Reliability-Oriented Automated Design of Double-Sided Cooling Power Module: A Thermo-Mechanical-Coordinated and Multi-Objective-Oriented Optimization Methodology

Compared with the traditional single-sided cooling (SSC) power module, owing to the decreased thermal resistance and packaging parasitics, the double-sided cooling (DSC) power module is a promising solution of the motor drive for electric vehicle (EV) implementation. However, the lack of the model to characterize the thermo-mechanical interaction mechanism in the DSC power module challenges the co-design methodology of the power module from the respective of multi-physics. In this paper, aiming at reliability improvement, to balance the tradeoff between thermal resistance and mechanical stress, a multi-objective-coordinated automated design methodology is proposed for the DSC power module. To characterize the influences of material properties and structure sizes, the mathematical models for the thermal and mechanical features of the DSC power module are proposed. These models are also examined by using the finite element analysis (FEA) tool. Besides, to promote the specifications of DSC power module, a multi-objective optimization model is proposed to coordinate thermo-mechanical metrics in multi-physics. The non-dominated sorting genetic algorithm II (NSGA-II) is implemented to automatically achieve the Pareto solutions of the proposed multi-objective optimization model with respect to thermal resistance and mechanical stress. Considering the influences of material properties, optimal structure sizes of the DSC power module are comprehensively presented. The proposed multi-objective-oriented optimization methodology provides a new routine to reshape the thermo-mechanical performance of the next-generation DSC power module toward more reliable EV implementation.