Simulation of Fracture Behavior in Asphalt Concrete Using a Heterogeneous Cohesive Zone Discrete Element Model

TLDR

Cracking in asphalt pavements remains a major cause of deterioration, especially in cold climates, and although design tools exist, fundamental knowledge of asphalt concrete fracture is lacking. A clustered discrete element method incorporating a bilinear cohesive zone model was verified against continuum theory and double‑cantilever beam tests, enabling mesoscale representation of material morphology through high‑resolution imaging. Simulations matched experimental results and provided new insights into fracture mechanisms, revealing how heterogeneity, crack path, and local material strength influence the global fracture response.

Abstract

With increasing traffic loads and changes in crude petroleum refining techniques, cracking in asphalt pavements continues to be a major cause of structural and functional deterioration of these systems, particularly in cold climates. Although modern design tools such as the AASHTO Mechanistic Empirical Pavement Design Guide have recognized the need to predict pavement cracking in pavement life cycle cost analyses, the development of true fracture tests and associated models is hampered by a lack of fundamental knowledge of the physical nature of cracking in asphalt concrete materials. A clustered discrete element method (DEM) was employed as a means to investigate fracture mechanisms in asphalt concrete at low temperatures. The DEM approach was first verified by comparing elastic continuum theory and the discontinuum approach using uniform axial compression and cantilever beam models. A bilinear cohesive zone model was implemented into the DEM framework to enable simulation of crack initiation and propagation in asphalt concrete. Verification of the cohesive zone fracture model was carried out using a double cantilever beam. The main advantage of the DEM approach was that a mesoscale representation of the morphology of the material could be easily incorporated into the model using high-resolution imaging, image analysis software, and by developing a relatively simple mesh generation code. The simulation results were shown to compare favorably with experimental results, and moreover, the simulations provide new insight into the mechanisms of fracture in asphalt concrete. The modeling technique can provide more details of the fracture process in laboratory fracture tests, the influence of heterogeneity on crack path, and the effects of local material strength and fracture energy on global fracture test response.