- Email: [email protected]
Effect of cold rolling ratio on the microstructure and recovery properties of Ti-Ni-Nb-Co shape memory alloys
Accepted Manuscript Effect of cold rolling ratio on the microstructure and recovery properties of Ti-Ni-NbCo shape memory alloys Bo Cui, Jian Yao, Ye Wu, Wei Cai PII:
S0925-8388(18)33353-X
DOI:
10.1016/j.jallcom.2018.09.113
Reference:
JALCOM 47532
To appear in:
Journal of Alloys and Compounds
Received Date: 23 May 2018 Revised Date:
5 September 2018
Accepted Date: 11 September 2018
Please cite this article as: B. Cui, J. Yao, Y. Wu, W. Cai, Effect of cold rolling ratio on the microstructure and recovery properties of Ti-Ni-Nb-Co shape memory alloys, Journal of Alloys and Compounds (2018), doi: 10.1016/j.jallcom.2018.09.113. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Effect of cold rolling ratio on the microstructure and recovery properties of Ti-Ni-Nb-Co shape memory alloys Bo Cui, Jian Yao, Ye Wu, Wei Cai∗
RI PT
School of Materials Science and Engineering, Harbin Institute of Technology, Harbin 150001, China
Abstract
SC
The Ti44.5Ni44.5Nb9Co2 alloys with different cold rolling ratios were annealed at
M AN U
650oC to investigate the influence of cold rolling ratio on the microstructure and recovery properties. The change in cold rolling ratio remarkably affected the distribution of β-Nb phase and the density of dislocations and precipitates. The density of dislocations increased with the increase of cold rolling ratio, which
TE D
facilitated preferential nucleation and growth of Ti2Co precipitates. Meanwhile, high density of dislocations and precipitates retarded the recovery and recrystallization process. Owing to the strengthening effect of dislocations and nanoscale precipitates,
EP
the 40% cold rolled Ti44.5Ni44.5Nb9Co2 alloy annealed at 650oC after 16% pre-strain
AC C
exhibited 25% enhancement in the total recovery strain and over 50% improvement for the recovery stress compared to the typical Ti-Ni-Nb alloys. Key words: Ti-Ni-Nb-Co alloys; wide hysteresis shape memory alloys; nanoscale precipitates; recovery and recrystallization; recovery properties.
∗
Corresponding author. E-mail address: [email protected] (W. Cai). 1
ACCEPTED MANUSCRIPT
1. Introduction In the past decades, shape memory pipe couplings have been successfully applied in the numerous fields, such as hydraulic lines and oil and gas pipelines,
RI PT
owing to simple installation and high fasten force [1, 2]. Earlier, Ti-Ni-Fe alloys have been used widely in the aerospace applications because of higher recovery stress [3]. But Ti-Ni-Fe shape memory pipe couplings are limited in engineering applications
SC
because of storage problem at room temperature [4, 5]. Nowadays, Ti-Ni-Nb alloys
M AN U
exhibit wide transformation hysteresis to bring more convenience for the engineering applications, which have attracted much interest as shape memory pipe couplings. However, Ti-Ni-Nb alloys with proper deformation can acquire wide transformation hysteresis at the expenses of recovery properties [6]. Therefore, improving recovery
TE D
properties while simultaneously maintaining wide transformation hysteresis is the crucial challenge for Ti-Ni-Nb alloys. Until now, some approaches have been developed to improve the recovery properties of Ti-Ni-Nb alloys, including
EP
thermal-mechanical treatment [7], Equal Channel Angular Pressing (ECAP) [8],
AC C
control of Nb content [9, 10] and fourth element addition [11-13]. Although the above-mentioned methods have been supposed to improve recovery properties in Ti-Ni-Nb alloys, the recovery performance is still not qualified for the requirements as shape memory pipe couplings. Thereby, searching for new strategies to substantially improve recovery properties is critical for Ti-Ni-Nb shape memory pipe couplings. Precipitation strengthening is an important method to improve the mechanical and recovery properties in the Ti-Ni-based alloys [14-16]. Recently, our group has 2
ACCEPTED MANUSCRIPT found a novel nanoscale Ti2Co precipitate formed in the annealed Ti-Ni-Nb-Co alloys [17]. Coherent GP zones and precipitates could produce higher resistance against deformation to significantly improve yield strength and recovery stress. In the
RI PT
previous researches, it was notable to find that the precipitation behavior had strong relationship with the Co content and annealing temperature in the annealed Ti-Ni-Nb-Co alloys. In the present work, the role of the dislocations density on the
SC
precipitation behavior has been systematically investigated. The microstructure
M AN U
evolution with cold rolling ratio has been also revealed in the annealed Ti44.5Ni44.5Nb9Co2 alloy. It is highly expected to further clarify the relationship between the density of dislocations and precipitates and transformation characteristics and recovery properties.
TE D
2. Experimental
The initially as-cast Ti44.5Ni44.5Nb9Co2 alloy was prepared by arc-melting method under Ar atmosphere. The ingots were re-melted six times for the composition
EP
homogeneity and flipped over after each melting. The homogenized ingot was sliced
AC C
into plates with different thickness for cold rolling process. These sliced plates were cold rolled up to 1mm in multiple steps with each step of 0.02mm and then annealed at 650oC for 2h. The final cold rolling ratios were given to be 10%, 20% and 40%, respectively. Hereafter, the samples were referred to as 10CRA, 20CRA and 40CRA according to their cold rolling ratio. The as-cast alloy without cold rolling process was also annealed at 650oC for 2h, which was regarded as reference sample labeled by 0CRA. All the heat treatments were done in Ar filled quartz tubes followed by water 3
ACCEPTED MANUSCRIPT cooling without crushing the tubes. The phase structure was determined by X-ray diffraction analysis (XRD, PANalytical X’Pert Pro) using Cu-Kα radiation with wavelength of 0.154178 nm.
RI PT
The morphology and microstructure were observed by scanning electron microscopy (SEM, FEI Quanta 200F) and transmission electron microscopy (TEM, FEI Talos 200X). The detailed preparation of TEM samples was given by Ref [17]. The
SC
transformation temperatures were detected by differential scanning calorimeter (DSC,
M AN U
PerkineElmer Diamond) with constant heating/cooling rate of 20oC/min. The mechanical and recovery properties were carried out on ThermoMechanical Analyzer (TMA, HRJ WDW-1D) at the temperature of Ms+30oC. The dimensions of all the samples for mechanical and recovery tests were 20mm×1mm×1mm and the
TE D
deformation rate is 0.2mm/min. The recovery properties tests include two separated experiments: one was recovery strain test, the other one was the recovery stress test. Two kinds of samples both were loaded up to 16% and then unloaded. Afterwards, the
EP
samples were heated over reverse martensite transformation finishing temperature.
AC C
The starting and finishing temperatures of reverse transformation temperature after deformation were marked as As' and A'f . The recovery strain test was measuring the length after heating to calculate the recovery strain. The recovery stress test was keeping crosshead displacement still to record the relationship between stress and heating temperature.
3. Results and Discussions Fig. 1 shows room temperature XRD profiles of 0CRA, 10CRA, 20CRA and 4
ACCEPTED MANUSCRIPT 40CRA samples. It is obvious that all the profiles can be indexed as a multiple structure, comprising of B2 parent phase, β-Nb phase and Ti2Co phase. These results are in accordance with the previous phase constitution of annealing samples [17].
RI PT
From Fig. 1a, it can be seen that the diffraction peak of B2 {110} and {200} has little difference on peak intensity in the 0CRA alloy. Distinctly, the diffraction peak intensity corresponding to B2 {110} becomes stronger while the relative intensity of
SC
B2 {200} diffraction peak decreases in the XRD profiles of sample with cold rolled
M AN U
deformation. This demonstrates that deformation texture remains after cold rolling and annealing process in the Ti44.5Ni44.5Nb9Co2 alloy. Besides, the zoom-in XRD patterns between 40o 48o are illustrated in Fig. 1b. As shown in Fig. 1b, the relative peak intensity of the Ti2Co phase further increases with the increase of cold rolling
TE D
ratio, which means the size or quantity of Ti2Co phase increases somewhat. The quantitative analysis on the Ti2Co precipitates in all the samples with different cold rolling ratios will be discussed in detail.
EP
Fig. 2 shows back-scattered SEM images of 0CRA, 10CRA, 20CRA and 40CRA
AC C
samples. The morphology of all the samples clearly shows typical eutectic microstructure. As shown in Fig. 2, the dark regions are Ti-Ni matrix while the white regions are composed of both β-Nb phase particles and stripe β-Nb/B2 eutectic colonies. But no indication of Ti2Co-type precipitates can be found. From Fig. 2a, β-Nb phase with network structure is dispersedly distributed on the Ti-Ni matrix of the 0CRA alloy. Compared to this microstructure, the morphology of β-Nb phase exhibits significant difference after cold rolled deformation: β-Nb phase has been 5
ACCEPTED MANUSCRIPT heavily elongated into strips and particles along with the cold rolling direction (Figs. 2b-d). To be more exact, the distance between β-Nb strips becomes smaller and the amount of broken β-Nb particles increases with cold rolling ratio increasing. Contrast
RI PT
to severe deformation microstructure of β-Nb phase, the Ti-Ni matrix shows no obvious change. It can generally be recognized that β-Nb is soft phase suffering more plastic deformation than Ti-Ni matrix during cold rolling process.
SC
Fig. 3 displays the TEM bright field images of 0CRA 10CRA, 20CRA and
M AN U
40CRA samples. It can be obviously seen that spherical precipitates are homogeneously distributed in the 0CRA sample, as shown in Fig. 3a. The microstructure of 10CRA sample reveals the presence of low density dislocations (marked by red arrows in Fig. 3d). The precipitates are preferentially nucleated on the
TE D
heterogeneous dislocations provided by cold rolled deformation. As cold rolling ratio increases up to 20%, the density of precipitates and dislocations increases and simultaneously the size of the precipitates increases (Fig. 3c). Interesting point can be
EP
noticed: for the 10CRA and 20CRA samples, only occurs dislocations rearrangement
AC C
annealing at 650oC. Normally, the recrystallization temperature of Ti-Ni-Nb alloys was previously reported to be between 400oC and 500oC [7]. It is noticeable that nanoscale precipitates nucleate in the defect and dislocation sites, which can be strong cause of depression in recovery and recrystallization process. The higher annealing temperature is needed to eliminate the relatively stable defects for the recovery and recrystallization process. Nevertheless, in the case of 40CRA sample, the situation is different. Instead of the dislocations rearrangement in the 10CRA and 20CRA samples, 6
ACCEPTED MANUSCRIPT the recovery and recrystallization process of the 40CRA sample occurs at 650oC shown in Fig. 3d. The recovery microstructure and new distortionless equiaxed grains can be observed, with an average grain size of 150nm. A large number of
RI PT
nano-precipitates (marked by red arrows in Fig. 3d) are distributed in the grain interiors and boundaries. The corresponding SAED pattern of the area circled by red square shows the precipitates are determined to be Ti2Co phase, which has
SC
face-center-cubic structure with lattice parameter of 0.6730nm. Moreover, the Ti2Co
M AN U
precipitates possess orientation relationship with B2 parent phase: (110)TiNi//(220)Ti2Co, [100]TiNi//[100]Ti2Co, as discussed in our previous research [17]. Meanwhile, the Moiré patterns (marked by white arrows in Fig. 3d) of several nanometers in size also can be seen around the Ti2Co precipitates, which are ascribed to coherent strain field between
TE D
Ti-Ni matrix and Ti2Co precipitates. In contrast to the 10CRA and 20CRA samples, the recovery and recrystallization process of the 40% cold deformed sample is inclined to occur. The heavily deformed samples, such as the 40CRA sample, store
EP
more energy during cold rolled deformation and release more driving force to easily
AC C
perform the recovery and recrystallization process in the lower annealing temperature [18]. Therefore, the deformation extent and the distribution of the precipitates have significant impact on the recovery and recrystallization process. To further verify the accurate size and volume fraction of Ti2Co precipitates, the
high magnification microstructure of all the samples with different cold rolling ratios are shown in Fig. 4. The spherical Ti2Co precipitates with Moiré patterns can be easily identified within the grain interior. The statistical precipitates size distributions of all 7
ACCEPTED MANUSCRIPT the samples are plotted in the inserts of Fig. 4 and they may be approximated to be a log-normal distribution. The average precipitates size of 0CRA, 10CRA, 20CRA and 40CRA samples are determined to be 6.57nm, 8.97nm, 11.09nm and 11.18nm,
RI PT
respectively. Compared with the 0CRA sample, the cold rolled samples have the Ti2Co precipitates with the larger size, in which the heterogeneous dislocations facilitate the precipitation growth. For the cold rolled samples, the precipitates size
SC
slightly increases with the increase of cold rolling ratio. Meanwhile, the volume
M AN U
fraction V p of the spherical precipitates can be calculated by the following equation when the truncation and overlap correction are considered [19]: 2D V p = − ln (1 − A ) D + 3T
where the A is the projected area fraction of the Ti2Co precipitates determined by
TE D
using the commercial software Image J, D is the average diameter of the spherical precipitate and T is the foil thickness determined by the convergent-beam electron
EP
diffraction (CBED) patterns [20]. In this way, the volume fraction of Ti2Co precipitates in the 0CRA, 10CRA, 20CRA and 40CRA samples can be calculated to
AC C
be 0.47vol.%, 0.54vol.%, 0.92vol.% and 1.25vol.%, respectively. Consequently, the calculated results indicate the volume fraction of Ti2Co precipitates increases with the cold rolling ratio increasing, which is in accordance with the previous XRD results. This is because more dislocations introduced by larger cold-rolled amount process produce more nucleation sites for Ti2Co precipitates. Fig. 5 shows the DSC curves of 0CRA, 10CRA, 20CRA and 40CRA samples. As shown in Fig.5a, the 0CRA sample exhibits two-stage transformation during heating 8
ACCEPTED MANUSCRIPT and cooling process, which is different with the previous results of Ti-Ni-Nb alloys [21]. It is noteworthy to find that the additional endothermic peak and exothermic peak appear to be observed prior to the normal transformation peak of B2B19’. The
RI PT
extra phase transformation has much lower and broader peak than that of B2B19’. It can possibly be indexed as two-stage transformation B2RB19’. The occurrence of R-phase transformation can be seen to be considered a result of
SC
homogeneous nanoscale precipitation. Dispersed Ti2Co precipitates produce higher
M AN U
coherent strain field to cause strong resistance to martensite transformation. Nevertheless, the formation of R-phase is related to the smaller lattice deformation, which is less vulnerable to strong strain field. Consequently, the presence of coherent Ti2Co precipitates can separate the R-phase transformation from the martensite
TE D
transformation. Similar phenomenon has been reported in aged Ni-rich Ti-Ni alloys, suggesting that fine and uniform precipitation induced internal stress field leading to the occurrence of R-transformation [22-24]. The confirmation of R-phase
EP
transformation will be performed in our other studies later. The DSC curves of the
AC C
samples with cold rolled deformation are given in Figs. 5b-d. It can be clearly observed that there is only one step phase transformation during the heating and cooling process in the samples with cold rolled deformation. The introduction of dislocations via cold rolled deformation can relax much more misfit in the interface between matrix and precipitates, resulting in the disappearance of R-phase transformation. Besides, the martensite transformation peak shows lower, whereas transformation hysteresis gradually increases with the cold rolling increasing. 9
ACCEPTED MANUSCRIPT Fig. 6 shows stress-strain curves of 0CRA, 10CRA, 20CRA and 40CRA samples performed at the temperature of Ms+30oC. All the samples show a typical stress vs. strain response of a stress-induced martensite (SIM) state. As shown in Fig. 6, the
RI PT
critical stress for SIM can be determined by the tangent line method. It can be seen from Fig. 6 that the critical stress for SIM and yield stress gradually increase with increased cold rolling ratio. Furthermore, the 40CRA sample exhibits the highest
SC
critical stress for SIM of 320MPa and the highest yield stress of 1280MPa owing to
M AN U
high density of dislocations and precipitates. Meanwhile, the elongation ratio of samples slightly increases from 20% to 28% as the cold rolling ratio increases. The highest elongation ratio of the 40CRA sample results from a large population of dispersed β-Nb formed during prior cold rolled deformation, which can confirm the
TE D
previous findings by Zheng et. al [25].
From the stress-strain curves, another interesting point can be noticed: no flat stress plateau can be seen in the 0CRA and 10CRA samples and simultaneously the
EP
stress plateau gradually becomes flat with the increase of cold rolling ratio. The
AC C
40CRA sample exhibits a flat stress plateau. The flatness seems to indicate as if Lüders-like deformation from Ti-Ni-based alloys. In the case of the typical Ti-Ni-based alloys, the flat stress plateau means some favorably oriented martensites first form and then increase the localized internal stress to trigger neighboring martensite transformation, indicating self-propagation of SIM or detwining under constant load [1, 26, 27]. Yet, for the 0CRA and 10CRA samples, there is a large population of β-Nb phase with slight deformation distributed in the Ti-Ni grain 10
ACCEPTED MANUSCRIPT boundaries (Figs. 2a and b). The self-propagation of SIM can be constrained by β-Nb phase at the grain boundaries. Thus, the stress is not enough to induce neighboring martensite transformation so that the external load is required to assist the further
RI PT
process of SIM in this stage. However, β-Nb phase is distributed as narrow stripes or particles manner in the 40CRA sample (Fig. 2d). The self-propagation of SIM is much easier than that of the 0CRA and 10CRA samples. Basically constant external stress
M AN U
the SIM mechanism of typical Ti-Ni-based alloys.
SC
can be observed in such a stress plateau of the 40CRA sample, which is similar with
On the basis of the effect of the deformation temperature and strain on the transformation hysteresis and strain recovery ratio [6], the transformation hysteresis slightly increases with deformation temperature at the deformation strain of 16%. As
rapidly to
TE D
the deformation temperature near Ms+30oC, the transformation hysteresis increases 150oC and simultaneously recovery strain remains high. However, with
the deformation temperature and strain further increasing, the transformation
EP
hysteresis does not further increase, while the recovery properties significantly
AC C
deteriorate. The aforementioned analysis indicates the transformation hysteresis can be effectively realized at the characteristic temperature (Ms+30oC) and strain range ( 16%). Herein, the recovery properties and transformation characteristics of all the samples with different cold rolling ratios after 16% pre-strain at Ms+30oC are shown in Fig. 7 and Table 1. Fig. 7a shows the loading-unloading curves of 0CRA, 10CRA, 20CRA and 40CRA samples with 16% pre-strain at the temperature of Ms+30oC. The curving arrows represent shape recovery after heating above A'f . As shown in Fig. 7a, 11
ACCEPTED MANUSCRIPT it can be found that the total recovery strain increases with the increased cold rolled amount. The 40CRA sample shows the largest total recovery strain of 13.52% (including elastic recovery strain of 3.81% and shape memory recovery strain of
RI PT
9.71%), which exhibits 25% enhancement in the total recovery strain compared to the previous Ti-Ni-Nb alloys [6] owing to the strengthening effect of high-density nano-precipitation and dislocations. Fig. 7b shows the recovery stress dependent
SC
heating temperature of all the samples. The maximum recovery stress shows an
M AN U
increase with the cold rolling ratio increasing likewise. As known, the maximum recovery stress is related with the reversible recovery strain and yield strength. As for the 40CRA sample, a large population of precipitates and dislocations leads to a larger recovery strain and the higher yield strength. The highest recovery stress of 584MPa
TE D
can be obtained in the 40CRA sample, which displays over 50% improvement in contrast to the typical Ti-Ni-Nb alloys (386MPa) [28]. As shown in Fig. 7b and Table 1, the starting temperature of reverse transformation ( As' ) decreases with the increase
EP
of cold rolling ratio. Simultaneously, the slope of the stress-heating temperature
AC C
curves becomes large with the cold rolled amount increasing. This indicates that the dislocations introduced by cold rolled deformation can be increased to bring more preferential sites for the martensite nucleation, which facilitates the formation and growth of reverse martensite transformation. Thus, the cold rolled deformation has significant impact on the reverse transformation of SIM. Concurrently, the transformation hysteresis of the samples shows almost unchanged upon different cold rolling ratios, shown in Table 1. The transformation hysteresis of all the samples with 12
ACCEPTED MANUSCRIPT different cold rolling ratios maintains about 150oC. It can be concluded that the transformation hysteresis of the samples highly depends on the prior deformation strain instead of cold rolling ratio. As a result, Ti-Ni-Nb-Co alloys with wide
RI PT
transformation hysteresis and simultaneously better recovery properties can be achieved, which can be the most promising candidate as shape memory pipe couplings.
SC
4. Conclusion
M AN U
In our research, the effect of cold rolling ratio on the microstructure and recovery properties of an annealed Ti44.5Ni44.5Nb9Co2 alloy has been systematically investigated. The main results of this study are summarized as followed:
1. As cold rolling ratio increases, the β-Nb phase has changed from network
TE D
structure to stripes or particles and simultaneously the Ti2Co precipitates size slightly increases and the volume fraction of precipitates ranges from 0.47vol.% to 1.25vol.%. 2. Precipitates nucleated at the dislocations remarkably retard the recovery and
EP
recrystallization process, but the greater cold rolled deformation facilitates occurrence
AC C
of the recovery and recrystallization process. 3. The Ti44.5Ni44.5Nb9Co2 alloy with 40% cold rolled amount annealed at 650oC
can obtain the highest yield stress and the best recovery properties owing to high-density nanoscale precipitates and dislocations.
Acknowledgements This research was financed by National Natural Science Foundation of China (Grant Nos. 51731005, 51501049 and 51471060). 13
ACCEPTED MANUSCRIPT
References [1] K. Otsuka, X. Ren, Physical metallurgy of Ti–Ni-based shape memory alloys, Prog. Mater Sci. 50 (2005) 511-678.
RI PT
[2] J.M. Jani, M. Leary, A. Subic, M.A. Gibson, A review of shape memory alloy research, applications and opportunities, Mater. Des. 56 (2014) 1078-1113.
[3] T.W. Duerig, A.R. Pelton, Materials properties handbook: Ti-Ni shape memory
SC
alloys, 683 (1994) 1035-1048.
M AN U
[4] R. Basu, M. Eskandari, L. Upadhayay, M.A. Mohtadi-Bonab, J.A. Szpunar, A systematic investigation on the role of microstructure on phase transformation behavior in Ni–Ti–Fe shape memory alloys, J. Alloys Compd. 645 (2015) 213-222.
TE D
[5] J. Frenzel, J. Pfetzing, K. Neuking, G. Eggeler, On the influence of thermomechanical treatments on the microstructure and phase transformation
635-638.
EP
behavior of Ni–Ti–Fe shape memory alloys, Mater. Sci. Eng. A s481–482 (2008)
AC C
[6] C.S. Zhang, L.C. Zhao, T.W. Duerig, C.M. Wayman, Effects of deformation on the transformation hysteresis and shape memory effect in a Ni47Ti44Nb9, Scripta Metallurgica Et Materiala 24 (1990) 1807-1812.
[7] M. Piao, S. Miyazaki, K. Otsuka, Characteristics of deformation and transformation in Ti44Ni47Nb9 shape memory alloy, Mater. Trans. 33 (1992) 346-353. [8] Y.X. Tong, P.C. Jiang, F. Chen, B. Tian, L. Li, Y.F. Zheng, D.V. Gunderov, R.Z. 14
ACCEPTED MANUSCRIPT Valiev, Microstructure and martensitic transformation of an ultrafine-grained TiNiNb shape memory alloy processed by equal channel angular pressing, Intermetallics 49 (2014) 81-86.
shape memory alloys with low Nb content
RI PT
[9] C. Ying, H.C. Jiang, L.J. Rong, X. Li, X.Q. Zhao, Mechanical behavior in NiTiNb Intermetallics 19 (2011) 217-220.
[10] K. Uchida, N. Shigenaka, T. Sakuma, Y. Sutou, K. Yamauchi, Effect of Nb
SC
Content on Martensitic Transformation Temperatures and Mechanical Properties
M AN U
of Ti-Ni-Nb Shape Memory Alloys for Pipe Joint Applications, Mater. Trans. 48 (2007) 445-450.
[11] Y. Wang, Y.F. Zheng, Y.X. Tong, B. Tian, F. Chen, L. Li, Microstructure and martensitic transformation of TiNiNbB shape memory alloys, Intermetallics 64
TE D
(2015) 32-36.
[12] C. Wang, Q. Gao, Y. Yuan, H. Zhang, J. Zhang, Q. Wang, F. Qu, Microstructure evolutions of Ni-Ti-Nb-Al alloys with different Al addition, J. Alloys Compd. 695
EP
(2017) 2923-2929.
AC C
[13] V.Y. Abramov, N.M. Aleksandrova, D.V. Borovkov, I.Y. Khmelevskaya, A.V. Korotitskiy, S.Y. Makushev, N.A. Polyakova, N.N. Popov, S.D. Prokoshkin, Martensitic transformations and functional properties of thermally and thermomechanically treated Ti–Ni–Nb-based alloys, Mater. Sci. Eng. A 438 (2006) 553-557. [14] M.I. Khan, H.Y. Kim, F.A. Khalid, S. Miyazaki, Effect of cold rolling ratio on the nanoscale precipitation behavior of TiNiPdCu based high temperature shape 15
ACCEPTED MANUSCRIPT memory alloys, J. Alloys Compd. 599 (2014) 212-218. [15] S.U. Rehman, M. Khan, A.N. Khan, M.I. Khan, L. Ali, S.H.I. Jaffery, Effect of precipitation hardening and thermomechanical training on microstructure and
RI PT
shape memory properties of Ti50Ni15Pd25Cu10 high temperature shape memory alloys, J. Alloys Compd. 616 (2014) 275-283.
[16] H.E. Karaca, S.M. Saghaian, G. Ded, H. Tobe, B. Basaran, H.J. Maier, R.D.
SC
Noebe, Y.I. Chumlyakov, Effects of nanoprecipitation on the shape memory and
M AN U
material properties of an Ni-rich NiTiHf high temperature shape memory alloy, Acta Mater. 61 (2013) 7422-7431.
[17] B. Cui, J. Yao, Y. Wu, X. Zhang, F.L. Wang, J.H. Sui, W. Cai, Precipitation behavior and mechanical properties of Ti-Ni-Nb-Co alloys, Intermetallics 95
TE D
(2018) 40-47.
[18] W.D. Callister, Fundamentals of Materials Science and Engineering: An Integrated Approach, 2nd Edition, John Wiley and Sons Ltd, 2005.
EP
[19] E.E. Underwood, Stereology, or the quantitative evaluation of microstructures,
AC C
Journal of Microscopy 89 (1969) 161-180. [20] D.B. Williams, C.B. Carter, Transmission Electron Microscopy: A Textbook for Materials Science, Springer, New York, USA, 2009.
[21] Y. Zheng, W. Cai, Y. Wang, Y. Luo, L. Zhao, Effects of Heat Treatment on the Transformation Temperature and the Microstructure of Ni-Ti-Nb Shape Memory Alloy, J. Mater. Sci. Technol. 14 (1998) 37-40. [22] G. Fan, W. Chen, S. Yang, J. Zhu, X. Ren, K. Otsuka, Origin of abnormal 16
ACCEPTED MANUSCRIPT multi-stage martensitic transformation behavior in aged Ni-rich Ti–Ni shape memory alloys, Acta Mater. 52 (2004) 4351-4362. [23] L.J. Chiang, C.H. Li, Y.F. Hsu, W.H. Wang, Age-induced four-stage
RI PT
transformation in Ni-rich NiTi shape memory alloys, J. Alloys Compd. 458 (2008) 231-237.
[24] D. Xue, Y. Zhou, X. Ren, The effect of aging on the B2-R transformation
SC
behaviors in Ti-51at%Ni alloy, Intermetallics 19 (2011) 1752-1758.
M AN U
[25] Y. Zheng, W. Cai, Y. Luo, Effects of heat treatment on tensile properties and shape memory effects of Ni-Ti-Nb alloys, T. Nonferr. Metal. Soc. (1997) 93-97. [26] L. Orgéas, D. Favier, Stress-induced martensitic transformation of a NiTi alloy in isothermal shear, tension and compression, Acta Mater. 46 (1998) 5579-5591.
TE D
[27] Y. Liu, Detwinning process and its anisotropy in shape memory alloys, Proc Spie 4234 (2001) 82-93.
[28] W. Cai, C. Zhang, L. Zhao, Recovery Stress in a Ni-Ti-Nb Shape Memory Alloy
AC C
EP
with Wide Transformation Hysteresis, J. Mate. Sci. Technol. 17 (1994) 27-30.
17
ACCEPTED MANUSCRIPT
Figures and Table Captions Fig. 1 a) Room temperature XRD profiles of all the samples with different cold rolling ratios, b) the zoom-in XRD patterns between 40o 48o.
RI PT
Fig. 2 Back-scattered SEM images of all the samples with different cold rolling ratios; a) 0CRA, b) 10CRA, c) 20CRA, d) 40CRA; cold rolling direction (RD) and transverse direction (TD) are marked. Fig. 3 TEM bright field images of all the samples with different cold rolling ratios and corresponding SAED pattern; a) 0CRA, b) 10CRA, c) 20CRA, d) 40CRA.
M AN U
SC
Fig. 4 High magnification TEM bright field images of all the samples with different cold rolling ratios, a) 0CRA, b) 10CRA, c) 20CRA, d) 40CRA; The statistical precipitates size distributions of all the samples are in the inserts. Fig. 5 DSC curves of all the samples with different cold rolling ratios; a) 0CRA, b) 10CRA, c) 20CRA, d) 40CRA. Fig. 6 Stress-strain curves of all the samples with different cold rolling ratios preformed at Ms+30oC; a) 0CRA, b) 10CRA, c) 20CRA, d) 40CRA.
TE D
Fig. 7 Recovery properties of all the samples with different cold rolling ratios after 16% pre-strain at Ms+30oC, (a) the loading-unloading curves, (b) recovery stress curves dependent heating temperature.
AC C
EP
Table 1 The recovery properties and transformation characteristics of all the samples with different cold rolling ratios after 16% pre-strain at Ms+30oC.
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
Figure 1
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
Figure 2
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
Figure 3
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
Figure 4
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
Figure 5
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
Figure 6
AC C
EP
TE D
M AN U
SC
Figure 7
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Table 1 Recovery Strain (%)
Recovery Stress (MPa)
(oC)
(oC)
Transformation Hysteresis ( - ) (oC)
0CRA 10CRA 20CRA 40CRA
11.65 12.24 12.80 13.52
464 492 523 584
-87 -98 -105 -111
60 54 42 35
147 152 149 146
AC C
EP
TE D
M AN U
SC
RI PT
Samples
ACCEPTED MANUSCRIPT Prior cold rolled deformation plays important role on the microstructure of alloy. High-density precipitates and dislocations can improve recovery properties. The recovery strain of the alloy annealed at 650oC after 16% strain shows
RI PT
13.25%. The recovery stress of the alloy annealed at 650oC after 16% strain shows 584MPa.
SC
The alloy possesses wide transformation hysteresis and better recovery
AC C
EP
TE D
M AN U
properties.
S0925-8388(18)33353-X
DOI:
10.1016/j.jallcom.2018.09.113
Reference:
JALCOM 47532
To appear in:
Journal of Alloys and Compounds
Received Date: 23 May 2018 Revised Date:
5 September 2018
Accepted Date: 11 September 2018
Please cite this article as: B. Cui, J. Yao, Y. Wu, W. Cai, Effect of cold rolling ratio on the microstructure and recovery properties of Ti-Ni-Nb-Co shape memory alloys, Journal of Alloys and Compounds (2018), doi: 10.1016/j.jallcom.2018.09.113. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Effect of cold rolling ratio on the microstructure and recovery properties of Ti-Ni-Nb-Co shape memory alloys Bo Cui, Jian Yao, Ye Wu, Wei Cai∗
RI PT
School of Materials Science and Engineering, Harbin Institute of Technology, Harbin 150001, China
Abstract
SC
The Ti44.5Ni44.5Nb9Co2 alloys with different cold rolling ratios were annealed at
M AN U
650oC to investigate the influence of cold rolling ratio on the microstructure and recovery properties. The change in cold rolling ratio remarkably affected the distribution of β-Nb phase and the density of dislocations and precipitates. The density of dislocations increased with the increase of cold rolling ratio, which
TE D
facilitated preferential nucleation and growth of Ti2Co precipitates. Meanwhile, high density of dislocations and precipitates retarded the recovery and recrystallization process. Owing to the strengthening effect of dislocations and nanoscale precipitates,
EP
the 40% cold rolled Ti44.5Ni44.5Nb9Co2 alloy annealed at 650oC after 16% pre-strain
AC C
exhibited 25% enhancement in the total recovery strain and over 50% improvement for the recovery stress compared to the typical Ti-Ni-Nb alloys. Key words: Ti-Ni-Nb-Co alloys; wide hysteresis shape memory alloys; nanoscale precipitates; recovery and recrystallization; recovery properties.
∗
Corresponding author. E-mail address: [email protected] (W. Cai). 1
ACCEPTED MANUSCRIPT
1. Introduction In the past decades, shape memory pipe couplings have been successfully applied in the numerous fields, such as hydraulic lines and oil and gas pipelines,
RI PT
owing to simple installation and high fasten force [1, 2]. Earlier, Ti-Ni-Fe alloys have been used widely in the aerospace applications because of higher recovery stress [3]. But Ti-Ni-Fe shape memory pipe couplings are limited in engineering applications
SC
because of storage problem at room temperature [4, 5]. Nowadays, Ti-Ni-Nb alloys
M AN U
exhibit wide transformation hysteresis to bring more convenience for the engineering applications, which have attracted much interest as shape memory pipe couplings. However, Ti-Ni-Nb alloys with proper deformation can acquire wide transformation hysteresis at the expenses of recovery properties [6]. Therefore, improving recovery
TE D
properties while simultaneously maintaining wide transformation hysteresis is the crucial challenge for Ti-Ni-Nb alloys. Until now, some approaches have been developed to improve the recovery properties of Ti-Ni-Nb alloys, including
EP
thermal-mechanical treatment [7], Equal Channel Angular Pressing (ECAP) [8],
AC C
control of Nb content [9, 10] and fourth element addition [11-13]. Although the above-mentioned methods have been supposed to improve recovery properties in Ti-Ni-Nb alloys, the recovery performance is still not qualified for the requirements as shape memory pipe couplings. Thereby, searching for new strategies to substantially improve recovery properties is critical for Ti-Ni-Nb shape memory pipe couplings. Precipitation strengthening is an important method to improve the mechanical and recovery properties in the Ti-Ni-based alloys [14-16]. Recently, our group has 2
ACCEPTED MANUSCRIPT found a novel nanoscale Ti2Co precipitate formed in the annealed Ti-Ni-Nb-Co alloys [17]. Coherent GP zones and precipitates could produce higher resistance against deformation to significantly improve yield strength and recovery stress. In the
RI PT
previous researches, it was notable to find that the precipitation behavior had strong relationship with the Co content and annealing temperature in the annealed Ti-Ni-Nb-Co alloys. In the present work, the role of the dislocations density on the
SC
precipitation behavior has been systematically investigated. The microstructure
M AN U
evolution with cold rolling ratio has been also revealed in the annealed Ti44.5Ni44.5Nb9Co2 alloy. It is highly expected to further clarify the relationship between the density of dislocations and precipitates and transformation characteristics and recovery properties.
TE D
2. Experimental
The initially as-cast Ti44.5Ni44.5Nb9Co2 alloy was prepared by arc-melting method under Ar atmosphere. The ingots were re-melted six times for the composition
EP
homogeneity and flipped over after each melting. The homogenized ingot was sliced
AC C
into plates with different thickness for cold rolling process. These sliced plates were cold rolled up to 1mm in multiple steps with each step of 0.02mm and then annealed at 650oC for 2h. The final cold rolling ratios were given to be 10%, 20% and 40%, respectively. Hereafter, the samples were referred to as 10CRA, 20CRA and 40CRA according to their cold rolling ratio. The as-cast alloy without cold rolling process was also annealed at 650oC for 2h, which was regarded as reference sample labeled by 0CRA. All the heat treatments were done in Ar filled quartz tubes followed by water 3
ACCEPTED MANUSCRIPT cooling without crushing the tubes. The phase structure was determined by X-ray diffraction analysis (XRD, PANalytical X’Pert Pro) using Cu-Kα radiation with wavelength of 0.154178 nm.
RI PT
The morphology and microstructure were observed by scanning electron microscopy (SEM, FEI Quanta 200F) and transmission electron microscopy (TEM, FEI Talos 200X). The detailed preparation of TEM samples was given by Ref [17]. The
SC
transformation temperatures were detected by differential scanning calorimeter (DSC,
M AN U
PerkineElmer Diamond) with constant heating/cooling rate of 20oC/min. The mechanical and recovery properties were carried out on ThermoMechanical Analyzer (TMA, HRJ WDW-1D) at the temperature of Ms+30oC. The dimensions of all the samples for mechanical and recovery tests were 20mm×1mm×1mm and the
TE D
deformation rate is 0.2mm/min. The recovery properties tests include two separated experiments: one was recovery strain test, the other one was the recovery stress test. Two kinds of samples both were loaded up to 16% and then unloaded. Afterwards, the
EP
samples were heated over reverse martensite transformation finishing temperature.
AC C
The starting and finishing temperatures of reverse transformation temperature after deformation were marked as As' and A'f . The recovery strain test was measuring the length after heating to calculate the recovery strain. The recovery stress test was keeping crosshead displacement still to record the relationship between stress and heating temperature.
3. Results and Discussions Fig. 1 shows room temperature XRD profiles of 0CRA, 10CRA, 20CRA and 4
ACCEPTED MANUSCRIPT 40CRA samples. It is obvious that all the profiles can be indexed as a multiple structure, comprising of B2 parent phase, β-Nb phase and Ti2Co phase. These results are in accordance with the previous phase constitution of annealing samples [17].
RI PT
From Fig. 1a, it can be seen that the diffraction peak of B2 {110} and {200} has little difference on peak intensity in the 0CRA alloy. Distinctly, the diffraction peak intensity corresponding to B2 {110} becomes stronger while the relative intensity of
SC
B2 {200} diffraction peak decreases in the XRD profiles of sample with cold rolled
M AN U
deformation. This demonstrates that deformation texture remains after cold rolling and annealing process in the Ti44.5Ni44.5Nb9Co2 alloy. Besides, the zoom-in XRD patterns between 40o 48o are illustrated in Fig. 1b. As shown in Fig. 1b, the relative peak intensity of the Ti2Co phase further increases with the increase of cold rolling
TE D
ratio, which means the size or quantity of Ti2Co phase increases somewhat. The quantitative analysis on the Ti2Co precipitates in all the samples with different cold rolling ratios will be discussed in detail.
EP
Fig. 2 shows back-scattered SEM images of 0CRA, 10CRA, 20CRA and 40CRA
AC C
samples. The morphology of all the samples clearly shows typical eutectic microstructure. As shown in Fig. 2, the dark regions are Ti-Ni matrix while the white regions are composed of both β-Nb phase particles and stripe β-Nb/B2 eutectic colonies. But no indication of Ti2Co-type precipitates can be found. From Fig. 2a, β-Nb phase with network structure is dispersedly distributed on the Ti-Ni matrix of the 0CRA alloy. Compared to this microstructure, the morphology of β-Nb phase exhibits significant difference after cold rolled deformation: β-Nb phase has been 5
ACCEPTED MANUSCRIPT heavily elongated into strips and particles along with the cold rolling direction (Figs. 2b-d). To be more exact, the distance between β-Nb strips becomes smaller and the amount of broken β-Nb particles increases with cold rolling ratio increasing. Contrast
RI PT
to severe deformation microstructure of β-Nb phase, the Ti-Ni matrix shows no obvious change. It can generally be recognized that β-Nb is soft phase suffering more plastic deformation than Ti-Ni matrix during cold rolling process.
SC
Fig. 3 displays the TEM bright field images of 0CRA 10CRA, 20CRA and
M AN U
40CRA samples. It can be obviously seen that spherical precipitates are homogeneously distributed in the 0CRA sample, as shown in Fig. 3a. The microstructure of 10CRA sample reveals the presence of low density dislocations (marked by red arrows in Fig. 3d). The precipitates are preferentially nucleated on the
TE D
heterogeneous dislocations provided by cold rolled deformation. As cold rolling ratio increases up to 20%, the density of precipitates and dislocations increases and simultaneously the size of the precipitates increases (Fig. 3c). Interesting point can be
EP
noticed: for the 10CRA and 20CRA samples, only occurs dislocations rearrangement
AC C
annealing at 650oC. Normally, the recrystallization temperature of Ti-Ni-Nb alloys was previously reported to be between 400oC and 500oC [7]. It is noticeable that nanoscale precipitates nucleate in the defect and dislocation sites, which can be strong cause of depression in recovery and recrystallization process. The higher annealing temperature is needed to eliminate the relatively stable defects for the recovery and recrystallization process. Nevertheless, in the case of 40CRA sample, the situation is different. Instead of the dislocations rearrangement in the 10CRA and 20CRA samples, 6
ACCEPTED MANUSCRIPT the recovery and recrystallization process of the 40CRA sample occurs at 650oC shown in Fig. 3d. The recovery microstructure and new distortionless equiaxed grains can be observed, with an average grain size of 150nm. A large number of
RI PT
nano-precipitates (marked by red arrows in Fig. 3d) are distributed in the grain interiors and boundaries. The corresponding SAED pattern of the area circled by red square shows the precipitates are determined to be Ti2Co phase, which has
SC
face-center-cubic structure with lattice parameter of 0.6730nm. Moreover, the Ti2Co
M AN U
precipitates possess orientation relationship with B2 parent phase: (110)TiNi//(220)Ti2Co, [100]TiNi//[100]Ti2Co, as discussed in our previous research [17]. Meanwhile, the Moiré patterns (marked by white arrows in Fig. 3d) of several nanometers in size also can be seen around the Ti2Co precipitates, which are ascribed to coherent strain field between
TE D
Ti-Ni matrix and Ti2Co precipitates. In contrast to the 10CRA and 20CRA samples, the recovery and recrystallization process of the 40% cold deformed sample is inclined to occur. The heavily deformed samples, such as the 40CRA sample, store
EP
more energy during cold rolled deformation and release more driving force to easily
AC C
perform the recovery and recrystallization process in the lower annealing temperature [18]. Therefore, the deformation extent and the distribution of the precipitates have significant impact on the recovery and recrystallization process. To further verify the accurate size and volume fraction of Ti2Co precipitates, the
high magnification microstructure of all the samples with different cold rolling ratios are shown in Fig. 4. The spherical Ti2Co precipitates with Moiré patterns can be easily identified within the grain interior. The statistical precipitates size distributions of all 7
ACCEPTED MANUSCRIPT the samples are plotted in the inserts of Fig. 4 and they may be approximated to be a log-normal distribution. The average precipitates size of 0CRA, 10CRA, 20CRA and 40CRA samples are determined to be 6.57nm, 8.97nm, 11.09nm and 11.18nm,
RI PT
respectively. Compared with the 0CRA sample, the cold rolled samples have the Ti2Co precipitates with the larger size, in which the heterogeneous dislocations facilitate the precipitation growth. For the cold rolled samples, the precipitates size
SC
slightly increases with the increase of cold rolling ratio. Meanwhile, the volume
M AN U
fraction V p of the spherical precipitates can be calculated by the following equation when the truncation and overlap correction are considered [19]: 2D V p = − ln (1 − A ) D + 3T
where the A is the projected area fraction of the Ti2Co precipitates determined by
TE D
using the commercial software Image J, D is the average diameter of the spherical precipitate and T is the foil thickness determined by the convergent-beam electron
EP
diffraction (CBED) patterns [20]. In this way, the volume fraction of Ti2Co precipitates in the 0CRA, 10CRA, 20CRA and 40CRA samples can be calculated to
AC C
be 0.47vol.%, 0.54vol.%, 0.92vol.% and 1.25vol.%, respectively. Consequently, the calculated results indicate the volume fraction of Ti2Co precipitates increases with the cold rolling ratio increasing, which is in accordance with the previous XRD results. This is because more dislocations introduced by larger cold-rolled amount process produce more nucleation sites for Ti2Co precipitates. Fig. 5 shows the DSC curves of 0CRA, 10CRA, 20CRA and 40CRA samples. As shown in Fig.5a, the 0CRA sample exhibits two-stage transformation during heating 8
ACCEPTED MANUSCRIPT and cooling process, which is different with the previous results of Ti-Ni-Nb alloys [21]. It is noteworthy to find that the additional endothermic peak and exothermic peak appear to be observed prior to the normal transformation peak of B2B19’. The
RI PT
extra phase transformation has much lower and broader peak than that of B2B19’. It can possibly be indexed as two-stage transformation B2RB19’. The occurrence of R-phase transformation can be seen to be considered a result of
SC
homogeneous nanoscale precipitation. Dispersed Ti2Co precipitates produce higher
M AN U
coherent strain field to cause strong resistance to martensite transformation. Nevertheless, the formation of R-phase is related to the smaller lattice deformation, which is less vulnerable to strong strain field. Consequently, the presence of coherent Ti2Co precipitates can separate the R-phase transformation from the martensite
TE D
transformation. Similar phenomenon has been reported in aged Ni-rich Ti-Ni alloys, suggesting that fine and uniform precipitation induced internal stress field leading to the occurrence of R-transformation [22-24]. The confirmation of R-phase
EP
transformation will be performed in our other studies later. The DSC curves of the
AC C
samples with cold rolled deformation are given in Figs. 5b-d. It can be clearly observed that there is only one step phase transformation during the heating and cooling process in the samples with cold rolled deformation. The introduction of dislocations via cold rolled deformation can relax much more misfit in the interface between matrix and precipitates, resulting in the disappearance of R-phase transformation. Besides, the martensite transformation peak shows lower, whereas transformation hysteresis gradually increases with the cold rolling increasing. 9
ACCEPTED MANUSCRIPT Fig. 6 shows stress-strain curves of 0CRA, 10CRA, 20CRA and 40CRA samples performed at the temperature of Ms+30oC. All the samples show a typical stress vs. strain response of a stress-induced martensite (SIM) state. As shown in Fig. 6, the
RI PT
critical stress for SIM can be determined by the tangent line method. It can be seen from Fig. 6 that the critical stress for SIM and yield stress gradually increase with increased cold rolling ratio. Furthermore, the 40CRA sample exhibits the highest
SC
critical stress for SIM of 320MPa and the highest yield stress of 1280MPa owing to
M AN U
high density of dislocations and precipitates. Meanwhile, the elongation ratio of samples slightly increases from 20% to 28% as the cold rolling ratio increases. The highest elongation ratio of the 40CRA sample results from a large population of dispersed β-Nb formed during prior cold rolled deformation, which can confirm the
TE D
previous findings by Zheng et. al [25].
From the stress-strain curves, another interesting point can be noticed: no flat stress plateau can be seen in the 0CRA and 10CRA samples and simultaneously the
EP
stress plateau gradually becomes flat with the increase of cold rolling ratio. The
AC C
40CRA sample exhibits a flat stress plateau. The flatness seems to indicate as if Lüders-like deformation from Ti-Ni-based alloys. In the case of the typical Ti-Ni-based alloys, the flat stress plateau means some favorably oriented martensites first form and then increase the localized internal stress to trigger neighboring martensite transformation, indicating self-propagation of SIM or detwining under constant load [1, 26, 27]. Yet, for the 0CRA and 10CRA samples, there is a large population of β-Nb phase with slight deformation distributed in the Ti-Ni grain 10
ACCEPTED MANUSCRIPT boundaries (Figs. 2a and b). The self-propagation of SIM can be constrained by β-Nb phase at the grain boundaries. Thus, the stress is not enough to induce neighboring martensite transformation so that the external load is required to assist the further
RI PT
process of SIM in this stage. However, β-Nb phase is distributed as narrow stripes or particles manner in the 40CRA sample (Fig. 2d). The self-propagation of SIM is much easier than that of the 0CRA and 10CRA samples. Basically constant external stress
M AN U
the SIM mechanism of typical Ti-Ni-based alloys.
SC
can be observed in such a stress plateau of the 40CRA sample, which is similar with
On the basis of the effect of the deformation temperature and strain on the transformation hysteresis and strain recovery ratio [6], the transformation hysteresis slightly increases with deformation temperature at the deformation strain of 16%. As
rapidly to
TE D
the deformation temperature near Ms+30oC, the transformation hysteresis increases 150oC and simultaneously recovery strain remains high. However, with
the deformation temperature and strain further increasing, the transformation
EP
hysteresis does not further increase, while the recovery properties significantly
AC C
deteriorate. The aforementioned analysis indicates the transformation hysteresis can be effectively realized at the characteristic temperature (Ms+30oC) and strain range ( 16%). Herein, the recovery properties and transformation characteristics of all the samples with different cold rolling ratios after 16% pre-strain at Ms+30oC are shown in Fig. 7 and Table 1. Fig. 7a shows the loading-unloading curves of 0CRA, 10CRA, 20CRA and 40CRA samples with 16% pre-strain at the temperature of Ms+30oC. The curving arrows represent shape recovery after heating above A'f . As shown in Fig. 7a, 11
ACCEPTED MANUSCRIPT it can be found that the total recovery strain increases with the increased cold rolled amount. The 40CRA sample shows the largest total recovery strain of 13.52% (including elastic recovery strain of 3.81% and shape memory recovery strain of
RI PT
9.71%), which exhibits 25% enhancement in the total recovery strain compared to the previous Ti-Ni-Nb alloys [6] owing to the strengthening effect of high-density nano-precipitation and dislocations. Fig. 7b shows the recovery stress dependent
SC
heating temperature of all the samples. The maximum recovery stress shows an
M AN U
increase with the cold rolling ratio increasing likewise. As known, the maximum recovery stress is related with the reversible recovery strain and yield strength. As for the 40CRA sample, a large population of precipitates and dislocations leads to a larger recovery strain and the higher yield strength. The highest recovery stress of 584MPa
TE D
can be obtained in the 40CRA sample, which displays over 50% improvement in contrast to the typical Ti-Ni-Nb alloys (386MPa) [28]. As shown in Fig. 7b and Table 1, the starting temperature of reverse transformation ( As' ) decreases with the increase
EP
of cold rolling ratio. Simultaneously, the slope of the stress-heating temperature
AC C
curves becomes large with the cold rolled amount increasing. This indicates that the dislocations introduced by cold rolled deformation can be increased to bring more preferential sites for the martensite nucleation, which facilitates the formation and growth of reverse martensite transformation. Thus, the cold rolled deformation has significant impact on the reverse transformation of SIM. Concurrently, the transformation hysteresis of the samples shows almost unchanged upon different cold rolling ratios, shown in Table 1. The transformation hysteresis of all the samples with 12
ACCEPTED MANUSCRIPT different cold rolling ratios maintains about 150oC. It can be concluded that the transformation hysteresis of the samples highly depends on the prior deformation strain instead of cold rolling ratio. As a result, Ti-Ni-Nb-Co alloys with wide
RI PT
transformation hysteresis and simultaneously better recovery properties can be achieved, which can be the most promising candidate as shape memory pipe couplings.
SC
4. Conclusion
M AN U
In our research, the effect of cold rolling ratio on the microstructure and recovery properties of an annealed Ti44.5Ni44.5Nb9Co2 alloy has been systematically investigated. The main results of this study are summarized as followed:
1. As cold rolling ratio increases, the β-Nb phase has changed from network
TE D
structure to stripes or particles and simultaneously the Ti2Co precipitates size slightly increases and the volume fraction of precipitates ranges from 0.47vol.% to 1.25vol.%. 2. Precipitates nucleated at the dislocations remarkably retard the recovery and
EP
recrystallization process, but the greater cold rolled deformation facilitates occurrence
AC C
of the recovery and recrystallization process. 3. The Ti44.5Ni44.5Nb9Co2 alloy with 40% cold rolled amount annealed at 650oC
can obtain the highest yield stress and the best recovery properties owing to high-density nanoscale precipitates and dislocations.
Acknowledgements This research was financed by National Natural Science Foundation of China (Grant Nos. 51731005, 51501049 and 51471060). 13
ACCEPTED MANUSCRIPT
References [1] K. Otsuka, X. Ren, Physical metallurgy of Ti–Ni-based shape memory alloys, Prog. Mater Sci. 50 (2005) 511-678.
RI PT
[2] J.M. Jani, M. Leary, A. Subic, M.A. Gibson, A review of shape memory alloy research, applications and opportunities, Mater. Des. 56 (2014) 1078-1113.
[3] T.W. Duerig, A.R. Pelton, Materials properties handbook: Ti-Ni shape memory
SC
alloys, 683 (1994) 1035-1048.
M AN U
[4] R. Basu, M. Eskandari, L. Upadhayay, M.A. Mohtadi-Bonab, J.A. Szpunar, A systematic investigation on the role of microstructure on phase transformation behavior in Ni–Ti–Fe shape memory alloys, J. Alloys Compd. 645 (2015) 213-222.
TE D
[5] J. Frenzel, J. Pfetzing, K. Neuking, G. Eggeler, On the influence of thermomechanical treatments on the microstructure and phase transformation
635-638.
EP
behavior of Ni–Ti–Fe shape memory alloys, Mater. Sci. Eng. A s481–482 (2008)
AC C
[6] C.S. Zhang, L.C. Zhao, T.W. Duerig, C.M. Wayman, Effects of deformation on the transformation hysteresis and shape memory effect in a Ni47Ti44Nb9, Scripta Metallurgica Et Materiala 24 (1990) 1807-1812.
[7] M. Piao, S. Miyazaki, K. Otsuka, Characteristics of deformation and transformation in Ti44Ni47Nb9 shape memory alloy, Mater. Trans. 33 (1992) 346-353. [8] Y.X. Tong, P.C. Jiang, F. Chen, B. Tian, L. Li, Y.F. Zheng, D.V. Gunderov, R.Z. 14
ACCEPTED MANUSCRIPT Valiev, Microstructure and martensitic transformation of an ultrafine-grained TiNiNb shape memory alloy processed by equal channel angular pressing, Intermetallics 49 (2014) 81-86.
shape memory alloys with low Nb content
RI PT
[9] C. Ying, H.C. Jiang, L.J. Rong, X. Li, X.Q. Zhao, Mechanical behavior in NiTiNb Intermetallics 19 (2011) 217-220.
[10] K. Uchida, N. Shigenaka, T. Sakuma, Y. Sutou, K. Yamauchi, Effect of Nb
SC
Content on Martensitic Transformation Temperatures and Mechanical Properties
M AN U
of Ti-Ni-Nb Shape Memory Alloys for Pipe Joint Applications, Mater. Trans. 48 (2007) 445-450.
[11] Y. Wang, Y.F. Zheng, Y.X. Tong, B. Tian, F. Chen, L. Li, Microstructure and martensitic transformation of TiNiNbB shape memory alloys, Intermetallics 64
TE D
(2015) 32-36.
[12] C. Wang, Q. Gao, Y. Yuan, H. Zhang, J. Zhang, Q. Wang, F. Qu, Microstructure evolutions of Ni-Ti-Nb-Al alloys with different Al addition, J. Alloys Compd. 695
EP
(2017) 2923-2929.
AC C
[13] V.Y. Abramov, N.M. Aleksandrova, D.V. Borovkov, I.Y. Khmelevskaya, A.V. Korotitskiy, S.Y. Makushev, N.A. Polyakova, N.N. Popov, S.D. Prokoshkin, Martensitic transformations and functional properties of thermally and thermomechanically treated Ti–Ni–Nb-based alloys, Mater. Sci. Eng. A 438 (2006) 553-557. [14] M.I. Khan, H.Y. Kim, F.A. Khalid, S. Miyazaki, Effect of cold rolling ratio on the nanoscale precipitation behavior of TiNiPdCu based high temperature shape 15
ACCEPTED MANUSCRIPT memory alloys, J. Alloys Compd. 599 (2014) 212-218. [15] S.U. Rehman, M. Khan, A.N. Khan, M.I. Khan, L. Ali, S.H.I. Jaffery, Effect of precipitation hardening and thermomechanical training on microstructure and
RI PT
shape memory properties of Ti50Ni15Pd25Cu10 high temperature shape memory alloys, J. Alloys Compd. 616 (2014) 275-283.
[16] H.E. Karaca, S.M. Saghaian, G. Ded, H. Tobe, B. Basaran, H.J. Maier, R.D.
SC
Noebe, Y.I. Chumlyakov, Effects of nanoprecipitation on the shape memory and
M AN U
material properties of an Ni-rich NiTiHf high temperature shape memory alloy, Acta Mater. 61 (2013) 7422-7431.
[17] B. Cui, J. Yao, Y. Wu, X. Zhang, F.L. Wang, J.H. Sui, W. Cai, Precipitation behavior and mechanical properties of Ti-Ni-Nb-Co alloys, Intermetallics 95
TE D
(2018) 40-47.
[18] W.D. Callister, Fundamentals of Materials Science and Engineering: An Integrated Approach, 2nd Edition, John Wiley and Sons Ltd, 2005.
EP
[19] E.E. Underwood, Stereology, or the quantitative evaluation of microstructures,
AC C
Journal of Microscopy 89 (1969) 161-180. [20] D.B. Williams, C.B. Carter, Transmission Electron Microscopy: A Textbook for Materials Science, Springer, New York, USA, 2009.
[21] Y. Zheng, W. Cai, Y. Wang, Y. Luo, L. Zhao, Effects of Heat Treatment on the Transformation Temperature and the Microstructure of Ni-Ti-Nb Shape Memory Alloy, J. Mater. Sci. Technol. 14 (1998) 37-40. [22] G. Fan, W. Chen, S. Yang, J. Zhu, X. Ren, K. Otsuka, Origin of abnormal 16
ACCEPTED MANUSCRIPT multi-stage martensitic transformation behavior in aged Ni-rich Ti–Ni shape memory alloys, Acta Mater. 52 (2004) 4351-4362. [23] L.J. Chiang, C.H. Li, Y.F. Hsu, W.H. Wang, Age-induced four-stage
RI PT
transformation in Ni-rich NiTi shape memory alloys, J. Alloys Compd. 458 (2008) 231-237.
[24] D. Xue, Y. Zhou, X. Ren, The effect of aging on the B2-R transformation
SC
behaviors in Ti-51at%Ni alloy, Intermetallics 19 (2011) 1752-1758.
M AN U
[25] Y. Zheng, W. Cai, Y. Luo, Effects of heat treatment on tensile properties and shape memory effects of Ni-Ti-Nb alloys, T. Nonferr. Metal. Soc. (1997) 93-97. [26] L. Orgéas, D. Favier, Stress-induced martensitic transformation of a NiTi alloy in isothermal shear, tension and compression, Acta Mater. 46 (1998) 5579-5591.
TE D
[27] Y. Liu, Detwinning process and its anisotropy in shape memory alloys, Proc Spie 4234 (2001) 82-93.
[28] W. Cai, C. Zhang, L. Zhao, Recovery Stress in a Ni-Ti-Nb Shape Memory Alloy
AC C
EP
with Wide Transformation Hysteresis, J. Mate. Sci. Technol. 17 (1994) 27-30.
17
ACCEPTED MANUSCRIPT
Figures and Table Captions Fig. 1 a) Room temperature XRD profiles of all the samples with different cold rolling ratios, b) the zoom-in XRD patterns between 40o 48o.
RI PT
Fig. 2 Back-scattered SEM images of all the samples with different cold rolling ratios; a) 0CRA, b) 10CRA, c) 20CRA, d) 40CRA; cold rolling direction (RD) and transverse direction (TD) are marked. Fig. 3 TEM bright field images of all the samples with different cold rolling ratios and corresponding SAED pattern; a) 0CRA, b) 10CRA, c) 20CRA, d) 40CRA.
M AN U
SC
Fig. 4 High magnification TEM bright field images of all the samples with different cold rolling ratios, a) 0CRA, b) 10CRA, c) 20CRA, d) 40CRA; The statistical precipitates size distributions of all the samples are in the inserts. Fig. 5 DSC curves of all the samples with different cold rolling ratios; a) 0CRA, b) 10CRA, c) 20CRA, d) 40CRA. Fig. 6 Stress-strain curves of all the samples with different cold rolling ratios preformed at Ms+30oC; a) 0CRA, b) 10CRA, c) 20CRA, d) 40CRA.
TE D
Fig. 7 Recovery properties of all the samples with different cold rolling ratios after 16% pre-strain at Ms+30oC, (a) the loading-unloading curves, (b) recovery stress curves dependent heating temperature.
AC C
EP
Table 1 The recovery properties and transformation characteristics of all the samples with different cold rolling ratios after 16% pre-strain at Ms+30oC.
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
Figure 1
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
Figure 2
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
Figure 3
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
Figure 4
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
Figure 5
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
Figure 6
AC C
EP
TE D
M AN U
SC
Figure 7
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Table 1 Recovery Strain (%)
Recovery Stress (MPa)
(oC)
(oC)
Transformation Hysteresis ( - ) (oC)
0CRA 10CRA 20CRA 40CRA
11.65 12.24 12.80 13.52
464 492 523 584
-87 -98 -105 -111
60 54 42 35
147 152 149 146
AC C
EP
TE D
M AN U
SC
RI PT
Samples
ACCEPTED MANUSCRIPT Prior cold rolled deformation plays important role on the microstructure of alloy. High-density precipitates and dislocations can improve recovery properties. The recovery strain of the alloy annealed at 650oC after 16% strain shows
RI PT
13.25%. The recovery stress of the alloy annealed at 650oC after 16% strain shows 584MPa.
SC
The alloy possesses wide transformation hysteresis and better recovery
AC C
EP
TE D
M AN U
properties.