The impact of heterogeneous microstructural features on crystal plasticity modeling of plastic anisotropy

Crystal plasticity finite element models typically assume each grain starts with a spatially uniform dislocation distribution, meaning each grain is initialized with a single lattice orientation and all grains share the same slip resistance. This study assessed that assumption by comparing crystal plasticity simulations of tensile tests against experiments on an extruded aluminum 7079 alloy. The simulations utilized computational microstructures with 'pancaked' grains and a significant degree of crystallographic texture, consistent with experimental measurements. In one case, the computational microstructures were assigned a single slip resistance and a single lattice orientation for each grain prior to the tension tests. In another case, the extrusion process prior to the tension tests was simulated to cause nonuniform slip resistances, nonuniform crystal orientations within each grain, and nonuniform slip resistances at each material point. Both approaches produced reasonably accurate predictions of the measured yield stress and lateral strain ratio anisotropy, but the predictions considering the initial nonuniformity exhibited less prediction error. Other factors, such as latent vs self hardening of slip systems and fine vs coarse grains, did not have as large an impact. These results suggest that dislocation heterogeneity should be experimentally characterized and fed into future crystal plasticity simulations.

Automated summary

Crystal plasticity finite element models typically assume uniform dislocation distribution within grains, but this study found that considering the initial nonuniformity of dislocations led to more accurate predictions in simulations comparing to experiments.