A Multi-Scale Computational Mechanics Model for Predicting Rutting in Asphaltic Pavement Subjected to Cyclic Mechanical Loading

摘要

This paper presents a multi-scale computational approach for predicting damage in asphaltic pavement subjected to cyclic mechanical loading. The approach utilized herein employs continuum mechanics on three simultaneous lengths scales: the global scale of the roadway (m); the local scale of the aggregate (cm); and the micro-scale of micro-cracking (mm). Failure is predicted due to two forms of evolutionary and interacting energy dissipation: material viscoelasticity and micro-cracking in the asphalt binder. The model utilizes damage dependent homogenization techniques to construct averaged properties at the two smaller length scales, thus providing two-way coupling between the three length scales (global, local and micro). The general formulation of the modeling approach is described herein, together with a simplified methodology for predicting pavement failure due to rutting. As a means of illustrating the model, example problems are presented for typical roadways. These examples demonstrate how this technique may be utilized by pavement engineers to control design variables both at the global scale (such as layer depths) and the local scale (such as aggregate volume fraction and size distribution) so as to improve pavement design.