Date of this Version
Computational microstructure models have been actively pursued by the pavement mechanics community as a promising and advantageous alternative to limited analytical and semi-empirical modeling approaches. The primary goal of this research is to develop a computational microstructure modeling framework that will eventually allow researchers and practitioners of the pavement mechanics community to evaluate the effects of constituents and mix design characteristics (some of the key factors directly affecting the quality of the pavement structures) on the mechanical responses of asphalt mixtures. To that end, the mixtures are modeled as heterogeneous materials with inelastic mechanical behavior. To account for the complex geometric characteristics of the heterogeneous mixtures, an image treatment process is used to generate finite element meshes that closely reproduce the geometric characteristics of aggregate particles (size, shape, and volume fraction) that are distributed within a fine aggregate asphaltic matrix (FAM). These two mixture components, i.e., aggregate particles and FAM, are modeled, respectively, as isotropic linear elastic and isotropic linear viscoelastic materials and the material properties required as inputs for the computational model are obtained from simple and expedited laboratory tests.
In addition to the consideration of the complex geometric characteristics and inelastic behavior of the mixtures, this study uses the cohesive zone model to simulate fracture as a gradual and rate-dependent phenomenon in which the initiation and propagation of discrete cracks take place in different locations of the mixture microstructure. Rate-dependent cohesive zone fracture properties are obtained using a procedure that combines laboratory tests of semi-circular bending specimens of the FAM and their numerical simulations. To address the rate-dependent fracture characteristics of the FAM phase, a rate-dependent cohesive zone model is developed and incorporated into the mainframe of ABAQUS in the form of a customized user element (UEL) subroutine. The applicability of the rate-dependent microstructure fracture model is demonstrated and a parametric analysis is performed to evaluate the effects of different mixture parameters on the mechanical behavior of virtually generated hot-mix asphalt (HMA) microstructures. The results presented in this research demonstrate that computational microstructure models, such as the one developed in this study, have a great potential to become efficient design tools for asphalt mixtures and pavement structures.