Efficient 3D porous microstructure reconstruction via Gaussian random field and hybrid optimization

摘要: Summary Obtaining an accurate three-dimensional (3D) structure of a porous microstructure is important for assessing the material properties based on finite element analysis. Whereas directly obtaining 3D images impractical under many circumstances, two sets methods have been developed in literature to generate (reconstruct) from its 2D images: one characterizes certain statistical descriptors, typically two-point correlation function and cluster function, then performs optimization process build that matches those descriptors; other method models using stochastic like Gaussian random field generates function. The former obtains relatively microstructure, but computationally can be very intensive, especially problems with large image size; latter quickly sacrifices accuracy due issues numerical implementations. A hybrid approach modelling isotropic two-phase materials proposed this paper, which combines hence maintains correlation-based improved efficiency. technique verified reconstructions silica polymer composite different volume fractions. comparison reconstructed microstructures histories both original our demonstrates efficiency approach.