Experimental comparison of two-phase and single-phase cooling methods of a power module

Abstract: Power modules of an electric drivetrain generate some of the highest heat fluxes in the system and are therefore challenging to cool. State-of-the-art cooling methods use single-phase water-glycol as coolant. Two-phase cooling with refrigerants can be a way to improve the heat transfer rate as high heat transfer coefficients are expected during boiling. A downside of two-phase cooling is the possibility of dryout when the heat flux goes above the critical heat flux, which is therefore important to be able to determine. Traditional refrigerants used in two-phase cooling (e.g. hydrofluorocarbons and perfluorocarbons) are being replaced by newer environmentally friendly refrigerants (e.g. hydrofluoroethers and fluoroketons). One such fluid is studied experimentally in this work, FK-649. Both heat transfer rates and critical heat flux are determined for varying refrigerant saturation temperature. The experimentally achievable heat flux of the two-phase cooled power module is compared to that of alternative methods (cold plate, direct contact and pin fin water-glycol cooling) on the same power module. It is also compared to experimental data of two-phase cooling with other refrigerants from literature. The results show that boiling with FK-649 is comparable in performance to cold plate cooling, but performs worse than direct contact baseplate and pin fin water-glycol cooling. Also, boiling heat transfer with R134a and HFE-7100 is superior to that of FK-649. Using correlations for boiling heat transfer and critical heat flux, other new refrigerants are assessed. This analysis shows that other refrigerants have more potential than FK-649 but these refrigerants require higher operating pressures