I am currently preforming a CFD validation on a 6:1 Spheroid with incompressible viscous flow. I am attempting to illustrate vortex sheading using the realizable k-epsilon model. However I have found that the standard setting and those settings suggested by my colleagues do not provide a stable solution. Fluent returns the following error message "turblent viscousity limited to 1e5 in X-number of cells". My question is, what settings have worked sucessfully in the past for this type of flow, and what type of near wall modelling have bee used? Currently enhanced wall treatment, with pressure gradient effects are being used. With thanks Stephen Mar

I have used all the models of k-e, with the default values, and different wall functions, and the final results were so acceptable, even more than what was supposed.

The default limit of turbulent vicosity ratio may be exceeded in cases of strong turbulence.

Hence it is not a fixed limit which must be kept by all means.

Why do you think that it is the settings that is the problem ?

As Volker says, it is not given that the ratio should be below 1e5.

A poor solution can be obtained due to many operator-errors and not just because of ill-chosen model constants.