Passive Oscillatory Heat Transport Systems

An underdeveloped class of oscillatory passive heat transport cycles are discussed that have the potential to transport significantly higher heat loads than current heat pipes. Prototype cycles employing inferior working fluids have demonstrated transport of higher heat loads over significantly greater distances than similarly sized heat pipes (including CPLs and LHPs) employing ammonia. Most of the proposed cycles do not require capillary forces to circulate the working fluid. They are also relatively insensitive to gravity and might best be compared to thermal systems using mechanical pumps. The history of development to date of such cycles is presented in relation to other approaches under consideration for various semi/passive thermal control applications. Specific operational characteristics of a select loop are presented. The obvious pros and cons of these systems are discussed as well as potential applications—particularly as regards electronics cooling. REVIEW The extrapolated challenges for advanced electronics thermal control have been well articulated (Mudawar 2000). As the spectrum of cooling requirements continues to widen, designers are forced to remain receptive to new approaches with unique or niche performance advantages over existing and/or competing systems. From smallto large-scale systems, traditional heat pipes (including CPLs and LHPs), thermosyphons, and mechanically pumped loops represent the majority share of the phase-change thermal control/transport solutions. Traditionally, for relatively small heat loads < O(10kW) and transport distances, heat pipes have proven excellent options for simple, reliable, and quiet thermal energy transport and temperature control at low to moderate heat fluxes. The common limitation for these devices is a capillary pressure limit, which continues to be extended through the development of higher performance wick materials and geometries. Of course, the capillary driving force is precisely the advantage of the traditional heat pipe in that no pump power and no moving parts are required for the essentially silent operation routinely achieved. High heat loads may be transported using heat pipe networks and large high power designs are being considered (Ottenstein, 2000). Thermosyphons, like heat pipes, are excellent passive cycles and can be applied at small and large scales, but at typically low heat fluxes. Thermosyphons exploit buoyancy-driven convection made possible by an acceleration field (i.e. gravitational, centrifugal, etc.). Thus it is the acceleration field strength and orientation that limits the performance of such systems, for example, preventing them from being able to transport heat ‘against gravity’ on Earth, or completely preventing their operation onboard low-gravity spacecraft. For high heat loads ~O(MW) mechanically pumped loops are arguably unchallenged. Despite a pump power penalty and the ‘blemish’ of moving parts, mechanically pumped loops have demonstrated their value in thermal control systems for over 100 years. It is traditionally undisputed that mechanically pumped loops are the method of choice for large heat loads transported over long distances. Even at ‘low’ total heat loads, mechanically pumped cycles may be designed to transfer heat at incredibly high heat fluxes (Mudawar, 2000) by exploiting forced convection. Microscale pumps have been designed and are being developed rapidly. Regular improvements in pump design continue to decrease pump wear and increase pump life, and unique pump designs, particularly on the microscale, indeed possess a ‘minimum’ of moving parts (Forster, 1999). A POTENTIAL NICHE FOR OSCILLATORY SYSTEMS There are numerous metrics employed to rate the performance of thermal solutions, to which one must add economic considerations. Such metrics are usually only of value when restricted to a specific cooling application with specified requirements, i.e. terrestrial silicon-based electronics cooling, reactor thermal control systems (TCS), solar heating, etc. It is becoming increasingly difficult to make a quick and best selection of a cooling or heating cycle due to the increasing variety of approaches available and the fact that, in general, traditionally large scale mechanically pumped cycles are being miniaturized while the traditionally small scale passive cycles are being scaled up. Due also to the increasing criticality of the thermal design, it is natural now to categorize systems by multiple criteria such as total heat load and expected heat flux at the source/sink. In doing so one might imagine a regime map of heat pipe and mechanically pumped loop thermal solutions as shown in FIGURE 1. In this conceptual map the upper limits of heat flux are set by physical performance limitations such as pool boiling heat flux for heat pipes and forced convection phase change heat flux for mechanically pumped loops, which can differ by more than two orders of magnitude. The lower limits for the different regimes might be set by economics. For example, a mechanically pumped cycle may not be cost competitive to a heat pipe at intermediate-to-low powers and fluxes. Similarly, a heat pipe will not compete against a passive solid conductor at low powers and fluxes. It may also be impractical for a heat pipe system to transport high heat loads (right vertical limit, FIGURE 1). For moderate-to-low powers at moderate-to-high heat fluxes it is possible to envision a regime in which heat pipes are incapable and mechanically pumped cycles are uneconomical—region A, FIGURE 1. (Similar arguments may be constructed comparing transport length versus total heat rate, or pitting system mass, volume, reliability, etc. against both heat flux and/or total power, etc.) It appears to be this regime of low-to-moderate heat rate, moderate-to-high heat flux, and moderate-to-high transport distance that several novel oscillatory thermal device concepts might grow to occupy. In this paper a selection of oscillatory thermal systems will be briefly reviewed. At smaller scales such devices are sometimes called ‘nontraditional heat pipes’ and have been referred to as oscillating or pulsating heat pipes in the limited literature discussing them. However, because such loops can be enormous, the term heat pipe seems inappropriate and the use of ‘loop,’ ‘cycle,’ and even ‘system’ is more fitting and more frequently used. An early distinction is made between oscillatory loops that are in part capillary controlled and those that are not. The discussion then focuses on the latter where characteristic performance data for a prototype loop is provided to support the obvious attractive features of an oscillatory approach. The potential for a niche for oscillatory cycles is then discussed followed by a list of obvious limitations and concerns. REVIEW OF SELECT OSCILLATORY CYCLES Oscillartory means for pumping fluids date back at least as far as 1698, with Savery’s fire engine devised to pump water against gravity (Balmer, 1990). Since then, a fair number of techniques to exploit a temperature difference to circulate a working fluid have been conceived and demonstrated. The class of devices discussed here are those that develop a pressure difference in the system resulting from the condensation and/or evaporation of a working fluid in " max q " max q Heat pipes Mech. loops