Synthesis of multiple cross-section pin fin heat sinks using multiobjective evolutionary algorithms

Abstract

This paper demonstrates multiobjective optimization of a multi cross-section pin fin heat sink (MCSPFHS) for use in electronic devices. The design problem is assigned to optimize junction temperature and fan pumping power of the heat sink. Design variables are encoded to shape the fin geometry with several pin fin cross-sections. The heat sink is set to be side-inlet-side-outlet (SISO) while several multiobjective evolutionary algorithms (MOEAs) including hybrid real code population-based incremental learning and differential evolution (PBIL-DE), second version of strength Pareto evolutionary algorithm (SPEA2), and unrestricted population size evolutionary multiobjective optimization algorithm (UPSEMOA) are applied to solve the bi-objective optimization problem. The results obtained are superior to those conventional pin fin heat sinks.