We work on the practical problems that appear when research-grade simulation has to become dependable engineering output.

CFD consulting and simulation studies

Consulting and simulation work across axial and radial turbomachinery, hydraulic turbines, external aerodynamics, aeroacoustics and hydroacoustics. We work with RANS where it is the right tool, and with high-order LES when the physics demands it.

Uncertainty quantification — KAOS

Quantifying how uncertain inputs — Reynolds number, inlet turbulence, incidence, turbulence- and transition-model coefficients — propagate to the results, so a CFD study comes with confidence bounds and a sensitivity ranking rather than a single unqualified number. This is our KAOS capability: non-intrusive polynomial chaos, global sensitivity and Bayesian calibration, wrapped around your solver.

Simulation workflow automation

Turning manual, error-prone CFD, post-processing and reporting steps into repeatable pipelines — so the same case runs the same way every time, with less effort and fewer mistakes.

Scientific software development

Building robust, maintainable tools from prototypes, research codes and published numerical methods, with attention to testing, documentation and reproducibility.

HPC and GPU computing

Performance-oriented implementation — parallelisation, GPU acceleration and profiling — for simulations that are too large or too slow to run comfortably on a single machine.

Research-code industrialization

Taking fragile research prototypes and making them stable enough for day-to-day engineering use: clearer structure, regression tests and a maintainable codebase.

Validation and post-processing pipelines

Reproducible workflows for verifying solvers against canonical and industrial cases, and for extracting consistent quantities from large result sets.