Selection of material models for predicting necking in superplastic forming

The mechanical behaviour of a superplastic alloy is characterised by both the sinh-law, (epsilon)over dot(p) proportional to sinh(betasigma), and power-law,(epsilon)over dot(p) proportional to (sigma/K)(1/m), constitutive equations together with the hardening variables containing significant physical meanings. The material constants in the power-law equation are determined using two different methods: the genetic algorithm (GA)-based optimization technique and the more recently developed physically-based multiple step method. The two techniques give different values for the material constants and thus result in different magnitudes for the strain rate sensitivity parameter m. Together three SETs of determined unified constitutive equations are obtained and all enable the experimental stress-strain curves and grain growth data of the material to be well captured. In this paper, studies are carried out to investigate the effects of the three determined material models on necking predictions for the superplastic gas-blow forming of a structural component. In addition the predicted gas pressure cycles, required to ensure that the maximum strain rate over the workpiece approximately equals the pre-defined target deformation rate, are compared and analysed for the three cases. (C) 2001 Elsevier Science Ltd. All rights reserved.