Thermal -fluid behavior in parallel boards with discrete heat -generating blocks and its thermal optimization using the entropy generation minimization method.

Abstract

Developing and periodically fully developed laminar flows in parallel boards with discrete heat-generating blocks were numerically investigated. The system simulates cooling passages in a stack of electronic circuit boards with heat-generating chips. Some unique problems in electronic cooling were examined, such as non-uniform thermal conductivity of the board and thermal contact resistance between the chip case and the board. Numerical results were obtained for velocity vectors, isotherms, the friction factor, local and average heat transfer coefficients, and thermal resistance. It was found that at low Reynolds number flows, a developing flow might achieve a fully developed flow state at certain block number from the entrance, and that a cavity-bypass flow interaction significantly affected the thermal performance of the channel flow under some flow conditions. The study discovered that the fluid flow and heat-transfer performance in this channel flow was similar to that in ribbed channel flows. The conventional method of thermal optimization suffers from complications involved in simultaneously optimizing a number of parameters which govern the process. This is particularly true when thermal-fluid phenomena are involved, for example, in the optimization of electronics-cooling devices. The entropy-generation rates in parallel boards with discrete heat-generating blocks were calculated. It was disclosed that local entropy generation distribution could identify the region of significant increase in local irreversibility, thus providing information important for designing optimal geometry by reducing the entropy generation in the system. It was attempted for the first time to optimize the board spacing of a stack of printed circuit boards using the entropy generation minimization (EGM) method. The results showed that entropy generation reached a minimum with a certain range of board spacing. The optimal board spacing was correlated approximately by the compact expression. It was discovered that the optimal board spacing increased with an increase of the Reynolds number. Results obtained from the EGM method were confirmed through a comparison with results obtained from a conventional thermal optimization method. Finally, this study confirmed that the entropy-generation rate from a system could be used as the sole factor in determining the hydrodynamic and thermal performance of the system.