COA-CFD closes its final review with an “Excellent” rating

The COA-CFD project, Cloud-based Online Access to Computational Fluid Dynamic Simulations, run under the CELTIC-NEXT EUREKA Cluster, has completed its final review with an “Excellent” rating.

The review panel comprised Xavier Priem, Director of CELTIC-NEXT, Harold Linke as technical reviewer, and Michael Walch of the Austrian Research Promotion Agency (FFG) representing the local funding agency. Beyond the rating, the reviewers encouraged the consortium to submit a follow-up proposal to exploit the project’s results and extend them into new application areas.

What the consortium set out to do

CFD simulation has long belonged to large corporations with supercomputers and in-house specialists. COA-CFD set out to change that: to make high-fidelity fluid simulation affordable and accessible to SMEs, delivered on the cloud, and usable without expert knowledge.

Four years on, from March 2022 to August 2026, that mission is what the reviewers responded to most directly.

What was delivered

A family of solvers built for accessibility. Lattice-Boltzmann (LBM) for general flow and conjugated heat transfer, Lagrangian Differencing Dynamics (LDD), a numerical method developed by ESS researchers for violent free-surface and non-Newtonian flows, a finite-difference solver for electrostatics, a Lagrangian particle solver for droplet and dust clouds, and a thin-film solver. All of them avoid complex mesh preparation, all are built for massive parallelism, and all are GPU-accelerated. A hybridised solver combines them for genuinely multiphysical industrial problems.

A modular user experience. Rather than one generic package, COA-CFD produced application-specific software modules, assembled from reusable components through a graphical GUI Creator, so a new application does not mean a new interface built from scratch.

An on-demand cloud platform. Beyond the software packages themselves, the project delivered a cloud platform on which every application can be reached from anywhere in the world, by non-experts, with no dedicated hardware. Built on microservices, GDPR-compliant hosting, ESS in-house GPU runners and European HPC capacity, and aligned with Gaia-X principles for data sovereignty. More than 15,000 simulations have been completed through it to date.

Six use cases from six industries

The reviewers highlighted that the results are mature enough for practical industrial and engineering application, and that they reach well beyond automotive:

• Top coating: accurate simulation of the final paint layer, validated against production measurements and delivering a 22% paint reduction in a bumper painting application
• Cargo bicycles: water management and dust management simulation feeding directly into prototype design
• Wind-farm aerodynamics: siting and performance for renewable wind energy. We wrote about it in Where should you build your wind farm? CFD simulation has the answer
• Offshore scour: seabed erosion around offshore structures
• Water management and water resources
• Chemical reactors, covered in Inside a reactor you can’t open: CFD simulation for trickle-bed chemical reactors

Thank you to our partners

COA-CFD was the work of a consortium spanning Europe, North America and Asia. Our thanks go to every partner and every person who carried a piece of it:

• ESS Engineering Software Steyr (Austria): project coordination, solvers, cloud platform, optimisation
• Citkar (Germany): cargo bicycle design and prototyping
• AUDI AG (Germany): top coating knowledge transfer, user needs and testing
• MYB Yazılım Mühendislik (Turkey): sonar and suction atomizer, top coating
• Scientific Solutions Systems (SSS) (Poland): cloud solution
• IONOS (Germany): cloud and HPC infrastructure
• Mazlite (Canada): droplet size measurement and prediction
• Biosimo (Switzerland): chemical reactor use case
• Institute for Advanced Engineering (IAE) (South Korea): Korean consortium lead
• Soda System (South Korea)
• Dtonic (South Korea)
• DOHWA Engineering (South Korea)
• Unison (South Korea)

Our thanks also to Steinbeis for the work on design optimisation and algorithmic differentiation, and to ABB Robotics for the testing and laboratory validation that kept the top coating work anchored in real production conditions.

And our thanks to CELTIC-NEXT and to the national funding agencies, including FFG in Austria, whose support made a consortium of this shape possible.

What comes next

The Spray Washing and Sealing modules are already in commercial use, and the cloud platform continues to run. Following the reviewers’ encouragement, the consortium is preparing a follow-up proposal to exploit the results further and open new application areas.

If you would like to discuss a specific deliverable or use case, the partner contacts are listed at the end of our Deliverables page.