Geometry-considered 3D pseudorandom grain-scale modelling for crystalline material miniature parts

Abstract

Grain-scale modelling and simulation is of vital importance during the full lifecycle of crystalline material miniature parts, whose mechanical responses depend on the actual material microstructure, part geometry and their correlation. Unfortunately, current studies fail to comprehensively consider indispensable microstructure information, and lack the consideration regarding part geometry. Therefore, this paper proposes and implements a geometry-considered 3D pseudorandom grain-scale modelling method, which takes into account the necessary actual microstructure information, part geometry and their mutual relations to generate accurate and efficient microstructure-dependent crystal plasticity finite element (M−CPFE) models. Firstly, the 3D actual grain-scale microstructure was acquired efficiently by combining 2D actual microstructures. Then the geometry-considered grain-scale microstructure was created pseudorandomly based on the actual 3D grain size, orientation and misorientation distributions. On this basis, the M−CPFE model was constructed accurately. The proposed method has been proved to have higher accuracy and robustness than the Voronoi random method on macroscopic response prediction, microstructure generation and microstructure evolution prediction. Furthermore, the applicability in micro-manufacturing and service has been demonstrated by micro sheet bending and micro impeller rotation simulations. There is a great potential for the full lifecycle tracking of crystalline material miniature parts with the help of the proposed method.