Feasibility of using a physics engine to virtually compact asphalt mixtures in a gyratory compactor

Optimization of asphalt mixtures in the laboratory is essential to maximize its performance in a pavement structure and reduce the likelihood of expensive premature failures. Several material and mixture variables affect the performance of an asphalt mixture. Evaluating the innumerable combinations of these variables in the laboratory to optimize an asphalt mixture is onerous. Computational models can supplement such laboratory based optimization studies due to their capability of executing a broad range of parametric and probabilistic analyses. This study demonstrates the potential of using a physics engine to simulate the asphalt mixture compaction process. Physics engines were originally developed for animation and video games, however, with improvement in their accuracy, they are currently being used in many research areas. This study presents a computational model developed using an open source physics engine to virtually compact asphalt mixtures in a gyratory compactor. This model uses 3D laser-scanned representations of real aggregate particles and incorporates the viscous and cohesive nature of the asphalt mortar through customized subroutines. To demonstrate the feasibility of this model, the effect of the three key compaction parameters (i) angle of gyration, (ii) compaction pressure, and (iii) specimen height on the compaction characteristics were analyzed and compared with laboratory findings from various studies. Results suggest that the developed model can be used as a reasonably accurate means to simulate asphalt mixture compaction. This study also demonstrates that such simulations can be carried out using realistic and representative particle shapes in 3D, in a reasonable time using desktop computers.