Deformation Behavior of Binary Mg-Y Alloy Under DynamicCompression Loading

MASAKI NAGAO,1 TOMOFUMI TERADA,1 HIDETOSHI SOMEKAWA,2

ALOK SINGH,2 and TOSHIJI MUKAI1,3

1.—Department of Mechanical Engineering, Graduate School of Engineering, Kobe University,Kobe, Hyogo 657-8501, Japan. 2.—National Institute for Materials Science, Sengen, Tsukuba,Ibaraki 305-0047, Japan. 3.—e-mail: mukai@mech.kobe-u.ac.jp

This study examines Mg-Y binary alloys at a high strain rate of approximately1 9 103 s�1 in compression by using a split Hopkinson pressure bar to eluci-date the effect of yttrium in magnesium on mechanical anisotropy and otherproperties. As a result of high strain rate compression, Mg-0.6 at.%Y alloyshowed less mechanical anisotropy, a lower strain hardening rate, and alarger compressive strain to failure of approximately 0.4, as compared withpure magnesium. Microstructure analysis by scanning electron microscopy/electron backscatter diffraction revealed that the addition of yttrium couldrelease the stress concentration at the interface between the matrix and the{10�12} c-axis tension twins by the formation of subgrains and lattice rotationaround the c-axis during dynamic compression.

INTRODUCTION

Weight reduction of automobiles and aircraft im-proves fuel economy and reduces greenhouse gasemissions. The use of Mg alloys may allow weightreduction because of their low densities,1 but adop-tion is hindered because they exhibit limited duc-tility at ambient temperatures.2 In a previous studyof fracture toughness in Mg alloy, a crack readilypropagated near twin boundaries and resulted inpoor ductility/toughness.3 It has been shown that apile-up of dislocations at the interface between thematrix and deformation twins caused stress con-centration at cracks.3 Another study suggested thatthe ductility of Mg alloys is further limited underdynamic loading due to lowered activity of disloca-tions.4 It has also been reported that a Mg-Al-Mnalloy had pronounced mechanical anisotropy at highstrain rates of around 1.0 9 103 s�1.5 Therefore, themechanical properties of Mg alloys should be eval-uated accurately for applications involving possibledynamic loading.

It has been reported that the addition of rare-earth elements such as yttrium may improvemechanical properties over conventional Mg-Al-Znalloys by minimizing intensity of the basal tex-ture.6–12 However, the deformation behavior of Mgalloys containing yttrium is not sufficiently under-

stood at strain rates exceeding 1 9 103 s�1. Thisstudy examines a binary Mg-Y alloy at high strainrates in compression to clarify the effect of yttriumaddition on mechanical properties.

MATERIALS AND EXPERIMENTALPROCEDURE

This study investigates a binary Mg-0.6 at.%Yalloy. A billet was fabricated by gravity casting intoa steel mold. The cast binary alloy was homogenizedat 773 K for 24 h, followed by quenching in water. Itwas then extruded at 673 K at a 25:1 ratio. Theextruded rod having a diameter of 8 mm was an-nealed at 673 K for 16.5 h to form an equiaxed grainstructure with an average grain size of 42 lm. As areference material, extruded magnesium with99.95% purity and an average grain size of 50 lmwas prepared. Figure 1 shows grain structures forthe examined Mg-0.6 at.%Y alloy (hereafter desig-nated as Mg-0.6Y) and commercial-grade puremagnesium. Cylindrical specimens having a heightof 8 mm and a diameter of 4 mm were machinedparallel to the extrusion direction (ED). To examinethe mechanical anisotropy, cylindrical specimens ofheight 6.5 mm and diameter 4 mm were also pre-pared along two directions, parallel and perpendic-ular to the ED.

JOM, Vol. 66, No. 2, 2014

DOI: 10.1007/s11837-013-0854-2� 2014 The Minerals, Metals & Materials Society

(Published online January 3, 2014) 305

High-strain-rate compression tests were per-formed by using a split Hopkinson pressure bar(SHPB) at ambient temperature (298 K). In theSHPB setup used in this study, the specimen issandwiched between the incident bar (diameter:16 mm; length: 1300 mm) and the transmission bar(diameter: 16 mm; length: 1050 mm). These barsare made of a tool steel, SKD11, for which theYoung’s modulus, density, and sound wave velocityare 205 GPa, 8144 kg/m3, and 5017 m/s, respec-tively. The contact end of the specimen/pressure barwas covered with a silicone grease to minimizefriction. A high-pressure chamber was used tolaunch a striker bar, which impacted one end of theincident bar and generated an elastic wave pulse.The specimen was loaded when the stress wavereached the contact end with the incident bar.Strain gauges were bonded on the surface of theincident and transmission bars to detect the prop-agation of stress waves. Because the stress equilib-rium in the specimen was confirmed by themeasured stress waves, one wave analysis wasconducted in this study.13 Strain rate _e tð Þ, stress

r(t), and strain e(t) were calculated using the fol-lowing equations:

_e tð Þ ¼ 2cber tð Þl0

(1)

r tð Þ ¼ AEet tð ÞA0

(2)

e tð Þ ¼Zt

0

_e sð Þds; (3)

where A and E are the cross-sectional area andYoung’s modulus of the bar materials, respectively;and cb is the speed of sound in the bar; l0 and A0 arethe length and cross-sectional area of the specimen,respectively; er(t) and et(t) are the reflected andtransmitted axial strain pulse, respectively; and s isthe time variable used for integration. The average

Fig. 1. Microstructure of (a) Mg-0.6Y, (b) pure magnesium (average grain size: 50 lm), and (c) pure magnesium (average grain size: 125 lm)extrusions perpendicular to the extrusion direction, obtained by optical microscopy (OM). Micrographs of (a) and (b) were observed by OM, and(c) was observed by SEM.

Fig. 2. Appearance of a Mg-0.6Y specimen (a) before loading and (b) loaded at a nominal strain of 0.27 under a high-strain-rate compression.

Nagao, Terada, Somekawa, Singh, and Mukai306

strain rate was measured to be 1.4 9 103 s�1 for aspecimen of 8.0 mm height and to be 1.7 9 103 s�1

for a specimen of 6.5 mm height. The shape of thespecimen was detected as a series of images by anultrahigh-speed camera (Shimadzu HPV-1; Shima-dzu Corporation, Kyoto, Japan) with a samplingtime of 4 ls.

RESULTS AND DISCUSSION

Figure 2 shows the appearance of the compressedspecimen. The image, captured by high-speed video,suggests that Mg-0.6Y can be compressed over anominal strain of 0.27 with little appearance of

barreling from friction at the contact with thepressure bars. A similar tendency was also observedin pure magnesium. Thus, the nominal stress–nominal strain relation detected by the SHPB wasconverted into a true stress–true strain relation byassuming constant volume during the compression.

Figure 3 shows true stress–true strain curves forthe Mg-0.6Y alloy and pure magnesium at a strainrate of 1.3 9 103 s�1 at a temperature of 298 K. Thestress–strain curves for pure magnesium exhibit aconcave shape and increased gradient with grainsize due to c-axis deformation twinning.14–17 Thestress–strain curves show that the Mg-0.6Y alloyhas a reduced strain hardening rate, a larger com-pressive strain to failure of around 0.4, and lessmechanical anisotropy than pure magnesium. Fig-ure 4 shows true stress–true strain curves for theMg-0.6Y alloy and pure magnesium having theaverage grain size of 50 lm at a strain rate of1.7 9 103 s�1 at a temperature of 298 K for the EDand the transverse direction (TD). Strain hardeningexponents, n = dlogr/loge, of Mg-0.6Y at a strain of0.055 are calculated to be 0.51 and 0.53 for the EDand TD, respectively. On the contrary, strain-hardening exponents of the pure magnesium at thefixed strain are 0.89 and 0.74, respectively. Thestrain-hardening exponents of pure magnesium,especially those for the ED direction, increasedrapidly with increasing strain. This is likely due to(I) the dominant deformation mechanism of {10�12}twins leading to a little work hardening at an earlystage of deformation and (II) the basal slip in thematrix and twins inducing an interaction aroundthe matrix/twin boundary, which causes a stressconcentration due to dislocation pile-up between thematrix grains and the {10�12} twins.3,18 However,the work-hardening exponents of the Mg-0.6Y alloy,

0

50

100

150

200

250

300

350

400

0 0.1 0.2 0.3 0.4

Φ4, h8400

350

300

250

200

150

100

50

00 0.1 0.2 0.3 0.4 0.5

Mg-0.6Y alloy

pure magnesium 125µm

pure magnesium 50µm

Tru

e st

ress

, MP

a

True strain

Fig. 3. True stress and true strain relations in compression for Mg-0.6Y alloy and pure magnesium with two different average grainsizes at a dynamic strain rate of 1.3 9 103 s�1. Axial directions of thecompression specimen are parallel to the extrusion direction.

Fig. 4. True stress and true strain relations in compression for (a) Mg-0.6Y alloy and (b) pure magnesium at a dynamic strain rate of1.7 9 103 s�1. Axial directions of the compression specimen are denoted by ED (parallel) and TD (transverse) with reference to the extrusiondirection.

Deformation Behavior of Binary Mg-Y Alloy Under Dynamic Compression Loading 307

independent of compression direction, had lesssensitivity to strain. This may be explained by thefact that addition of yttrium into magnesiumreduces the intensity of basal texture.

Figure 5a and b show the deformed microstruc-ture of pure magnesium (d: 125 lm) at a true strainof 0.059 and Mg-0.6Y at a true strain of 0.045 asanalyzed by scanning electron microscopy (SEM)/

50µm50µm

Compressive

(a) (b)

direction

Fig. 5. Deformed microstructure of (a) pure magnesium, average grain size 125 lm, and (b) Mg-0.6Y at true strain of 0.045 and 0.059,respectively.

50µm

A

50µm

50µm

Compressivedirection

{1012}twin-

µm(a)

50µm

(b)

(c) (d)

Fig. 6. Deformed microstructure of Mg-0.6Y at true strain of (a) 0.0, (b) 0.03, (c) 0.23, and (d) 0.41, observed by SEM/EBSD.

Nagao, Terada, Somekawa, Singh, and Mukai308

electron backscatter diffraction (EBSD). The in-verse pole figure (IPF) image in Fig. 5 revealed that{10�12} twins were formed in both materials atan early stage of deformation. The fraction ofdeformation twins in the pure magnesium is noted

to be larger than that in Mg-0.6Y, which supportsthe hypothesis of the strain-hardening behaviordescribed above. The IPF images in Fig. 6c and dsuggest that no crack is initiated near the twinboundaries, unlike in a conventional Mg alloy ofAZ31.3 The grains colored in red convert to those ingreen or blue, which suggests the grain orientednear [0001] is rotated by the c-axis tension twin-ning. Instead of crack initiation, fine grains orsubgrains less than 20 lm are formed in the de-formed specimen beyond a nominal strain of 0.23. Asimilar tendency of the dynamic recrystallizationhas been reported for Mg-Y binary alloys under aquasi-static loading.19 Another interesting featurecan be seen in several grains having gradation inthe IPF map at true strains of 0.23 and 0.41. Fig-ure 7 shows a magnified image of the Mg-0.6Y alloyat a true strain of 0.41, analyzed by a kernel aver-age misorientation (KAM) map (Fig. 7a) and IPF(Fig. 7b). The KAM is indicative of average localizedstrain. The KAM map suggests that the strain washighly localized around grain boundaries. The IPFimage of the center grain in Fig. 7b indicates adefinite color gradation. Therefore, variations of themisorientation angle were measured as shown inFig. 8 in the grain along the black arrow A inFig. 6d and along arrow B in Fig. 7. According toFig. 8a, the misorientation angle varies at intervalsof several micrometers, which reveals that subgrainstructure is formed in the Mg-0.6Y alloy. This sug-gests that the Mg-0.6Y alloy has recovery mecha-nisms that may assist in releasing the stressconcentrations which develop during deformation.In contrast, the variation in misorientation alongarrow B increased gradually, and the difference at12 lm from the origin of the arrow is measured to beapproximately 25�. Hexagonal column images byEBSD in Fig. 7b clearly indicate c-axis rotation ofthe hexagonal lattice along arrow B. The KAM mapin Fig. 7a suggests that the localized strain is notvery high in this region. Figure 9a shows a latticeimage obtained by high-resolution transmissionelectron microscopy (HRTEM). This micrograph is

Fig. 7. Magnified image of Mg-0.6Y alloy at true strain of 0.41analyzed by (a) kernel average misorientation (KAM) and (b) inversepole figure (IPF). Hexagonal column representations in (b) clearlyindicate c-axis rotation of the hexagonal lattice along the arrow B.

(b)

point-to-pointpoint-to-origin

point-to-pointpoint-to-origin

Distance from the origin, μm

Mis

orie

ntat

ion

angl

e, d

egre

e

Distance from the origin, μm

Mis

orie

ntat

ion

angl

e, d

egre

e

(a)

Fig. 8. Variations in misorientation angle of the grain along the black arrow A in Fig. 6d and arrow B in Fig. 7b.

Deformation Behavior of Binary Mg-Y Alloy Under Dynamic Compression Loading 309

recorded along a h11�20i zone axis, showing basalplane and a first-order {10�10} prismatic plane edge.The larger spacing planes are basal, nearly hori-zontal in the micrograph. Local strain variations areevident at first by variation of contrast overall.Looking more closely, the lattice contrast showsunevenness too, which is indicative of the presenceof dislocations. The white ovals indicate regions oflocalized strain that contains dislocation cores. Thebinary Mg-0.6Y alloy was annealed at 673 K, but nosecond phase was found by transmission electronmicroscopy (TEM) investigations. The distributionof yttrium (points in the map) in Mg-0.6Y detectedby a three-dimensional atom probe (3DAP) is shownin Fig. 9b (top). The contour map (bottom) shown asa gray color zone indicates a dense concentrationregion of yttrium over 1.0 at.% in Mg-0.6Y. Inter-estingly, spacing between dense pockets of yttriumatoms in the annealed alloy corresponds to thespacing of the ovals in Fig. 9a. Thus, the yttriumdistribution seems to correlate with the dislocationspacing. This result is similar to that obtained byStanford et al.20 in their study of a Mg-Gd alloy. Itappears that the large rare-earth alloying elementsare strongly attracted to dislocations, which wouldcertainly induce additional resistance against dis-location motion. First-principles calculations havedemonstrated that yttrium addition could reducethe critical resolved shear stress for nonbasalslip.21,22 It has been also reported that the additionof yttrium enhances the activity of cross slip.23,24 Ithas been proposed that a dislocation cell structurewith c-axis rotation could form by the operation oftwo prismatic slip systems at once during deforma-tion.25 Another possible mechanism of the largec-axis rotation is screw dislocations aligned in a

basal plane that provide a small axial rotation be-tween neighboring parallel planes. A similararrangement of screw dislocations in the neighborplane provides additional axial rotation around thec-axis. Further microstructure analysis with TEM isunderway to clarify the dislocation structure. Insummary, subgrain formation or lattice rotationaround the c-axis may reduce localized strain, pre-venting the initiation and propagation of cracksunder dynamic loading by dense localization of yt-trium in the matrix.

SUMMARY

By investigating a Mg-0.6 at.%Y alloy and com-paring with a commercial purity magnesium sam-ple, we have found that yttrium solute contributedto enhanced compressive ductility, reduced strain-hardening rate, and minimized deformation asym-metry in magnesium, even under dynamic loading.Although {10�12} twins were the predominantdeformation mechanism for the Mg-0.6Y alloy in theearly stage of deformation, cracks did not initiatenear twin boundaries, unlike in the conventionalMg alloy AZ31. This suggests stress relaxationduring deformation. Subgrain formation and c-axisrotation possibly release stress concentrations dur-ing high-strain-rate deformation in Mg-Y alloy.

ACKNOWLEDGEMENTS

This work was supported in part by Toyota MotorCorporation and by a Grant-in-Aid for ScientificResearch (No. 25246012) from the Ministry ofEducation, Culture, Sports, Science and Technologyof Japan. The author (T. M.) thanks Profs. KazuhiroHono and Tadakatsu Ohkubo at National Institutefor Materials Science, Japan, for the use of 3DAP.

Fig. 9. (a) Lattice image in Mg-0.6Y inspected by high-resolution transmission electron microscopy (HRTEM). White ovals indicate the disorderin the lattice in (a). (b) Distribution of yttrium (blue point in the above map) in Mg-0.6Y detected by a three-dimensional atom probe (3DAP). In thecontour map, gray color zones indicate dense concentration regions of yttrium over 1.0 at.% of Mg-0.6Y (Color figure online).

Nagao, Terada, Somekawa, Singh, and Mukai310

REFERENCES

1. K. Halada and K. Ijima, Mater. Jpn. 43, 264 (2004).2. T. Mukai, H. Watanabe, and K. Higashi, Mater. Sci. Tech-

nol. 16, 1314 (2000).3. H. Somekawa, A. Singh, and T. Mukai, Philos. Mag. Lett. 89,

2 (2009).4. T. Mukai, M. Yamanoi, H. Watanabe, and K. Higashi, Ma-

ter. Trans. 42, 1177 (2001).5. P. Mao, Z. Liu, and C. Wang, Mater. Sci. Eng. A 539, 13

(2012).6. E.A. Ball and P.B. Prangnell, Scr. Metall. Mater. 31, 111

(1994).7. S.R. Agnew, M.H. Yoo, and C.N. Tome, Acta Mater. 49, 4277

(2001).8. S. Miura, S. Imagawa, T. Toyoda, K. Ohkubo, and T. Mohri,

Mater. Trans. 49, 952 (2008).9. R. Ninomiya, H. Yukawa, M. Morinaga, and K. Kubota, J.

Alloys Compd. 215, 315 (1994).10. S. Miura, S. Yamamoto, K. Ohkubo, and T. Mohri, Mater.

Sci. Forum 350, 183 (2000).11. J. Bohlen, M.R. Nurnberg, J.W. Senn, D. Letzing, and S.R.

Agnew, Acta Mater. 55, 2101 (2007).12. J. Geng, Y.B. Chun, N. Stanford, C.H.J. Davies, J.F. Nie,

and M.R. Barneet, Mater. Sci. Eng. A 528, 3659 (2011).13. G.T. Gray, ASM Metals Handbook, 8th ed. (Materials Park,

OH: ASM International, 2000), pp. 462–476.

14. M.R. Barnett, Z. Keshavarz, A.G. Beer, and D. Atweel, ActaMater. 52, 5093 (2004).

15. J. Koike, Metall. Mater. Trans. A 36A, 1689 (2005).16. D. Ando and J. Koike, J. Jpn. Inst. Met. 71, 684 (2007).17. V. Livescu, C.M. Cady, E.K. Cerreta, B.L. Henrie, and G.T.

Gray III, Magnesium Technology 2006, ed. A.A. Luo, N.R.Neelameggham, and R.S. Beals (New York: Wiley, 2006), pp.153–158.

18. M. Tsushida, K. Shikada, H. Kitahara, S. Ando, and H.Tonda, Mater. Trans. 49, 1157 (2008).

19. H. Somekawa, Y. Osawa, A. Singh, K. Washio, A. Kato, andT. Mukai, Mater. Trans. 55, 182 (2014).

20. N. Stanford, I. Sabirov, G. Sha, A. La Fontaine, S.P. Ringer,and M.R. Barnett, Metall. Mater. Trans. A 41, 734 (2010).

21. J.A. Yasi, L.G. Hector, and D.R. Trinkle, Acta Mater. 58,5704 (2010).

22. T. Tsuru, Y. Udagawa, M. Yamaguchi, M. Itakura, H. Ka-buraki, and Y. Kaji, J. Phys. Condens. Mater. 25, 022202(2013).

23. S. Sandlobes, S. Zaefferer, I. Schestakow, S. Yi, and R.G.Martinez, Acta Mater. 59, 429 (2011).

24. S. Sandlobes, M. Friak, S. Zaefferer, A. Dick, S. Yi, D. Let-zing, Z. Pei, J.F. Zhu, J. Neugebauer, and D. Raabe, ActaMater. 60, 3011 (2012).

25. J.P. Hadorn, K. Hantzsche, S. Yi, J. Bohlen, D. Letzig, J.A.Wollmershauser, and S.R. Agnew, Metall. Mater. Trans.43A, 1347 (2012).

Deformation Behavior of Binary Mg-Y Alloy Under Dynamic Compression Loading 311