Generation of Models of Materials with Complex Structure

This study focuses on the generation of surrogate finite element models of the materials with complex microstructures produced by modern additive manufacturing. These materials often exhibit microscopically non-homogeneous structures characterized by irregularly shaped grains, porosity, and microcracks between grains. Traditional finite element modeling of such materials relies on advanced imaging techniques, followed by a labor-intensive process of geometry restoration and meshing. In contrast, our innovative approach leverages properties and statistics derived from real images and extensive empirical data to create simplified models that accurately represent the overall morphology of the original material. By rigorously testing our models against actual microstructural images and empirical datasets, we ensure a high degree of fidelity in capturing essential statistical properties, including volume fractions, porosity, and distributions of microcrack parameters (length, width, orientation). The code generates models with tens of millions of elements in minutes. This capability enables the mass generation of models across a range of parameters. This facilitates extensive parametric studies of material behavior, as discussed in the companion presentation. This streamlined model generation process enhances our understanding of how microstructural features influence material behavior, paving the way for improved material design and performance optimization.