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
