No abstract

No abstract

No abstract

There is renewed interest in passive 2-phase immersion for cooling power electronics and high performance computers. This can be attributed to recent research showing its performance potential compared with more complex and costly techniques, to innovations that simplify its application and to a general trend toward higher power densities.Though the thermal performance capabilities of passive 2-phase immersion cooling are well documented, the technique is not widely practiced and system designers will find little published information concerning subtler and very critical aspects of system design. There is no manual, for example, concerning practical details like material compatibility, electrical signal integrity (SI), fluid decomposition, management of moisture and light gases, and so on. This paper presents a useful material compatibility test method and explains the mechanisms of distillation and extraction that are intrinsic to a refluxing 2-phase system and by which wetted materials interact with the fluid and each other. It discusses sources, implications and techniques for removal of organic contaminants, water, non-condensable air and fluid thermal decomposition products. Data are presented from sub-20GHz SI experiments conducted with backplane connectors and microstrip transmission lines submerged in two classes of environmentally sustainable working fluids. It is hoped that this overview will demystify these subjects for designers unfamiliar with passive 2-phase immersion cooling and encourage more widespread adoption of this elegant and proven technology.

In this work, a new micro porous metallic coating method is used to produce two compact boiler assemblies that are practical for use in thermosyphons in electronics systems. These assemblies were tested using a segregated hydrofluoroether (HFE) working fluid C 3 F 7 OCH 3 on 1.0, 4.0 and 13.7 cm 2 heat sources at atmospheric pressure. The best of these yielded sink-to-fluid resistances, R sf , at 230W of 0.092, 0.047 and 0.024ºC/W, respectively. This boiler assembly was incorporated into a thermosyphon with a remote 100x100x2.2mm finned-tube condenser that produced a condensation resistance, R fa , of ~0.1ºC/W with a 7.6W 92x38mm fan. That thermosyphon was tested on the aforementioned heat sources yielding much improved R sf values of 0.083, 0.026 and 0.014ºC/W at 150W. Data gathered with larger condensers or more powerful fans suggest that it is quite practical to produce a compact thermosyphon that can cool a 20mm device generating 250W with R sa <0.10C/W, a performance level competitive with forced water systems.