Annealing-induced evolution of transformation characteristics in TiNi shape memory alloys

Annealing-induced evolution of transformation characteristics in TiNi shape memory alloys

ARTICLE IN PRESS Physica B 353 (2004) 9–14 www.elsevier.com/locate/physb Annealing-induced evolution of transformation characteristics in TiNi shape...

261KB Sizes 0 Downloads 44 Views

ARTICLE IN PRESS

Physica B 353 (2004) 9–14 www.elsevier.com/locate/physb

Annealing-induced evolution of transformation characteristics in TiNi shape memory alloys Z.G. Wanga, X.T. Zua,, X.D. Fengb, S. Zhuc, J.M. Zhoub, L.M. Wangc a

Department of Applied Physics, University of Electronic Science and Technology of China, Chengdu 610054, People’s Republic of China b Department of Physics, Sichuan University, Chengdu 610064, People’s Republic of China c Department of Nuclear Engineering and Radiological Sciences, University of MI, Ann Arbor, MI 48109, USA Received 7 February 2004; received in revised form 21 August 2004; accepted 25 August 2004

Abstract The effect of annealing on transformation characteristics of TiNi shape memory alloys (SMAs) was investigated by differential scanning calorimetry (DSC) and the evolution of the microstructure was studied using positron annihilation technology (PAT) and transmission electron microscopy (TEM). The results showed that transformation characteristics depend on the annealing temperature. The R-phase transformation appeared at low annealing temperature. The Rphase disappeared and austensite transformed into martensite directly as the annealing temperature exceeded 550 1C. With increasing annealing temperature, the vacancy cluster and dislocation related positron lifetime decreased. Changes in transformation characteristics can be attributed to the evolution of the microstructrue of the TiNi specimen. r 2004 Elsevier B.V. All rights reserved. Keywords: Shape memory alloy; Annealing; Differential scanning calorimetry (DSC); Positron annihilation technology (PAT); Transmission electron microscopy (TEM)

1. Introduction TiNi shape memory alloys (SMAs) of nearequiatomic composition are of technological importance, because the TiNi SMA combines good functional and structural properties [1–3]. There is an interest in Ni-rich NiTi alloys because phase transformation temperatures can be controlled Corresponding author. Tel: 86-28-83201939; fax: 86-2883201939. E-mail addresses: [email protected], [email protected] (X.T. Zu).

through heat treatment [4,5]. With different thermal heat treatments, the TiNi alloy exhibits either a onestep martensitic transformation (MT) from the high temperature B2 (cubic) to the B19’ (monoclinic) phase, or a two-step MT from the B2 to the R to the B19’, even a multiple-step transformation (MST) [6–8]. The MST is a special case of MT, which is characterized by the appearance of several transformation steps during the transformation of the Rphase into the martensite phase. Positron annihilation and, in particular, positron lifetime spectroscopy, is a very useful and

0921-4526/$ - see front matter r 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.physb.2004.08.021

ARTICLE IN PRESS 10

Z.G. Wang et al. / Physica B 353 (2004) 9–14

effective technique in studying defects in condensed materials. More specifically, in recent years, positron lifetime and Dopper-broadening measurements have been performed to study the behavior of quenched in vacancies and ordering transformations in Cu–Zn–Al [9,10]. The same technique has also been used to study ageing effects in the Cu–Al–Ni [11] and Cu–Al–Be [12] shape memory alloys. Recently, we have analyzed defects generated by small fluence electron irradiation in TiNiCu [13], TiNi [14] and CuZnAl [15] shape memory alloys using positron lifetime measurements. In this work the effect of heat treatment on transformation characteristics of TiNi SMA were investigated by differential scanning calorimetry (DSC), positron annihilation technology (PAT) and transmission electron microscopy (TEM).

2. Experimental procedure Investigations were carried out on a commercial Ti–50.7 at% Ni SMA with a thickness of 0.35 mm, which was provided by the Northwest Institute of Non-Ferrous Metal of China. The specimen was annealed at 400, 450, 500, 550 and 600 1C for 1 h in an evacuated silica tube followed by air cooling. Samples of 12 mm  12 mm  0.35 mm were for positron lifetime measurements and samples of 5 mm  5 mm  0.35 mm were for DSC measurements. The positron annihilation lifetime was measured at room temperature using a conventional fast–fast coincidence setup (ORTEC) with NEIII scintillators. The time resolution was 235 ps. A 22 Na source was sandwiched between two identical samples. The computer program, Positronfit, was used for the lifetime spectrum fitting with a total count over 106. Transformation temperatures of the samples were measured using DSC with a scanning rate of 10 1Cmin under nitrogen atmosphere. The specimens were placed in an aluminum pan. The pans were sealed and placed in the measuring chamber of a differential scanning calorimeter (DSC131, France).

The microstructure of the TiNi sample annealed at 400 and 500 1C was studied by TEM. Ex situ TEM analysis was performed with JEOL 2010 FEG TEM at the University of Michigan. The TEM specimens were prepared as follows: the specimens were ground with a series of waterproof abrasive papers to a thickness of 300 mm, then the disc specimens with 3 mm diameter were cut by a slurry drill core cutter. Then the disc specimens were mechanically polished to 100–140 mm and twin-jet electro-polished at 45 V with 10% H2SO4 and 80% CH3OH by volume at a temperature of 283 K.

3. Experimental results 3.1. Transformation behavior Fig. 1 shows the DSC curves of both the asreceived sample and the samples heat treated at different temperatures. It can be seen from the figure that no phase transformation happens in the temperature range between 60 and 140 1C for the as received sample. For samples annealed at temperature below 500 1C, two transformation processes from austenite to R-phase and further to martensite have taken place upon cooling, onestep reverse transformation occurs upon heating. When the annealing temperature increases to 550 1C, the B2-R-phase transformation almost disappears, which just appears as a shoulder on the high temperature side of martensitic transformation. When the annealing temperature reaches 600 1C, as shown in Fig. 1(f), the R-phase disappears and austenite transforms into martensite directly. Transformation temperatures for samples annealing at different temperatures are shown in Table 1. Transformation temperatures are determined by determining the onset point of slope change in the DSC curves. In the cooling process, As, Af, Rs, Rf and Ms temperatures decreases with increase in annealing temperature and this results is consistent with the previous result [16]. Mf reaches its minimum for samples annealed below 500 1C, above this temperature, it seems to increase with increasing annealing temperature.

ARTICLE IN PRESS Z.G. Wang et al. / Physica B 353 (2004) 9–14

11

Latent heat of transformation (J/g)

20

,

,

19 18 17 16 15 14 400

450

500

550

600

Annealing temperature (˚C)

Fig. 2. Effect of annealing temperature on the latent heat of transformation upon heating for TiNi specimen.

The values of latent heat of transformation for the heating curves are shown in Fig. 2. As can be seen from the figure the latent heat increases with increasing annealing temperature. , ,

3.2. Positron lifetime measurements

Fig. 1. DSC curves of both the as-received sample Ti–50.7 at%Ni shape memory alloy flake (a) and the samples heat treated at 400 (b), 450 (c) 500 (d), 550 (e) and 600 1C (f) for 1 h.

Table 1 Transformation temperatures of the Ti–50 at%Ni samples with different annealing temperatures Heat treat As(1C) Af(1C) Rs(1C) Rf(1C) Ms(1C) Mf(1C) temperature (1C) 400 450 500 550 600

44 34 20 13 13

65 57 45 48 39

56 41 25 23 —

39 31 18 — —

24 12 7 — 8

15 18 26 14 12

A three-lifetime-component fitting model was used to analyze the spectrum, the lifetimes being t1, t2, and t3, whilst the corresponding intensities are I1, I2 and I3. t3 is rather large and has a low intensity of about 1–3% which reflects those annihilation processes occurring on the sample surfaces. Fig. 3 shows the changes of t1 and t2 against the annealing temperature. It is seen that t2 has a value ranging from about 450 to 300 ps. It is known [17] that the positron lifetime is a function of the size of the vacancy cluster and that a positron lifetime of about 450 ps corresponds with a quite large open volume of about 20 vacancies. This implies that there are many vacancy clusters in the TiNi samples. t1 has a value ranging from 180 to 130 ps, which is near the value of positron trapping at monovacancies [18]. t1 reflects the positron trapping at dislocations. The two positron lifetimes decrease with increasing annealing temperature. This implies a decrease in size of the trapping centers after annealing.

ARTICLE IN PRESS Z.G. Wang et al. / Physica B 353 (2004) 9–14

12

3.3. TEM results

annealing temperature the dimension and the density of dislocations decrease.

Fig. 4(a) and (b) show the bright field TEM image of the samples annealed at 400 and 500 1C at room temperature. The selected area electron diffraction pattern of the samples annealed at 400 1C is also shown as an insert. The bright field image in Fig. 4(a) shows the typical self-accommodation structure. In the electron diffraction pattern the satellite reflections located at near 1/3 position of the reciprocal lattice are very sharp and strong. This confirms that the microstructure is Rphase at room temperature. This is consistent with the DSC results. The highest dislocation density is distinctive for the two samples, but with increasing

Positron second lifetime (ps)

430

τ2

380 330 280 230

τ1

180 130 350

400

450

500

550

600

650

Annealing temperature T (°C)

Fig. 3. Positron lifetime measured after the Ti–50.7 at%Ni shape memory alloy flake had been annealed at different temperatures for 1 h.

4. Discussion It is well known that severe cold work inhibits the martensitic transformation by the introduction of defects, which are essentially dislocations [19]. The SMAs are in a cold worked state in the present investigation, so no transformation happens in the temperature range between –60 and 140 1C for the as-received sample. There are many dislocations in the cold worked TiNi specimen. Annealing induces annihilation of dislocation and recrystallization of the cold worked specimen. At low annealing temperature, the dislocation density is high, and the associated internal stress generated by these dislocations restrict the specimen martensite from transforming into austenite. So the As and Af are in the hightemperature region. With increasing annealing temperature, the dislocation density reduces and the internal stress reduces too, which leads to a decrease in As and Af. The latent heat of transformation increases with increasing annealing temperature. This is consistent with previous results [20], which show that dislocations associated with high levels of plastic deformation generate an internal stress, which restricts the martensite from transforming into austenite. This

Fig. 4. Bright field TEM image of TiNi alloy annealed at (a) 400 and (b) 500 1C.

ARTICLE IN PRESS Z.G. Wang et al. / Physica B 353 (2004) 9–14

martensitic phase remains ‘‘pinned’’ in the microstructure until the dislocations are removed through an annealing process. Therefore, with increasing annealing temperature, more dislocations are removed and more martensite transforms into austenite, so the latent heat increases. Several studies have reported that the appearance of the R-phase could be due to dislocations [21,22]. The R-phase nucleates preferentially in the internal-stresses-concentrated regions [7,22]. With increasing annealing temperature, the dislocation density reduces and the internal stress reduces too, thus both Rs and Rf decrease. The disappearance of the R-phase transformation can be attributed to the disappearance of the internal stresses in recrystallized grains, which has been verified by Khelfaoui et al. [19]. So the parent to R-phase transformation disappears when the internal stresses disappear from the grains. The positron is a positively charged particle, and thus it is repelled by the atomic nuclei and tends to stay in the lowest atomic density zones of the solid, the defects. In a perfect crystal, the e+ (in a nonlocalized state) would have the maximum probability of being annihilated, and that state corresponds to the lifetime for a given metal. On the other hand, when the e+ is in a low density zone (localized state) it will interact with a lower electron density and its mean life will increase. Low atomic density regions in alloy materials, such as grain boundary, phase interface, dislocation, vacancy and vacancy cluster, are known to be positron trap sites. In the present work, dislocation and vacancy clusters should be the two key positron trap centers. Annealing induces annihilation of the dislocation and recrystallization of the cold worked TiNi specimen [23,24]. With increasing annealing temperature, the size of the dislocation decreases. This has been proved by the TEM observations, thus leading to the second positron lifetime t1 decreasing from 183 to 140 ps. The positron lifetime measurements also show that the vacancy-cluster-related t2 decreases from 430 to 310 ps. The vacancy cluster also has an effect on the transformation temperatures [25,26]. The size of the vacancy clusters decreases with increasing annealing temperature. Thus the evolution of the microstructures during annealing

13

is the key factor to affect transformation characteristics.

5. Conclusions Transformation characteristics depend on annealing temperature. The R-phase transformation appears at low annealing temperature. The Rphase disappears and austenite transforms into martensite directly as the annealing temperature exceeds 550 1C. With increasing annealing temperature, the vacancy cluster and dislocationrelated positron lifetime decreased from 430 to 310 ps and from 183 to 140 ps, respectively. The evolution of the microstructure is the key factor to result in the change of transformation characteristics during annealing.

Acknowledgements This study was financially supported by the National Natural Science Foundation of China (10175042).

References [1] J. van Humbeeck, Mat. Sci. Eng. A 273–275 (1999) 134. [2] T. Duerig, A. Pelton, D. Stoeckel, Mat. Sci. Eng. A 273–275 (1999) 149. [3] Jafar Khalil-Allafi, Antonin Dlouhy, Gunther Eggeler, Acta Mater. 50 (2002) 4255. [4] T. Saburi, in: K. Otsuka, CM. Wayman, (Eds.), Shape Memory Materials, Cambridge Unversity Press, Cambridge, 1998, p. 49–96. [5] W. Tang, B. Sundmann, R. Sandstro¨m, C. Qiu, Acta Mater. 47 (1999) 3457. [6] J. Khalil Allafi, X. Ren, G. Eggler, Acta Mater. 50 (2002) 793. [7] L. Bataillard, J.-E. Bidaux, R. Gotthardt, Philos. Mag. 78 (1998) 327. [8] Antonin Dlouhy, Jafar Khalil-Allafi, Gunther Eggeler, Philos. Mag. 83 (2003) 339. [9] G.M. Lin, J.K.L. Lai, C.Y. Chung, Scripta Metall. Mater. 32 (1995) 1865. [10] R. Romero, A. Somoza, Mater. Sci. Eng. A 273–275 (1999) 572. [11] I. Hurtado, D. Segers, L. Dorikens-Vanpraet, C. Dauwe, J. van Humbeeck, J. Phys. IV 5 (1995) C8–949.

ARTICLE IN PRESS 14

Z.G. Wang et al. / Physica B 353 (2004) 9–14

[12] R. Romero, A. Somoza, M.A. Jurado, A. Planes, Li. Manosa,, Acta Mater. 45 (1997) 2101. [13] X.T. Zu, L.M. Wang, Y. Huo, L.B. Lin, Z.G. Wang, T.C. Lu, L.J. Liu, X.D. Feng, Appl. Phys. Lett. 80 (1) (2002) 31. [14] X.T. Zu, L.B. Lin, Z.G. Wang, S. Zhu, L.P. You, L.M. Wang, Y. Huo, J. Alloy. Comp. 351 (2003) 87. [15] Z.G. Wang, X.T. Zu, J.H. Wu, L.J. Liu, H.Q. Mo, Y. Huo, J. Alloy. Comp. 364 (1–2) (2003) 171. [16] X. Huang, Y. Liu, Scripta Mater. 45 (2001) 153. [17] P. Hautoja¨rvi, J. Heinio, M. Manninen, R. Nieminen, Philos. Mag. 35 (1977) 973. [18] P. Donner, R. Wurschum, E. Hornbogen, H.-E. Schaefer, Scripta Metall. 25 (1991) 1875. [19] F. Khelfaoui, G. Thollet, G. Gue´nin, Mater. Sci. Eng. A 338 (2002) 305.

[20] D.A. Miller, D.C. Lagoudas, Mater. Sci. Eng. A 308 (2001) 161. [21] S.K. Wu, H.C. Lin, Y.C. Yen, Mater. Sci. Eng. A 215 (1996) 113. [22] T. Fukuda, T. Saburi, K. Dio, S. Nenno, Mater. Trans. JIM 33 (1992) 271. [23] R. W. Siegel, in: P.G. Coleman, S.C. Sharma, L.M. Diana (Eds.), Positron Annihilation, North-Holland Publishers, Amsterda, 1982, p. 351. [24] J. Hurtado, D. Segers, J. van Humbeeck, L. DorikensVanpraet, C. Dauwe, Scripta Metall. Mater. 33 (1995) 741. [25] Z.G. Wang, X.T. Zu, Y. Huo, S. Zhu, X.W. Wei, L.M. Wang, Nucl. Instrum. Methods B 215 (2004) 436. [26] Z.G. Wang, X.T. Zu, L.J. Liu, S. Zhu, Y. Huo, L.B. Lin, X.D. Feng, L.M. Wang, Nucl. Instrum. Methods B 211 (2003) 239.