Use this URL to cite or link to this record in EThOS:
Title: The developments of multi-level computational methodologies for discrete element modelling of granular materials
Author: Zhao, Tingting
ISNI:       0000 0004 7972 4000
Awarding Body: Swansea University
Current Institution: Swansea University
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Granular materials are prevalent in this world while their non-trivial behaviour, which may resemble solid, liquid and/or gas under different circumstances, is still poorly understood. The challenging mechanics and dynamics of granular materials combined with their ubiquity have made this topic especially interesting to study. The discrete element method (DEM) is a reliable and effective numerical technique to model many scientific and engineering problems involving granular materials but it is still not a fully mature method. Considering the unique properties of granular materials and the inadequate features of the DEM, this thesis improves the current DEM from three different aspects and scales. On the micro scale at the particle level, a novel contact model is developed by introducing the statistical Greenwood Williamson (GW) model which can consider the stochastic surface roughness of particles. Two non-dimensional forms of the original formulations are derived which can reduce the computational costs significantly. A Newton-Raphson based numerical solution is proposed which can solve the inter-dependence problem involved. A theoretical inconsistency of the classic GW model is deduced which leads to the development of the extended elastic GW (E-GW) model. An empirical normal contact law is obtained by the curve-fitting method and can be incorporated into the DEM code to conduct the one and three dimension compression tests. An extended elastic-plastic GW (EP-GW) model is developed to allow the plastic deformation at the asperities. Furthermore, the tangential contact model and thermal conductivity model are proposed. On the meso scale at the sample level, a new packing characterisation method is proposed based on the digitalised image matrix of a packing and the subsequent application of the principal component analysis (PCA) with which the configuration of the particle assemblies can be evaluated quantitatively. The procedures of the packing digitalisation and formation of packing image are established for both 2D and 3D cases. The obtained PCA results of the packing image matrix can be revealed by the proposed principal variance function (PVF) and dissimilarity coefficient (DC). The values of PVF and DC can indicate the magnitude of effects on a packing caused by the configuration randomness, the particle distribution, the packing density and the particle size distribution. The uniformity and isotropy of a packing can also be investigated by this PCA based approach. On the macro scale at the level of real industrial applications, the existing coarse graining methods are carefully analysed by the exact scaling law and the effective thermal properties of particulate phase change materials are derived by the homogenisation method. An enthalpy based discrete thermal modelling framework for particulate systems with phase change materials is developed which can consider both the heat conduction process and the phase change transition. This proposed methodology is assessed by solving a particle version of the classic one-phase Stefan melting problem. Additional numerical simulations are also conducted to illustrate the effectiveness of this modelling framework.
Supervisor: Feng, Yuntian T. Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral