[luckyluke]

What is the difference? If I model the RT instability in a box, what conditions should be set to left,right,top and bottom boundaries for gamma?

Thanks.

[luckyluke]

Who give me a few words? ^_^

[henry]

zeroGradient sets the boundary value to the near-wall cell value whereas a symmetryPlane is a symmetry-plane which is the same thing for scalars but not for vectors or tensors.

[RygeltheXVI]

Hail,

I'm trying to work out description of what he symmetryPlane boundary condition actually is. I haven't had any luck with either the user nor programmers guide.

This thread gives the most mathematical description that I can find online but is still incomplete:

Originally Posted by henry

zeroGradient sets the boundary value to the near-wall cell value whereas a symmetryPlane is a symmetry-plane which is the same thing for scalars but not for vectors or tensors.

What is it for vectors? (If it was just normal component zero, then it would be the same as a slip boundary condition).

I have tried but have no luck interpreting the source code: and even then I don't know which one to look at (I don't know which of /src/finiteVolume/fields/fvPatchFields/basic/basicSymmetry or /src/finiteVolume/fields/fvPatchFields/constraint/symmetry [if either, it could be another one] corresponds to symmertyPlane).

If anyone knows which, or preferably a description of what the symmetryPlane boundary condition does, it would be very much appreciated.

Many Thanks
Rygel

[guitarbren]

I would also like to know the difference if there is anyone out there with an idea?
Thanks

[MOHAMMAD67]

Dear Foamers
Is there anyone here help us to have more information about that? I obserevd some differences between the results of the model with and without symmetryPlane?

[aliqasemi]

too late but, I had a quick look, apparently the boundary field is assigned in function: Foam::basicSymmetryFvPatchField<Type>::evaluate. It makes use of transform boundary condition, and I didn't decode it thoroughly. However, my speculation is as follows:

For vector fields, the value of the field on the patches with type "symmetryPlane" are calculated by projecting the patchInternalField to the patch. They are calculated as the average of the patchInternalField values and their mirrors in the patch faces.

[duongquaphim]

Hi,

For zeroGradient, both scalar and vector field are set as the value of the cell near the patch by setting surface normal gradient to 0: grad(f) = (f_P - f_N)/delta = 0 (programmerGuide page 40).

With symmetryPlane, for a scalar, the computation is similar to zeroGradient. For a vector, to make it simple, all components parallel to the patch is mirrored while the normal component is set to zero. Looking at the code of OpenFOAM at src/finiteVolume/fields/fvPatchFields/basic/basicSymmetry, I think they use transformation matrix to compute the new value on the BC. The method can be find in http://en.wikipedia.org/wiki/Householder_transformation

Regards,

Duong

[sam.dj]

Originally Posted by henry

zeroGradient sets the boundary value to the near-wall cell value whereas a symmetryPlane is a symmetry-plane which is the same thing for scalars but not for vectors or tensors.

Hi,
in tutorials,lagrangian,DPMFoam,goldschmidt, this BC is used for p and u, so we can use it for vectors.