[f0208secretx]

Hello,

I'm running 2D LES simulation for classical flow over cylinder problem. Comparing the result with the literature, I found the vorticity contour was quite different and I couldn't help but start wondering if the problem was caused by the algorithm of Fluent in 2D LES.

The reference of the calculation was M. Tutar, Computational modelling of flow around a circular cylinder in sub-critical flow regime with various turbulence models, 2001. The set up was almost identical to the paper. Except for my mesh point was about ten folds than his. (Thanks to the computer technology, I got even lesser computational time)

The results was quite different and confusing, as the figure I attached. (vorticity.jpg was from me, while vorticity2.jpg was from Tutar) Although the separation angle was about the same, much stronger activities were in the near wake region in my case (as the presence of many small vortices) and the Cd was nonsensed 2.0. Since there aren't much parameters to set up in the LES models, I would like to know if it is good for Fluent making prediction in 2D cases? Is there anyone successfully render a sensible results for the similar problems?

For your reference, the operating condition:
Reynolds number 1.4E+5
Structured mesh about 340k points.
The LES model was activated through command text (rpsetvar 'les-2d? #t) and the SGS model is Smargorinsky-Lilly
CPL is about 1