Materials Transactions, Vol. 47, No. 3 (2006) pp. 518 to 522 Special Issue on Shape Memory Alloys and Their Applications #2006 The Japan Institute of Metals
Effect of Nb Addition on Shape Memory Behavior of Ti–Mo–Ga Alloys Hee Young Kim1 , Yoshinori Ohmatsu1 , Jae Il Kim1;2 , Tomonari Inamura3 , Hideki Hosoda3 and Shuichi Miyazaki1; * 1
Institute of Materials Science, University of Tsukuba, Tsukuba 305-8573, Japan Materials Science and Engineering, Dong-A University, Busan 604-714, Korea 3 Precision and Intelligent Laboratory, Tokyo Institute of Technology, Yokohama 226-8503, Japan 2
Effect of Nb addition on mechanical properties and shape memory behavior of Ti–Mo–Ga alloys was investigated by cyclic loadingunloading tensile tests, tensile tests at various temperatures and Vickers hardness tests. The martensitic transformation start temperature (Ms) decreased by 20 K with 1 at% increase of Nb content in the Ti–6Mo–3Ga(at%) alloy. The shape memory effect was observed in the Ti–6Mo– 3Ga(at%) alloy. The superelastic strain increased with increasing Nb content. The increase of the yield stress during the cyclic deformation decreased the strain recovery rate in the Ti–6Mo–3Ga(at%) alloy. Aging at intermediate temperatures resulted in increase in hardness of the Ti– 6Mo–3Ga–(0–4)Nb(at%) alloys. The increase in hardness decreased significantly with increasing Nb content because the addition of Nb was effective to suppress the ! phase hardening. The yield stress decreased and the strain recovery rate increased with increasing number of cycle in the Ti–6Mo–3Ga–4Nb(at%) alloy. (Received October 11, 2005; Accepted December 16, 2005; Published March 15, 2006) Keywords: shape memory alloy, superelasticity, biomaterial, smart material, titanium based alloy
1.
Introduction
Ni-free Ti-based shape memory alloys have attracted much interest in recent years because of their potential applications for biomedical devices. Ti-based alloys have been developed for biomedical applications such as artificial bones, joints and dental materials due to their high biocompatibility and good mechanical properties.1,2) It has been well known that a martensitic transformation form (disordered BCC) to orthorhombic martensite (00 ) and its reverse transformation are related to the shape memory behavior of the -type Ti alloys.3,4) The shape memory effect and superelastic behavior have been reported in Ti-based alloys.5–15) We have reported that Ti–Mo–Ga alloys exhibited excellent superelastic behavior with a recovery strain of 4%.9) But, the Ti–Mo based alloys are susceptible to ! phase embrittlement. It is generally acknowledged that the formation of ! phase leads to severe loss of ductility. A premature failure occurred in the Ti–Mo–Ga alloys subjected to annealing at intermediate temperatures.16) In this study, the effect of Nb addition on mechanical properties and shape memory behavior of Ti– Mo–Ga alloys was investigated. The effect of Nb addition on ! phase embrittlement is also reported. 2.
Experimental Procedures
Ti–6Mo–3Ga–(0–4)Nb(at%) alloy ingots were prepared by the Ar arc melting method. Hereafter, Ti–xMo–yGa– zNb(at%) is abbreviated to Ti–xMo–yGa–zNb. The ingots were sealed in vacuum into a quartz tube and homogenized at 1373 K for 86.4 ks, and then cold-rolled with the reduction of 95% in thickness. Specimens for the tensile tests and Vickers hardness measurements were cut from the cold-rolled sheet by an electro-discharge machine. The specimens were cleaned with ethanol, wrapped in Ti foils and encapsulated in quartz tubes under a 3.3 kPa partial pressure of high-purity Ar, and then annealed at 1073 K for 120 s. Some specimens *Corresponding
author, E-mail:
[email protected]
were aged at temperatures between 573 and 1073 K for 60 s after the annealing. The specimens were quenched into water by breaking the quartz tubes. The oxidized surface was removed by mechanical polishing followed by electropolishing. The final thickness of specimens was about 0.4 mm. Bending tests were carried out at room temperature where the specimens were deformed into a round shape with a diameter of 10 mm and heated up to approximately 500 K. Tensile tests were carried out at a strain rate of 1:67 104 s1 at various temperatures between 223 and 383 K using a Shimadzu AG-A universal testing machine. The gage length of the specimens was 20 mm. The shape memory behavior was investigated by strain increment cyclic tensile tests at room temperature. Vickers hardness was measured more than ten points where applied load and time were 200 g and 10 s, respectively, using an Akashi HM-102 Vickers hardness tester. 3.
Results and Discussion
Figure 1 shows a result of bending test for a Ti–6Mo–3Ga alloy subjected to heat treatment at 1073 K for 120 s. It is clearly seen that the specimen bent at room temperature recovered to the original straight shape by heating up to 500 K. The alloy exhibits a perfect shape memory effect within a strain of 3.5% by the bending test. The shape memory behavior was also investigated by cyclic loadingunloading tensile tests. Figure 2 shows stress–strain curves of the Ti–6Mo–3Ga alloy subjected to heat treatment at 1073 K for 120 s. Each stress–strain curve and strain recovery were obtained at room temperature by a loading and unloading cycle followed by heating: broken lines with an arrow indicate the shape recovery by heating. The similar measurement was repeated by increasing the maximum strain upon loading in a same sample. The starting points of the stress– strain curves are shifted in order that they are separated each other. Almost perfect shape memory effect was observed at the first and the second cycle. The plastic strain remained after heating increased with increasing tensile strain. It is
Effect of Nb Addition on Shape Memory Behavior of Ti–Mo–Ga Alloys
Fig. 1
A result of bending test for the Ti–6Mo–3Ga alloy: (a) after deformation and (b) after heating up to 500 K.
500
800
Ti–6Mo–3Ga Ti–6Mo–3Ga–1Nb Ti–6Mo–3Ga–2Nb Ti–6Mo–3Ga–3Nb Ti–6Mo–3Ga–4Nb
600
450
400 200 0
Strain, ε (%)
2%
Fig. 2 Stress–strain curves obtained by cyclic loading-unloading tensile tests for the Ti–6Mo–3Ga alloy.
Vickers hardness / Hv
Stress, σ / MPa
519
400
350
300
250
noted that the apparent yield stress indicated by an arrow increased with increasing number of cycle until the 3rd cycle. The apparent yield stress was defined by the 0.2% offset yield stress in this study. It was reported that the shape memory effect in the Ti–6Mo–3Ga alloy was mainly due to the stress induced martensitic transformation during tensile deformation followed by the reverse transformation during heating.9) Thus, the increase in the apparent yield stress means that the critical stress to induce the martensitic transformation increases during the cyclic deformation in the Ti–6Mo–3Ga alloy. This is opposite tendency observed in Ti–Ni alloys during cyclic deformation. It has been reported that the critical stress to induce the martensitic transformation decreases with increasing cyclic deformation in the Ti–Ni alloys, since the internal stress formed by a slip deformation assists the formation of the stress induced martensites.17,18) It is supposed that the increase in yield stress during cyclic deformation in the Ti–6Mo–3Ga alloy is due to the ! phase formed during heating for each cycle. The Vickers hardness was measured in Ti–6Mo–3Ga–(0– 4)Nb alloys which were subjected to aging treatment at temperatures between 373 and 1073 K for 60 s. Figure 3 shows the change in hardness as a function of heat treatment temperature. Aging the specimens at intermediate temperatures resulted in increase in hardness of the Ti–6Mo–3Ga– (0–4)Nb alloys. The rate of increase in hardness is distinctly more rapid in the Ti–6Mo–3Ga alloy. For the Ti–6Mo–3Ga alloy, the hardness increased rapidly with increasing aging temperature up to 673 K. This result is consistent with previous works that ! phase forms on aging at temperatures up to about 723 K in Ti–Mo based alloys.19,20) The rate of increase in hardness decreased significantly with increasing Nb content. This result obviously indicates that the addition of Nb suppressed the formation of thermal ! phase in the Ti– 6Mo–3Ga alloy.
200 200
400
600
800
1000
1200
Aging temperature, T/ K Fig. 3 Vickers hardness of the Ti–6Mo–3Ga–(0–4)Nb alloys which were subjected to aging treatment at temperatures between 373 and 1073 K for 60 s.
Figure 4 shows stress–strain curves of the Ti–6Mo–3Ga– (1–4)Nb alloys subjected to heat treatment at 1073 K for 120 s. The Ti–6Mo–3Ga–(1–3)Nb alloys exhibit the shape memory effect at the first cycle. A small amount of strain recovered superelastically during unloading in the Ti–6Mo– 3Ga–3Nb alloy. The strain mainly recovered superelastically in the Ti–6Mo–3Ga–4Nb alloy. This indicates that the addition of Nb decreased the martensitic transformation temperature. The yield stress increased slightly at the second cycle in the Ti–6Mo–3Ga–1Nb and Ti–6Mo–3Ga–2Nb alloys. But, no increase of yield stress was observed in the Ti–6Mo–3Ga–3Nb and Ti–6Mo–3Ga–4Nb alloys during the cyclic deformation. The yield stresses were evaluated using the stress–strain curves of the Ti–6Mo–3Ga–(0–4)Nb alloys and they are plotted as a function of number of cycle in Fig. 5. The yield stress increased during cyclic deformation reaching a maximum yield stress at the 3rd cycle, and then the yield stress decreased gradually with increasing number of cycle for the Ti–6Mo–3Ga alloy. The maximum yield stress was obtained at the 2nd or 3rd cycles for the Ti–6Mo– 3Ga–1Nb and Ti–6Mo–3Ga–2Nb alloys, respectively. On the other hand, the yield stress decreased monotonously with increasing number of cycle for the Ti–6Mo–3Ga–4Nb alloy. This result also confirms that the Nb is effective to suppress the ! phase hardening. In order to clarify the effect of Nb addition on shape
520
H. Y. Kim et al. 100
Ti-6Mo-3Ga-1Nb
600 400
Strain recovery rate (%)
95
200 0 Ti-6Mo-3Ga-2Nb
600 400
Stress, σ / MPa
200 0
90
85
80
75
Ti-6Mo-3Ga Ti-6Mo-3Ga-4Nb
Ti-6Mo-3Ga-3Nb
600
70 0
400
1
2
3
4
5
6
Number of cycle
200
Fig. 6 Comparison of recovery rates between the Ti–6Mo–3Ga and Ti– 6Mo–3Ga–4Nb alloys during the cyclic deformation.
0 Ti-6Mo-3Ga-4Nb
600 400
600
200
243K
263K
283K
297K
323K
343K
363K
383K
400
0
Strain, ε (%)
200
2%
600
Ti–6Mo–3Ga Ti–6Mo–3Ga–1Nb Ti–6Mo–3Ga–2Nb Ti–6Mo–3Ga–3Nb Ti–6Mo–3Ga–4Nb
500
0
Stress, σ / MPa
Fig. 4 Stress–strain curves obtained by cyclic loading-unloading tensile tests for of the Ti–6Mo–3Ga–(1–4)Nb alloys.
Yield stress, σ / MPa
223K
600 400 200 0 600 400
400
200 0
300
1
2
3
1
2
3
1
2
3
Strain, ε (%)
200
Fig. 7 Stress–strain curves obtained at various temperatures for the Ti– 6Mo–3Ga–2Nb alloy.
100 1
2
3
4
5
6
7
8
Number of cycle Fig. 5 Change of yield stress of the Ti–6Mo–3Ga–(0–4)Nb alloys during the cyclic deformation.
memory properties, the strain recovery rates of Ti–6Mo–3Ga and Ti–6Mo–3Ga–4Nb alloys are compared in Fig. 6 as a function of number of cycle. The strain recovery rate of Ti– 6Mo–3Ga alloy decreased between the first and the third cycles. This is caused by the increase of the critical stress to induce the martensitic transformation in the Ti–6Mo–3Ga alloy during the cyclic deformation. Slip occurs if the stress level for slip deformation becomes lower than the critical stress to induce the martensite. Thus, it is suggested that the
increase of the critical stress by the formation of ! phase increased the permanent strain by slip deformation, and resulted in the decrease of the strain recovery rate during the cyclic deformation. On the other hand, the strain recovery rate increased during the cyclic deformation in the Ti–6Mo– 3Ga–4Nb alloy. This is consistent with the decrease of the critical stress to induce the martensitic transformation in the Ti–6Mo–3Ga–4Nb alloy during the cyclic deformation as shown in Fig. 5. Figure 7 shows a series of stress–strain curves obtained at various temperatures for the Ti–6Mo–3Ga–2Nb alloy. Tensile stress was applied until strain reached about 3% and then the stress removed followed by heating to measure the strain recovery. The shape memory effect was observed in the Ti–
Effect of Nb Addition on Shape Memory Behavior of Ti–Mo–Ga Alloys 243K
600
263K
521
400
283K
200 0 298K
Stress, σ / MPa
600
323K
343K
400 200
Ms temperature, T/K
400 350
300
250
0 363K
600
200 0
400
1
2
3
4
Nb content(at%)
200
Fig. 10 Effect of Nb content on Ms temperature for the Ti–6Mo–3Ga alloy.
0 1
2
3
4
Strain, ε (%) Fig. 8 Stress–strain curves obtained at various temperatures for the Ti– 6Mo–3Ga–4Nb alloy.
500 Ti-6Mo-3Ga-2Nb Ti-6Mo-3Ga-4Nb
Stress, σ / MPa
400
300
200
100 200
250
300
350
400
Temperature, T/K Fig. 9 Temperature dependence of yield stress for the Ti–6Mo–3Ga–2Nb and Ti–6Mo–3Ga–4Nb alloys.
6Mo–3Ga–2Nb alloy in the temperature region between 223 and 383 K. The residual strain after unloading was almost completely recovered by heating to about 500 K for the specimens deformed at temperatures between 263 and 297 K. On the other hand, the superelastic behavior was observed in the Ti–6Mo–3Ga–4Nb alloy as shown in Fig. 8. This result is reasonable since the Nb decreases the martensitic transformation temperature in Ti-based alloys. The temperature dependence of yield stress for the Ti–6Mo–3Ga–2Nb and Ti– 6Mo–3Ga–4Nb alloys is plotted in Fig. 9. The minimum critical stress is observed at 298 and 263 K in the Ti–6Mo– 3Ga–2Nb and Ti–6Mo–3Ga–4Nb alloys, respectively. The temperature exhibiting the minimum yield stress is taken as the martensitic transformation start temperature (Ms).11–13) The critical stress increased with decreasing temperature
from the minimum stress point. This is because the stress for the rearrangement of martensite variants increases with decreasing temperature. On the other hand, the critical stress increases with increasing temperature from the minimum stress point since the stress for inducing martensites increases with increasing temperature, which is in accordance with the Clausius–Clapeyron relationship. It is noted that the superelastic behavior was observed in the temperature range between 243 and 283 K in which the minimum critical stress was obtained for the Ti–6Mo–3Ga– 4Nb alloy as shown in Fig. 8. For the Ti–Ni shape memory alloy, the superelastic behavior does not appear at the Ms temperature because the start temperature for the reverse transformation (As) is higher than the Ms. It is supposed that the As temperature is lower than Ms in the Ti–6Mo–3Ga– 4Nb alloy due to the large temperature difference between Ms and Mf. The large temperature difference between Ms and Mf has been also reported in the Ti–Nb based alloys.11,13,15) The superelastic strain decreased with increasing temperature for the Ti–6Mo–3Ga–4Nb alloy. This result can be explained by the difference in the temperature dependence of the critical stress for inducing martensites and the stress for slip deformation. The critical stress to induce the martensites increases with increasing temperature, while the stress for slip deformation decreases with increasing temperature. Thus, the permanent strain by slip deformation increases with increasing temperature, causing the recoverable strain to decrease. Figure 10 shows the Nb content dependence of the Ms temperature in the Ti–6Mo–3Ga alloy. The temperature exhibiting the minimum critical stress was taken as Ms. The Ms linearly decreases with increasing Nb content: the Ms decreased by 20 K with 1 at% increase of Nb content. The Mo and Ga also decrease Ms in Ti-based alloys.9) This means that the shape memory effect or superelastic behavior is achievable at body temperature by controlling Mo and Ga contents, suggesting that the Ti–Mo–Ga–Nb alloy can be a candidate material for the biomedical shape memory alloys.
522
4.
H. Y. Kim et al.
Summary
(1) The critical stress to induce martensitic transformation increased with increasing number of cycle until the 3rd cycle due to the ! phase formed during the cyclic deformation followed by heating in the Ti–6Mo–3Ga alloy. The increase of the critical stress increased the permanent strain by slip deformation, resulting in the decrease of the strain recovery rate. (2) Aging at intermediate temperatures resulted in the increase in hardness of the Ti–6Mo–3Ga–(0–4)Nb alloys. The increase in hardness decreased significantly with increasing Nb content. (3) The addition of Nb was effective to suppress the ! phase hardening during the cyclic deformation followed by heating. The critical stress to induce martensites decreased and the strain recovery rate increased with increasing number of cycle in the Ti–6Mo–3Ga–4Nb alloy. (4) The martensitic transformation start temperature (Ms) decreased by 20 K with 1 at% addition of Nb content in the Ti–6Mo–3Ga alloy. Acknowledgments This work was partially supported by the ILC Project from University of Tsukuba the 21 Century Center of Excellence Program and the Grants-in-Aid for Fundamental Scientific Research (Kiban A (1999–2001), Kiban A (2002–2004)) from the Ministry of Education, Culture, Sports, Science and Technology, Japan.
REFERENCES 1) M. Niinomi: Mater. Sci. Eng. A 243 (1998) 231–236. 2) Y. Okazaki, Y. Ito, A. Ito and T. Tateishi: Mater. Trans., JIM 34 (1993) 1217–1222. 3) C. Baker: Metal Sci. J. 5 (1971) 92–100. 4) T. W. Duerig, J. Albrecht, D. Richter and P. Fischer: Acta Metall. 30 (1982) 2161–2172. 5) T. Grosdidier and M. J. Philippe: Mater. Sci. Eng. A 291 (2000) 218– 223. 6) E. Takahashi, T. Sakurai, S. Watanabe, N. Masahashi and S. Hanada: Mater. Trans. 43 (2002) 2978–2983. 7) T. Maeshima and M. Nishida: Mater. Trans. 45 (2004) 1096–1100. 8) T. Maeshima and M. Nishida: Mater. Trans. 45 (2004) 1101–1105. 9) H. Y. Kim, Y. Ohmatsu, J. I. Kim, H. Hosoda and S. Miyazaki: Mater. Trans. 45 (2004) 1090–1095. 10) Y. Fukui, T. Inamura, H. Hosoda, K. Wakashima and S. Miyazaki: Mater. Trans. 45 (2004) 1077–1082. 11) H. Y. Kim, H. Satoru, J. I. Kim, H. Hosoda and S. Miyazaki: Mater. Trans. 45 (2004) 2443–2448. 12) J. I. Kim, H. Y. Kim, H. Hosoda and S. Miyazaki: Mater. Trans. 46 (2005) 852–857. 13) J. I. Kim, H. Y. Kim, T. Inamura, H. Hosoda and S. Miyazaki: Mater. Sci. Eng. A 403 (2005) 334–339. 14) H. Y. Kim, T. Sasaki, K. Okutsu, J. I. Kim, T. Inamura, H. Hosoda and S. Miyazaki: Acta Mater. 54 (2006) 423–433. 15) H. Y. Kim, S. Hashimoto, J. I. Kim, T. Inamura, H. Hosoda and S. Miyazaki: Mater. Sci. Eng. A 417 (2006) 120–128. 16) Y. Ohmatsu, H. Hosoda and S. Miyazaki: 4th Pacific Rim Inter. Conf. Adv. Mater. Proc., (The Japan Inst. Metals, 2001) pp. 1627–1629. 17) S. Miyazaki, T. Imai, Y. Igo and K. Otsuka: Metall. Trans. A 17A (1986) 115–120. 18) S. Miyazaki: Engineering Aspects of Shape Memory Alloys (Ed. by T. W. Duerig et al.), Butterworth-Heineman (1990) 452–469. 19) F. Langmayr, P. Fratzl, G. Vogl and W. Miekeley: Phys. Rev. B 49 (1994) 11759–11766. 20) F. H. Froes, C. F. Yolton, J. M. Capenos, M. G. H. Wells and J. C. Williams: Metall. Trans. A 11A (1980) 21–31.
