Constrained Voronoi models for interpreting surface microstructural measurements

Abstract

Measurement and analysis of microstructures is an essential aspect of materials design and structural performance. In the case of surface experimental measurements such as digital image correlation (DIC), it is beneficial to know the subsurface microstructure to interpret the surface observations accurately. However, subsurface microstructures are expensive to obtain through three-dimensional (3D) tomography. Hence, it is of interest to generate these structures computationally. In this work, a generalized inverse Voronoi problem is used to grow an approximate representation of the 3D microstructure from a surface electron backscatter diffraction (EBSD) image. The novelty of the approach is that the surface microstructure is retained during the simulation. This technique is employed for the reconstruction of a recrystallized magnesium alloy microstructure. Crystal plasticity finite element modeling (CPFEM) was employed for comparing the predicted surface strains in the reconstructed 3D microstructures against experimentally measured data. It is observed that the surface strains of different 3D reconstructions are qualitatively similar to the experiment. However, strong basal slip activation in some subsurface grains can influence the choice of activated slip systems on surface microstructures. The results show the implications of performing a full 3D crystal plasticity analysis of measured surface data as compared to only analyzing a two-dimensional extruded microstructure.