No slip condition post processing

[igergian]

Hi all,

I have a problem with the post processing of my cases regarding the velocity near the walls. I have used both paraview and CFD post (ANSYS) and I get the same result. If I take the velocity-z profile on a cross section of a tube (axis z) although I have a parabolic profile, the velocity is not equal to zero on the wall.

Any ideas?

[Yvan Fournier]

Hello,

This is normal. If you are looking at the values on the boundary mesh, those values are "non-reconstructed" values, which may provide useful information, but are not the exact wall values. If you are looking at the values in cells, the value near the wall is the mean value in the cell, so it is not zero. Add to that the fact that if you are using a RANS or RSM model, you will not have a zero value at the wall.

Best regards,

Yvan

[igergian]

My flow in laminar so I expect the velocity to be zero. Is there any solution to this?

[Yvan Fournier]

If under ParaView, you use the "cell data to point data" filter, your value near the boundary should be close to zero, depending on the way ParaView interpolates this (actually, this should be true interpolating with gradients, but if ParaView averages using neighboring cell values, it won't work).

The solution would be to interpolate from cell values to point values inside the code, using the the contribution from cells using their gradient (plus use boundary condition info near boundaries), then average the resulting value. but this would require advanced user programming.

Best regards,

Yvan