Computational Analysis of Thermal Conductivity of Asphalt Mixture Using Virtually Generated Three-Dimensional Microstructure

This study developed computational models for virtually generating three-dimensional (3D) random asphalt mixture microstructure (3D RAMM) using a Voronoi diagram, and analyzed the thermal conductivity of heterogeneous microstructure. The mixture was simulated as a three-phase composite material including fine aggregate matrix (FAM), coarse aggregate, and air void. The 3D RAMM was generated considering different morphological characteristics (size distribution, sphericity, orientation) of the dispersed phases (coarse aggregate and air void) using virtual sieve analysis and an innovative shrinking algorithm. Finite-element (FE) modeling of the steady heat transfer process was used to calculate the effective thermal conductivity of an asphalt mixture. The thermal simulation results were validated with experiment data reported in the literature. The results show that the heat flux in the asphalt mixture is not uniformly distributed. The sphericity and the orientation angle of the dispersed phase have a combined effect on the effective thermal conductivity of asphalt mixture. The developed computational model provides an effective way to analyze the microstructure effect on the bulk properties of composite materials.