CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Bugs (http://www.cfd-online.com/Forums/openfoam-bugs/)
-   -   sample interpolation error in inflation layer (http://www.cfd-online.com/Forums/openfoam-bugs/66468-sample-interpolation-error-inflation-layer.html)

eelcovv July 16, 2009 06:05

sample interpolation error in inflation layer
 
Dear OpenFoamers,

I am trying to use the 'sample' utility in order to sample a data plane from my data set as defined in a sampleDict file. Eventually I would like to do this during a run using libsamling.so function call in the controlDict.

Unfortunately, there seems to be an interpolation error in the sampling routing when the plane intersects prisms cells such are found in the inflation layer in my grid (tetrahedra cells are no problem). If I run 'sample', the following error message is generated:

--> FOAM Warning :
From function Foam::cuttingPlane::walkCell
in file cuttingPlane/cuttingPlane.C at line 205
Did not find closed walk along surface of cell 1742616 starting from edge 1656820 in 1001 iterations.
Collected cutPoints so far:
1001

and further:

--> FOAM Warning :
From function Foam::cuttingPlane::walkCell
in file cuttingPlane/cuttingPlane.C at line 189
Did not find closed walk along surface of cell 1877596 starting from edge 7 in 0 iterations.

The resulting interpolated data plane shows erroneous sampled values in all cells in the inflation layers that consists of prisms cells.

If have already tried the swhich on/off the interpolation (true/false) and triangulate (true/false) and also all different interpolation schemes (cell/cellPoint/cellpointFace) in the samleDict file, but this gives the same results.

Is there anybody who has found the same problem and hopefully also has a solution for this?

Any hints appreciated!

Many thanks,

With kind regards,

Eelco

eelcovv July 16, 2009 07:02

Found solution
 
Their all,

I found that slightly shifting the planesolve the problem. Probably the interpolation is sensitive to the location where it intersects the cells. Thanks anyway.

Eelco

mattijs July 28, 2009 17:20

in 1.6 you can try the cuttingPlane type in the sample utility. It does a tet decomposition so should work on non-convex cells.


All times are GMT -4. The time now is 04:11.