Experimental Investigation and Nonlocal Modeling of Compaction of Hot Asphalt Mixtures

Abstract

Compaction is the most critical step in the construction process of asphalt pavements, which has a significant consequence on the durability of asphalt pavements. However, people’s understanding on compaction of asphalt mixture is still very limited. The objective of this thesis is therefore to explore physical mechanisms of the compaction of asphalt mixtures and to model the compaction process based on its physical mechanisms. Before exploring mechanisms of compaction, a statistical analysis of the field density data is performed to achieve a thorough understanding of the current situation of field compaction. Then, physical mechanisms are investigated in different length scales. In the length scale of the thickness of asphalt binder films (∼ 10 µm), the lubricating behavior of asphalt binder between rough aggregates is believed as the mechanism that affects the compactability of asphalt mixtures. A tribological test is developed to measure the lubricating effect of asphalt binder between rough surfaces. The results show a strong effect of asphalt binder lubrication on improving the compactability of asphalt mxitures. In the length scale of asphalt mixtures (∼ 10 cm), two mesoscopic mechanisms (aggregate rearrangement and aggregate-binder interaction) are proposed to explain the macroscopic phenomena observed in gyratory compaction process. Based on the mechanism of aggregate rearrangement, a 1D nonlocal model is then proposed to simulate the gyratory compaction process of asphalt mixtures. The results show that the model is able to simulate the compaction curve, the density profile, and the size effect observed in gyratory compaction tests. Lastly, the 1D nonlocal model is employed to study the effects of randomness in the initial density profile and compaction effort on the compaction process of asphalt mixtures. The results can explain some key features of the probability distribution of field density data. The results of this study have the potential to be used for the design of more compactable asphalt mixtures and the prediction of asphalt pavement compaction.