Research and optimization of thermophysical properties of sic oil-based nanofluids for data center immersion cooling

Thermal management of high-power density equipment in data center is a key obstacle to power densification, and submerged cooling considered as an effective mean. Considering operability and economic factors, mineral oil is selected as the research object. In this study, the insulating silicon carbide oil based nanofluids are used as a new type of immersion cooling working fluid to replace the traditional mineral oil. The thermophysical properties of silicon carbide oil based nanofluids with wide concentration range from 0.1-10.3 vol% are characterized, and the optimal concentration of nanofluids are selected by Ansys Computational Fluid Dynamics finite volume method. Moreover, the heat transfer performance at Reynolds number of 250-3000 is analyzed. As a result, compared with the basic mineral oil, the heat transfer is significantly enhanced by silicon carbide based nanofluids. In particular, the enhanced effect cannot improve monotonously as the concentration increase. When the nanofluids are in the range of 0.3vol % and 3.7vol %, the maximum enhancements are 11.4% and 11.7% towards the basic fluid respectively. Last but not least, the 0.3 vol% nanofluids have better heat dissipation effect at low Reynolds number while 3.7 vol% nanofluids perform better at high Reynolds number. These results indicate that silicon carbide oil based nanofluids have a promising application in data center immersion cooling.