An assessment of polarized light microscopy as a characterization method for crystal plasticity simulations

Abstract

Polarized light microscopy is a large-area, high-resolution, high-throughput microstructural characterization method for polycrystalline materials comprised of hexagonal close-packed crystals. Due to the fact that polarized light microscopy only determines the orientation of a crystal’s c axis, it is necessary to assess the applicability of this new characterization modality for use in integrated computational materials engineering workflows for simulating the deformation response of polycrystalline materials. We present a computational study in which the effect of this orientation ambiguity on the predictions of crystal plasticity finite element simulations is quantified. We focus on an idealized polycrystalline sample with random texture, from which a number of c axis-similar samples are generated, each with random rotations about each constituent crystal’s c axis. We scrutinize the differences in stress field predictions between the reference sample and the randomly altered samples during monotonic tensile tests, as well as the spatial differences in predictions. Our findings indicate that differences are lowest in the elastic regime, and increase dramatically during the macroscopic elastic–plastic transition. We further find that results do not exhibit a strong spatial dependence, indicating that orientation and neighborhood are the primary causes of differences in stress field predictions.