Study On The Data Center Airflow Simulation Improvement And Startup Strategies

In the fast-developing information age,the high energy consumption caused by the high demand of data centers needs to be greatly improved.In the research and development of data center energy conservation and consumption reduction by using ventilation and cooling technology,good airflow organization and IT equipment startup deployment strategy are the key to improving the heat dissipation performance of the data center equipment room,and further research is needed.Therefore,in this paper,the steady-state and transient numerical simulation methods are used to seek the energy-saving operation startup strategy,and the two main problems of improving the airflow organization and the opening and distribution mode of the overall load of the equipment room are studied.Firstly,energy conservation and consumption reduction are achieved by improving airflow organization mode.A closed channel cold channel closure model is proposed to further curb the cooling and hot air flow mixing,effectively improving the cooling capacity utilization of the air conditioning system;on this basis,through the comparative analysis of multiple sets of numerical experiments,it was found that a local hot zone will be formed if the air supply size is too small.The maximum temperature of the server will decrease first and then increase with the change of the air supply angle,and the optimal air outlet size and angle are obtained.In the air conditioning layout mode,when the heat is symmetrically arranged,the ambient temperature is the lowest.The more dispersed the air conditioning distance is,the worse the cooling effect has;Also,the power variation is linearly related to the temperature change,and the corresponding functional relationship is calculated.Secondly,after rationalizing the airflow organization,the load distribution is improved to further explore the energy-saving space.Based on the numerical study with thermal steady state,a power distribution model based on temperature averaging is established.The results show that this distribution method can effectively eliminate local hot spots due to different airflow distribution so as to improve the utilization of cooling capacity.The middle layer is the best,and the lower layer is the worst.At the same time,the thermal environment performance index is used for evaluation.It is found that for the steady-state power-on strategy selection,50%and above of the medium-high startup load state,the Model5 has the best cross-over-distribution mode.Under the low starting load state of 50%or less,the distribution mode of the Model2 concentrated middle layer has obvious advantages.Finally,the transient temperature rise characteristics of data center cooling failure and dynamic variable operating conditions are analyzed to provide guidance for the safe and energy-saving operation of IT equipment.Through the thermal transient numerical study,the influences of air conditioning parameters and IT equipment power changes on the transient temperature rise of the startup are analyzed.The differences of heat dissipation deterioration after different airflow containment models are compared,and the power of the cold channel closed mode is obtained.The relationship between the power density and temperature rise rate;the temperature change characteristics of transient thermal environment under variable working conditions are analyzed.The influence of power change is fast,but the change of cooling capacity plays a leading role.For the dynamic change of starting quantity,the direct increase and phased transition are adopted.When the power-on amount is reduced as well as the power-on position changes dynamically,the cabinet server startup position is sequentially turned on from the top to the bottom,and the deployment scheme is undoubtedly an ideal booting strategy that is safe and feasible.