01. Experimental/Computational Fluid Dynamics

Two-Phase Immersion Cooling is highlighted as an effective solution for thermal management in data centers and high-performance computing (HPC) packages. By boiling the coolant directly from electronic components, it significantly enhances heat convection compared to single-phase cooling. This method allows for rapid heat dissipation and eliminates the need for air circulation, making it highly efficient for cooling densely packed electronic components.. This study investigates the thermal and flow characteristics of two-phase immersion cooling for electronic systems using Computational Fluid Dynamics (CFD) simulations with STAR-CCM+. A system-level model of a PCB with multiple heat-generating chips was developed to analyze the cooling performance under varying power densities. The dielectric liquid FC72 was used as the coolant, and the boiling effects were confined to the chip surfaces, creating an efficient heat dissipation cycle. Key findings include the observation that the highest chip temperature reached 69.95°C at a heating power of 50W, with temperatures increasing as power levels rose. The Critical Heat Flux (CHF) was identified as a crucial parameter for system cooling capacity, with the upper chip experiencing CHF at 250W and the lower chip at 300W due to the vertical placement of the chips and vapor coverage effects. The optimal cooling efficiency was achieved at a power density of 12.25 W/cm² when the chip heating power was 250W. Overall, the results confirm the effectiveness of the CFD model for analyzing thermal management and chip-to-system interactions, providing valuable insights for designing efficient cooling systems in high-performance electronic applications.