Relative Performance of Two-Phase vs Solid Conductive Heat Spreaders for High Heat Flux Applications

摘要

State of the art integrated circuit devices operate at ultrahigh heat fluxes with local hotspots generating well above $1,000 W/cm^2$. Heat spreaders attach to devices and transform their heat fluxes to lower levels for rejection to heat sinks with given surface areas, heat transfer coefficients and boundary temperatures. Thermal engineers design heat spreaders to minimize device hotspots by optimizing material selection, geometry, and heat spreader type. Heat spreader types include solid conductive units as well as twophase solutions; i.e., wick-based heat pipes (or vapor chambers) and pressure-driven oscillating heat pipes (OHPs). Two-phase spreaders more evenly diffuse heat across the heat sink but also add superheat at the interface of their solid wall and working fluid (i.e., convection boundary). For two-phase spreaders, thinner walls lower the temperature rises through the wall but increase heat fluxes, superheats and thus temperature rises at the convection boundary. The goal of the following study is to establish the optimum wall thickness of a heat spreader that decreases the temperature rise from the device through the spreader to the convection boundary-or in the case of a solid conductor to the heat sink boundary. In this paper, exact conduction solutions for this optimization problem are presented for both rectangular and radial geometries. For corroboration, the results are compared to finite element solutions for several sample problems with excellent agreement. The utility of the solutions is that they can readily be used in a spreadsheet format for rapid thermal trades to identify the optimum heated wall thickness and provide the minimum device temperature.