Prediction of Distributed Air Leakage in Raised-Floor Data Centers

In raised-floor data centers, the distributed air leakage, which refers to airflow through the gaps between the floor panels, reduces the amount of cooling air available for direct cooling of computer equipment. In this paper, we discuss a methodology for predicting this air leakage, using an airflow
model based on computational fluid dynamics, and validate this methodology by comparing the predicted results for air leakage with measured values. Subsequently, we use the airflow model to study the effect of air leakage on airflow rates through the perforated tiles and the pressure under the raised floor. Finally, we present a simplified model for estimating air leakage and assess its validity.