Large strain flow curve identification for sheet metals under complex stress states

Abstract

Strain hardening behaviours at large strain under various loading conditions are the basic but the most important input for reliable numerical simulation of plastic deformation processes, such as sheet metal forming and crash. However, neither the flow curve at large strain beyond necking nor the strain hardening under stress states different from uniaxial tension can be reasonably characterized by the widely employed tensile tests of dogbone specimens. In this study, various experimental methods are investigated to characterize strain hardening behaviours up to large deformation under different stress states for an aluminium alloy sheet of AA5182-O. The experiments conducted include tensile tests of four different specimens (e.g. dog-bone specimen, notched specimen, specimen with a central hole and in-plane shear specimen), bulge tests and twin bridge shear tests. These tests cover wide stress states ranging from shear to equibiaxial tension. Strain hardening is obtained by both analytical computation and an inverse engineering approach under different loading conditions of shear, uniaxial tension, plane strain tension and equibiaxial tension. These two approaches are evaluated by comparing the obtained stress-strain curves under different loading conditions. The evaluation shows that the inverse engineering approach is an effective method to characterize the stress-strain curve up to large plastic deformation till fracture for tests with inhomogeneous deformation. The results also reveal that it needs to develop advanced yield functions to model yielding and hardening behaviours under complex stress states.