Adaptive Three-Dimensional Aggregate Shape Fitting and Mesh Optimization for Finite-Element Modeling

Abstract

Numerical representations of aggregates that fit the shape of realistic grains reasonably with optimized mesh are preferred in numerical simulations of asphalt mixtures. This paper proposes a three-dimensional (3D) finite-element (FE) modeling approach for realistic aggregates to obtain FE models that represent the shape of grains adaptively with tetrahedron elements of good quality. The methodology comprises two main steps: (1) 3D solid models are developed from X-ray computed tomography (CT) images of aggregates, and the aggregate surface is transferred to triangle facets with a predefined surface deviation tolerance; and (2) tetrahedron elements are adaptively constructed based on the obtained facets under the control of element quality and quantity. More than 270 grains were reconstructed to obtain FE representations using the proposed method, and the quality and quantity of tetrahedrons in FE models were analyzed. Results indicate that the proposed approach supports adaptive FE modeling for aggregates with controllable shape-fitting accuracy, element quality, and quantity, which facilitates reliable and efficient FE modeling in micromechanical simulations of asphalt composites.