One of the challenges in simulating aerodynamic flows is the accurate prediction of onset and progression of separation. In earlier STO research, it was found that various predictions for three-dimensional airfoils at angle of attack showed different pitch moments due to differences in the predicted location of separation areas. Therefore, a diamond shaped wing was developed within the NATO STO AVT-183 research group, to isolate the blunt leading edge separation and understand the underlying physical mechanisms. In the present paper, two viscous-flow solvers dedicated to hydrodynamic applications are used to predict the flow around the diamond wing. Both codes solve the incompressible flow using unstructured grids and finite volume discretisation. The results comprise different grid set-ups, with/without peniche and different turbulence models ranging from isotropic or anisotropic statistical turbulence closures to hybrid LES turbulence models. Moreover, the role played by the discretisation error is assessed by using anisotropic automatic grid refinement procedures based on flow-related criteria. A detailed discussion is made, highlighting the similarities and differences of the results, the respective influence of modeling and discretisation errors and showing the potential of CFD tools to predict the type of flow under consideration.
Team leader CFD development / Senior Researcher
manoeuvring and nautical studiescfd developmentcfd/simulation/desk studiestime-domain simulationsmanoeuvringinfrastructuredefencepassengers and yachtingtransport and shippingresearch and developmentsimulationsresearch