Computer Modelling of First Stage Solid State Sintering at Particle Scale

Sintering, originated in ancient civilization thousands of years ago, is a typical process applied to ceramic and metal component manufacture by heating material powders until particles are fully densified and adhere to each other. With the development of controlled sintering of metal and ceramics in the early twentieth century, the technological background for modern theory and practice of sintering was established, and then experienced fast growth after the mid-1940s. In sintering study, empirical and theoretical ways are used to understand sintering behaviours and connect behaviours to sintering variables. The experimental studies and theoretical analyses have been performed quite well over the last 60 years when providing an outstanding qualitative understanding of sintering in terms of the driving forces, the mechanisms, and the influence of major sintering variables such as the particle size, temperature and applied pressure. However, the analytical solutions of experiments and the models of sintering are not so successful when providing a quantitative description for most sintering. With the development of computer technology, the powerful computing capability of computer allows us to build a complex sintering model with more sintering variables that provide a more accurate description on the effects of sintering variables. In this thesis, a computational model is firstly presented to validate and correct the previous assumption of fast surface diffusion for modelling first-stage sintering by coupled grain-boundary and surface diffusion. There are two main achievements in this study. To be specific, the first one is providing a numerical solution for the curvature at a triple junction (pore tip), and the second one is demonstrating the effect of surface diffusion on first-stage sintering by coupled grain-boundary and surface diffusion. Then a temperature-dependent model is presented to study the effect of the fast heating rate on sintering. Based on the analysis of numerical results, the effect of spark plasma sintering on densification can be partially attributed to its fast heating rate. Finally, a computational model is presented to qualify and quantify the effects of inhomogeneity concerning sintering variables on sintering kinetics. The work in this thesis is entirely computational and guided by existing experimental data and observations in the literature.