Yield locus prediction using statistical and RVE-based fast Fourier transform crystal plasticity models and validation for drawing steels

Abstract

A methodology is presented to predict yield loci of drawing steels with virtual mechanical tests using (i) the statistical ALAMEL crystal plasticity model (VEF software of KU Leuven) and (ii) an approach based on representative volume elements (RVE’s) solved with a spectral numerical Fast Fourier Transform (FFT) technique (DAMASK software of MPIE). In the latter case, the RVE’s are defined in the form of a regular mesh of grid points in 3D-space with a crystal orientation assigned to each grid point. The experimental texture is derived from XRD pole figure measurements. Only tensile curves in the rolling direction are used to calibrate the hardening model included in the crystal plasticity models. For an interstitial free steel the predictions are compared to experimental yield loci, as obtained from uniaxial tests, plane strain tests, shear tests and through-thickness compression tests. Analytical FACET yield loci are calibrated to the crystal plasticity (CP) models and these are used in FE simulations of a square cup deep drawing process, validated by experiments.