Finite Element Modeling of Hot-Mix Asphalt Performance in the Laboratory

Abstract

The theoretical investigation of Hot-Mix Asphalt (HMA) response in the dynamic modulus and the semi-circular bending (SCB) laboratory test procedures is necessary to understand the influence of various design parameters on the performance of the mix. In addition, laboratory tests such as the dynamic complex modulus assume that this material can be dealt with as a homogeneous material while overlooking the particulate nature of this composite. The ultimate goal of this study is to develop an advanced theoretical framework, based on three-dimensional (3D) finite element (FE) methods and digital image analysis techniques to describe the behavior of HMA in two laboratory test methods: the dynamic complex modulus test and the SCB test. The developed models were validated against experimental testing results and were used to investigate the mix constitutive behaviors in these important laboratory test methods. In addition, the cracking and damage propagation in the SCB test was investigated using cohesive zone modeling techniques. The SCB test process as well as the propagation of damage was successfully simulated using 3D FE and cohesive elements. Based on the results of the FE model, it was determined that damage propagates in the vicinity of the notch and is mainly caused by a combination of vertical and horizontal stresses in the specimen. The effect of shear was negligible in progressing damage in the specimen. In addition, the results of FE model indicated that dissipated energy due to fracture is the predominant factor controlling failure in the SCB test. A fatigue crack propagation model was developed to predict the number of cycles to failure based on a cyclic SCB test and the generalized J-integral approach. A three-dimensional (3D) heterogeneous model was developed to describe the response of asphalt mixtures in the dynamic complex modulus test using an X-ray computed tomography (CT) image-based FE modeling approach. Results of the model showed that most of the deformations during the dynamic modulus test are derived from the mastic, which controls the viscoelastic behavior of the composite. In addition, the influence of mastic is more pronounced than the aggregates in the behavior of the mixture in the dynamic modulus test.