Packing density (PD) of aggregates mixture is one the important and effective parameters in concrete mix design. Aggregate blends with high packing density (PD) require low amounts of paste to fulfill voids between aggregates. If the paste volume is fixed in concrete, the increase in packing density of the aggregate could increase the workability of the concrete at the same water/cementitious materials ratio (W/CM); or while maintaining the same workability, it could reduce W/CM and consequently increase the strength and durability of concrete. Different numerical models have been presented by researchers according to experimental tests to predict PD of aggregates mixture. During the two recent decades, researchers tried to predict PD of aggregates mixture through computer simulation. However, most performed computer simulations are not realistic and the rare accurate researches were accompanied with noticeable errors. In the current thesis, computer simulation was used to predict PD of aggregates mixture. Solving packing density problems with computer simulation is usually a time-consuming procedure and in some cases it is impossible. In this thesis, by analyzing the results obtained from computer simulation through PFC^3D software, a simplified method is presented to predict PD of aggregates mixture. In the proposed method, each graded aggregate that its grading and loose packing density are available from experiments is converted to monosized aggregates with equalized friction coefficient in such a way that PD of actual aggregates is equal to PD of monosized aggregates. Accuracy of the suggested method has been approved by different experimental tests. In this study, 700 independent runs have been done through PFC^3D software and a rich database was prepared for practical usage; in which, PD of each monosized aggregate with different friction coefficients is presented. Through this database, it is possible to determine friction coefficient for each monosized aggregate without any computer simulation. The suggested method has been compared with other numerical methods suggested by different researchers and the efficiency of the method has been approved. The developed formulations in this thesis which are based on computer simulation results make possible the prediction of packing density of aggregates mixture without computer simulation and only through simple calculations or programming.

Keywords: Aggregate, Packing Density, Computer Simulation.