Design and Fabrication of a Micro-cpl for Chip-level Cooling

This paper summarizes the development of a micro-CPL designed for chip-level cooling. This research indicates that there are numerous fundamental issues regarding the thermofluid dynamics at small scales that are poorly understood. These issues, which can often be ignored at macroscales, will be critical as devices such as micro-CPL’s and heat pipes are optimized in order to handle the increasingly large thermal loads of modern technology. INTRODUCTION As the power supplied to electronic packages and Microelectromechanical Systems (MEMS) type devices increases, the question of thermal management becomes critical. The current mandate for power is reflected in everything from personal computer micro-processors to electronic packages required for unmanned aircraft. This, coupled with the demand for smaller systems, creates heat fluxes which become the most limiting factor in the production of micro-devices. Conventional methods of heat removal are simply not capable of dealing with the thermal gradients which lead to material failure. It therefore becomes necessary to look beyond traditional thermal management schemes, and develop techniques to deal with the cooling of micro-devices. A micro-cooler based on MEMS technologies could be placed in direct contact with, or integrated directly into the micro-processor, sensor, or other electronic chip for which cooling is required and could maintain an optimal temperature. The advantages of this approach are three-fold. First, it allows for precise temperature control at the chip-level. Second, the overall cooling is more efficient because specific heat sources within the electronics package may be targeted. Third, the overall size of the electronic system can be kept small. This paper presents an overview of several years of work that have been described previously as a framework to discuss interesting, but poorly understood, phenomena associated with micro-scale heat transfer. CAPILLARY PUMPED LOOPS (CPL) Capillary Pumped Loops (CPL) are similar to heat pipes in that the large effective thermal conductivity of a CPL is derived from the vaporization and condensation of its working fluid. The main difference between the two approaches is that in a heat pipe the vapor and liquid flows are in contact and run counter current to each other whereas in a CPL the vapor and liquid flows are separated into individual channels. This provides three advantages: first, this increases thermal isolation between the vapor and liquid (Dickey and Peterson, 1994), secondly, this provides for significantly improved geometric freedom by separating the evaporator and condenser, and finally, because the wicking structure has been removed from a majority of the device, the pressure drops are reduced along the vapor and liquid lines, allowing for larger mass flow rates under the capillary pumping limit. Research on CPL’s started in 1961 when Laub and McGinness began to investigate a two-phase thermal control system called a Capillary Pumped Loop (CPL). While they only examined a capillary pumped vapor generator, their work was later expanded on by Stenger in 1966, who reported on two CPLs capable of transporting more than 800 W over 50 ft. Since then, the CPL has been rigorously examined, and observed to be governed by the same limits as heat pipes. One– dimensional paradigms to design and predict their performance have been established (Dickey and Peterson, 1994). There are many challenges in transitioning MEMS technologies to this specific application. While some of the challenges relate to micro-fabrication design including material and geometric characteristics and overall flow dynamics. Many of the technical challenges bring up extremely interesting problems relating to the physics of micro-heat transfer and fluids mechanics. A few of these issues that have been identified