Refrigeration systems are capable of removing high heat fluxes and are able to maintain the temperature of the processing units at cryogenic levels thereby enhancing their performance. Therefore, such systems are commonly used for cooling of high-performance computers. The performance of a refrigeration system depends upon the two -phase flow characteristics of the individual components, the refrigerant charge, and the properties of the refrigerant. System-level computational analysis is very useful the design of such systems. In the present study, a computational methodology is presented for the analysis of the fluid mechanics and thermodynamics of a refrigeration system. It is based on a two level approach. First, the one-dimensional equations for two-phase momentum and energy transport are solved within each component to derive the overall characteristics. These characteristics are, in turn, used for the solution of the mass, momentum, and energy conservation equations for the entire system. Linearized form of the component characteristics is used to account for the interdependence of pressure and temperature within the two-phase region. The system-level equations are solved using a direct solution technique. The computational method is very efficient and enables a rigorous analysis of the interaction among the components in determining the performance of the refrigeration system. The computational technique has been applied for the prediction of the performance of a practical single loop refrigeration system to illustrate the engineering guidance it provides for the design and optimization of refrigeration systems used in electronics cooling.
展开▼
