SerlptaMetailurglca et Materi& zyxwvutsrqponmlkjihgfedcbaZYX Vol. 32, No. 7, pp. 1079-1084.1995 Copylight8 1995 lslsevier scie.nce Ltd Rinted in the USA. All rights reserved 0956-716X/95 $9.50+ Ml 0956-716X(94)00003-4 EFFECT OF LIQUID PHASES ON THE TENSILE ELONGATION OF SUPERPLASTIC ALUMINUM ALLOYS AND COMPOSITES K. Higashi*, T.G. Nieh**, M. Mabuchi***, and J. Wadsworth** * Department of Mechanical Systems Engineering, University of Osaka Prefecture, Gakuen-cho, Sakai, Osaka 593, Japan ** Lawrence Livermore National Laboratory, P.O. Box 808, L-350 Livermore, CA 94550, USA *** National Industrial Research Institute of Nagoya, 1 Hirate-cho, Kita-ku, Nagoya 462, Japan (Received October 2 1, 1994) (Revised November 4,1994) Introduction One of the major drawbacks of conventional superplastic forming is that the phenomenon is only found at relatively low strain rates, typically about 10-4 to 10-3 s-l [ 11. Recent studies, however, have demonstrated that su V rplasticit can in fact be found at considerably higher strain rates than l&3 s-l, i.e. at strain rates of up to 10 y_. to lo2 s- . .fhts high-strain-rate superplasticity (HSRS) phenomenon was originally observed in metal matrix composites [2] and then found in mechanically alloyed materials [3]. The phenomenon was then studied in some detail, principally in metal matrix composites [4-71 but was also pursued in the mechanically alloyed materials [S]. More recently the effect has also been observed in metallic alloys produced by more conventional methods [9-121. Technologically, HSRS in metal matrix composites is expected to result in a viable, near-net-shape forming technique for the automobile, aerospace, and even semi-conductor industries. It was initially pointed out by Nieh et al. [13, 141 that the observed HSRS phenomenon may be related to the presence of some liquid phases at interfaces or grain boundaries as a result of, or at least accompanied by, the segregation of solutes to such regions. Recently, Mabuchi and Higashi [15], using an in zyxwvutsrqponmlkjihgfedcba situ TEM technique, have directly observed solute segregation at interfaces or grain boundaries and the preferential melting of these enriched grain boundaries. In the current paper, we present further evidence, through the analysis of existing data, to illustrate that the elongation to failure of many superplastic materials is associated with the presence, or the likelihood of the presence, of liquid phases at boundary interfaces. The materials and experimental procedures have been described previously in various papers. A summarv of the sunernlastic orouerties of a number of advanced structural materials is nresented in Table 1. The materials mclude mechanically-alloyed (MA) materials (IN9021, IN9052, IN905XL, aid IN9021/SiC/15p), metal-matrix composites (Al-Cu-Mg/Si3N4/20p, Al-Mg/Si3N4/20p, Al-Mg-Si/Si3N4/20p, and AI-Mg- Si/Si3N4/20w), as well as powder metallurgy (PM) and ingot metallurgy (IM) metallic alloys (PM/Al-Mg-Mn, PM/Al-Ni-Misch metal, PM/Al-Ni-Zr-Misch metal, Al-Cr-Fe deposit, and IM/7475). In Table 1. T,, renresents the optimum superplastic temperature, Ti is the incipient melting point, Ts is the solidus temperature; ~-IS the strain rate, cr is the flow stress at a true strain of 0.1, m is the strain rate sensitivity, El. represents the percentage elongation to failure, and d is the grain size. All the materials are noted to be fine-grained (-1 urn), except IM/7475 which has a grain size of approximately 20 pm. Since the thermal stability of materials can be influenced by thermal mechanical processing, a direct experimental DSC technique was used to determine the incipient melting point in most of the materials listed in Table 1. Examples are shown in Fig. 1 of DSC curves for the IN9021 and IN9052 alloys. The curves indicate that: the incipie.nt melting point is 754K (caused by the eutectic CuMgAl2 phase) compared to the solidus 1079