数据中心冷却失效引起的机房温升研究
|
摘要
总结了对数据中心冷却失效引起的温升研究成果,并进行了基本结论的推导,给出了数据中心传热模型的描述方程式。应用6SigmaDC软件对一个100m~2的数据机房进行了3D建模和CFD仿真,模拟其冷却失效后的机房温升情况。基本结论为:数据中心的冷却失效时间必须与数据中心的电源后备时间匹配,否则需设计专门的蓄冷装置做一定时间(一般为15min)的冷源备份;在功率密度小于10kW/机柜时,数据中心冷却失效后引起的温升速率与功率密度基本呈线性关系;数据中心冷却失效引起的温升在平均进风温度达到80℃前,温升速率基本恒定。
Summarizes the current research results on room air temperature rise caused by cooling failure in data centers,derives some basic conclusions,and obtains the description equations of heat transfer model for data centers.Taking a 100 m~2 data center as an example,establishes a 3D model and performs CFD simulation using 6SigmaDC software,and simulates the temperature rise after cooling failure.The basic conclusions are as follows:in the data center,the duration of cooling failure must be compatible with power backup response time;otherwise a special cool storage device should be designed as backup cooling source(generally 15 minutes).When power density is less than 10 kW per rack,the relationship of temperature rising rate and rack power density is basically linear after cooling failure.When the average inlet air temperature is lower than 80℃,the temperature rising rate caused by cooling failure is approximate constant.
Summarizes the current research results on room air temperature rise caused by cooling failure in data centers,derives some basic conclusions,and obtains the description equations of heat transfer model for data centers.Taking a 100 m~2 data center as an example,establishes a 3D model and performs CFD simulation using 6SigmaDC software,and simulates the temperature rise after cooling failure.The basic conclusions are as follows:in the data center,the duration of cooling failure must be compatible with power backup response time;otherwise a special cool storage device should be designed as backup cooling source(generally 15 minutes).When power density is less than 10 kW per rack,the relationship of temperature rising rate and rack power density is basically linear after cooling failure.When the average inlet air temperature is lower than 80℃,the temperature rising rate caused by cooling failure is approximate constant.
引文
[1]ASHRAE TC9.9.Thermal guidelines for data processing environments[M].3rd ed.USA:ASHRAE,2011:14
[2]Emerson Network Power and Ponemon Institute.National survey on data center outages[J].Ponemon Institute Journal,2010(1):12
[3]Emerson Network Power.Choosing system architecture and cooling fluid for high heat density cooling solutions[J].Emerson White Paper,2011(1):8
[4]MOSS D L.Facility cooling failure:how much time do you have?[J].Dell White Paper,2011(1):5
[5]MENUET R,TURNER W P.Continuous cooling is required for continuous availability:data center managers need to match their availability expectations to the appropriate class of continuous cooling[J].Uptime White Paper,2006(1):7
[6]Active Power.Data center thermal runaway:a review of cooling challenges in high density mission critical environments[J].Active Power White Paper,2007(1):8-9
[2]Emerson Network Power and Ponemon Institute.National survey on data center outages[J].Ponemon Institute Journal,2010(1):12
[3]Emerson Network Power.Choosing system architecture and cooling fluid for high heat density cooling solutions[J].Emerson White Paper,2011(1):8
[4]MOSS D L.Facility cooling failure:how much time do you have?[J].Dell White Paper,2011(1):5
[5]MENUET R,TURNER W P.Continuous cooling is required for continuous availability:data center managers need to match their availability expectations to the appropriate class of continuous cooling[J].Uptime White Paper,2006(1):7
[6]Active Power.Data center thermal runaway:a review of cooling challenges in high density mission critical environments[J].Active Power White Paper,2007(1):8-9