scholarly journals THE RESEARCH ON DYNAMIC CYCLE MONITORING OF TRANSFORMER SUBSTATION MODULAR DATA CENTER CONFIGURATION BASED ON INTELLIGENT ASSISTANT

2020 ◽  
Vol 34 (14n16) ◽  
pp. 2040123
Author(s):  
Yong-Liang Chen ◽  
Zi-Qiang Qin ◽  
Yao Li ◽  
Hai-Bo Wang ◽  
Sheryar Muhammad ◽  
...  

In high-density data center, energy consumption is increasing dramatically. For reducing the energy consumption, CFD software, Fluent 15.0, is used to simulate the flow and temperature field distribution with [Formula: see text] turbulence model and fluid–solid coupling method. Fans on the back of racks are simplified as walls with a certain pressure jump. Severs are treated as solid heat sources and porous media. Simulation results reveal that the temperature distribution on the back of racks is not uniform when air conditioners are arranged face-to-face, and local high temperature points emerge near the side wall of air conditioners. Factors affecting cooling efficiency, such as location of air conditioners, speed of inlets, distance of racks, etc., need to be improved. Geometric model is optimized by using a diagonal rack arrangement and drilling holes on the side wall. Based on this, four different cases with various hot aisle distance are proposed. Single and double modular data center are both simulated. Results of new model are better than those of baseline model.


2015 ◽  
Vol 138 ◽  
pp. 258-275 ◽  
Author(s):  
Sang-Woo Ham ◽  
Min-Hwi Kim ◽  
Byung-Nam Choi ◽  
Jae-Weon Jeong

Author(s):  
Amip J. Shah ◽  
Van P. Carey ◽  
Cullen E. Bash ◽  
Chandrakant D. Patel

Recent work has proposed an exergy-based strategy to achieve optimal system-wide performance via localized control of individual data center thermal management components. This paper presents the results of a case study where the proposed approach is applied to a data center with two rows of computing racks and two Computer Room Air-Conditioning (CRAC) units. The formulated model is used to predict the optimal data center configuration in terms of supply temperatures, flowrates, and rack heat load configurations. Two extreme cases are chosen: one with the maximum experimental heat load in the data center, and one with a minimal experimental heat load. For each case, the optimal settings for each CRAC unit were predicted using the model and using temperature + flow measurements in the data center. The setpoints predicted by the model for optimal CRAC flow and supply temperature were within 25% of the experimentally determined optima.


Sign in / Sign up

Export Citation Format

Share Document