Modeling and simulation of the second Sandia challenge problem using phantom paired shell element: characterization of material anisotropy and rate-dependency

Abstract

A phantom paired solid shell toolkit for Abaqus explicit solver (XSHELL) coupled with accumulated plastic strain and maximum plastic strain direction is applied for the blind prediction and recalibrated analysis of the ductile rupture of a Ti–6Al–4V sheet under both quasi-static and modest-dynamic loading. The complexities of this second Sandia challenge problem include the rate dependent and anisotropic material behaviour and co-existence of multiple round notches and holes to activate a combined tensile and shear driven failure behaviour. Given the nature of this thin-walled component, both the response and kinematic description of damage initiation and propagation can be captured using a relatively coarse model with degrees of freedom in the order of 30,000. A simplified modelling strategy is employed during the blind prediction without considering the material anisotropy, shear failure data, and free rotation at its pin connections. While both the crack path and critical load at crack initiation and propagation can be predicted reasonably well especially for the quasi-static case, its initial stiffness and peak load has a deviation from the test data. In order to capture the anisotropic material behaviour, a Hill’s constitutive model along with its reduction for a 2D plane stress case is implemented in XSHELL and its model parameters are determined using the shear calibration data. By allowing a free rotation at the pin connection via the definition of a contact surface, a recalibrated analysis is performed for the ductile failure prediction of the Ti–6Al–4V sheet under both quasi-static and modest-rate dynamic loading. A much improved rupture prediction is achieved in terms of crack path, an entire load deflection curve, and peak loads associated with the crack initiation and propagation. The exercise of this Sandia challenge problem reveals that a correct data calibration and accurate representation of boundary condition is essential to perform the ductile failure prediction in the presence of limited test data and available material properties.