Increasing power dissipation and chip densities in the rapidly evolving electronics cooling industry are causing an ever increasing need for the tools and methods necessary for electronic systems design and optimisation. Modern electronic systems have the capacity to produce significant amounts of heat which, if not removed efficiently, could lead to component failure. The most common technique of heat removal is by making use of a heat spreader, or so¬-called heat sink. These devices are excellent heat conductors with a large surface area to volume ratio, and cooled through either natural or forced convection. Despite the advantages of these devices, there are serious consequences involved in the application of heat sinks. The required size of a heat sink may limit the miniaturisation of a product, while inadequate design, due to a lack of understanding of the flow physics, may lead to premature component failure. It is therefore crucial that an optimal heat sink design is achieved for every particular application. In the past, both heat sink design and optimisation have occurred mostly through experimental characterisation of heat sinks, which was not always particularly successful or accurate. Recent rapid developments in computer technology have led to the availability of various computational fluid dynamics or CFD software packages, with the capability of solving the discretized form of the conservation equations for• mass, momentum, and energy to provide a solution of the flow and heat fields in the domain of interest. This method of using the fundamental flow physics is currently the most complete way to determine the solution to the heat sink design and optimisation problem. It does unfortunately have the drawback of being computationally expensive and excessively time consuming, with commercial software prices being financially restrictive to the average designer. The electronics cooling community has subsequently identified the need for so-called "compact models" to assist in the design of electronic enclosures. Compact models use available empirical relations to solve the flow field around a typical heat sink. Current models require significantly less computational power and time compared to CFD analysis, but have the drawback of reduced accuracy over a wide range of heat sink geometries and Reynolds numbers. This is one of the reasons that compact modelling of heat sinks remain an international research topic today. This study has focused on the CFD modelling of a variety of forced flow longitudinal fin heat sinks with tip clearance. Tip clearance allows the flow to bypass the heat sink and downgrade its thermal performance. The flow bypass phenomenon, general flow behaviour, and pressure loss characteristics were investigated in detail. Thermal modelling of the heat sinks was left for future study. The flow information provided by the CFD analysis was combined with data available from literature to develop an improved compact flow model for use in a variety of practical longitudinal fin heat sinks. The new compact model leads to a 4.6 % improvement in accuracy compared to another leading compact model in the industry, and also provides more localised flow information than was previously available from compact modelling. <p The study therefore contributed significantly towards the general understanding and prediction of forced flow behaviour in longitudinal fin heat sinks with tip bypass, using both CFD analysis and the compact modelling approach. The new improved compact model may now be extended and incorporated together with the relevant flow details from the CFD analysis in a total package, solving for the flow and heat fields of forced flow longitudinal fin heat sinks. The study therefore assists in the global effort of making the confident and accurate use of compact modelling in modem electronic systems design and optimisation a practical reality.
Dissertation (M Eng (Mechanical Engineering))--University of Pretoria, 2007.