Mesomechanical Prediction of Viscoelastic Behavior of Asphalt Concrete Considering Effect of Aggregate Shape

Yiren Sun,Zhuang Zhang,Xin Wei,Cong Du,Mingyang Gong,Jingyun Chen,Hongren Gong

DOI: https://doi.org/10.1016/j.conbuildmat.2020.122096

IF: 7.4

2021-01-01

Construction and Building Materials

Abstract:Accurate prediction of effective (overall) viscoelastic properties and mesoscopic responses of asphalt concrete is crucial to investigating the mechanical behavior and damage mechanisms of asphalt pavement. The present study evaluated the effect of aggregate shape on mesomechanical prediction of viscoelastic behavior of asphalt concrete. Two two-dimensional (2D) computational mesomechanics methods, i.e., a random aggregate generation (RAG)-based method and a digital image processing (DIP)-based method, and a three-dimensional (3D) analytical two-layer built-in (TLB) mesomechanics method were adopted. In the RAG-based approach, four regular aggregate shapes (circle, hexagon, pentagon and square) and irregular aggregate shapes were considered. The results showed that when asphalt concrete was treated as a composite consisting of coarser aggregates larger than 2.36 mm, asphalt mortar and air voids, the volume fraction of coarse aggregates was a more critical factor than their shapes for the effective dynamic modulus prediction. The 3D TLB method performed better in predicting the macroscopic effective dynamic modulus of asphalt concrete than the 2D methods, whereas the 2D methods generated superior effective phase angle predictions. The RAG-based model with irregular aggregates and the DIP-based model exhibited significantly higher maximum principal stresses on the mesoscale than those from the RAG-based models with regular aggregates under the compressive loading condition, indicating the importance of applying the realistic aggregate shapes in the mesoscopic stress and damage analysis of asphalt concrete. (C) 2020 Elsevier Ltd. All rights reserved.