| The data center,as an important digital transmission hub,not only brings significant economic benefits,but also causes huge energy consumption.The construction,operation and maintenance and transformation of green data centers have become new goals.The raised-floor data center,as the mainstream data center layout form,has the advantages of precise air supply and separation of cold and hot aisles.However,due to the uneven air distribution in the cold aisle of the data center,local hot spots or excessive cold air are likely to exist in the room,which causes potential safety hazards and energy waste.Therefore,this thesis is dedicated to the optimization of the thermal environment and air distribution of data center.A scheme to improve the heat flow environment of data center is proposed,and the influence of various factors is considered,which can provide a design reference for the energy-saving transformation of the existing data center.The main research content of this article includes the following aspects:First.Research on the improvement of the thermal environment and airflow organization of the data center.The article uses computational fluid dynamics(CFD)numerical simulation as the main research method.Based on the Georgia Institute of Technology’s data center,a simplified CFD model is established by the commercial software Flotherm,and the accuracy and reliability of the model are verified.The research analyzes the problems existing in the heat flow field of high-density data centers.The rectifier baffle layout strategy is introduced in the high-density data center,and the full-factorial design method(DOE)is used for research.The two-way ANOVA method was used to analyze the correlation between the length and angle of the rectifying baffle.Scecond,Analysis of the influence of multiple factors on the heat flow environment of the data center.Based on the entransy dissipation principle,the loss in the heat transfer process of the data center is analyzed,and the rectifier baffle layout strategy is optimized based on this.Then,the influence of the depth of the plenum,the opening rate of the perforated tiles and the air supply angle of the CRAC on the heat flow environment of the data center and the layout strategy of the rectifying baffle are studied.The changes in the rack inlet temperature and the flow rate of the perforated tiles clarify the influence of the above-mentioned factors.In addition,the temperature cloud map and airflow cloud map are used to compare the changes in the heat flow field of the data center before and after optimization.Third,Optimization and thermal management evaluation of data center influencing factors.In order to find the best solution,this thesis uses genetic algorithm(GA)to optimize the main influencing factors of the rectifier baffle layout strategy.Use the CurveFitting toolkit to fit the objective function with 25 sets of discrete data points,use the center and scale method(CAS)to improve the fitting accuracy,and use the sum of squared errors(SSE)and coefficient of determination(R-Square)indicators to evaluate the fitting function.Finally,the best optimization plan is evaluated through the data center thermal management index.The room optimization plan proposed in this article has improved in three aspects:airflow organization management,mixing degree of hot and cold air,and energy consumption of CRAC and refrigeration. |
