Hybrid-Cooled Data Center Server Layout Optimization for Air-Side Heat Recovery

Abstract

The rapid increase in energy demand for data center requires improved cooling techniques. This study investigates, numerically and experimentally, the energy efficiency optimization based on server level air flows and also identifies the potential for waste heat recovery from the air stream for a hybrid air/liquid cooled server. Multi-objective genetic algorithm and entropy generation minimization are chosen as tools to address the multiple objectives involved in the problem and examine the cooling performance and waste heat recovery potential. Selected experimental measurements have successfully validated the CFD simulation model.