M Fakhri - Assistant Professor, K.N.Toosi University of Technology, Department of Civil Engineering
T Kheiry - M.Sc. Student, K.N.Toosi University of Technology
A.A Mirghasemi - Associate Professor, University of Tehran

To gain further insight into the interaction among the mix constituents, researchers utilized numerical simulation techniques, such as the Finite Element Method (FEM) and the Discrete Element Method (DEM). The Discrete Element Method was primarily introduced for the analysis of rock-mechanics and granular materials. In recent years, DEM has also been applied to simulate the properties of asphalt concrete. For the purpose of discrete element modeling of the materials two different mechanisms have been extensively employed, namely random geometry generation and importing pre-generated models. In order to gain more realistic simulation it is preferred to enhance more precise initial geometric models which are capable to represent the shape and arrangement of the constituents. For this purpose, two and three dimensional imaging techniques have been investigated by some researchers. In this paper, a userdefined 2D image processing technique was developed to generate the initial geometry of asphalt concrete mixtures. Asphalt concrete specimens were prepared in the laboratory and cut smoothly prior to optical scanning. The digital images were then processed and de-noised by defining the appropriate thresholds. The processed images were employed to generate the initial model for DEM. The presented methodology in this paper enables researchers to have a better comparison between the simulations and the experimental laboratory tests. It also facilitates defining clusters of circular particles to provide better representation of aggregate shape and angularity which yields significant benefits for numerical modeling techniques

Discrete element, Image processing, Numerical modeling, Asphalt Concrete