Abstract
Numerical modeling of asphalt concrete and other asphalt mixes used in road engineering is an actively developed research field. In this study, a framework combining the following aspects is presented: (1) reliable reconstruction of the real samples; (2) using realistic material models of the microstructure constituents; and (3) providing high numerical efficiency. Asphalt concrete microstructure was reconstructed using image processing. The Burgers material model was applied to the subdomains identified as the mastic, and the linear elastic model was used for the aggregate particles. In order to increase the numerical efficiency, the developed homogenization method was used to accelerate the finite element analysis. The main novelty of this study is the integration of the Burgers material model with the numerical homogenization in the small strains range. A homogenization error measured in the maximum norm was smaller than 7% in the presented numerical examples (6.8% for the elasticity and 6.9% for the viscoelasticity problem, respectively). Simultaneously, the observed reduction in the number of degrees of freedom was larger than 510 times. The obtained results confirmed the applicability of the developed methodology to the analysis of the viscoelastic materials in the range of the small strains.