Tundish Technology for Casting Clean Steel: A Review YOGESHWAR SAHAI With increasing demand of high-quality clean steel, cleanliness is of paramount importance in steel production and casting. Tundish plays an important role in controlling the continuously cast steel quality as it links a batch vessel, ladle, to a continuous casting mold. Tundish is also the last vessel in which metal flows before solidifying in mold. For controlling the quality of steel, flow and temperature control of the melt are critical, and these are presented in this paper. Use of proper flux, design of flow control devices, and gas injection in tundish become important factors in casting clean steel. Recycling of hot tundish, centrifugal flow tundish, H-shaped tundish, etc. are some of the developments which were implemented to cast clean steel and these are discussed. DOI: 10.1007/s11663-016-0648-3  The Minerals, Metals & Materials Society and ASM International 2016

CONTINUOUS casting of steel is a widely used process and an important step in steel production. Over 90 pct of the steel produced in the entire world is continuously cast. At the same time, the quality requirements of steel are also increasing. Thus, steel cleanliness and strict composition control are now becoming the primary concerns of steelmakers. Tundish is the last metallurgical vessel through which molten metal flows before solidifying in the continuous casting mold. During the transfer of metal through the tundish, molten steel interacts with slag, refractories, and the atmosphere. Thus, the proper design and operation of a tundish are important for delivering steel of strict composition and quality. Two to three decades of the last century have seen major advances in tundish technology for clean steel production. Growth in tundish technology has slowed down as it is becoming a mature technology now and as a result, there has been slow advance in this century or so. Some of the technological advancements are discussed in this paper. The term quality of steel refers to steel which has strict composition control and small number of evenly distributed small-sized non-metallic inclusions. High-quality clean steel is necessary to meet the stringent requirements of better mechanical properties, particularly ductility and durability of steel. Non-metallic inclusions include oxides, sulfides, nitrides, carbides, and their compounds or composites. Sulfides, carbides, and nitrides precipitate under normal conditions during cooling of steel below solidus temperatures. Small particles of particular oxide inclusions, sulfides, carbides, and nitrides, have been utilized to control

YOGESHWAR SAHAI, Professor Emeritus, is with the Materials Science & Engineering Department, The Ohio State University, Columbus, OH 43210. Contact e-mail: [email protected] Manuscript submitted October 30, 2015. METALLURGICAL AND MATERIALS TRANSACTIONS B

microstructure for improving steel properties. However, most of the large oxide inclusions and some sulfide inclusions form while the steel is in liquid state. If these large inclusions are not removed before solidification, they give rise to processing difficulties and degrade product quality. Sulfide inclusion can be minimized by desulfurizing melt in ladle prior to its transfer to the tundish. Oxide inclusions are of two types; exogenous and indigenous. Exogenous inclusions form either by reoxidation of the deoxidized and refined melt by air and/or oxidizing slag, or by entrainment of slag or refractory particles. Indigenous inclusions form in ladle from reactions between dissolved oxygen and deoxidizers such as aluminum or silicon. Indigenous inclusions are much smaller in size and are less harmful unless they agglomerate and form larger inclusions. Indigenous inclusions are generally smaller than ~50 lm in diameter, while exogenous inclusions are larger than this size. Functions of a tundish include delivery of clean melt to different strands at required flow rate, temperature, and composition. So in terms of quality improvements, a tundish is required to prevent the generation of large inclusions of exogenous origin; prevent the formation of large agglomerates of indigenous inclusions; and remove any remaining inclusions during the melt transfer through the tundish. To achieve these objectives, the melt transfer must avoid reoxidation by ambient air and ladle slag. Transfer of ladle slag into the tundish, and tundish flux into the mold must be avoided to prevent their emulsification and entrainment into the melt. Detailed discussion of non-metallic inclusions and clean steel technology can be found in a book.[1] Zhang and Thomas[2] published state-of-the-art review of evaluation and control of steel cleanliness. Over 20 methods of evaluation of steel cleanliness are discussed in this paper. The paper also reviewed operating practices to improve steel cleanliness.

I.

REOXIDATION DURING LADLE-TO-TUNDISH MELT TRANSFER AND SLAG ENTRAINMENT

The pre-eminence of the reoxidation phenomenon in the generation of large-sized inclusions can be seen from Figure 1. Ohno et al.[3] found that metal stream reoxidation increased the amount of large inclusions by a factor of 2.5 between the ladle and the tundish, and that the reoxidation products were bigger than 100 lm in size. The teeming stream can be protected from the oxidizing atmosphere by physical shrouding where the melt stream is enclosed in a refractory pouring tube or in an envelope of Ar gas. Figure 2[4] schematically shows argon gas shrouding inside of a refractory tube and the use of long-nozzle melt stream shrouding techniques. The oxidizing components, FeO, MnO, and SiO2, in ladle slag carried over to the tundish react with aluminum in steel and form alumina clusters in steel slabs. Part of the slag transferred from the ladle to tundish ends up in the mold which may get trapped in the solidifying shell, leading to the formation of macro inclusions or slag spots. Slag detection devices can be used at the ladle-to-tundish outlet that can detect the onset of slag transfer from the ladle to the tundish. An electromagnetic method in which the slag sensor electromagnetically detects slag in the pouring streams is very efficient and is commonly employed in steel industry. This was first developed by a company named AMEPA.[5] The AMEPA slag detection sensor is now being used extensively in many steel companies. The sensor signal can directly operate the slide gate nozzle, and thus, can minimize slag carryover. Figure 3 shows the slag signal as detected by the AMEPA sensor.[6] The figure demonstrates that it takes less than 2 seconds for the sensor to activate the closing of slide gate nozzle aperture.

Fig. 1—Relative contribution of various factors causing inclusions in steel.[3]

II.

EFFECT OF TUNDISH SIZE

Tundish size has been found to have a significant effect in improving the quality of cast steel. Increasing the tundish size is the most obvious way of increasing the average residence time of steel in the tundish. For a given casting rate, which determines the volumetric flow rate of the melt through the tundish, a larger volume results in a longer average residence time in the tundish. Increasing the volume of the tundish leads to a dramatic decrease in the number of macro inclusions (alumina clusters) obtained in the cast product particularly at ladle change. The results of a plant trial with a larger tundish by Tozaki et al.,[7] where the tundish capacity was increased from 65 to 85 T, are shown in Figures 4 and 5. The number of alumina clusters decreased during both the steady-state and non-steady-state casting operations. Another benefit of increasing the depth of the tundish was in maintaining the casting rate without any decrease during ladle change. Thus, the productivity and quality both were improved by increasing the depth of the tundish. Similar plant trial results made at Kobe steel were published by Ishikura et al.[8] In this study, the performance of an old caster (No. 3) with a 50-T tundish is compared with their new caster (No. 4) with an 80-T tundish. They found that the caster with the 80-T tundish can operate at 2.0 m/minute casting speed and cast the same quality of slabs as the caster with the 50-T tundish at 1.4 m/minute speed. Thus, the productivity of the caster was increased without sacrificing the quality. They also found that the quality of the slabs cast during the ladle change period was significantly improved with the 80-T tundish caster, as shown in Figure 6.

III.

EFFECT OF FLOW CONTROL DEVICES

Many attempts have been made to improve melt flow characteristics in existing tundishes by the installation of various flow control devices, such as weirs, dams, baffles with holes, pour pads, and turbulence suppressers within the tundish. The beneficial effects of various flow modification devices have been borne out by actual industrial trials as well as physical and mathematical modeling studies (e.g., References 9 through 18). Majumdar and Guthrie[19] reviewed physical and mathematical modeling of continuous casting systems. A review of modeling of tundish operations was later published recently by Chattopadhyay et al.[20] Optimum placement of dams and weirs has been found to result in an increase in the average residence time of fluid as well as an increase in the plug flow volume in the tundish. These flow control devices, properly installed, may create localized mixing in contained regions, which may help in inclusion agglomeration and hence their removal. The metal stream from the ladle, especially when it is shrouded and has no entrained gas, enters into the tundish at a very high velocity and turbulence. The impact of the stream at the tundish bottom may cause severe refractory erosion problems. Pour pads are designed and placed at the bottom to withstand the METALLURGICAL AND MATERIALS TRANSACTIONS B

Fig. 2—Ladle stream protection by (a) Ar gas shrouding and (b) long nozzle.[4]

Fig. 3—Slag carryover signal detected by AMEPA sensor.[6]

Fig. 5—Effect of bath depth on size of alumina clusters.[7]

Fig. 4—Effect of bath depth on alumina clusters.[7]

erosive force of the ladle stream. They are made of very dense, chemically stable refractory material. Corrugated impact pads were found to be effective in reducing the turbulence of the incoming stream. A pour pad design which appears to reduce the stream turbulence significantly has been reported by Bolger and Saylor.[21] Their pour pad and associated fluid flow are shown schematically in Figure 7. The plunging metal stream, after diverging back to the free surface, becomes quiescent. Such surface-directed flow is considered favorable for inclusion flotation. Of course, the surface-directed flow

METALLURGICAL AND MATERIALS TRANSACTIONS B

Fig. 6—Quality improvement in 80 T caster.[8]

depends upon the original shape of the pad which may gradually change with time due to refractory erosion. An interesting approach was recently studied by Morales-Higa et al.[22] in which ladle shroud was used as a flow control device. A new design of ladle shroud was developed and studied in a one-third scale water model.

depths of bubble column penetration by 5 to 15 pct. Extensive modeling work covered studies of fluid flow, determination and optimization of RTD, and effects of FCD on flow pattern in tundish. Role of thermal effects on flow patterns and inclusion removal were also modeled. However, sag-melt interactions were not modeled earlier. Solhed et al.[26,27] successfully considered slag–steel interactions in their modeling efforts.

IV.

Fig. 7—Schematic of a pour pad and associated melt flow patterns.[21]

A computer modeling study by Mukhopadhyay et al.[23] to optimize the placement of flow control devices is summarized here. They studied the effects of different flow control devices in tundishes on flow and turbulence profiles and finally on inclusion flotation. They modeled flow in a tundish with a baffle with holes, a pour pad, or a pour pad with a dam. These three tundish configurations are shown in Figure 8. Top-in and top-out in Figure 8 and later in Figure 11 refer to inclusions floated to the top surface on the inside and outside, i.e., upstream and downstream of the baffle location. Velocity vectors colored by velocity magnitude and turbulent kinetic energy iso-surfaces in the tundish with a baffle are shown in Figures 9 and 10. Similar predictions were also made with the other two configurations. The authors studied the flotation of 10- to 500-lm-sized inclusions. Figure 11 shows the percentage of inclusions which floated to the top surface and escaped through the outlet to the cast steel. It can be seen that inclusions greater than 100 lm were completely trapped at the top surface and did not exit through the tundish outlet. Further analysis of the data in Figure 11 was carried out to show the statistical bounds of the residence time data, which is shown in Figure 12. It shows that very small inclusions are subject to maximum influence of fluid turbulence and that the residence time varied from 1200 to 150 seconds. The residence time variation decreases with increasing particle size, and becomes indistinguishable at sizes greater than 200 lm. The unsteady behavior of inclusions in a turbulent metal bath was very well simulated in this study. Thus, mathematical modeling using CFD can be very effectively used for tundish design analysis and optimization. Odenthal et al.[24] physically and numerically modeled flow and RTD in water model tundish and validated the results by LDA and DPIV measurements. Tundish filling and ladle change operations were simulated by volume of fluid (VoF) model. This multiphase model successfully predicted gas bubbles motion and transient surface waves during tundish filling. Chattopadhyay et al.[25] modeled inert gas shrouding in a tundish using DPM approach. The mathematical model was found to over-estimate the

GAS INJECTION IN TUNDISH

Several water modeling studies and plant trials (e.g., References 28 through 31) have demonstrated the beneficial effects of Ar gas injection in a tundish. Yamanaka et al.[28] studied the effects of Ar gas injection in one arm of a V-shaped tundish. Double dams were installed in both arms of the tundish, but only one side was equipped with a porous plug for Ar gas injection as shown in Figure 13. The results show that the number of large inclusions was reduced, while the number of small inclusions was found to increase. The index of ultrasonic defects that represent the number of macro inclusions reduced significantly. In a study at CRM,[29] a gas bubbling device was embedded in the porous refractory at the bottom of a tundish (Figure 14). They also used a water model of the system to study the fluid flow pattern and its effect on simulated inclusion flotation. Gas injection also increased the removal rate of simulated inclusions in the size range of 20 to 100 lm from 43 to 65 pct. Inclusions larger than 100 lm were being nearly completely removed even without gas injection. Ramos-Banderas et al.[32] mathematically modeled multiphase steel flow with gas bubbling in tundish using Eulerian–Eulerian model. Inclusion trajectories were modeled using Lagrangian particle-tracking approach. The results were compared with PIV measurements in a water model. Garcia-Hernandez et al.[33] summarized similar studies using discrete phase modeling (DPM) for particle tracking and studied vortex and short circuit flow effects on inclusion removal in a slab tundish.

V.

LARGE TUNDISH WITHOUT FLOW MODIFIERS

The benefits of implementing various flow modifiers, mentioned above, on the cleanliness of melt have been changing lately. Significant developments in secondary refining processes and their installation in many steel plants have achieved acceptable levels of steel melt cleanliness for demanding applications, especially during steady-state casting. Even with careful control of secondary refining, macro inclusions do occur during the non-steady state of casting. Large tundishes that offer sufficient residence time for macro inclusion flotation and that prevent vortexing at non-steady state have been found satisfactory for sequential casting of high-quality steels, even at a relatively high casting rate. In addition, flow modifiers are inconvenient to sustain hot cycle tundish practice, which has been found to be quite economical. A large tundish of simple design, METALLURGICAL AND MATERIALS TRANSACTIONS B

Fig. 8—Tundish with a baffle with holes (top), a pour pad (middle), and with a pour pad and dam (bottom).[23]

which can cast at a high throughput rate and sustain long sequential casting, would result in high productivity and high quality at low cost.

VI.

MELT TEMPERATURE CONTROL

Metal temperature in a tundish has a strong influence on the quality and properties of cast product, caster operation, and refractory life. A large ladle may pour metal into the tundish for as long as one hour at the caster. Temperature losses are always inherent in the system, specifically in the ladle and by the tundish from the melt surface and through the refractory walls. Thus, the temperature of the ladle-to-tundish stream METALLURGICAL AND MATERIALS TRANSACTIONS B

continuously changes. Coupled with the temperature loss in the tundish, temperature changes in the stream result in melt temperature variation within the tundish. Figure 15(upper)[34] is a result of continuous temperature measurement at the Bethlehem Steel Corporation. This also shows the classical dome-shaped temperature profile with a variation of about 20 K (20 C). Figure 15(lower)[34] shows the tundish melt temperature variation for casting of 4 ladles. ‘A’ in this figure represents the ladle change, and points to the fact that each ladle may start with a different temperature, and that the highest and lowest melt temperatures in a tundish may vary from ladle to ladle. The melt temperature in the tundish in each of these cases first rises and reaches a maximum value in about 20 minutes, followed by a gradual decrease for the

Fig. 9—Velocity vectors colored by velocity magnitude in the tundish with baffle.[23]

Fig. 10—Turbulent kinetic energy iso-surfaces in the tundish with baffle.[23]

Fig. 11—Distribution of inclusion particles on the top surface and in the outlet stream.[23]

Fig. 12—Distribution of particle residence time with a baffle.[23]

METALLURGICAL AND MATERIALS TRANSACTIONS B

Fig. 13—Gas injection in a V-shaped tundish.[28]

0 minute) and after 47 minutes of casting (i.e., near the end of the casting), respectively. Figure 17 shows flow profiles in two vertical planes (a) near the plane of symmetry and (b) near the back wall in symmetrical half of the tundish. In the beginning (Figure 17), incoming melt from ladle to tundish is much hotter than the melt in the tundish, so buoyancy aided with a dam in the tundish directs the steel to the free surface and a clock-wise flow pattern is formed. Flow patterns shown in Figure 18 are counterclock-wise because the incoming steel melt is cooler than the steel melt in the tundish after 47 minutes of casting. Even the presence of the dam is not able to direct melt flow to the surface. So toward the end of the casting, incoming steel melt is predicted to flow near the tundish bottom and directly goes to the outlet. This short-circuiting may result in less efficient inclusion flotation. A well-preheated, well-covered, and well-insulated ladle will have less stream temperature loss, and in turn, will affect the tundish melt temperature. Quality of steel depends on the melt superheat is shown in Figure 19. The effect of tundish steel temperature on the inclusion index and centerline segregation index in a billet is plotted in Figure 19.[36] Matsumoto et al.[37] also studied the influence of tundish melt superheat on the index of inclusion in cast slab. Their results are shown in Figure 20, which demonstrates that the inclusions increase at low as well as high superheat. Thus, it is important to maintain melt temperature in the tundish in a very narrow range close to the optimum temperature. Many companies have developed plasma or induction heating of melt in tundish, and they have been able to maintain the melt temperatures within a very narrow range.

VII. Fig. 14—Ar gas bubbling device embedded in the porous refractory.[29]

rest of the heat. Thus, at the end of a ladle, metal in the tundish is at its lowest temperature. Chakraborty and Sahai[35] developed mathematical models for coupling heat transfer and fluid flow in a typical ladle and a slab casting tundish over a complete casting sequence. Figure 16 shows melt temperature change in a tundish during 47 minutes heat from one ladle. Curve A in Figure 16 shows the predicted ladle stream temperature drop during 48 minutes of casting. The stream temperature drops by about 40 K (40 C) in this period. The rate of stream temperature drop is smaller in the beginning but increases toward the latter part of the casting. This is due to the smaller volume of metal in the ladle which suffers a relatively higher heat loss toward the end of the casting period. Curves B, C, and D represent predicted melt temperatures in the tundish at three different monitoring points. The melt temperature is predicted to rise to a maximum value in about 25 minutes, and is followed by a gradual decline. Predicted flow profiles in tundish are shown in Figures 17 and 18[35] at the start of a new heat (i.e., at METALLURGICAL AND MATERIALS TRANSACTIONS B

EFFECT OF CALCIUM ADDITION

Calcium addition in ladle and tundish has been successfully used by many steel companies. Calcium transforms large alumina inclusion clusters into liquid calcium aluminate inclusions, which after solidification and during subsequent hot and cold rolling are elongated and fragmented into smaller sizes which are not detrimental to steel properties. These liquid inclusions reduce nozzle clogging which may even cause nozzle blockage, and thus maintains the desired melt flow. Calcium also improves other mechanical properties of steel. Many industrial studies have reported benefits of calcium addition. For example, Yoshii et al.[38] at Kawasaki Steel found that the Ca-treated steel had a significantly reduced number of manganese sulfide inclusions as well as large oxide inclusions with the addition of 50 ppm Ca (shown in Figure 21). They also found that the number of inclusions in the transition slabs cast during ladle change was reduced, and the overall cleanliness was improved in the treated steel.

VIII.

RECYCLING OF A HOT TUNDISH

After a sequence casting, the removal of skull from a tundish sometimes damages the working lining. This

Fig. 15—Tundish metal temperature profile during normal casting sequence.[34]

Fig. 17—Steady-state profile in (a) vertical symmetrical plane which has ladle stream and outlet to mold, and (b) in a plane closer to back wall.[35]

Fig. 16—Metal temperature variation in a tundish at four monitoring points.[35]

tundish lining repair is very expensive in terms of refractory cost and is a labor intensive task. So, many steel companies, especially in Japan, have adopted hot tundish recycling, pioneered by Kobe Steel.[39] In this operation, soon after one sequence casting, the tundish is tilted to pour out any remaining metal and slag in it, followed by an automatic change of the old pouring nozzle with a new and preheated nozzle, replacing the slide gate valve, and repairing any refractory damage. Finally, the tundish may be preheated to a desired temperature before returning for another casting sequence. In a hot tundish recycling operation, it is not necessary to perform all the above steps. The

Fig. 18—Unsteady-state profile developed after 47 min of teeming in (a) vertical symmetrical plane which has ladle stream and outlet to mold, and (b) in a plane closer to back wall.[35]

preheating step is generally avoided as it causes oxidation of any remaining metal in the tundish, which contributes significantly to the formation of non-metallic inclusions in the next sequence. Refractory repairs are only performed if they are deemed necessary. As METALLURGICAL AND MATERIALS TRANSACTIONS B

Fig. 21—Effect of calcium treatment on large inclusions.[38] Fig. 19—The effect of melt temperature on the inclusion index and centerline segregation index.[36]

super-clean steel. It can be seen that ladle-to-tundish melt transfer has a slag detector and the melt oxidation is prevented by a long nozzle and nozzle shroud. The tundish has a tight seal cover and is filled with argon gas. Melt is also covered by tundish flux. Residence time of the melt is increased by large tundish capacity and by employing flow control devices. Pour pad and dam also provide surface-directed flow which helps in inclusion flotation. Induction heating in tundish may be employed to correct melt temperature. Argon bubbling in tundish to mold SEN prevents nozzle clogging. The tundish is lined with a high-quality refractory which is coated with magnesia. This system is shown to cast super ultra-low carbon (SULC) steel of high quality. In general, a tundish may not employ all these devices at the same time, but one may select the devices based on the type of steel being cast and the quality requirements.

X. Fig. 20—The influence of tundish melt superheat on the index of inclusion in cast slab.[37]

Figure 22 shows, therefore, the tundish may return to operation after pouring the remaining slag off, or it may return after the slide gate valve change, or it may go for refractory repair and preheating before returning for the next casting sequence. The hot recycle process has reduced 8 hours of tundish preparation time for conventional process to only 25 minutes, and reduced refractory cost to less than one tenth.

IX.

CONTINUOUS CASTING SYSTEM FOR HIGH-QUALITY STEEL CASTING

Figure 23 by Okimori[40] shows a continuous casting system i.e., ladle, tundish, and mold, in which typical devices and procedures are shown which help in casting

METALLURGICAL AND MATERIALS TRANSACTIONS B

NON-TRADITIONAL TECHNOLOGIES

The H-shaped tundish is not a new technology, but it is a unique one. The tundish has traditionally focused on preventing the clean ladle melt from forming exogenous macro inclusions as a result of reoxidation and slag emulsion particularly during the transient periods of operation by (1) Separating the inlet and outlet compartments of the tundish by a refractory wall, and connecting the two compartments by a tunnel placed through the bottom of the wall; and (2) Making overlapping teeming of the melt into the tundish possible from the preceding ladle and the succeeding ladle simultaneously, via a long nozzle without decreasing bath depth in the tundish during the ladle change. The H-shaped tundish was later equipped with DC-arc plasma heaters as shown by Kimura et al. in Figure 24[41] to produce more demanding clean steels,

Fig. 22—Recycling operation of a hot tundish (SGV: Slide Gate Valve, TD: Tundish).[39]

Fig. 23—Continuous casting system (ladle, tundish, and molds) with all devices and procedures for casting clean steel. (SULC: Super ultra-low carbon).[40]

such as deep drawn and ironing plastic film-coated steel sheet. DC plasma heater made it possible to control the tundish melt temperature within ±5 K or ±5 C even during the ladle change, which in turn reduced clogging of the tundish and SEN. The quality of cast steel is significant better compared to traditional casting.

In centrifugal tundish, steel melt is rotated in first part of the tundish by application of a magnetic field. Centrifugal force acts differently in inclusions and steel melt which causes inclusions to separate from the melt. It was implemented in production at Chiba Works of Kawasaki Steel Corporation (later JFE Steel) first with a METALLURGICAL AND MATERIALS TRANSACTIONS B

10-T CF Tundish to confirm the results obtained earlier with an experimental 600 kg CF Tundish. It was later scaled up to a 30-T CF tundish that consisted of a 7-T rotation chamber (inlet compartment) and a 23-T rectangular chamber (outlet compartment) as shown by Miki et al. in Figure 25.[42] Aluminum-killed stainless steel melt or high-carbon steel melt (S45C) was poured from a 180-T ladle into the rotation chamber where it was rotated by a progressive magnetic field generated by a semi-cylindrical linear motor-type electromagnetic stirrer attached outside the compartment. The two compartments were connected by a through hole placed

at the bottom. Production data indicated that macro inclusions during the ladle change of Al-killed and VOD-treated ferritic stainless steel melts (SUS430) decreased almost by 50 pct by rotating the melt at 45 rpm. Alex McLean[43] very wisely reminded the steel industry that a turbulent tundish if not used properly could act as a contaminator rather than continuous refiner.

XI.

SUMMARY AND CONCLUSIONS

Some important aspects of tundish technology for controlling quality of the cast metal are presented in this paper. For cleaner steel casting, the following aspects are considered important.

Fig. 24—Schematic view of a 65-ton H-shaped tundish with two 2.35MW DC transfer Ar-N2 plasma torches installed at the outlet compartments.[41]

(1) Melt reoxidation by oxygen in air or by oxidizing slag should be prevented or minimized during its flow from tundish to mold; (2) The ladle-to-tundish metal stream should have in a long nozzle or enclosed in a ceramic shrouding pipe with Ar gas injection; (3) Slag transfer from the ladle to the tundish and from the tundish to the mold should be prevented by slag sensing and transfer prevention technologies; (4) A larger and deeper tundish with a covered top lid filled with argon gas atmosphere or covered by tundish flux should be used; (5) Calcium alloy addition helps in inclusion modification; (6) Use of flow control devices depending upon the situation may provide the desired melt flow; (7) Proper melt temperature control by induction or plasma heating produced high-quality steel;

Fig. 25—Schematic view of a 30-T centrifugal flow tundish (CF Tundish) with a semi-cylindrical progressive magnetic field stirrer on a 7-ton inlet compartment (rotation chamber).[42]

METALLURGICAL AND MATERIALS TRANSACTIONS B

(8) Hot tundish recycling reduces refractory cost, saves energy, and casts cleaner steel; and (9) H-shaped tundish and centrifugal tundish are two novel technologies for casting high-quality steel.

ACKNOWLEDGMENT The author gratefully acknowledges contributions made by Dr. Toshihiko Emi, coauthor of the book[1], which has been extensively used in preparing this manuscript.

REFERENCES 1. Y Sahai and T Emi: Tundish Technology for Clean Steel Production, World Scientific Publishers, Hackensack, 2008. 2. L.F. Zhang and B.G. Thomas: ISIJ Int., 2003, vol. 45, pp. 271–91. 3. T. Ohno, T. Ohashi, H. Matsunaga et al.: Trans. ISIJ, 1975, vol. 15, pp. 407–16. 4. Y. Shirota: Nishiyama Kinen Kouza, NMS-ISIJ, 1990, 143–144, p. 167. 5. E. Julius: Stahl u. Eisen, 1987, vol. 107, pp. 397–402. 6. A. Sato, H. Nakajima, T. Sakane et al.: CAMP-ISIJ, 1991, vol. 4, p. 1293. 7. Y. Tozaki, T. Hirata, A. Satoh, K. Sekino, et al.: Steelmaking Conference Proceedings, ISS AIME, 1993, pp. 377–82. 8. T. Ishikura, T. Saito, T. Yasui, K. Matsuo, et al.: 9th PTD Conference Proceedings, ISS AIME, 1990, pp. 115–21. 9. K.-H. Tacke and J.C. Ludwig: Steel Res., 1987, vol. 58, pp. 262– 70. 10. A.K. Sinha: Ph.D. Thesis, The Ohio State University, USA, 1990. 11. F. Kemeny, D.J. Harris, A. McLean, et al.: 2nd Process Technology Conference, ISS-AIME, 1981, pp. 232–45. 12. D.J. Harris and J.D. Young: Steelmaking Conference Proceedings, ISS-AIME, 1982, 65, pp. 3–16. 13. J. Knoepke and J. Mastervich: Steelmaking Conference Proceedings, ISS-AIME, 1986, vol. 69, pp. 777–88. 14. M. Schmidt, T.J. Rosso, and D.J. Bederka: Steelmaking Conference Proceedings, ISS-AIME, 1990, vol. 73, pp. 451–60. 15. M.L. Lowry and Y. Sahai: Steelmaking Proceedings. I.S.S. Publication, 1989, vol. 72, pp. 71–79. 16. J.D. Dorricott, L.J. Heaslip, and P.J. Hoagland: Tundish Metallurgy, ISS-AIME Publication, II, 1991, pp. 71–76. 17. T. Emi and Y. Habu: Proc. Phys. Chem. and Steelmaking, S.F.M., IRSID, ATS, 1978, pp. 126–31. 18. H. Nakajima, F. Sebo, S. Tanaka et al.: Steelmaking Conference Proceedings, ISS-AIME, 1986, vol. 69, pp. 705–16.

19. D. Majumdar and R.I.L. Guthrie: ISIJ Int., 1999, vol. 39, pp. 524– 47. 20. K. Chattopadhyay, I. Mihaila, and R.I.L. Guthrie: ISIJ Int., 2010, vol. 50, pp. 331–48. 21. D. Bolger and K. Saylor: Steelmaking Conference Proceedings, ISS-AIME, 1994, 77, pp. 225-33. 22. K Morales-Higa, RIL Guthrie, and M Isac: Metall. Mater. Trans. B, 2013, vol. 44, pp. 63–79. 23. A. Mukhopadhyay, H.L. Gilles, and B. Kocatulum: Electric Furnace Conference, I.S.S. Publication, 2002, vol. 60, pp. 343–51. 24. HJ Odenthal, R Bolling, and H Pfeifer: Steel Res. Int., 2003, vol. 74, p. 44. 25. K. Chattopadhyay, M. Isac, and R.I.L. Guthrie: ISIJ Int., 2011, vol. 51, pp. 573–80. 26. H. Solhed, L. Jonsson, and P. Jonsson: Metall. Mater. Trans. B, 2002, vol. 33B, pp. 173–85. 27. H. Solhed, L. Jonsson, and P. Jonsson: Steel Res. Int., 2008, vol. 79, pp. 348–57. 28. H Yamanaka, T Terajima, K Nakada, T Koshikawa, N Ueda, and Y Yoshii: Tetsu-to-Hagane, 1983, vol. 69, p. S213. 29. C. Marique, A. Dony, and P. Nyssen: Steelmaking Conference Proceedings, ISS-AIME, 1990, vol. 73, 461–67. 30. T. Saeki, H. Tsubakihara, A. Kusano, K. Umezawa, and I. Suzuki: Tetsu-to-Hagane, 1987, vol. 73, pp. A207–10. 31. M. Hanmyo, M. Ishikawa, Y. Ogura, S. Matsumura, S. Miyahara, and Y. Okubo: Tetsu-to-Hagane, 1987, vol. 73, pp. A215–18. 32. A. Ramos-Banderas, R.D. Morales, L. Garcia-Demecies, and M. Diaz-Cruz: ISIJ Int., 2003, vol. 43, pp. 653–62. 33. S. Garcia-Hernandez, J. de J. Barreto, J.A. Ramos-Banderas, and G. Solorio-Diaz: Steel Research International, 2010, vol. 81, pp. 453–60. 34. T.J. Russo and R.M. Phillippi: Steelmaking Conference Proceedings, 1990, vol. 73, pp. 237–46. 35. S. Chakraborty and Y. Sahai: Metall. Trans. B, 1992, vol. 23B, pp. 152–67. 36. I. Suzuki, S. Noguchi, Y. Kashiwakura, et al.: Steelmaking Conference Proceedings, 1988, vol. 71, pp. 125–31. 37. K. Matsumoto, Y. Hoshijima, K. Ishikura, et al.: The Sixth International Iron and Steel Congress, The Iron and Steel Institute of Japan Publication, 1990, vol. 3, pp. 222–29. 38. Y. Yoshii, Y. Habu, T. Emi et al.: Tetsu-to-Hagane, 1978, vol. 64, p. S626. 39. M. Maeda, T. Saito, K. Ebato et al.: CAMP-ISIJ, 1990, vol. 3, p. 199. 40. M. Okimori: Nippon Steel Technical Report, 1996, No. 361, pp. 67–76. 41. H. Kimura, A. Uehara, M. Mori, et al.: Shin-Nittetsu Giho (Nippon Steel Corp. Techn. Report, in Japanese), 1994, vol. 351, pp. 21–6. 42. Y. Miki, H. Shibata, N. Bessho et al.: Tetsu-to-Hagane, 2000, vol. 86, pp. 239–46. 43. A. McLean: 71st Steelmaking Conference, Toronto, Canada, 17–20 April 1988, pp. 3–23.

METALLURGICAL AND MATERIALS TRANSACTIONS B