Abstract

Two-phase refrigerant pumped loop cooling systems have seen wider use in recent years because of higher power density electronics and the need for increased heat transfer capability. However, two-phase systems are difficult to model because they operate in a closed-loop and both liquid and vapor phases exist and change with time throughout the system.

A two-phase “test-bed” system was constructed in order to explore the various operating regimes and limitations of this technology. Additionally, physical results from this test-bed were used to determine if the system performance could be accurately modeled using a commercially available software package such as SINDA/FLUINT. The refrigerant circuit uses R134a as the cooling fluid that is pumped around a complete loop using a positive displacement pump. Heat is added at the evaporator using ceramic heaters on the outside of the housing that boils the fluid, resulting in a liquid/vapor mix in the transport tube. Heat is removed at the condenser which is a commercial flat plate heat exchanger that uses building water for cooling. A closed-loop fluid-thermal model of the test-bed was created in Thermal Desktop®, which is the pre and post-processor for the SINDA/FLUINT analyzer, and compared to several tests conducted on the physical system at varying heat loads. The model successfully predicted the system performance after adjustments to the model for refrigerant mass charge and condenser performance. The maximum difference between the measured evaporator temperature and the modeled temperature was only 6.1°F; which occurred at the highest heat input level (1400 Watts). Explanations for differences between model and test results will be discussed. In particular, modeled system temperatures were found to have a very non-linear sensitivity to the total refrigerant mass charge.

This content is only available via PDF.
You do not currently have access to this content.