In-Situ Observation on the Formation Behavior of the Deformation Kink Bandsin Zn Single Crystal and LPSO Phase

Koji Hagihara1, Masahito Honnami1, Ryosuke Matsumoto2, Yoshihiro Fukusumi1, Hitoshi Izuno1,Michiaki Yamasaki3, Takuya Okamoto1, Takayoshi Nakano4 and Yoshihito Kawamura3

1Department of Adaptive Machine Systems, Graduate School of Engineering, Osaka University, Suita 565-0871, Japan2Department of Mechanical Engineering and Science, Graduate School of Engineering, Kyoto University, Kyoto 615-8246, Japan3Magnesium Research Center/Department of Materials Science, Kumamoto University, Kumamoto 860-8555, Japan4Division of Materials and Manufacturing Science, Graduate School of Engineering, Osaka University, Suita 565-0871, Japan

The deformation microstructure and deformation process of the Mg-based long period stacking ordered (LPSO) phase accompanied by theformation of deformation kink bands were examined through dynamical observations during compression tests, and the features were comparedto those in Zn single crystals. In both crystals, the formation of deformation kink bands was confirmed in specimens compressed along thedirection parallel to the basal plane. The deformation kink bands formed in the directionally solidified (DS) LPSO phase crystals and the Znsingle crystals had similar morphologies. Their formation induced plastic strain almost along the c-axis. However, their formation behaviorsshowed some different features. In the LPSO phase DS crystal, two different migration behaviors of the deformation kink band boundariesexisted. The slower migration process was comparable to that observed in Zn single crystals. However, a migration process more than 105 timesfaster than the slower process was also monitored. The results imply that two different formation mechanisms of the deformation kink bandmight exist in the LPSO phase crystal. [doi:10.2320/matertrans.MH201412]

(Received January 6, 2015; Accepted February 19, 2015; Published April 3, 2015)

Keywords: long period stacking ordered (LPSO) phase, deformation kink band, dynamical observation

1. Introduction

Mg alloys containing long-period stacking ordered phases,the so-called LPSO phase, have recently received attentionowing to their excellent mechanical properties.1­18) To clarifythe strengthening mechanisms acting in Mg/LPSO two-phase alloys, the plastic deformation behavior of the LPSOphase has been of increasing research interest.19­24) Wepreviously clarified that (0001) basal slip is the dominantlyoperative deformation mode in the LPSO phase. In addition,the formation of deformation kink band effectively accom-modates the plastic strain under the condition where theoperation of basal slip was macroscopically hindered.19)

The formation of deformation kink band is a less commondeformation mode than slip and twinning. However, it is oneof the important deformation mechanisms not only in theLPSO phase but also in other materials that show strongplastic anisotropy, such as hcp-Cd, Zn,25,26) Ti3SiC2 ce-ramics,27­29) and mica (muscovite),30,31) in which only oneslip system was predominantly operative. However, theformation mechanism of the deformation kink band has notbeen sufficiently clarified yet. One of the famous modelsexplaining the formation of deformation kink band in the hcpcrystal was proposed by Hess and Barrett for a Zn singlecrystal in 1949,26) as shown in Fig. 1. In this model,corporative initiation and/or operation, and followingarrangement of basal dislocations to be aligned perpendicularto the slip plane are thought to be the basic processes for theformation of the deformation kink band boundary. However,experimental observation and analysis of the deformationkink bands have not been adequately conducted, and thusthe detailed formation mechanism has not yet been fullycomprehended.

Based on the above background, in this study, theformation behavior of the deformation kink bands in the

LPSO phase was experimentally examined by in-situobservation during the compression tests. In addition to theLPSO phase, Zn single crystal, in which the model for theformation of the deformation kink bands shown in Fig. 1 wasproposed, was also studied. The differences and similaritiesbetween characteristic features of deformation kink formed inthem were examined, and the mechanism of formation of thedeformation kink band was discussed.

2. Experimental Procedure

Mother ingots with a composition of Mg85Zn6Y9 (at%)were prepared by an induction melting for LPSO phasealloy. Directional solidification (DS) was conducted withthe mother ingot by using the Bridgeman method (furnace:NEV-DS2, Nissin giken, Japan) at a growth rate of 10mm/hunder Ar atmosphere in a carbon crucible. The same methodwas used to grow a Zn single crystal from its raw ingots.From the obtained LPSO phase DS crystal and Zn single

Fig. 1 Schematic illustration showing the formation mechanism ofdeformation kink bands in a Zn single crystal proposed by Hess andBarrett.26)

Materials Transactions, Vol. 56, No. 7 (2015) pp. 943 to 951Special Issue on Long-Period Stacking Ordered Structure and Its Related Materials II©2015 The Japan Institute of Metals and Materials

crystal, rectangular specimens with an approximate size of2mm © 2mm © 5mm were cut by electrical dischargemachining for compression tests. The loading axis was setparallel to the growth direction in the LPSO phase crystal,and parallel to ½10�10� in the Zn single crystal. Hereafter, theyare called 0°-orientation. In both specimens, the compressivedeformation was expected to occur accompanied by theformation of deformation kink bands. Since, the operation ofbasal slip is hindered owing to the development of basal fibertexture in the LPSO phase DS crystal;19,32) the basal plane isparallel to the loading axis in many of the grains in the LPSOphase DS crystal. Compression tests were conducted at roomtemperature in air at a nominal strain rate of 1.67 © 10¹4 s¹1.During the compression test, the development of deformationmicrostructure was dynamically monitored using digitalmicroscopes: VHX-200, Keyence, Japan at a frame rate of15 s¹1, and Fastcam-SA3, Photron, Japan at a frame rate of40000 s¹1 with a zoom lens (VH-Z100R, Keyence, Japan).The deformation microstructure after the compression testwas observed by optical microscopy with Nomarskiinterference contrast (BX60M, Olympus, Japan). In someobservations, the crystal orientation in the LPSO phase DScrystal was determined by using scanning electron microsco-py equipped with back scatter electron diffraction detector(SEM-EBSD) (SEM: JEM-6500F, JEOL, Japan, EBSD: TSLsolutions K.K., Japan). The details of the specimenpreparation method for SEM-EBSD observation are referredin our previous papers.33,34)

3. Results and Discussion

3.1 Morphology of the deformation bandsFigure 2(a), (b) shows the typical stress-strain curve for Zn

single crystal and LPSO phase DS crystal deformed along the0°-loading orientation, respectively. In Fig. 2(a), (b), thestress-strain curves at the loading orientation inclined to 45°from ½10�10� along the [0001] direction in Zn single crystaland inclined to 45° to crystal growth direction in the LPSOphase DS crystal (referred to as 45°-orientation, hereafter) arealso indicated for comparison purposes. As seen in thefigures, the stress-strain curve showed significant differencesdepending on the loading orientation. In both crystals, thebasal slip was known to be the predominant slip system.19,26)

Indeed, the specimens exhibited low flow stress at 45°-orientation by easy operation of the basal slips in bothcrystals. Although a relatively high work hardening rate wasmonitored in the LPSO phase crystal compared to that in theZn single crystal, it may be owing to the DS microstructure.In the LPSO phase DS crystals the (0001) basal planes areprecisely inclined to 45° with respect to the loading axisin only some of grains,19) different from that in Zn singlecrystal. Thus, the influence of the deformation of other grainscould not be fully ruled out on the shape of the stress-straincurve after yielding. The detailed features of the deformationbehavior of the LPSO phase 45°-oriented crystal are nowunder survey with the help of computer simulation. Theresults will be described elsewhere.

On the other hand, in the deformation at 0°-orientation,both crystals showed much higher yield stress than thoseat 45°-orientation, since the operation of basal slip was

macroscopically hindered owing to their negligible Schmidfactors. Instead of the operation of basal slip, the formationof deformation bands was confirmed in them as expected,though the operation of f11�22g pyramidal slip was alsoobserved in Zn single crystal. Higher yield stress wasobserved in both crystals at 0°-orientation, but their stress-strain curve profiles were slightly different. In Zn singlecrystal, the flow stress showed a plateau after the initialstrong work-hardening and then large drop in flow stressfrequently followed at around 10% plastic strain. In theLPSO phase crystal, on the other hand, the flow stress alsoshowed a plateau after yielding but it showed a continuousincrease as the deformation proceeded. These differences inthe stress-strain curves suggest that the deformation micro-structure features developed differently in them.

Figure 3(a), (b) shows the change in the shape of the Znsingle crystal and LPSO phase DS crystal specimensdeformed to ³20% plastic strain at RT along the 0°-orientation. In this paper, the loading axis was set to beparallel to the vertical direction in all the figures. In the Znsingle crystal shown in Fig. 3(a), the observed direction wasparallel to the ½1�210�. The vertical black lines observed in thespecimen correspond to the traces of basal planes which wereintroduced by chemical etching in specimen preparationbefore compression tests. The Zn single crystal specimen didnot show a uniform deformation but it locally deformedmainly at both the ends of the specimen accompanied by theformation of deformation bands as shown in Fig. 3(a). On theother hand, in the LPSO phase DS crystal, small deformationbands were relatively homogeneously introduced in entire

(a) (b)

Fig. 2 Stress-strain curve of the (a) Zn single crystal and (b) LPSO phaseDS crystal compressed at 0°- and 45°-loading orientations at RT.

(a) (b)

Fig. 3 Appearance of ³20% plastically deformed specimens. (a) Zn singlecrystal observed on ð1�210Þ, and (b) LPSO phase DS crystal, compressedat 0°-orientations at RT.

K. Hagihara et al.944

specimen as shown in Fig. 3(b). The morphology of thedeformation bands were examined in detail with the highermagnification images. Figure 4(a), (b) shows higher magni-fication images of the deformation bands observed on ð1�210Þand (0001) side surfaces in the Zn single crystal deformed at0°-orientation to ³5% plastic strain. As seen in Fig. 4(a), (b),the contrasts of the deformation band boundaries was ratherweak when observed on ð1�210Þ compared to those on (0001)surface. Macroscopically, the deformation bands wereintroduced nearly perpendicular to the loading axis, butwhen observed on ð1�210Þ surface the formed deformationband boundaries were shown to be inclined to the loadingaxis. As shown in Fig. 4(a), the inclined angles of thedeformation bands were not constant but they were differentin each of the bands. This feature is coincident with themodel shown in Fig. 1, demonstrating that formed deforma-tion bands were not the deformation twin but predominantlydeformation kink bands as reported in 1949.26,34) In Fig. 4(a),(b), the change in the shape of specimen surface edges by theformation of deformation kink bands in the Zn single crystalcan be confirmed. In Fig. 4(a), it can be clearly seen that onthe ð1�210Þ surface large ridges were formed along the [0001]direction by the formation of deformation kink bands. While

the flatness of surface edge was almost unchanged along½�12�10� direction on (0001) surface even by the formationof deformation kink bands as shown in Fig. 4(b). Thisdemonstrates that the formation of deformation kink bands inZn single crystal brings about the plastic strain almost onlyalong the c-axis. This feature was also coincident with themodel of the deformation kink band shown in Fig. 1. Notethat although the trace of the deformation kink band wasshown to be relatively straight on the ð1�210Þ surface, it wasfrequently observed to be wavy on the (0001) surface asshown in Fig. 4(b). Further details on the 3-dimensionalmorphology of the deformation kink bands were now undersurvey using SEM-EBSD analysis, which will be describedelsewhere.

Figure 5(a) shows the higher magnification images of thedeformation bands observed in the LPSO phase DS crystaldeformed at 0°-orientation to ³5% plastic strain. Thedeformation bands formed in the LPSO phase frequentlyshowed the characteristic beak-like shapes, as previouslyreported.19,34) Macroscopically the beak-like deformationbands were introduced almost perpendicular to the basalplane, i.e. loading axis, but the deformation band boundarieswere formed to be inclined to the loading axis. The inclined

(a) (b)

Fig. 4 Higher magnification images of the deformation traces introduced in the Zn single crystal by compressive deformation at0°-orientations at RT to ³5% plastic strain, on (a) ð1�210Þ surface, (b) (0001) surface.

(a) (b) (c)

Fig. 5 Higher magnification images of the deformation traces introduced in the LPSO phase DS crystal by compressive deformation at0°-orientation at RT to ³5% plastic strain. (a) The morphology of the frequently observed deformation bands. (b) The morphology of thedeformation bands in the grain whose surface was parallel to �ð�5 �1 6 10Þ [near prism plane]. (c) The deformation band observed on the�ð5 1 �6 900Þ surface [near basal plane with the deviation angle of ³6°] in the same grain shown in Fig. 5(b).

In-Situ Observation on the Formation Behavior of the Deformation Kink Bands in Zn Single Crystal and LPSO Phase 945

angle of the deformation band boundaries was different ineach deformation band, similarly to those observed in Znsingle crystal. To examine the variations in morphology ofthe deformation bands depending on the observed plane, thecrystal orientation in the grains of the LPSO phase DS crystalwas determined by the SEM-EBSD method. Figure 5(b), (c)shows that the deformation bands observed on �ð�5 �1 6 10Þsurface that is nearly parallel to the prism plane, and on�ð5 1 �6 900Þ surface that is nearly parallel to the basal planewith deviation angle of ³6°, respectively, in the same grain.By the observation shown in Fig. 5(b), (c), the deformationbands were found to show many similar features to thoseobserved in Zn single crystal. That is, the deformation bandboundaries showed relatively straight morphology on theprism plane, while they showed wavy morphology on thebasal plane. In addition, the surface ridge brought aboutby the formation of deformation bands were significantlyappearing along the [0001] c-axis direction when observedon the prism plane as shown in Fig. 5(b), while little changein the flatness of the specimen edge was observed on basalplane surface as shown in Fig. 5(c). This indicates that theplastic deformation occurred almost along the c-axis by theformation of deformation bands. Owing to this feature, thecontrast of the deformation bands was considerably faintwhen observed on the �ð�5 �1 6 10Þ near prism plane surface asshown in Fig. 5(b). These features suggest that the formeddeformation bands in the LPSO phase DS crystal werepredominantly deformation kink bands as the same as thosein Zn single crystal, as previously discussed.19,34) However,some differences of the morphology of the deformation kinkbands were found in them, such as the size, shape anddistribution of the deformation kink bands. The size of thedeformation bands was much smaller, and they wereintroduced in the LPSO phase specimen with relativehomogeneity compared to those in the Zn single crystal. Inthe LSPO phase crystal, most of the deformation kink bandsshowed characteristic beak-like shapes. As one of the reasonsto induce such differences, instead of the nature of thematerials themselves, the influence of the microstructure inspecimens, whether single crystal or polycrystal, is consid-ered. In polycrystalline Zn crystal, the formation of thedeformation kink bands was reported to be more homoge-neous.34) To elucidate further, the help of computersimulations must be utilized in addition to the experimenta-tion. The research related to this has been conducted by somegroups.35)

3.2 Formation behavior of the deformation kink bandsin Zn single crystal and LPSO phase DS crystal

As described in the previous section, the deformation kinkbands formed in Zn single crystal and LPSO phase DS crystalshowed many similarities. However, in-situ observationrevealed that their formation behaviors were slightly differentin them. Figure 6 shows multiple snapshot images of thedeformation microstructure development process during thecompression test on the Zn single crystal along 0°-orientationat RT, monitored with VHX-200 (Keyence) digital micro-scope at a frame rate of 15 s¹1. The observed part was nearthe bottom end of the specimen on the ð1�210Þ surface. Inthe early stage of the deformation at around yielding, the

operation of the f11�22g pyramidal slip was first confirmed bydetailed observation using optical microscope, but it couldnot be clearly detected in this in-situ observation since thecontrast of their slip traces was considerably weak. Theoperation of the f11�22g slip induced slight crystal rotationand hence the operation of basal slip was locally induced asshown in Fig. 6(a). As the deformation proceeded, a faintline showing the deformation kink band boundary appearedfrom the end part of the specimen as indicated by the arrowin Fig. 6(b). The contrast of the deformation kink bandboundary gradually became to sharp and the boundary movedtoward the central region of the specimen as the deformationproceeded as shown in Fig. 6(c)­(h). The traces of basal slipwere largely bent by the deformation kink band formation,which indicates the occurrence of large crystal rotations. Thebent angle became larger from a small angle to significantlylarge angles as the deformation proceeded as shown inFig. 6(c)­(h). This is attributed to the increasing in thedensity of dislocations that form the deformation kink bandboundary and the accumulation of new deformation kinkband boundaries, as Abe et al. previously reported the relatedresults in the LPSO phase.36) The increase in the deformationkink band’s rotation angle enabled the operation of the basal

(a) (b)

(c) (d)

(e) (f)

(g) (h)

Fig. 6 Multiple snapshot images showing the deformation process of theZn single crystal observed on the ð1�210Þ surface in compression at ½10�10�0°-orientation at RT. The amount of applied plastic strain was indicated inthe figures.

K. Hagihara et al.946

slip, since the crystal rotation increased the Schmid factor ofthe basal slip from its negligible value. Indeed, strong linecontrast ascribed to the basal slip traces was confirmed in thedeformation kink bands as shown in Fig. 6(e)­(h). Thiscrystal rotation by the formation of deformation kink bandsinduced the so-called geometrical softening, i.e. the signifi-cant activation of basal slips inside the deformation kinkbands, leading to a large stress drop on the stress-strain curvein deformation of Zn single crystal at 0°-orientation as shownin Fig. 2(a). As the deformation proceeded further, manydeformation kink bands accumulated in the local area andinduced a significant deformation as shown in Fig. 6(e)­(h).In some deformation kink bands, secondary deformation kinkbands were also developed to accommodate the complicatedinternal strain field as indicated by arrows in Fig. 6(e), (g).

From the snapshot images shown in Fig. 6 and others,the average migration speed of the deformation kink bandboundaries was evaluated to be ³1.6 µm/s along the

direction perpendicular to the kink band boundary. Themigration speed of the boundary along the horizontaldirection, i.e. the elongation speed of the kink band alongits boundary, could not be precisely determined from anin-situ observation. The entire region of the kink boundarysimultaneously appeared on ð1�210Þ side surface and thefeature of the boundary to “elongate” could not be clearlydetected in the observation, as shown in Fig. 6.

Figure 7 shows multiple snapshot images of the develop-ment process of the deformation microstructure during thecompression test in the LPSO phase DS crystal along the0°-orientation at RT. By applying stress, deformation kinkbands suddenly formed in the grain in small beak-like shapes,which rapidly grow up to the grain boundary (Fig. 7(a), (b)).Some of these bands stopped at the grain boundaries, whilesome grow further beyond, as indicated by arrows inFig. 7(c). To clarify the criteria for the propagation ofdeformation kink band beyond the grain boundary, the

(a) (b)

(c) (d)

(e) (f)

(g) (h)

Fig. 7 Multiple snapshot images showing the deformation process of the LPSO phase DS crystal in compression at 0°-orientation;perpendicular to the c-axis in most of grains, at RT. The amount of applied plastic strain was indicated in the figures.

In-Situ Observation on the Formation Behavior of the Deformation Kink Bands in Zn Single Crystal and LPSO Phase 947

relation between the propagation behavior of the deformationkink bands and the geometry of grain boundary wasexamined as shown in Fig. 8. The vertical and horizontalaxis in Fig. 8 indicate the misorientation angle of the c-axesand a-axes in the adjacent two grains examined by theSEM-EBSD analysis, and whether the propagation of thedeformation kink band proceeds beyond the correspondinggrain boundary or not are indicated by white circle andblack rectangle, respectively. The results indicate that largemisorientations around the c-axis over 20­30° stronglyprohibit the propagation of deformation kink bands beyondthe grain boundary. On the other hand, under a misorientationof 20­30° with respect to the c-axis, the misorientationaround the a-axis does not strongly affect the propagationbehavior. This feature may be related to the high arbitrarinessof the deformation kink band in which the crystal rotationaxis adaptively varied on the [0001] zone axis by changingthe kind of basal dislocations that construct the deformationkink band boundary.33,34) Further study is required toconclude this.

Figure 9 shows the another example of multiple snapshotsdemonstrating the formation behavior of the deformationkink bands. It can be seen in Fig. 9(a), (b) that once adeformation kink was formed, formation of new kink bandswere frequently followed nearby as indicated by arrows. Thiswas a frequently observed feature in the LPSO phase crystal.In Fig. 9(c), it can be seen that a relatively large deformationkink band was introduced at the lower part of the imageas indicated by the arrow. Around the kink bands, theintroduction of traces of basal slip was clearly seen, causedby the local geometrical softening as observed in Zn singlecrystal. As the deformation proceeded, the bent angle of thebasal slip traces, i.e. the rotational angle of the deformationkink band increased, as similarly to those observed in Znsingle crystal. In Fig. 9(d), a large deformation kink bandwas formed in the right-side grain. Interestingly, from theboundaries of the deformation kink band, new boundaries

were continuously generated and many deformation kinkbands accumulated in the grain as shown in Fig. 9(f )­(h). Bytheir formations, the plate-like shape of the LPSO phase grainwas largely bent at the bottom of the large kink bands, sincethe formation of deformation kink band brings about the largeplastic strain almost along the c-axis as described in previoussection. It is noteworthy that even after large deformation,microcracks were not frequently initiated around the bentgrains, and the generated strain was effectively accommo-dated by the plastic deformation in the LPSO phase grains,i.e. by the bending of the grain and/or the formation of somenew deformation kink bands.

As described above, it was found that deformation kinkbands formed in Zn single crystal and LPSO phase DS crystalshow many similarities such as their geometrical morphol-ogies. However, differences were observed on the formationbehavior of the deformation kink bands in the LPSO phaseDS crystal compared to those in the Zn single crystal. Asshown in Fig. 7(a)­(c) and Fig. 9(a)­(c), many of thedeformation kink bands suddenly form within the 1/15 s,i.e. faster than the frame rate limit for dynamical observationusing VHX-200 (Keyence) digital microscope, and once theirformation was completed their growth rarely proceededagain. However, some other deformation kink bands thatshowed much slower formation speed were also observed. Insuch bands, the movement of kink band boundary alongthe lateral and perpendicular directions with respect to theboundary was sometimes detected. An example of suchdeformation kink bands were seen in Fig. 9(d)­(f ). Asdescribed above, a large deformation kink band formed in theright-side grain, and it gradually expanded as the deformationproceeded. In Fig. 9(e), (f ), the position of the deformationkink band boundaries observed in Fig. 9(d) was inserted asdot lines. By comparing the boundary position, the slowmigration of the kink band boundaries can be recognized.From the snapshot images, the average migration speed of thedeformation kink boundary along the direction perpendicularto the kink band boundary, Vn, was estimated to be ³1.9µm/s; i.e. the migration speed was similar to that observed inZn single crystal. In addition, the average migration speed ofthe boundary along its parallel direction, Vs, was measured tobe ³10.7 µm/s, although the speed showed large scatters ineach band. That is, the migration speed of the deformationkink boundary was found to show anisotropy between thedirections parallel and perpendicular to the boundary. Thelarge scatter detected in the value of Vs might suggest that themigration rate shows anisotropy also in-plane direction on thedeformation kink band boundary. Further study is required toclarify such details. However, at least the obtained resultssuggest that the migration speed of the “slower type” ofdeformation kink band was rather comparable to thoseobserved in Zn single crystal.

To clarify the migration speed of the “faster type” of thedeformation kink bands, the other digital microscope, i.e. theFastcam-SA3 (Photron) that can capture the 40000 frameper second was prepared, and in-situ observation was againconducted. Figure 10 shows the multiple snapshot imagesshowing the growth behavior of the “faster type” ofdeformation kink bands. Figure 10(a)­(c) shows the exper-imentally obtained snapshot images and Fig. 10(d), (e)

Fig. 8 Relation between the geometry of the grain boundary and thepropagation of deformation kink bands beyond the grain boundary. Thevertical axis and horizontal axis indicate the angle of the c-axis and a-axisin the adjacent two grains examined by the SEM-EBSD analysis, and theoccurrences of the propagation of the deformation kink band beyond thecorresponding grain boundary or not are indicted by white circle andblack rectangle, respectively.

K. Hagihara et al.948

display the explanation on the images. As shown inFig. 10(b), (e), a new deformation kink band was rapidlyformed nearby the pre-existing deformation kink band withinthe 1/40000 s. In the next frame (Fig. 10(c)), the contrast ofthe newly formed deformation kink band boundary becamestronger and it further extended as compared to the initialposition shown in Fig. 10(b). Owing to their faint contrasts,precise evaluation of the migration speed of the “faster type”boundary was still difficult, but if assuming that theirformation and elongation process occurred during eachobserved frame indicated in Fig. 10(d), (e), the migrationspeed could be roughly estimated to be ³4.6m/s along thehorizontal direction and ³0.6m/s along vertical directionwith respect to the deformation kink band boundary,respectively.

To summarize, the migration speed of the deformationkink band boundary evaluated in this study was shownin Fig. 11. In the LPSO phase DS crystal, two types offormation behaviors of the deformation kink band wereobserved. The migration speed of the “faster type”deformation kink band was more than 105 times greater thanthose of the “slower type” of deformation kink band,although significant differences were not observed in theirappearance. The reason behind such an extremely largedifference in the formation behavior has not been clarifiedyet, but the result might imply that two different mechanismsfor the formation of deformation kink band exist in the LPSOphase. In the recent computer analysis using atomisticsimulation, the possibility of the existence of some differentformation mechanisms (processes) of the deformation kink

(a) (b)

(c) (d)

(e) (f)

(g) (h)

Fig. 9 Another example of the multiple snapshot images showing the deformation process of the LPSO phase DS crystal in compressionat 0°-orientation at RT. The amount of applied plastic strain was indicated in the figures.

In-Situ Observation on the Formation Behavior of the Deformation Kink Bands in Zn Single Crystal and LPSO Phase 949

band has been suggested.37,38) As one of the suggestedmechanisms, the local operation of pyramidal dislocationplays a role to induce the formation of pair of basaldislocations, and it leads to the development of deformationkink bands. In this case, since the motion of “one” pyramidaldislocation governs the formation behavior of the deforma-tion kink band, the formation speed of the deformation kinkband is expected to be much higher than that formed bythe so-called “Hess-Barrett mechanism”.26) Further study isrequired to clarify the formation and developing mechanismsof the deformation kink band in the LPSO phase.

4. Conclusions

(1) The deformation kink bands formed in the Zn single

crystals and LPSO phase DS crystals showed manysimilarities in their morphologies. Their formationinduced plastic strain almost along the c-axis.

(2) Differences were observed between the deformationmicrostructure, such as large deformation kink bandsbeing locally formed especially at the ends of thespecimen in Zn single crystals, while many beak-likeshapes of the deformation kink bands being entirelyintroduced after a large deformation in the LPSO phasecrystals. The sizes of deformation kink bands wererather small in the LPSO phase DS crystal since thegrain boundary frequently hinders the growth ofdeformation kink bands.

(3) The migration speed of the deformation kink bandboundary was estimated by in-situ observation. It wasfound that in the LPSO phase DS crystal two differentmigration behaviors exist. The slower migration processis comparable to that observed in Zn single crystal, butmore than 105 times faster migration process was alsoconfirmed. Further studies are required for under-standing the formation mechanism of the deformationkink band in the LPSO phase.

Acknowledgement

This work was supported by a grant-in-aid for ScientificResearch on Innovative Areas (Project: “Materials Scienceon Synchronized LPSO Structure³The Evolution of theMaterial Science for Innovative Development of the Next-generation Lightweight Structure Materials³”) from theMinistry of Education, Culture, Sports, Science and Tech-nology of Japan. This work was also supported by“Innovative Design and Production Technology Project”under the Strategic Innovation Promotion (SIP) programfrom the New Energy and Industrial Technology Develop-ment Organization (NEDO).

REFERENCES

1) Y. Kawamura, K. Hayashi, A. Inoue and T. Masumoto: Mater. Trans.42 (2001) 1172­1176.

2) M. Yamasaki, T. Anan, S. Yoshimoto and Y. Kawamura: Scr. Mater. 53(2005) 799­803.

3) M. Matsuda, S. Ando and M. Nishida: Mater. Trans. 46 (2005) 361­364.

4) S. Yoshimoto, M. Yamasaki and Y. Kawamura: Mater. Trans. 47 (2006)959­965.

5) Y. Kawamura and M. Yamasaki: Mater. Trans. 48 (2007) 2986­2992.6) M. Yamasaki, M. Sasaki, M. Nishijima, K. Hiraga and Y. Kawamura:

Acta Mater. 55 (2007) 6798­6805.7) T. Itoi, K. Takahashi, H. Moriyama and M. Hirohashi: Scr. Mater. 59

(2008) 1155­1158.8) K. Hagihara, A. Kinoshita, Y. Sugino, M. Yamasaki, Y. Kawamura,

H. Y. Yasuda and Y. Umakoshi: Acta Mater. 58 (2010) 6282­6293.9) K. Hagihara, A. Kinoshita, Y. Sugino, M. Yamasaki, Y. Kawamura,

H. Y. Yasuda and Y. Umakoshi: Trans. Nonferrous Met. Soc. China 20(2010) 1259­1268.

10) M. Yamasaki, K. Hashimoto, K. Hagihara and Y. Kawamura: ActaMater. 59 (2011) 3646­3658.

11) E. Oñorbe, G. Garcés, P. Pérez, S. Cabezas, M. Klaus, C. Genzel, E.Frutos and P. Adeva: Scr. Mater. 65 (2011) 719­722.

12) E. Oñorbe, G. Garcés, P. Pérez and P. Adeva: J. Mater. Sci. 47 (2012)1085­1093.

13) J. Wang, P. Song, X. Zhou, X. Huang and F. Pan: Mater. Sci. Eng. A

(a)

(b)

(c)

(d)

(e)

Fig. 10 (a)­(c) Multiple snapshot images showing the deformation processof the LPSO phase DS crystal in compression at 0°-orientation monitoredwith ultra-high speed digital microscope. (d), (e) Explanation on theposition of the deformation kink band boundaries observed in Fig. 10(b),(c) by dot lines. The interval between the Figs. 10(a)­(c) were the 2.5 ©10¹5 s (frame rate: 40000 s¹1).

Fig. 11 Migration speed of the deformation kink band boundary evaluatedby in-situ observations. In the LPSO phase DS crystal, two differentmigration behaviors of “slower” and “faster” processes were observed.

K. Hagihara et al.950

556 (2012) 68­75.14) K. Hagihara, A. Kinoshita, Y. Fukusumi, M. Yamasaki and Y.

Kawamura: Mater. Sci. Eng. A 560 (2013) 71­79.15) E. Oñorbe, G. Garcés, F. Dobes, P. Pérez and P. Adeva: Metall. Mater.

Trans. A 44 (2013) 2869­2883.16) L. B. Tong, X. H. Li and H. J. Zhang: Mater. Sci. Eng. A 563 (2013)

177­183.17) H. Liu, F. Xue, J. Bai and Y. Sun: Mater. Sci. Eng. A 585 (2013) 261­

267.18) T. Itoi, T. Inazawa, M. Yamasaki, Y. Kawamura and M. Hirohashi:

Mater. Sci. Eng. A 560 (2013) 216­223.19) K. Hagihara, N. Yokotani and Y. Umakoshi: Intermetallics 18 (2010)

267­276.20) K. Hagihara, A. Kinoshita, Y. Sugino, M. Yamasaki, Y. Kawamura,

H. Y. Yasuda and Y. Umakoshi: Intermetallics 18 (2010) 1079­1085.21) X. H. Shao, Z. Q. Yang and X. L. Ma: Acta Mater. 58 (2010) 4760­

4771.22) K. Hagihara, Y. Sugino, Y. Fukusumi, Y. Umakoshi and T. Nakano:

Mater. Trans. 52 (2011) 1096­1103.23) K. Hagihara, Y. Fukusumi, M. Yamasaki, T. Nakano and Y. Kawamura:

Mater. Trans. 54 (2013) 693­697.24) A. Inoue, K. Kishida, H. Inui and K. Hagihara: MRS Symp. Proc. 1516

(2013) 151­156.25) E. Orowan: Nature 149 (1942) 643­644.

26) J. B. Hess and C. S. Barrett: Met. Trans. 185 (1949) 599­606.27) M. W. Barsoum and T. El-Raghy: Met. Mater. Trans. A 30 (1999) 363­

369.28) M. W. Barsoum, L. Farber and T. El-Raghy: Met. Mater. Trans. A 30

(1999) 1727­1738.29) F. Barcelo, S. Doriot, T. Cozzika, M. Le Flem, J. L. Béchade, M.

Radovic and M. W. Barsoum: J. Alloy. Compd. 488 (2009) 181­189.30) S. Basu, A. Zhou and M. W. Barsoum: J. Structural Geology 31 (2009)

791­801.31) S. Misra and J. P. Burg: Tectonophysics 548­549 (2012) 22­33.32) M. Tane, Y. Nagai, H. Kimizuka, K. Hagihara and Y. Kawamura: Acta

Mater. 61 (2013) 6338­6351.33) M. Yamasaki, K. Hagihara, S. Inoue, J. P. Hadorn and K. Kawamura:

Acta Mater. 61 (2013) 2065­2076.34) K. Hagihara, M. Yamasaki, M. Honnami, H. Izuno, M. Tane, T. Nakano

and Y. Kawamura: Philos. Mag. 95 (2015) 132­157.35) R. Ueta and K. Shizawa: Key Eng. Mater. 626 (2015) 281­286.36) D. Egusa, M. Yamasaki, Y. Kawamura and E. Abe: Mater Trans. 54

(2013) 698­702.37) R. Matsumoto, M. Uranagase and N. Miyazaki: Mater. Trans. 54

(2013) 686­692.38) R. Matsumoto and M. Uranagase: Mater. Trans. 56 (2015) doi:

10.2320/matertrans.MH201408.

In-Situ Observation on the Formation Behavior of the Deformation Kink Bands in Zn Single Crystal and LPSO Phase 951