[StefanW]

Dear Foamers

I try to simulate a turbulent twophase pipe-flow (water/air) to get the upcomming wall Wall Shear Stresses and i got really in trouble to get the things sorted.
I use OF 2.01 and the solver compressibleInterFoam with a k-w-SSt model.

As discussed in many threads the OF utility wallShearStress dosn't work for multiphase according some problems with transportProperties.

So I look for a way to compute them otherwise.

I made a solution for multiphase like the following which works well and recognizes the two phases
tau=twoPhaseProperties.mu()*symm(fvc::grad(U))

I did some testings with different mesh-sizes with the mentioned function
and the values of Tau got nearly double if I half the size of the elements.
But the behavior of the "main" flow in the middle of the pipe is nearly the same (vel,p) in both cases, so I would guess the mesh size is fine enough in both cases.

I also found this Thread
http://www.cfd-online.com/Forums/ope...interfoam.html
where is mentioned that we have to use different kinds of calculations if the flow is laminar or turbulent.....

I got a turbulent flow therefore it is suggested to use and not the one above:
wallShear = (-mesh.Sf().boundaryField()[patchID]/(mag(mesh.Sf().boundaryField()[patchID])) ) & (turbulence->R()().boundaryField()[patchID])
Could somebody please explane the single parts of the equation?
Is there mu already attached? So I have to multiply it with mu?


Additionally I got a generall question according the influence of the mesh-size near the wall on the calculated values of the shear stresses.

So far I wasn't able to check the y+ values (yPlusRas) cause the utility doesn't work in multiphase as well. (I use wall functions but i cant figure out how to extract y+; help here would also be appreciated; I found the variable "ypl" in the wallfunction of omega
https://github.com/OpenFOAM/OpenFOAM...hScalarField.C
but i'm not sure if that would help and how to extract it)

So should my function work correctly just in the viscous sublayer when y+ < 5 so we see laminar behavior between the wall and the first cell? And if I enter the <5 the further mesh size doesn't matter and the shear stresses will be the same values???

Would the function for turbulent cases solve this problem so I can use meshes with
y+>30 (logarithmic region) and get still similar results? (Kind of the same principle the wall functions are working with the blending factor of Mentner k-w-sst)

So basically that leads to the question what setup is needed so OF calculates the normal gradient du/dy correctly. And how OF interpolates the velocity between the wall(0) and the first cell.

I hope anybody can follow my explanations and can give me some hints.

Thanks a lot in advance

Stefan

[StefanW]

nobody got any ideas according this??