UFR 3-01 Best Practice Advice
Boundary layer interacting with wakes under adverse
pressure gradient - NLR 7301 high lift configuration
Underlying Flow Regime 3-01 © copyright ERCOFTAC 2004
Best Practice Advice
Best Practice Advice for the UFR
The key physics of boundary layer-wake interaction in context of this UFR can be summarized as follows. The wake of the flow over one element interacts with the boundary layer developing on a downstream element. Both the wake and the boundary layer are subject to curvature and adverse pressure gradient effects, which generate higher turbulence intensities and enhance the merging of the wake with the boundary layer developing on the downstream element.
Numerical modelling issues
Grids and grid resolution
- y+ values should be around 1 near solid walls, and 20 to 30 grid points should be placed up to y+=100.
- A 2D structured grid of 36,000 cells seems to be sufficient to predict the lift coefficient and surface Cp values in agreement with experimental data. Finer grids are needed to predict accurately the drag coefficient.
Computational Domain and Boundary Conditions
- Place the far field boundary at least 50 chords from the body, or use a far field circulation correction method.
- Use a scheme with as little numerical dissipation as possible.
- Use the full, compressible, Navier Stokes formulation (no Thin layer approximation, no Euler+boundary layer, and do not use the incompressible formulation).
- Ensure that the transition location on all elements is correctly specified.
- Use algebraic turbulence models when the interest is in the prediction of the aerodynamic coefficients only.
- Use the Spalart-Allmars 1 equation turbulence model or linear 2 equation turbulence models to get the structure of the merging of the wake accurately. Among the linear two-equation models, use the k-ω model since it performs slightly better in terms of comparison with experimental data than other linear 2 equation models. There is no need to use a SST model. No improvements were found using a non-linear two-equation or differential Reynolds stress model.
The advice on turbulence models should be exercised with caution when analysing practical high-lift systems for which the boundary-layers and wakes may be subjected to higher adverse pressure gradients and greater levels of streamline curvature than considered here.
Recommendations for future work
Much progress has been made in non-linear turbulence models since the ECARP project. In particular EARSM models are now becoming increasingly popular, and it is recommended to run the NLR 7301 using one of these turbulence models.
© copyright ERCOFTAC 2004
Contributors: Jan Vos - CFS Engineering SA