[CIG-SHORT] Cubit tet mesh not being read into pylith

[Oliver Boyd]

I wanted to locally refine around the fault and was having trouble doing
this in Cubit with hexes and a paving scheme. I didn't have trouble with
tets, hence the initial post with tets. Now, with Steve's help, I've settled
on using a mapping scheme and am able to locally refine hexes without a
problem. Hence the current posts with hexes.

The mesh looks fine in Cubit and Paraview, but my limited skill with both of
these programs is keeping me from a thorough analysis.


On 10/22/09 1:56 PM, "Brad Aagaard" <baagaard at usgs.gov> wrote:

> Oliver-
> 
> Another potential source for getting this error is that there is an
> error in the topology of the mesh.  My guess is that locally refining
> the mesh is introducing inconsistencies in the topology of the mesh.
> Does the refined mesh look okay in ParaView (it can read Exodus files)?
> Also use the quality filters available in ParaView to try to find any
> topology issues.
> 
> The response above applies to both tet and hex meshes, but it is not
> clear whether you are using a tet mesh or a hex mesh. Your lastest
> message refers to a hex mesh but your earlier email refers to a tet mesh.
> 
> Brad
> 
> 
> Oliver Boyd wrote:
>> As you suggest (hopefully), I removed all references to faults in
>> dislocation.cfg. The error disappears. But I only have one fault, and I
>> don't quite understand how the fault is the problem. I am now doing this
>> with hex meshes. The problem appears when I start to locally refine the
>> mesh. And the local refinement can be far from the fault. How is refinement
>> far from the fault affecting the fault?
>> 
>> 
>> On 10/21/09 5:31 PM, "Brad Aagaard" <baagaard at usgs.gov> wrote:
>> 
>>> Oliver-
>>> 
>>> This error is related to creating cohesive cells. First, try reading the
>>> mesh without any faults (this error should disappear). Then try with one
>>> fault. If you have multiple faults, try to isolate which fault is
>>> causing the error. The restrictions are: (1) no vertex can be on more
>>> than one fault and (2) more than one of a cell's faces cannot be on a fault.
>>> 
>>> Brad
>>> 
>>> 
>>> Oliver Boyd wrote:
>>>> Having trouble reading a cubit generated tet mesh into pylith. I get the
>>>> following output and error.
>>>> 
>>>> $pylith dislocation.cfg
>>>> 
/Users/buildbot/install/x86_darwin_single_nosched_binbot/lib/python2.3/site>>>>
-
>>>> packages/pylith/utils/PetscManager.py:48:initialize
>>>>  -- petsc(info)
>>>>  -- Initializing PETSc.
>>>> 
/Users/buildbot/install/x86_darwin_single_nosched_binbot/lib/python2.3/site>>>>
-
>>>> packages/pylith/meshio/MeshIOObj.py:46:read
>>>>  -- meshiocubit(info)
>>>>  -- Reading finite-element mesh
>>>>>> meshio/MeshIOCubit.cc:151:<unknown>
>>>>  -- meshiocubit(info)
>>>>  -- Reading 37250 vertices.
>>>>>> meshio/MeshIOCubit.cc:207:<unknown>
>>>>  -- meshiocubit(info)
>>>>  -- Reading 199518 cells in 2 blocks.
>>>>>> meshio/MeshIOCubit.cc:283:<unknown>
>>>>  -- meshiocubit(info)
>>>>  -- Found 7 node sets.
>>>>>> meshio/MeshIOCubit.cc:303:<unknown>
>>>>  -- meshiocubit(info)
>>>>  -- Reading node set 10 with 788 nodes.
>>>>>> meshio/MeshIOCubit.cc:303:<unknown>
>>>>  -- meshiocubit(info)
>>>>  -- Reading node set 11 with 85 nodes.
>>>>>> meshio/MeshIOCubit.cc:303:<unknown>
>>>>  -- meshiocubit(info)
>>>>  -- Reading node set 12 with 85 nodes.
>>>>>> meshio/MeshIOCubit.cc:303:<unknown>
>>>>  -- meshiocubit(info)
>>>>  -- Reading node set 13 with 85 nodes.
>>>>>> meshio/MeshIOCubit.cc:303:<unknown>
>>>>  -- meshiocubit(info)
>>>>  -- Reading node set 14 with 85 nodes.
>>>>>> meshio/MeshIOCubit.cc:303:<unknown>
>>>>  -- meshiocubit(info)
>>>>  -- Reading node set 15 with 590 nodes.
>>>>>> meshio/MeshIOCubit.cc:303:<unknown>
>>>>  -- meshiocubit(info)
>>>>  -- Reading node set 16 with 590 nodes.
>>>> faults/TopologyOps.cc:111: failed assertion `classifySize <
>>>> vReplaceCells.size() + vNoReplaceCells.size()'
>>>> [0]0:Return code = 0, signaled with Abort trap
>>>> 
>>>> Thanks for your help,
>>>> Oliver
>>>> 
>>>> 
>>>> 
>>>> 
>>>> ------------------------------------------------------------------------
>>>> 
>>>> _______________________________________________
>>>> CIG-SHORT mailing list
>>>> CIG-SHORT at geodynamics.org
>>>> http://geodynamics.org/cgi-bin/mailman/listinfo/cig-short
>> 
>> 
>> 
>