Impact of different coolants on performance of two-phase immersion cooling data centre

Abstract Two-phase immersion liquid cooling (TPIC) technology provides an effective solution for managing the high heat flux density in data centre. The coolant selection is crucial to the operation of TPIC data centre. To comprehensively evaluate the performance of different coolants, this study established a validated computational fluid dynamics (CFD) model. Results show that, SF-33 and Novec 7000, which possess low boiling points and high latent heats of vaporization, demonstrate superior cooling performance. Moreover, the temperature uniformity coefficient shows the opposite tendency. Under the same heat flux and maintaining the same average temperature, the required flow rate of SF-33 is 0.387m 3 /h lower than the highest Novec 649, where the PUE difference is 0.021.