Skip to main content

Thermal Management of High Power Microsystems Using Multiphase Flows

Final Report Summary - THERMAPOWER (Thermal Management of High Power Microsystems Using Multiphase Flows)

Increased functionality and power consumption of micro-devices and high power electronics has come at a cost: power dissipation and heating. This heat must be dissipated to ensure reliable operation of such devices in both earthly and reduced gravity environments (e.g. space industry), without adversely affecting their performance. With a highly competitive world market, worth tens of billions of Euros, it is imperative for EU to gain a competitive position in this field (currently led by USA and China). Advances in manufacturing technologies and subsequent increased use of small-scale electronic devices operating at high power densities have brought about a dramatic demand for thermal management systems to provide extremely intensive localised cooling.
The project allowed significant advancements in the following areas, pool boiling heat transfer, flow boiling and multiphase flows, condensation, phase change and wetting phenomena. These were achieved using diverse tools, experimental, numerical and theoretical. The flow maps for flow boiling in micro-spaces generated via experiments in EPFL and Edinburgh will be with no doubt of great significance in designing future cooling solutions for high power systems. Measurements of local heat transfer in pool boiling, performed by Maryland University (USA) was very useful in developing theoretical models implemented by teams in Nottingham and Shanghai, this has clarified and advanced our knowledge of controlling and limiting physical mechanisms. Theoretical and numerical models developed by Shanghai Jiao Tong (China) for critical heat flux (CHF) in pool boiling and condensation are unique and will be translated in optimising thermal management options. The work of Nottingham (UK) on numerical models for two phase flows in micro passages will be further refined to be more predictive in the future.
The partners brought to the project experience and top international expertise in analytical modelling, global and localised experimental measurements in microchannel boiling, phase change and condensation, expertise in the fabrication of micro instrumentation with outstanding facilities for microfabrication in silicon, and expertise in the numerical simulation of multiphase flow with deformable interfaces. A particular strength was the combination of boiling and condensation in one project, which led to cross-fertilisation of ideas relating to processes with shared characteristics.
The exchange between the groups has been very productive (see outputs). A very extensive list of scientific publications in international journals and text books were produced as a result of this project. These latter will be leading references on this topic for the international community for many years to come.
Owing to its success with the international community, our consortium was approaches by research groups from Japan, Canada and Brazil to join our activity. This is a testimony to the high standing and quality of research developed during this project.
Overall the exchange of researchers between UK, Switzerland, USA, China, Japan, Brazil and Canada (these three last partners joined later on in the project) allowed over the four years period to have a very fruitful and beneficial exchange of results and expertise. The flow boiling experiments in microscale which were undertaken in EPFL, SJTU and Edinburgh allowed an exchange of ideas on using optical techniques for measuring flow pattern in real time. The exchange between Edinburgh and Maryland in USA enabled access to the state of the art infra-red thermography which has been used by researchers from Edinburgh visiting Maryland to explore phase change phenomena. One of these researchers is appointed at MIT and continuing the exploration of observations obtained in the context of this project. The numerical work performed in Edinburgh, Nottingham and SJTU and the collaboration between researchers allowed to investigate the use of various numerical tools such as CFD and LBM to model two phase flows both in pool boiling and flow boiling situations. The outcome was a successful model which can predict the complete boiling curve without fitting parameters. This is a significant achievement in studying boiling heat transfer phenomenon (see publications).
Because of space limitation, we presented few of the interactions and results obtained during this project. The breadth and depth of the full results can be deduced from the substantial list of publications generated throughout this project.