Computer Simulation And Property Analysis Of Random Close Packing Algorithm Of Unequal Ellipses

In deep investigation of properties of materials, computer simulation plays a very important role. In consideration of some existing problems in random close packing of elliptical particles, the aspects on which we focus are as:An effective computer numerical method, including equations and a numerical algorithm, is developed to determine the probability density function of elliptical particles. The results are analyzed by statistics methods, and the findings indicate that the algorithm can simulate the random close packing of elliptical particles.Random close packing of elliptical particles whose areas obey log-normal distribution is investigated. A Monte Carlo algorithm is used to simulate the packing procedure. A numerical algorithm, based on the relationship between the orientation of an ellipse and the overlapping area with its neighbors, is developed to reduce or eliminate the overlap between intersection ellipses by rotating the orientations of the ellipses. Except the overlap elimination module, the other significant part which is composed of numerical algorithm is the overlap detection module. The module judge whether overlap occurs by the number of real roots of the simplified equation set of ellipses, and the coordinates of common points of intersection ellipses is its byproduct. By the transformation of coordinates, the simplification process changes two intersection ellipses into one standard ellipse and one circle. The aim is to implement the solution procedure by computer languages. The simulation is an iteration process which consists of three steps: initialization, compression and rearrangement, and the exit of iteration is the compression rate.Statistics methods are employed to analyze the simulation results. The relationship between the packing density and the aspect ration is investigated. The Kolmogorov-Smirnov test confirms statistically that the packing is isotropic. Finally, we employ the time series analyses technique to verify the randomness of the packing results.