Kazunari Uchida 1, Naoto Shigenaka 1, Toshio Sakuma 2, Yuji Sutou 3 and Kiyoshi Yamauchi 3

Transcription

1 Materials Transactions, Vol. 49, No. 7 (28) pp. 165 to 1655 #28 The Japan Institute of Metals Effects of Pre-Strain and Heat Treatment Temperature on Phase Transformation Temperature and Shape Recovery Stress of Ti-Ni-Nb Shape Memory Alloys for Pipe Joint Applications Kazunari Uchida 1, Naoto Shigenaka 1, Toshio Sakuma 2, Yuji Sutou 3 and Kiyoshi Yamauchi 3 1 Nuclear Plant Service Department, Hitachi-GE Nuclear Energy, Ltd., Hitachi , Japan 2 Faculty of Engineering, Oita University, Oita , Japan 3 Tohoku University Biomedical Engineering Research Organization, TUBERO Aobayama Material Science Branch, Sendai , Japan Effects of pre-strain and heat treatment on shape recovery stress and phase transformation temperature in Ti-Ni-Nb shape memory alloys with various Nb content (6, 9, and 12 mol%) were investigated by tensile testing. The recovery stress of Ti-Ni-Nb alloys increased with increasing the pre-strain and then decreased after reaching the maximum recovery stress of around 45 5 MPa at about 9% pre-strain. Martensitic transformation temperatures (M s, A s, A f ) also increased with increasing the applied pre-strain, while the relation between A s and prestrain did not depend on the Nb contents and Ni/Ti ratio. Moreover, the recovery stress in the Ti-Ni-Nb alloy heat treated at 673 and 773 K after solution treatment was slightly higher than that in the solution-treated alloy. These behaviors were examined relative to the increased defect density due to pre-straining and the elimination of defects by heat treatment. [doi:1.232/matertrans.mra27266] (Received January 22, 28; Accepted April 4, 28; Published June 25, 28) Keywords: shape memory alloy, titanium-nickel-niobium alloy, wide hysteresis transformation, pre-deformation, recovery stress, pipe joint, shape memory alloy (SMA) joint ring 1. Introduction Among all new materials now being developed, much public attention is being focused on shape memory alloys (SMAs) whose functions can be used as machine elements. In particular, Ti-Ni SMA exhibits excellent mechanical properties and corrosion resistance, and is therefore already used in industrial, medical, and other applications. 1 3) When shape recovery and recovery stress due to the shape memory effect (SME) are utilized, the SMA is usually deformed in the martensite phase (M phase) and then heated above the reverse transformation finishing temperature (A f ). However, since the high deformation of the SMA causes internal slip deformation, perfect shape recovery dose not occur even by heating above the A f. Moreover, the recovery stress and the reverse transformation starting (A s ) and the A f temperatures also depend on the deformation conditions in the M phase state. 4 1) Loading the SMA at a temperature slightly higher than the A s causes the M phase having lower critical stress in slip deformation to undergo slip deformation on a priority basis. 11) Even when heated, the slip-deformed M phase does not revert to its original state, the austenite phase (A phase), but remains unchanged. The coauthors have defined the volume fraction of the M phase remaining in the A phase as the remaining M phase fraction. They then demonstrated that it could be used as an indicator of dislocation introduced by pre-strain. 12,13) In M transformation, the material stores elastic strain energy. This elastic strain energy reportedly assists in reverse transformation while resisting M transformation. 7 1) Consequently, the release of elastic strain energy in response to slip deformation when deformation occurs in the M phase state will cause a rise in transformation temperature. The Ti-Ni-Nb shape memory alloy has a high transformation temperature hysteresis, and has thus been used in pipe joints and similar applications ) The pipes of SMA joint rings are expanded to the specified dimensions, and then stored and controlled accordingly. When installed, the SMA joint rings are heated and used to couple pipes. The storage and control of SMA joints require that the temperature is controlled to prevent the expanded joint rings from reverting to their previous shapes. The SMA reverts to its previous shape due to its shape memory effect upon reaching the start temperature of its reverse transformation or higher. Therefore, the start temperature of reverse transformation of SMA joints must be no less than the normal storage and control temperature. In other words, these SMA joints should maintain their specified functions under storage and control temperature conditions, and exhibit those specified functions under operating conditions. Therefore, the SMA composition, fabrication, heat treatment, and other manufacturing and fabrication conditions must be optimized in order for the SMA joints to achieve the specified performance. The present study is therefore intended to impose various pre-strains on Ti-Ni-Nb shape memory alloys, heat the alloys by using the strain constraints generated after prestrain removal, examine the amounts of recovery strain and recovery stress, the transformation temperature, and other conditions after imposing the pre-strain, and finally determine the effects of pre-strain on the transformation and deformation characteristics after the imposition of pre-strain. 2. Experiment 2.1 Specimens The specimens of Ti-Ni-Nb SMAs were fabricated in a high-frequency induction vacuum furnace. Two kinds of chemical compositions were selected: (1) one having a atomic ratio of Ni and Ti (Ni/Ti) as a constant of 1.88 and with several Nb contents (6, 9, and 12 at%), and (2) one

2 Effects of Pre-Strain and Heat Treatment Temperature on Phase Transformation Temperature 1651 Table 1 Chemical Composition of Ni-Ti-Nb Alloys. Nominal Analytical Nb (mol%) Ni/Ti Ni (mol%) Ti (mol%) C (ppm) O (ppm) Nb (mol%) Ni/Ti having a Ni/Ti atomic ratio of 1.1 and with Nb contents of 6 and 9 at%. A total of five alloys were fabricated. The alloy compositions were analyzed using inductively coupled plasma atomic emission spectroscopy (ICP-AES). Table 1 summarizes the analysis results of said alloy compositions. No serious deviations were recognized between the nominal and analytical compositions. The nominal compositions were subsequently used for expressing the alloys and handling data. After alloying, the ingots were hot-forged and hot-extrude to make coils with a diameter of 7.8 mm. Next, the manufacturing process of cold-drawn and solid solution treatment at 1173 K for seconds was repeated to make wires with a diameter from 7.8 mm to 1 mm, and solid solution treatment was given final process. Then visual and eddy current inspections were conducted to find wire defects, and no cracked parts were used for specimens. The specimen used in this study is wire with 1 mm in diameter, and 7 mm gage length. After solid solution treatment at 1173 K for seconds, thermal aging treatment was conducted at 673 or 773 K for 3.6 ks to investigate the effect of heat treatment on shape recovery stress and transformation temperature. Moreover, pre-straining from to 15% was conducted prior to heat treatment for investigating the effects of pre-strain on shape recovery stress and transformation temperature. 2.2 Experimental procedure Figure 1 shows a schematic drawing of the stress-strain curve used in testing the amount of elastic recovery after prestrain imposition, the recovery stress, and transformation temperature. In this figure, the strain is strain obtained by the extensometer. The strain was measured with the extensometer installing in the loading shaft outside of heating/cooling device. Moreover, the loading shaft is cooled, and removes the influence of the strain of the test machine itself at the temperature change. An expansion of the loading shaft, slipping of the specimen chuck, and an expansion of the specimen not uniform are enumerated as an item that influences the accuracy of the strain excluding these. It corresponds about this as follows. The expansion of the loading shaft doesn t influence it because the loading shaft is tungsten alloy of high strength, and the loading shaft diameter about 2 times the diameter of the specimen. Moreover, both ends of the specimen are fixed with a V- shaped chuck, and when slipping is caused between the specimen and the chuck even if it is a little, the test data is excluded. In addition, reduction of area in fracture of the Stress A B F Pre-strain C Elastic Recovery Strain Fig. 1 Schematic drawing of the stress-strain curve for shape memory alloys. specimen is very small, and an expansion of the specimen not uniform hardly influences it. Specimens used were completely transformed to the M phase by cooling down to point M f -3 K. Specimens were loaded from point A at a given temperature of point A s -2 K. Elastic deformation (A!B), reorientation of the martensite variants (B!C), and work hardening (C!D) occurred. Pre-strain is the total strain from A to D at a strain rate of 1.2 mm/min, and as a result, some elastic and plastic strains were introduced in the specimen. After a specified strain (pre-strain) was imposed, the test pieces were then unloaded from point D to point E, with the difference between the strain at D and that at E becoming the amount of elastic recovery. Next, the test pieces were heated up at a heating rate of 3 K/min with the strain constrained at point E. After reaching point A f or higher, the test pieces were cooled down to reach point M s or less. The strain is constrained in both heating and cooling processes. Figure 2 shows a schematic drawing of the stress-temperature curve used in the heating and cooling processes. Upon heating the test pieces at stress after unloading, the stress remained unchanged until the temperature reached point A s. When the temperature exceeded point A s, the stress increased. Beyond point A f, the stress remained constant and did not increase under further heating. down from the temperature range beyond point A f caused no change in stress up to point M s. Further cooling resulted in a E D

3 1652 K. Uchida, N. Shigenaka, T. Sakuma, Y. Sutou and K. Yamauchi Stress Shape Recovery Stress Loading M s Temperature Fig. 2 Schematic drawing of the stress-temperature curve for shape memory alloys. A s A f Elastic Recovery Strain, ε e Stress, MPa Fig Pre-Strain:7.5% Pre-strain, One example of the stress-strain curve for shape memory alloys. decline in stress. These heating and cooling tests were conducted to measure the shape recovery stress and transformation temperatures of A s, A f and M s, respectively, with regard to pre-strain. Elastic Recovery Strain, ε e Fig Pre-Strain, ε p 12. Variation in elastic recovery strain with pre-strain. Pre-Strain : 1% Results and Discussion 3.1 Elastic recovery strain The amount of elastic recovery compared to the amount of pipe expansion occurring at the joint rings is an important factor in determining the inner ring diameter. Figure 3 shows one example of the stress-strain curve for shape memory alloys. This curve was conducted to measure the elastic recovery strain with pre-strain of 7.5%. Moreover, the elasticity recovery strain of other specimens was measured by the method similar to this curve. Figure 4 shows how elastic recovery strain varies with the pre-strain of a solution heat treatment material. As shown in the figure, elastic recovery strain increases along with the rise in prestrain. Moreover, as the pre-strain rises, the percentage of elastic recovery strain also increases. About 2% of the material recovers elasticity at pre-strain of 1%. At 15%, prestrain of the material recovers elasticity at a rate nearly double that of the level achieved at 1% pre-strain. Hardly any difference due to the Nb content was observed Heat Treatment Temperature, T h Fig. 5 Relation between elastic recovery strain and heat treatment temperature. Figure 5 shows the effect of heat treatment temperature on elastic recovery strain at 1% pre-strain. As shown in this figure, the heat treatment temperature exerts hardly any effect at all. 3.2 Recovery stress Joint rings are used to fasten pipes caused by their shape recovery stress, and this requires achieving a certain level of shape recovery stress, such as a level much greater than the pull-out strength. Figure 6 shows one example of the stress-temperature curve for shape memory. This curve was conducted to measure the shape recovery stress and transformation

4 Effects of Pre-Strain and Heat Treatment Temperature on Phase Transformation Temperature 1653 Stress, MPa Loading Temperature, K Pre-Strain:7.5% Recovery Stress, σ R /MPa Fig. 6 One example of the stress-temperature curve for shape memory alloys. Pre-Strain:1% Heat Treatment Temperature, T h Recovery Stress, σ R /MPa Pre-strain, ε p Fig. 7 Variation in recovery stress with pre-strain. temperatures, with pre-strain of 7.5%. Moreover, the shape recovery stress of other specimens was measured by the method similar to this curve. And also, the transformation temperature measurement result is described later. Figure 7 shows how recovery stress varies with the prestrain in a solution heat-treated material. As shown in the figure, recovery stress of both specimens rises along with the increase in pre-strain. They peak at about 9% of pre-strain, thus generating recovery stress between 45 and 5 MPa in Ti 44:6 Ni 49:5 Nb 5:9. Conversely, applied pre-strain higher than that level will reduce the recovery stress. This decline in recovery stress is presumably due to plastic deformation, since applying a higher pre-strain reduces the strain of shape recovery after heating. Recovery stress of low Nb specimen is larger than that of rich Nb specimen in almost all region of pre-strain because of the yield stress of Nb particle lower than that of the matrix. Figure 8 shows the effect of thermal aging treatment temperature on recovery stress at 1% of pre-strain. Aging Fig. 8 Effect of heat treatment temperature on recovery stress. increases recovery stress slightly more than the solution treatment. It is considered to mean precipitation hardening by aging treatment. Those compositions of specimens indicate only Ti-Ni -phase from Ti-Ni-Nb ternary system on the 173 K, 16) but less temperature makes Ti-Ni -phase region vary narrow. So, it is assumed that aging treatment generates precipitations in Ti-Ni matrix, however, there is no data of crystal structure by XRD. In Ti 44 Ni 47 Nb 9 alloy, J. Di et al. reported that heat treatment generated (Ni, Nb) 3 Ti and materials were hardened. 17) Therefore, it is presumed that reverse transformation strain will change the internal stress field, and result in increase in recovery stress. 3.3 Transformation temperature Effect of pre-strain As a specimen is cooled and reaches the M phase, its self-adjustment function stores elastic strain energy in the M phase. And, if the stored elastic strain energy is released through slip deformation or other factors, the transformation temperature will rise. 1) Figure 9 shows how the transformation temperature varies with pre-strain. As the pre-strain rises, slip deformation releases the elastic strain energy, thus raising the transformation temperature. Figure 1 shows how the start temperature of reverse transformation (point A s ) varies with pre-strain, by using the Nb content as a parameter. It could be said that point A s remains constant even if the Nb content changes. With regard to the changes occurring at point Ms with pre-strain, one could also conjecture that hardly any differences exist among the four compositions, though the related diagrams are not shown here Effect of heat treatment temperature Figure 11 shows how the transformation temperature varies with heat treatment temperature at 1% pre-strain. Heat treatment reduces the transformation temperature more than in a solution-treated material. The effect of heat treat-

5 1654 K. Uchida, N. Shigenaka, T. Sakuma, Y. Sutou and K. Yamauchi Transformation temperature, T h M s A f A s Transformation temperature, T A f A s M s Ni/Ti = 1.88 Nb = 6 at% Pre-strain:1% Pre-strain, ε p Heat treatment temperature, T h Fig. 9 Variation in transformation temperature with pre-strain. Effect of heat treatment temperature on transformation temper- Fig. 11 ature. Reverse Transformation Temperature, A S Nb=12 at% Pre-strain, ε p Fig. 1 Relation between reverse transformation temperature and prestrain. Increment of transformation temperature, T h M s A s Pre-strain:1% Ni/Ti ratio A f ment temperature on the transformation temperature is presumably due to the generation of precipitates. The changes to be noted here are at points M s and A s. As shown in Fig. 9, the heat treatment temperature exerts only a small effect on point M s. Point A s, on the other hand, can be reduced by 3 to 4 K through heat treatment. Therefore, the effect of heat treatment can be used to adjust points M s and A s in joint rings as required Effect of the Ni/Ti ratio Figure 12 shows the relation between the rise in transformation temperature and the Ni/Ti ratio. Here, the prestrain is 1% with Nb content set at 6 at% and 9 at%. When Ni/Ti = 1, point M s increases by at least 6 K, while Fig. 12 Relation between the rise in transformation temperature and Ni/Ti ratio. point A s rises by a value in the 65-to-8 K range. When Ni/Ti = 1.88, the comparable values rise by values in the 85-to-1 K range, respectively. As the Ni/Ti ratio increases, the transformation temperature increases at a higher rate. Therefore, the rise in transformation temperature due to the Ni/Ti ratio should be considered when selecting a composition. Moreover, since changes in transformation temperature also depend on pre-strain, it is necessary to select several compositions and examine the changes with the pre-strain in determining the appropriate composition.

6 Effects of Pre-Strain and Heat Treatment Temperature on Phase Transformation Temperature Conclusion SMA joint rings must exhibit certain functions under operating temperature conditions, such as storage and control temperatures, and fastening force. This study examined the effects of pre-strain on the elastic recovery strain, recovery stress, and transformation temperature. The results obtained are summarized below. (1) The elastic recovery strain rises as the pre-strain increases. At 1% pre-strain, about 2% of the material recovers elasticity. Moreover, the amount of elastic recovery is hardly affected by the Nb content or heat treatment temperature. (2) The recovery stress peaks at pre-strain of about 9%. Moreover, heat treatment slightly increases the recovery stress. (3) The transformation temperature rises as the pre-strain increases. However, point M s hardly increases after the pre-strain exceeds about 5%. (4) Even when pre-strain is applied, heat treatment proves effective, resulting in lower transformation temperature. (5) The greater the Ni/Ti ratio, the higher the transformation temperature. Acknowledgment This work was conducted as a joint study involving Japanese BWR utilities (Tokyo Electric Power Co., Inc., Chubu Electric Power Co., Inc., Chugoku Electric Power Co., Inc., and Japan Atomic Power Company), Hitachi, Ltd., and Toshiba Corporation. REFERENCES 1) S. Miyazaki, T. Sakuma and T. Shibuya: Properties and Application Development of Shape Memory Alloy, (CMC, Japan 21). 2) T. Honma: Jour. Jpn. Soc. Mech. Eng. 786 (1984) ) K. Yamauchi: Jpn. Inst. Met 32 (1993) ) K. Kaneko, H. Aoki, M. Kubo, T. Suzuki, M. Uehara and A. Yoshida: J. Japan Inst. Metals 58 (1994) ) P. Lin, H. Tobushi, K. Tanaka, T. Hattori and M. Makita: Trans. Jpn. Soc. Mech. Eng. A 569 (1994) ) P. Lin, H. Tobushi, K. Kimura, H. Iwanaga and T. Hattori: Trans. Jpn. Soc. Mech. Eng. A 569 (1994) ) H. G. Yong and C. M. Wayman: Acta Met. 22 (1974) ) G. B. Olson and M. Cohen: Scripta Met. 9 (1975) ) R. J. Salzbrenner and M. Cohen: Acta Met. 27 (1979) ) M. Piao, K. Otsuka, S. Miyazaki and H. Horikawa: Mater. Trans. JIM 34 (1993) ) S. Miyazaki, T. Imai, Y. Igo and K. Otsuka: Met. Trans., A, 17 (1986) ) T. Sakuma, M. Hosogi, N. Okabe, U. Iwata and K. Okita: Mater. Trans. 43 (22) ) K. N. Melton, J. Simpson and T. W. Duerig: Proc. Int. Conf. on Martensitic Transformations (ICOMAT-86). Japan (1986) pp ) K. N. Melton, J. L. Proft and T. W. Duerig: Inst. MRS Meeting, (1988). 15) L. C. Zhao, T. W. Duerig, S. Justi, K. N. Melton, J. L. Proft, W. Yu and C. M. Wayman: Scripta Met. 24 (199) ) Y. Guanjun and H. Shiming: Rare Metal Mater. Eng. 26 (1997) ) J. Di, L. Wen-xi, D. Zhi-zhong, H. Ming and W. De-fa: Trans. Nonferrous Met. Soc. China, 13 (23)