COOLING ARCHITECTURE
A dynamic approach Sudlows’ Sam Wicks explains how Computational Fluid Dynamics can improve the efficiency of cooling architecture
Computational Fluid Dynamics (CFD) produces efficiency predictions for facilities yet to be built and provides scientific validation for cooling architecture designs. When absorbed into the data centre design phase, CFD drives design, unlocking the most efficient data centre cooling designs possible.
C
Sam Wicks is CFD applications engineer at Sudlows.
FD is a powerful tool that data centre managers, designers and auditors have outsourced in the past to predict the behaviour of air flow within a data centre, but recent advances in computational power, CFD algorithms and software have enabled advanced, in-house capability within the data centre industry. To build the most energy and cost efficient data centre cooling solution, it is
necessary to be able to predict 38 | Architects Choice |
ArchitectNews.co.uk
the cooling performance at the design stage. This allows the designer to explore the countless permutations of data centre layouts and cooling architectures in order to determine the best solution. There are many methods for predicting the behaviour of a data centre during the design phase. For an air-cooled data centre, current cooling design practices are at the scale of an entire room, with the heat outputs from rack equipment offset by the cooling capacity of the computer room air conditioners, coupled with volume flow rate
recommendations. This method is illustrated by Figure 1. It is known that a volume of air exits the CRAC unit, per second, at a specified temperature. It is also known that a server has a volume of air passing through it per second,
and that the server will transfer an amount of heat to this air. However, there is no guarantee that the air that leaves the CRAC ever reaches the rack. Instead, it is assumed that a sufficient volume of the cooled air reaches the server inlets, without knowledge of the exact volume. The placement of racks and CRACs is heuristically determined to reach this goal, with general rules of thumb guiding the design.
There are further guidelines added on top of this approach, increasing in complexity due to increased server densities, including the use of hot or cold aisle containment. By separating the hot exhaust air from cold inlet air, it is assumed that racks will always draw cold air. The architecture is an improvement compared to uncontained solutions, but the design assumption ignores a
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52