Crystal plasticity simulations using nearest neighbor orientation correlation function

Abstract A probabilistic scheme is presented for simulating evolution of polycrystalline microstructures during deformation. Microstructure images are described using a compact descriptor called the nearest-neighbor conditional orientation correlation function, defined as the probability density of occurrence of a crystal orientation at one pixel distance from a known orientation. The neighborhood information obtained from this function is used to correct a Taylor-based formulation of crystal plasticity. A finite differencing scheme is developed to capture equilibrium of each orientation in an average sense. The predictions of textures and stresses using our approach are compared against crystal plasticity finite element model of a planar polycrystalline microstructure. We find that the new descriptor is able to capture texture components that are otherwise missed by the Taylor model and provides consistent improvements in the prediction of reorientation and stresses. The simulation speed is significantly faster than crystal plasticity finite element method and is more comparable to that of Taylor models.