Air-cooled heat sinks remain a simple and reliable technology for heat dissipation in high-power-density servers, computing, and electrified transportation applications. The power dissipation capabilities of air cooling can be improved by utilizing advanced heat-spreading techniques to increase the finned surface area, often by incorporating vapor chambers in the heat sink base. Vapor chambers offer phase-change-based heat spreading at lower thermal resistances compared to solid metallic heat spreaders, thereby allowing heat transfer from small, concentrated heat sources (> 100 W over 100 mm(2)) to a large rejection area. Accurate thermal performance and dryout limit prediction in such vapor-chamber-embedded heat sinks is critical for their design. However, models that describe vapor chamber transport typically do not account for nucleate boiling in the evaporator wick, which is increasingly likely to occur as applications trend toward higher heat fluxes. Due to the substantial increase in two-phase flow pressure drop in the presence of nucleate boiling, model-based predictions of dryout limits considering only single-phase flow of liquid in the evaporator wick can be significantly off the mark. This paper introduces a physics-based modeling approach for predicting the thermal resistance and dryout heat flux for vapor chambers that accounts for the occurrence of boiling in the evaporator wick. The heat transfer model considers the thermal coupling between a solid heat spreader representing the vapor chamber wall and the transport in the core of the vapor chamber to calculate the temperature fields. The onset of boiling is characterized using a wall superheat criterion that determines the area of the wick undergoing boiling. Using the mass fluxes from the heat transfer model, the pressure fields in the wick are solved. In the boiling region of the wick, fluid transport is modeled using the Darcy-Ergun equation corrected for the relative permeabilities of the liquid and vapor. A case study illustrating the usage of the model is demonstrated for an example vapor chamber of dimensions 50 mm x 50 mm x 5 mm, with a sintered copper wick structure. A purely evaporation-based pressure drop prediction for such a vapor chamber with the assumption of single-phase flow in the wick leads to large overprediction of the dryout heat flux. The modeling framework developed herein highlights the need to account for boiling in the evaporator wick for vapor chambers used in such high-heat-flux applications.
原文链接
分享
引用
