Open AccessComparative analysis of thermal environment between raised-floor and row-based cooling in a campus data center
Author(s) -
Chaoqiang Jin,
Xian–Xu Bai,
Ya’nan An,
Xin Zhang
Publication year - 2019
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/609/3/032058
Subject(s) - aisle , blanking , data center , airflow , inlet , rack , thermal comfort , environmental science , air cooling , containment (computer programming) , thermal , engineering , mechanical engineering , simulation , marine engineering , automotive engineering , computer science , meteorology , structural engineering , physics , programming language , operating system
In recent years, the modular data center, which adopts row-based cooling strategy, is becoming popular and widely applied. Also, cold aisle containment is integrated to reduce the mixing of hot and cold air. Even so, similar to the raised-floor cooling, the cold air bypass and hot air recirculation still affect the thermal environment. This paper presents the comparison of thermal environment between raised-floor and row-based cooling, through changing the factors of servers’ arrangement, aisle containment and blanking panel. Experiment and simulation were employed to investigate the inlet air temperature of racks and thermal environment of the data room, and a total of 15 models were simulated. The results showed that combining the cold aisle containment with blanking panels can bring uniform airflow, and the servers installed in the upper part of the rack was the best choice for the both cooling strategies. Especially, the blanking panels could further improve the airflow for all servers’ arrangement using row-based cooling strategy, which can reduce the mean inlet air temperature by 16.7 °C, and the minimum SHI value can reach to 0.089. These results could guide the thermal environment management to guarantee the effective utilization of cooling capacity in data centers.