Multi-fidelity Surrogate Modelling of Wall Mounted Cubes

Abstract This paper focuses on the application of multi-fidelity surrogate modelling to characteristics a flow as it changes with parameter. provides insight into potential combining approaches varying fidelities computational fluid dynamics methods parameter space exploration. A limited number trusted high-fidelity large eddy simulation data points, in combination an extended study using lower-fidelity Reynolds averaged Navier–Stokes is used input for model. Multi-fidelity models are implemented bridge low-fidelity and providing improved model over single fidelity alone. The around tandem wall mounted cubes at inlet yaw angle aerodynamic test case this methodology. Results presented show that significant improvement prediction global properties. methodology then combine multiple local features build up fuller descriptions angles not included training results both one-dimensional line plots range locations along center two-dimensional slices velocity field. produces away from high match closely results.