[MOAB-dev] Polygon->polyhedra adjacencies bug?

[Timothy Tautges]

Ok, this is a general bug with polyhedra-polygon adjacencies. See the
attached program; if the program doesn't find polyhedra in the mesh, it
generates them using the DualTool. I ran this on the
MeshFiles/unittest/tet_brick.vtk file and got the following output:

Mesh has 274 polyhedra and 1513 faces.
There were 1513 faces that failed cell adjacency test in this mesh.

I'm hoping somebody there can put this into the bug reporting system you
use, I forget where that is (prob bitbucket, but I'm being lazy atm).

Thanks.

- tim


On Tue, Aug 25, 2020 at 4:15 PM Vijay S. Mahadevan <vijay.m at gmail.com>
wrote:

> Hi Tim,
>
> I think you may be referring to the bug that Johannes Probst raised here
> [1].
>
> I am not sure whether this is a logical or a design issue since the
> behavior charted out appears to be what I would have expected, but
> clearly not what Johannes wanted.
>
> Did you have a different discussion in mind ?
>
> Vijay
>
> [1]
> https://bitbucket.org/fathomteam/moab/issues/117/incorrect-adjacencies-in-polyhedral-mesh
>
> On Tue, Aug 25, 2020 at 5:02 PM Timothy Tautges <ttautges at simscale.com>
> wrote:
> >
> > Hi all, I've run across a bug, where a polygon doesn't know about its
> > adjacent polyhedra. But, something rings a bell about that being a known
> > bug that got discussed during the last few months. Does that ring a bell
> > with anybody here?
> >
> > Thanks.
> >
> > - tim
>