Detwinning behavior of Mg–3Al–1Zn alloy at elevated temperatures

Detwinning behavior of Mg–3Al–1Zn alloy at elevated temperatures

Materials Science & Engineering A 617 (2014) 24–30 Contents lists available at ScienceDirect Materials Science & Engineering A journal homepage: www...

3MB Sizes 0 Downloads 28 Views

Materials Science & Engineering A 617 (2014) 24–30

Contents lists available at ScienceDirect

Materials Science & Engineering A journal homepage: www.elsevier.com/locate/msea

Detwinning behavior of Mg–3Al–1Zn alloy at elevated temperatures Huihui Yu, Yunchang Xin n, Hua Zhou, Rui Hong, Lingyu Zhao, Qing Liu nn School of Materials Science and Engineering, Chongqing University, Chongqing 400044, China

art ic l e i nf o

a b s t r a c t

Article history: Received 29 July 2014 Received in revised form 13 August 2014 Accepted 15 August 2014 Available online 24 August 2014

Detwinning behavior of a strongly textured magnesium alloy AZ31 at different temperatures was investigated, with a focus on the effect of temperature on twin boundaries (TBs) migration. A hot extruded Mg alloy AZ31 rod was subjected to a 3% pre-compression along the extrusion direction (ED) to generate f1012g twins. Subsequent tensions along the ED at temperatures of 30 1C, 100 1C, 150 1C, 200 1C and 250 1C, respectively, were carried out to initiate detwinning. Mechanical behavior and microstructure evolution during detwinning deformation were examined. Our results show that the activation stress for f1012g TBs migration decreases with the increasing of temperature from 30 1C to 250 1C. Even at 250 1C, detwinning rather than non-basal slips dominates tension along the ED of pre-strained specimens. All strain hardening curves of detwinning deformation at temperatures of 30–250 1C contain three stages that often appear in a f1012g twinning predominant deformation. The length of Stage II is predominantly related to f1012g twin volume fraction. It is found that the peak hardening rate in Stage II decreases with increased temperature. & 2014 Elsevier B.V. All rights reserved.

Keywords: Mg alloy Detwinning Temperature Twin boundary migration

1. Introduction Strong directional anisotropy and low ductility at room temperature have hampered general adoption of Mg alloys as structural materials [1]. The activity of different deformation modes is highly dependent on temperature [2]. For example, the critical resolved shear stresses (CRSSs) of slips on basal plane and extension twinning are about 0.5–0.7 MPa [3] and 2.0–2.8 MPa, respectively [4], while the CRSS ratio of basal slip, prismatic slip, first order pyramidal slip, second order pyramidal slip in a single crystal magnesium is reported to be about 1:38:50:100 [5]. Therefore, basal slip and f1012g twinning constitute the main deformation modes at room temperature. The onset of non-basal slips in a large amount generally takes its appearance at temperature above 200 1C. Deformation behavior of Mg alloys is also highly dependent on texture [6–9]. Strong texture often generates intensive mechanical anisotropy. For example, f1012g twinning dominates compression along the ED of an extruded Mg alloy AZ31, while prismatic slip of 〈a〉 dislocations is the main deformation mode during tension along the ED [10]. An increase in tensile or compressive load will even favor contraction twinning [1]. Compression or tension along the rolling direction, transverse direction and normal direction of a hot-rolled plate often induce quite different deformation behavior, too [11].

n

Corresponding author. Tel./fax: þ 86 23 65106407. Corresponding author. Tel./fax: þ 86 23 65111295. E-mail addresses: [email protected] (Y. Xin), [email protected] (Q. Liu).

nn

http://dx.doi.org/10.1016/j.msea.2014.08.034 0921-5093/& 2014 Elsevier B.V. All rights reserved.

For Mg alloys that contain pre-strained twins, detwinning is the main deformation mechanism under specific loading conditions. For example, detwinning following twinning can occur under strain-path changed reloading [1,4], cyclic loading [12–14] and unloading [12]. As noted by many publications, detwinning is a TBs migration process, i.e. the pre-existing twins shrink and transform back into the matrix orientation, leading to increasing of matrix volume fraction [4]. As detwinning does not need nucleation, the activation stress for detwinning is suspected to be much lower than that for twinning nucleation. Increasing evidences have demonstrated that the twinning–detwinning process greatly affects mechanical behavior of Mg alloys. The low activation stress of detwinning often results in quite low yield stress in a detwinning predominant deformation. A twinning–detwinning process during cyclic loading often generates an asymmetric sigmoidal-shaped hysteresis loop in stress–strain curves [10,15]. This asymmetry is more pronounced with a higher strain or a higher stress amplitude [14,16]. Similar to twinning, detwinning also reorients crystallographic orientation of the twinned region by about 86.31 [17]. Previously, there are intensive studies about detwinning behavior of Mg alloys at room temperature [1,4,12–14,18,19]. Detwinning at elevated temperature is seldom investigated and, therefore, not well understood. It is well known that mobility of high angle grain boundaries with arrays of dislocations rises with increasing of temperature. Compared to high angle grain boundaries, a coherent TB often possesses a lower interfacial energy and, thus, a lower mobility at elevated temperature. However, how the elevated temperature influences the activation of TBs migration is

H. Yu et al. / Materials Science & Engineering A 617 (2014) 24–30

25

still not well elucidated. Recently, Nie et al. [20] found that solute atoms would segregate to the coherent TBs of Mg alloys during annealing treatment and pin TBs migration, which makes detwinning behavior at elevated temperature more complex. As stated above, CRSSs of non-basal slips of Mg alloys decrease rapidly with increasing of deformation temperature [15,21,22]. How does the elevated temperature affect competition between non-basal slips and detwinning? Unfortunately, no publications have addressed this issue yet. Recently, the applications of pre-strained twins to harden Mg alloys attract increasing attentions [23]. As detwinning is one of the main deformation mechanisms of pre-twinned samples, a deep understanding about detwinning behavior at elevated temperature is of great importance. Such a result (i.e. temperature dependent CRSS data) can also be used to improve and validate crystal plasticity constitutive models for slip and twinning in metallic crystals, for example as described theoretically in [24,25]. In this paper, detwinning behavior of a strongly textured Mg alloy AZ31 at temperatures of 30–250 1C was investigated, with a focus on the effect of temperature on TBs migration. The influence of temperature on the activation stress of detwinning and the competition between detwinning and non-basal slips were systematically addressed.

2. Experiments and methods 2.1. Sample preparation and mechanical tests The as-received material was a commercial Mg alloy AZ31 in the as-cast condition. The as-received ingots were cut into cylinders of 80 mm in diameter, which were homogenized at 400 1C for 5 h in a muffle furnace and air cooled. The as-homogenized cylinders were kept at 400 1C for 40 min followed by extrusion at 400 1C immediately using an extrusion ratio of 10:1. The final extruded rod has a diameter of about 25 mm. In order to acquire a fully recyrstallized microstructure, the extruded rod was annealed at 450 1C for 4.5 h. The resulting material has a fully recrystallized grain structure with an average grain size of about 26 μm (Fig. 1a) and a typical extrusion texture with basal poles largely normal to the ED (Fig. 1b) and a preferred distribution of prismatic planes. Cylinders (70 mm in height and 25 mm in diameter) cut from the extruded rod were pre-compressed along the ED by 3% to generate f1012g twins. The pre-compressed cylinders (the designated PC sample) were subsequently machined into dog-bone shaped specimens for tension test as shown in Fig. 1c. Tension tests at temperatures of 30 1C, 100 1C, 150 1C, 200 1C and 250 1C, respectively, were performed using a constant strain rate of  2.2  10  3 s  1. Recently, Nie et al. [20] have reported that periodic segregation of solute atoms at TBs of Mg alloys will appear during annealing treatment. Xin et al. [26] further confirm that an annealing at 170 1C for 15 min can induce a full segregation of Al and Zn atoms at TBs of Mg alloy AZ31, leading to annealing hardening in detwinning deformation. In current study, the tension tests at high temperatures (100 1C, 150 1C, 200 1C and 250 1C) generally started after the specimens were kept at testing temperature for 15 min. It is suspected that solute segregation at TBs will occur during heating of specimens. However, no solute segregation would appear during tension tests at room temperature. To eliminate the influence of solute segregation at TBs on detwinning at different temperatures, prior to tensile tests, all specimens were annealed at 250 1C for 1.5 h to achieve a full solute segregation at TBs without damaging the twin structure. As a comparison, tension tests along the ED of samples without pre-straining (the designated AR sample) were also carried out using the same conditions. All mechanical tests were repeated three times.

Fig. 1. (a) Optical micrograph and (b) relevant pole figures of the extruded AZ31 with annealing at 450 1C for 4.5 h. (c) A schematic diagram showing preparation of the specimens with pre-strained twins (the designated PC samples) for tension test along the ED. ED and TD refer to the extrusion direction, and the tangential direction, respectively.

2.2. Microstructure and texture For microstructural examination in a optical microscope, the specimens were carefully ground with a series of SiC sand papers (800 grit and 1000 grit) and chemically etched in an acetic picral solution (2 ml acetic acid þ1 g picric acid þ2 ml H2O þ16 ml ethanol). Pole figures of samples were determined using an X-ray diffraction (XRD, Rigaku D/max-2500PC) analysis. Microstructures and crystallographic orientation of samples were further studied using an electron back-scattered diffraction (EBSD) technique. EBSD mapping using a step size of 1 μm was performed on an FEI NOVA400 scanning electron microscope (SEM) equipped with an HKL-EBSD system. Samples for EBSD mapping were mechanical ground followed by electro-chemical polishing in a AC2 electrolyte solution at 20 V for 90 s.

26

H. Yu et al. / Materials Science & Engineering A 617 (2014) 24–30

3. Results 3.1. Mechanical behavior The true stress–strain curves obtained from tensile tests along the ED are given in Fig. 2. The curves of PC samples at all testing temperature are sigmoidal shaped, while all AR samples exhibit parabolic shaped curves. Mechanical properties (ultimate tensile strength (UTS), yield strength (YS) and elongation to failure (ε)) derived from true stress–strain curves are given in Tables 1 and 2. YSs of both PC and AR samples decrease with the rising of deformation temperature. YSs of PC samples drop from 118 MPa at 30 1C to 86 MPa at 250 1C, compared to that of AR samples dropping from 160 MPa at 30 1C to 93 MPa at 250 1C. At temperatures below 200 1C, all YSs of PC samples are much lower than that of AR samples. However, YS of PC sample at 250 1C is only slightly lower than that of AR sample. Strain hardening rate (the slope of true stress–strain curve) as a function of strain is calculated and is given in Fig. 3. The strain hardening rate was calculated by differentiation of true strain– strain curves. All curves of PC samples have similar three stages (as denoted by arrows in Fig. 3a): a fast drop of hardening rate in Stage I, a quick increasing of hardening rate to a maximum in Stage II, a decreasing in hardening rate again in Stage III. However, all AR samples have a continuously decreasing straining hardening rate (Fig. 3b) and a further low reduction tendency of hardening associated with strain hardening of plastic deformation. It can be noticed that the lengths of Stage II at different deformation temperatures are similar to each other, while the peak hardening rate in Stage II drops with the rising of deformation temperature.

3.2. Microstructure and texture Fig. 4 shows inverse pole figure map and corresponding boundary misorientation map of sample after a 3% pre-compression along the ED and subsequent annealing at 250 1C for 1.5 h. Obviously, the twin structure is well kept after annealing at 250 1C for 1.5 h. Boundary misorientation map in Fig. 4b and the relevant twin area fraction in Table 3 clearly show that the red lamellas in Fig. 4a are mainly f1012g twins with an area fraction of about 21.8%. A peak around 861 in Fig. 4c further confirms the presence of f1012g twins in a large number. Contraction twins (f1011g or f1013g) and double twins (f1011g–f1012g or f1013g–f1012g) are not detected.

Pole figures of PC samples after tension along the ED at 30 1C and 250 1C, respectively, are present in Fig. 5. Fig. 5a shows where the tension tests are terminated. As shown in Fig. 5b, after a 3% pre-compression along the ED, a part of basal poles rotate from the direction perpendicular to the ED to that parallel to the ED. This inclination of basal poles is extensively reported to be a consequence of f1012g twinning that rotates the basal poles by about 861 [27,28]. After tension along the ED by the same plastic strain, the basal poles close to the ED in both PC samples tensioned at 30 1C and 250 1C, respectively, nearly disappear. Microstructure of PC samples terminated at the locations of 1 and 2 in Fig. 5a is further examined by EBSD and the results are shown in Fig. 6. In PC sample tensioned at 30 1C, most twin lamellas disappear. Similar phenomenon takes place in PC sample that were tensioned at 250 1C. The twin area fractions in Fig. 6a and b are around 3.9% and 5.5% respectively (see Table 3), respectively. As extensively reported in many publications [1,15,29], detwinning dominates the room-temperature tension along the ED of an extruded Mg alloy that is pre-compressed along the ED. In current study, after a tension along the ED by the same plastic strain, PC sample at 250 1C has similar twin area fraction to that of PC sample at 30 1C. It can be inferred that detwinning of f1012g twins is the predominant deformation mode during tension of PC sample at 250 1C.

Table 1 Yield stress (YS), ultimate tensile stress (UTS) and elongation to fracture (ε) of PC samples during tension along the ED at temperatures from 30 to 250 1C. Temperature (1C)

30

100

150

20

250

YS (MPa) UTS (MPa) ε (%)

118 313 20.5

112 290 24.9

99 252 29.5

91 198 38.0

86 149 42.8

Table 2 Yield stress (YS), ultimate tensile stress (UTS) and elongation to fracture (ε) of AR samples during tension along the ED at temperatures from 30 to 250 1C. Temperature (1C)

30

100

150

200

250

YS (MPa) UTS (MPa) ε (%)

160 304 19.7

147 291 26.0

148 258 25.4

108 202 34.5

93 158 42.5

Fig. 2. The true stress–strain curves of (a) PC samples and (b) AR samples under tension along the ED at temperatures from 30 to 250 1C.

H. Yu et al. / Materials Science & Engineering A 617 (2014) 24–30

27

Fig. 3. The strain hardening rates as a function of true strain of (a) PC samples and (b) AR samples under tension along the ED at temperatures from 30 to 250 1C.

Fig. 4. (a) Inverse pole figure map, (b) corresponding boundary misorientation map and (c) misorientation angle distribution of Mg alloy AZ31 rod with a 3% precompression along the ED. The red boundaries in (b) and the peak around 861 in (c) show that a pre-compression along the ED generates a large number of f1012g twins. RD – radial direction, TD – tangential direction. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Table 3 Twin area fractions of PC samples and the PC samples after 2.5% tension along the ED at 30 1C and at 250 1C, respectively.

Twin area fraction (%)

PC sample

Tension at 30 1C

Tension at 250 1C

21.8

3.9

5.5

4. Discussion 4.1. The effect of temperature on TBs migration f1012g Deformation TBs, special high-angle boundaries which possess significantly lower energies than the random boundaries, consist of sequential f1012g coherent twin boundaries (CTBs) and parallel Basal-Prismatic planes serrations (BPs) [30]. Detwinning, a process of TBs migration, involves the glide of twinning dislocations (TDs) on twinning planes. These defects can be described by a Burgers vector b and a step of height h that encompasses more than one crystallographic plane. As discussed in a previous publication [31], the elementary TD of f1012g twins in hcp structured metals, generally has a Burgers vector of btw ¼ λ[1011] (λ ¼c/a ratio). This can be schematically illustrated in Fig. 7. Gliding of these individual partials on many consecutive planes creates a

twinned or detwinned domain within which the previously hcp lattice is rearranged, sheared and reoriented significantly, e.g. 86.31 for Mg. Regarding the coexistence of CTBs and BPs in TBs, the possible migration mechanisms of TBs include a migration of CTBs via the glide–shuffle mechanism of TDs on the twinning plane, a subsequent formation of stacking faults (SFs) and a final migration of BP interfaces via climb of the interface dislocations (IDs). The IDs is equal to the difference in the inter-planar spacing resulting from converting a basal plane into a prismatic plane [32]. Some researchers consider that TBs mobility in fcc metals enhances with elevated temperature [33–35]. However, no direct experiment evidences have been reported. To disclose the influence of temperature on activation of detwinning, it is essential to judge the main deformation mechanisms during tension along the ED of PC samples at different temperatures. The possible deformation modes include prismatic 〈a〉 slip in matrix and detwinning of f1012g twins. Generally, tension along the ED of AR sample is a prismatic 〈a〉 slip predominant deformation. Thus, the activities of prismatic 〈a〉 slip and detwinning in PC samples can be deduced from the YSs of PC and AR samples at the same testing temperature. Both YSs of PC and AR samples as a function of temperature are plotted and shown in Fig. 8. Obviously, at temperature below 200 1C, YSs of AR samples are much higher than that of PC samples. This implies that, compared to prismatic 〈a〉 slip, detwinning is preferred to occur at temperature below 200 1C. The EBSD

28

H. Yu et al. / Materials Science & Engineering A 617 (2014) 24–30

Fig. 5. (a) Tension stress–strain curves of PC samples at 30 1C and 250 1C. The arrows in (a) showed the locations (1 and 2) where the tension along the ED were terminated. Note that the tensions along the ED at 30 1C and 250 1C were terminated at the same plastic strain. (b) Pole figures of sample with a 3% pre-compression along the ED. (c, d) The pole figures of PC samples after tension along the ED to locations 1 and 2 in (a), respectively.

Fig. 6. Inverse pole figure maps and the corresponding boundary misorientation maps of PC samples after tension along the ED at (a, b) 30 1C and (c, d) 250 1C by the same plastic strain (locations 1 and 2 in Fig. 5a). RD – radial direction, TD – tangential direction.

result in Fig. 6 also demonstrates that detwinning is the predominant deformation of PC sample even at 250 1C. YSs of PC sample drop from 118 MPa at room temperature to about 86 MPa at 250 1C. This clearly shows that raising deformation temperature significantly reduces the activation stress of f1012g TBs migration. That is, the CRSS of TDs in f1012g TBs decreases with the increasing of temperature. We therefore believe that this is a proof that, at 250 1C, the stress to activate gilding of TDs is lower than that to initiate prismatic 〈a〉 slip. According to the publication from Nie et al. [20], solute segregation at twin boundary will occur during

tension test of PC sample at 150 and 300 1C. The pinning of TDs by solute atoms generally leads to a higher activation stress for TDs gilding. Therefore, the intrinsic CRSS to start gilding of TDs should be lower. 4.2. Strain hardening in detwinning dominated deformation As seen in Fig. 3, strain hardening curves of detwinning dominated deformations at different temperatures all have similar three stages to that of deformation dominated by f1012g twinning.

H. Yu et al. / Materials Science & Engineering A 617 (2014) 24–30

29

Fig. 7. A schematic diagram showing structure of a f1012g twin in Mg alloy AZ31. The Al or Zn atoms at twin boundaries result from solute segregation by annealing treatment and will pin twin boundary migration.

hardening rate at the end of detwinning. Tension of PC samples at elevated temperature leads to a pronounced drop in CRSS to activate non-basal slips, as a consequence, the strain hardening rate at the end of detwinning (corresponds to the peak hardening rate in Stage II) decreases, too.

5. Conclusion Detwinning behavior of a strongly textured magnesium alloy AZ31 at temperatures of 30 1C, 100 1C, 150 1C, 200 1C and 250 1C was investigated. Mechanical behavior and microstructure evolution during detwinning deformation were examined. Several conclusions can be reached as follows:

Fig. 8. Tension yield stresses of PC and AR samples as a function of temperature.

Presently, it is accepted that three main mechanisms are suspected to contribute to strain hardening of a f1012g twinning predominant deformation [36]: (i) a Hall–Petch-like effect that arises from sub-division of grains by twin lamellas, (ii) a glissile-to-sessile transformation of dislocations already present in the region experiencing the twinning shear transformation, and (iii) changes in crystal lattice orientations to harder orientation. As detwinning transforms the twins into the matrix, the matrix can be defined as a hard orientation structure for detwinning. Therefore, similar to twinning, detwinning also rotate the twinned regions to a harder orientation. Therefore, the mechanism (iii) also works in strain hardening of detwinning dominated deformation. However, compared to twinning that can undergo by both twin nucleation and twin growth, detwinning only proceeds by TBs migration. TBs migration cannot further subdivide grains. It is therefore very reasonable to postulate that the strain hardening during detwinning is unlikely to be a consequence of the mechanism (i). In strain hardening curve of a f1012g twinning controlled deformation, the length of Stage II is found to be closely related to the volume fraction of grains favorable for twinning [37]. In the present study, it is found that the length of Stage II in detwinning controlled deformation is dependent on the volume fraction of f1012g twins favorable for detwinning. As seen in Fig. 3a, the peak hardening rate in Stage II drops with elevation of temperature. Regarding the peak hardening rate, Wu et al. consider that the consequent hard orientation after a complete detwinning require the activation of non-basal slips and the hardening rate can be increased owing to the exhaustion of detwinning [1]. Therefore, the activation stress of non-basal slips will determine the

(1) The activation stress for f1012g TBs migration drops with the elevated deformation temperatures. Even at 250 1C, detwinning rather than non-basal slips is the predominant deformation mode during tension along the ED. (2) All strain hardening curves of detwinning deformations at temperatures of 30–250 1C consist of three stages that generally appear in a f1012g twinning dominated deformation. The length of Stage II is highly dependent on the f1012g twin volume fraction and the peak hardening rate in Stage II drops with increased temperature.

Acknowledgment The current study is co-supported by the National Natural Science Foundation of China (51371203, 51101175 and 51131009) and National Key Basic Research Program of China (2013CB632205). References [1] L. Wu, A. Jain, D.W. Brown, G.M. Stoica, S.R. Agnew, B. Clausen, D.E. Fielden, P.K. Liaw, Acta Mater. 56 (2008) 688–695. [2] M. Yuasa, K. Masunaga, T. Yoshida, M. Mabuchi, Y. Chino, Acta Mater. 61 (2013) 4714–4725. [3] B. Raeisinia, S.R. Agnew, Scr. Mater. 63 (2010) 731–736. [4] C.J. Geng, B.L. Wu, X.H. Du, Y.D. Wang, Y.D. Zhang, F. Wagner, C. Esling, Mater. Sci. Eng. A 559 (2013) 307–313. [5] G. Bajargan, G. Singh, D. Sivakumar, U. Ramamurty, Mater. Sci. Eng. A 579 (2013) 26–34. [6] A. Fernández, M.T. Pérez-Prado, Y. Wei, A. Jérusalem, Int. J. Plast. 27 (2011) 1739–1757. [7] R. Gehrmann, M.M. Frommert, G. Gottstein, Mater. Sci. Eng. A 395 (2005) 338–349. [8] T. Al-Samman, G. Gottstein, Mater. Sci. Eng. A 488 (2008) 406–414. [9] I. Ulacia, N.V. Dudamell, F. Gálvez, S. Yi, M.T. Pérez-Prado, I. Hurtado, Acta Mater. 58 (2010) 2988–2998. [10] L. Wu, S.R. Agnew, D.W. Brown, G.M. Stoica, B. Clausen, A. Jain, D.E. Fielden, P.K. Liaw, Acta Mater. 56 (2008) 3699–3707.

30

H. Yu et al. / Materials Science & Engineering A 617 (2014) 24–30

[11] R. Bhattacharya, B.P. Wynne, Mater. Sci. Technol. 27 (2011) 461–477. [12] W. Wu, S.Y. Lee, A.M. Paradowska, Y.F. Gao, P.K. Liaw, Mater. Sci. Eng. A 556 (2012) 278–286. [13] C.F. Gua, L.S. To0 th, Scr. Mater. 67 (2012) 673–676. [14] J.B. Jordon, J.B. Gibson, M.F. Horstemeyer, H.E. Kadiri, J.C. Baird, A.A. Luo, Mater. Sci. Eng. A 528 (2011) 6860–6871. [15] F.A. Mirza, D.L. Chen, D.J. Li, X.Q. Zeng, Mater. Sci. Eng. A 575 (2013) 65–73. [16] L. Wu, S.R. Agnew, Y. Ren, D.W. Brown, B. Clausen, G.M. Stoica, H.R. Wenk, P.K. Liaw, Mater. Sci. Eng. A 527 (2010) 7057–7067. [17] S.H. Park, J.H. Lee, Y.H. Huh, S.G. Hong, Scr. Mater. 69 (2013) 797–800. [18] D. Sarker, D.L. Chen, Scr. Mater. 67 (2012) 165–168. [19] V.S. Boĭko, R.I.G. Garber, A.M. Kosevich, Reversible Crystal Plasticity, American Institute of Physics, New York, 1994. [20] J.F. Nie, Y.M. Zhu, J.Z. Liu, X.Y. Fang, Science 340 (2013) 957–960. [21] E. Karimi, A. Zarei-Hanzaki, M.H. Pishbin, H.R. Abedi, P. Changizian, Mater. Des. 49 (2013) 173–180. [22] Q. Ma, B. Li, A.L. Oppedal, W.R. Whittington, S.J. Horstemeyer, E.B. Marin, P.T. Wang, M.F. Horstemeyer, Mater. Sci. Eng. A 559 (2013) 314–318. [23] Y.C. Xin, M.Y. Wang, Z. Zeng, M.G. Nie, Q. Liu, Scr. Mater. 66 (2012) 25–28. [24] A. Staroselsky, L. Anand, Int. J. Plast. 19 (2003) 1843–1864.

[25] J.D. Clayton, Nonlinear Mechanics of Crystals, Springer, Netherlands, Dordrecht, 2011. [26] Y. Xin, X. Zhou, H. Chen, J.F. Nie, H. Zhang, Y. Zhang, Q. Liu, Mater. Sci. Eng. A 594 (2014) 287–291. [27] Y.C. Xin, M.Y. Wang, Z. Zeng, G.J. Huang, Q. Liu, Scr. Mater. 64 (2011) 986–989. [28] S.H. Park, S.G. Hong, C.S. Lee, Mater. Sci. Eng. A 578 (2013) 271–276. [29] C. Chen, T.M. Liu, C.L. Lv, L.W. Lu, D.Z. Luo, Mater. Sci. Eng. A 539 (2012) 223–229. [30] J. Wang, I.J. Beyerlein, J.P. Hirth, C.N. Tomé, Acta Mater. 59 (2011) 3990–4001. [31] J. Wang, I.J. Beyerlein, C.N. Tomé, Scr. Mater. 63 (2010) 741–746. [32] J. Tu, X. Zhang, J. Wang, Q. Sun, Q. Liu, C.N. Tome, Appl. Phys. Lett. 103 (2013) 051903. [33] Y.S. Wang, W.H. Lee, S.C. Chang, J.N. Nian, Y.L. Wang, Thin Solid Films 544 (2013) 157–161. [34] T. Sinha, Y. Kulkarni, J. Appl. Phys. 109 (2011) 114315. [35] S. Qu, G. Wang, H. Zhou, Z. Huang, Comput. Mater. Sci. 50 (2011) 1567–1570. [36] M. Knezevic, A. Levinson, R. Harris, R.K. Mishra, R.D. Doherty, S.R. Kalidindi, Acta Mater. 58 (2010) 6230–6242. [37] B. Wang, R. Xin, G. Huang, Q. Liu, Mater. Sci. Eng. A 534 (2012) 588–593.