An image-based method for modeling the elasto-plastic behavior of polycrystalline microstructures based on the fast Fourier transform

An efficient full-field method of computing the local and homogenized macroscopic responses of elasto-plastic polycrystalline microstructures based on the fast Fourier transform (FFT) algorithm is presented. This approach takes realistic microstructure images as the input and estimates the mechanical response/properties of polycrystal microstructures under periodic boundary conditions without requiring complex discretization. Effective stressstrain response, local mechanical response fields and crystallographic texture of deformed microstructures are examined. Special interest is given to the fatigue indicator parameters (FIPs) of IN100 (nickel-based superalloy). This approach accounts for both intergranular and intragranular interactions of an elasto-plastic heterogeneous medium and therefore provides accurate prediction of the mechanical response and texture evolution. The elastic and plastic responses are computed separately by satisfying two forms of the equilibrium equations. A multi-grid strategy is also adopted to capture the heterogeneous deformation of microstructures. The obtained results are compared with a widely used crystal plasticity finite element approach. The gained computational efficiency of full-field polycrystalline microstructure simulation is prominent.