Experimental and numerical investigation of differential hardening of cold rolled steel sheet under non-proportional loading using biaxial tensile test

Abstract

Strain path change is common phenomenon during industrial sheet metal forming processes. These strain paths can be proportional and non-proportional in nature. To enable strain path change in a single experiment is key challenge faced by many researchers. To predict deformation behavior accurately during such strain paths, it is required to represent initial yielding, the evolution of subsequent yield surface and hardening behavior precisely. This work deals with the mechanical characterization of Interstitial Free High Strength Steel (IFHS) during multi-axial proportional and non-proportional loading for the determination of yield surface evolution and strain hardening behavior. A set of tensile tests at various orientations with respect to the rolling direction and in-plane biaxial tests using cruciform specimens for various linear and bi-linear stress ratios were performed on as-received materials. It was found that at the onset of stress path change, the stress strain curve shows a change in the hardening rate for rolling and transverse directions of the biaxial specimen due to the change in active slip systems. To predict this behaviour various yield criterion like von Mises, Hill's 48 and Yld2000-2d were used. The experimental results show that the mechanical behavior of the subjected specimen deviates from isotropic behavior predicted by the von Mises and anisotropic behavior predicted by the Hill's 48 yield criterion. The initial yield loci measured by biaxial testing machine is in good agreement with what is predicted by Yld2000-2d yield criterion. However, it was observed that such phenomenological models were unable to capture the observed deformation behaviour. It is also shown that classical Crystal Plasticity Finite Element Method (CPFEM) can capture change in strain hardening rate during non-proportional deformation for few cases reasonably well while in few cases it could not predict.