Twinning characterization of fiber-textured AZ31B magnesium alloy during tensile deformation

Author’s Accepted Manuscript Twinning characterization of fiber-textured AZ31B magnesium alloy during tensile deformation Linghui Song, Baolin Wu, Li Zhang, Xinghao Du, Yinong Wang, Claude Esling www.elsevier.com/locate/msea

PII: DOI: Reference:

S0921-5093(17)31381-3 https://doi.org/10.1016/j.msea.2017.10.055 MSA35658

To appear in: Materials Science & Engineering A Received date: 13 July 2017 Revised date: 18 October 2017 Accepted date: 19 October 2017 Cite this article as: Linghui Song, Baolin Wu, Li Zhang, Xinghao Du, Yinong Wang and Claude Esling, Twinning characterization of fiber-textured AZ31B magnesium alloy during tensile deformation, Materials Science & Engineering A, https://doi.org/10.1016/j.msea.2017.10.055 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 galley proof before it is published in its final citable 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.

Twinning characterization of fiber-textured AZ31B magnesium alloy during tensile deformation Linghui Song1, Baolin Wu2,, Li Zhang2, Xinghao Du2, Yinong Wang1, Claude Esling3 1

School of Materials Science and Engineering, Dalian University of Technology, Dalian 116024, China

2

School of Materials Science and Engineering, Shenyang Aerospace University, Shenyang 110136, China 3

LEM3UMR CNRS 7239, Université de Lorraine, 57045 Metz, France;

Abstract The twinning behavior of an AZ31B magnesium alloy with weak fiber texture was investigated and characterized in this work. It was found that both {10-11} contraction and {10-12} extension twinning can be activated during uniaxial tensile process. Numbers of both types of twins increase as tensile deformation continues. However, the increasing rate of contraction twins is much higher than that of extension twins due to the fact that Schmid factor of grains for contraction twinning distributes mainly in the higher value range than that for extension twinning, which results in a more rapid increment of resolved shear stress when loading stress increases. Primary twinning type selection depends on the integrate effect of Schmid factor and critical resolved shear stress. Extension twin variant selection is highly linked to the SF rank although Schmid factor criterion is not absolute and suffers some exceptions. Compared with extension twin, contraction twin variant selection suffers more exceptions of Schmid factor criterion.

Keywords Magnesium alloy; Fiber texture; Twinning; variant; Schmid factor



Corresponding author at: School of Materials Science and Engineering, Shenyang Aerospace University, South Avenue of Daoyi, 110136, Shenyang, China Tel. : +86-024-89723876; fax: +86-024-89723876. Email address: [email protected] 1

1. Introduction Magnesium alloys have been drawing much attention as a potential light-weight structural material [1, 2], but the limited formability at room temperature limits their extensive applications. According to Von Mises criterion that five independent slip systems should accommodate homogeneous deformation [3], only two independent basal slip systems of magnesium alloys could not satisfy this requirement. In some cases, other slips such as prismatic , pyramidal and pyramidal can be also activated. Nevertheless, their critical resolved shear stresses (CRSS) are much higher than that of basal slip at ambient temperature [4]. Except for pyramidal slip, none of slips could be activated to accommodate c-axis strain of a grain for homogeneous deformation. So twinning, as a deformation mode, usually plays an important role during the plastic deformation. Unlike slip, twinning makes a sudden reorientation of the crystal [5–8]. For example, {10-11} contraction twin re-orientates with <11-20> 56º with respective to the parent grain, which can stimulate the activation of the basal slip [8]. So twinning can afford a change of crystallite orientation to satisfy Von Mises criterion to continue deformation. Magnesium alloys are plentiful of twinning modes. In addition to {10-11} contraction twining, various twinning types were reported in previous researches [9–13]. {10-11} contraction twinning and {10-12} extension twining are the two dominant modes in general. Twinning is a unidirectional shear process. Under uniaxial loading, {10-11} contraction twinning or {10-12} extension twining always occurs depending on whether there is a contraction or tension strain component parallel to the c-axis of the 2

deformed grain. As compared with {10-12} extension twinning with CRSS of 2~2.8 MPa, it is more difficult for {10-11} contraction twinning to activate in magnesium alloys due to its higher CRSS which could reach to 76-153MPa[14]. The activation of a twining mode depends not only on its CRSS, but also on the orientation factor, i.e., Schmid factor (SF) of the grain undergoing deformation. Generally, texture always forms during processing. Textured magnesium alloys exhibit strong anisotropic behaviors because of SF-related deformation mechanisms. In the past decades, there have been a lot of investigations of mechanical behaviors of magnesium alloys [15–22]. It was found that twinning type selection determines the anisotropic mechanical behavior. Although the CRSS difference between {10-12} extension twinning and {10-11} contraction twinning is quite large, the two types of twinning could be activated in the same grain if the grain orientation were convenient [23]. It was noted that the lower CRSS is the main reason for {10-12} extension twins to activate in the grains with a wide orientation range, and {10-11} contraction twins with the higher CRSS activate in a narrow orientation range. As twinning begins at the early stage of deformation, to compare behaviors of the two types of twinning during the whole deformation process is of interest for understanding twinning type selection and its effect on the mechanical behaviors. In this work, twinning behaviors of a fiber-textured AZ31B magnesium alloy during tensile process were investigated at room temperature. Twinning characteristics was discussed in terms of type selection and variant selection.

3

2. Experimental In the present work, an as-cast AZ31B magnesium alloy (Al 3.0wt%+Zn 1.0wt%+Mn 0.3wt%, Bal. Mg) ingot was used. The ingot was extruded at 350oC with an extrusion rate of 5.0mm/min and an extrusion ratio of 12.8:1. The final bars with diameter of 15.0mm were obtained. After then, the extruded bars were annealed at 520 oC for 2h and cooled down in the air. For tensile test, the samples with a gauge size of Ф5.0mm30.0mm were prepared from the annealed bars with the tensile axis parallel to the extrusion direction (ED). The tensile test was conducted on an electro-hydraulic servo multi-function material testing machine (MTS landmark 100kN) at a strain rate of 5.610-4s-1 at room temperature. In the tensile test, a sample was tensioned to failure firstly, and then the following samples were tensioned and then stopped respectively at logarithmic strain of 7.7%, 10.9%, 14.1% and 17.0% for investigating the microstructure evolution. For the microstructure observation and electron backscattered diffraction (EBSD) measurement, specimens were cut from the tensile samples with an observed (measured) surface along the longitudinal-section. An Olympus GX71 type microscope was used for metallographic analysis, and the specimens were etched with acetic picral (5g of picric acid, 10 ml of acetic acid, 70 ml of ethanol and 10 ml of water) for 6-12s. An HKL Channel 5 system (Oxford instruments) equipped on Zeiss-Sigma scanning electron microscope (SEM) was used for EBSD analysis. Prior to observation and measurement, specimen surface was ground with 1000~5000 SiC papers and then 4

mechanically polished with diamond paste in size of 0.5μm. In order to get a more smooth surface, electrolytic polishing was conducted as the final step before EBSD measurement. The electro polishing was carried out at 15V for 60s in a solution of 10% perchloric acid (HClO4) and 90% absolute ethyl alcohol under -30 ºC. EBSD measurement was conducted under 20kV, and a step size of 0.7 μm was used. The boundaries were identified with misorientation of 86.4±5o <11-20> ± 5o, 56±5o <11-20> ±5o and 38±5o <11-20> ±5o, respectively, for {10-12} extension twins, {10-11} contraction twins and {10-11}-{10-12} double twins.

3. Results and discussion 3.1 Initial microstructure and texture Fig.1 presents the microstructures of the as-extruded and annealed bars. It can be seen that the as-extruded bar was not completely recrystallized (Fig.1 (a)). Most grains are very small (less than 10μm), and some large and elongated grains are parallel to the extrusion direction. After annealing, a complete recrystallization took place (Fig.1 (b)). Grains are equiaxed but not very homogeneous with the average size of 30m. Fig.2 (a) presents the orientation map of the annealed bars. The corresponding ED-inverse pole figure is displayed in Fig.2 (b). From Fig.2 (a), it is seen that most grains are colored in green. This demonstrates that most grains orientates with their c-axis perpendicular to ED, forming a weak <11-20> fiber texture that is consistent with Fig.2 (b). Figs.2 (c) and (d) show respectively the relative frequencies of misorientation angle and grain size. It is seen that the misorientation distribution is 5

random, and the grain size distributes mainly between 10m and 60m.

3.2 Mechanical behavior and microstructure evolution In Fig.3, 5 logarithmic stress-strain curves were obtained by tensile tests on samples to 7.7%, 10.9%, 14.1%, 17.0 and 20.1% total strain, respectively. The corresponding stresses are listed in table 1. The corresponding microstructures were also displayed in the figure. The yield strength, ultimate tensile strength and elongation to failure of the annealed bar is 112.0 MPa, 275.6 MPa and 18.9%, respectively. The stress-strain curve exhibits generally a concave down shape, i.e., strain hardening rate decreases with increasing tensile strain, which is consistent with the stress-strain response of the magnesium alloy with fiber texture [21, 24]. The concave-down shape of the stress-strain curve was explained in terms of contraction twining and slip-dominated mechanisms [25]. In the present work, c-axis of most grains is perpendicular to the tensile direction due to the weak texture, which is favorable for the occurrence of contraction twinning. However, CRSS of contraction twinning is very high, so it would be not easy to activate in grains deviating much from the fiber orientation. In contrast, extension twining with much lower CRSS would be easily activated in these grains. Table 1 Maximum stress of the samples at different tensile strains Strain / %

7.7

10.9

14.1

17.0

20.1

Maximum stress / MPa

234.1

249.8

265.2

274.3

275.6

When the sample was tensioned to 7.7%, a few twins appeared in some large grains 6

(as shown in Fig.3 (a)). Some of these twins exhibit a thin-long morphology that is typical feature of {10-11} contraction twins [17]. With increasing tensile strain, twins increase in number (see Figs.3 (b), (c), (d) and (e)). This means that twinning contributes to the strain in the whole tensile process. The previous researches indicated that the contribution of contraction twinning to strain is limited, its significance is to promote basal slip via the reorientation. Contraction twining is responsible for the concave down shape of stress-strain curve, and extension twinning is responsible for the sigmoid shape [16, 26]. So the stress-strain response is affected by the competition between contraction twinning and extension twinning. It was reported that contraction twins are usually the fracture origin at room temperature [8, 17, 27]. Thus texture which usually exerts an effect on the primary twinning type selection would be an important parameter to be concerned for estimation of mechanical behaviors. In the present study, the texture is weak, so in addition to contraction twinning, extension twinning is another important deformation mode.

3.3 Twinning characterization Fig.4 presents the orientation band contrast maps of the microstructures at the different tensile strains. It can be seen that both {10-11} contraction (colored in red) and {10-12} extension (colored in blue) twins were activated during tension. In addition, there exist lots of {10-11}-{10-12} double twins (colored in yellow). Obviously with increasing strain, numbers of both {10-11} contraction and {10-12} extension twins increase, but the number of contraction twins increases more rapidly 7

than that of extension twins. It is known that twinning activation strongly depends on the orientation of the deformed grain with respect to the direction of applied stress [16, 17], which has been usually considered as Schmid factor (SF) related [28]. Although SF criterion is not absolute and suffers some exceptions for twin variant selection, the twinning activation is highly linked to the SF rank [29, 30]. In the experiment, the orientation of the deformed grains with twins was identified in inverse pole figures as shown in Fig.5. It was found that at all the stress levels, the inverse pole of the deformed grains with extension twins distributes in a lower polar angle range, in contrast, that of the grains with contraction twins distributes in a higher polar angle range. So the above result suggests that the primary twinning selection is orientation factor related. High SF distributes in the crystallographic orientation space with high inverse polar angle for contraction twining, and that for extension twining distributes with the low polar angle [23]. Based on the EBSD analysis of the original microstructure, we collected the data of Euler angles of 156 grains (Fig.6 (a)) and calculated SFs for all six variants. The highest SF among the six variants distributes as shown in Fig.6 (b). For {10-12} extension twinning, the highest SF of 156 grains ranges from -0.15 to 0.5 and more than half of them have values from 0 to 0.2; For {10-11}contraction twinning, the highest SF of 156 grains ranges from 0 to 0.5 and overwhelming majority of these values ranges from 0.3 to 0.5. So the texture would be more convenient for the alloy deformed by contraction twinning in terms of SF. However, it is well known that in addition to SF, CRSS is another parameter to determine the activation of a deformation mechanism. CRSS of contraction twinning is much higher than that of extension twinning, so resolved shear stress applied on contraction 8

twinning systems (twin variants) should be much higher for their activation. According to the Schmid law, the resolved shear stress applied on the twinning systems is s=m, where  is tensile stress, m is SF. In this study, the resolved shear stresses applied on contraction and extension twinning systems of all 156 grains were calculated with the highest SFs under 5 loading stress levels which correspond to the different tensile strains (see table 1). Fig.7 presents the distributions of the calculated resolved shear stresses on contraction and extension twinning systems. CRSS for extension twinning and contraction twinning is 2~2.8MPa and 76~153MPa [14], respectively. We took respectively their median values of 2.4MPa and 115.0 MPa (indicated with vertical blue lines in Fig.7) as CRSSs to give a general estimation whether the twinning systems would be activated. From Fig.6, it is seen that with increasing of stress or strain, the number of grains with resolved shear stress greater than 2.4 MPa on the extension twinning system increases very slowly. When stress is 234.1MPa, 44.87% of grains would be subjected to resolved shear stress greater than 2.4MPa on the extension twinning system (Fig.7 (a)), and this probability maintains at this value until the tensile stress reaches to 249.8MPa (Fig.7 (b)), and then increases slightly to 45.51% when the stress increases to 265.2MPa. As loading stress increases further, the probability remains unchanged (see Fig.7 (c), (d) and (e)). In contrast, the number of grains with resolved shear stress greater than 115.0MPa on the contraction twinning system increases more rapidly. When stress is 234.1MPa, there are merely 4.49% grains subject to resolved shear stress greater than 115.0MPa (Fig.7 (f)). When the loading stress increases to 249.8MPa (Fig.7 (g)), 265.2MPa (Fig.7 (h)), 274.3MPa (Fig.7 (i)) and 275.6MP (Fig.7 (j)), correspondingly, the probability is respectively 22.44%, 38.46%, 51.92% and 51.92%.

9

Analyzed as the above, it suggests that during tension, both contraction and extension twinning can be activated, the number of twins increases with increasing loading stress (or strain), but the number of contraction twins increases more rapidly than that of extension twins because SF of grains for contraction twinning distributes mainly in the higher value range than that for extension twinning, which results in a more rapid increment of resolved shear stress when loading stress increases. This conclusion is consistent with the present experiment result (Fig.4). According to Euler angles of the deformed grains and the activated twins in them, we determined variants of all the twins and calculated SF for these variants. Fig.8 presents percentages of the activated twins with SF in the range of 0-0.1, 0.1-0.2, 0.2-0.3, 0.3-0.4 and 0.4-0.5, respectively, at different tensile strains. It was found that SF distributions for extension and contraction twin variants are quite different, but both are approximately similar to those in Fig.6 (b). SF of the activated extension twins mostly distributes in the range of 0-0.2, while SF of the activated contraction twins mostly distributes in the range of 0.1-0.4. This means that under the same loading stress, extension twins are easier to be activated than contraction twins. In the present study, due to the weak fiber texture, although most grains orientate with <11-20> parallel to the tensile axis, which is favorable for contraction twinning, there exist some other grains with orientation deviating from the fiber texture. These grains have lower SF for extension twinning than contraction twinning at average. However, due to the lower CRSS, extension twins can be easily activated in the tensile process. So to predict the primary twinning type selection, the consideration of both SF and CRSS is the precondition. The average SF for the activated twins in the deformed grains is shown in Fig.9. It can 10

be seen that the average SF value of contraction twins is higher than that of extension twins. In Fig.9, two lines are approximately parallel to each other, though slope (-0.00262) of line for extension twins is slightly lower than that (-0.00217) for contraction twins. It is seen that the average SF value for both extension and contraction twins decreases with increasing tensile strain. The reason is that the increment of loading stress results in activation of more twins which has low SF and cannot be activated at lower stress.

3.4 Twin variant selection As shown above, the primary twinning type selection is determined by considering simultaneously SF and CRSS. However, for the same twining type, SF is not a decisive criterion to judge whether a variant is activated. Barnett et al. [31] concluded that primary {10-12} extension twinning departs more from SF-type predictions than {10-11} contraction twinning. Jonas et al. [32] found that about half of the observed {10-11} contraction twins were of the “high SF” (0.3–0.5) type, while nearly half had “low” SFs (0.15–0.30). Furthermore, 5% of the observed twins were of the “very low SF” type. For extension twinning, similar results were obtained in the references [23, 33, 34]. In the present experiment, it was found that all six variants of both extension and contraction twins were activated in the grains. Fig.10 presents the percentages of activated twin variants at different strains. In Fig.10, it seems that the six variants were activated randomly, no obvious character can be concluded, but as shown in Fig.11, extension twins were more likely activated in SF rank order. In Fig.11, Rank 1 11

stands for the highest and rank 6 the lowest SF. Most of extension twin variants were activated with rank1, secondly rank2 (Fig.11 (a)). This result is consistent with those in Ref. [29, 35, 36]. It demonstrates that extension twin variant selection is highly linked to the SF rank although SF criterion is not absolute and suffers some exceptions. In contrast, contraction twin variants were activated with ranks 2-4 (Fig.11 (b)). It suggests that contraction twin variant selection suffers more exceptions. The plastic deformation process of polycrystalline materials is quite complicated. This process includes strain accommodation between grains. Based on the strain accommodation process, Jonas et al [32, 37] proposed a strain accommodation model explaining the deviation from SF-type prediction of twin variant selection. In addition to SF, the strain accommodation between grains plays a key role. Based on this model, an accommodation factor was proposed in our previous research [38], which predicted extension twin variant selection well. In this case, Schmid factor, CRSS of deformation systems and misorientation between the twinned grain and the accommodation grain were quantitatively integrated into one parameter, i.e., the accommodation factor:

A 

 1 1  cp  FtFl t Ft

(1)

where t and cp is respectively the CRSS of the twinning system and the accommodative deformation system; Ft and Fl is respectively Schmid factor for twinning and misorientation factor for accommodation deformation mechanism. Under this consideration, the average misorientation factor Fl for extension- and 12

contraction-twinning was calculated, respectively. It is 0.56 for extension- and 0.58 for contraction-twinning if accommodated by basal slip. Accommodated by prismatic slip, the average misorientation factor is respectively 0.66 for extension- and 0.69 for contraction-twinning. It is seen that the average misorientation factor for contraction twinning is higher than that for extension twining accommodated by both basal and prismatic slips which are the easier accommodation mechanisms to be activated. This means that in tension, twin variant selection for contraction twinning would be more affected by strain accommodation between grains with lower accommodation factor as compared with extension twin variant selection at the same level of SF. This leads to the result that contraction twin variant selection deviates more from SF-type prediction.

4. Summary and conclusions The twinning behavior of an AZ31B magnesium alloy with weak fiber texture was investigated and characterized in this work. Both {10-11} contraction and {10-12} extension twinning was activated during uniaxial tensile process. Some conclusions were derived as follows: (1) The matrix grains with activated contraction twins orientate in a higher inverse polar angle range than those with extension twins. The average Schmid factor for the activated contraction twins is higher than that for the extension twins. Primary twinning type selection should be determined by simultaneous consideration of Schmid factor and critical resolved shear stress. 13

(2) Numbers of twins increase as tensile deformation continues. However, the increasing rate of contraction twins is much higher than that of extension twins. The reason is that Schmid factor of grains for contraction twinning distributes mainly in the higher value range than that for extension twinning, which results in a more rapid increment of resolved shear stress when loading stress increases. (3) Extension twin variant selection is highly linked to the SF rank although Schmid factor criterion is not absolute and suffers some exceptions. Compared with extension twin, contraction twin variant selection suffers more exceptions of Schmid factor criterion.

Acknowledgements The authors gratefully acknowledge the financial support from the National Foundation of Natural Science (No.51371121) of China.

References [1]

E. Aghion, B. Bronfin, Magnesium Alloys Development towards the 21st Century, Materials Science Forum. 350–351 (2000) 19–30.

[2]

J. Hirsch, T. Al-Samman, Superior light metals by texture engineering: Optimized aluminum and magnesium alloys for automotive applications, Acta Materialia. 61 (2013) 818–843.

[3]

R. V. Mises, Mechanik der plastischen Formänderung von Kristallen, ZAMM Zeitschrift Für Angewandte Mathematik Und Mechanik. 8 (1928) 161–185.

[4]

N. Zhou, Z. Zhang, L. Jin, J. Dong, B. Chen, W. Ding, Ductility improvement by twinning and twin-slip interaction in a Mg-Y alloy, Materials and Design. 56 (2014) 966–974. 14

[5]

S.H. Park, S. Hong, S.C. Lee, In-plane anisotropic deformation behavior of rolled Mg – 3Al – 1Zn alloy by initial { 10 – 12 } twins, Materials Science & Engineering A. 570 (2013) 149–163.

[6]

D. Sarker, D.L. Chen, Texture transformation in an extruded magnesium alloy under pressure, Materials Science and Engineering A. 582 (2013) 63–67.

[7]

O. Muránsky, M.R. Barnett, D.G. Carr, S.C. Vogel, E.C. Oliver, Investigation of deformation twinning in a fine-grained and coarse-grained ZM20 Mg alloy: Combined in situ neutron diffraction and acoustic emission, Acta Materialia. 58 (2010) 1503–1517.

[8]

K. Yoshida, Prediction of ductile fracture induced by contraction twinning in AZ31 sheet subjected to uniaxial and biaxial stretching modes, International Journal of Plasticity. 84 (2015) 102–137.

[9]

H. Yoshinaga, T. Obara, S. Morozumi, Twinning deformation in magnesium compressed along the C-axis, Materials Science and Engineering. 12 (1973) 255–264.

[10] B.C. Wonsiewicz, W.A. Backofen, Plasticity of magnesium crystals, Trans. TMS-AIME, 239 (1967), 1422-1431. [11]

E.W. Kelley, W. F. Hosford, Plane-strain compression of magnesium and magnesium alloy crystals, Trans Met Soc AIME. 242 (1968) 5–13.

[12] R.. Reed-Hill, W.. Robertson, Additional modes of deformation twinning in magnesium, Acta Metallurgica. 5 (1957) 717–727. [13] R.. Reed-Hill, W.. Robertson, The crystallographic characteristics of fracture in magnesium single crystals, Acta Metallurgica. 5 (1957) 728–737. [14] J. Koike, Enhanced deformation mechanisms by anisotropic plasticity in polycrystalline Mg alloys at room temperature, Metallurgical and Materials Transactions A. 36 (2005) 1689–1696. [15] D. Ando, J. Koike, Y. Sutou, Relationship between deformation twinning and surface step formation in AZ31 magnesium alloys, Acta Materialia. 58 (2010) 4316–4324. [16] M.R. Barnett, Twinning and the ductility of magnesium alloys. Part I: ‘Tension’ 15

twins, Materials Science and Engineering A. 464 (2007) 1–7. [17] M.R. Barnett, Twinning and the ductility of magnesium alloys. Part II. ‘Contraction’ twins, Materials Science and Engineering A. 464 (2007) 8–16. [18] S.B. Yi, C.H.J. Davies, H.G. Brokmeier, R.E. Bolmaro, K.U. Kainer, J. Homeyer, Deformation and texture evolution in AZ31 magnesium alloy during uniaxial loading, Acta Materialia. 54 (2006) 549–562. [19] Z. Keshavarz, M.R. Barnett, EBSD analysis of deformation modes in Mg-3Al-1Zn, Scripta Materialia. 55 (2006) 915–918. [20] L. Jiang, J.J. Jonas, R.K. Mishra, A.A. Luo, A.K. Sachdev, S. Godet, Twinning and texture development in two Mg alloys subjected to loading along three different strain paths, Acta Materialia. 55 (2007) 3899–3910. [21] Y.N. Wang, J.C. Huang, The role of twinning and untwinning in yielding behavior in hot-extruded Mg-Al-Zn alloy, Acta Materialia. 55 (2007) 897–905. [22] M.R. Barnett, M.D. Nave, A. Ghaderi, Yield point elongation due to twinning in a magnesium alloy, Acta Materialia. 60 (2012) 1433–1443. [23] B.L. Wu, Y.D. Zhang, G. Wan, M. Humbert, F. Wagner, C. Esling, Primary twinning selection with respect to orientation of deformed grains in ultra-rapidly compressed AZ31 alloy, Materials Science and Engineering: A. 541 (2012) 120–127. [24] Y. Chino, K. Kimura, M. Hakamada, M. Mabuchi, Mechanical anisotropy due to twinning in an extruded AZ31 Mg alloy, Materials Science and Engineering A. 485 (2008) 311–317. [25] M. Gzyl, R. Pesci, A. Rosochowski, S. Boczkal, L. Olejnik, In situ analysis of the influence of twinning on the strain hardening rate and fracture mechanism in AZ31B magnesium alloy, Journal of Materials Science. 50 (2015) 2532– 2543. [26] S.G. Hong, S.H. Park, C.S. Lee, Role of {10-12} twinning characteristics in the deformation behavior of a polycrystalline magnesium alloy, Acta Materialia. 58 (2010) 5873–5885. [27] P. Cizek, M.R. Barnett, Characteristics of the contraction twins formed close to 16

the fracture surface in Mg-3Al-1Zn alloy deformed in tension, Scripta Materialia. 59 (2008) 959–962. [28] D.W. Brown, S.R. Agnew, M.A.M. Bourke, T.M. Holden, S.C. Vogel, C.N. Tomé, Internal strain and texture evolution during deformation twinning in magnesium, Materials Science and Engineering A. 399 (2005) 1–12. [29] S. Godet, L. Jiang, A.A. Luo, J.J. Jonas, Use of Schmid factors to select extension twin variants in extruded magnesium alloy tubes, Scripta Materialia. 55 (2006) 1055–1058. [30] M.R. Barnett, A. Ghaderi, J. Quinta Da Fonseca, J.D. Robson, Influence of orientation on twin nucleation and growth at low strains in a magnesium alloy, Acta Materialia. 80 (2014) 380–391. [31] M.R. Barnett, Z. Keshavarz, A.G. Beer, X. Ma, Non-Schmid behaviour during secondary twinning in a polycrystalline magnesium alloy, Acta Materialia. 56 (2008) 5–15. [32] J.J. Jonas, S. Mu, T. Al-Samman, G. Gottstein, L. Jiang, E. Martin, The role of strain accommodation during the variant selection of primary twins in magnesium, Acta Materialia. 59 (2011) 2046–2056. [33] I.J. Beyerlein, L. Capolungo, P.E. Marshall, R.J. McCabe, C.N. Tomé, Statistical analyses of deformation twinning in magnesium, Philosophical Magazine. 90 (2010) 2161–2190. [34] Y. Pei, A. Godfrey, J. Jiang, Y.B. Zhang, W. Liu, Q. Liu, Extension twin variant selection during uniaxial compression of a magnesium alloy, Materials Science and Engineering A. 550 (2012) 138–145. [35] S.H. Park, S.G. Hong, C.S. Lee, Activation mode dependent {1 0 -1 2} twinning characteristics in a polycrystalline magnesium alloy, Scripta Materialia. 62 (2010) 202–205. [36] L.H. Song, B.L. Wu, L. Zhang, X.H. Du, Y.N. Wang, Y.D. Zhang, C. Esling, Cyclic deformation behaviors of AZ31B magnesium alloy in two different asymmetric loading manners, Materials Science and Engineering: A. 689 (2017) 134–141. 17

[37] S. Mu, J.J. Jonas, G. Gottstein, Variant selection of primary, secondary and tertiary twins in a deformed Mg alloy, Acta Materialia. 60 (2012) 2043–2053. [38] B. L. Wu, G. S. Duan, X.H. Du, L.H. Song, Y.D. Zhang, M.J. Philippe, C. Esling, In situ investigation of extension twinning-detwinning and its effect on the mechanical behavior of AZ31B magnesium alloy, Materials & Design. 132 (2017) 57–65.

18

阿)

阿)

Fig.1 Microstructures of the AZ31B magnesium alloy as (a) extruded and (b) annealed.

19

(b)

(c)

(d)

Fig.2 Characteristics of the annealed alloy. (a) inverse pole orientation maps; (b) inverse pole figure; (c) misorientation angle distribution and (d) grain size distribution.

20

Fig. 3 Logarithmic stress-strain curves along with correlative microstructures of samples at tensile strain and strain hardening rate of failure sample.

21

Fig.4 Band contrast maps of the samples at different tensile strains. 14.1%; (d) 17.0% and (e) 20.1%.

22

(a) 7.7%; (b) 10.9%; (c)

(a)

(f)

(b)

(g)

(c)

(h)

(d)

(i)

(e)

(j)

Fig.5 Scattered inverse pole figures of twined grains in the samples at different tensile strains.

(a) with

extension twins, 7.7% ; (b) with extension twins, 10.9% ; (c) with extension twins, 14.1%; (d) with extension twins, 17.0%; (e) with extension twins, 20.1%; (f) with contraction twins, 7.7%; (g) with contraction twins, 10.9%; (h) with contraction twins, 14.1%; (i) with contraction twins, 17.0%; (j) with contraction twins, 20.1%. 23

(a)

(b)

Fig. 6 156 collected grains in the origin microstructure under EBSD (a) and distribution of the highest of Schmid factor among the six twin variants (b).

24

25

Fig. 7 Distribution of resolved shear stress of 156 collected grains on extension twinning (a)-(e) and contraction twinning (f)-(j) systems under different tensile stresses.

(a)

(b)

阿)

阿)

Fig.8 Schmid factor distribution for activated twin variants at different tensile strains. extension twin variants; (b) contraction twin variants.

26

(a)

Fig.9 Averaged Schmid factor for activated twin variants at different strains.

27

(a)

(b)

阿)

阿)

Fig.10 Frequency of twin variants at different tensile strain。 (a) extension twin variants; (b) contraction twin variants.

28

(a)

(b)

阿)

阿)

Fig.11 Frequency of Schmid factor rank of activated twin variants at different tensile strains. extension twin variants; (b) contraction twin variants.

29

(a)