Font Size: a A A

Research On The Influence Of Air-conditioning Multi-tuyere Layouts And Air Flow Pattern On Thermal Environment Of Data Centers

Posted on:2021-09-19Degree:MasterType:Thesis
Country:ChinaCandidate:W ChenFull Text:PDF
GTID:2492306353462994Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
Due to the rapid development of the Internet,data centers are springing up in every corner of the world,which not only bring benefits to the development of science and technology and economy,but also consume a lot of energy.Among the energy consumption of data centers,the cooling system occupies a high energy consumption ratio.Currently,in the field of green data centers,more and more scholars begin to pay attention to the optimization of cooling system.With the need of data communications equipment is stable and reliable operation,the inlet temperature of the servers needs stability in a certain range,and the cooling capacity through each server rack in a raised-floor data center is non-uniform,the cooling capacity of the server racks on the side near the computer room air-conditioning(CRAC)units is far less than that on the side far away from the CRACs,which will cause the waste of cold air and raise the inlet temperature of the cabinets.In order to reduce the inlet temperature of the cabinets,CRACs need to lower the supply cold air temperature or improve the flow rate,which can cause a huge waste of energy in the data center.How to optimize the cold air supply distribution in data centers is an important task for the optimization of cooling system.The research contents of this paper mainly include the following aspects:(1)Firstly,computational fluid dynamics(CFD)method is used to analyze air flow pattern at the outlet of CRACs,that is,the influence of the air outlet mode on the thermal environment of the data center,which mainly studies the inlet temperature of the cabinets,the airflow characteristics of the plenum under the floor and the flow distribution through the perforated tiles,and also analyzes the applicability of the influence laws of the air outlet modes under different cabinet thermal density.The results show that the inlet temperature of the cabinet decreases obviously when the air supply mode is sheared,which provides a useful scheme for optimizing the cooling performance of the data center.When the supply air is normal,the amount of cold air required for cabinet level cooling comes from the hot air appearing in the upper half and bottom of the cabinet(leakage at the bottom of the cabinet),which will lead to the mixture of hot and cold air in the cold aisle,resulting in the waste of energy.However,this thermal trend can be improved when sheared air flow is used.(2)Then,CFD method is used to analyze the influence of CRACs layouts on the thermal environment of the data center.During the study,it is found that among the three basic layouts in the computer room,the flow distribution through the perforated tiles of the diagonal layout is more uniform,which not only reduces the cold air bypass in the middle rack,but also reduces the hot air recirculation in the end rack of the rack row.In the process of research,the layout of CRACs in the computer room should follow the rules:(i)the CRACs in the computer room should face the hot aisle rather than the cold aisle,because the static pressure under the raised-floor of the cold aisle should be the largest;(ii)when the CRACs are arranged in parallel,their air supply direction should increase the static pressure of the space under the floor,rather than make the airflow coil flow conflict,resulting in the static pressure drop in these areas,resulting in the overall loss of cold supply.(3)Finally,the response surface method(RSM)is used to optimize the influencing factors,find out the optimal combination scheme,and evaluate the scheme reasonably.First,according to the Box-Behnken Design,the three-factor and three-level experimental design is carried out,and the cost function is selected as the response.Then,the response surface is fitted with the Design-Expert software and nonlinear GRG method is used to optimize the response surface equation.The table of the analysis of variance(ANOVA)shows that the regression equation obtained has a good fitting degree,and the response surface equation can be used to replace the test point to analyze the experimental results.The optimal combination of factors is as follows:the air flow pattern at CRACs outlet is sheared,the CRACs layout is diagonal and the spacing between the two CRACs is 1.45 m,the depth of the plenum is 800 mm,and the opening rate of the perforated tiles is 25%.The data center with the best combination is simulated.And the numerical results are compared comprehensively with the results of the basic model from three aspects,such as the thermal environment of the data center,the evaluation index of the thermal environment,and the cooling energy consumption of the CRACs.It is concluded that the optimal combination scheme can effectively reduce the energy consumption and improve the energy efficiency of the data center.
Keywords/Search Tags:data center, CRAC, thermal environment, CFD, response surface method
PDF Full Text Request
Related items