A two-dimensional random aggregate structure generation method: Determining effective thermo-mechanical properties of asphalt concrete

This study presents a 2D random aggregate structure generation algorithm for asphalt concrete. It involves two steps: first, the "convex-concave surface method" is proposed to generate random aggregate particles; second, the "double background grid method" is developed to place the generated aggregates and form the asphalt concrete mesostructure. The convex-concave surface method can produce aggregate particles randomly with both irregular convex and concave surfaces, enabling the generated aggregates to look realistic. Very few parameters are required during the particle generation, significantly simplifying the operation of input parameters. The double background grid method employs two types of background grids: the coarse grids are used to determine the placing position of the particle, whereas the fine grids are used to inspect if the particle overlaps with the placed ones; in this way, it effectively improves the placing efficiency. Besides, it allows for a predefined minimum gap between the adjacent particles, successfully guaranteeing the quality of finite element meshing. Based on the developed 2D random aggregate structure of asphalt concrete, a unified computational methodology for determining the effective thermo-mechanical properties is presented. The effectiveness of these methods is demonstrated using experimental data and typical values reported in the literature.