Automated Thermal Optimization of Pinfin Heatsinks for Three-Phase Multichip SiC Power Modules

With the growing demands for compact and high-efficiency electric vehicle (EV) drive systems, effective thermal management of Silicon Carbide (SiC) power modules has become a critical design challenge. This paper proposes an optimization methodology for Pinfin heatsinks, aiming to effectively reduce the maximum junction temperature and improve thermal uniformity among parallel-connected chips in three-phase multichip SiC power modules. Through analysis of the thermal model, a parametric automated optimization framework for circularly regular Pinfin heatsinks was established, utilizing the Lattice Boltzmann Method (LBM) for efficient thermal performance evaluation. Based on the proposed automated optimization approach, optimizations for reducing thermal redundancy chips and minimizing temperature difference were respectively carried out for designed three-phase SiC power module. Furthermore, a novel irregular layout of elliptical Pinfin with randomized orientations and aspect ratios is proposed and optimized. Experimental validation and simulation results demonstrated the effectiveness of the proposed method.