Abstract:
One of the major reasons for high energy consumption in data centers is due to less focus (e.g.,
intermixing of hot and cold air, creation of stagnation points in cold-aisle etc) on the proper flow
of air inside the data center. To enhance energy efficiency and improve air flow movement in
data center requires well verified CFD model. The research in this field conducted so far majorly
emphasis on thermal behavior of a data center. Thermal study verifies the temperature
distribution in a data center, but it does not authentically assures the airflow patterns inside the
data center which play important role in energy efficiency of a data center. This thesis
summarizes effects of aisle configurations on energy efficiency of a data center. Experimental
data of point velocity and temperature distribution in a real data center is obtained. CFD analysis
of the data center is then performed and results are compared with the experimental data. In
order to analyze the effect of hot and cold aisle configuration effect on energy efficiency of a
data center, a numerical study for two even numbered rows of racks data center is performed.
Based on results obtained, we conclude that for any even numbered rows of racks data center,
configuration with one more cold aisle than number of hot aisles gives energy efficiency in the
data center.
