Investigation On Experiment And Simulation Of Flat-microchannel Heat Pipe Based On Data Center Heat Dissipation

At present,information technologies such as cloud computing,Internet+,mobile Internet and mega data services are developing rapidly,and the scale of data center construction continues to expand.The accompanying high energy consumption problem of data center cooling systems has become increasingly prominent.The traditional data center cooling system is difficult to achieve accurate heat dissipation,which also causes the problem of excessive local temperature.Therefore,improving the energy efficiency of the data center and solving local hot issues are of great significance to the energy saving and emission reduction of the data center.In this paper,experimental and theoretical methods are used to study the performance of the microchannel flat loop heat pipe(MCFLHP)heat dissipation system.This research innovatively installs two micro-channel flat heat pipe evaporators inside the data cabinet,which can achieve efficient and precise heat dissipation at the server level.The micro-channel flat heat pipe evaporator increases the heat exchange area between the evaporator and the IT equipment,which is beneficial for the micro-channel flat evaporator to directly absorb heat from the surface of the IT equipment and output heat.The evaporator and condenser of the proposed system can be installed tens of meters apart,and the condenser can be installed outside the data center,which contributes to achieve passive heat dissipation and ultra-low carbon emissions of data center cooling system.The main content of this article is as follows:First,the working principle of the proposed system based on the heat dissipation of the data center and the construction of the experimental platform are introduced,and the material selection of MCFLHP,the selection of the refrigerant,the processing technology of the experimental device and the determination of the experimental test plan are described.Secend,the experimental platform of the MCFLHP system based on the heat dissipation of the data center is tested and the key factors affecting the heat recovery efficiency of the system are studied,such as filling rate,coolant water supply temperature,coolant water flow rate,and simulated heat source heating power.The average heat recovery efficiency of the system first increases and then decreases with the increase of the filling rate and the cooling water flow rate,and decreases with the increase of the coolant water supply temperature and the heating power of the simulated heat source.The optimal filling rate of the system is 25%,the optimal coolant water flow rate is 300 L/h,the optimal coolant water supply temperature is 15 °C,and the optimal simulated heat source heating power is 600 W.At last,the heat transfer process of the system is analyzed,and the dynamic heat transfer model of the system is established.Under different filling rates,different coolant water supply temperatures,different coolant water flow rates and different simulated heat source heating powers,the relative error range of the average heat recovery efficiency obtained from the experiment and the simulated value is 1.10%-9.54%.The error is caused by the model simplification and a certain accuracy error in the experiment,which proves the accuracy and reliability of the dynamic heat transfer model.The heat recovery performance of the system under different evaporator and condenser height differences is predicted by the dynamic heat transfer model.It is found that the height difference between the evaporator and the condenser has little effect on the heat recovery performance of the system,and the optimal evaporator and condenser height difference of the system is determined to be 1.0 m.Under the optimal working conditions(e.g.the filling rate is 25%,the coolant water flow rate is 300 L/h,the coolant water supply temperature is 15 °C,the heating power of the simulated heat source is 600 W,and the height difference between the evaporator and the condenser is 1.0 m),the average heat recovery efficiency is the highest,with an experimental test value of 88.03% and a simulated value of 89.50%,and the error is only 1.64%.