PlotOverLine: boundary value ?
 

Hi all,

There is a behavior I find really annoying in some cases: it is the fact that after a plotOverLine boundary values (on patches) ​​are not routinely used.

For example it's not possible to have on the curve of a plotOverLine the value of zero for the velocity at a stationnary wall .

Do you watch also this behavior?
How can we fix it ?

Thanks
Francois

Greetings Francois,

I've tested just now with ParaView 3.12.0 and 4.1.0, on OpenFOAM 2.2.x and 2.3.x respectively, using the tutorial "incompressible/icoFoam/cavity" as reference. And used both the internal reader (file extension ".foam") and the official reader (".OpenFOAM").
At first, I didn't have any problem what so ever. The values at the boundaries were represented without any problems.
But them I remembered to move ever so slightly one of the ends of the sampling line and pufft... there it went out the door, the value at the boundary.

Therefore, there are 2 issues at work here:

1. The end points must be exactly or nearly exactly on top of the boundary in question. Or if not one of the end points, then one of the sampling points in between.
2. Sampling is done through interpolation, based on the point data. This means that if your sampling point is not exactly in the same plane as sampling points that make up the face of the boundary, then the sampling data isn't the one your looking for.
   • Worse case scenario: the distances between points is much closer to points that are inside the mesh, instead of on the surface.

If neither one of these apply to your case, then I'll need to see some images to assess what might be going on!

Best regards,
Bruno

Hi Bruno,

Thank you very much for your kind and helpful answer. :)

What you described is exactly what I met and when playing a little on the positions of the ends point I can get the values at the boundaries.

The most sensitive cases beings when the sampling point is not exactly in the same plane as sampling points that make up the face of the boundary.

So apparently no systematic solution except to play on the positions of the end points especially if the surface of the patches are curved surfaces as it is the case for me.

Any way thanks you very very much for your advices.

Regards,
François

Hi François,

I forgot the other day to mention that there is another possibility, explained here: http://www.cfd-online.com/Forums/ope...tml#post476212 post #2

Best regards,
Bruno

Thanks for the advice Bruno :)
Will try this method to see if I get better results

Best regards
François