Investigating the mechanical and electrical behaviour of particulate systems using discrete element method

Abstract

Electro-mechanical behaviour of granular materials represents a significant multi-physical problem, garnering considerable interest and extensive application across various industries. A multitude of experimental and discrete element method (DEM) studies have investigated the interaction mechanisms of mechanical and electrical behaviour in particulate systems from both microscopic and macroscopic perspectives. However, acquiring data at the grain-scale through experiments is challenging, and comprehensive analysis and formulation of models for overall electrical resistance between contacting objects (particle-to-particle or particle-to-wall) under mechanical loading are still needed. This research aims to develop understanding of the interaction between mechanical and electrical behaviour in particulate systems across both micro and macro scales. An electromechanical contact model is developed to represent both particle-to-particle and particle-towall interactions, characterised by particle resistances and contact resistances in overlapping regions. The modelling of contact resistance follows Hertz and Holm’s theories, while the particle resistance is estimated by considering the geometric transformations of particles under mechanical force. The model is verified and validated against analytical solutions and experimental data in the literature. The high pressure torsion (HPT) and drained triaxial monotonic compression tests are utilised in this thesis to systematically study electrical response characteristics under different mechanical loading conditions. A DEM model of HPT test is developed to examine the effect of sand particle characteristics on tribological behaviour at the wheel-rail interface, complemented by an in-house Python code to simulate realistic fragment size distribution following particle breakage. The electro-mechanical contact model is employed in the HPT test to examine the variations in electrical resistance at the wheel-rail interface when conductive and non-conductive sands are applied individually, as well as when a mixture of both particles is utilised. Furthermore, three samples of varying inherent anisotropies are compressed under drained triaxial monotonic condition to investigate the effect of anisotropy on the electrical responses of the particulate system.The findings indicate that the particle size and mass, the number of particles, and the fragment size distribution (FSD) have a significant effect on the tribological behaviour at the wheel-rail interface. Two commercial conductive particles can reduce HPT electrical resistance to below 10 Ω, and that mixing 5% conductive particles to silica sand dramatically drops the HPT resistance from 55 kΩ to 220 Ω. Drained triaxial monotonic compression tests on loose samples reveal that those with predominantly vertical contact orientations (Ver_L) show the fastest growth in deviator stress ratio (q/p’) and pronounced dilative behaviour, while those with predominantly horizontal contact orientations (Hor_L) exhibit the largest decrease in coordination number (CN) and minimal increase in mean contact area (A). Differences in bulk resistance trends across samples at small to medium strain levels converge to similar decreasing trends at residual state (45% axial strain).