Performance prediction of asphalt mixture based on dynamic reconstruction of heterogeneous microstructure

Abstract This paper proposed an algorithm to reconstruct the microstructure of asphalt mixture considering dynamic arrangement process of aggregate particles. In this method, asphalt mixture was considered to be composed of dispersed phase including coarse aggregate and air voids, and continuous phase including fine asphalt matrix (FAM). The dispersed phase materials were characterized by high-dimensional polyhedral particles. In order to reconstruct the microstructure by 1:1 in volume fraction, three basic parameters that determine the heterogeneous structure of asphalt mixture were abided in the design of digital specimen. Then, impulse-based rigid collision response was introduced to simulate the rearrangement of aggregate particles in the compaction process of asphalt mixture. In particular, the applicability of the proposed method to typical structural types of asphalt mixture was discussed. For the asphalt mixture with high air void content, the reconstructed specimen may deviate from the real structure. Furthermore, the dynamic modulus and thermal conductivity of the generated specimens were predicted by finite element method (FEM) and compared with the experimental values. The predicted dynamic modulus and thermal conductivity coincide with the experimental results well, which indicates that the proposed method is reliable for reconstructing the heterogeneous structure of asphalt mixture with relatively low content of air voids.