Dynamic performance analysis and thermal modelling of a novel two-phase spray cooled rack system for data center cooling
Cooling systems in information technology (IT) equipment account for the greatest share of direct energy consumption in data centers. Conventional air-cooling schemes are thermally inefficient and energy intensive as they require large temperature difference between heat source and cooling medium. T...
Saved in:
Main Authors: | , , , , |
---|---|
Other Authors: | |
Format: | Article |
Language: | English |
Published: |
2023
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/169068 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
Summary: | Cooling systems in information technology (IT) equipment account for the greatest share of direct energy consumption in data centers. Conventional air-cooling schemes are thermally inefficient and energy intensive as they require large temperature difference between heat source and cooling medium. To overcome the aforementioned drawbacks of air-cooling, a “chillerless” novel spraying architecture which has the capability of performing high heat flux cooling, is highly scalable and easily adaptable by modern data centers is proposed in this paper. To demonstrate the scalability and performance of this new cooling scheme, a full-scale spray cooled rack system which has the capacity of housing up to 12 server units is developed. The dynamic performance of the spray cooled rack system under variable ambient temperature is studied. It is found that the ambient temperature strongly affects the spray cooling performances and the chamber pressure stability. A grey box model of the system is established. The newly developed model has successfully predicted the temperature variation on both the water side and spray cooling side with reasonable accuracy. Our model sets the foundation for predictive control of the chamber pressure and workload management of the spray cooling scheme for application in large-scale data centers. |
---|