دانلود رایگان مقاله ارزیابی عملکرد سیستم خنک کننده در مراکز داده

1. Introduction

Data center is a facility containing computer equipment and related devices such as storage systems of a telecommunication system. These devices consist of power supply, air conditioner, fire-fighting devices and security systems [1]. Air conditioner of data centers should work full time all over the year to remove the heat added to the system by operation of different devices. Development of telecommunication industry, therefore, made data center a common element for either general or dedicated usage [2]. Heat propagation increases the temperature and causes damage to rack servers; therefore protection of system and prolonging their life by cooling the inherent devices is of crucial importance. In other words, it's very significant to design an appropriate air conditioner to protect data center from destruction caused by heating. For proper operation of devices, they should be held on a certain and proper temperature. Each IT server should receive certain amount of cool air enough for the removal of the produced heating [3]. Appropriate operation of equipment depends on proper distribution of cool air on data center and the response of supply air to the demand of components. Computational fluid dynamics is an assisting tool to configure the hall of devices. Some indexes are used for quantitative assessment of efficient performance of cooling systems. In practice, influence of the design factors is quantified using cooling indexes such as SHI, RHI, RCI and RTI (described in detail later in Section 2) for different models where each index indicates specific aspect of operation in the cooling systems; however in some cases the mismatch among existing approaches might end up with irrelevant results which makes decision making process a challenge.

6. Conclusion

A new index is introduced here as comprehensive cooling index (CCI) to identify the best design of data centers. The verification of the developed method is sought on meaningful case study in Iran where Marand data center is assessed in more detail. Different configurations of the data center (15 configurations) were analyzed to find out the most desirable design of the cooling system. Based on the obtained results, the following conclusion is derived: 1) Cooling systems by downward blowing have more cooling efficiency in comparison to upward blowing packages. 2) Not only closing the chamber of cool aisles but also local channeling of hot aisles improves the efficiency of cooling system. 3) Closed chamber cooling system has better performance than cooling system with normal under floor. 4) Conducting cool air to the IT equipment by either channeling or partitioning is useful for improving the efficiency of cooling system where channeling is more effective than partitioning. 5) If the cool air flow is across the aisle direction better results are obtained than perpendicular configuration.