[Rozie123]

I am trying to simulate a wall-bounded compressible channel flow (M=2) using sonicFoam. I set synthetic turbulent generator at the inlet, however my result does not show turbulent velocity near the wall and its rather laminar. my inlet boundary condition is turbulentInlet with 10% intensity and the outlet is inletOutlet. The result is like following: Anyone has a clue what could be wrong with my simulation?

[shock77]

Hi,


would do you mean by turbulent velocites? The velocity at the wall is zero since you apply a noSlip boundary condition. It looks like what I would expect.

[Rozie123]

Originally Posted by shock77

Hi,


would do you mean by turbulent velocites? The velocity at the wall is zero since you apply a noSlip boundary condition. It looks like what I would expect.

Thank you, I mean the fluctuations at the wall is zero. The mean velocity and the instantaneous velocities are about the same. There shouldn't be eddies near the wall since I am imposing turbulence at the inlet?

In the literature, the DNS of compressible channel flow has many fluctuations near the wall. I am deploying WM-LES though.

[shock77]

Didnt know you are trying to do an LES. I dont know a WM-LES Model. As far as I know only Smagorinsky and WALE are implemented in OpenFOAM. And if you are using a LES model with a sub grid scale viscosity, you have to make sure, that your mesh is fine enaugh to resolve the big eddies.


Also, if there is no boundary seperation, I dont know where turbulence should come from. Has the pipe some sort of roughness? How does your synthetic turbulence work? Is it like a pseudo random number generator?

[Rozie123]

Originally Posted by shock77

Didnt know you are trying to do an LES. I dont know a WM-LES Model. As far as I know only Smagorinsky and WALE are implemented in OpenFOAM. And if you are using a LES model with a sub grid scale viscosity, you have to make sure, that your mesh is fine enaugh to resolve the big eddies.


Also, if there is no boundary seperation, I dont know where turbulence should come from. Has the pipe some sort of roughness? How does your synthetic turbulence work? Is it like a pseudo random number generator?

Yes, OpenFoAM has two or three WL-LES models.

The mesh should be good for at least some larger eddies. No, I have plane walls and just generation artificial turbulence at the inlet which is similar to a tutorial case in incompressible\pimpleFoam\channel. replicating the tutorial case for incompressible case, it shows the eddies. However for my compressible case I only get the mean velocity. Here is what the literature shows for compressible channel flow, the near wall is highly turbulent.

But I do not know how realistic would be having turbulent region near wall like this.

[stiffler_014]

Hey which research Papier did you refer for the compressible supersonic for the channel at what Reynolds number ?