Conclusion
This comparison clearly indicates the influence of the turbulence model in a RANS calculation.
Standard models tend to give poor results, whereas elaborate models, e.g. the EASM return fairly accurate results.
The EASM calculation shows all the flow features of the LES.
A closer view reveals the superior quality of the LES.
Compared to the EASM the LES can reproduce more of the vortex dynamics of the flow.
Higher rms values, correct drag and Strouhal number and are an indicator for the satisfactory overall agreement with the experimental data.
The capability of the EASM to capture the important flow features correctly makes them a useful tool in industrial design,
particulary with respect to the computing time of only 5 % of the LES.
Unfortunately the quality of the LES cannot be recovered.
|
|
|
|
|
Fig. 8 u-velocity isolines on a near-wall plane at transition strip
|
|
Fig. 9 w-velocity isosurfaces at trailing edge
|
In a further study of an airfoil (NACA 4412, alpha = 12°, Re = 1.6*106) with only weak unsteadiness (RANS steady)
the LES fails to predict the separation at the trailing edge, as depicted in the pressure distribution in fig. 10.
In contrast to this, the RANS agrees with the experimental data.
Fig. 10 pressure distribution
Webmaster
Last modified: Fri Nov 17 13:03:31 CET 2000
