Trans. Nonferrous Met. Soc. China 25(2015) 4063-4071

Influence of Ni₄Ti₃ precipitates on phase transformation of NiTi shape memory alloy

Shu-yong JIANG¹, Yan-qiu ZHANG¹, Ya-nan ZHAO¹, Si-wei LIU², Li HU³, Cheng-zhi ZHAO³

1. College of Mechanical and Electrical Engineering, Harbin Engineering University, Harbin 150001, China;

2. Hubei Water Resources Research Institute, Wuhan 430070, China;

3. College of Materials Science and Chemical Engineering, Harbin Engineering University, Harbin 150001, China

Received 19 April 2015; accepted 20 October 2015

Abstract: NiTi shape memory alloy samples were aged for 2 h at 573, 723 and 873 K, respectively. Two R-phase variants are observed in the NiTi samples aged at 573 and 723 K, where the orientation relationship between the two R-phase variants and the B2 matrix is determined. In the NiTi samples aged at 573 and 723 K, fine and homogeneous Ni₄Ti₃ precipitates are coherent with the B2 austenite matrix. The Ni₄Ti₃ particles precipitate in the grain interior and at the grain boundaries, where the heterogeneous Ni₄Ti₃ precipitates are coherent, semi-coherent and incoherent with the B2 matrix in the NiTi sample aged at 873 K. As for the NiTi sample aged at 873 K, one-stage phase transformation from B19' martensite to B2 austenite occurs on heating, but two-stage phase transformation of B2-R-B19' arises on cooling. The NiTi sample aged at 723 K shows two-stage phase transformation of B2-R-B19' on cooling as well, but exhibits two-stage phase transformation of B19'-R-B2 on heating. The NiTi sample aged at 573 K exhibits three-stage transformation on cooling due to local stress inhomogeneity and local composition inhomogeneity around the Ni₄Ti₃ precipitates.

Key words: shape memory alloy; NiTi alloy; martensitic transformation; heat treatment; microstructure

1 Introduction

NiTi shape memory alloy has been widely used in the engineering field because of shape memory effect [1-3]. Ni₄Ti₃ precipitate is a metastable phase as well as an important phase, which usually precipitates in Ni-rich NiTi shape memory alloy below about 923 K. Ni₄Ti₃ phase belongs to R3 space group and it possesses a rhombohedral unit cell structure, in which there are six Ti atoms and eight Ni atoms [4,5]. The habit plane of Ni₄Ti₃ phase is {111}_B2 plane in the B2 austenite matrix and there exists a certain orientation relationship between the precipitate and the matrix. Furthermore, there are four {111}_B2 planes in the parent phase and every two variants possess the same habit plane, so the Ni₄Ti₃ phase has eight variants [6,7]. The size, the density, the distribution and the shape of the Ni₄Ti₃ precipitate phase are related closely to aging temperature, aging time and composition of NiTi alloy, which further have an effect on the phase transformation of NiTi shape memory alloy [8-10]. In general, the size of the Ni₄Ti₃ precipitates increases with the increase of aging temperature and aging time, and the smaller Ni₄Ti₃ particle (for example, less than about 100 nm) is coherent with the B2 austenite matrix [11-14]. The presence of the Ni₄Ti₃ precipitates can suppress the occurrence of martensitic transformation and thus facilitates the formation of R phase. Therefore, the occurrence of two-stage phase transformation of B2-R-B19' is considered to be rational in the aged Ni-rich NiTi shape memory alloy. However, the abnormal three-stage phase transformation often occurs in the aged Ni-rich NiTi shape memory alloy [15-21].

In the present study, Ni-rich binary NiTi shape memory alloy with a nominal composition of 50.9% Ni (mole fraction) was subjected to solution treatment and subsequent aging at low temperature (573 K), intermediate temperature (723 K) and high temperature (873 K), respectively. In the case of three different aging temperatures, the influence of Ni₄Ti₃ precipitates on microstructural evolution and phase transformation behavior of NiTi shape memory alloy was systematically investigated.

2 Experimental

The binary NiTi alloy with a nominal composition of 50.9% Ni is used in the present study and it belongs to the drawn cylindrical bar with a length of 1.5 m and a diameter of 12 mm. The NiTi bar was heated at 1123 K for 2 h and was then quenched into ice water. The NiTi samples with a diameter of 6 mm and a height of 9 mm were cut from the solution-treated NiTi bar by means of electro-discharge machining (EDM). The solution- treated NiTi samples were aged for 2 h at 573, 723 and 873 K, respectively, and subsequently were cooled to room temperature in the atmosphere. During all the aforementioned heat treatment, the specimens were placed into a silica tube, which was flushed with high purity argon gas three times and was then evacuated to a pressure of about 0.1 Pa.

Microstructural evolution of the NiTi samples subjected to solution treatment and aging was investigated by means of transmission electron microscopy (TEM). Foils for TEM observation were mechanically ground to 70 μm and then thinned by twin-jet polishing in an electrolyte consisting of 6% HClO₄, 34% C₄H₁₀O and 60% CH₃OH (volume fraction). TEM observations were conducted on an FEI TECNAI G2 F30 microscope with a side-entry and double-tilt specimen stage with angular range of ±40° at an accelerating voltage of 300 kV. Differential scanning calorimetry (DSC) test was performed at the heating and cooling rates of 10 K/min in order to obtain the phase transformation temperatures of all the NiTi samples.

3 Results

3.1 Microstructures

Figure 1 shows TEM image and the corresponding selected area electron diffraction (SAED) pattern of the microstructure of the solution-treated NiTi samples. It can be seen from Fig. 1 that there is an extremely small amount of TiC phase in the NiTi matrix, in which no Ni₄Ti₃ precipitates arise.

Fig. 1 TEM image (a) and corresponding SAED patterns (b,c) of solution-treated NiTi specimen

Figures 2-4 indicate TEM images of the NiTi specimens aged at 573, 723 and 873 K, respectively. It can be seen from Figs. 2 and 3 that the fine Ni₄Ti₃ precipitate phases are homogeneously dispersed in the grain boundary and grain interior, but the size of the Ni₄Ti₃ precipitates in the NiTi specimen aged at 723 K is obviously greater than that in the NiTi specimen aged at 573 K. However, it can be found from Fig. 4 that the Ni₄Ti₃ precipitates exhibit the inhomogeneous distribution in the matrix of NiTi sample aged at 873 K, where the Ni₄Ti₃ precipitates are characterized by the typical lenticular shape and they are relatively much coarser than those aged at 573 and 723 K. It can be concluded that the size of the Ni₄Ti₃ precipitates increases with the increase of the aging temperature. In addition, it can be observed from Fig. 4(b) that the Ni₄Ti₃ precipitates arise at the triple-junction grain boundaries and they are obviously coarser in the grain interior than those at the grain boundary. However, for the NiTi samples aged at 573 and 723 K, the distribution and the size of the Ni₄Ti₃ precipitates at the grain boundary show no difference from those in the grain interior, as shown in Figs. 2(b) and 3(b), respectively.

Fig. 2 TEM images of NiTi alloy aged at 573 K

Fig. 3 TEM images of NiTi alloy aged at 723 K

Fig. 4 TEM images of NiTi alloy aged at 873 K

3.2 DSC analysis

Figure 5 indicates the DSC cuves of NiTi samples which were solution-treated and subsequently aged at 573, 723 and 873 K, respectively. It can be found from Fig. 5 that the solution-treated NiTi sample shows one-stage phase transformation from B2 austenite to B19' martensite on cooling and from B19' martensite to B2 austenite on heating. As for the NiTi sample aged at 873 K, the one-stage phase transformation from B19' martensite to B2 austenite on heating occurs, but the two-stage phase transformation of B2-R-B19' on cooling arises. The NiTi sample aged at 723 K shows the two-stage phase transformation of B2-R-B19' on cooling and the two-stage phase transformation of B19'-R-B2 on heating. However, for the NiTi sample aged at 573 K, the two-stage phase transformation of B19'-R-B2 still takes place on heating, but the three-stage phase transformation can be observed on cooling, where the first and the second DSC peaks correspond to the transformation of B2-R1 and B2-R2, respectively, and the third DSC peak stands for the transformation of R-B19'.

4 Discussion

4.1 Precipitation kinetics of Ni₄Ti₃

It is generally accepted that the precipitation of Ni₄Ti₃ phase is a function of aging temperature and aging time in the aged Ni-rich NiTi alloy, where the size of the Ni₄Ti₃ precipitates increases with the increase of the aging temperature and the aging time. KIM and MIYAZAKI [11] found that in the NiTi sample aged at the low temperature of 373 K for 3000 h, the Ni₄Ti₃ precipitates cannot be detected by means of TEM, and even in the NiTi sample aged at the low temperature of 473 K, the Ni₄Ti₃ precipitates cannot be observed at the aging time less than 10 h. However, the Ni₄Ti₃ precipitates have only 2-3 nm in size even though the NiTi sample is subjected to aging at 473 K for 100 h [11]. In addition, the Ni₄Ti₃ precipitates with 3-4 nm in size can be detected in the NiTi sample aged at 573 K for 1.2 ks, and the size of the Ni₄Ti₃ precipitates increases from 10 to 40 nm when the aging time increases from 100 to 3000 h [11]. It can be concluded that the Ni₄Ti₃ precipitates do not occur until both the aging temperature and the aging time are greater than a critical value, which provides a sufficient driving force for nucleation of Ni₄Ti₃ phase. In the present study, the aging temperature was determined as the variable and the aging time was constant, where the aging temperatures were selected as 573 K (low temperature), 723 K (intermediate temperature) and 873 K (high temperature), respectively, and the aging time was determined as 2 h. The purpose of the three aging treatments was to generate the Ni₄Ti₃ precipitates with different sizes. The size of the Ni₄Ti₃ precipitates was estimated by means of TEM images, in which the diameter of the Ni₄Ti₃ precipitates can be obtained by combining digital micrograph software with equal perimeter method. In the NiTi sample aged at 573 K, the Ni₄Ti₃ precipitates show the uniform size, where the average size of the Ni₄Ti₃ precipitates is about 5 nm. In the NiTi sample aged at 723 K, the Ni₄Ti₃ precipitates also show the uniform size, where the average size of the Ni₄Ti₃ precipitates is about 50 nm. However, in the NiTi sample aged at 873 K, the Ni₄Ti₃ precipitates exhibit the very inhomogeneous size, where the average size of the Ni₄Ti₃ precipitates is about 200 nm, while the maximum precipitation size is about 1200 nm, and the minimum precipitation size is about 40 nm. In the NiTi sample aged at 573 K, the intensity of the diffraction patterns of the Ni₄Ti₃ precipitates is very weak due to small size and small volume fraction, while in the NiTi sample aged at 873 K, the electron diffraction patterns exhibit the basic characteristic of the Ni₄Ti₃ precipitate, where the diffraction spots are located at 1/7 positions along <321> reciprocal vectors of B2 austenite matrix as shown in Fig. 6, which agrees with the results reported in Refs. [6,17,22]. Furthermore, it can be found from Fig. 6 that the orientation relationship between the Ni₄Ti₃ phase and the B2 matrix is determined as  and  where it is convenient to describe the crystal structure of Ni₄Ti₃ phase by means of hexagonal system. It is of great importance to investigate the size of the Ni₄Ti₃ precipitates because whether the Ni₄Ti₃ phase is coherent with the B2 matrix depends on the size of the Ni₄Ti₃ precipitates. For the first time, ZOU et al [10] provided the experimental evidence by means of high resolution electron microscopy that the fine Ni₄Ti₃ precipitate is coherent with the B2 matrix in the Ni-rich NiTi sample aged at 723 K for 3 h, where the size of the Ni₄Ti₃ precipitate is approximately 60 nm. However, in the Ni-rich NiTi sample aged at 873 K for 20 h, large Ni₄Ti₃ precipitate possesses a semi-coherent interface with the B2 matrix, where the size of the Ni₄Ti₃ precipitate is not measured [12]. GALL et al [13,14] proposed that as the Ni₄Ti₃ precipitates grow from nanometer to micrometer in size, they undergo a transition from coherent interface to incoherent interface with the B2 matrix. They further conclude that the Ni₄Ti₃ precipitates up to approximately 100 nm in size remain coherent with the B2 matrix, while the Ni₄Ti₃ precipitates with 500 nm lose coherency with the B2 matrix completely [13,14]. In our study, it can be inferred that the Ni₄Ti₃ precipitates keep the coherent interface with the B2 matrix in the NiTi samples aged at 573 and 723 K, while the Ni₄Ti₃ precipitates which are coherent, semi-coherent and incoherent with the B2 matrix coexist in the NiTi sample aged at 873 K.

Fig. 5 DSC curves of NiTi alloys

Fig. 6 Diffraction patterns of NiTi alloy aged at 873 K showing orientation relationship between Ni₄Ti₃ phase and B2 austenite

It is well known that the distribution of the Ni₄Ti₃ precipitates in the microstructures plays an important role in the transformation behavior of the aged Ni-rich NiTi sample. The distribution characteristic of the Ni₄Ti₃ precipitates depends upon several factors, such as external stress, aging temperature, aging time and chemical composition [9,17,18,20,22,23]. For instance, as compared to stress-free aging, stress-assisted aging results in the microstructures with a homogeneous distribution of Ni₄Ti₃ precipitates in terms of number density [17]. In the case of aging at the low temperature, the longer aging time contributes to obtaining the microstructures with a homogeneous distribution of Ni₄Ti₃ precipitates [20]. Furthermore, high Ni content leads to more homogeneous distribution of the Ni₄Ti₃ precipitates across the whole microstructure in the aged NiTi sample than low Ni content [18], where the grain boundary plays a predominant role. In general, according to the theory of phase transformation kinetics, the Ni₄Ti₃ precipitates preferentially nucleate and grow in the grain boundary region, where the nucleation rate is greater than that in the grain interior due to the decrease of nucleation barrier at the grain boundary. In addition, the grain boundary energy plays a significant role in promoting the nucleation of the Ni₄Ti₃ precipitates at the grain boundary. At the early stage of aging at 873 K, the Ni₄Ti₃ precipitates nucleate and grow at the grain boundary and thus Ni concentration in the grain boundary region is lower than that in the grain interior. Accordingly, Ni atoms in the grain interior diffuse towards the grain boundary region, but the Ni₄Ti₃ precipitates in the grain boundary region are unable to form the coherent interface with the B2 matrix, so the Ni₄Ti₃ precipitates in the grain boundary region are difficult to grow up towards the grain interior due to the impediment of the grain boundary. With the increase of the aging time, the Ni₄Ti₃ precipitates nucleate and grow in the grain interior along the coherent interface with the B2 matrix. Furthermore, the Ni₄Ti₃ precipitates in the grain interior grow larger and larger by absorbing Ni atoms from the B2 matrix around them and the larger Ni₄Ti₃ precipitates gradually merge the smaller Ni₄Ti₃ precipitates. As a result, the Ni₄Ti₃ precipitates lose coherency with the B2 matrix when they grow up to a certain value, and finally the microstructure with an inhomogeneous distribution of Ni₄Ti₃ precipitates is established. In the NiTi sample aged at 573 K, the Ni₄Ti₃ precipitates can also nucleate preferentially at the grain boundary at the beginning, but such a low temperature cannot provide a sufficient driving force for the long-range diffusion of Ni atoms from the grain interior to the grain boundary. Therefore, the supersaturation of Ni atoms in the B2 matrix promotes the homogeneous nucleation of the Ni₄Ti₃ precipitates in the grain interior due to no crystal defects in the solution-treated NiTi sample. In addition, it can be seen from Fig. 2 that the Ni₄Ti₃ precipitates uniformly distribute in the matrix of NiTi alloy, but the local inhomogeneous stress field and the local inhomogeneous chemical composition coexist due to the lower aging temperature and the shorter aging time. The high stress and the low Ni content occur near the Ni₄Ti₃ precipitates, while the low stress and the high Ni content arise away from the Ni₄Ti₃ precipitates. In the NiTi sample aged at 723 K, the appropriate aging temperature and aging time lead to the homogeneous distribution of the Ni₄Ti₃ precipitates in the matrix of NiTi alloy, where the stress field and the chemical composition are relatively homogeneous.

4.2 Influence of Ni₄Ti₃ on phase transformation

It is well known that the Ni₄Ti₃ precipitates have an important influence on the phase transformation behavior of NiTi shape memory alloy. In general, the Ni₄Ti₃ precipitates contribute to the occurrence of the R phase transformation. The R phase belongs to P3 space group and possesses a rhombohedral unit cell structure [24,25]. It can be found from Fig. 7 that in the NiTi samples aged at 573 and 723 K, the R-phase is indexed by means of the electron diffraction patterns, where the superlattice reflections are located at 1/3 positions along <110> and <111> directions of the B2 austenite matrix in the reciprocal space. In addition, the two R-phase variants are observed in the NiTi samples aged at 573 and 723 K. In the NiTi samples aged at 573 and 723 K, the orientation relationship between the two R-phase variants and the B2 austenite matrix is determined as ,, and   , respectively.

Fig. 7 Diffraction patterns of aged NiTi alloy showing orientation relationship between R phase and B2 austenite

It can be seen from Fig. 5 that the two-stage phase transformation of B2-R-B19' on cooling occurs in the NiTi samples aged at 723 and 873 K, while the NiTi samples aged at 573 K exhibit the three-stage phase transformation, which is in accordance with the results reported by KIM et al [19] and ZHOU et al [20,21]. So far, the mechanism of the two-stage phase transformation of B2-R-B19' has been uncontroversial in the aged Ni-rich NiTi shape memory alloy, where the presence of the Ni₄Ti₃ precipitates can impede the occurrence of martensitic phase transformation, which thus contributes to the formation of R phase. In general, the R-phase nucleates at the interface between the Ni₄Ti₃ precipitates and the B2 austenite matrix, and subsequently B19' martensite nucleates at the interface between the Ni₄Ti₃ precipitates and the R-phase [22,23]. The mechanism of the three-stage phase transformation on cooling has attracted great attention over the last decade, but it has been controversial as well. For the NiTi samples aged at high temperatures, three different explanations can be generalized, according to the corresponding experimental observations. One explanation is that the three-stage phase transformation is attributed to the local stress heterogeneity around the Ni₄Ti₃ precipitates, where the transformation of B2-R first occurs, while the second transformation of R-B19' arises in the high stress region near the Ni₄Ti₃ precipitates, and the third transformation of R-B19' arises in the low stress region away from the Ni₄Ti₃ precipitates [15]. Another explanation is that the local composition heterogeneity around the Ni₄Ti₃ precipitates is responsible for the three-stage phase transformation, where the transformation of B2-R first occurs as well, while the second transformation of R-B19' arises in the low Ni region near the Ni₄Ti₃ precipitates, and the third transformation of R-B19' arises in the high Ni region away from the Ni₄Ti₃ precipitates [16]. The last explanation is that the grain boundary effect leads to the large-scale heterogeneity of the Ni₄Ti₃ precipitates between the grain boundary region and the grain interior, which is responsible for the three-stage phase transformation [17,18], where KHALIL-ALLAFI et al [17] and FAN et al [18] gave the different transformation sequences. The transformation sequence obtained by KHALIL-ALLAFI et al [17] is the transformation of B2-R in the grain boundary region with the Ni₄Ti₃ precipitates, the transformation of R-B19' in the same grain boundary region, and the transformation of B2-B19' in the precipitate-free grain interior, while the transformation sequence observed by FAN et al [18] is the transformation of B2-R in the grain boundary region with a high density of Ni₄Ti₃ precipitates, the transformation of B2-B19' in the almost precipitate-free grain interior region, and the transformation of R-B19' in the previous grain boundary region. For the NiTi samples aged at low temperature, ZHOU et al [20,21] proposed that the grain boundary effect also plays a critical role in the three-stage phase transformation, where the grain boundary region with low Ni content first undergoes the transformation of B2-R, and then the grain boundary region with high Ni content undergoes the transformation of B2-R, which finally is followed by the transformation of R-B19'. In the present study, we do not obtain the experiment evidence that the presence of the grain boundary has a predominant influence on the three-stage phase transformation. We suggested that the three-stage phase transformation is attributed to the combination of the local stress heterogeneity and the local composition heterogeneity in the NiTi matrix. In the NiTi samples aged at 573 K for 2 h, the shorter aging time and the lower aging temperature are insufficient to establish an equilibrium Ni concentration between the region near the Ni₄Ti₃ precipitates and the region away from the Ni₄Ti₃ precipitates. Therefore, there is a low Ni content near the Ni₄Ti₃ precipitates, while there is a high Ni content away from the Ni₄Ti₃ precipitates. In addition, the fine Ni₄Ti₃ precipitate is coherent with the B2 matrix, so there is a coherent stress field at the interface between the Ni₄Ti₃ precipitate and the B2 matrix. As a consequence, there is a high stress region near the Ni₄Ti₃ precipitate, while there is a low stress region away from the Ni₄Ti₃ precipitate. In the case of cooling, the R-phase first nucleates at the interface between the Ni₄Ti₃ precipitate and the B2 matrix so as to relax the coherent stress [10], and with the progression of cooling, the transformation of B2-R occurs away from the Ni₄Ti₃ precipitate. Finally, all the R phases are transformed into the B19' martensite. The mechanism of the influence of the inhomogeneous composition on the three-stage phase transformation will be investigated in the future.

5 Conclusions

1) In the NiTi samples aged at 573 and 723 K, the two R-phase variants are observed, and the orientation relationship between the two R-phase variants and the B2 austenite matrix is determined as  and     respectively.

2) The Ni₄Ti₃ precipitates preferentially nucleate and grow at the grain boundary. In the NiTi sample aged at 873 K, the Ni₄Ti₃ precipitates in the grain boundary region are unable to form the coherent interface with the B2 matrix, and thus they are difficult to grow up towards the grain interior due to the impediment of the grain boundary. In the NiTi sample aged at 573 K, the low temperature cannot provide a sufficient driving force for the long-range diffusion of Ni atoms from the grain interior to the grain boundary. In the NiTi sample aged at 723 K, the appropriate aging temperature and aging time lead to the homogeneous distribution of the Ni₄Ti₃ precipitates in the matrix of NiTi alloy.

3) In the NiTi samples aged at 573 and 723 K, the size of the Ni₄Ti₃ precipitates is uniform, where the former is about 5 nm, and the latter is about 50 nm. However, in the NiTi sample aged at 873 K, the Ni₄Ti₃ precipitates exhibit very inhomogeneous size, where the average size of the Ni₄Ti₃ precipitates is about 200 nm, while the maximum precipitation size is about 1200 nm, and the minimum precipitation size is about 40 nm. In the NiTi samples aged at 573 and 723 K, the Ni₄Ti₃ precipitates are coherent with the B2 matrix, while the Ni₄Ti₃ precipitates which are coherent, semi-coherent and incoherent with the B2 matrix coexist in the NiTi sample aged at 873 K.

4) The two-stage phase transformation of B2-R-B19' on cooling occurs in the NiTi samples aged at 723 and 873 K, while the NiTi samples aged at 573 K exhibit the three-stage phase transformation on cooling. The three-stage phase transformation is attributed to the combination of the local stress heterogeneity and the local composition heterogeneity in the NiTi matrix, where the two-stage phase transformation of B2-R occurs in the low Ni and high stress region near the Ni₄Ti₃ precipitates and in the high Ni and low stress region away from the Ni₄Ti₃ precipitates, respectively, which is followed by one-stage transformation of R-B19'.

References

[1] JIANG Shu-yong, ZHANG Yan-qiu, ZHAO Ya-nan. Dynamic recovery and dynamic recrystallization of NiTi shape memory alloy under hot compression deformation [J]. Transactions of Nonferrous Metals Society of China, 2013, 23(1): 140-147.

[2] JIANG Shu-yong, HU Li, ZHAO Ya-nan, ZHANG Yan-qiu, LIANG Yu-long. Plastic yielding of NiTi shape memory alloy under local canning compression [J]. Transactions of Nonferrous Metals Society of China, 2013, 23: 2905-2913.

[3] JIANG Shu-yong, TANG Ming, ZHAO Ya-nan, HU Li, ZHANG Yan-qiu, LIANG Yu-long. Crystallization of amorphous NiTi shape memory alloy fabricated by severe plastic deformation [J]. Transactions of Nonferrous Metals Society of China, 2014, 24(6): 1758-1765.

[4] SABURI T, NENNO S, FUKUDA T. Crystal structure and morphology of the metastable X phase in shape memory Ti-Ni alloys [J]. Journal of the Less-Common Metals, 1986, 125: 157–166.

[5] TADAKI T, NAKATA Y, SHIMIZU K, OTSUKA K. Crystal structure, composition and morphology of a precipitate in an aged Ti-51at.%Ni shape memory alloy [J]. Transactions of Japan Institute of Metals, 1986, 27: 731–740.

[6] NISHIDA M, WAYMAN C M, KAINUMA R, HONMA T. Further electron microscopy studies of the Ti₁₁Ni₁₄ phase in an aged Ti-52at.%Ni shape memory alloy [J]. Scripta Metallurgica, 1986, 20: 899–904.

[7] NISHIDA M, WAYMAN C M. Electron microscopy studies of precipitation processes in near-equiatomic TiNi shape memory alloys [J]. Materials Science and Engineering A, 1987, 93: 191–203.

[8] JIANG Shu-yong, ZHAO Ya-nan, ZHANG Yan-qiu, HU Li, LIANG Yu-long. Effect of solution treatment and aging on microstructural evolution and mechanical behavior of NiTi shape memory alloy [J]. Transactions of Nonferrous Metals Society of China, 2013, 23(12): 3658-3667.

[9] HEINEN R, HOPPE U, HACKL K. Prediction of microstructural patterns in monocrystalline shape memory alloys using global energy minimization [J]. Materials Science and Engineering A, 2008, 481-482: 462-465.

[10] ZHOU N, SHEN C, WAGNER M F X, EGGELER G, MILLS M J, WANG Y. Effect of Ni₄Ti₃ precipitation on martensitic transformation in Ti-Ni [J]. Acta Materialia, 2010, 58: 6685-6694.

[11] KIM J I, MIYAZAKI S. Effect of nano-scaled precipitates on shape memory behavior of Ti-50.9at.%Ni alloy [J]. Acta Materialia, 2005, 53: 4545-4554.

[12] ZOU W H, HART X D, WANG R, ZHANG Z, ZHANG W Z, LAI J K L. TEM and HREM study of the interphase interface structure of Ti₃Ni₄ precipitates and parent phase in an aged TiNi shape memory alloy [J]. Materials Science and Engineering A, 1996, 219: 142-147.

[13] GALL K, JUNTUNEN K, MAIER H J, SEHITOGLU H, CHUMLYAKOV Y I. Instrumented micro-indentation of NiTi shape-memory alloys [J]. Acta Materialia, 2001, 49: 3205-3217.

[14] GALL K, MAIER H J. Cyclic deformation mechanisms in precipitated NiTi shape memory alloys [J]. Acta Materialia, 2002, 50: 4643-4657.

[15] BATAILLARD L, BIDAUX J E, GOTTHARDT R. Interaction between microstructure and multiple-step transformation in binary NiTi alloys using in-situ transmission electron microscopy observations [J]. Philosophical Magazine A, 1998, 78: 327-344.

[16] ALLAFI K J, REN X, EGGELER G. The mechanism of multistage martensitic transformations in aged Ni-rich NiTi shape memory alloys [J]. Acta Materialia, 2002, 50: 793-803.

[17] KHALIL-ALLAFI J,  A, EGGELER G. Ni₄Ti₃-precipitation during aging of NiTi shape memory alloys and its influence on martensitic phase transformations [J]. Acta Materialia, 2002, 50: 4255-4274.

[18] FAN G, CHEN W, YANG S, ZHU J, REN X B, OTSUKA K. Origin of abnormal multi-stage martensitic transformation behavior in aged Ni-rich Ti-Ni shape memory alloys [J]. Acta Materialia, 2004, 52: 4351-4362.

[19] KIM J I, LIU Y N, MIYAZAKI S. Ageing-induced two-stage R-phase transformation in Ti-50.9at.%Ni [J]. Acta Materialia, 2004, 52: 487-499.

[20] ZHOU Y M, ZHANG J, FAN G L, DING X D, SUN J, REN X B, OTSUKA K. Origin of 2-stage R-phase transformation in low-temperature aged Ni-rich Ti-Ni alloys [J]. Acta Materialia, 2005, 53: 5365-5377.

[21] ZHOU Y, FAN G, ZHANG J, DING X, REN X, SUN J, OTSUKA K. Understanding of multi-stage R-phase transformation in aged Ni-rich Ti-Ni shape memory alloys [J]. Materials Science and Engineering A, 2006, 438-440: 602-607.

[22] MICHUTTA J, SOMSEN C, YAWNY A,  A, EGGELER G. Elementary martensitic transformation processes in Ni-rich NiTi single crystals with Ni₄Ti₃ precipitates [J]. Acta Materialia, 2006, 54: 3525-3542.

[23]  A, BOJDA O, SOMSEN C, EGGELER G. Conventional and in-situ transmission electron microscopy investigations into multistage martensitic transformations in Ni-rich NiTi shape memory alloys [J]. Materials Science and Engineering A, 2008, 481-482: 409-413.

[24] GORYCZKA T, MORAWIEC H J. Structure studies of the R-phase using X-ray diffraction methods [J]. Journal of Alloys and Compounds, 2004, 367: 137-141.

[25] KHALIL-ALLAFI J, SCHMAHL W W, TOEBBENS D M. Space group and crystal structure of the R-phase in binary NiTi shape memory alloys [J]. Acta Materialia, 2006, 54: 3171-3175.

Ni₄Ti₃沉淀相对镍钛形状记忆合金相变的影响

江树勇¹，张艳秋¹，赵亚楠¹，刘思维²，胡 励³，赵成志³

1. 哈尔滨工程大学 机电工程学院，哈尔滨 150001；

2. 湖北省水利水电科学研究院，武汉 430070；

3. 哈尔滨工程大学 材料科学与化学工程学院，哈尔滨 150001

摘  要：镍钛形状记忆合金样品分别在573、723和873 K时效2 h。在573和723 K时效的镍钛样品中，可以观察到2个R相变体，而且这2个R相变体与B2基体之间的取向关系可以确定。在573 和723 K时效的镍钛样品中，细小均匀的Ni₄Ti₃沉淀相与B2奥氏体基体之间保持共格关系。在873 K时效的镍钛样品中，可以观察到Ni₄Ti₃沉淀相在晶粒内部和晶界处析出，不均匀的Ni₄Ti₃沉淀相与B2基体之间保持共格、半共格和非共格关系。对于873 K时效的镍钛样品，加热时发生B19'马氏体向B2奥氏体转变的一阶相变，但在冷却时发生B2-R-B19'的二阶相变。723 K时效的镍钛样品在冷却时也表现出B2-R-B19'的二阶相变，但在加热时发出B19'-R-B2的二阶相变。对于573 K时效的镍钛样品，由于Ni₄Ti₃沉淀相周围的局部应力不均匀和局部成分不均匀，则其发生三阶相变。

关键词：形状记忆合金；镍钛合金；马氏体相变；热处理；显微组织

(Edited by Xiang-qun LI)

Foundation item: Projects (51305091, 51305092, 51475101) supported by the National Natural Science Foundation of China

Corresponding author: Shu-yong JIANG; Tel: +86-13936266338; E-mail: jiangshy@sina.com

DOI: 10.1016/S1003-6326(15)64056-0