Parametric study by means of 3D CFD to improve the airflow pattern and temperature distribution in data center rooms

Macedo, Diogo Gonçalves Costa e Silva de; Godina, Radu; Gaspar, Pedro Dinis; Silva, Pedro Dinho da; Covas, Miguel Trigueiros

http://hdl.handle.net/10400.6/7529

Use this identifier to reference this record.

Name:	Description:	Size:	Format:
Paper_ICR2019_1770.pdf		803.74 KB	Adobe PDF	Download

Send Feedback

Authors

Macedo, Diogo Gonçalves Costa e Silva de

Godina, Radu

Gaspar, Pedro Dinis

Silva, Pedro Dinho da

Covas, Miguel Trigueiros

Abstract(s)

The processing efficacy and energy efficiency of data centers (DC) is directly related with the cooling performance of the air-conditioning systems. The cooling effectiveness depends on airflow pattern and temperature distribution. This paper presents the parametric study of three-dimensional (3D) computational fluid dynamics (CFD) models to improve these technical characteristics. In this case study the most suitable thermal load, as deemed by the organization, of the heat flow boundary condition is simulated and studied for maximum and minimum airflow. The analysis of the numerical results provides a prediction of the thermal performance, i.e., the lower values and spatial gradients (hot spots) of air temperature and velocity while maintaining the computational processing requirements. These results allow the development of an optimized CFD model for setting up the operational parameters of CRAC systems for DC rooms.