REDUCED-ORDER MODELLING OF PARAMETERIZED TRANSIENT FLOWS IN CLOSED-LOOP SYSTEMS

In this paper, two Galerkin projection based reduced basis approaches are investigated for the reduced-order modeling of parameterized incompressible Navier-Stokes equations laminar transient flows. The first approach solves only momentum equation with additional, physics-based approximations dynamics pressure field. On other hand, second both and continuity equations. bases velocity fields generated using method snapshots combined Proper Orthogonal Decomposition (POD) data compression. To remedy stability issues two-equation model, is enriched supremizer functions. Both have been implemented in GeN-Foam, an OpenFOAM-based multi-physics solver. A numerical example presented a two-dimensional axisymmetric model Molten Salt Fast Reactor (MSFR) dynamic viscosity as uncertain parameter. results indicate that one-equation slightly more accurate terms velocity, while built same amount modes far pressure. speed-up factors models 3060 2410 model.