Computation of Leishman-Beddoes model parameters using unsteady RANS simulations

The determination of the Leishman-Beddoes (LB) model parameter values from experimental measurements and Computational Fluid Dynamic (CFD) simulations is presented. Two-dimensional unsteady RANS simulations are carried out for experimental test cases of two airfoils oscillating in pitch in the wind tunnel at Reynolds numbers of the order of Re=1.8e4. The RANS results are first compared directly to the experimental measurements and it is shown that the simulations cannot represent some important aspects of the physics of the phenomenon, such as the effect of a laminar separation bubble occurring at low angles of attack and the effect of the resulting leading edge vortex once it has started travelling over the surface of the airfoil. The flowfields computed from the CFD simulations are then used to estimate the values of three parameters for the Leishman-Beddoes model. The aim is to explore the possibility of using the LB model as a reduced order model for CFD simulations. It is shown that the inaccuracies of the CFD simulations lead to inaccurate parameter values, such as an overestimation of the leading edge vortex shedding time. Nevertheless, the resulting LB model can smooth the oscillations in the post-stall load responses predicted by the CFD. It is concluded that higher-fidelity simulations are necessary, involving a boundary transition model or even Large Eddy Simulation schemes.