(66d) Multi-Objective Optimization of a Novel Conductive Path for Area-to-Point Heat Conduction Problem

The increase of power density in electronics has prompted researchers to investigate how to effectively remove the generated heat to avoid the damage of devices. Designing a highly conductive path to remove the heat to the heat sink is a feasible method which can be abstracted to an area-to-point heat conduction problem. Such a problem can also be found in other field like thermal conductivity enhancement of energy storage media. It has been proved that the tree-network structure may be the optimal conductive path which is hardly to be manufactured practically. Therefore, an effective and easy-to-manufacture conductive path, refined from the tree-network structure, is designed including a platform adjacent to the heat sink and several branches extending to the corner. Its geometry can be described by four design parameters, involving the width and the height of the platform, the number of branches and the level of branches extending to the boundary. A multi-objective optimization is necessary to obtain the most effective heat conduction performance at the lowest cost of the quantity of high conductive material. Hence, response surface methodology and CFD method were applied to acquire the relationship between the heat conduction performance (average/maximum temperature) and the quantity of high conductive material (average thermal conductivity). The Pareto front and the corresponding optimal design parameters of the conductive path were then obtained. Ultimately, comparisons with existing conductive path verified that the optimized structure has better performance.