[CIG-LONG] Gale in 3D

Walter Landry walter at geodynamics.org
Mon May 14 13:10:08 PDT 2012


Rainer Nerlich <rainer at simula.no> wrote:
> Hi Walter,
> 
> yes, I get the same Triquadratic error message for extension3D.json. The
> results for yielding.json or thermal.json in 2D look ok - it took me so
> long to figure that out because first I used paraview and the results were
> always overlain by some weird looking one colored layer, which I was able
> to filter out, though. However, I tested Visit in addition now and
> everything looks automatically normal and fine.
> 
> I was also trying to figure out where the changes between my old Gale
> installation and the new repository version have been made. I found several
> changes and narrowed it down to some potential candidates that could cause
> my 3D issue, but of course I am not sure at all if those changes truly have
> anything to do with it (the first part is the new version, the second the
> old one):
> 
> ./StgDomain/Mesh/src/Mesh_Algorithms.cxx
> <       for( unsigned d_i = 0; d_i < nDims; d_i++ ) {
> <           if( point[d_i] < minCrd[d_i] - (maxCrd[d_i]-minCrd[d_i])*0.1
> <               || point[d_i] > maxCrd[d_i] + (maxCrd[d_i]-minCrd[d_i])*0.1
> )
> ---
>>       for( d_i = 0; d_i < nDims; d_i++ ) {
>>               if( point[d_i] < minCrd[d_i] || point[d_i] > maxCrd[d_i] )

This is the only change that could possible have anything to do with
it.  The change should make it more lenient, so you would not get
these errors.  You can try reverting this change and see if it makes a
difference.

Cheers,
Walter Landry


More information about the CIG-LONG mailing list