[petsc-dev] 6-node wedge cells in DMPlex

[Matthew Knepley]

On Mon, Apr 14, 2014 at 8:44 PM, Adrian Croucher
<a.croucher at auckland.ac.nz>wrote:

> hi
>
> I just discovered that DMPlexInterpolate() complains if I try to
> interpolate a mesh containing 6-node wedge cells in 3-D:
>
> [0]PETSC ERROR: Argument out of range
> [0]PETSC ERROR: Cone size 6 not supported for dimension 3
>
> Looking in the code, it appears that DMPlexGetRawFaces_Internal() only
> supports cells with 3, 4 or 8 nodes in 3-D.
>
> At first I thought it would be simple enough to add another case in there
> for cone size 6. But then I noticed that DMPlexInterpolateFaces_Internal()
> will throw an error for any type of cell with faces that don't all have the
> same number of nodes - and there's a comment in there saying "I see no way
> to do this if we admit faces of different shapes".
>
> I don't know the routine well enough to understand what the difficulty is,
> but are there any plans to generalise DMPlexInterpolate() so that it can
> handle commonly-used cell types (like 6-node wedge) that have faces with
> different numbers of nodes? We use wedge cells quite a bit, so it would
> limit the usefulness of DMPlex if it can't support them. Unless there is
> some kind of workaround?
>

There is certainly a workaround. You give me the interpolated mesh. The
only thing this code does is
allow you to input just cells and vertices. Since you are using these
elements now, you know the faces,
so just use something like DMPlexCreateFromDAG() to input the fully
interpolated mesh.

The problem here is that if I allow faces of different shapes, I have to
keep track of the shape of each
face when I make it. This is possible,  and I have added it to the TODO
list, but I think it will be a while
before I get to it. If you want to try doing it, I would help.

  Thanks,

     Matt


> Cheers, Adrian
>
> --
> Dr Adrian Croucher
> Department of Engineering Science
> University of Auckland
> New Zealand
> tel 64-9-373-7599 ext 84611
>
>


-- 
What most experimenters take for granted before they begin their
experiments is infinitely more interesting than any results to which their
experiments lead.
-- Norbert Wiener
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20140415/90eb9ae7/attachment.html>