Abstract
The influence of Pr addition on the microstructure, phase transformation, and hardness of Ni50Ti50-xPrx (x=0, 0.1, 0.3, 0.5, 0.7, 0.9, at%) shape memory alloys was investigated. Results show that Ni50Ti50-xPrx alloys consist of NiTi matrix and NiPr precipi-tates. Ni50Ti49.5Pr0.5 alloy has the optimal properties of martensitic transformation start temperature of 73 °C, the thermal hysteresis as narrow as 37 °C, and the Vickers hardness of 2850 MPa. Pr addition obviously decreases the martensitic transformation temperature of NiTi alloys, but the composite alloy maintains a relatively narrow thermal hysteresis and relatively high Vickers hardness, compared with other NiTi-based shape memory alloys.
Science Press
NiTi-based shape memory alloys (SMAs) have unique shape memory effects and superelasticity behavior, which have been applied in various fields, especially in engineering and medical applications[1]. Current researches about SMAs mainly focus on controlling the transformation temperatures and improving the shape memory effect. The phase transfor-mation behavior of NiTi-based SMAs is strongly dependent on composition, alloying, precipitates, and heat treatment history[2]. Experimental researches suggest that the influence factors of phase transformation temperature are the elastic properties of the parent austenite and certain microstructural features, such as precipitates[3]. Alloying is a frequently used approach for tuning the transformation temperatures of NiTi-based SMAs. Alloying elements change not only the transformation temperatures but also the transformation product and the transformation route of NiTi-based SMAs[4]. To date, transition metals such as Fe[5], Cu[6], Nb[7], Ta[8], Zr[9], Hf[10], Au[11], and Pd[12] have been adopted as alloying elements for NiTi-based SMAs, and their influences on microstructure, phase transformation, and mechanical properties have been comprehensively studied. It is found that most transition metals lower the transformation temperatures, but some transition metals, such as Zr, Hf, Au, and Pd, increase the transformation temperatures by 100 °C[9]. Therefore, NiTiZr, NiTiHf, NiTiAu, and NiTiPd are considered as candidates for high temperature SMAs[2]. However, Au and Pd are noble metals, indicating that the large-scale applications of NiTiAu and NiTiPd are restricted because of their high cost. Considering the high transformation temperatures and low cost, NiTiHf SMA is a favorable candidate of high tempe-rature SMAs for engineering applications[13]. However, NiTiHf SMAs always exhibit a wide thermal hysteresis (>50 °C), which is not beneficial to fast actuation applications (such as Micro-electronmechanical system and robotics)[13,14]. Thus, quaternary NiTi-based SMAs such as NiTiHfTa[15] are designed for improving the functional properties through microstructural modifications. However, the thermal hyste-resis of Ni49Ti36Hf15-xTax (x=0, 3, 6, 9, 12, at%) SMAs is still wide, ranging from 50.0 °C to 80.8 °C.
Rare earth (RE) elements are often added into various alloys to tune their properties. The influence of RE elements (such as Ce[16], Gd[17], Dy[18], Nd[19], and La[20]) addition on the microstructure and phase transformation behavior of binary NiTi SMAs has also been studied. The addition of Ce, Gd, and Dy into NiTi alloy can increase the transformation temperature because some new type second precipitates exist in the TiNiRE SMAs, changing the Ni/Ti ratio of the matrix[21]. Recently, the influence of RE elements addition on micro-structure, phase transformation, and mechanical properties of TiNiHfY SMAs has been systematically investigated[22]. It is found that the Y addition slightly decreases the transformation temperature, whereas the thermal hysteresis is still wide and almost keeps constant at ~50 °C as the Y content changes. Pr is another widely applied RE element, especially in the use of magnetic materials. However, few studies have been conducted on the addition of Pr into NiTi SMA. Therefore, the influence of Pr addition on microstructure, phase transfor-mation, and mechanical properties of TiNiPr SMAs remains unclear.
In this research, NiTiPr SMAs with Pr content of 0.1at%~0.9at% were prepared via arc-melting, and the microstructure, phase transformation behavior, and hardness were studied.
The NiTiPr alloys were prepared by melting raw materials of Ti (purity 99.9wt%), Ni (purity 99.7wt%), and Pr (purity 99.95wt%) with different nominal composition in a vacuum non-consumable arc-melting furnace using a water-cooled copper crucible. The Ni50Ti50-xPrx alloys with x=0, 0.1, 0.3, 0.5, 0.7, 0.9 are Ni50Ti50, Ni50Ti49.9Pr0.1, Ni50Ti49.7Pr0.3, Ni50Ti49.5Pr0.5, Ni50Ti49.3Pr0.7, Ni50Ti49.1Pr0.9 alloys in atom percentage, respec-tively. Arc-melting was conducted for four times to ensure the uniformity of composite. The as-prepared alloys were spark-cut from the ingots and solution-treated at 850 °C for 1 h in a vacuum quartz tube furnace. Subsequently, the alloys were quenched with water. Thereafter, the specimens were mecha-nically and lightly polished to obtain a plain surface for the observation of microstructure and phase transformation behavior.
The morphology of Ni50Ti50-xPrx alloys was observed using scanning electron microscope (SEM, TM3030, Hitachi, Tokyo, Japan) equipped with energy dispersive spectrometer (EDS, Oxford). X-ray diffraction (XRD) patterns were obtained using a D/MAX-2500PC X-ray diffractometer (Rigaku, Tokyo, Japan). The phase transformation tempera-tures of the Ni50Ti50-xPrx alloys were measured by differential scanning calorimeter (DSC, Q2000, TA Instrument, New Castle, USA) with a scanning range of 0~550 °C and a scan-ning rate of 10 °C/min during heating and cooling procedure. The microstructure was in-situ observed under a transmission electron microscope (TEM, JEM-2010, JEOL Ltd., Tokyo, Japan) at 200 kV. TEM images were recorded with a Gatan 1 k×1 k CCD and processed using the DigitalMicrograph software. The hardness of Ni50Ti50-xPrx and NiTi SMAs were measured using a HXD-1000TMC/LCD Vickers hardness tester.
Fig.1 depicts SEM back-scattered electron (BSE) images of all six alloys. For NiTi alloy without Pr, the morphology is uniform and featureless (Fig.1a). For Ni50Ti50-xPrx alloys with x=0.1, 0.3, 0.5, 0.7, 0.9, white particles and gray matrixes can be identified in Fig.1b~1f, and the number of the particles shows an evident increase with the increase in Pr content. The chemical composition of Ni50Ti50-xPrx alloys was analyzed by EDS and summarized in Table 1. The Ni:Ti ratio in the matrix of each alloy is very close to 1. Thus, the matrix may be a NiTi binary alloy[23]. According to the isothermal section of ternary alloy phase diagram of Ni50Ti50-xPrx alloys at 773 K, no intermetallic compounds can be found in the TiPr binary alloy system[24]. Furthermore, the phase diagram shows 7 kinds of intermetallic compounds: PrNi, Pr3Ni, Pr7Ni3, PrNi2, PrNi3, Pr2Ni7, and PrNi5[24]. The EDS results show that the Ni:Pr:Ti ratio in the white particles is approximately 48:40:12 and they may be NiPr precipitate with Ti solute.
Fig.1 SEM-BSE images of Ni50Ti50-xPrx alloys: (a) Ni50Ti50, (b) Ni50Ti49.9Pr0.1, (c) Ni50Ti49.7Pr0.3, (d) Ni50Ti49.5Pr0.5, (e) Ni50Ti49.3Pr0.7, and
Table 1 Chemical composition of Ni50Ti50-xPrx SMAs (at%)
| x | Component | Ti | Ni | Pr |
|---|
|
0 |
Matrix |
49.5 |
50.5 |
- |
|
0.1 |
Matrix |
49.7 |
50.3 |
0.0 |
|
Precipitate |
11.2 |
48.5 |
40.3 |
|
0.3 |
Matrix |
49.1 |
50.9 |
0.0 |
|
Precipitate |
11.9 |
48.2 |
39.9 |
|
0.5 |
Matrix |
49.1 |
50.9 |
0.0 |
|
Precipitate |
10.8 |
48.7 |
40.5 |
|
0.7 |
Matrix |
49.4 |
50.6 |
0.0 |
|
Precipitate |
12.5 |
48.2 |
39.3 |
|
0.9 |
Matrix |
49.7 |
50.3 |
0.0 |
|
Precipitate |
12.6 |
48.2 |
39.2 |
Fig.2a depicts the XRD patterns of Ni50Ti50-xPrx alloys with x=0, 0.1, 0.3, 0.5, 0.7, 0.9 at room temperature. The diffraction peaks are identified as NiTi B19′ martensite, NiTi B2 austenite, NiTi2, NiTi, and NiPr after comparison with JCPDF cards of No.65-0145, No.65-4572, No.72-0442, No.65- 2038, and No.19-0837, respectively. The detailed crystal plane indices of Ni50Ti50 and Ni50Ti49.5Pr0.5 at room temperature are presented in Fig.2b. The relative intensity of XRD curves markedly varies owing to the fraction difference of martensite and austenite phases. The letter M and A in Fig.2a denotes the NiTi B19′ martensite and NiTi B2 austenite, respectively. Thus, it can be confirmed that the matrix in Ni50Ti50-xPrx alloys is near-equiatomic NiTi binary alloy, whereas the white particles in Ni50Ti50-xPrx alloys are NiPr precipitate with Ti solute. This microstructure is highly similar to that of NiTiCe[16], NiTiGd[17], NiTiDy[18], NiTiY[25], and NiTiNd[19].
Fig.3 depicts the DSC curves of Ni50Ti50-xPrx alloys and NiTi SMAs. Each DSC curve shows only one peak during the cooling process, which indicates a one-step B2↔B19′ phase transformation[14]. DSC curves of Ni50Ti50-xPrx alloys with x=0.3, 0.5, 0.7, 0.9 show two peaks during the heating process, which indicate a two-step B19′→R→B2 transformation. Ac-cording to the DSC curves, the martensitic transformation start temperature Ms, martensitic transformation peak tempe-rature Mp, and austenite transformation peak temperature Ap are obtained and summarized in Table 2. The Ms of Ni50Ti49.9Pr0.1 alloy reaches to 77 °C, which is the highest Ms of the Ni50Ti50-xPrx alloys and NiTi SMAs. Meanwhile, with the increase of Pr content, the R→B2 transformation separates from B19′→R transformation, which is similar to the result of Ni4Ti3 precipitate in the NiTi binary alloy. Therefore, the NiPr precipitate introduces an extra transformation step and results in a two-step phase transformation. For NiTi-based alloys, the addition of a third element often changes the transformation temperature, transformation product, or the transformation route[2] Ni50Ti49.5Pr0.5 alloy has a Ms of 73 °C, and a thermal hysteresis of 37 °C.
Table 2 Transformation temperature and hysteresis HM of diffe-rent NiTi-based SMAs (°C)
| Alloy | Ms | Mp | Ap | HM | Ref. |
|---|
|
Ni50Ti49.9Pr0.1 |
77 |
63 |
97 |
34 |
|
|
Ni50Ti49.7Pr0.3 |
74 |
63 |
97 |
34 |
|
|
Ni50Ti49.5Pr0.5 |
73 |
61 |
98 |
37 |
|
|
Ni50Ti49.3Pr0.7 |
63 |
54 |
92 |
38 |
|
|
Ni50Ti49.1Pr0.9 |
53 |
46 |
84 |
38 |
|
|
Ni50Ti50 |
77 |
67 |
99 |
32 |
|
|
Ni50.3Ti49.7 |
21 |
10 |
48 |
38 |
[26] |
|
Ni50.45Ti49.05Ce0.5 |
26 |
20 |
60 |
40 |
[16] |
|
Ni50.7Ti49.3Gd1 |
65 |
25 |
60 |
35 |
[17] |
|
Ni50.2Ti48.8Dy1 |
40 |
24 |
62 |
38 |
[18] |
|
Ni50.2Ti48.8Y1 |
36 |
25 |
55 |
30 |
[25] |
|
Ni50Ti49.5La0.5 |
57 |
40 |
80 |
40 |
[20] |
|
Ni50Ti48W2 |
-37 |
- |
- |
39 |
[4] |
|
Ni50Ti45Ta5 |
44 |
- |
- |
53 |
[4] |
It is well known that the Ms of NiTi binary SMA is strongly dependent on Ni content[2]. Ni content increases by about 0.1at% resulting in the decrease of Ms of NiTi binary SMAs by more than 10 °C when Ni content is higher than 50at%[27]. As shown in Table 2, the Ms of Ni50Ti50 SMA is 77 °C, whe-reas the Ms of Ni50.3Ti49.7 SMA decreases to 21 °C because of the decrease in Ni content. The martensitic transformation finish temperature Mf in all Ni50Ti50-xPrx alloys is lower than the room temperature (25 °C). Thus, the martensite transfor-mation cannot completely finish at room temperature, and both NiTi B19′ martensite and NiTi B2 austenite exist in the matrix of Ni50Ti50-xPrx alloys. These findings are in agreement with the XRD results. Although the actual Ni content of Ni50Ti50-xPrx alloys has a slight difference, the Ms is not notably affected by Ni content, and drops steeply when Ni content increases above the equiatomic content (50at%)[13]. The Ni contents in the matrix of Ni50Ti50-xPrx alloys are very close to each other and not higher than 50at%. Therefore, the Ni content is not the main reason for the decrease of Ms of Ni50Ti50-xPrx alloys. The martensitic transformation as a coordinated atom movement related to the distance to the habit plane occurs at the crystalline state. Energy accommo-dated around vacancies, dislocation, and precipitate decreases driving force of the martensitic transformation. The stress generated in these zones requires additional energy for triggering the occurrence of martensitic transformation. In consequence, further overcooling is required for continues transformation. Therefore, the Ms decreases to lower temperature region. It is also well known that the precipitate such as Ni4Ti3 in binary NiTi SMA significantly influences the Ms due to the stress state and the Ni depletion in the matrix around the precipitates[2]. In this work, the NiTi matrix and NiPr precipitates form during the preparation process. After heat treatment, the chemical composition of the matrix and precipitates should be uniform; meanwhile the size and quantity of the NiPr precipitates increase with the increase of Pr content. Therefore, the stress around NiPr precipitates is responsible for the decrease of Ms with the increase of Pr content in Ni50Ti50-xPrx alloys. A better crystallographic compatibility and a narrow thermal hysteresis are obtained for the Ni50Ti50-xPrx alloys, compared to other NiTi-based alloys[28].
Fig.4 depicts the TEM bright-field (BF) images and sele-cted area electron diffraction (SAED) patterns of martensitic transformation of the Ni50Ti49.5Pr0.5 alloy. At room temperature, two different morphologies are identified in the TEM images (Fig.4a). According to the related SAED and XRD patterns, the observed area in Fig.4a comprises two phases: the austenite (A) located at the upper left side and the martensite (M) with strip area, which shows the process of martensitic transformation of Ni50Ti49.5Pr0.5 alloy clearly. With increasing the temperature, the martensite changes to austenite. However, during the cooling process, the austenite changes to martensite again, which is verified by the EDS and XRD results.
Fig.4 TEM BF images and SAED patterns of martensitic transformation (a~d) and reverse transformation (e, f) of Ni50Ti49.5Pr0.5 alloy at 25 ℃ (a), 120 ℃ (b), 150 ℃ (c), 170 ℃ (d), 40 ℃ (e), and 25 ℃ (f)
Fig.5 depicts the XRD patterns of Ni50Ti49.5Pr0.5 alloy at 25 and 150 ℃. The diffraction peaks are identified as the NiTi B19′ martensite and NiTi B2 austenite after comparison with JCPDF cards of No.65-0145 and No.65-4572, respectively. At 25 ℃, the Ni50Ti49.5Pr0.5 alloy contains NiTi B19′ martensite, NiTi B2 austenite, and NiPr precipitate, whereas at 150 ℃, the Ni50Ti49.5Pr0.5 alloy contains NiTi B2 austenite and NiPr precipitate, which is in agreement with the in-situ TEM observation (Fig.4).
Fig.5 XRD patterns of Ni50Ti49.5Prf0.5 alloy at 25 and 150 ℃
Hardness is closely related to the material yield strength and has been regarded as an indicative parameter of mechanical properties of NiTi-based SMAs[14,22,26,29,30]. Thus, the hardness value of Ni50Ti50-xPrx alloys was measured, as shown in Fig.6. The hardness value of Ni50Ti49.5Pr0.5 alloy is 2850 MPa, which is larger than that of binary Ni50Ti50 SMA. This can be explained by the solid solution strengthening effect. The bigger Pr atoms and substitution of Pr for Ti in Ni50Ti50-xPrx alloys lead to the linear increase in alloy hardness[31]. Compared with that of Ni50Ti49.5Pr0.5 alloy, the hardness of Ni50Ti49.3Pr0.7 and Ni50Ti49.1Pr0.9 decreases by 200 (7%,) and 300 MPa (10.5%,), respectively. Previous reports demonstrate that the hardness of NiTi-based SMAs can be influenced by precipitates[14], thermal treatment[30], and quater-nary alloying[22], as shown in Table 2. SEM-BSE observation (Fig.1) confirms that the size of NiPr precipitates increases with the increase of Pr content in Ni50Ti50-xPrx alloy. Conse-quently, the yield strength decreases with the increase of Pr content in Ni50Ti50-xPrx alloys[14]. The decline in hardness of Ni50Ti50-xPrx alloys can be explained by the Orowan mecha-nism[30]. The martensitic transformation start temperature decreases gradually with the increase of Pr content. However, Ni50Ti50-xPrx alloys keep a relatively narrow thermal hysteresis and high Vickers hardness, compared with other NiTi-based SMAs.
1) The microstructure of Ni50Ti50-xPrx alloys consists of the NiTi matrix and NiPr precipitates. The Ni50Ti50-xPrx alloys have a two-step phase transformation during heating process and one-step transformation during cooling process.
2) Ni50Ti49.5Pr0.5 alloy has a martensitic transformation start temperature of 73 °C, a thermal hysteresis of 37 °C and a Vic-kers hardness as high as 2850 MPa.
3) The martensitic transformation start temperature decrea-ses gradually with the increase of Pr content. However, Ni50Ti50-xPrx alloys keep a relatively narrow thermal hysteresis and high Vickers hardness, compared with other NiTi-based shape memory alloys.
References
1 Otsuka K, Ren X. Intermetallics[J], 1999, 7(5): 511
[Baidu Scholar]
2 Otsuka K, Ren X. Prog Mater Sci[J], 2005, 50(5): 511
[Baidu Scholar]
3 Michutta J, Somsen C, Yawny A et al. Acta Mater[J], 2006, 54(13): 3525
[Baidu Scholar]
4 Zarinejad M, Liu Y. Adv Funct Mater[J], 2013, 18: 2789
[Baidu Scholar]
5 Frenzel J, Pfetzing J, Neuking K et al. Mat Sci Eng A[J], 2008, 481-482: 635
[Baidu Scholar]
6 Tomozawa M, Kim H Y, Miyazaki S. Acta Mater[J], 2009, 57(2): 441
[Baidu Scholar]
7 Zheng Y F, Zhao L C, Ye H Q. Mat Sci Eng A[J], 1999, 273-275: 271
[Baidu Scholar]
8 Gong C W, Wang Y N, Yang D Z. Mater Chem Phys[J], 2006, 96(2-3): 183
[Baidu Scholar]
9 Bertheville B. J Alloy Compd[J], 2009, 398(1-2): 94
[Baidu Scholar]
10 Meng X L, Cai W, Fu Y D et al. Intermetallics[J], 2008, 16(5): 698
[Baidu Scholar]
11 Casalena L, Coughlin D R, Yang F. Microsc Microanal[J], 2015, 21(S3): 607
[Baidu Scholar]
12 Liu Y, Kohl M, Okutsu K et al. Mat Sci Eng A[J], 2004, 378(1-2): 205
[Baidu Scholar]
13 Ma J, Karaman I, Noebe D R. Int Mater Rev[J], 2010, 55(5): 257
[Baidu Scholar]
14 Karaca H E, Saghaian S M, Ded G et al. Acta Mater[J], 2013, 61(19): 7422
[Baidu Scholar]
15 Prasad R V S, Park C H, Kim S W et al. J Alloy Compd[J], 2017, 697: 55
[Baidu Scholar]
16 Cai W, Liu A L, Sui J H et al. Mater Trans[J], 2006, 47(3): 716
[Baidu Scholar]
17 Liu A L, Cai W, Gao Z Y et al. Mat Sci Eng A[J], 2006, 438-440: 634
[Baidu Scholar]
18 Liu A L, Gao Z Y, Gao L et al. J Alloy Compd[J], 2007, 437(1-2): 339
[Baidu Scholar]
19 Dovchinvanching M, Zhao C W, Zhao S L et al. Adv Mater Sci Eng[J], 2014, 2014: 489 701
[Baidu Scholar]
20 Zhao C W, Li W Y, Zhao S L et al. Vacuum[J], 2017, 137: 169
[Baidu Scholar]
21 Cai W, Meng X L, Zhao L C. Curr Opin Solid St M[J], 2005, 9(6): 296
[Baidu Scholar]
22 Yi X Y, Gao W H, Meng X L et al. Alloy Compd[J], 2017, 705: 98
[Baidu Scholar]
23 Itin V I, Bratchikov A D, Merzhanov A G et al. Combust Explos Shock[J], 1982, 18(5): 536
[Baidu Scholar]
24 Zhong X P, Yang Z, Liu J Q. J Alloy Compd[J], 2001, 316(1-2): 172
[Baidu Scholar]
25 Liu A L, Sui J H, Lei Y C et al. J Mater Sci[J], 2007, 42(12): 5791
[Baidu Scholar]
26 Prasher M, Sen D, Tewari R et al. Mater Chem Phys[J], 2020, 247: 122 890
[Baidu Scholar]
27 Frenzel J, George E P, Dlouhy A et al. Acta Mater[J], 2010, 58(9): 3444
[Baidu Scholar]
28 Frenzel J, Wieczorek A, Opahle I et al. Acta Mater[J], 2015, 90: 213
[Baidu Scholar]
29 Yi X, Sun K, Gao W et al. J Alloy Compd[J], 2018, 735: 1219
[Baidu Scholar]
30 Moshref-Javadi M, Seyedein S H, Salehi M T et al. Acta Mater[J], 2013, 61(7): 2583
[Baidu Scholar]
31 Suresh K S, Kim D I, Bhaumik S K et al. Intermetallics[J], 2014, 44: 18
[Baidu Scholar]
