Research On Airflow Organization And Influencing Factors In Data Center

With the rapid development of new technologies such as Internet of Things,cloud computing,big data,artificial intelligence,and blockchain,the data center has received more and more attention and attention.The rapid development of data centers in China has led to a sharp increase in energy consumption in data centers.In 2015,the energy consumption of data centers in China reached 100 billion kWh.It is expected that the power consumption of data centers in China will exceed 250 billion kWh in 2020.The increasing energy consumption of data centers year after year has not only caused a heavy burden on enterprises,but also brought enormous challenges to the energy and environment of the entire society.The unreasonable design of the data center air flow organization may lead to local overheating in the operation stage of the server equipment.In order to reduce the local hot spot temperature,it is necessary to lower the setting temperature of the entire date center,thereby causing a huge waste of energy.The unreasonable design of data center air flow organization is one of the main reasons for the high energy consumption of data centers.Therefore,the rational optimization of design data center air flow organization has become the key to the design of data room air conditioning systems.Domestic and foreign scholars have done a lot of research on optimization of data center air flow organization and evaluation methods.Some of these studies have analyzed the influence of certain factors on the temperature field and velocity field within the data center through CFD simulation.Excellent operating conditions,finally combining the optimal operating conditions of each factor to obtain an overall optimized design plan for the airflow center of the data center.The lack of analysis of airflow distribution and pressure distribution uniformity in the raised floor,and the distribution of the airflow and the pressure distribution uniformity in the raised floor.It is the fundamental reason for the uneven distribution of temperature and speed in the equipment room.The existing research lacks research on the primary and secondary sequences of the influence of various influencing factors on the airflow organization in the data center.The simple combination of optimal conditions obtained by single factor optimization may not be the optimal combination.Some scholars have proposed data center airflow organization evaluation methods based on air-conditioning supply air temperature and rack inlet and outlet air temperature deviations.These evaluation methods cannot reflect the air-conditioning status of the entire engine room from the system and energy level,and have a certain degree of practical application.limitation.Exergy loss is the physical quantity describing the energy quality in the energy flow system,but the data center heat transfer process does not involve the heat transfer process of the heat transfer,and the loss of its heat transfer capacity is not equivalent to the loss of energy to make the power.Therefore,it is not applicable to evaluate the data center thermal environment with the exergy loss.Based on the current problems in data center air flow organization and evaluation methods,this paper applies the entransy loss to the data center air flow organization evaluation,and conducts the following research work on the standardized data center as the object:1.Through the professional CFD simulation software,a standardized data center model for air distribution on the air duct,underfloor air supply,and air return before air conditioning between rows of air conditioners is established.The temperature field and velocity field for the three types of airflow organization are established.A simulation analysis was carried out to evaluate the three types of airflow organization through the thermal principle plot loss.It was found that the degree of mixing of hot and cold air was air supply on the air duct> air supply under the floor> air return after the front air supply.Verify the rationality of the application of entransy loss to the assessment of airflow organization in the data center.2.As the "Data Center Design Specification" GB50174-2017 recommends that a host cabinet with a heat output more than 4 kW is used for a single cabinet,it is recommended to use the underfloor ventilation/upper return air supply,inter-row cooling air conditioning before the air supply,and after air return.And the high cost of air conditioning between the columns,applicable to single-rack power 10~30kW date center.At present,data centers mostly use the form of underfloor air supply.Therefore,in this paper,a standardized data center in the form of under-floor ventilation is the main research object.Through the use of CFD numerical simulation methods,a comprehensive analysis of air distribution,pressure distribution uniformity,floor grille air volume,and temperature distribution in the equipment room of the raised floor is conducted.The basic causes of the uneven temperature distribution in the room are: the influence of the height of the raised floor,the open rate of the floor grille,and whether the closed hot and cold aisle are closed to the influence of the air distribution in the data center.The entransy loss evaluate the three factors from the system and energy levels,quantify the degree of mixing of hot and cold air in various working conditions,and then optimize the design parameters of data center airflow organization.3.Taking the entransy loss as the test index,the orthogonal test is used to analyze the combination of three factors that affect the air floor structure of the data center,such as the height of the raised floor,the open rate of the floor grille,and whether the closed hot and cold passages are closed.The significance of the three factors on the test results was analyzed by range analysis and analysis of variance.The order of importance of the three factors on the thermal loss was: whether the closed hot and cold aisle> overhead floor height> floor grille opening rate,combined with the calculation of the entransy loss,shows that the height of the raised floor is 800 mm,and the floor grille opening rate is combined and the closed cold aisle scheme is the optimal air distribution scheme of the standardized data center.4.By putting up a data center lab bench,test the optimal combination of orthogonal experiments to test the accuracy of the simulation results.By optimizing the design parameters of the data center,the energy consumption of the data center can be saved,and the theoretical basis and method guidance for the design of data center air flow organization can be provided.Data center air-conditioning system energy consumption accounts for about 40% of the total energy consumption of the data center,and data center air-flow organization is directly related to the cooling effect of the cabinet and the energy consumption of the air-conditioning system,thus optimizing the data center air-conditioning organization has great potential for energy saving.There are very important social and economic benefits for energy conservation,consumption reduction and green environmental protection.