The effect of different random number distributions on the porosity of spherical particles

A geometrical loose sphere packing model is initiated by randomly generating initial image coordinates of every sphere based on a specific distribution. In general, a normal distribution is applied in such models. However, other distributions, such as exponential, log-normal, and arbitrary distributions, may give better packing of particles, resulting in lower porosity values. In this study, the log-normal distribution provides the lowest porosity (40.72%) for mono-sized sphere packing. The same behavior is also shown in multi-sized sphere packing using a log-normal distribution.