COOLING ARCHITECTURE
location, floor tile location, plenum depth) can be tuned such that the design is optimised to meet to cooling requirements in the most efficient or cost effective way. The accuracy of CFD also allows for rigorous design, with facility running conditions set to tighter tolerances, allowing for higher densities than previously possible. The heat distribution within the hot aisle shown in Figure 3 is an illustration of how CFD can be used to position and size in-row coolers. Even though they draw their air from the same hot aisle contained region, the CRAC inlet temperatures are not necessarily uniform. This is due to a number of interrelating variables, including the IT load
distribution, server flow rates, CRAC fan set point and the heat distribution in the region exterior to the hot aisle. Each variable can be isolated and its dependencies tested, allowing optimisation of the design. CFD is also useful for existing facilities, as it makes the invisible visible. After taking key measurements within the data centre, a CFD model provides a complete description of temperatures and flow rates at any location within the facility, allowing data centre operators to identify hot spots,
inefficiencies and crucially their causes, as well as appraise their current capacity with a view to future requirements. The model can easily be adjusted to show the facility cooling behaviour with
Figure 3 – plan view of in row cooling with hot aisle containment.
additional IT load, or changes to infrastructure and layout, without the logistics of relocating racks before the benefits are clear, with a full before-and-after description with efficiency gains and cost benefits shown by means of pressure, velocity and temperature contour plots, graphs, 3D visualisation and streamline analysis. The effects of a different CRAC running point can be tested safely, without incurring either high costs or
“When utilised early and regularly in the design phase, CFD can be used to drive design, unlocking the most efficient data centre cool- ing designs possible.”
Figure 2 – CFD temperature plot of Figure 1 scenario.
risk to mission critical systems. As a tool, CFD has an unmatched capability to help data centre operators locate sources of inefficiency within their existing facilities. As the method deals directly with the underlying principles of data centre cooling, CFD also brings to light the causes of these inefficiencies, providing operators with a path toward efficiency gains. For the data centre designer, CFD allows temperature distributions to be predicted for facilities yet to be built, providing scientific validation for cooling architecture designs. When utilised early and regularly in the design phase, CFD can be used to drive design, unlocking the most efficient data centre cooling designs possible. Although CFD software is advancing rapidly, a thorough understanding of the underlying physics is still necessary on the part of the CFD engineer to ensure the modelling assumptions and initial data inputted into the simulation, as well as the results, are accurate. This process must be performed in tandem with validation and verification to ensure CFD results are fit for purpose. n
40 | Architects Choice |
ArchitectNews.co.uk
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