Simulating the future of data centre thermal management
May 28, 2026
Flownex and ES2 collaboration
Flownex has recently joined the US National Science Foundation (NSF) Centre for Energy Smart Electronic Systems (ES2) – an Industry/University Cooperative Research Centre (I/UCRC) (https://www.binghamton.edu/es2/).
Across the industry, there is a clear move towards tightly integrated infrastructure, higher rack densities, predictive controls, and increasingly complicated thermal and power management.
For Flownex, this creates a natural opportunity for collaboration at the intersection of hardware, controls, and simulation. The company’s background in system-level thermal-fluid simulation — often referred to as flow network modelling (FNM) — together with its experience in controls integration and physics-based digital twins, originated in the nuclear industry. Those same capabilities now align strongly with many of the challenges being explored within ES2, particularly as physics-based prediction becomes just as important as AI and data analytics in managing next-generation cooling systems.
andre van der walt
Head of Global Sales
vincent britz
Principal Application Engineer
why simulation and flow network modelling is important for modern data centres?
Flow network modelling has been part of the nuclear industry for decades, where accurate transient thermal-hydraulic analysis is essential. Flownex itself was developed within the quality framework required by the US nuclear industry, including compliance with NQA-1 standards associated with the U.S. Nuclear Regulatory Commission. That background brings a level of simulation rigour and quality assurance that is uncommon outside the nuclear sector.
Today, Flownex has evolved far beyond a conventional flow network modelling tool. It serves as a simulation platform for thermal-fluid systems, allowing engineers to study systems at multiple levels of detail while integrating controls, transient analysis, CFD models (through API connection), and real equipment performance into a single environment. The platform can solve multiple fluids within the same network and supports both single- and two-phase modelling, making it particularly well suited to modern data centre cooling challenges and optimising new cooling technologies.
These capabilities have made Flownex increasingly valuable in the analysis of data centre thermal management systems. In air-cooled facilities, it has helped engineers evaluate transient conditions during power failures and thermal ride-through events, reducing the risk of breaching SLA temperature limits. As AI-driven facilities push the industry toward higher-density liquid cooling systems, Flownex has also proven effective in identifying control issues and refining operational strategies before deployment. Flownex’s contribution to the ARPA-E funded Omnicool project further demonstrated how physics-based digital twins can support the development of next-generation cooling technologies.
Through ES2, Flownex now has the opportunity to collaborate with academic partners including Villanova University, Binghamton University, and The University of Texas at Arlington across several important research areas to promote faster development of technologies for next-generation data centres.
conclusion
The challenges facing modern data centres are becoming increasingly interconnected. Cooling systems, controls, AI workloads, energy storage, and operational optimisation can no longer be treated as isolated problems. Through its collaboration with ES2, Flownex brings a proven physics-based simulation approach that can help bridge these areas within a unified digital twin environment.
As the industry continues moving toward higher power densities, liquid cooling, and autonomous operation, the combination of predictive simulation, controls integration, and real-time digital twins is likely to become an essential part of how next-generation facilities are designed and operated.
To find out more about our data centre applications, please visit our industry page https://flownex.com/industries/data-centre/ or contact our sales team at sales@flownex.com
how flownex can support the es2
research themes:
holistic it load management and control
Flownex’s capability to serve as a platform for physics-based digital twins could be used to develop predictive workload-aware cooling strategies by linking IT load behaviour directly to cooling system dynamics.
Air cooling systems
By combining 1D system-level thermal modelling with detailed CFD analysis of data halls, Flownex provides a framework for co-simulation to study both air and liquid cooling systems together at very high fidelity. This allows researchers to investigate transient events and operational scenarios without relying heavily on simplified boundary condition assumptions.
control and optimisation
This is one of the strongest areas of alignment between Flownex and the ES2 program. The platform’s integrated controls functionality enables research into optimised control strategies for air, liquid, and two-phase cooling systems. It also creates opportunities to investigate digital twins for virtual commissioning and advanced operational optimisation.
warm water liquid cooling
Flownex can model warm-water cooling loops, CDUs, cold-plate networks, and transitions between free cooling and mechanical cooling modes. This creates opportunities to study compressor-less cooling strategies under dynamic AI workloads and varying environmental conditions.
two-phase cooling
With strong two-phase modelling capabilities, Flownex can support research ranging from cold plate behaviour to full facility integration. Areas of interest include loop stability, two-phase pressure drops, condenser interaction, transient thermal response, and the integration of two-phase systems into broader plant cooling infrastructure.
waste energy harvesting
Because Flownex is used across multiple industries, it offers a broader perspective on heat recovery and energy reuse opportunities. Research could explore district heating integration, thermal storage systems, and dynamic waste heat recovery strategies under changing operating conditions.
power distribution and energy storage
Flownex’s transient simulation capability makes it well suited to evaluating the interaction between cooling infrastructure, backup power systems, and energy storage technologies. An integrated modelling approach could provide valuable insights into resilience, emergency load management, and operational reliability.
embedded cooling
The platform’s multi-scale modelling capability allows researchers to connect chip-level cooling to rack and facility-level infrastructure. This supports research into advanced embedded cooling architectures, including both single and two-phase direct-contact cooling technologies.
data centre metrics and system energy optimisation
Physics-based operational digital twins could support optimisation of PUE, water usage, resiliency, thermal margins, and equipment performance. Research in this area could focus on predictive operational optimisation and on combining AI with physics-informed simulation models.