Accelerate CFD

This section collects research posts related to CFD acceleration, adaptive prediction, surrogate modelling, reduced-order modelling, and data-driven workflows for high-fidelity simulations.

Speed-up CFD using adaptive hybrid ROMs: unsteady solutions

Category: Adaptive Methodologies

A fast adaptive prediction method that uses a lightweight AI model, calls CFD solver and retrain the model again when the prediction becomes unreliable.

ROMIA: A methodology to accelerate RANS

Category: Adaptive Methodologies

A divergence-aware adaptive CFD-surrogate framework that alternates between fast POD-DL forecasting and targeted OpenFOAM recalls for robust long-horizon prediction of unsteady flows.