Increasing chip densities and high-powered processors are forcing data centers to consider what technologies they will need to implement to overcome the heat that new IT equipment produces. Liquid cooling, what once seemed like an impossible concept, is fast becoming an accepted and sought-after technology in the data center industry.
Is your data center ready to operationalize liquid cooling? Is there a verifiable way to test this technology before committing to it? Is there any way to compare air-cooled and liquid cooled data centers?
The Digital Twin is the only tool that can answer all of the above. This blog examines a simple liquid/ hybrid cooled data center model in our product suite 6SigmaDCX to demonstrate how 1D/3D co-simulation uses the best of both 1D and 3D simulation to overcome the unique modeling requirements that liquid cooling presents.
The Digital Twin is a virtual replica of the data center that can simulate its physical operations under any working conditions.
In turn, the Digital Twin can simulate the liquid cooling loop and airflow from IT to facility level to accurately model liquid cooling implementation in the data center. For a more detailed look into how our software does this, we walk you through a “chip to chiller” thermal design methodology in this blog.
If you are looking to evaluate the impact of your chilled water system on data center whitespace, we suggest reading about our 1D Liquid Side Network to better understand how water flow is directed around your facility.
Instead of a lengthy and expensive experimental analysis, the Digital Twin enables you to virtually test how liquid cooling would work in your data center. While some CFD models may have gotten a bad reputation for having lengthy solve times, our software is driven by the fastest, most powerful CFD solver on the market.
Even with our solver, however, liquid cooling is intricate enough that solely solving in a 3D model could be computationally excessive. To combat this issue, we have connected the 3D model with the 1D network, which makes for quick solve times and easy thermal analysis.
To obtain the heat capture ratio, or the ratio of heat that was picked up by liquid over the total heat, our software uses the static pressure drop and thermal resistance curves of cold plates, which can be obtained from their detailed CAD model. After completing the IT level analysis, the IT model can be used in the rack level model and, in turn, the rack level model can be used in the 6SigmaRoom model to complete the solve.
By incorporating the 1D network with the 3D model, we’ve saved you grid cells and therefore solve time. This connection is unique to our software.
Figure 1. 3D simulation of liquid cooling in a data center model (above) and a 1D flow network (below)
The only constant in the data center industry is changing IT. IT equipment will continue to develop, accelerate, and specialize. Data centers, from edge to legacy, will have to adapt. Data center simulation is one sure-fire way to test out new technology before committing to it and helps you stay flexible in new and turbulent times.
Other Recent Posts
Data Center Simulation Case Study: CFD Software Supports New DCMS
“A sovereign, eco-responsible data center offering the best level of security and availability at a…
Kao Data: Using CFD Analysis to Design and Build the UK Innovation Corridor’s Home for HPC and AI
Wholesale and high-performance colocation provider Kao Data worked together with Future Facilities,…
22 February, 2021