A cyclic twin bridge shear test for the identification of kinematic hardening parameters

Abstract

A twin bridge cyclic shear test with in-plane torsion is proposed for the identification of kinematic hardening parameters for metallic sheets. Besides its simplicity, noteworthy advantages of the test are (a) reduced loads on the experimental device as compared to a one-sided shear test, (b) identical orientation of the principal stresses with respect to the rolling direction in both of the shear bridges, e.g. which cannot be realized by the classical Miyauchi shear test, and (c) no premature termination by instability mechanisms such as buckling or necking. Two main disadvantages appear to be (a) a preclusion of the use of analytically solved initial value problem in parameter identification due to a diffusivity of the plastic region around the shear bridges, (b) smeared out anisotropic material response proportional with the width of the shear bridge. As a remedy for the former, an inverse parameter identification methodology is used to determine the hardening parameters using an objective function devising the measured moment and rotation angle. For the latter, an optimum shear bridge width is selected which also minimizes the edge effect where a shear equilibrium is not possible. A combined non-linear isotropic and kinematic hardening model respectively based on Voce and Armstrong–Frederick is selected as the material model which is implemented as a VUMAT subroutine for ABAQUS/EXPLICIT. Strain-controlled tests are conducted using three different classes of steel sheet materials, namely a mild steel DC06, a dual phase steel DP600 and a Transformation Induced Plasticity steel TRIP700. These tests with one cycle including a forward shearing and a reverse shearing phase merely focus on the Bauschinger effect. Variations with different stress and strain based loading cycles for phenomena like shakedown, ratcheting, mean stress relaxation, cyclic hardening and softening are not explored and left beyond the scope of the current study. The results, besides showing the applicability of the test to the kinematic hardening parameter identification purposes, also show that the Armstrong–Frederick model falls short to capture the cyclic response of the selected materials, especially advanced high strength steels DP600 and TRIP700.