Utilization of a 3D CFD model to simulate airflow behaviour towards optimization of cooling within data center

Abstract

In recent years, efficient cooling management systems in data centers have been a high priority to ensure that the optimum operating condition is maintained under the current high level of digitalization. However, numerous data centers still encounter cooling inefficiencies due to factors such as CRAC unit arrangement, poorly positioned perforated tiles, and inadequate return air patterns where supply air mixes with return air. Addressing these challenges is essential for enhancing airflow and improving cooling efficiency. Therefore, in this research, the quality of temperature distribution and the airflow pattern in the DFCU data center were analyzed using the computational fluid dynamics (CFD) approach. CFD simulations were used to quantify the facility’s thermal cooling by developing a graphical user interface for component description and viewing. There was a need to compare the flow field and temperature distribution in the data center under different layouts since using CFD models describing solid bodies with thermal mass and energy balance principle makes it possible to visualize temperature and airflow patterns. From the results, the DFCU Bank data center showed temperature variations; hence there was a timely need for efficient cooling management to ensure the temperatures were within the limits recommended by ASHRAE. The simulation results had projections of the high-temperature zones and a comprehensive 3D examination of airflow distribution, and optimization methods established the best airflow arrangement. In addition to that, detailed temperature examinations of the individual racks and three-dimensional thermal mapping would establish the problematic regions and also inform how well the computer room air conditioning units were working. The various conclusions drawn from the research reveal that the cooling capacity would be improved, by adding two additional CRAC units each of 18kw, incorporating a return air duct system for better air distribution, and reducing exhaust air and supply air mixing through arterial blanking. These would result in a streamlined airflow in the data center, and a high cooling efficiency will be noted, which would eventually lead to low levels of energy consumed.