Cyclic Deformation Behavior of Ultra Fine Grained Copper Produced by Equal Channel Angular Pressing

Cyclic Deformation Behavior of Ultra-Fine Grained Copper

Produced by Equal Channel Angular Pressing

Yoji Furukawa

_{, Toshiyuki Fujii}

_{, Susumu Onaka}

_{and Masaharu Kato}

1

Department of Materials Science and Engineering, Tokyo Institute of Technology, Yokohama 226-8502, Japan

2_{Department of Innovative and Engineered Materials, Tokyo Institute of Technology, Yokohama 226-8502, Japan}

Cyclic deformation behavior of ultra-fine grained (UFG) Cu of 99.99% purity processed by equal-channel angular pressing was investigated. In tension-compression fatigue tests under strain control, UFG Cu showed cyclic softening. Shear bands were formed along the direction inclined by about 45_{from the loading axis. Observations using an electron backscattering diffraction technique and transmission} electron microscopy revealed that local grain growth took place in the shear bands and overall grains elongated along the shear direction. Cyclic softening can be understood as a result of dynamic grain coarsening occurred intensively in the strain localized shear bands.

[doi:10.2320/matertrans.MD200806]

(Received July 14, 2008; Accepted September 2, 2008; Published October 16, 2008)

Keywords: ultra-fine grained copper, equal channel angular pressing, cyclic deformation, cyclic softening, shear bands, grain coarsening

1. Introduction

In recent years, much attention has been given to the mechanical properties of ultra-fine grained (UFG) materials produced by severe plastic deformation (SPD) techniques such as equal channel angular pressing (ECAP),1–4) accu-mulative roll-bonding (ARB)5,6) and high-pressure torsion (HPT).7,8) In considering practical applications, the cyclic deformation and fatigue properties of UFG materials have been studied extensively. In particular, the following characteristic features in cyclic deformation of UFG Cu processed by ECAP have been sporadically reported: (1) cyclic softening is frequently detected as a stress-strain response,9–13)_{(2) macroscopic shear bands (SBs) are formed}

inhomogeneously,9,10,13–18) _{(3) coarse grains are locally}

developed.9,12,14,15,19–21) _{For example, Wu et al.}14,15)

ob-served no grain coarsening, while they found apparent SBs by scanning electron microscopy (SEM) observation. There-fore, they concluded that there is no relationship between the SB formation and the grain coarsening. Contrary to this result, Mughrabi and Ho¨ppel19) suggested that the grain coarsening is closely related to the formation of the SBs and proposed two possible SB formation mechanisms. As far as the authors know, there has been no consistent explanation to understand all the above features of the cyclic deforma-tion of UFG Cu. To obtain a thorough understanding of the microstructural changes during cyclic deformation of UFG materials, it is essential to conduct precise observations with the combined use of SEM and transmission electron microscopy (TEM).

In this study, plastic-strain-controlled cyclic deformation tests were performed using high purity UFG Cu produced by ECAP. Based on the results of SEM and TEM observations, the relationship among cyclic softening, SB formation and local grain coarsening was discussed.

2. Experimental Procedure

UFG copper (99.99% purity) billets of 10 mm in diameter and 60 mm in length were processed by the ECAP technique. Each billet was subjected to pressing for 8 passes in such a way that after each pressing the billet was rotated by 90_in longitudinal direction, usually denoted as route Bc.3,4)

Specimens for fatigue tests with the gauge dimensions of

4mm6mm10mm were taken from the central part of the ECAP billets by spark erosion in the direction parallel to the rod axis. The specimens were electrolytically polished to avoid any influence of mechanical ptreatment. Fully re-versed tension-compression low-cycle fatigue tests were car-ried out at room temperature (RT) under the constant plastic-strain amplitudes ranged from "pl¼2104 to "pl¼2 103 _{using an electro-hydraulic testing machine (Shimadzu}

Servopet). Strain was measured with an extensometer mounted directly on the gauge section and a constant strain rate of 1103_s1 _{was employed using a triangular}

com-mand signal. The stress response was monitored on a digital oscilloscope. For microstructural observation, the fatigue tests were interrupted when the stress amplitude decreased and reached 85% of the maximum stress amplitude.

Surface observations were performed on both as-pressed (as-ECAPed) and fatigued specimens by using a Hitachi S-4300 scanning electron microscope equipped an electron backscattering diffraction (EBSD) system (Oxford INCA crystal). Grain maps were constructed with an individual step size of 40 nm and a lower cut-off misorientation angle of 5_. The fatigued specimens were sliced into 3 mm discs and ground down to 0.2 mm using silicon-carbide paper. Final thin foils were prepared by electrolytic polishing, and then TEM observations were carried out on a JEOL 2011 microscope at an accelerating voltage of 200 kV.

3. Results and Discussion

3.1 Cyclic stress-strain response

Cyclic hardening/softening curves of UFG Cu specimens

*1_{Graduate Student, Tokyo Institute of Technology. Present address:} Nippon Steel Corporation, Kimitsu 299-1141, Japan

*2_{Corresponding author, E-mail: fujii.t.af@m.titech.ac.jp}

Special Issue on Severe Plastic Deformation for Production of Ultrafine Structures and Unusual Mechanical Properties: Understanding Mechanisms

are shown in Fig. 1. Cyclic softening apparently occurs at all plastic-strain amplitudes as previously reported in the literature.9–12,19)To discuss the strain-amplitude dependence of the amount of cyclic softening, a cyclic softening ratio was defined as1 ð"3=MaxÞ. Here,"3is the stress amplitude at

the late stage of "cum¼3 and Max is the maximum stress

amplitude. A higher ratio indicates a higher degree of cyclic softening. Figure 2 shows a plot of the cyclic softening ratio against the plastic-strain amplitude. Although the cyclic softening ratio at "pl¼5104 is exceptionally small, the

cyclic softening becomes more remarkable at the higher plastic-strain amplitude. Therefore, it is reasonable to conclude that the amount of cyclic softening increases with increasing the plastic-strain amplitude.

Since the cyclic hardening/softening curves shown in Fig. 1 did not exhibit stress saturation, the pseudo-cyclic stress-strain curve was drawn in Fig. 3 by using the values of stress amplitudes at the cumulative plastic strain of"cum¼1.

The results obtained from strain-controlled tests on UFG Cu

by Ho¨ppelet al.12)_{and those on conventional coarse-grained}

polycrystalline Cu by Mughrabi and Wan22)_{are also shown in}

Fig. 3. It is found that there is a good agreement between the present result and that obtained by Ho¨ppelet al.12)For the cyclic stress-strain response of polycrystalline cubic metals, the power-law relationship

a¼k"npl ð1Þ

has been established,23) _{where k is the cyclic strength}

coefficient and n is the cyclic strain hardening exponent. Applying eq. (1) to the experimental results of the UFG Cu, the values of k¼1060MPa and n¼0:168 were received. Similarly, the values of k¼376MPa and n¼0:175 were obtained in the case of the conventional coarse-grained polycrystalline Cu. Although the stress levels are quite different between the two kinds of specimens, the values of

n are almost the same. This implies that the multiple slip deformation may occur even in the UFG Cu specimens as expected in the coarse-grained polycrystalline Cu.

3.2 Microstructural characterization

Figure 4 shows a set of SEM images taken from specimens cyclically deformed at various plastic-strain amplitudes. On the surfaces of all deformed specimens, so-called SBs were developed. The length of the SBs varied widely from 1mmto 100mmin each specimen. The morphology of the SBs was quite similar to the extrusions frequently formed in persistent slip bands of fatigued Cu single crystals.24–26)_{The SBs were}

also inclined by about 45_{from the stress axis. This indicates} that the SBs are formed parallel to the planes of maximum resolved shear stress. Since micro-crack initiation along the SBs was detected, strain localization indeed occurs in the SBs during cyclic deformation. These structural features of the SBs characterized by the present study are consistent with those reported in the literature.9,11,14,16,18)

Changes in grain size and grain morphology of the fatigued specimens were observed using the EBSD method. Grain maps obtained from an as-ECAPed material and a specimen

Stress amplitude

,

Cumulative plastic strain,

ε

0.1 1 10

200 250 300 350 400

εpl=5 × 10 pl

ε =2 × 10

εpl=1 × 10

εpl=2 × 10-3 -3 -4 -4

σ

Fig. 1 Cyclic hardening/softening curves of specimens cyclically de-formed under plastic-strain amplitudes ranged from "pl¼2104 to "pl¼2103.

0 0.1 0.2

10-4 10-3 10-2

Cyclic softening r

atio

Plastic strain amplitude,

ε

Fig. 2 Plastic-strain-amplitude dependence of the amount of cyclic soft-ening. 60 80 100 200 400 600 800

10-4 10-3 10-2

Present study Höppel et al. (2002)

Mughrabi and Wang (1981) UFG Cu

Polycrystalline Cu

Plastic strain amplitude,

ε

Stress amplitude , a / MP a

σ

fatigued at"pl¼2104 are shown in Figs. 5(a) and 5(b),

respectively. It can be seen from Fig. 5(a) that fine grains with an average size of approximately 500 nm were formed after 8 passes of ECAP. As shown in Fig. 5(b), coarse grains with sizes of about 10mmlong were locally developed in the cyclically deformed specimen. These grains were banded along the direction almost parallel to the trace of SBs (see Fig. 4(a)). Here, it is needed to pay attention to the actual

sizes of the coarsened grains since small-angle boundaries with misorientations of up to 5 _{are indistinguishable in} our SEM observation. The results of TEM observations to elucidate the grain interior in the fatigued specimens will be shown later in this paper.

It also turns out from Fig. 5(b) that the other major grains slightly grow up to 1–2mm and elongate into the same direction as that observed in the coarse grains. Although Wu et al.,14) _{Kunz et al.,}13) _{Luka´sˇ et al.}17,27) _{and Xu et al.}18)

conducted SEM observations in fatigued commercial-purity UFG Cu (99.8%–99.9%) produced by ECAP, they all concluded that no detectable grain coarsening took place during cyclic deformation. This result is obviously in disagreement with the present result. On the other hand, Ho¨ppel et al.12,20,28) _{have reported that shear banding and}

pronounced grain coarsening take place during cyclic deformation of high-purity UFG Cu (99.99%), while they have not observed the grain elongation as represented in the present study. Therefore, the results of our study accord qualitatively with those obtained by Ho¨ppel et al.12,20,28) From all these findings, it is evident that the stability of UFG structure during cyclic deformation strongly depends on the purity of materials.

The relationship between the SB formation and the local grain coarsening is one of the most interesting problems in cyclic deformation of UFG Cu. In the previous studies mentioned above,13,14,17,18,27)_{it has been concluded that there}

is no relationship between the SB formation and the grain coarsening since as-ECAPed grain structures are kept in fatigued specimens with clear SBs. On the contrary, Mughrabi and Ho¨ppel,19)_{and Ho¨ppelet al.}20,28)_{have insisted}

that the grain coarsening plays a dominant role in the formation of the SBs. From Fig. 4(a) and Fig. 5(b), it is reasonable to conclude in the present study that the grain coarsening is closely related to the SB formation. The following experimental results of TEM observations will give strong evidence for this conclusion.

(a) (b)

(c) (d)

2 mµ

stress axis

Fig. 4 SEM images of shear bands formed during cyclic deformation at (a)"pl¼2104, (b)"pl¼5104, (c)"pl¼1103and (d)"pl¼2103. Stress axis is in the vertical direction.

(a)

(b)

2

µ

m

stress axis s

tr

e

s

s

a

x

is

e

xtrusion direction x

tr

u

s

io

n

d

ir

e

c

ti

o

n

The grain morphology in a specimen after ECAP through 8 passes is shown in Fig. 6. Dense dislocations are found to be tangled in the grain interiors. In contrast, it is obvious from Fig. 7 that local grain growth occurred characteristically in a specimen cyclically deformed at"pl¼2104. It is

note-worthy that the coarse grains having a low dislocation density are aligned along the direction approximately parallel to the SBs as represented in Fig. 5(b). The average size of the coarse grains was measured as about 2mm, and this value is

much smaller than that obtained from the grain map shown in Fig. 5(b). It can be readily understood from Figs. 5(b) and 7 that each elongated grain about 10mm in length actually consists of subgrains whose size is in the range of 1–3mm. From our multi-scale observations of the SBs by SEM and TEM, we can justify the conclusion that the plastic strain localization during cyclic deformation of UFG Cu induces the dynamic grain/subgrain growth and causes the develop-ment of the SBs.

500nm

Fig. 6 TEM photograph taken from an as-ECAPed specimen.

1µm

3.3 Correlation among cyclic softening, shear banding and local grain coarsening

In the present study, isolated coarse grains with sizes of 1–3mm were also observed frequently in specimens cycli-cally deformed at various strain amplitudes. A typical result is shown in Fig. 8. Since the dislocation wall structure is well developed in this isolated coarse grain as originally reported by Agnew and Weertman,9) _{it may be expected}

that the strain localization takes place even in the isolated grain. This is consistent with the fact that the length of the SBs varied widely from 1mm to 100mm. To explain the formation mechanism of the SBs, Mughrabi and Ho¨ppel19) _{have proposed two possible scenarios. One is}

that grain/subgrain coarsening is locally initiated and leads to SBs in a large scale. The other is that a catastrophic extended SB firstly happens and subsequently triggers the formation of a new coarsened microstructure within the SB. A reasonable implication of our observations is that the former scenario is more likely to occur as the SB formation mechanism.

As shown in Fig. 1, the cyclic softening was observed at all plastic-strain amplitudes. The SB formation and the grain growth were also detected by the SEM and TEM observa-tions. These findings inevitably raise the question of whether the development of the SBs is related to the cyclic softening. Xuet al.18)_{and Kunzet al.}13)_{have concluded that there is not}

a one-to-one correlation between the SB formation and the cyclic softening since they both have found distinct SBs even in the case of cyclic hardening. Moreover, they could provide no microstructural evidence of the cyclic softening. On the other hand, the present study gives strong evidence for the direct relationship between the SB development and the local grain coarsening. Therefore, it is quite reasonable to suppose that the cyclic softening is closely related to the strain localization and the resultant shear-banding phenomenon, while the purity of materials strongly affects the stability of the UFG structure.

4. Conclusions

Plastic-strain-controlled cyclic deformation of high purity UFG Cu (99.99%) produced by ECAP was performed at RT. The stress-strain response and the microstructural change were investigated. The conclusions can be summarized as follows.

(1) UFG materials show cyclic softening at all plastic-strain amplitudes.

(2) SBs are developed along the planes of maximum resolved shear stress. Grain coarsening occurs inten-sively in the SBs and gives rise to subgrain formation. (3) Cyclic softening can be understood as a strain local-ization phenomenon and is closely related to the SB formation and the dynamic grain coarsening.

Acknowledgments

The authors wish to express their gratitude to Professor Zenji Horita, Kyushu University, Professors Hisashi Sato and Yoshimi Watanabe, Nagoya Institute of Technology, for performing the ECAP processing of the samples used in this study. This research was supported by a Grant-in-Aid for Scientific Research on Priority Areas (18062002) by the Ministry of Education, Culture, Sports, Science and Tech-nology of Japan.

REFERENCES

1) V. M. Segal, V. I. Reznikov, A. E. Drobyshevskiy and V. I. Kopylov: Russi. Metall.1(1981) 99–105.

2) R. Z. Valiev, N. A. Krasilnikov and N. K. Tsenev: Mater. Sci. Eng. A 137(1991) 35–40.

3) Y. Iwahashi, Z. Horita, M. Nemoto and T. G. Langdon: Acta Mater.46 (1998) 3317–3331.

4) M. Furukawa, Y. Iwahashi, Z. Horita, M. Nemoto and T. G. Langdon: Mater. Sci. Eng. A257(1998) 328–332.

5) Y. Saito, H. Utsunomiya, N. Tsuji and T. Sakai: Acta Mater.47(1999) 579–583.

6) N. Tsuji, Y. Saito, H. Utsunomiya and S. Tanigawa: Scr. Mater.40 (1999) 795–800.

7) N. A. Smirnova, V. I. Levit, V. I. Pilyugin, R. I. Kuznetsov, L. S. Davydova and V. A. Sazonova: Fiz. Metal. Metalloved. 61(1986) 1170–1177.

8) A. P. Zhilyaevb, G. V. Nurislamovac, B.-K. Kimd, M. D. Baro´a, J. A. Szpunard and T. G. Langdon: Acta Mater.51(2003) 753–765. 9) S. R. Agnew and J. R. Weertman: Mater. Sci. Eng. A 244(1998)

145–153.

10) S. R. Agnew, A. Vinogradov, S. Hashimoto and J. R. Weertman: J. Electro. Mater.28(1999) 1038–1044.

11) A. Vinogradov and S. Hashimoto: Mater. Trans.42(2001) 74–84. 12) H. W. Ho¨ppel, Z. M. Zhou, H. Mughrabi and R. Z. Valiev: Phil. Mag. A

82(2002) 1781–1794.

13) L. Kunz, P. Luka´sˇ and M. Svoboda: Mater. Sci. Eng. A424(2006) 97–104.

14) S. D. Wu, Z. G. Wang, C. B. Jiang, G. Y. Li, I. V. Alexandrov and R. Z. Valiev: Scr. Mater.48(2003) 1605–1609.

15) S. D. Wu, Z. G. Wang, C. B. Jiang, G. Y. Li, I. V. Alexandrov and R. Z. Valiev: Mater. Sci. Eng. A387–389(2004) 560–564.

16) M. Goto, S. Z. Han, T. Yakushiji, C. Y. Lim and S. S. Kim: Scr. Mater. 54(2006) 2101–2106.

17) P. Luka´sˇ, L. Kunz and M. Svoboda: Metall. Mater. Trans. A38A (2007) 1910–1915.

18) C. Xu, Q. Wang, M. Zheng, J. Li, M. Huang, Q. Jia, J. Zhu, L. Kunz and M. Buksa: Mater. Sci. Eng. A475(2008) 249–256.

19) H. Mughrabi and H. W. Ho¨ppel: Mat. Res. Sym. Proc.634(Mat. Res.

1

µ

m

Soc., 2001) pp. B2.2.1–B2.2.11.

20) H. W. Ho¨ppel, M. Brunnbauer, H. Mughrabi, R. Z. Valiev and A. P. Zhilyaev: Proc. Materials Week 2000, (Wiley-VCH, 2001) http:// www.materialsweek.org/proceedings.

21) S. D. Wu, Z. G. Wang, C. B. Jiang and G. Y. Li: Phil. Mag. Lett.82 (2002) 559–565.

22) H. Mughrabi and R. Wang: Proc. 2nd Risø Int. Sym. Metall. Mater. Sci., ed. by N. Hansenet al., (Risø National Lab., 1981) pp. 87–98. 23) S. Suresh: Fatigue of Materials, (Cambridge, Cambridge University

Press, 1998) p. 94.

24) J. M. Finney and C. Laird: Phil. Mag.31(1975) 339–366.

25) Z. S. Basinski and S. J. Basinski: Scripta Metall.18(1984) 851–856. 26) B.-T. Ma and C. Laird: Acta Metall.37(1989) 325–336.

27) P. Luka´sˇ, L. Kunz, M. Svoboda, M. Buksa and Q. Wang: Proc. H. Mughrabi Honorary Symposium, ed. by K. J. Hsia, M. Go¨ken, T. Pollock, P. D. Portella and N. R. Moody, (TMS, 2008) pp. 161–166. 28) H. W. Ho¨ppel, C. Xu, M. Kautz, N. Barta-Schreiber, T. G. Langdon and