hi folks, just wondering if anyone here has experience with modelling transition using the 2 eqn sst model. i am trying to simulate the laminar turbulent transition in a mach 7.5 flow over a flat plate. my residuals decrease and oscillate. the omega and the 2 transitional closure equations are especially bad with residuals oscillating around 1e-2. that is my first problem. second problem is when i view the intermediate results, transition occurs very far downstream compared to the experimentally observed transition point. a strange thing is that this position seem to move further downstream with increasing number of iterations. i am using steady state simulation, 1% free stream turbulence degree. i have checked the converged Yplus and it is a little below 1 at the wall, which i guess should not be a problem. would be grateful for some insight. regards, ben.

You may need to solve it as a transient problem. There was a monster thread about the transition model some time back. Search the forum for "transition wing" etc.

Well, I have no experience with supersonic stuff since I work in water, but this is what I've learned. YMMW.

- It seems not to like structured meshes at all.

- It seems to like y+ < 1.0 even at max shear.

- It seems to like mesh expansion lower than 1.2 and lots of layers.

- It seems to like a lot finer mesh than what is suggested in the manual (manual says >70 streamwise nodes are fine, I usually the double.

- It seems to work best with Intensity/Length Scale as the inlet turbulence spec, but it is still very sensitive to the applied values.

- It seems to do a better job at higher angles of attack.

/C

Oh, and one more thing: It's sensitive to mesh resolution and smoothness at the front of the foil. Like how the fluid hits the leading edge will influence how far down transition occurs.