unsteady RANS-flow around cylinder

omierka

Hi there,

our finite element based flow solver was extended with high-reynolds k-epsilon model which was later modified with Chien's low-reynlods model. It was also succesfully tested on a benchmark problem of channel flow and backward facing step. My aim is to simulate flow around cylinder at Re=10000. I know, that there is a problem with Chien's model to predict separation and reattachment, for this reason the friction velocity has an alternative scale, which is widely used in CFD (u_tau=srt{sqrt(c_mu)*k}), but the behaviour of the flow is still not what the experiments predict, because the solution is oscillation free (i.e. no vortex shedding is present). Do You know, if Chien's model is usable or not in this case, and if not what kind of model should I use? In fact I do not think, that I have problems with the resolution of the mesh, because y+ varies around the cylinder in a range of 0.1-0.4.

thanks, otto