Hello,

I am trying to model the flow through various supersonic nozzles with the jet flowing over a flat plate. My main area of interest is the point of separation on the plate and also the shear stress distribution along the plate.

At the moment I am using the Realizable k-epsilon model, but is there a better model to use?

Thanks for your help.

Chris Rand

(1)."is there a better model to use?" I guess, there is always a possibility. But, you need to define the better model first. I mean, how do you measure the model to see whether it is better? I am sure that, if you try different models, you will get different results.

(2). So, why not pick another turbulence model, run the calculations, and let us know the results. If you have flow separation, then it is a good idea to use a low Reynolds number model. (vs wall function)

Thanks I will try that.

Chris Rand

Why do you state that a low Reynolds number should be used if flow is found to separate? And why the wall function?

Wall-functions are only valid in "well-behaved" boundary layers, ie the boundary layers should be attached, fully-turbulent and have a small streamwise pressure-gradient. If you have a complex flow with separations etc. the wall-functions give incorrect predictions and you should use a low-Re model.