next up previous
Next: Performance Issues Up: High-Lift Flows Previous: Three-Element Aerofoil

Wing-Body Configuration at High Angle-of-Attack

Figure 13: WBC: Surface pressure distribution and streamlines
\includegraphics[width=120mm]{wbc_surf.eps}

This testcase is a wing-body configuration (WBC) in the subsonic regime at a high angle-of-attack, where the flow on the upper surface of the wing is partially separated, see Fig. 13. The inflow conditions are set to $ Ma=0.2$, $ Re=2.7\cdot10^{6}$ and $ \alpha=10^{\circ}$. This case is particularly interesting as a streakline picture of the wing is available. Computations are performed with both FLOWer and TAU. The structured grid used consists of 48 equal-sized blocks with $ 25\cdot45\cdot45$ points and 33,882 surface points, yielding approximately 2.4 million points. The hybrid grid also contains about 2.4 million points. Again, for both grids, which were supplied by Airbus Germany, a $ y^{+}\approx1$ is guaranteed over most of the wing surface. Since no transition locations are given, all computations are performed in a fully turbulent manner. Results presented here include Wilcox and LLR $ k$-$ \omega $ with FLOWer as well as RQEVM + Wilcox $ k$-$ \omega $ with TAU.

Figure 14: WBC: streaklines on the wing, experimental data
\includegraphics[width=63mm]{wbc_exp.eps}

Figure 15: WBC: Surface pressure distribution and streaklines on the wing for various turbulence models
\includegraphics[width=170mm]{wbc_cmp.eps}

Figs. 14 and 15 compare the computed surface pressure distribution and streamlines with an experimental oil flow picture. The measurements show a pocket of separated flow in the outer third of the wing. Both LLR and RQEVM predict the extent of the separation quite correctly, whereas Wilcox $ k$-$ \omega $ only displays a tiny separation near the trailing edge. However, not even the more advanced models deliver the correct flow topology, thus, a quantitative agreement cannot be expected, see Fig. 16. As the grid has to be considered as comparatively coarse for such computations and the effects of transition are neglected, no final conclusion on the accuracy of the turbulence models can be drawn here. Furthermore, it becomes evident that correct predictions are harder to achieve in high-lift flows as compared to the transonic regime, which mainly has to be attributed to the more complex transition patterns which can hardly be prescribed (as seen in the oil flow picture, Fig. 14) and the fact that larger laminar regions are encountered. It should be noted, though, that even here, the advanced models' predictions do outperform the standard $ k-\omega$ approach.

Figure 16: WBC: Pressure distribution in selected wing sections
\includegraphics[width=170mm]{wbc_cp.eps}


next up previous
Next: Performance Issues Up: High-Lift Flows Previous: Three-Element Aerofoil
Martin Franke 2003-10-22