- Email: [email protected]
Kink formation and concomitant twin nucleation in Mg–Y
Scripta Materialia xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Scripta Materialia journal homepage: www.elsevier.com/locate/scriptamat
Kink formation and concomitant twin nucleation in Mg–Y L. Wang a,⇑, J. Sabisch a,b, E.T. Lilleodden a a b
Institute of Materials Research, Helmholtz-Zentrum Geesthacht, 21502 Geesthacht, Germany Department of Materials Science and Engineering, University of California, Berkeley, CA 94709, USA
a r t i c l e
i n f o
Article history: Received 22 June 2015 Revised 13 August 2015 Accepted 16 August 2015 Available online xxxx Keywords: Magnesium alloys Tensile testing Kink band Twinning
a b s t r a c t During in situ tensile loading of a Mg–2.5 wt%Y alloy, a large kink band was observed. The orientation of the kink band varies continuously with a maximum misorientation angle of 18° with respect to the 2g matrix grain. The formation of the kink band is accompanied by the nucleation of numerous f1 0 1 twins at the kink boundaries. A mechanistic model is proposed to account for this unique kink band and concomitant twin nucleation via dissociation of basal dislocations. This mechanism is aided by the 2g twinning and simultaneously decreases the stacking fault presence of Y, which suppresses f1 0 1 energy. Ó 2015 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
Mg is considered a promising structural material for automobile and aerospace industries due to its high specific strength compared with Al and steels. A critical technical barrier against a widespread application of Mg alloys, however, is the lack of ductility and poor formability [1]. At room temperature, the deformation of Mg is 2g tensile twinning, while other dominated by basal slip and f1 0 1 slip or twin systems are generally suppressed due to their much higher critical resolved shear stresses (CRSSs) [2–6]. One approach to alleviate this problem is by adding a small amount of solute rare earth elements, which have been shown to affect both relative CRSS and stacking fault energy (SFE). Stanford et al. [7] estimated 2g twinning the CRSS of basal slip, pyramidal hc + ai slip, and f1 0 1 in Mg–2.2 wt%Y by coupling in situ neutron diffraction experiment and elastoplastic self-consistent (EPSC) simulation. The result suggests that Y hardens all three deformation modes, but the harden 2g twinning than for slip modes. ing effect is stronger for f1 0 1 Sandlöbes et al. [8] used transmission electron microscopy (TEM) to examine the deformation microstructure in Mg–3 wt%Y and pure Mg after cold rolling. A high density of I1 stacking faults were found, which served as nucleation sites for hc + ai dislocations. It suggests that Y reduces the stacking fault energy (SFE) in Mg, as was subsequently verified by first-principles studies [9,10]. In a more recent study, Stanford et al. [11] reported the observation 1g twins in a single-phase Mg–10 wt%Y alloy after comof f1 1 2 pression. While this twin mode has been frequently observed in other hexagonal materials such as Ti [12,13] (c/a = 1.587), Zr
⇑ Corresponding author. E-mail addresses: [email protected], [email protected] (L. Wang).
[14,15] (c/a = 1.593), Re [16] (c/a = 1.614), and Co [17] (c/ a = 1.628), it is rarely observed for Mg (c/a = 1.623). A general argument for the surprising observation of this twin in a Mg–Y alloy is 2g twin mode which leads to the that Y hardens the primary f1 0 1 1g twins, but details of the twin preferential nucleation of f1 1 2 nucleation process is still unclear because the characterization was made ex situ. Here, we present the results from an in situ straining study of a coarse-grained Mg–Y alloy, using grain-level characterization of the deformation-induced microstructural evolution. An unexpected kink band was observed to form in a grain whose c-axis was almost parallel to the tensile direction. Basal slip lines were 2g twins were identified found crossing the kink, and small f1 0 1 in the vicinity of the kink. Based on the observed microstructure evolution, a dislocation reaction model is proposed to account for the kink formation. A binary Mg–2.5 wt%Y solid solution alloy was produced by casting at the Magnesium Innovation Centre (MagIC) of Helmholtz-Zentrum Geesthacht. To achieve a homogeneous composition distribution, the melt was taken to 700 °C for 10 min before casting. A relatively small cylinder die (D = 18 mm, H = 100 mm) was used. The as-cast piece of cylinder material was cut into two half-cylinders by electrical discharge machining (EDM). Dog-bone tensile specimens with a gauge dimension of 5 mm 2 mm 1 mm (L W T) were extracted from one halfcylinder, with the tensile axis being parallel to the vertical direction of the half-cylinder. The specimen to be tested was close to the top of the half-cylinder. The surface was prepared using conventional grinding and polishing procedures. The final thickness was reduced to 0.4 mm. An EBSD scan of the initial
http://dx.doi.org/10.1016/j.scriptamat.2015.08.016 1359-6462/Ó 2015 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: L. Wang et al., Scripta Mater. (2015), http://dx.doi.org/10.1016/j.scriptamat.2015.08.016
2
L. Wang et al. / Scripta Materialia xxx (2015) xxx–xxx
microstructure close to the gauge center was performed using a Zeiss Gemini Ultra 55 scanning electron microscope (SEM). Grain sizes range from 100 lm to more than 1 mm. The specimen was deformed by tensile test using a MICROTEST 200N (Deben, UK Limited) module at a constant crosshead speed of 0.2 mm/min, which corresponds to a nominal strain rate of 6.67 104 s1. The microstructure evolution during the loading was monitored using an Olympus SZX10 optical microscope connected with a video recording system. Fig. 1 shows the engineering stress–strain curve of the test at a strain of about 10%, along with the microstructure evolution near the gauge center associated with points A–H labeled on the stress–strain curve. Based on the initial EBSD measurement, crystallographic orientations of four large grains (d > 500 lm) as well as the approximate grain boundaries between them were overlaid on the micrograph of the initial structure prior to loading (point A). Grain 3 (Euler angles = (181°, 83°, 5°)) is a large grain with its c-axis almost parallel to the tensile axis. This is a favorable 2g twin nucleation, and our initial focus was on condition for f1 0 1 the 1–3 grain boundary in order to observe potential slip transfer induced twin nucleation events [18]. During the loading, slip lines were first observed in Grain 1 and Grain 2, favorably oriented grains for basal and prismatic slip, respectively. At an applied stress of 70 MPa (point C), slip lines appeared inside Grain 3. Using the standard trace analysis [18], these slip lines were found to correspond to basal slip, as marked by the white dashed line in the micrograph for C. At point D, a lenticular kink band, evidenced by the bright contrast, quickly developed in Grain 3 from the region where basal slip lines had concentrated. The original basal
slip lines were redirected by about 12° relative to their initial direction at point C, suggesting an orientation rotation inside the kink band. In addition, small features show up at the upper boundary of the kink band, which are later identified as newly nucleated 0 1 2Þ ½1 0 1 1 twins. With further straining (from E–H), the kink ð1 0 1 2Þ twins were found in its vicinity, grew laterally and more ð1 and a few ð1 1 0 2Þ ½1 1 0 1 twins with dark contrast were identified in the left part of Grain 3 (point F). The video recording the microstructure evolution during loading can be found in the Supplementary material (loading.mp4). After the peak load was reached, the sample was gradually unloaded at the same displacement rate, and the unloading process was recorded in a second video (unloading.mp4). Detailed EBSD characterization was subsequently performed for the kink in Grain 3, as shown in Fig. 2. From the optical micrograph, the kink band grew across the entire Grain 3. There are also some small kinks parallel to the big kink, as evidenced by their bright contrast. The direction of basal slip lines are deviated when crossing these kinks. Furthermore, twins are observed adjacent to the kink boundary, having likely nucleated from the upper boundary of the big kink. These are marked by red arrows in the figure. According to EBSD based misorientation analysis (Fig. 2(b)), as well as twin plane trace analysis (Fig. 2(c)), they belong to the twin system 0 1 2Þ ½1 0 1 1 with a Schmid factor of 0.482. Likewise, twins that ð1 nucleated from the grain boundary between Grains 3 and 4 (Fig. 1, 1 0 2Þ ½1 1 0 1 twins with a Schmid facpoint F) were identified as ð1 tor of 0.493.
Fig. 1. Engineering stress–strain curve of the tensile test. Optical microscopy tracks the microstructure evolution. Basal slip lines first appeared in Grain 3 (C), followed by the 0 1 2Þ twins nucleated from the upper boundary of the kink. With further straining, the kink grew laterally and more (1 0 1 2Þ twins are formation of a large kink band (D). ð1 1 0 2Þ twins also developed in the left part of Grain 3 (F). found in its vicinity. A few ð1
Please cite this article in press as: L. Wang et al., Scripta Mater. (2015), http://dx.doi.org/10.1016/j.scriptamat.2015.08.016
L. Wang et al. / Scripta Materialia xxx (2015) xxx–xxx
3
0 1 2Þ twins in its vicinity as mark by red arrows. (b) EBSD inverse pole figure map of the small box region in (a) that identifies Fig. 2. (a) Optical micrograph of the kink and ð1 0 1 2Þ twins. (c) Plane trace for the six f1 0 1 2g planes. The ð1 0 1 2Þ twins are cross-verified by the trace analysis. (d) EBSD inverse pole figure map of the big box region in the ð1 0 1 2Þ and ð0 1 1 2Þ twins were found both inside and outside the kink. (e) {0 0 0 1} pole figure (a) that covers the kink. The kink has varied crystal orientations inside. Small ð1 0 1 2Þ and ð0 1 1 2Þ twins are evident. (f) Misorientation profile along the vertical arrow in (d). (For interpretation of the corresponding to the same area of (d), where ð1 references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 2(d) shows the EBSD inverse pole figure (IPF) map for the big box region in Fig. 2(a) that covers the kink band. Local crystal orientations are schematically presented as small hexagonal cells overlaid on it. Crystal orientations inside the kink varied from place 2g twins to place, but being distinct from the matrix grain. f1 0 1 were found both inside and outside the kink. By misorientation 0 1 2Þ and ð0 1 1 2Þ twins are present, with their analysis, both ð1 respective orientations indicated in Fig. 2(d). Fig. 2(e) shows the {0 0 0 1} pole figure for the same area of Fig. 2(d). Orientation 0 1 2Þ and ð0 1 1 2Þ spread due to the kink is evident. Poles from ð1 twins are also marked. Fig. 2(f) shows the misorientation profile along the vertical arrow in Fig. 2(d), calculated by the OIMTM software (EDAX, Draper, UT, USA). Crystal orientation changed continuously from the matrix into the kink and then back to the matrix on the other side. The maximum misorientation from the matrix to the center of the kink was about 18°, smaller than the theoretical 2g 86 , misorientation for all possible twin modes in Mg: f1 0 1
1g 56 , f1 0 1 3g 64 , f1 1 2 1g 35 [19]. This misorientaf1 0 1 tion analysis confirms that the kink band is not a twin. The crystal orientation of Grain 3 with respect to the loading direction results in low Schmid factor for basal slip (0.12) and 2g twinning (0.48). If the material high Schmid factor for f1 0 1 were pure Mg with a ratio of CRSStwin/CRSSbasal in the range of 1.5–2 [20], we would expect the grain to primarily deform by 2g twinning. The fact that basal slip lines appeared prior to f1 0 1 2g twins in Grain 3 during in situ loading indicates that the f1 0 1 ratio of CRSStwin/CRSSbasal is probably greater than 4 (=0.48/0.12) for this alloy. This is consistent with the conclusion made by Stanford et al. [7], who estimated the ratio to be about 5.6 in Mg–2.2 wt%Y. Nonetheless, basal slip in Grain 3 has a relatively small resolved shear stress. Development of the kink turns out to be an effective way to facilitate further deformation in this grain. The volume inside the kink is more favorably oriented for basal slip than the
Please cite this article in press as: L. Wang et al., Scripta Mater. (2015), http://dx.doi.org/10.1016/j.scriptamat.2015.08.016
4
L. Wang et al. / Scripta Materialia xxx (2015) xxx–xxx
2g twins from the kink boundaries via dissociation of basal dislocations. Fig. 3. A mechanistic model to account for the kink formation and concomitant nucleation of f1 0 1
matrix; kink boundaries also serve as heterogeneous sites for 2g twin nucleation. Taking a closer look at the video of specf1 0 1 imen loading, twin nucleation was almost concomitant with the kink propagation. Many of these twins were unable to grow substantially and were swept by the moving kink boundary during further straining. As a result, some twins are found inside the kink after the test, as shown in Fig. 2(d). On the other hand, twins in the vicinity of the kink were also likely to impede its further growth after point F (Fig. 1). A mechanistic model is proposed in Fig. 3 to explain the process of kink formation and concomitant twin nucleation. The key to the model is the dissociation of basal dislocations at the kink boundary, with a number of twinning dislocations being generated: 0
ba ¼ ba þ xbtw
ð1Þ
This reaction is partially supported by previous atomistic simulation studies [21–23], which suggest basal dislocations may disso 2g twin interface. ciate into twinning dislocations at the f1 0 1 Having described the mechanism by which this unique kink band occurs, it is now important to consider the reasons for which the kink is typically not observed in binary Mg alloys, but is uniquely observed here for the solid solution Mg–Y alloy. The viability of this kink mechanism in Mg–Y depends on two independent aspects of how Y influences the microscale deformation in 2g twinning so that Mg. First of all, Y increases the CRSS for f1 0 1 alternative deformation mechanisms, such as kinking, become necessary to accommodate c-axis extension in grains whose c-axes are close to the tensile direction. Secondly, Y is one of the most effective elements to reduce the SFE of Mg according to density function theory (DFT) studies [10]. Lowered SFE generally increases the spacing between the two partials of a full dislocation on the basal plane, which provides a favorable condition for localized dislocation reaction to occur that is necessary for kink to form. Interestingly, kinking is a major deformation mechanism in recently studied long-period stacking ordered (LPSO) phase in Mg–Zn–Y ternary alloys [24–26]. Inferred from the result in the present study, it is reasonable to believe that solute Y is also responsible for the LPSO phase to deform by kinking and increase the overall material ductility. Future design of Mg alloys may use the ability of kink formation as a possible mechanism to improve material ductility. In summary, formation of a large, unique kink was observed in a binary Mg–Y alloy during in situ tensile straining. The kink boundary served as a heterogeneous nucleation site for numerous
2g twinning events. A mechanistic model is proposed to f1 0 1 account for the kink formation and concomitant twin nucleation via dissociation of basal dislocations. This mechanism is enabled by the presence of Y. This work was funded by the Helmholtz Association. L.W. is supported by the Alexander von Humboldt Foundation. The authors would like to thank Dr. Chamini Mendis for useful discussions and the fabrication of the cast Mg–Y alloys used herein.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.scriptamat.2015. 08.016.
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26]
B.C. Suh, M.S. Shim, K.S. Shin, N.J. Kim, Scr. Mater. 84–85 (2014) 1. S.R. Agnew, O. Duygulu, Int. J. Plast. 21 (2005) 1161. J. Koike, Metall. Mater. Trans. A 36 (2005) 1689. W.B. Hutchinson, M.R. Barnett, Scr. Mater. 63 (2010) 737. X.Y. Lou, M. Li, R.K. Boger, S.R. Agnew, R.H. Wagoner, Int. J. Plast 23 (2007) 44. C.M. Byer, B. Li, B. Cao, K.T. Ramesh, Scr. Mater. 62 (2010) 536. N. Stanford, R. Cottam, B. Davis, J. Robson, Acta Mater. 78 (2014) 1. S. Sandlöbes, S. Zaefferer, I. Schestakow, S. Yi, R. Gonzalez-Martinez, Acta Mater. 59 (2011) 429. S. Sandlöbes, M. Friák, S. Zaefferer, A. Dick, S. Yi, D. Letzig, Z. Pei, L.-F. Zhu, J. Neugebauer, D. Raabe, Acta Mater. 60 (2012) 3011. S.L. Shang, W.Y. Wang, B.C. Zhou, Y. Wang, K.A. Darling, L.J. Kecskes, S.N. Mathaudhu, Z.K. Liu, Acta Mater. 67 (2014) 168. N. Stanford, R.K.W. Marceau, M.R. Barnett, Acta Mater. 82 (2015) 447. N. Bozzolo, L.S. Chan, A.D. Rollett, J. Appl. Crystallogr. 43 (2010) 596. L. Wang, R.I. Barabash, T.R. Bieler, W. Liu, P. Eisenlohr, Metall. Mater. Trans. A 44 (2013) 3664. R.E. Reed-Hill, W.H. Hartt, W.A. Slippy, Trans. AIME 242 (1968) 2211. L. Capolungo, P.E. Marshall, R.J. McCabe, I.J. Beyerlein, C.N. Tomé, Acta Mater. 59 (2009) 6047. J. Kacher, A.M. Minor, Acta Mater. 81 (2014) 1. S. Vaidya, S. Mahajan, Acta Metall. 28 (1980) 1123. L. Wang, Y. Yang, P. Eisenlohr, T.R. Bieler, M.A. Crimp, D.E. Mason, Metall. Mater. Trans. A 41 (2010) 421. N. Stanford, Philos. Mag. Lett. 88 (2008) 379. N. Stanford, M.R. Barnett, Int. J. Plast 47 (2013) 165. A. Serra, D.J. Bacon, Philos. Mag. A 73 (1996) 333. J. Wang, L. Liu, C.N. Tomé, S.X. Mao, S.K. Gong, Mater. Res. Lett. 1 (2013) 81. H. El Kadiri, C.D. Barrett, J. Wang, C.N. Tomé, Acta Mater. 85 (2015) 354. K. Hagihara, N. Yokotani, Y. Umakoshi, Intermetallics 18 (2010) 267. X.H. Shao, Z.Q. Yang, X.L. Ma, Acta Mater. 58 (2010) 4760. J.-K. Kim, S. Sandlöbes, D. Raabe, Acta Mater. 82 (2015) 414.
Please cite this article in press as: L. Wang et al., Scripta Mater. (2015), http://dx.doi.org/10.1016/j.scriptamat.2015.08.016
Contents lists available at ScienceDirect
Scripta Materialia journal homepage: www.elsevier.com/locate/scriptamat
Kink formation and concomitant twin nucleation in Mg–Y L. Wang a,⇑, J. Sabisch a,b, E.T. Lilleodden a a b
Institute of Materials Research, Helmholtz-Zentrum Geesthacht, 21502 Geesthacht, Germany Department of Materials Science and Engineering, University of California, Berkeley, CA 94709, USA
a r t i c l e
i n f o
Article history: Received 22 June 2015 Revised 13 August 2015 Accepted 16 August 2015 Available online xxxx Keywords: Magnesium alloys Tensile testing Kink band Twinning
a b s t r a c t During in situ tensile loading of a Mg–2.5 wt%Y alloy, a large kink band was observed. The orientation of the kink band varies continuously with a maximum misorientation angle of 18° with respect to the 2g matrix grain. The formation of the kink band is accompanied by the nucleation of numerous f1 0 1 twins at the kink boundaries. A mechanistic model is proposed to account for this unique kink band and concomitant twin nucleation via dissociation of basal dislocations. This mechanism is aided by the 2g twinning and simultaneously decreases the stacking fault presence of Y, which suppresses f1 0 1 energy. Ó 2015 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
Mg is considered a promising structural material for automobile and aerospace industries due to its high specific strength compared with Al and steels. A critical technical barrier against a widespread application of Mg alloys, however, is the lack of ductility and poor formability [1]. At room temperature, the deformation of Mg is 2g tensile twinning, while other dominated by basal slip and f1 0 1 slip or twin systems are generally suppressed due to their much higher critical resolved shear stresses (CRSSs) [2–6]. One approach to alleviate this problem is by adding a small amount of solute rare earth elements, which have been shown to affect both relative CRSS and stacking fault energy (SFE). Stanford et al. [7] estimated 2g twinning the CRSS of basal slip, pyramidal hc + ai slip, and f1 0 1 in Mg–2.2 wt%Y by coupling in situ neutron diffraction experiment and elastoplastic self-consistent (EPSC) simulation. The result suggests that Y hardens all three deformation modes, but the harden 2g twinning than for slip modes. ing effect is stronger for f1 0 1 Sandlöbes et al. [8] used transmission electron microscopy (TEM) to examine the deformation microstructure in Mg–3 wt%Y and pure Mg after cold rolling. A high density of I1 stacking faults were found, which served as nucleation sites for hc + ai dislocations. It suggests that Y reduces the stacking fault energy (SFE) in Mg, as was subsequently verified by first-principles studies [9,10]. In a more recent study, Stanford et al. [11] reported the observation 1g twins in a single-phase Mg–10 wt%Y alloy after comof f1 1 2 pression. While this twin mode has been frequently observed in other hexagonal materials such as Ti [12,13] (c/a = 1.587), Zr
⇑ Corresponding author. E-mail addresses: [email protected], [email protected] (L. Wang).
[14,15] (c/a = 1.593), Re [16] (c/a = 1.614), and Co [17] (c/ a = 1.628), it is rarely observed for Mg (c/a = 1.623). A general argument for the surprising observation of this twin in a Mg–Y alloy is 2g twin mode which leads to the that Y hardens the primary f1 0 1 1g twins, but details of the twin preferential nucleation of f1 1 2 nucleation process is still unclear because the characterization was made ex situ. Here, we present the results from an in situ straining study of a coarse-grained Mg–Y alloy, using grain-level characterization of the deformation-induced microstructural evolution. An unexpected kink band was observed to form in a grain whose c-axis was almost parallel to the tensile direction. Basal slip lines were 2g twins were identified found crossing the kink, and small f1 0 1 in the vicinity of the kink. Based on the observed microstructure evolution, a dislocation reaction model is proposed to account for the kink formation. A binary Mg–2.5 wt%Y solid solution alloy was produced by casting at the Magnesium Innovation Centre (MagIC) of Helmholtz-Zentrum Geesthacht. To achieve a homogeneous composition distribution, the melt was taken to 700 °C for 10 min before casting. A relatively small cylinder die (D = 18 mm, H = 100 mm) was used. The as-cast piece of cylinder material was cut into two half-cylinders by electrical discharge machining (EDM). Dog-bone tensile specimens with a gauge dimension of 5 mm 2 mm 1 mm (L W T) were extracted from one halfcylinder, with the tensile axis being parallel to the vertical direction of the half-cylinder. The specimen to be tested was close to the top of the half-cylinder. The surface was prepared using conventional grinding and polishing procedures. The final thickness was reduced to 0.4 mm. An EBSD scan of the initial
http://dx.doi.org/10.1016/j.scriptamat.2015.08.016 1359-6462/Ó 2015 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
Please cite this article in press as: L. Wang et al., Scripta Mater. (2015), http://dx.doi.org/10.1016/j.scriptamat.2015.08.016
2
L. Wang et al. / Scripta Materialia xxx (2015) xxx–xxx
microstructure close to the gauge center was performed using a Zeiss Gemini Ultra 55 scanning electron microscope (SEM). Grain sizes range from 100 lm to more than 1 mm. The specimen was deformed by tensile test using a MICROTEST 200N (Deben, UK Limited) module at a constant crosshead speed of 0.2 mm/min, which corresponds to a nominal strain rate of 6.67 104 s1. The microstructure evolution during the loading was monitored using an Olympus SZX10 optical microscope connected with a video recording system. Fig. 1 shows the engineering stress–strain curve of the test at a strain of about 10%, along with the microstructure evolution near the gauge center associated with points A–H labeled on the stress–strain curve. Based on the initial EBSD measurement, crystallographic orientations of four large grains (d > 500 lm) as well as the approximate grain boundaries between them were overlaid on the micrograph of the initial structure prior to loading (point A). Grain 3 (Euler angles = (181°, 83°, 5°)) is a large grain with its c-axis almost parallel to the tensile axis. This is a favorable 2g twin nucleation, and our initial focus was on condition for f1 0 1 the 1–3 grain boundary in order to observe potential slip transfer induced twin nucleation events [18]. During the loading, slip lines were first observed in Grain 1 and Grain 2, favorably oriented grains for basal and prismatic slip, respectively. At an applied stress of 70 MPa (point C), slip lines appeared inside Grain 3. Using the standard trace analysis [18], these slip lines were found to correspond to basal slip, as marked by the white dashed line in the micrograph for C. At point D, a lenticular kink band, evidenced by the bright contrast, quickly developed in Grain 3 from the region where basal slip lines had concentrated. The original basal
slip lines were redirected by about 12° relative to their initial direction at point C, suggesting an orientation rotation inside the kink band. In addition, small features show up at the upper boundary of the kink band, which are later identified as newly nucleated 0 1 2Þ ½1 0 1 1 twins. With further straining (from E–H), the kink ð1 0 1 2Þ twins were found in its vicinity, grew laterally and more ð1 and a few ð1 1 0 2Þ ½1 1 0 1 twins with dark contrast were identified in the left part of Grain 3 (point F). The video recording the microstructure evolution during loading can be found in the Supplementary material (loading.mp4). After the peak load was reached, the sample was gradually unloaded at the same displacement rate, and the unloading process was recorded in a second video (unloading.mp4). Detailed EBSD characterization was subsequently performed for the kink in Grain 3, as shown in Fig. 2. From the optical micrograph, the kink band grew across the entire Grain 3. There are also some small kinks parallel to the big kink, as evidenced by their bright contrast. The direction of basal slip lines are deviated when crossing these kinks. Furthermore, twins are observed adjacent to the kink boundary, having likely nucleated from the upper boundary of the big kink. These are marked by red arrows in the figure. According to EBSD based misorientation analysis (Fig. 2(b)), as well as twin plane trace analysis (Fig. 2(c)), they belong to the twin system 0 1 2Þ ½1 0 1 1 with a Schmid factor of 0.482. Likewise, twins that ð1 nucleated from the grain boundary between Grains 3 and 4 (Fig. 1, 1 0 2Þ ½1 1 0 1 twins with a Schmid facpoint F) were identified as ð1 tor of 0.493.
Fig. 1. Engineering stress–strain curve of the tensile test. Optical microscopy tracks the microstructure evolution. Basal slip lines first appeared in Grain 3 (C), followed by the 0 1 2Þ twins nucleated from the upper boundary of the kink. With further straining, the kink grew laterally and more (1 0 1 2Þ twins are formation of a large kink band (D). ð1 1 0 2Þ twins also developed in the left part of Grain 3 (F). found in its vicinity. A few ð1
Please cite this article in press as: L. Wang et al., Scripta Mater. (2015), http://dx.doi.org/10.1016/j.scriptamat.2015.08.016
L. Wang et al. / Scripta Materialia xxx (2015) xxx–xxx
3
0 1 2Þ twins in its vicinity as mark by red arrows. (b) EBSD inverse pole figure map of the small box region in (a) that identifies Fig. 2. (a) Optical micrograph of the kink and ð1 0 1 2Þ twins. (c) Plane trace for the six f1 0 1 2g planes. The ð1 0 1 2Þ twins are cross-verified by the trace analysis. (d) EBSD inverse pole figure map of the big box region in the ð1 0 1 2Þ and ð0 1 1 2Þ twins were found both inside and outside the kink. (e) {0 0 0 1} pole figure (a) that covers the kink. The kink has varied crystal orientations inside. Small ð1 0 1 2Þ and ð0 1 1 2Þ twins are evident. (f) Misorientation profile along the vertical arrow in (d). (For interpretation of the corresponding to the same area of (d), where ð1 references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 2(d) shows the EBSD inverse pole figure (IPF) map for the big box region in Fig. 2(a) that covers the kink band. Local crystal orientations are schematically presented as small hexagonal cells overlaid on it. Crystal orientations inside the kink varied from place 2g twins to place, but being distinct from the matrix grain. f1 0 1 were found both inside and outside the kink. By misorientation 0 1 2Þ and ð0 1 1 2Þ twins are present, with their analysis, both ð1 respective orientations indicated in Fig. 2(d). Fig. 2(e) shows the {0 0 0 1} pole figure for the same area of Fig. 2(d). Orientation 0 1 2Þ and ð0 1 1 2Þ spread due to the kink is evident. Poles from ð1 twins are also marked. Fig. 2(f) shows the misorientation profile along the vertical arrow in Fig. 2(d), calculated by the OIMTM software (EDAX, Draper, UT, USA). Crystal orientation changed continuously from the matrix into the kink and then back to the matrix on the other side. The maximum misorientation from the matrix to the center of the kink was about 18°, smaller than the theoretical 2g 86 , misorientation for all possible twin modes in Mg: f1 0 1
1g 56 , f1 0 1 3g 64 , f1 1 2 1g 35 [19]. This misorientaf1 0 1 tion analysis confirms that the kink band is not a twin. The crystal orientation of Grain 3 with respect to the loading direction results in low Schmid factor for basal slip (0.12) and 2g twinning (0.48). If the material high Schmid factor for f1 0 1 were pure Mg with a ratio of CRSStwin/CRSSbasal in the range of 1.5–2 [20], we would expect the grain to primarily deform by 2g twinning. The fact that basal slip lines appeared prior to f1 0 1 2g twins in Grain 3 during in situ loading indicates that the f1 0 1 ratio of CRSStwin/CRSSbasal is probably greater than 4 (=0.48/0.12) for this alloy. This is consistent with the conclusion made by Stanford et al. [7], who estimated the ratio to be about 5.6 in Mg–2.2 wt%Y. Nonetheless, basal slip in Grain 3 has a relatively small resolved shear stress. Development of the kink turns out to be an effective way to facilitate further deformation in this grain. The volume inside the kink is more favorably oriented for basal slip than the
Please cite this article in press as: L. Wang et al., Scripta Mater. (2015), http://dx.doi.org/10.1016/j.scriptamat.2015.08.016
4
L. Wang et al. / Scripta Materialia xxx (2015) xxx–xxx
2g twins from the kink boundaries via dissociation of basal dislocations. Fig. 3. A mechanistic model to account for the kink formation and concomitant nucleation of f1 0 1
matrix; kink boundaries also serve as heterogeneous sites for 2g twin nucleation. Taking a closer look at the video of specf1 0 1 imen loading, twin nucleation was almost concomitant with the kink propagation. Many of these twins were unable to grow substantially and were swept by the moving kink boundary during further straining. As a result, some twins are found inside the kink after the test, as shown in Fig. 2(d). On the other hand, twins in the vicinity of the kink were also likely to impede its further growth after point F (Fig. 1). A mechanistic model is proposed in Fig. 3 to explain the process of kink formation and concomitant twin nucleation. The key to the model is the dissociation of basal dislocations at the kink boundary, with a number of twinning dislocations being generated: 0
ba ¼ ba þ xbtw
ð1Þ
This reaction is partially supported by previous atomistic simulation studies [21–23], which suggest basal dislocations may disso 2g twin interface. ciate into twinning dislocations at the f1 0 1 Having described the mechanism by which this unique kink band occurs, it is now important to consider the reasons for which the kink is typically not observed in binary Mg alloys, but is uniquely observed here for the solid solution Mg–Y alloy. The viability of this kink mechanism in Mg–Y depends on two independent aspects of how Y influences the microscale deformation in 2g twinning so that Mg. First of all, Y increases the CRSS for f1 0 1 alternative deformation mechanisms, such as kinking, become necessary to accommodate c-axis extension in grains whose c-axes are close to the tensile direction. Secondly, Y is one of the most effective elements to reduce the SFE of Mg according to density function theory (DFT) studies [10]. Lowered SFE generally increases the spacing between the two partials of a full dislocation on the basal plane, which provides a favorable condition for localized dislocation reaction to occur that is necessary for kink to form. Interestingly, kinking is a major deformation mechanism in recently studied long-period stacking ordered (LPSO) phase in Mg–Zn–Y ternary alloys [24–26]. Inferred from the result in the present study, it is reasonable to believe that solute Y is also responsible for the LPSO phase to deform by kinking and increase the overall material ductility. Future design of Mg alloys may use the ability of kink formation as a possible mechanism to improve material ductility. In summary, formation of a large, unique kink was observed in a binary Mg–Y alloy during in situ tensile straining. The kink boundary served as a heterogeneous nucleation site for numerous
2g twinning events. A mechanistic model is proposed to f1 0 1 account for the kink formation and concomitant twin nucleation via dissociation of basal dislocations. This mechanism is enabled by the presence of Y. This work was funded by the Helmholtz Association. L.W. is supported by the Alexander von Humboldt Foundation. The authors would like to thank Dr. Chamini Mendis for useful discussions and the fabrication of the cast Mg–Y alloys used herein.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.scriptamat.2015. 08.016.
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26]
B.C. Suh, M.S. Shim, K.S. Shin, N.J. Kim, Scr. Mater. 84–85 (2014) 1. S.R. Agnew, O. Duygulu, Int. J. Plast. 21 (2005) 1161. J. Koike, Metall. Mater. Trans. A 36 (2005) 1689. W.B. Hutchinson, M.R. Barnett, Scr. Mater. 63 (2010) 737. X.Y. Lou, M. Li, R.K. Boger, S.R. Agnew, R.H. Wagoner, Int. J. Plast 23 (2007) 44. C.M. Byer, B. Li, B. Cao, K.T. Ramesh, Scr. Mater. 62 (2010) 536. N. Stanford, R. Cottam, B. Davis, J. Robson, Acta Mater. 78 (2014) 1. S. Sandlöbes, S. Zaefferer, I. Schestakow, S. Yi, R. Gonzalez-Martinez, Acta Mater. 59 (2011) 429. S. Sandlöbes, M. Friák, S. Zaefferer, A. Dick, S. Yi, D. Letzig, Z. Pei, L.-F. Zhu, J. Neugebauer, D. Raabe, Acta Mater. 60 (2012) 3011. S.L. Shang, W.Y. Wang, B.C. Zhou, Y. Wang, K.A. Darling, L.J. Kecskes, S.N. Mathaudhu, Z.K. Liu, Acta Mater. 67 (2014) 168. N. Stanford, R.K.W. Marceau, M.R. Barnett, Acta Mater. 82 (2015) 447. N. Bozzolo, L.S. Chan, A.D. Rollett, J. Appl. Crystallogr. 43 (2010) 596. L. Wang, R.I. Barabash, T.R. Bieler, W. Liu, P. Eisenlohr, Metall. Mater. Trans. A 44 (2013) 3664. R.E. Reed-Hill, W.H. Hartt, W.A. Slippy, Trans. AIME 242 (1968) 2211. L. Capolungo, P.E. Marshall, R.J. McCabe, I.J. Beyerlein, C.N. Tomé, Acta Mater. 59 (2009) 6047. J. Kacher, A.M. Minor, Acta Mater. 81 (2014) 1. S. Vaidya, S. Mahajan, Acta Metall. 28 (1980) 1123. L. Wang, Y. Yang, P. Eisenlohr, T.R. Bieler, M.A. Crimp, D.E. Mason, Metall. Mater. Trans. A 41 (2010) 421. N. Stanford, Philos. Mag. Lett. 88 (2008) 379. N. Stanford, M.R. Barnett, Int. J. Plast 47 (2013) 165. A. Serra, D.J. Bacon, Philos. Mag. A 73 (1996) 333. J. Wang, L. Liu, C.N. Tomé, S.X. Mao, S.K. Gong, Mater. Res. Lett. 1 (2013) 81. H. El Kadiri, C.D. Barrett, J. Wang, C.N. Tomé, Acta Mater. 85 (2015) 354. K. Hagihara, N. Yokotani, Y. Umakoshi, Intermetallics 18 (2010) 267. X.H. Shao, Z.Q. Yang, X.L. Ma, Acta Mater. 58 (2010) 4760. J.-K. Kim, S. Sandlöbes, D. Raabe, Acta Mater. 82 (2015) 414.
Please cite this article in press as: L. Wang et al., Scripta Mater. (2015), http://dx.doi.org/10.1016/j.scriptamat.2015.08.016