- Email: [email protected]
Cold workability and shape memory properties of novel Ti–Ni–Hf–Nb high-temperature shape memory alloys
Available online at www.sciencedirect.com
Acta Materialia 65 (2011) 846–849 www.elsevier.com/locate/scriptamat
Cold workability and shape memory properties of novel Ti–Ni–Hf–Nb high-temperature shape memory alloys Hee Young Kim,a,⇑ Takafumi Jinguu,a Tae-hyun Namb and Shuichi Miyazakia,b,c,⇑ a
Institute of Materials Science, University of Tsukuba, Tsukuba, Ibaraki 305-8573, Japan School of Materials Science and Engineering & ERI, Gyeongsang National University, 900 Gazwadong, Jinju, Gyeongnam 660-701, Republic of Korea c Center of Excellence for Advanced Materials Research, King Abdulaziz University, P.O. Box 80203, Jeddah 21589, Saudi Arabia b
Received 7 June 2011; revised 26 July 2011; accepted 26 July 2011 Available online 30 July 2011
The effect of Nb content on the cold workability, microstructure and shape memory properties of Ti–Ni–Hf–Nb alloys is investigated. The addition of Nb to Ti–Ni–Hf improves the cold workability due to the formation of a soft Nb-rich phase. The Ti–Ni– Hf–Nb alloys exhibit a high-temperature shape memory effect during heating and cooling cycles under various stress levels. As the Nb content increases, the shape memory properties of the Ti–Ni–Hf–Nb alloys become more stable although the shape recovery strain decreases. Ó 2011 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. Keywords: Shape memory alloys; Phase transformation; Cold working; Microstructure
Since the discovery of methods for achieving superelasticity and for stabilizing the shape memory effect in Ti–Ni alloys [1–3], these alloys have over the last couple of decades been increasingly used in various fields of industry including home appliances, medical devices and electronic devices owing to their unique properties of shape memory effect and superelasticity [4]. However, most practical applications of shape memory alloys are limited to temperatures below 373 K because of the low transformation temperatures of commercially available shape memory alloys such as Ti–Ni and Cubase alloys. Recently shape memory alloys operating at high environmental temperatures (above 373 K) have attracted attention as solid-state actuators for use in aircraft engines, automobile engines and energy-generation systems due to their large stroke, high recovery force and high work output [5,6]. Up to now several types of shape memory alloys have been reported to exhibit a martensitic transformation start temperature above 373 K [5–12]. Among these, Ti–Ni–X alloys (where X = Pt, Pd, Hf or Zr) have
⇑ Corresponding
authors. Address: Institute of Materials Science, University of Tsukuba, Tsukuba, Ibaraki 305-8573, Japan. Tel./fax: +81 29 853 6942 (H.Y. Kim). Tel./fax: +81 29 853 5283 (S. Miyazaki); e-mail addresses: [email protected]; [email protected]
received the most research attention [13–19]. The martensitic transformation temperatures increase with the replacement of Ni with Pd and Pt above a critical concentration [13–15]. Ti–Ni–(Pd, Pt) alloys exhibit a relatively stable shape memory effect at above 373 K with a small hysteresis. However, the high cost of precious metal alloying elements has the limited practical applications of Ti– Ni–(Pd, Pt) shape memory alloys. It has also been widely recognized that substitution of Zr or Hf for Ti increases the transformation temperature of Ti–Ni shape memory alloys [16–19]. The Ti–Ni–(Zr, Hf) alloys are regarded as more practical systems compared with the Ti–Ni– (Pd, Pt) alloys because of the relatively low price of the raw materials required; however, for several reasons no practical devices have yet been developed. One of the most serious problems with Ti–Ni–(Zr, Hf) alloys is their low workability: their hardness is increased and their cold workability and ductility are decreased significantly by the addition of Zr and Hf [6,20–22]. Improvement of the cold workability of Ti–Ni–(Zr, Hf) alloys is vital for manufacturing them into suitable sizes and shapes and for microstructure control. The introduction of a ductile phase is an effective way to improve the cold workability of brittle materials. It has been reported that the addition of certain alloying elements such as Nb, W and Ag to Ti–Ni binary alloys results in the formation of a ductile phase [23–26]. However, up to now there has been no
1359-6462/$ - see front matter Ó 2011 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.scriptamat.2011.07.049
H. Y. Kim et al. / Scripta Materialia 65 (2011) 846–849
report on Ti–Ni–(Zr, Hf) shape memory alloys containing a ductile phase. In this study, a new alloy system, Ti–Ni–Hf–Nb, was designed, and the effect of Nb content on the cold workability, microstructure and shape memory properties of Ti–Ni–Hf–Nb high-temperature shape memory alloys was investigated. (Ti–49.5Ni–15Hf)100 X–XNb (X = 0–20 at.%) alloys were fabricated by the argon arc-melting method. The ingots were sealed in a quartz tube under vacuum and homogenized at 1223 K for 7.2 ks. Specimens for all measurements were cut by an electrodischarge machine. The oxidized or damaged surface was removed by mechanical polishing. Cold workability was evaluated by measuring the minimum thickness attained before fracture upon cold rolling at room temperature (RT) where specimens were prepared in the following dimensions: 30 mm in length 10 mm in width 1 mm in thickness. Microstructure and phase constituent were investigated by Xray diffraction (XRD) and scanning electron microscopy (SEM). XRD measurements were conducted at RT and 573 K with Cu Ka radiation. The surfaces of specimens for SEM observation were polished using an alumina slurry (0.3 lm). Shape memory properties were investigated by thermal cycling tests (with a heating and cooling rate of 10 K min 1) under various stress levels. The thermal cycling tests were carried out using a thermomechanical analyzer incorporating a linear variable differential transformer to measure the strain. The tests were conducted in helium gas atmosphere in order to prevent oxidation and improve thermal conductivity. The dimensions of the specimens were 13 mm in gauge length, 0.7 mm in width and 0.15 mm in thickness. Figure 1 shows the effect of Nb content on the cold workability of Ti–49.5Ni–15Hf–(0–20)Nb alloys. A Ti– 49.5Ni–15Hf specimen fractured instantly after cracking at 19% reduction of thickness (see the image inserted at the lower-left corner of Fig. 1). The total reduction in thickness until fracture increased with increasing Nb content. In particular, the addition of 10 at.% Nb and more allowed the Ti–Ni–Hf alloy to be cold rolled up to 60% reduction in thickness without fracture, although small cracks were introduced from side surfaces as shown in the inserted image.
80
60
Figure 2 shows XRD profiles of (Ti–49.5Ni– 15Hf)100 X–XNb (X = 0–20 at.%) alloys heat treated at 1223 K for 7.2 ks. The XRD profiles were obtained at 573 K (above the reverse transformation finish temperatures of the alloys) in order to provide better clarity and allow for detailed analysis of the secondary phase because the peaks from the secondary phase overlapped with the peaks from the martensite structure (B19’ with a monoclinic structure) at RT. Hereafter, each alloy is referred to according to its Nb concentration, i.e. 0Nb, 10Nb, etc. The 0Nb alloy is mainly composed of a B2 phase with a small amount of Ti2Ni type phase, which can be identified by the weak reflection at 2h = 40.5°. In addition to B2 and Ti2Ni type phases, an additional phase (b phase), which has a disordered body-centered cubic (bcc) structure was observed in the alloys containing 5 at.% Nb and more. The lattice constant of the b phase is determined to be 0.3315 nm, which is very close to that of pure Nb. It is also noted that the lattice constant of the b phase is insensitive to Nb content, while the peak intensity of the b phase increased with increasing Nb content. Figure 3 shows scanning electron images of (Ti– 49.5Ni–15Hf)–(0–20)Nb alloys. Observation was carried out in back-scattered electron (BSE) mode since the BSE image efficiently distinguishes different phases. For the 0Nb alloy (Fig. 3a), the Ti2Ni type phase can be seen as relatively dark particles. In addition to Ti2Ni type phase, the b phase (which appears white on the images) is seen in the Nb-containing alloys. It is noted that the b phase was observed in the 1Nb alloy (Fig. 3b) although the peaks from the b phase could not be detected by the XRD analysis due to the small volume fraction. This result implies that the solution limit of Nb in the matrix phase, i.e. B2 phase, is less than 1 at.%. The amount of the b phase increased with increasing Nb content. The 15Nb alloy (Fig. 3e) reveals a fully lamellar microstructure, which is a characteristic of eutectic solidification. It is also noted that the 10Nb alloy exhibits a typical microstructure of off-eutectic solidification, which is composed of a primary dendrite phase (B2) and interdendritic lamellar structure. On the other hand, the primary dendrite phase in the 20Nb alloy is the b phase as shown in Figure 3f. A similar change in microstructure by the addition of Nb was reported in a Ti–Ni binary alloy where the Nb content in the eutectic composition was 20 at.% [23]. The chemical composition of the b phase analyzed by EDS in the 20Nb alloy is 80.2Nb–12.9Ti–4.9Ni–2.0Hf
20 10 0 0
5
10
15
20
(200) B2
Si
Si
30
(110)β (511) Ti2Ni (110)B2
40
(200)β
50
Intensity (AU)
Cold reduction ratio (%)
70
847
20Nb 15Nb 10Nb 5Nb
25
1Nb
Nb content (at.%) 30
Fig. 1. The effect of Nb content on the cold workability of Ti–49.5Ni– 15Hf–Nb alloys. Images of cold-rolled specimens are included in the lower-left corner (Ti–49.5Ni–15Hf) and at the center ((Ti–49.5Ni– 15Hf)–15Nb).
35
40
45
50
55
60
0Nb 65
2θ (degree)
Fig. 2. XRD profiles obtained at 573 K for (Ti–49.5Ni–15Hf)–(0– 20)Nb alloys. See text for details.
848
H. Y. Kim et al. / Scripta Materialia 65 (2011) 846–849 (b)
Strain
εp 300MPa 200MPa
500MPa
2%
εa
400MPa
400MPa
Strain
2%
(a) 500MPa
300MPa 200MPa 100MPa
As Af
100MPa
Ms
Mf Ms
200
300
400
500
600
200
700
300
Temperature (K)
500
600
700
600
700
Temperature (K)
(c) 500MPa
(d) 500MPa
2%
2%
400
400MPa
Strain
Strain
400MPa 300MPa 200MPa
200MPa
100MPa
100MPa
Ms
200
300
400
Ms
500
Temperature (K)
Fig. 3. Back-scattered scanning electron images of (Ti–49.5Ni–15Hf)– (0–20)Nb alloys. (a) Ti–49.5Ni–15Hf, (b) (Ti–49.5Ni–15Hf)–1Nb, (c) (Ti–49.5Ni–15Hf)–5Nb, (d) (Ti–49.5Ni–15Hf)–10Nb, (e) (Ti–49.5Ni– 15Hf)–15Nb and (f) (Ti–49.5Ni–15Hf)–20Nb.
(at.%), confirming that the b phase is a Nb-rich phase and contains more Ti than Ni. This result is also consistent with previous electron probe microanalysis in (Ti– Ni)–Nb alloys: the solubility of Ni is smaller than that of Ti in a Nb-rich b phase. As a result, it is concluded that (i) the addition of Nb forms a ductile phase in the Ti–Ni–Hf high-temperature alloy; and (ii) this ductile phase can improve the cold workability of the alloy. The shape memory properties of (Ti–Ni–Hf)–Nb alloys were investigated by thermal cycling under various constant stresses. After heating the specimen above its Af temperature (see below), a stress of 100 MPa was applied, then a cooling and heating cycle was carried out under the constant stress. After each thermal cycling, the applied stress was increased stepwise. Strain– temperature curves obtained from the 0Nb, 5Nb, 10Nb and 15Nb alloys are shown in Figure 4. The changes in strain during cooling and heating cycles, which are represented as solid and dashed lines, respectively, confirm the shape memory effect. The temperatures for martensitic transformation start, its finish, reverse transformation (i.e. austenitic) start and its finish are denoted by Ms, Mf, As and Af, respectively. The transformation stresses, hysteresis (Af–Ms), recovery strain ea, plastic strain ep, and recovery ratio (ea/ (ea + ep)) are listed in Table 1. For the 0Nb alloy, Ms and Af were determined to be 462 and 537 K, respectively, at 100 MPa, and the transformation temperatures increased with increasing applied stress, which is consistent with the Clausius–Clapeyron relationship. The addition of Nb caused a decrease in transformation temperatures: Ms (at 100 MPa) was 441, 406 and 345 K for the 5Nb, 10Nb and 15Nb alloys, respectively. The decrease in the transformation temperatures is explained by a change in matrix composition as follows. As shown in Figure 3, an increase in Nb content directly resulted in
300MPa
600
700
200
300
400
500
Temperature (K)
Fig. 4. Strain–temperature curves under various stresses for (a) Ti– 49.5Ni–15Hf, (b) (Ti–49.5Ni–15Hf)–5Nb, (c) (Ti–49.5Ni–15Hf)–10Nb and (d) (Ti–49.5Ni–15Hf)–15Nb alloys.
an increase in the volume fraction of b phase because the solubility of Nb in the matrix phase is very limited. The b phase contains more Ti than Ni as mentioned above, and thus the increase in the volume fraction of b phase raises the Ni content of the matrix. This Ni enrichment in the matrix causes the decrease in transformation temperatures. The results suggest the possibility of tailoring the transformation temperature by changing Ni content of the alloy. It is noted from Table 1 that the addition of Nb caused an increase in temperature hysteresis defined by (Af–Ms): the hysteresis (at 100 MPa) was increased from 75 to 109 K by the addition of 15 at.% Nb. It is also noted that the hysteresis increased with increasing plastic strain. Both results are consistent with previous reports that the addition of Nb to Ti–Ni increases the temperature hysteresis considerably [27] and the plastic deformation further increases the temperature hysteresis in Ti–Ni–Nb alloys [28]. The recovery strain ea and plastic strain ep was also affected by the Nb content. For the 0Nb alloy, ea increased with increasing stress up to 400 MPa, then decreased slightly at 500 MPa due to an increase in plastic deformation ep. The maximum ea of 4.6% was obtained at 400 MPa in the 0Nb alloy. It is seen that ea was slightly decreased by the addition of Nb. The maximum ea was measured to be 4.1%, 3.9% and 3.5% for the 5Nb, 10Nb and 15Nb alloys, respectively. It can also be clearly seen that ep decreased with increasing Nb content when comparing strain–temperature curves obtained at a fixed stress, e.g. 400 MPa, resulting in an increase of the shape recovery ratio. The decreases in both ea and ep can be explained by the change in microstructure. As mentioned above, the volume fraction of b phase increased with increasing Nb content. This caused the decrease in the volume fraction of the matrix phase undergoing martensitic transformation, resulting in the decrease of ea. In addition, the formation of the b phase causes the decrease in the grain size of the matrix phase. In particular, the 15Nb alloy exhibited an almost perfect
H. Y. Kim et al. / Scripta Materialia 65 (2011) 846–849
849
Table 1. Transformation temperatures and shape memory properties of Ti–49.5Ni–15Hf, (Ti–49.5Ni–15Hf)–5Nb, (Ti–49.5Ni–15Hf)–10Nb and (Ti–49.5Ni–15Hf)–15Nb alloys. Specimen
Stress (MPa)
Mf (K)
Ms (K)
As (K)
Af (K)
Hysteresis (K)
Recovery strain
Plastic strain
Recovery ratio (%)
0Nb
100 200 300 400 500
431 430 436 446 452
462 465 482 499 508
502 507 516 527 536
537 544 564 597 625
75 79 82 98 117
0.74 2.42 4.08 4.55 3.95
0.08 0.15 0.45 1.16 2.66
90 94 90 80 60
5Nb
100 200 300 400 500
411 394 398 411 414
441 437 442 461 484
483 474 478 492 507
518 513 549 587 621
77 76 107 126 137
0.56 1.45 3.28 4.09 3.62
0.06 0.17 0.27 0.92 1.90
90 90 92 82 66
10Nb
100 200 300 400 500
367 359 362 360 391
406 403 411 435 453
452 441 445 458 479
489 484 512 552 571
83 81 101 117 118
0.65 1.59 3.04 3.87 3.64
0.11 0.17 0.29 0.50 1.03
86 90 91 89 78
15Nb
100 200 300 400 500
331 325 327 328 333
345 356 363 390 422
433 426 424 417 419
454 451 461 493 534
109 95 98 103 112
0.22 0.99 2.06 2.88 3.45
0.03 0.15 0.19 0.21 0.24
88 87 92 93 93
shape memory effect with very small ep because of the fine lamellar structure. In this work we investigated the possibility of the development of cold-workable Ti–Ni–Hf-base high-temperature shape memory alloys. The addition of Nb to a Ti–49.5Ni–15Hf high-temperature shape memory alloy was effective at improving the cold workability through the formation of a ductile phase. The shape memory effect was confirmed for (Ti–Ni–Hf)–(0–15)Nb alloys during thermal cycling under stress. As the Nb content increased in the (Ti–Ni–Hf)–(0–15)Nb alloys, the plastic strain decreased and the shape recovery ratio increased while the transformation temperature decreased. This work was partially supported by the Grants-in-Aid for Fundamental Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan and by the WCU (World Class University) Program from the National Research Foundation of Korea (No. R32-2008-000-20093-0). [1] S. Miyazaki, Y. Ohmi, K. Otsuka, Y. Suzuki, J. de Phys. 43 (Suppl. 12) (1982) 255–260. [2] S. Miyazaki, K. Otsuka, ISIJ Inter. 29 (1989) 353–377. [3] S. Miyazaki, in: T.W. Duerig et al. (Eds.), Engineering Aspects of Shape Memory Alloys, Butterworth-Heinemann, 1990, pp. 394–413. [4] K. Yamauchi, I. Ohkata, K. Tsuchiya, S. Miyazaki, Shape Memory and Superelastic Alloys: Technologies and Applications, first ed., Woodhead Publishing, Cambridge, 2011. [5] G.S. Firstov, J. Van Humbeeck, Y.N. Koval, J. Intelligent Mater. Syst. Struct. 17 (2006) 1041–1047. [6] J. Ma, I. Karaman, R.D. Noebe, Inter. Mater. Rev. 55 (2010) 257–315. [7] Y.Q. Ma, S.Y. Yang, Y. Liu, X.J. Liu, Acta Mater. 57 (2009) 3232–3241. [8] J. Dadda, H.J. Maier, I. Karaman, Y.I. Chumlyakov, Acta Mater. 57 (2009) 6123–6134.
[9] L. Kovarik, F. Yang, A. Garg, D. Diercks, M. Kaufman, R.D. Noebe, M.J. Mills, Acta Mater. 58 (2010) 4460– 4473. [10] Y. Li, Y. Xin, L. Chai, Y.Q. Ma, H.B. Xu, Acta Mater. 58 (2010) 3655–3663. [11] P.J.S. Buenconsejo, H.Y. Kim, S. Miyazaki, Scripta Mater. 64 (2011) 1114–1117. [12] P.J.S. Buenconsejo, H.Y. Kim, H. Hosoda, S. Miyazaki, Acta Mater. 57 (2009) 1068–1077. [13] G.S. Bigelow, S.A. Padula, A. Garg, D. Gaydosh, R.D. Noebe, Metal. Mater. Trans. A 41A (2010) 3065–3079. [14] R. Noebe, D. Gaydosh, S. Padula, A. Garg, T. Biles, M. Nathal, in: SPIE Conference Proceedings of the Smart Structures and Materials 2005: Active Materials: Behavior and Mechanics, vol. 5761, 2005, pp. 364–375. [15] H.Y. Kim, Y. Yuan, T.H. Nam, S. Miyazaki, Inter. J. Smart Nano Mater. 2 (2011) 9–21. [16] S.F. Hsieh, S.K. Wu, Mater. Charact. 41 (1998) 151–162. [17] S.F. Hsieh, S.K. Wu, Mater. Charact. 45 (2000) 143–152. [18] B. Kockar, I. Karaman, J.I. Kim, Y.J. Chumlyakov, Scripta Mater. 54 (2006) 2203–2208. [19] X.L. Meng, Y.F. Zheng, W. Cai, L.C. Zhao, J. Alloys Compd. 372 (2004) 180–186. [20] N. Okada, Y. Fujii, Y. Ishikawa, M. Onoda, H.Y. Kim, S. Miyazaki, J. Jpn. Inst. Met. 72 (2008) 152–157. [21] P. Olier, J.C. Brachet, J.L. Bechade, C. Foucher, G. Guenin, J. de Phys. IV C8 (1995) 741–746. [22] D.R. Angst, P.E. Thoma, M.Y. Kao, J. de Phys. IV C8 (1995) 747–752. [23] M. Piao, S. Miyazaki, K. Otsuka, Mater. Trans. JIM 33 (1992) 346–353. [24] L.C. Zhao, T.W. Duerig, C.M. Wayman, Proc. MRS Int. Mtg. Adv. Mats., Tokyo 9 (1989) 171–176. [25] S.F. Hsieh, S.K. Wu, H.C. Lin, C.H. Yang, J. Alloys Compd. 387 (2005) 121–127. [26] Y.F. Zheng, B.B. Zhang, B.L. Wang, Y.B. Wang, L. Li, Q.B. Yang, L.S. Cui, Acta Biomater. 7 (2011) 2758–2767. [27] K.N. Melton, J.L. Proft, T.W. Duerig, Proc. MRS Int. Mtg. Adv. Mats., Tokyo 9 (1989) 165–170. [28] C.S. Zhang, L.C. Zhao, T.W. Duerig, C.M. Wayman, Scripta Metall. 24 (1990) 1807–1812.
Acta Materialia 65 (2011) 846–849 www.elsevier.com/locate/scriptamat
Cold workability and shape memory properties of novel Ti–Ni–Hf–Nb high-temperature shape memory alloys Hee Young Kim,a,⇑ Takafumi Jinguu,a Tae-hyun Namb and Shuichi Miyazakia,b,c,⇑ a
Institute of Materials Science, University of Tsukuba, Tsukuba, Ibaraki 305-8573, Japan School of Materials Science and Engineering & ERI, Gyeongsang National University, 900 Gazwadong, Jinju, Gyeongnam 660-701, Republic of Korea c Center of Excellence for Advanced Materials Research, King Abdulaziz University, P.O. Box 80203, Jeddah 21589, Saudi Arabia b
Received 7 June 2011; revised 26 July 2011; accepted 26 July 2011 Available online 30 July 2011
The effect of Nb content on the cold workability, microstructure and shape memory properties of Ti–Ni–Hf–Nb alloys is investigated. The addition of Nb to Ti–Ni–Hf improves the cold workability due to the formation of a soft Nb-rich phase. The Ti–Ni– Hf–Nb alloys exhibit a high-temperature shape memory effect during heating and cooling cycles under various stress levels. As the Nb content increases, the shape memory properties of the Ti–Ni–Hf–Nb alloys become more stable although the shape recovery strain decreases. Ó 2011 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. Keywords: Shape memory alloys; Phase transformation; Cold working; Microstructure
Since the discovery of methods for achieving superelasticity and for stabilizing the shape memory effect in Ti–Ni alloys [1–3], these alloys have over the last couple of decades been increasingly used in various fields of industry including home appliances, medical devices and electronic devices owing to their unique properties of shape memory effect and superelasticity [4]. However, most practical applications of shape memory alloys are limited to temperatures below 373 K because of the low transformation temperatures of commercially available shape memory alloys such as Ti–Ni and Cubase alloys. Recently shape memory alloys operating at high environmental temperatures (above 373 K) have attracted attention as solid-state actuators for use in aircraft engines, automobile engines and energy-generation systems due to their large stroke, high recovery force and high work output [5,6]. Up to now several types of shape memory alloys have been reported to exhibit a martensitic transformation start temperature above 373 K [5–12]. Among these, Ti–Ni–X alloys (where X = Pt, Pd, Hf or Zr) have
⇑ Corresponding
authors. Address: Institute of Materials Science, University of Tsukuba, Tsukuba, Ibaraki 305-8573, Japan. Tel./fax: +81 29 853 6942 (H.Y. Kim). Tel./fax: +81 29 853 5283 (S. Miyazaki); e-mail addresses: [email protected]; [email protected]
received the most research attention [13–19]. The martensitic transformation temperatures increase with the replacement of Ni with Pd and Pt above a critical concentration [13–15]. Ti–Ni–(Pd, Pt) alloys exhibit a relatively stable shape memory effect at above 373 K with a small hysteresis. However, the high cost of precious metal alloying elements has the limited practical applications of Ti– Ni–(Pd, Pt) shape memory alloys. It has also been widely recognized that substitution of Zr or Hf for Ti increases the transformation temperature of Ti–Ni shape memory alloys [16–19]. The Ti–Ni–(Zr, Hf) alloys are regarded as more practical systems compared with the Ti–Ni– (Pd, Pt) alloys because of the relatively low price of the raw materials required; however, for several reasons no practical devices have yet been developed. One of the most serious problems with Ti–Ni–(Zr, Hf) alloys is their low workability: their hardness is increased and their cold workability and ductility are decreased significantly by the addition of Zr and Hf [6,20–22]. Improvement of the cold workability of Ti–Ni–(Zr, Hf) alloys is vital for manufacturing them into suitable sizes and shapes and for microstructure control. The introduction of a ductile phase is an effective way to improve the cold workability of brittle materials. It has been reported that the addition of certain alloying elements such as Nb, W and Ag to Ti–Ni binary alloys results in the formation of a ductile phase [23–26]. However, up to now there has been no
1359-6462/$ - see front matter Ó 2011 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.scriptamat.2011.07.049
H. Y. Kim et al. / Scripta Materialia 65 (2011) 846–849
report on Ti–Ni–(Zr, Hf) shape memory alloys containing a ductile phase. In this study, a new alloy system, Ti–Ni–Hf–Nb, was designed, and the effect of Nb content on the cold workability, microstructure and shape memory properties of Ti–Ni–Hf–Nb high-temperature shape memory alloys was investigated. (Ti–49.5Ni–15Hf)100 X–XNb (X = 0–20 at.%) alloys were fabricated by the argon arc-melting method. The ingots were sealed in a quartz tube under vacuum and homogenized at 1223 K for 7.2 ks. Specimens for all measurements were cut by an electrodischarge machine. The oxidized or damaged surface was removed by mechanical polishing. Cold workability was evaluated by measuring the minimum thickness attained before fracture upon cold rolling at room temperature (RT) where specimens were prepared in the following dimensions: 30 mm in length 10 mm in width 1 mm in thickness. Microstructure and phase constituent were investigated by Xray diffraction (XRD) and scanning electron microscopy (SEM). XRD measurements were conducted at RT and 573 K with Cu Ka radiation. The surfaces of specimens for SEM observation were polished using an alumina slurry (0.3 lm). Shape memory properties were investigated by thermal cycling tests (with a heating and cooling rate of 10 K min 1) under various stress levels. The thermal cycling tests were carried out using a thermomechanical analyzer incorporating a linear variable differential transformer to measure the strain. The tests were conducted in helium gas atmosphere in order to prevent oxidation and improve thermal conductivity. The dimensions of the specimens were 13 mm in gauge length, 0.7 mm in width and 0.15 mm in thickness. Figure 1 shows the effect of Nb content on the cold workability of Ti–49.5Ni–15Hf–(0–20)Nb alloys. A Ti– 49.5Ni–15Hf specimen fractured instantly after cracking at 19% reduction of thickness (see the image inserted at the lower-left corner of Fig. 1). The total reduction in thickness until fracture increased with increasing Nb content. In particular, the addition of 10 at.% Nb and more allowed the Ti–Ni–Hf alloy to be cold rolled up to 60% reduction in thickness without fracture, although small cracks were introduced from side surfaces as shown in the inserted image.
80
60
Figure 2 shows XRD profiles of (Ti–49.5Ni– 15Hf)100 X–XNb (X = 0–20 at.%) alloys heat treated at 1223 K for 7.2 ks. The XRD profiles were obtained at 573 K (above the reverse transformation finish temperatures of the alloys) in order to provide better clarity and allow for detailed analysis of the secondary phase because the peaks from the secondary phase overlapped with the peaks from the martensite structure (B19’ with a monoclinic structure) at RT. Hereafter, each alloy is referred to according to its Nb concentration, i.e. 0Nb, 10Nb, etc. The 0Nb alloy is mainly composed of a B2 phase with a small amount of Ti2Ni type phase, which can be identified by the weak reflection at 2h = 40.5°. In addition to B2 and Ti2Ni type phases, an additional phase (b phase), which has a disordered body-centered cubic (bcc) structure was observed in the alloys containing 5 at.% Nb and more. The lattice constant of the b phase is determined to be 0.3315 nm, which is very close to that of pure Nb. It is also noted that the lattice constant of the b phase is insensitive to Nb content, while the peak intensity of the b phase increased with increasing Nb content. Figure 3 shows scanning electron images of (Ti– 49.5Ni–15Hf)–(0–20)Nb alloys. Observation was carried out in back-scattered electron (BSE) mode since the BSE image efficiently distinguishes different phases. For the 0Nb alloy (Fig. 3a), the Ti2Ni type phase can be seen as relatively dark particles. In addition to Ti2Ni type phase, the b phase (which appears white on the images) is seen in the Nb-containing alloys. It is noted that the b phase was observed in the 1Nb alloy (Fig. 3b) although the peaks from the b phase could not be detected by the XRD analysis due to the small volume fraction. This result implies that the solution limit of Nb in the matrix phase, i.e. B2 phase, is less than 1 at.%. The amount of the b phase increased with increasing Nb content. The 15Nb alloy (Fig. 3e) reveals a fully lamellar microstructure, which is a characteristic of eutectic solidification. It is also noted that the 10Nb alloy exhibits a typical microstructure of off-eutectic solidification, which is composed of a primary dendrite phase (B2) and interdendritic lamellar structure. On the other hand, the primary dendrite phase in the 20Nb alloy is the b phase as shown in Figure 3f. A similar change in microstructure by the addition of Nb was reported in a Ti–Ni binary alloy where the Nb content in the eutectic composition was 20 at.% [23]. The chemical composition of the b phase analyzed by EDS in the 20Nb alloy is 80.2Nb–12.9Ti–4.9Ni–2.0Hf
20 10 0 0
5
10
15
20
(200) B2
Si
Si
30
(110)β (511) Ti2Ni (110)B2
40
(200)β
50
Intensity (AU)
Cold reduction ratio (%)
70
847
20Nb 15Nb 10Nb 5Nb
25
1Nb
Nb content (at.%) 30
Fig. 1. The effect of Nb content on the cold workability of Ti–49.5Ni– 15Hf–Nb alloys. Images of cold-rolled specimens are included in the lower-left corner (Ti–49.5Ni–15Hf) and at the center ((Ti–49.5Ni– 15Hf)–15Nb).
35
40
45
50
55
60
0Nb 65
2θ (degree)
Fig. 2. XRD profiles obtained at 573 K for (Ti–49.5Ni–15Hf)–(0– 20)Nb alloys. See text for details.
848
H. Y. Kim et al. / Scripta Materialia 65 (2011) 846–849 (b)
Strain
εp 300MPa 200MPa
500MPa
2%
εa
400MPa
400MPa
Strain
2%
(a) 500MPa
300MPa 200MPa 100MPa
As Af
100MPa
Ms
Mf Ms
200
300
400
500
600
200
700
300
Temperature (K)
500
600
700
600
700
Temperature (K)
(c) 500MPa
(d) 500MPa
2%
2%
400
400MPa
Strain
Strain
400MPa 300MPa 200MPa
200MPa
100MPa
100MPa
Ms
200
300
400
Ms
500
Temperature (K)
Fig. 3. Back-scattered scanning electron images of (Ti–49.5Ni–15Hf)– (0–20)Nb alloys. (a) Ti–49.5Ni–15Hf, (b) (Ti–49.5Ni–15Hf)–1Nb, (c) (Ti–49.5Ni–15Hf)–5Nb, (d) (Ti–49.5Ni–15Hf)–10Nb, (e) (Ti–49.5Ni– 15Hf)–15Nb and (f) (Ti–49.5Ni–15Hf)–20Nb.
(at.%), confirming that the b phase is a Nb-rich phase and contains more Ti than Ni. This result is also consistent with previous electron probe microanalysis in (Ti– Ni)–Nb alloys: the solubility of Ni is smaller than that of Ti in a Nb-rich b phase. As a result, it is concluded that (i) the addition of Nb forms a ductile phase in the Ti–Ni–Hf high-temperature alloy; and (ii) this ductile phase can improve the cold workability of the alloy. The shape memory properties of (Ti–Ni–Hf)–Nb alloys were investigated by thermal cycling under various constant stresses. After heating the specimen above its Af temperature (see below), a stress of 100 MPa was applied, then a cooling and heating cycle was carried out under the constant stress. After each thermal cycling, the applied stress was increased stepwise. Strain– temperature curves obtained from the 0Nb, 5Nb, 10Nb and 15Nb alloys are shown in Figure 4. The changes in strain during cooling and heating cycles, which are represented as solid and dashed lines, respectively, confirm the shape memory effect. The temperatures for martensitic transformation start, its finish, reverse transformation (i.e. austenitic) start and its finish are denoted by Ms, Mf, As and Af, respectively. The transformation stresses, hysteresis (Af–Ms), recovery strain ea, plastic strain ep, and recovery ratio (ea/ (ea + ep)) are listed in Table 1. For the 0Nb alloy, Ms and Af were determined to be 462 and 537 K, respectively, at 100 MPa, and the transformation temperatures increased with increasing applied stress, which is consistent with the Clausius–Clapeyron relationship. The addition of Nb caused a decrease in transformation temperatures: Ms (at 100 MPa) was 441, 406 and 345 K for the 5Nb, 10Nb and 15Nb alloys, respectively. The decrease in the transformation temperatures is explained by a change in matrix composition as follows. As shown in Figure 3, an increase in Nb content directly resulted in
300MPa
600
700
200
300
400
500
Temperature (K)
Fig. 4. Strain–temperature curves under various stresses for (a) Ti– 49.5Ni–15Hf, (b) (Ti–49.5Ni–15Hf)–5Nb, (c) (Ti–49.5Ni–15Hf)–10Nb and (d) (Ti–49.5Ni–15Hf)–15Nb alloys.
an increase in the volume fraction of b phase because the solubility of Nb in the matrix phase is very limited. The b phase contains more Ti than Ni as mentioned above, and thus the increase in the volume fraction of b phase raises the Ni content of the matrix. This Ni enrichment in the matrix causes the decrease in transformation temperatures. The results suggest the possibility of tailoring the transformation temperature by changing Ni content of the alloy. It is noted from Table 1 that the addition of Nb caused an increase in temperature hysteresis defined by (Af–Ms): the hysteresis (at 100 MPa) was increased from 75 to 109 K by the addition of 15 at.% Nb. It is also noted that the hysteresis increased with increasing plastic strain. Both results are consistent with previous reports that the addition of Nb to Ti–Ni increases the temperature hysteresis considerably [27] and the plastic deformation further increases the temperature hysteresis in Ti–Ni–Nb alloys [28]. The recovery strain ea and plastic strain ep was also affected by the Nb content. For the 0Nb alloy, ea increased with increasing stress up to 400 MPa, then decreased slightly at 500 MPa due to an increase in plastic deformation ep. The maximum ea of 4.6% was obtained at 400 MPa in the 0Nb alloy. It is seen that ea was slightly decreased by the addition of Nb. The maximum ea was measured to be 4.1%, 3.9% and 3.5% for the 5Nb, 10Nb and 15Nb alloys, respectively. It can also be clearly seen that ep decreased with increasing Nb content when comparing strain–temperature curves obtained at a fixed stress, e.g. 400 MPa, resulting in an increase of the shape recovery ratio. The decreases in both ea and ep can be explained by the change in microstructure. As mentioned above, the volume fraction of b phase increased with increasing Nb content. This caused the decrease in the volume fraction of the matrix phase undergoing martensitic transformation, resulting in the decrease of ea. In addition, the formation of the b phase causes the decrease in the grain size of the matrix phase. In particular, the 15Nb alloy exhibited an almost perfect
H. Y. Kim et al. / Scripta Materialia 65 (2011) 846–849
849
Table 1. Transformation temperatures and shape memory properties of Ti–49.5Ni–15Hf, (Ti–49.5Ni–15Hf)–5Nb, (Ti–49.5Ni–15Hf)–10Nb and (Ti–49.5Ni–15Hf)–15Nb alloys. Specimen
Stress (MPa)
Mf (K)
Ms (K)
As (K)
Af (K)
Hysteresis (K)
Recovery strain
Plastic strain
Recovery ratio (%)
0Nb
100 200 300 400 500
431 430 436 446 452
462 465 482 499 508
502 507 516 527 536
537 544 564 597 625
75 79 82 98 117
0.74 2.42 4.08 4.55 3.95
0.08 0.15 0.45 1.16 2.66
90 94 90 80 60
5Nb
100 200 300 400 500
411 394 398 411 414
441 437 442 461 484
483 474 478 492 507
518 513 549 587 621
77 76 107 126 137
0.56 1.45 3.28 4.09 3.62
0.06 0.17 0.27 0.92 1.90
90 90 92 82 66
10Nb
100 200 300 400 500
367 359 362 360 391
406 403 411 435 453
452 441 445 458 479
489 484 512 552 571
83 81 101 117 118
0.65 1.59 3.04 3.87 3.64
0.11 0.17 0.29 0.50 1.03
86 90 91 89 78
15Nb
100 200 300 400 500
331 325 327 328 333
345 356 363 390 422
433 426 424 417 419
454 451 461 493 534
109 95 98 103 112
0.22 0.99 2.06 2.88 3.45
0.03 0.15 0.19 0.21 0.24
88 87 92 93 93
shape memory effect with very small ep because of the fine lamellar structure. In this work we investigated the possibility of the development of cold-workable Ti–Ni–Hf-base high-temperature shape memory alloys. The addition of Nb to a Ti–49.5Ni–15Hf high-temperature shape memory alloy was effective at improving the cold workability through the formation of a ductile phase. The shape memory effect was confirmed for (Ti–Ni–Hf)–(0–15)Nb alloys during thermal cycling under stress. As the Nb content increased in the (Ti–Ni–Hf)–(0–15)Nb alloys, the plastic strain decreased and the shape recovery ratio increased while the transformation temperature decreased. This work was partially supported by the Grants-in-Aid for Fundamental Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology, Japan and by the WCU (World Class University) Program from the National Research Foundation of Korea (No. R32-2008-000-20093-0). [1] S. Miyazaki, Y. Ohmi, K. Otsuka, Y. Suzuki, J. de Phys. 43 (Suppl. 12) (1982) 255–260. [2] S. Miyazaki, K. Otsuka, ISIJ Inter. 29 (1989) 353–377. [3] S. Miyazaki, in: T.W. Duerig et al. (Eds.), Engineering Aspects of Shape Memory Alloys, Butterworth-Heinemann, 1990, pp. 394–413. [4] K. Yamauchi, I. Ohkata, K. Tsuchiya, S. Miyazaki, Shape Memory and Superelastic Alloys: Technologies and Applications, first ed., Woodhead Publishing, Cambridge, 2011. [5] G.S. Firstov, J. Van Humbeeck, Y.N. Koval, J. Intelligent Mater. Syst. Struct. 17 (2006) 1041–1047. [6] J. Ma, I. Karaman, R.D. Noebe, Inter. Mater. Rev. 55 (2010) 257–315. [7] Y.Q. Ma, S.Y. Yang, Y. Liu, X.J. Liu, Acta Mater. 57 (2009) 3232–3241. [8] J. Dadda, H.J. Maier, I. Karaman, Y.I. Chumlyakov, Acta Mater. 57 (2009) 6123–6134.
[9] L. Kovarik, F. Yang, A. Garg, D. Diercks, M. Kaufman, R.D. Noebe, M.J. Mills, Acta Mater. 58 (2010) 4460– 4473. [10] Y. Li, Y. Xin, L. Chai, Y.Q. Ma, H.B. Xu, Acta Mater. 58 (2010) 3655–3663. [11] P.J.S. Buenconsejo, H.Y. Kim, S. Miyazaki, Scripta Mater. 64 (2011) 1114–1117. [12] P.J.S. Buenconsejo, H.Y. Kim, H. Hosoda, S. Miyazaki, Acta Mater. 57 (2009) 1068–1077. [13] G.S. Bigelow, S.A. Padula, A. Garg, D. Gaydosh, R.D. Noebe, Metal. Mater. Trans. A 41A (2010) 3065–3079. [14] R. Noebe, D. Gaydosh, S. Padula, A. Garg, T. Biles, M. Nathal, in: SPIE Conference Proceedings of the Smart Structures and Materials 2005: Active Materials: Behavior and Mechanics, vol. 5761, 2005, pp. 364–375. [15] H.Y. Kim, Y. Yuan, T.H. Nam, S. Miyazaki, Inter. J. Smart Nano Mater. 2 (2011) 9–21. [16] S.F. Hsieh, S.K. Wu, Mater. Charact. 41 (1998) 151–162. [17] S.F. Hsieh, S.K. Wu, Mater. Charact. 45 (2000) 143–152. [18] B. Kockar, I. Karaman, J.I. Kim, Y.J. Chumlyakov, Scripta Mater. 54 (2006) 2203–2208. [19] X.L. Meng, Y.F. Zheng, W. Cai, L.C. Zhao, J. Alloys Compd. 372 (2004) 180–186. [20] N. Okada, Y. Fujii, Y. Ishikawa, M. Onoda, H.Y. Kim, S. Miyazaki, J. Jpn. Inst. Met. 72 (2008) 152–157. [21] P. Olier, J.C. Brachet, J.L. Bechade, C. Foucher, G. Guenin, J. de Phys. IV C8 (1995) 741–746. [22] D.R. Angst, P.E. Thoma, M.Y. Kao, J. de Phys. IV C8 (1995) 747–752. [23] M. Piao, S. Miyazaki, K. Otsuka, Mater. Trans. JIM 33 (1992) 346–353. [24] L.C. Zhao, T.W. Duerig, C.M. Wayman, Proc. MRS Int. Mtg. Adv. Mats., Tokyo 9 (1989) 171–176. [25] S.F. Hsieh, S.K. Wu, H.C. Lin, C.H. Yang, J. Alloys Compd. 387 (2005) 121–127. [26] Y.F. Zheng, B.B. Zhang, B.L. Wang, Y.B. Wang, L. Li, Q.B. Yang, L.S. Cui, Acta Biomater. 7 (2011) 2758–2767. [27] K.N. Melton, J.L. Proft, T.W. Duerig, Proc. MRS Int. Mtg. Adv. Mats., Tokyo 9 (1989) 165–170. [28] C.S. Zhang, L.C. Zhao, T.W. Duerig, C.M. Wayman, Scripta Metall. 24 (1990) 1807–1812.