Experimental and numerical study of structural parameters on the heat transfer performance of submerged cooling system

Abstract

Immersion direct contact cooling effectively transfers heat from electronic components to a cooling solution without additional thermal resistance, due to its high energy density, low power consumption, system simplicity, and integrated thermal management for electronic devices. Current scholarly research on immersed structures predominantly focuses on surface temperature effects, with a scant investigation into internal flow and temperature field distributions concerning boundary layers. This study addresses structural optimization in immersion cooling systems, employing experimental and simulation methods to elucidate the thermal mechanisms by which structural optimization impacts the performance of chip immersion cooling systems. The research examines the impact of panel spacing and liquid level height on the thickness of velocity and thermal boundary layers, as well as the combined effects of natural and forced convection under three distinct cooling modes. Analysis reveals that heat pipe immersion cooling can reduce the heat source junction temperature to 40.7 °C under a 60 W power dissipation, a 47.2 % decrease from the bare heat source temperature. Reducing panel spacing disrupts the thermal boundary, enhancing forced convection and lowering the junction temperature by up to 10.1 % (6.5 °C). As liquid level height increases, the junction temperature initially decreases and then rises, with an optimal height of 175 mm for the best overall heat transfer effect. At low flow velocities (0.04 m/s), the vertical buoyancy-induced flow velocity at the heat source surface (24.2 × 10-6 m/s) exceeds the horizontal forced convection velocity (0.78 × 10-6 m/s), indicating that buoyancy changes due to density differences are significant and must be considered.