Project outline
Vasileios’ doctoral research project involves the numerical analysis of granular media, employing the Discrete Element Method and non-spherical particles. The Discrete Element Method, originally developed to investigate the behaviour of geomaterials, is now utilised in several industries, like process engineering, mining and packaging of pills, but to name a few. Most 3D DEM software packages still use spherical particles; yet in reality particles are everything but spheres.

Recently, a new class of algorithms based on linear programming have been introduced for non-spherical convex particles, called ‘potential particles’, that are described by a combination of flat surfaces with either sharp or smooth edges. These algorithms have been proven successful in a variety of rock mechanics problems, where blocks tend to be ‘sharp’ polyhedra. The challenge still ahead is to employ these algorithms for geomaterials (such as ballast, sands and silts) and the irregular particles generated by mining and process engineering.

During this research project, a methodology is being developed, in order to determine the mathematical description of the potential particle that best fits particle shapes from literature data and work out what is the required fit to replicate the main mechanical properties at the scale level of a bulk assembly. To assess the bulk properties, numerical DEM analyses on the reconstructed bulk assembly are ran. The methodology is coupled to state of the art image recognition techniques, currently utilised to scan particles, to obtain a seamless tool from image recognition to DEM simulations.