Transient analysis for contained-cold-aisle data center

The objective of this paper is to conduct a transient analysis for a contained-cold-aisle data center using Computational Fluid Dynamics (CFD) modeling. Containing the cold aisle reduces the inherent challenges of predictability and energy consumption of data center cooling systems by separating the hot and cold air streams. Transient analysis is crucial to further show the benefits of this methodology and investigate the potential drawbacks. In this work, we investigate the effect of variable power and flow rate in time on a raised-floor data center with specific geometry. First, a base case numerical model is established. Then, we conduct a transient analysis for the uncontained base case and the three containment configurations. The containment configurations under consideration are ceiling-only, doors-only, and fully-contained cold aisle. We hold a comparison between different geometrical configurations of the containment system under certain transient operating conditions. Computer Room Air Conditioning (CRAC) failure and change in IT-demand are chosen to represent conventional transient scenarios in a data center. The transient analysis shows overshoots of cabinet inlet air temperatures beyond the final steady state, which cannot predicted through a simple steady state analysis. In addition, the uniform temperature distribution inside the cold aisle is affected by the change in air supply and level of containment. The temperature distribution in the cold aisle changes with time. Guidelines can be recommended based on the conclusion of this study.