Materials Transactions, Vol. 45, No. 4 (2004) pp. 1261 to 1265 #2004 The Japan Institute of Metals
Superplasticity at Room Temperature in Zn-22Al Alloy Processed by Equal-Channel-Angular Extrusion
Tsutomu Tanaka* and Kenji Higashi
Department of Metallurgy and Materials Science, Osaka Prefecture University, Sakai 599-8531, Japan
Equal-Channel-Angular extrusion (ECAE) has been conducted at room temperature in order to develop an ultrafine-grained structure in a Zn-22 mass% Al eutectoid alloy. The microstructures had an equiaxed and homogeneously distributed Al-rich and Zn-rich duplex-phase and its average grain size was about 0.35 mm. This alloy exhibited a large elongation of 240% albeit at room temperature and a high strain rate of 10 2 s 1. A grain-size-dependent phenomenon was also observed at lower strain rates. The activation energy was close to that for grain boundary diffusion. It was apparent that the deformation behavior at room temperature is consistent with that observed in conventional superplasticity. From the results, it was concluded that the dominant deformation mechanism was grain boundary sliding accommodated by dislocation movement controlled by grain boundary diffusion.
(Received November 28, 2003; Accepted February 20, 2004) Keywords: equal-channel-angular extrusion; superplasticity; mechanical properties; constitutive equations; Zn-22Al alloy
1. Introduction temperature of 373 K and initial grain size before ECAE of 1.0 mm or greater. More recently, Tanaka et al.21) has reported Superplasticity refers to the ability of a material to pull out that the equiaxed and homogeneous microstructure with the to a high tensile elongation without the development of average grain size of about 0.3 mm was attained by ECAE at necking.1) It has been well known that superplastic deforma- room temperature during phase transformation. tion requires a fine grain size, typically below 10 mm, and In this study, tensile tests of Zn-22Al eutectoid alloy reasonably high testing temperature above 0.5 Tm, where Tm processed by ECAE were carried out in a wide rage of strain is the absolute melting temperature of the material in degrees rates and at room temperature, and deformation mechanism Kelvin.2) Superplasticity has been utilized for the fabrication at room temperature was compared with that at high of complex parts from sheet metal because of the large temperature. elongation.3) There is considerable current interest in devel- oping materials with ultrafine grain sizes, since experimental 2. Experimental Procedures evidences have shown that reducing grain size increases the superplastic strain rate and/or decreases the superplastic The material used in present investigation was Zn- temperature.4,5) Such discoveries have enormous practical 22 mass%Al alloy produced by Kobe Steel Ltd. At first, the significance; for instance, it was reported recently that the alloy was machined into rod with a diameter of 19.5 mm and ultrafine grained Zn-Al alloy could be applied as a seismic a length of 90 mm. Then, the alloy solution treated in air at damper for high-rise buildings by attaining high strain rate 623 K for a day and subsequently water-quenched. superplasticity at room temperature.6–11) ECAE was conducted by the die with a 90 angled channel Recently, severe plastic deformation (SePD) techniques through the die using, so-called, processing route Bc, in such as equal-channel-angular extrusion (ECAE) or torsion which the specimen was removed from the die and then straining (TS) have been seen as methods for producing bulk rotated by +90 in the same direction between each pass.22) materials with ultrafine-grains.12,13) The principle behind This route was selected for the majority of the tests because it these techniques is to produce a very large strain in the leads a homogeneous microstructure of fine grains with high sample; the rearrangement of the dislocations introduced by angle boundary.23) Kato et al. has reported that the phase straining leads to a substantial grain refinement down to the transition from 0-Al single phase to -Al and -Zn two submicrometer or even the nanometer scale. Materials phases occurred from 5 seconds to 20 minutes at room processed by ECAE have the potential for superplasticity, temperature after quenching.24) In the previous report, then, especially at very high strain rates and low temperatures. The we compared the microstructure formed by ECAE during the validity of this proposal has been demonstrated by several phase transformation with that formed by ECAE after the recent reports of high strain rate superplasticity above phase transformation, and demonstrated that ECAE during 10 2 s 1 and low temperature superplasticity in aluminum phase transformation results in almost equivalent fine micro- and magnesium alloys processed by ECAE.14–17) However, structure compared with ECAE after phase transformation in there have been few attempts to improve mechanical proper- spite of fewer passages.21) The alloy performed ECAE during ties at room temperature, and to refine the microstructure of the phase transformation was used in the present investiga- Zn-Al alloys by means of ECAE.18–20) In existing inves- tion. All of the ECAE were conducted using a hydraulic press tigations, the microstructure of such alloys has relatively operating at 0.2 mm/s for the first passage and 1 mm/s after large (0.6–1 mm) or elongated grains, due to an operating the second passage. Microstructures of materials processed by ECAE were *Graduate Student, Osaka Prefecture University observed using transmission electron microscopy. After the 1262 T. Tanaka and K. Higashi samples were polished to a thickness of 150–200 mm strain rate. From Fig. 2, it is obvious that the flow curves in all parallel to the extrusion direction, small disks with a diameter investigations could be divided into two regions: that in of 3 mm were punched out. A twin-jet electro-polishing unit which the m-value was about 0.25 in the low strain-rate range was then used with a solution of 10% HClO4, 20% C3H8O3 and that in which the m-value was above 0.2 in the high and 70% C2H5OH. The samples were thinned at 263 K until strain-rate range. In general, it is well known that the m-value perforation.19) All of the microstructural observations were in the superplastic region, suggesting that the dominant conducted using a JEOL JEM-200CX transmission electron deformation mechanism could be grain boundary sliding,1) microscope operating at 200 kV and selected area electron exhibits near 0.5. Therefore, the m-value at room temperature diffraction (SAED) patterns were recorded from areas. The was slightly lower than that at high temperature. Similar grain size d was estimated using the equation d ¼ 1:74L: L is behaviors can be seen in the results of the previous reports the linear intercept length. shown in Fig. 2. The results in the present investigation also Tensile specimens machined directly from the extruded exhibit the lowest flow stress at all strain rates; it was bars had tensile axes parallel to the extrusion direction. The apparent at lower strain rates (below 1 10 2 s 1) the flow gauge length of the specimen was 5.0 mm, and the diameter stress particularly increases with an increase in grain size. was 2.5 mm. Tensile tests were carried out on an Instron-type This grain-size-dependent phenomenon is a common char- testing machine controlled by a computer. These tests were acteristic for superplastic materials.2) In order to characterize carried out with a range of strain rate from 1:0 10 5 s 1 to the effect of grain size on flow stress, therefore, the variation 1 1s and at room temperature (0:44Tm). The flow stress for in strain rate at ¼ 90 MPa is plotted as a function of the each strain rate was determined at small strain of 0.1. In such reciprocal grain size in Fig. 3, which also includes the results a small strain, grain growth occurring during superplastic of earlier reports.7,10,20,24,25) The grain size exponent, p-value, flow would be negligible. which is estimated from the slope of the line, was found to be 2 at lower strain rates. Several experimental results on 3. Results and Discussion superplasticity at high temperature in Zn-22Al alloys have also indicated p ¼ 2.27) This result suggests that the defor- Figure 1 shows (a) TEM bright-field image and (b) the mation behavior at room temperature resembles the general associated electron diffraction pattern of Zn-22Al alloy after superplastic behavior. ECAE. It is apparent that the equiaxed and very fine grains The variation in elongation-to-failure as a function of are distributed, and the average grain sizes after ECAE were about 0.35 mm. In the previous reports, the average grain sizes in Zn-22Al alloys processed by only quenching were about 0.6–0.8 mm.24–26) The grain refinement during ECAE at room temperature was verified in the present investigation. The diffraction spots corresponding to Al and Zn were also observed as shown in Fig. 1(b). In addition, the electron diffraction pattern shows a well-defined ring pattern indica- tive of the presence of boundaries having high angles of misorientation. Owing to the very fine and equiaxed grains, high-strain-rate superplasticity at room temperature was expected to occur in the Zn-22Al alloy at room temperature. Variation in flow stress as a function of strain rate at room temperature in the Zn-22Al alloy is shown in Fig. 2, which includes the results of earlier reports,7,20,24–26) as well. The strain rate sensitivity exponent (m-value) is defined in the Fig. 2 The variation in flow stress as a function of strain rate at room equation (d ln =d ln "_) where is the flow stress and "_ is the temperature in previous studies and in this study.
Fig. 1 (a) TEM bright-field image and (b) associated SAED pattern for Zn-22Al alloy after ECAE. Superplasticity at Room Temperature in Zn-22Al Alloy Processed by Equal-Channel-Angular Extrusion 1263
Fig. 5 The variation in "_ðT=GÞðd=bÞ2 as a function of 1=T for Zn-22Al alloy.
appearance or fracture surfaces of the specimens tested in the regions exhibiting m-values of 0:25 0:3.7,29) Therefore, it Fig. 3 The variation in strain rate as function of the reciprocal grain size at room temperature and at a flow stress of ¼ 90 MPa in Zn-22Al alloy. was concluded that superplasticity occurs even at room temperature, regardless of the lower m-value near 0:25 0:3. It was considered that the apparent m-value decreased with decreasing testing temperature because of the existence of a threshold stress. To determine a threshold stress, a plot of against "_1=n (n ¼ 2, 3, 5, 8) on a double-linear scale was adopted. Use of n ¼ 2 gave the best linear fit among the assumed stress exponents for all materials. Subsequently, all values of threshold stress could be estimated by extrapolation to the zero strain rate of a line, which the data gave as against "_1=2 on a double-linear scale.30) Thus, the threshold stress estimated in the present investigation was about 19.9 MPa. The variation in "_ðT=GÞðd=bÞ2 as a function of the reciprocal temperature is shown in Fig. 5 for the results at 423–523 K,18) 373 K10) and 303 K (this study), where ð 4 0Þ=G ¼ 3 10 . As shown in Fig. 5, the activation energy estimated the slop of the line was about 54 kJ/mol and the datum in the present investigation was successfully included Fig. 4 The variation in elongation-to-failure as a function of strain rate at into the line without an inflection. In general, it has been room temperature in previous studies and in this study. reported that the activation energy for Zn-22Al alloy in superplasticity equal to the activation energy for grain 31) boundary diffusion in zinc (Qgb ¼ 60:5 kJ/mol). This strain rate in the present study is plotted in Fig. 4, which also suggests that the diffusion of zinc operates dominantly albeit includes comparative data taken from Ref. (7, 24, 25, 28). at room temperature as well as the superplastic deformation The region exhibiting large elongations was roughly in good at high temperature. agreement with the region where high m-values were In order to compare the present data with other high achieved. As shown in Fig. 4, it is apparent that the ductility temperature superplastic behaviors,7,18,19,24–26,32–36) normal- 2 at room temperature is improved by using ECAE at room ized strain rate, "_ðkT=DgbGbÞðd=bÞ , as a function of temperature. It should be noted that the large elongation of normalized stress, ð 0Þ=G, is plotted in Fig. 6, where 240% in the present alloy was attained at a higher strain rate Q ¼ 60:5 kJ/mol. From Fig. 6, it is clearly observed that the of 1 10 2 s 1. It is also important that high-strain-rate material in the present investigation behaved identically to superplasticity was attained despite room temperature con- other superplastic materials, and the constitutive equation for ditions. The optimum superplastic strain rate in the present superplastic flow of Zn-22Al at room temperature is given as alloy was faster than that in the earlier results. It appears that 2 2 Gb b 0 the optimum strain rate increases with decreasing grain size. "_ ¼ 114 Dgb High-strain-rate superplasticity at room temperature in the kT d G present alloy may be attributed to the very small grain size of where G is the shear modulus, b is the Burgers vector, k is the 0.35 mm. Furthermore, direct evidence of grain boundary Boltzmann constant, T is the absolute temperature, d is the sliding was confirmed by SEM observations on the surface grain size, 0 is the threshold stress and Dgb is the grain 1264 T. Tanaka and K. Higashi
(3) The grain size exponent of 2 was obtained in the lower strain rate region, and the activation energy of the deformation condition in the present investigation was close to that for grain boundary diffusion as well as the superplastic condition at high temperature. (4) The normalized plot suggested that the dominant deformation mechanism for superplasticity at room temperature is grain boundary sliding accommodated by dislocation movement controlled by grain boundary diffusion.
Acknowledgements
The authors are grateful to Dr. H. Watanabe (Osaka Municipal Technical Research Institute) for their helpful comments and Prof. R. Oshima and Dr. H. Nishizawa (Osaka Prefecture University) for their helpful aids on the TEM analysis. This investigation is a part of projects in Innovation Plaza Osaka of Japan Science and Technology Corporation.
REFERENCES
1) T. G. Langdon: Metall Trans. 13A (1982) 689–701. 2 Fig. 6 The variation in "_ðkT=DgbGbÞðd=bÞ as a function of ð 0Þ=G in 2) O. D. Sherby and J. Wadsworth: Prog. Mater. Sci. 33 (1989) 166–211. the superplastic region. 3) A. J. Barnes: Mater. Sci. Forum. 170–172 (1994) 701–714. 4) H. Somekawa, T. Tanaka, H. Sasaki, K. Kita, A. Inoue and K. Higashi: Acta Meter. 52 (2004) 1051–1059. 5) H. Somekawa, H. Watanabe, T. Mukai and K. Higashi: Scr. Mater. 48 boundary diffusion coefficient. Based on the constitutive (2003) 1249–1254. equation, it is concluded that the dominant deformation 6) K. Makii, Y. Seki, A. Kushibe and K. Higashi: The Fourth Pacific Rim International Conference on Advanced Materials and Processing mechanism for superplasticity at room temperature can be (PRICM4) (2001) 1977–1978. GBS accommodated by dislocation movement involving the 7) T. Tanaka, K. Makii, A. Kushibe and K. Higashi: Mater. Trans. 43 sequential steps of glide and climb. (2002) 2449–2454. Grain boundary sliding causes stress concentrations at 8) T. Tanaka, K. Makii, A. Kushibe and K. Higashi: Key Engineering triple points of grain boundaries. Therefore, it is required to Materials 233–236 (2003) 719–724. 9) H. Ueda, T. Okada, M. Kato, Y. Mimura, K. Makii, M. Tabuchi, K. relax the stress concentrations in order to attain large Aoki, T. Okamoto, H. Okumoto and A. Kushibe: The second Japan 37) elongations. Diffusional flow and diffusion-controlled National Symposium on Structural Control., (2000) 545–551. dislocation movement38–40) have been considered to accom- 10) T. Tanaka, K. Makii, A. Kushibe, M. Kohzu and K. Higashi: Scr. modate grain boundary sliding. However, it may be difficult Mater. 49 (2003) 361–366. to accommodate sliding at room temperature by the diffu- 11) T. Tanaka, K. Makii, A. Kushibe, M. Kohzu and K. Higashi: Int. J. Mech. Sci. 45 (2003) 1599–1612. sion-related processes. This is because the diffusion coef- 12) K. Nakashima, Z. Horita, M. Nemoto and T. G. Langdon: Acta Mater. ficient for grain boundary diffusion, Dgb, reduces with a 46 (1998) 1589–1599. reduction of temperature. Accordingly, it is considered that 13) R. S. Mishra, R. Z. Valiev and A. K. Mukherjee: NanoStructured the m-value at room temperature may be slightly lower than Mater. 9 (1997) 473–476. that at high temperature. The details will be described in our 14) S. W. Chung, H. Somekawa, T. Kinoshita, W. J. Kim and K. Higashi: Scr. Mater. 50 (2004) 1079–1083. next report. 15) P. B. Berbon, S. Komura, A, Utsunomiya, Z. Horita, M. Furukawa, M. Nemoto and T. G. Langdon: Mater. Trans., JIM 40 (1999) 772–778. 4. Summary 16) H. Watanabe, T. Mukai, K. Ishikawa and K. Higashi: Scr. Mater. 46 (2002) 851–856. The deformation mechanism at room temperature was 17) H. Watanabe, T. Mukai, M. Mubuchi and K. Higashi: Scr. Mater. 41 (1999) 209–213. investigated in Zn-22 mass%Al alloy with very fine grains by 18) S. M. Lee and T. G. Langdon: Mater. Sci. Forum. 357–359 (2001) 321– means of ECAE at room temperature. The following results 326. were obtained. 19) M. Furukawa, Y. Ma, Z. Horita, M. Nemoto, R. Z. Valiev and T. G. (1) The equiaxed and fine microstructure with the average Langdon: Mater. Sci. Eng. A. 241 (1998) 122–128. grain size of 0.35 mm was attained. 20) Y. Kaneko, A. Vinogradov, K. Kitagawa and S. Hashimoto: Inves- tigations and Applications of Severe Plastic Deformation, in: T.C. (2) The largest elongation of 240% was obtained even at Lowe and R.Z. Valiev (Eds.) (2000) 289–295. room temperature and at a high strain rate of 21) T. Tanaka, H. Watanabe and K. Higashi: Mater. Trans. 44 (2003) 1 10 2 s 1, where an m-value of about 0.25 was 1891–1894. obtained. The optimum superplastic strain rate in the 22) Y. Iwahashi, J. Wang, Z. Horita, M. Nemoto and T. G. Langdon: Scr. present alloy was faster than that in the other Zn-Al Mater. 35 (1996) 143–146. 23) K. Oh-ishi, Z. Horita, M. Furukawa, M. Nemoto and T. G. Langdon: alloy, which was processed by recrystallization only. Superplasticity at Room Temperature in Zn-22Al Alloy Processed by Equal-Channel-Angular Extrusion 1265
Metall. Mater. Trans. A. 29 (1998) 2011–2013. (Pargamon Press, Oxford, 1982) 44. 24) M. Kato, E. Kawai, K. Mutsuzaki and J. Miyagawa: J. Japan. Inst. 32) K. Takeoka, D. Oelschla¨gel and O. Izumi: J. Japan Inst. Metals. 35 Metals. 38 (1974) 539–545. (1971) 780–787. 25) M. Hirohashi, E. Kawai, K. Mutsuzaki and M. Miyagawa: J. Japan. 33) A. Ball and M. M. Hatchison: Metal Sci. J. 3 (1969) 1–7. Inst. Metals. 42 (1978) 101–106. 34) M. L. Vaidya, K. L. Murty and J. E. Dorn: Acta. Metall. 21 (1973) 26) H. Naziri, R. Pearce, M. H. Brown and K. F. Hale: Acta Metall. 23 1615–1623. (1975) 489–496. 35) Y. Motohashi, T. Shibata and M. Miyagawa: J. Japan Inst. Light 27) P. Yavari and T. G. Langdon: Mater. Sci. Eng. 57 (1983) 55–65. Metals. 33 (1983) 270–282. 28) G. Torres-Villasen˜or and J. Negrete: Mater. Sci. Forum. 243–245 36) P. K. Chaudhury, V. Sivaramakrishnan and F. A. Mohamed: Metall. (1997) 553–556. Trans. A. 19 (1988) 2741–2752. 29) J. Negrete, L. Valdes and G. Torres-Villasen˜or: Metall. Trans. A. 14 37) M. F. Ashby and R. A. Verrall: Acta Metall. 21 (1973) 149–163. (1983) 1931–1934. 38) R. C. Gifkins: Metall. Trans. A. 7 (1976) 1225–1232. 30) H. Watanabe, T. Mukai, M. Mabuchi and K. Higashi: Acta Mater. 49 39) J. H. Gittus: J. Eng. Mater. Tech. 99 (1977) 244–251. (2001) 2027–2037. 40) A. K. Mukherjee: Mater. Sci. Eng. 8 (1971) 83–89. 31) H. J. Frost and M. F. Ashby: in Deformation-mechanism Maps,