Masters Research - Master of Philosophy (Engineering) / Thermal conduction properties are of major concern for those metal/ceramic composite materials having applications in semiconductor devices and electronic packaging materials. A higher thermal conductivity to coefficient of thermal expansion ratio is an advantage for such materials employed in electronic devices due to the subjective high thermal loads. It is well known that the shape, size and distribution of the insulating phase have an effect on the overall thermal conductivity properties. But the details are lacking and well deserving of study. Metal/ceramic oxide interfaces are important in the strengthening mechanisms of dispersion strengthened materials. Accordingly, considerable attention has been given to recent investigations of oxygen diffusion characteristics and the bonding mechanisms at such interfaces. Susceptibility to oxidation can be studied by analysing several thicknesses of material. As an example, studying a thin film and a semi-infinite material subjected to a high oxygen partial pressure environment and a vacuum condition would help to determine the oxidation (in-diffusion) and de-oxidation (out-diffusion) processes respectively. Since metal/ceramic internal interfaces play a very important role in controlling the mechanical, thermal and electrical properties, it is timely to consider these diffusion processes for detailed study. In this Thesis, the two areas mentioned above were selected for detailed investigation. The Thesis also addresses the further development of a method for solving complex phenomenological diffusion problems. This method makes use of lattice-based random walks of virtual particles, directed according to the Monte Carlo method (the Lattice Monte Carlo method) which is then used to address various mass and thermal diffusion processes. Chapter 2 is concerned with using this method to determine the thermal conductivity of model composites. In that chapter, the Lattice Monte Carlo method is used to calculate the effective thermal conductivity of several models of a composite, where inclusions are arranged in square planar and cubic arrangements with periodic boundary conditions. Excellent agreement is found of the effective thermal conductivity with the century-old Maxwell-Garnett Equation. Chapter 3 is concerned with a phenomenological representation of oxygen diffusion and segregation in a model composite based on Ag/MgO. The Lattice Monte Carlo method is employed to address mass diffusion in this composite. Square and randomly distributed multiple inclusions were considered as shapes of the MgO inclusion phase. The time-dependence of oxygen concentration depth profiles and contour maps were determined. First, oxygen in-diffusion is considered from a constant surface source solely into the Ag metal matrix: oxygen depth profiles were in excellent agreement with exact results. Next, oxygen in-diffusion/segregation is simulated in the composite by permitting and restricting the mobility of oxygen in different scenarios involving the Ag-MgO interface. The (higher temperature) out-diffusion of oxygen from the composite was also simulated and corresponding results obtained for the oxygen depth profiles. In both cases, very good agreement was found between the results from the Lattice Monte Carlo method and analytical expressions.
| Identifer | oai:union.ndltd.org:ADTP/280707 |
| Date | January 2008 |
| Creators | Muthubandara, Nilindu |
| Source Sets | Australiasian Digital Theses Program |
| Language | English |
| Detected Language | English |
| Rights | Copyright 2008 Nilindu Muthubandara |
Page generated in 0.0018 seconds