Integrating CFD and Experiment in Aerodynamics - CFD4Aircraft
Integrating CFD and Experiment in Aerodynamics - CFD4Aircraft
Integrating CFD and Experiment in Aerodynamics - CFD4Aircraft
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
3<br />
2. Surface flow visualisation. Oil flow visualisation gives an <strong>in</strong>dication of surface streaml<strong>in</strong>es, but<br />
only <strong>in</strong> a time averaged sense. Tufts also give an <strong>in</strong>dication of surface streaml<strong>in</strong>es <strong>and</strong> can reveal<br />
flow separation <strong>and</strong> reattachment, but are limited with the response time <strong>in</strong> unsteady flows <strong>and</strong><br />
can also be <strong>in</strong>trusive.<br />
3. Off surface flow visualisation (smoke/dye). This can provide useful <strong>in</strong>formation on shear layer<br />
structures <strong>and</strong> vortex breakdown, but extra care should be taken <strong>in</strong> <strong>in</strong>terpret<strong>in</strong>g the streakl<strong>in</strong>e<br />
patterns <strong>in</strong> unsteady flows.<br />
4. Multi-hole velocity probes. These can measure three-components of mean velocity, but are<br />
<strong>in</strong>trusive <strong>and</strong> can cause premature breakdown.<br />
5. Hot-wire anemometry. This can provide unsteady velocity components but can be <strong>in</strong>trusive.<br />
6. LDV <strong>and</strong> PIV. These are non-<strong>in</strong>trusive po<strong>in</strong>t <strong>and</strong> field measurements respectively of velocity<br />
vectors <strong>in</strong> a plane. Seed<strong>in</strong>g of vortical flow near the axis becomes problematic with <strong>in</strong>creas<strong>in</strong>g<br />
speed <strong>in</strong> air flows.<br />
2.2 <strong>CFD</strong> Techniques<br />
It has been well documented that <strong>CFD</strong> has developed at a rapid pace over the past 30 years. With<br />
developments <strong>in</strong> algorithms <strong>and</strong> computers it is possible to simulate complex flows on real aircraft<br />
us<strong>in</strong>g cheap computers. A recent NATO technical organisation (RTO) work<strong>in</strong>g group has exam<strong>in</strong>ed<br />
the predictive capability for vortical flows on generic delta w<strong>in</strong>g configurations (AVT 80) [2].<br />
1. Euler simulations can predict vortex breakdown <strong>and</strong> vortical <strong>in</strong>teractions when a sharp lead<strong>in</strong>g<br />
edge is used, fix<strong>in</strong>g the separation po<strong>in</strong>t. No secondary separation can be predicted s<strong>in</strong>ce this is<br />
due to boundary layer separation hav<strong>in</strong>g the effect of shift<strong>in</strong>g the primary vortex closer to the<br />
w<strong>in</strong>g lead<strong>in</strong>g edge. In addition the strength of the lead<strong>in</strong>g edge vortex is strongly dependent on<br />
the grid used. However, for sharp lead<strong>in</strong>g edges this level of modell<strong>in</strong>g is useful for evaluat<strong>in</strong>g<br />
qualitative behaviour at a low cost.<br />
2. Unsteady RANS simulations can give good prediction for the secondary separation although<br />
the prediction of primary separation <strong>and</strong> vortex formation for rounded lead<strong>in</strong>g edge w<strong>in</strong>gs has