Hallo Everybody

I am trying to model a geoemtry where my flow starts out as a fully turbulent velocity profile and then as the geometry chnages shape the flow undergoes transition to laminar. I am trying to use a low k-e model. However, it seems to me that transition takes place way to early. Is there any explanation for this?

Regards

Bo

Yes, there is a short explanation : As far as I know, there is no tubulence model that can predict transition from laminar to turbulent or from turbulent from laminar.

In practice, when we are familiar with the problem, we fix the transition point....

Hallo Seb

Very nice of you to give me an answer. I would very much like to get a little bit more information on fixing the transition - is it possible to do in a commercial CFD code??

Regards

Bo

http://www.cfd-online.com/Forum/star...0.cgi?read=253

I do not know if any commercial code is offering such a model right now. But I don't think so. Anyway, a lot of work is ongoing on that topic in the research community. So a literature survey should bring up a lot of material

I cannot comment on commercial packages today but the k-e turbulence model was developed over 30 years ago within a project on laminarisation in strongly accelerated boundary layers.

It is possible to fix he transition point with some commercial codes (such as the flow solver sold by the compagny I am working for). Sorry, I don't have much information about the major software vendors (Fluent, Star-CD, ...).

I think there is much more information on fixing the transition point with the Spalart-Almaras turb than with the other well-known models. I suggest doing some research in the litterature.