A Study on the Optimization of EV Inverter Power Module Considering Tolerances

This paper was produced for the 2019 NAFEMS World Congress in Quebec Canada

Resource Abstract

To face with the environmental problems, vehicle electrification is rapidly on the progress. The traction motor of electric vehicle, which replaces engine in vehicle, requires high power and high efficiency during driving. The optimization of traction motor performance comes from the inverter efficiency. The role of the inverter is to control the traction motor by converting the DC(Direct Current) electricity from high voltage battery pack to AC(Alternating Current) that can drive the motor. The power module in the inverter plays main role of converting the electricity. The power module in the inverter consists of power conversion chips and the cooling structure. During converting electricity, heat is generated from power conversion chips in the power module. So the cooling strategy and optimization of cooling structure is the key of high performance / high efficiency traction motor system.



It was found that the micro dimensions of the power conversion chips in the power module significantly affect the cooling efficiency. So the optimization of the cooling structure, which can cool down the conversion chips in any tolerance condition of chips and structures, is required in the early stage of developments. The 3D tolerance analysis has been performed to get the tolerance combination of 6 chips. The maximum tolerance values from chips and structures are considered in structural analysis of cooling structures. The cooling performance is coming from minimized gap between chips and cooling structure. So the purpose of structure analysis is to minimize the gap between chips and the cooling structure. At that time, the pressure on the module should be limited to avoid any damage on chips. The worst condition of tolerance combination is considered in the structural analysis. The optimal design of cooling structures among all combinations of tolerances has been derived. To check the side effect of the optimized structure, the modal analysis and random vibration fatigue analysis are performed. The optimized cooling structure has been verified through cooling analysis and by checking the gap between module and structure from the test samples.