ÓASM International 1059-9495/$19.00

JMEPEG DOI: 10.1007/s11665-014-0937-4

Hot Workability of CuZr-Based Shape Memory Alloys for Potential High-Temperature Applications Carlo Alberto Biffi and Ausonio Tuissi (Submitted October 17, 2013; in revised form February 28, 2014) The research on high-temperature shape memory alloys has been growing because of the interest of several potential industrial fields, such as automotive, aerospace, mechanical, and control systems. One suitable candidate is given by the CuZr system, because of its relative low price in comparison with others, like the NiTi-based one. In this context, the goal of this work is the study of hot workability of some CuZr-based shape memory alloys. In particular, this study addresses on the effect of hot rolling process on the metallurgical and calorimetric properties of the CuZr system. The addition of some alloying elements (Cr, Co, Ni, and Ti) is taken into account and their effect is also put in comparison with each other. The alloys were produced by means of an arc melting furnace in inert atmosphere under the shape of cigars. Due to the high reactivity of these alloys at high temperature, the cigars were sealed in a stainless steel can before the processing and two different procedures of hot rolling were tested. The characterization of the rolled alloys is performed using discrete scanning calorimetry in terms of evolution of the martensitic transformation and scanning electron microscopy for the microstructural investigations. Additionally, preliminary tests of laser interaction has been also proposed on the alloy more interesting for potential applications, characterized by high transformation temperatures and its good thermal stability.

Keywords

advanced characterization, intermetallics, non-ferrous metals, rolling

1. Introduction Since last years the development of high-temperature shape memory alloys (HTSMAs) has been growing, thanks to the great interest in many industrial fields, such as automotive, aerospace, mechanical, and control systems (Ref 1, 2). Among the different SMA systems with high martensitic transformation temperatures (Ref 3) (above 100 °C), CuZr-based one can be an interesting candidate (Ref 4). The binary system has been studied (Ref 3, 5) but very few indications in non-equiatomic alloys are reported (Ref 6). Other studies on the addition of some alloying elements, such as Cr (Ref 5), Al (Ref 7), Ni, Co, and Ti (Ref 8) have been discussed. One of the most evident problems of the system CuZr is the martensitic transformation stability during thermal cycling and the lack of information in the open literature about the production of semi-finished products, such as wires and thin tapes, probably because of their brittleness (Ref 9). For these reasons, the goal of this work is the evaluation of the hot workability properties of some representative CuZr-based This article is an invited paper selected from presentations at the International Conference on Shape Memory and Superelastic Technologies 2013, held May 20-24, 2013, in Prague, Czech Republic, and has been expanded from the original presentation. Carlo Alberto Biffi and Ausonio Tuissi, National Research Council, Institute for Energetics and Interphases, Unit of Lecco, Corso Promessi Sposi, 29, Lecco, Italy. Contact e-mail: carlo.biffi@ieni.cnr.it.

Journal of Materials Engineering and Performance

systems; some alloying elements were added to the binary system and the effect of their presence has been taken into account on the calorimetric and microstructural properties before and after the hot rolling. The effect of the hot working has been studied on the characteristic temperatures and the heat exchanged during short thermal cycling by the stabilization of the martensitic transformation. After the realization of the plates, preliminary results of laser interaction have been also proposed on the quaternary alloy, as representative of a potential alloy suitable for industrial applications, thanks to the high transformation temperatures and its good thermal stability.

2. Experimental Pure metals, Zr, Cu, Cr, Co, Ni, and Ti, were melted by means of a vacuum arc furnace (Leybold mod. LK 6/45) with a nonconsumable tungsten electrode. The furnace was equipped with a water-cooled copper crucible in order to avoid the contamination of the liquid pool. The nominal atomic compositions of the investigated alloys were the following: Zr50Cu45Cr5, Zr50Cu40 Co10, Zr50Cu32Ni5Co13, and Zr50Cu27Ni7Co9Ti7 (at.%). Small billets (30 mm in width, 80 mm in length, and about 100 g in weight) were prepared. The elements were melted in Ti-gettered high-purity inert atmosphere (Ar at 400 mbar pressure) in order to avoid the material oxidation. The obtained billets were six times re-melted, being flipped over after each melting step in order to ensure a high homogeneity in composition. After melting, the specimens were annealed at 850 °C for 3 h with final water quenching in water at room temperature. The billets were covered with a stainless steel can (1 mm thickness) under low vacuum for protecting from oxidation at

Fig. 1 Sample preparation for the hot rolling test: stainless steel for covering the small billet (a) and the enclose of the can under low vacuum for protecting from oxidation (b)

Table 1 Main characteristics of the fiber laser, used during the experiments Max. average power, W 1000

Max. frequency, kHz 5

Pulse duration range, ms 0.1-30

high temperature; moreover, the adoption of the protective can is also considered for a better control of the plastic deformation imposed during the process. The steps of the billet preparation and sealing are depicted in Fig. 1. The billets were hot rolled using the following procedure: after a first maintaining in furnace at 800 °C for 30 min for its homogenization and a preliminary hot forging (10% in thickness reduction); then, hot rolling at 800 °C with a thickness reduction of about 5% for each rolling step with intermediate annealing of 5 min. The global thickness reduction during the complete rolling procedure was about from 10 to 2 mm in 30 passes. At the end of the rolling, the tapes were water quenched and the stainless steel has been mechanically removed. A differential scanning calorimeter, Seiko model 520C, equipped with a liquid nitrogen cooling system and calibrated with a standard indium reference, was used to perform the calorimetric scansions. During the calorimetric measurements, the temperatures associated during thermal cycling (by the 10th cycle) were taken into the range between 0 and 450 °C using 10 °C/min heating/cooling rate and the weight of the samples was about 40 mg. Complete thermal scans were given by a first heating and then by the cooling. Characteristic temperatures and the corresponding transformation heats were evaluated for both direct/reverse transformations upon cooling/heating scans. The microstructure of the alloys was observed by means of scanning electron microscopy (SEM); quantitative analysis of the secondary phases was performed by means of energydispersive spectroscopy (EDS). Finally, preliminary test of laser machining on the quaternary alloy has been proposed using a continuous wave fiber laser (mod. YLR-1000 from IPG Photonics), whose the main characteristics are listed in Table 1. During the laser process, a flow of high-purity inert gas (N2 at pressure of 4 bar) was adopted as shielding gas to create a shielding atmosphere and to

Laser beam diameter, lm

Beam quality factor

42

< =1.05

Laser polarization Random

blow away the melted and partially vaporized material. Single spots (drilling ‘‘on the fly’’) were performed on the rolled plate of Zr50Cu28Ni7Co15 with a pulse duration of 10 ms and a power of 750 W.

3. Results and Discussion The external surfaces of the hot rolled ingots, after the removal of the protective can, are shown in Fig. 2. The final thickness of all the rolled tapes was about 2 mm, so the total thickness reduction was about 80%. A positive effect of the stainless steel can, able to protect the alloys from oxidation during high-temperature rolling, can be observed. The aspect of rolled tapes looks to be quite brittle; both the irregular morphology of both top and bottom surfaces and lateral fractures are indicative of the quite poor workability of the intermetallics based CuZr (Ref 9). It is clear that the integrity of the hot rolled samples is very poor for the ternary alloys (see Fig. 2a); it was observed in a previous work that the addition of 5% of Cr or 10% of Co was enough to improve the hot workability of this system, if compared with the binary one (Ref 9). It looks that higher Cu content is associated to larger irregularities on the rolled surfaces, because of higher hardness. Considering the quaternary alloy, the lateral edges are less irregular than in the case of the ternary alloys and the surface morphology appears to be more regular and uniform. The uniformity of the rolled tape is good; few cracks are laterally present only and both top and bottom surfaces are crack free. Similar result was obtained for the ZrCuNiCoTi alloy, even if that tape seems to show lower stiffness. After the cutting of the samples, metallographic analysis was done to evaluate the microstructure of the alloys. Figure 3

Journal of Materials Engineering and Performance

Fig. 2

Hot rolled tapes of the ternary CuZrCr (on the left) and CuZrCo (on the right) (a), quaternary CuZrNiCo (b), and quinary CuZrNiCoTi (c)

Fig. 3 SEM images of the microstructure of the CuZrNiCo alloy before and after hot rolling: cross section of the condition as cast (a) and transversal section (b) of the rolled tape. In the first picture (a) the results of the compositional analysis in atomic percentage are also reported

shows representative SEM pictures of the quaternary alloy in the condition as cast and in the rolled condition. This quaternary alloy has been firstly characterized from the compositional point of view using EDS measurements; two

Journal of Materials Engineering and Performance

Fig. 4 SEM images of the microstructure of the rolled ZrCuCr (a) and ZrCuCo (b) alloys

secondary phases were detected, quite equally distributed in the matrix, whose composition (Zr49Cu29Ni7Co15 in atomic percentage) is given by the mean chemical composition of the alloy. The two secondary phases are characterized by Zr/Cu ratio, respectively, higher and lower than the ratio representative of the matrix. The first phase recognized (Zr53Cu32Ni5Co10) is

ZrCuNiCo as cast ZrCuNiCo hot rolled

Heat flow [au]

Heat flow [au]

ZrCuCr as cast ZrCuCr hot rolled

-100

0

(a)

100

200

100

300

150

200

300

350

ZrCuNiCoTi as cast ZrCuNiCoTi hot rolled

Heat flow [au]

Heat flow [au]

ZrCuCo as cast ZrCuCo hot rolled

-100

(b) Fig. 5

250

Temperature [°C]

(c)

Temperature [°C]

-50

0

50

100

0

150

Temperature [°C]

50

100

150

200

250

300

350

Temperature [°C]

(d)

DSC scans of the ZrCuCr (a), ZrCuCo (b), ZrCuNiCo, (c) and ZrCuNiCoTi (d) before and after hot rolling during thermal cycling

Table 2 Characteristic temperatures and heats exchanged of the investigated SMAs before and after hot rolling process at the 1st and at the 10th thermal cycle Composition, at.% Zr50Cu45Cr5

Zr50Cu40Co10

Zr50Cu28Ni7Co15

Zr50Cu27Ni7Co9Ti7

Condition

N. of cycle

Ms, C

Mf, C

As, C

Af, C

QM>A, J/g

QA>M, J/g

Cast Rolled Cast Rolled Cast Rolled Cast Rolled Cast Rolled Cast Rolled Cast Rolled Cast Rolled

1th 1th 10th 10th 1th 1th 10th 10th 1th 1th 10th 10th 1th 1th 10th 10th

67 33 3 26 23 8 2 20 248 235 229 228 124 96 128 96

14 7 81 66 27 54 63 73 167 141 133 153 57 17 14 2

136 208 183 223 33 23 45 47 237 232 227 234 146 98 150 75

161 234 249 276 82 89 94 105 316 304 304 310 232 201 242 200

10.7 8.5 4.6 4.4 10.7 8.3 2.9 4.1 20.4 10 14 9.6 2.5 7.1 3.7 8.3

10.5 12 12.6 11.7 10.1 11.2 8.4 6.8 10.5 7.6 13.3 8.4 6.3 6.3 5.9 7.2

given by a small increase of the amount of Zr and Cu and by the same Ni/Co ratio, present in the matrix. On the contrary, the other phase (Zr40Cu46Ni7Co7) is Cu rich; moreover, the Ni/Co

ratio is increased by 0.5 to 1 but the amount of these two elements is decreased in comparison of their content in the matrix.

Journal of Materials Engineering and Performance

It can be observed that these secondary phases look to be casually orientated in the condition as cast (see Fig. 3a) while these are orientated after the plastic deformation process. In particular, the secondary phase is elongated along the rolling direction, as shown in Fig. 3b, and compressed along the direction of the plate thickness, thanks to the applied compression stress. No sensible variation of the chemistry of these phases was detected after the hot rolling process. Moreover, no evident cracks are visible in the rolled material and a uniform distribution of the secondary phase can be observed in the matrix. This could be associated to a discrete workability at high temperature of such alloy. Similar results were also obtained for the other alloys. In Fig. 4, the SEM images of the orientated microstructure, representative of the two ternary CuZr-based alloys, are shown. The secondary phase is elongated along the rolling direction; even the matrix looks to be quite ductile, due to the absence of cracks inside. It can be affirmed that, in comparison with the workability of the binary system, the addition of some elements can improve the deformability of the material, saving it from crack generation. The calorimetric analysis puts in evidence that the martensitic transformation (MT) is not suppressed after the hot rolling for all the alloys. As reported in some works in literature, the thermal stability of the MT can be generally a drawback of such

525

ZrCuCo

ZrCuCr

ZrCuNiCo

as system (Ref 3, 5). In Fig. 5, the DSC scans of the investigated alloys are depicted, putting in evidence the evolution of the peaks of the MT during thermal cycling by their stabilization. It can be seen that the stabilization of the MT strongly depends on the chemical composition while the hot rolling does not seem to improve its stabilization of the MT. The ternary alloys require about 8 thermal cycles for their thermal stabilization as the quaternary and the quinary ones reach a stable behavior at the 2nd cycle. Moreover, the characteristic temperatures and the heat exchange of the two ternary alloys are significantly shifted with higher thermal hysteresis, probably due to the plastic deformation imposed to the material; on the contrary, the quaternary alloy does not show a sensible modification of the characteristic temperatures (see Table 2). An estimate of the mechanical properties of the alloys is given by the microhardness, measured on the samples before and after the hot rolling. Figure 6 shows the effect of the hot plastic deformation for each alloy. It seems that the hot rolling procedure, tested in this work, does not influence the hardness of the alloys. On the contrary, it was observed that the chemical composition of the alloys can modify the hardness in the range 350-520 HV. In particular, the ternary alloys show the lowest hardness values (about 360 and 400 HV, respectively) while the

ZrCuNiCoTi

500

Hardness [HV]

475 450 425 400 375 350 325

as cast

as cast

hot rolled

as cast as cast hot rolled hot rolled

hot rolled

Fig. 6 Hardness of the CuZr-based alloys before and after hot rolling

Fig. 7

Fig. 8 Magnification of the lateral view of the hole, in the proximity of the entrance surface

SEM images of the entrance (on the left) and exit (on the right) sides of a hole laser machined with a single pulse

Journal of Materials Engineering and Performance

quaternary and the quinary ones seem to be similar (about 475 HV). Finally, a preliminary study of the interaction between a laser beam and the quaternary CuZr-based SMA has been proposed. During the experiments, through hole of 2 mm thickness was obtained using a single laser pulse. The entrance and the exit surfaces of the hole are characterized by the presence of a corona of melted material, as shown in Fig. 7, because laser pulses in ms order usually operate in fusion because of the relevant contribute of the heat conduction. The circularity of both the entrance and exit diameters is nice and the hole diameter is approximately equal to 230 lm. The aspect ratio of the hole, indicated as ratio of the depth over the diameter, is about 10 and it is interesting to underline that the productivity of the laser machining in this configuration is in the order of 100 holes per second on such as thickness. The optimization of the process parameters has to be improved but significant evidences in the potentialities of this process on the CuZr-based alloys are clearly visible. Figure 8 shows a magnification of the transversal section of the hole laser machined in the proximity of the entrance surface. It can be seen that the inner surface appears to be quite smooth and no relevant amount of melted material is visible; this could be due to the high beam quality, which allows the main of the energy to be focalized in a small laser spot.

4. Conclusions In this paper, hot workability of CuZr-based SMAs has been evaluated. The investigated procedure consists in a preliminary hot forging and then hot rolling at 800 °C. The alloys were sealed in a stainless steel can to preserve them from oxidation at high temperature; soft thickness reduction for each rolling step was considered. With respect to the binary alloy, some alloying elements in CuZr system were added to improve thermal stability of the martensitic transformation as well as for increasing the operating temperatures. An improvement of the hot workability was also observed. The evolution of the martensitic transformation has been reported during thermal cycling, showing the effect of the alloying elements in the role of stabilizing the characteristic temperatures. Otherwise, no significant effect has been found associated to the plastic deformation on the calorimetric response of these alloys. The mechanical properties of the rolled alloys have been investigated, too; the microhardness of the samples is influenced only by the chemical composition of the alloys but

they are not modified after the rolling process, probably due to the limited work hardening involved. Finally, a preliminary test of laser machining on the quaternary alloy, selected for its better thermal stability and the highest operating temperatures, has been also reported. Due to the quite brittle behavior of such alloys, the possibility of their machining by means of a thermal process, i.e., laser material processing, is certainly a challenging issue. For this reason, a representative example of a through hole laser machined, of 2 mm thickness, with one single laser pulse has been shown. The hole size is in the order of about 200 lm with low amount of melted material and limited taper. Next investigation will be further reported in future works for studying the laser machinability of these SMAs.

Acknowledgments This work was developed in the framework of ‘‘Mind in Italy’’ project and sponsored by Regione Lombardia. The authors wish to thank Mr. Marco Pini, Nicola Bennato, and Giordano Carcano for their technical support.

References 1. J. Ma, I. Karaman, and R.D. Noebe, High Temperature Shape Memory Alloys, Int. Mater. Rev., 2010, 55–5, p 257–315 2. S. Firstov, J. Van Humbeeck, and Y.N. Koval, High Temperature Shape Memory Alloys Problems and Prospective, J. Intel. Mat. Syst. Str., 2006, 17, p 1041–1047 3. G.S. Firstov, J. Van Humbeeck, and Y.N. Koval, Pecularities of the Martensitic Transformation in ZrCu Intermetallic Compound—Potential High Temperature SMA, J. De Phys. IV France, 2001, 11 Pr-8-481-486 4. G.S. Firstov, J. Van Humbeeck, and Y.N. Koval, High-Temperature Shape Memory Alloys: Some Recent Developments, Mater. Sci. Eng. A, 2004, 378, p 2–10 5. C.A. Biffi, A. Figini Albisetti, and A. Tuissi, CuZr Based Shape Memory Alloys: Effect of Cr and Co on the Martensitic Transformation, Mater. Sci. Forum, 2013, 738–739, p 167–171 6. C.A. Biffi, A. Figini, and A. Tuissi, Influence of Compositional Ratio on Microstructure and Martensitic Transformation of CuZr Shape Memory Alloys, Intermetallics, 2014, 46, p 4–11 7. F. Qiu, P. Shen, T. Liu, Q. Lin, and Q. Jiang, Influence of Al Content on Martensitic Transformation Behavior in Zr50Cu50-xAlx, J. Alloys. Compd., 2012, 522, p 136–140 8. G.S. Firstov, J. Van Humbeeck, and Yu.N. Koval, Comparison of High Temperature Shape Memory Behavior for ZrCu-Based, Ti-Ni-Zr and Ti-Ni-Hf alloys, Scripta Mater., 2004, 50, p 243–248 9. C.A. Biffi, P. Bassani, R. Casati, M. Vedani, and A. Tuissi, Processing of CuZr Based Shape Memory Alloys, Mater. Sci. Forum, 2014, 773–774, p 542–548

Journal of Materials Engineering and Performance