Random packings via mechanical contraction

Abstract

Random packings of particles are size independent, exhibiting the same features in the meso and in the nanoscale. These particle configurations are widely present in nature and technology, and their study constitutes an interesting and challenging topic.

One of the ways of inducing random configurations of granular matter is via computer simulations using a Mechanical Contraction Method (MCM), explained and analyzed in details in the present thesis. A more robust base has been provided to the method and its applicability has been extended to hard boundary conditions and superquadric shapes. The contact detection developed for superquadrics finds applicability beyond the results presented in this project.

An optimization of the set of parameters inherent to the MCM has been performed. For rods, a physical relation for the amount of overlap removed by means of translations or rotations has been developed. The outcome from the algorithm has been used to construct a glass-transition equation of state for rods of any aspect ratio. The influence of hard boundary conditions in random packings of rods has been connected with experimental results. Finally, random configurations of superballs have been generated and characterized.

The roots of the MCM has been revised. This has provided insight about its limitations and advantages.