From macro to micro: investigation of three-dimensional particle-scale responses of asphalt mixtures under non-uniform rolling tire loads via coupled FEM-DEM simulations

In response to escalating traffic volumes and climate-induced challenges, road infrastructure faces considerable transformations. The optimisation of asphalt pavement surface design, crucial for managing road network costs tied to wearing layer deterioration, necessitates a thorough comprehension of the interaction mechanism between tires and asphalt mixtures. Existing research predominantly adopts a continuum mechanics approach to study this interaction mechanism, yet it falls short in capturing adequately the inherent discontinuity nature of asphalt mixtures. In this study, a novel framework, employing a Finite Element Method-Discrete Element Method (FEM-DEM) coupling strategy, is developed for examining particle-scale responses of an asphalt mixture layer under realistic non-uniform tire loads. Through meticulous FEM-DEM coupling simulations, internal particle-scale responses are assessed precisely under rolling tire loads for both free rolling and full braking conditions. The robustness of this coupling strategy is validated by analyzing particle displacement distributions and the evolution of particle displacement and contact force networks. The results indicate that tire braking exerts notable effects on asphalt mixture responses, both in terms of particle displacement and particle contact force characteristics. Additionally, tire rolling velocity is shown to minimally affect particle contact force distribution but significantly influence average particle velocity and residual particle displacement. By seamlessly integrating FEM and DEM strengths, the proposed coupling strategy presents a promising avenue for micro-mechanical analysis of asphalt mixtures under rolling tire loads. This enhances an understanding of intricate mechanisms governing asphalt mixture deterioration, with substantial potential for refining asphalt pavement surface design.