONENT STEELS: M.S. Bhat, D.R. Poirier, P.M.N. Ocansey, Departruent of Materials Science and Engineering, The University of Arizona, Tucson, AZ 85721 The equilibrium partition ratios of alloy elements in two multicomponent steels were detennined. The ratios can be used in models to simulate dendritic solidification. In one system, the equilibrium partition ratios of Mn, Si and C were detennined to cover a scope of concentrations pertaining to the solidification path of a high carbon steel (91125). In the second, the equilibrium partition ratios of C and Cr were measured for concentrations relevant to the solidification of a bearing steel (52100). The experimental method employed and the effect of sample preparation on the quantitative analyses are also discussed. 2:50pm MODELLING OF LEDGES AND SKULLS: VaughanR. Voller, DepartruentofCivil Engineering, 500 Pilsbury Drive, University of Minnesota, Minneapolis, MN 554550220 In many high temperature processes involving liquid melts the common practice is to solidify a fraction of the melt adjacent to the containment walls; a practice that can provide protection of the structural components of the process. The position, shape and transient behavior of these so called "skulls" and/or "ledges" are directly dependent on the thermal process conditions. This paper presents a model of a generic ledge process that can predict both the steady shape of the ledge and its transient behavior driven by time dependent process conditions. Two numerical implementations of the model are presented and compared: (I) based on a space grid of continuously deforming control volume finite elements, and (2) based on a conjugate grid of fixed control volumes and deforming boundary elements. 3:15 pm BREAK 3:25pm MODELING THE COLUMNAR-EQUIAXED TRANSITION IN CASTINGS: S. Sundarraj, G. Upadhya, U. Chandra, Concurrent Technologies Corporation, 1450 Scalp Avenue, Johnstown, PA 15904 The morphological transition between columnar and equiaxed structures is an important solidification phenomenon frequently observed in castings, especially complex parts with varying cross-section. The mechanical properties and the occurrence of defects such as hot tears are strongly dependent on the structure formed in the casting. Hence, it is important to understand the conditions leading to columnar or equiaxed structures. Extensive work has been done in modeling solidification systems. Almost all of these models are based on the assumption that the structure formed in the casting is either purely columnar or purely equiaxed. The prediction of the Columnar-Equiaxed Transition (CET), however, requires a solidification model which accounts for the formation of both columnar and equiaxed structures, along with related microscopic and macroscopic solidification phenomena. Such a model is presented which: (i) takes account of both columnar and equiaxed structures; (ii) accounts for microscopic phenomena, such as solidification kinetics and microsegregation, coupled with macroscopic heat transfer; (iii) satisfies solute conservation; and (iv) is easy to implement in commercial finite element and finite difference software. The proposed model has been validated with available experimental results. A sensitivity analysis has also been carried out to detennine the effects of the model pararueters on predicting the CET position. This work was conducted by the National Center for Excellence in Metalworking Technology, operated by Concurrent Technologies Corporation under contract to the U. S. Navy, as part of the U. S. Navy Manufacturing Science and Technology Program. 3:50pm A MONTE CARLO SIMULATION OF THE DEVELOPMENT OF MICROSTRUCTURE IN AN ALLOY SOLIDIFYING WHILE FLOWING OVER A CHILL: RA. Stoehr, Materials Science and Engineering, 848 Benedum Hall, University of Pittsburgh, Pittsburgh, PA 15261

B77

A Monte Carlo simulation of solidification of a metal alloy flowing over a chill has been developed. Velocity, temperature, and solute concentration profiles in the bulk flow regime are calculated on a fairly coarse grid by macroscopic techniques. Bulk flow conditions are presumed to be externally determined (i.e., quasi-forced convection), and a given distribution of nuclei is carried with the entering fluid. In the interface region, Monte Carlo simulation is used to model solidification on a much finer scale. Microscopic longitudinal and transverse velocity fluctuations are addressed by a shuffling routine. Growth, shrinkage, and possible annihilation of solid regions are modeled based on probability factors related to the local supersaturation. Capillarity and crystallographic considerations can be included in the supersaturation. The model has been related to the type and orientation of microstructures observed in aluminumcopper alloys solidified during flow over a water cooled chill. 4:15pm NEW DIMENSIONLESS HEAT TRANSFER CORRELATIONS FOR LIQUID METALS: Stavros A. Argyropoulos, Department of Metallurgy and Materials Science, University of Toronto, Toronto,Ontario, Canada M5S IA4; Anthony C. Mikrovas, Cominco Ltd., Trail, British Columbia, P.O. Box 2000, Canada VIR 4S4; Don. A. Doutre, Alcan International Ltd. Kingston Research and Development Centre, PO Box 8400, Kingston, Ontario, Canada K7L 5L9 This paper presents a method for deducing dimensionless heat transfer correlations in liquid metals for natural and forced convection conditions. This method was applied to liquid metals with different Prandtl numbers, namely, aluminum and steel. Solid spheres of aluminum and steel, initially at room temperature, were immersed into liquids of aluminum and steel, respectively. In the ftrst set of tests the spheres were immersed under natural convection conditions, and in the second, under forced convection conditions. The melting time of the spheres was carefully monitored and it was related tu the average heat flux from the metal bath to the immersed sphere. New dimensionless heat transfer correlations will be presented for natural and forced convective heat transfer in liquid metals. The validity of these new dimensionless heat transfer correlations was examined with a different set of experiments in an independent setting. This validation work will be shown.

MATERIALS SCIENCE OF CONTACTS AND METALLIZATION AND INTERCONNECTS III Sponsored by: EMPMD Thin FUm & Interfaces Committee Program Organizers: Jagdish Narayan, Dept of Matls Sci & Engrg, Box 7916, NC State University, Raleigh, NC 27695; John Sanchez, Advanced Micro Devices, M3 160, PO Box 3453, Sunnyvale, CA 94088 Tuesday, PM February 6, 1996

Room: Orange County 5 Location: Anaheim Marriott Hotel

Session Co-Chairs: F. Meyer, University of Paris, Orsay Cedex,France; R J. Nemanich, North Carolina State Universitiy, Raleigh, NC 27695

2:00 pm Invited METALSILICIDES-ACTIVEELEMENTSOFULSICONTACTS:C,M.Osbum. J.Y. Tasi, Department of Electrical and Computer Engineering. North Carolina State University, Raleigh, NC 27695-7911 As device dimensions scale to the 0.1 J.Un regeme the self-aligned silicide (SALICIDE) contact technology increasingly becomes an integral part of both the ultra-shallow junction and the MOSFET device itself. This paper will discuss the effect of silicide materials and formation processes on silicide stability, junction consumption, the ability to accurately profile shallow junctions, and contact resistance in series with the channel. The use of silicides as diffusion soorces (SADS) provide an important pathway towards optimization silicide technology. Diffusion of boron and arsenic from nearly epitaxial layers of CoSi2' formed with bilayers ofTI and Co, offer good silicide stability, ultra-shallow,low-leakage junctions, and low contact resistance. 2:30pm STUDIES ON GRAIN GROWTH AND WETfING BEHAVIOR OF AN Co-Sn ALLOY AND PURE Cu FOR ADVANCED CHIP INTERCONNECTS: J. Zhang, H.K. Kim, K.N. Tu, Department of Materials Science and Engineering, University of California, Los Angeles, CA 90095 Copper-Tin alloys have been studied as interconnects for advanced device applications where a high resistance to electromigration is required. A bulk CU2 at. % Sn alloy was prepared and used to study the effect of Sn on Cu grain growth. The alloy was heat-treated at 700, 800, and 900°C and the average grain sizes, grain morphology and orientation were quantified and compared with those of pure Cu. No secondary grain growth in the Cu-Sn alloy was observed, in contrary to pure Cu. The grain growth rate and mechanism in both systems will be discussed. A comparison of wetting behavior of eutectic SnPb solder on both materials will also be given. 3:00 pm Invited EPITAXIAL TIN BASED CONTACTS FOR SILICON DEVICES: J. Narayan, R. D. Vispute, Department of Materials Science and Engineering, North Carolina State University, Raleigh, NC 27695-7916 Fabrication of modem integrated circuits heavily relies on TISi2 compound (C-54 phase with specifiC resistivity of 15-cm) and an orthorhombic structure (a=O.825, b= 0.478, c=O.854 nm). However, this structure is incompatible with diamond cubic, and, therefore, it is not be able to grow epitaxially on silicon (a=O.543 nm). The polycrystal-

B78

line TISi2 films have the following problems: (i) interface roughness, and (ii) grain boundaries along which dopants diffuse out. Recently, we have grown epitaxial TINI Si(I00) and multilayerd epitaxial CulTIN/si(IOO) heterostructures by domain matching epitaxy where four lattice units of TIN matc:h with three of silicon, and six units of TIN match with seven units of copper (Narayan, US Patent # 5406123, Apnl 11, 1995). We also found that epitaxial TiN has 15 -cm resistivity with excellent barrier properties. In addition, Schottky barrier height of TIN is dose to that ofTISi2 (O.SeV). Thus for next generation device CMOS structures, CuffiN/Si(lOO) contacts hold considerable promise, particularly since Cu is a low resistivity metal (2-cm) and is resistant to electromigration. 3:30 pm BREAK 4:00 pm Invited MICROSTRUCTURE AND ELECTRICAL PROPERTIES OF SrRu03 THIN FILMS ON LaAI03 SUBSTRATES: F. Chu ,Q'x' Jia, X.D. Wu, T.E. Mitchell, Materials Science and Technology Division, Los Alamos National Laboratory, Los Alamos NM87545 Conductive oxide SrRu03 thin ftlms have been deposited using pulsed laser deposition on LaAl03 substrates at different substrate temperatures. Structural and microstructural properties of the SrRuOJILaAI03 system have been studied using xray diffraction and scanning electron microscopy. Electrical properties of SrRu03 thin films have been measured. It was found that the film deposited at 250"C is amorphous, showing semiconductor-like temperature dependence of electrical conductivity. The filmdeposited at 425°C is crystalline with very fine grain size (IOO-200A), showing both metallic and semiconductor-like temperature wlpendence of the electrical conductivity in different temperature regions. Epitaxial [CI() I) growth of the ftlms takes place above 6500C deposition temperature. The film deposited at 775°C shows a resistivity of 280 jUl-cm at room temperature. The optimized deposition conditions to grow SrRu03 thin films on LaAI03 substrates have been found. Possible engineering applications of SrRu03 thin films deposited at different temperatures will be discussed. 4:30pm LaN103 AND CU3Ge CONTACTS IN YBazCu30,_ FILMS: D. Kumar, R.D. Vispute, O. Aboelfotoh, S. Oktyabrsky ,J. Narayan, Department of Materials Science and Engineering, North Carolina State University Raleigh, NC 27695-7916; P.R Apte, RPinto, Tata Institute of Foundamental Research, Bombay, India Epitaxial LaNi03 (LNO) metallic oxide lihin films have been grown on YBa2Cu3D7_ x(YBCO) films at 700'C using pulsed ia.'ler evaporation technique. The sheet resistance of LNO films was measured to be 201square at 300 K which decreases to 0.7501 square at lOOK indicating good metallic~y of LNO films. The contact resistance of LNO-YBCO thin film interface was found to be reasonably low (of the order of 10-4 Ocm2 at 77 K) which suggests that interface formed between the two films is quite clean and LNO can emerge as a promising metal electrode-material to YBCO films. A preliminary investigation related to chemi(:al compatibility of Cu3Ge as an alternative metallic electrode to YBCO films would also be presented. 5:00pm THERMOMECHANICAL FATIGUE IBEHAVIOR OF LAP SHEAR JOINTS: R.K. Mahidhara, S. Hariprasad, High Performance Materials Inc., Campus Box 1087, B. Goldstein, S.M.L. Sastry, K.L. Jerina, Mechanical Engineering Department, Campus Box 1185, Washington University, St. Louis, MO 63130 Abstract not available.

MEGATRENDS IN MANUFACTURING

Sponsored by: MaterlaIs Design and Manufacturing Division ProgramOrganizers: Paul S. Follansbee, Office of Economic Competitiveness, DOE Defense Program DP-14, 1000 Independence Ave. SW, Washington, DC 20585; Krishna Vedula, Iver E. Anderson, Iowa State Univ., Matls Sci & Engrg, 3053 Gilman, Ames, IA50011; William Frazier, NavaAir Development Ctr, Code 606, Street Road & Jacksonville Road, Warmister, PA 18974; Carole Trybus, Argonne National Labs., PO Box 2528 Bldg T-2, Idaho Falls, ID 83403-2528 Tuesday, PM February 6, 1996

Room: Marquis 1&2 Location: Anaheim Marriott Hotel

Session Chair: rBA

The importance of manufacturing to the economic well being of a nation is well estab-

lished. Manufacturing industries, however, are facing ever increasing challenges posed by technological innovation, stiff foreign competition, environmental challenges, and workforce issues. The reaction to these challenges is changing the face of manufacturing across the world. Materials are an essential component of every manufactured component, and as such, materials scientists and engineers must be knowledgeable of and responsive to the rapid changes throughout the manufacturing sector. The porpose of the Megatrends '96 Symposium is to provide a forum for the dissemination of crucial information from recognized leaders regarding emerging directions within select materials industries. Education leaders will describe related transformations within colleges and universities, and examples of effective: partnerships between industry, universities, and Federal laboratories will be highlighted.

1996 TMS ANNUAL MEETING AND EXHIBITION

MELT·SPINNING, STRIP CASTING, AND SLAB CASTING IV: Microstructures And Open Forum

Sponsored by: MDMD Solidification and SCAMP CommIttees Program Organizers: E.F. Matthys, Mechanical Engineering Department, University of California, Santa Barbara, CA 93106; w.G. Truckner, Technical Director-Product Development, Alcoa Technical Center, Alcoa Center, PA 15069

Tuesday, PM February 6, 1996

Room: B3 Location: Anaheim Convention Center

Session Chairperson: E.F. Matthys, University of California, Mech Engrg Dept., Santa Barbara, CA 93106

2:15pm EXPERIMENTAL INVESTIGATION ON THE SPACING OF DENDRITE ARRAYS: Q. Han, X. Wan, and J.D. Hunt, Department of Materials, University of Oxford, Parks Road, Oxford OXI 3PH, United Kingdom Experiments have been carried out to measure the arm spacing of dendrite arrays and the spacing distribution. The results show that there is a minimum spacing under which overgrowth takes place and a maximum spacing above which arm splitting occurs. The minimum spacing agrees well with recent predictions. Primary dendrite arm spacing lies between the minimum and the maximum spacings, but the spacing distribution varies randomly as the freezing front advances. When a dimensionless velocity exceeds a critical value, the dendrite array becomes irregular. The critical value is investigated both experimentally and theoretically. 2:40pm EFFECT OF GRAIN SIZE UPON OXIDATION RESISTANCE IN RAPIDLY SOLIDIFIED Fe-20Cr-5A1-O.05La ALLOYS: K.S. Lee, Y.C. Shin, K.H. Oh, H.Y. Ra, Department of Metallurgical Engineering, Seoul National University, 56-1 Shilimdong, Kwanak-ku, Seoul 151-742, Korea The strip of Fe-2OCr-5AI-0.05La alloys was produced by Planar Flow Casting. Microstructures evolved pure ferritic phase and grain size was less than 101l.m. Preferred orientation of strip was (110) texture. The effect of the preferred orientation and grain size on oxidation resistance were studied comparing with the conventionally rolled alloys, which has grain size of about 20 - 6OOj.lm and (200) texture. Oxidation resistance was remarkably improved in melt spun strip. This improvement is related with good adherence of oxide scale, the purpose of this study is to find the effect of grain size and crystal orientation on the adherence of oxide scale. Addition of La element was favorite to the supression of grain growth at high temperature and showed good oxidation resistance. Effect of crystal orientation seemed to be related with recrystallization and regarded to growth of (200) orientation at high temperature. 3:05pm MELT SPINNING OF A SHAPE MEMORY ALLOY: A.A Tseng, S.Y. Yang, Department of Mechanical Engineering and Mechanics, Drexel University, Philadelphia, PA 19104; C.L. Un, J.S. Chen, Materials Research Center, Temple University, Philadelphia, PA 19122; M.V. Nathal, Advanced Metallics Branch, NASA-Lewis Research Center, Cleveland, OH 44135 Rapidly solidified TINi ribbons for shape memory applications have been produced by melt spinning under an argon atmosphere. The effects of rapid solidification with different processing conditions on the structure and transformation characteristics of the alloy have been investigated. The martensitic transformation characteristics of meltspun specimens are described in relation to the fine-grained structure and the ordered state of the material. The paper details the melting of the master alloys, production of rapidly solidified ribbons, and processing of the ribbon to sheet by hot pressing and hot rolling. 3:30 pm BREAK 3:50pm MICROSTRUCTURAL DEFECTS IN THIN SHEET TWIN-ROLL CAST ALUMINUM ALLOYS: Ming Yun, S.A. Lockyer, J.D. Hunt, Department of Materials, University of Oxford, Parks Road, Oxford, United Kingdom 1\vin-roll casting has been used to produce thin sheet (1-4 mm thick) in aluminum alloys. The microstructural defects found in thin sheet twin-roll cast material have been investigated. The defects include surface bleeds, channel segregates, deformation segregates and inhomogeneity in grain structure. This paper describes an experimental observation and metallographical examination of the defects and anlyzes the relationShip to the operating parameters. Various generalization can be made for occurrence of the defects in different alloys and mechanisms are proposed for their formation. The defects are illustrated in a number of commercial alloys cast at different gauges and with very different loads. 4:15pm MICROSTRUCTURE EVOLUTION OF AISI 430 STAINLESS STEEL PRODUCED BY TWIN ROLL STRIP CASTER: J.T. Choi, Y.H. Kim, K.H. Oh, H.Y.Ra, Department of Metallurgical Engineering, Seoul National University, 56-1 Shilim-dong, Kwanak-ku. Seoul 151-742. Korea AISI 430 stainless steel strip was fabricated by twin roll strip caster and the thickness of as cast strip was 0.6 - 2.0 mm. The microstructure of as cast strip consists of ferritic matrix, carbides and martensite at grain boundary. observed in the transmission elec-

1996 TMS ANNUAL MEETING AND EXHIBITION

tron microscopy. Martensite was transformed from retained austenite at grain boundary because austenite to ferrite transformation was suppressed due to rapid cooling of strip casting. The thinner thickness of strip, the larger volume fraction of martensite. The texture evolution of the strip was investigated by the analysis of pole figure and orientation distribution function. In as cast strip. strong component of g fibre. {III} /I roiling plane. was observed and typical soidification texture component. <100> texture, was rarely observed. Crystallographic rotation of surface texture component was observed, which showed the asymmetry of pole figure and development of rolling texture. This rotation is due to shear and plane strain deformation near the kissing point during strip casting. 4:40pm A METHOD FOR SOLUTION OF THE LAMELLAR EUTECTICS GROWTH PROBLEM: Ljubomir Andreev Anestiev. Institute for Metal Science. I 574-Sofia. Bulgaria A mudel for the explicit solution of the steady state growth problem of lamellar eutectics has been developed. The proposed model takes into account the complex form of the old-new phases interface. the capillarity effects and the influence of the growth velocity on the solute redistribution between the phases. The obtained results compare favorably with the ones obtained with other mudels. 5:05pm OPEN FORUMNIDEO PRESENTATIONS

METAL MATRIX COMPOSITES

TMS General Abstract Tuesday. PM February 6, 1996

Room: B4 Location: Anaheim Convention Center

Session Chairperson: TBA 2:00pm ON ELASTOPLASTIC RESPONSE OF PARTICULATE-REINFORCED METAL-MATRIX COMPOSITES UNDER MULTI-AXIAL LOADINGS: M. V. S. Ravisankar. Department of Mechanical and Aerospace Engineering and Engineering Science. N. Yu. P. K. Uaw. Department of Materials Science and Engineering. The Univ. of Tennessee, Knoxville, TN 37996 A micromechanics model based on periodic microstructure is developed to predict the overall elastoplastic response of and to examine the microstructure-property interactions in particulate-reinforced metal-matrix composites (MMCs) under multi-axial loadings. The model explicitly accounts for not only the material properties of the matrix and the particulates but also vital microstructural parameters. such as the shape. size and packing density of the particles. Analytic solutions of the average stress and strain increments in the MMCs are obtained by using homogenization techniques along with Fourier series(J). The elastoplastic behavior predicted by the present model are compared with the uniaxial tensile testing results of silicon carbide particulate reinforced (SiCp) aluminum (AI) matrix composites. fabricated by a powder metallurgy extrusion process[2]. The effects of particle volume fraction (0-30-%). matrix properties (2124. 6061 and 7091 AI). and particle aspect ratio on the effective elastoplastic response of AUSiCp composites are also studied. 2:20pm THE DUCTILITY OF METAL MATRIX COMPOSITES UNDER TRANSVERSE LOADING CONDITIONS: Leon L Shaw. Department of Metallurgy and Institute of Materials Science. University of Connecticut. Stom. CT 06269 Continuous-fiber reinforced titanium alloys exhibit superior properties along the fiber axis. However. the transverse properties of these composites have been shown to be lower than desired for some applications. In this study. an effort is made to develop an understanding of the dependencies of the composite ductility on the matrix ductility and the fiber/matrix bond strength under transverse loading conditions. Localized ductility around the fiber was evaluated through a laser interferometer, and compared with the overall ductility of the composites under the transverse loading. Numerical modeling (elastic/plastic) was also carried out to extend the experimental results to accommodate composites with different matrix ductilities and interfacial properties. The experimental and modeling results will be presented and the approaches to improving the transverse ductility of the composites will be proposed. 2:40pm MECHANICAL BEHAVIOR OFAN AI-6092 ALLOY REINFORCED WITH 17.5 VOLUME PER CENT OF SiC PARTICULATES: Kedarnath Poduri. Yan Ma. Terence G. Langdon (Departments of Materials Science & Mechanical Engineering. University of Southern California, Los Angeles, CA 90089-1453) Experiments were conducted to evaluate the mechanical characteristics of an AI-6092 metal matrix composite containing 17.5 vol % of SiC particulates. Creep and tensile tests were performed in air at temperatures in the range from 723 to 823 K. Maximum tensile elongations in excess of 100% were recorded at the lower testing temperatures when using very high strain rates. This paper describes the results of these experiments. including the values of the stress exponent and the activation energy. 3:00pm HIGH-TEMPERATURE DEFORMATION OF A SiC--2124 AI COMPOSITE AND 2124 AI: Yong Li. Farghalli A. Mohamed. Materials Science and Engineering, Dept. of Chemical and Biochemical Engineering University of California. Irvine. CA 92717 The effect of stress and temperature on the high-temperature deformation of S vol. %

B79

silicon carbide particulate reinforced 2124 Al (SiCp - 2124 AI), produced by powder metallurgy, has been studied. 1be experimental data of the composite are examined in reference to those of the matrix alloy, 2124 AI, that was tested under similar experimental conditions. 3:20 pm BREAK 3:30pm CREEP IN SIC-REINFORCED PIM 6061 ALUMINIUM COMPOSITE: S. L. I. Chan, J. L. Hwang, Institute of Materials Science and Engineering, National Taiwan University, Taipei, Taiwan, 100 China 1be mechanical behaviour of PIM 6061 aluminium matrix composite at elevated temperatures (3S0-450"C) has been studied. 1bese composites contained 0 to 20% of SiC reinforcements of different sizes.1be results show that for the monolithic 6061 Al alloy, the solutes in the alloy impinged the movement of dislocations, consequently the first stage of creep was absent or reduced. A threshold stress was found for the onset of creep, but this threshold stress diminished as the temperature increased. With the addition of SiC reinforcement, the resistance of the material to creep was improved, and the activation energy for creep and stress exponent increased correspondingly. However, no threshold stress was detected in the composites. In contrast to the monolithic specimens, all three stages of creep were found for the composites, but the extent of steady creep (Stage II) was reduced by the increasing amount of ceramic reinforcement. 3:50pm THERMOMECHANICAL FATIGUE OF A METAL MATRIX COMPOSITE: D. R. Lesuer, C. K. Syn, T.G. Nieh, L-342, Lawrence Livermore National Laboratory, Livermore, CA 94551 1be thermo mechanical fatigue (IMF) behavior of a discontinuously reinforced aluminum (6090/SiCl25p) has been studied over the temperature range of 25-450"C. TMF can occur in metal matrix composites because of the internal stresses produced during thermal cycling. 1bese stresses result from the thermal expansion mismatch between matrix and reinforcement. It is well established that during temperature cycling these internal stresses can produce local plastic flow near the interface of the matrix and reinforcement. Damage can also result that leads to failure. In this study we have evaluated the influence of cyclic mechanical loading and thermal loading on fatigue life. 1be work reported here was conducted with a unique experimental apparatus that permits independent control over both the thermal and the mechanical loading histories applied to the sample. This presentation will describe the results of these studies including the influence of cycling temperature and cycling thermal frequency. 1be origins offracture will be described and possible mechanisms discussed. This work was performed under the auspices of the U.S. Department of Energy by LLNL under contract No. W-7405-Eng-48. 4:10pm THE COMBINED EFFECT OF SOLUTIONIZING TIME AND VOLUME FRACTION OF PARTICLES ON THE AGING CURVES AND MICROSTRUCTURESIN A COMPOSITE WITH 6061 ALLOY MATRIX REINFORCED WITH AI103 PARTICLES: Javier Ponce, Daniel Salas, Michael Solis, Shane Andrews, S. K. Varma, Department of Metallurgical and Materials Engineering, 1be University of Texas at EI Paso, El Paso, TX 79968-0520 1be composites with 6061 aluminum alloy matrix and 0.10, 0.15 and 0.20 volume fractions of AI203 particles have been solutionized at 540'C for different times up to 20 hours to produce various grain sizes after quenching. 1be aging response of these solutionized composites have been examined with the help of microstructural evolution and aging curves. 1be relationship between the grain size developed during the solutionization process by grain growth and volume fraction of the particles on the aging curves has been investigated. 1be influence of CTE dislocations on the precipitation sequence and peak hardness values will be compared with the behavior of 6061 aluminum alloy in its monolithic form under identical experimental conditions. This research has been supported by the National Science Foundation through the grant number HRO-9353547. 4:30pm LASER BEAM PROCESSING OF PARTICULATE REINFORCED 6061 ALUMINIUM METAL MATRIX COMPOSITE: Li Hong, Laser Lab, Department of Materials, Institute Superior Teenico Av. Reviseo Pais, 1096 Lisbon, Portugal; Wang Youming, Department of Metal Forming, University of Science & Tecbnology Beijing, Beijing 100083, China SiClAl203 particulate reinforced 6061 Al metal matrix composites were laser beam cut and welded using a 3kw continuous wave COz laser. 1be influence of of laser processing parameters such as cutting/welding speed, laser power, and shielding gas rate on the quality of cuts and welds were investigated. Optical microscopy, scanning electron microscopy and X-ray diffraction were used to analyze the laser treated zone. Experimental results had shown that 6061 Al MMCs can successfully laser cut and weld. 1be AI4C3 plates were formed in HAZ due to the chemical reaction between Si and Al ocurring during laser processing. 4:50pm MICROSTRUCTURE AND PROPERTIES OF A SIC FIBER-REINFORCED "ORTIlORHOMBIC"TITANlUMALVMINIDE COMPOSITEPRODVCED BY PHYSICAL VAPOR DEPOSITION: S. Krishnamurthv, P. R. Subramanian, A. McCormick, UES, Inc., Dayton, OH 45432 Several methods are currently available for fabricating fiber reinforced titanium matrix composites which are candidate materials for structural aerospace applications at moderately high temperatures. 1be present work deals with the processing of a SiC fiber reinforced "orthorhombic" titanium aluminide composite using physical vapor deposition (PVD). 1be PVD process consisted of magnetron sputtering of a TI3AI-Nb alloy matrix on SCS-6 SiC fibers. 1be matrix-coated fibers were consolidated by hot pressing to form a unidirectionally reinforced composite containing about 35 vol.% fibers. Unreinforced matrix alloy samples were also fabricated using identical sputtering and

B80

consolidation conditions. The microstructural changes in these materials following different heat treatments was studied. Preliminary tensile tests were also conducted on matrix and composite specimens at room temperature. The microstructures of all the specimens were characterized by opticallnicroscopy, SEM, electron microprobe, and x-ray diffraction tecbniques. 1be results of microstructural charactelization and tensile testing of these PVD processed materials will be discussed. This research was performed under AF Contract F33615-94-C-S214.

PHYSICAL CHEMISTRY OF HIGH TEMPERATURE COMPOSITES AND INTERMETALLICS IV: Defects In Intermetallics

Sponsored by: EPD Process Fundamentills Committee, MSD Thermodynamics & Phase Equilibria Committee, Japan Institute of Metals Program Organizers: Prof. R.Y. Lin, University of Cincinnati; Prof. Y. Austin Chang, University of Wisconsin-Madison; Prof. R. Reddy, University of Navada-Reno and Dr. C.T. Liu, Oak Ridge NL Tuesday, PM February 6, 1996

Room:B2 Location: Anaheim Convention Center

Session Chairpersons: Y. A. Chang. Dept. of Material Science And Engineering,University of Wisconsin-Madison, Madison, WI 53706; Herbert Ipser, Institute ofinorganic Chemistry, Universilty of Vienna, Wahringerstr. 42, A-I090 Wien, Austria

2:30 pm Invited ON THE SITE REFERENCES OF TERNARY ADDITIONS TO TRIPLE -DEFECT B2 INTERMETALLIC COMPOUND: L. M. Pike, C. R. Kao, S. L. Chen, Y. A. Chang, Dept. of Material Science And Engineering, University of Wisconsin-Madison, Madison, WI, 53706 A knowledge of the site preference of ternary solute additions is essential to developing and understanding of how these sollutes affect the properties of B2 intermetallic compounds. A quasi-chemical model will be presented which is able to predict the site preferences of dilute solution additions to triple-defect B2 compounds. The only parameter required is the enthalpy of formation at the stoichiometric composition. General equations are developed which can be used to determine site occupancies and defect concentrations for dilute as well as non dilute solution additions. 1bese equation use atom pair bond enthalpies as the parameters. It is found that the site preferences of dilute additions are not always in agreement with predictions based on the solubility lobes in ternary Gibbs isotherm. Predictions for dilute additions to NiAl, CoAl and FeAl are compared to experimental resulls found in the literature. Satisfactory correlation is found between the predictions and the experimental results. In addition, the predictions from the model on site occupations and defect concentrations in the (Ni, Fe )AI system are compared to recent experimental results by the authors. 2:55 pm Invited A MODEL FOR INTERMETALLlClCERAMIC COMPOSITE MICROSTRUCTURES FROM SOLID STATE DISPLACEMENT REACTIONS: C. H. Henager, Ir., Pacific Northwest Laboratory, Richland, WA 99352; L. N. Brush University of Washington, Dept. of Materials Science and Engr., Seattle, WA 98195 1be formation of intermetallic/ceramic composite microstructures from solid state displacement reactions is reviewed and II diffusion/nucleation model for the development and growth of these microstructures is discussed. While a wide variety of complex microstructures are produced from displacement reactions, the most likely microstructure to form is a particulate-type miicrostructure where one of the phases is continuous and the second phase is discontinuous and dispersed homogeneously within the first phase. 1be formation of these microstructures can be rationalized by a model that allows interdiffusion and nucleation at moving boundaries to simulate solid state displacement reactions in diffusion couples. 1be implications of the results for tailoring complex microstructures for various applications, such as ceramic joining will also be addressed. 1be discussion and model will focus on research in the Mo-Si-C and TI-SiC ternary systems. PNL is Operated for the U.S. DOE by Battelle Memorial Institute under Contract DE-AC06-76RLO 1830. 3:20 pm Invited PROCESSINGIMICROSTRUCTURE CONTROL FOR HIGH STRENGTH, DUCTILITY, AND TOUGHNESS AT ROOM TEMPERATURE IN PIM FeAI: P.I. Maziasz, V. Sitka, D. J. Alexander, A.Goyal, J.L. Wright. Metals and Ceramics Division, Oak Ridge National Laboratory, P.O. Box 2008 Oak Ridge, TN 37831-6114 A B2-phase FeAl (Fe-38 AI, at%) alloy with minor alloying additions of Mo, Zr, and C was produced as N2 atomized powder, and consolidated by extrusion at 950, 1000 and 1100'C. Both the properties and the processing-induced microstructures were quite sensitive throogh the extrusion conditions. PIM FeAl extruded at llOO"C had an equiaxed grain structure (6-30 mgrain size): which produced total elongation of 9.2%, YS of 500 Mpa, UTS of 940 Mpa and Charpy impact energy of 20 J for test in air at room temperature. Extrusion at 950·C significantly refined the grain size (1-71lID) and improved the ductility to 13 %, YS to 670 Mpa, UTS to 1080 Mpa and the impact toughness to 100 J. More in depth analysis of microstructure and texture will be presented, and effect of microstructural Changes on property improvement in FeAI will be discussed. 3:45 pm Invited COMPUTATIONAL THERMODYNAMICS IN MATERIALS SCIENCE: Hans lurgen Seifert, Hans Leo Lukas, Gunter Petzow, Max-Planck-Institut f. Meta:Iforschung, Inst. f. Werkstoffwissenschaft, Heisenbergstr. 5, 0-70569 Stuttgart, Germany

1996 TMS ANNUAL MEETING AND EXHIBITION

The CALPHAD method (CALculation of PHAse Diagrams) has been extensively developed in the last 25 years. Thermodynamic descriptions for the functions of state of unary, binary and ternary phases are derived by the optimization of phase diagrams. These functions are stored in computerized thermochemical databases and can be combined to extrapolate to multicomponent systems. Depending on special demand, tbe actually most advantageous presentations of the calculated results can be chosen: tables or in a clear graphical form as phase diagrams, phase amount diagrams or volatility diagrams. The method will be described by the system TI-Si-C-N-O including the ceramic subsystem Si3N4-SiC-TICxN I_x-C-N-O which is most important for the development of ceramic composites. The Wagner-Schottky model expressed by the compound energy formalism proves to be an efficient description for solution phases of this system. The calculations give the thermodynamic stabilities of the condensed phases as function of temperature, concentration and partial pressure of different gas species. Adjusting the sintering conditions to these results the formation of unfavorable silicides or liquid phases can be suppressed. 4:10 pm BREAK 4:20 pm Invited STATISTICAL THERMODYNAMICS A TOOL FOR UNDERSTANDING POINT DEFECTS IN INTERMETALLIC COMPOUNDS: Herbert Ipser. Regina Krachier, Institute of Inorganic Chemistry, University of Vienna, Wahringerstr, 42, A1090 Wien, Austria Statistical-thermodynamic methods are used to interpret the composition dependence of partial thermodynamic properties in non-stoichiometric intermetallic compounds. With these it is possible to derive the defect mechanism responsible for non-stoichiometry, but also to obtain concentrations of individual point defects. Two examples are given: I) For p' -FeAI with the cubic eso-structure different experimental values for defect concentrations, determined by different independent methods, are compared with those obtained from a statistical treatment of thermodynamic data. 2) For Rh3Te4 with the hexagonal NiAs-structure defect mechanism and vacancy concentrations are derived from tellurium vapor measurments. 4:45 pm Invited FORMATION AND STABILITY OF AI3TI IN THE HYPER-PERITECTIC AITI ALLOYS AFTER MECHANICAL ALLOYING. Dong-Wha Kum, Hye-Sung Kim, Gyeungho Kim and Dong-Su Shur, Korean Institute of Science and Tech., 136791 P.O. Box 131, Cheongryang, Seoul, Korea Mechanical alloying is an effective way to finely distribute inert dispersoids in AITM(TM is a transition metal element) systems. It is considered that high melting point alurninides are formed by precipitation from supersaturated aluminum matrix. The analysis is based on the fact tbat much higher content of TM than the solubility limit can be dissolved in alpha aluminum during the bigh energy ball milling. Nonetheless, the role of undissolved TM particles from mechanically alloyed AI-TI alloys in the hyperperitectic region. This study showed tbat LI2-type Al3TI (a metastable form) formed first during annealing at elevated temperatures. The results indicated undissolved TI particles of nanosize should have played an important role initiating the formation of AI3TI phase. Thermal stability of the tri-aluminide phase is also discussed. 5:10pm DIFFUSION GROWTH OF TITANIUM CARBIDE FILM ON NI-TI BINARY SUBSTRATES: S. Lesokhin, L. Levin, Dept. of Mat. Eng., Israel Institute of Techno1ogy -Technion, Haifa 32000, Israel The formation of titanium carbide during hot filament assisted carburizing of Ni-TI alloys was investigated. Phase content and morphology of the diffusion zone were examined by x-ray diffraction, x-ray microanalysis, SEM and Auger electron spectroscopy. The growth of stable and adherent TIC film was observed. A non-planarity of the interphase boundaries was noticed. The applicability of the chemical diffusion approach to the analysis of the experimentally observed interface instabilities is discussed. A new criterion, taking into account the influence of growth stresses on the stability of the flat interface, is developed by means of linear perturbation analysis. The predictions of the new criterion are compared with the experimental observations. 5:35pm THERMODYNAMICS OF THE TERNARY SYSTEM Fe-NI-Cr: iosefTomiska, Jan Vrestal, Institute of Physical Chemistry, University of Vienna, Waehinger-Strasse 42, A-I09O Vienna, Austria Thermodynamic investigations on liquid as well as on fcc solid ternary Fe-Ni-Cr alloys have been performed by means of the computer-aided Knudsen cell mass spectrometry. The "Digital Intensity-Ratio" -method in combination with an overall best fit technique has been applied for the thermodynamic evaluation. Ternary T.A.P.-series (Thermodynamically adapted power series) are used for the algebraic representation of the thermodynamic excess properties. Based upon these data, and on the results of mass spectrometric investigations on the binary boundary systems, the phase diagram of the system Fe-Ni-Cr has been recalculated.

PROCESS MINERALOGY I: Pyrometallurgy, Flotation and Environment Sponsored by: EPD Process Mineralogy Committee Program Organizer: Dr. W. Petruk, CANMET, Natural Resources Canada, 555 Booth St., Ottawa, Ontario, Canada KIA 001 Tuesday, PM February 6, 1996

Room: AI4-15 Location: Anaheim Convention Center

1996 TMS ANNUAL MEETING AND EXHIBITION

Session Chairperson: Richard D. Hagni, Department Geology and Geophysics, 125 McNutt Hall, Univ. Missouri-Rolla, Rolla, MO 65401

2:00pm PROCESS MINERALOGY OF REACTION PRODUCTS BETWEEN NICKEL CONCENTRATE AND Fe-Mg SILCATE SLAGS: Richard D. Hagni, Department Geology and Geophysics, 125 McNutt Hall, University of Missouri-Rolla, Rolla, MO 65401; Maria K. Kojo, H.Y. Sohn, Department Metallurgy and Metallurgical Engineering, 412 Wm. C. Browning Bldg, University of Utah, Salt Lake City, UT 84112 Pelletized nickel sulfide concentrates submerged into Mg-Fe silicate slags were studied by ore microscopy and electron microprobe to characterize the mineralogy of the reactions. The original concentrate, which consisted of pyrrhotite, pentiandite, pyrite, chalcopyrite, and violarite, formed a matte with original sulfide particle outlines rounded to obliterated, and recrystallized into two phases: pentlandite, Feo.58Nio.51 CII().03S and pyrrhotite, Feo.75Nio.I 8CII().0IS solid solutions. Those portions of the sulfide pellets that sloughed off into the slag were melted to form liquid spheres of pentlandite and pyhrrhotite. Reaction with the slag resulted in the formation of magnetite in the pyrrhotite portions of the sulfide spheres. The slag consisted of ferrohypersthene, Mg-AI-Fe spinel, and glass. 2:30pm MINERALOGICAL CHARACTERIZATION OF ZINC RETORT RESIDUES OF THE MISSOURI-OKLAHOMA-KANSAS TRI-STATE REGION: Dhiren K. Panda, David G.C. Robertson, Generic Mineral Technology Center for Pyrometallurgy, 215 Fulton Hall, University of Missouri-Rolla, Rolla, MO 65401 The zinc retort residues from old pyrometallurgical sites (in the Oklahoma-KansasMissouri tri-state area) pose an environmental problem, since they contain a significant amount of lead and zinc, among other contaminants. The residue is heing treated by the SCIBS (Submerged Combustion in-hath S.melting) process at the University of Missouri-Rolla. Results of the mineralogical characterization of the residue and the resulting slag are presented in this paper. 3:00pm SLURRY PUMP PARTS DRILLED AND TAPPED WITH HARDNESS IN EX· CESS OF 700HBN: George A. Calbomanu, A.R. Wilfley & Sons, Denver, CO In order to overcome drilling and lapping of high hardness wear resistent material in slurry pumps, most mild steel inserts are mounted in the mold and after pouring become an integrated part of the casting. This process limits the design possibilities of slurry pumps and efforts are made to drill and tap the hard iron itself. The paper will present practical results obtained in drilling and tapping a wear resistent material with hardness in excess of 700HBN. The life time of the slurry part pumps in service increased 2.5 times. 3:30 pm BREAK 3:50pm MINERALOGICAL CHARACTERIZATION AND SUPERFICIAL PRETREAT· MENT OF XENOTIME AND ZIRCON FROM THE PITINGA MINE (AMAZON, BRAZIL): M.L. Torem, LB. Scorzelli, L.C. Bertolino, Department of Material Science and Metallurgy, Catholic University of Rio de Janeiro - Rua Marques de Slio Vicente 225, 22453 900 Rio de Janeiro, Brazil Significant amounts of a middling xenotime - YP04 (90%) and zircon ZrSi04 (10%) are produced by the flow sheet of Taboca Mining, one of the largest world tin concentrate producers. Due to the lack of information about the complexity of the xenotime and zircon no successful separation process of the middling has been put forward until now. Detailed characterization was performed on the xenotime and zircon, using optical microscopy, SEMIEDS and XRD techniques. In addition, superficial pretreatment with HCL, HN03, HF and NaOH was done to evaluate modifications on surface composition and the morphology of the particles for flotation purposes. The studies showed that both minerals contain very small inclusions (1 to 5 J.lm) of thorite and iron oxides, and the pretreatment showed alterations of the minerals on the particle surfaces. 4:20pm CHARACTERIZATION OF PARTICLES IN AIR-BORNE DUSTS: William Perruk, CANMET, 555 Booth St., Ottawa, Ontario, Canada KIAOGI; H.C.W. Skinner, Dept of Geology and Geophysics, Yale University, New Haven, CT 06520-8109 Airborne dusts contain particles that are up to 35 J.lR1 in size and have aspect ratios that range from Ito >10. The mean size and mean aspect ratio is about 3.5 J.lmand 2.7 respectively. The particles are composed of a wide variety of minerals including tremolite, talc, serpentine, hypersthene, kaolinite, dolomite, calcite, quartz, etc. A method of quickly identifying the minerals and determining quantities, grain sizes, and aspect ratios was developed by using an Kontron image analyzer interfaced with a Jeol microprobe and Tracor Northern EDS. The mineral in each particle is identified automatically by scanning with an X-ray beam and correlating the X-ray signal to signals from mineral standards. Grain sizes and aspect ratios are measured with the image analyzer. 4:50pm MINERALOGY APPLICATION TO METALLURGICAL WASTES DISPOSAL AS BINDING MATERIALS: Dr. Professor Y.A. Mymrin, Scientific Centre for Engineering Geology and the Environment, Russian Academy of Sciences, P.O. Box 145, Ulanski per. 13, Moscow 101000, Russia This paper is devoted to the role of all types of ferrous minerals (iron, steel and casting) and some kinds of nonferrous dump slags in combination with liquid or sludge wastes as binding materials. These materials are used for binding other industrial wastes (incinerator's and heat power station ash and slag, burnt coal mining refuses and burnt foundry sands, etc). These New materials work like poor concrete and, in comparison with natural traditional inert materials (sand, gravel, broken rock, etc), have numerous

B81

advantages in strength, elasticity, construction rate, long service, cost-price, environment protection, etc. All the hazardous elements of wastes are chemically bonded in nonsoluble compositions and were tested in acid, alkaline and neutral solutions. Ten roads and one airfield were constructed in rigorous climate conditions, including the Russian North and Siberia and during 17 years of exploitation they show high performance.

RADIATION MATERIALS SCIENCE IN TECHNOLOGY APPLICATIONS II: Radiation Effects In Ceramics

Sponsored by: Jt. SMDIMSD Nuclear Materials Committee Program Organizers: L. K. Mansur, Metals and Ceramics Division, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831-6376; C. L. Snead, Jr., Applied Technologies Division, Brookhaven National Laboratory, P.O. Box 5000, Upton, NY 11973-5000 liIesday,PM February 6, 1996

Room: Grand H Location: Anaheim Marriott Hotel

Session Chairperson: Y. Chen, Department of Energy, Division of Materials Sciences, ER-I3I, 19901 Germantown Road, Germantown, MD 20874-1290

2:00 pm invited

FLUENCE AND TEMPERATURE DEPENDENCE OF ION-BEAM-INDUCED AMORPHIZATION IN a-8IC: W. J. Weber, Pacific Northwest Laboratory, P.O. Box 999, Richland, WA 99352; N. Yu, Los Alamos National Laboratory, P.O. Box 1663, MS K-762, Los Alamos, NM 87545; L. M. Wang, Department of Earth and Planetary Sciences, University of New Mexico, Albuquerque, NM 87131 Single crystals of a-SiC have been irradiated at 173,300, and 373 K with 360 keV Ar+ ions using the facilities within the Ion Beam Materials Laboratory at Los Alamos National Laboratory. The fluence dependence of the amorphization process was investigated as a function of temperature using in situ Rutherford backscattering spectroscopy in channeling geometry. The disordered fraction increased sigmoidally with ion fluence, consistent with a defect accumulation process. The rate of disordering decreased with increasing temperature. Post-irradiation HRTEM and Raman spectroscopy conftrmed the amorphous nature of the damage state. Nanoindentation indicated decreases in hardness and elastic modulus with increasing disorder. 2:30pm EVIDENCE FOR IONIZATION ENHANCED DIFFUSION IN ION-IRRADIATED CERAMIC INSULATORS: S. J. Zinkle, Metals and Ceramics Division, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831-6376 Cross section transmission electron microscopy was used to investigate radiationinduced microstructural changes in polycrystalline specimens of alumina (A1203), magnesium aluminate spinel (MgAI204) and silicon nitride (Si3N4) following irradiation at room temperature and 650·C with ions of varying mass and energy (I MeV H+ to 4 MeV Zr). The density of interstitial dislocation loops decreased and the mean loop size increased with decreasing mass of the bombarding ion. In addition, cavity formation could be induced in some specimens at irradiation temperatures as low as 20·C, which corresponds to a homologous irradiation-temperature of only 0.12 Tm. Irradiation of silicon nitride with 3.6 MeV Fe ions at 20·C produced amorphization in the Feimplanted regions, but amorphization did not occur in specimens that were simultaneously irradiated with I MeV He ions. Possible physical mechanisms responsible for these effects will be diSCUssed, with particular emphasis on ionization enhanced diffusion. Research sponsored by the Office of Fusion Energy, U.S. Department of Energy, under contract DE-AC05-840R21400 with Lockheed Martin Energy Systems. 2:50pm IN SITU BEAM CHARACTERIZATION OF RADIATION DAMAGE IN YTTRIUM STABIUZED ZIRCONIA INDUCED BY 400 keV XENON IONS: Nina Yu, K. E. Sickafus, Michael Nastasi, Materials Science and Technology Division, Los Alamos National Laboratory, Los Alamos, NM 87545 Single crystals of cubic yttrium stabilized zirconia (YSZ) were irradiated with 400 ke V Xe ion beams at cryogenic temperatures and room temperature. Radiation damage accumulation in YSZ crystals as a function of irradiation dose was in situ monitored with Rutherford backscattering spectrometry in conjunction with ion channeling techniques using a 2 MeV He ion beam. It was observed that the amount of lattice disorder in the peak damage region increased with the increasing dose and finally saturated at about 70% of the random level up to a dose of 3x 1016Xe/crrIJ.· Cross-section transmission electron microscopy was used to determine the microstructures of irradiated YSZ crystals. 3:10pm THE MECHANICAL AND STRUCTURAL RESPONSE OF SPINEL FOLLOWING 12 MeV Au ION IRRADIATION: Ram D~anathan, Ning Yu, K. E. Sickafus, Michael Nastasi, Materials Science and Technology Division, Los Alamos National Laboratory, Los Alamos, NM 87545 Magnesium aluminate spinel (MgAl204) is a prOmising candidate for insulating and structural applications in fusion-reactors because of its radiation resistance. We have studied the structure-property relationships in MgAl204 subjected to high-energy ion irradiation. Single-crystals of spinel were irradiated with 12 MeV Au3+ at 100 K. The projected range of the implanted Au ions, determined by TRIM, was 1.5 11m. The microstructure of the irradiated samples was studied using cross-sectional transmission electron microscopy, while the mechanical properties were determined using the nanaindentation technique. For each dose, the Young's modulus and nana-hardness were

B82

determined for three different indentation depths. The changes in these mechanical properties with dose will be discussed in light of the observed structural changes. These results will be compared with the findings of our earlier study of spinel subjected to 400 ke V Xe ion irradiation. 3:30 pm BREAK 3:50pm RADIATION DAMAGE IN ILMENIT E-GROUP MINERALS: Jeremy N. Mitchell, Ning Yu, K. E. Sickafus, Michael Nastasi, Thomas N. Taylor, Kenneth J. McClellan, Materials Science and Technology Divi.sion, Los Alamos National Laboratory, Los Alamos, NM 87545; Gordon L. Nord, USGS, Reston, VA 22092 The radiation damage response of the rhOmbohedral oxides ilmenite (FeTI03) and geildelite (MgTI03) were studied using Ithe in situ and ex situ ion irradiation facilities at Los Alamos National Laboratory. The MgTI03 crystals studied were grown synthetically, whereas the FeTI03 sample is II natural crystal collected from the Adirondack Mts., NY. In our experiments, single crystals of these materi~were implanted with Ar and Xe at varying doses and the surface composition and structure were monitored using ion channeling combined with Rutherford backscattering spectrometry (RBS). RBS and ion channeling measurements of these crystals implanted at varying doses at -IOO·C indicate that FeTI03 becomes fully random at -lxlQ15 Ar cm-2 and MgTI03 approaches a fully random level at a dose of 2xloI5. Analysis of the FeTI03 crystal using XPS suggests that the implantation process may have oxidized the target. Thus, the poor radiation resistance of FeTI03 may be due to chemical and crystallographic changes caused by implantation. In this presentation we will also explore the effects of self ion irradiations of FeTI03. The behavior of MgTI03 holds promise for the use of this material and other titanates in nucle:n: reactor applications. 4:10pm REDOX REACTIONS IN TRANSITION METAL CATION-BEARING OXIDES AND SILICATES: KEY TO UNDERSTANDING THERMODYNAMICS OF METAL COLLOID FORMATION? R. F. Cooper, N. Yu, J. N. Mitchell, K. E. Sickafus, Materials Science and Technology Division, Los Alamos National Laboratory, Los Alamos, NM 87545 Self-cation implantation into oxides has been demonstrated to result in the internal formation of metal colloids: e.g., Mg+ or AI+ implantation into MgAl204 results in the formation of colloidal AI+ in the region at which the implant species comes to rest [e.g., Evans et aI., Mat. Res. Soc. Symp. Proc . 373, (1995)]. The result clearly suggests that the relative oxidation potentials of cation components is critical in such reactions. We are addressing this question directly through self-ion implantation experiments in Fe2+doped spinel ([MgO.999FeO.001]AI204) and in natural olivine ( [MgO 88' FeO I212Si04). Redox potentials make spedfic predictions about the Chemical/structural reactions that should occur. Implantation of oxygen into olivine produces backscattering spectra that is directly correlated wiUI that of dynamic chemical oxidation. Implantation of "inert" species may produce a similar reaction. Mg+ implantation in either solid results in metal colloid formation. We are presently evaluating the nature of these colloids by analytical electron microscopy. 4:30pm THERMAL CONDUCTIVITY DEGRADATION OF GRAPHITES IRRADIATED AT INTERMEDIATE TEMPERATURE: L L Snead, T. D. Burchell, Metals and Ceramics Division, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831-6376 Several graphites and graphite composites (ClC's) have been irradiated in the temperature range of 200 to 800·C to determine the absolute reduction in thermal conductivity for a dose range between 0.01 te> 1.0 dpa. The unirradiated room temperature thermal conductivity of these materials varied from 114 W/m-K for H-45 I isotropic graphite, to 670 W/m-K for unidirectional FMI-ID ClC composite. As expected, the amount of thermal conductivity degradation is a strong function of irradiation temperature with a substantially higher reduction occurring for the lower temperature irradiations. As an example, the saturated resi.dual thermal conductivity for those materials irradiated in the lower temperature region was approximately 20% of the original thermal conductivity measured at the irradiation temperature while approximately 50% of the thermal conductivity was retained for the elevated temperature irradiations. Data from these measurements will be compared to a design algorithm which gives the thermal conductivity of high conductivity ~raphite as a function of irradiation temperature and displacement level. Research spom;ored by the Office of Fusion Energy, U.S. Department of Energy, under contract DE-AC05-840R21400 with Martin Marietta Energy Systems.

STRUCTURE AND MORPHOLOGY OF EPITAXIAL THIN FILMS SESSION IV: Defects I Sponsored by: EMPMD Thin FUms & Interfaces Committee Program Organizer: Dr. David E. Jesson, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831-6030 liIesday,PM February 6, 1996

Room: Orange County 4 Location: Anaheim Marriott Hotel

Session Chairperson: R. C. Pond, Department of Materials Science and Engineering, University of Liverpool, Liverpool, L69 3BX

2:00 pm Invited

MISFIT DISLOCATION FORMATION INDUCED SURFACE MODIFICA-

1996 TMS ANNUAL MEETING AND EXHIBITION

TIONS STUDIED BY UHV-SCANNING TUNNELING MICROSCOPY: G. Springholz, N. Frank, Y. Ueta, V. Holy, G. Bauer. Institut fur Halbleiterphysik, Universitaet Linz, A-4040 Austria Surface modifications caused by misfit dislocation formation in MBE-grown EuTe on (Ill) PbTe buffer layers (lattice mismatch: 2.1%) were studied using ultra-high-vacuum scanning tunneling micrsocopy. In agreement with the Matthews Blakeslee mechanism for strain relaxation, at the critical layer thickness straight monolayer step lines appear due to the glide of preexisting threading dislocations, grown-in from the PbTe buffer. Apart from these steplike featores at the surface local wave-like deformations with an amplitude equal to the normal component of the Burgers vector are induced above the misfit dislocations at the epilayerlsubstrate interface, caused by the local lattice deformations around the misfit dislocation segments. The exact form of these surface deformations depends on the orientation and magnitude of the Burgers vector and fortbermore dislocation reactions have their direct signature in the accompanying surface deformations. 2:30 pm Invited A THERMALLY-INDUCED STRESS-ASSISTED COOPERATIVE DISLOCATION GENERATION MECHANISM: APPLICATION TO STRAINED EPITAXIAL FILMS: M. Khantha, Department of Materials Science and Engineering, University of Pennsylvania, Philadelphia, PA 19104-6272 There are many phenomena where dislocation generation occurs suddenly over laIge length scales. Some well-known examples include the massive dislocation activity which occurs near a crack tip at the briUle-to-ductile transition, the yielding of whiskers and the generation of misfit dislocation network in epitaxial films. The mechanism by which dislocations are generated when a suitable control parameter such as temperatore, stress, or film thickness is varied is not well understood. We have recently proposed a new thermally-induced stress-assisted cooperative dislocation generation mechanism, hased on the Kosterlitz-Thouless concept of dislocation screening, and used it to explain the general characteristics of the brittle-ductile transition. We discuss how this mechanism can he adapted to explain the generation of misfit dislocations in epitaxial films above a critical film thickness for a given temperature. 3:00 pm InvIted DISLOCATION MECHANICS IN ATOMISTIC DETAIL: THE CASE OF SILICON: Vasily V. Bulalov, MIT, Dept. of Mechanical Engineering, 77 Massachusetts Ave., Cambridge, MA 02 139 Results are presented of extensive computer simulations of full and partial dislocations of the III-guide subsystem in silicon. New principal findings include a prominent role in dislocation motion of the (2x I) core reconstruction and the reconstruction defects, and a strong left-right asymmetry of dislocation kink propagation (in 30-partial dislocation). Computer simulation techniques are developed to examine atomistic structure effects on short-range interactions between two dislocations. The results show that certain pre-assumptions common in the classical (continuum) dislocation mechanics have to be re-examined and that details of the dislocation core structure have implications for dislocation mobility and interactions, including processes of cross-slip and cutting of the forest dislocations. Such core aspects can be explicitly addressed only within a fully atomistic approach. This work clearly shows both necessity and feasibility of atomistic simulations for the development of a quantitative theory of dislocations in silicon. 3:30 pm BREAK DEFECTS II Session Chairperson: C. B. Carter, Department of Chemical Engineering and Materials Science, Amundson Hall, University of Minnesota, Minneapolis, MN 55455 3:50 pm InvIted INTERFACIAL DEFECT ARRAYS AND MISORIENTED EPITAXIAL LAYERS: M. Aindow, School of Metallurgy and Materials, The University of Birmingham, Elms Road, Edgbaston, Birmingham, B 15 2TT; R. Beanland, 2GMMT Caswell, Towchester, Northants, NN 12 8EQ; R. C. Pond, Department of Materials Science and Engineering, University of Liverpool, Liverpool, L69 3BX The evolution of the microstructure and surface morphology of epitaxial deposits is strongly influenced by the degree of misfit and the condition of the substrate surface. In particular, epitaxial films may be misoriented by up to a few degrees away from a state where low-index directions in the substrate. Such misorientationed layers are referred to as being tilted, and this paper is a review of experimental observations reported over the last twenty years or so. It will be shown that the tilting layers can be explained comprehensively by topological considerations of defects at the film/substrate interface. Such defects arise at interfaces either to accommodate mismatch at singular or vicinal substrate surfaces, or as a consequence of deformation processes initiated in epitaxial layers to relieve long-range strain. When the total interfacial dislocation content is assessed, a component causing tilting, and not contributing to misfit relief may accrue. The topological theory of interfacial defects and arrays of defects will be reviewed briefly and subsequently used to understand the occurrence of tilted layers. 4:20 pm Invited INDENTATION INDUCED DISLOCATION NUCLEATION AT EPITAXIALINTERFACES: W. W. Gerberich, C. J. Palmstrl'lm, Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, MN 55455; P. O. Cohen, Department of Electrical Engineering, University of Minnesota, Minneapolis, MN 55455; P. M. Anderson, Materials Science and Engineering, Ohio State University, Columbus, OH 43210 Recently, Page at Newcastle, Michalske at Sandia National Laboratories, and ourselves have detected the onset of plasticity in bulk ceramic and metallic systems using nanoindentation. Our measurements have principally dealt with Fe-3wt%Si single crystals and in those, it bas been shown that dislocation loops initiate at tip forces in the range of 30 to 300 N. The measurements are sufficiently accurate to isolate the exact tip forces at which the first dislocation loop nucleates. We have analyzed the conditions

1996 TMS ANNUAL MEETING AND EXHIBITION

and found that tip forces, as balanced by image forces and the Peierl's barrier are greatly affected by tip radius and oxide film thickness. Epitaxial films become doubly interesting because of the two interfaces at which the dislocation may nucleate and the epitaxial stress as a potential driving force for nucleation. A series of interrnetallic and rare-eartb group V semi-metallic films have been grown on InGaAs and GaAs with various misfit strains. The films have B2 and rock salt crystal structures, respectively. First order theoretical analysis of the various contributions to film stability have been made, and it is demonstrated, experimentally, that the stored elastic energy in the thin film lowers the onset of dislocation nucleation as expected. 4:50pm EPITAXIAL GROWTHS OF HIGHLY CONDUCTIVE Ru02 TIUN FILMS ON YSZ AND LaAI03: Shihong G. Song, Quan Xi Jia, Xin Di Wu, Los Alamos National Laboratory, Los Alamos, NM 87545 Epitaxial growths of highly conductive Ru02 thin films on YSZ and LaAlOJ substrates have been observed using cross-sectional HRTEM techniques. The matrixlfilm interface in either case is seen to be atomically sharp. The epitaxial lattice relationships were determined using X-ray and electron diffractions. Although both films exhibit diagonal epitaxy, behavior details of the lattice relationship are distinctively different. Crystallographic variants are seen in the thin films as a result of the diagonal lattice relationShip. Small lattice misfit is present in both cases, and a dislocation mechanism is provided to interpret the interface structures. 5:10pm CHARACTERIZATION OF DEFECTS IN GaPlSl HETEROEPITAXY: Srikanth B. Samavedam, E. P. Kvam, School of Materials Engineering, Purdue University; T. P. Chin, J. M. Woodall, School of Electrical Engineering, Purdue University, W. Lafayette, IN 47907 Integration of III-V based devices with silicon-based technology may be possible through the growth of good quality gallium phosphide layers on silicon (GaP/Si). GaPI Si films were grown using molecular beam epitaxy (MBE) under various processing conditions and substrate orientations, and were characterized primarily using transmission electron microscopy ClEM). The observations seemed to indicate that islanding during initial periods of GaP film growth might be responsible for the high density of stacking defects observed. The problem was magnified by low GaP fault energy and surface energy anisotropy. There were strong indications that anti-phase domain boundaries (APBs) could be avoided by resorting to miscut Si(IOO) substrates for growth.

SYNTHESIS AND PROCESSING OF NANOCRYSTALLINE POWDER III: MIlling Techniques

Sponsored by: MDMD Powder Metallurgy Committee and FEMS (Federation of European Materials Societies) Program Organizers: Dr. David L. Bourell, The University of Texas at Austin, Materials Science & Engineering, MC C2201, Austin TX 78712. Dr. Liisa Kuhn-Spearing, Laboratory for the Study of Skeletal Disorders and Rehabilitation, Harvard Medical School, Children's Hospital, 300 Longwood Avenue, Boston MA02115. Professor Dr. Herbert Gleiter, Karlsruhe Research Center, P.O. Box 3640, D-76021 Karlsruhe, Federal Republic of Germany Tuesday, PM February 6, 1996

Room: Grand G Location: Anaheim Marriott Hotel

Session Chairperson: Robert D. Shull, Magnetic Materials Group, Materials Science and Engineering Laboratory, National Institute of Standards and Technology, Gaithersburg MD 20899

2:00pm SYNTHESIS. COMPACTION AND PROPERTIES OF NANOCRYSTALLINE ATTRITED POWDER: C. Schmidt, C. Moelle, K. MUller, H.1. Fecht, Technical University Berlin, 10623 Berlin Germany Mechanical attrition has been developed as a versatile a1temative to other processing routes in preparing nanocrystalline structures. In this process, the intemal refining process with a reduction of the average grain size by a factor of about 104 results from the creation and self-organization of dislocation to high-angle grain boundaries within the powder particles during the high energy milling process. These powder particles have been compacted to solid wire shaped samples with a density of about 93% by a hydrostatic extrusion process. Recent results for aluminum alloys are discussed with a special emphasis on the achieved mechanical properties. 2:25pm SYNTHESIS OF NANOSTRUCTURED TITANIUM BASED MATERIALS BY PLASMA PROCESSING OR MECHANICAL ALLOYING AND HOT ISOSTATIC PRESSING: F.H. Froes, Institute for Materials and Advanced Processes. P.R. Taylor, Metallurgical and Mining; C. Suryanarayana, Institute for Materials and Advanced Processes, University of Idaho, Moscow, ID 83844-3026; J. Hebeisen, Industrial Materials Technology, 155 River St., Andover MA 01810 Both plasma proceSSing (PP) and Mechanical Alloying (MA) allow the product of "large" amounts of equiaxed nanostructured materials; in contrast to many of the "Iaboratury curiosity" processes being evaluated. This paper will discuss both the PP and MA techniques as applied to the titanium system. It will also discuss the critical step of compaction to a usable article using hot isostatic pressing. 2:50pm SYNTHESIS AND CHARACTERIZATION OF BULK NANOCRYSTALLINE M50 STEEL: Maggy L. Lau, Benlih Huang, Robert J. Perez, Enrique J. Lavemia,

B83

Department of Chemical Engineering and Materials Science, University of Califomia, Irvine CA 92717-2575 M50 steel (4.0 wt% Cr, 4.5 wt% Mo, 1.0 wt% V, 0.8 wt% C, balance of Fe) is a technologically important alloy due to its good resistance to tempering, wear, and roIling contact fatigue, and bas been used extensively in the bearings in gas-turbine engines. TIle present study bas been conducted using both mecbanical alloying of elemental powders as well as spray atomization of pre-alloyed MSO with subsequent ball milling to synthesize MSO steel powders. In order to improve the thermal stability of the nanocrystalline MSO, incorporation of AI and subsequent cryomilling in liquid nitrogen resulted in thermally stable bulk nanocrystalline M50 steel. Characterization of the bulk nanocrystalline MSO is being conducted to study the mecbanical and damping properties. TIle authors would like to acknowledge the financial support provided by the Office of Naval Research under grants NOOOI4-93-1072 and NOOOI4-94-0017. 3:15pm PRODUCTION OF TITANIUM NITRIDE AND CARBONITRIDE POWDERS BY REACTIVE MILLING IN DIFFERENT ATMOSPHERES: S. QuintanaMolina, J.G. Cabanas-Moreno, H. Balmori-Ramfrez, R. Martinez-S4nchez, Instituto Politechnico Nacional, ESIQIE, Apdo, Postal 75-373, 07300 M~xico, D.P., M~xico TItanium powders, with or without graphite additions, have been milled under atmospheres of NH3, N2 or laboratory air. In all cases, the titanium powders have become reactive during milling, with the main products of the reaction being TIN, TI(N ,0) and TI(N,C). These products have been characterized by XRD, SEM, 1EM and chemical analysis. Milling of mixtures of TI and graphite powders have yielded TI(N,C) compounds with lattice parameters between those of pure TIN and TIC. In some occasions, the as-milled powders have ignited simultaneously after removal from the mill. Milling in air bas mainly produced a TI(N,O) phase, while the pure N2 atmosphere has yielded a product which is closer to pure TIN. TIle crystallite size of all the products is within the nanometer range, with some exceptions in those mixtures which ignited. TIle results of additional sintering experiments will also be described. This work was supported by DEPI-PIFI and COFAA-IPN and CONACYT. 3:40 pm BREAK 3:50pm OXIDE-METAL NANOCOMPOSITES BY BALL MILLING: Laszlo Takacs, TIle University of Maryland Baltimore County, Department of Physics, Baltimore MD 21228 High energy ball milling is a promising low-tech method to produce large quantities of oxide-metal nanocomposites. However, the price for the apparent simplicity of the process is difficulty of control. TIle possibilities to monitor the process are limited, prior information on the expected kinetics is crucial. Some nanocomposites can be prepared by milling the final components until the desired particle size and grain structure is reached. Excessively long milling times may result in contamination and deteriorating properties. Additional control may be possible via in situ chemical reactions between an oxide and a more reactive metal. TIle process usually occurs in three overlapping steps: An activation period of attrition, mixing, and defect formation; a relatively fast reaction when an intermediate sub-oxide is produced; and the decompoSition of this oxide into a fine dispersion of the final oxide and metal pbases. Several examples including the preparation of iron particles on A1203, MgO, and ZnO will be discussed. 4:15pm PROCESSING OF NANOCRYSTALLINE MATERIALS BY SEVERE PLASTIC DEFORMATION CONSOLIDATION: Ruslan Z Va/iev, Institute of Physics of Materials, Ufa State Aviation Technical University, Ufa 450000 Russia Recent investigations show !bat severe plastic deformation, i.e. intense plastic straining under high hydrostatic pressure, is not only a method to refine microstructures effectively, but also is an innovative method for consolidation of powders into bulk samples. TIle present paper focuses on three types of severe plastic deformation consolidation (SPDC) methods. First is a compaction of ordinary sized powders, cbaracterized by a formation of supersaturated nanocrystals, because second pbases get dissolved during SPDC. Second is a consolidation of metal and ceramic powders in order to produce nanocomposites with high strength and thermostability. Third is the consolidation of ball-milled powders. This latter SPDC process is capable of producing quasi-amorphous nanosolids, formed due to a presence of a very high density of dislocations (up to 10 17 m-2) situated at grain boundaries. Atomic structure and some attractive properties of these unusual materials are considered. 4:40pm STRUCTURE AND PROPERTIES OF MECHANICALLY ALLOYED NI -AI AND Ag-AI POWDERS BEFORE AND AFTER LEACHING OF ALUMINUM: B. Zeifert, R. Esquivel, D. Jaramillo-Vigueras, J.G. Cabafias-Moreno, Instituto Politechnico Nacional, ESIQIE, Apdo, Postal 75-373, 07300 M~xico, D.P., M~xico Elemental powder mixtures of AI+Ni and AI+Ag, in a variety of nominal compositions, have been milled in an attritor mill and a horizontal ball mill. TIle mechanically alloyed powders bave been characterized by XRD, SEM, 1EM, DSC and chemical analysis. Ni-AI mixtures with nominal aluminum contents in the range from 55 to 70 atomic % mainly containing the NiAI-type intermetallic phase after sufficiently long milling times. On the other hand, the composition Ag65AI35 was identified as the yAg2AI phase. Leaching of aluminum from the mecbanically alloyed powders has been successfully accomplished and the leached materials are currently being tested as potential catalytic solids. This work was funded by IPN (DEPI, COFAA, PIFI) and CONACYT. 5:05pm COMPACTION OF MECHANICALLY ALLOYED IRON POWDERS: J.G. Rawers, D. Alman, U.S. Bureau of Mines, Albany Research Center, Albany OR 973212198; J. Groza, Materials Science, University of California Davis, Davis CA, G. Korth, Idaho National Engineering Laboratory, Idaho Falls ID Although much has been written about high-energy Milling of powders and the prop-

B84

erties of the resulting powder, there bas been limited information about consolidation of these materials. This study describes and compares the results of several techniques recently attempted to consolidate mechanically alloyed, micron size, nanograin iron alloys. Powers were all prepared using It high energy attrition ball-mill, and analyzed for phase, grain size, and internal strain. Consolidation included: (i) explosive compaction, (ii) high pressure (>3GPa) cold compaction, and (iii) low pressures (-50MPa) compaction at temperatures from 600 to lOOO·C with hold times from a few minutes to an hour. Consolidated samples were examined for pbase, grain size, and hardness.

TANTALUM: MICROSTRUCTURE, PROPERTIES AND MODELING: Session IV

Sponsored by: SMD Refractory Metals and Materials Committee and Jt. MDMDI EPD Synthesis, Control and Analysis In Materials Processing Program Organizers: Andrew Crowson, U.S. Army Research Office, Research Triangle Park, NC; Edward S. Chen, U.S. Army Research Office, Research Triangle Park, NC; Prabhat Kumar, Cabot Corp, BoyeItown, PA; Willam Ebibara, Picatinny Arsenal, Picatinny, NJ; Enrique J. Lavernia, UC Irvine, Irvine, CA Tuesday, PM February 6, 1996

Room:A4-5 Location: Anaheim Convention Center

Session Chairpersons: William Ebiharu, Picatinny Arsenal, Picatinny, NJ. David H. Lassila, Lawrence Livennore National Laboratory, Livennore, CA

1:30pm FACTORS EFFECTING THE MECHANICAL PROPERTIES AND TEXTURE OF TANTALUM: Christoper A. Michaluk, Project Leader, Cabot Performance Materials, County Line Road, Boyertown, PA, 19512 A comprehensive review of factors which influence the microstructure, strength, and texture of tantalum is presented. Data demonstrating the roles of melting procedure (Electron Beam versus Vacuum Arc Rc:melting), alloy additions, and annealing temperature on mecbanical properties, metallurgical structure, and texture is provided. Results reveal that material chemistry ba.~ an overbearing effect on the bebavior of the final product. 2:00pm MULTIPLE MECHANISMS THE THERMALLY ACTIVATED FLOW OF TANTALUM: William H. Gourdin, David H. Lassila, University of California, Lawrence Livermore National Laboratory, PO Box 808, L-355, Livennore, CA 94550 We propose a unified, physically based, approach to modeling of BCC and FCC materials which incorporates a two-component Peierls-type mechanism and an FCC obstacle mecbanism in series. We have applied our model to unalloyed tantalum with reasonably good success over a wide range of strain rates and a modest range of temperatures. We find !bat the proposed two component Peierls term is necessary to properly reproduce the initial flow stress oyer the entire range of strain-rates and temperatures studied. We will discuss the rell:vance of the proposed model to modeling of shock-prestrained tantalum. 2:lO pm

WORK-HARDENING AND DISLOCATION BEHAVIOR OF TANTALUM AND TANTALUM ALLOYS: W. Wasserbach, Max-Planck-Institut fur Metallforschung, Institut fur Physik, HeisenberstraMe I, D-70569 Stuttgart, Germany Higb-purity tantalum and tantalum alloy single crystals have been deformed in the temperature region between 17K and 470K. TIle deformation bebavior bas been studied by Slip-line morphology, Laue X-ray backreflexion, X-ray topography, transmission electron microscopy and meaSUl1:ments of the flow stress. At low deformation temperatures the work-hardening cum:s are parabolic and the dislocation arrangement consists of primary and secondary screw dislocations. At intermediate temperatures the work-hardening curves exhibit two regimes which are analogous to stage I and stage III of the work-hardening curves of face-centered cubic metals. In stage I bundles of edge dipoles of primary dislocations are observed. In stage III the dislocations are arranged in symmetric pairs of dislocation sheets consisting of crossed grids of primary and secondary dislocations parallel to the primary slip plane. In spite of the different dislocation arrangements, in both stages th,: macroscopic slip proceeds by glide over large slip distances. 2:40pm POLYCRYSTAL ELASTO-PLASTICITY: APPLICATION TO PREDICTION OF EARING DURING CUP DRAWING OF TANTALUM SHEET: Dr. Lallit Anand, Manish Kothari, 1-310, Department of Mechanical Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139 Constitutive equations for polycrySllll elasto-plasticity, which are capable of modeling the initial and evolving anisotropy in ductile materials due to the evolution of crystallographic texture are reviewed, and the implementation of these equations in a finite element program is described. The constitutive equations and computational procedures are used to simulate the formation of earing defects during cup-drawing of bodycentered-cubic tantalum sheet. Comparisons of the results of the numerical simulations against existing experimental results on earing shows that the predictions of the number of ears and their positions are in excellent agreement with the experiments. TIlese results should have excellent bearing on the formation of fins and flares in explosively formed penetrators. 3:00 pm BREAK

1996 TMS ANNUAL MEETING AND EXHIBITION

3:15pm INFLUENCE OF LARGE-STRAIN DEFORMATION ON THE MICROSTRUCTURE TEXTURE AND MECHANICAL RESPONSE OF TANTALUM BAR: G. T. (RuslV) Gray IIi. Stuart l. Wright, Shuh Rong Chen, Shem Bingert, Paul L. Maudlin Los Alamos National Laboratory, Los Alamos, NM 87545 Numerous studies to date have established the influence of impurities, crystallographic texture, temperature, and strain rate separately or collectively on the cons~tu­ tive response of annealed tantalum, in particular plate s!ock. However, fewer de~led studies have examined the evolution of crystallographic texture and the mechanical response of tantalum bar or rod material following prestrains to strains> I. In this talk the influence of large plastic prestraining on the microstructure evolution, texture evolution, and mechanical response of high-purity tantalum bar material will be ~ese.nted. Tantalum cylinders annealed at 1200·C were quasi-statically upset forged, With m~r­ mediate lubrication to true strains of 0.4, I, and 2. Microstructural and texture banding within the starting ~talum bar was characterized in detail and found to evolve during large-strain forging leading to significant differences in mechanical response. Aspects of defect storage, work-hardening response, and texture evolution in tantalu~ bar as a function of forging strain will he discussed. Work performed under the auspices of the U.S. Department of Energy. 3:45pm MODELING DYNAMIC BEHAVIOR AND TEXTURE EVOLUTION IN PURE 1'8: Seoll E. Schoenfeld, U.S. Army Research Laboratory, AMSRL-WT-TD, Aberdee.n Proving Ground, MD 21005; Said Ahzi, Clemson University, Department of Mechanical Engineering, Clemson, SC 29634 In order to model high temperature, high strain rate deformation and textu,: e~olu­ tion in commercially pure Ta, two distinct descriptions for the thermal elasuc VISCOplastic behavior of Ta single crystals will be considered along with associated polycrystal averaging schemes. The first deSCription will i?corporate a temperature dependent description for pencil glide on the planes of maximum resolved shear stress. Calculated stress-strain data and texture evolution for these two models will be compared with experimental data taken over a range of temperature and stain rates. 4:05pm GRAIN SUBDIVISION AND THE DEVELOPMENT OF LOCAL ORIENTATIONS IN ROLLED TANTALUM: DA. Hughes, Center for Materials and Appl. Mechanics Sandia National Laboratory, Livermore, CA 94550 The fo~tion of dislocation boundaries and their local surroundings determine the properties of deformed tantalum including texture, recrystallization, flow stress and formability. The deformation microstructure formed in tantalum has been o~erved ~d quantitatively characterized using transmission electr~n microscopy followmg rOlli.ng to large reductions. The observed structure includes higher angle lamellar boundarie.s parallel to the rolling plane surrounding either groups of cells or arrays of loosely kmt dislocations and equiaxed subgrains. The distribution of local orientations between .individual dislocation boundaries and their angle/axis pairs have been measured usmg convergent beam Kikuchi analysis and are compared to the rolling texture. The sequence of near neighbor orientations shows that individual grain~ subdivide into three to four different texture components separated by sharp boundanes. The arrangement of local orientations is much more diverse than suggested by simple models and single crystal studies. This work supported by U.S. DOE under contract No. DE-AC0494AL8500. 4:25pm SITE COMPETITION OF IMPURITIES AND GRAIN BOUNDARY STABILITY IN TANTALUM: Genrich L. Krasko, U.S. Army Research Laboratory, Materials Directorate, AMSRL-MA-CC, Aberdeen Proving Ground, MD 21005-5069 Metalloid impurities have a very low solubility in tantalum, and therefore prefer to segregate at the grain boundaries (GBs). In order to analyze the energetics of the impurities on the tantalum GB, the LMTO calculations were performed on a simple 8-atom supercell emulating a typical (capped trigonal prism) GB environment, and the socalled "environment-sensitive embedding energies" were calculated for B, C, N, 0, P, and S, as a function of the electron charge density due to the host atoms at the impurity site. The calculations showed that, at the electron density typical of a GB, C has the lowest energy (followed by N and B) and thus would compete with the other i~urities for the site on the GB, tending to displace them from the GB. The above energl~s were then used in a modified Finnis-Sinclair embedded atom approach for calculating the cohesive energies and the eqUilibrium interplanar distances in the vicinity of a (1\1) 3 tilt GB plane, both for the clean GB and that with a? impurity. ~~e ~istances ~er~ found to oscillate, returning to the value correspondmg to the eqUilibrium spacmg 10 b~Ik BCC tantalum by the IOth- 12th plane off the GB. The results on both the GB relaxation and GB cohesive energies suggest that C, N and B are intergranular cohesion enhancers, while 0, P and S result in decohesion effects. 4:45pm EFFECT OF DATA SET SIZE ON CONSTITUTIVE MATERIAL MODELING OF TANTALUM: Craig M. Lopatin, Craig L Willman, Alliant Techsystems Inc, 600 2nd Street N.E., Hopkins, MN 55343 Analytical and numerical techniques were used to derive the constants for the JohnsonCook material model from data sets of various sizes for pure tantalum. The numerical curve fitting routine was based on a commercial graphiCS package which adjusts all the constants simultaneously to obtain the best fit. Numerical solutions are then compared to those given by solving for each term in the model separately in a standar~ analytical approach. Recommendations are given for the minimum number of data pomts needed to obtain an accurate material model.

1996 TMS ANNUAL MEETING AND EXHIBITION

TECHNICAL, ECONOMIC COMMERCIAL & ENVIRONMENTAL CONSEQUENCES OF SULFUR RECOVERY IN COPPER PRODUCTION SYSTEMS II: Part 1: Management And Market Implications Of S02 Pollution Abatement; Part 2: Pyrometallurgal In-Situ Sulfur Fixation:

Sponsored by: EPD Copper, Nickel, Cobalt Committee Program Organizers: Norbert L. Piret, Piret & Stolberg Partners, 1m Licht 12, D47279 Duisburg, Germany; Ivan M. Santos Moraes, Carafba Metais, Oias D' Avila, Bahia, Brazil Tuesday,PM February 6, 1996

Room: AlI-12 Location: Anaheim Convention Center

Session Chairpersons: J.F. Castle, RTZ Technical Services Ltd., PO Box 50, Castlemead, Lower Castle Str., Bristol, BS99 7YR, UK; J.E. Hoffmann, Jan H. Reimers & Associates, PO Box 420545, Houston, TX 77242-0545

2:00 pm Invited CODELCO-CHILE'S CORPORATE DECONTAMINATION PLAN FOR ITS SMELTERS: Gerardo Muiioz, Codelco-Chile, Corporaci6n Nacional del Cobre de Chile, Huerfanos 1270, Casilla 150-D, Santiago de Chile All along this century Codelco's smelters have been operating on conventional technology, which Codelco has streamlined by using oxygen, developing oxy-fuel burners for reverberatory furnaces and introducing oxygen smelting of concentrates in converters (Teniente Converters). Chuquicarnata began capturing sulfur from its first Teniente Converter in 1985. Since then, it has implemented a program allowing to gradually increase such capture. In 1994 a Frame Law on the Environment was enacted in Chile. By virtue thereof and of Codelco's corporate environmental policy, an investment program aimed at capturing more than 95 % of gaseous pollutants in. all its smel~rs ~n a term not exceeding twelve years was established. The above entails a substantial 1mprovement of the operational management as regards fugitive gases capture and treatment. This paper describes the way Codelco is to deal with this challenge with a view to fully complying with environmental regulations and obtaining the required environmental certifications of its operations. 2:30pm ENVIRONMENTAL MANAGEMENT MEASURES RELATING TO A GREENFIELD COPPER SMELTER: Norbert L Piret, Karl H Schmidt, Piret & Stolberg Partners, 1m Licht 12, 0-47279 Duisburg/Germany From the copper smelter operation various outputs are likely to exert negative impacts on the environment. Particularly, emissions like SD2 and dusts, solid wastes like contaminated gypsum and arsenic sludge, effluents and heated cooling water are generated. Most countries have stipulated an environmental legislation to provide the framework for an efficient environmental management As part of the planning activities for a major greenfield copper smelter, the following elaborations are required and were drawn up: The Environmental Impact Assessment, which is the first step of environmental management and mandatory for every new project; The Environmental Monitoring Plan, which describes the monitOring procedures for the main impacts identified and assessed; The Environmental Management Plan, which deals with the minimization of environmental impacts, and the methods of its implementation. 2:55 pm Invited SMELTER ACID IN THE BROADER MARKET PERSPECTIVE: Michael F. KillO, Research Manager, Sulphur, British Sulphur Consultants, 31 Mount Pleasant, London WCIX DAD, Great Britain The production of sulfuric acid at base metals smelters is predominantly an ~nvolun­ tary activity and now comprises approximately 17 % of all-forms sulfur supply 1R world terms. The majority of elemental sulfur is also produced involuntarily in response to environmental regulations regarding emissions from hydrocarbon processing plants. The paper examines the points of similarity and difference in the sulfur and smelt~ acid markets and reviews their interactions. The merchant acid market collapsed 1R 1993 but has 'subsequently rallied. The reasons for these developments are considered, and the likelihood of a repeat of 1993 is assessed against the projected growth of overall acid demand and of smelter acid production in the period to the year 2005. 3:25 pm BREAK 3:45pm REACTION OF ruGH STRENGTH SULFUR DIOXIDE WITH LIMESTONE AND DOLOMITE AT ELEVATED TEMPERATURE: Ryan Cavanagh, N. A. Warner, School of Chemical Engineering, The University of Birmingham, Edgbaston, Birmingham, BI5 2TT, UK Studies relating to the fixation of sulfur by reaction of 3 to 8 rom diameter cylinders of calcined limestone or dolomite with high strength sulfur dioxide in the temperature range 880·C to 1020·C were undertaken. At relatively low SD2 concentrations gaseous diffusion through the reaction product layer has a marked effect on the rate. As SD2 concentrations are increased up to 100 percent more complex rate controlling mechanisms unfold. By sectioning partially reacted pellets a moving interface was observed but SEM analysis indicated that substantial amounts of unreacted calcium oxide were retained in the reaction product layer. Reaction ceased because of pore plugging for the most reactive limestone at 60-65 % conversion, whereas for dolomite virtually

B85

complete sulfation of the calcium was observed. Lower conversions in the less reactive limestones are related to pore structure and internal BET surface area. 4:10pm SULFUR FIXATION AT HIGH TEMPERATURE WITmN THE COPPER SMELTING PROCESS: N A Warner, School of Chemical Engineering, The University of Birmingham, Edgbaston, Birmingham, B 1521T, UK Fundamental studies on the reaction of limestone or dolomite at high temperature are extended into the realm of large scale primary copper smelting. Whereas flue gas desulfurisation in the electric power utility industry is concerned with S~ concentrations normally less than 0.5 percent, high temperature smelting gases can have S~ levels approaching 100 percent. The reaction involved is very exothermic so proposals are made in this paper for returning the heat generated in sulfur fixation directly back to the copper smelting process. This integration leads to metal production with lower energy requirements than current technology. The anhydrite reaction product can either be disposed of locally with minimum adverse environmental impact or upgraded to a value-added by-product. It is assumed, of course, that no viable market exists to support sulfuric acid production. 4:35pm MINIMIZATION AND CONTROL OF S~-EMISSIONS BY APPLICATION OF THE CIRCULATING FLUID BED: H. Hirsch/elder, E. Stolarski, Lurgi Metallurgie GmbH. Lurgiallee 5, 0-60295 Frankfurt am Main, Germany The Circulating Fluid Bed (CFB) has now surpassed a quarter of a century of commercial operations. There are more than 80 CFB units operating in the fields of alumina, coal combustion and gold roasting. The paper describes how S~ is captured utilizing the CFB technology in coal combustion and gold roasting plants. The process, by means of improved mass transfer, enhances the capture of S~ in the vessel. Depending upon the feed material analyses, additives may also be required to minimize S~ emissions. An additional benefit of this process is that it permits the installation of a simpler, less costly off-gas treatment plant.

TRANSIENT THERMAL PROCESSING OF MATERIALS: SESSION III: Non Silicon Processes

Sponsored by: EMPMD Thin FIlms & Interfaces Committee Program Organizers: N. M. Ravindra, New Jersey Institute of Technology, Newark, NJ; R. K. Singh, University of Florida, Gainesville, FL Tuesday, PM February 6, 1996

Room: Grand J Location: Anaheim Marriott Hotel

Session Chairmen: Kevin S. Jones, University of Florida, Gainesville, FL; James C. Sturm, Princeton University, Princeton, NJ

2:00 pm Invited RAPID THERMAL PROCESSING OF REFRACTORY SILICIDES FOR THE GaAs DEVICE TECHNOLOGY: A. Cbristou, Department of Materials and Nuclear Engineering, University of Maryland, College Park, MD 20742 Abstract not available. 2:30 pm Invited TRANSIENT THERMAL PROCESSING IN III-V SEMICONDUCTOR TECHNOLOGY: S. J. Pearton, Department of Materials Science & Engineering, University of Florida, Gainesville, FL 32611 Transient processing is employed at numerous points in the processing onlI-V electronic and photonic devices. Examples include implant activation of both donor and acceptor dopants, alloying of ohmic contacts, maximization of sheet resistance in implant-isolated regions and RTCVD of dielectrics, metallization and semiconductors. Recently there has been tremendous interest in GaN and related alloys for blue emitters. We have achieved formation of both n-type and p-type regions in GaN using rapid thermal annealing at - llOO'C on Si or MgIP implanted material. We will review the use of transient thermal processing in III-V semiconductor technology and identify areas for future research. 3:00 pm Invited INFLUENCE OF RTP ANNEAL ON DISTRIBUTION OF OXYGEN 16118 IN Ge AND GaAs: M. Dubey, R. Lareau, M. Ervin, Army Research Lab, Advanced Device Fabrication Div, Ft. Monmouth, Nl07703 Abstract not available. 3:30 pm BREAK 3:45 pm Invited EFFECT OF RAPID THERMAL ANNEAL ON THE ELECTRICAL PROPERTIES OF POLY CRYSTALLINE DIAMOND: Eng Kie Souw, Department of Applied Technologies, Brookhaven National Lab, Bldg. 820M, Upton, NY 11973; F.M. Tong, N.M. Ravindra, Department of Physics, New Jersey Institute of Technology, University Heights, Newark, NJ 07102; R.I. Meilunas, Northrop Grumman Corpn., MS AOI-26, Bethpage, NY 11714 Abstract not available. 4:15pm PROCESSING OFmGH TEMPERATURE STABLE CONTACTS TO SINGLE CRYSTAL DIAMOND: A. Vescan, P. Gluche, W. Ebest, T.H. Borst, E. Kohn, Department of Electron Devices and Circuits, University of Ulm, D-89081, Ulm, Germany

B86

In principle, the wide bandgap of diamond allows operation of the crystal at very high temperatures in the order of 1200 K. One of the key requirements for such an operation is an appropriate metallization system for Ohmic and rectifying contacts. In a recent experiment!I], we have shown that Si on p-type diamond can form chemically and electronically stable contacts up to approximately 850 deg C. On moderately boron-doped layers, it forms a Schottky batTier of 1.2 eV barrier height, whereas on p+ doped material, the contact is Ohmic by tunneling. Thus, Si on diamond is a very promising candidate for extremely high temperature operation of diamond device structures. However, for technological use, this contlet needs to be complemented by a high conductivity overlayer of Au. But Au will alloy with Si at low temperatures destroying the Si-diamond interface. Therefore, a high temperature stable diffusion barrier needs to be inserted. Here WSi- alloys have been invc:stigated. All contact layers have been deposited by sputtering and they are therefore amorphous. But, during high temperature operation, recrystallization and reactions with the Si-contact layers may occur. Therefore, the deposition process has to he accompanied by an appropriate temperature procedure and nitrogen treatment to control and stuff grain boundaries. Details on the processing procedure and chemical and electrical stability will be discussed in this presentation. Reference:!I]. A. Kescan et.a!., 4th Intemational Symposium on Diamond Materials, Reno, Nevada (1995). 4:35pm RTPPROCESSING OF MAGNETIC ALLOYS: J. Viatella & R.K. Singh, Department of Materials Science & Engineering, University of Florida, Gainesville, FL 32611 Abstract not available. 4:55pm REDISTRIBUTION OF ZN IN INDIUM PHOSPHIDE UNDER PULSED LASERANNEALING ACCOMPANIED BY PHASE TRANSITION: M.1. Markevich, A.M. Chaplanov, F.A. Piskounov Institute of Electronics BAS, 220090, 22 Logoisldi Trakt, Minsk, Belarus The report presents the physical and mathematical models of the redistribution of implanted metal impurities in A3B5 serrdconductors with due regards for phase melting and crystallization areas. The model consists of three differential equations. The irradiation transfer equation allowed for the absorption of heat by the non-transparent components of the transformation of laser irradiation into heat and the absorption of the later at phase transition in the dual phase area and during melting and recrystallization. The mass stability equation for the impurity allowed for its diffusion; the diffusion coefficients were different for single and dual phase areas. The redistribution profiles of implanted zinc in indium phosphide following the nanosecond laser treatment have also been presented. 5:15pm SPECTRA STUDY OF PULSED YAG LASER SPUTTERING PROCESS FOR CARBON NITRIDE (CNX) FILM DEPOSITION: Zhi-Feng Ying, Zhong-Min Ren, Yuan-Cheng Du, Fu-Ming Li, State Key Lab. for Material Modification by Laser, Ion and Electron Beams, Fudan University, Shanghai 200433, China A high purity graphite target was sputtered by a Pulsed-Laser YAG with various wavelengths (532, 355 nm and 1.06 urn) during Carbon nitride (CNx) films deposition co-processed with low energy nitrogen ion beams. In pulsed laser transient sputtering processing for graphite, the optical emission spectra have shown that the pulsed laser transient sputtering will lead to stronger peaks such as CII: 392.1,407.5 and 426.7 nm etc. for active carbon atoms with 532 nm YAG Laser than the one with both of 355 nm and 1.06 urn YAG laser. In experiments, it is also shown that crystallization structure according to Electron diffraction (ED) patterns of Carbon nitride (CNx) films with 532 nm YAG has heen improved. For potential applications, it will be beneficial to the compound films growth, such as Beta - C3N4. In this paper, the synthesis processes of carbon nitride (CNx) films deposition due to the produced carbon and nitrogen atoms interaction will he discussed briefly.

WEDNESDAY AM ADVANCES IN THE SCIENCE AND TECHNOLOGY OF TITANIUM ALLOY PROCESSING-V: Heilt Treatment I

Sponsored by: SMD Titanium Committee, MDMD Shaping and Forming Committee

Program Organizers: Prof. Isaac Weiss, Prof. Raghavan Srinivasan, Mechanical and

Materials Engineering Dept, Wright State University, Dayton, OH 45435; Dr. Paul Bania, TImet Corporation, TImet-Henderson Technical Laboratory, P.O. Box 2128, Henderson, NV 89009; Prof. Daniel Eylon, Graduate Materials Engineering, University of Dayton, Dayton, OH 45409 Wednesday, AM February 7, 1996

Room: B5-6 Location: Anaheim Convention Center

Session Chairpersons: T. Broderick, ProceSSing Science Group, Wright Laboratory, Materials Directorate, Wright Patterson AFB, OH 45433; l.R. Wood, RMI TItanium Company, 1000 Warren Avenue, Niles, OH 44446 8:30 am Invited HEAT TREATMENT OF TITANIliM ALLOYS: OVERVIEW: Rodney R. Boyer, Boeing Commercial Airplane Group, Seattle, WA 98124 There are several types of heat treatments used for titanium alloys, depending on the alloy type and specific application. Heat treatments include stress relief, annealing,

1996 TMS ANNUAL MEETING AND EXHIBITION

duplex-, triplex-, recrystallization-, and beta-annealing, solution treat and aging, duplex aging etc. Each of these heat treatments, in conjunction with the prior processing history, provides a unique microstructure and properties combination. The heat treatments for the various alloy types and their rationale-effect on the microstructure and properties - will be discussed. Different heat treatments are utilized to optimize certain properties, such as fatigue, toughness, strength, etc. 8:55 am Invited HEAT TREATMENT CHALLENGES FOR TITANIUM IN GAS TURBINE ENGINE APPLICATIONS: James Hall, Allied Signal Engines, Phoenix, AZ 85072 In modem gas turbine, aero propulsion, engines, the concurrent needs for increased perfonnance, improved efficiency and greater reliability have placed heavy demands on titanium rotating components. These demands push titanium alloys closer to their capacity and thus, in tum, creates the need for better and tighter property control. It is the purpose of this paper to review these challenges with respect to microstructure control and the role that heat treatment and related processing play in affecting improved perfonnance and reliability related properties. 9:20am SELECTION OF OPTIMUM TECHNOLOGIES OF HEAT TREATMENT OF INTRICATELY SHAPED TITANIUM ALLOY ARTICLES: A.N. Lozhko, G.Z. Malkin, State Metallurgical Academy of Ukraine, Dnepropetrovsk, Ukraine Reheating of intricately shaped titanium alloy blanks was studied. Optimum heating schedules have been determined which provide for minimum oxidation, gas saturation, and the required final blank temperatures. Cooling of intricately shaped titanium alloy blanks during the thermal treatment procedure will be described. The optimum cooling scbedules necessary to provide the required microstructure of the articles due to controlled and variable cooling of the sections of the articles with different mass was determined. Oxidation and gas saturation of titanium alloys in air and in combustion products of coke and natural gases have been studied experimentally at different composition of the fuel mix, within the temperature range of 900-1200°C. A search of the scientific technology of cooling (or heating) of an intricately shaped article has been carried out on a computer model which visualized the distribution of temperatures, distribution of heating (cooling) rates, gas saturation layer and scale layer during all the heat treatment stages. An original mathematical apparatus based on orthogonal conformal transformations has been used for the simulation of the non-stationary temperature field. This enabled to solve the problem with higher speed and accuracy in comparison with such methods as finite or bottleneck elements. 9:35am EFFECT OF COOLING RATE ON PHASE TRANSFORMATIONS IN TI-t'iAI4V: T. Ahmed, H.J. Rack. Materials Science and Engineering Program, Clemson University, Clemson, SC 29634-0921 The effect of cooling rate on the microstructure ofTI-6AI-4V has been examined. At the highest cooling rate examined, 523°Cls, S transformed to acicular hexagonal martensite, a'. Decreasing the cooling rate to 41O°Cls resulted in the introduction of a competitive massive transformation reaction, S ~ml whose volume fraction increased with decreasing cooling rate. Finally, at slow cooling rates (=2°Cls) the microstructure was comprised of Widmanstiittan These results will be discussed with relation to the thermodynamic stability of a' , am and aw as represented in the form of a ccr diagram.

aw.

9:50am TITANIUM PRODUCTION TECHNOLOGIES AND THE AUTOMOTIVE MARKET: Christopher Sommer, Timet, 8615 Richardson Road, Suite 100, Walled Lake M148390 Due to the downsizing of the military aerospace industry and the fixed production overcapacity in the titanium industry, many new markets need to be pursued. One outstanding case in point is the O.E.M. automotive world. This industry is faced with a daunting task. That is to design and build, on a consumer affordable basis, a lightweight, energy efficient automobile. The recent approval of stricter C.A.F.E. requirements and the coming push for further clean air legislation ie; the electric car, has certainly sparked new intrest in titianium as a material of choice for automotive components. With this new interest increased focus has been placed on finished part costs, bypassing to a degree, the age old stigma of high input material cost. The focus of the titanium industry today should be placed on these finisbed component opportunities and the problems presented with their manufacture. This paper will present some of these components and so far successful methods of manufacture used to overcome these obstacles. 10:05 am PROCESSING, MICROSTRUCTURE, AND PROPERTIES of 8-CEZ: G. Lutjering, 1.0. Peters, TU Hamburg-Harburg, 21071 Hamburg, Gennany The influence of different processing routes (a+S forging, "trough-transus" forging) and different heat treatments on the microstructure and mechanical properties of 6CEZ will be discussed. The advantages and disadvantages of the so-called "necklace" microstructure achieved by "trough-transus" deformation will be outlined in comparison to the bi-modal microstructure resulting from conventional a+S deformation. The discussion of the mechanical properties is based on tensile tests, fracture toughness tests, and fatigue tests (S-N curves, daldN-K curves of micro-and macro- cracks). 10:20 am

PRINCIPLES OF TITANIUM ALLOYS' STRUCTURE CONTROL WITH A PURPOSE OF INCREASING THEIR MECHANICAL PROPERTIES: M. Brun, G. Shachanova, All-Russia Institure of Light Alloys, VILS, Moscow, Russia Titanium alloys typical have diverse microstructures due to the presence of several phases, differences in their morphology, wide range of volume fractions and sizes of various structural constituents. The diversity of these alloys (especially of a+S alloys) has a significant effect on their mechanical properties. The following "structural contribution to the properties" in annealed condition was obtained from the experimental

1996 TMS ANNUAL MEETING AND EXHIBITION

data: strength -150-200 MPa, elongation -10-15%, reduction of area -40-40%, fracture toughness -50 -60 MPa Vm, high- cycle fatigue strength 120-150 MPa, creep resistance - 80-120 MPa, etc. These values are extremely high and exceed the so called "superiority's quota" which is used as one of the conditions forthe development of new alloys. Another important aspect of the "structure-properties" problem for titanium alloys is that mechanical properties (ductility, fatigue resistance, etc.) are influenced by many structural parameters. The task of selecting the preferable structure selectionfor a specific component develops in 3 directions: I. revealing of the mechanical properties which determine serviceability of various components; 2. determination of the correlation between structure and properties; 3. establishment of the conditions necessary to obtain required structures. This presentation is devoted to the consideration of titanium alloy structure formation's generality at deformation and heat treatment, as a scientifical base for the regulation of semiproducts structure. It will consider the general conceptions and quantitutive data for the influence of deformation and heat treatment regimes to various parameters of a+6- alloys' structure. 10:35 am EFFECTS OF HEAT TREATMENT ON MATRIX MICROSTRUCTURE, INTERFACIALREACTIONS AND FATIGUE CRACK GROWTH RESISTANCE OF TITANIUM METAL MATRIX COMPOSITES: S. V.Sweby, A.Dowson, and P.Bowen, School of Metallurgy and Materials IRC in Materials for High Perfonnance Applications, The University of Birmingham, Edgbaston, B15 2IT UK A study has been undertaken to assess the effects of heat treatment on two titanium alloy metal matrix composites (MMC).TI-6AI-4V (IMI 318) and TI-5.8AI-4Sn-3.5Zr0.Nb-0.5Mo-O.3Si (IMI 834) were chosen for this study, reinforced with SiC fibres. Beta transus approach curves were generated for the composite matrices. Shifts in beta transus temperature of up to 55°C have been recorded for the matrix regions of the composite over the monolithic alloys. Two fibre coating systems have also been evaluated: SMI240, which consists of a llUI1 carbon and a llUI1 titanium boride layer and SMI140+, which consists of a 4.511m carbon layer. Considerable differences in the morphology of the fibre/matrix interfacial regions and their effects on the fatigue crack growth resistance have also been recorded. The effects of matrix microstructure and interfacial reactions on the fatigue crack growth resistance of the two alloys will be discussed. 10:50 am BREAK

11:00 am MlCROSTRUCTURFJFATIGUE CRACK PROPAGATION RELATIONSHIPS FOR SP700 TITANIUM SHEET: C. N. Kimura. Boeing Defense & Space Group, Materials & Processes Technology, P.O. Box 3999, MS 3E-ffi, Seattle, WA 98124; T. F. Archbold, University of Washington, Department of Materials Science & Engineering, FB-IO, Seattle, WA 98195 SP700 is a beta-rich, alpha-beta titanium alloy (TI-4.5AI-3V-2Mo-2Fe) specially developed by NKK (Japan) for superplastic forming (SPF). The alloy has a fme grain microstructure well-suited for SPF, exhibiting lower flow stresses and processing temperatures than TI-6AI-4V. These characteristics provide the potential for less tooling wear, less accumulation of oxidation contamination (alpha case) on surfaces of SPF parts, and ultimately, reduced cost of SPF titanium parts. In addition, NKK has reported SP700's improved hot and cold formability, fatigue strength, and toughness. The alloy's heat treat response and age bardenability result in a wide range of microstructures which includes elongated primary alpha. The microstructure/properties relationship of similar microstructures in TI-6AI-4V is well-documented. This investigation seeks to identify microstructure vs. fatigue crack propagation relationships for SP700 sheet material. The microstructures described above will be evaluated for this study. A quantitative assessment of each microstructure in terms of primary alpha grain morphology and other factors will be related to tensile properties, daldN vs. delta K, and fracture surface morphology. These relationships will be compared with those for TI-6A1-4V. 11:15 am EFFECT OF THE COOLING RATE FROM MILL ANNEALING TEMPERATURE ON STRESS CORROSION THRESHOLD OFTI-6AI-4V ELI BETA ANNEALED: Bob Briggs, Boeing Company, PO Box 3707, MIS 73-44, Seattle, WA 99124 The cooling rate from the mill anneal heat treatment following the beta anneal was shown to have a dramatic effect on the stress corrosion threshold ofTI 6AI-4V ELI beta annealed. The stress corrosion threshold was reduced by more than half when slow cooling was performed. The stress corrosion threshold versus seven different cooling rates (from air cool to furnace cool) was measured and a correlation between stress corrosion threshold and cooling rate obtained. Additional stress corrosion specimens were exposed to various times and temperatures, and the resulting change in stress corrosion threshold used to create a TIme-Temperature-Transformation (TIT) diagram to indicate the range in which the stress corrosion susceptibility was most acute, attributed to the formation of TI3Al. The fracture surface of the stress corrosion specimens were examined metallographically and with SEM and TEM. Microstructural features such as colony size of transformed beta and the width of lamellar alpha were correlated to details observed on the fracture surface. TEM was used to examine specimens for the presence of TI3Al. The formation of TI3AI was explained using the TI-Al phase diagram and the measuring aluminum content of the alpha phase due to partitioning.

11:30 am ENHANCEMENT OF STRENGTH AND CREEP PROPERTIES OF TITANIUM ALLOYS BY SPECIAL a2-PHASE PRECIPITATION: P.E. Markovsky, Institute for Metal Physics National Academy of Sciences of the Ukraine, 36 Vemadsky str., 252142, Ukraine An additional way of mechanical and creep property enhancement for titanium alloys with 6 wt. % aluminum and more the hardening by precipitation of very fine particles of a2 (TI3AI)-phase is shown. The process of a2-phase precipitation is quickened by special deformation before the ageing, that allows to shorten ageing time from 24 hours and more to 2-3 hours. Since a2-phase precipitation leads to decreasing of

B87

titanium alloys' ductility this effect may be reduced by alloys' structure refinement.

The best balance of tensile and creep properties have alloys with average grain size less

than 2 IllD reinforced by a2 - phase precipitates. For example, TI-6A1-4V alloy with average grain size I-211m after this treatment has UTS > 1300 MPa, AS > 9% and creep resistance characteristics at SOO'C better than special high temperature titanium alloy TI6242 in bimoda1 state tested at the same conditions. 11:45 am HEAT TREATMENT OF LARGE-SCALE SEMIPRODUCTS FROM TITANIUM ALLOYS: A.A. Ilyin, M. Yu. Kollerov, A.A. Krastilevsky, Moscow State University of Aviation Technology, Metals Science Dep., Petrovlca St. 27, Moscow, K-31, Russia, 103767 Specific strength is among the first ordered requirements for the materials utilized in commercial industry (especially in aerospace). Thus on each stage of semiproducts and products manufacture they need various types of strengthening, including heat treatment. Strengthening heat treatment has not found wide application towards the largescale semiproducts due to temperature distribution inhomogenity through the crosssection of semiproduct. It causes the irregular phase composition and structure through the semiproduct reansverse, as well as causes the high level of temporary residual micro and macro-stresses. The new type of diagrarus are suggested to predict a titanium alloy phase composition formed during the 8-phase decomposition at continuous cooling. There was found some regular trends of cooling rates from various temperatures and future ageing influence on mechanical properties of TI alloys of diverse classes. It is shown that the maximum level of strength and ductility is observed after cooling from 8-area with the rate near as called third critical rate (Vc3). The formation of residual and temporary microstresses appeared as a result of inhomogenity of temperature field and the phase transformation taking place during heating and COOling through the cross-section of plates and bars from TI-6A1-4V; TI-6AI-2Mo-4V-ICr-IFe; TISAI-4M0-4V-ICr-lFe. The critical size of semiproducts, which could be heated and cooled in various natures without residual microstresses, was received. On the hase of formed microstructure, phase composition, properties and microstresses level through the cross-sections of large-scale semiproducts (bars and plates) from investigated alloys, the new types heat treatment, that leads to the maximum strength properties with simultaneous ductility and minimum level of microstresses, as well as minimum irregularity of mechanical properties along the cross-section of semiproducts, were learned out. 12:00 noon TITANIUM AND ITS ALLOYS-AN OVERVIEW: S.C. Sharma, B.M. Girish, Rathnalcar Kamath, Centre for Manufacturing Research and Technology Utilisation R. V. College of Engineering, Banglore - S60 OS9, India In the recent years, titanium has found wide applications as a structura1 material, mainly due to its high mechanical properties, low specific weight and outstanding corrosion resistance. It can be satisfactorily worked, both hot and cold and has good weldability. Formation of a stable oxide film on its surface helps provide high corrosion resistance. TItanium which has a melting point of I72S'C, has two allotropic forms namely a-titanium and 8-TItanium. The titanium alloys are of special intrest since they have better mechanical properties than pore titanium. One important factor about the titanium alloys is that they have the highest specific strength among all of the structural materials. The present paper reviews the status of titanium and its alloys.

ALUMINA & BAUXITE TECHNOLOGY I: Red Mud

Sponsored by: LMD AlumInum CommIttee

Program Organizer: C.G. Hendriks, Kaiser Aluminum and Chemical Corp.,913 I Interline Drive, Baton Rouge, LA 70809

Wednesday, AM February 7, 1996

Room:A6 Location: Anaheim Convention Center

Session Chairperson: Dr. R.S. Thakur, Regional Research Laboratory, Bhubaneswar, Orissa, India 8:30am RED MUD AND SANDS HANDLING: NEW THOUGHTS ON AN OLD PROBLEM: N.C.R. Oeberg, Consultant, Kirschweg 37, CH - Oberengstringen, Schweix, Austria; A.E. Steinlechner, Maschinenfabrik Andritz Actiengesellschaft, Statteggerstrasse 18, A-804S Graz, Austria The Alumina industry is facing worldwide a growing disposal problem. Red Mud creates increasingly technical, economic and environmental problems. Available solutions are reviewed, and an old idea using a new approach with Pressurized disk or Drum Filters for Red Mud is herewith presented and submitted for consideration as a potential and prospective solution to the problem. 9:00am VALUE ADDED PRODUcrS FROM BAYER PROCESS BYPRODUCTS: D. B. Kirkpatrick, Kaiser Aluminum & Chemical Corp., P. O. Box 337, Gramercy, LA 700S2 Kaiser Aluminum & Chemical Co. impounds the byproducts of alumina production in lakes at the Gramercy, La. facility. Kaiser recognizes that this action cannot be maintained indefinitely. Therefore, the physical benification to recover value from the impounded material and the reduction of stored residue is being investigated. The Gramercy site faces two major challenges, Louisiana has a net negative evaporation rate, meaning the mud cannot be dried by passive means. Second, Kaiser exclusively uses Jamaican North Coast bauxite. This material is very fine which causes problems with the

B88

handleability and dustiness of potential products. The product development strategy addresses environmental and product liability issues as well as meeting customer needs. Engineers at Kaiser have developed techniques to dewater and dry the existing mud lakes to high solid levels. Currently, two products are being tested in solid-waste landfills. Other products are being developed for use outside of these controlled areas. 9:30am IMPROVING THE OPERATION OF IlED MUD PRESSURE FILTERS: John T. Malito, Ph. D., Research Scientist, Alumina Processing Chemicals, Nalco Chemical Company, One Nalco Center, Naperville, IlL 60563-1198 Almost all alumina refineries clarify bauxite digestion slurry using a combination of sedimentation and filtration. Typically, greater than 99% of the mud solids are removed in the settler, while the majority of the remaining solids suspended in the settler overflow is removed by pressure filtration or less commonly, sand filtration. The performance of the pressure filter, with regards to filtration rate, filtrate clarity, and filter cycle time, is strongly influenced by sevef1l11 important factors. In this study, the effects of suspended mud, quantity of filter aid, 8l~d residual flocculent and starch in the filter feed slurry on filtration rate are quantified. This information is useful in optimizing the operation of pressure filters, as well as troubleshooting filtration problems which occur periodically. The use of a reagent specifi(:ally for increasing filtration rate is also discussed. 10:00 am CHARACTERIZATION OF INDIAN RED MUDS AND RECOVERY OF THEIR METAL VALUES: J. Pradhan, S. N. Das, J. Das, S. B. Rao, R. S. Thakur, Regional Research Laboratory, Bhubaneswar 751013, Orissa,India The huge reserve of bauxite, large scale industries and good export potential have made the aluminum industry a promising enterprise in India. At the same time growing concern for environment and raw material conservation has intensified the research and development work on the safe disposal and bulk utilization of red mud, and alkali enriched solid waste generated in alumina manufacturing plants. However, in India, the problem has received scant attention especially from aluminum industries. The present paper outlines the physico-chemical characteristics and mineralogical phases of some Indian red muds. Attempts have been made to study the dissolution behavior of red muds, which of course, depends upon the chemical composition, for the recovery of metal values in mineral acids of different concentrations. 10:30 am BREAK 11:00 am EFFECT OF CALCINATION OF PHYSICO-CHEMICAL PROPERTIES OF RED MUD: M. G. SUjana, R. S. ThaJj:ur, B. C. Acharya, S. N. Das & S. B. Rao, Regional Research Laboratory, Bhubaneswar 751013, Orissa, India Physico-chemical studies were carried out on a sample of red mud collected from NALCO, Orissa after calcination up to 1200'C at an interval of 1000C. The observed change in properties like surface area, particle size, bulk density and mineraI phases etc. can be profitably used while finding an appropriate application of this polluting waste from alumina plants. 11:30 am USE OF RED MUD IN CONSTRUCTION MATERIALS: C. Femadez, P. Garcia, J. M. Ava/a, F. Blanco, J.P. Sancho; Departmento de Ciencia de los Materiales e Ingenieria Metalurgica, Escuela de Minas, Universidad de Oviedo, Indepondencia 13. 33004, Oviedo, Spain Most of the industrial processes generate, at the same time as the prinCipal product by-products or residuals of minor or no economical value. The storage of which represent, from the environmental point of view, problems of different magnitude depending on the type of residual. Solving or minimizing these problems would involve, on one hand the improvements or optimization of the production process, with the purpose of generating the minimum possible amount of such materials for obtaining same profit of them. The residuals utilized in this work are fly ashes from powder stations burning powdered coal and red muds generated itn Bayer Plants during bauxite digestion. The possibility of getting construction materials such as "raised or bare face red-brown bricks" or red-brown poIishable hard tales has been studied.

ALUMINUM ALLOYS-DEFORMATION TMS Gene... ' Abstract Session Wednesday, AM February 7, 1996

Room: Grand 0 Location: Anaheim Marriott Hotel

Session Chairperson:TBA 8:30am METALLURGICAL ASPECTS OF ALUMINIUM FOR CANMAKING: P. D. C. Roqers, CarnaudMetalbox Technology pIc, Downsview Road, Wantage, Oxon, OXI2 9BP, England The properties of alluminium alloYI, used for canmaking have been the object of much research, as the demands of the Canmakers for zero defects and trouble-free running have grown, The main requirement-is consistency, but the puzzle still remains why coils will run one day, but will not the next. Work has shown the importance of controlling earing, mechanical properties, coil shape and thickness, both from coil to coil and within the coil. This means that traditional testinq is inadeguate, and more emphasis has to be placed on process control. It is not difficult to write a specification which describes the main metallurgic,lII and dimensional features of body, end and tabstock but it will not describe why material from one mill will run better, whilst the

1996 TMS ANNUAL MEETING AND EXHIBITION

reverse is true elsewhere. While we do not have clear answers to all problems, our understandinq is growing. 8:50am INFLUENCE OF TOOLING GEOMETRY ON QUALITY OF THE SHEARED SURFACE OF ALUMINUM SHEETS: S. F. Golovashchenko, R. G. Davies, W.S.Stewart, P.O. Box 2053IMD3182 SRL, Dearborn, MI 48121 In order to achieve vehicle weight reduction, stamped aluminum outer panels, such as hoods are being used in automotive production. The present research is concerned with the process of shearing blanks for stamping. Experiments were done to analyse the influence of the gap between cutting edges and the geometry of cutting blades on the quality of the sheared surface. The results are presented in form of the macrostructures of polisbed samples and force displacement curves. To predict the quality of the sheared surface a numerical code based on solid mechanics equations, elastoplastic flow tbeory and cummulative theory of damaged was created. The die was assumed as rigid and friction between sheet and tooling was simulated in accordance to the Coloumb friction law with the restriction that its value be equal to the maximum shear stress. Fracture of material was simulated at every element of the numerical grid wbere the amount of plastic deformation corresponding to the fracture criterion was reached. The fracture criterion included the influence of hydrostatic pressure on material plasticity. 9:10am MODELLING OF SUPERPLASTIC FORMING PROCESSES FOR ALUMlNUM ALLOYS: Yong Nam Kwon, Young-Won Chang, Center for Advanced Aerospace Materials(CAAM) POSlECH, Pohang 790-784, Korea Superplastic fonning of thin sheet into complex components is an important manufacturing process especially in aerospace industry. The main interest in modelling the superplastic forming process is to predict the fonning pressure cycle to maintain optimum formability and the resulting thickness distribution. The ABAQUS finite element code was used for the prediction of process parameter for axisymmetric cup forming and rectangular pan fonning for Supral 100 and 7075 aluminum alloy. To characterize mechanical parameters of these alloys, a series of load relaxation and variable angle cone tests were used. The performance of the finite element model in comparison with the experimental results was diSCUSsed. 9:30 am BREAK 9:40am ANISOTROPY OF DYNAMIC GRAIN GROWTH OF SUPERPLASTIC ALUMlNUM ALLOYS: F. Li, A. K. Ghosh, Department of Materials Science & Engineering, The University of Michigan, Ann Arbor, MI 48109-2136 Grain morphology of several superplastic aluminum sheet materials during uniaxial deformation is studied. The grains are found to undergo elongation with strain along the direction of applied stress, while both static and dynamic grain growth occur leading to an increase in the lateral dimensions of the grains. The growth behavior is examined and kinetics analyzed. Simultaneous static grain growth is also measured and subtracted from the overall grain growth to detennine the extent of net dynamiC grain growth. The anisotropy of the dynamic grain growth in the various directions is analyzed, and corrected with the extent of dislocation creep and that due to mantle creep process such as grain boundary sliding. An attempt is made to develop a microstructural model for dynamic grain growth. 10:00 am STRUCTURAL EVOLUTION AND THE HALL-PETCH RELATIONSHIP IN AN AI-3% Mg ALLOY WITH SUBMICRON GRAIN SIZE: Minoru Furukawa, Zenji Horita, Minoru Nemotu, Dept. of Materials Science & Engineering, Kyushu University, Fukuoka 812, Japan; Ruslan Z. Valiev, Institute for Metals Superplasticity Problems, Ufa 450001, Russia; Terence G. Langdon, Dept. of Materials Science & Mechanical Engineering, University of Southern California, Los Angeles, CA 90089-1453 A submicron grain size was produced in an AI-3% Mg solid solution alloy by introducing intense plastic deformation using either equal-channel angular pressing or torsion straining. These procedures produced grain sizes of the order of -0.2 and -0.09 m, respectively. This paper describes the structural evolution during static annealing at high temperatures and the variation of the microhardness as a function of the grain size. The results demonstrate an increase in microhardness with decreasing grain size down to the smallest grain size examined experimentally (90 nm). 10:20 am MECHANICAL CHARACTERISTICS OF AN AI-Mg-LI-Zr ALLOY SUBJECTED TO EQUAL-CHANNEL ANGULAR PRESSING: Patrick Berbon, Van Ma, Dept. of Materials Science, University of Southern California, Los Angeles, CA 90089-0241; Minoru Furukawa, Zenji Hurita, Minoru Nemoto, Dept. of Materials Science & Engineering, Kyushu University, Fukuoka 812, Japan; Ruslan Z. Valiev, Institute for Metals Superplasticity Problems, Ufa 450001, Russia; Terence G. Langdon, Depts of Materials Science & Mechanical Engineering, University of Southern California, Los Angeles, CA 90089-1453 An AI-5.5% Mg-2.2% Li-O.l2% Zr alloy, with an initial grain size of -400 m, was subjected to equal-channel angular (ECA) pressing to produce an average grain size in the region of -1-2 m. This paper describes the results obtained after tensile testing of this material, both with and without ECA pressing, at temperatures from 298 to 403 K, and the variation in grain size of the ECA pressed alloy after static annealing to temperatures above 800 K.

ALUMINUM ALLOYS FOR PACKAGING II Session III

Sponsered by: SMD Non-Ferrous Metals Committee and Ught Metals Division Program Organizer: Professor James G. Morris, Chemical & Materials Engineering, Director, Light Metals Research Labs., University of Kentucky, Lexington, KY 40506 Wednesday, AM February 6,1996

Room:A2-3 Location: Anaheim Convention Center

Session Chairperson: Gyan Jha, ARea Aluminum, Inc., 2900 First National Tower, P.O. Box 32860, Louisville, KY 40232

8:30 am Invited THE lNFLUENCE OF ALLOYING AND PROCESSlNG ON THE DEEP DRAW· ING PERFORMANCE OF FOOD CAN MATERIALS: S.E. Naess, Hydro Aluminium, N-3080 Holmestrand, Norway The effect of some variations of processing parameters as well as alloy selection on the stretch formability and earing behaviour are illustrated by using industrially fabricated cans. Stretch forming fracture problems may be experienced in the upper part of the comer region in the second step of two-step deep-drawn conical rectangular cans. Of the two alternative allOYS, AAJ005 and AAS050A, the problem was confined to the latter, even though this alloy has the best work hardening properties as sheet metal. However, by tensile testing, a much more pronounced decrease in the work hardening for the 5050A compared to 3005 was measured after the strong predeformation (first draw). Range cracking may also be a problem with rectangular cans. The defect is avoided by using a material with not a too highly developed deformation texture. Round cans produced from AI-Mg-alloys may develop ears parallel to the rolling direction. By adding Mn, the Al-Mg alloys may change to low 45' earing materials. The rivets of easy open ends are produced by biaxial stretchforming, and alloys from the 3000 series exhibit superior properties compared to 5000 series alloys. 9:00am COMPUTER ASSISTED METALLURGY FOR PACKAGlNGALLOYS: C. Sigli, H. Vichery, Pechiney CRV, BP 27,38340 Voreppe, France A software has been developed in order to predict the yield strength of 3000 and 5000 alloys as a function of the process parameters. The main subroutines perform the calculation of: • microstructure in the as-cast, homogenized and hot rolled conditions, • yield strength as a function of microstructure and cold rolling reduction, • yield strength as a function of the baldng treatment. A graphic module enables to readily visualize the stress-strain curve and recovery kinetics of a given alloy. An inverse modeling module (simplex method) has also been implemented in order to perform backward calculation, i.e. to optimize alloy composition and rolling parameters as a function of the required properties. Several examples of application will be demonstrated. 9:20am NUCLEATION AND GROWTH OF RECRYSTALLIZATION IN AN AI-IMn ALLOY DURING THERMOMECHANICAL PROCESSING: P.L Orseni Rossi, C.V.G. Industria Venezolana de Aluminio, C.A., Zona Industrial Matanzas, PO BOX 302, Puerto Ordaz, Venezuela; C.M. Sellars, The University of Sheffield, Department of Engineering Materials, Mappin Street, Sheffield S I 310, UK A procedure based on quantitative optical microscopy has been used to investigate nucleation and growth of recrystallization in hot rolled and annealed AI-IMn (AA3003) as a function of strain (0.4-0.71) and temperature (420' C and 460' C) at a constant mean strain rate of about 5 s-I. Despite the experimental scatter, site saturation seems to be a reasonable approximation for modelling purposes. During recrystallisation, the average growth rate decreases continuously probably due to concurrent recovery in the unrecrystallised fraction although the magnitude of the reduction suggests that a non uniform distribution in stored energy and concurrent precipitation must also play an important role. 9:40am SINGLE STEP PREHEATING OF AA3104 CAN BODY STOCK INGOTS: Scon L Palmer, RavenswoodAiuminumCorporation, P.O. Box 98, Ravenswood, WV 26164; Zhong Li, Light Metals Research Labs., Dept. of Chemical and Materials Engineering, University of Kentucky, 177 Anderson Hall, Lexington, KY 40506 Comparison of a single-step versus two-step (controlled cooling) preheating practice has been investigated for AA3104 can stock ingots. Earing and tensile tests performed on final gauge H19 sheet revealed that earing tendencies were very similar for both processes, while tensile properties were slightly lower for the single-step practice. The difference in tensile properties was correlated to a difference in Mn supersaturation. Manganese precipitated in the form of a fine « I J.UD) second phase dispersion was found to be a more effective hardener than if it was retained in solid solution. Texture development in the single step preheated material followed a different path than the two step process, but with the same final earing result in the HI9 sheet. Manganese dissolution curves were developed for a variety of thermal histories based on resistivity measurements. 10:00 am BREAK

1996 TMS ANNUAL MEETING AND EXHIBITION

B89

10:20 am A TECHNIQUE TO MEASURE INTERMETALLIC SIZE DISTRIBUTION IN ALUMINUM CAN BODY STOCK: X. Wang, Corporate R&D, Reynolds Metal Co., Muscle Shoals, AL 35661; R. G. Kamat, Corporate R&D, Reynolds Metals Co., Richmond, YA 23261 Quantification of the intermetallic constituents in aluminum alloys is important in evaluating the metal performance during fabrication of products. For example, in aluminum D&I can body alloys the intermetallics play an important role in cleaning aluminum build-up on the ironing dies during the can-making process. On the other hand, heyond a certain size these constituents could potentially contribute to the tool wear and/or tear-offs. Hence, optimizing the size distribution of the constituents is very important. Differentiating the intermetallics by conventional etching and metallographic techniques has drawbacks. This technique is time consuming and the results can he subjective due to poor etching response for color differentiation of particles. In this study, a technique for quantifying the intermetallics using a scanning electron microscope coupled with an energy dispersive X-ray spectrometer integrated with an image analyzer is descrihed. The intermetallic phase distribution in aluminum can body alloys is measured in an efficient manner. Particularly, the IX-AI12(Fe,Mn)3Si phase fractions and size distributions were determined. The technique has been successfully used for evaluation of industrial homogenization practices and for intermetallic particle distribution in can body sheets from various producers. 10:40 am BUCKLE STRENGTH LOSS IN 10Z DIAMETER BEVERAGE CAN ENDS: M. S. Nagorka, T. C. Sun, Kaiser Aluminum and Chemical Corp., Center for Technology, 6177 Sunol Blvd., Pleasanton, CA 94566 202 diameter heverage can ends show a significant time-dependent drop in buckle strength. This drop, which can be over 83 kPa (12 psi) after 30 days at 38'C (100'F), is of concern to beverage packagers. This presentation will examine the temperature dependence of the strength loss phenomenon and how different alloys affect the magnitude of the strength loss. The process appears to he thermally activated--activation energies calculated from strength loss data obtained at different temperatures will be compared with those of mechanisms that might be responsible for the strength loss. 11:00 am PERFORMANCE EVALUATION OF CANSTOCK ALLOYS USING THE DRAW BEAD SIMULATOR: Jean Y. Reid, Rajeev G. Kamat, Corporate Research and development, Reynolds Metals Company, P. O. Box 21003, Richmond, YA 23261 The DBS test has been found useful to evaluate lubrication conditions in the drawbead during sbeet metal forming operations. A study was undertaken to evaluate the usefulness of this test in differentiating metallurgical, surface texture, and lubrication conditions for canstock. The test allows the determination of a friction coefficient under plastic deformation conditions. In addition, the frictional force can be separated from the deformation force. In this study, performance in the test is based on friction coefficient and forces generated during the test as well as an evaluation of drawbead and sample galling. The performance data is correlated with canstock mechanical properties, metallurgical structure, surface texture characterization and lubrication mode.

ALUMINUM REDUCTION TECHNOLOGY V: Physical And Numerical Modelling

Sponso~d by: LMD Aluminum Committee Program O'llanizer: Ms FionaJ Stevens, Comalco Research and Technology, Comalco Research Centre, PO Box 316, Thomastown, Victoria 3074, Australia

Wednesday, AM February 1, 1996

Room:A9 Location: Anaheim Convention Center

Session Chairperson: Martin Segatz, VAW Aluminium-Technologie GmbH, Friedrich - Woehler - Str.2, 53111 Bonn, Germany

8:30am A STUDY OF BATH VELOCITY DISTRIBUTION IN A 3-D WATER MODEL: J.J. Chen. Xianchun Shen, B.I. Welch, Chemical & Materials Engineering Department, The University of Auckland, Private Bag 92019, Auckland, New Zealand; M.P. Taylor, New Zealand Aluminium Smelter, Private Bag 90110, Invercargill, New Zealand A probe consisting of a sphere and an embedded thermocouple was used for measuring point velocities in a 3-D air-water model. Prior to testing in the model, the probe was calibrated in a water flow channel. Results measured in a full-scale 3-D model covering various parts of the cell including various positions of the anode/anode ledge slot and the anode/sidewall gap will be reported. How directions may also be detected using this measurement technique. 8:55am A HYDRODYNAMIC MODEL FOR THE BATH FILM BETWEEN METAL AND SIDE LEDGE IN ALUMINUM CELLS: Torstein Hansen, Kemal Nisancioglu, Department of Electrochemistry, Norwegian Institute of Technology, N-1034 Trondheim, Norway; Asbj0m Solheim, SINTEF Materials Technology, N-1034 Trondheim, Norway Existence of a thin bath film between the side ledge and metal has been suggested to affect heat and mass transfer in aluminum cells. The present mathematical model, based on the simultaneous solution of the equations of momentum, energy and mass transfer shows that such a film can be maintained by surface-tension driven flow along the metal-film interface, provided that the bottom sludge is more acidic than the top electrolyte. The film contributes to the resistance against heat flow between the metal and the side ledge. Moreover, the variation in the xs AIF3 and AI203 concentrations along

B90

the film leads to decreasing liquidus temperature and increasing heat flux in the vertical position toward the bottom of the cell. 9:20am A WATER·MODEL STUDY OF THE LEDGE HEAT TRANSFER IN AN ALUMINIUM CELL: John J.J. Chen, Chuck C. Wei, Chemical & Materials Engineering Department, The University of Auckland, Private Bag 92019, Auckland, New Zealand; Anthony Ackland, Comalco Research Cenltre, PO Box 316 Thomastown Victoria 3074, Australia A heat transfer probe was developed for studying the ledge heat transfer in a fullscale 3-D air-water model. Quantitative measurements were conducted to determine the batMedge heat transfer characteristics at various positions and different operating conditions. A similitude analysis was carried out to relate the measured point results to data available in the literature. A suggested range of heat transfer coefficients for the reduction cell is presented. Variation of the heat transfer due to anode bottom inclination and position on the side ledge relative to the anode slot as well as in the vertical direction, were also examined. 9:45am A STUDY OF THE SURFACING BUlIBLES SPEED IN A PHYSICAL MODEL REPRESENTING THE LAYER OF· LIQUID COAL PITCH: Begunov A.I., Yakovleva A.A., Irkutsk State Technical University; Irkutsk, Lermontov Str., 83, Russia Many plants in Russia operate cells with self-baking anodes. Gas and fluid dynamic processes, occurring in the formation of anodes under the electrolysis conditions, influence the volume of cancer causing hydr()(:arbonic wastes and they were not sufficiently investigated. This paper investigates the motion specifics of gas bubbles, penetrating from the baked part of the anode and barbotting through the layer of liquid anodic paste. A model has been used, in which paraffine represents pitch, both having very similar thermal and physical properties. The column of the liquid part of the anode is modelled to its original size and with thc;, similar vertical distribution of temperatures. The speed of gas bubbles, surfacing in !:he model medium, has been studied and the numerical values of the coefficients of resistance for the surfacing of bubbles have been determined. 10:10 am BREAK 10:25 am ELECTRIC FIELDS IN THE ELECTROLYTES OF ALUMINIUM CELLS: A.I. Begunov, I.S. Grinberg, Y.N. Kulkov, Irkutsk State Technical University; Irkutsk, Lermontov Str., 83, Russia By means of physical modelling the 1:lectric fields for an aluminium cell have been constructed for the operating modes with a stable wall freeze and bottom freeze, with the anode effect and with a bottom freeze without a wall one. The bottom freeze, projecting over the metal surface, incn:ases the voltage of the electric field and the current density on the wall plate when the distribution over the electrodes is unchanged. When the bottom freeze length doubles, there arises a redistribution of the current shares between the side and bottom surfaces of the anode, leading to the decrease of the load on the side surface. The electric field at different stages of anode effect development has been investigated. When the anode bottom conducting surface share changes from 0,016 to 0,025, there occurs a sharp redistribution of current from the side surface of the anode to the bottom one. 10:50 am FINITE ELEMENT MODELLING FOR STRUCTURAL ANALYSIS OFCATHODE CASINGS: Mohammad M. Megahed, Dept Mechanical Design & Production, Cairo University, Egypt; Hesham S. Sayed, Structural Engineering Dept, Cairo University, Egypt; Mohammed M. Hasan and Shaher A. Mohammad, Aluminium Co of Egypt, Nagi-Hammadi, Egypt In order to conduct parametric studies of design changes, a desirable prerequisite of the F.E model is the ease by which design changes can be evaluated. In this paper, three levels of F.E models are built for a certain design of cathode casing; viz. two types of simplified models and a detailed F.E model. The flfSt simplified model employs a combination of shell elements and 3-d beam elements connected by rigid-bar elements to simulate welding between shell and sdffeners. The second model employs a larger number of shell elements together witll 3-d truss and/or beam elements to represent beam webs. In the third F.E model, the whole casing structure including all beam flanges and webs is represented by much larger number of thin shell elements. Compared to the all-shell detailed model, simplified models allow quick and reasonably accurate assessment to be made of the effects of various spatial arrangements of different sizes of beam stiffeners on overall cathode stresses and deformations. The simplified models have been successfully used to conduct design improvement studies which resulted in weight reduction by about 1% while improving its structural performance significantly. This result is confirmed by conducting non-linear F.E analysis. 11:15 am ASSESSMENT OF CATHODE SWELLING PRESSURE USING INVERSE NONLINEAR FINITE ELEMENT TECHNIQUE: Hesham S. Sayed, Structural Engineering Dept, Cairo University, Egypt; Mohammad M. Megahed, Dept Mechanical Design & Production, Cairo University, Egypt. Hamad H. Omar and Ismael M. Ismael, Aluminium Co of Egypt, Nagi·Hammadi, Egypt The paper presents a new technique to determine the magnitudes of swelling pressures acting on the sidewalls of cathode casings. A nonlinear finite element model for the cathode structure is developed taking into consideration the material non-linearity effect and influence of temperature on mechanical properties of steel. A system identification technique (inverse problem) is used to estimate the swelling pressure by solving the forward non-linear problem in conjugate with the deformation field measurements of side walls of an operating cathode. The outcome of the inverse problem revealed that the proposed technique is efficient and can capture the essential features of the swelling phenomenon and the cathode structure response using a simple field de-

1996 TMS ANNUAL MEETING AND EXHIBITION

formation measurements. 11:40 am THE FURTHER STUDIES ON CALCULATION METHOD OF MAGNETIC FIELD IN ALUMINIUM REDUCTION CELL: Li GlWhua, Li Dexiang and Li Dianfeng, Department of Nonferrous Metallurgy of Northeastern University, Shengyang, 110006, China On the basis of the theory of using two scalar potentials method for calculation magnetic fields in aluminium reduction cells, we developed an applied software. In the paper, the function blocks of the software are described and some key problems in using this software for calculation magnetic fields in reduction cells are introduced. The software has been debugged step by step and taken cold simulated test in laboratory. The results show that the method and software are very good. In the paper, the result using this software for calculation magnetic field in a certain industry cell is also given.

CARBON TECHNOLOGY III: Anode Quality Control Sponsored by: LMD Aluminum Committee Program Organizer: Lise Castonguay, Alcan International Ltd., Arvida Research and Development Centre, P.O. Box 1250, 1955 Mellon Blvd, Jonqui~re, Qu~bec, Canada, G7S4K8 Wednesday, AM February 7, 1996

Room:A7 Location: Anaheim Convention Center

Session Chairperson: Mr. Berthold HohI, Maschinenfabrik Gustav Eirich, Walldiirner Strae 50, 0-74736 Hardheim, Germany

8:30am A METHOD FOR CHARACTERIZATION OF ANODE PORE STRUCTURE BY IMAGE ANALYSIS: S. Rorvik. H. A. Oye, Institute of Inorganic Chemistry, Norvegian Institute of Technology, University ofTrondheim, N-7034 Trondheim, Norway A fully automatic method for image analysis of porosity in baked carbon anodes has been developed. The method is ~ased on optical microscopy, and is capable of analysing large sample areas (several cm ). It provides various statistics of pores in the range of 10 11m to 10 mm radius. In addition to standard measurements like pore size distribution, some alternative techniques have been used to give information on how the pores are positioned relative to each other. These techniques include determination of the fractal dimension of the pore surface, the degree of pore clustering, and the degree of pore connection. The results from application of the present method point to the importance of using additional ways to characterize pore structure; the pore size distribution alone does not provide sufficient information tu explain variation in anode physical properties such as permeability and electrical resistance. 9:00am REDUCTION OF AHEAD OF SCHEDULE ANODES THROUGH ANODE ROD QUALITY CONTROL: IVon Menard, Fthan~ois Baillargeon, Alcan Smelters and Chemicals Ltd., Grande-Baie Plant, 6000, 6 Avenue, P.O. Box 900, La Baie, Qu~bec, Canada, G7B 4G9; Andre L Proulx, Claude Perron, Alcan International Ltd., Arvida Research & Development Centre, 1955 Mellon Blvd, P.O. Box 1250, Jonqui~re, Qu~bec, Canada, G7S 4K8 One of the major causes of ahead of schedule anodes reported in recent years by P155 potroom operation was related to an inherent weakness in the rod welded joints. The development and implementation of an apparatus and a procedure to measure, detect and reject the faully rods prior to anode rudding, will be presented and discussed. The technology has not only Significantly reduced the number of ahead of schedule anodes, it has also provided useful information concerning other process improvements. 9:20am METHODS OF VERIFYING NET CARBON: Marllou McClung, Ravenswood Aluminum, P.O. Box 98, Ravenswood, WV 26164 Problems currently exist with using net carbon as an industrial standard to gauge smelter performance. First, throughout the industry there are a number of different methods used for determining net carbon. Also, until recently there has not been a viable method to cross check or predict change in net carbon. This inherently leads to differences and most likely inaccuracies when comparing performances of different plants using a net carbon number. Ravenswood uses specific methods when calculating the net carbon balance. The R&D Carbon, Ltd. formula developed by Verner Fischer, et aI., to predict and cross check net carbon based on baked carbon core analysis has also been successfully used. Another method is used, as a cross check, which is based on the raw materials (cokes and pitch) usage as related to the metal produced. The combination of these methods gives a definitive representation of the carbon performance in the reduction cell. This report details the methods Ravenswood Aluminum uses and the information derived from it. 9:40am IN·LINE INSPECTION OF CARBON ANODES FOR USE IN ALUMINUM PRO· DUCTION: Fawzi Emad, Paul Haldemann, Electrical Engineering Department, University of Maryland, College Park, MD 20742; Art Nispel, Eastalco, 5601 ManorWoods Road, Frederik, MD 21701; Larry Logan, Formerly with Eastalco, now at C&K Publications, PO Box 340, Hancock, MD 21750 In this paper, methods of measuring the quality of carbon anodes which are produced and continually consumed in the commercial production of aluminum are dis-

1996 TMS ANNUAL MEETING AND EXHIBITION

cussed. The electrical and mechanical properties of these anodes are very important for the economic operation of the aluminum plant. At least six different ways of checking some of the most important anode characteristics were investigated and those methods selected as the most suitable for testing the quality of the anodes are elaborated. Results from tests made at the plant are included. Future work planned in this area is also discussed. The methods considered include: I. Direct resistivity measurement using core samples; 2. Ultrasonic measurement; 3. Hammer method; 4. Hall effect; 5. 4probe method; 6. Magnetic coils (various arrangements). It will be shown that some of these methods give inaccurate results while others are not practical due to operating conditions andlor production parameters. 10:00 am BREAK 10:20 am STANDARDIZATION OF METHODS FOR TESTING CARBONACEOUS MATERIAL FOR THE PRODUCTION OF ALUMINIUM: Wolfgang SchmidtHatting, International Standardization, Taeusistrasse 48, CH-8630 Rueti (ZH), Switzerland When determining whether material conforms to specifications the mean value of the property under test must be compared to upper and lower limits. In order to do this comparison, standardized test methods including precision (repeatability and reproducibility) figures must be available. Since 1973, many standards prepared by the ISO Subcommittee 7 of the ISO Technical Committee 47 (chemistry) have been published. A large number of methods are being prepared for publication as a standard. The work program of subcommittee 7 will be presented. A close cooperation between ISO and ASTM has been established to aim for universally valid Standards on the American continent and worldwide. The so-called Vienna Agreement makes it possible for ISO standards to become directly CEN (European) Standards. Therefore, active participation in ISO seems to be essential for industrialized countries. 10:40 am TOTAL QUALITY MANAGEMENT: HOLISTIC APPROACH AT THE ALUCHEMIE CARBON ANODE PLANT: Jan Engels, Ton Halsema, Aluchemie, Oude Maasweg 80, 3197 KI Botlek-Rotterdam, The Netherlands In June 1991, Ton Halsema, general manager of Aluchemie Rotterdam, introduced Total Quality Management (TQM) at the Aluchemie carbon anode plant. First a new business plan (\991 - 1996) was made and discussed. Then the Total Quality Program was introduced in February 1992. Since 1992, the implementation ofTQM atAluchemie has by its holistic approach radically changed the bottom line results, as shown by some examples of 1994 results against those of 1991: 15% less labour hours per ton produced,60% less Nett Working Capital, 40% less absenteeism due to illness. Most of the results have been reached by systematic approach. The crux of the holistic approach is the recognition of the existing links between «the company's strategy, its structure and culture». Progress in achieving TQM is measured against the criteria of the European Quality Award. These criteria, which build up TQM, help to structure the business, to measure it and to manage it for better results.

CAST SHOP TECHNOLOGY V: Modelling, Solidification, and Alloy Development

Sponsored by: LMD Aluminum Committee Program Organizer: Julian V. Copenhaver, Technical/Quality Manager, NSA A Division of Southwire, POBox 500, Hawesville, KY 42348

Wednesday, AM February 7, 1996

Room: AIO Location: Anaheim Convention Center

Session Chairperson: Dr. Madhu Nilmani, University of Melbourne, Department of Chemical Engineering, Parkville, Victoria 3052, Australia 8:30am MODELING THE FEEDING OF ALUMINUM ALLOY CASTINGS: Alauddin Ahmed, Dr. D. Apelian, Worcester Polytechnic Institute, Aluminum Casting Research Laboratory, 100 Institute Road, Worcester, MA 01609; A. K. Dahle, Dr. Lars Amberg, The Norwegian Institute of Technology, Department of Metallurgy, N-7034, Trondheim, Norway Aluminum alloy castings, in general solidify with the growth of equiaxed dendrites followed by eutectic precipitation. Dendrite growth and eutectic precipitation are influenced by a complex combination of macroscopic transport phenomena and microscopic solidification kinetics. The feeding process, i.e. the shrinkage driven material transport and solidification mechanism are critically dependent on the local solid fraction. Consequently, it is necessary and possible to divide the mushy zone into several distinct regions in order to successfully model the solidification and feeding process. These regions have different mechanical and solidification characteristics and are governed by different transport mechanism and solidification kinetics. A mathematical model for solidification and feeding is developed coupling the macroscopic transpon process and microscopic solidification kinetics taking into account the various regions of the mushy zone. A finite element computer program is developed to solve the model and predict several important solidification variables such as dendritic coherency, grain size distribution, feeding and shrinkage. The phenomenology of feeding castings and pertinent results of the model will be reviewed. 8:50am MODELLING OF ELECTROMAGNETIC PHENOMENA IN A MASSIVELY PARALLEL PROCESSOR COMPUTING ENVIRONMENT: Dan P. Cook, Reynolds Metals Company, Corporate Research & Development, 4th and Canal Streets,

B91

Richmond. VA 23261; Dr. D. MacKay. Intel Corporation. Super Computer Divisions. Oalc Ridge National Lab. Building 4500N. Room 1114. Oalc Ridge. TN 37831-6203 A computer model which calculates the electromagnetic phenomena in electromagnetic casting has been ported to the Intel Paragon. a distributed memory. massively parallel processor (MPP) supercomputer. A parametric study of the mesh density requirements was carried out over a wide range of current frequencies and electrical conductivities. These results were validated by several analytic solutions of tbe vector potential form of the induction equation. Finally. a study was conducted to determine the optimum number of processors to use for the computations for any given problem size. 9:10am INSTABILITY OF A LIQUID METAL SURFACE IN AN ELECTROMAGNETIC FIELD AND RELEVANCE TO EMC: Ryuichi Kageyama. Graduate Student, Dr. J. W. Evans. Professor of Metallurgy. Department of Materials Science and Mineral Engineering. University of California. 577 Evans Hall # 1760. Berkeley. CA 94720-1760 In electromagnetic casting (EMC) the surface of the molten metal. at the solidification front around the periphery of the melt pool. is not confined by a solid mold (as in. say. direct chill casting) but is free to move. Consequently. distorbances of the melt surface are reflected in defects (waviness) in the solid ingot. The present paper examines the dynamics of a liquid metal surface in an electromagnetic field comparable to that of EMC. Numerical calculations of the flow of metal and motion of the melt surface have been accompanied by laboratory experiments in which a laser vibrometer has been used to measure the oscillations of the free surface of a mercury pool. Surface oscillations grow with increasing electromagnetic field strength in both the computations and the experiment. probably originating from the torbulent flow in the melt. The implications for EMC are discussed. 9:30am THERMO-MECHANICAL MODELLING OF THE 3C ROLL CASTING OF ALLOYS: Phillippe Jarry. Senior Scientist. Denis Toitot. Pierre-Yves Menet. Pechiney Centre de Recherches de Voreppe. BP 27. F-38340 Voreppe. France In the roll casting process. coupling between the therrnokinetics of solidification and mechanics of deformation is governed by the rolling operation. This is the basic prinCiple of a numerical simulation of the 3C process based on the two-dimensional large deformation dedicated FORGE2 program. Heat transfer coefficients are calculated from the contact pressure. which is itself calculated from the temperature field; this causal loop governs ~ convergence of the computation. The model provides both global results in terms of rolling force. exit temperature. etc ...• and local mappings of relevant variables within the roll gap; it allows predictive computation of different casting configurations (including down gauging) as well as it helps the metallurgical understanding of microstructure and defect formation. 9:50am SIMULATING AND OPTIMIZING AN AI-SIC COMPOSITE MIXING TANK: D. Kocaefe. Rung T. Bui. Departement des sciences appliqu~s. Universitl! du Qu~bec aChicoutimi. 555. boulevard de I'Universitl!. Chicoutirni. ~bec. Canada G7H 2B 1; R. Provencher. T. Bourgeois. Alcan International Limited. P. O. Box 1250. Jonqui~re. Qu~bec. Canada G7H 4K8 Quality of SiC reinforced aluminum composites depends on the uniformity of the particles in the matrix which in tum is defined by the good mixing conditions in the holding furnace. A 3D. two-phase model is developed to simulate the mixing of AI-SiC mixtures in a holding tank which has a shape of truncated cone and equipped with an off-center mixer. The objective is to evaluate the mixing process using various operational and design parameters and to determine the conditions that will give the best mixture in terms of homogeneity. Due to complex geometry. the commercial code FLOW3DTtd which has a multi-block option was used for this model. An optimization study is carried out by changing the various design and operational parameters such as the shape and dimensions of the mixing tank. impeller dimension. rotational speed. cone angle and melt height. It was observed that the mixture becomes more uniform with increasing impeller diameter. rotational speed. and decreasing melt height and cone angles. However. a certain increase in angle might be preferable. Even though this results in a slight decrease in homogeneity of the system, there is a significant augmentation in heat transfer efficiency due to the increase (as much as 30%) in the top surface area of the cone.

The effect of solidification direction on buoyancy-induced convection in a binary alloy and the resulting concentration mac:ro segregation are numerically simulated in a three-dimensional rectangular cavity. Th,~ system of equations governing the transport of momentum. heat. and species is based on the modified continuum transport equation and is solved using the SIMPLE computational scheme with a multi-grid method to solve the pressure correction equation. The freezing of AI7%Si is simulated and results are discussed. The temperature, concenb-ation. and velocity fields during solidification and. consequently. the distribution of solute in the solid. are found to be strongly affected by the solidification direction.

11:00 am INVESTIGATION OF RAPID MOLD HEATING AND COOLING TECHNIQUES TO OPTIMIZE PART QUAUTY AND CYCLE THROUGHPUT: MaryJo Mallhews. Makhlouf Makhlouf. Aluminum Casting Research Laboratory. Worcester Polytechnic Institute. 100 Institute Road. Worcester. MA 01609-2280 A continuous tilt pour casting process at an aluminum foundry was investigated. modeled. and simulated with a commercial solidification software package. In order to reduce the cycle time and optimize the ,quality of the chosen casting. several cooling channels were designed and modeled using the software. Due to enhanced temperature gradients in the cycle. an improvement in the quality of the casting is expected. One channel design that significantly reduced the cycle time. while maintaining or improving the quality of the part. was run under real manufacturing conditions. The mold was machined. and any other process change:s were made to match the conditions stated in the simulation. The casting produced under the forced cooling conditions was then checked for quality by x-ray inspection. 11:20 am DEVELOPMENT AND COMMERCIALIZATION OF A 3000 SERIES BODY STOCK ALLOY USING THE LAUNER BWCK CASTER: Bill Newlon. Manager. Technology Development. GoldelB Aluminum Company. 1600 Jackson Street. Golden. CO 80401 Paper will describe the development and commercialization of a 3000 series body stock using the Launer block caster located in San Antonio. Starting with a 5000 series alloy. 5349. and describing its entry into the can-malcing industry. and its differences from mainstream 310413004 alloys. The decision to change the alloy to one which was closer to the regular 300413104 alloys we use; the evolution of this alloy and its subsequent acceptance by the can malcers. 11:40 am MAGSIMAL-59, AN AIMgMnSI-TYPE SQUEEZE-CASTING ALLOY DESIGNED FOR TEMPER F: Ulrich Hielscher. Horst Sternau. Dr. Hubert Koch. Alois J. Franke. Aluminium Rheinfelden GmbH. P. O. Box 11 40. D-79601 Rheinfelden. Germany To get high mechanical properties using standard squeeze casting alloys (for example. A356). it is indispensable to malce a beat treatment. That means solution heat treatment and quenching and artificially aging. For this reason. we were challenged to develop an alloy that provides sophisticated mechanical properties without any heat treatment. Compared to A356 T6. value s in brackets. the new alloy has yield strength > 21 ksi [>32 ksi). tensile strengths> 42. ksi [43 ksi). and elongation> 15% [10%) in temper F. Fatigue strength (1'=-1. high frequency pulsation test) is > ± 16 ksa [13.5). To meet these properties. a casting procesl! with high solidification velocity like squeeze casting or high pressure die casting is necessary. Magsimal-59 is of the AlMgSiMntype. The microstructure consists of -AI and a very fine dispersed ternary eutectic. The microstructure and the influence of cooling rate on the mechanical properties will be discussed including some examples of ,~astings.

DEFORMATION I: F~TIGUE TMS General Abstract SessIon Wednesday. AM February 7. 1996

AND FRACTURE:

Room: Orange County 1 Loclltion: Anaheim Marriott Hotel

Session Chairperson: TBA

10:10 am BREAK 10:20 am SOLUTE REDISTRIBUTION DURING STEADY STATE DIRECTIONAL SOLlDiFICATION OF A TERNARY ALUMINUM-BASE ALLOY: M. Chen. Dr. T. Z. Kallamis. Dr. H. D. Brody. Department of Metallurgy. University of Connecticut. 97 No. Eagleville Road. U-136. Storrs. cr 06269-3136 Several dendritic monocrystals of AI-3.75%Cu-1.5%Mn were direc~onally solidified at various growth rates under a constant thermal gradient of 6.8xlO Kim. Solidification was interrupted at a given time by quenching the remaining liquid. Copper distribution profiles within the dendrites and the surrounding liquid were determined in transverse sections corresponding to various temperatures at the moment of quench and were compared with computed ptedictions of a solute tedistribution model. Copper concentration monotonically increased with increasing fraction solid and decreasing temperature. Manganese concentration ftrst increased with increasing fraction solid. reached a plateau and subsequently decreased. Below the ternary eutectic temperature copper concentration at a given fraction solid decreased with increasing growth rate. Measured solute concentrations agree reasonably well with computed predictions. 10:40 am A STUDY OF THE EFFECT OF COOLING DIRECTION ON THE MACRO SEGREGATION OF A BINARY METALALLOYDURING SOLIDIFICATION: Kianxi Jin. Dr. Will Schreiber. The University of Alabama. DepartrnentofMechanical Engineering. P. O. Box 870276. Thscaloosa. AL 35487-0276

B92

8:30am AN INTERNAL VARIABLE APPROACH TO SUPERPLASTIC DEFORMATION OF Pb-Sn EUTECTIC ALLOY: Tae Kwon Ha. Young Won Chang. Center for Advanced Aerospace Materials. Pohang University of Science and Technology. Pohang. 790-784. Korea A series of load relaxation and tensile tests were conducted on superplastic Pb-62wt%Sn eutectic alloy at room temperature. The internal variable theory of structural superplasticity proposed recently was employed to analyze the results. The flow curves obtained from load relaxation tests were shown to consist of the contributions from boundary sliding. such as grain boundary sliding (GBS) and phase boundary sliding (PBS). and the accommodating plastic deformation. A new attempt to deduce the optimum strain rate. one of the most important parameters of superplastic (SPF) forming process. has been also made. which malces it possible to quantitatively introduce the nature of boundary sliding into the SPF process. 8:50am SUPERPLASTIC BEHAVIOR OF STAINLESS STEEL NITRONIC 19D: C. K. Syn. D. R. Lesuer. Lawrence Livermore National Lab .• P. O. Box 808. Livermore. CA 94550; O. D. Sherby. Dept. of MSE. Stanford University. Stanford. CA 94305 Superplastic forming of stainless s~eels can be an economically viable manufacturing process for complicated structures for applications in aggressive environment such as automotive exhaust parts. A commercial stainless steel known as Nitronic 19D with a duplex microstructure and a nominal chemistry of 21Cr-4Mn-2Ni-0.3C-0.2N-Fe was

1996 TMS ANNUAL MEETING AND EXHIBITION

repeatedly hot-rolled to develop a micro-structure containing fine elongated grains. Tensile tests were performed at various strain rates and temperatures to examine the plastic deformation behavior at elevated temperature. A tensile ductility in excess of 250% at 950°C was obtained at constant true strain rate of 3x JO-4sec-l. Cavitation was observed at ferrite-austenite grain boundaries. During tensile testing, the elongated grains were broken up into randomly oriented equiaxed grain structure with high angle grain boundaries. A detailed description of microstructural evolution and plastic deformation behavior will be presented. Work performed under the auspices of the U.S. Dept. of Energy by the Lawrence Livermore National Laboratory under contract No. W-7405-ENG-48.

exhibited similar Cbarpy impact properties. The small heats containing higher concentrations of Cr and TI, in a fully recrystallized condition exhibited a DBTT at around -100°C, whereas the V-3Cr3TI alloy failed by pure ductile shear at liquid nitrogen temperatures. Research sponsored by the Office of Fusion Energy, U.S. Dept. of Energy, under contract DE-AC05-840R21400 with Martin Marietta Energy Systems, Inc.

9:10am EXPERIMENTAL AND NUMERICAL INVESTIGATION OF FRACTURE IN DOUBLE EDGE NOTCHED STEEL PLATES: R. Knockaert, I. Doghri, Universite6 Catholique de Louvain, CESAME, Bitiment Euler, 4-6 Avenue G. LeIlla1"tre, B1348 Louvain-La-Neuve, Belgium Double edge notched (DENT) steel plates were pulled until complete rupture and several experimental observations were made (including profilometry and electronic microscope measurements). The essential work to fracture (EWF) model was found to be well verified; indeed the specific work to ropture, the maximum load and the maximum displacement are linear functions of the ligament length. Numerical simulations of the experiments taking into account geometric and material non-linearities (large deformation elasto-plasticity) were performed using the finite element method (FEM). Excellent agreement was found between experimental observations and numerical computations of load-displacement curves, thickness reduction and stored plastic work. The FEM also allowed to check the shape of the plastic zone against the one assumed in the EWF model. In order to determine the size of the plastic zone surrounding the crack tip, several models have been tested, ranging from simple criteria to a fully coupled approach based on continuum damage mechanics and using the Lemaitre-Chaboche ductile damage model.

Sponsored by: EMPMD Thin Film & Interfaces CommIttee Program Organizers: Professor Krishna Rajan, Materials Science and Engineering Department, Rensselaer Polytechnic Institute, Troy, NY 12180-3590; Dr. Michael Fury, Rodel Inc., 451 Bellvue Road, Newark, DE 19713

9:30 am BREAK 9:40am FATIGUE PROPERTIES OF GRAPHITE-EPOXY COMPOSITEAFTERPLASTIC MEDIA BLASTING PAINT REMOVAL: M. Shoo, A. Inchekel, J. E. Talia, Wichita State University, Department of Mechanical Engineering, Wichita, KS 67208 An experimental study was conducted in order to investigate the effects of plastic media blasting (PMB) on the fatigue properties of woven graphite-epoxy composite material. Plastic media blasting was used to remove paint from the samples. A procedure consisting of fatigue cycling, blasting, and repainting (FCBP) was devised. Specimens subjected to one to four cycles of FCBP were tested for their fatigue life and residual strength. The results showed that plastic media blasting (PMB) reduced the fatigue life of the composite. Furthermore, FCBP cycles reduced significantly the fatigue life of samples and this effect is more detrimental as the number of FCBP increases. 10:00 am STRESS MAPPING AT FATIGUE FRACTURE INTERFACES: David/.Dingley, H. H. Wills Pbysics Laboratory, University of Bristol, Bristol BS8 1TI.. UK; David P. Field, TSL, Inc., 226 W. 2230 North Provo, UT 84604 Through the use of the new technique of Orientation Imaging Microscopy, OIM, it is possible to construct micrographs depicting the changing cryrstallography over the surface of single crystal or polycrystalline bulk materials. This technique has been used in the investigation of fatigued nickel base alloy single crystals. The specimens were cut and polished normal to the fracture surface and examined in an SEM fitted with the OIM system. Orientation image micrographs were obtained over three sections of the fatigue crack and over an area sufficiently far from the fracture surface as to represent undamaged material. Two types of micrograph were produced. The first showed the variation of crystal damage as measured by the contrast observed in diffraction patterns obtained at each point. The second contoured regions across which there were small lattice rotations of 0.5- or more. These contours depict the locations of dislocation clusters. It was observed that the fatigue damage extended only microns from the fracture surface. 10:l0am EFFECT OF SCRATCH AND SHOT PEENING ON FATIGUE CRACK GROWTlI OF ALlO14-T3: M. Shoo, J. E. ThIia, A. Inchekel, Wichita State University, Department of Mechanical Engineering, Wichita, KS 67208 The effects of scratch and shot peening on fatigue crack growth in AI 2024-T3 alloy specimens have been investigated. The results indicated that the scratch accelerated the crack growth. However, shot peening reduced the fatigue crack growth in scratched specimens to those of as-received specimens at low K. As K increased, crack growth rates gradually merged with those of scratched specimens. The specimens with the across-width scratch showed small crack growth behavior. 10:40 am INFLUENCE OF Cr-TI CONTENT AND MICROSTRUCTURE ON IMPACT BEHAVIOR OFVANADIUMALLOYS:A. N, Gubbi, A, F. Rowcliffe, D.J.Alexander, M. L. Grossbeck, Metals and Ceramics Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831 A vanadium alloy containing 4 wt. % Cr and 4 wt. % TI has been identified as a prOmising candidate structural alloy for liquid metal cooled blankets in advanced fusion systems. Cbarpy impact testing was carried out on a large heat (-500-kg melt) of Y-4Cr-4TI and small heats (-15-kg melt each) of compositional variants with Cr and TI contents ranging from 3 to 6 wt. % which were used for optimization studies. One-third-size Charpy impact specimens were employed for impact testing. In a fully recrystallized condition with a grain size of -16 m the large heat of Y-4Cr-4TI exhibited a high level of resistance to cleavage failure with a DBTT at --190°C. The small (15 kg) heat of Y-4Cr4TI heat treated to produce the same microstructural condition

1996 TMS ANNUAL MEETING AND EXHIBITION

EMERGING TECHNOLOGIES SYMPOSIUM: Engineering Science Of Chem-Mechanlcal Planarization:

Wednesday, AM February 7,1996

Room: Grand K Location: Anaheim Marriott Hotel

Session Chairperson: Dr. Michael Fury, Rodel Inc., 451 Bellvue Road, Newark, DE 19713

8:30am A MATERIALS SCIENCE PERSPECTIVE OF CHEM-MECHANICAL PLANARIZATION: Krishna Rajan, Department of Materials Science and Engineering, Rensselaer Polytechnic Institute, Troy, NY 12180-3590 The ever decreasing dimensions in multi-level interconnects dictates the need to incorporate the microstructural and atumistic level issues in understanding the mechanisms controlling chem-mechanical planarization technologies. In this presentation, we present an overview of where the "science-technology gap" exists in this emelging field and how fundamental materials science research can playa pivotal role in enhancing and controlling planarlzation technology in multi-level metallization. 9:00 am Invited OXIDE·METALINTERACTIONSAND THEIR EFFECTS ON MECHANICAL POLISHING: R. Gibala, Department of Materials Science and Engineering, University of Michigan, Ann Arbor, MI 48109-2136 In the chemical/mechanical polishing (CMP) for planarization of insulator and conductor layers in multi-level metallization interconnect structures, many of the resulting oxide-metal interface microstructures are subject to mechanical stresses that can have large effects on device performance and reliability. This paper examines mechanisms by which surface oxides can alter the mechanical properties of refractory metal and intermetallic alloy substrates. Experiments demonstrate, for example, that some oxidecoated metals and intermetallics (e.g., Nb, Ta, Mo, W, NiAI, FeAI, MoSi2) can exhibit enhanced plasticity and reduced flow stresses relative to the uncoated material. Such effects depend significantly on oxide properties, interface stresses, surface topography of the substrate and dislocation mobilities and source densities. As the dislocation mean free path increases relative to the material dimensions, as in CMP-processed systems, these effects can become especially important. 9:30 am Invited MATERIALS SCIENCE RESEARCH OPPORTUNmES IN CHEM-MECHANI· CAL PLANARIZATION (CMP): Mukesh Desai, Chris Karlsrud, Inki Kim, Jim Schlueter, SpeedFam Corp., 7406 W. Detroit, Chandler, AZ 85226 CMP is an old technology with practical applications in making lenses, silicon wafers, bard disks (storage media), etc. Materials used in fabrication of CMP equipment and consumables are as old as these applications. Recent application of CMP in IC fabrication has brought a new level of attention from researchers. Application of CMP in IC fabrication requires a higher level of performance than traditional CMP applications. The challenge of meeting these needs will require a concurrent improvement in materials used in fabrication of equipment and consumables used in CMP along with reduction of their cost. This paper will attempt to review the current understanding of the technology and point out material properties issues which are dominating current and future performance needs of CMP. 10:00 am BREAK 10:l0am SURFACE MEASUREMENTS OF CMP PADS WITHIN A SINGLE POLISH CYCLE: D.L Hetherington, S.M. Kempka, M.T. Dugger, Sandia National Laboratories, Albuquerque, NM 87185; DJ. Stein, University of New Mexico, Chemical and Nuclear Engineering Department, Albuquerque, NM 87131 Time variation in pad topology during Chemical-Mechanical PoliShing (CMP) is examined experimentally. Topology measurements, in situ, are made using a WYKO non-contact optical surface profilometer. A prinCipal undesirable feature of CMP is that the removal rate decreases with time during a single polish cycle. Polishing increases the surface area of contact between the wafer and pad, resulting in a reduction of the pressure each asperity applies to the surface. We characterize the effects of pressure and rotation speeds on the surface topology of an ICl000 pad as a function of time within a polishing cycle. Numerical volume and surface area data are determined from 300 11m x 300 IlJI1 imaged surface profiles. We discuss the implications of the change in these values as a function of polishing time. The results of this study provide a partial mechanistic basis for the development of planarization models to allow for optimization of CMP processes. 10:50 am InvIted CHARACTERISTICS OF SCRATCHING ON BEOLCMP: Thom Sandwick, Bill Cote, William Landers, IBM Microelectronics Division, Hopewell Junction, NY 12533

B93

Chemical Mechanical Polishing (CMP) of oxides is commonly used for BEOL planarization of semiconductors. There is also a growing interest in the polishing of metals to form damascene studs and wiring. This type of process has a unique set of problems and requirements that have not traditionally been addressed. A key problem of damascene processing is CMP scratching, which leaves a unique defect signature. Since damascene patterning leaves metal in the recessed areas by design, and a scratch is a random recessed area, the scratch can cause electrical shorts at subsequent damascene or RIE defined line levels. Scratching may be caused by the pad, slurry tool, environment of even possibly the wafers. This paper will review the characteristics and possible mechanisms to understand the effects of CMP and scratching. The effects of many of these variables on scratching will also be discussed. 11:20 am Invited EFFECT OF ELECTROCHEMICAL PARAMETERS ON SURFACE MORPHOLOGY: R. Ricker, National Institute of Standards and TechnolOgy, Gaithersburg, M020899 Abstract not available.

EXPERIMENTAL METHODS FOR MICROGRAVITY MATERIALS SCIENCE: SESSION IV

Sponsored by: MSD Thermodynamic and Phase Equllbrla Committee, NASA Mlcrogravlty Sciences Program Organizers: Dr. RobertA.Schiffman, R. S. Research Inc., Barton, VT05822; Professor Carlo Patuelli, Universita di Bologna, 1-40126 Bologna, Italy Wednesday, AM February 7, 1996

Room: Orange County 2 Location: Anaheim Marriott Hotel

Session Chairperson: J. Smith, Jr., The Department of Chemical and Materials Engineering, The University of Alabama in Huntsville, Huntsville, Alabama 35899

8:30am AMPOULE DESIGN AND TESTING FOR MICROGRAVITY EXPERIMENTATION ON COUPLED GROWTH IN HYPERMONOTECTICS: J.B. Andrews, J.S. O'Dell, A.B. Cheney, Y. Arikawa, L.B. Hayes, Department of Materials Science and Engineering, University of Alabama at Birmingham, Birmingham, AL 35294-4461 Microgravity experimentation on immiscible alloys presents several unique problems concerning ampoule design. This presentation will cover the steps being taken in an attempt to minimize some of these difficulties for the AI-In alloys that will be directionally solidified aboard the Life and Microgravity Space Lab during the Summer of 1996. As an example, caution must be used when selecting the ampoule material in order to avoid undesirable wetting behavior between the ampoule and the alloy. In addition, a means must be provided to accommodate the thermal contraction and solidification shrinkage during processing in order to avoid free surface formation on the melt. Steps must also be taken to control thermal end effects in order to maintain a constant growth rate during processing. The final design results in a relatively complicated ampoule assembly containing insulating segments, dummy samples, moving pistons and a high temperature spring assembly. The details of this design and the results of ground based testing on the Advanced Gradient Heating Facility engineering model in preparation for flight experimentation will be discussed. 8:50am MASS-TRANSFER CHARACTERISTICS OF A MULTISTAGE BIPHASIC AQUEOUS PARTITIONING SYSTEM DESIGNED FOR USE IN LOW GRAVITY: Mark Deuser, John Vellinger, John Weber, Space Hardware Optimization Technology (SHOT), Inc., Floyd Knobs, IN 47119; Martin Guinn, Paul Todd, University of Colorado, Boulder, CO 80309 Extraction of solutes from one liquid phase to another is a common method of solute purification. A 22-stage biphasic aqueous extractor has been designed and tested for applications in low gravity and as a prototype for ground-based commercial extractions. This device operates on the principle of sliding half-chambers and will test concepts of gravity dependent mass transfer in the purification of particulate products. A mass-halance model predicts the potential effects of low-gravity operations on mixing and demixing of the immiscible fluids used in the extraction process. A magneticallydriven, programmable mixing system has been designed to operate in each of the 22 chambers to assure effective mass transfer between the phases. This mixing system was successfully tested for comparison with the mass-transfer model. The overall system designed for space-flight applications includes interchangeable "cassettes" consisting of 22-stage disks preloaded with immiscible aqueous solutions and samples to be separated. The first space-flight applications of this multistage extractor will occur on Space Shuttle flight STS-77 in the SPACEHAB-04 payload module. 9:10am DESIGN AND DEVELOPMENT OF ASAMPLFJAMPOULE CARTRIDGE ASSEMBLYTOSUPPORTHIGH-TEMPERATUREMICROGRAVITYCRYSTAL GROWTH EXPERIMENTS: Michael R. Fiske, John O. Stark, Dean L. Heston, Teledyne Brown Engineering, 300 Sparkman Drive, MS 128, Huntsville, AL 35807 Microgravity processing of semiconductor materials has long been of interest to the science community. The microgravity environment of space holds promise for reduction of buoyancy-driven convection and sedimentation and associated structural defects. In addition, there has been significant interest in quantitatively assessing the effects of Shuttle thruster firings, crew movements, and other payload operations on the morphology of a growing crystal front. Both of these research areas can be addressed through on-orbit Current Pulse Interface Demarcation (Peltier pulsing) during

B94

the growth process to identify the shape and location of the solid-liquid interface. This paper address key issues related to the development of a Sample!Ampoule Cartridge Assembly (SACA) which incorporates a Current Pulse Interface Demarcation (CPID) capability for microgravity crystal growth experiments. These experiments have been processed in the Crystal Growth Furnace (CGp) on the second United States Microgravity Laboratory (USML-2) in late 1995 and are planned for use in CGF as part of the Space Station Furnace Facility (SSFF). Specifically, issues such a materials compatibility, cartridge fabrication, thermocouple design, electrical and mechanical interfaces, ampoule failure detection sensor design, and assembly techniques will be discussed. Results of analysis of the USML-2 SACA performance and development of SACAs for planned flights of the CGF in the Space Station Furnace Facility (SSFF) will be presented, including design approaches allowing operation at temperatures up to 1600·C. 9:30am SPACE-DRUMS(TM) - CONTAINE1[{LESS PROCESSING OF MATERIALS USING ULTRASONICS: EdwarrlA. Sioot, Barry L. Wetter, Canadian Space Agency, 6767 rue de I' Aeroport, St-Hubert, Quetec, Canada BY 8Y9 Space-ORUMS(TM) is a new Canadian technology capable of processing materials without a container in both gravity and microgravity environments using ultrasonics. The prototype system uses twenty eql~ivalently poSitioned ultrasonic beams to levitate or position large samples from 5mm to 6cm in diameter inside a closed dodecahedron container. This new technological approach relies on direct acoustic radiation forces to position the sample dynamically with real-time adjustments in response to the behaviour of the material undergoing processing. A high level of sample stability is provided by the multiplicity of energy enveloping the sample specimen. Precision control of the sample is maintained by the response characteristics of the feedback mechanisms built into a novel multi procc~ssor/controller. Space-DRUMS(TM) can containeriessly process a variety of matl~rials with temperatures as high as 3500·C. 9:50 am BREAK 10:10 am RECENT RESULTS OF THE GERMAN MICROGRAVITYPROGRAM: Rainer Kuhl, Horst Binnenbruck, German Space Agency, OARA, 53227 Bonn, P'O.B. 3003 64

Main results of fundsmental and application-oriented research topics in the field Materials and Physical Sciences using m~crogravity conditions are outlined. This refers to precise measurements of thermophysical properties of metallic melts and nonmetallic liquids, crystal growth of semiconductor materials with external magnetic forces, the quantitative confrrmation of the divergence of the specific heat of fluids near the critical point, a deeper understanding of phase separation and agglomeration processes in immiscible alloys which resulted in a new ground-based pilot production of bearing metals, and others. The results were gained by using a broad spectrum of microgravity flight opportunities such as Drop Tower Bremen, Sounding rockets TEXUS, unmanned FOTON satellites, Space Shuttle carried Spacelab and the MIR space station. 10:30 am UTILIZATION OF A MICROGRA'VITY ENVIRONMENT TO FABRICATE UNIFORMLY COMPOSITE SPHERES FROM IMMISCIBLE ALLOY SYSTEMS: R. N. Grugel, Universities Space Research Association, Marshall Space Flight Center, MS-ES75, Huntsville, AL 35812; L.N. Brush, University of Washington, Department of Materials Science and En~ineering, Seattle, WA 98195 Uniform microstructures can evolve during solidification of monotectic composition alloys if the liquid reaction product (L-I1) wets the primary solid (S-I). However, highly segregated macrostructures tend to develop during processing of hyperrnonotectic alloys. Here the interfacial surface tension can not accommodate the excess LII which then, because of Earth's gravity, runs down and hangs off the solid as is demonstrated by melting silver foil (L-I1) over a suspended nickel sphere (S-I). A 1-0 numerical solidification model was developed to track multiple interfaces of the bi-material drop configuration. Experimental processing, parameters dictated by the model were utilized to promote and insure a uniform silver thickness on precast nickel spheres within the limited microgravity time provided by Marshall Space Flight Center's 105 meter drop tube. Results limitations, and possible applications are presented and discussed. 10:50 am A HYBRID GLASS CRYSTAL OBSERVATION CHAMBER FOR MICROGRAVITY ENVIRONMENTS: Deall S. Schrage; AOF, Inc., Lewis Research Center Group, Brook Park, OH 44142 The Space Shuttle experiment, Isothermal Dendritic Growth Experiment (lOGE) will be flown on United States Micro-Gravity Payload (USMP)-4, in 1997, to evaluate the morphology of dendrite growth of a carboxilic acid, Pivalic acid (PVA). Earlier micro-gravity dendritic test programs, using Succinonitrile (SCN) as the working fluid, were successful in achieving nominal maximum supercoolings of 2.0 C. The dendrite growth with SCN were successfully pc~rformed within the confines of a crystal observation chamber (COC), ruggedized for space flight, constructed primarily of stainless steel. However, when these same chambers are filled/tested with PYA, the maximum supercooling that can be achieved, prior to auto-nucleation, decreases to 0.1 C. It is postulated that the stainless steel chamber wall material catalyzes nucleation. As this supercooling is insufficient to capture the desired dendritic growth phenomenon, a development program has been initiated to design and build a COC that will achieve, functionally the same operation as the conventional stainless steel chambers, yet harbor the PYA within the confines of inert glass material interfaces. This paper presents a detailed discussion of the development of this hybrid glass COC, presents high-level conceptual drawings and a detailed diiICussion of the operational characteristics. 11:10 am LIQUID-METAL PROCESSING - ZARM: Reginald Smith, Mark Tunnicliffe, Mark Gallemeault, Evangelos Vekris, Queen's University, Kingston, Canada K7L 3N6 Queen's University is collaborating with the University of Bremen to investigate the freezing of liquid metals during free fall in the "ZARM" drop tower. Of particular

1996 TMS ANNUAL MEETING AND EXHIBITION

interest is the manner in which the undercooling at which nucleation occurs influences the resulting microstructure. In order to produce specimens which will begin to freeze over a particular temperature interval. 5 rom diameter samples are first "trained" by repeatedly melting and freezing at Ig under a fused glass flux. During the melting! freezing operation the flux strips out nucleant so that reproducible nucleation behaviour may be obtained. When trained. the flux-covered droplet is placed in an experiment capsule where it is melted in an optical furnace. Just prior to release. the multi-specimen assembly is plunged into a molten tin bath to effect a progressive quench so that solidification has been completed before the fall of the drop capsule is arrested. The equipment used and the results obtained during an experimental campaign in September 1995 will be described.

EXTRACTION, PROCESSING, AND RECYCLING

TMS General Abstract Session

Wednesday. AM February 7. 1996

Room: B3 Location: Anaheim Convention Center

Session Chairperson: TBA 8:30am A MATHEMATICAL MODEL FOR ELECTRICAL ENERGY, HEAT TRANSFER AND FLUID FLOW IN FALCONBRIDGE'S SMELTING FURNACE: Y. Y. Sheng. Noranda Technology Centre. Pointe Claire. Quebec. H9R IG5. Canada; G. A. Irons, Department of Materials Science & Engineering, McMaster University, Hamilton, Ontario, LSS 4L7, Canada; D. G. TIsdale, Falconbridge Limited, Falconbridge, Ontario, POM 1 SO, Canada Falconbridge smelts nickel calcine in a 36 MVA immersed arc smelting furnace. A three-dimensional model for the electrical energy dissipation, heat transfer and fluid flow in the matte and slag was developed. To validate the model, the voltage and temperature distribution were measured in the furnace. Analysis of the results revealed a significant voltage drop at the Soderberg electrode interface due to arcing. The implications for the operation of the furnace are discussed. 8:50am HIGH TEMPERATURE INTERACTION STUDIES TO SCREEN MELT CRUCmLE MATERIALS FOR STAINLESS STEEL-ZIRCONIUM ALLOYS AND URANIUM: Sean M. McDeavitt, Chemical Technology Division, Argonne National Laboratory, 9700 S. Cass Ave., Argonne. IL 60439; Garth W. Billings, Thermal TechnOlogy, Inc., 1911 Airport Blvd., Santa Rosa, CA 95403 Two important steps in the electrometallurgical treatment of spent nuclear fuel, as developed by Argonne National Laboratory, are the melt consolidation of stainless steelzirconium (Ss-Zr) alloys and of uranium. The sS-Zr alloys are generated as waste forms for the metallic cladding hulls and fission product metals. The uranium is collected as a dendritic mass and melted into a dense ingot. Both processes occur under a molten salt. Cold-crucible melting is not a viable option for either of these processes, so stable ceramic and coated materials are being evaluated as melt containment crucibles. Thermodynamic stahility was used to select candidate compounds consisting of oxides, carbides, nitrides. borides and sulfides. Elevated temperature sessile drop experiments were performed using zirconium. sS-Zr alloys and uranium under inert as at temperatures up to -2100°C, -1900°C and -1600°C, respectively. The sessile drop experiments were observed and recorded in situ through an external video camera. 9:10am THE KINETICS OF SELENIUM REMOVAL FROM MOLTEN COPPER BY POWDER INJECTION: Y. F. Zhao, G A.lrons, Department of Materials Science and Engineering, McMaster University, Hamilton, Ontario. L8S 4L7, Canada The injection of reagent powders is a rapid means to remove impurities from copper. In this study. it was found that selenium and oxygen were simultaneously removed when sodium carbonate and calcium carbide were injected simultaneously into 70 kg heats of copper. A mathematical model was developed to describe diffusioncontrolled selenium and oxygen removal rates at calcium carbide particle interfaces. It was found that the utilization of the powder varied over a wide range, depending on the selenium and oxygen contents. The industrial implications of the work are discussed. 9:30 am BREAK 9:40am BARRIER REFACTORIES FOR ALUMINA REDUCTION CELLS: George Oprea, Clayburn Refractories Ltd., 33765 Pine Street, Abbotsford, B.C. Canada V2S 5Cl Several fireclay refractories, such as refractory bricks and dry-barrier powders were investigated using different testing procedures for their corrosion resistance to cryolite bath in oxidizing and reducing atmospheres. Although it is difficult to simulate in laboratory experiments the real conditions of an electrolytic cell, all test results verified a similar corrosion mechanism. Accordingly, the "barrier layer" the product of reaction of cryolite bath and refractory material, has multiple roles in the corrosion evolution. The chemical, mineralogical and thermal characteristics were investigated in order to estimate the life in service for a potlining refractory. 10:00 am VSS CELL ELECTRIC, MAGNETIC AND HYDRODYNAMIC FIELDS DISTRIBUTION MODELLING: V. V. Ovchinnikov, O. G. Provorova, T. V. PiskazhovaDept of Higher Mathematics, Krasnoyarsk University, Russia; V. V. Pingin, D.A. Gorin, New Technologies and Equipment, Krasnoyarsk Aluminum Smelter, Russia; V. Y. Buzunov, Scientific & Technological Light Metals Center, Russia

1996 TMS ANNUAL MEETING AND EXHIBITION

The paper describes mathematical models of VSS cell electric, magnetic and hydrodynamic fields (Moro-Evance method). The computations were carried out using the following inputs: stud current distribution, collector bar current distribution, inducted B gradients along the cell perimeter. The models allow to develop a two - dimensional current density map of the anode, bath, metal, cathode bottom and also inducted B vectors in metal, metal flow patterns and metal-bath boundary contiguration. Based on experimental metal flow patterns data a method has been developed to determime a turbulent action coefficient used to compute the metal movement velOCity. Computation results are shown for C-8bM cell design. 10:20 am INNOVATIVE EQUIPMENT TO DECOAT SCRAP ALUMINUM: Robert De Saro, Energy Research Company, 12 Dartmouth Road, Annandale, NJ 08801 Energy Research Co. (ERCo), the New York State Energy Research and DevelopmentAuthority (NYSERDA), Roth Bros. Inc., Stein Atkinson Stordy (SAS), and O'Brien & Gerc are demonstrating a technology to decoat oily and painted scrap aluminum for the secondary and primary aluminum industries. Compared to conventional equipment this equipment, termed IDEX, reduces energy use, emissions, and dust formation, and increases production. The IDEX consists of a sealed rotating kiln, afterburner, and associated equipment. Scrap is fed into the kiln and the scrap oils or paint are vaporized by a 1500° counter-flowing gas stream. The gas stream has an oxygen level of 6-9% which prevents combustion from occurring in the kiln. The gas stream is then passed to an incinerator where the organics are incinerated apart from the scrap. An energy use reduction of 56% is possible. Additionally, there is a major reduction in emissions of VOC's, particulates, CO, Nox and C03, and a production increase due to the reduction in dross.

GENERAL PYROMETALLURGY I Sponsored by: EPD Pyrometallurgy Committee Program Organizer: Jerome P. Downey, Hazen Research, Inc., Golden, CO 80403; Ed Mounsey, Ausmelt Limited, 2/13 Kitchen Road, Dandenong, Victoria 3175 Australia Wednesday, AM February 7, 1996

Room: AI6-17 Location: Anaheim Convention Center

Session Chairmen: J. D. McCain, Vice President, Operations, Magma Metals Company, San Manuel Division, PO Box M, San Manuel. AZ 85631; H. Y. Sohn, Professor, Department of Metaliurgical Engineering, The University of Utah, 412 Browning Building, Salt Lake City, UT 84112

8:30am OXIDATION OF FeCI3 USING RADICALS PRODUCED IN A NON-EQUILIBRIUM OXYGEN PLASMA: Quiling Yu, D. C. Lynch, Arizona Materials Laboratory, Dept. of Materials Science & Engineering, University of Arizona, Tucson, AZ 85721 FeCI3 particulate was reacted with radicals produced in an oxygen plasma at pressures between 1 to 20 Torr. Reaction between the iron chloride and the gas pbase occurred at temperatures between 66 and 226'C. The experimental evidence indicates that the extent of reaction is independent of the temperature of the particulate, but is a function of the square root of the power absorbed by the plasma divided by the plasma pressure. This parameter is a reflection of the average energy of free electrons in the plasma. Electron energy establishes the collision cross section for inelastic collisions, collisions which are responsible for producing radicals such as monatomic oxygen. The radicals react with surface material with the maximum amount of Fea3 reacted being 67 wt%. 8:55am THERMOGRAVIMETRIC ANALYSIS OF THE REDUCTION OF MIXED OXIDES: Dhiren K. Panda, Kent D. Peaslee, David G. C. Robertson, Generic Mineral Technology Center for Pyrometallurgy, University of Missouri-Rolla, 215 Fulton Hall, Rolla, MO 65401-0249 Several pyrometallurgical processes for the treatment of waste oxides involve the addition of charge material in the form of pellets containing carbon (coko) as the reductant. Waste oxide materials usually contain a mixture of several oxides, such as iron oxide(s), zinc oxide, and lead oxide. Results ofTGA experiments are presented in this paper highlighting the complex reduction reactions (of various oxides) occurring within the pellets during the heating cycle. These experiments were carried out with zinc retort residues from old zinc production sites in the Missouri-Oklahoma-Kansas tri-state areas. This material contains oxides of iron, lead, and zinc with carbon content as high as 30%. 9:20am COPPER LOSSES IN COPPER SMELTING SLAGS: R. Sridhar, J. M. Toguri, S. Simeonov, Department of Metallurgy and Materials Science, University of Toronto, 184 College Street, Toronto, Ontario, Canada M5S IA4 A relationship between copper in slag and copper in matte during copper sulfide smelting has been derived using industrial data from 40 plants employing blast furnaces, reverberatory furnaces, flash furnaces, and Mitsubishi smelting furnaces, together with the available thermodynamiC eqUilibrium data for Cu-Fe-S-O, FeO-Si~, and Cu-Fe-S systems and laboratory slag-matte equilibrium information. A "Copper Smelting Diagram" showing oxygen potential, sulfur potential, copper. magnetite, and sulfur contents in slag during the smelting of different grades of copper mattes is developed. The data presented can be used to determine entrained copper losses in slag. The dissolved sulfur content of slag was derived using calculated matte entrainment and the corresponding plant data for sulfur content of slag. The sulfur content thus derived corresponded closely to the experimental determined sulfide capacity of fayalite slags. The review of available and derived information suggest that tbere is no need to assume the presence of dissolved copper sulfide species in industrial slags. The existing equi-

B95

Iibrium data which relate copper content of slag to oxygen potential adequately describe the copper losses in industrial slags. 9:45 am BREAK 10:15 am COBALT EQUILIBRIUM DISTRIBUTION BETWEEN NICKEL-COPPER MATTE AND FAYALITE SLAGS: S. Simeonov, R. Sridhar, J. M. Toguri, Department of Metallurgy and Materials Science, University of Toronto, 184 College Street, Toronto, Ontario, Canada M5S lA4 In recent years, the interest in the understanding of cobalt losses in nickel-copper smelting and converting has increased due to higher cobalt prices and the need for better cobalt recovery. Knowledge of the thermodynamic behaviour of cobalt in pyrometallurgical smelting and converting stages of nickel processing is essential to understand the behaviour of cobalt and to develop or modify processing steps to increase the cobalt recovery, which averages only about 50% in present day practice. A literature review indicates that only limited data are available under controlled oxygen and sulphur potentials corresponding to that encountered in the plant. The purpose of this study is to obtain such thermodynamic data for cobalt eqUilibrium distribution between nickel-copper matte and fayalite type slags. Cobalt equilibrium distribution measurements between fayalite slags and nickel copper matte at specific oxygen and sulphur pressures were carried out in the temperature range 1473 to 1623 K. In the actual operation, MgO is inevitable, because of tbe erosion of refractory. It is therefure of practical importance to also know how MgO addition to fayalite based slags influences cobalt behaviour. 10:40 am FLOW AND SETTLING PHENOMENA IN MATTE SMELTING ELECTRIC FURNACES: A. Andreas Hejja, R. Hurman Eric, School of Process and Materials Engineering, Branch of Extractive Metallurgy, University of the Witwatersrand, Johannesburg, Private Bag 3, WITS 2050, South Africa Flow of slag and settling of matte in slag resistance electric furnaces are discussed in conjunction with furnace geometries, stirring action of the electrodes, buoyancy and electromagnetic effects. The rectangular electric furnace was regarded as a channel reactor and the mode of slag flow as a channel flow. The flow is affected both by buoyancy and electromagnetic forces, however the effect of electromagnetic forces is minimal. The flow velocity is made up of two components: (i) that developed by the action of the electrodes and (2) the bulk flow velocity determined by the smelting rate and the volume of the reactor. Slag samples were taken from various depths of the molten bath and were analyzed microscopically to determine the size and distribution of matte particles. The action of the electrodes was described by a simple mathematical formula indicating a balanced settling rate in the area of electrode action. 11:05 am WETTING BEHAVIOR AND INTERFACIAL REACTION BETWEEN FAYALITE SLAG AND MgO: H. Fu/cuyama, J. Donald, J. M. Toguri, Department of Metallurgy and Materials Science, University of Toronto, 184 College Street, Toronto, Ontario, Canada M5S IA4 Extension of refractory life is of great interest in the copper and nickel industry. In pyrometallurgical processes, infiltration of slag melts into refractory materials affects the thermal and mechanical properties of the bricks and it may lead to early failure. The height of infiltration of the melts is generally govemed by the surface tension of the melts and the contact angle between the melts and the bricks. MgO is a main component of commercial refractories, constituting between 30 and 98 percent of the brick. There is very little information regarding the interfacial phenomena between this refractory material and slag. Thus, in the present work, the sessile drop technique was employed to investigate the wetting behavior of single crystal MgO by the fayalite type slag melts. This provided the basic data for the infiltration phenomena of commercial bricks. The reactions at the interface between the melts and the MgO were examined by Electron Probe Microanalysis. From these results, phase relationships between the slag and the MgO were determined. The inteIdiffusivity of Fe and Mg in Magnesiowustite was also determined.

A SYMPOSIUM DEDICATED TO PROFESSOR .JOHANNES WEERTMAN'S SEVENTIETH BIRTHDAY: SESSION V: HIGH TEMPERATURE DEFORMATION II

Sponsored by: JL SMDIMSD Nuclear Materials and MSD Flow and Fracture Committees and FEMS (Federation of European Materials Societies) Program Organizers: R.I. Arsenault, Deptartrnent of Materials Science and Nuclear Engineering, University of Maryland, College Park, MD 20742-2115; David Cole, CRREL, 72 Lyme Rd., Hanover, NH 03755; Todd Gross, Department of Mechanical Engineering, University of New Hampshire, Durham, NH 03824; Gemot Kostorz, Institut fUr Angewandte Physik, EfH Honggerberg, CH-8093 ZUrich, Switzerland; Peter Liaw, Department of Materials Science and Engineering, University of Tennessee, Knoxville, TN 37996-2200; Sivan Parameswaran, NRC-Institute for Aerospace Research, Ottawa, Canada KIA OR6; Howard Sizek, Inco Alloys International Inc., Huntington, WV 25705-1771 Wednesday, AM February 7, 1996

Room: Orange County 3 Location: Anaheim Marriott Hotel

Session Chairpersons: K.A. Heck, Inco Alloys International, Inc., 3200 Riverside Dr., Huntington, WV 25705-1771. T.G. Langdon, Department of Material Science, University of Southern California, Los Angeles, CA 90089-0241

8:30 am Invited DISLOCATION STRUCTURE EVOLUTION AND DEFORMATION KINETICS OF PURE MATERIALS: W. B114m, Institut fUr Werkstoffwissenschaften LS I, Universitlit Erlangen-Nilrnberg, MartensstraBe 5, 0-91058 Erlangen, Germany Results from work hardening tests at constant strain rate £ and from creep tests at constant stress a for aluminum and other pure materials are combined to discuss the mechanical equation of state relating t ,a and structure at constant temperature. The dislocation structure is mainly determined by a. Subgrain boundaries play an important role as hard regions and as dislocation sinks. Despite the complications related with subgrain and grain boundaries, Weertruan's natural law of steady state deformation is a good approximation of the steady state £ -a relation in the limit of low stresses. At high stresses the interaction hetween climb and glide must be taken into account. Information on glide kinetics is available from tests at constant structure ( £ -change, a-change and a-relaxation tests). The relative decrease of the apparent activation area of glide, which is found as the steady state is approached, is interpreted as a result of concurrent recovery. 9:00 am Invited CREEP MECHANISMS IN DISPERSION ON STRENGTHENED MATERIALS: A.K. Mukherjee, R.S. Mishra, Department of Chemical Engineering and Materials Science, University of California, Davis, CA 95616 The steady state creep mechanisms in dispersion strengthened materials are reviewed. The influence of microstructural parameters like interparticle spacing and grain size on the steady state dislocation creep are highlighted. The analysis shows that the steady state dislocation creep mechanism depends on the ratio of interparticle spacing and grain size. In materials with fine particles, the individual dislocation-particle interaction creep dominates. The dislocation creep data on Ni-ThD2 and AI-AI203-AI,jC3 alloys shows a grain size dependence. A rationale is developed to explain this behavior. A comparison of the kinetics constant shows that the creep kinetics is slower in dispersion strengthened materials even after compensating for the threshold stress. This is explained on the basis of geometrical details of dislocation climb bypass'of particles. On the other hand, the dominant steady state dislocation creep mechanism in several aluminum alloy composites, where the particles are quite coarse, is constant substructure creep. The possible reasons for the variation of true stress exponent for aluminum alloy composites from a value of 5 to 8 are discussed. 9:30am TORSIONAL BEHAVIOR AND DAMAGE MECHANISMS IN TI-6AI-4V ALLOY AND TiCffl-6AI-4V COMPOSITE: J.H. Zhu, P.K. Liaw, Departruent of Materials Science and Engineering, The University of Tennessee, Knoxville, TN 379962200; J.M. Corum, M. Ruggles, Engineering Technology Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831 The IOOO°F torsional and creep properties under torsional loading of both baseline TI-6AI-4AI and TI-6AI-4V reinforced with 10 wt%TIC particulate were investigated in this paper. It has been found that the torsional yield strength, the modulus of elasticity in shear, the modulus of rupture, and torsional creep resistance of the composite are greater than the unreinforced matrix alloy at the test temperature, whilst the ductility of the composite is much lower than that of the matrix alloy. A replication technique was used to monitor the microstructural change and damage development during the various stages of testing, which was assisted by scanning electron microscopy. Furthermore, analytical transmission electron microscopy was used to assess the nature and extent of the interactions between TIC: particulate and matrix of the TI-6AI-4Vmc composite during the torsional creep testing. 9:50am A MODEL FOR DISLOCATION CREEP FOR FINE GRAIN METALS: A.K. Ghosh, Department of Materials Science and Engineering, University of Michigan, Ann Arbor, MI 48109-2136 Grain boundary sliding has been known to influence creep behavior in metals deforming primarily by dislocation g1idf,-climb process. It has been realized in recent times that grain boundary sliding itself results from localized glide-climb creep in the vicinity of the grain boundaries, i.e., tile grain mantle region. In a recent model, we have proposed that the number of dislocation sources activated at grain boundaries depends on the strength of obstacles arId the applied stress. At low stresses, the emission and absorption of dislocations are confined to the grain boundary region. At large stresses, dislocation creep of the grain ·core becomes dominant, but due to the relative ease of grain boundary sliding, stress c()ncentration at triple points also builds up. The resulting acceleration of creep in the grain comer regions provides an additive creep component in the conventional power law creep regime. This model will be discussed and illustrated with data on metals and intermetallics. Research sponsored by AFOSRURI Grant No. F 49620-93-1-0289 and DoEER program. 10:10 am DEFORMATION MODES OF THE SUPERPLASTICAL-I0wt% TiALLOV AT HIGH STRAIN-RATES: Dong-Wha Kum, Hye-Sung Kim, Sang-Hee SUh, Korea Institute of Science and Technology, Cheongryang P.O. Box 131, Seoul 130-650, Korea; Won-Tac Kim, Department of Physics, Chongju University, Naedok-dong 36, Chongju 360-764, Korea Gas-atomized AI-IOwt% TI powders were consolidated by power metallurgy process involving hot extrusion, and two-phase microstructure consisted of micro-size aluminum grains and fine flake-shaped AI3TI particles was obtained. The fine-grained AI-IOwt% TI alloy showed very high ductility at strain-rate of O.lIsec at temperatures above 600°C. This high strain-rate superplastic(HSRS) behavior is correlated with that of whisker-reinforced aluminum composites processed by powder metallurgy, referring to the microstructural similarities. Because local melting did not occur in the AI-TI binary system at the test temperatures, requirement of incipient melting of HSRS is questioned. The HSRS behavior is intc:rpreted by an extension of the microstructural superplasticity which has been well ch,lII1Icterized in fine-grained metallic alloys. 10:30 am BREAK

B96

1996 TMS ANNUAL MEETING AND EXHIBITION

10:40 am EFFECT OF GRAIN SIZE IN POWER-LAW CREEP: K. Sadananda, C.R. Feng, Code 6323, Naval Research Laboratory, Washington D.C. 20375 Grain size effects are generally observed during Newtonian viscous flow with stress exponents equal to one, but not during power-law creep, where the exponents are of the order of 3 to 5. Our recent studies of creep of molybdenum disilicides and their composites showed significant grain size effects even during power-law creep, and these effects are strain-history dependent. We examine these effects to evaluate the generality and factors that govern the behavior. 11:00 am THE SIGNIFICANCE OF HARPER-DORN CREEP IN METALS, CERAMICS AND GEOLOGICAL MATERIALS: M. Zhou, Department of Materials Science, University of Southern California, Los Angeles, CA 90089-0241; D.M. Owen, Graduate Aeronautical Laboratories, California Institute of Technology, Pasadena, CA 91125; T.G. Langdon, Department of Materials Science and Mechanical Engineering, University of Southern California, Los Angeles, CA 90089-1453 Harper-Dorn creep is now established as an important deformation mechanism which occurs at low stresses with a stress exponent, n, of I. Recently the possible significance of Harper-Dorn creep in a range of f.c.c. and b.c.c. metals. This paper examines this controversy and further considers the possible role of Harper-Dorn creep in ceramic and geological materials. 11:20 am INDENTATION CREEP: O. Prakiish, Department of Materials Science and Engineering, McMaster University, Hamilton, Ontario, Canada L8S 4M7 The issues concerning the analysis of indentation creep data are discussed. A model for analyzing indentation creep data in terms of hack/internal stresses resulting from the interaction of mobile dislocations with microstructural inhomogeneities is proposed. The model enables us to rationalize creep indentation data to uniaxial creep results for a range of materials including metals and metal-type organic crystals and their particulate composites. The effect of recovery in indentation creep is considered, and it is shown that under conditions of slow recovery, indentation creep can be considered similar to the creep of a "constant structure" material subjected to a continuously decreasing applied stress. At high homologous temperatures, stress exponents close to 3 are obtained in agreement with the natural power law theories of creep deformation first proposed by Professor Weertrnan. 11:40 am A MECHANISTIC MODEL FOR TRANSIENT SUPERPLASTIC DEFORMATION: K. Zhang, C.H. Hamilton and H.M. Zbib, Department of Mechanical and Materials Engineering, Washington State University, Pullman, WA 99164-2920 The primary focus of research on superplasticity has been on the characterization and modeling under deformation conditions considered to be "steady-state", as a result, many models have been developed for the superplastic flow and constitutive relations have been proposed to represent the mechanistic characteristics of the relationship among flow stress, grain size, diffusion parameters, and other physical parameters. The general form of the mechanistic models has been used in formulating reasonable constitutive relations for superplastic forming (SPF) process. In earlier investigation, a strong transient behavior has been found in superplastic deformation of model material Ph-Sn eutectics, and results in difference in the stress/strain rate relation under loading and unloading(relaxation), both of which are different from the steady-state relationship. In this study, a mechanistic model, with consideration of internal stress in superplastic deformation, will be proposed to account for the transient effects with superplastic deformation. The theoretic calculation based on this model shows good representation of experimental results with Ph-Sn eutectic and engineering materials. The mechanism with the transient superplastic deformation will be considered.

HIGH TEMPERATURE COATINGS II: Physical Properties And Protection Coatings

Sponsored by: MDMD Surface Modification and Coatings Committee Program Organizers: Narendra B. Dahotre, Center for Laser Applications, University of Tennessee Space Institute, Tullahoma, TN 37388; Janet M. Hampikian, School of Materials Science & Engineering, Georgia Institute of Technology, Atlanta, GA 30332; Jacob J. Stiglich, PO Box 206, Sierra Madre, CA 91025 Wednesday, AM February 7,1996

Room:BI Location: Anaheim Convention Center

Session Chairpersons: P. P. Patil, School of Physical Sciences, North Maharashtra University, Jalgaon, 425001, India; Narendra B. Dahotre, Center for Laser Applications, University of Tennessee Space Institute, Thllahoma, TN 37388

8:30am PHYSICAL PROPERTIES OF AND PHASE TRANSFORMATION IN V-Cr-TIo SOLID SOLUTIONS: I.-H. Park, Energy Technology Division, Argonne National Laboratory, Argonne, IL 60439. E.-T. Park, Metallurgy and Materials Engineering, Illinois Institute of Technology, Chicago, IL60616 Incorporation of 0 in the surface ofV-TI-Cr alloys has been investigated by adding o to the alloys in controlled environments at 550-750·C. Several tests were performed to determine the physical properties of V-Cr-TI-O solid solutions. The amount of 0 in the alloys has been determined by the weight change before and after incorporation of the O. Microhardness was measured to define the 0 depth profile across the alloys. xray diffraction studies indicated a cubic-to-tetragonal phase transformation and a highly stressed lattice because of 0 incorporation. 0 atoms occupied face-centered interstitial

1996 TMS ANNUAL MEETING AND EXHIBITION

sites and sites between comers in the BCC sublattice. Elastic modulus and Vickers hardness also increased in the O-enriched V-Cr-TI alloys. Work supported by the U.S. Department of Energy, under Contract W-31-109-Eng-38. 8:50am EFFECT OF MECHANICAL PROPERTIES ON EROSION RESISTANCE OF WELD OVERLAY COATINGS: B.P. Levin, J.N. DuPont,A.R. Marder, Lehigh University, Energy Research Center, Bethlehem, PA 18015 The elevated temperature (400°C) erosion behavior of eleven commercially available weld overlay coatings, deposited on 1020 steel substrates, was evaluated. A relative ranking of their erosion resistance has been developed. It was found that as a result of erosion, one group of coatings did not deform plastically while another group experienced significant plastic deformation. For coatings that did not deform plastically, increased hardness led to a decrease in their erosion resistance. For plastically deformed coatings, a new toughness parameter has been determined which shows good correlation with erosion resistance. In order to further investigate the relation between mechanical properties and erosion resistance of plastically deformed materials, erosion tests and elevated temperature mechanical tests were conducted on wrought alloys of similar compositions. Their erosion resistance also showed good correlation with the new toughness parameter. The physical significance of the new toughness parameter is discussed along with relationships between hardness, tensile properties and erosion resistance. 9:10am STRUcruRALAND MECHANICALPROPERT1ES OF PULSED LASER PROCESSED fiGH TEMPERATURE COATINGS: Ashok Kumar, U. Ekanayake, D. Kjendal, Department of Electrical Engineering, University of South Alabama, Mobile, AL 36688; RB. Inturi, J .A. Bamard, Department of Metallurgical and Materials Engineering, The University of Alabama, Thscaloosa, AL 35487 Over the past few years, the pulsed laser deposition (PLD) technique has emerged one of the simplest and the most versatile methods for the deposition of thin films and coatings of a wide variety of materials. We are currently investigating the thin film coatings of nitrides (TIN, Si3N4, BN, CN, NbN etc.), carbides (SiC, TIC, B4C etc.), and also polymer coatings of polytetraflouroethylene (PTFE) on various commercial useful substrates by PLD method. The coatings of these materials have many desirable properties for application at elevated temperatures because of extremely high melting point, hardness, high temperature strength, good thermal shock resistance, and high thermal conductivity. The mechanical properties of the films will be evaluated at different processed conditions by nanoindentation. The in-situ UVNIS emission spectroscopy technique will be used to understand the nature of excited atoms, molecules, and ions in the ablated plume. The films will be characterized by X-ray diffraction, scanning electron microscope and FfIR techniques. The chemical and structural properties of these coatings will be correlated to the mechanical properties. Optimization of laser deposition parameters (energy density, repetition rate etc.,) to obtain high quality thin films will be discussed in detail. This research was supported by Alahama NASA EPSCoR Program. 9:30am FINITE ELEMENT MODELING OFTHE STRAIN AND TEMPERATURE DISTRIBUTION ON TURBINE BLADES AND ITS RELATIONSHIP WITH THE DIFFUSIONAL PHENOMENA UNDER REAL WORKING CONDITIONS: Alejandro Sanz, J.-P. Bemadou, Laboratoire de Metallugrgie, 10 Avenue E. Belin, 31055 Toulouse cedex, France; L. Llanes, M. Anglada, M. Gomez, UPC ETSII, Depde Ciencia de loc Materiales y Metalurgia, Av, Diagonal 647, 08028 Barcelona, Spain; M. B. Rapaccini, Polytecnico di Milano, Dipartimento de Ingenieria Aeroespaziale, Piazza Leonardo da Vinci, Milano, Italy Computer similation of temperature and stress field associated with turbine blades under different working conditions are presented. The calculation problem was solved using CATIA system finite element code. The improvement of the gas turbine engine efficiency requires to have the highest possible Thrbine Inlet Temperature (T.I.T.). Turbine's turbine blades and nozzle vanes are commonly protected against high temperature degredation with NiCoCrAlYTa coatings. Important element diffusion is observed across the substrate/coating interfaceleading to substantial changes in the chemical and mechanical behavior of the coated system. The computed results were used to compae the diffusional phenomena in real AM-3 superalloy single-crystal blades under real working conditions, and laboratory samples after creep tests and four points bending fatigue test. Blades as well as laboratory samples were single crystals, [001] oriented, coated with a Low Pressure Plasma Spray (LPPS) NiCoCrAlYTacoating. Samples were submitted to four points bending isothermal low cycle fatigue during 1000 cycles until fracture. The stress controlled test was performed usind a 10-10-10 second cycle at 1323K (l050°C) and 140 Mpa; other laboratory samples were tested under creep conditions underthe same conditions of stress and temperature during different times. Analyzing directly the concentration profiles the information of the interdiffusion fluxes of all componentscan be obtained fomm Energy Dispersive microprobe Spectroscopy (EDS) the diffusional phenomena between the coating and the substrate is characterized and correlated with finite element calculated field stresses and temperature at the point in which the EDS data was obtained. 9:50am SYNTHESIS AND CHARACTERIZATION OF POLY (2-ANISIDINE) TmN FILMS: P. P. Palil, J. R. Mahajan, V. V. Taiele, P. P. Kumbhar, M. A. More, School of Physical Sciences, North Maharashtra University, Jalgaon 425 001, India Poly (2-anisidine) thin films have been synthesized by electrochemical deposition technique on transparent Sn~: F conducting glass substrates under galvanostatic conditions. The effect of current density, monomer concentration, deposition temperature and time on the UV-visible absorption and mOIphology of the films has been investigated. The UV-visible absorption spectra clearly indicate a peak at-850 nm which is attributed to the presence of conducting emeraldine salt phase. In addition the spectra show a shoulder like structure at-450 nm corresponding to the formation of lattice polaron. The extent of the sharpness of the peaks decreases with increase in monomer

B97

concentration, thereby suggesting that a lower monomer concentration favors the formation of conducting emeraldine salt phase. The morphology of the films strongly depends on the monomer concentration and reaction temperature. which may be due to the difference in the growth speeds of the films. Some characteristics of synthesized polymer were examined by Infrared spectroscopy (IR) Nuclear Magnetic Resonance (NMR) spectrometry and X ray diffraction (XRD) measurements. 10:15 am BREAK 10:25 am UFEASSESSMENTOFSERVICE STRESSED IDGH TEMPERATURE COAT· INGS BY PERMEABILITY MEASUREMENTS: N. Czech, F. Kirchner, W. Stamm, Siemens AG, Power Generation Group (KWU), Materials Technology, D-45466 Mulheim a.d. Ruhr, Germany During service blades and vanes of stationary gas turbines are subjec~d to di.ffe~nt kind of loadings like mechanical and thermal stresses as well as corrOSIon, OXIdation and erosion. The increase in surface temperature causes a more severe attack to the blade coating which commonly is an MCrAlY-type. The time for refurbishing the blade must be detennined depending on the effective life of the protecting coating. This should be done by a simple measurement and an equipment which is portable. In the new method applied by Siemens the physical effect used is th~ change of magn~tism?f the coating during service: By the loss of p-phase the magnetic state of the coating WIll change from para-to ferromagnetism. By measuring the magnetic permeability it is possible to assess the coating. We will present permeability measurements on different bladings which will be compared with calibration curves. 10:45 am CHARACTERIZATION OF THE INTERFACIAL TOUGHNESS A SUPERAL· LOY PROTECTED BY A NICoCrAIYTa COATING UNDER CREEP: Alejandro Sanz, J.-P. Bernadou, Laboratoire de Metallugrgie, 10 Avenue E. Belin,31055 Toulouse cedex, France; L. Llanes, M. AngIada, M. Gomez, UPC ETSII, Dep de Ciencia de loc Materiales y Metalurgia, Av, Diagonal 647, 08028 Barcelona, Spain; M.B. Rapaccini, Polytecnico di Milano, Dipartimento de Ingenieria Aeroespaziale, Piazza Leonardo da Vinci, Milano, Italy One of the main parameters in the quality of a coating is its adhesion t? the substrate. NiCoCrAIYTa coatings frequently used to protect the components m the hot path of the gas turbine engines are commonly deposited by the Low Pressure Plasma Spray (LPPS) technique. Unfortunately, the tests which are commonly used measure the adhesion have a number of drawbacks and thus do not provide useful data for the study of the adhesion mechanisms. The interface indentation test, based on the fractore mechanics method, is an attempt to detennine the interface thoughness or interface cracking energy. Although some questions arise about the physical meaning of such an interface tougltness, this test may be used fruitfully to compare t!te cracking resis~ce at the substrate/coating interface. AM-3 single-crystals, [001] onentated, coated WIth a LPPS NiCoCrAIYTa coating were tested under 1323 K (105OC) and 140 Mpa for different times and until fracture. The paper reports the evolution of the interface toughness relating it with the microstructural and chemical evolution of both the substrate and the coating. 11:05 am Invited INTERMETALUC PHASE FORMATION AT THE SOLID-UQUID INTERFACE IN TIlE "BASE METAL-THIN MELT LAYER" SYSTEM: L P. Ejimenko, E.A. Antonova, L.P. Petrova, Institote of Silicate Chemistry of Russia Academy of Sciences, uI.Odoevskogo,24, korp.2, St. Petersburg, 199155, Russia The general regularities of liquid phase formation were studied in the binary multicompound Fe-TI, Ni-TI, Ni-Zr, etc. systems in the temperature range involving several eutectics (950-135OC). It was shown that in the "solid metal-thin melt layer" system the eutectic is formed, which is closest to the base metal ad in the system studied the partial thermodynamic equilibrium is attained. The growth kinetics of intermetallic and solid solution layers formed by multiphase diffusion were studied at various temperatures in different solid-liquid diffusion couples. It was found that the thickness of the grown layers varies parabolically with time. The diffusion coe~ficients were detennined. The activation energies of the diffusion were compared WIth those of .growth. expirimental results obtained were summarized and analyzed on the baSIS of eqUlbbrium phase diagrams.

!he

11:30 am MICROSTRUCTURE OF BONDING ZONE IN LASER CLAD NI·ALLOY BASED COMPOSITE COATINGS REINFORCED WITH VARIOUS CERAMIC PARTICLES: J. H. Ouyang, Y. T. Pei, T. C. Lei, Department of Metals and Technology, P.O. Box 433, Harbin Institute of Technology, Harbin,15~1, China . Microstructure of the bonding zones between laser clad NI-alloy based coatings and their substrates was studied by means of SEM and TEM techniques. Observation indicated that for pure Ni-alloy coating the changes of laser par~ters selected ~or good interface fusion has no effect on the microstructure of the ~ndmg zone .except I~S thickness. However, the addition of ceramic particles (TIC, SIC, or ZrOz) m the NIalloy changes the compositional or constitutional undercooling of the alloy melt near the solid/liquid interface and consequently leads to.the obvious changes of micr~stru~­ ture of the bonding zones. For TIClNi-alloy coating, the morphology of Ii-NI sobd solution in the bonding zone changes from dendritic to planar form with increasing scanning speed. A colony structure of eutectic is found in the bonding zone of SiC~i­ alloy coating in which the complete decomposition of SiC particles takes place dunn.g laser cladding. The immiscible melting of ZrQ and Ni-alloy powders indu~e the stratification of ZrOzlNi-alloy coating which consists of a pure ZrOz layer m the upper region and a bonding zone composed mainly of ~-Ni dendrites adj~cent to ~ S?bstrate. All the bonding zones studied in this study gIve good metallurgIcal combmation between the coating and the substrate. 11:50 am THE OXIDATION RESISTANCE AND IDGH THERMAL, CONDUCTIVITY

B98

ON NOVEL CU BASED ALLOYS CONTAINING SI, RE ELEMENTS: Shou Hong Huang, Wei He, Wen Liu Huang, S. China Univ. of Tech. Guangzbou, China; TIe Fan Li, Inst. of Cor. and Prot. of Metals, Academia Sinica, Shenyang, China Using the chemical heat-treatment process, we diffuse Si, Re etc elements into Cu and its alloys which has an excellent thermal conductivity and a better oxidation resistance at 600°C. The electrodeposition of Ni onto Cu followed by co-diffusion of Si etc, can greatly improve liquid metal corrosion resistance of samples in molten Sn-Pb alloy at 250°C. Diffusing Re into Cu, the micmhardness and layer depth are increased. The oxidation behavior was characterized by weight-gain-time data. The samples appeared CU5Y, A1203, SiOz etc, it showed an c:xcellent stability at~. The compositi~n, phases, microstructures were analyzed by means of electron trucroanalyzer, x-ray diffraction and SEM respectively.

HIGH TEMPERATURE SUPERCONDUCTORS· SYNTHESIS, PROCESSING, AND LARGE SCALE APPLICATIONS V: Applications and Conductor Developmlent

Sponsored by: Jt. EMPMD/SMD SUF>erconductlng Materials Committee, MSD Electronic, Magnetic and Opt1caI Phenomena Committee, and FEMS (Federation or European Materials Societies) Program Organizers: U. Balachandran, Energy Technology Division, Argonne National Laboratory, Argonne, IL 60439; Paul J. McGinn, University of Notre Dame, Notre Dame, IN 46556; Stuart Abell, University of Binningham, Edgbaston, Binningham BI52TT,UK Wednesday, AM February 7, 1996

Room: B7 Location: Anaheim Convention Center

Session Chairpersons: U. Balachandran, Argonne National Laboratory, Argonne, IL 60439; WoK. Chu, Texas Center for Superconductivity at the University of Houston, Houston, TX 77204 8:30 am Invited DYNAMICS AND FRICTIONAL ENERGY LOSS ON HYBRID SUPERCONDUCTING MAGNETIC BEARINGS: WoK. Chu, Z. Xia, K. B. Ma, Q. Y. Chen, R. Cooley, P. C. Fowler, C. K. McMichael, Texas Center for Superconductivity at the University of Houston, Houston, TX 77204-5932 Rotor dynamics and analysis of frictional energy loss have been conducted for a flywheel energy storage (FES) prototYIll:, in which a 42 lb. flywheel is levitated by the hybrid superconducting magnetic bearings (SMB). The flywheel currently can rotate up to 6,000 RPM with kinetic energy of 8 Wh stored. The previous spindown tests indicate an average frictional energy lOllS <2% per hour in a vacuum of 10-5 torr. To improve overall system performance for a further industrial module design, we installed a sensor system for vibration measurements and did mechanical modifications to minimize the friction loss. In this talk, we will report recent characterization of our bearings and extrapolate its performance onto a larger flywheel system. 8:50 am Invited A ruGH TEMPERATURE SUPERCONDUCTOR ruGH VOLTAGE GENERA· TOR: J. X. ]in, S. X. Dou, H. K. Liu, Center for Superconducting and Electronic Materials, University ofWollongong, Wollongong, NSW 2522, Australia; C. Grantham, Dept. of Electric Power, University of NSW, Sydney 2052, Australia A high voltage generator has been developed based on a R, L, C resonant circuit. The generated high voltages are simulated, and also tested with the built .contro.lIer. Resistance in the circuit limits the voltages that can be built up. Therefore, usmg a hlghtemperature superconducting (IITS) inductor is considered .. The (Bi,Pb)2Sr2Ca2Cu3010+x HTS wires are tested and used to make a HTS COIl. The results show that this HTS wire is a good candidate to make a HTS inductor. By replacing a normal copper winding inductor with a HTS inductor, the magnitude of achievable voltages is substantially increased. 9:10am APPLICATIONS OF SOIL MECHANICS TO THE EXTRUSION PROCESS· ING OF SUPERCONDUCTING WIRE: C. H. Schilling, R. W. McCallum, D. Emery, Ames Laboratory and Department of Materials Science and Engineering, Iowa State University, Ames, IA 50010 Key requirements for extrusion processing of superconducting composite wire are dimensional unifonnity of the cross section and high packing densities within the core to enhance the sinterlreaction step. To meet these requirements, the core must exhibit uniform plastic flow, which is in turn regulated by two critical factors: (i) the stress state, which is inherent in the fonning operation, and (ii) powder processing variables including the particle morphology, size distribution, and interparticle friction. This presentation will begin with a review of soil mechanics prinCiples that are used to characterize these factors and enhance the predictability of the deformation response of the superconducting core. Soil mechanics !fiSts will be discussed that were used to characterize the compressibility and critical state flow behavior as a function of the particle size distribution of model powder systems. 9:30am QUANTITATIVE ANALYSIS OF THE EFFECTS OFSTRAIN·STATE ON TIlE MICROSTRUCTURE AND J c OF BSCCO TAPES: W. R. Blumenthal, Y. T. Zhu, R. J. Sebring, T. C. Lowe, Materials Science and Technology Division, Los Alamos National Laboratory, Los Alamos, NM 87545; R. J. Asaro, Dept. of Applied Mechanics and Engineering Sciences, University of California at San Diego, La Jolla, CA 92093 The effect of strain-state on the microstructure and critical current density (Jc) Of silver-sbeathed, Bi 1.9Pbo 4Pbo.4Sr2Ca:uCu3010+x tape was investigated by rolling

1996 TMS ANNUAL MEETING AND EXHIBITION

tapes embedded within steel blocks. Pure compression and combined compressionshear deformation states were generated in different planes of the blocks. In addition, high hydrostatic loading was investigated by confining the edges of tapes to prevent free lateral deformation. Quantitative image analysis was performed on digital backscattered electron images of transverse sections of tape using standard gray-level threshold analysis techniques. Image analysis showed highly porous oxide core regions at the edges of unconfined tapes that could explain low Jc values compared to tapes rolled under plane-strain constraint. Image analysis was unable to detect significant differences in the volume of nonsuperconducting phases which could have explained the superior Jc Of confined compression-shear processed tapes compared to pure compression processed tapes. 9:50 am Invited RECENT DEVELOPMENTS IN FABRICATION AND PROPERTIES OF AgCLAD BSCCO CONDUCTORS: U. Balachandran, A. N. Iyer, R. Jammy, J. Y. Huang, Argonne National Laboratory, Argonne, IL 60439; P. Haldar, Intermagnetics General Corporation, Latham, NY 12110 Long lengths of Ag-clad BSCCO superconductors with consistent current transport properties were fabricated by the powder-in-tuhe technique. A high-Tc magnet, fabricated from 770 m of monofilament conductor, generated a record-high field of I Tat 4.2 K in a field of 20 T. Strain tolerance of the conductors was evaluated by in-situ tensile and hend tests. Superconductingjoints have been developed to connect Ag-clad BSCCO conductors. Typical Ie through a monocore butt joint was 23 A. An analysis of the results will he presented, along with the effects of various parameters and microstructural studies. Work at ANL and part of the work at IGC is supported by the U.S. Department of Energy (DOE), Energy Efficiency and Renewable Energy, as part of a DOE program to develop electric power technology, under Contract W-31-109-Eng38. 10:lOam MASS DENSITY EVOLUTION OF (BI,Pb)+SrlCalCu30IotAg WIRE AND TAPE DURING MECHANICAL DEFORMATION: Q. Y. Hu, H. K. Liu, S. X. Dou, Centre for Electronic and Superconducting Materials, University of Wollongong, Wollongong, NSW 2500, Australia The mass densities of the Ag-clad (Bi,PbnSr2Ca2CU30to wire and tape varies during the mechanical deformation process, as one of the steps of the oxide-powder-intube technique used to fabricate the composite superconductor. Results show that the rolling has more significant effect on densifying tape core. Whereas the drawing process can only densify the core to about 75% of the theoretical denSity. SEM observations conftrm the mass density calculations. SEM observation also show that with increasing the deformation extent, the average grain size is reduced. It is proposed that although the rolling densifies the tape core much significantly, it also destroy the crystallinity of the superconducting phases and results in the formation of amorphous phase. Since the textured Bi-2223 phase forms by the epitaxial growth on the textured Bi2212 seed crystals, the deformation induced texture is critical. Appropriate deformation extent is necessary, since too high extent deformation may change the well aligned grains into amorphous phase. The formation of the amorphous phase is harmful to the texturing formation of the Bi-2223 phase, which finally leads to a critical current degrading. Microhardness measurement results are also identical to the mass density calculations. 10:40 am CONTROL OF MECHANICAL DEFORMATION PROCESSING IN FABRICATION OF BI-BASED SINGLE AND MULTIFILAMENT TAPE: W. G. Wang, H. K. Liu, S. X. Dou, Centre for Electronic and Superconducting Materials, University of Wollongong, Wollongong, NSW 2500, Australia Different strain between Ag and oxide core is measured to describe interface between Ag and oxide core. The sausage effect is induced by different deforming resistance during mechanical deformation process which comes from difference in mechanical properties between silver or silver alloy and oxide material. The thickness of 0.1 mm without sausage effect is achieved by controlling processing parameters, such as smaller diameter of roller, variable reduction from 30% to 5% depending on density of oxide core from low to high, smaller powder particle size, lower rolling speed, doping pure silver powder, lower and homogenized packing density and reducing annealing times in fabrication of single filament tape. For the multifilament tape, the satisfied packing factor and eliminating inhomogeneity of internal stress are very important. A model is put forward to explain the effect of various parameters of material and mechanical deformation to interface between silver and oxide. The improvement of interface is achieved by processing control according to this model for fabricating Bi-2223 long multifilament tape. 11:00 am HIGH TEMPERATURE PRESSING OF MONO AND MULTIFILAMENTARY BI SUPERCONDUCfING TAPES: P. A. Bain, D. Yu, Q. Y. Hu, T. Chandra, H. K. Liu, S. X. Dou, Centre for Electronic and Superconducting Materials, University of Wollongong, Wollongong, NSW 2500, Australia Silver clad (Bi,Pb)2Sr2Ca2CU30to superconducting monofilamentary and multifilamentary tapes have been uniaxially pressed at various temperatures ranging from 700°C to 850°C under various pressures ranging from 5 MPa to 20 MPa. The uniaxial pressing was performed at various stages of thermomechanical processing consisting of alternate cold pressings and sinterings. To determine complementary or similar mechanism benefits, high temperature vacuum treatments were also performed at various processing stages. SEM, TEM, Jc behaviour in magnetic field, AC susceptibility and microhardness measurements were performed at each stage of processing to assess induced changes. These were compared with measurements of tape undergoing the same thermomechanical treatment to the exclusion of high temperature pressing and vacuum treatments. All the obtained microstructural and electromagnetic properties of high temperature pressed and vacuum treated tapes show significant improvements over non-treated tapes. In particular, powder texture and consequently Jc show outstanding improvements. A substantial improvement Of Jc behaviour in magnetic

1996 TMS ANNUAL MEETING AND EXHIBITION

field also indicates a beneficial effect of high temperature pressing on powder density and possibly flux pinning properties. All aspects of both treatments have been thoroughly investigated and analyzed. 11:20 am MICROSCOPIC MECHANISMS OF THE SUPER CURRENT TRANSPORT IN (BI, Pb)2Sr2Ca2Cu301O SILVER-SHEATHED TAPES: B. Hensel, G. Grasso, R. Aiikiger, Universite de Geneve, Departement de Physique de la Matiere Condensee, 24, quai Ernest-Anserrnet, 1211 Geneve 4, Switzerland We have proposed the "railway-switch" model to describe the supercurrent transport in (Bi,PbnSr2Ca2Cu30to silver-sheathed tapes. Primary limitations for the critical current are imposed by the small effective cross-section for the current, by the low intragrain critical current density jcC along the c-axis (compared to jcab) and by the even lower critical current density jc I across twist boundaries or intergrowths within individual grains and colonies of grains. In magnetic fields a secondary limitation for ic is given by the lack of effective pinning that is omnipresent in Bi-based superconductors. A limitation by weak links is only of minor importance. In this work we give an overview of the experimental evidence in favour of the "railway-switch" model and present ways to overcome the secondary level of limitation for icIrradiation with high energy protons generates splayed columnar defects that serve as effective pinning centers and enhance the irreversibility field by one order of magnitude at T = 77 K. 11:40 am LOCAL MEASUREMENTS OF CRITICAL CURRENT DENSITY IN BSCCOPHASE CONDUCTORS: M. Johnston, J. Everett, M. Dhalle, A. D. Caplin, Centre for High Temperature Superconductivity, Blackett Laboratory, Imperial College, London SW7 2BZ, UK We describe scanning Hall probe studies of the current distribution in BSCCOphase tape and ribbon conductors. The pattern of current flow is affected by temperature, applied magnetic field and also the current level itself. We correlate these local studies with macroscopic information from transport and magnetization experiments, and also with the microstructure. Preliminary studies have been made also of the current distribution between filaments in multi-filamentary BSCCO-phase tapes.

INTERMETALLICS AND TITANIUM TMS General Abstract Wednesday, AM February 7, 1996

Rooom: B4 Location: Anaheim Convention Center

Session Chairperson: TBA 8:30am THE USE OF ATOMIC SIZE EFFECTS TO TUNE PHASE STABILITY IN FE3ALXSI I-x: L. Anthonv, University of Toledo, Department of Pbysics and Astronomy, Toledo, OH 4360-3390 It has been known for some time that the addition of certain ternary elements such as TI and Si can result in huge increases in the DOJ-B2 transition temperature (TcD03-B2) of Fe3Al. Recent high-temperature calorimetry work by Anthony and Fultz [Acta Metall. Mater., in press~ has shown that the varying efficacies of early transition metal solutes in raising TcDo -B2 in Fe3AI is related to their metallic radii. This size-effect phenomenon may prove useful in extending the stability range of other D03-ordered interrnetallics. In this talk, I shall present results of such an attempt at controlling the stability range of the D03-ordered pseudobinary alloy Fe3AlySil_y' By adjusting the relative concentrations of AI and Si in Fe3AIySi I-y, it should be possible to tune the size matching to a particular early transition meW, and thereby optimize the stability of the DOJ structure in this alloy system. I shall discuss the relevance of my results for alloy design and engineering in these and other intermetallic systems. 8:50am INTERDIFFUSION AND PHASE FORMATION IN THE Mo-S! SYSTEM: P. C. Tortorici, M. A. Dayananda, Purdue University School of Materials Engineering 1289 MSEE Bldg. West Lafayette, IN 47907-1289 Diffusion studies were conducted in the Mo-Si system to investigate the formation of Mo silicide compounds during interdiffusion of Mo and Si. Solid solid couples were assembled with pure Si, Mo or MoSi2 as terminal disks and isothermally annealed at selected temperatures between 900°C and l200°C for various times. Optical Microscopy, SEM, TEM, and x-ray diffraction were utilized to characterize the resulting phase layers and microstructures. Phase layers observed include MosSi3 and MoSi2' The MoSi21ayer exhibited a columnar microstructure. integrated diffusion coefficients were calculated for the various silicide layers. Results on the relative diffusion behavior of Mo and Si in the silicide layers will also be discussed. 9:10am ALLOYING TO ENHANCE THE SHAPE MEMORY PROPERTIES OFNIMn: M. L. Johnson, D. E. Mikkola, Dept. of Metallurgical and Materials Engineering, Michigan Technological University, 1400 Townsend Dr., Houghton, MI 49931 Earlier work has shown that the shape memory behavior of binary NiMn can be altered with higher order alloying. Changes in shape recovery, transformation temperature for the B2 to Llo transformation, and mechanical properties can be linked to changes in the cIa ratio of the low temperature phase. A series of alloys Ni50Mn50_ xXx has been studied that covers a range of transformation temperatures from 500·C to below room temperature. Efforts to relate changes in shape memory behavior to changes in transformation and deformation behavior will be discussed. 9:30am FRACTURE AND FATIGUE-CRACK GROWTH RESISTANCE OF MoSll COMPOSITES: K. Badrinarayanan, K. T. Venkateswara Rao, R. 0. Ritchie, Dept. of

B99

Materials Science and Eng., University of California, Berkely, CA 94720-17(1.) Recent research ofMoSi2 alloys for structural use at temperatures in excess of 1200'C has been motivated by their excellent oxidation resistance; however, they invariably suffer from poor fracture resistance. In light of this, the current work was aimed at comparing the role of ductile vs. brittle reinforcements in influencing the fracture toughness (Kk) and fatigue-behavior of MoSi2 at both ambient and elevated temperatures; specifically using 20 vol.% of Nb and SiC. It was found that the addition of the ductile phase as chopped Nb wire mesh resulted in a significant increase to Kk from 2MPa+m for unreinforced MoSi2 to 5.5 MPa+m for the composite, toughening that was associated with uncracked ligament bridging and crack deflection. However, this beneficial effect was not found under fatigue loading, as the Nb ligaments failed prematurely, thereby limiting any shielding. With brittle SiC p reinforcements, there was no detectable effect at room temperature, although at 1200'C, cyclic crack growth was observed between a threshold K of 4.1 MPa+m. The latter phenomenon is considered in terms of a mutual competition between intrinsic cavitation damage by localized grain-boundary void formation (which promotes crack advance) and extrinsic crack-tip shielding by viscous bridging (which acts to oppose it). 9:50am THE (J.HY PHASE TRANSFORMATION DURING RAPIDLY QUENCHING AND HEATING IN TI-rlch TIAI ALLOY. T. Kumagai. E. Abe, M. Nakamura. National Institute for Metals, Ibaraki 305, Japan The phase transformation from high-temperature (J.-TI (disordered h.c.p) to y-TIAI (ordered f.c.t. with a Llo structure) and the reverse transformation from to in TI48al. %AI alloy have been examined. The extremely fine a2 (ordered b.c.p with D>!9 structure) I y lamellae together with the y grains with a massive morphology (ym) due to (J.--.yreaction are observed in the sample solution-treated in the (J. phase field (I 683K) followed by ice water quenching. On the other hand the similar fine (J.2/ylamellae due to (J.--.yreaction are formed within the ym grain interiors by up-heating to the temperature range above 1473K, although these fme lamellae in the ym grain interiors are not recognized in the samples up-heated to the temperature range below 1273K. The formation process of the extremely fme (J.2/Y lamellae by rapidly quenching and heating will be discussed. 10:10 am BREAK 10:20 am SMALL FATIGUE CRACK GROWTH IN y-BASED TITANIUMALUMINIDES: J. P. Campbell, S. Lillebridge, K. T. Venkateswara Rao, R. 0. Ritchie, Department of Materials Science and Mineral Engineering, University of California, Berkeley, CA 94720-17(1.) Recently, y-based titanium alumindes have received considerable attention in the aerospace community as candidate materials for several high-temperature structural applications. However, in part due to their limited room-temperature ductility and toughness, these alloys suffer from extremely steep daldN v. AI( crack-growth relationships. Since this results in a severely limited crack-propagation life in damage-tolerant lifetime predictions, the structural use of y-TIAI alloys may necessitate the use of fatigue threshold and near-threshold small-crack behavior in such design and life calculations. In this study, the fatigue behavior of "small" surface cracks (<500 -I1TIAI alloys, including the TI-48A1-2Cr-2Nb ("GE") alloy in both equiaxedduplex and lamellar (a2 +y) microstructures. In addition to delineating the principal microstructural features controlling crack growth, the intent is to characterize differences between long and small crack behavior, specifically in terms of the role of crack size on crack-tip shielding and on statistical sampling considerations. Work supported by the U. S. Air Force Office of Science Research. 10:40 am MICROSTRUCTURAL INFLUENCE ON THE FATIGUE CRACK PROPAGATION OF GAMMA nAI ALLOY: C. Choi, C. S. Lee, Center for Advanced Aerospace Materials, Pohang Univ. of Sci. & Tech., Pohang, 790-784, Korea The role of microstructure on the fatigue crack growth resistance of TI-48AI-ICr-IV-2.5Nb alloy was examined in this study. Fully lamellar microstructure showed superior fatigue properties as compared to the duplex and near gamma microstructures because of large extrinsic effect such as crack closure rather than intrinsic property. In the fully lamellar microstructore, colony boundary acted as strong or weak barriers for the crack propagation depending on the relative orientation of the lamellar lath to the crack propagation direction. Lamellar structures in the duplex microstructures were not the effective obstacles due to a small colony size. 11:00 am THE EFFECTS OFTIJAI AND SILICIDE PRECIPITATES ON THE TENSILE, FATIGUE CRACK GROWTH, AND CREEP BEHAVIOR AT ROOM TEMPERATURE AND 593°C IN THE NEAR-ALPHA TITANIUM ALLOY, TI-11OO: A. Madsen, Department of Materials Science and Engineering, University ofWashington, Seattle, WA 98\05; A.H. Rosenberger, H. Ghonem, Mechanics of Solids Laboratory, Dept. of Mechanical Engineering, Univ. of Washington, Seattle, WA 98105 Long term aging of near-alpha titanium alloys, at temperatures typical during service in a gas turbine engine, results in the formation of silicide precipitates along the alpha platelet boundaries and the formation ofTI3A1 homogeneously distributed throughout the matrix. A post aging heat treatment that relies on the phenomenon of the critical ordering temperature is used to remove the TI3AI precipitate while leaving the silicide intact. Three materials-unaged (precipitate free), overaged (TI3AI+silicide). and post aging heat treated (silicide-only)-are compared, with the objective of identifying the separate effects of the TI3AI and silicide precipitates on tensile, fatigue crack growth, and creep behavior at room temperature and 593°C. The TI3AI precipitate is shown to be largely responsible for the increase in tensile yield stress and for the decrease in ductility at both test temperatures. In contrast, the increase in room temperature fatigue crack growth rate associared with aging is attributed to the silicide, with TI3AI playing only a minor role. Aging produces a slight improvement in fatigue crack resistance at 593°C, which also appears to be due to the presence of the silicide precipitate. Acceler-

B100

ated creep tests indicate that the presence of TI 3AI has a beneficial infl uence on the creep rcsistance ofTI- 1100. Since precilpitation occurs over both short and long time periods (800 to (1.),000 minutes), aging effects should be considered when characterizing high temperature mechanical behavior in this and similar near-alpha titanium alloys. 11:20 am

LOW CYCLE FATIGUE PROPERTIES OF A TITANIUM ALLOY: B. Said, A. Inchekel, N. Arakere, Wichita State University, Department of Mechanical Engineering, Wichita, KS 67208 Low cycle fatigue tests at moderately elevated temperatures were conducted on round titanium alloy specimens that were extracted from aircraft turbine disks. These tests were performed in order to investigate the effect of moderate temperatures on the cyclic stress strain curves, mechanical properties, and the cyclic life of the material. Furthermore, microstructural changes with temperature and fractographic analysis were carried out. Results showed that the te!;ting temperatures softened the material, and reduced the fatigue life, yield point, and ultimate tensile strength.

INTERNATIONAL SYMPOSIUM ON REACTIVE METALS: Processing & Applications III: Separation Processes

Sponsored by: LMD Reactive Metals Committee Co-sponsored by: Canadian Institute of Mining, Metallurgy & Petroleum, Montreal, Canada; The Japan Institute of Metals, Sendai, Japan; Mining & Materials Processing Institute of Japan, Tokyo, Japan; Society for Mining, Metallurgy & Exploration, Littleton, CO Program Organizers: B. Mishra, Department of Metallurgical & Materials, Engineering, Colorado School of Mines, Golden, CO 80401; GJ. Kipouros, Department of Mining & Metallurgical Engineering, Technical University of Nova Scotia, Halifax, Nova Scotia, Canada B3J 2X4; R.G. Reddy, Department of Chemical & Metallurgical Engineering, University of Nevada, Reno, NV 89557 Co-Organizers: W.A. Averill, Rocky Flats, Inc., Golden. R.G. Bautista, University of Nevada - Reno, Reno; M.C. Bronson, Lawrence Livermore Natl. Lab., Livermore; J.A. Sommers, Teledyne Wah Chang Albany, Albany; C.B. Wilson, The Dow Chemical Company, Freeport Wednesday, AM February 7, 1996

Room: B9 Location: Anaheim Convention Center

Session Chairperson: Dr. R.G. Baustista, University of Nevada - Reno, Department of Chemical & Metallurgical Engineering, Mackay School of Mines, Reno, NV 89557; Dr. K. Liddell, Department of ChemiCltl Engineering, Washington State University, Pullman, WA 99164

8:45am PROCESSING OF LOW GRADE ZINNWALDITE (LITHIUM MICA) CONCENTRATE: P. Alex, A.K. Suri, Metallurgy Division, BhabhaAtomic Research Center, Trombay, Bombay 400 085, India Zinnwaldite, lithium bearing mica, is obtained as one of the byproduct during the beneficiation of wolframite ore of Degana (Rajasthan). It contains about 0.29 to 0.60% lithium, and some amounts of rubidium and cesium. Initially various processing schemes involving a combination of pyro- and hydro techniques were examined. This paper is concerned with the studies carried out on sulfuric acid leaching and pugging for a lithium concentrate analyzing 0.59% lithium. It was observed that acid leaching resulted in recovery of about 95% of lithium but the leach liquor contained a high proportion of iron. Recoveries in the acid pugging on the other hand were almost the same as in leaching but the iron contamination of the leach liquor was substantially low (Li to iron ratio of about 8: I). Lithium from the leach liquor was recovered as lithium carbonate by adopting a combination of ion exchange purification and chemical precipitation. 9:10am ROLE OF ION EXCHANGE REAC-nONS IN THE IN SITU LEACHING OF URANINITE BY NII4HC03 -(NII4>2C03 -H202: J. Brantuer, K. Liddell, Department of Chemical Engineering, Washington State University, Pullman, WA 99164 Ion exchange on clay mineral surfaces can alter the concentrations of solution species, the soluhility of other minerals thaI share a common cation, and the porosity and permeability of the host formation. Equilibrium cation exchange reactions involving NH4+, Na+, K+, H+, Ca2+, and Mg2+ were incorporated in a mathematical model for in-situ leaching ofU02 by solutions ofNH4HC03, (NH4nC03, and H202.The clay surface was assumed to be in equilibrium with groundwater initially; contact with leach solution resulted in displacement of the Na+, K+, H+, Ca2+, and Mg2+ cations from the clay, often accompanied by precipitation of calcite and magnesite. Development of the ion exchange model is described and implications for successful in-situ uranium recovery and groundwater restoration are discussed. 9:35am REMOVAL OF RADIONUCLIDES USING ZEOLITES: R.G. Reddy, Z. Cai, Department of Chemical and Metallurgical Engineering, MS 170, University of Nevada, Reno, NV 89557 Adsorption of uranium (VI) from aqueous solutions on natural zeolites,i.e. chabazite, clinoptilolite, erionite and mordenite was investigated. The influence of time and pH of the solution were studied. The results showed that uranium (VI) species are strongly adsorbed on the zeolites between pH 6 to 9. The amount of uranium adsorption is strongly dependent on pH and to some extent, on the type of zeolites. For pH> 6 and at

1996 TMS ANNUAL MEETING AND EXHIBITION

25°C, more than 92% of uranium from solution was removed in 10 minutes. Adsorption mechansim of uranium and application of the results to some industrial systems are discussed. 10:00 am BREAK 10:15 am NITRIDING PROCESS FOR THE RECOVERY OF VANADIUM FROM FERROVANADIUM: K. Singh, A.K. Suri, C.K. Gupta, Metallurgy Division, Bhabha Atomic Research Center, Trombay, Bombay 400 085, India A three step process based on nitriding-Ieaching-pyrovacuum decomposition has been developed to prepare vanadium-nitrogen alloy from ferrovanadium. Recent investigations in our laboratory have vanadium-nitrogen alloy analyzing about 5.76%N and 0.28%C. Adopting this process a few hundred grams of V-N alloy has been prepared for use as soluble anode in fused salt electrorefining cell to prepare vanadium metal. 10:40 am DISTILLATION MODELING FOR A URANIUM REFINING PROCESS: B.R. Westphal, Argonne National Laboratory, P.O. Box 2528, Idaho Falls, ID 83403 Apart of the spent fuel treatment program at Argonne National Laboratory, a vacuum distillation process is being employed for the recovery of uranium following an electrorefining process. Distillation of a saIt electrolyte, containing a eutectic mixture of lithium and potassium chlorides, from uranium is achieved by a simple batch operation and is termed "cathode processing". The incremental distillation of electrOlyte salt will be modeled by both an equilibrium expression and on a molecular basis since the operation is conducted under moderate vacuum conditions. As processing continues, the two models will be compared and analyzed for correlation with actual operating reSUlts. Possible factors that may contribute to aberrations from the models include impurities at the vapor-liquid boundary, distillate reflux, anomalous pressure gradients, and mass transport phenomena at the evaporating surface. Ultimately, the purpose of either process model will be to enable the parametric optimization of the process. 11:05 am A STUDY OF THE THERMAL DECOMPOSITION OF BaC03 AND SrC03: I. Arvanitidis, D. Sichen and S. Seetharaman: Theoretical Metallurgy, Royal Institute of Technology, S-IOO 44 Stockholm, Sweden In the present work, the following decomposition reaction were studied: BaC03(s) =BaO(s) + C02(g) and SrC03(s) =SrO(s) + C<>2(g). The rate of the thermal decomposition of the carbonates was followed by thermogravimetric analysis (TGA) and differential thermal analysis (DTA) simultaneously during heating in argon. Shallow powder beds and high argon flows were employed. The DTA curves show a single phase transformation for SrC03 from rhombohedral to hexagonal form occurring at 1204 K. In the case of BaC03, two phase transformations could be observed, viz. from orthorhombic to hexagonal at 1079 K and hexagonal to cubic at 1237 K. The mechanisms of the reactions are examined in the light of the kinetic studies. In the case of BaC03 decomposition, BaC03 forms an eutectic melt with the BaO, which is formed as the product. The formation of the melt appears to play an important role in the reaction kinetics. The activation energies for the decomposition of SrC03 and BaC03 were evaluated to be 290 and 305 kllmot., respectively. The results are discussed in the light of the thermodynamic stabilities of alkaline earth carbonates.

MATERIALS SCIENCE OF CONTACTS AND METALLIZATION AND INTERCONNECTS IV

Sponsored by:EMPMD Thin FIlm & Interfaces Committee Program Organizers: Jagdish Narayan, Dept of Matls Sci & Engrg, Box 7916, NC State University, Raleigh, NC 27695; John Sanchez, Advanced Micro Devices, M3 160, PO Box 3453, Sunnyvale, CA 94088 Wednesday, AM February 7,1996

Room: Orange County 5 Location: Anaheim Marriott Hotel

Session Chair: J. Sanchez, Advanced Micro Devices, OneAMD Place, Sunnyvale, CA 94088

8:30 am Invited ALLOYING EFFECTS ON THE ELECTROMIGRATION BEHAVIOR IN NARROW Al INTERCONNECTS: O. Kraft, G. Bauer, E. Arzt, Max-Planck-Institut fur Metallforschung and Institit fur Metallkunde der Universitat Stuttgart, Seestrasse 92, 70174 Stuttgart, Germany AI-based alloys are widely used in ULSI technology as interconnect materials. About I WI.% Cu is commonly added to the Al in conductor lines in order to improve the electromigration resistance. However, a satisfying interpretation of the mechanism by which Cu effects electromigration has not yet been given. Many other alloying additions have been investigated in the past without finding an alloy with further improved properties. In this paper, we present results of detailed analyses of electromigrationinduced damage in lines of different alloys (pure AI, AI-Si-Cu, AI-Cu, and AI-Sc). These investigations reveal that the different additions influence the microstructure, and the mechanical and electronic properties of the material. A damage-based electromigration model is presented which takes these features into account and which is able to predict the electromigration-induced damage and the lifetime as a function of stress conditions. Finally, possible routes for the design of new alloys are outlined. 9:00 am Invited ELECTROMIGRATION BEHAVIOR OF Cu(Sn) TIDN FILM INTERCONNECTS:

1996 TMS ANNUAL MEETING AND EXHIBITION

K .N. Tu, Department of Materials Science and Engineering, University of California,

Los Angeles, CA 90095; K.L. Lee, C. K. Hu, IBM, T. J. Watson Research Center, PO Box 218, Yorktown Heights, NY 10598 In-situ SEM has been used for real time comparison studies of the electromigration behavior of Cu and Cu-O.5 to 2 WI. % Sn alloy thin film interconnects. Drift velocity test structures were powered at current density of 5 x 105 to 2.1 x 106 amp/cm2 over the temperature range of 250 to 450°C. The measured electromigration activation energy forCu and Cu (Sn) alloys are 0.73 eV and 0.95 to 1.25 eV, respectively. The measured resistivity values of the Cu-0.5 wt.% Sn and Cu-I wt. % Sn alloys are 2.4 and 2.9 -crn, respectively. These characteristics show that Cu (Sn) alloys are potentially good candidates for future interconnect metallization where high conductivity and good electromigration resistance are required. 9:30 am Invited STRESS AND ELECTROMIGRATION IN TIDN FILM CONDUCTORS: J. R. Lloyd, Digital Equipment Corporation, 77 Reed Road, Hudson, MA 01749; M. A. Korhonen, Dept. Materials Science and Engineering, Cornell University, Ithaca, NY 14853 Electromigration failure, due to a divergence in the mass flux responding to current flow, is significantly affected by stress and stress gradients. These gradients are caused by electromigration itself and, for instance, by thermal history and the choice of and process used to deposit interlevel dielectric materials and passivation layers. The time to failure can be treated as the time necessary to reach a critical stress or a critical vacancy/atomic concentration to form a void or an extrusion. Whether the process is best described as the diffusion of vacancies, atoms or stress will be discussed. The interaction of electromigration induced and other stresses in the conductor will be explored. 10:00 am BREAK 10:30 am A VIA SHORTING MECHANISM RELATED TO THE ROUGHNESS OF THE METAL LAYERS: 1. Maniaci, N. de Lanerolle, Standard Microsystems Corporation, 35 Marcus Blvd., Hauppauge, NY 11788 Excessive failures due to electrical shorts were experienced in via contacts between the flfSt level metal layers (MI) and the second level metal layers (M2). Scanning electron micrograph studies showed holes in the interlevel dielectric on completed wafers using decorative wet etching technique. Holes were caused by particles and possibly hillocks protruding through the photoresist and being cut off by plasma etch erosion of the resist. High particle counts in the underlying layer in MI exacerbated the problem. If alloying is performed after the interlevel dielectric deposition the via contact resistance was lowered. In this case, profilometric measurements showed a decrease in roughness of the second level metal layer (M2). It is proposed that there is a relationship between increased roughness, and susceptibility to electrical shorts and increased contact resistance. 10:50 am MICROSTRUCTURE MAPPING OF INTERCONNECTS USING ORIENTA· TION IMAGING MICROSCOPY: David J. Dingley, D. P. Field, TSLInc., 226 West 2230 North Provo, UT 84604; H. H. Wills Physics Laboratory, University of Bristol, Bristol BS8 ITL, UK The new technique of Orientation Imaging Microscopy, OIM, provides a detailed description of the variation in crystallography over the surface of single crystal or polycrystalline bulk materials. This technique has been used in the investigation of interconnect lines and contact pad material in a number of aluminum metallizations on silicon oxide/silicon substrates. The specimens were examined in a SEM fitted with the OIM system. The technique has provided information regarding the local texture variations between the different regions of a metallized silicon wafer between contact pad and interconnect line. It could be concluded that significant modification of the microstructure had occurred subsequent to initial metal deposition and that this modification differed between contact pad and interconnect line. 11:10 am THE EFFECTS OF DEPOSITION CONDITIONS ON THE CRYSTALLOGRAPHIC TEXTURE OF SPUTTERED TI, TIN and Al ALLOY TIDN FILMS: Paul R. Besser. J.E. Sanchez, Jr., Advanced Micro Devices, One AMD Place, Sunnyvale, CA 94088; D. Knorr, Materials Engineering Department, Rensselaer Polytechnic Institute, Troy, NY 12180-3590 Abstract not available.

MODELING OF COMPOSITES: PROPERTIES AND PROCESSING SESSION I: Sponsored by: MSD Computer Simulation Committee Program Organizers: S.P. Chen, Los Alamos National Lab., Los Alamos, NM 87545; M.P. Anderson, Exxon Research Center, Clinton Township, Route 22 East, Annandale, NJ 08801 Wednesday, AM February 7,1996

Room: Marquis 1&2 Location: Anaheim Marriott Hotel

Session Chairman: S. P. Chen, Los Alamos National Laboratory, Los Alamos, NM 87545; M. P. Anderson, Exxon Research Center, Clinton Township, Route 22 East, Annandale, NJ 08801

9:00am THE YIELD STRESS OF THE COHERENT NANOLAYER COMPOSITE CuNI:S. I. Rao, P. M. Hazzledine, UES Inc., 4401 Dayton-Xenia Road, Dayton, OH 45432

B10l

There are good experimental and theoretical reasons for believing that epitaxial multilayers have a yield stress which is low both when the layers are ve~ thin ~nd when they are very thick. At intermediate thickness there is a large peak 10 the Yield stress which, in the case of Cu-Ni, occurs at a double-layer thickness of about 20 nm. At thickness below 100m., the material is expected to be fully coherent and therefore to contain large alternating biaxial stresses caused by the strains required to match the lattice parameters of Cu and Ni. In order for the multilayer to yield, dislocations must cross interfaces from regions of biaxial tension to regions of biaxial compression a.nd vice versa. The interfaces between these regions present two forms of obstacle to gbding dislocations, the Koehler (elastic modulus ~smatch) obs~le and a '~hemi~a1' obstacle caused by the fact that neither the stacking fault energtes nor the disloca.tt~n core structures are the same in Cu and Ni. This paper reports the results of atorrusttc simulations of dislocations crossing Cu-Ni interfaces in the presence of coherency stresses. The calculations employ the embedded atom method with suitable Cu, Ni an~ Cu-Ni potentials. The results are used to discuss the variation of the strengths of Cu-NI multilayers with thickness when the layers are very thin and fully coherent. 9:30am FINITE ELEMENT MODELING OF PARTICLE CRACKING IN SIC-REINFORCED ALUMINUM ALLOYS: P. Scarber, Jr., G. M. Janowski, Dept. of Materials Science and Engineering, The University of Alabama at Birmingham, Birmingham, AL 35294-4461 Simulations of an aluminum matrix composite reinforced with 9 vol% SiC particles possessing varying surface defect geometries were performed using a nonlinear axisymmetric finite element model. Preliminary results have shown that the tendency for a reinforced particle to crack is not only dependent upon the existence of surface flaws, but also on the types of flaws. Model results have shown that long slender surface defects do not promote particle cracking during tensile loading. The residual stres~es created during cooling from solutionizing temperatures have been shown to be a major factor in the tendency for a reinforced particle to crack. These results will be analyzed using composite strengthening theory with consideration being given to the effects of matrix strength and particle volume fraction. This research was supported by AlabamaNSF EPSCoR and the Alabama Space Grant Consortium under contract number NGT40010. 10:00 am MODELING OF STATISTICAL TENSILE STRENGTH OF SHORT-FIBER COMPOSITES: Y. T. Zhu, W. R. Blumenthal, M. G. Stout, T. C. Lowe, Materials Science and Technology Division, Los Alamos National Laboratory, Los Alamos, NM 87545 This paper develops a statistical strength theory for thre~-dimentionally (~-D) oriented short-fiber reinforced composites. Short-fiber COmpoSiteS are usually remforced with glass and ceramic short fibers and whiskers. These reinforcements are brittle. and display a range of strength values, which can be statistically characteriz~ by a Welb~1I distrihution. This statistical nature of fiber strength needs to be taken mto account 10 the prediction of composite strength. In this paper, the statisti~a1 nature of fil~r strength is incorporated into the calculation of direct fiber stre~gthemng, and a maxlmum-I~ad composite criterion is adopted to calculate the composite strength. Other strengthemng mechanisms such as residual thermal stress, matrix work hardening, and short-fiber dispersion hardening are also briefly discussed. 10:30 am THE ROLE OF GRAIN BOUNDARY SLIDING AND REINFORCEMENT MORPHOLOGY ON THE CREEP DEFORMATION BEHAVIOR OF DISCONTINOUSLY REINFORCED COMPOSITES: S. B. Biner, Ames Laboratory, Iowa State University, Ames, IA 50011 . . The role of grain boundary sliding behavior on the creep deformatton characte~s­ tics of discontinously reinforced composites is investigated numerically together With the other influencing parameters: reinforced aspect ratio, grain size and interfacial behavior between the reinforcement and the matrix. The results indicate that the stress enhancement factor for the composites is much higher than the one observed for the matrix material and its value increases with increasing reinforcement aspect ratio, reduction in the matrix grain size and sliding interfacial behavior. Experimentally observed higher creep exponent values or stress dependent.creep eXP:Dnent values f?I. the composites could not be explained solely by the mechamsm of &:-"n bou~dary shdm~. This work was supported by USDOE, Office of Basic Energy SCiences, Div. of Matena1s Science under contract no. W-7405-ENG-82. 11:00 am CONSTRAINED METAL DEFORMATION IN AN INTERMETALLIC/METALLIC MICROLAMINATE COMPOSITE: J. Heathcote, G. R. Odette, G. E. Lucas, . Materials Department University of California, Santa Barbara, CA 93106 The stress-displacement function sigma(u), is one of the fundamental properties that detennines mechanical properties in ductile phase reinforced composites. This function is determined for the constrained metal layers of an intermetallic/metallic microlaminate composite by a fracture reconstruction/modeling technique. First, quantitative topographical measurements are made on conjugate fracture surfaces by confocal microscopy. Then, height profiles from the two surfaces are overlaid and sequentially separated to recreate the deformation fracture process. The rec~ation sequc:nce provides a direct assessment of the large scale geometry changes dunng ~onstramed deformation. This information is used to guide the development of a fimte element method (FEM) simulation of constrained deformation to model the sigma(u) function of the metal layers. The model predictions are compared to other independent evaluations of sigrna(u). Knowledge of the sigma(u) function allows for the prediction of resistance curves and fracture strengths for various testing geometries. 11:30 am MODELING OFBRITTLEIBRITTLEAND BRiTTLElDUCTILELAMINATES: THE EFFECTS OF THE INTERFACIAL COHESION: S. P. Chen, Los Alamos National Laboratory, Los Alamos, NM 87545

B102

A spring-network model was used to simulate the deformation and fracture behaviors of brittlelbrittle and brittle/ductile lantinates. The effects of the interfacial cohesion, moduli, grain boundary cohesion, yield stress will be presented and comparted with available experiments.

PHYSICAL CHEMISTRY OF HIGH TEMPERATURE COMPOSITES AND INTERMETALLICS V: CERAMICS AND COMPOSITES

Sponsored by: EPD Process Fundamentals Committee, MSD Thermodynamics & Phase Committee, Japan Institute of Metals Program Organizers: Prof. R.Y. Lin, Ul1iversity of Cincinnati; Prof. Y. Austin Chang, University of Wisconsin-Madison; Prof. R. Reddy, University of Navada-Reno and Dr. C.T. Liu, Oak Ridge NL Wednesday, AM February 7, 1995

Room:B2 Location: Anaheim Convention Center

Session Chairperson: Y. Iguchi, Dept. Metallurgy, Tohoku Univ., Sendai, JAPAN; Michael J. Koczak, Department of Materials Engineering, Drexel University, Philadelphia, PA 19104 8:30 am Invited EFFECTS OF INTERGRANULAR NANODISPERSIONS OF HIGH TEMPERATURE PROPERTIES OF CERAMIC MATRIX COMPOSITES: K. Niihara, T. Ohji, N. Kusunose and K. Akamatsu, The Research Institute of Scientific and Industrial Research, Osaka University,8-1 Mihogaoka, Ibaraki, Osaka 567, Japan Ceramic based nanocomposites, in which nano-sized ceramics and metals were dispersed within matrix grains and/or at grain boundaries, were fabricated by an usual powder metallurgical method. The striking finding in the nanocomposite sys~ems is that the mechanical properties such as fracture toughness, strength and creep resistance are significantly improved by the nano-sized dispersions. Specifically, the creep rates of alumina/silicon carbide, magnesia/silicon carbide and silicon nitride/silicon carbide composites were 3 to 4 orders improved by the nano-sized dispersions. In this presentation, the process-structure-properties reaction will be first described, and finally the roles of intergranular nanodispersions in the creep resistance improvement will be discussed. 8:55 am Invited CMC'S RESEARCH IN EUROPE AND THE FUTURE POTENTIALS OF CMC'S IN INDUSTRY: M.Van de ~borde, European Commission-Joint Research Center, P.O. Box 2, 1755ZG Petten, The Netherlands CMC's have been developed for high Temperature Applications in Aerospace and Military Industries. In general, the CMC's were capable to withstand high temperatures (up to 1700°C) in an inert atmosphere for short durations. Great efforts are being given to pinpoint spin-off technologies i.e. applications in aero engines, modem fossil fuel power plants, gas-turbines, petrochemistry etc. In these applications: the CMC's .~ve to operate at up to 1400°C, in corrosive environments for long duraltons. In addltton, the CMC's process should become cost effective. The development of such CMC's will provoke a breakthrough for this new group of materials. The paper will I) give an overview of the CMC's research in Europe; processing-characterization of physical, chemical and engineering properties at H.T.; 2) pinpoint the R&D needs to achieve the potential growth; 3) evaluate the market potentials; and 4) review the industrial applications. 9:20 am Invited HIGH TEMPERATURE MECHANICAL BEHAVIOR OF MGO-BASED NANOCOMPOSITES: Y. H. Choa, K. Niihara, The Research Institute of Scientific and Industrial Research, Osaka University, 8-1 Mihogaoka, Ibaraki, Osaka 567, Japan MgO/SiC nanocomposites were found to exhibit the excellent high fracture strength and toughness even at high-temperatures . These significant improvements of mechanical properties were contributed to the displ,rsion of nano-sized second phase. In this presentation, the effects of micro-sized SiC particulate dispersions on the fracture strength, toughness, Young's modulus, and effeclive surface energy up to 1400°C were investigated. 9:45am COMPLEX OXIDE FILMS FORMED DURING THERMAL OXIDATION OF LIQUID ALUMINUM ALLOYS: Y.Iguchi, T.Narushima, M.Maruyama, T.Shiozawa Department of Metallurgy, Tohoku University, Sendai 980-77, Japan; H.Arashi, Department of Machine Intelligence and System Engineering, Tohoku University, Sendai 980-77, Japan Oxidation behavior of liquid aluminum alloys is important in the view point of crack formation during continuous casting of aluminum alloys or syntheses of ceramic matrix composites. In the present work, the in situ observation system for thermal oxidation using Raman spectroscopy was constructed, and oxide films formed on liquid pure AI and AI-Mg alloys (Mg contents =1 and 4 mass%) were analyzed at room temperature to 1273 K in dry oxygen with this system. Oxidation films formed on the specimens depended on the composition of alloys. In situ Raman spectra of the surface of a liquid AI-4 mass% Mg indicated the fonnation of MgO$B !&(BAI203 (spinel) as oxidation products, while e$B&A(B-A1203 and MgO$B!&(BAI203 (spinel) were observed on a liquid AI-lmass%Mg alloy. It was suggested that the formation of MgO film affected oxidation behavior of these alloys. The oxidation mechanisms were discussed thennodynamically. 10:10 am BREAK

1996 TMS ANNUAL MEETING AND EXHIBITION

10:20 am INTERNAL OXIDATION OF LIQUID ALUMINUM-MAGNESIUM ALLOYS IN MOIST AIR: Laurence Rault, Laboratoire de Thermodynamique et Physico-Chimie Metallurgiques, ENSEEG, BP 75, Domaine Universitaire, 38402 Saint Martin d'Heres Cedex, FRANCE Reactions occurring during oxidation of liquid aluminum-magnesium alloys inmoist air at 850C are studied by thermogravimetric analysis. In humid air, oxidation rate curves exhibit a maximum at a time and a rate both depending on water vapor pressure. Surface alternations corresponding to the oxidation curves are monitored by an in situ observation device. Experiments show that oxidation takes place inside the liquid metal just under the surface oxide film formed during heating. The reaction with water vapor produces magnesia and hydrogen bubbles in the melt. Rising bubbles create blisters and growths on surface. Spinel is obtained on surface and inside the melt for exposure times longer than the maximum rate time. Consistent explanations are proposed for the observed phenomena and their kinetics. 10:45 am Invited IN-SITU PROCESSING TECHNIQUES FOR FUNCTIONAL GRADED MATERIALS: Michael I. Koczak, Department of Materials Engineering, Drexel University, Philadelphia, PA 19104 It is now widely recognized that Functionally Graded Materials (FGMs) are of interest for structural and electronic applications. The properties of FGMs are controlled by the size and volume fraction of the reinforcement phase, as well as the microstructure and properties of the matrix/reinforcement interface. Traditionally, FGMs have been produced by such processing techniques as powder metallurgy, preform infiltration, spray deposition and various casting technologies, i.e. squeeze casting, rbeocasting and compocasting. In the last decade, new in situ processing, analysis and evaluation technologies have emerged. For processing FGMs, in-situ techniques can be applied and involve a chemical reaction resulting in the formation of very fine thermodynamically stable reinforcing ceramic phases during processing, e.g. DIMOXTM, XDTM, and reactive gas infiltration (ROI) and can be combined with densification techniques, e.g. reactive infiltration, centrifugal casting, SHS rolling approaches. The application of new processing and analysis approaches should low for more cost-effective productions of FGM for wear and structural applications. The author acknowledges from the Office of Naval Research - Materials Division. 11:10 am LAYERED METAL MATRIX COMPOSITES: Li Lu, Dept. Mechanical & Production Engineering, The National University of Singapore, 10 Kent Ridge Crescent, Singapore 0511 Metal matrix composite (MMC) with composition of AI-4.5wt. %Cu reinforced by IOwt.%TIB2 or 15wt.%SiC particulate has been successfully fabricated using powder metallurgy technique. Disk and beam shaped specimens were prepared with (a) layered structure containing ductile AI-4.5wt. %Cu with brittle AI-4.5%CulTIB2 or 4.5wt. %Cu/ SiC layers, and (b) homogenous constituent with composition of AI4.5wt%Cul IOwt.%TIB2. The normal direction of the layers in the layered specimens, being built through three ductile-brittle-ductile alternating layers was parallel to the axis of the disk specimen for the beam specimens, the layers were built through brittle-ductilebrittle sequence. The brittle layer consisted of AI-4.5%Cu with fixed amount of reinforcement. Fracture toughness testing revealed an increase for layered composites although the total amount of the reinforcement was the same for all disk specimens. The increase in the fracture toughness is due to the existence of ductile layers which manifests a strong influence on delaying the occurrence of fracture. It acts as a barrier to resist the propagation of crack and bence increases the fracture toughness of the layered MMC specimens. Delamination in the beam specimens were observed due to shear stresses. Fracture surface of both disk and beam specimens wer characterized using the SEM. 11:35 am PHASE RELATIONS IN THE Sm203-Ce02-Cuo SYSTEM INVOLVING THE N-TYPE SUPERCONDUCTOR Sm2 x Cex CU04 y: Zhivu Qiao, Weidong Zhuang, Shoukun Wei, Lin Cheng, Jingkui Liang. Department of Pllysical Chemistry, University of Science and Technology Beijing, Beijing 100083, China The subsolidus phase relations in the Sffi203-Ce~-CuO system involving the ntype Sffi2-x Cex Cu04-y supercooductorwere investigated by x-ray diffraction at lOOO°C in air. The system can be divided into two ternary phase regions and three binary phase regions. the solid solution Sm2, CexCu04 Y was found to be the only ternary phase, and the maximum solid solubility was confirmed to be x=O.20. The solid solution crystallizes in a tetragonal unit cell, space group I4/mmm. The c value of the solid solution is smaller than that of Sm2Cu04, but the a value is larger than that of Sffi2CU04. The average valence of copper ion of the solid solution decreases linearly with the increase Ce content.

PROCESS MINERALOGY II: Image Analysis

Sponsored by: EPD Process Mineralogy Committee Program Organizer: Dr. W. Petruk, CANMET, Natural Resources Canada, 555 Booth St., Ottawa, Ontario Canada KIAOGI

Wednesday, AM February 7,1996

Room: AI4-15 Location: Anaheim Convention Center

Session Chairperson: William Petruk, CANMET, Department of Natural Resources Canada, 555 Booth St., Ottawa, Ontario, Canada KIA OG I

8:30am TECHNIQUES FOR MEASURING MINERAL LIBERATIONS BY IMAGE ANALYSIS: W. Petrul<, R. Lastra, CANMET, Department of Natural Resources Canada, 555 Booth St., Ottawa, Ontario, Canada KIA OG I Several image analysis techniques are used to measure mineral liberations observed in polished sections of crushed ores. The techniques were compared to determine whether the same results are obtained by each, and to establish a base for developing a standard technique for measuring mineral liberations. The simplest and least valuable technique involves separating images of liberated grains from images of unliberated grains and measuring areal fractions coverd by grains in each occurrence. A second technique involves measuring the areal fraction of mineral in each particle and classifying the particles according to areal fraction of mineral in the particle. A third technique insoles measuring the areal fractions of all minerals in each particle and calculating (using standard S.G.) the weight fractions of the minerals in the particles. The resulting data are classified in the same manner as for the second technique, but can be readily correlated to assays. 9:00am MULTISPECTRALIMAGE BASED APPROACH APPLIED TO INDUSTRIAL ORE MODELLING: MINERALS IDENTIFICATION AND LIBERATION: G. BoniJazi, Dipartimentu di Ingegneria Chimica, dei Materiali, delle Materie Prime e Metallurgia, Universita' Degli Studi di Roma "La Sapienza", Via Eudossiana, 18-00184 Roma, Italy The possibility of recognizing different mineral species in an ore by an imaging multispectral approach is examined. The procedure involves acquiring microscopical optical images from polished andlor thin sections and applying new segmentation strategies to correctly identify mineral species which are not easily identified by classical B&W image analysis techniques. The main aim of this paper is to introduce some of the new techniques and to stimulate the use of such procedures in the field of applied mineralogy and mineral processing. The adoption of these procedures, would in many cases contribute to better identification of the minerals, permit a better and more accurate quantitative characterization of mineralogical textures and consequent better set up of the further beneficiation strategies to apply to the ore. 9:30am A DIGITAL MULTISPECTRAL APPROACH APPLIED TO CHARACTERIZE A SWEDISH WOLLASTONITE ORE AND THE RESULTING PRODUCTS OF BENEFICIATION: Girolamo Belardi, Roberto Gorga, Istituto per iI Trattamento dei Minerali Consiglio Nazionale delle Ricerche, Via Bolognola, 7-00138 Roma, Italy; Guiseppe Bonifazi, Dipartimento di Ingegneria Chimica, dei Materiali, delle Materie Prime e Metallurgia, Universita' Degli Studi di Roma "La Sapienza", Via Eudossiana, 18-00184 Roma, Italy Minerographic and petrochemical research was carried out on a wollastonite ore from Sweden to establish a correlation between ore mineralogy and petrographic characteristics and the resulting products of beneficiation. A structural and textural characterization of the ore was numerically developed, with emphasis on morphological and morphometric features of the products according to their mineral compositions. All the procedures tested and developed were multispectral digital image based. Reflected and transmitted light microscopy images were utilized. The results can be applied to assess future strategies for designing beneficiation processes, involving similar ores. 10:00 am A COMPUTER DESIGN APPROACH TO TOPOLOGICALLY CHARACTER· IZEANDMODELAHYDROTHERMALSULPHIDEVEINDEPOSIT(JALES, VILA REAL, NORTHERN PORTUGAL): G. Bonifazi, Dipartimento di Ingegneria Chimica, dei Materiali, delle Materie Prime e Metallurgia, Universita' Degli Studi di Roma "La Sapienza", Via Eudossiana, 18-00184 Roma, Italy; S. Serranti, Dipartimento di Scienze della terra, Universita' Degli Studi di Roma "La Sapienza", P. Ie Aldo Moro 5,00185 Roma, Italy A digital procedure has been developed and discussed in order to identify, recognize and characterize mineral associations and textural properties of a sulphide deposit containing electrum and gold. The deposit is located in the vila Real region in northern Portugal. A computerized full color image analysis system was used to define structural and textural characteristics of the ore, with particular reference to electrum distribution. The procedures were based on a combined use of reflected and transmitted light microscopy techniques. All the analyses and determinations were carried out in order to obtain detailed information for the further beneficiation processes.

RADIATION MATERIALS SCIENCE IN TECHNOLOGY APPLICATIONS III: Embrittlement of Steels and Amorphization of Compounds

Sponsored by: Jt. SMDIMSD Nuclear Materials Committee Program Organizers: L. K. Mansur, Metals and Ceramics Division, Oak Ridge Na-

tional Laboratory, PO Box 2008, Oak Ridge, TN 37831-6376; C. L. Snead, Jr., Applied Technologies Division, Brookhaven National Laboratory, PO Box 5000, Upton, NY 11973-5000 Wednesday, AM February 7, 1996

Room: Grand H Location: Anaheim Marriott Hotel

Session Chairperson: C. Lewis Snead, Jr., Applied Technologies Division, Brookhaven National Laboratory, PO Box 5000, Upton, NY 11973-5000

1996 TMS ANNUAL MEETING AND EXHIBITION

B103

8:30 am Invited STRENGTH SUPERPOSITION IN STRUCTURALALLOYS: IMPLICATIONS TO THE IRRADIATION HARDENING AND ANNEALING OF PRESSURE VESSEL STEELS: G. Tedesky. G. R. Odene. G. E. Lucas. Departments of Mechanical Engineering and Materials. University of California. Santa Barbara. CA 93106 A model describing how various individual contributions «Ji. i = 1.2•... ) combine to produce the net yield stress ((In) is described. Typically. n contains both linear sum LS: (In = (JI + (J2 + ... ) and rootsumofthe squares (RSS: (In =vaI2+(J22 +.... ) as well as intermediate contributions. For discrete dislocation pinning points (PP). both (Ji and their combined strength contribution depend on the PP number densities (N). diameters (d) and strength factors (~). The ~ directly govern the maximum force per PP and indirectly influence the effective PP spacing as mediated by N. d and the shapes of the stressed dislocation lines. Computer simulations of dislocation motion through random point arrays of PP with varying ratios (NWIN s) of a range of weak (~w) and strong (~s) obstacles extends earlier work of Foreman and Mald!!.,1ielding an general expression S = «Jnsirn-(Jnrss) I «Jnls - (Jnrss)V~s- -(A - B~s)(~~"'- C - DNwIN s where A to D are fit constants (note. the best fit values depend on the application range of the parameters and weighting of the simulation data). Model predictions are in good agreement with increases of (In in reactor pressure vessel steels caused by small copper rich precipitates (CRP) that form under irradiation. The as-irradiated CRP are weak PP and combine with a pre-irradiation population of strong carbide PP with as> O. The model also predicts the large annealing (A) recovery observed in irradiated steels; as the CRP coarsen. their ~ increases. hence. S decreases. approaching 0 at large sizes. Thus. decreases in (In under PIA are much larger than the corresponding recovery of the isolated CRP strength increment alone. Supported by the Office of Fusion Energy. U.S. Department of Energy Grant: E-FG03-94ER54275. 9:00am THE EFFECT OF SOLUTES ON DEFECT DISTRmUTIONS IN ION IRRADIATED MODEL LWR PRESSURE VESSEL STEELS: P. M. Rice. R. E. Stoller. Oak Ridge National Laboratory. Oak Ridge. TN 37831-6376 A series of nine model LWR pressure vessel steels were ion irradiated at -300·C using 2.5 MeV He ions. to a dose of 1.4 x l()l7 ionlcm2. which corresponds to -0.1 dpa at a depth of 211 and -3.5 dpa at the peak damage region which occurs at about 411 deep. The resultant changes in hardness as a function of depth were measured using a nanoindenter. TEM is being used to measure defect distributions down to -1 nm. The effects of the various solutes. Cu and N in particular. but C and Mn as well. on the number and size distribution of the defect clusters caused by the ion irradiation will be discussed. Research sponsored by the Division of Materials Science. U.S. Department of Energy and the Office of Nuclear Regulatory Research. U.S. Nuclear Regulatory Commission under inter-agency agreement DOE I 886-8109-8L with the U.S. Department of Energy under contract DEAC05-840R21400 with Lockheed Martin Energy Systems and by an appointment to the Oak Ridge National Laboratory Postdoctoral Research Program administered by the Oak Ridge Institute for Science and Education. 9:10am EVOLUTION OF FINE SCALE PRECIPITATES IN IRRADIATED PRESSURE VESSEL STEELS: G. R. Odene. B. L. Chao. B. Wirth. Departments of Mechanical Engineering and Materials. Universil)' of California. Santa Barbara. CA 93106 Formation of a high density (» l(j23/m3) of fine scale (-1 nm) precipitates leads to irradiation hardening and embrittlement of light-water reactor pressure vessel (RPV) steels. Copper rapidly precipitates from supersaturated solutions at around 300·C as a consequence of radiation enhanced diffusion. The precipitates are typically copper rich (CRP) but contain manganese and nickel; however. at high alloy contents of these elements the precipitates are manganese and nickel rich (MNP). Detailed thermodynamic and kinetic precipitation model predictions are compared to small angle neutron scattering data. The quaternary thermodynamiC model includes the effect of the interface. which is significant at small sizes. Kinetics models. are based on either: I) a full treatment of overlapping regimes of nucleation. growth and coarsening kinetics by integration of the full set of clustering equations; or 2) a simpler treatment of coupled timedependent nucleation and multigroup growth and coarsening. In both cases. copper clustering is assumed to be rate controlling. While involving a minimum number of adjustable parameters. model predictions are shown to be in remarkable agreement with experiment. particularly at higher copper contents. Preliminary results of atomistic lattice Monte Carlo simulations that mitigate imitations of the model. including the assumption of a discrete interfaces and neglect of the potential role of cascades assisted nucleation are discussed. Supported by the U.S. Nuclear Regulatory Commission Contract: NRC-04-94-049. 9:40am A MODEL FOR POST IRRADIATION ANNEALING PRESSURE VESSEL STEELS: E. Mader. G. R. Odette. Departments of Mechanical Engineering and Materials. University of Califomia. Santa Barbara. CA 93106 Irradiation embrittlement of reactor vessels can be reduced by post irradiation annealing (PIA). Hardening and embrittlement results from a high density of fine nm scale irradiation induced features including: a) Cu (Mn-Ni) rich precipitates (CRP); b) small defect clusters that are unstable (UMD) and larger clusters that are stable (SMD) under irradiation; and c) a variety of other possible precipitates. A detailed thermodynamic-kinetic model for the evolution of these features is presented and compared to both small angle neutron scattering and microhardness recovery data. The CRP recovers by diffusion controlled precipitate dissolution (mostly Mn and Ni) and coarsening at rates are about 10 to 100 times faster than predicted using extrapolated diffusion coefficients. The UMD and SMD are modeled as a size distribution of vacancy clusters. They dissolve during PIA at rates controlled by self-diffusion and the Gibbs-Thompson effect. Recovery of hardness is based on a detailed treatment of the superposition of contributions from the irradiation features with pre-existing sources of the alloy strength. The model is in good agreement with the a variety of observations. including a recent semiempirical correlation developed for the data base PIA of transition temperature shifts. Detailed understating of the fate of the key alloy constituents and structure-property relations can be used to optimize PIA treatments. For example. treat-

B104

ments that result in significant recovery while leaving copper in solution would suffer greater re-embrittlement rates compared to those resulting in copper sequestered in the large precipitates. Supported by the U.S. Nuclear Regulatory Commission Contract: NRC-04-94-049. 10:00 am BREAK 10:20 am Invited AMORPHIZATION OFINTERMETALLIC COMPOUNDS UNDER IRRADIATION-A REVIEW: Arthur T. MollO. Department of Nuclear Engineering. The Pennsylvania State University. University Park. PA 16802 A review is presented of the field of irradiation-induced amorphization of interrnetallic compounds. with special attention to the kinetic aspects of the transformation. A fun update will be given of recent experimental results using in-situ particle irradiation showing the effects of dose rate. tempernture. crystal orientation. electron energy and the presence of stacking faults. Results firom neutron. ion and electron irradiation will be reviewed. in the context of a kinetic description of the amorphization process. where the rate-limiting step is the accumulation of enough radiation damage in the lattice when opposed by thermal annealing. It will be shown how stability criteria. thermodynamic or otherwise. can be combined with kinetics of radiation damage and annealing to provide an overall description of the amorphization process. and how the experimentally measured quantities of the critical dose and the critical temperature fit in the model. Finany a comprehensive approach is proposed where a combination of computer simulation. in-situ irradiations and theoretical modeling can be used to glean a deeper understanding of the amorphization procesl> under irradiation. 10:50 am Invited AMORPHIZATION OF UJSI AND UJSI2 BY ION OR NEUTRON IRRADIATION: R. C. Birtcher. J. W. Richardson. Jr.• M. H. Muener. Materials Science Division. Argonne National Laboratory. Argonne. IL 60439 Electron diffraction (at ANL-HVEM National User Facility) during in situ 1.5 Me V Kr ion irradiation was used to determine lhe temperature dependence of amorphization doses for both U3Si and U3Si2. Neutron diffraction (at ANL-IPNS National User Facility) was used to fonow crystallographic changes produced by neutron irradiation at 3O·C. Both irradiations result in direct anlorphization. The temperature limits for complete amorphization are 280°C for U3Si ailld 250·C for3 U2Si . Lattice strains from the localized amorphous zones develop in both materials at the same rate. however the total dilation upon amorphization is +2.3% in U3Si and -2.2% in U3Si2. At high doses. plastic flow in the amorphous volume fraction of U3Si relieves strain in the remaining crystalline volume fraction. The floVi rate of U3Si2 is very small. Complete amorphization occurs between 0.88 and 1.1 3 x \0 17 fissions/cm 3 or between 0.29 to 0.38 dpa. This work supported by the U.s. Department of Energy. BES-Materials Sciences. under Contract W-31-109-Eng-38. 11:10 am CRYSTAL AND MICROSTRUCTURE OF UJSI2 FUEL PLATES AFTER REACTOR IRRADIATION: G. L Hofman. J. W. Richardson. Jr.• R. C. Birtcher. Argonne National Laboratory. Argonne. IL 60439 Neutron diffraction (at ANL-IPNS National User Facility) and SEM have been used to determine irradiation induced changes in reactor fuel plates containing U3Si2. The fuel plates were fabricated by hot rolling low enriched U3Si2 (45 volume %) and Al powders. Plates were irradiated at 150"C in the ORR reactor to 7% and 16% bumup of the uranium. Amorphization of U3Si2 by 4 x \0-6 burnup results in a 3.3% volume contraction. SEM exantination shows the development of a severalj.lm thick shell around each fuel particle due to Al diffusion into the particles. The shell thickness varies as the square root of the irradiation dose. Neutron diffraction shows that after the lower dose the fuel is amorphous but that after the higher dose it also contains a crystalline aluminide phase. The stability of U3Si2 fuel plates is believed to be associated with the volume contraction upon amorphization and the :subsequent formation of the aluminide shen. This work supported by the U.S. Department of Energy. BES-Materials Sciences. under Contract W-31-109-Eng-38.

STRUCTURE AND MORPHOLOGY OF EPITAXIAL THIN FILMS SESSION V: Steps and Adatoms

Sponsored by: EMPMD Thin FUms & Interfaces Committee Program Organizer: Dr. David E. Jesson. Oak Ridge National Laboratory. PO Box 2008. Oak Ridge. TN 37831-6030 Wednesday. AM February 7. 1996

Room: Orange County 4 Location: Anaheim Marriott Hotel

Session Chairperson: J. Tersoff. IBM T. J. Watson Center. PO Box 218. Yorktown Heights. NY 10598

8:30 am Invited ATOMIC EVENTS IN EPITAXIAL GROWTH AT LOW TEMPERATURES: Gert Ehrlich. Materials Research Laboratory and Department of Materials Science and Engineering. University of I11inois at Urbana-Champaign. Urbana. IL 61801 Using the field ion microscope. it is IIOW routinely possible to examine individual atomic events that dictate the morphology of epitaxial films. At low temperatures. transport of atoms over crystal facets. diffusion on top of surface clusters. and transport between adjacent crystal layers are decisive in determining whether growth is three-dimensional or layer-by-Iayer. Direct observations of these atomic processes on c10sepacked metal surfaces win be reviewed. with special emphasis on the role of tran-

1996 TMS ANNUAL MEETING AND EXHIBITION

sient events. Supported by the Department of Energy under Grant DEFG02-9IER-4539. 9:00 am Invited DIRECT MEASUREMENTS OF THE KINETICS OF SIADDIMERS ON SI(OOI) USING ATOM-TRACKING SCANNING TUNNELING MICROSCOPY: B. S. Swartzentruber, Sandia National Laboratories, Albuquerque, NM 87185-1413 The rotation and diffusion of Si addimers on the Si(OO I) surface at temperatures between room temperature and 128°C is measured using a novel atom tracking technique. The atom tracker employs lateral-positioning feedback to lock the scanning tunneling microscope (STM) probe tip into position above selected atoms with sub-Angstrom precision. Once locked the S1M tracks subtle changes in the addimer configuration as well as their position as they migrate over the crystal surface. The addimers perform random walks on tup of the substrate dimer rows and are reflected by defects. The average hopping rate increases from 0.001 sec-I at 25°C tu 10 sec-I at 128°C yielding an activation barrier for diffusion of 0.94 eV. This work performed at Sandia National Laboratories is supported by the U. S. Department of Energy under contract DE-AC0494AL85000. 9:30 am Invited ADATOM CONCENTRATION DURING GROWTH OF GaAs: Brad Orr, Mark Johnson, Dept. of Physics, The University of Michigan, Ann Arbor, MI 48109-1120 We present a study of the concentration of adatoms on GaAs(OOI) during MBE annealing. By rapidly cooling the sample from typical growth temperatures and typical As overpressures, the thermal concentration of Ge adatoms can be frozen into small islands on the terraces. The area of the resulting islands which are far from terrace steps is measured with STM giving an estimate of the concentration of adatoms during equilibrium. We find a large concentration of adatoms (0.18 monolayer coverage when cooled from 600°C) is present for typical growth temperatures. Interestingly, Monte-Carlo simulations which are successful at predicting growth dynamics show less than 1120 of the adatom concentration as measured with S1M. 10:00 am BREAK 10:10 am Invited EPITAXY AND STRESS AT SURFACE DEFECfS: J. M. Gibson, R. D. Twesten, 0, Pobland, University ofIllinois, Urhana, IL; D. Loretto, Lawrence Berkeley National Laboratory, Berkeley, CA We use an ultra-high vacuum transmission electron microscope to study in-situ deposited thin films on clean surfaces using transmission electron microscopy. Samples are observed in the plan-view geometry. Of particular interest is the observation of long-range stress fields originating from surface defects such as steps and domain boundaries in reconstructed surfaces. These originate from fundamental surface stresses, which have been measured by image matching. When layers of Ge are deposited on such surfaces, using molecular beam epitaxy, the step stress field is radically altered. This alteration can be useful in identifying the nature of the interfacial and surface steps for epitaxial films. In addition, we consider the significant impact of these long-range stress fields on the epitaxial growth process. 10:50 am Invited DYNAMICS OF GROWTH AND ETCH STRUCTURES ON CRYSTAL SURFACES: Georg Rosenfeld, Karina Morgenstern, George Comsa, Institut fur Grenztlachenforschung und Vakuumphysik, KFAlForschungszentrum Julich GmbH, D-52425 Julich, Germany The dynamics of growth or etch morphologies created by deposition on or removal from crystal surfaces are discussed and illustrated by experimental examples including motion and decay of islands and pits on Ag(I II) studied by high-speed STM. In particular, it will be shown that at room temperature, vacancy islands on Ag( III) perform a random walk according to the laws of Brownian motion of free particles. The diffusion coefficient of the vacancy islands scales with the inverse of the island area, indicating that the microscopiC mechanism behind the motion is adatom diffusion across the vacancy island rather than along the island boundary. This interpretation is confirmed by the fact, that at the same temperature, adatom islands decay via evaporation of adatoms. The decay rate of adatom islands is studied in detail and is found to increase with decreasing island size. However, the time dependence of the island decay does not follow a simple power law. The possibility of estimating characteristic parameters like diffusion coefficients, step edge barriers, and step free energies from these experiments is examined. l1:lO am Invited STEP DYNAMICS OF THE RELAXATION OF FABRICATED STRUCTURES ON SI(111): M. D. Johnson, Department of Physics, University of Maryland, College Park, Maryland 20742 The relaxation of step bunched Si( III) surfaces has been investigated using scanning tunneling microscopy (STM). Initially, meta-stable step bunch regions containing approximately 10 steps separated by 2 m terraces are prepared by direct current heating. Subsequent heating by alternating current, direct current in the step up direction, or electron emission heating at 900°C allowed the surface to relax slowly to evenly spaced steps. The time evolution of this process was characterized by quenching the surface at selected points during the relaxation and imaging with S1M. These results are compared with several models based on previous results of step-step interactions on Si( III) in order to better understand the microscopic mechanism primarily responsible for the relaxation. The models, which should be generally applicable to a variety of materials, can be used to predict the relaxation rate of artificially created structures such as quantum dots and devices created with selected area epitaxy. 11:50 am MEASUREMENT OF INTRA- AND INTERLAYER DIFFUSION BARRIERS IN METAL EPITAXY AND THEIR SENSITIVITY TO STRAIN: H. Brone, K. Bromann, K. Kern, Institut de Physique Experimentale, EPF Lausanne, CH-IOI5 Lausanne, Switzerland

1996 TMS ANNUAL MEETING AND EXHIBITION

A general concept is shown how to measure both, the barriers for terrace and step-down diffusion for an epitaxial system with high accuracy by application of mean-field nucleation theory to variable temperature STM data. Terrace diffusion is studied via the variation of saturation island density with temperature at a critical nucleus size of one. The additional barrier for an adatom to descend the step edge (first discovered by Ehrlich and Hudda, as well as by Schwoebel in 1966) can be determined through measurement of the nucleation rate on top of islands as a function of island size and temperature. This approach is applied to the growth of Ag on the (III) surfaces of Ag and Pt. The first Ag layer on PI( III) grows pseudomorphic and is thus under a considerable compressive strain of 4.2%. This compression drastically lowers, first, the barrier for terrace diffusion (60 meV, compared to 97 meV for Ag/Ag(lll) and 157 meV for Ag/Pt(I11», and second, the additional step-edge barrier for interlayer diffusion from Es=(12015) meV for Ag(III) homoepitaxy to Es=(305) meV for jump down from the frrst Ag layer on Pl(111). These examples illustrate the effect of the layer dependent strain on the intra and interlayer masstransport and thus on the film morphology in the kinetic growth regime. Potential origins of this pronounced effect will be discussed. t:z:l0 pm RATE AND DIFFUSION EQUATION ANALYSES OF MASK DEPOSITION PROCESSES: J. A. Venables, School of MaPS, University of Sussex, Brighton BNI 9QH, UK; A. Sugawara, Box 871504, Department of Physics and Astronomy, Arizona State University, Tempe, AZ 85287-1504 In many low temperature deposition processes, the diffusion length of adatoms on the surface before evaporation can approach macroscopic dimensions. In this case the actual adatom diffusion length is determined by competing surface processes, such as nucleation of islands, interaction with steps, or surface reactions such as segregation or interdiffusion. Experiments can be performed which use a deposition mask to define the area where atoms are deposited; the mask width is a constant length, which can be compared with the other temperature-dependent lengths. Examples will be given of recent UHV-SEM experiments involving deposition through 20-100 micron masks, where temperature dependent diffusion lengths are observed. The present state of the analysis will be described, and activation energies for specific surface processes extracted. The scope for mask deposition experiments at higher spatial resolution will he discussed.

SYNTHESIS AND PROCESSING OF NANOCRYSTALLINE POWDER IV: Sintering and Densiflcation Sponsored by: MDMD Powder Metallurgy Committee and FEMS (The Federation of European Materials Societies) Program Organizers: Dr. David L. Bourell, The University of Texas at Austin, Materials Science & Engineering, MC C2201, Austin TX 78712; Dr. Liisa Kuhn-Spearing, Laboratory for the Study of Skeletal Disorders and Rehabilitation, Harvard Medical School, Children's Hospital, 300 Longwood Avenue, Boston MA 02115; Professor Dr. Herbert Gleiter, Karlsruhe Research Center, P.O. Box 3640, 0-76021 Karlsruhe, Federal Republic of Germany Wednesday, AM February 7, 1996

Room: Grand G Location: Anaheim Marriott Hotel

Session Chairperson: B.T. Fultz, Mail 138-78, California Institute of Technology, Pasadena CA 91125

8:30am SYNTHESIS, PROCESSING AND MECHANICAL PROPERTIES OF NANOCRYSTALLINE MATERIALS PRODUCED BY SOLUTION PHASE SYNTHESIS: Shankar M.L Sastry, W.E. Bubro, Washington University, St. Louis M063130 Solution phase synthesis (SPS) is an extremely promising approach for producing nanometer-scale particles of metals, intermetallics and ceramics. In this method, solution-phase chemical reactions are conducted at low temperatures « 200°C) in solution to form solid precursors which are subsequently annealed in the solid state at temperatures of 500 - lOOO°C, to produce by nucleation and growth 10 - 100 nm crystallites. We have synthesized by SPS nanocrystallites of Cu, TIB2, ZrB2, TIC, TIN, MoSi2, Ni3AI, NiAl, TIAI, and TIAI3. The consolidation characteristics, room and elevated temperature strength, high temperature creep and room temperature fracture toughness of several of the above nanocrystalline materials will be discussed. This research was conducted under NSF Grant No. ECS-9119006 and AFOSR Grant No. F 49620-93-10131. 8:50am DENSIFICATION OF NANOCRYSTALLINE POWDER PRODUCED BY SOLUTIONPHASE SYNTHESIS: THEORETICAL MODELING AND COMPARISON WITH EXPERIMENTS: R. Suryanarayanan, S.M.L Sastry, Department of Mechanical Engineering. T.1. Trentler, B.E. Waller, W.E. Bubro, Department of Chemistry, Washington University, St. Louis MO 63130 A theoretical model for the densification of nanoparticles by hot pressing and/or hot isostatic pressing is developed. Based on the Helle-Easterling-Ashby model for hot isostatic pressing, the present model attempts to take into account the effect of agglomeration, surface imporities, and particle size distribution in nanoparticle systems. Precision density measurements are performed on nanocrystalline copper and nanocrystalline molybdenum disilicide, that have been compacted by a combination of hot pressing or hot isostatic pressing. Measurements are performed on compacts that have been pr<>duced with and without exposure to air during precompaction powder handling. The

B105

model results are compared with the experimental data. The model is used to predict the dominant densification mechanism and these results are compared with experiments by performing microstructural characterization of the nanocompacts. Preliminary results show that impurities on particle surfaces lead to dramatic decrease in densification rates, requiring greater temperatures and/or pressures to achieve full densification. The results of this model may partially explain why attempts to densify nanoparticle systems by several researchers have been unsuccessful, in-spite of expectations of high densification rates due to orders of magnitude increase in diffusivities.

erties, especially hardness - fracture toughness relationship, reduction of the grain size to the nano-scale is suggested. The resellrch concentrates on investigating various synthesis processes of nanocrystalline WC and selecting the optimal process with respect to grain size, microstructure and composition. A novel approach is the investigation of sintering kinetics by means of dilatomeuy. The study consists of the characterization of the microstructore, the composition and the mechanical properties; the major goal is to understand the microstructure to the finest details, which in turn is expected to lead toward improvement of the hardness and plasticity of the final products.

9:15am HIGH PRESSURE CONSOLIDATION OF 'NANO-NANO' ALUMINA COMPOSITES: R.S. Mishra, Department of Chemical Engineering and Materials Science. C.E. Lesher, Department of Geology. A.K. Mukherjee, Department of Chemical Engineering and Materials Science, University of California, Davis CA 9S616 Fully dense 'nano-nano' alumina composites have been obtained by sintering at a nominal pressure of 1 GPa in the temperature range of 1073-1223 K. The temperature of sintering depends on the volume fraction and chemistry of the second phase. Results are presented on A1203-SiC and A1203-TI02 composites. Application of high pressure makes it possible to achieve 'nano-nano' structure, where both phases are nanocrystalline (the reported literature results for Al203-SiC show large AI203 grains with intragranular nanocrystalline SiC). The 'nano-nano' alumina composites show high toughness and hardness (>26 GPa). The sintering kinetics agree with the theoretical prediction of densification by dislocation creep mechanism. The implication of the present results on the diffusional densification rate is discussed.

11:15 am GRAIN SIZE STABILIZATION PRODUCED BY DISPERSED FULLERENES: E. V. Barrera, J. Sims, D.L. Callahan, Department of Mechanical Engineering and Materials Science, Rice University, Houston, TX 77251-1892 Recent research on processing coppc:r, tin, aluminum, iron, and stainless steel with fullerenes has focused on several issues: dispersing fullerenes on a nanometer size molecular level both in and at grain boundaries in polycrystalline materials, maintaining their stability throughout the processilng and subsequent anneals, and verifying their contribution to strengthening. Nanocrystalline copper and stainless steel have been processed with fullerene additives where well dispersed fullerenes contributed to the control of the starting grain size and to their stability. Fullerenes have been shown to effectively stabilize grain size and are currently being studied as a phase stabilizing agent as well. Films of co-deposited aluminum and fullerenes have been studied by transmission electron spectroscopy where nanometer size particles have been observed in grains and less frequently at grain boundaries where grain size has also been stabilized. Discussion will be given as to the role of thl: fullerenes in metals and to the advantages in using fullerenes for mechanical property enhancement. Support has been provided by the National Science Foundation, under grant number DMR-93S7S0S.

9:35am PROCESSING OF NANOCRYSTALLINE ZIRCONIA AND ZIRCONIA-ALUMINA COMPOSITE POWDER BY HOT PRESSING AND HIP: G. Prabhu, D.L. Bourell, Center for Materials Science and Engineering, University of Texas at Austin, MC C22OO, Austin, TX 78712 Three mol% Yttria Stabilized Tetragonal Zirconia (YSTZ) and a YSTZ - alumina nanocomposite were synthesized using a co-precipitation technique using chloride precursors. The effect of calcination on both phase characteristics and particle size was assessed. Xray diffraction studies revealed that crystallization ofYSTZ began at SSO'C. The tetragonal phase appears and is retained after sintering to 122S'C. The nanocomposite had to be calcined at lOOO'C to observe any crystallinity. Alumina peaks were observed only after calcination at 14OO·C. As TGA showed that the attached water of hydration in these ceramics was lost at SSO'C, the powders were calcined at this temperature and used as starting material in the sintering studies. The powders used had particle sizes of approximately 10 to IS nm. YSTZ sintered to greater than 99% theoretical density at temperatures higher than 122S'C while the nanocomposite sintered to similar density at 13OO·C. Powders were also cold pressed and hot isostatically pressed (HIP) using argon as the pressurizing medium. Sintering parameters to minimize coarsening and maximize density were established. 10:00 am A MICROSTRUCTURAL STUDY OF CONSOLIDATION OF FE AND CU NANOPOWDERS: O. Dominguez, Y. Champion, J. Bigot, CECM-CNRS, IS rue Georges Urbain, 94407 Vitry sur Seine, France Previous studies of compacted nanocrystalline metal powders suggest that for certain metals, high green density cannot be obtained even at high consolidation pressures due to their intrinsic high yield strength. This study concerns the different behaviour of nanometric powders as a function of their crystal structure and of the nature of the compaction process. l\vo initial nanometric powders of Fe and Co with equivalent particle size distributions, were prepared by levitation melting in liquid nitrogen. TEM analysis of the as-prepared nanopowders shows, for Cu, planar defects, thought to be twins, in a substantial fraction of the particles, whilst the iron particles are totally free of any defect. A comparative analysis of the consolidation process of nanometric Fe and Cu powders, using die-compaction and isostatic pressing, shows a remarkable difference in density and calculated yield strength between the two bulk specimens. The evolution of the defects for both nanopowders was studied as a function of pressure. We discuss these changes in relation to the observed consolidation and mechanical behaviour. 10:25 am SINTERING OF NANOCRYSTALLINE POWDERS: EXPERIMENTS AND COMPUTER SIMULATIONS: R. S. Averback, J.M. Gibson, D. Olynick, R. Tao, H. Zhu. Department of Materials Science and Engineering, University of Illinois at UrbanaChampaign, Urbana, IL 61801 Sintering of nanocrystalline materials has been investigated at three levels of observations. On the atomistic scale, molecular dynamics computer simulations have been used to study the initial stages of sintering when two or more nano-particles first come into contact. They show that because of the high shear stress developed in the nanoparticle contacts, the particles undergo extensive densification in matters of picoseconds by dislocation motion. In situ electron microscopy of similarly small assemblies of particles has enabled us to follow the sintering to longer times and at somewhat larger length scales and observe the role of surface diffusion and particle contamination. Finally, measurements of densification on bulk specimens undergoing shear deformation provide information on the macroscopic behavior of sintering involving large assemblies of nanocrystals. Attempts will be made to relate the macroscopic sintering behavior of nanocrystals to the microscopic processes elucidated by the simulation and TEM. 10:50 am THE RELATIONSHIP BETWEEN MICROSTRUCTURE AND MECHANICAL PROPERTIES OFNANOCRYSTALLINE WC BASED HARD METALS: R. Porat, S. Berger, A. Rosen, Technion, Israel Institute of Technology, Haifa 32000, Israel WC belongs to the family of cemented carbides, widely used in industry due to unique combination of mechanical, physical, and chemical properties. To improve prop-

B106

11:35 am A MODEL FOR FULLERENE GROWTH MECHANISM: H. Yoshida, National Institute for Advanced Interdisciplinary Research, 1-1-4 Higashi, Tsukuba, Ibaraki 3OS, Japan A model for fullerene growth is proposed that describes stochastical formation of its five and six membered rings. For the model, the following is assumed: (1) The intermediates consist of only pentagons and hexagons; (2) The isolated pentagon rule (IPR) is applied; (3) Structore transformations are forbidden; (4) All pentagons and hexagons adjacent to the intermediate form at once. The expressions of creation probability of C6() and C70, P(C60) and P(C70), with P are obtained; P is the mean probability of sticking of five membered ring to intermediate. P(C60) is formed to be a monotonically increasing function of P tor OPSI. The limit as P goes to 1 of the quantity P(C60) is equal to 0.5, while P(C60)=O for P=O. P(C60) of P has a singUlarity at P=1 that is caused by IPR. P(C70) has a peak atabout P=O.9, and P(C70)1 ~ for P~ and 1. P(C70) has no singularity at P=1. For P(C70)IP(C60)=O.l; the present model gives P=O.7. This means that pentagon ill more easily created than hexagon except for the restriction on the creation of pentagon owing to IPR.

TANTALUM: Applications: Session V

Sponsored by: SMD Refractory Metals and Materials Committee and Jt. MDMDI EPD Synthesis, Control and Analysis in Materials Processing Program Organizers: Andrew Crowson, U.S. Army Research Office, Research Triangle Park, NC; Edward S. Chen, U.S. Army Research Office, Research Triangle Park, NC; Prabhat Kumar, Cabot Corp, Boyertown, PA; Willam Ebihara, Picatinny Arsenal, Picatinny, NJ; Enrique J. Lavernia, UC Irvine, Irvine, CA Wednesday, AM February 7, 1996

Room:A4-S Location: Anaheim Convention Center

Session Chairpersons: Robert Balliet, H.C. Stark Inc., Newton, MA; George J. Korinek, Tantalum-Niobium International Study Center, Brussels, Belgium

8:30am EVOLUTION OF APPLICATIONS O]FTANTALUM: Prabhat Kumar, Cabot Performance Materials, County Line Road, Boyertown, PA 19S12 This presentation will review the evolution of several applications of tantalum, such as in capacitors and in electronic industry, corrosion-resistant components for chemical processing industries, additives in superalloys, ballistics and carbides. Emphasis will be on critical requirements for these applications. Unique combination of properties of tantalum, such as high melting point, thermodynamic stability of its oxides and carbides,low temperature of coefficient of capacitance, high density and inertness of its oxide film, which make it suitllble for these diversifi~d applications will be reviewed. 9:00am TANTALUM INDUSTRY IN CmNA: MA Fukang, Chu Youyi, The Nonferrous Metals Society of China, 12B Fuxing Road, Beijing 100814, China; Zeng Fangping, Ningxian Nonferrous Metals Smeltery, Shizuishan, Ningxia 753000, China The explored Ta reserves in 14 main mines in China are estimated to be of about 40,000 tons containing Ta20S' Various Ta compounds and metal products are produced form S major plants, such as tantalum pentoxides, potassium tantalum fluoride, tantalum carbides, tantalum powder, elec:tron beam melted tantalum ingot, tantalum wires, plates, strips and tubes, etc. Recently, the application of Ta in capacitors provided in the form of powder and wires is dominant in China, which accounts for more than half of total domestic consumption of Ta content. The application of Tantalum carbides in cemented carbide industry oc
1996 TMS ANNUAL MEETING AND EXHIBITION

tronies, chemical and metallurgical industries. In order to meet the increasing requirements for capacitors, importation of Ta powder and wires has been made since 1991. China has seen a prospective future for the development of tantalum industry. 9:20am NEW DEVELOPMENTS IN TANTALUM CAPACITOR MATERIAL TECHNOLOGY: Hongiu Chang, Cabot Performance Materials, Boyertown, PA 19512 Tantalum is an excellent capacitor material. Conventional powders are produced form sodium reduction of potassium tantalum fluoride, and ball milling of EB melted ingots. This paper will examine new technologies which hold promises for the future. These include fiberous materials, extruded anodes, and fine particle engineering. 9:40am MICROWAVE SINTERING OF TANTALUM FOR CAPACITOR APPLICATIONS: April D. McMillan, R.E. Clausing, R.J. Lockbeed Martin Energy Systems, P.O. Box 2008, Oak Ridge, TN 37831-6087; w.F. Vierow, AVX Tantalum Corporation, 69 Landry St., Biddeford, ME 04005 Tantalum capacitor anodes can be sintered using microwave heating provided the casket contains lossy materials to initiate heating from ambient temperatures. The use ofTa powder in the casket, along with the green capacitor bodies, improves the heating and eliminates contamination from the vacuum environment. Microwave sintered anodes often show unusual surface ripples on the individual Ta particles; their exact origin is presently unknown. This work is supported by the U.S. Department of Energy, Office of Eoergy Research, Laboratory Technology Applications Program, and conducted under Cooperative Research and Development Agreement No. ORNL 91'()()56 between Lockheed Martin Energy Systems, Inc., and AVX Tantalum Corporation. 10:00 am BREAK 10:15am THE PRODUCTION OF CAPACITOR GRADE TANTALUM WIRE: Robert W. Balliett, H.C. Starck Inc., 45 Industrial Place, Newton, MA 02161-1951 Capacitor grade tantalum wire is a critical component in tbe manufacture of capacitors for the electronics industry. Todays capacitor grade wire is produced from tantalum powder using various powder metallurgical processes. Sodium reduced powders are microalloyed with elements that form a fine dispersion of a thermally stable second phase. This dispersion inhibits the motion of grain boundaries during elevated temperature sintering, thus reducing the wire's tendency to become embrittled. This paper briefly discusses the history of wire manufacturing and the major alloying techniques used to produce grain growth resistant wire. Current generation capacitor grade wire is also characterized in terms of chemical, physical and mechanical properties as well as surface chemistry and structure. The structure of the dielectric film formed on the tantalum wire during the anodization process is examined. 10:45 am EFFECT OF PHOSPHORUS OOPANTS ON THE SINTERING OF TANTALUM POWDERS: Sophie G. Dubois, Andrew A. Marszalek, Randall M. German, PIM Lab, The Pennsylvania State University, 118 Research Building West, University Park, PA 16802-6809 The performance of tantalum capacitors is directly proportional to the sintered powder surface area; therefore, dopants such as phosphorus compounds are commonly added to the tantalum powder as a mean to preserve surface area during sintering. The fundamental explanation of the effects of these dopants is still lacking. In this study, the effect of phosphorus on tantalum sintering has been investigated. The effect of the amount of dopants and the sintering temperature have been studied and the mechanism responsible for the preservation of the surface area is being researched. 11:05 am AN OVERVIEW OF THE APPLICATION OF TANTALUM AND ITS ALLOYS AS WARHEAD LINER MATERIALS: James F. Muller, Engineering Manager-Armament Systems, GenCorp-Aerojet, P.O. Box 13222, Sacramento, CA 95813-6000 Modern anti-armor warheads require the use of high density metals as explosive cavity liners to achieve required performance levels against state-of-the-art reactive and composite armors. Those high density materials which possess dynamic ductility at high strain rates (I0ysec) are appropriate candidates for these warhead liners. Tantalum and its alloys is one of the few classes of materials that have the requisite properties, and are not either economically or environmentally untenable. This paper will discuss the use of tantalum materials for various warhead liner applications. Results of laboratory materials properties testing and of actual warhead test firings from several development efforts will be presented. Data and information associated with the use of pure tantalum as the liner material for the TOW 2B warhead production program will also be discussed. 11:25 am TANTALUM FOR HIGH TEMPERATURE APPLICATIONS: Tomoo Izumi, Showa Cabot Supermetals, Higashinagahra Factory, Kawahigashi-Machi, Kawanumagun, Fukushinma-Ken Japan It is desirable to have tantalum of ultra-high purity for capacitors. Two potential sources of impurities in the manufacturing of capacitors, heating elements in the sintering furnace and the lead wire, are also expected to maintain ductility and the fine grainsize after being exposed to high temperature. These two somewhat contradictory requirements are satisfied by micro-alloying tantalum with highly stable oxides, such as Y203· Effect of Y203 on mechanical and metallurgical properties are discussed.

TECHNICAL, ECONOMIC COMMERCIAL & ENVIRONMENTAL CONSEQUENCES OF SULFUR RECOVERY IN COPPER PRODUCTION SYSTEMS III: Sulfur Removal from Gaseous and Liquid Effluents Sponsored by: EPD Copper, Nickel, Cobalt Committee Program Oranizers: Norbert L. Piret, Piret & Stolberg Partners, 1m Licht 12, D-47279 Duisburg German; Ivan M. Santos Moraes, Carafba Metais, Dias D' AVila, Bahia, Brazil

Wednesday, AM February 7, 1996

Room:AII-12 Location: Anaheim Convention Center

Session Chairpersons: N. Steiner, Degussa Corporations, 4 Pearl Court, Allendale, NJ 07402; I.M. Santos Moraes, Caralba Metais, Dias D' Avila, Bahia, Brazil

8:30 am Invited REDUCTIONOFS~-EMISSIONSBEHINDACOPPERCONVERTERINTIIE

COPPER SECONDARY INDUSTRY: Jorg Wallner, Montanwerke Brixlegg AG, Postfach 19, A-5630 Brixlegg In order to reduce S02 emissions from its copper converter, Montanwerke Brixlegg Aktiengesellschaft, a secondary copper smelter, commissioned a flue gas scrubber designed to reduce peak emissions with a significant ecological impact. The system is of the wet scrubber type, employing an oxidative process with a magnesium hydroxide solution produced from magnesium oxide as the absorption agent. The average desulphurisation rate is more than 97,5 %. The requirements, the process itself and the results are presented. 9:00am TOPSOE WSA TECHNOLOGY PROVIDES EFFICIENT DESULFURIZATION OF OFF·GASES FROM NON·FERROUS ROASTERS: Ole Rud Bendixen, Hans Kiilerich Hansen, Haldor Topsoe AIS, Nymollevej 55, DK-2800 Lyngby, Denmark Topsoe has developed a Wet gas Sulfuric Acid (WSA) process which effectively recovers sulfur dioxide in off-gases from non-ferrous smelters and roasters. The product is concentrated sulfuric acid of typically more than 96 wt % irrespective of the sulfur dioxide concentration of the off gas. The process does not generate any waste products or waste water and does not use any absorbents or other chemicals. In the process, up to 99 % of the sulfur dioxide is converted catalytically into sulfur trioxide which reacts with water in the gas, hereby forming gaseous sulfuric acid. The acid is condensed in a WSA condenser of proprietary design. Excellent heat economy and practically no sensitivity to the water balance mean that no extra sulfur burning is required at any sulfur dioxide concentration. The process is especially well suited for treating off-gases where the sulfur dioxide content is lower than about 5-6 %, and it is a good alternative to a complete renovation of a smelter. The process has found good use in both molybdenum and lead roasters. An example from a molybdenum roaster in Chile shows that 97-98 % acid is produced from a 2 3 % sulfur dioxide gas. The plant is situated on the outskirts of Santiago and represented a major local source of sulfur dioxide. The WSA plant was started in May 1993. 9:25am TAILSTACK OPACITY CONTROL FOR METALLURGICAL SULFURIC ACID PLANTS: Dan McLaws, Acid Plant Coordinator, Magma Metals Company, P.O. Box M, San Manuel, AZ 85631 Environmental control is always an important part of operating a copper smelter acid plant. A very important part of environmental control is the visibility of the stack plume; that not only affects public opinion, but also local inversion conditions. Even though these metallurgical acid plants were designed and built to be able to handle specified minimum and maximum gas feed rates and %SD2 concentrations, the feed gas flows and gas grades from copper smelting furnaces are relatively unsteady in comparison to sulfur burning plants. A good fundamental understanding of what causes a visible stack is a great benefit to the metallurgical acid plant operator, as frequent adjustments are required from the highly variable feed gas conditions. This information is presented from an operating point of view to aid in trouble-shooting problem stacks and in training operating personnel. 9:50am FLUE GAS DESULFURIZATION WITH HYDROGEN PEROXIDE: Norbert Steiner, Degussa Corporation, 4 Pearl Court, Allendale, NJ 07401 For the desulfurization of weak S02 gas streams, Degussa developed a process using dilute H2S04 solution containing H2~' instead of conventional gas scrubbing with all solid absorbent medium (limestone, MgO). The process is based on the fast reaction of H2D2 with SD2 yielding H2S04 up to 70 % strength, depending on temperature. As compared to the conventional system with typically 95 % efficiency, independent on the raw gas S02 concentration, residual S02 less than 20 mg/Nm3 can be easily achieved, corresponding to nearly 100 % removal. These features render the Degussa process a very attractive,low capital cost technology to retrofit existing conventional desulfurization plants. 10:15 am BREAK 10:35 am FACTORS AFFECTING THE UTILIZATION POTENTIAL OF GYPSUM PRO. DUCED FROM ACID NEUTRALIZATION: Lydia Luckevich, Ortech, 2395 Speakman Dr., Mississauga, Ontario Canada, L5K I B3

1996 TMS ANNUAL MEETING AND EXHIBITION

B107

Gypsum is a product from the neutralization of sulfuric acid by lime or limestone. 1bis paper will discuss specifications for gypsum for use in wallboard, cement and agriculture and the factors affecting its utilization potential. Process conditions during the neutralization which enhance or inhibit utilization potential will be included in this discussion. 11:05 am BIOLOGICAL SULFATE REMOVAL AND HEAVY METAL RECOVERY FROM MINE WATERS: Andre L de Vegt, Paques, Inc. 486 Thomas Jones Way, Suite 100, Exton, PA 19341; CeesJ. N. Bulsman, Paques B. V, P. O. Box 52, 8560AB Balk, The Netherlands For the past eight years Paques has been engaged in developing and installing treatment systems based on biotechnological processes to remmove sulfur compounds and metals from water and gaseous streams, using two biological steps: 1) Sultate reducing bacteria convert sulfate to hydrogen sulfide, which precipitates soluble metals as insoluble metal suifides; 2) Sulfide oxidizing bacteria convert excess H2S to elemental sulfur. Paques bas designed and installed in May 1992 a 5,000 m3/d groundwater treatment system for the Budelco zinc refmery in the Netherlands. A 50,000 Nm3/h commercial size biological fluegas desulfurization plant using a scrubber and the two bIological steps has been operated since July 1994. The biotechnological methods, the full scale experience and initial operational results of a large US copper mine groundwater remediation pilot plant for the removal of sulfate and heavy metals, started in September 1995, will be presented in this paper.

TRANSIENT THERMAL PROCESSING OF MATERIALS SESSION IV- Process Sensors, Equipment Issues & Manufacturing Perspective

Sponsored by: EMPMD Thin FUms & Interfaces Committee Program Organizers: N.M. Ravindra, New Jersey Institute of Technology, Newark, NJ; R.K. Singh, University of Florida, Gainesville, FL

Wednesday, AM February 7,1996

Room: Grand J Location: Anaheim Marriott Hotel

Session Chairman: Krishna C. Saraswat, Stanford, University, Stanford, CA; Walter F. Kosonocky, New Jersey Institute of Technology, Newark, NJ 8:30 am Invited MANUFACTURING FEASIBILITY OF RIPPLE PYROMETRY IN RAPID THERMAL ANNEALING (RTA): Minseok Oh, Anthony T Fiory, AT&T Bell Lab0ratories, 600 Mountain Ave., Murray Hill, NJ 07901; Binh Nguyenphu AT&T Microelectronics, 9333 S. John Young Parkway, Orlando, FL 32819 Single wafer processing ofRfA requires a temperature sensing method that is independent of wafer-to-wafer process variations. Ripple pyrometry utilizes an AC ripple component of radiation produced by heating lamps and converts it to a temperature in situ. Therefore, it can compensate for the wafer-emissivity variations induced by batch processes prior to an RTA step. A method based on AC lamp ripple was tested extensively for a long-term period at process temperatures of 650 C to 9SO'C. Device wafers as well as blanket-film wafers were examined to study the accuracy and repeatability of the temperature measurement. Statistical analysis with sheet resistance data will be presented to confirm the feasibility of pyrometry using AC lamp ripple in microelectronics manufacturing. 9:00 am Invited DEVELOPMENT OF A NON-CONTACf SENSOR FOR SIOz LAYER TWCKNESS AND TEMPERATURE IN A RAPID THERMAL OXIDATION REAC· TOR: Stuart Farquharson and Peter A. Rosenthal of Advanced Fuel Research, Inc. 87 Church Street, East Hartford, cr06 108; Peter R Solomon, On-Line technologies, Inc., 87 Church Street, East Hartford, cr06108; Nuggehalli M. Ravindra, F.M. Tong, New Jersey Institute of Thchnology, Department of Physics, 323 Martin Luther King Blvd., Newark, NJ 07102 Recently, rapid thermal reactors have been identified as ideal modules for building a programmable factory capable of flexible production of submicron ultra large scale integration devices. Critical to the development and employment of these reactors is the real-time measurements of process parameters to institute process control. Recently, Advanced Fuel Research Inc. and On-Line Technologies, Inc., have developed a Fourier transform infrared system capable of determining film temperature, thickness and composition within process reactors. The system was interfaced to a rapid thermal oxidation reactor at the New Jersey Institute of Technology, to monitor temperature and layer formation rate for one of the most important steps in integrated circuit manufacturing, the formation of thin, high quality silicon dioxide layers on silicon. A matrix method employing the complex dielectric functions for the Si~ layer and the Si substrate allowed the fICSt known determination of Si~ layer thickness AND interstitial (intrinsic) oxygen in the substrate from infrared reflectance spectra, independent of back-side wafer roughness. Absorption bands allowed layer thickness measurements to 0.2 nm (2A) and intrinsic oxygent concentration simultaneous measurement of infrared spectral radiance allowed real-time wafer temperature measurements (200 to 800 +- 2 C). Integration of real-time temperature and layer formation rate into process models to control final product properties will be presented. 9:30 am Invited MULTI-WAVELENGTH IMAGING PYROMETER FOR NON·CONTACT TEMPERATURE MONITORING: M.B. Kaplinsky, J. Li, NJ. McCaffrey, W.F. Kosonocky, E.S.H. Hou, N.M. Ravindra, Electronic Imaging Center, New Jersey Institute of Technology, University Heights, Newark, NJ 07102 Abstract not available.

Bl08

10:00 am BREAK 10:15 am REAL·TIME PROCESS CONTROL OF MOLECULAR BEAM EPITAXIAL GROWTH OF MERCURY CADIMIUM TELLURIDE FILMS BY FOURIER TRANSFORM INFRARED SPECTliOSCOPY: Stuart Farquharson, Martin D. Carangelo, James R Marlcham, Advanced Fuel Research. Inc., 87 Church Street, East Harford, cr 06108; Peter R. Solomon, John R. Haigis, On-Line technologies, Inc., 87 Church Street, East Hartford, cr 06108; Nuggehalli M. Ravindra, Fei-Ming Tong, New Jersey Institute of Technology, Department of PhysiCS, 323 Martin Luther King Blvd., Newarlc, NJ 07102; Glenn Wesphal, Texas Instruments Incorporated, 13536 North Central Expressway, Dallas, TX 75265 The ternary semiconductor mercury cadmium telluride (Hgl-xCdxTe) has become the most widely used material in infrared detection devices. 1bis is primarily due to the ability to tailor its spectral response (band gap energy or cut-off wavelength) and efficiency to the desired application. 1bis is :lCcomplished by controlling the relative alloy concentration (x) of mercury and cadmium. Nevertheless, stringent process control of fih~ temperature and film growth rate is necessary to attain the required alloy concentrattons. Recently, Advanced Fuel Research Inc. and On-Line Technologies, Inc., have developed a Fourier transform infrared system which performs real-time measurement of these properties, and interfaced it to a molecular beam epitaxial growth reactor at Texas Instruments for dedicated process: control. The system performs simultaneous measurement of the film infrared reflectance and emittance. The former is used to determine film thickness, the latter to determine the cut-off wavelength, and the combination to determine the film temperature. Instrument design, spectral analysis, and implementation of process control will be presented. 10:35 am ESTIMATION OF TRANSPORT LOSSES IN RTP SYSTEMS: R.K. Singh, Department of Materials Science & Engineering, University of Florida, Gainesville, FL 32611 Abstract not available. 10:55 am THREE DIMENSIONAL MODELLING OF RTP SYSTEMS: R.V. Nagabhusnam, R.K. Singh, Department of Materials Science & Engineering, University of Florida, Gainesville, FL 32611 Abstract not available. 11:15 am REAL·TIME IMPLEMENTATION OF AN ADAPTIVE CONTROL SYSTEM FOR A 3·Z0NE RTP STATION: Serl~ey Belikov, David Hur, Bernard Friedland, N.M. Ravindra, New Jersey Institute ofThchnology, Newark, NJ 07102 1bis presentation concerns with real-time implementation of a model-based adaptive RTP control algorithm developed in NJIT. An adaptive controller with measured temperatures of the wafer as the inputs and control voltages for the heating lamps as the outputs was designed and implemented on a transputer parallel computing system. An intermediate step of the process of installation of the controller into the 3-zone RTP station, under investigation at NJIT, involves its testing with an RTP simulator, which has the same inputs and outputs as the actual RTP station. Due to natural and artificial noise of the RTP simulator, the simulabld inputs for the controller become close to actual real-time outputs of temperature sensors used for RTP. Using the RTP simulator allowed us to solve various real-time implementation problems before the actual installation of the controller into the RTP static,n. During this step, we also improved quality and robustness of parameter estimation. Instrumental wafers with embedded thermocouples have been used for testing the adaptive control system on real-time RTP processes. 2-dimensional infrared and fiber optic sensors are planned to be used for the controller after its successful testing witll thermocouples. 1bis investigation was supported by the National Science Foundation under Grant ERSC-9312451 and ARPA under contract F33615-92-C-5817. 11:35 am COMPUTER WORKSTATION FOR MULTIWAVELENGTH IMAGING PY· ROMETER: J. Li, E.S.H. Hoo, M.B. Kaplinsky, NJ. McCaffrey, W.F. Kosonocky, Electronic Imaging Center, New Jersey Institute of Technology, University Heights, Newarlc, NJ 07102 Abstract not available.

WEDNESDAY PM ADVANCES IN THE SCIENCE AND TECHNOLOGY OF TITANIUM ALLOY PROCESSING - VI: Heat Treatment II

Sponsored by: SMD TItanium Committee, MDMD ShapIng and FormIng Com· mlttee Program Organizers: Prof. Isaac Weiss, Prof. Raghavan Srinivasan, Mechanical and Materials Engineering Dept., Wright State University, Dayton, OH 45435; Dr. Paul Bania, TImet Corporation, TImet-Henderson Technical Laboratory, P.O. Box 2128, ~enderson, NY 89009; Prof. Daniel Eylon, Graduate Materials Engineering, UniverSIty of Dayton, Dayton, OH 45409 Wednesday, PM February 7,1996

Room: B5-6 Location: Anaheim Convention Center

Session Chairperson: RR Boyer, Boeing Commercial Airplane Group, Seattle, WA 98124; H. Rack, Materials Science and Engineering Program, Clemson University, Clemson, SC 29634-0921

1996 TMS ANNUAL MEETING AND EXHIBITION

2:00 pm Invited HEAT TREATMENT OF TITANIUM ALLOYS: J. R. Wood. P. A. Russo. RMI 11tanium Company. 1000 Warren Avenue. Niles. OH 44446 11tanium and titanium alloys are heat treated for a variety of purposes. Examples are: stress relieving to relieve residual stresses during fabrication; process annealing to produce an acceptable combination of ductility. machinability. and dimensional stability; and solution heat treating and aging to optimize strength. ductility. fracture toughness. fatigue strength. creep resistance. etc. A review of these processes and a discussion on specialized heat treating processes are described in this paper for alpha-beta titanium alloys and metastable beta titanium alloys. 2:25 pm Invited ROLE OF OXYGEN ON TRANSFORMATION KINETICS OF TIMETAL 21S TITANIUM ALLOY: M.A. Imam. C.R. Feng. Materials Science and Technology Division. Naval Research Laboratory. WaShington. DC 2037S-S343 Tbe 11metal 2 I S is one of the newest beta titanium alloy which has shown excellent resistance to corrosion and heat compared to other beta titanium alloys. Interstitial elements such as oxygen. nitrogen. carbon. and hydrogen affect the transformation characteristics of the 11metal 21 S resulting in the variations of the mechanical properties. Although the stable phase at room temperature is alpha. the measureable beta is retained even after slow cooling from above the temperature. Tbe transformation of beta phase is sluggish and can be controlled to produce desired properties. Tbe structure of the transformed product and their distribution depend on the aging conditions and interstitial contents. The present investigation deals with the role of oxygen on the phase transformation kinetics and establish a time-temperature and transformation curves of timetal 21S. 11-ISMo-3AI-2.8Nb-0.2Si, Oxygen contents in the alloy was systematically varied from 0.9wt% to 0.33wt%. 2:50pm PROCESSING· MICROSTRUCTURE· PROPERTY RELATIONSHlPS IN NI MODIFIED CORONA 5: J. C. Fanning, Henderson Technical Laboratory, 11met, P.O. Box 2128, Henderson, NV 8901S; P.L Martin Rockwell Science Center, 1049 Camino Dos Rios, Thousand Oaks, CA 91360 Three Alloys were defined to modify the relative strength and volume fraction of the IX and 8 phases in Corona 5 (11-4.5AI-5Mo-I.5Cr). The AI, Mo, Cr, and oxygen levels were varied and Ni was added to increase the volume fraction of 8. Ingots weighing 36 kg of three compositions were double VAR melted then 8 forged followed by aJ8 forging to produce 50mm slab. Sections were cross-rolled to a thickness of 19mm using two starting temperatures; Ta ±28°C. Tbe microstructure at each stage of processing was documented using backscattered electron metalography on electropololished sections. lWo step heat treatments, beginning both above and below the a transus, were used to modify the primary and secondary morphology in order to maximize the tensile strengthltoughness trade-off. Ambient temperature fracture toughness values as high as 70 MPa m112 were measured at a corresponding UTS of 1120 MPa. Tbe variation of K Ic with tensile strength as influenced by the microstructure will be shown and discussed in light of the probable mechanisms of fracture. 3:05pm PRODUCTION AND HEAT TREATMENT OF Ti·15·3 ALLOY TUBE: Hong Quan, Luo Guogzhen, Wu Yiqing, Northwest Institute for Nonferrous Metal Research, P.O. Box 51 Xian, China 710016 This paper described the process of11-15-3 alloy tube production and its processing behavior, as well as the effect of heat treatment on its mechanical properties are studied. Tbe microstructure and texture type of the tube has been investigated by optical microscope and X-ray diffraction methods. Tbe results show that11-IS-3 tube has good processing properties: flattening is 4s, expanding> 25%, which is similar to pure 11. Its mechanical properties can be easily controlled by adjusting heat treatment parameters. Cold rolling and ageing is suitable for 11-15-3 tube because of its exellent properties and low cost. This tube can be applied to the field of sport material.

phology due to the coalescence of the single variant precipitates. Additionally, at triple points of the high-angle grain boundaries, very few precipitation is observed because of difficulties of the variant selection to three 8 grains at once. On the other hand, the triple points of sub-boundaries are observed to be the preferred nucleation site for a precipitation. 4:00pm MECHANICAL BEHAVIOR AND TRANSFORMATIONS IN TI-60wt%18: Robert W. Margevicius, James D. Cotton, Los Alamos National Laboratory, Los Alamos, New Mexico 87545 Of the a-stabilized 11 alloy systems, 11-Ta has not been extensively investigated. In a broader research program, a range of 11-Ta alloy compositions was investigated. Tbe results for one alloy, 11-6Owt%11I (11-28at%11I), will be presented here. Tbe alloy was plasma arc-melted and cross-rolled at 900°C in air to approximately 80% reduction. Tbe rolled plate was subjected to heat treatment at 700°C, 800°C, or 900°C for one hour in argon followed by a water quench. Tensile tests on the heat-treated material showed a plateau in the stress-strain curve at a relatively low stress, indicative of a stress-assisted transformation of metastable a-xx", as confirmed with optical and electron microscopy and texture measurements. Tbe stress required to trigger the transformation increased with increasing heat-treat temperature, suggesting that stored work removed by annealing processes stabilizes the metastable a phase. Mechanisms for this phenomena are discussed. 4:15pm EFFECTS OF HEAT TREATMENT ON MICROSTRUCTURE, STRENGTH, AND TOUGHNESS OF TlMETAL 21S SHEET: R. Kieth Bird. Terry A. Wallace, William D. Brewer, NASA Langely Research Center, Hampton, VA Because of their versatility, Beta titaium alloys are of interest for a variety of aecospace applications, including subsonic and supersonic airframe structures. Generally, they have high strength to density ratios, deep hardenability, can be heat treated to a wide range of properties, and have potential for lower cost because of being strip producible. In partlcular,11metal2lS has been given considerable attention because of the additional benefits of improved oxidation, and corrosion resistance. In this study, the effects of heat treatment on the tensile properties and toughness of 0.022 inch thick, 11metal 21 S sheet were studied over a range of temperatures, with the prupose of developing appropriate microstructures and phase distribution to maximize strengthltoughness combinations. The heat treatments included beta and alpha·beta anneals with various quenches and ages, as well as direct aging. Optical, SEM., TEM, and x-ray analyses were used to define the microstructures. Toughness was measured by the J-integral method using compact tension specimens. Heat treatments, microstructure, tensile and fracture properties were correlated from -65°F to +350°F. 4:30pm ISOTHERMAL DECOMPOSITION OF TIMETAL-21S: Qiong Li, Roy Crooks Analytical Services and Materials, Inc., Metallic Materials Branch NASA Langley Research Center, Hampton, VA Tbe decomposition of the metastable 6-phase titanium alloy TIMETAL-21S was studied by transmission electron microscopy after 6-annealing, quenching and isothermal aging at temperatures ranging from 450-565°C. Uniformly distributed isothermal IDphase was observed after short aging times at lower temperatures; while was the dominent phase after further aging. Several stages were observed in the precipiation and growth of al and these were characterized by their morphologic and crystallographic changes. Colonies of fine "sideplate" IX were observed to extend from some grain boundaries, and much coarser was present in grain interiors. Tbe intragranular IX was associated with ID-phase in the early stages of aging. Tbe sideplate IX grew from a polycrystalline prism structure formed by the junction of three a variatiants. After pr0longed aging times, IX further decomposed into type II . Nucleation theories for these phases are discussed.

3:35 pm BREAK

4:45pm EFFECTS OF MICROSTRUCTURE AND OXYGEN CONTENT ON THE FRACTURE BEHAVIOUR OF THE a+ 8 TlTANIUMALLOYTI-4AI-4Mo-2Sn0.5SI wL % (lM1 550): LJ. Hunter, M. Strangwood, P. Bowen, School of Metallurgy and Materials IIRC in Materials for High Performance Applications, Tbe University of Birmingham, Edgbaston, Birmingham, B 15 2TT, UK It is well establisbed that both fracture toughness and strength of IX + 8 titanium alloys are sensistive to variations in microstructure and alloy composition. Hence, in this and previous studies, the roles of microstructure and oxygen content on the balance of strength and fracture toughness for the 11-4AI-4Mo-2Sn-0.5Si alloy have been investigated. A detailed investigation to elucidate the micromechanism of ductile failure has also been undertaken. Both fracture toughness and tensile strength levels revealed a strong dependence on microstructure and oxygen content. In all fracture toughness specimens, the ductile micromechanism of failure is dominated by interfacial decohesion. TEM characterization of interfacial regions has been undertaken and related to variations in the microstructure and the amount of oxygen present.

3:45pm A STUDY OF GRAIN BOUNDARY NUCLEATED IX PRECIPITATES IN AN (a+8) TITANIUM ALLOY: Kei Ameyama, Dept. Sci. & Eng. Ritsumeikan University, Kusatsu, Shiga, Japan 525-77; Hiroshi Fujiwara, Hajime Kawakami Graduate School, Ritsumeikan University, Allec Mitchell, Dept. Metals & Mater. Eng., University of British Columbia, Vancouver, B.c., Canada V6T IZ4 Tbe Morphology and crystallography of a grain boundary IX (HCP) precipitates formed on the a (BCC) matrix grain boundaries have been studied with TEM and SEM-ECP ina a 11-22V-4A1 alloy. Most ofa precipitates have a near-Burgers orientation relationship, i.e., (OOOI)a, with respect to, at least, one of the adjacent 8 grains. Variant selections of precipitates are made so as to minimize the misorientation with respect to the "opposite" a grain from the Burgers one. Such rules in variant selection frequently result in the formation of a precipitates with a single variant at a planar grain boundary and thus the precipitates appear as a film like mor-

5:00pm PHASE STABILITY IN A TI-45.5AI·2Nb-2Cr ALLOY: V.Seetharaman, Materials and Processes Division, UES Inc., Dayton, OH 45432-1894; C.M.Lombard, S.L.Semiatin, Wright Laboratory, Materials Directorate, WUMLLN, Wright-Patterson AFB, OH 45433-7817 High temperature phase equilibria in a 11-45.5AI-2Nb-2Cr alloy were investigated via X-ray diffraction and metallographic techniques. Tbe alloy was hot worked through a sequence of isothermal forging and pack rolling operations coupled with intermediate, recrystallization heat treatments. Specimens from the rolled sbeets were subjected to isothermal heat treatments in the temperature range 900-1 320°C for durations ranging up to 100 hours, followed by water quenching as well as slow cooling. Tbe results obtained indicate that a three phase equilibrium involving 'Y, aJIX2, and Il:2 phases exists in the temperature range II OO-I300°C. While the volume fraction of Il:2 appears to reach a maximum at -1200°C, that of 'Y decreases progressively with an increase in

3:20pm DIRECT AGING OF Ti·6.8Mo·4.5Fe· 1.5AI: S. Azimzadeh, H. J. Rack, Materials, Science and Engineering Program, Clemson University, Clemson, SC 29634-0921 Tbe kinetics of phase transformations occurring in TiMET LCB (11-6.8Mo-4.5FeI.SAI) upon a solution treatment and direct aging were investigated over the temperature range 34S-6S0oC utilizing hardness, x-ray diffraction, optical and electron microscopy. Aging below 400°C resulted in precipitation of ellipsoidal OlisothermaJ either through a nucleation and growth mechanism or by displacive controlled growth of Cllthermal, depending upon the aging temperature relative to the Cllstart temperature. At higher temperatures Widmanstlitten a precipitated directly from the a matrix, extended aging times resulting in a platelet growth and spherodization. Tbese observations will be discussed and contrasted with studies of solution treated, water quenched and aged TIMETLCB.

1996 TMS ANNUAL MEETING AND EXHIBITION

B109

temperature in the range lIOO-I300°C. The influence of temperature and time of annealing and cooling rate on microstructure of this alloy will be discussed in the light of available information on phase stability in TtAI-X systems, where X is a beta stabilizingelement. 5:15pm MICROSTRUCTURE EVOLUTION IN AN ORTHORHOMBIC TITANIUM ALUMINIDE AS A FUNCTION OF TEMPERATURE-STRAIN -TIME: R.W. Hayes, Metals Technology Inc.; 19801 Nordhoff St., Northridge, CA 91324; C. G. Rhodes, Rockwell International Corp., 1049 Camino Dos Rios, Thousand Oaks, CA 91360 Uniaxial tension creep tests have been pedonned on the orthothrombic titanium aluminide alloy Tt-22AI-23Nb (alo) over the temperature range 650°C-760°Cat constant initial applied stress levels ranging from 69 MPa up to 172 MPa. During the course of creep testing within this temperature-stress regime, the development of a phase at the primary alpha-2-orthorhombic + ordered beta phase matrix intedace appears to occur. This phase appears in the grip section of creep specimans as well as in the gage section. The formation of this phase appears to be driven by temperature, strain and time. This paper will address through detailed microstructural characterization, the nature of this phase and the parameters which are primarily responsible for its formation. 5:30pm HYDROGEN TECHNOLOGY AS NEW PERSPECTIVE TYPE OF TITANIUM ALLOY PROCESSING: B.A. Kolachev, A.A. Ilyin, V.K. Nosov, Moscow State University of Aviation Technology after K. Tsilkovsky, Metals Science Dep. Petrovka St. 27, K-31, Moscow, Russia, 103767 In studying the "metal-hydrogen" problem it was found that though hydrogen caused embrittlement, it also had some effects which might be used for improvement of the processing technology of metals, for example titanium alloys. These effects include hydrogen induced decrease in flow stresses, increase of maximum deformation prior to the occurence of the flfSt crack, structural and phase transformation, hydrogen enhanced adhesion phenomena, and hydrogen influence on thermo-physical properties. The above mentioned effects have fonned the basis of hydrogen technology of titanium alloys which includes the following processing: a) hydrogen plasticization; b) thermohydrogen treatment; c) mechanical-hydrogen treatment; d) processes based on hydrogen strengthening of adhesion phenomena; e) hydrogen technology of casting. In the paper there are given the examples of successful using of hydrogen technology for solving the following problems: a) refinement of micro- and macrograins of castings; b) improvement of quality welded bonds; c) increasing of heat treatment effects in pseudo IXalloys; d) producing of high quality active semiproducts from difficult to deform alloys; e) improvement of fasteners technology production; f) easier PM processing and diffusion bonding of half-finished pressings from titanium chip without its remelting; h) improvement machining of titanium alloys; i) enhanced of quality of castings; j) utilization of cast scrap. Hydrogen technology allows the increase in part-to-scrap ratio, to raise labor productivity, to reduce energy and metal consumption.

ALUMINA & BAUXITE TECHNOLOGY IV: Process

Sponsored by: LMD Aluminum Committee Program Organizer: e.G. Hendriks, Kaiser Aluminum and Chemical Corp.,913 I Interline Drive, Baton Rouge, LA 70809 Wednesday, PM February 7, 1996

Room:A6 Location: Anaheim Convention Center

Session Chairman: J.L. Anjier, Kaiser Aluminum and Chemical Corp., Box 337, Gramercy, LA 70052 2:00pm MORE COMPLETE DESILICATION OF ALUMINATE SOLUTION IS THE KEY-FACTOR TO RADICAL IMPROVEMENT OF ALUMINA REFINING: Dr. Victor Rayzman, 7329-D Alicante Rd., Carlsbad, CA 92009 In the nepheline process for alumina production, alumina is precipitated from the liquor by carbonation. This also co-precipitates all the dissolved silica which, if allowed to remain at a high level, unacceptably contaminates the product. This is controlled by desilicating the liquor prior to precipitation to a level in which the Al203:Si~ ratio is 1000-1200:1. In the Bayer process this ratio is typically 300-350:1, as most of the silica in solution passes through precipitation without co-precipitating with the alumina. This residnal silica has serious effects on the process, leading to the formation of DSP which encrusts heat exchange sudaces and increases the viscosity of the liquor thus reducing settling rates. There are a serious reasons to suppose that the rise in the rate of aluminum hydroxide precipitation and, consequently, AI203 yield are limited due to excessive Si~ concentration. This supposition is well-founded by thermodynamic calculations. Some processes for the more complete desilication of aluminate solution are described. 2:30pm WORLDWIDE BAUXITE AND ALUMINA PRODUCTION CAPACITY-I996: Errol D. Sehnke, U.S. Bureau of Mines, 810 7th. st., NW, Washington DC 20241-0002 A comprehensive review indicates that total worldwide "in place" capacity for the production of bauxite and bauxite-alternative ores currently is approximately 133 million metric tons per year. This includes the annual capacity to produce 123 million metric tons of metallurgical grade bauxite, 7 million metric tons of bauxite used for non-metal applications, and 3 million metric tons of alumina equivalent non-bauxitic materials (alunite and nepheline syonite). Current total world alumina production capacity amounts to about 51 million metric tons per year. This includes the annual ca-

B110

pacity to produce approximately 48 million metric tons per year of smelter-grade alumina for the production of primary aluminum metal and about 3 million metric tons for non-metal specialty applications. 3:00pm OPTIMIZATION OF MULTISTAGE PLANT AT C.V.G. BAUXILUM: Karel Fort, Ph.D., KF Engineering Services, P. O. Box 2111, 8604 Volketswil, Switzerland; Saul Escalonia, Alberto Parra, Paolo Murgila, C.V.G. Bauxilum, Cuidad Guiana, Venezuela; Jose Angel Sorrentino, Universidad Central de Venezuela, Caracas, Venezuela The three thermal areas, Digestion, HID and evaporation are at BAUXILUM designed as Multi Stage Hash Units. Whereas the digestion Hash Unit has been operating from the begining with the expected hydraulic pedormance and is producing process condensate of excellent quality, the HID and Evaporation Units had to be optimized since notorious high levels in the flash vessels and high entrainment of the process condensate caused operational problems. A pilot plant conSisting of two typical flash modules was built and operated as a "ide stream to the Evaporation Area at different temperature levels. Detailed measurements of the two-phase flow in the flash device, its analysis and comparison with models provided a sound basis for later development of modifications. The capacity of t1K: Evaporation Unit increased by 25% after the modification whereas the process condensate quality improved by a factor of 10. 3:30pm ALUMINA CALCINATION WITH THE ADVANCED CIRCULATING FLUID BED TECHNOWGY: A DESIGN WITH INCREASED EFFICIENCY COMBINED WITH OPERATING FLEXIBILITY: Dr. Hans-Werner Schmidt, Vice President, Lurgi Metallurgie GmbH, Lurgi·AlIee 5, 60295 Frankfurt am Main, Germany; Alan N. Silberberg, Senior Vice President, Lurgi Corporation, West 115 Century Road, Paramus, NJ 07652; Dip!. Eng. Werner Stockhausen, Lurgi Mettalurgie GrnbHLurgiAllee 5, 60295 Frankfurt am Main, Germany The Circulating Huid Bed (CFB) te<:hnology has now been applied for alumina calcination for a quarter of a century. Th~ combined capacity of the 32 units installed is greater than 10 million metric tons per year. The paper highlights the consistency of the product quality which is based upon the operating experience of the last decade and improvements to the calcination system which in addition provide lower heat consumption. The principal modifications are i.ncorporated in the preheating and cooling sections of the plant. These design modifications have also reduced capital cost. Overall the plant retains its proven features of high flexibility, unique temperature control, high availability, reliable performance, and low maintenance cost. The design is applicable to single train units up to a capacity of 3,000 MTPD (alumina). 4:00 pm BREAK 4:20pm MEASUREMENT OF MUD LEVEL INTERFACES: A TOOL FOR OPTIMIZATION OF RED MUD WASHING AT C.V.G. BAUXILUM: Ami/car Pineda, C. V.G. Bauxilum, C.A. Gerenciade Desarrollo, Zona Industrial Matanzas, Edo. Bolivar, Venezuela; Guillermo Adrian, C. V.G. Bauxilum, Ciudad Guayana, Venezuela For the expansion to 2.0 MTPY of the alumina plant, the area of clarification and red mud washing was rearranged from four parallel trains of 5 washing stages each to two trains of 7 stages each. The tanks being of the flat bottom type. As a result of this, the specific capacity of the tanks for handling mud was tripled, increasing the necessity for an on-line system that would allow better and faster control of mud handling. With this scope, it was developed and installed a new continuous mud level detector that gives the measurement of both mud level and. turbid zone in the tanks. The development of the new instrument started from commercial instruments for density measurements and was completely re-engineered in order to obtain the maximum reliability. Actually twenty eight tanks are equipped with these instruments and are working satisfactorily. 4:50pm THE MULTIPLE APPLICATIONS OF HYDROCYCLONES IN ALUMINA PRODUCTION: Nils Oeberg, Gerald Kelton, Christian Besendoder, Hank Rawlins; 1205 Chrysler Drive, Menlo Park, CA 94025-1178 Hydrocyclones are used in many alumina refineries to enhance efficiency. A partial list of applications includes bauxite grinding, bauxite refinement, sands separations from red muds, sand washing, hydrate classification and product size control, seed size control, hydrate thickening and deliquoring, agglomeration, preferential separation/reduction of oxalates, spent liquor solids recovery, cooling tower water clean-up, and lime slurry degrining. This paper review!: these applications including viable f10wsheets and operating experience; and summarizes basic hydrocyclone selection including the main factors affecting pedormance. 5:20pm INTEGRATED TECHNOWGY FOR COMBUSTION OF ALUMINUM-CONTAINING COAL AND THE RECOVERY OF ALUMINA AND BYPRODUCTS FROM THE FORMED ASH: Dr. Vicl'Or Rayzman, 7329-D Alicante Rd., Carlsbad, CA9200 Coal is a major contributor to power and heat generation in the U.S .. Consumption is projected to steadily increase through 2010, with consequent waste disposal problems. The alumina contained in this waste, 2 3 million tons annually, if recovered, would satisfy the annual U.S. need. Numerous processes have been proposed and investigated to recover alumina and other products from coal ash. None of them have been developed completely because in previous processes the conditions for both coal combustion and alumina recovery are economically unattractive. An integrated and environmentally friendly coal combustion ash-sintering sinter-leaching process, which completely uses the thermal and chemical potential of the boiler wastes, bottom ash, fly ash, and flue gases, is proposed. Alumina or alumina concentrate and material suitable for construction use is produced. Key aspc:cts of the proposed method have been benchscale and pilot-plant tested in the former USSR. The next stage of the integrated pr0cess development would be a pilot-plant b:st of the entire cycle, and design of a demonstration project.

1996 TMS ANNUAL MEETING AND EXHIBITION

ALUMINUM ALLOYS FOR PACKAGING II: Session IV

Sponsored by: SMD Non-Ferrous Metals Committee and Light Metals Division Program O'8anizer: Professor James G. Morris, Chemical & Materials Engineering, Director, Light Metals Research Labs., University of Kentucky, Lexington, KY 40506 Wednesday, PM February 7,1996

Room:A2-3 Location: Anaheim Convention Center

Session Chairperson: Subodh K. Das, ARCO Aluminum, Inc., 2900 First National Tower, P.O. Box 32860, Louisville, KY 40232

2:00 pm Invited TRENDS IN ALUMINUM CAN STOCK NEEDS: Speaker to be announced. An analysis of the aluminum industry's needs for future use of can stock materials and an indication of the trends that this technology will take in the coming years. 2:30pm DEVELOPMENT OF 5000 SERIES ALLOYS FOR RCS APPLICATIONS: J. K. McBride, 253 Highland Avenue, New Kensington, PA 15068; R. E. Sanders, Alcoa Technical Center, 100 Technical Drive, Alcoa Center, PA 15069; H. G. Reavis, Aluminum Company of America, Warrick Operations, Newburgh, IN 47629 The demands for formability and consistency placed on aluminum rigid container sheet (RCS) used in packaging applications are among the most stringent in the metal-working industry. The material characteristics necessary for making aluminum ends for beverage cans have changed dramatically as the industry has evolved over the past 30 years. This paper traces the evolution of alloys used in end and tab stock applications from the 1950s to the present. Early experiences with H19 RCS alloys which resulted in the development of alloy 5182 are reviewed with emphasis on worlc hardening and partial annealing behavior. Changes in the industry, particularly coil coating and new end deSigns, which drove metallurgical changes to the product are discussed. 2:50pm THE EFFECT OF ALUMINUM SURFACE PROPERTIES ON BLEEDTHROUGH: R. J. Rigge, Kaiser Aluminum and Chemical Corporation Center for Technology, 6177 Sunol Blvd., Pleasanton, CA 94566 The surface properties of aluminum can stock are important factors in contrOlling the defect of " bleedthrough" on commercial beverage cans. The nature of the organo-metallic monolayers generated and the surface topography imprinted during the final rolling steps are key properties that need to be controlled to minimize bleedthrough. It is believed some types of bleedthrough may be ascribed to deficiencies in these areas. 3:10pm STRIP CAST ALUMINIUM FOIL: P. Charlier, Granges Eurofoil, Dudelange, Luxembourg, H.E. Ekstrom, Granges Technology, Finspang, Sweden Aluminium foil is an ideal product for strip casting. The advantages with the process-short production route and uniform materials properties-are utilized at the same time as the disadvantages-limitations in alloy composition and microstructural control-are avoided. The reasons for using aluminium foil in general and the threat from competing materials are discussed in the paper. The foil properties which are important for the manufacturing into different products are listed and examples are given of how the specific demands on each product can be met by the choice of process conditions and alloy composition. 3:30 pm BREAK 3:50pm METALLURGICAL AND MECHANICAL EFFECTS ON CAN BOTTOM WRINKLING: William M. BellS, Senior Research Metallurgist, KAISER Aluminum Corporation, 6177 Sunol Boulevard, Pleasanton, CA 94566 Bottom wrinkling has always been an issue for the D&I can but has become more troublesome with the advent of the 202 stackable bottom. This wrinkling occurs during redraw when the unsupported tapered area essentially buckles. Various 0&1 stock was tested to determine the effect of the following parameters on bottom wrinkling: Strength, worlc hardening exponent, gauge, and redraw radius. The investigation involved forming cans without doming the bottom, then measuring wrinkling by a stylus profiler. The results indicate that wrinkling can be reduced by both metal supplier and canmaker.

ing, MaJaviya Regional Engineering College, Jaipur-17, India With the advancement in 'high-density integrated packaging' and 'interconnection technology' electronic industries have made a rapid improvement in multidimensional arena. The electronic packaging and interconnections refer to the physical link between computers and other electronic systemslintegrated circuits. It leads to a drastic progress, in increasing the computing capabilities and speed of high-performance workstations, minicomputer, mainframe computers as well as to miniaturization of consumer products as explained. Electronic-packaging sorely depends on compatibility between conductorldielectrics (as a substrate) and circuitization and integration processes. Aluminium and its alloys are being used as a conductor lines of .001 inch in a multilayer structure by patterning technique due to high performance in achieving necessary physio-mechanical properties. Fabrication of these patterning lines is carried out by the latest technique of plasma etching which is discussed with the practical problems of its process variables. It has numerous applications like computer Chip manufacturing of sensors, high-end computing etc. It is observed by analyzing the associated data's that this AI-alloys gives a viable approach to the optimization of desired properties for 'compatibility' with dielectric substrate for ego electrochemical migration, a low defect densitylimpurity content, low electrical resistance and high thermal conductivity. Paper also deals with optoelectronics packaging which have a significant impact on telecommunication (fibre-optic lines), optical reading devices (scanners) and supercomputers.

ALUMINIUM REDUCTION TECHNOLOGY VI: Cell Emissions

Sponsored by: LMD Aluminium Committee Program O'8anizer: Ms Fiona J Stevens, Comalco Research and Technology, Comalco Research Centre, PO Box 316, Thomastown, Victoria 3074, Australia Wednesday, PM February 7, 1996

Room:A9 Location: Anaheim Convention Center

Session Chairperson: Margaret Hyland, Chemical and Materials Engineering, University of Auckland, Private Bag 92019, Auckland, New Zealand

2:00pm THE NORWEGIAN ALUMINIUM INDUSTRY AND THE LOCAL ENVIRONMENT: Caro [sene Stoll, Hydro Aluminium a.s. P.O. Box 80, 1321, Stabekk. Norway Connected to a possible expansion of the Norwegian aluminium industry, a comprehensive environmental impact assessment has recently been carried out for the seven primary aluminium plants in Norway. The project, employing more than 30 scientists from 13 independent research institutes during four years, is the largest interdisciplinary environmental research program within this industry to date. The main objective was to establish scientific knowledge about the relationships between emissions and effects and to establish critical loads and levels. Scientific relationShips between varying levels of emissions and the corresponding environmental effects (dose/response relationships) in relation to human beings, terrestrial animals, soil and vegetation and marine life were investigated. As a consequence of major reduction in emissions over the last years, the environmental condition around the aluminium plants have improved significantly, but further reductions are needed for some plants. 2:30pm IMPROVED EFFICIENCIES IN THE DRY SCRUBBING PROCESS: Geir Wedde, ABB Environmental, PO Box 6260 Etterstad, 0603 Oslo, Norway Only moderate changes over the last decades has been implemented into the dry scrubbing process consisting of the concurrent reactor and filtering process. An advanced process for removal and recovery of fluorides and particulates from aluminum pot gas has been developed and tested in pilot and full scale installation. The new features of the system include: Two-step process arranged concurrently for high fluoride removal efficiency at high fluoride load; New filter design and fabric material (bag) design for low pressure drop and low dust emission; Compact filter allowing for 55% more filtering area in same sized compartment; New pulse valve and nozzle pipe design, Potential for dry S02 removal in second step of process, Integrated impurity removal. Discussions of the results from the development and plant experiences are presented in this paper. 3:00 pm BREAK

4:10pm MICROSTRUCTURAL EVOLUTION AND MECHANICAL PROPERTIES OF ALUMINUM ALLOYS PREPARED BY EQUAL CHANNEL ANGULAR PRESSING (ECAP): Jimin Zhang, Department of Materials Science and Engineering, University of California, Los Angeles, CA 90095 Equal Channel Angular Pressing (ECAP) was used to produce uniform and ultra-fine grain size microstructure for several aluminum alloys. The uniform and large deformation of the process, combined with subsequent thermal treatment, was utilized to manipulate the microstructure of the alloys. The microstructure evolution, particularly changes in shapes and sizes of the grains and micropores, was quantitatively studied. Mechanical properties of the alloys were measured and analyzed. Deformation characteristics, apparatus design, and potential applications of the ECAP will be discussed.

3:10pm REMOVAL OF IMPURITIES FROM DRY SCRUBBED FLUORIDE ENRICHED ALUMINA: Lolhar Schuh, ABB COlJlOrate Research Centre, Speyerer Strasse 4, 0-69115 Heidelberg, Germany. Geir Wedde, ABB Environmental, PO Box 6260 Etterstad, 0603 Oslo, Norway The pot gas from an aluminum electrolytic cell is cleaned by a dry scrubbing process using fresh alumina as the scrubbing agent. This alumina is enriched with fluorides and other trace impurities in a closed loop system with the pots. The only Significant removal of the accumulated impurities is through the metal tapping. An improved technique has been developed that is more effective than earlier stripper systems. The impurity rich fine fraction «I Oum) of the enricbed alumina is partly attacbed to the coarser alumina. This attachment has to be broken. Impact-milling under special moderate conditions and air classifying has shown to be a cost effective process for removal of impurities. For iron (FE) and phosphorous (P) about 30-50% can be removed by a separation of 0.5% of the enriched alumina. Full scale tests has successfully confl1ll1ed these results.

4:30pm PROCESSING AND PROPERTY CHARACTERIZATION OF ALUMINIUM PACKAGING MATERIALS: Mohil Sisodia, Department of Metallurgical Engineer-

3:40pm MODELLING OF BATH AND IMPURITY BALANCE FOR AN ALUMINIUM SMELTER: Weidong Zhang, Xiaoling Liu, Comalco Research Centre, PO Box 316,

1996 TMS ANNUAL MEETING AND EXHIBITION

B111

Thomastown, Vic. 3074, Australia; Peter McMaster, Mark Taylor, New Zealand Aluminium Smelter Ltd, Private Bag 90110, Invercargill, New Zealand Modelling of bath and impurity balance for an aluminium smelter was conducted to assist the smelter's upgrade program. A flowsheet model was developed for the smelter and was validated against an interim scenario. The purpose of the modelling was to examine the effects on metal purity and compositional control of a change from primary to reacted alumina and to determine operational strategies to minimise the impact on high purity production. 4:10pm MEASUREMENTSOFCF4ANDC2F6EMISSIONSFROMPREBAKEDPOTS: G. Bouzat, J.C. Carraz, M. Meyer, Aluminium Pechineny - LRF. B.P. 114,73303, St Jean De Maurienne Cedex, France. Emissions of CF4 from 300 kA Centre Worked Pre-Baked Point Fed (CWPB-PF) pots and emissions of CF4 and C2F6 from 130 kA Side Worked Pre-Baked (SWPB) pots were measured with an Infrared Photoacoustic Spectrometer. Gases were collected from gas ducts connecting the pots to the dry process treatment centres. Numerous measurements of CF4 and C2F6 generation were carried out and related to anode effects. No CF4 ~or C2F6 was observed during normal pot operation for a detector limit of 0.1 mglNm . The relation between fluorocarbon generation and anode effect characteristics is discussed; the anode effect frequency and anode effect duration are not fully representative of the fluorocarbon generation. It is shown that anode effect overvoltage is the relevant parameter since it integrates the fluctuation in voltage during anode effect treatment.

CARBON TECHNOLOGY IV: Cathode/Plant Operation

Sponsored by: LMD Aluminum Committee Program Organizer: Lise Castonguay, Alcan Intemational Ltd., Arvida Research and Development Centre, PO Box 1250, 1955 Mellon Blvd, Jonqui~re, Qu~bec, Canada G7S4K8

Wednesday, PM February 7, 1996

Room:A7 Location: Anaheim Convention Center

Session Chairperson: Dr. B~n~dicteAllard, Carbone Savoie, 30, rue Louis Jouvet, B.P. 16, F-69631, V~nissieux Cedex, France

2:00pm RESIN BINDERS IN RAMMING PASTE: K. Kvam, H. A. Oye, Institute of Inorganic Chemistry, Norwegian Institute of Technology, University ofTrondheim, N-7034 Trondheim, Norway; J. A. Johansen, R. Ugland, Elkem als Carbon, PO Box 40 Vaagsbygd, N-4602 Kristiansand, Norway Resin bonded carbon refractories avoid the emission ofPAH associated with tar based binders. Five prototype resin binders were tested and compared with two reference binders, one tar and one resin based. The binders were tested in the laboratory as well as in actual operation. For the optimum formulation, the mechanical strength was reasonably high and the baking shrinkage was low. Even if resin binders have a glassy structure, the sodium resistance was satisfactory. The viSCOSity of the resin binder can be adjusted to provide the desired range of temperature of use for the ramming paste. 2:30pm OPTIMIZATION OF ANTHRACITE CALCINATION PROCESS IN A VERTICAL ELECTRIC ARC FURNACE: Jean Perron, Alcan International Ltd .. PO Box 1250,1955 Mellon Blvd, Jonqui~re, ~bec, Canada, G7S 4K8; Jean-Fran~ois Bouvette, Alcan Smelters and Chemicals, Centre des Produits Cathodiques, PO Box 1500, Jonqui&e, Qul!bec, Canada, G7S 4L2; Marc Dupuis, Gl!nisim, 3111 Alger Street, Jonqui~re, Qul!bec, Canada, G7S 2M9 Anthracite used for cathode block manufacturing must be calcined prior its use to eliminate volatile matter and reduce its electrical resistivity. Anthracite calcination in vertical electric arc furnaces is widely used in the industry. It is well known that this technology leads to a radial temperature gradient in the furnace which results in nonhomogeneous calcined anthracite properties. Also, operation experience has shown that production of calcination furnaces can be difficult to stabilize if process changes occur, which may lead to variable quality of the calcined material. To optimize the calcination process, an improved control strategy and a mathematical model of E1kem-type electric arc furnace were developed. Using these, it was successfully demonstrated that temperature gradient in the furnace can be reduced and that furnace productivity can be substantially increased. 2:50pm INTERFACIAL OXIDATIVE PROCESSES OF CARBON ELECTRODES AND LINERS AT LOW OXYGEN POTENTIALS -AN EXPERIMENTAL METHOD: B. J. James, M. M. Hyland, B. J. Welch, Department of Chemical and Materials Engineering, The University of Aukland, Private Bag 92019, Auckland, New Zealand; J. R. Mittag, SGL Carbon GmbH, Griesheim, 0.65933, Frankfurt, Germany Oxidative reactions of carbon are frequently characterised by an ignition temperature determined for example, by thermogravimetric analYSis. However, the heterogeneous nature of carbon materials makes this method insensitive. Ignition in furnace liners is complicated by select oxidation. An experimental apparatus has been designed using a low thermal mass furnace in which samples may be heated in an atmosphere with controlled low oxygen partial pressures, similar to those that might exist in the cathode cavity of a reduction cell. Product gases are led to a mass spectrometer for sensitive detection of oxides; a combined thermogravimetric apparatus is then used in order to perform a continuous materials balance. Results show slow reaction occurs well before

B112

accepted ignition temperature, the rat'~ displaying a time/temperature interdependence. After a certain temperature the oxygen partial pressure appears to become the limiting factor (e.g. 650°C for a fully graphitic sample). This paper presents description of the method and apparatus used with results and comparison with reported ignition temperatures. 3:10pm AUTOGENOUS GRINDING FOR BATH SCRAPS RECYCLING: Andre Pinoncely and Philippe Podda, FCB Mineral ProceSSing and Carbon Plant Division, B.P. 24 - 69702 Givors Cedex, Franct' In the early 80's, FCB designed an original process for the recycling of bath scraps in aluminium smelters, using a single stage fully air-swept autogenous mill. Since then, the 9 industrial references confirmed and even exceeded the expectation in terms of dust-free and easy to run operation, high recovery ratio of bath among the metallic scraps, and low maintenance cost. Ptoblems encountered on conventional processes belong to the old days and new projects tend to give an increasing importance to classiftcation and storage of crushed products, autogenous grinding being already recognised as the most suitable, simple and reliable process route. Present paper describes this original process and draws up the overall performances of ten years of experience. 3:30 pm BREAK 3:50pm AN IMPROVED APPROACH TOl'HE CONSTRUCTION OF AIR CLEANING ARRANGEMENTS FOR CARBON ANODE PLANTS: L. L. Basov, S. F. Egorov, G. N. Kuz'min, H. Post, V. A. Sverdlin, Russia; 198147, St-Petersburg, Malodetskoselsky av, 26-7, HACA Technocentr, St-Petersburg Carbon anodes for aluminum production are fabricated with coal-tar pitch. When pitch is heated to 200"C, volatiles mainly composed of polynuclear aromatic hydrocarbons (PAH) are emitted. Central ventilation with a scrubber at the exit is the usual way of cleaning the air of the working environment. The existing air-cleaning systems are massive, technically complicated, metal consuming and their efficiency is poor. The duct arrangement from the working area to the scrubber cannot guarantee the required reduction of PAH concentration in the working areas and any failure of the system necessitates interruption of the anode production. An improved concept of air cleaning system which includes the same local equipment based on a plasmachemical method is proposed. Experience has shown that 'Iir-cleaning efficiency is improved and energy consumption is reduced. This new coocept has the following advantages: simplicity, small size and weight and high reliability. Modules can be installed at the source of gas emission with individual air intake system. Due to the high cleaning efficiency of the system, air can be recirculated in workiing areas or emitted directly to the atmosphere.

CAST SHOP TECHNOLOGY VI: D C Casting

Sponsored by: LMD Aluminum Committee Program Organizer: Julian V. Copenhaver, Technical/Quality Manager, NSA A Division of Southwire, PO Box 500, Hawesville, KY 42348 Wednesday, PM February 7,1996

Room:AIO Location: Anaheim Convention Center

Session Chairperson: Dr. Wolfgang Schneider, YAW Aluminium AG, George-VonBoeselager-Strasse, 250.53117 Bonn, Germany

2:00pm REAL-TIME VIDEO MONITORING OF INGOT CASTING: Ray T. Richter, P. R. Adomaitis, G. J. Hildeman, Alcoa Company of America, Alcoa Technical Center, Alcoa Center, PA 15069 During start-up and steady-state casting of ingots, a number of effects related to ingot casting variables can be observed on the surface of an ingot. For example, the amount of curl at the bottom of the ingot, surface laps, Iiquation, cracks, and folds, as well as extreme events such as "bleed-outs" whl:n molten metal melts through the shell of the ingot crater, are significant surface effects which could be viewed. Unfortunately, observation of these surface effects is difficult since the as-cast surface is hidden below the casting table, and direct visual obsef1lations of ingot surfaces are limited due to the proximity of the molds, cooling water sprays, and the ability of casting operators to safety view ingot surface details at close range. The purpose of this paper is to describe a unique video monitoring capability which is being used at Alcoa's Advanced Vertical Casting facility, to observe in real-time, surface effects of an ingot during casting. A deSCription of the development of the video monitoring system, including camera,lighting, and video recording capability will be presented. Examples of various surface conditions on direct chill (DC) and electromagnetic cast (EMC) ingots will also be discussed. 2:20pm A PRACTICAL MODEL FOR DETI:RMINING HEAT LOSS IN A MOLTEN METAL DISTRIBUTION SYSTEM: 1. Martin Ekenes, Casting Consultant, North 3418 Arden Road, Otis Orchards, WA 99027 The casting of aluminum ingots for wrought and other applications necessitates transferring molten metal from furnace to mold. This is ordinarily accomplished via refractory-lined troughS. In a multi-strand casting system, substantial heat losses can occur even between casting positions. In order to conserve energy and minimize process and product variability, it is desirable to minimize heat losses at each stage of molten metal transport. Troughing design plays an im(lortant role in minimizing these losses. This paper describes a practical model for determining heat losses in multi-strand casting systems. Predicted results are presented for various trough designs and compared with operating data.

1996 TMS ANNUAL MEETING AND EXHIBITION

2:40pm NON·CONTACT LEVEL CONTROL CASTING OF AA5182: Dr. Bjorn Rune Henriksen. I. J. Thorvaldsen. J. Hayes. Elkem Aluminium Mosjoen. PO Box 566. N· 8651 Mosjoen. Norway Elkem Aluminium Mosjoen has developed a non-contact metal level control system using lasers. The goal was to control the metal level in each mould within ±2 mm. The metal level is continuously monitored and controlled during casting. lYpically. the metal level varies within iOJ mm and iO.25 mm at maximum. The system has the ability to cast with low metal level in order to improve surface quality. and hence reduce the edge cracking of ingots during hot rolling. The automatic start-up of the cast· ing improves the safety of the operators. The total cost of the project was approxi· mately $400.000. 3:00pm CHANGING AN INGOT PROFILE BY COMPRESSING THE MOLD TO A DIFFERENT DIMENSION: Brad S. While/ealher Krumme. Senior Casting Metal· lurgist. Ravenswood Aluminum Corporation. PO Box 98. Ravenswood. WV 26164 For the past few years. RAC has had a problem with a high number of rescalps on our 24" x 55" ingot size. It was discovered that our molds were designed to cast at faster speeds than where we are currently casting. Instead of our ingots have a nice flat profile (24" thick). this design problem is giving us an ingot which is 114" bigger in the center than on the sides. We are in the process of having some of our molds compressed to a smaller dimension in order to get a flatter ingot profile and reduce the rescalp problem. This paper will compare the ingot profile before and after the molds were compressed and the impact on rescalp rates. 3:20pm QUANTITATIVE EVALUATION OF INGOT BUTT COOLING CONDITIONS REQUIRED FOR THE OPTIMIZATION OF D.C. SHEET INGOT START·UP PRACfICES: A CASE STUDY: }\res Caron. Andr~ Larouche. Alcan International Limited. Arvida Research and Development Centre. 1955 Mellon Boulevard. PO Box 1250. Jonquiere. Qu~bec. Canada G7S 4K8 During a cast start-up. the ingot butt is subjected to severe quench conditions resulting from direct water cooling. These quenCh conditions. which were determined through experience acquired by technical and production personnel. are based on safety and the production of a defect-free ingot butt. In spite of the precautions taken to ensure a flawless cast start-up. unforeseen defects may occur. The effectiveness of the corrective measures to be brought to the casting practice. therefore. greatly depends on our understanding of tbe phenomenon in question. This paper presents an experimental approach to understand the formation of unusual subsurface ingot butt defects observed on can body stock. It describes the quantitative technique developed to evaluate ingot butt cooling conditions during cast start-ups as well as their correlations to the subsurface defect observed. Finally. it describes how quantification of the ingot cooling conditions is essential in implementing the appropriate corrective actions required to eliminate the defect. 3:40 pm BREAK 3:50pm 3-D MODELING OF THE START·UP PHASE OF DC CASTING OF SHEET INGOTS: G.-U. Griin. Dr. Wolfgang Schneider. YAW. aluminium AG. Georg-vonBoeselager-Strasse 25. 0-53117 Bonn. Germany Three-dimensional transient simulations of the coupled fluid flow and heat transfer problem during the filling of the mold and the beginning of starting block movement have been used to investigate the influence of distributors and start-up programs on temperature and solidification front. As butt curl and possible hot tearing of the ingot are critically controlled by the temperature development of the metal in the mold. and the temperature itself depends on mold filling and start-up strategy. quality is strongly connected to appropriate metal distribution conditions in this early start-up phase. The calculated filling behavior and temperatures are discussed in dependency of parameters like distributor geometry or ingot size. The simulations support the development of adjusted metal distribution systems. 4:10pm 3-D SOLIDIFICATION ANALYSIS OF THE INITIAL STATE OF D.C. CAST· ING PROCESS: Yoshio Walanabe. Norifumi Hayashi. Teruo Uno. Sumitomo Light Metal Ind .• Ltd .• Technical Research Laboratory. Process Technology Department. 112. 3-chome. Chitose. Minato-Ku. Nagoya. 455. Japan The computer simulation program which considered motion of the ingots was developed to solve thermal problem associated with the initial non-steady state of D.C. casting process. Boundary conditions were investigated by casting tests of pure AI slabs. Heat transfer coefficient between ingots and cooling water (H). which was important to simulate the initial state. was affected by boiling phenomena and depended on surface temperature of the ingots. However. temperature dependency of H in the region where cooling was impinged on the ingot surface differed froni that below the region. It became clear that H should be defined as different functions of temperature in two regions. Sump profiles which were calculated with H proposed in this study airly agreed with measured profiles in both the initial and steady states. The agreement between calculation and measurement under the different casting conditions was also confirmed. 4:30pm INGOT BUTT CURL CONTROL OF ALUMINUM INGOTS THROUGH THE USE OF SHIM MATERIAL: Brad S. Whilelealher Krumme. Senior Casting Metallurgist. Ravenswood Aluminum Corporation. PO Box 98. Ravenswood. WV 26164 RAC has experienced a reduction in butt curl by as much as three inches through the use of shim tape. A factor that affects butt curl is how the shim material is applied. For example. the amount of butt curl is significantly less if shim tape is applied to all four faces of the molds as compared to only two faces of the mold. This paper will compare the amount of butt curl versus (I) different types of shim material and (2) application (four faces versus two faces) of the shim material.

1996 TMS ANNUAL MEETING AND EXHIBITION

4:50pm RECENT DEVELOPMENT IN LARGE FORMAT MAGNESIUM CASTING: Bj¢rn Killilsen. Norsk Hydro a.s .• Research Centre Porsgrunn. PO Box 2560. N-3901 Porsgrunn. Norway; Peter M.D. Pinfold. Hydro Magnesium Marketing. 5. Place Ville Marie. Suite 1537. Mon~al. Quebec. Canada H3B 2G2 Safety concerns from customers led to a decision to replace magnesium sows in the mid-eighties. The introduction of DC-cast T-bars and cylinders overcame these safety concerns. but DC-casting imposed productivity limitations in the magnesium foundry. as casts were limited to one or two strands only. The transition to today's multi-strand. hot-top. level pour. DC-casting technology. which addresses these productivity needs and also produces an improved product surface. is documented. This technology. when combined with efficient particle refining. offers a wide range of cavity free products. with high degree of intemal cleanliness. than was the case with sows. 5:10pm DEVELOPING OF MATHEMATICAL MODEL FOR CASTING 2024·2025 ALLOY BILLETS: A. Zorrilla. Heclor Medina. CVGVENALUM. Gerencia de Tecnologia de Producci6n. Ciudad Guayana Estado Bolivar. Apdo. 312 Actually. the 2024-2025 alloys have a high demand on the world market for automobile and aircraft's aluminum parts. Venalum has developed process parameters for producing 2024-2025 alloy billets in a conventional vertical casting process. In order to obtain the casting parameters the ternary diagrams AI-Cu-Mo and AI-Cu-Mg were studied to know eutectic temperatures and solidification phases. Also the solidification isotherms were obtained for horizontal casting to observe the thermal profile of these alloys. Using a mathematical finite difference model. the theoretical casting parameters were established as a reference to melting at experimental level in the D.C. process.

CONCURRENT ENGINEERING APPROACH TO MATERIALS PROCESSING

Sponsored by: Jt. EPDIMDMD Synthesis, Control and Analysis In Materials Pro· cesslng Program Organizer: Anand J. Paul. Concurrent Technologies Corporation. 1450 Scalp Avenue Johnstown. PA 15904 Wednesday. PM February 7. 1996

Room:BI Location: Anaheim Convention Center

Session Chairperson: Anand J. Paul. Concurrent Technologies Corporation. 1450 Scalp Avenue. Johnstown. PA 15904

8:30am MICROSTRUCTURE PATH PLANNING DURING CONSOLIDATION OFTI· TANIUM METAL MATRIX COMPOSITES: Ravi \tmcheeswaran. David G. Meyer. Haydn N.G. Wadley. IPM Laboratory. School of Engineering and Applied Science. Charlottesville. VA 22903 The consolidation of fiber reinforced metal matrix composites is a key step in the processing of high performance composite systems. A simulation tool has recently been developed for predicting the evolution of the composites internal state microstructural variables (like relative density. fiber microbendinglfracture. and chemical reaction layer thickness at the fiber-matrix interface) during consolidation processing. Here. a Generalized Predictive Control (GPC) scheme is used in conjunction with the simulation tool to design process paths that minimize fiber damage and the reaction product thickness while simultaneously maximizing the composites relative density. The microstructural evolution depends nonlinearly on the process conditions and therefore the scheme uses a constantly updated linearization of the model to optimize the input process variable vector. A control simulation is presented to illustrate the performance of the control scheme and to assess the feasibility of its use for the achievement of goal state microstructures at the completion of consolidation cycles. Electrical and Computer Engineering Dept.. University of Colorado. Boulder. CO 80309-0425. 8:55am MICROSTRUCTURE TRAJECfORY OPTIMIZATION DURING CONSOLI· DATION OF POROUS TITANIUM MATRIX COMPOSITE MONOTAPES: R. Vancheeswaran. R. Gampala. H. N. G. Wadley. School of Engineering and Applied Science. University of Virginia. Charlottesville. VA 22903 The high temperature consolidation of fiber reinforced titanium matrix composites (TMC's) seeks to reduce the concentration of matrix pores (Le .• increase relative density). while simultaneously minimizing fiber microbendinglfracture. and the thickness of reaction products and the fiber-matrix interface. These three "microstructure" goals have conflicting dependencies upon the consolidation processes variables (time varying temperature and pressure) and it has been difficult to identify process pathways by "trial and error" that lead to composites of acceptable microstructure. The availability of high performance computers. powerful non-linear programming tools and experimentally validated unified dynamic consolidation models have paved the way for finding a "globally optimal" trajectory of process variables that will eliminate porosity. limit both fiber damage and reaction product thickness while minimizing the process· ingtime. 9:20am METAL MATRIX COMPOSITE MONOTAPE DENSIFICATION DUE TO COMBINED CREEP AND PLASTICITY OF ASPERITES: R. Gampala. Concurrent Technologies Corporation. 1450 Scalp Avenue. Johnstown. PA 15904; R. Vancheeswaran. H.N.G. Wadley. School of Engineering and Applied Science. University of Virginia. Charlottesville. VA 22903 Previous models of the consolidation of metal matrix composite monotapes have

B113

been based upon micromechanical analysis of contact deformation by rate independent plasticity and power law creep (i.e. steady state creep). 'These models have recently been utilized to simulate and optimize consolidation process cycles so that they steer the MMC microstructure to a user defined goal state. However, the instantaneous response of the plasticity mechanism caused finite discontinuities in the state vector field and thus the large body of results for trajectory optimization scbemes that required continuous vector fields could not be utilized. To overcome this problem, a unified creep/plasticity constitutive model is used to derive the contact deformations for monotape consolidation. 'The contact blunting of a bemispherical asperity is first analyzed and its response represented by relating (i) the contact area to the blunting strain and (ii) the mean contact pressure to the blunting rate. This unified blunting problem is than numerically solved using finite element analysis, and results presented for two composite systems of interest (Tt-24Al-llNblSCS-6 and Tt-6AI-4V/SCS-6). 'The resulting model is amenable to the application of non linear programming methods to "globally" optimize the consolidation trajectory. 9:45am MULTI-LEVEL SIMULATION OF BRIDGMAN GROWTH OF ~-NIAl AND CdTe CRYSTALS: Hong Ouyand, Wei Shyy, Department of Aerospace Engineering, Mechanics and Engineering Science, University of Aorida, Gainesville, FL 32611 A computational model has been developed for the Bridgman growth processes of~­ NiAl and CdTe crystals. 'The model accounts for beat transfer around the heater and the ampoule, and phase change dynamics. To handle the geometrical and physical complexities of the crystal growth processes, a two-level approach has been developed. At the global furnace level, combined convection/conduction/radiation calculations with realistic geometrical and thermal boundary conditions supplied by the global furnace simulations. 'The present multi-level model can help improve the predictive capabilities for crystal growth techniques by optimizing the use of the computing resources; it allows one to probe the effects of different physical and geometrical variables on the crystal qUality. 10:10 am COMPUTATION OF MAGNETICALLY SUPPORTED FREE SURFACE BY A COUPLED FINITEIBOUNDARY ELEMENT METHOD: S. S. Ping, B. Q. U, Dept. of Mechanical Engineering, Louisiana State University, Baton Rouge, LA 70803 A computational methodology based on the coupled finite element and boundary/ element methods is presented for predicting the free surface shapes of liquid droplets and cylinders supported by the electromagnetic field. This is an extension of our previous FEmE model for induction calculations. 'The free surface algorithm is developed based on the weighted residual method, the same that forms as a basis for nonlinear finite element formation. 'The computational scheme is iterative but robust and efficient and is capable to simulating complex shape deformations. Detailed algorithms will be given. Computational results are presented for magnetic levitation under both normal and microgravity conditions and compared with available experimental measurements. 10:35 BREAK 10:45 am CONCURRENT ENGINEERING OF DIE CASTING PROCESS PARAMETERS AND TOOLING: Yong K. Park, Paul Paliani, Jerald Brevick, Dept. of Industrial Welding and Systems Engineering, 'The Ohio State Univeristy, Room 210, Baker Systems Building, 1971 Neil Avenue, Columbus, OH 43210 'The horizontal cold chamber die casting process can produce a wide range of nearnet shape, non-ferrous castings with small dimensional variation, excellent surface finish and very complex geometries. However, top quality die casting can be achieved only when optimal performance of the cold chamber and plunger is maintained during the molten injection phase of the process. Cold chambers (shot sleeves) are known to deteriorate fast and sometimes fail catastrophically because of poor design. Thermoelastic distortion and catastrophic fracture are leading causes to degrade quality, increase downtime and thereby decrease productivity/profitability. 'Therefore, a systematic approach to minimize irregular, adverse conditions which prevent an optimal injection profile, is highly desirable. To eliminate them, design criteria for shot sleeves (mechanical and thermal) were developed using analytical tools, numerical process simulation and physical elements. Also, the relationship between die casting process parameters and deteriorationlfailure mechanisms were carefully evaluated. 11:10 am DEEP DRAWING OF CONICAL FRUSTUM SHAPED CUPS - A COMPUTER SIMULATION: Ch. A. V. Prasad, R. C. Chaturvedi, Dept. of Mechanical Engineering, Indian Institute of Technology, Bombay 400076, India The authors have developed a theory of deep drawing of conical frustum shaped cups. This theory has been used in the development of a dedicated software for modeling the process. The software developed provides full information on the state of stress and strain throughout the formed sheet at successive stages of forming under a specified conditions of processing. It also provides a graphical display of the progress of the forming process. Different modules of the software such as those for input of geometry, tools, process, and material parameters, and calculations of geometry, forces, stresses, and strains at different stages of the process are described along with those for graphical display of the forming process and for displaying the results. 'The paper also gives the validation procedure used and sample outputs showing the influence of some of the process parameters on the formability of such shapes. 11:35 am ANALYSIS OF THERMAL SPRAY SYSTEMS BY COMPREHENSIVE NUMERICAL SIMULATIONS: Xiaolong Yang, Shmuel Eidelman, Science Applications International Corporation, 1710 Goodridge Drive, McLean, VA 22102 A comprehensive numerical capability for the analysis of Thermal Spray (TS) systems consisting of a three-dimensional compressible, turbulent flow Navier-Stokes model and a multi-phase particle flow model in Lagrangian formulation, was developed. This capability is validated for a set of flow conditions typical for TS. The developed meth-

B114

odology is used for analysis of the JP-5000 TS gun gas and coating powder flow conditions. Modeling and analysis of particle/turbulent eddies interaction is performed. The simulations allow evaluation of particle density, velocity, and temperature distribution in the TS gun's flow field and optimization of these parameters for particular coating conditions. This methodology can be used for process control and reliability improvement, and has the potential to signifi(:antly improve TS gun performance through control of its parameters. 12:00 pm COMPUTER MODELS FOR ROLLING, BENDING, AND FORGING OF SINGLE CRYSTAL SUPERALLOYS: D. Zhao, S. Cheng, R. Thomas, Concurrent Technologies Corporation, 1450 Scalp Avenue, Johnstown, PA 15904 Single crystal superalloys have long been used in cast form to make gas turbine engine components. However, the possibility of hot forming cast single crystal superalloys has been ignored for the most pwrt because of their low ductility. Successful forming of single crystals would open other areas of application in aerospace, marine and land based powder generation. Forming of single crystals could provide single crystal sheets for applications such as combustor components. Furthermore, larger components could be fabricated out of single crystal superalloys using these enhanced technologies, opening up the possibility of their use in more powerful combustors. 'The hot forming technologies for single crystals which are being developed include hot rolling, bending and forging. To significantl:1 shorten the time between the development of optimized manufacturing routes and their implementation on the shop floor, computer modeling was performed for these fabrication processes as applied to gas turbine engine components. Modeling offorming processes was closely integrated with materials testing. The results of computer modding, materials testing, and laboratory scale trial runs are presented.

DEFORMATION II

TMS General Abstract Session Wednesday, PM February 7, 1996

Room: Orange County I Location: Anaheim Marriott Hotel

Session Chairperson:TBA 2:00pm TENSILE BEHAVIOR OF ALLOY:800HAS AFFECTED BYAGEING,STRAIN RATE AND NOTCHES: A. M. Eleiche, Mech. Eng. Dept. of Mech. Design & Production, Cairo University, Egypt; M. F. Aly, National Center for Atomic Safety, Cairo, Egypt; M. A. Adly, Dept. of Mech. Design & Production, Cairo, Egypt; M. M. Ghoneim, Atomic Energy Authority, Egypt 'The objective of this work was characterizing the mechanical behavior in tension of INCOLOY 800H at the amhient tempc:rature. Quasi-static tests were performed using an electro-mechanical universal testing machine, while a tensile split-Hopkinson-bar apparatus was used in the dynamic test~. 'The testing program was deSigned to elucidate the following: (a) 'The effect of material ageing: by carrying out tests at strain rates of 7.5 E-5 and 7.5 E-3/s on solution-treated material and on material thermally-aged at 9OO'C for 24 hrs. (b) The effect of strll.in rate: by carrying tests in the range 7.5E-5 to 400 /s on solution-treated material. (c) The effect of strain-rate history: by quasi-statically deforming solution-treated specimens to strain amplitudes of 5, 7.7, 11.3 and 13.6 % foHowed by dynamic reloading.(d) The effect of notches: by carrying tests at 7.S E-S/ s on specimens, in both the solution-trf'ated and the aged conditions, having two types of notches, namely circular and B.S. notch. Reference: M. F. Aly, The behaviour of Alloy 800H at 20' C using a novel tensile split-Hopkinson bar apparatus, M.Sc. Thesis, Dept. of Mechanical Design and Production, Faculty of Engineerinq, Cairo University, 1993. 2:20pm TEMPERATURE AND GRAIN SIZE DEPENDENCE OF MICROSTRUC· TURALCREEPMECHANISMS IN APIM NI-BASE SUPERALLOY: Dr. Charles P. Blankenship. Jr., Dr. Ernest L. Hall, GE Corporate Research & Development, P.O. Box 8, Schenectady, NY 12301 Creep deformation studies were conducted in polycrystalline Rene 88 at 1200,1400, and 1500°F. Samples with grain sizes of 7, 15, and 9D\.Lm and identical 1 sizes and distributions were tested at each temperature in order to isolate the grain size contribution to creep resistance. After deformation to 0.5% and to failure, transmission electron microscopy was used to characterize the microstructure. Individual dislocation segments and extended faults are observed under all creep conditions. The density of these intergranular defects generally decreases with test temperature at a given grain size, but is not strongly affected by grain size. 'The amount of mass transport along grain boundaries, as evidenced by discontinuous 'Y' precipitation and grain boundary precipitate coarsening, increases with increasing temperature and decreasing grain size. The nature of the matrix dislocations and associated faults appears to be similar at 1400 and 1500°F, but different from those at 1200°F and room temperature. 'The relationship of these results to the observed creep behavior wiII be discussed.

2:40pm DEFORMATION BEHAVIOR ANn MECHANICAL PROPERTIES OF L12 ALLOYS BASED ON (NI,pt)3AI: J. L. Kamm, w: w: Milligan, Department of Metallurgical and Materials Engineering, Michigan TechnolOgical University, Houghton, MI49931 Ternary alloys based on Ni3AI, containing up to 25 atomic percent Pt substituted for Ni, were prepared. 'The alloys possessed the L12 structure, and were Single-phase. The temperature dependence of the mechanical properties was similar to that of Ni3AI, although the peak strength was higher and occurred at much lower temperatures. Dislocation behavior was also similar to Ni3AI, except tbe density of trapped screws at room

1996 TMS ANNUAL MEETING AND EXHIBITION

temperature was much higher for the Pt-bearing alloys. 1be results are consistent with increasing ease of cross slip with increasing Pt content, and not with solution hardening. Sponsored by the National Science Foundation under grant DMR-9257465. 3:00pm EFFECT OF COMPOSITION ON CRYSTALLOGRAPIDCTEXTURE IN HOT ROLLED AI-LJ-Cu ALLOYS: M. A. Przystupa, Dept. of Materials, Science and Engineering, University of California, Los Angeles, CA 90024; A. K. Vasud van, Office of Naval Research, Code-332, 800 N. Quincy St., Arlington, VA 22217; W. G. Fricke, Jr. (Deceased), Alcoa Technical Center, Alcoa Center, PA 15069 1be hot-rolling textures of ternary AI-Li-Cu alloys have been characterized to determine their effects on yield strength anisotropy.1be alloys used in the study had (LilCu) ratios varying from 0 to infinity and were hot-rolled in the 520-400'C temperature range. All alloys developed beta-fibers, with maximum intensity at Brass, and a weak background alfa-fiber. 1be Brass components were the strongest for the alloys with intermediate(LilCu)ratios. All alloy compositions showed weak Cube or rotated Cube recrystallization texture components. 1be texture results were used to predict yield strength anisotropies. 1be predictions were in agreement with the experimental results for the high (LifCu) alloys. In the low (LilCu) alloys, the discrepancies between the predicted and the measured yield strength anisotropies have been attributed to the modifying effect of the directional precipitates on deformation. 3:20 pm BREAK 3:30pm HIGH-TEMPERATURE RUPTURE OF 7075 AL ALLOY UNDER MULTIAXIAL STRESS STATES: AhmaJaIi Yousejiani, AI-Badrawy Abo Elnasr, FarghalJi A. Mohamed, James C. Earthman, Materials Science aud Engineering, Dept. ofChemicaI and Biochemical Engineering, University of California, Irvine, CA 92717 Experiments were performed on 7075 AI alloy under three different stress states: uniaxial tension, biaxial shear and triaxial tension. Rupture times are compared for the three states with respect to the maximum principal stress, von Mises effective stress, and the principal facet stress. 1be results of this comparison along with microstructural observations regarding the cavitation behavior of the alloy are discussed. 3:50pm CHARACTERIZATION OF THE DEFORMATION MECHANISMS IN PURE ALUMINUM AND COPPER AND Cu-2.2%AI AND Cu-4.5%AI ALLOYS DUR· ING TENSILE TESTING AT ROOM TEMPERATURE: Veronica Caballero, Arturo De La Cruz, S. K. Varma, Department of Metallurgical and Materials Engineering, 1be University of Texas at EI Paso, EJ Paso, TX 79968-0520 1be differences in the deformation mechanisms for the pure metals aluminum and copper and - the Cu-22%AI aud Cu-4.5%AI alloys during the room temperature tensile testing has been explored. 1be choice of materials provides for a stacking fault energy values in a range from 4 to 166 rnJfm2. 1be influence of grain size and strain rate on the development of microstructures during the deformation has been investigated. While pure aluminum, with a SFE value of 166 rnJfm2, deforms by a process of slip leading to the cell formation and cell size refinement, the mechanisms change from the development of cellular structure to slip band, microband and deformation twin formation as the SFE decreases, corresponding to those for pure copper to the very low values (up to 4 rnJfm2) for the dilute alloys. This research has been supported by the National Science Foundation through the grant number HRO-9353547. 4:10pm INVESTIGATION OF CD SEGREGATION AT BOUNDARIES IN THEPB-62% Sn EUTECTIC: Souping Yan, FarghaJli A. Mohamed, Materials Science and Engineeling, Department of Chemical and Biochemical Engineering, University of California, Irvine, CA 92717 A detailed microstructure investigation was performed on the Pb-62%Sn eutectic doped with 890ppm Cd. In conducting the investigation, SEM, EDS and AES surface analysis techniques were used. 1be data have shown that Cd segregation occurs along Sn-Sn intercrystalJine boundaries and Ph-Sn interphase boundaries. The data are discussed in reference to recent results regarding the effects of impurities on superplastic flow at low stresses. 4:30pm ELEVATED-TEMPERATURE DEFORMATION OF (Lal x,Srx)Mn03: J. Wolfensrine, Department of Chemical and Biochemical Engineering, University of California, Irvine, CA 92717-2575; M. A. Boling-Risser, K. C. Goretta, J. L. Routbort, Energy Technology Division, Argonne National Laboratory, Algonne, IL 60439-4838 1be compression creep behavior of fine-grained (La 1-xSrx)Mn03 (with x from 0.1 to 0.25) materials with relative theoretical densities hetween 85-90% was investigated over the temperature range 1150-1300°C. 1be fine-grain size, brief creep transients, stress exponent close to unity and absence of deformation induced dislocations, suggested that the deformation is controlled by a diffusional creep mechanism. A comparison of the activation energy for creep with existing diffusion and creep data for pervoskite oxides suggested, that the diffusional creep of (La l_x,Srx)Mn03 is controlled by lattice diffusion of either lanthanum or manganese. This is confirmed by the dependence of the steady-state creep rate on Sr composition.

EMERGING TECHNOLOGIES SYMPOSIUM: Engineering Science Of Chem-Mechanlcal Planarlzatlon

Sponsored by: EMPMD Thin FUm & Interfaces Committee Program Organizers: Professor Krishna Rajan, Materials Science and Engineering Department, Rensselaer Polytechnic Institute, Troy, NY 12180-3590

1996 TMS ANNUAL MEETING AND EXHIBITION

Wednesday, PM February 7, 1996

Room: Grand K Location: Anaheim Marriott Hotel

Session Chairperson: Dr. Michael Fury, Rodel Inc., 451 Bellvue Road, Newark, DE 19713

2:00 pm Invited A PROGRESS REPORT ON THE SCIENCE OF CMP: WHERE DO WE GO FROM HERE?: Frank B. Kaufman, IBM Research Division (FBK @ Watson.Res.Com), P.O. Box 218, Yorktown Heights, NY 10598 1be implementation of Chemical-Mechanical Polishing (CMP) processes into high volume semiconductor manufacturing on a schedule (for some users) that is significantly more aggressive than the SIA roadmap, means that scientific investigations into this novel technology must he both accelerated in time and more focussed in scope. Drawing from published work in the open scientific literature, and from recent results in the author's laboratory, we will highlight the research areas which currently appear to offer significant leverage in terms of (a) enhancing manufacturability of the CMP technology, (b) identifying potential paradyne shifts to current CMP processing, and, (c) providing a foundation for robust modelling capabilities in the area of polishing pads, and oxide and metal CMP. We will conclude by suggesting ways in which suppliers, universities and the national laboratories can work more effectively together to do targeted research on the relevant science behind CMP. 2:30pm PAD DEGRADATION AND CONDITIONING STUDIES FOR CMP OF INTERLAYER DIELECTRIC FILMS: K. Achurhan, S.V. Babu, Department of Chemical Engineering, Clarkson University, Potsdam, NY; J. Curry, R.H. Strasbaugh Inc., San Luis Obispo, CA; D.R. Campbell, Center for Advanced Materials Processing, Clarkson Univ., Potsdam, NY As integrated circuit technology approaches the era of Ultra Large Scale Integration (ULSI), chemical mechanical polishing (CMP) is being introduced into the mainstream of semiconductor processing to help accommodate the sub half-micron design ground rules. In this paper, an attempt to understand different aspects of CMP process with emphasis on the polish pad degradation and its conditioning on Silica films is sought. Chemical mechanical polishing is found to result from a complex interplay of many mechanical and chemical parameters. Degradation of polishing pad is one of the major shortcomings of the CMP process and significantly affects the overall process stability and cost of ownership. Pad conditioning helps maintain removal rate stability and thereby enhances the longevity of the pad. In this study, the effect of different pad conditioners and conditioning processes are discussed. 3:00 pm Invited CHALLENGES IN CMP OF MATERIALS FOR SEMICONDUCTING DEVICES AND CIRCUITS: Shyam P. Murarka, Center for Integrated Electronics and Electronics Manufacturing, Rensselaer Polytechnic Institute, Troy, NY 12180 Planarized surfaces alleviate problems associated with the lithography, overetching, thinning of the films during deposition on non-planarized surfaces, and asperities at the surface leading to potentially damaging effects on the electrical behavior of devices. In the recent past, the chemical mechanical polishing or planarization (CMP) has been demonstrated to be the preferred technique to achieve so-called global planarization at all levels. CMP, although practiced for thousands of years, has remained an art made useful, practical, aud beautiful more by experience than by scientifiC understanding. This paper will briefly review the CMP processes used or needed in microelectronics fabrication and the variables in the CMP process. This will be followed by an assessment of the scientific issues, including those associated with post CMP cleaning and passivation and the electrochemical phenomena, that must be addressed to make the art of CMP into a scientifically based teChnOlogy, useful for both small and large area planarization. 3:30 pm BREAK 3:50pm MECHANISMS OF TmN FILM OXIDE POLISIDNG: Krishna Rajan, Department of Materials Science and Engineering, Rensselaer Polytechnic Institute, Troy, NY 12180-3590 Chern-mechanical polishing of oxides is a critical component in the manufacturing steps of multi-level interconnects. 1be role of chemical and mechanical interactions on thin oxide films is fundamentally different than in bulk oxides. In this presentation, we outline these differences and discuss the fundamental mechanisms controlling material removal and its effect on properties of thin film dielectric oxides. 4:20pm THE IMPORTANCE OF PARTICLE SIZE TO THE PERFORMANCE OF ABRASIVE PARTICLES IN THE CMf PROCESS: Michael C. Pohl, Horiba Instruments, Inc., 17671 Armstrong Avenue, Irvine, CA 92714 One of the critical components in the CMP process is the abrasive used to achieve the planarization. 1be speed of the process, the degree of scratching and the ability to subsequently perform the post-cleaning operation are all determined by the abrasive particle size. A variety of techniques exist to measure the particle size, aud they must be carefully evaluated. It is critical that the size reported is indicative of the particles in use in the process and not the particles seen by the instrument. Several areas are critical to measuring particle size even hefore the sample is introduced to the instrument First is obtaining a representative sample for analysis. Another critical area is the dispersion of the particles in the liquid phase. This application is particularly critical since it is a concentrated slurry of very small particles. This issue will be discussed both theoretically and practically. Finally, application of particle size information to abrasive slurries will be discussed, as well as aging phenomena.

B115

GENERAL PYROMETALLURGY II

Sponsored by: EPD Pyrometallurgy Committee Program Organizer: Jerome P. Downey, Hazen Research, Inc., Golden, CO 80403; Ed Mounsey, Ausmelt Umited, 2113 Kitchen Road, Dandenong, Victoria 3175 Australia Wednesday, PM February 7,1996

Room:AI6-17 Location: Anaheim Convention Center

Session Chairmen: J. M. Toguri, Professor, Department of Metallurgy and Materials Science, University of Toronto, 184 College Street. Toronto, Ontario, Canada M5S lA4; Ed Mounsey, Marketing Manager, AusmeIt Umited. 2/13 Kitchen Road. Dandenong, Victoria 3175, Australia

2:00pm INTERFACIAL PHENOMENA BETWEEN ALUMINA, QUARTZ AND FAYALITE SLAG: J. R. Donald. H. Fukuyarna, J. M. Toguri, Department of Metallurgy and Materials Science, University ofToronto, 184 College Street, Toronto, Ontario, Canada M5S IA4 Traditionally in the nonferrous industry. refractory expenses were considered to be fixed costs because they constitute a relatively small portion of the total production expenditures and the product value is inherently high. However, in the current global economy, in order to remain competitive, all costs must be minimized. Hence, the extension of refractory life in the nonferrous sector has become an important issue. Surface and interfacial phenomena playa significant role in refractory performance. For example, the wettability of the refractory material with slag, matte, and metal influences infiltration and interfacial reactions. Yet, very little fundamental information regarding the surface properties of refractory materials and slag is available. Thus. in the present work, the sessile drop technique incorporating a high temperature furnace equip with x-ray fluorescence was employed to investigate the surface and interfacial phenomena of fayalite slag on solid alumina and quartz (Si~). Contact angles, interfacial tensions and surface tension of the fayalite slag are reported. In addition, electron microprobe analysis (EPMA) was utilized to determine the products of the interfacial reactions as well to investigate the diffusion of elements into and out of the refractory material. 2:25pm PHYSICAL MODELLING OF THREE-PHASE MIXING IN A COUNTERCURRENT REACTION LAUNDER (CCRL) PROCESS FOR METAL REFINING: K. Narayana Swamy, David G. C. Robertson, Kent D. Peaslee, Center for Pyrometallurgy, Department of Metallurgical Engineering, University of Missouri-Rolla, 215 Fulton Hall, Rolla, MO 65401-0249 Experimental results of three-pbase mixing in a physical model of a CCRL for metals refining, at high levels of bottom gas injection (up to a Q/A = 25 cm/min), are presented. The experimental results at low levels of bottom gas injection through a line of central, single tuyeres (up to a Q/A = 7.5 cm/min) have already been discussed in a previous paper. A thermal tracer technique was used to elucidate both longitudinal mixing and interphase heat transfer. The results indicate that even under conditions of significant bottom gas injection, it should be possible to maintain acceptably low longitudinal mixing (Os < 20 cm2/s and Ds/uL < 0.1) in the CCRL. Interphase mass-transfer coefficients of the order of 0.02 cm/s and higher are predicted from the measured interphase heat transfer coefficients using the mass and heat transfer analogy. Adequate rates of refining are predicted for an industrial CCRL operated under such conditions.

under the most oxidizing conditions with the finest size fraction. The matte with Ni/Cu = I and high iron content was foulld to be most reactive. The least reactive matte was with NilCu = 10 and low iron content; the coarse fraction particles did not react at all under the conditions studied. 4:10pm MINIMIZATION OF DUST GENERATION IN OUTOKUMPU FLASH SMELTING: D. M. Jones, Magma Metals Company, P. O. Box M. San Manuel, AZ 85631; W. G. Davenport, University of Arizona, Dept. of Materials Science, Tucson, AZ 85711 Dust generation data have been obtained from 19 industrial Outokumpu flash furnaces. Statistical analyses of the data have been carried out and the most significant factors affecting dust generation evaluated. The results indicate that dust production may most effectively be minimized by I) minimizing the distance between the molten slag and the furnace roof and 2) optimizing burner design and combustion parameters. 4:35pm VOLATILIZATION KINETICS OF MINOR ELEMENTS FROM MATTE: Xu Zhong and D. C. Lynch, Arizona Materials Laboratory. Dept. of Materials Science & Engineering, University of Arizonll, Tucson, AZ 85721 The theoretical limit to smelting rate and possibly the size of a smelter is dependent on the rate of heat transfer and the chemical kinetics of both sulfur removal and minor element volatilization. The kinetics of volatilization of minor elements such as As, Sb. Bi. and Pb from matte and white DlI~ta1 is being investigated. It has been shown that the molecular form of trace elements ill matte depends on the chemical potential of sulfur and the propensity for the minor ell~ment to form a sulfide. The impact of the variable molecular form of the minor elements on the volatilization kinetics is unknown and is under investigation in this study. The experimental technique is reviewed and results are reported.

A SYMPOSIUM DE,DICATED TO PROFESSOR JOHANNES WEERTMAN'S SEVENTIETH BIRTHDAY: Session VI: Fatigue & Fracture II

Sponsored by: Jt. SMDIMSD Nuclt'ar Materials and MSD Flow and Fracture Committees and FEMS (Federation of European Materials Societies) Program Organizers: R.J. Arsenault, Department of Materials Science and Nuclear Engineering, University of Maryland. College Park, MD 20742-2115; David Cole, CRREL, 72 Lyme Rd.• Hanover, NH 03755; Todd Gross. Department of Mechanical Engineering, University of New Hampshire. Durham, NH 03824; Gemot Kostorz, Institut fUr Angewandte Physik. ETH H6nggerberg. CH-8093 Zilrich, Switzerland; Peter Liaw, Department of Materials Scilence and Engineering, University of Tennessee, Knoxville, TN 37996-2200; Sivan Parameswaran, NRC-Institute for Aerospace Research, Ottawa, Canada KIA OR6; Howard Sizek. Inco Alloys International Inc., Huntington. WV 25705-1771 Wednesday. PM February 7.1996

Room: Orange County 3 Location: Anaheim Marriott Hotel

Session Chairpersons: Gemot Kostorz. Institut fUr Angewandte Physik, ETH H!lnggerberg, CH-8093 Zilrich, Switzerland; Peter Liaw, Department of Materials Science and Engineering, University of Tennessee, Knoxville, TN 37996-2200

2:50pm INTERFACIAL PHENOMENA IN THE COPPER-CALCIUM FERRITE SYSTEM: T. Sakai, S. W. Ip, J. M. Toguri, Department of Metallurgy and Materials Science, University of Toronto, 184 College Street, Toronto, Ontario, Canada MSS IA4 The Mitsubishi process is an innovative pyrometallurgical operation for the processing of copper-containing sulphide ores. In the converting stage, a calcium ferrite slag is used to ahsorb the iron oxide generated during converting. To better understand the operation variables in the furnace, the surface and interfacial tensions of the copper and calcium ferrite slag were studied at 1,3OO'C using a high temperature x-ray apparatus. The results indicated that dissolution of copper increases with oxygen potential. The surface tension of the slag decreased with decreasing calcium oxide content. However, the effect of calcium oxide on the interfacial tension between copper and slag is negligible. Increasing oxygen pressure has a negative effect on the interfacial tension. Using these reSUlts, the entrainment behavior of copper by gas bubble attachment in the calcium ferrite slag was determined. The results indicate that copper entrainment is unavoidable.

2:00 pm Invited EFFECTS ON THE"fty' RAFT STRUCTURE ON THE CREEP AND FATIGUE BEHAVIOR OF MONOCRYS'll'ALLINE SUPERALLOYS: H. Mughrabi, Universitlit Erlangen-Nilmberg, Martensstrasse 5, 0-91058 Erlangen, Federal Republic of Germany Nickel-base superalloys subjected to high temperatures under the action of an applied stress exhibit a marked directional coarsening of the "f fy' microstructure. The latter is characterized by the transfonnation of the original "f ty • microstructure, consisting of cuboidal "f' particles in a "f matrix, into the plate-like "f ty. raftstructure. The present work summarizes some of the results obtained in recent years in the author's research group in high-temperature c:reep and fatigue studies on monocrystalline, "f 'hardened nickel-base superalloys. In particular, attention shall be focussed on the following: 1) Development and consequences of"f ty • rafting during creep. 2) Formation of I' rafts during isothermal and thermomechanical fatigue. 3) Effect of different "f fy' microstructures (. cubes. rafts parallel or perpendicular to the stress axis) on the isothermal fatigue behavior.

3:15 pm BREAK

2:30 pm Invited FATIGUE BEHAVIOR OF SINGLE CRYSTALS AND POLY CRYSTALS: P. Lukds. L. Kunz. Institute of Physics of Materials. Academy of Sciences of the Czech Republic, 616 62 Bmo, Zizkova 22, Czech Republic Fatigue characteristics (S-N curves. cyclic stress-strain curves and Coffin-Manson curves) of single crystals and polycrystals are compared from the point of view of their convertibility using a simple conversion factors both for the stress and the plastic strain values. The S-N curves for single crystals of f.c.c. and b.c.c. metals were found to be in very good agreement with the S-N curves for polycrystals converted by the 1l1ylor factor M=3.06. The cyclic stress-strain curves for f.c.c. metals can be converted only to a fltst approximation, while the limited data for b.c.c. metals show a reasonable convertibility. The application of the conversion factor to the Coffin-Manson curves of f.c.c. metals does not lead to any satisfactory agreement. The explanation is based on the conSiderably different volume fraction of zones exhibiting higher cyclic slip activity in single crystals and in polycrystals.

3:45pm OXIDATION OF SYNTHETIC NICKEL MATTES UNDER SIMULATED FLASH SMELTING CONDITIONS: Satu JyrkOnen, M. Nopanen, E. Peuraniemi, S. Str!lmberg. A. Jokilaakso, Department of Materials Science and Rock Engineering, Helsinki University of Technology, FlN-02150 Espoo, Finland A single-particle laminar-flow technique was used to study oxidation of four synthetic nickel mattes, which were screened into three size fractions. Experimental conditions were: reaction gas from N2 + 21 to 75 vol% ~, gas preheating temperatures ranging from 500 to 1, lOOC. The four mattes were synthesized so that NilCu ratios of I and 10 were obtained with high and low iron contents. Chemical analysis of the particles was used to determine the oxidation degree of the materials, and optical and scanning electron microscopy were used to detect their morphology, internal composition and structure. The best sulfur removal from the synthetic nickel mattes was achieved

B116

1996 TMS ANNUAL MEETING AND EXHIBITION

3:00pm PHYSICAL PROCESSES OF FATIGUE CRACK CLOSURE OBSERVED NONDESTRUCTIVELY IN THE INTERIOR OF AI-LI 2090 SAMPLES: A. Guvenilir. S.R. Stock. School Materials Science and Engineering. Georgia Institute of Technology. Atlanta, Georgia 30332-0245 Understanding the physical basis of macroscopic. indirect measurements interpreted as fatigue crack closure requires nondestructive. microscopic quantification of the crack face separations as a function of applied load. Nondestructive sectioning by high resolution computed tomography allows in situ observation of the crack faces under applied load. and measurements of physical crack openings as a function of position on the crack face are summarized for three samples of AI-Li 2090. In situ loading during x-ray tomography allows direct observation of the position of crack faces at different loads. These measurements on the interior of the sample are discussed in terms of the grain morphologies and orientations. the crack face geometry. macroscopic measurements of crack growth rate and changes in load displacement curve. lbis research was supported by the US Office of Naval Research. and the tomography measurements were made in collaboration withI.H. Kinney's group at Lawrence Livermore National Laboratory and with M.D. Barker's group at Lockheed Missiles and Space Company. Some of the results were obtained at CHESS which is supported by the National Science Foundation. 3:20pm SOLID SOLUTION SOFTENING REVISITED: M.E. Fine. M. Meshii. Department of Materials Science and Engineering. Northwestern University. Evanston. IL 60208 Intrinsic solid solution softening has been well established experimentally in bee and hcp metals and ionic crystals. It has been pointed out by Weertman that the stress field of a solute atom exerts a couple force on a screw dislocation and. thus. reduces the activation energy of the screw dislocation overcoming the Peierls stress barner by the double kink mechanism. It is known that intermetallic compounds possess high Peierls stress increasing the flow stress. When the flow stress is near or greater than the fracture Slress. brittle hehavior results. Thus reducing the Peierls stress may be a way to make intermetallics less brittle. Alloy softening has been reported in some intermetalIics. The possibilities for ductilizing intermetallics by solid solution softening will be discussed, applying what has been leamed with bec and hcp metals. 3:40pm CORROSION FATIGUE BEHAVIOR OF CARBON FIBER REINFORCED EPOXY MATRIX COMPOSITES UNDER COMBINED HYDROSTATIC AND AXIAL LOADING: J. Stolk. S.T. Mear. H.G. Wheat. Offshore Technology Research Center and the Center for Materials Science and Engineering. The University of Texas at Austin, Austin, TX 78712; H.L. Marcus, Institute of Materials Science, University of Connecticut. Storrs. CT 06269 Stress relaxation and load control fatigue tests of unidirectional. [90]12. carbon fiber-epoxy matrix composites were performed in air. seawater. and distilled water on the as-received composite material to obtain baseline daldN vs. AK crack growth data. Samples which were soaked to saturation in seawater and distilled water were fatigued at ambient pressure in order to evaluate the effects of both moisture uptake and seawater chemistry on crack growth rates. In addition. both the as-received and saturated samples were fatigue tested inn elevated pressure (up to 4000 psi) seawater or distilled water to determine what effect. if any. the superimposed hydrostatic pressure had on the fatigue crack growth rate. Finally. post-failure scanning electron microscopy (SEM) was performed to determine the effects of the various test conditions on the failure modes of the specimens. The feasibility of using nuclear magnetic resonance spectroscopy (NMR) and magnetic resonance imaging (MRI) to characterize moisture uptake at the specimen crack tip was also examined. 4:00pm NEAR-THRESHOLD FATIGUE CRACK GROWTH BEHAVIOR OF NICKELTITANIUM SHAPE MEMORY ALLOYS: R.L. Holtz,Geo-Centers. Inc .• 109033 Indian Head Highway, Fort Washington, MD 20744; K.Sadananda, M.A. Imam, C.A. Greene. CODE 6323. U.S. Naval Research Laboratory. 4555 Overlook Avenue SW, Washington. DC 20375 Traditional interpretations of fatigue crack growth (FCG) mechanisms rely heavily on "crack closure" concepts. Among other things. plasticity and phase transformations have been invoked as contributing to crack closure. In order to delineate more clearly the role of plasticity and phase transformations on FCG. a comprehensive study of the FCG behavior of aNi 49.9% -li 50.1% alloy with a martensitic transformation temperature (Ms) of 80°C was undertaken. Fatigue crack growth rate measurements were performed at frequencies of 10 Hz for stress ratios. R. from 0.1 to 0.9. in both vacuum and air. at temperatures ranging from room temperature up to 120°C. spanning the thermoelastic and pseudoelastic regimes. The threshold and near-threshold crack growth behaviors are discussed in the context of the two-parameter model of Vasudevan. Sadananda, and Louat [Materials Science and Engineering. A188. 1 (1994)] which avoid ad-hoc crack closure arguments. 4:20pm MICROMECHANISMS OF ROOM AND ELEVATED-TEMPERATURE FATIGUE CRACK GROWTH IN GAMMA TITANIUMALUMINIDES: C. Mercer. M. Mehdirad. W.O. Soboyejo. Department of Materials Science and Engineering. The Ohio State University. 2041 College Road. Columbus. OH 43210-11 79 The micromechanisms of room-and elevated-temperature fatigue crack growth in a wide range of gamma-based titanium aluminnides are elucidated in this paper. The differences between the fatigue crack growth behavior at room- and elevated-temperature are explained by considering the implications of different crack-tip deformation and shielding modes. The possible influence of alloying on oxide-induced wedging! closure is also examined within a micromechanics framework. The implications of the results are discussed for alloy/microstructure design and fatigue life prediction.

1996 TMS ANNUAL MEETING AND EXHIBITION

4:40pm SELF-ORGANIZATION AND SPATIO-TEMPORAL DISSIPATIVE STRUCTURES IN FATIGUED METALS: M. Glazov. C. Laird. Department of Materials Science and Engineering. School of Engineering and Applied Science. University of Pennsylvania. Philadelphia. PA 19104-6272 Strongly nonequilibrium systems of different physico-chemical nature may exhibit a tendency to self-organization either in the form of emerging temporal rhythms. or spatial scales. or both. We demonstrate that fatigued metallic alloys fall into this behavior as well. In the first part of the work the "classical" Portevin-Le-Chatelier(PLC) effect will be discussed as example of a temporal dissipative structure. The existing models of the PLC-effect (Kubin and Estrin. Ananthakrishna. McCormick) will be discussed along with their limits of applicability. A-non-Iinear dynamical model explaining the cyclic analog of the PLC-effect-the "Yan-Hong-Laird bursts" will be developed. We will also explain the "Neumann bursts" occurring in fatigued metals under the conditions of spatial coupling into the system of differential equations modeling the temporal instabilities in fatigued metals allows one to describe different types of dislocation patterning. Using the so-called "reaction+diffusion" approach (the finite-size analog of the Walgraef-Aifantis model) we will demonstrate that dislocation patterning in fatigue metals can be understood as a result of competition between the grain size and the characteristic length scales of the model. Finally. the prospects of future work will be briefly discussed. 5:00pm EVALUATION OF FRACTURE CHARACTERISTICS FOR SICISIC COMPOSITES UNDER ruGH TEMPERATURE CONDITIONS BY ACOUSTIC EMISSION METHOD: S.T. Kim. Y.T. Lee. Department of Mechanical Engineering. Yeungnam University. 214-1 Daedong. Gyongsan. Kyongpook. Korea The temperature dependency of mechanical properties on woven-cloth SiC-fiber reinforced SiC composite materials. which were fabricated by injection molding method. was examined under slow speed (O.02mm1min) tensile test inn various temperatures and atmospheric condition. Acoustic emission (AE) signal was detected by piezoelectric AF sensor in real time. It was found that the AE source location method. utilizing the longitudinal wave velocity. was effective in identifying fracture points of the specimens tested at room temperature. However. this method was not effective at elevated temperatures. because the fracture occurred so fast by the oxidation of the materials. The strength degradation of these composites at elevated temperature was turned out to be due to the formation of Si~ and li~ oxide particles.

HIGH TEMPERATURE SUPERCONDUCTORS· SYNTHESIS, PROCESSING, AND LARGE SCALE APPLICATIONS VI: Processing

Sponsored by: Jt. EMPMDISMD Superconductlng Materials Committee, MSD Electronic, Magnetic and Optical Phenomena Committee, and FEMS (Federation of European Materials Societies) Program Organizers: U. Balachandran. Energy Technology Division. Argonne National Laboratory. Argonne. IL 60439; PaulI. McGinn. University of Noire Dame. Notre Dame. IN 46556; and Stuart Abell. University of Birmingham. Edgbaston. BirminghamB152TT, UK Wednesday. PM February 7. 1996

Room:B7 Location: Anabeim Convention Center

Session Chairpersons: S. X. Dou. University ofWollongong. Wollongong. NSW. Australia; A. Bhargava. The University of Queensland. Australia

2:00pm HOT-ROLLING AND HOT-COMPRESSING TREATMENTS OF Y123 AND BI2l23 BULKS: Q. Y. Hu. D. Yu. H. K. Liu. T. Chandra. S. X. Dou. Centre for Electronic and Superconducting Materials. University of Wollongong. Wollongong. NSW 2500, Australia ¥BCO 123 and Bi2223 bulks were hot compressed isothermally in a temperature range of 780-850°C under constant stress of 15-50 MPa for 60 minutes. The materials have also been hot rolled in a same temperature range by given a small amount of reduction. Hot worked specimens then were examined using scanning electron microscopy and hardness testing. It was found that the mass density of the bulk Y123 and Bi2223 superconductors has been increased by both hot compressing and hot rolling. lbis was achieved by reducing the enclosed voids in the materials under the pressure at high temperature. In addition of high mass density. texture structure in the superconductors has also improved by hot compressing and hot deformation which in tum improved critical current of both materials. However. the texture extent of the Bi-2223 compound was enhanced by the hot working more significantly compared to compound Y123. Isothermal hot compressing was found more efficient compared to hot rolling in terms of increasing the mass density and texture structure of the superconducting materials. 2:20pm SEARCH FOR NEW SUPER CONDUCTING MATERIALS: P. Majewski, Helsenbergstr. 5. 0-70569 Stuggart. Germany A summary of the research efforts. aiming the discovery of new superconducting materials. of the Max-Planck-Institut fur Metalforschung. Pulvermetallurgisches Laboratorium will be given. Results of our investigations of the systems M-Ba-Ca-Cuo with M =La. Sb, AI. and Bi-Sr-Y-Ca-Cu-O. as well as Y-Ni-B-C will be presented. 2:40 pm Invited MAGNETO-OPTIC IMAGING OF HIGH-Tc SINGLE CRYSTALS REVEALING FLUX FRONT INSTABILITIES AND MACRO-TURBULENCE: T. H. Johansen, M. Baziljevich. II. Bratsberg. Department of Physics. University of Oslo.

B117

PO Box 1048 Blindern, 0316 Oslo, Norway; B. Hauglin and G. P. Lafyatis, Department of Physics, Ohio State University, Columbus, OH 43210-1106 We have used the advanced real-time magneto-optic method, where a doped irongarnet film with in-plane anisotropy serves as flux indicator, to study the behavior of flux front dynamics in high-Tc single crystals. Aux front instabilities that develop into macro-turbulent behavior of the flux system was observed in plate-Shaped YBCO crystals when the field is applied normal to the plate (along the c-axis). The appearance of the instability is seen as a growing meandering of the flux annihilation zone separating the invading flux from an initially frozen-in distribution of opposite polarity. The characteristic fingering structure is seen to develop with a maximum amplitude near the middle of the sides of the sample leaving the comers essentially unperturbed. We report on our observation of this phenomenon, displaying a variety of new effects which include the influence of a transport current on the turbulent behavior. 3:00pm BI-BASED 2213 SUPER CONDUCTING CERAMICS PREPARED BY THE GLASSY "MATRIX" PRECURSOR METHOD: S. Stassen, P. Rouxhet, A. Vanderschueren, R. C1oots, A Rulmont, M. Ausloos, S.U.P.R.A.S.-University of LiegeB6-Sart-TIlman-4000 Liege-Belgium Glassy "matrix" precursor method has been followed in order to prepare Bi l.7Pbo.3Sr2Ca2Cu30y superconducting materials. This method involves the mixture of a glass precursor and a crystalline precursor. Five crystalline precursors with two starting stoichiometries have been studied. We have first analyzed the successive formation of various phases during the thermal treatment which showed that glass compound crystallizes frrst, 2201 phase is than formed, followed by 2212 and then 2223. Quality of final materials is influenced by the presence of free calcium and copper ions during the process. The role of both crystalline precursors and glassy matrix has been evaluated by X-ray diffraction, EDX analysis, as well as by resistivity versus temperature measurements. srCuOz is found to be the best crystalline precursor and 2234 starting stoichiometry leads to 2223 richer materials. Thermal treatment has been modified in order to optimize the 2223 phase formation. 3:20 pm BREAK 3:30pm CONTINUOUS PROCESSING OF BSCCO-22121Ag DIP-COATED CONDUCTORS: J. W. Burgoyne,D. DewHughes, Dept. of Engineering Science; C. J. Eastell, C. G. Morgan, C. R. M. Grovenor, M. J. Goringe, Dept. of Materials; R. G. Jenkins, H. Jones, Clarendon Laboratory, University of Oxford, Parks Road, Oxford, OXI 3PJ, UK A continuous process has been developed for the production of unlimited lengths of BSCCO-2212 open-surface conductor using a long multi-zone furnace. Dip-coated tape on an Ag-alloy substrate is pulled through a partial-melting/annealing temperature profile at speeds of the order I mh-I. Phase-pure, well-aligned 2212 is formed with Tc-85 K. Initial results of Ic up to 92.5 A at 4.2 K, 0 T demonstrate the viability of this method, and compare well with an Ie of 95 A for identical tape in a conventional heattreatment. Higher critical currents should be achievable through improved coating and precursor preparation. Highly homogeneous conductors are obtained sipce every part of the tape receives an identical heat-treatment, avoiding the thermal gradients experienced across wind-and-react coils. Optimisation of the temperature profile and process speed to reduce impurity phases and increase Ic will be discussed, and data on reactand-wind coils from continuously-processed conductor will be presented. 3:50pm MANUFACTURE OF TmCK FILMS AND BULK COMPONENTS OF BSCCO BY SLIP CASTING: Atit Bhargava. I. D. R. Mackinnon, A. llusbech1tin, T. Yamashita, J. A. Alarco, J. C. Barry, Centre for Microscopy and Microanalysis, The University of Queensland, Q 4072, Australia Room temperature slip-casting is a simple and economical process to fabricate bulk components and thick films ofYBCO and BSCCO. When BSCCO precursor powder transforms into HTS phase, such as during sintering, it expands causing cracking and deformation in a fabricated body. We have tailored sintering parameters to convert a slip-cast "green" body into a HTS ceramic of high density, high strength and high current carrying capacity. The process for bulk components has been extended for thick films on polycrystalline substrates. SEM, TEM, XRD, Jct and a.c. magnetic susceptibility are used to characterize the films and components. The process has applications in power, telecommunications, magnetic shielding and imaging industries. 4:10pm HOT ROLLING EFFECT ON THE CRITICAL CURRENT DENSITY IMPROVEMENT OF THE Ag-SHEATHED BI-2223 TAPE: Q. Y. Hu, D. Yu, H. K. Liu, T. Chandra, S. X. Dou, Centre for Superconducting and Electronic Materials, University of Wollongnng, Wollongong, NSW 2500, Australia Ag-sheathed Bi2223 wires were hot-rolled into thin tapes in a temperature range of 500.800°C at a four-high laboratory rolling machine with attachment of a tube furnace. It was found that hot-rolled tapes had higher core hardness compared to cold rolled tapes indicating that hot rolling has increased core mass density of the superconductor more significantly than COld-rolling. This was conftrmed by scanning electron microscopic observation which has clearly shown that the hot rolled tapes have a lower void density compared to cold rolled specimens. The enhanced core densification has resulted in the improvement of the critical current of the superconductor tapes. 4:30pm CHEMICAL COMPOSITION, MICROSTRUCTURE AND PHYSICAL PROPERTIES OF MAGNETICALLY MELT-TEXTURED BULK BezSr2Cao.sDYO.2CU20S_y SUPER-CONDUCTING CERAMICS: S. Stassen, R. Cloots, A. Vanderschueren, Ph. Vanderbemden, M. Ausloos, SUPRAS, Univeristy of Liege. Sart TIlman, B-4000 Liege. Belgium Dysprosium doped Bi-based 2212 ceramics were synthesized by a melt-texturing growth process in a 1.2 T magnetic field, applied perpendicularly to the main axis of a

B118

cylindrical shape sample. Well oriented and (four) chemically defined phases forming successive layers were observed and characterized by X-ray diffraction, scanning electron microscopy and electron dispersive X-ray analysis. The origin and composition of such layers can be described. A nO'vel aspect is the calcium solubility at the strontium site in 2212 phases. and the strontium solubility at the calcium site in 8250 stmctures. The texturing process was OptimiZI~ with respect to electrical resistivity and magnetic susceptibilty. 4:50 pm Invited SIMILAR DEPRESSION OF Tc BY ZnAND NI SUBSTITUTION AND DEPARTURE FROM THE UNIVERSAL THERMOPOWER BEHAVIOR IN CaLaBaCuO: D. Goldschmidt, Y Eckstein, Department of Physics, Technion, Haifa 32000. Israel In the new tetragonal 1-2-3 CaLaBaCuO family. electron concentration becomes constant by Ca2+ and La3+ charge-compensating substitutions and by keeping the oxygen content constant. Nevertheless. Tc and transport properties change substantially indicating that doping (hand filling) occurs. These materials have the same Tcmax and a single doping parameter y-YM-I, independent of Ca and La concentration (where YYM-I is the oxygen content y measured relative to the metal-insulator transition), suggesting that they have essentially \he same band structure. Our results are interpreted within the simple band picture modified to consider the existence oflow-mobility states near EF. besides the CuOz band. 5:10pm THE EFFECTS OF PRECURSOR POWDER ON THE TEXTURE FORMATION IN (BI,Pb)zSr2Ca2Cu30xlAg TAPES: J. Jiang. J. S. Abell. School of Metallurgy and Materials, University of Birmingham, Birmingham BI5 2TT. UK Three kinds of precursor powders with the same nominal composition were obtained by co-decomposition method. powder A with Bi-2212 as major phase. powder B with Bi-220l as major phase. and powder C with Bi-2223 and 2212 as major phase. The silver-sheathed Bi-system tapes were made by the powder-in-tube technique. The texture measurement was made by XRD and SEM. It was found that the mechanical deformation during fabricating process induced significant texture of BPSCCO grains in tbe as-pressed tapes and that this texture degree depends on the phase content and particle size of the precursor. The final tapes with 2223 phase made from the three kinds of precursor powders exhibited almost the same texture degree. The texture degree in the as-pressed tapes is not the determining factor for the texturing of the Bi-2223 grains in the final tapes with high Jc. The experiment results indicates that texture degree was improved significantly during the phase transformation. both in the process from 220 I to 2212. and from 2212 to 2223. The intermediate mechanical deformation enhanced the texture degree. The possible mechanisms of the texture formation are discussed.

HIGH TEMPERATURE SUPERCONDUCTORS· SYNTHESIS, PROCESSING, AND LARGE SCALE APPLICATIONS VII: Characterization

Sponsored by: Jt. EMPMD/sMD Superronductlng Materials Committee, MSD Electronic, Magnetic and Optical Phenomena Committee, and FEMS (Federation of European Materials Societies) Program Organizers: U. Balachandran. Energy Technology Division. Argonne National Laboratory, Argonne. IL 60439; Paul J. McGinn, University of Notre Dame. Notre Dame. IN 46556; and Stuart Abe:ll, University of Birmingham, Edgbaston, BirminghamBI52TT, UK Wednesday. PM February 7.1996

Room:B8 Location: Anaheim Convention Center

Session Chairpersons: S. Abell. University of Birmingham. Edgbaston. UK; M. Dhalle. Imperial College. London, UK

2:00p.m HYSTERESIS MOTOR AND LINEAR CARRIER SYSTEM CONSTRUCTED WITH BULK MELT TEXTURED YBCO: W. Gawa/ek, T. Habisreuther. T. Straer. M. Wu, D. Litzkendorf. P. GOmert. Institut filr Physikalische Hochtechnologie e.V., Helmholtzweg 4. 0-07743 Jena. IGermany; K. V. Ilushin. L. K. Kovalev. Moscow State Aviation Institute. Volokolarnskce shosse. 125871 Moscow. Russia Bulk cylindrical blocks of YBCO are prepared by a modified Melt Textured Growth batch process. Multi-grain samplles are characterized by integral levitation force measurements at 77 K and 18 K in the zero field cooled and field cooled state. Field mapping of the frozen flux distribution was carried out at 77 K using an axial Hall probe. In maximum the blocks trap 0.4 T (extrapolated to the surface). With blocks of our melt textured material the rotor of a hysteresis motor was constructed. The motor was tested in liquid nitrogen for some hours with 3.000 rpm and an output power of about 80 W is reached. With the material also II linear carrier system for application relevant investigations at 77 K was constructed. The carrier equiped with 4 permanent magnet rails is placed below the dewar containing the HTSC blocks before the system was field cooled. The carrier can be moved paralll:1 to the rails with very low friction. for the remaining five degrees of freedom we found high stiffness and damping. 2:20pm THE EFFECT OF BUFFER LAYERS ON SLOW WAVE PROPAGATION IN SUPERCONDUCTING PLANAR TRANSMISSION LINES: Famat Abbas, L. E. Davis, Dept. Elec. Eng. & Elec. UMIST POBox 88 Manchester MOO IQD UK; J. C. Gallop, National Physical LabOltory. Teddington Middlesex TWll OLW UK Attempts to grow YBCO films on both sapphire and Si directly have encountered some serious problems due to Ilirge mismatches of both lattice constants and thermal

1996 TMS ANNUAL MEETING AND EXHIBITION

expansion coefficients of the YBCO ftlms and sapphire, and Si. For cryogenic microwave microelectronics application it is necesS8l)' to seek proper buffer layers which, prevent the interdiifusion, better thenna) expansion coefficient matching, better lattice matching, confinement of the field into substrate and enhance the fact that a wave in a superconducting transmission line would be slowed due to the penetration of the magnetic field into the superconductor. In this study an analysis has been completed to explore the slow mode of propagation which is dependent on the material properties and thicknesses of the superconductors, substrate, and buffer layers. Relative velocity with respect to that in vacuo is computed for various combinations of material properties. The dependence on the dielectric constant of the buffer layers is discussed. 2:40pm CURRENT-VOLTAGE CHARACTERISTICS OF BSCCO CONDUCTORS: M. Dhalle, J. Everett, M. Johnston, J. V. Thomas, A. D. Caplin, Centre for High Temperature Superconductivity, Blackett Laboratory, Imperial College, London SW7 2BZ, UK Transport and magnetization measurements of current-voltage (I-V) characteristics of conductors complement each other in two important respects: The magnetic studies extend the characteristics downward by several orders of magnitude in electric field, and the topology of the important current paths is very different in the two cases. Furthermore, current paths that are associated with weak-links are suppressed in rather low magnetic fields. We report detailed transport and magnetic studies of a range ofBSCCOphase conductors, and interpret the data within a model that incorporates a spectrum of current paths, and includes both inter- and intra-granular current limitation. 3:00pm CRYOGENIC MICROCALORIMETER FOR MEASUREMENT OF MICROWAVE SURFACE IMPEDANCE OFHTS-THIN FILMS: 1.. Brunetti, M. Petrizzelli, Isituto E1ectrotecnico Nazionale "G Ferraris" Strada delle Cacce 91, 10135 Torino, Italy Since the discovery of superconductivity, the measurement of the surface impedance Z of thin ftlms represents one of the most important paces for the application of the materials at microwave frequencies. In this paper we describe a cryogenic microcalorimeter for measurements of microwave surface impedance in HTS thin ftlms. The microcalorimeter technique, that is an alternative approach to other methods, is based on the temperature rise measurement produced by microwave power dissipation in a thermally isolated thin ftlm. Basically, the HTS thin ftlm is the dissipative element of a self-resonant microwave structure, that is a tuned bolometer mount. It is placed inside an undervacuum container and supplied with microwave power through a thermal insulating waveguide section while a copper-constantan thermopile senses its temperature. This structure allows to know the real power dissipated in the ftlm and then the related quantities, like R 3:20 pm BREAK 3:30pm MICROSTRUCTURE OF 123-211 MELT-TEXTURED COMPOSITES STUDIED BY POLARIZED LIGHT MICROSCOPY: P. Dileo, Institute of Experimental Physics, SAS, Watsonova 47, 04353 Kosice, Slovakia; W. Gawalek, T. Habisreuther, P. GlImert,IPHT, Helmholtzweg 4, 100239 Jena, 0-07702, Germany; N. Pellerin, P. Odier, CCRPHT, CNRS, 4S071 Orleans Cedex 2, France; R. aoots, M. Ausloos, SUPRAS, University of Li~ge, B-4OOO Li~ge, Belgium The microstructure features of melt-textured YBaCuO (123) such as observed phases, 123 domain and subgrain boundaries, macro and microcracking and twin structure have been characterized. The linear dependence between a-b microcrack spacing, Ie, and d211, VI23N2ll(d2ll-mean 211 particle size, V123-123 volume fraction, V211-211 volume fraction) was observed. Shorter mean free distance between 211 particles depresses the subgrain thickness, the twin complex size and twin spacing. The level of residual tensile stress in the 123 a-b plane (50-100 MPa) was estimated from detwinning observed around 211 particles. A maximum residual tensile stress of 290 MPa in the "c" direction was estimated on the basis of the 123 fracture toughness. Oxygenation of the melt-textured samples was shown to he a combination of oxygen volume diffusion, microcracking and penetration of oxygen through the cracks. A new secondsi)' phase having a composition close to Ba2Cu5Ox has been found. The essential influence of gas bubbles, and unifonnity of 211 particles in the melt, on the microstructure of 123 domains has been shown. 3:50pm XPS AND A.C. SUSCEPTIBILITY SruDY OF SURFACE DEGRADATION OF YBa2Cu307_x POWDERS: I. Sargamkova, P. Diko, Institute of Experimental Physics, Slovak Academy of Science, Watsonova 47, 043 53 Kosice, Slovakia; J. D. 1\veed,C. A. Anderson, N. M. D. Brown, Joint Ceramics Research Centre, Surface Science Laboratory, University of Ulster, Cromore Road, Coleraine, Co. LondondCfl)', Northern Ireland, BT52 ISA The preparation of powder fractions ofYBa2Cu3O,_x by sedimentation in 2-butanol raises questions of particle surface degradation of particles during the sedimentation and the possibility of regeneration by re-annealing. In this work we studied this behaviour by XPS and a.c. susceptibility measurements. Characteristic peaks for monitoring changes due to degradation and regeneration were found in the Y3d (c. 160 e V) and as (c. 282 eV) regions. Our results suggest that annealing close to the sintering temperature can regenerate the particle surfaces. 4:10pm MAGNETIC BEHAVIOR OFYBa2Cu3O,_x (123)-Y2BaCu05 (211) SINTERED COMPOSITES PREPARED BY SOL-GEL: P. Dileo, I. Sargankovli, Institute of Experimental Physics, Slovak Academy of Sciences, Watsonova 47, 04353 Kosice, Slovakia; M. Reissner, W. Steiner, Institut fur Angewandte und Technische Physik, Technische Universitat Wien Wiedner Hauptstrasse 8-101137, a-104O Wien, Austria The grain growth at 980°C in 123-211 composites prepared by sol-gel technic is controlled by pinning of 123 grain boundaries by 211 particles. It produces a narrow 123 grain size distribution. The secondsi)' grain growth in pure 123 leads to the struc-

1996 TMS ANNUAL MEETING AND EXHIBITION

ture composed by large and smail grains (bimodal grain size distribution). The quality of contacts between 123 grains is related to the deviation from 123-211 line and to the grain size unifonnity. 4:30pm IN SITU PREPARATION OF EXPITAXIAL T1Ba2Ca2Cu30, FILMS BY LASER ABLATION IN THE PRESENCE OF Tl20 VAPOR: N. Reschauer, U. Spreitzer, W. Brozio, A. Piehler, K. F. Renk, Institut fUr Angewandte Physik, Universillit Regensburg, 93040 Regensburg, Germany We report an in situ growth method of epitaxial T1Ba2Ca2Cu3D9 (11-1223) high Tc films. By laser ablation from a BaCaCuO target we deposited Ba, Ca, and Cu simultaneously with 11 in an oxygen atmosphere. The 11 was delivered as 1120 vapor by thenna) evaporation of 11203 powder. lYpical film thicknesses were 80-400 nm. The films had transition temperatures near 100 K. X-ray diffraction patterns of 9 - 29 scans indicated that the films were single phase and highly oriented with the c axis perpendicular to the substrate surfaces. Epitaxial growth of the films was proved by 9 scan measurements. Scanning electron microscope investigations showed smooth surfaces without platelike structures.

INTERNATIONAL SYMPOSIUM ON REACTIVE METALS: Processing & Applications IV: Radioactive Metals Processing

Sponsored by: LMD Reactive Metals COmmittee Co-sponsored by: Canadian Institute of Mining, Metallurgy & Petroleum, Montreal, Canada; The Japan Institute of Metals, Sendai, Japan; Mining & Materials Processing Institute of Japan, Tokyo, Japan; Society for Mining, Metallurgy & Exploration, Littleton, CO Program Organizers: B. Mishra, Department of Metallurgical & Materials Engineering, Colorado School of Mines, Golden, CO 80401; GJ. Kipouros, Department of Mining & Metallurgical Engineering, Technical University of Nova Scotia, Halifax, Nova Scotia, Canada B3J 2X4; R.G. Reddy, Department of Chemical & Metallurgical Engineering, University of Nevada, Reno, NV 89557 Co-organizers: W.A. Averill, Rocky Aats, Inc., Golden. R.G. Bautista, University of Nevada - Reno, Reno; M.C. Bronson, Lawrence Livermore Natl. Lab., Livermore; J.A. Sommers, Teledyne Wah Chang Albany, Albany; C.B. Wilson, The Dow Chemical Company, Freeport Wednesday, PM Febru3l)' 7, 1996

Room:B9 Location: Anaheim Convention Center

Session Chairperson: Dr. B.R. Westphal, Argonne National Laboratory, PO Box 2528, Idaho Falls, ID 83403; Dr. D.L. Olson, Department of Metallurgical & Materials Engineering, Colorado School of Mines, Golden, CO 80401

2:00pm TREATMENT OF SPENT NUCLEAR FUEL FOR THE SEPARATION OF REACTIVE METALS: B.R. Westphal, K.M. Goff and RW. Benedict, Argonne National LaboratOl)', PO Box 2528, Idaho Falls, ID 83403 One of the key features to the treatment of Experimental Breeder Reactor-II (EBRII) spent fuel at Argonne NationallaboratOl)' is the separation of reactive metals (Na, Cs, Sr, and rare earths) from an actinide product. Metallic fuel, burdened with a multitude of fission products, is bonded with the sodium for ease of heat transfer from the fuel to the reactor coolant. The removal of sodium and fission products from the actinides by electrometallurgical means enables the recovery of a product that can be isotopically diluted and stored. The primary separation of the reactive metals occurs during an electrorefining operation although additional recoveries of the actinide product are realized by subsequent concentration steps. As it currently stands, the EBR-I1 spent fuel treatment process has definite potential for application to other fuels in the DOE inventory. 2:25pm DIFFERENTIAL SCANNING CALORIMETRY STUDY OF SOLID STATE PHASE TRANSFORMATIONS IN PLUTONIUM: P.C. Lopez, J.R Cost, K.M. Axler, Nuclear Materials Technology Division, Los Alamos National Laboratory, Los Alamos, NM 87545 Allotropic transformations between the six solid phases of plutonium have been studied both isothermally and during temperature scans. Initially, the j!-ty and 'l'~j! transformations have been investigated, mostly at a rate of 1.0'C per minute. At this rate the transformation on heating occurs over the temperature range from 215'C to 225'C. The 'l'~j! reaction upon cooling transforms over the range 145'C to 125'C indicating typical undercooling hysteresis. Studies of the higher temperature transformations, 6HI>' and 6He' will also be reported. 2:50pm EVALUATION OF ERBIA AS MOLTEN PLUTONIUM CONTAINMENT MATERIAL: C. Lensing, D.L. Olson, B. Misbra, J. Selle, Department of Metallurgical & Materials Engineering, Colorado School of Mines, Golden, Colorado 80401 The compatibility between erbium oxide and molten cerium, as a surrogate for plutonium, was investigated to understand the high temperature corrrosion mechanisms and to provide kinetic data. The corrosion kinetic data was compared with the data obtained from compatibility tests of erbium oxide in molten plutonium. High density erbium oxide was immersed into molten cerium at temperatures ranging from 850 to 975'C for 16 to 128 hours. For both cerium and plutonium systems, a parabOlic rate dependence was observed as well as intergranular penetration of cerium into erbium oxide was noted. Activation energies calculated for cerium and plutonium systems are 38.5 and

B119

24.2 kCalImole, respectively. Two reaction layers formed for the cerium system, viz. cerium oxide particle layer and a layer of erbium oxide-cerium oxide solid solution. An additional third layer, possibly a ternary, was observed for the erbia immersion in molten plutonium, TIle work was intended to develop material systems for the containment and processing of radioactive materials. 3:15 pm BREAK 3:30pm PYROCHEMICAL PROCESSES FOR THE RECOVERY OF WEAPONS GRADE PLUTONIUM EITHER AS A METAL OR AS PuOl FOR USE IN MIXED OXIDE REACTOR FULE PELLETS: C.A. Colmenares, B.E. Ebbinghaus, M.C. Bronson, University of California, Lawrence Livermore National Laboratory, PO Box 808, Livermore, CA 94550 We have developed two processes for the recovery of weapons grade plutonium, as either Pu metal or Pu~, that are strictly pyrochemical and do not produce any liquid waste. Large amounts of Pu metal ( up to 4 kg), in various geometric shapes, have been recovered by a hydrideldehydride/cast process (HYDEC) to produce metal ingots of any desired shape. TIle three processing steps are carried out in a single compact apparatus. TIle experimental technique and resulsts obtained will be described. We have prepared Pu~ powders from weapons grade Pu by a process that hydrides the Pu metal followed by the oxidation of the hydride (HYDOX Process). Experimental details of the best way to carry out this process will be presented, as well as the characterization of both hydride and oxide powders produced. This work was performed under the auspices of the US Department of Energy by the Lawrence Livermore National Laboratory under contract number W-7405-ENG-48. 3:55pm TREATMENT OF WASTE SODIUMlNaK; DECOMMISSIONING HARDWARE AND FACILITIES: A.E. Stewart, Pyropacific Processes, 10447 Amwood Road, Sylmar, CA 91342-6907; J.D. Navratil, RUST Federal Sevices, 1597 Cole Boulevard, Bldg 15- Suite 350, Golden, CO 80401-3314 TIle decline in government interest (both USA and abroad) in further developing liquid metal cooled reactors has left a legacy of unneeded and unused government owned sodium and Nak component test facilities, including nuclear reactors. TIlese facilities will have to have their inventories of sodium drained and the equipment and piping treated to passivate the residual reactive metal. TIle bulk drained sodium (estimated at 1.4 million kg that will be radioactive, and 0.5 million kg that will be nonradioactive) will also have to be treated to convert the chemically active metallic &0dium into a nonhazardous chemical form or possibly reused (non-radioactive sodium). A review of past experience with various techniques for safely handling and disposing of reactive metals and facilities and equipment containing these materials is to be presented with emphasis on both the technical basis and practical requirements involved. 4:10pm SEPARATION OF ACTINIDES FROM LANTHANIDES UTILIZING MOLTEN SALT ELECTROREFINING: D.L. Grimmett, S.P. Fussselman, JJ. Roy, R.L. Gay, Rockwell IntemationallRocketdyne Division, 6633 Canoga Avenue, Canoga Park, CA 91309-7922; C.L. Krueger, T.S. Storvick, University of Missouri-Columbi and Missouri University Research Reactor, Columbia, MO 65211; T. Inoue, T. Hijikata, Central Research Institute of Electric Power Industry, Komae Research Laboratory, Tokyo, Japan; N. Takahashi, Kawasaki Heavy Industries, Ltd., Nuclear Systems Division, TokYo,Japan TRUMP-S (TRansUranic Management through Pyropartitioning Separation) is a pyrochemical process being developed to separate actinides from fission products in nuclear waste. A key process step involving molten salt electrorefining to separate actinides from lanthanides has been studied on a laboratory scale. E1ectrorefining of U, Np, Pu, Am, lanthanide mixtures from molten cadmium at 450·C to a solid cathode utilizing a molten chloride electrolyte resulted in >99% removal of actinides from the molten cadmium and salt phases. Removal of the last few percent of actinides is accompanied by lowered cathodic current efficiency and some lanthanide codeposition. Actinide/lanthanide separation ratios on the cathode are ordered U>Np>Pu>Am and are consistent with predictions based on eqUilibrium potentials.

LEAD ZINC TIN PRODUCTION

Sponso"d by: EPD Lead, ZInc, TIn Committee Program Organizer: Eogin Ozberk, Sherrill Inc., Fort Saskatchewan, Alberta, Canada Wednesday, PM February 7, 1996

Room:AI4-15 Location: Anaheim Convention Center

Session Chair: Eogin Ozberk, Sherrill Inc., Fort Saskatchewan, Alberta, Canada

1:00pm PYROMETALLURGICAL TREATMENT OF ZINC RETORT RESIDUE BY SUBMERGED COMBUSTION IN-BATH SMELTING: Dhiren K. Panda, Kent D. Peaslee, David G.C. Robertson, Generic Mineral Technology Center for Pyrometallurgy. 215 Fulton Hall. University of Missouri-Rolla. Rolla, MO 65401 TIle zinc retort residues from old pyrometallurgical zinc production sites (in the Oklahoma-Kansas-Missouri tri-state area) pose an environmental problem, since they contain a significant amount of lead and zinc, among other contaminants. TIlese residues were treated at the University of Missour-Rolla by the submerged combustion in-bath smelting (SClBS) technique to fume off lead and zinc and obtain a non-hazardous slag. Details of the thermodynamic and kinetic modelling of the process are presented. Results of laboratory scale batch type experiments for studying the different operating parameters such as coke size, feed rate, fuming rate, (gas) blowing rate, etc. are high-

B120

lighted. 1:15pm ECO-TECHNO-ECONOMIC SYNTIIESIS OF PROCESS ROUTES FOR THE PRODUCTION OF ZINC USING SIMULATED ANNEALING: MarkusA. ReUler, Sabina C. Sudholter, Mintek, Measurement & Control Division, Private Bag X3015, 2125 Randburg, South Africa TIle demands placed on the environmental and social acceptability of metallurgical processing technology is rising steadily. Into the spotlight come the production techniques, products and disposal of residues. TIlese aspects are affected by varying compositions of primary and secondary raw materials processed in these plants and the rapidly changing market situations in the metallurgical industry. The metallurgical engineers have to select "optimal" processes from a vast number of existing technologies for the primary production of zinc and for the processing of zince containing residues. To enable the engineer to compllre these techniques and to choose the right combination of unit operations for this purpose on the most general basis, a process design methodology is presented, whicb has been adapted from methodologies developed in chemical engineering and minerals processing. 1:50pm THE USE OF OXYGEN IN LEAD SMELTING AND REFINING OPERATIONS: R. David Prengaman, Senior Vke President, Research & Development, RSR Corporation, 2777 Stemmons Freeway, Suite 1800, Dallas, TX 75207 In modem continuous lead smelting processes such as the QSL, Kivcel, Isasmelt and Ausmelt tonnage oxygen is used to oxidize sulfidic concentrates and residues to provide the major part of the energy for the process. TIle sulfur is evolved as S~, and recovered as H2S04 from a rich S02 gas. In secondary lead smelting processes, using rotary or reverberatory fumancl~s oxygen is used to accelerate smelting rates and reduce dust carryover by reducing gas flow. in these processes, however, care must be taken to restrict oxygen access to the charge to prevent evolution of S~. Oxygen is also used to accelerate lead refining in both primary and secondary operations. This paper outlines the benefits of using oxygen in lead smelting and refining processes. 3:15pm LEAD RECYCLING PROCESS FOR SELECTIVE RECOVERY OF PRECIOUS METALS FROM ARGENTOPYRITE CONCENTRATE: J. R. Parga, G.G. Muzquiz, Technologico de Saltillo, A.P. 84C, 25000 Saltillo Coah Mexico A shortage of landfill space and growing environmental concerns in Mexico are renewing interests in the recyclability of scrap from automobiles. This work describes a pyrometallurgical process for lhe recovery of metallic lead from waste lead paste. Argentopyrite concentrate is commingled with the lead paste residue and smelted in a fuming process therby transferring the precious metals into lead bullion. TIle process consists of smelting lead past and argentopyrite, sodium carbonate, and metallic iron at smelting temperatures of about \IOO·C. TIle reaction products are slags and molten lead bullion enriched with precious metals. A major feature of this process is the removal of precious metals from the bullion by means of zinc, whereby a silver-zinc intermetallic compound is formed and floats out of the lead. TIle recoveries of precious metals are about 98%.

MATERIALS SCIENCE OF CONTACTS AND METALLIZATION AND INTERCONNECTS V

Sponsored by:EMPMD ThIn nm & Interfaces Committee Program Organizers: Jagdish Narayan, Dept of Matis Sci & Engrg, Box 7916, NC State University, Raleigh, NC 27695; John Sanchez, Advanced Micro Devices, M3 160, PO Box 3453, Sunnyvale, CA 94088 Wednesday, PM February 7,1996

Room: Orange County 5 Location: Anaheim Marrioll Hotel

Session Chair: K.N. Tu, Department of Materials Science and Engineering, University of Califumia, Los Angeles, CA 90095

1:00 pm Invited SCHO'ITKY BARRIER HEIGHTS TO IV-IV SEMICONDUCTORS: F. Meyer, M. Mamor, V. Aubry-Fortuna, IEF CNRS URA 22, Bal. 220, Universite Paris Sud, 91405 Orsay Cedex, France; S. Bodnar, J.L. Regolini, France Telecom CNET BP98, 38243 Meylan Cedex, France TIle variations of Schottky barrier heights to Sil_x_)'GexCy films with composition and strain have been investigated and compared to those expected for the band gap energy. TIle results suggest that the Fermi level at the interface between tungsten and Sil_x_yGexCy alloys (xO) is pinned relative to the conduction-band. For Sil_ xGex,the barrier on p-type follows the same trends than the band gap. For the ternary alloys, the variations of the barrier on p-type seems to be too large to be only due to a variation of the band-gap. In addition, we have investigated the role of the strain retained in the sputtered-W-gate on the barrier to silicon. Our results show that the barrier on n-type follows the same trends than the stress retained in the W-films as a function of the deposition pressure. The absence of change in the barrier height to p-type silicon suggests that the Fermi level at 11K: interface with Si is pinned relative to the valence-band. 1:30 pm InvIted OHMIC CONTACTS IN SI-Ge ALLOYS: R. J. Nemanich, D. Aldrich, Z. Wang, D. Sayers, Department of Physics, North Carolina State University, Raleigh, NC 27695 This study addresses the interface stability and electrical properties of contacts on SiGe alloys. Reactive contact formation has been explored for a series of different

1996 TMS ANNUAL MEETING AND EXHIBITION

metals (n, Co, and Zr) on MBE prepared SiGe epitaxial films grown on Si substrates. The results indicate that the reactions are significantly more complex than solid phase

reactions of the same metals on Si (or Ge). 1\vo different interface instabilities are identified in the n-SiGe and Co-SiGe materials systems, and these instabilities are modeled in terms of the thermodynamic properties and surface and interface energies of the materials. The intrinsic electrical property of the interface is the Schottky barrier, and we have measured the alloy dependence of the Schottky barrier of Co on both strained and unstrained SiGe alloys using angle resolved UV photoemission spectroscopy. The results indicate that the p-type Schottky barrier is the same for both strained and unstrained materials. Since the bandgap of the materials varies significantly with strain it is concluded that the n-type Schottky barrier shows significant strain dependence. 3:00pm CONTACT FORMATION OF Cu3Ge ON Ge WAFERS: l.S. Huang, K.N. Tu, Department of Materials Science and Engineering, University of California, Los Angeles, CA 90095-1595; F. Deng, S. S. Lau, Department of Electrical and Computer Engineering, University of California at San Diego, La Jolla, CA 92093-0407 We have carried out a study of formation of CU3Ge by reacting Cu tiIms with Ge wafers. Phase identification and reaction kinetics were performed by a combination of Read camera X-ray diffraction, Rutherford backscattering spectroscopy, and in-situ resistivity measurement. In the as-deposited samples, a trace of Cu3Ge phase was detected. The CU3Ge grew rapidly at temperatures above 200°C. We have found that the CU3Ge phase is very stable again oxidation in air, in contrast to the fast oxidation behavior of Cu and CU3Si films. The oxidation behavior will be discussed. 3:30pm PLENARY SESSION

Leaders: J.M. Woodall, Purdue University; J. Narayan, North Carolina State University; J.E. Sanchez, Jr., Advanced Micro Devices Panel Members: D. Crawford, National Science Foundation; H. Morkoc, University of Illinois at Urbana-Champaign; P.H. Holloway, University of Florida; M. Murakami, Kyoto University; F. Meyer, University of Paris; K.N. TIl, University of California at Los Angeles; T. Sands, University of California at Berkeley; C.M. Osburn, North Car0lina State University

MODELING OF COMPOSITES: PROPERTIES AND PROCESSING SESSION II:

Sponsored by: MSD Computer Simulation Committee Program Organizers: S.P. Chen, Los Alamos National Lab., Los Alamos, NM 87545; M.P. Anderson, Exxon Research Center, Dinton Township, Route 22 East, Annandale, NJ08801 Wednesday, PM February 7, 1996

Room: Marquis 1&2 Location: Anaheim Marriott Hotel

Session Chairman: S. P. Chen, Los Alamos National Laboratory, Los Alamos, NM 87545; M. P. Anderson, Exxon Research Center, Clinton Township, Route 22 East, Annandale, NJ 0880 I

2:00pm ON THE CRITICAL CONDITIONS OF KINK BAND FORMATION UNDER COMPRESSIVE LOADING IN FIBER COMPOSITES: Ming Dao, RobertJ. Asaro, Department of Applied Mechanics and Engineering Sciences, University of California, San Diego, La Jolla, CA 92093-0411 Compressive loading along fiber direction in fiber composites usually results in kink band formation, which dramatically reduce the compressive strength of the fiber composites. A rate independent elasto-plastic, finite strain model is applied to study the critical conditions of kink band formation of fiber composites under compressive loading. Orientation dependent mechanical behavior of fiber composites is modeled using a simple slip model. Unlike many previous models that give separate analysis for the maximum load and the kink band angle, our model gives a unified explanation for both the critical load and the kink band angle. Elastic anisotropy id found to playa very important role in terms of the kink band angle. The predicted values of the critical load and the kink band angle are in very good agreement with available experimental data. 2:30pm TEMPERATURE·DEPENDENT ELASTIC MODULI OF TWO PHASE COM· POSITE, CALCULATED USING SIMPLE UNIT CELL MODELS: K. S. Ravichandran, Department of Metallurgical Engineering, 412 WBB, The University of Utah, Salt Lake City, UT 84112 Temperature dependence of elastic moduli are of interest to evaluate structural performance and residual stresses of composites used in structural and electronic applications. In the present research, prediction of elastic moduli of composites as a function of temperature will be made using simple, microstructure-based unit cell models. The models are based on the concept of dividing the unit cell into isostress and isostrain elements and determining the overall composite behavior by using appropriate volume fractions for these elements at different stages of division. Composites having discrete second phases in continuous matrix materials is considered for this purpose. Comparisons of predicted trends with experimental data on several composites is made. The accuracy of the present method is evaluated in the light of other models. 3:00pm LOAD TRANSFER DURING LOADING IN PARTICLE-REINFORCED

1996 TMS ANNUAL MEETING AND EXHIBITION

METAL MATRIX COMPOSITES: N. Shi, M. A. Bourlee, J. A. Goldstone, Los Alamos National Lab., Los Alamos, NM 87545, J. E. Allison, Ford Motor Company, Dearborn, M148121 Internal stresses in the matrix and reinforcement of a 2219AVTIC/15p-T6 metal matrix composite were measured in situ via neutron diffraction during loading. The evolution of stresses in individual phases (phase stresses) with the applied load was also modeled using a finite element method model. The numerical predictions are in good agreement with the neutron diffraction stress measurements. Both modeling and measurements indicate that while internal normal stress parallel to the loading direction continues to transfer from the matrix to the reinforcement as applied load increases, the internal particle normal stress perpendicular to loading decreases at small loads, and reaches a minimum at a load well below the composite macroscopic yield. As the load further increases, the perpendicular particle normal stress relaxes until marcroscopic yield is reached. This kind of behavior can be numerically reproduced only when the effects of thermal residual stresses are included. It was demonstrated that thermal residual stresses result in alteration of matrix plastic flow initiation site, which is critical in affecting load transfer before macroscopic yield. 3:30pm MODELING THE CONSOLIDATION BEHAVIOR OF METAL MATRIX COATED FIBERS: Joseph M. Kunze, Haydn N. G. Wadley, Department of Materials Science and Engineering, School of Engineering and Applied Science University of Virginia, Charlottesville, VA 22903 An experimental study has been conducted on the consolidation of randomly packed arrays of silicon carbide fibers coated via physical vapor deposition (PVD) with a n6AI-4V matrix. The densification behavior of the metal matrix coated fibers was recorded in situ using an eddy current sensor. A model has subsquently been developed describing the densification of continuously reinforced metal matrix composites from fibers coated with the matrix material. The densification mechanisms considered include plasticity, power law creep and diffusion accommodated grain sliding. The model utilizes previous work on the blunting of contacts (stage I) and the closure of isolated voids (stage II) in determining the overall densification behavior. The results showed diffusion accomodates grain sliding to be the dominant densification mechanism. Modelbased simulations using the process schedules of the IDP consolidation experiments were found to agree well with experimental results. 4:00pm THE ROLE OF INTERFACES AND MATRIX VOID NUCLEATION MECHA· NISM ON THE DUCTILE FRACTURE PROCESS OF DISCONTINUOUS RE· INFORCED COMPOSITES: S. B. Biner, Ames Laboratory,Iowa State University, Ames,IA5001l The role of fiber morphology, interface failure and void nucleation mechanisms within the matrix on the deformation and fracture behavior of discontinuous reinforced composites was numerically investigated. The matrix was modeled usin8 a constitutive relationship that accounts for strength degradation resulting from nucleation and growth of voids. For the matrix, two materials exhibiting identical strength and ductility but having different void-nucleation mechanisms (stress-controlled and strain controlled) were considered and reinforcements were assumed to be elastic. The results indicate that in the absence of interface failure, the void nucleation in the matrix is the key controlling parameter of the composite strength and ductility, hence, of the fracture toughness. The weak interfacial behavior between the fibers and the matrix can significantly increase the ductility without sacrificing strength for certain reinforcement morphology and certain matrix void-nucleation mechanisms. This work was supported by USOOE, Office of Basic Energy Sciences, Div. of Materials Science under contract no. W-7405-ENG-82. 4:30pm THE MECHANICAL PROPERTIES OF INTERMETALLICJMETALLIC MICROLAMINATE COMPOSITES: J. Heathcote, G. R. Odette, G. E. Lucas, Materials Department, University of California, Santa Barbara, CA 93106 A self-consistent procedure is used to determine the stress-displacement function, sigma(u), of constrained metal layers in an intermetallic/metallic microlaminate composite. A fitting procedure is used with a large scale bridging model to determine the sigma(u) function that corresponds to measured resistance curve (R-curve) data. In this evaluation u·, the peak stress, sigma(p), and failure extension, u·, of the metal layers are fixed to allowable ranges by independent measurements. Microhardness tests are used to estimate sigma(p), and a fracture reconstruction procedure is used to determine u·. Finally, as a test of self-consistency, the sigma(u) functions are used to determine R-curves for tensile test samples for a bridging-crack stability analysis of fracture strength controlled by pre-exisiting defects. Strength predictions from this analysis are compared to measured data. The evaluation of sigma( u), a fundamental composite property, and of how it controls mechanical behavior gives insight into how to develop these composites to optimize this behavior.

PHYSICAL CHEMISTRY OF HIGH TEMPERATURE COMPOSITES AND INTERMETALLICS VI: Synthesis And Processing Sponsored by: EPD Process Fundamentals Committee, MSD Thermodynamics & Phase Equilibria Committee, Japan InstItute or Metals Program Organizers: Prof. R.Y. Un, University of Cincinnati; Prof. Y. Austin Chang, University of Wisconsin-Madison; Prof. R. Reddy, University of Navada-Reno and Dr. C.T. Uu, Oak Ridge NL Wednesday, PM February 7, 1996

Room: B2 Location: Anaheim Convention Center

8121

Session Chairpersons: P.R. Taylor, Dept. of Met. & Min. Eng., College of Mines, Univ. of Idaho, Moscow,lH 83843; S.G. Kumar, Dept. of Chern. & Met. Eng., Univ. of Nevada, Reno, NV 89557

2:00 pm Invited PROCESSING OF A NbJAI-ll TERNARY ALLOY FROM ELEMENTAL AND PREALLOYED POWDERS: S. G. Kumar. R. G. Reddy, Department of Chemical and Metallurgical Engineering, University of Nevada, Reno, NV 89557 Niobium aluminides offer the advantage of higher high temperature capability compared to other intermetallic aluminides due to the refratory nature of niobium. However the reaction sequence in Nb-AI system is slugish and currently there is a great interst in the processing aspects of these materials. In the present study, a systematic investigation was undertaken to compare the effect of different powder metallurgy processing routes on the phase relations and microstructure of 66Nb-22AI-12TI (in atom %) alloy. 1be three different processing routes adapted: (I) cold pressing followed by reaction sintering; (ii) cold pressing and hot pressing followed by sintering; and (iii) arc melting, bydride-dehydride process to make the alloy powder followed by cold pressing and then sintering. 1be samples were analyzed by X-ray diffraction, optical microscopy, scanning electron microscope with energy dispersive spectrometer and electron microprobe analyzer. 1be analysis showed that alloyed pellets made from prealloyed powders i.e, arc melting followed by hydride-debydride process, gave good densification. 2:25pm SYNTHESIS OF ZIRCONIUM CARBIDE FROM ZIRCON CONCENTRATES IN A THERMAL PLASMA REACTOR: E. E. Vidal I , M. Manrique, Department of Materials Science, Simon Bolivar University Caracas, Venezuela; Patrick R. Taylor, Department of Metallurgical and Mining Engineering, College of Mines, University of Idaho, Moscow, Idaho 83843 Ultra fine powder of zirconium carbide has been synthesised from zircon concentrate and methane using a thermoplasma reactor. The effect of several experimental variables sucb as powder used and zircon particle size has bee studied. Acomplete thermodynamic analysis of possible reactions and stable compounds was performed. A mathematical that describes in-flight decomposition of zircon particles was developed based on a description of the temperature and velocity profiles calculated in the reactor. Based on the results obtained, it is thermodynamically possible to form zirconium carbide from zircon and methane. Complete vaporization of zircon in the reactor is predicted when the particle size is below 45m and the effective power entering the reactor is 15 kW. The powders conected was cbaracterized by XRD, SEM, TEM and chemical analysis using ICP. 1be results indicated the presence of zirconium carbide, zirconium oxide, silicon carbide and grapbite. 2:50pm THE PHYSICAL CHEMISTRY OF NUCLEATION OF SUB-MICROMETER NON-OXIDE CERAMIC POWDERS VIA SUB-OXIDE VAPOR PHASE REDUCTION REACTION: Animesh Jha, Department of Material Technology, Brunei University, Kingston Lane, Uxbridge UB8 3PH, England Fine ceramic powders « 500 nm) exhibit exceptional physical and mecbanical properties in engineered structural ceramics. The production of fine powders, in particular the non-oxide ceramics, via a cbeaper route than the organic solvent route bas been rather elusive. This paper exarninines the physical chemistry of sub-oxide vapour-phase reduction reaction for the nucleation of non-oxide ceramic phase. Well known vapour species eg SiO and BO in the production of technical ceramic powders (SiC, BN) are particularly discussed for understanding the nucleation process of SiC and BN ceramic phases respectively. 1be regimes of partial pressures and temperatures are parlicularly identified and, the calculaled nucleation rate as a function of the temperature is compared with the experimental results on powder morphology. The produclion of amorpbous and nanocrystalline h-BN powders is discussed in the context of substrate structure and thermodynamic parameters. 3:15pm KINETIC STUDIES OF REDUCTION OF SOME TRANSITION METAL TUNGSTATES: J. A. Bustness, Do Siehen, S. Seetharaman, Royale Institute ofTecbnology, Division Tbeoritical Metallurgy, S-IOO 44 Stockholm, Sweden The present study deals witb the kinetics studies of reduction of some selected transistion metal tungstates with bydrogen. The measurement were carried by the isothermal thermogravimetric method. Shallow powder beds were employed in order to avoid the effects due to transport phenomena affecting the reaction rate. In this way, the cbemical reaction rates could be selectively stidied. The experiments were carried out in the temperature range 873-1173K. From the experimental data the activation energies were evaluated and were found to be around 88kJ/moi in the case of both Fe as well as Co-tungstates indicating the similarities in WoO bonds in both cases. The activation energies obtained are discussed in the light of the position of the transient metal in the periodic table as well as the thermodynamic stabilities of the various tungstates. 3:40 pm BREAK 4:00pm EVALUATION OF THERMODYNAMICS OF SOLUTES OXYGEN AND SULPHUR IN TERNARY LIQUID ALLOYS USING INTERACTION PARAMTER FORMALISM: Singareddy R. Reddy, Ramana G. Reddy, Department of Chemical and Metallurgical Engineering, University of Nevada, Reno, NV 89557 A thermodynamic formalism for interpreting excess Gibbs energy of a ternary system is used in the present study to interpret activity coefficient of a dilute solute, 2, in a binary solvent 1-3. The formalism is based on Gibbs energy interaction parameters is deduced from the integral function of a ternary system. The derived expression is used to interpret logarithmic activity coefficient of oxygen in seven metallic systems an of sulfur in three metallic systems. The present model calculation of logarithmiC activity

B122

coefficient of solute are compared with experimental data. Self interaction parameter of the solute is derived from the partial function of solute using constant compositional path. The predicted thermodynamic values are in excellent agreement with the experimental data and also our predictllOns are far more accurate than the predictions made by other models. The derived equation is used successfully to deduce thermodynamics of self interaction of oxygen as a interstitial atom in two metallic systems. 4:25pm MECHANISMS OFREDUCll0N - NITRIDATION OF SYNTHETIC KAOLINITE IN THE PRESENCE OF CARBON: Catherine Davey, Animesh Jha, Department of Material Technology, Bmnel University, Kingston Lane, Uxbridge, Middlesex UB8 3PH, England Carbothernic reduction and nitridation reactions using synthetic kaolinite mineral were investigated for the formation of silicon based nitrogen ceramics. The temperature dependence of the reaction rate and its effect on phases formed are reported. The effect of the reactivity of carbon on the kinetics of formation of AIN and Si3N4, is also discussed. Using the standard Gibbs free energy values for the formation of nitrides and oxides of aluminium and silicon, a log P02 versus Iff diagram is constructed. The experimental results are discussed and compared with the predicted phase equilibria. We also report the influence of 8ub-oxide gaseous species on the microstructure of the ceramic pbases produced. 4:50pm AN EVOLUll0N METHOD ON DEGREE OF GRAPmTIZATION OF CAR· BON MATERIALS UNDER mGH TEMPERATURE AND mGH PRESSURE: Yuan-Haun Lee,Jonq-Ren Tarng, Ke-MingLin, Gin-Haw Chen, Wen-Ku Chang, Institute of Materials Science and Engineering, National Taiwan University, I Roosevelt Rd. Sec 4, Taipei, Taiwan The majority of early studies used X-ray diffraction, but this requires a model to interpret the diffraction proces&es. A turbostratic model bas been applied by Warren, Franklin and Bacon to a wide range of materials including both soft and bard carbons. Maire and Mering applied the same metbods to the interpretation of the (001) lines of a series of graphitic carbons. In our case, graphitic carbons undergone high temperature and high pressure graphitization processes have been evaluated by the same methods. It were found that most technological XRD data were obtained by a Rietveld method with the least-squares refinmenbl could be in the simplest way since the full correction technique is very laborious. The behaviour of materials as it procedes througb tbe graphatization process is characterized by changes in the measured interlayer spacings (d) as a function of temperature. A series of graphitic carbons including the transformation of diamond to graphite through the process under high temperature and bigh temperature have been evaluated. A simple calibration equation has also been established. 5:15pm FLUX ASSISTED DISPERSION OF CERAMIC PHASES IN MOLTEN ALU· MINUM ALLOYS: Chris Dometakis, A. Jha, Dept of Matis Tech., Brunei Univ., Uxbridge, Middlesex, UBS 3PH England Without the use of a flux, only a poor dispersion of ceramic phase can be acbieved. A molten-flux-assisted ceramic pbase dispersion technique for producing aluminum alloy metal-matrix composites bas been developed. In this paper, we report the physical chemistry for the dispersion of TiB2 in molten aluminum. In particular, aspects of the interfacial energy and the themodynamics of the dispersion reactions are briefly discussed for enhancing the volume: of dispersion phase in the molten metal and for controlling their morphology. This investigation also addresses the role of molten flux in the fabrication of AI-alloy mmc using a casting technique. The morphological aspects of the dispersion are explained on the basis of the classical theory of nucleation.

RADIATION MATERIALS SCIENCE IN TECHNOLOGY APPLICATIONS IV: Radiation Effects In Alloys

Sponsored by: Jt. SMDIMSD Nuclear Materials Committee Program Organizers: L. K. Mansur, Metals and Ceramics Division, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831-6376; C. L. Snead, Jr., Applied Technologies Division, Brookha,ven National Laboratory, PO Box 5000, Upton, NY \1973-5000 Wednesday, PM February 7,1996

Room: Grand H Location: Anaheim Marriott Hotel

Session Chairperson: K. G. Lynn, Physics Department and Department of Applied Science, Brookhaven National laboratory, PO Box 5000, Upton, NY \1973-5000

2:00 pm Invited MECHANISMS OF RADIATION-INDUCED DEGRADATION OF VESSEL MATERIALS: L K. Mansur. K. Farrell, Metals and Ceramics Division, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831-6376 In reactor components, fast neutrons are typically considered the source of various radiation effects caused by atom displacements such as ernbrittlement, swelling and irradiation creep. However, in principle, other reactions can contribute to atom displacements, such as those causedl by thermal neutron capture recoils,-induced electron interactions, and transmutation maction recoils. In reactor pressure vessels, special circumstances may be encountered where these reactions may become significant with respect to fast neutron induced displacements. General considerations governing relative contributions are described,and recent knowledge gained by evaluating a particular case, the HFIR reactor vessel is reviewed.Research sponsored by the Division of Materials Sciences, U.S. Department of Energy under Contract DE-AC05-840R21400

1996 TMS ANNUAL MEETING AND EXHIBITION

with Lockheed Martin Energy Systems. 2:30pm FLUENCE DEPENDENCE OF TENSILE PROPERTIES OF FERRI TIC STEELS IRRADIATED AT 5060'C IN A HIGH NEUTRON FLUX: S. T. Mahmood, K. Farrell, Metals and Ceramics Division, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831-6376 (Present address: General Electric Vallecitos Nuclear Center, Pleasanton, CA 94566) Five ferritic steels representing materials used for nuclear reactor pressure vessels, nozzle forgings and vessel support structures were irradiated at 50-60°C in a very high flux to fluences in the range 1.5xl020 to 3.5xl023 nlm2 (0.05 dpa ) at nine fluence intervals. Tensile strengths were increased and ductilities reduced in a consistent pattern, with little or no dependen,ce on chemical composition. An exponential expression of the form t.ay '" A( I_e-~l)n with n' equal to about 0.7, gives a singular description of the fluence dependence of the change in yield strength. This is shown to be consistent with a new model of radiation strengthening based on clusters of self interstitial atoms. Comparison of these high flux data for A212B steel with literature data show no significant effects of damage rate over a range of flux spanning five orders of magnitude. Research sponsored by the Division of Materials Sciences, U.S. Department of Energy under contract DE-AC05-84OR21400 with Lockheed Martin Energy Systems. One of the authors (S. T. Mahmood) was supported by an appointment to the Oak Ridge National Laboratory Postdoctoral Research Program administered by the Oak Ridge Institute for Science and Education. 2:50pm METASTABLE REVERSAL OF THE MARTENSITIC PHASE TRANSITION IN Nb3Sn INDUCED ENERGETIC-ELECTRON IRRADIATION: C. L. Snead, Jr., Applied Technologies Division, Brookhaven National Laboratory, P.O. Box 5000, Upton, NY 11973; R. C. Birtcher, M. A. Kirk, Materials Science Division, Argonne National Laboratory, Argonne, IL 60439 NbJSn upon cooling undergoes a structural martensitic phase transition from the cubic A-15 structure to a tetragonal one commencing at about Tm - 50 K, with the cIa of the tetragonal structure increasing as the temperature is decreased until the ratio becomes constant at Tc. In conducting electron irradiations between 0.4 and 3.0 MeV and measuring resistivity changes at 20 K in order to determine the damage threshold energy, it was observed that in all irradiations the resistivity initially decreased, contrary to all experience with irradiated metals. For electron energies above the energy determined to be threshold for the production of Frenkel defects(about 0.4 MeV) the resistivity eventually increased with increasing dose as expected. For energies below threshold, however, the decrease in resistivity was seen to approach a limiting value of about 0.4% of the initial resistivity. 1be resistivity vs. temperature plot shows a deviation of the curve to higher-resistivity values commencing at Tm and increasing the deviation with decreasing temperature as the tetragonality increases. We ascribe the observed decrease in resistivity with electron irradiation as a metastable reversal of the phase toward a state of reduced tetragonality(reversal of the phase transition). This resistivity decrease is completely recovered in annealing to slightly above Tm. Highvoltage electron microscopy at 20 K confirms the decrease in tetragonality by a disappearance of the twin boundaries associated with the tetragonal phase. The microstructure responsible for the resistivity changes will be discussed. Work performed under the auspices of the Department of Energy. 3:10pm RADIATION DAMAGE PROBED BY POSITRONS: Bent Nielsen. K. G. Lynn, Departments of Applied Science and Physics Department, Brookhaven National laboratory, PO Box 5000, Upton, NY 11973-5000; O. W. Holland, Solid State Division, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831-6033; A. van Veen, Delft University of Technology. 2629 JB Delft,1be Netherlands 1be senSitivity of the positron annihilation technique to vacancies and small vacancy clusters, not yet resolvable by the electron microscope, is one of the major potentials of the technique in studies of radiation damage. Defects formed by energetic ions, electrons and neutrons in metals and semiconductors have been studied using positrons. Defect evolution during annealing has been followed (vacancy migration, void nucleation, etc.) 1be role played by impurities (N and H) has been examined - including the observation of effects such as enhanced defect production and strongly reduced vacancy mobility. Further examples of secondary defect formation are presented and the underlying mechanisms are discussed. Supported by the U.S. Department of Energy, Division of Materials Sciences, Office of Basic Energy Sciences under Contract No. DE-AC02-76CHOOOI6 and Contract No. DE-AC05-84-0R21400. 3:30 pm BREAK 3:50pm TECHNOLOGICAL DEVELOMENTS CONNECTED WITH RADIATION EFFECTS RESEARCH ON LIGHT-WATER REACTOR VESSEL STEELS: R. K. Nanstad, Metals and Ceramics Division, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831-6151 1be effects of radiation on ferritic pressure vessel steels have been studied for over four decades and are driven primarily by commercial light-water reactor applications. Work has focused on two themes: the ability to predict the effects of radiation on existing reactor vessels and the fabrication of radiation-resistant steels. 1be need for integrity analyses of vessels with radiation-sensitive steels has resulted in advances in the development of elastic-plastic fracture mechanics, which have found widespread application in other technologies.1be need to understand compositional factors has led to improvements in microstructural characterization techniques at the atomic level. Advances in fabrication of heavy-section steel components for miclear reactors have also found application in other technologies. All these developments are discussed and descriptions of current research are presented.

1996 TMS ANNUAL MEETING AND EXHIBITION

SOLIDIFICATION, JOINING AND HEAT TREATMENT TMS General Abstract Wednesday, PM February 7,1996

Room: B3 Location: Anaheim Convention Center

Session Chairperson: TBA 2:00pm THERMAL EXPANSION AND SOLIFICATION BEHAVIOR OF Mo-V ALLOYS: J. Stephens, B. K. Damkroger, S. L. Monroe, R. H. Moore, Sandia National Laboratories, Albuquerque, NM 87\85-0340 Abstract not available. 2:20pm MICROSTRUCTURES OF SENSITIZED ALLOY 600 SURFACE-MELTED BY C02 LASER: Joung Soo Kim, Jeong Hun Suh, Yun Soo Lim, n Hyun Kuk, Korea Atomic Energy Research Institute, P. O. Box 105, Yusung, Taejun, Korea Microstructures of sensitized Alloy 600 the surface of which had been melted by a C~ laser was examined under transmission electron microscope(TEM) and analyzed using EDX attached to the TEM.1be molten metal solidified epitaxially from the boundary between the melted and unmeted(matrix) regions. The cells formed in the grains during rapid solidification were observed to have different orientations depending on the orientation of each grain. 1be boundaries of the cell and the grains formed during the rapid solidification were with higher density of dislocations than that in the cells, and very fine precipitates (less than 50 nm) were often observed to form along the boundaries. Also, it was interestingly observed that the concentration of Cr along the boundaries of the cells and the grains was higher than that in the cells. From these observations, it can be concluded that the prevention oflGSCC in sensitized and subsequently laser-melted Alloy 600 might be attributed to the desensitization and the segregation of Cr along the boundaries of the cells and the grains occurred during the rapid melting and solidification processes. 2:40pm MICROSTRUCTURAL STUDY OF THE BEAM WELDS ACROSS DISSIMI· LAR TI·ALALLOYS: O. Umezawa. S. Tsukamoto, National Research Institute for Metals, 1-2-1 Sengen, Tsukuba, Ibaraki 305, Japan 1be microstructures of TI-AJ alloys produced by electron beam or laser re-melting on TI3AI (0C2) and ')'-TiAI substrates have been investigated in the equiatomic range. 1be substrate was overlaid with samples (TI- 43, 49, 53, 55 at.%AI) 0.5 mm thick. Lap weldings were then carried out to allow the rapid solidification from the bottom of partial penetration beam welds. 1be beam penetration into the substrate was minimized just below the dissimilar alloy interface. Phases observed are in agreement with those produced by other rapid solidification methods. An optically featureless, planar zone was observed at an interface between the solidified zone and the substrate; the zone gave way to a dendritic structure away from the interface. Effects of the substrate on the microstructures have been also examined. Epitaxial growth from the ')'-TIAI su)). strate into the melt was not detected. This work was done under auspices of the AADC project, a collaboration between University of Cambridge and the Research and Development Corporation of Japan. 3:00pm ROLE OF THE MICROIMACRO STRUCTURE OF WELDS IN CRACK NUCLEATION AND PROPAGATION IN AL·LI ALLOYS: G. E. Talia, Wichita State University, Department of Mechanical Engineering, Wichita, KS 67208; A. C. Nunes, NASA, Marshall Space flight Center, Huntsville, AL 35812 A model relating the microlmacrostructure of welds of aluminum alloys with the fracture event and fracture criterion capable of predicting the mechanical behavior of these welds are sought. A study was therefore conducted on the microstructure aspects of crack nucleation and propagation in welds of 2195 AI-Li AIIoy. These welds were produced by the Variable Polarity Plasma Arc (VPPA) process. The critical role of grain boundaries. phases, and other features of the microstructure were identified and compared with the crack propagation paths. Analysis of weld fusion zone revealed instances of intergranular porosity and the segregation of a Cu-rich second phase. Optical and electron micrograph of deformed specimens displayed a ductile grain material and a Cu-rich phase brittle in nature. 1be fracture strengths appear to be controlled by planes of weakness dominated by the second phase material. 3:20 pm BREAK 3:30pm ANALYSIS OF AGING HARDENING IN NI·Cr·AI ALLOYS: Uu Gui/an. Liu Wei, Zhu Liming, Northeastern University, Shenyang, China In the present work, the microstructure morphologies and the micro-zoned composition in the wire specimen with a diameter of 0.3 mm of Ni-Cr-3%AI-0.2%Ce aIloy, aged at 500-600·C, have been examined by Transmission Electron Microscopy (TEM), Scanning Electron Microscopy (SEM) and Programmed Image Analysis as well as Quantitative Extraction Technique (QET) etc. It has shown that y' precipitated phase with a fcc superstructure is a primary strengthening one which is separated in a coherent pattern with respect to (100) planes of the matrix yand assumes granular form. A great quantity of supersaturated Cr(a)-rich phase, it may be termed combination strengthening.

B123

3:50pm EFFECTOFPOST-SOLUTION TREATMENT ON THE MICROSTRUCTURE OF Zn-ll'il>AI: Kimberly Duong. Souping Yan.FarghalliA. Mohamed. Materials Science and Engineering. Dept of Chemical and Biochemical Engineering. University of California. Irvine. CA 92717 An examination of the microstructure of Zn-22'i1>AI after quenching from the single phase region has revealed that groups of fine and li-phases are surrounded by traces of former a-boundaries. During superplastic deformation. the former a-boundaries hehave as independent domains. The presence and density of these former a-boundaries have been investigated as a function of impurity level and annealing conditions. 4:10pm STRUCTURE-PROPERTY RELATIONS IN A ruGH-STRENGTH CUPRONICKEL ALLOY: R. J.Orylls. M. H. Loretto. I.R.C. in Materials. The University of Birmingham, Edgbaston. Birmingham. BI5 2TI. England; C. D. S. Tuck. Langley Alloys Ltd.• Alloys House. COrdwallis Park. Maidenhead. Berkshire. SL6 7BU. England Marinel is the latest alloy in a series of age-hardening. wrought cupronickels. It combines high strength with excellent resistance to corrOSion and hydrogen embrittlement. and is designed for use in critical applications such as subsea bolting. Strengthening is provided by additions of Mn. AI. Fe. Nb and Cr. The aim of the present work is to understand the nature of the a1loy's microstructure and hardening mechanisms such that mechanical properties can be optimised. The phases present in Marinel have been identified using electron microdiffraction and analytical electron microscopy. The microstructure of the alloy has been studied after various heat-treatments. which has led to a detailed understanding of the strengthening mechanisms. Production of trial ingots with differing compositions has established the relationships between alloying additions. microstructure and mechanical properties. This paper will describe the initial results of this alloy development program, indicating the routes whereby mechanical property enhancement has been achieved.

STRUCTURE AND MORPHOLOGY OF EPITAXIAL THIN FILMS SESSION VI: Characterization Of Epitaxial Systems

Sponsored by: EMPMD ThIn Films & Interfaces Committee Program Organizer: Dr. David E. Jesson. Oak Ridge National Laboratory. P.O. Box 2008. Oak Ridge, TN 37831-6030 Wednesday. PM February 7.1996

Room: Orange County 4 Location: Anaheim Marriott Hotel

Session Chairperson: Brad Orr. Dept. of Physics. The University of MiChigan. Ann Arbor. MI 48109-1120

2:00 pm Invited AN AFM STUDY OF THE GROWTH MECHANlSMSAND MORPHOLOGIES OF SOLUTION-BASED CRYSTALS. J. J. De Yoreo. A. J. Malkin. T. A. Land. Lawrence Livermore National Laboratory. Livermore. CA. 94550; Yu G. Kuznetsov. J. D. Lee. A McPherson. Department of Biochemistry. University of California. Riverside, CA. 92521 Most studies of the nanometer scale morphology of crystalline surfaces have focused on the growth of thin ftIms by molecular beam epitaxy or sputter deposition where growth proceeds at high supersaturation and low flux through 20 nucleation and multi-layer growth and where the critical nucleus consists of as few as one atom. In contrast, many bulk single crystals. both organic and inorganic. are grown from vapors or solutions at much lower supersaturations and higher fluxes where stable islands consist of tens to hundreds of molecules. We describe the results of both freeze-and-look and in situ atomic force microscopy (AFM) on single crystal surfaces of both inorganic and macromolecular systems grown from solution. The lattice spacings in these systems range from less than 5 A for KH2P04 (KDP) and CaC03 to over 80 A for the protein Canavalin and 160 Afor the Sattelite Tobacco Mosaic Virus (STMV). We observe multiple growth mechanisms including step flow on hillocks formed by screw dislocations as well as 2D and 3D nucleation of islands. We compare the growth morphologies to the predictions of classic growth models and from measurements of step and island dynamics we calculate the step edge free energies and the kinetic coefficients of elementary steps. This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under contract No. W-7405-ENG-48. 2:30pm FASCINATING PHYSICS OF THE PblGe(111) INTERFACE:. J. M. Carpinelli. Department of Physics. University of Pennsylvania, Philadelphia, PA 19104; H.H. Weitering. Department of Physics and Astronomy. The University of Tennessee. Knoxville. TN 37996; E.W. Plummer. Solid State Division, Oak Ridge National Laboratory. Oak Ridge. TN 37831; Roland Stumpf. Sandia National Laboratory. Albuquerque. NM 87185 The morass of different phases of thin Pb films on Ge(l1l) exhibit a tremendous assortment of physical phenomena. We have investigated the structural and electronic properties of these films as a function of temperature and coverage using STM. LEED. and EELS. 1\vo metal-nonmetal transitions are observed for the ('''3 x -.J13)R3O° T4 adatom structure. The first occurs at room temperature with increasing Pb coverage. as this interface evolves from the mosaic phase to the alpha phase. The second occurs gradually and reversibly with decreasing temperature (onset near -50°C) on the alpha phase. accompanying the formation of a (3 x 3) surface charge density wave and corresponding structural rippling. First principles calculations of the alpha phase indicate

B124

that there is a basic instability in this interface that could lead to the observed low-temperature structure. However. this predicted instability is too low in energy and produces atomic displacements that are too small to fully explain our observations. 2:50pm PREPARATION AND CHARACTERIZATION OF Tl02 THIN FILMS BY PLASMA ENHANCED CHEMICAL VAPOR DEPOSITION: In-Soon Lee. Yoon-Bong Hahn. Department of Chemical Engineering & Technology. Chonbuk National University. Duckjin-Dong: IGa. Chonju 561-756. Korea Dielectric TI~ thin films were prepared on p-Si( I00) substrates by plasma enhanced chemical vapor deposition using titanium isopropoxide and oxygen. Experiments were carried out under various operation conditions. Properties of the films were characterized by Ellipsometer. Auger electron spectroscopy. X-ray diffraction. FT1R. SEM. EDX. and LCR meter. The deposition rate was little affected by oxygen flow rate. but much influenced by RF power. substrate temperature. gap distance between electrndes. and feed rate of precursor. Morphology of the film became coarser with increasing the deposition time. and the film showed less uniformity at high deposition rates. It was also found that the deposition rate is controlled by surface reaction below 200°C. but controlled by mass transfer at higher temperatures. The effect of plasma annealing on the electrical properties of as-deposited TI02 films was also investigated. 3:10pm ATOMIC STRUCTURES AND GROWTH MORPHOLOGIES OF ELECTRODEPOSITED METAL TmN FILMS ON SINGLE CRYSTAL GOLD SURFACES OBSERVED BY IN-SITU ATOMIC FORCE MICROSCOPY: N. Ikemiya.S. Haca. Department of Materials Science and Processing, Faculty of Engineering. Osaka University. Yarnadaoka 2-1. Suita 565. Japan We have investigated atomic .structures as well as nucleation and growth mechanisms of electrndeposited Ag. Cu. and Te thin films on single crystal Au surfaces using in-situ atomic force microscopy. We have determined atomic structures of underpotentially deposited metal monolayers in acidic electrolytes. We have found the growth morpbologies of bulk deposited metal thin films are controlled mainly by the surface diffusion processes of metal adatoms on substrates. We will discuss epitaxial relationships between metal thin films and substrates and the relaxation mechanisms of strained thin Te films with large lattice misfits. 3:30 pm BREAK 3:50 pm Invited LOW TEMPERATURE EPITAXIAL GROWTH OF SI: S. M. Yalisove. Department of Materials Science and Engineering. University of Michigan. Ann Arbor. MI 48109-2136 Homoepitaxial growth of Si has: been shown to occur at temperatures far below traditional epitaxial temperatures via a limited thickness epitaxy pathway. The atomistic mechanisms which control limited thickness epitaxy are not yet understood. Many factors. such as impurities. surface roughness. temperature. and energy of the arriving species can alter the kinetics of the mechanism. Work will be presented which suggests that the phenomenon is intrinsic and controlled via a specific nucleation site. Data from studies on patterned substrates are consistent with the idea that the nucleation site for the crystalline to amorphous phase is at the intersection of (III) microfacets. The density of the microfacets can be controlled and shown to have a strong effect on the process of limited thickness epitadal growth. These data will be discussed in the context of other systems which display similar phenomena. 4:20 pm Invited EVOLUTION OF SURFACE MORPHOLOGY OF EPITAXIAL CALCIUM FLUORIDE FILMS: L J. Schowalter, Byong Kim. Dept. Physics and Ctr. Integrated Electronics and Electronis Manufacturing. Rensselaer Polytechnic Institute. Troy. NY 12180 The growth of CaF2 on Si(l1l) vicinal substrates (tilted towards the [I I -2) azimuth) has been studied using RlIEED. AFM. and SEM. For growth at 770°C. the CaF2 grows in a layer-by-Iayer fashion for the first two monolayers as indicated by RHEED intensity oscillations. After two monolayers. CaF2 strongly prefers to nucleate at the Si step edges and grow laterally with constant height. The Si step edges will form step bands for this substrate orientation (-5 nm for 20 miscut substrates). Interestingly. once the CaF2 nucleates at a step band. it immediately grows to the same height as the step hand and then grows out and along the step band will maintaining a uniform height. Eventually these thick islands coalesce and form a relatively flat overlayer. These results can be understood if the relative energy (per molecule) to form a third layer is very high and can not be explained as dl~e to relative mobility differences between thick and thin CaF2. 4:50pm CHARACTERIZATION OF LOW TEMPERATURE GROWTH OF A HETEROEPITAXIALSYSTEM: G. Vidali. Hong Zeng. and Chi liu. The Solid State Science and Technology. Program and Physics Department. Syracuse University. Syracuse. NY 13244 In-situ and in real time characteIization by atum beam scattering of deposition of Pb on Cu(OOI) at 150 K gives the opportunity to study the growth of a heteroepitaxial system with large lattice mismatch from submonolayer coverage to hundreds of layers. The first Pb layer forms a high order commensurate square symmetry phase; this is fOllowed by double-stepped islands growth in the (001) Pb direction. From the fourth layer to the 16th layer. there is a quaSi layerby-Iayer growth of Pb(111); this is to be contrasted with cluster growth for deposition at room temperature. Then the film roughens, with the vertical width growing as - ,0.3; the average domain size decreases and Pb( 111) mounds start to form. After the 40th layer the growth of the vertical width saturates. This is attributed to the fact that the size of the base of the mounds approaches the typical size of the coherent domains imposed by the mosaic structure of the first Pb layers. At this point the lateral correlation length has reached the lateral dimension of domains. The growth then continues with no further significant change of mounds upto

1996 TMS ANNUAL MEETING AND EXHIBITION

450 layers. S:10pm THE STRUCTURE AND MORPHOLOGY OF EPITAXIAL DyP ON GaAs: L P. Sadwick, P. P. Lee, M. Patel, M. Nikols, R. J. Hwu Depanment of Electrical Engineering, The University ofUT, 3280 MEB, Salt Lake City, Utah; R. Alvis, One AMD Place, MS 32, Sunnyvale, CA 94088; R. Gedridge, Jr., T. Groshens Naval Air Warfare Center, China Lake, CA The structure and morphology of epitaxial dsyprosium phosphide (DyP) on (001) gallium arsenide (GaAs) has heen studied by double crystal and two theta X-ray diffraction coupled with high resolution transmission electron microscopy (TEM). DyP was grown by molecular beam epitaxy (MBE) as a function of substrate temperature and phosphorus overpressure. The room temperature lattice constants ofDyP and GaAs are 5.6534 and 5.6533 Angstroms, respectively. DyP on GaAs could be easily identified as it has significantly different structure factors than that of GaAs. As an example, the 002 peak is larger by a factor of 20 compared to the 004 peak for DyP; whereas the 004 peak is a 100 times stronger than the 002 peak for GaAs. X-ray, TEM, SEM, and high resolution microscope results show a strong dependence of the crystal structure and orientation and morphology of DyP as a function of substrate growth temperature and phosphorus pressure. Details of the DyP structure and morphological properties will be presented.

STRUCTURE STUDIES AND PHASE TRANSFORMATION

TMS General Abstract Wednesday, PM February 7, 1996

Room: Orange County 2 Location: Anaheim Marriott Hotel

Session Chairperson: TBA

1:00pm TOMOGRAPHY OF POLYCRYSTALLlNE MATERIALS: S. R. Stock, D. P. Piotrowski, A. Guvenili, C. P. Patterson, J. D. Haase, School of Materials Science and Engineering, Georgia Institute of Technology, Atlanta, GA 30332-0245; Z. U. Rek, Stanford Synchrotron Radiation Laboratory, Stanford University, Stanford, CA 943090210 In understanding the macroscopic response of polycrystalline structural materials to loading, it is frequently essential to know both the three-dimensional distribution of strain and of micro-texture. The methods must be nondestructive, however, if the evolution of quantities such as strain at a fatigue crack tip are to be studied. This paper describes approaches for high resolution synchrotron x-ray diffraction tomography of polycrystalline materials. Preliminary experiments are reported on partially cracked compact tension samples of AI-Li 2090 and on model samples of randomly-packed, millimeter-sized pieces of single crystals. Polychromatic beams collimated to diameters as small as 30 urn have been used, and collecting the spatial distribution of diffracted intensity on image storage plates as a function of sample to detector separation allowed inference of the depth of the volume elements contributing to diffraction. This research was supported by the US Office of Naval Research and was partially done at SSRL which is operated by the Department of Energy, Office of Basic Energy Sciences. 1:20pm NANO-STRUCTURALINVESTIGATION OF NICKEL-BASE SINGLE CRYSTAL SUPERALLOYS USING MONTE-CARLO SIMULATIONS AND ATOM PROBE MICROANALYSES: H. Murakami, Y. Saito, H. Harada, National Research Institute for Metals, 1-2-1 Sengen, Tsukuba-City, 305, Japan The atomic structure of some multicomponent Ni-base multicomponent single crystal superalloys has been numerically estimated using the Monte-Carlo Simulation (MCS), and experimentally analysed using Atom-Probe field ion microscopy (APFIM), respectively. The MCS was programmed so as to produce ladder diagrams, composition profiles etc. which are also experimentally obtained using APFIM analysis. The verification of MCS was thus carried out by comparing and contrasting the two simultaneously-formatted graphs, obtained by the two methodologies. In this study, the chemistry of yand '( interface, partitioning behaviour of alloying elements and substituting behaviour of alloying elements in the y phase will mainly be discussed. This work was carried out under the auspices of the U.K.-Japan cooperative Science program, organized by the Royal Society and the Japan Society fo the Promotion of Science. 1:40pm SURFACE SHORT-RANGE ORDERlNGOF CU;JAU ABOYE TC IN THE NEAR SURFACE REGION (001); M. Kimura, Advanced Materials & Technology Research Laboratories, Nippon Steel Co., 1618 Ida, Nakahara-ku, Kawasaki 211 Japan; J. B. Cohen, Department of Materials Science and Engineering, Northwestern University, Evanston, IL 60208; S. Chandavarkar, Depanment of Physics, University of Tennessee, Knoxville, TN, 37966; K. Liang, Exxon Corporate Research Laboratory, Annadale, NJ08801 The shon-range order in the near surface region of the CU3Au (001) face was investigated above the critical temperature by glancing-incidence X-ray diffraction, measuring the diffuse intensity throughout a two-dimensional region of reciprocal space. This intensity was analyzed quantitatively to obtain the two-dimensional Cowley-Warren shon-range-order parameters and atomic displacements. Monte-Carlo simulation based on these values have revealed that the atomic configurations in the surface consist of ordered domains and clusters in a disordered matrix. There is a large number of [10] anti-phase domain boundaries (APDB).

1996 TMS ANNUAL MEETING AND EXHIBITION

3:00pm THE LOCAL ATOMIC STRUCTURE OF FE-31 % PO ABOYE THE TRANSFORMATION TEMPERATURE: J. J. Felten, Containerless Research, Inc., 910 University Place, Evanston, IL 60201; J. B. Cohen, Department of Materials Science & Engineering, Northwestern University Evanston, IL, 60208 The diffuse x-ray scattering from an Fe-Pd single crystal quenched from 900°C containing 31 atomic percent palladium was analyzed to obtain information about the local atomic environment around each atom in the form of the shon-range order parameters and the average atomic displacements from the normal lattice positions. A computer program was used to simulate the lattice based on experimentally observed shon-range order parameters and displacements. We found an interconnected distribution of ordered and clustered Fe regions no more than a few atoms wide which appeared similar to spinodal decomposition and small distoned regions that may be related to the formation of the face-centered tetragonal (PeT) embryos surmised by Oshima et al. and others. 3:20 pm BREAK 3:30pm QUANTITATIVE EVALUATION OFyi PRECIPITATE SHAPES IN BlNARY NI-BASE ALLOYS: T.-H. Cha, A. J. Ardell, Department of Materials Science and Engineering, University of California, Los Angeles, CA 90095-1595 We introduce a new parameter, I= 4~, to characterize the morphology of 'y precipitates in Ni-base binary alloys; A is the projected area and P the perimeter of the image of a dark-field TEM image of a '( precipitate in a [IOO]oriented thin foil. The main advantage of using 1:, which decreases from I to 0.785 as the shape changes from spheres to cubes, is that 1: is conveniently measured using available image-analysis software. In addition to re-examining some previous data on Ni-Al, Ni-Ge and Ni-Ga alloys, we examined the morphology of Ni3Si precipitates in alloys containing 6.50 and 7.42 wt. % Si. As expected, decreases as the "radius", r, increases in all the alloys. The rate at which changes with r, the "Inorphological evolution rate" M =OdldrO, is smallest in Ni-Si alloys (for which li is the smallest), but there is no systematic relationship between 1: and M in any of the other alloys. M is independent off'( in Ni-Si alloys, but there is some evidence that M is larger the larger the value of f' in the other alloys; more data are needed to make a definitive statement, however. We fmd a very good correlation between and a modified elastic energy parameter, W'f =s2r/f113, for all the alloys. This work is supponed by the National Science Foundation. 3:50pm THROUGH THICKNESS MICROSTRUCTURAL GRADIENTS IN THE THICK PLATE ALUMINUM 70S0-T7451 ALLOYS: M. A. Przystupa, J.Zhang,A. J. Luvano, Depanment of Materials Science and Engineering, University ofCalifomia, Los Angeles, CA 90024 The purpose this work was to characterize through-thickness microstructural gradients in the six-inch-thick new and old variants of the aluminum 7050-T7451 plate alloys. The features characterized included micropores, constituent particles and grain structures. They were all measured at the plates surface, quarter depth and center locations on the TL, ST and LS planes. In all cases the size distributions, shapes and directional alignment were analyzed using the linear intercept method. For micropores and constituent particles the characterizations also involved measurements of the volume fractions and spatial distributions. The latter were quantified using the nearest neighbor spacing obtained via Voronoi (Dirichlet) tessellation techniques. Results of all above measurements will be discussed in details during the talk. Work supported by Office of Naval Research under Grant No. NOOOI4-91-J-1299. 4:10pm DISSOLUTION OF THETA PHASE PRECIPITATES IN AL-CU: C. M. Malam, B. R. Patterson, R. G. Thompson University of Alabama at Birmingham 1150 10th Ave. So. BEC 360 Birmingham, AL 35294-4461 There have been seversl studies done on the solution kinetics of the theta phase precipitate in a binary AI-Cu system. Each of these studies employs a rapid up-quench from an equilibrium two phase condition, followed by an isothermal treatment of varying time. The presentation of these data bas been in various formats which do not allow for ready comparison of experimental results. Some studies have used a quantitative microscopy approach while others have used measurements of single particles to obtain kinetic data. The present study seeks to combine each of the previous experiment's results into the readily useable format of radius vs. time. Such a single format will allow for comparison of results between studies. 4:30pm KINETIC PARAMETERS FOR PRECIPITATION IN AI-L1 ALLOY: AN 1MPROYED ANALYTICAL METHOD: W. V. Youdelis, Dept. of Mechanical & Materials Engineering, University of Windsor, Windsor, Ontario, Canada N9B 3P4; A. Luo, Institute of Magnesium Technology, Inc., 357, rue Franquet, Sainte-Foy, Quebec, Canada GlP4N7 Using a modified Avrami-Johnson-Mehl equation, a new analytical method is developed for the study of precipitation reactions using differential scanning calorimetric (DSC) methods. The kinetic parameters for GPB zone formation and S' (Al2CuMg) precipitation in Al-Li 8090 alloy, determined using the new method, are in good agreement with previously published data. The new method has the advantage of economy over the conventional method (which necessitates using several heating rates) in that all the required kinetic parameters for the transformation are obtained from a single heating rate. 4:50pm KINETIC STUDIES OF NANOCRYSTALLINE GROWTH AND THE ONSET OF EQUILIBRIUM PHASES IN A40FE5GDS METALLIC GLASSES: A. A. Csontos, W. A. Jesser, G. J. Shiflet, Department of Materials Science and Engineering, University of Virginia, Charlottesville, VA 22903

B125

Observations of nanocrystalline growth from an amorphous AlooFesGdS matrix and the subsequent transformation to the equilibrium phases were conducted. AlooFesGds metallic glasses were melt spun and heat treated from one minute to two weeks through a temperature series from 483 to 773 Kelvin. Kinetic growth data were established for the nanocrystalline growth and the onset of the equilibrium phases determined by conventional transmission electron microscope (TEM). Clarification regarding the stability and cessation of the nanocrystalline growth before and after the formation of the equilibrium phases were further analyzed and correlated by use of high resolution in-situ electron microscopy with a high resolution transmission electron microscope equipped with a hot-stage. We thank the University ofVrrginia Academic Enhancement Program which made this research possible. 5:10pm DISLOCATION MECHANISM OF PHASE TRANSITION IN POTASSIUM CHLORIDE, E. Zaretsky, Pearlston Center for Aeronautical Engineering Studies, Department of Mechanical Engineering, Ben-Guiron University of the Negev, P. O. Box 653, Beer-Sheva, 841OS, Israel A dislocation mechanism of shock induced phase transition in potassium chloride is suggested. The mechanism employs a high-speed dislocation source described in (I). Triggering of this source (a pinned segment of partial dislocation) by a strong shock induced shear stress results in the generation, motion and multiplication of transformation dislocations in the initial BI lattice (rocksalt-type; two face-centered cubic sublattices of ions of opposite sign). The first stage of the transformation results in the development of an intermediate structure B* containing two ionic sub lattices of the same-sign, the unchanged face-centered cubic sub-lattice and a transformed simple cubic (or orthorhombic) sub-lattice. At the second stage the transformation of B* into the final B2 (cesium chloride) structure with two simple cubic sub-lattices of ions of opposite sign, takes place. In the first BI to B* stage, the unrelaxed shear stress favors easy dislocation climb and the high-speed dislocation multiplication. In this stage the transformation rate is very high, about 1-10 nsec- I . In the second B* to B2 stage, easy crossing of the dislocations and their forward motion are inhibited because of the decrease of shear stress and overlapping of the regions of new phase. In this stage the transformation rate is comparatively low, about S-25 sec-I. The suggested mechanism is in agreement with results of experimental measurements of the kinetics of phase transformation in potassium chloride (2), and flash X-ray diffraction study of structural changes accompanying the transformation (3).

SYNTHESIS AND PROCESSING

TMS Genera. Abstract Wednesday, PM February 7,1996

Room: B4 Location: Anaheim Convention Center

Session Chairperson: TBA

2:00PM FACTORS LEADING TO THE MICROSTRUCTURAL REFINEMENT AND ELIMINATION OF MICROPOROSITY IN DROPLET-BASED FREEFORM FABRICATION OF STRUCTURAL MATERIALS: M. Orrne,J. Courter, C. Huang, Department of Mechanical And Aerospace Engineering,University Of California,lrvine, CA 92717-3975 The development of a droplet-based freeform fabrication technique for the fast manufacture of metallic structural components directly from CAD information is presented. The technique relies on the use of precisely controlled molten nano-liter droplets as the deposition element. Encouraging experimental results indicate that typical material porosities as low as 0.03% and microstructures with grain sizes on the order of a micron can be achieved with molten droplet throughPUt rates as high as .18 mils with a single droplet stream. Extremely high control of the droplet trajectories (measured angular dispersions of the order of I micro-radian), and droplet speeds (measured speed dispersions ofthe order of 3xlo-7 times the average stream speed) provides the ability for precise control over the solidification characteristics and hence microstructure. Experimental results from parametric studies combined with results from numerical analyses on the solidification characteristics of the droplets determine a control scheme for achieving a controlled droplet deposited part with minimum porosity and microstructure. 2:20pm FABRICATION OF TI-Al INTERMETALLIC SHEETS VIA HOT ROLLING OFTI AND Al MULTILAYER FOILS: W. J. Park, W. C. Lee, S. Lee, NackJ. Kim, Center for Advanced Aerospace Materials, Pohang University of Science and Technology, San 31, Hyojadong, Pohang 790-784, Korea 11-Al based interrnetallics possess high specific strength and specific modulus, and high resistance to oxidation, which make them attractive for a variety of applications. However, these intermetallics are difficult to fabricate into near net shape using conventional processing due to their lack of ductility. It has recently been shown that a self-propagating, high temperature synthesis (SHS) can be used to produce near net shape intermetallic parts based on 11 and AI. Such processes usually involves the press and sinter or powder forging of elemental powders or the hot pressing of elemental foils. The present study is aimed at fabricating the 11-AI interrnetallic sheets via hot rolling of 11 and AI foils. It has been shown that the microstructure of hot rolled sheets consists of three phases; llAI, 113AI and AI311. Subsequent heat treatment of hotrolled sheets results in the changes in the relative amounts of these three phases. The microstructural development will be discussed as functions of hot rolling and heat treatment conditions. 2:40pm PREPARATION OF SIC POWDER BY SELF-PROPAGATING HIGH TEMPERATURE SYNTHESIS PROCESS FROM SILICA: Chang Whan Won, Seong

B126

Seock Cho, Byong Sun Chun, Engineering Research Center for Rapidly Solidified Materials Chungnam National University, Taejon 30S-764 The SiC powder was prepared by a SHS processing using Si~, Mg and C as raw materials to determine the optimum process variables. The results obtained from this study may be summarized as follows; I) The mole ratio between C and Mg affected the magnesiothermic reduction reliction most and the highest yield of SiC was obtained at the mole ratio of Si~: Mg: C = 1.0: 2.S: 2.2. 2) The propagation speed of combustion reaction was about 10mmlsec increasing with the decrease of Si~ particle size. 3) The amount of unreacted silicon decreased with increasing compaction pressure. 4) Impurities were found to decrease by evaporation during the high temperature reaction or by solution during leaching process. S) The particle size of prepared SiC powders was in the range of 0.2 - 1.0 microns with the purity of about 98.88S weight percent. 3:00pm ALTERNATIVE PROCESSING METHODS FOR TUNGSTEN-BASE COMPOSITE MATERIALS: E. K. Ohriner, V. K. Sikka, Metals and Ceramics Division, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831-6083 Tungsten composite materials contain large amounts of tungsten distributed in a continuous matrix phase. Current commercial materials include the tungsten-nickel-iron and related materials and also tungsten-copper materials. In the past few years, there has been interest in a wider nlllge of matrix materials, which offer the potential for superior properties. These must be processed by solid-state processes and at sufficiently low temperatures to avoid undesired reactions between the tungsten and the matrix phase. These processes, in order of decreasing process temperature requirements, include hot-isostatic pressing, hot extrusion, and explosive compaction. The results of laboratory-scale investigations of solid-state consolidation of a variety of matrix materials including titanium, hafnium, copper, nickel aluminide, and steels are reviewed, including the limited available mechanical property data. The potential advantages and disadvantages of each of the po!sible alternative consolidation processes are identified. Managed by Lockheed Martin Energy Systems under contract DE-ACOS-840R2I400 with the U.S. Department of Energy. 3:20 pm BREAK 3:30pm SINTERING OF CUBIC BORON NITRIDE PARTICLES COATED WITH ULTR-FINE TITANIUM-METAL PARTICLES: H. Yoshida, National Institute for Advanced Interdisciplinary Research (NAIR), 1-1-4 Higashi, Tsukuba, Ibaraki 3OS, Japan; S. Kume, National Research Institute of Nagoya (NIRIN), Japan; Y. Yamada, T. Fuyuki, Nisshin Aour Milling Co., Ltd. Technical Research Center, Japan; Y. Sumita, K. Kimura, Read Co., Ltd., Japan Pine-grained cBN whose grain-size is Ol2l1m was successfully coated with ultra-fine titanium-metal particles whose grain-size is less than O.ll1m by PVD without creation of hBN. Tablets of the coated cBN were processed by a cubiC-type high pressure apparatus under S.5GPa at 1723K for 30 min. That is not too severe ultra-high pressure sintering condition but typical condition for industrial production of cBN sinter bodies. Reaction of cBN and titanium-metal during sintering resulted in dense cBN sintered body whose microstructure consists of more-fine-grained cBN than raw materials, and of binding material ofllN-TtB2' Hardness of the sintered body containing 83vol%cBN was fairly higher than that of the cBN-cutting tools on the market. 3:50pm A MONTE CARLO SIMULATION OF POWDER COATING USING TUM· BLING FLUIDIZED BED: Hiroyuki Nakamura,Eiichi Abe, Noriyuki Yamada, Kyushu National Industrial Research Institute, Shuku-machi, Tosu, Saga, 841 Japan In a preparation of coated porticIes, homogeneous distribution of coating mass is often desired and a lot of effort is made to accomplish it. If it is possible to reproduce the effect of operating conditions on coating mass distributions by computer simulations, the effort will be reduced. From this point of view, in this study, Monte Carlo simulations for coating mass distribution of particles were performed to reproduce the results obtained by using tumbling fluidized bed. As the result, this method was found to be applicable to reproducing the experimental tendency of the effect of operating conditions on the coating mass distribution, showing more homogeneous coating mass distribution was achieved under these conditions; long coating time, small hold up of core particles, large mixing rate of particles etc. Furthermore, some predictions of the effect of operating conditions on c:oating mass distributions using this simulation method were performed. 4:10pm COATING MASS DISTRIBUTIONS OF THE CORE PARTICLES IN A TUMBING FLUIDIZED BED COATER: Eiichi Abe, Noriyuld Yamada, Hideharu Hirosue and Hiroyuki Nakamura, Kyushu National Industrial Research Institute, Shukumachi, Tosu, Saga 841 Japan In order to control the extent of the function of coated particles, it is necessary to clarify how each individual core particle is coated quantitatively. In this work, coating mass distribution of core particlc~s in a tumbling fluidized bed coater, having a rotary disk at the bottom of the coating chamber, was investigated experimentally. As the results, coating mass per unit surface area of core particles was found to obey a normal distribution by number. Coefficient of the variation of coating mass distribution, cv' varied depending upon the operating variables of the coater. It was found that Cv decreased with increasing coating time and size of core particle and with decreasing holdup of core particles and concentration of the coating substance in a spraying solution. However, rotational speed of the:, disk, effective to force aggregative coated particles apart, hardly affected cv' 4:30pm TI(ON)-WO-ThC-NI-Co CERMETS; PROCESSING & MICROSTRUCTURE: S. Kang, Seoul National University, Seoul, Korea (02) 880-7167 Some issues in the dissociation behavior of a carbonitride solid solution, the transport and precipitation phenomena of second phases in the binder phase were investi-

1996 TMS ANNUAL MEETING AND EXHIBITION

gated with a TI(ON)-WO-TaC-Ni-Co system. It was shown experimentally that the dissociation of the cubic phase occurred at a fixed ratio of carbon to nitrogen. This indicated that the nitrogen in the cubic phase has a stronger affinity with the metallic species that expected, making the system of interest relatively stable at higher temperatures. For the reactions occurring in the binder phase, the effects of sintering conditions on the binder structure were measured with coercive force (Ho) and magnetic saturation (a). This study showed some evidence that the precipitation in the heating stage influence the binder microstructure extensively and nitrogen can transport to the surface in the form of voids during sintering. The issues were discussed with the results of an analytical electron microscopy and thermodynamic principles. 4:50pm FOAMING PROPERTIES OF ALUMINA HYDRATE-COATED FINE VOLCANIC GLASS: Kunio Kimura, Hiroyuki Nakamura, Kyushu National Industrial Research Institute, Calceed Co., Ltd., Shuku-machi, Tosu, Saga 841 Japan; Hiromi Okada, 4611-1 Isa, Isa-machi, Mine, Yamaguchi 759-22 Japan The hollow glass microspheres larger than a few decades microns have heen produced from natural volcanic glass particles (VGP) by heat treatment. However, it was difficult to produce hollow glass microspheres less than 10 microns resulting from the insufficient amount of water in VGP for foaming. Homogeneous precipitation technique was applied to coating VGP with alumina hydrate to inhibit the release of water vapor from the particle surface under heating. VGP ranging from 5 to 10 microns were suspended in the aqueous solution of aluminum sulfate and urea. Alumina hydratecoated VGP were obtained by heating the suspension at 80T for 4 hours. Fine hollow glass microspheres of around 10 microns could he produced from alumina hydrate coated VGP by heat treatment of 1040'C for a few seconds. 5:10pm ADVANCED CERAMIC: Rajeev Agrawal, Rajendra Kumar Goya, Department of Metallurgical Engineering Malaviya Regional Engineering College, Jaipur, India Advanced Ceramics are a relatively new class of high performance material with a significant potential for future economic impact. This paper deals with processing and applications of sialons. By substituting some of the silicon and nitrogen in silicon nitride by aluminium and oxygen the sialon ceramics are formed. During sintering, silica (Si~) on the surface of the silicon nitride particle reacts with alumina and yttria, to form a low melting liquid. The silicon nitride reacts with this liquid to form sialon. Sialons have been shown to work efficiently both at low and high speeds. Sialon has been successful in machining super alloys at considerably high speeds.

SYNTHESIS AND PROCESSING OF NANOCRYSTALLINE POWDER V: Thermal Stability And Coarsening

Sponsored by: MDMD Powder Metallurgy Committee and FEMS (Federation of European Materials SocIeties) Program Organizers: Dr. David L. Bourell, The University of Texas at Austin, Materials Science & Engineering, MC C220l, Austin TX 78712; Dr. Liisa Kuhn-Spearing, Laboratory for the Study of Skeletal Disorders and Rehabilitation, Harvard Medical School, Children's Hospital, 300 Longwood Avenue, Boston MA02115; Professor Dr. Herhert Gleiter, Karlsruhe Research Center, PO Box 3640, 0-76021 Karlsruhe, Federal Republic of Germany Wednesday, PM February 7, 1996

Room: Grand G Location: Anaheim Marriott Hotel

Session Chairperson: E.V. Barrera, Rice University, Mechanical Engineering 101, PO Box 1892, Houston TX 77251

2:00pm GRAIN GROWTH OFNANOCRYSTALLINE MATERIALS - A REVIEW: T.R. Malow, c.c. Koch, Department of Materials Science and Engineering, North Carolina State University, PO Box 7907, Raleigh NC 27695-7907 Quantitative studies of grain growth in nanocrysta11ine materials (nc) will he reviewed. Methods for experimentally measuring. nc grain size and their limitations are presented. The grain growth exponents n from J)lIn - Dolin = kt for nc and conventional polycrystalline (0 > 111m) materials are compared. Grain growth data on nc materials from the literature and the authors'laboratory are analyzed by the above equation and one which takes grain boundary pinning forces into account. Activation energies for grain growth are also determined by both methods where applicable. Grain growth in nc Fe is emphasized in work from the literature and the authors'laboratory. Grain growth kinetics, in terms of the grain growth exponent, n, and the activation energy for grain growth, Q, are compared for nc Fe and conventional grain size (~O 11m) wne-refined Fe. An activation energy for grain growth in nc Fe is found to compare closely with that for grain boundary self diffusion in Fe. 2:30pm DESIGN OF NANOCRYSTALS WITH STABILITY AGAINST GRAIN GROWTH: B. Fultz, L.B. Hong, Z.-Q. Gao, Mail 138-78, California Institute of Technology, Pasadena CA 91125 Nanocrysta11ine alloy powders were prepared by high energy ball milling, and their stabilities against grain growth during low temperature annealings were tested. Grain growth was measured by transmission electron microscopy dark field imaging, in combination x-ray diffraction lineshape analysis. Our nanocrystalline alloys were solid solutions in the as-milled state. Mtlsshauer spectrometry and x-ray diffractometry were used to measure the chemical ordering or chemical segregation that occurred during annealing. The grain growth in alloys of Fe3Si and Fe3(AI,Ge) slowed considerably after short-time annealings, which was coincident with the development of 003 chemi-

1996 TMS ANNUAL MEETING AND EXHIBITION

cal order. The addition of Nb to these alloys suppressed further their grain growth, evidently because the Nb segregated to grain boundaries. We argue that grain boundary chemical segregation can be useful for suppressing grain growth, as is the development of chemical order. An exception occurs for bec Fe-Cu, however, in which the segregation of Cu atoms to grain boundaries has little effect on the suppression of grain growth. This work was supported by the NSF under contract DMR-9213447. 2:50pm MEASURING THE GRAIN SIZE DISTRIBUTION IN NANOCRYSTALLINE MATERIALS BY INDIRECT DECONVOLUTION OF THE BRAGG REFLECTION PROFILES: J. Weissmiiller, National Institute of Standards and Technology, Bldg 223 RmA153, Gaithersburg MD 20899; J. Lilffier, Paul Scherrer-Institut, Villigen, Switzerland; c.E. Krill, R. Birringer, Fachbereich Physik, Universitiit des Saarlandes, 0-66 Saarbriicken, Germany; H. Gleiter, Forschungszentrum Karlsruhe, Karlsruhe, Germany It is often desirable to characterize the grain size distribution of nanocrystalline solids, in addition to the average values for the grain size. The grain size distribution can be obtained from scattering data by solving the linear set of equations relating the experimental Bragg reflection profile of a sample to the profiles of the particles from different size classes. The influence of the grain size on the atomic distribution and interference functions can be described by the intra-grain correlation function. The Bragg reflection profile is the cosine transform of this function. Each profile is also broadened due to instrumental effects and to the mean square lattice strain. The reflection profile of a nanocrystalline solid with a distribution of grain sizes is a weighted sum of the profiles of the individual grains. The determination of the size distribution involves no assumption on the functional form of the size distribution. The grain size distribution of nanocrystalline Palladium obtained in this way from X-ray diffraction data is compared. 3:10pm DIFFERENTIAL MOBILITY ANALYZER FOR NANO PARTICLE SIZING: H. Fissan, Gerhard-Mercator-Universitlit Duisburg, ProzeS- und Aerosol-meStechnik, Bismarckstr. 81,0-47057 Ouisburg Germany; D. Pui, Particle Technology Laboratory, Mechanical Engineering Department, University of Minnesota, III Church St. SE, Minneapolis MN 55455 Nano-particles are commonly synthesized from the gas phase. For certain applications it is of great interest to be able to produce monodisperse nano-particles. However, nucleation, condensation and coagulation processes in the gas phase lead to polydisperse particles. Differential Mobility Analyzers (DMAs), based on the electrostatic classification principle, have been used with great success to select monodisperse fractions from the polydisperse particles and to perform accurate sizing of nanometer particles. The transfer function of the DMA provides a full description of the instrument performance and is used in the DMA data analysis. Up to now it has been assumed that the transfer function is only dependent on the flow and electrostatic effects and that there are no particle losses on the flow path within the instrument. While the assumption is valid for large particles, it is not applicable for particles below 100 nm diameter. It is important to take into account the diffusional effect of the nanometer particles, which causes the transfer function to broaden and significant particle losses in the instrument. Four commercially available DMAs of different designs have been evaluated for the present study. They are Challenged with defined nano-particles and their size resolution and losses have been determined. In all cases the diffusional effect is too large to be negligible. Based on these results, we will discuss the requirements for an instrument especially designed for nano-particles and how an ideal instrument can be achieved. 3:30 pm BREAK 3:50pm DENSIFICATION VERSUS GRAIN GROWTH IN NANOCRYSTALLINE ZrO:z3 MOL% Y203: HOW TO WIN: M.J. Mayo, Department of Materials Science and Engineering, Pennsylvania State University, University Park, PA 16802 The kinetics of densification and grain growth during the pressureless sintering of nanocrystalline (15 nm) Zr02-3mol% Y203 were studied for samples with green densities from 30-58%, over temperatures ranging from 700- I3OO°C, times from 0-64 hours, and heating rates from 2-2oo°C/min. The results show that a rapid rate sintering approach, which uses fast heating rates to bypass a thermodynamically unfavorable, low temperature surface diffusion regime, cannot be applied to nanocrystalline zirconia due to thermal gradient problems unique to that material. With more ordinary heating rates, the densification kinetics show a strong dependence, not only on grain size and temperature, as expected, but also on pore size. Conversely, grain growth shows no dependence on pore size. By decreasing the pore size, then, it is possible to accelerate the densification kinetics relative to those of grain growth and achieve 99.9% dense zirconia with 80-90 nm grain diameters. Successful methods of decreasing the pore size include increasing the dry compaction pressure or, alternatively, using wet consolidation techniques. Simple manipulations of the temperature/time protocol during sintering have no effect on pore size evolution and thus no effect on the final denSity/grain size combinations that can be reached. 4:10pm SOL-GEL SYNTHESIZED AND HOT PRESSED ALUMINA AND ALUMINA· ZIRCONIA NANOCOMPOSITES: J-F. Kuo, D.L. Bourell, Center for Materials Science and Engineering, University ofTexas at Austin, MC C22oo, Austin, TX 78712 Alumina and a co-precipitated nanocomposite ofYttria Stabilized Tetragonal Zirconia (YSTZ, 3 mol% Yttria) were synthesized using a sol-gel synthesis technique using chloride precursors. The effect of calcination on both phase characteristics and particle size was assessed. After crystallization of the amorphous gel, YSTZ remained tetragonal after calcinationlsintering to l250'C, but the alumina underwent a series of phase transformations involving the phases, bayerite, boehmite, gamma, eta, theta and alpha. Particle size for ceramics and composites after calcination was as fine as 10 to 15 nm with agglomerates of the size 0.5 mm. Powders were cold and hot pressed. Sintering parameters were established to minimize coarsening and maximize density.

B127

4:30pm THERMAL STABILITY OF DYNAMICALLY CONSOLIDATED BALLMILLED NANOCRYSTALLINE FE-2C POWDER: G.E. Korth, T.M. Lillo, Lockheed Martin Idaho Technologies, P.O. Box 1625, Idaho Falls ID 83415-2218; J.C. Rawers, U.S. Bureau of Mines, 1450 Queen Avenue SW, Albany, OR 97321-2198 Nanocrystalline material was synthesized by ball milling Fe-2C powder in an argon environment attritor which resulted in micron size powder particles with a substructure (grains) of 8-15 nm. TEM analysis was performed on dynamically consolidated material after I hour heat treatments up to 7OO·C. Results showed the 10-40 nm post-consolidated nanograins to he stable up to approximately 550·C, but after I hour at 7OO·C the grains had grown over an order of magnitude to approximately 500 nm. The grain size stability up to 550·C has real Significance since preliminary property results indicate possible onset of superplastic behavior at approximately 5OO·C. Work supported by the U.S. Bureau of Mines under Subcontract No. JO I 345035 with the Department of Energy Idaho Operations. 4:50pm EFFECT OF IN-SITU FORMATION OF NANOSCALE y-AL203 AND AIN DURING CRYOMILLING ON THE THERMAL STABILITY OF NANOCRYSTALLINE III AT ELEVATED TEMPERATURES: Robert J. Perez, Benlib Huang, EJ. Lavernia, Department of Chemical Engineering and Materials Science, University of California, Irvine, CA 92717-2575 The cryogenically attritor milled (cryomilled) Fe-IO wt.'II>AI powders in liquid nitrogen have heen found to retain nanocrystalline structures following consolidation at 550·C for 30 minutes using bot pressing and subsequent annealing at elevated temperatures for 60 minutes. Thermal stability of these heat treated powders was studied using dark field imaging of TEM. The crystallite size of the as-consolidated material was 11±5 nm. Subsequent annealing of the consolidated powders at 8OO·C and 950·C resulted in the crystallite size of 13±6 nm and 16±7 nm, respectively. Selected area diffraction (SAD) indicated the formation of nanoscale y.A1203 and AIN, whicb accounted for at least 2 volume percent in the cryomilled powders. Extraction replicas of the materials are being studied to determine the distribution and interparticle spacings of the y.A1203 and AIN. Zener's model is applied to provide insigbt into the interaction between the particles and the grain growth of the nanocrystalline materials. The authors would like to acknowledge the financial suppon provided by the Office of Naval Research under grants Nooo 14-93-1 072 and Nooo 14-94-0017. 5:10pm COMPUTER SIMULATIONS OF NANOPARTICLE INTERACTIONS: H.L. Heinisch, Pacific Northwest Laboratory, Ricbland, WA 99352 Interactions of idealized individual metal nanoparticles are simulated at the atomic scale using molecular dynamics with an embedded atom potential for nickel. The sintering of two identical, crystalline particles in free space, eacb initially at 900 K, is simulated for an interaction time of 250 ps. The particles reorient themselves to acbieve registry of their crystal planes, wbile necking occurs by surface diffusion. Surface layering, surface diffusion coefficients and particle melting temperature compare favorably with measured and calculated information. For simulations of more than two particles interacting simultaneously, particle-particle interfaces are examined, as well as the impact of the additional constraints on relative nanoparticle motion. Pacific Nonbwest Laboratory is operated for the U.S. Department of Energy by Battelle Memorial Institute under contract DE-AC06- 76RLO 1830.

TANTALUM: Applications And New Concepts: Session VI

Sponsored by: SMD Refractory Metals and Materials Committee and Jt. MDMDI EPD Synthesis. Control and Analysis In Materials Processing Program Organizers: Andrew Crowson, U.S. Army Researcb Office, Researcb Triangle Park, NC; Edward S. Chen, U.S. Army Researcb Office, Researcb Triangle Park, NC; Prabbat Kumar, Cabot Corp, Boyenown, PA; Willam Ebihara, Picatinny Arsenal, Picatinny, NJ; Enrique J. Lavernia, UC Irvine, Irvine, CA Wednesday, PM February 7,1996

Room:A4-5 Location: Anaheim Convention Center

Session Chairpersons: Jerry Wittenauer, Lockheed Missiles & Space Co., Palo Alto, CA; Edward S. Chen, U.S. Army Researcb Office, Researc Triangle Park, NC

1:30pm THE VAPOR DEPOSITION OF 1'8 AND 1'8 COMPOUNDS: Erik Randich, Lawrence Livermore National Laboratory, Livermore, CA 94550 Ta and many of its compounds can be deposited as coatings by techniques ranging from pure thermal Chemical Vapor Deposition (CVD) to pure Physical Vapor Deposition (PVD). There are many vapor deposition techniques wbicb are hybrids of tbe two methods. A review of the various deposition tecbniques wbicb are more closely related to CVD tban PVD will be presented. The advantages and limitations of the techniques will be compared. The need for new lower temperature processes and hence new precursor chemicals will be examined and explained. Areas for new research and development of these processes and materials will be explored. 2:00pm TANTALUM USAGE IN NICKEL-BASED SUPERALLOYS: Gary L. Erickson, Cannon-Muskegon Corporation, 2875 Lincoln Street, Muskegon, MI 49441 Tantalum bas been an imponant superalloy constituent for nearly fifty years. Similar to W and Re, it is a refractory element panly characterized by it's relatively large atomic radius and contribution to superalloy strength through coherency strain. Unlike W and

B128

Re which primarily partition 10 the gamma solid solution, tantalum contributes to both the gamma and gamma prime superalloy constituents. Additionally, tantalum is beneficial to superalloy directionally solidified component castability and promotes good environmental properties resistance. Superalloy tantalum usage has continually increased since the commercialization of the directional solidification casting process for gas turbine engine blade and vane manufacture, partly because it helps prevent freckle defect formation in single crystall cast components. Furthermore, its usage has increased due to the ability to increase superalloy strengthener content in directionally solidified articles as opposed to equiaxe aJloys. The narrative reviews the usage of tantalum in superalloys with a primary emphasis being placed upon directionally solidified compositions. It presents the rationale behind superalloy tantalum usage and the likely trend for its continued application in I,merging materials for gas turbine engine hardware. 2:30pm APPLICATION OF PM TANTALUM FOR EFP LINER: T.K. Chatterjee, J Cross, U.S. Army Research Development & Engineering Center, Picatinny Arsenal, NJ 07806-5000, C. Michaluk, Cabot Corporation, Boyenown, PA 19512 High Oxygen (250-450 ppm) content in PIM forged materials are characterized as having a higher flow stress and greater work hardening is greater at a low strain rate then at a higher rate. This behavior may be due to the intergranular tantalum oxide particles. In spite of the higher oxygen content, PIM forged tantalumexhibits favorable ductility. TEM study was perfor~ on extruded and forged PM tantalum annealed at different temperatures. A comparison was made between these microstructures and in other microstructres previously revealed in tantalum that has been processed differently. 2:50pm NEW BARRIER LAYER OF lliSlN FOR AI INTERCONNECTION LAYER AND Ta205 CAPACITOR: TohlU Hara, Department of Electrical Engineering, College of Engineering, Hosei University, 3-7-2 Kajinocho, Loganei City, Tokyo 184 Barrier effect of TaSiN layer h,u been studied. This barrier layer witb different compositions of Ta, Si and N is deposited, where precise composition is determined by Rutherford Backscattering spectroscopy measurement. This layer shows much hetter thermal stability when compared! with other barrier layers. Fundamental properties of this layer and barrier effect of tins layer for Al interconnection layer and for Ta205 dielectric layer is described. Other Ta layers, for instance, CVD Ta205 layer form memory capacitor and TaSi2 for sallicide gate in MOS LSls are also described. 3:10 pm BREAK 3:20pm CHEMISTRY MODIFICATIONS OF TANTALUM AND TANTALUM-TITANIUM-SCANDIUM ALLOYS BY PLASMA ARC MELTING: Paul S. Dunn, Deniece Korzekwa, Fermin Garcia, Materials Science and Technology: Metallurgy, P.O. Box 1663, MS G770, Los Alamos National Laboratory, Los Alamos, NM 87545 State-of-the-art melting of tanll~um and titanium alloys have relied on electron beam (EB) or vacuum arc remelting (VAR) for ingot fabrication. The limited melting techniques for tbese materials are th,e result of high melting temperatures and reactivity with mold/crucible materials. In addition, the vacuum levels typically used in EB and VAR processing limit the alloy addition candidates to those elements with low vapor pressures. Plasma arc melting (PAM) provides an alternative for melting tantalum materials with the advantage of processing under ambient pressures. The current work examines processing parameters for melting two classes of materials; I) high oxygen sintered tantalum powder and 2) a tantalum-titanium-scandium alloy. High oxygen tantalum powder: Attempts to EB melt the high oxygen tantalum resulted in continuous extinguishment of the electron guns due to the large amount of outgassing as the material melted. Plasma arc melting w,as used to consolidate the material and correlate oxygen levels with standard argon helium and argonlhelium/hydrogen cover gases. Tantalum-titanium-scandium: The melting of the Ta-TI-Sc alloy posed a major technical problem because of the boiling point of scandium, 2836·C, and the melting point of the alloy, 2650·C. Data and processing parameters will be presented that show scandium was maintained in the melt and ingot segregation is minimal. 3:50pm SINTERING OF fiGH SURFACE AREA TANTALUM POWDER: Sophie G. Dubois, Randall M. German, P/M Lab, The Pennsylvania State University, 118 Research Building West, University Park, PA 16802-6809 To achieve the higb surface ama required for the fabrication of higb performance electronic capacitors, tantalum powders are usually plate-like or agglomerated. Understanding the sintering behavior of these powders is essential to improve the manufacturing process; bowever, solid-state sintering models and simulations have only been developed for spheres. Using the concept that the driving force for neck growth is provided by the curvature gradienl, three different particle geometries were simulated: spherical, plate-like, and agglome,rated. Computer simulations and experimental sintering results of plate-like and agglomerated powders were combined to identify the critical sintering parameters for high surface area tantalum powders. Sintering maps for tantalum powders were createdl to predict surface area reduction and shrinkage during sintering. 4:10pm PROCESSING TECHNIQUES FOR TANTALUM-CARBON COMPOSITES: Keith Axler, Patrick Rodriauez, Ramiro Pereyra, Gerald DePooner, Los Alamos National Laboratory, Los Alamos, NM; Charles West, Patrick Doheny, BIRL Industrial Laboratory, Nonbwestem University, Evanston, !L. A tantalum-based alloy with Ta2C inclusions has been developed and demonstrated as a construction material for use in severely corrosive metallurgical processing environments. Alloy development involved multi-step thermal processing to invoke specific microstructural features. Grain boundary preCipitation of carbide from supersaturated conditions provides corrosion resistant attributes. The kinetics of carbide formation from supersaturated solid solutions of carbon in tantalum were established. Due to

1996 TMS ANNUAL MEETING AND EXHIBITION

its exceptional corrosion resistance, this engineered material is well-suited for containment of molten metals and salts. 4:30pm HIGH TEMPERATURE DEFORMATION IN Ta·2.5 wi. % W ALLOY: Yong li, Weidong Cai, Enrique J. Lavernia, Farghalli A. Moham~d, Department of Chemical Engineering and Materials Science, University of California, Irvine, CA 92717 1be effect of stress on the high temperature deformation behavior of Ta· 2.5 wt. % alloy has been studied by conducting a series of compressive creep tests at 1350°C.1be creep tests were performed using a specially designed materials testing system in Ar· gon atmosphere to prevent the oxidation of test specimens at elevated temperatures. 1be stress dependence of the steady-state creep rate in the alloy is discussed in reference to the creep behavior of metals and alloys.

TECHNICAL, ECONOMIC COMMERCIAL & ENVIRONMENTAL CONSEQUENCES OF SULFUR RECOVERY IN COPPER PRODUCTION SYSTEMS IV: Part 1: Economic and Environmental Effects of Sulfur Recovery Options; Part 2: Sulfur In Hydrometallurglcal Copper Production Systems Sponsored by: EPD Copper. Nickel. Cobalt Committee Program Oranizers: Norbert L. Piret, Piret & Stolberg Partners, 1m licht 12,0-47279 Duisburg German; Ivan M. Santos Moraes, CanUba Metais, Dias D' Avila, Bahia, Brazil

Wednesday, PM February 7, 1996

Room: AIl-12 Location: Anaheim Convention Center

Session Chairpersons: I. C. Knight, CRU International, 31 Mount Pleasant, London WCIX OAD Great Britain; P. B. Queneau, Hazen Research, Inc., 4601 Indiana Street Golden, CO 80403

2:00 pm Invited EFFECTIVE SULFUR RECOVERY IN COPPER SMELTERS A BLESSING ORA BURDEN?: P. G. Gocht, H. J. K6nig, P. Luedtke, D. Werner, Lurgi Metallurgie GmbH., Lurgiallee 5, 0-60295 Frankfurt am Main Germany In days gone by recovered SO:z from smelters was an appreciated valuable by-product. Once converted into H2S04, it commanded good prices from industries producing fertilizer, sodium sulfate, glass, gypsum and detergents. Even the application of sophisticated technologies to reduce stack emissions did not adversely affect the cost of metal production. But the situation changed. 1be paper deals with last five years' market development of sulfur bearing commodities, reviewing their availability, quality, pricing, output opportunities and production costs effective parameters, as well as highlighting the difficulties in forecasting the future. It discusses the impact of current smelter operations on SO:z recovery including the smelter's sulfur balance and handling of fluctuating sulfur-bearing streams. It describes recently developed processes and some advanced and long-known developments which might be applicable in the future, their progress, costs and by-product treatment. 2:30pm PRODUCTION OF ELEMENTAL SULFUR. LIQUID S02 AND SULFURIC ACID FROM HIGH STRENGTIlMETALLURGICALGAS • PROCESSES AND ECONOMICS: LeonanJ J. Friedman, Acid Engineering & Consulting, Inc., 4619 Kings Point Court, Lakeland, FL 33813 With the increasing use of oxygen smelting, economic production of elemental sulfur, liquid S02 and sulfnric acid are viable choices for handling and disposing of the sulfur values rejected by the smelter. This paper reviews the modem processes for pro. duction of each product, discusses advantages and disadvantages of each route, and presents capital and operating costs showing the effect of variables, i.e. SQ:z gas strength, energy cost, natural gas or other S02 reductant cost, oxygen cost, etc. 2:55pm TIlE SULFUR PROBLEM IN THE METALLURGICAL INDUSTRY: Fathi Habashi, Department of Mining and Metallurgy, Laval University, Quebec City, Canada Gl K7P4 During the recovery of copper, nickel, zinc, and lead from their sulfide ores, large amounts of sulfur dioxide are generated. A fraction of this gas is captured and transformed into sulfuric acid and liquid SO:z and the remainder is emitted to the atmosphere. Numerous proposals were made to solve this problem. Hydrometallurgy offers the only valid solution by which elemental sulfur instead of sulfur dioxide can be produced. Sulfur can be easily stored, transported long distances, or transformed into S02 and H2S04 when desired. 3:20 pm BREAK

3:40 pm Invited BIOOXIDATION OF COPPER SULFIDES·ENVIRONMENTAL AND ECO· NOMIC CONSIDERATIONS: RichanJ Poulin, Richard W. Lawrence, Dept of Mining and Mineral Process Engineering, University of British Columbia, Vancouver, Canada Oxidation of copper sulfides using biological leaching offers a potential route for

1996 TMS ANNUAL MEETING AND EXHIBITION

copper recovery. Oxidation of sulfide sulfur can proceed either to sulfate or to elemental sulfur depending on mineralogy and process conditions. 1be advances in commercial application of biooxidation for sulfidic gold ores has made available operating and cost data which allow the derivation of conceptual costs for copper ore processing for comparison with conventional processing routes. Although costs for copper biooxidation generally remain high relative to conventional smelting, application might be competitive in specific cases. Biooxidation becomes more competitive in the general case if the cost of SQ:z emissions in pyrometallurgical processing were to be internalized to a greater degree than currrently required. This paper describes the technical aspects of biological processing of copper sulfide ores and concentrates, specifically with respect to the fate of sulfur. 1be environmental and economic implications of the biooxidation route are then discussed in relation to conventional processing. 4:10pm BEHAVIOR OF SULFUR DURING LEACHING OF SULFIDIC MINERAL: Thomas Havlik. Roland Kammel, Department of Non-ferrous Metallurgy, Faculty of Metallurgy, Technical University, Lelna 91A, 04385 Kosice, Slovak Republik It is proposed to realise the direct leaching, during which will be generated elemental sulfur and not sulfate or sulfuric acid as endproduct, under conditions that sulfur will not cover the leaching surface and not hinder the leaching course. 1be following improvements have been considered; use of organic sulfur solvents, ultra high-frequency leaching, ultrasonic leaching, use of surface detergents or surfactants in leaching processes, simultaneously crushing and leaching. 1bese proposals are under investigation and preliminary results are promising. 4:35 pm Invited SULFUR IN THE INTEC COPPER PROCESS: P K Everett, Intec Ply. Ltd, 21 Smith Street, Chatswood NSW 2067 Australia One of the major driving forces behind the development of the Intec Copper Process has been the increasingly stringent regulations governing SO:z emissions and, in some parts of the world, a declining H2S04 market. Other incentives include operating cost (92 US cflb Cu), capital cost (US$ 1,400/annuai tonne) and the ability to treat low grade and contaminated concentrates. With such advantageous features, widespread use of the process would lead to the production of large amounts of residues containing around 25 % elemental sulfur together with akaganeite (FeOOH) and the originai gangue. This paper briefly describes the process chemistry involved in the formation of elemental sulfur and the investigations to either extract it or otherwise tum the residue into a valuable by-product and minimise any adverse environmental effects. 1be process is being developed jointly by a group of 15 major Mining Companies.

TRANSIENT THERMAL PROCESSING OF MATERIALS SESSION V: Process Sensors, Equipment Issues & Manufacturing Perspective Sponsored by: EMPMD Thin FIlms & Interfaces Committee Program Organizers: N.M. Ravindra, New Jersey Institute of Technology, Newark, NJ; R.K. Singh, University of Florida, Gainesville, FL Wednesday, PM February 7, 1996

Room: Grand J Location: Anaheim Marriott Hotel

Session Chairman: S. J. Pearton, University of Florida, Gainesville, FL

2:00 pm Invited TECHNO·ECONOMICALASPECTS OF TRANSmNT THERMAL PROCESS· ING OF ELECTRONICS MATERIALS USING SINGLE·WAFER TECHNOL· OGY: Ahmad Kerrnani, CVC Products, Inc., Fremont, CA 94539 1be drive for higher density and superior performance memory and logic integrated circuits balanced with the economics of the device manufacturing process is the fundamental requirement for large-scale acceptance of single- wafer technology. Controlled thermal budget, shallow junction maintenance, reliable multi-level metallization scheme with reduced RC delay, and ultra· high density storage cells implemented on larger than 200 mm silicon wafers are among the primary technological drivers requiring introduction of new materials and processes. In this paper, the acceptance methodology of Transient 1bermal ProceSSing for both front- and back-end-of-the-line processing in integrated circuits manufacturing environment based upon various metrics of the manufacturing process namely yield, throughput, cleanliness, processing equipment utilization, and cost-of-ownership is reviewed. 2:30 pm Invited PROCESS MONITOR WAFERS FOR RAPID THERMAL PROCESSING: Terrence J. Riley, Arun K. Nanda, SEMATECH, Austin, TX 78741 SEMATECH has an ongoing effort in evaluating and benchmarking commercial Rapid 1bermal Processors (RTP). 1be primary method used to evaluate the thermal performance of an RTP is the use of monitor wafers. 1be methods include: rapid thermal anneal (RTA) monitor wafers, thermocouple instrumented wafers, rapid thermal oxidation (RTO) monitor wafers, and nSi anneal monitor wafers. 1be main subject to be covered is SEMATECH's use of ion implanted wafers for monitoring rapid thermal annealing, concentrating on source/drain anneals. A goal of this effort is to create a monitor wafer which is sensitive to the temperature of the anneal while being insensi· tive to other process conditions. This type of monitor allows for accurate assessment of the temperature uniformity and repeatability of the RTP. Both Boron and Arsenic are used as dopants for the ion implant. Combinations of various screen oxides, implant energies, and dopant doses are used to create significantly different monitors. Each wafer fabrication process results in a monitor which displays a unique temperature

B129

sensitivity at each RTA temperature. Thus, certain monitors display optimal sensitivity for a given anneal process. This presentation will discuss the method of fa~cation of monitor wafers which are sensitive to IITA processes and the range of operation of such monitors, based on experimental results. 3:00pm EMISSMTYMEASUREMENTS & MODELLING·AMATERIALS PERSPEC· T1VE: F.M. Tong, N. M. Ravindra, Department of Physics, New Jersey Institute of Technology, Newark, NJ 07102 Abstract not available. 3:10 pm BREAK

3:35 pm PANEL DISCUSSION· "FUTURE OF TRANSIENT PROCESSING TECHNOLOGIES"

THURSDAY AM ADVANCES IN THE SCIENCE AND TECHNOLOGY OF TITANIUM ALLOY PROCESSING· VII: Joining of Titanium Alloys Sponsored by: SMD Titanium Committee, MDMD Shaping and Forming Com· mlttee Program Organizers: Prof. Isaac Weiss, Prof. Raghavan Srinivasan, Mechanical and Materials Engineering Dept., Wright State University, Dayton, OH 45435; Dr. Paul Bania, TImet Corporation, TImet-Henderson Technical Laboratory, P.O. Box 2128, Henderson, NV 89009; Prof. Daniel Eylon, Graduate Materials Engineering, University of Dayton, Dayton, OH 45409 Thursday, AM February 8,1996

Room: B5-6 Location: Anaheim Convention Center

Session Chairpersons: G. Welsch, Case Western Reserve University, Cleveland, OH 44106; V.K. Sikka, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831-6083 8:30 am Invited JOINING OF ADVANCED TITANIUM·BASED MATERIALS· PRESENT CHALLENGES AND APPROACHES: W.A. Baeslack Ill, College of Engineering, The Ohio State University, Columbus, OH 43210; T.J. Kelly, GE Aircraft Engines, Cincinnati, OH 45215-6301 The effective application of advanced titanium-based materials such as gamma and orthorhombic titanium aluminides, metal-matrix composites and conventional alloys in aerospace and non-aerospace applications will require the development and application of joining processes, procedures, and postweld processing tech?iques that ~oduce high integrity (i.e. defect free) joints exhibiting acceptable mechaniCal prope~e~. The inherent metallurgical nature of these materials commonly precludes the application of conventional welding processes and/or procedures currently utilized in the joining of monolithic TI-6AI-4V, and has required new joining approaches. This presentation will describe current requirements for the joining of several advanced titanium-based materials and approaches utilized by investigat?~ in the USA. to meet these ~hall~n~es. Material/process combinations described wlllmciude the high-energy denSity, fncuon and diffusion welding of TI-48A1-2Nb-2Cr gamma titanium aluminide, the resistance and conventional diffusion welding and diffusion brazing of monolithic (neat) and SiCreinforced TI-20AI-23Nb orthorhombic titanium aluminide, the gas tungsten-arc and electron-beam welding of Alloy C and the gas tungsten-arc and laser welding ofTI-622-22Si. The dissimilar joining of these materials to titanium and ferrous-based materials using principally solid-state joining processes will also be presented. Approac.hes to joining process selection and process optimization via the development B?d ap~lIca­ tion of physical and welding metallurgy principles for each of these matenals will be described. 9:00am JOINING OF DISSIMILAR TITANIUM ALLOYS FOR DUAL ALLOY COM· PRESSOR STAGE CONSTRUCTION: Prabir Bhowal, James A. Hall, Allied Signal Engines, Phoenix, AZ 85072 The call for increased stage performance in the high pressure compressor of advanced gas turbine engines has led to increasing demands on the strength and temperature capabilities of the materials from which the disks in these rotating stages are constructed. The use of titanium alloys in these applications is met with a challenge that alloys and microstructures idealized for the high strength requirements of the bore material are divergent with the alloys and microstructures suitable for the higher temperature creep requirements at the outer, or rim regions of a disk or impeller. u.se of dual alloy impeller construction is one possible solution to the problem. Allied Signal work on the development of effective joining methods for dissimilar titanium alloys is discussed. Demonstration of effective joining methods for conventional alloy pairs in impeller configuration will be presented. Also, work in progress, on the joining of advanced alloy pairs incorporating TI3AI base a2, TI2AINb base a(orthorhombic) and TIAI base ')(gamma) combinations, will be discussed.

!he

9:15am A STUDY OF DIFFUSION BRAZING OF DISSIMILAR TITANIUM ALLOYS: Gopal Das, Carter Barone, Pratt & Whitney, PO Box 109600, West Palm Beach, FL 33410-9600 The diffusion brazing of dissimilar titanium alloys-a beta titanium alloy (trade name

B130

Alloy C) and a gamma titanium ,uuminide (trade name XD) was assessed using commercially available TIcuni filler metal. The effect of process parameters such as temperature, time as well as initial filler metal intedayer thickness on the microstructure of brazements have been evaluated. Post-braze heat treatments were employed to achieve homogeneity across the brazements both in terms of microstructure and chemical composition. Phase identification including composition gradient across the brazements for as-brazed and post-braze heat treated samples were achieved by TEM and EMPA studies. Microhardness technique wa,s used to study the effects of process parameters and post-braze heat treallnents on the mechanical properties of the brazements. In addition. three-point bend strength and tensile strength of brazed samples were determined at lIT and 1200°F. Attempts will be made to establish microstructure/property relationships and the results will be discussed in terms of present understanding of diffusion brazing of dissimilar materials. 9:30am INFRARED TRANSIENT LIQUID-PHASE JOINING OF SCS·6I821S TITA· NIUM MATRIX COMPOSITE: C.A. Blue, V.K. Sikka, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831-6083; R.A. Blue, R. Y. Lin, Department of Materials Science and Engineering, ML 12, University of Cincinnati, Cincinnati, OH 45221 Fiber reinforced titanium matrix composites (TMC) are one of the advanced materials being considered for use in the aerospace industry due to its light weight, high strength, and high modulus. In o~:ler to use TMC as structural components with complicated geometry, the successful development of joining techniques which provide strong bonds and do not weaken the base materials being joined is essential. A successful TMC joining method must be developed which will limit the reaction between the matrix and reinforcement during joining. This may be accomplished by minimizing the liquid pool during liquid state processes and minimizing time in the case of solid and liquid-solid state processing. In an effort to minimize processing time in transient-liquid-phase, solid-liquid state, joining of TMC, a rapid infrared joining (RU) technique has been developed. The application of the RU technique for joining a 16 ply SCS-6I 821S titanium matrix composite will be discussed. 9:45am TI75 ALLOY AND ITS WELDING BEHAVIOR: Zhao Yongqing, Li Zuochen, Li Changliang, Wu Qingzhi, Northwest Institute for Nonferrous Metal Research, P.O. BOX 51 Xian, China, 710016 In order to meet the requirement for new materials for marine, chemical industry and genetic engineering applications, the TI75 alloy, a new middle strength, high toughness and corrosion resistant near -a tita~ium alloy, was designed and developed by NIN. Its main properties are as follows:ai»730MPa, as>630MPa,1) > 13%, 1jI >30%, ak>6OOKJ m-2, KJO 2>IOOMPa m- l12 and KIscc>85MPa ml12 .. The alloy used in this experiment was hot rolled commercial plate of 14mm thickness. The results indicated that TI5 alloy had good welding behavior. The k value of welding seam changed with increasing distance between melting line and welding seam. In any conditions at was greater than 50010 m-2. Tensile properties of welding seam was as follows: ab ~ 750MPa, as ~ 650MPa, I) ~ 10%, 1jI > 25%. The microstructure of the weld and fractograph were examined by means of SEM and optical microscopy. 10:00 am EXPLOSION WELDED TRANSITION JOINTS FOR WELDS BETWEEN TI· TANIUM AND DISSIMILAR METALS: John G. Banker, Dynamic Materials Corp., Explosive Fabricators Div., 551 Aspen Ridge Drive, Lafayette, CO 80026 Explosion welded transition joints provide a means for making fully welded connections between titanium and dissim~ar metals such as aluminurns, stainless steels and nickel alloys. The manufacturing process is briefly discussed. Mechanical test data is presented on several metal combinations. Frequently thin intedayers of other metals are employed in the bonding operation to enhance product features, such as strength, toughness or leak tightness. Data on performance of specific intedayer metals is presented. Potential applications for both the aerospace and process industries are reviewed. 10:15 am TITANIUMISTEEL EXPLOSION BONDED CLAD FOR AUTOCLAVES AND REACTORS: John G. Banker, Dynamic Materials Corp., Explosive Fabricators Div., 551 Aspen Ridge Drive, Lafayette, CO 80026 TItanium provides excellent performance in many highly corrosive process industry environments. TItanium clad steel offers a cost effective means for manufacture of vessels for high pressure and high temperature applications. Explosion bonded titanium clad steel has been used exteRlively for reliable clad vessel fabrication for nearly 30 years. The explosion bonding process and characteristics of titanium explosion clad are discussed. TItanium clad reactor fabrication methods and experience are reviewed in detail. Specific applications of titanium explosion clad in hydrometallurgical autoclaves are discussed. 10:30 am LASER WELDING OF TIMETALLCB ALLOY: I. Weiss, N. Stefansson, M. Saqib, Mechanical and Materials Engineering Dept., Wright State University, Dayton, OH 45435; W.A. Baeslack Ill, College of Engineering, The Ohio State University, Columbus, OH 43210 Laser welding was used to obtain high integrity welds in TImetal LCB sheet. These welds are characterized by narrow fusion and heat affected zones, as well as finer microstructure. The objective of this paper is to evaluate the laser welding of 0.060 in. LCB sheet in the as-solution treated condition. The effect of heat treatment on weld microstructure and mechanical properties will also be discussed. 10:45 am BONDING OF PORCELAIN TO TITANIUM COMPONENTS FOR DENTAL RESTORATION: O.R. Monteiro, I. G. Brown, Lawrence Berkeley Laboratory, Berkeley, CA 94720; R.R. Wang, School of Dentistry, Case Western Reserve University,

1996 TMS ANNUAL MEETING AND EXHIBITION

J. Chung and G. Welsch, Materials Science & Engineering, Case Western Reserve University, Cleveland, OH 44106 . The bonding of porcelain to titanium components used in dental restorations requires heating to around IOOO·C. Direct application. of por~elain ~o titanium res~lts in poor adherence and spallation because of a weak mterfaclal ruttle layer. S~lmens were coated with thin layers of chromium and silicon using sputter and plasma I~r­ sion deposition methods. The purpose of the coatings is prevention of rutile formation. A chromium oxide or silica surface film is to be formed with good adherence to both the substrate and the porcelain overlayer. Results of microstructure and mechanical evaluations of titanium/coating/porcelain specimens will be presented. 11:00 am MICROSTRUCTURE EVOLUTION IN THE FUSION ZONE OFTI-48AL-2NB2CR GAMMA TITANIUMALUMINIDE WELDMENTS: J. S. Lee, W.A. Baeslack III, College of Engineering, The Ohio State University, Columbus, OH 43210;. T.J. Kelly, GE Aircraft Engines, Cincinnati, OH 45215 Achieving acceptable mechanical properties in gamma titanium aluminide weldments requires optimization of the weld fusion zone microstructure. Microstructure. e~olu­ tion in this weld region is complex due to the inherently complex nature of sohdlfication and on-cooling solid-state phase transformations in ~s alloy sy~tem and theyotentially wide range of solidification and cooling rates expenenced dunng arc and highenergy density fusion welding processes. This presentation will describe fusion zone solidification and on-cooling solid-state phase transformations for laser welds produced in a series ofTI-(45-49)AI-2Nb-2Cr alloys that experienced fusion zone cooling rates ranging from about 50 to 5000°C/s. Effects of aluminum content B?d wel~ cooling rate on the fusion zone solidification mode (BCC vs. HCP) and on-cooling solid-state phase transformations will be described. Effects of postweld heat treatment on the modification and normalization of the as-welded fusion zone microstructures will also be presented. 11:15 am INFRARED TRANSIENT-LIQUID-PHASE JOINING OF GAMMA TITANIUM ALUMINIDE: C.A. Blue, V.K. Sikka, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831-6083 . Ordered intennetallic compounds based on aluminides such ~ TI~A1, TIAI, and NI3.AI are emerging as attractive materials for high-temperature apphca~ons. The a~~ual 10dustrial use of these intermetallic alloys has been slow due to therr poor ductihty and problems of weld cracking while being joined into a complicated geometry. In this study, the reaction between titanium aluminide and a liquid titanium allo!, 70TI: 1.5Cu15Ni in wt%, was investigated in order to reveal the behavior ofTIAI dunng solidification. Issues of concern in this study include: (I) the solidification process of the melt upon termination of heating, and (2) the diffusion of Cu ~d Ni in~ the s.ubstan~e during processing. To accurately control the extent of reac~on, an mnovative raptd infrared joining (RU) technique was used. Infrared processmg can produce heating rates exceeding 100°C/sec up to the processing temperature. The cooling rate is also rapid due to it being a cold wall process. Such rapid rates o~ processi.ng decreases ~r eliminates the adverse effects associated with prolonged heating. In this study, expenments were conducted at various heating times in order to investigate to solidified zone microstructure melt affected zone thickness, and base material microstructure. The system investi~ated in this study is directly related to joining of gamma titanium aluminide with a titanium brazing alloy. Results show that the joint show strengths as high as 220 MPa can be achieved with minimal effects on the mechanical properties of the gamma titanium aluminide. 11:30 am THE RELATIONSHIP BETWEEN MICROSTRUCTURE AND CRACKING IN GAMMA TITANIUM ALUMINIDE WELDS: Mary C. Juhas, Edison Welding Institute, 1100 Kinnear Road, Columbus, OH 43212; William A. Baeslack, The Ohio State University, College of Engineering, 2070 Neil Ave., Columbus, OH 43210; ~sh L. Fraser, The Ohio State University, Department of Materials Science and Engmeering, 116W. 19th Avenue, Columbus, OH432JO Gamma titanium aluminides have recently emerged as strong candidates to supplant titanium and nickel-base superalloys in high temperature gas turbine engines. Solid state joining of titanium aluminides is often accompanied by cracking in the weld heat and deformation zone (HDZ) under certain conditions. The focus of this work was to determine the relationship between weld HDZ microstructures and cracking behavior. Preliminary pilot welds using the continuous drive friction process were frrst prepared in order to characterize the location and nature of cracking. The themtaJ profiles across the joint interface during welding were then measured by thermocoupling wel~ sarnple~. The temperature excursions derived from the thermocoupled weld were used 10 combInation with the parameters total plastic strain and strain rate to simulate vat?0us regi~ns of the HDZ. The corresponding microstructures were evaluated for relative crackmg behavior and deformation features using optical metallography and transmission electron microscopy (TEM). In addition to the HDZ simulations, friction welded samp~es were tensile tested at ambient and elevated (700°C) temperatures and compared WIth similarly tested base material. The fracture surfaces and underlying defo~tion microstructures were characterized as a function of temperature and deformation level. It was observed that the HDZ cracking occurred along gamma grain boundaries and lath boundaries as observed in the elevated temperature tensile tests. Ambient temperature fractures were transgranular in nature.

ADVANCES IN THE SCIENCE AND TECHNOLOGY OF TITANIUM ALLOY PROCESSING· VIII: Casting of Titanium Alloys

Sponsored by: SMD Titanium Committee, MDMD Shaping and Forming Com· mlttee Program Organizers: Prof. Isaac Weiss, Prof. Raghavan Srinivasan, Mechanical and Materials Engineering Dept., Wright State University, Dayton, OH 45435;. Dr. Paul Bania, TImet Corporation, TImet-Henderson Technical Laboratory, P.O. Box 2128, Henderson, NV 89009; Prof. Daniel Eylon, Graduate Materials Engineering, University of Dayton, Dayton, OH 45409 Thursday, AM February 8, 1996

Room: B8 Location: Anaheim Convention Center

Session Chairpersons: J. Ault, Precision Castparts Corp., Portland, OR 97206; D. E. Larsen, Howmet Corporation, Whitehall, MI 49461

8:30 am Invited THE CORRELATION OF CREEP RESISTANCE TO HEAT TREATED MICRO·

STRUCTURE IN INVESTMENT CAST GAMMA ALLOYS: D. E. Larsen, Howmet Corporation, Whitehall, MI 49461, D.Y. Sec and T.R. Bieler, Michigan State University, East Lansing, MI 48824 Cast gamma titanium aluminides are gaining acceptance as replacements for superalloy and steel components in aerospace, automotive and industrial applications. Components cast from these alloys can operate at temperatures up to 15000F at half the weight of the components they replace. Most applications of cast gamma components require good creep resistance to meet component life requirements. Four heat treatments were developed and applied to investment cast Ti-47AI-2Nb-2Mn (at%}+.8v%TIB2 XD"', TI-45AI-2Nb-2Mn(at%)+O.8v%TIB2 XD and TI-47AI-2Nb2Cr (at%) alloys in an effort to enhance creep properties with a decrease in heat treatment time compared to current practice. Results show that the creep resistance of the TI-47AI-2Nb-2Mn+O.8v%TIB2 XD and TI-47AI-2Nb-2Cr alloys can be significantly altered, up to 3X, through heat treatment. However, the creep resis~ct; of the TI45A1-2Nb-2Mn+O.8v%TIB2 XD alloy was virtually unaffected. The vanaUon, or lack of variation, in creep resistance with heat treatment can be explained by differences observed in the microstructures produced by the various heat treatments. Both creep resistance data and microstructural observations obtained using optiCal, SEM, EDS, and TEM techniques will be presented and discussed. 9:00am DEVELOPMENT OF CAST TI-llOO STRUCTURAL COMPONENTS FOR ADVANCED GAS TURBINE ENGINES: James Ault, Precision Caslparts Corp., Portland, OR 97206; Paul Allen, TItanium Metals Corp., Henderson, NV 89009; Robert Testin, Allison Engine Co., Indianapolis, IN 46206 Advanced gas turbine engines require more temperature capable lightweight structures to achieve aggressive performance goals. Allison Engine Company's AE2loo/ AE3OO71T406 family of engines features a TI-6AI-2Sn-4Zr-2Mo-.08Si (TI-6242S) diffuser/combustor case with a cast diffuser. Growth versions of this engine will require titanium alloy capability beyond that of TI-6242S, if the weight penalty of ~ nickelbase diffuser/combustor case is to be avoided. TIMEfAL 1100 (TI-Iloo) IS a new generation, high temperature titanium alloy offering improved te~ture ca~bility over TI-6242S. This paper reviews the development of a TI-lloo diffuser casting for evaluation in an Allison Engine Company AE30IXI Advanced T406 gas generator core. Casting and manufacturing issues, as well as mechani~~ properties of ~ast TI-Iloo relative to TI-6242S, are discussed. Foundry produceablhty and mechanICal property data suggest that TI-II 00 is readily castable and offers up to 100°F additio?aI temperature capability over TI-6242S. Engine durability testing of an all TI-lloo diffuser/combustor prototype in an AE 30 I X core will commence in early 1996.

9:20am SIMULATION OF MULTI COMPONENT EVAPORATION IN ELECTRON BEAM MELTING AND REFINING: A. Powell, J. Szekely, Massachusetts Institute of Technology, 77 Massachusetts Ave., Rm. 8-135, Cambridge, MA 02139;.J. V~Den Avyle, B. Damkroger, Liquid Metal Processing Laboratory, MS 1134, SandIa National Laboratory, PO Box, 5800, Albuquerque, NM 87185 Experimental results and a mathematical model are presented to describe differential evaporation rates in electron beam melting of titanium alloys containing aluminum and vanadium. Experiments characterized the evaporation rate of commercially pure titanium and vapor composition over titanium with up to 6% AI and 4.5% V content as a function of beam power, scan frequency and background pressure. This model is made up of a steady-state heat and mass transport model of a melting hearth and a model of transient thermal and flow behavior near the surface. Activity coefficients for aluminum and vanadium in titanium are roughly estimated by fitting model parameters to experimental results. Based on the ability to vary evaporation rate by 10-15% using scan frequency alone, we discuss the possibility of on-line composition control by means of intelligent manipulation of the electron beam. 9:40am RAPID SOLIDIFICATION PROCESSING OF TITANIUM ALLOYS: Mohit Sisodia, Department of Metallurgical Engineering Malaviya Regional Engineering College J aipur-17, India Paper lucidly explains and compares the major Rapid Solidification (RS) of TI-al-

1996 TMS ANNUAL MEETING AND EXHIBITION

B131

loys techniques currently available. These are classified in two broad categories i) drop formation (which includes all atomization techniques) ii) stabilization of liquid, steam or surface. Salient features of more than 25 techniques are discussed at length, with much emphasis on arc plasma and free jet, melt spinning. Later one analytically comprises wheel and crucible materials, process atmosphere, crucible design, heating methods process parameters and their relationsbip to melt composition are described. Practical solutions to Ibis process problems are also analysed. Besides it, guidelines for selection of alloying additives along with a new approach of its processing in the devitification of TItanium rich glasses are introduced. Apart from the microstructure of RS TI alloys its preliminary results on the age-hardening response and thermal stability are critically reviewed. 10:00 am RECENT PROCESS DEVELOPMENT OF TITANIUM SPONGE PRODUC· TION IN INDIA: Mohit Sisodia, Department of Metallurgical Engineering, Malaviya Regional Engineering College Jaipur-17, India This review contemplates the recent advances and innovations in titanium technology in India. TItanium and its alloys have emerged as the most important non-ferrous metal and finds extensive applications in the field of aerospace and general engineering due to its physical properties. Its pure form is prepared by TItanium tetrachloride which is obtained by chlorination of oxide followed by reduction in presence of liquid Mgt Na. The process is complex and is restricted to a few countries like U.S.A., France, China, etc., In fact, India has a few rich deposits of Ibis metal in eastern and southern beach sand region along with the production facilities. The present paper discusses the statistically analyses of the efforts of Research and Development with emphasis on the current status of commercial production. 10:20 am BREAK 10:30 am CASTING TECHNOLOGY FOR GAMMA TITANIUMALUMINIDE VALVES: D. Ey/on, P. Jones, M. Keller, WJ. Porter Ill, Graduate Materials Engineering, University of Dayton, Dayton, OH 45409 Gamma titanium a1uminide alloys are considered for automotive applications in exhaust valves. The use of such valves has already been demonstrated in automobile engines. The challenge at the moment is to develop a cost effective mass production method using casting technology. This paper will discuss technologies investigated in the EMTEC program. 10:50 am THERMOHYDROGEN TREATMENT OF SHAPE CAST TITANIUM ALLOYS: A.A.llyin, V.K. Nosov, A.M. Mamonov, V.N. Uvarov, A.V. Alexandrova, Moscow State University of Aviation Technology, K. Tsilkovsky, Metals Science Dep., Petrovka St 27, K-31, Moscow, Russia, 103767 Shape casting is one of the most economical and technologically easy processing methods for the production titanium alloy components. However, titanium shape casting application in an industrial scale is limited by low level of their mechanical properties, especially fatigue. We developed new technological methods of cardinal improvement of structure and casting porosity limitation for TI-6A1-4V (VT6) and high-strength TI-5.5AI-4V-2Mo-l Cr-O.5Fe (VT23) alloys. This method is based on combinations of HIP and thermohydrogen treatment (11IT), they are directed on fatigue properties of shape castings improvement The results of Ibis testing shown, that suggested treatment strongly increase the cyclic endurance of both alloys. Thus, the number of cycles before fracture for VT6 samples at 500MPa loading increased in more than two times, in comparison with as-cast conditions and secure about 105 cycles. The endurance level for VT23 alloy on 107 base grew from 300 MPa in as-cast condition to 580 MPa after HIP and TIlT, and it exceeded the level of properties, achieved after strengthening heat treatment (quenching + ageing) of deformed semiproducts. The best fatigue resistance is ensured at the stressed not higher 0.6-0.8 of the ultimate tensile strength. In Ibis suggested technology the castings sudace oxidation is eliminated, the rates of heating and cooling are limited to 1-3 Kls. This is very important for near net shape castings of complicated configuration. 11:10 am HYDROGEN TECHNOLOGY OF TITANIUM ALLOYS ON TlJAI BASE: A.A. Ilyin, V.K. Nosov, A.M. Mamonov, V.N. Uvarov, A.V. Alexandrova, Moscow State University of Aviation Technology, K. Tsilkovsky, Metals Science Dep. Petrovka St. 27, K-31, Moscow, Russia, 103767 In this work the effect of hydrogen on the ductility and resistance to deformation of the three titanium intermetallic (TI3A1) based alloys in system TI-A1-Mo-Zr, TI-A1-Nbv-Zr and TI-A1-Nb-Mo-Zr with different content of 8-stabilizing elements was investigated. Deformation temperature and hydrogen concentration in alloys were chosen according to the alloy-hydrogen equilibrium phase diagrams. Drop forging and rolling of the TI3AI base alloys were carried out in service conditions with the use of hydrogen plastification, that gives a possibility to deformation conditions applied at a-and (a+8)heat-resistant titanium alloys - T=850-950°C, instead of T=1l00-1250oC for alloys which are not hydrogen alloyed. The hydrogen plastification using not only increases the alloys plasticity but together with therrnohydrogen treatment (TIlT) allows to increase their ductility at room and working temperature. Extruded with the use of hydrogen plastification process followed by thermo-hydrogen treatment compressor blades made of TI-A1-Nb-v-Zr alloy showed the following properties at room temperature: tensile strength 1330 MPa, yield strength 1196 MPa, elongation 3.3%, contraction 2.8%. At a temperature of 650°C tensile strength 885 MPa, yield strength 802 MPa, elongation 33.6%, contraction 53.5% were achieved. 11:30 am FOUNDRY INVOLVEMENT IN ALLOWABLES: Dale L McLellan, McLellan & Associates, Bellevue, WA 98008 Founders are faced with various obstacles when marketing their products for aerospace structural applications. They must convince airframers that casting offer cost and

B132

weight benefits over traditionally fabricated and machined parts. Even before a candidate part is manufactured assurances must be made for dimensional capabilities and control of tolerances. Consideration must also be given to the subject of mechanical properties suitable for design. Such properties, referred to as allowables, have always been established by airframers since they are responsible for certification of all structures to regulatory agencies. llitis paper concludes that statistical development of a1lowables for any material, should preclude the necessity of applying any other reduction factors. Otherwise, there would be no reason for the cost and time required to obtain such properties. FurthemlOre, founders should be involved in determining the a1lowables as they pertain to the development of new castings.

ALUMINA & BAUXITE TECHNOLOGY VI: Process II

Sponsored by: LMD A1umlnW11I Committee Program Organizer: C.G. Hendriks, Kaiser Aluminum and Chemical Corp., 9131 Interline Drive, Baton Rouge, LA 70809 Thursday, AM February 8, 1996

Room:A6 Location: Anaheim Convention Center

Session Chairman: Gerald Keltoll, Krebbs Engineers, 1205 Chrysler Dr., Menol Park, CA 95025-1178

8:30am THE FUTURE OF ALUMINAIBAUXITE MINING IN GUINEA: New Ap· proaches Towl!I'd Growth and Competitiveness: Siala Coulibaly, Department Inspector Ministry of Mines and Geology, PO Box 295, Conakry, GUINEA Guinea is one of the richest countries in Africa in view of its bauxite mining and energy potential. It has nearly olle-third of the world's bauxite reserves and it is the second world producer of this mineral, coming after Australia. The mining sector plays a key role in the Guinea economy. The state will continue to depend highly on Ibis sector fiscal revenue and foreign exchange for the coming years at least. Guinea is currently facing serious economic problems. Some of these problems are the result of a hostile external environment over which it has little or no influence, let alone control. It is therefore a prime importance that government should adopt effective policies with regard to management and making viable regulation to govern the alurninalbauxite industry, as well as acquire strong capacity to implement them. Since the market of these commodities remains extremely depressed, the Guinea a1uminalbauxite industry is subject to decline. In order to BlTest Ibis situation, the country wants to embark on a path toward diversifying its mineral industry and supporting efforts for global economic liberalization, gond communication, transport, energy and agriculture. This paper will survey the basic parametl~rs such as resources and competitiveness, Guinea's prospective in terms of efficiency technology equipment rehabilitation, environmental consideration. The article will emphasize on Guinea position in the market, the problem of pursing the production tools, the matters of competitiveness related to quality of bauxite, production cost reductiorl, fiscal regime, market arrangements, increase production capacities by opening new bauxite mines. 8:50am OPERATION OF AZ4 MTPD SOLID·LlQUID CALCINER (SLC) FOR LIQUOR PURIFICATION IN A BAYER PLANT: lens Fenger. EL. Smidth & Co. AIS, Research & Development Departmellt, Vigerslev AIle 77, DK 2500 Valby Copenhagen, Denmark; Joergen I1kjaer, FLS Minerals, Alumina & Bauxite Technology, Liam Fleming, Aughinish Alumina Limited; Jose IPulpeiro, Alumina Espanols, S.A.; Andy Carruthers, Alcan International Limited as Minerals, in association with Aughinish Alumina, Alumina Espanola, andAican International, has developed a liquor purification system for alumina refineries. Laboratory work has been carried out to characterize the behavior of different feed slurries, and several tests have been carried out on a 24 MTPD pilot plant at the FLS Research Center in Denmark. This paper reports the results of the laboratory work and the operation of the pilot unit. Examples of start-up and shut-down sequences are described. 9:10am USING OF PETROLEUM COKE DUST AS A PARTIAL SUBSTITUTE FOR NATURAL GAS AND FUEL OIL IN THE ALUMINA TRIHYDRATE CALCI· NATION PROCESS: B. Desgrosseilliers, l. Perron, Alcan International Ltd., PO Box 1250, 1955 Blvd. Mellon, Jonquiere (Qc.), Canada G7S 4K8; Pierre Laurier, Odette Menard, Jocelyn Martin, Alcan Smelters & Chemicals Ltd., Jonquiere (Qc.), Canada G7S4L2 In the Saguenay region, A1can operates two rotary kilns for calcining petroleum coke in order to provide the pre-bake anode and Soderberg paste production plants with calcined coke. The annnal production is about 220,000 metric tons. For several years, under-calcined dust has been escaping from the kilns with the hot gases; this undesirable product amounts to about 16,0:)0 mt per year. A maximum of 6,000 mt of dust has been recycled in the fabrication process. In the search for different recycling or sale possibilities for the remaining 10,000 mt, there arose the possibility of using this coke dust as a source of calorific energy in other processes and more particularly in the calcination of alumina trihydrate. A combustion feasibility study and a plant trial demonstrated the benefit of burning thill dust in two calcination units with a capacity of 60 mt per hour. In effect, an appreciable economy of natural gas or fuel oil as well as increased energy efficiency (gain of 6%) was observed in the calcination units. Moreover, the effect on the quality of t1H~ metallurgical alumina is negligible. These results may be explained in part by the use of solid fuel for which the combustion time is of the order of 1000 times greater than dlat of natural gas. The combustion zones are thus elongated, favoring the transfer of heat towards the solid alumina during calcination.

1996 TMS ANNUAL MEETING AND EXHIBITION

9:30am ALCAN ESP DUST MICRO-GRANULATION TECHNOLOGY: Kwa/ The, Claude Boulanger and Leopold Tremblay, Alcan International Limited, Arvida Research and Development Centre, Jonquiere (Qc), Canada; Robert Clegg, Stephan Girouard, Societe d'electrolyse et de chimie Alcan, Vaudreil Works, Jonquiere (Qc), Canada; Andrew Carruthers, Alcan International Limited, Engineering and Technology Group, Montreal (Qc), Canada Stringent pollution regulations and the increasingly competitive nature of the market demand that fines generated during the alumina calcination operation must all be collected and used in an environmentally and economically acceptable manner. Alcan has successfully developed a process for recycling ESP dust to the potrooms. This paper describes the novel Alcan ESP dust micro-granulation technology, its concept and pr0cess, as well as fundamental chemical reaction involved. Results of test on the production of ESP dust micro-granules on a semi-industrial scale and on the quality of these micro-granules with respect to calcination and electrolysis will be presented and discussed. 9:50am APPLICATION OF HYDROCYCLONE IN BAYER CIRCUIT: S. K. Patnaik, R. Brahma, P. Das, National Aluminum Co. Ltd., Darnanjodi-763 008, Koraput Dist., Orissa, India Due to its operational simplicity, effectiveness, low space requirement and maintenance cost, hydrocyclones of different sizes and material of construction are progressively finding more and more application in Bayer circuit for producing alumina from bauxite. In alumina refining process, hydrocyclones are most effectively used in the areas like closed circuit grinding, sand separation, sand washing, removal of residual solid from spent liquor and second stage filtrate to enbance plant productivity and classification of hydrate particles for product and seed granulometry control. The paper described the main features of the hydrocyclones installed at different units, its operational experience and performance achieved during the last nine years of plant operation, paper also covers the role of spent liquor solid separation cyclones a project under implementation, for augmenting productivity and plant capacity. 10:10 am BREAK

10:30 am IMPROVED FLOW CHARACTERISTICS OF BAUXITE BY TREATMENT WITH BAUXITE HANDLING CHARACTERISTIC MODIFIER: Carlos Cardounel, Reynolds Metals Company, PO Box 9911, Corpus Christi, TX 78469; Kevin O'Brien, NALCO Chemical Co., 13333 Blanco Rd.,Suite 316, San Antonio, TX 78216 Sherwin Plant bauxite handling facilities are capable of processing relatively free flowing dry bauxites. The processing of wet bauxite through these facilities was challenging and required the use of back-hoes and bulldozers. The use of a bauxite handling characteristic modifier to enhance flow properties of wet bauxite was investigated in the laboratory and subsequently tested in the plant. A significant improvement in flow characteristics was achieved. The bauxite became free flowing, eliminating the need for back-hoes and bulldozers to reclaim the bauxite. This paper relates how laboratory testing led to successful plant application of the bauxite handling characteristic modifier. 10:50 am OPTION OF BAYER DIGESTION TECHNOLOGY AND EQUIPMENT FOR CIDNESE BAUXITE: Ma Chao-jian, TIan Xin-jiu, Shenyang Aluminum and Magnesium Engineering and Research Institute, 184 Hepingbei Street, Shenyang LN 110001, Liaoning, China China is rich in bauxite with a proven reserve of more than two billion tones, which almost is the diaspore with complex and various minerals. To treat this kind of ore the Bayer digestion possesses it's own characteristic because of the high temperature and caustic concentration. Based on the bebavior of silicon and titanium minerals digested in the lab and plant, the digestion technology and equipment for Chinese bauxite be discussed; and based on the Chinese construction condition, the economic comparison of several typical and industrial digestion devices be estimated in the paper. As a conclusion, a proper digestion technology and device for Chinese bauxite be proposed. 11:10 am ALUMINA PRODUCTION IN RUSSIA: AN OVERVIEW: Dr. V. S. Smirnov, Chief Consultant, Alumina, AO VAM!, 86, Sredni pr, St. Petersburg, 199026, Russia Alumina production in Russia: heritage of the system relying on its internal resources. History in brief of establishing Russia's alumina plants: a pre-war period - constructing new facilities, commissioning new capabilities during the second world war retnrning old capacities into service after the war and expansion of the plants built in the following years. Introduction of new local raw materials and construction of major processing facilities. List of Russia's alumina plants after a collapse of the Soviet Union and formation of the CIS. Data covering the plants including production capacities,locations, power supply, a source and quality of a raw material processed. Schematic f1owsheets, size of major types of equipment. Difference of bauxite processing parameters and of the process employed between Russia's plants and Western plants. Innovations introduced during the plants' history. Switch to a market economy and new challenges of survival in a competitive world. The main ways of cutting production costs at the plants. Forecast for an alumina production in Russia.

11:30 am DRESSING-BAYER PROCESS FOR CHINESE BAUXITE TO PRODUCEALUMINA: Wang En-fu, Ma Chao-jian, Shenyang Aluminum and Magnesium, Engineering and Research Institute, Shenyang, Liaoning, China China is rich in bauxite with a proven reserve of more than two billion tones, which almost is the diaspore with high alumina and silica (AI OlSiO of weight: 4.5-7) and can not be economically processed with Bayer process. Since 1950's the Combination and Sinter Process, with high recovery and low soda consumption, have been used in Chinese alumina industry. However, both process need extremely high energy at 2.5 even

1996 TMS ANNUAL MEETING AND EXHIBITION

4 times as that of Bayer process. With the inflation of energy cost, to research new process for this bauxite gradually become a strategic task to develop Chinese aluminum industry. Based on the effects of SiO in Bayer process, the dressing process for the high-silica bauxite, Dressing-Bayer Process and their economic estimation have been described in the paper. As a conclusion, the Dressing-Bayer Process has more advantages than Combination and Sinter Process, which would be widely used in China. 11:50 am TECHNOLOGICAL FEATURES OF THE PROCESSING OF THE ALUMOSILICATE RAW MATERIAL WITH POTASSIUM'S HIGH CONTAIN: V. A. Upin, N. N. Ttkhonov, VAM!, 86 Sredny Dr. LTG, St. Petersburg 199025 Russia The deposits of alumosilicate raw material with high portion of potassium alkali are revealed in Russia, China, Pakistan, Northern Korea, Iran, Venezuela and other. These ores are studied with the purpose of the industrial processing by the sintering method for production of alumina, cement, potassium compounds, some rare elements. This process has been in use at the Russian alumina plants. The equations of the dependence of technological indexes from chemical composition of processing area were made on the basis of numerous experimental data using the original programs. This enabled us to choose the most effective conditions during our research on nepheline raw materials in the main technological operations.

ALUMINIUM REDUCTION TECHNOLOGY VII: Cell Dynamics And Process Control Sponsored by: LMD Aluminum Committee Program Organizer: Ms FionaJ Stevens, Comalco Research and Technology, Comalco Research Centre, PO Box 316, Thomastown, Victoria 3074, Australia Thursday, AM February 8, 1996

Room:A9 Location: Anaheim Convention Center

Session Chairperson: Jennifer Purdie, Boyne Smelters Ltd, PO Box 524, Boyne Island, Gladstone, Queensland 4680, Australia

8:30am ON THE MECHANISMS OF ALUMINA DISSOLUTION WITH RELEVANS TO POINT FEEDING ALUMINIUM CELL: Ove Kobbeltved/, Jomar Thonstad Department of Electrochemistry, Norwegian Institute of Technology, N-7034 Trondheirn, Norway; Sverre Rolseth, SINTEF Materials Technology, N7036 Trondheirn. Norway The dissolution of alumina in cryolite melts was studied in a laboratory apparatus intended to simulate the conditions in industrial cells with point feeding. Rapid dissolution is achieved if (i) a large proportion of the batch of alumina added is dispersed in the bath upon addition, and (ii) the agglomerates which form, disintegrate quickly. Bubbling of gas through the melt increased the dissolution rate by promoting the dispersion of alumina grains as well as breaking up the agglomerates. Preheating of the alumina also resulted in faster dissolution, mainly due to effective dispersion of the alumina grains. High superheat of the bath enhanced the dissolution rate of the agglomerates. Measurements on a point fed industrial cell showed that the dissolution rate of the alumina decreased when the feeding hole was plugged compared to the situation when the feeding hole stayed open. 9:00am TRANSPORT OF DISSOLVED ALUMINA IN POINT-FED SODERBERG REDUCTION CELLS: K. Torklep, K. Kalgraf, T. Nordbo, Elkem a1s Research, PO Box 40, Vagsbygd, N-4602 Kristiansand, Norway Knowledge on how dissolved alumina is distributed in the interpolar region is required to determine the optimum number of point feeders and their location, and the particular demands of Soderbergs make a straightforward application of experiences obtained with prebak~ technology suspect. We have studied the transport of a radioactive, isotopic tracer ( 4NaCI ) injected molten in the bath in gram amounts at existing and proposed point feeder locations. A system of air-cooled Geiger-Muller detectors placed in the pot permitted continuous monitoring of the relative tracer concentration at six different locations. Convection velocities, patterns, turbulent diffusivities, and mixing times were obtained. Results are used in two-dimensional modelling of the oxide concentration in the bath. 9:30am SOME PITFALLS IN USING CELL RESISTANCE TO CONTROL ALUMINA CONCENTRATION: Richard Brown, Collin MacPherson Alcan Australia Ltd - Kurri Kurri Smelter, PO Box I, Kurd Kurri, NSW 2327, Australia Modem aluminium cells use sophisticated computer control algorithms to control the alumina concentration of the cell. The alumina concentration is inferred from the bath resistance as it cannot be measured directly. Stable and efficient cell feeding is dependent on the accurate interpretation of the cell resistance changes over time. This paper describes how, in certain conditions, mobile sludge can mask the true bath alumina concentrations. Under these conditions it is possible for the alumina feed control algorithm to "overfeed" the cell. This situation can have a negative effect on cell stability and cell operation maldng good alumina control more difficult than is often portrayed. 10:00 am THERMAL EFFECTS BY ANODE CHANGING IN PREBAKE REDUCTION CELLS: Frank Anune, Institute of Inorganic Chemistry, University ofTrondheim, N7034 Trondheim-NTH, Norway; Marvin Bugge, Norsk Hydro a.s, Research Centre, N3901 Porsgrunn, Norway; Halvor Kvande, Trygve Ringstad, Hydro Aluminium a.s, Karm0y Plant, N-4265 Havik, Norway; Sverre Rolse/h, SINTEF Materials Technol-

B133

ogy, N-7034 Trondheim-NTIl, Norway Thermal effects caused by anode changing in prebake alumina reduction cells were studied by temperature measurements using thermocouples located at different positions inside the anodes. Anode changing caused a temperature reduction in the nearest neighbour anode of typically 20 to 4O'C, as measured close to its working surface. The minimum temperature was recorded about 5 hours after the anode changing, and it could take up to 24 hours before the old anode again showed a stable temperature reading. The cooling effect is due to heat radiation losses to the new anode, and thus the heat balance of the neighbour anodes is altered significantly after an anode changing. Temperatures were also measured inside the newly set anode. If the new anode voluntarily was set too low by 20 mm to provoke a formation of a spike on the working surface, the measured temperatures inside this anode could reach as high as 1080°C. 10:25 am BREAK 10:35 am BATH AND LIQUIDUS TEMPERATURE SENSOR: Paul Verstreken, Heraeus, Electronite Int. N. V. Grote Baan 27a, 3530 Houthalen, Belguim; Siegfried Benninghoff, Hoogovens Aluminium Huttenwerk GmbH, ScheusensstraBe, 46562 Voerde, Germany A newly developed sensor which measures both bath and liquidus temperature is described. The difference between both is known as superheat. Determination of the liquidus temperature is based on interpretation of the time-temperature trace during cooling down of a sample taken from the bath. Supercooling is avoided by vibrating the sample. Accuracy and reproducibility of the sensor is discussed. Examples of measurements and series of measurements in point-feed and center-worked cells are given. Evolution of bath and liquidus temperature is shown on short time bases, with respect to feed strategies and on longer time bases. To get better information out of temperature measurements, samples taken, or liquidus measurements performed, it is necessary to relate to timing of the feed strategy in use. Whether analyzing bath samples can be replaced by superheat measurements for cell control purposes will follow from ongoing experiments. 11:00 am FURTHER DEVELOPMENT OF THE TEMPERATURE MODEL: Peter M. Enmer, Alusuisse Lonza Services Ltd, Research and Process Development, 3965 Chippis, Switzerland; Gunnar A. Gudmundsson, Icelandic Aluminum Company Ltd., Straumsvik, PO Box 244, 222 Hafnarfordur, Iceland At TMS in 1995 a model was presented to control the bath temperature of electrolytic pots. Regression methods calculate the parameters of model equations using historical data. Optimal set values of the pot voltage and aluminum fluoride feed are then determined. A pot may be in the reactive or inactive pot state. In the reactive state it responds readily to changes, e.g. an increase AlF3 feed causing an increased AlF3 concentration.1n the inactive state however the bath temperature or the AlF3 concentration for instance may stay constant even when pot operation makes strong efforts to change the pot state. Especially interesting is the transition from one state to the other. To predict the moment of transition the temperature model uses indicators, derived from the time behaviour of pot parameters like metal height, bath height or alumina feeding. With the knowledge of a coming transition the model adjusts the optimal pot parameters. A mechanism is presented to explain the reactive and inactive pot states and the transitions. 11:30 am PROCESS SIMULATION OF ALUMINUM REDUCTION CELLS: [mad Tabsh, CompuSIM Inc. 1003055 Avenue N.E. Calgary, Alberta, Canada TIE 6WI; Marc Dupuis, GeniSim, 3111 Alger, Jonquiere, ~bec, Canada G7S 2M9; Alexander Gomes, Alcan Aluminio do Brasil A program was developed to model the dynamic behavior of an aluminum reduction cells. The program simulates the physical process by solving the heat and mass balance equations that characterize the behavior of eleven chemical species in the system. It also models operational events (such as metal tapping, anode change, etc.) and the process control logic including various alumina feeding policies and anode effect quenching. The program is a PC based Windows applications that takes full advantage of the Windows user interface. This paper describes the implementation of the process model and the control logic. Various results using the simulation are also presented.

CAST SHOP TECHNOLOGY VII: Metal Treatment-Fluxing

Sponsored by: LMD Aluminum Committee Program Organizer: Julian V. Copenhaver, TechnicallQuality Manager, NSA - A Division of Southwire, PO Box 500, Hawesville, KY 42348

Thursday, AM February 8,1996

Room:AIO Location: Anaheim Convention Center

Session Chairperson: Mr. Ray T. Richter, Aluminum Company of America, Alcoa Laboratories, Alcoa Center, PA 15069

8:30am A COMPUTER MODEL FOR ALKALI REMOVAL FROM MOLTEN ALUMINIUM: Dr. Christian J. Simensen, SINTEF Materials Technology, PO Box 124, Blindern, N-0314 Oslo, Norway; Dr. Madhu Nilmani, University of Melbourne, Department of Chemical Engineering, Parkville, Victoria 3052 Australia A computer model has been developed for alkali removal from molten aluminium by reactive gas purging. The model takes into account the bubble size distribution and the bubble expansion as it ascends through the melt. Both bath furnace fluxing and

8134

fluxing in continuous series of in-line units are modeled. It was found that the following average bubble diameter could be introduced in the calculations:

_ d = (6

n

6. ) 41/7 r. ~ 1.75 I

d io

where diO is the initial diameter of the bubble in size class i and Gn and Gni are the total amount of gas entering the refinin,g unit per second and the amount of gas in size class i respectively. The diffusion coefficients of alkali metals were determined by holding aluminium melts and measuring ilhe decay of the alkali metals, in addition to using published values. The computer model was validated using oxygen desorption from water and measuring gas bubble distribution simultaneously. It was then applied to alkali removal from a 4.2-tonne h,~lding fumace. The experimentally determined values and the computed values from the model for the effective average bubble diameter are in reasonable agreement. 8:50am THE ALCAN COMPACT DEGASSER - A TROUGH-BASED ALUMINUM TREATMENT PROCESS. Part I: Metallurgical Principles and Performance: Peter Waite, Alcan International Limited, Arvida Research and Development Centre, 1955 Mellon Boulevard, PO Box 1250, Jonquiere, Quebec, Canada G7S 4K8; Roger Thiffault, Alcan Smelters and Chemicals Ltd., Grande Baie Works, La Baie, Quebec, Canada In-line degassing processes, employing rotary gas injectors, are used extensively in the aluminum industry for removing unwanted impurities from the liquid metal just prior to casting. Although these pmcesses are efficient, and provide the required level of metal treatment, a significant quantity of between I and 3 tons of metal is retained within the degassing unit between casts. Metal retention is a serious drawback for multi-alloy cast shops, due to the scrap metal produced by draining the degassing unit at alloy changes. The Alcan Comp.act Degasser is a new multi-stage in-line degassing process which treats the liquid meM directly in the trough. FollOwing the initial development period in the laboratory, two years of process optimization in the plant has proven that the Alcan Compact Del:asser provides metal treatment efficiencies equivalent to or better than existing technologies, while eliminating metal retention between casts. The metallurgical principles of this process are described, and quantitative plant data are presented which Charactelrize the metallurgical performance with respect to degassing, metal cleanliness, and alkali removal. 9:10am TIlE ALCAN COMPACT DEGASSER - A TROUGH-BASED ALUMINUM TREATMENT PROCESS. Part II: Equipment Description and Plant Experience: Serge Lavoie, Alcan International Limited, Arvida Research and Development <;entre, 1955 Mellon Boulevard, PO Box 1250, Jonquiere, Quebec, Canada G7S 4K8; Eric Pilote, Marc-Andre Thibault, STAS, 1846 Outarde, Chicoutimi, Quebec, Canada G7K !HI; Jean-Claude Pomerieau. Alcan Wire & Cable, Lapointe Works, Jonquiere, Quebec, Canada This Alcan Compact Degasser is an efficient trough-based molten metal treatment unit designed as an alternative to the existing in-iine "heated-box"-based degassers. It was shown in Part I of this paper that in spite of the Alcan Compact Degasser's small size, no concessions were made on its metallurgical performances. The Alcan Compact Degasser is installed directly in the casting trough, which eliminates the metal hold-up between casts. This has several significant advantages. Firstly, no heating system is required, which reduces the equipment and maintenance cost of the machine and simplifies it operation. as compared to the heated-box alternatives. Secondly, the floor space requirement is reduced, making it easier to retrofit in existing plants. F'mally, the absence of hold-up eliminates a si~:nificant source of metal loss when frequent alloy changes are required. Another key feature of the Alcan Compact Degasser is it modularity. This allows the casting flow Irate capability to be accurately adapted to the need and assures extremely good reliability. Various versions of the Alcan Compact Degasser have been used in Alcan for the last two years. This paper firstly describes the industrial unit and its main characteristics. The operational experience with the Alcan Compact Degasser in two Alcan plants is then presented. 9:30am HYDROGEN AND INCLUSION CONTENT IN RECYCLED ALUMINIUM AT HOLMESTRAND ROLLING MILL: Dr. Frede Frisvold, Eivind 0vrelid, SINTEF Materials Technology, N-7034 Trondheim, Norway; Per Bakke, Norsk Hydro a.s., Research Centre, PO Box 2560, N-3901 Porsgrunn, Norway; Niels Erik Hald, Thorvald A. Engh, The Norwegian Institute of Technology, Division of Metallurgy, N-7034 Trondheim, Norway Recycling of aluminium by remelting of lacquered scrap in Norway is only carried out at Hydro Aluminium's Holmestrand Rolling Mill plant. A study has been carried out in order to determine if recycled metal contains higher than normal contents of carbide inclusions and dissolved hydrogen. The inclusion number size distribution was determined using filter samplers and employing automatic image analysis. An exponentially decreasing number size distribution was obtained. It was found that carbide inclusion contents were at the same levels as from primary producers. Dissolved hydrogen - before degassing - was, on the other hand, found to be slightly higher than in primary aluminium. Metal refining by a two-stage Hycast degassing unit ensured that both inclusion and hydrogen content were at acceptably low levels befure casting. 9:50am ALLOYING OF LIQUID ALUMINUM: A. Shafyei, R.I.L. Guthrie, Department of Mining and Metallurgical Engineering, McGill University, 3450 University Street, Montreal, Quebec, Canada, H3A 2A 7 The addition of high melting point alloying elements is an essential step in the processing of molten aluminum. In this research, the kinetics of dissolution of Mn and Fe particles, freely dispersed in turbultmt-liquid aluminum, has been investigated on a

1996 TMS ANNUAL MEETING AND EXHIBITION

laboratory scale. First, the suspension behaviour of alloying elements in liquid aluminum was studied through water modelling experiments as well as dimensional analysis and similarity techniques. Second, mass transfer coefficients from solid Mn and Fe particles into stirred liquid aluminum were measured. Results of high temperature experiments showed that the measured mass transfer coefficients of Mn and Fe particles are strongly dependent on the intensity of the mixing, until the particles became fully suspended. Further increases in the rate of mixing produced negligible changes in mass Iransfer coefficients. Therefore, from a practical point of view, a very high rate of mixing of the melt is not recommended during alloying process. 10:10 am BREAK 10:20 am EXPERIMENTAL DETERMINATION OF BUBBLE SIZES IN MELT REFINING REACTORS: Stein T. Johansen, S. Gn\dahI, 0. Dahle, I.-R. Johansen, SIN1EF Materials Technology, N-7034 Trondheim-NTH, Norway; R. Tangen, MESTECH as, Fyrstikkalleen 21, 0661 Oslo, Norway; Erling Myrbostad, Hydro Aluminium, Metallurgical R&D Cenlre, N-6601, SunndalSiM1l, Norway An intrusive optical probe for experimental determination of bubble sizes in water models has been designed and tested. The system is using automatic image processing and gives information about bubble size distribution and bubble shapes. Using a standard in-line melt refining rotor, the distribution of bubble sizes has been determined. The paper will discuss the experimental findings for this system. The bubbles close to the rotor are very small and grow due to coalescence towards te reactor surface. The effects of rotor frequency, gas purging f10wrate and melt flow-rate will be discussed. Removal rates of dissolved oxygen will be presented and theoretical mass Iransfer theories will be discussed. Finally, the accuracy of this new bubble probe, compared to resistance probes, will be addressed. 10:40 am IMPROVEMENT OF IN-LINE MELT CLEANING SYSTEMS AT HOOGOVENS ALUMINIUM NV - SIDAL: Jan M. Rabenberg, R. Boterman H.B.M. Schulte, Corporate Research Laboratorium, Hoogovens Groep BV, Umuiden, The Netherlands; D. Van Nieuwerburgh, Hoogovens Aluminium NY, (A Company of the Hoogovens Groep), A. Stocletlaan 87, B-2570 Duffel, Belgium The cast shop of Hoogovens Aluminium NV - Sidal in Duffel, Belgium, annually produces 180,000 tonnes of slab and 30,000 tonnes of billets. The casting rate of the large EMC unit goes up to 40 tonnes per hour. A proprietary in-line system, called Sidafil, cleans the melt by means of a purge gas, diffused by rotating nozzles. In view of the increasing quality demands, Hoogovens Aluminium initiated a research project to improve the efficiency of the Sidafil. Degassing of oxygen from water in a full scale water model established the influence of main factors like melt flow rate, purge gas flow rate, rotor size, rotor speed, and baffle positions. Dimensionless groups describe bubble flow regime and rotor mixing characteristics. The general degassing theory, for example as formulated by Sigworth and Engh, provides insight in water model results and cast shop experiments. 11:00 am EFFECT OF POROUS PLUGS ON METAL QUALITY: Ronald Murga, Casting Quality Process Control Engineer, Ravenswood Aluminum Corporation, P. O. Box 98, Ravenswood, WV 26164 An initial look at what the effect of porous plugs have on molten metal quality in a tilting holding furnace. From a quality viewpoint, how do porous plugs compare to conventional non-porous plug technology in melt preparation? How is the quality of a typical charge of molten metal impacted by the use of porous plugs in the holding furnace for the introduction of process gas? 11:20 am OPERATING PRINCIPLES AND CAST SHOP APPLICATION OF THE NOTORP HYDROGEN ANALYZER: Luiz C. B. Martins, Leonard S. Aubrey, SELEE Corporation, 700 Shepberd Slreet, Hendersonville, NC 28792; George M. Grendell, TYK Refractories Company, 301 Brickyard Road, Oairton, PA 15025 The NOTORP Hydrogen Analyzer was developed recently to monitor dissolved hydrogen content in molten aluminum alloys. The Analyzer uses an electrochemical cell mounted inside a sampling tube, with one side of the cell exposed to the molten aluminum and the other side in eqUilibrium with a reference gas of known hydrogen content. The electrochemical potential measured across a solid state proton conducting ceramic element is proportional to the difference in hydrogen activity. The Analyzer was tested in several cast shops, demonstrating ease of use, reliability, and rugged design. Comparison of the "NOTORP" Analyzer with other techniques was also done during these tests, and the results are presented here. Hydrogen content results were consistent with other techniques. Advantages of the NOTORP Analyzer were ease of use, continuous readout capability, and longer probe life.

EMERGING TECHNOLOGIES SYMPOSIUM: Engineering Science Of Chem-Mechanical Planarization III

Sponsored by: EMPMD Thin FUm & Interfaces Committee Program Organizers: Professor Krishna Rajan, Materials Science and Engineering Department, Rensselaer Polytechnic Institute, Troy, NY 12180-3590; Dr. Michael Fury, Rodel Inc., 451 Bellvue Road, Newark, DE 19713 Thursday, AM February 8, 1996

Room: Grand K Location: Anaheim Marriott Hotel

Session Chairperson: Dr. Mukesh Desai, Speedfam Corporation, 7406 W. Detroit, Chandler,AZ

1996 TMS ANNUAL MEETING AND EXHIBITION

8:30 am Invited ELECTROCHEMICAL EFFECTS IN THE CHEMICAL-MECHANICALPOLISHING OF COPPER FOR INTEGRATED CIRCUITS: C.A. Sainio, D.J. Duquette, Rensselaer Polytechnic Institute, Troy, NY 12180-3590 Because of its high conductivity and high resistance to electromigration, copper has been proposed as a replacement for aluminum in silicon integrated circuit interconnections. Chemical-mechanical polishing (CMP) is being investigated as a means for patterning copper interconnects. The studies descrihed in this work have focused on understanding the electrochemical and mechanical parameters that control planarization during CMP. The role of dissolved Cu 2+ ion, oxidizers, and passivators in the polishing solutions has been investigated. Eleclrochemical potential, pH, and linear polarization resistance measurements have been used to characterize the removal of copper during CMP. The effects of the slurry on the diffusion barrier (titanium or tantalum) have been examined as well as possible interactions of copper with the diffusion barrier. It is shown that eleclrOChemicai techniques can be used to study the effects of chemical compositions of polishing solutions on the planarization process. 9:00 am Invited ADVANCES IN PHYSICALLY-BASED EROSION SIMULATORS FOR CMP: Scott R. Runnels, Southwest Research Institute, San Antonio, TX A physically-based simulation platform has been developed that is capable of simulating the complete erosion of multiple oxide features during chemical-mechanical polishing of semiconductor wafers. The platform separates the erosion model from the stress model, making it easy to use and versatile. In this talk, theories of planarization will be reviewed and their implementation into the simulator explained. Erosion simulation results will also be presented. 9:30am THE MECHANICS OF PAD DEFORMATION: Krishna Rajan, Materials Science and Engineering Department, Rensselaer Polytechnic Institute, Troy, NY 12180-3590 The development of pads in chern-mechanical planarization technologies is a critical component of this technology, yet is governed by a high degree of empiricism. While there is slrong need to develop customized pads for different types of polishing applications, there is at present no scientific framework to design and manufacture pads. In this presentation, we develop such a framework based on understanding the relationship between the mechanical properties of pad materials and their "microstructure". 10:00 am BREAK 10:30 am DISTRIBUTION SYSTEMS FOR CMP: Robert Korman, MEGA Systems and Chemicals, Dennis Capitanio, Pall Corporation, Scientific & Laboratory Services Dept., 25 Harbor Park Drive, Port Washington, NY 11050 Until recently slurry blending and distribution was done in small hatches in close proximity to the polisher. The modern FAB eqnipped with many polishing lines demand that a slurry be mixed, filtered and delivered to each polisher in a higbly automated fashion. The slurry must be of reproducible quality and concenlration as well as free of agglomerates and foreign contaminants. These distribution systems are referred to as Bulk Slurry Distribution Systems (BSDS™) which meet the unique requirements of the CMP process. The design of a distribution system is discussed in great detail to emphasize the need for solid engineering practice and field experience to meet every aspect of slurry management and distribution. Also presented will be an in depth study of CMP slurry filtration that was recently undertaken to insure that a contaminant free product is delivered to the polisher. 11:00 am PANEL DISCUSSION: MATERIAL SCIENCE ROADMAP FOR CHEM-MECHANICALPLANARIZATION: Moderators: Dr. K. V. Ravi, Applied Materials Corporation, MS-0204, 3100 Bowen Avenue, Santa Clara, CA 95054; Dr. Michael Fury, Rodel Inc., 451 Bellvue Road, Newark, DE 19713; Professor Krishna Rajan, Materials Science and Engineering Department, Rensselaer Polytechnic Institute, Troy, NY 121803590 Abstract not available.

GENERAL PYROMETALLURGY III

Sponsored by: EPD Pyrometallurgy Committee Program Organize.... Jerome P. Downey, Hazen Research, Inc., Golden, CO 80403; Ed Mounsey, Ausmelt Limited, 2113 Kitchen Road, Dandenong, Victoria 3175 Australia Thursday, AM February 8, 1996

Room: AI6-17 Location: Anaheim Convention Center

Session Chairmen: L. G. Twidwell, Professor, Montana College of Mineral Science & Technology, Department of Metallurgy, Butte, MT 59701; Osamu Ishikawa, Bechtel Corporation, 50 Beale Street, San Francisco, CA 94119-3965

8:30am MANGANESE BEHAVIOR AT MINIMAL SLAG OPERATION IN A COMBINED BLOWING CONVERTER: lkuhiro Sumi, Tomoo Isawa, Toshio Takaoka, Materials and Processing Research Center, NKK, I Kokan-Cho, Fukuyarna, 721 Japan A new steel making process, in which manganese ore was intentionally reduced, was developed to achieve a higher manganese yield using a combined blowing converter. In this study, manganese behavior was discussed through the inlroduction of a model. The model consists of two terms; one represents reduction rate of manganese ore and the other represents oxidizing rate of manganese by oxygen supply from top lance. Appar-

B135

ent reduction rates of manganese were calculated by the two terms. TIlese terms were determined experimentally, with 50 kg and 250 ton combined blowing converters. TIle model calculated manganese changes in a short duration and the integration of these changes gave an estimation of total manganese content change for the operation. We applied the model to the actual operation and found that manganese behavior was predicted fairly well by this model. 8:55am MELTING BEHAVIOR OF SILICA FLUX IN THE COPPER CONVERTER: GUO Xian Jian, DING Wei An, Beijing General Research Institute for Nonferrous Metals, Beijing, 100088, Cbina; James L. Hendrix, University of Nevada, Reno, Mackay School of Mines, Reno, Nevada In this paper, the effects of matte grade, flux grain size, blast flowrate and blast ~ content on the melting process of silica flux in a copper converter were investigated. TIle results showed that the blast flowrate and the blast ~ content were not significant factors in the melting process. However, with an increase in grain size of flux and grade of matte, the amount of the flux molten decreased. TIle data on the rate of the melting process correspond well with the mathematical model used which simulates the chemical reactions occurring at the interface. TIle kinetic equations of melting are as follows: X =1-(1 - 0.0206t1ro)3 (matte grade =69.57% Cu); X =1 - (I - 0.0169t1ro)3 (matte grade =74.95% Cu); X =melted fraction, t =time, ro =initial radius of grain. 9:20am AN IN-BATH LIQUIDUS TEMPERATURE MEASUREMENT FOR MOLTEN SALTS AND SLAGS: Eric J. Grimsey, Nan Li, X. Yong Yan, Western Australian School of Mines, PO Box 597, Kalgoorlie WA, Australia 6430; Andrew Warczok, Torstein A. Utigard, University of Toronto, Toronto, Ontario, Canada M5S lA4 A technique is being developed to provide a rapid liquidus measurement for salts and especially metallurgical slags. A copper or nickel cylinder is immersed into a molten bath, and a computer based data-logging system is used to record the heating profile of a type K thermocouple centered within the cylinder. TIle heat transfer is affected initially by the presence of a thin crust on the cylinder surface which freezes from the bath. When the crust melts, the rate of heating shows a subtle increase. Numerical differentiation is used to detect the on-set of this increase so as to provide an indication of the liquidus temperature of the melt. This paper outlines the technique and its application to a number of molten salts and slags. Factors which affect the senSitivity of the measurement are discussed also. 9:45 am BREAK 10:15 am CURRENT OPERATIONS IN SMM'S SLIME TREATMENT: C. Segawa, T. Kusalwbe, Niihama Copper Refinery, Sumitomo Metal Mining Co. Ltd., Niihama, Ehirne -792,Japan SMM operates an original slime treatment process for recovering precious metals and other metals (Te, Se, Sb, Bi) from the copper and the lead anode tankhouse slimes. In recent years, the amount of slimes had increased and the pyrometallurgical processing units capacity became the bottleneck of the entire plants. For the expansion of the plant, new pyrometallurgical facilities, incorporating a BBOC, was commissioned in 1992 and has now been successfully operated. In addition, unit operations were optimized to increase recovery and reduce operating cost. TIle results of modifications made in the plant are discussed in this paper. 10:40 am V-AI ALLOYS BY ALUMINOTHERMIC REDUCTION OF V105: Carlos A. Nunes, Univ. of Wisconsin - Madison, 1500 Eng. Drive, Madison, WI 53706 and Hugo R Z. Sandim, FAENQUIL - DEMAR, Polo Urbo-Industrial sinO, 12600-00, Lorena, SP, Brazil V-AI alloys have been used commercially for the production of pure vanadium by Electron Beam Melting (EBM), as well as for the obtainment of11-based alloys (ex. 116AI-4V). TIle aluminothermic reduction of the vanadium pentoxide (V205) is the primary method to obtain these alloys. Concerning the obtention of pure vanadium, some results in the literature have indicated a V-15 wt% Al alloy as the suitable alloy for direct EBM. We have studied the a1uminothermic reduction of V205 in order to define operational parameters which might lead to V-AI alloys presenting composition close to 15 wt% AI. In this work, we present results on the effect of Al excess, Cao addition and reaction atmosphere (air, argon) on the alloy composition (AI, 0, N) and metallic yield. In one of the experiments carried out in argon and using 30% excess AI, we obtained an alloy presenting the following composition: 13.5 wt% AI; 2,018 wt ppm 0, and 40 wt ppm N. TIle vanadium yield in this specific case was 93%. Results on microstructural analysis of selected alloys by SEM (BSE images) and EPMA (WDX) will also be presented.

A SYMPOSIUM DEDICATED TO PROFESSOR JOHANNES WE ERTMAN'S SEVENTIETH BIRTHDAY: Session VII: High Temperature Deformation III

Sponsored by: JL SMDIMSD Nuclear Materials and MSD Flow and Fracture Committees and FEMS (Federation of European Materials SocIeties) Program Organizers: RJ. Arsenault, Deptartment of Materials Science and Nuclear Engineering, University of Maryland, College Park, MD 20742-2115; David Cole, CRREL, 72 Lyme Rd., Hanover, NH 03755; Todd Gross, Department of Mechanical Engineering, University of New Hampshire, Durham, NH 03824; Gemot Kostorz, Institut fUr Angewandte Physik, ETII Honggerberg, CH-8093 ZUrich, Switzerland; Peter Liaw, Department of Materials Science and Engineering, University of Tennessee, Knoxville, TN 37996-2200; Sivan Parameswaran, NRC-Institute for Aerospace Research, Ottawa, Canada KIAOR6; Howard Sizek, IncoAlloys International Inc., Hun-

B136

tington, WV 25705-1771 Thursday, AM February 8, 1996

Room: Orange County 3 Location: Anaheim Marriott Hotel

Session Chairpersons: RJ. Arsenault, Metallurgical Materials Laboratory, Department of Materials and Nuclear Engi_ring, University of Maryland, College Park, MD 207422115; A.K. Mukherjee, Department of Chemical Engineering and Materials Science, University of California, Davis, CA 95616

8:30am ON THE DEFORMATION MECHANISM IN HARPER·DORN CREEP: J.N. Wang, Chemistry and Materials Sciences, L-370, Lawrence Livermore National Lab0ratory, PO Box 808, Livermore, CA 94551-9900 Harper-Darn creep, a Newtonian dislocation creep, has been observed in many metallic, ceramic and silicate materials. Based on an early Weertman's model for power law creep, a microphysical model for Harper-Dom creep is developed. Two assumptions are made. One is that a steady dislocation density is established not only by the applied stress but by the Peierls stress as well. The other is that the flow process is dislocation glide plus climb being rate-controlling. It is shown that the predicted dependencies of dislocation density and strain rate on the Peierls stress are in very good agreements with experimental dnta on a wide range of materials. 8:50am AN INVESTIGATION OF THE TRANSITION CONDITIONS BETWEEN DIF· FUSIONAL CREEP, HARPER-DORN CREEP AND POWER LAW CREEP: J.N. Wang, Chemistry and Materials Sciences, L-370, Lawrence Livermore National Laboratory, PO Box 808, Livermore, CA 94551-9900 TIle information about the transition condition between different creep regimes is essential both for developing new materials for engineering applications and for understanding and modelling geologi(:al deformation processes in the Earth. It is recently found that the Peierls stress of a crystal is an important factor determining the transitions between lattice diffusional ,creep, Harper-Dam creep and power law creep. Both theoretical predictions and experimental data consistently show that a high Peierls stress leads to a high stress marking the transition from power law creep to Harper-Dam and diffusional creep and to a small grain size representing the transition from Harper-Dam creep to diffusional creep. Experimentally confirmed theoretical or semi-empirical rate equations are used to construct go:neralized deformation mechanism maps. Such maps clearly illustrate that Harper-Dam creep may be an important mechanism controlling the high-temperature deformation of materials with high Peierls stresses in wide ranges of grain size and stress. 9:10am THE ROLE OF GRAIN BOUNDARY CHARACTER DISTRIBUTION ON CREEP DEFORMATION OF NI-16Cr·9Fe AT 0.38 T m: V. Thaveeprungstipom, Department of Nuclear Engineering, University of Michigan, Ann Arbor, MI 48109; G.S. Wae, Department of Nuclear Engineering and Materials Science and Engineering, University of Michigan, Ann Arbor, MI 48109 During high temperature creep deformation, lattice dislocations interact with grain boundaries creating extrinsic grain boundary dislocations (EGBDs) which may then be dissociated, absorbed or trapped in the boundaries depending on the type and structure of the grain boundaries. Evidence indicates that dislocation absorption in random high angle boundaries (HABs) is much easier than that in coincidence-site-lattice boundaries (CSLBs). Thus the rate of formation and absorption of EGBDs during creep deformation is of practical importarlce since it may be related to the rate at which backstresses on following dislocations. are relaxed and this relaxation may control the rate of deformation. This research reveals that by increasing CSLB populations in Ni-l6Cr9Fe alloy the flow stress was increased by 15%-30% and its steady-state creep rate was lowered by a factor of 40. A modified recovery creep theory that includes the effect of grain boundary character distribulion will also be discussed for a better understanding of high temperature creep deformation. 9:30am PRIMARY CREEP IN DISPERSION STRENGTHENED MATERIALS: R.S. Mishra, A.K. Mukherjee, Department of Chemical Engineering and Materials Science, University ofCalifomia, Davis, CA 95616 Primary creep is perhaps the most important stage for a majority of creep applications, yet it has received very little attention. TIle understanding of the influence of second phase particles on primary stage inn quite inadequate. TIle primary creep data of a number of dispersion strengtholned aluminum alloys is analyzed. The magnitude of primary creep scales with the interparticle spacing and it appears to be linked with the steady state creep mechanism. Based on the present results, the primary creep response of a dispersion strengthened matelial can be divided in three regimes; (a) low primary strain when the interparticle spacing is smaller than grain size. (b) medium primary strain when the interparticle spacilng is similar to the grain size, and (c) large primary strain when the interparticle spacing is larger than grain size. TIle primary decay constant for composites is significantly different from mechanically alloyed materials. A model is suggested to explain the interparticle dependence of the primary creep. 9:50am CREEP MECHANISMS AND A.CCELERATED COARSENING IN A LAMEL· LAR EUTECTIC: T. Plookphol, D.S. Stone, Department of Materials Science and Engineering, University of Wisconsin, MS&E Building, Madison, WI 53706; S-M. Lee, Samsung Electronics Co. Ltd, Suwon, South Korea TIle microstructure of a eutectic solder joint coarsens during thermal fatigue. To gain insight into this process, we investigate a related one: acceleration, due to prior deformation, of coarsening during an isothermal anneal. Eutectic lead-tin specimens are deformed to 10% at room temperature and strain rates 10-6/&< £ < Io-l/s, then annealed

1996 TMS ANNUAL MEETING AND EXHIBITION

at 12S·C. Beyond 50% volume fraction coarsened material (V = 0.5), the rate of coarsening can be expressed as Y=f ( )xY· where Y· is Y· in absence of prior deformation. The multiplicative factor f( peaks at =10-4/s. Based on evidence from creep, load relaxation, and tensile tests, we propose f( reflects the ability of the microstructure to soore dislocations during deformation. The peak in f( is likely related to colony boundary sliding.

e e)

e

e)

e)

10:10 am BREAK 10:30 am IDGHTEMPERATURE DEFORMATION MECHANISM OF A120JINIAI COMPOSITES: K. Xu, R.I. Arsenault, Metallurgical Materials Laboratory, Department of Materials and Nuclear Engineering, University of Maryland, College Park, MD 207422115 Long screw dislocations with jogs or super jogs were observed in NiAi matrix composites after high temperature deformation of NiAI matrix composites. Thermally activated motion of those jogged screw dislocations are believed to be the major mechanism to explain the mechanical behavior or NiAl matrix composites at high temperature. However, the jogged screw dislocation mechanism requires the existence of only vacancy-producing jog in the matrix during the deformation. Barret and Nix also proposed a creep model which indicated that the alternative arrangement of vacancy-producing and vacancy-absorbing jogs would make the mechanism break down. In this paper, a multiple cross slip of screw dislocation model due to the local interaction between screw dislocation and spherical particles based on deformation of NiAi matrix composites is proposed. As a result, groups of single sign jobs are produced in screw dislocation with large group interspacing. This multiple cross slip mechanism may give the survival condition for jogged screw dislocation mechanism for high temperature deformation. 10:50 am CREEP-FATIGUE CRACK PROPAGATION BEHAVIOR OF HAYNES HRI60 ALLOY: Weiju Ren, Peter K. Liaw, Department of Materials Science and Engineering, The University of Tennessee, Knoxville, TN 7996-2200; R.W. Swindeman, B.G. Gieseke, Metal and Ceramic Division, Oak Ridge National Lab, Oak Ridge, TN 378316155; G.Y. Lai, Haynes International, Inc., 1020 West Park Ave., Kokomo, ID 469029013 Creep-fatigue crack growth behavior was studied on a Haynes HR 160 alloy, a newly commercialized high· temperature superalloy. The test temperature investigated ranged from 850·C to lOOO·C. The results indicated that creep crack propagation rates can be uniquely described by the time-dependent fracture mechanics parameter, C*(t), in the form of daldt = A[C*(I))q. This correlation was found to be independent of test temperature, load level and specimen size. Using the time-dependent fracture mechanics parameter, the creep-fatigue crack propagation behavior has been successfully correlated with the creep crack growth rates. Microstructural characterization showed that the creep crack grew mainly in an intergranular fracture, while load cycling induced a mixed intergranular and transgranular crack growth mode.

Program Organizers: U. Balachandran, Energy Technology Division, Argonne National Laboratory, Argonne, lL 60439; Paul J. McGinn, University of Notre Dame, Notre Dame, IN 46556; and Stuart Abell, University of Birmingham, Edgbaston, Birming· hamBI521T, UK Thursday, AM February 8, 1996

Room:B7 Location: Anaheim Convention Center

Session Chairpersons: M. D. Sumption, The Ohio State University, Columbus, OH 43210; S. Abell, University of Birmingham, Edgbaston, Birmingham BI521T, UK

8:30am INFLUENCE OF LI DOPING ON FLUX PINNING IN BULK BI:2212: M. D. Sump/ion, E. W. Collings, Department of Materials Science, The Ohio State University, Columbus, OH 43210; P. Krishnaraj, N. G. Eror, Department of Materials Science, University of Pittsburgh, PA 15261 Magnetic Jc and Hr were measured for Bi:2212 sintered samples containing varying amounts of Li. The first set of samples consisted of: (I) Bi2Sr2CaCUI.9Li(1.l0x, (2) Bi2Sr2CaCuI.9Li(1.20x, (3) Bi2Sr2CaCu2Lio.20x, and (4) a control sample. It was found that Li generally depressed Tc but enhanced the pinning over a wide range of temperatures. This was found to be true even when Tc suppression was taken into account. For low concentrations of Li (up to Li = 0.4) the increase in pinning is proportionalto the concentration. For example, results at 57.5 K (without Tc normalization) show that Hr (0-4 Li) = 2.8 kOe; Hr(0.2 Li) = 2.6 kOe; and Hr(control = 1.8 kOe. For Li = 0-4, Tc is depressed 10 88.2 K (down from 91.2 K). Sample microstructure and the role of CuO precipitates in flux pinning are also addressed. 8:50am MODIFIED 10DODMETRIC TITRATION TECHNIQUE FOR HIGH SENSI· TIVlTY OXYGEN DETERMINATION IN CUPRATE SUPERCONDUCTORS: P. S. Whitfield, J. Bhakta, G. Yang, I. Gameson,J. S. Abell, School of Metallurgy and Materials, University of Birmingham, Birmingham, BI5 21T, UK The importance of oxygen content on the structure and properties of high Tc materi· a1s has long been known. The number of techniques available to routinely determine accurate oxygen contents of these materials is limited, especially where sample size is limited as with crystal growth. The oxygen contents of crystalline and amorphous mao terial have been determined by modified iodometric titration, neutron data refinement, and TGA. For titrations samples of 20-40 mg were dissolved in HBr and the evolved Br3-analyzed iodometrically using citrate complexation of Cu(ll). The results from each of the methods will be compared and the suitability of the iodometric titration technique for accurate oxygen content determination of small samples assessed.

were determined from the effects of temperature and strain rate cycling on the flow stress at a constant structure. Both i'l 0 and G increased with stacking fault energy. The rate controlling dislocation mechanism was deduced to be the cutting of forest dislocations. The derived dislocation jog and stacking fault energies were inn accord with those obtained by other methods of measurement.

9:10am ENHANCED PROPERTIES IN SCREEN PRINTED YBa2Cu307_1i THICK FILMS WITH DOPING ADDITIONS OF PLATINUM AND RHODIUM: J. Langhom, Y. J. Bi, J. S. Abell, Department of Metallurgy and Materials, University of Birmingham, Edgbaston, Birmingham BI5 21T Platinum group metals have been observed to react within the YBCO system 10 form phases which may act as heterogeneous nucleation sites for the deposition of refined Y2BaCu05 (211) precipitates. Additions have been made to thick films of YBCO, and have induced significant improvements in the physical properties of the processed films, in particular critical current densities (Jc). Controlled concentration additions have yielded Jc values in excess of 6 x 103 A/cm-2 at 77 K and zero applied field, this is with respect to Jcs of2 x 103 Acm-2 for undoped specimens. Optical and transmission electron microscopy bas shown a homogeneous distribution of sub-micron sized, and larger highly anisotropic 211. An increased density of dislocations associated with the 1231 211 interface with this refinement suggesting that smaller 211 precipitates may act as heterogeneous nucleation sites for flux pinning defects. Additions have also yielded increased homogeneity, and optimized Tcs in the system by aiding oxygen diffusion through the system

U:5Oam DISLOCATION LINE TENSION AND SPACE CURVES: C.S. Hartley, Depart· ment of Mechanical Engineering, Rorida Atlantic University, Boca Raton, fl. 334310991 A component of the restoring force on dislocations having the form of space is identified by differentiation of the vector line tension. This component acts in the direction of the binormal to the space curve and is proportional to the torsion, while the familiar "effective line tension" force is proporlionalto the curvature and acts in the direction of the principal norrnalto the curve. The component in the direction of the binormal vanishes for plane curves. These concepts are applied to analyze the formation and stability of dislocation helices from straight dislocations in the presence of non-equilibrium vacancy concentrations and general stress fields. It is shown that, in the isotropic approximation, the circular helix is stable with respect to a straight dislocation originally lying along the axis of the helix only if the dislocation is a screw. The equilibrium form of a general space curve immersed in an arbitrary stress field and non-equilibrium vacancy concentration is given implicitly by relations which equate the corresponding forces to the self force produced by the dislocation.

9:30am YTTRlA STABILIZED ZIRCONIA BARRIER LAYERS BY ATMOSPHERIC PLASMA SPRAYING: S. C. Wareham, J. S. Abell, School of Metallurgy and Materials, University of Birmingham, Edgbaston, Birmingham, B IS 21T, UK Thick films ofYBa2Cu307_x (YBCO) high temperature superconductor have p0tential use in microwave device and magnetic screening applications. Deposition onlO a substrate to support the film and/or be an integral part of a microwave device is necessary. YBCO films deposited onlO alumina substrates (the preferred dielectric material) show a large degree of interaction, unlike films on yttria stabilized zirconia (YSZ). Atmospheric plasma spraying (APS) is used commercially to produce high density ceramic coatings in thicknesses from several microns to a few millimeters. In the present work the production of YSZ barrier layers by APS onto alumina substrates has been studied. The results of the subsequent evaluation of these layers by optical and electron microscopy, and x-ray diffraction are presented. A comparison is made between the properties of plasma sprayed and screen printed barrier layers. Results of the physical and electrical behaviour of screen printed YBCO films prepared on these barrier layers will also be presented.

11:10 am PLASTIC DEFORMATION KINETICS OF FCC METALS AT LOW HOMOLOGOUS TEMPERATURES: W.D. Cao, H. Conrad, Materials Science and Engineering Department, North Carolina State University, Raleigh, NC 27695-7517 The plastic deformation kinetics of polycrystalline Ag, Cu and AI representing a range in stacing fault energy were investigated at 78 10 300 K. The parameters in the thermally activated plastic strain rate equation

HIGH TEMPERATURE SUPERCONDUCTORS· SYNTHESIS, PROCESSING, AND LARGE SCALE APPLICATIONS VIII: Processing Sponsored by: JL EMPMD/sMD Superconductlng Materials Committee, MSD Electronic, Magnetic and Optical Phenomena Committee, and FEMS (Federation of European Materials SocIeties)

1996 TMS ANNUAL MEETING AND EXHIBITION

9:50am CRUCIBLES FOR CRYSTAL GROWTH BY ATMOSPHERIC PLASMA SPRAYING: S. C. Wareham, D. Holdom, J. S. Abell, School of Metallurgy and Materials, University of Birmingham, Edgbaston, Birmingham, B15 21T, UK High quality crucibles for crystal growth are conventionally produced by the slip casting of alumina or zirconia. The success of slip casting is limited by the sintering characteristics of the material used. Potentially useful materials such as BaZr03 do not sinter easily, and so do not lend themselves to forming in this way. Recent studies have used hot isostatic pressing to achieve the required properties. Atmospheric plasma spray-

B137

ing (APS) is used to form near net shape artifacts from a number of materials including metals, ceramics, and plastics. The chosen material is sprayed to the required thickness onto a prepared mandrel. Subsequent removal of the mandrel leaves the free standing artifact. The fabrication of crucibles from yttria stabilized zirconia, alumina and barium zirconate by APS has been studied. These have been evaluated by optical and electron microscopy, and density measurements and compared to those commercially available. Characterization of cuprate superconducting crystals grown in these crucibles will be presented. 10:10 am BREAK 10:l0am MAGNETIC CHARACTERIZATION OF BULK MELT TEXTURED YBCO: LIMITS OF LEVITATION FORCES: W. Gawalek, T. Habisreuther, T. Straer, M. Wu, D. Litzkendorf, and P. Gllrnert, Institut fur Physikalische Hochtechnologie e.V., Helmholtzweg 4, 0-07743 Jena, Germany Bulk cylindrical blocks of YBCO, 30 mm in diameter and 17 mm in thickness, are prepared by a modified melt textured growth hatch process. Blocks are consisting of weak textured 1-2 crn3 grains. By vibration sample magnetometry intra-grain critical current densities up to 5·104 Ncffil. at 77 K are measured. Integral levitation force measurements are performed at 77 K and 18 K in the zero field cooled (zfc) and field cooled (fc) state. Measured values are compared with calculated theoretical limits for an ideal hard pinning superconductor material interacting with a given strong permanent magnet Using a 0.41 T SmCo magnet, 25 mm in diameter and 15 mm in thickness a zfC maximum levitation force of 36 N at 77 K was measured 0.5 mm ahove the block. Tbis is about 41 % of the theoretical limit in case of perfect shielding. At 18 K about 57% of the zfc limit is reached. In the fc state the multi grain material is more efficient. We found about 70% of the fc limit at 77 K and about 93% of the fc limit at 18 K. 10:40 am PROGRESS IN QUASI-CONTINUOUS DIRECTIONAL SOLIDIFICATION OF BULK YBazCu3O, CERAMICS WITH CeOlADDITIVES: X. Obradors, S. Piilol, V. Gomis, F. Sandiumenge, B. Martinez, N. Vilalta, Institute de Ciencia de Materials de Barcelona, C.S.I.C. Campus U. A. Barcelona, 80193 BeUaterra, Catalunya, Spain Directional solidification of bulk 1231211 ceramic composites has beel' • wed out by means of a vertical Bridgman technique allowing to reach a detailed control of the growth parameters. The use of small concentrations of Ce<>2 additives «1% wI) has been shown to have two different beneficial effects. First, the conditions under which the semisolid melt is stable at high temperature are largely widened. Second, these additives further contribute to the refinement of 211 precipitates in the composite thus allowing to increase the critical currents above 105 Ncm2 at 77 K. The Bridgman solidification technique has been shown to have a great potential for quasi-continous production of bulk rods of superconducting textured YBCO. Samples with diameters between 1 mm and 20 mm and lengths up to 120 mm have been fabricated with our methodology and the microstructure and superconducting properties have been investigated by means of several experimental techniques. A detailed discussion of our processing optimization activity will be reported. Finally, the applicability of these materials for several devices (current leads,levitators, etc.) will be stressed. 11:00 am A COMPARATIVE STUDY OF MAGNETIC AND TRANSPORT PROPERTIES OF YBalCu30 SINGLE CRYSTALS WITH A ruGH DENSITY OF CRITICAL CURRE~ S. N. Gordeev, Physics Department, University of Southampton, Southampton S017 IBJ, UK; W. Jahn, A. A. Zhukov, H. Kupfer, T. Wolf, Forschungszentrum Karlsruhe, Institut fur Technische Physik, 76021 Karlsruhe, Germany Sr-doped YBa2Cu30y single crystals were studied. Due to a high density of pointlike defects such samples have high critical current density J -105 Ncm 2 at T =77 K in magnetic field 1.5 T. Both magnetic and pulsed transport measurements result in a nonmonotonous J(B) dependence with a typical for peak-effect maximum at intermediate field Bp. It was found that the shape of current-voltage curves changes in the region of J(B)-peak. At low magnetic fields (B
RADIATION MATERIALS SCIENCE IN TECHNOLOGY APPLICATIONS V: Simulations and Modeling

Sponsored by: Jt. SMDIMSD Nuclear Materials Committee Program Organizers: L. K. Mansur, Metals and Ceramics Division, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 37831-6376; C. L. Snead, Jr., Applied Technologies Division, Brookhaven National Laboratory, PO Box 5000, Upton, NY 11973-5000 Thursday, AM February 8, 1996

Room: Grand H Location: Anaheim Marriott Hotel

Session Chairperson: T. A. Gabriel, Computational Physics and Engineering Division, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 378316364

B138

8:30 am Invited POINT DEFECT CLUSTER MOBILITY IN RADIATION DAMAGE MODELS: R. E. Stoller, Oak Ridge National Laboratory, PO Box 2008, Oak Ridge, TN 378316376 Results of recent molecular dynamics simulations of displacement cascades indicate that the mobility of small point dl~fect clusters may be relatively high. The impact of their mobility on the predictions of radiation damage models using the chemical rate theory will be discussed. Although this theory has been broadly applied in models simulating such radiation-induced phenomena as void swelling and embrittlement, the mobility of defects other than single vacancies and interstitials has generally been ignored. As a result, both the point defect concentrations and point defect cluster sink strengths can reach levels that are not physically reasonable at low temperatures. The results of the new rate theory simulations indicate that including the role of mobile clusters may extend the applicability of the kinetic models to lower temperatures. Research sponsored by the Office of Nuclear Re'gulatory Research, U.S. Nuclear Regulatory Commission under inter-agency agreement DOE 1886-8109-8L with the U.S. Department of Energy and by the Division of Materials Sciences, U.S. Department of Energy under contract DEAC05-840R21400 with Lockheed Martin Energy Systems. 9:00am MIGRATION OF INTERSTITIAL CLUSTERS IN a-IRON: B. D. Wirth, G. R. Odette, D. Maroudas, G. E. Lucas, Departments of Mechanical Engineering and Chemical Engineering, University of California, Santa Barbara, CA 93106 Energetic primary recoil atoms from fast neutron irradiations generate both isolated point defects and clusters of vacancies and interstitials. Cluster stability and mobility play key roles in the subsequent fate of the defects and, hence, in the overall microstructural evolution under irradiation. Single interstitials, di-interstitials, and probably larger interstitial clusters are highly mobile at low temperatures. Indeed, recent molecular dynamics simulations have demonstrated the mobility of larger interstitial clusters. Interstitial cluster mobility has profound significance to the kinetics of microstructural evolution: immobile clusters can serve as stable nuclei for interstitial-type loops while mobile clusters diffuse to and are annihilated at sinks. For example, a population of clusters that anneal during irradiation would mediate flux and temperature dependent recombination rates. In this study, the mobility of small interstitial clusters in -iron is evaluated by molecular··statics and molecular-dynamics simulations using interatomic potentials based on the embedded-atom method. Clusters of direction towards the direction which bas only a slightly higher energy. The interstitials can migrate by rotation and translation to equivalent sites, off-a:ds <110:> sites, or fully rearrange into a crowdion configuration; the crowdions migrate along their axis in what is often observed to be a correlated sequence of jumps befoCl~ returning to the configuration. The mixed mode process can be described by lID effective diffusion coefficient. The interactions between interstitials and single point defects as well as point-defect clusters are characterized by calculating binding ener~:ies and position-dependent reaction probabilities. These results are incorporated into lattice Monte Carlo simulations of residual defect production following intermediate-term cascade evolution and of defect-sink interactions. Supported by the U.S. Nuclear Regulatory Commission Contract: NRC-04-94049. 9:40am A MOLECULAR DYNAMICS STUDY OF DEFECT.INDUCED AMORPHIZATION OF INTERMETALLIC COMPOUNDS: Ram Devanathan. Materials Science and Technology Division, Los Alamos National Laboratory, Los Alamos, NM 87545; Nghi Q. Lam, Paul R. Okamoto, Materials Science Division, Argonne National Laboratory, Argonne, IL 60439 A molecular dynamics (MD) study of the thermodynamic and elastic properties of intermetallic compounds has led to a deeper understanding of the effects of simple lattice defects in destabilizing crystilline materials. In particular, MD simulations of defect-induced amorphization of in«~rmetallic compounds show that the static component of the relative mean square atomic displacements (rmsd) provides a quantitative measure of the excess enthalpy stored in the lattice. This static component produces a softening of the average shear modulus nearly identical in magnitude to that produced by the dynamic component associa«~d with heating. In fact, the MD results show that the average shear modulus is a unique function of the total (static+dynamic) rmsd. These results provide the first direct confirmation of a generalized version of the Lindemann melting criterion. In its simplest form, this criterion states that a metastable crystal will melt when the total rmsd exceeds a critical value. Thus, solid-state arnorphization can be understood as defect-induced melting of a critically-disordered crystal.

1996 TMS ANNUAL MEETING AND EXHIBITION

DE-AC04-94AL85ooo.

10:00 am BREAK 10:20 am Invited COLLISION CASCADES IN POLYATOMIC COMPOUNDS: K. E. Sickafus, Los Alamos National Laboratory, Los Alamos, NM 87545; W. J. Weber, R. E. Willford, Pacific Northwest Laboratory, PO Box 999, Richland, WA 99352; R. A. Sutton, CamegieMellon University, Pittsburgh, PA 15213 Irradiation damage is dependent on: (I) the number of secondary knock-on atoms (ska) displaced from their atomic sites by a primary knock-on atom (pka); (2) the kinetic energy imparted to ska atoms in a cascade; and (3) the spatial extent of cascades. We have solved for item I in polyatomic materials by calculating solutions to simultaneous integrodifferential transport equations, by the method of Parkin and Coulter (J. Nucl. Mater. 101 (1981)261). We have evaluated items 2 and 3 using the Monte Carlo code TRIM. Transport equation solutions allow us to evaluate the number of displaced atoms in a polyatomic compound, as a function of threshold displacement energies for individual atomic constituents, and mass disparity between constituent atoms. We use items 2 and 3 to evaluate the mean cascade energy density by tbe method of Walker and Thompson (Rad. Effects 37 (1978) 113). We have determined, however, that it is best to employ the geometric mean volume to evaluate the average cascade energy density from a compilation of numerous Monte Carlo ion simulations. We will present results for three similar ternary compounds, with and without mass disparities: spinel (MgAI204), chrysoberyl (BeAl204), and gahnite (ZnAI204). 10:50 am NET DISPLACEMENT FUNCTIONS FOR SIC, A1203, and MgO: W. 1. Weber, R. E. Williford, Pacific Northwest Laboratory, PO Box 999, Richland, WA 99352; K. E. Sickafus, Los Alamos National Laboratory, PO Box 1663, MS K-762, Los Alamos, NM87545 Solutions to polyatomic integrodifferential equations for net displacements were determined numerically for SiC, AI203, and MgO using the Ziegler, Biersack,:and Littrnark (ZBL) universal scattering potential. The effect of electronic stopping powers was investigated by comparing results calculated using electronic stopping powers derived from LSS and Bethe-Bloch models with results calculated using the electronic stopping powers provided by the TRIM-95 code. In both cases, the sum of the net displacement functions for each self-ion (pka) were compared to the results determined using the Monte Carlo code TRIM-95. The numerical results derived using the LSSI Bethe-Bloch electronic stopping powers were significantly higher than the net displacements calculated by TRTM-95. The numerical solutions calculated using the TRIM-95 electronic stopping powers were in closer agreement to the full cascade TRIM-95 calculations. 11:10 am DEVELOPING SI POTENTIALS FOR MODELLING ION IMPLANTATION: David F. Richards, James B. Adams, Jing Zhu, Lin H. Yang, Christian Mailhiot, Department of Materials Science and Engineering, University of Illinois, 105 S. Goodwin Avenue, Urbana,lL 61801 We have used the Force Matching Method to generate a new interatomic potential for silicon which is fit to both experimental properties and a quantum mechanical database of force and energy data for many atomic configurations. These configurations include surfaces, point defects, clusters, bulk solids (crystalline and amorphous), and dimers at very close separations. We discuss the reliability of this new potential and applications to modeling damage cascades produced by ion implantation.

STRUCTURE AND MORPHOLOGY OF EPITAXIAL THIN FILMS SESSION VII: Polycrystalline Films Sponsored by: EMPMD Thin Films & Interfaces Committee Program Organizer: Dr. David E. Jesson, Oak Ridge National Laboratory, P.O. Box 2008, Oak Ridge, TN 37831-6030 Thursday, AM February 8, 1996

Room: Orange County 4 Location: Anaheim Marriott Hotel

Session Chairperson: David J. Srolovitz, Department of Materials Science and Engineering, The University of Michigan, 2300 Hayward, Ann Arbor, Michigan 48109

8:30 am Invited INTERFACIAL FORCE MICROSCOPY MEASUREMENTS OF THE NANOMECHANICAL PROPERTIES OF MATERIALS, 1. E. Houston, T. A. Michalske, Sandia National Laboratories, Albuquerque, NM 87185-1413 Scanning probe techniques have shown dramatic growth over the last several years and are presently making significant impact on surface and interfacial problems in material science. Interfacial Force Microscopy (IFM) utilizes a self balancing force sensor which permits stable operation in both the attractive and repulsive modes with effectively a zero compliance. In recent studies, we have taken advantage of these unique sensor properties in applications to the study of the mechanical properties of materials at the nano-meter scale. In this presentation, we show results demonstrating the enhanced capabilities of the IFM in application to studies of the elastic and plastic deformation of singie nano-scale grains in a polycrystalline films of Au as well as singie-crystal Au surfaces. We show examples of superplastic behavior in polycrystalline films, the role of tip radius in nucleating dislocations leading to plasticity, and the effect on the measured nanomechanical properties of varying the crystalline surface for the singie-crystal work. This work is supported by the US Department of Energy, Office of Basic Energy Sciences, Division of Material Sciences under contract

1996 TMS ANNUAL MEETING AND EXHIBITION

9:00 am Invited A FINITE ELEMENT APPROACH TO SIMULATING EVOLVING MICROSTRUCTURES IN THIN FILMS: Z. Suo, Mechanical and Environmental Engineering Department, University of California, Santa Barbara, CA 93106 This presentation describes a class of simulation methods for structural evolution in materials. The method is formulated on the basis of an integral form of the nonequilibrium thermodynamics, uses finite elements to model structural features (e.g., grain boundaries), and results in a set of ordinary differential equations that govern the time evolution. The method has been used to simulate diverse phenomena, including grain growth, surface diffusion, and ferroelectric domain evolution. This presentation uses epitaxial grain growth in thin films as example to illustrate the method. When a polycrystalline film is heated, the grains with small free energies enlarge at the expense of the grains with larger free energies. The process may eventually produce a single crystal film. A number of events, however, can happen to prevent the film to evolve to a singie crystal. For example, surface grooving can break the film into droplets. The finite element method is used to study the competition among these and other kinetic processes. 9:30am EPITAXIAL DIAMOND FILMS ON SILICON (001) SURFACE: Chonglin Chen, Gary S. Song, Terence E. Mitchell CMS, MS-K765, Los Alamos National Laboratory, Los Alamos, NM 87545; Byungyou Hong, Robert Collins, Russel Messier Materials Research Laboratory, Pennsylvania State University, Pennsylvania, PA 16802 The nucleation and growth behavior of nanocrystalline diamond thin films on Si substrate, prepared by plasma-enhanced chemical vapor deposition in a ~2/~ mixture gas atmosphere, has been studied using high-resolution transmission electron microscopy (HRlEM). Results show that an amorphous interiayer of -150 Awas formed prior to diamond nucleation. Local epitaxial growth of nanocrystalline diamond grains in the thin films on Si (001) surface has been observed. HRlEM images reveal that the epitaxial orientation relationship follows either a 3:2-match cube on cube arrangement with an _7° tilt and a small angle rotation between the diamond lattice and the substrate lattice, or a I: I-match with (lII)diamon&(OIl)si and [OOI]diamond/[OIl]si' 9:50am THIN FILM TEXTURES OF Au,Ag, Pd,AND Au-Pd COUPLES: Young S. Chung, Keenan L. Evans, Motorola, 5005 E. McDowell Rd. POO4, Phoenix, AZ 85008 The crystal textures of polycrystalline films of Au, Pd, Ag and Au-Pd couples before and after annealing at 400°C were investigated via the x-ray diffraction pole figure method. The extent of a specific plane texture exhibited a significant dependence on metallic species and composition. Interaction between the substrate surface and condensed metals during the initial nucleation process is important in determining the ultimate film texture. The {III) texture formation on Au-Pd thin film couples displayed a strong dependence on the nature of the underlying seed layer. Gold films deposited on a palladium seed layer revealed much less (Ill) texture, than Au films deposited directly on a silicon dioxide surface. On the contrary, Pd films deposited on polycrystalline Au films showed higher degree of {III) texture, compared to Pd films deposited directly on silicon dioxide. Annealing the films greatly enhanced the degree of {III ) texture, and the order of deposition affects the texture of the Au-Pd alloy films after annealing 10:10 am BREAK DEFECTS III Session Chairperson: W. W. Gerberich, Department of Chemical Engineering and Materials Science, University of Minnesota, Minneapolis, MN 55455 10:30 am GROWTH AND CHARACTERIZATION OF ALUMINUM FILMS ON SAPPIDRE (0001): K. F. McCarty, D. L. Medlin, R. Q. Hwang, P. B. Mirkarimi, S. E. Guthrie, N. R. Moody, M. I. Baskes, Sandia National Laboratories, Livermore, CA 94551 We are studying aluminum films on a-alumina as a model system for film/substrate effects in electronics packaging. In particular, we are investigating the texture and possible epitaxy of aluminum films grown on sapphire (0001) surfaces. Aluminum films are deposited on sapphire (0001) surfaces by both an energetic process (pulsed laser deposition) and a thermal process (evaporation). The film structure is characterized both in situ (reflection high-energy electron diffraction) and ex situ (high-resolution and conventional transmission electron microscopy and atomic force microscopy). The observed interface structure is compared to predictions of semi-empirical, atomistic calculations. The work of adhesion of the films to the substrate is determined by micro-scratch testing. We emphasize the relationships between the conditions of growth, the structure of the interface, and the resulting mechanical properties. This work was performed under U.S. DOE contract DE-AC04-94AL85ooo. 10:50 am HETEROGENEOUS GENERATION OF STACKING FAULTS IN SEMICONDUCTORS:, Y. Chen, S. Ruvimov, J. Jasinski Z. Liliental-Weber, and J. Washburn, Lawrence Berkeley National Laboratory, University of California, Berkeley, CA 94720 In our experiments, stacking faults were observed frequently in semiconductors in which there was nonstoichiometry. For example, in low-temperature grown, and in annealed As-implanted GaAs, a high density of stacking faults have been obaerved. An atomic model is proposed to explain that a stacking fault is induced in this case by the excess As during crystal growth. In ZnSe epilayers grown on GaAs SUbstrates, stacking faults were also frequently observed, and the density of the stacking faults is influenced by the different types of interfacial structures (SeZnlAsGa or ZnSelGaAs). The different stacking fault densities are explained by the local bonding between GalSe or 'UlJAs. Different densities of stacking faults were also observed in GaN bulk crystal for growth along [0001] compared to growth along [30- 1]. In this case too the formation of stack-

B139

ing fault appears to be influenced by excess Ga. In general. the existence of nonstoichiometty. whicb can distort the local atomic bonding. seems to induce the formation of stacking faults during crystal growth. 11:10 am GROWTH AND RELAXATION OFHETEROEPITAXIALSII_x_ GexCy FILMS ON SI(IOO): Harald Jacobsson. Nicole Herbots. Sean Hearne. Joan !;li;g. Peibua Yeo Department of Pbysics and Astronomy. Arizona State University. Thmpe. AZ 85287 The new group IV-IV-IV ternary semiconductor Sil_x_>:GexCy offers new possibilities for strain modification and bandgap engineering on a Si subStrate. In this work the growth kinetics and onset of relaxation for a cubic diamond. solid solution of Sil_x_ yGexcy has been investigated. Epitaxial quality and defect nucleation has been studied using Rutherford backscattering combined with ion cbanneling (IC) in different axial directions. transmission electron microscopy (TEM) and infrared spectroscopy (lR). The results sbow that the films are epitaxial. The crystalline quality is fair at low carbon concentration but the defect density increases with increasing carbon concentration. The energy dependence of the minimum yield. combined with lEM observations. sbows that dislocations dominate at lower carbon concentrations. whereas stacking faults dominate at higber carbon concentrations. Both IC and lR spectra indicate an increase in bondlength distribution with increasing carbon concentration. These observations suggest that carbon is detrimental to the elastic properties of Si 1_ xGex causing relaxation and defect nucleation wben carbon concentrations exceed 2%. 11:30 am LATTICE MISMATCH INDUCED ROUGHENING IN THE MBE GROWTH OF NbxTII-XOz on TIOz (110) and (100): Y. Gao. S. A. Chambers. Environmental Molecular Sciences Laboratory. Pacific Northwest Laboratory. P.O. Box 999. MS K2-12. Ricbland. WA 99352 We have grown single-crystal. mixed rutile phases of NbxTIl_x02 on TI02 (110) and (100) by molecular beam epitaxy for x ranging from 0.02 to 0.50. The resulting epitaxial films have been characterized by means ofRHEED. LEED. UPS. XPS. XPD. AFM. and lEM. From the resulting analyses. we have determined that Nb goes in substitutionally to cation sites in the rutile lattice. displaces TI. and assumes a +4 oxidation state. Surface roughening occurs as x increases for both crystal orientations. However. significantroughening occurs at x =0.15 on (100) compared to a value ofx =0.30 for (110). This difference comes about because of the numerical values of the TI-O and Nb-O bond lengths. The irregular octrabedral M-O bond lengths. R1 and R2. differ by 1% and 12% in going from TI02 to Nb02. respectively. Significantly. only half the metal atoms in a given layer of the (110) growth surface have in-plane components of R2 wbereas every metal atom in the (100) growth plane has sucb components. Thus. there is a substantially larger in-plane lattice mismatcb wben the growth surface is (100) compared to (110). Therefore. misfit dislocation nucleation and subsequent step formation at the surface occurs for a lower value of x on (100). leading to surface rougbening. Pacific Northwest Laboratory is operated for the U.S. Department of Energy by Battelle Memorial Institute under contract De-Ac06-76RLO 1830. 11:50 am EVOLUTION OF MISFIT DISLOCATION NETWORK AND MORPHOLOGY IN OMVPE GROWN (001) ZnSe/GaAs: S. Ruvimov. E. D. Bourret. J. Wasbburn. Y. Cben. W. Swider. and Z. Liliental-Weber. Lawrence Berkeley Laboratory. Berkeley. CA94720 Transmission electron microscopy and X-ray diffraction techniques bave been applied to study strain relaxation and the evolution of the dislocation structure and morpbology of ZnSe epUayers grown by low pressure OMVPE on a (001) surface of semi-insulating GaAs. Before the ZnSe growth. the substrate surface was exposed at 6S0·C to a flow of tertiarybutyl arsine to form an As-terminated surface. This surface treatment is sbown to result in a drastic decrease in the SF density. For the first time formation of (f;f' misfit dislocations was observed at a layer thickness of 0.05 IUn. This agrees well with the theoretical critical thickness expected for misfit dislocation formation in the ZnSeJGaAs gystem. but is mucb lower than the experimental critical thickness of 0.15-0.2 11m reported earlier. Various mechanisms of dislocation generation were observed at different growth stages. The evolution of dislocation structure is discussed in relation with layer morpbology.

B140

1996 TMS ANNUAL MEETING AND EXHIBITION