The ordering of mid-face nodes and centroid that works with paraview is 26,
20, 25, 24, 22, 21, 23 but it is absolutely non-sense for me.

Your ordering is good for IO but for some reason I don't understand I need
to do a mesh.prepare_for_use() before adding the mesh to equation_systems.
Otherwise I get a seg fault.

I'm quite disappointed because even the prism18 is broken in paraview.
I think the problem is the vtk exodus reader.

Thanks
Lorenzo

2010/4/27 Derek Gaston <[email protected]>

> Moving discussion to devel list.
>
> Is that after you reverted my changes?
>
> We definitely read Hex27's without issues around here.  Writing is another
> story.
>
> That said... I'm not happy with the Hex27 numbering... I know it's not
> right with Paraview.  I worked on it for a while and as far as I can tell
> what I have in code right now matches both libMesh and Exodus documentation.
>  BUT... I must be wrong because it's obviously not quite working correctly.
>
> One issue is that it appears to be fine in Ensight... but not Paraview...
> and if I try to fix it in Paraview then it breaks when I try to read a mesh
> in.  I wondered a while ago if Cubit, Paraview and Ensight were all using
> the same Exodus standard... but there really isn't a way to know.  I highly
> doubt that's the problem though.
>
> Let's get this figured out once and for all.  If you want to take a swipe
> at an ordering... that would be awesome.  I'll definitely check any code you
> want me to.  I'll also work here and see if I can spot something that I
> haven't before.
>
> Derek
>
>
> On Apr 27, 2010, at 3:57 AM, Lorenzo Botti wrote:
>
>  If I try to read a hex 27 file written with exodusII I get a negative
>> jacobian...
>> There could be a problem with connectivity.
>> With hex20 everything works fine.
>>
>> Lorenzo
>>
>> ---------- Forwarded message ----------
>> From: Lorenzo Botti <[email protected]>
>> Date: 2010/4/27
>> Subject: Re: [Libmesh-users] problem with exodusII, hex27 and paraview
>> To: John Peterson <[email protected]>
>>
>>
>> Dear John,
>> thank you for the reply.
>>
>> I've already tried the old version of the connectivity but it doesn't work
>> with paraview.
>> The ordering seems correct so I guess you are able to read a exodusII mesh
>> with hex27 elements.
>>
>> Lorenzo
>>
>>
>> 2010/4/26 John Peterson <[email protected]>
>>
>> On Mon, Apr 26, 2010 at 2:18 PM, Lorenzo Botti <[email protected]>
>>
>>> wrote:
>>>
>>>> Hi all,
>>>> I noticed a problem in the visualization (with paraview) of hex27
>>>>
>>> elements written using exodusII_IO.
>>>
>>>> The connectivity is broken resulting in a corrugated surface.
>>>> I took a look at the patran documentation, nodes numbering as well as
>>>>
>>> face numbering seems correct to me.
>>>
>>>>
>>>> Is someone else aware of this problem? Could it be a vtk problem?
>>>> Can I read exodus files with others open source visualization tools?
>>>>
>>>
>>> We've had problems with the Hex27 numbering in ExodusII before.  It
>>> looks like Derek was the last one to change the indices around...here
>>> are the diffs from that version:
>>>
>>> svn diff -r3391:3462 exodusII_io_helper.C
>>>
>>> const int ExodusII_IO_Helper::ElementMaps::hex27_node_map[27] = {
>>>  // Vertex and mid-edge nodes
>>>  0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19,
>>>  // Mid-face nodes and centroid
>>> -  26, 21, 22, 23, 24, 25, 20};
>>> +  21, 25, 24, 26, 23, 22, 20};
>>>
>>> -const int ExodusII_IO_Helper::ElementMaps::hex27_face_map[6]   = {1,
>>> 0, 3, 5, 4, 2};
>>> +const int ExodusII_IO_Helper::ElementMaps::hex27_face_map[6]   = {1,
>>> 2, 3, 4, 0, 5};
>>>
>>> -const int ExodusII_IO_Helper::ElementMaps::hex27_inverse_face_map[6]
>>> = {2, 1, 6, 3, 5, 4};
>>> +const int ExodusII_IO_Helper::ElementMaps::hex27_inverse_face_map[6]
>>> = {5, 1, 2, 3, 4, 6};
>>>
>>> You might try reverting them and see if that helps?  Perhaps paraview
>>> is going off an older Exodus numbering scheme.
>>>
>>> --
>>> John
>>>
>>> ------------------------------------------------------------------------------
>> _______________________________________________
>> Libmesh-users mailing list
>> [email protected]
>> https://lists.sourceforge.net/lists/listinfo/libmesh-users
>>
>
>
------------------------------------------------------------------------------
_______________________________________________
Libmesh-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/libmesh-devel

Reply via email to