Annealing effect on martensitic transformation of
severely cold-rolled Ti
50
Ni
40
Cu
10
shape memory alloy
K.N. Lin and S.K. Wu
*Department of Materials Science and Engineering, National Taiwan University, Taipei 106, Taiwan
Received 27 September 2006; revised 18 December 2006; accepted 19 December 2006 Available online 11 January 2007
The annealing effect on the recovery of B2! B19 and B19 ! B190transformations of 40% cold-rolled Ti
50Ni40Cu10is studied by the differential scanning calorimetry (DSC) and dynamic mechanical analyzer (DMA) tests. The DSC test is only sensitive for mea-suring the recovery of B2! B19 transformation, while the DMA test is suitable for both transformations. The change of internal friction values of these two transformations affected by annealing is due mainly to the difference in the rate of change of the trans-formation volume between them, which is related to their different recovery behaviors and microstructures.
2007 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
Keywords: Annealing; Cold working; Differential scanning calorimetry (DSC); Shape memory alloys (SMA); Dynamic mechanical analyzer (DMA)
Ti50Ni50xCux shape memory alloys (SMAs), with
x 6 30 at.%, have been investigated extensively from various aspects, such as the shape memory effect [1–3], martensitic transformation behavior [4–7], mechanical characteristics[8–10], microstructures[11–16]and inter-nal friction (IF) [17–20]. The transformation sequences of Ti50Ni50xCux SMAs are B2 M B190, B2 M B19 M
B190 and B2 M B19 for x < 5, 5 6 x 6 20 and x > 20
at.%, respectively [21–24]. Here B2 is parent austenite, and B19 and B190 are orthorhombic and monoclinic
martensite, respectively.
Traditionally, differential scanning calorimetry (DSC) has been employed to investigate martensitic transformation behavior of SMAs. Lo et al. [17] and Ren et al. [25] showed that the transformation shear strain required in B2 M B19 transformation (8%) of Ti50Ni40Cu10 is larger than that in B19 M B190 (2%).
Thus, the transformation enthalpy, DH, of B2 M B19 transformation is significantly larger than that of B19 M B190 transformation and the DSC technique is
suitable for investigating B2 M B19 transformation
[17]. Dynamic mechanical analyzer (DMA) is another technique for investigating martensitic transformation of SMAs. DMA measurement indicates that Ti50Ni40
-Cu10 exhibits two significant IF peaks corresponding
to B2 M B19 and B19 M B190 transformations [17–20].
Thus, DMA is useful for investigating both B2 M B19 and B19 M B190transformations.
A cold-working process can generate high-density defects, large residual stresses and distortion in cold-worked alloys. These imperfections greatly affect mar-tensitic transformation behavior and can be reduced or released by annealing treatment. However, it is unclear which between B2! B19 and B19 ! B190
transforma-tions will recover first after annealing treatment for cold-worked Ti50Ni40Cu10 SMA. In this study, the
annealing effect on B2! B19 and B19 ! B190
transfor-mations of 40% cold-rolled Ti50Ni40Cu10is investigated
by DSC and DMA. The different recovery behaviors of B2! B19 and B19 ! B190 transformations under
different annealing conditions are also discussed. Ti50Ni40Cu10 ingot was prepared by vacuum
arc-remelting (VAR) method in which high-purity Ti (99.8 wt.%), Ni (99.9 wt.%) and Cu (99.99 wt.%) were remelted six times in a high-purity Ar atmosphere. The ingot was hot-rolled at 900C to a plate of 2 mm thick-ness, solution-treated at 900C for 1 h and subsequently quenched into water. The oxidation layer of the plate was chemical etched by a solution composed of HF: HNO3:H2O in a 1:5:20 volume ratio. After removing
the oxidation layer, the thickness of the plate became 1.9 mm, and the plate was cut into 90 mm· 20 mm strips with the longitude along the hot-rolling direction. Thereafter, the strips were cold-rolled along the hot-rolling direction to 1.15 mm at room temperature. The total reduction in thickness was about 40%. After
1359-6462/$ - see front matter 2007 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.scriptamat.2006.12.016
* Corresponding author. Tel.: +886 2 2363 7846; fax: +886 2 2363 4562; e-mail:[email protected]
Scripta Materialia 56 (2007) 589–592
cold-rolling, the strips were cut into 40 mm· 4.5 mm specimens, sealed in evacuated quartz tubes and then annealed at 500 and 650C for different time intervals. Transformation temperature and enthalpy of cold-rolled and annealed specimens were determined by TA Q10 DSC equipment with 10C min1cooling/heating rate.
The IF peaks and their tan d values of cold-rolled and annealed specimens were determined by TA 2980 DMA equipment with 3C min1 cooling/heating rate under constant frequency (1 Hz) and amplitude (2 lm, with a strain of 2· 105). The testing temperature range was from +150 to 150 C in both DSC and DMA tests. The abbreviations ‘‘tan d1’’ and ‘‘tan d2’’ are
em-ployed to represent the IF peak values of B2! B19 and B19! B190 transformations, respectively.
Figures 1 and 2show the DSC results of 40% cold-rolled Ti50Ni40Cu10 specimens annealed at 500 and
650C, respectively, for different time intervals. The annealing conditions for Figure 1a–c are 500C for 1, 24 and 96 h, respectively, and those for Figure 2a–c are 650C for 0.5, 24 and 72 h, respectively. After annealing at 500C for 1 h, as shown inFigure 1a, the B2! B19 transformation peak clearly recovers while the B19! B190 transformation peak is insignificant.
After further annealing, as shown in Figure 1b, the B2! B19 transformation peak becomes sharper and the B19! B190transformation peak appears but is still
not clear. Finally, after annealing for 500C at 96 h, as shown in Figure 1c, both B2! B19 and B19 ! B190
transformation peaks become clear. Under 650C annealing, both B2! B19 and B19 ! B190
transforma-tion peaks recover obviously within 0.5 h, as shown in
Figure 2a.Figure 3plots the relationship of DHcversus
annealing time for the specimens ofFigures 1 and 2, an-nealed at 500 and 650C, respectively. Here DHcis the
transformation enthalpy of the DSC cooling curve.
Figure 3 shows that the DHcvalue increases obviously
within 1 h and then becomes saturated when the anneal-ing time is prolonged to about 23.5 J g1under 500 and 650C annealing.
According to Figure 3, DHc values of specimens
an-nealed at 500C for 1 h and 650 C for 0.5 h, and those at 500C for 24 h and 650 C for 1 h are almost the same (about 21.6 and 23.4 J g1, respectively). Figure 4a and b shows the DMA results of these two pairs of specimens. It is obvious that, even if the DHc values
are approximately equal, the IF peaks of the B2! B19 (tan d1) and B19! B190 (tan d2) transformations
for different annealed Ti50Ni40Cu10 specimens are not
identical. From Figure 4, for specimens having equal DHc, their tan d1values are almost the same while their
tan d2values are quite different. This is because the DHc
value is contributed mainly by the B2! B19 transfor-mation, rather than the B19! B190 transformation
Figure 1. DSC results of 40% cold-rolled Ti50Ni40Cu10specimens annealed at 500C. (a) 1 h, (b) 24 h and (c) 96 h.
Figure 2. DSC results of 40% cold-rolled Ti50Ni40Cu10specimens annealed at 650C. (a) 0.5 h, (b) 24 h and (c) 72 h.
Figure 3. The relationship of DHcversus annealing time interval under
500 and 650C annealing. 590 K. N. Lin, S. K. Wu / Scripta Materialia 56 (2007) 589–592
[17]. In other words, the annealing effect on cold-rolled Ti50Ni40Cu10specimens is not sensitive for the recovery
of the B19! B190 transformation, as indicated by the
DSC results.
The DMA results of 40% cold-rolled and annealed Ti50Ni40Cu10 specimens are shown in Figures 4 and 5.
After annealing at 500C for 1 h, as shown in Figure 4a, the B2! B19 transformation peak recovers obvi-ously while the B19! B190transformation peak is not
so clear. After further annealing, as shown in Figures
4b and5a, the B19! B190transformation peak
gradu-ally becomes apparent but the change in the height of the B2! B19 transformation peak can be negligible. Under 650C annealing, both the B2 ! B19 and B19! B190 transformation peaks recover obviously
within 0.5 h annealing, as shown in Figure 4a. After further annealing, as shown in Figures 4b and 5b and c, tan d2 grows rapidly while tan d1grows only a little.
FromFigures 4 and 5, tan d1and tan d2versus annealing
time under 500 and 650C annealing are plotted in
Figure 6. It is obvious that B2! B19 transformation recovers rapidly at first and tan d1is saturated to about
0.09–0.1 under both 500 and 650C annealing. How-ever, the recovery of B19! B190transformation shows
quite different behavior. Under 650C annealing, tan d2
grows rapidly and is always larger than tan d1, while
under 500C annealing, tan d2 grows slowly and is
always smaller than tan d1.
Zu et al.[26]investigated the recrystallization behav-ior of heavy ion-irradiation-induced amorphized Ti50Ni43Cu7 SMA by in situ transmission electron
microscopic observation. They pointed out that the recrystallization of amorphized Ti50Ni43Cu7 started at
273C. Tsuji and Nomura[27]investigated the relation-ship between hardness and annealing temperature of
27% cold-rolled Ti50.5Ni40.5Cu9 SMA annealed at 450,
500, 550, 600 and 750C for 1 h. They suggested that annealing temperature up to 500C allows 27% cold-rolled Ti50.5Ni40.5Cu9to recover, and that higher
anneal-ing temperatures can rapidly remove the internal defects and distortion, and further cause recrystallization and grain growth. According to these reports, we propose that the recrystallization temperature of 40% cold-rolled Ti50Ni40Cu10 is lower but close to 500C. Therefore,
only recrystallization occurs under 500C annealing, whereas both recrystallization and grain growth occur rapidly under 650C annealing.
The aforementioned results indicate that different recovery behaviors of B2! B19 and B19 ! B190
trans-formations exhibit in between DHcand IF resulted from
different testing methods of DSC and DMA measure-ments. In the DSC test, the driving force of transforma-tion is attributed only to the change in temperature, and the DHc value can be employed to estimate the total
transformation volume. In the DMA test, the driving force of transformation comes not only from the tem-perature change but also from the applied stress. The IF of a first-order phase transformation can be predicted by the De Jonghe–Delorme model [28,29]:
tan d/1 x dWðVmÞ dVm oVm oT oT ot þ oVm or or ot ð1Þ where x is the angular frequency of applied stress, Vmis
the volume fraction of martensite, W(Vm) is a
monoto-nous function associated with transformation volume change or shape strain, T is temperature, t is time and r is the stress that induces martensitic transformation
Figure 4. DMA results of cold-rolled and annealed Ti50Ni40Cu10
specimens having the same DHc values. (a) Specimens annealed at
500C for 1 h and at 650 C for 0.5 h, and (b) specimens annealed at 500C for 24 h and at 650 C for 1 h.
Figure 5. DMA results of 40% cold-rolled and annealed Ti50Ni40Cu10specimens. Annealing at (a) 500C for 96 h, (b) 650 C for 24 h and (c) 650 C
for 96 h.
Figure 6. The relationship of tan d1and tan d2versus annealing time
interval under 500 and 650C annealing.
or reorientation of martensite variants. In this study, x and the temperature changing rate are kept as constants (1 Hz and 3C min1, respectively), and r is in the order of 1 MPa, which is not large enough to induce martens-itic transformation or to reorient martensite variants. In supposing that dW(Vm)/dVmis constant for all
thermo-elastic martensites [30], the De Jonghe–Delorme model can be simplified as
tan d/oVm
oT ð2Þ
The tan d value is now simply determined by dVm/dT,
that is, the volume of martensite formation per unit tem-perature. FromFigure 4, DHcand tan d1of the annealed
specimens show almost the same values but their tan d2
values do not. This implies that, for these specimens, their dVm/dT of the B2! B19 transformation, (dVm/
dT)B2!B19, can be regarded as almost equal, but that
of the B19! B190 transformation, (dV
m/dT)B19!B190,
cannot be. The different values of (dVm/dT)B2!B19 and
(dVm/dT)B19!B190are attributed to the different intrinsic
crystal structures associated with B19 and B190
martens-ites, that is, defect-free B19 and (0 0 1)B190twinned B190
martensites[4]. Under 500C annealing, only recrystal-lization occurs, the grain size is small and grain size distribution is not uniform, and thus B19! B190
trans-formation is more suppressed. Under 650C annealing, however, both recrystallization and grain growth occur quickly. At this time, tan d2 also increases in the same
pace as the annealing recovery because the grain size increases and the grain size distribution becomes more uniform.
The annealing effect on the recovery of B2! B19 and B19! B190 transformations of 40% cold-rolled
Ti50Ni40Cu10 is investigated by DSC and DMA
tech-niques in this study. Experimental results show that the recovery behaviors of B2! B19 and B19 ! B190
transformations are different. The DSC and DMA results of annealed specimens having equal DHc value
show that the DHcvalue of the DSC curve is mainly
con-tributed by B2! B19 transformation and so the test is not sensitive for measuring the recovery of B19! B190 transformation. However, the DMA test is useful
for investigating both B2 M B19 and B19 M B190
trans-formations. The recovery of B2! B19 transformation can proceed with annealing at 500C for 1 h, while that of B19! B190requires annealing at 650C for 1 h. The
recrystallization temperature of 40% cold-rolled Ti
50-Ni40Cu10 is lower but close to 500C. The different
recovery behaviors of B2! B19 and B19 ! B190
trans-formations are due mainly to the different microstruc-tures of defect-free B19 and (0 0 1)B190 twinned B190
martensites. The IF tan d values of B2! B19 and B19! B190 transformation peaks are proposed to be
proportional to dVm/dT values, which are different for
the B2! B19 and B19 ! B190 transformations. This
causes the different recovery behaviors of B2! B19 and B19! B190 transformations.
The authors gratefully acknowledge the financial support from the National Science Council (NSC), Taiwan, Republic of China, under Grant NSC95-2221-E002-163.
[1] T.H. Nam, T. Saburi, Y. Kawamura, K. Shimizu, Mater. Trans., JIM 31 (1990) 262.
[2] T.H. Nam, T. Saburi, K. Shimizu, Mater. Trans., JIM 33 (1992) 814.
[3] H. Sehitoglu, I. Karaman, X. Zhang, A. Viswanath, Y. Chumlyakov, H.J. Maier, Acta Metall. 49 (2001) 3621. [4] T. Fukuda, T. Saburi, T. Chihara, Y. Tsuzuki, Mater.
Trans., JIM 36 (1995) 1244.
[5] T. Fukuda, M. Kitayama, T. Kakeshita, T. Saburi, Mater. Trans., JIM 37 (1996) 1540.
[6] G. Mazzolai, A. Biscarini, B. Coluzzi, F.M. Mazzolai, A.R. Ross, T.A. Lograsso, Mater. Sci. Eng. A 370 (2004) 497.
[7] X. Ren, N. Miura, K. Taniwaki, K. Otsuka, T. Suzuki, K. Tanaka, Yu.I. Chumlyakov, M. Asai, Mater. Sci. Eng. A 273–275 (1999) 190.
[8] K.N. Melton, O. Mercier, Metall. Trans. 9A (1978) 1487. [9] T. Saburi, T. Takagaki, S. Nenno, K. Koshino, MRS Int.
Mtg. Adv. Mats. 9 (1988) 147.
[10] J.L. Proft, K.N. Melton, T.W. Duerig, MRS Int. Mtg. Adv. Mats. 9 (1988) 159.
[11] P.L. Potapov, A.V. Shelyakov, D. Schryvers, Scripta Mater. (2001) 1.
[12] T. Tadaki, K. Shmizu, C.M. Wayman, Mater. Trans., JIM 32 (1991) 43.
[13] R.H. Bricknell, K.N. Melton, O. Mercier, Metall. Trans. 10A (1979) 693.
[14] R.H. Bricknell, K.N. Melton, Metall. Trans. 11A (1980) 1541.
[15] T. Tadaki, C.M. Wayman, Metallography 15 (1982) 233. [16] T. Tadaki, C.M. Wayman, Metallography 15 (1982) 247. [17] Y.C. Lo, S.K. Wu, H.E. Horng, Acta Metall. Mater. 41
(1993) 747.
[18] S.K. Wu, H.C. Lin, J. Alloys Compd. 355 (2003) 72. [19] I. Yoshida, D. Monma, K. Otsuka, M. Asai, H. Tsuzuki,
J. Alloys Compd. 355 (2003) 79.
[20] I. Yoshida, D. Monma, K. Iino, K. Otsuka, M. Asai, Mater. Sci. Eng. A 370 (2004) 444.
[21] O. Mercier, K.N. Melton, Metall. Trans. 10A (1979) 387. [22] H. Miyamato, T. Taniwaki, T. Ohba, K. Otsuka, S.
Nishigori, K. Katc, Scripta Mater. 53 (2005) 171. [23] T.H. Nam, T. Saburi, K. Shimizu, Mater. Trans., JIM 31
(1990) 959.
[24] T.H. Nam, T. Saburi, Y. Nakata, K. Shimizu, Mater. Trans., JIM 31 (1990) 1050.
[25] X. Ren, N. Miura, J. Zhang, K. Otsuka, K. Tanaka, M. Koiwa, T. Suzuki, Yu.I. Chumlyakov, M. Asai, Mater. Sci. Eng. A 312 (2001) 196.
[26] X.T. Zu, S. Zhu, X. Xiang, L.P. You, Y. Huo, L.M. Wang, Mater. Sci. Eng. A 363 (2003) 352.
[27] K. Tsuji, K. Nomura, Scripta Metall. Mater. 24 (1990) 2037.
[28] J.F. Delorme, R. Schmid, M. Robin, P. Gobin, J. Phys. 32 (1971) C2–C101.
[29] W. Dejonghe, R. De Batist, L. Delaey, Scripta Metall. 10 (1976) 1125.
[30] O. Mercier, K.N. Melton, Y. De Pre´ville, Acta Metall. 27 (1979) 1467.
