Automated Optimization of Irregular Elliptical PinFin Heatsinks for SiC Power Module

Abstract:
Enhancing the heat dissipation performance of SiC power modules plays a crucial role in the overall performance of motor drives. In this paper, an automatic optimization design method for PinFin heat sink with irregular layout is proposed for SiC power modules. Firstly, a coding method is introduced to automatically generate candidate layouts of elliptical PinFin with random arrangements of rotation direction, size, and spacing. Secondly, the lattice Boltzmann method (LBM) is employed to evaluate the elliptical PinFin performance. Finally, genetic algorithm (GA) is utilized to select optimal layouts from candidate encodings, and the optimized heat sink is fabricated and welded onto the three-phase SiC power module. Simulation and experimental results demonstrate that the junction temperature of the power module with optimized Irregular Elliptical PinFin is reduced by more than 4 °Celsius under a three-phase series-connected continuous current of 180A, validating the effectiveness of the proposed optimization method in this paper.