The development of a boundary layer transition model for helicopter rotor CFD
A novel transition model has been developed for use in CFD simulations of helicopter rotor aerodynamics. The model includes significantly improved physical modelling of the transition processes occurring in the steady and unsteady flows found on helicopter rotors. The model has been coupled with the k-co and k-co SST two equation turbulence models using a novel adaptation of the technique developed by Wilcox for the low Reynolds number k-oa model. The method has been employed to calculate transitional flows occurring in three key ow regimes found in helicopter aerodynamics; that around steady and unsteady aerofoils and that around a hovering helicopter rotor. The performance of the k-co and the k-w SST turbulence models have been investigated for transitional flow simulations and the k-w SST shown to provide substantial improvements for transitional flows containing separations. Dramatic improvements in the computed pressure and skin friction distributions for several aerofoil flows have been observed over those computed using a conventional fully turbulent simulation. Corresponding improvements are observed in the computed lift and drag polars and transition on set is well predicted for both low and high Reynolds number flows. A novel structured/unstructured a priori adapted grid generation strategy has been developed for hovering rotor flows that provides improved rotor solutions for transitional flow analysis. The method offers vast improvements in the preservation of vorticity in the solution at greatly reduced computational expense. Tip vortices have been maintained to a Wake age of 1170 degrees with just 2 million cells per blade. The transition model has then been applied to the high quality rotor solutions and good agreement obtained between computed and experimental results, highlighting that three-dimensional effects have a relatively small effect on hovering rotor transition in-board of the blade tip. I addition, the first known verification of a Navier-Stokes rotor code against the Fogarty semi-analytical rotating at plate case was presented and excellent agreement obtained.