透過您的圖書館登入
IP:18.188.175.182
  • 學位論文

小型資料中心設計數值模擬與分析

Numerical Simulation and Analysis of Small-Scale Data Center

指導教授 : 崔燕勇 王啟川

摘要


資料中心致冷效力和能源效率的不同,歸根到底還是取決於氣流分配系統的差異。本文利用商用計算流體力學軟體,模擬天花板供風系統在小型資料機房的熱管理表現,主要針對不同的供風口直徑、供風角度和檔板設計進行研究,以優化機房系統,並提出環型的風管結構,而基於不同的溫度和流動分布模式,同時利用氣流管理指標如RCI、SHI等參數,定量評估熱量與溫度之變化關係,再分析機房內之冷卻效能及內部流場的影響。模擬結果顯示,供風口直徑的尺寸會影響流量分布的均勻度,因此當供風口孔徑為210 mm時,RCI提升到了70.79%。藉由改變供風角度,將能延遲供風口的偏移行為。而在環型風管結構時,噴流的動量將能有效的降低,不過需要妥善的氣流控制配置,否則流場會相當紊亂。當環型系統下方加裝檔板後,能有效地引導並控制氣流,讓RCI從46.12%提升到77.05%。再者,直線型系統半封閉後,在直徑335 mm時,讓RCI提升到94.93%,SHI降至0.12。而環型系統在半封閉的設計下,尤其在供風口直徑為210mm時,平均RCI甚至達到99.82%、SHI則是0.095。

並列摘要


Among the numerous methods to increase data center cooling performances and energy efficiencies, air distribution is the crucial determinant to many data centers. In this research, a commercial CFD software is conducted in order to evaluate the thermal behavior of a small-scale data center with overhead air distribution systems. A few practical designs have been studied for optimization, such as the geometry of ventilation tile, the angle of supply valve and the implementation of blanking panels. In addition, an alternative ceiling ducts configuration have been proposed. Then, the performance is evaluated using the air flow management indexes RCI and SHI. According to the simulation results, varying the ventilation geometry has an effect on the flow rate uniformity. Hence, the RCI reaches 70.79% in the case of 210 mm. Also, rotating the tilt angle of supply valve could delay the attachment effect of discharging airflow. Although the momentum of jet decreases in the alternative ducts system, it should be implemented with airflow management by deploying some physical panels. Moreover, the cooling performance of the alternative system improves from 46.12% to 77.05% after sealing the openings under racks. Furthermore, the RCI elevates to 94.93% and SHI descends to 0.12 in the 335 mm prototype system while enclosing the front and both sides of cold aisle. Also, the RCI achieves 99.82% and SHI mitigates to 0.095 in the 210 mm alternative system.

參考文獻


[1] Cisco, "Cisco Global Cloud index : Forecast and Methodology, 2015-2020 White Paper," 2016.
[2] United States Environmental Protection Agency, "Report to congress on server and data center energy efficiency: Public law 109-431," Energy Star Program, 2007.
[3] ASHRAE Technical Committee, "Thermal guidelines for data processing environments-expanded data center classes and usage guidance," American Society of Heating Refrigerating and Air conditioning Engineers, 2011.
[4] The Grid Green, "The Green Grid data center power efficiency metrics: PUE and DCiE," Green Grid report paper, 2007.
[5] S. Flucker, R. Tozer, "Data Centre Cooling Air Performance Metrics," CIBSE, 2011.

延伸閱讀