Hi, > >> Unfortunately, neither the number of nodes nor the > >> number of the elements are correct. according to my knowledge, Gmsh seems to firstly generate mesh in a lower dimension, which means if you try to create a 3D mesh, Gmsh will automatically generate a surface mesh for you, and your 3D mesh is actually created based on the surface mesh. However, Gmsh also stores those surface elements and nodes in its msh file. LibMesh does its element type identification while it reads a msh file. Those elements and nodes in a lower dimension are simply thrown away, therefore when you print information about the mesh, the numbers of elements and nodes are always less than you saw in the msh file. In order to check the data, I usually read the msh file and write the data into another msh file, then open the msh file with Gmsh again. I hope I didn't get it wrong and the little experience will do some help.
regards Ping On Tuesday 21 April 2009 14:46:46 Svilen Sabchevski wrote: > >> Dear All, > >> > >> Please see attached a simple test code called gyreoss.c. > >> It reads the mesh file test.msh produced by gmsh using > >> the geometry description in the file test.geo. Then the > >> code prints information about the mesh. Also attached > >> are files with the output of the code in optimized as well > >> as in the debugging modes > >> > >> Unfortunately, neither the number of nodes nor the > >> number of the elements are correct. > >> > >> Additionally there is an error when attempting to get data from > >> the active elements: > >> Assertion `_elem_data_closed' failed. > >> > >> > >> I wonder if this is a bug or improper usage of the library > >> from my side. > >> > >> Any hint would be greatly appreciated. > > > >Hi, > > > >For me at least, your attachment did not make it through to the list. > >I also did not find your message trapped in any of the sourceforge > >list filters. If the attachment was larger than 40K, it may have been > >deleted automatically... You can send it to me directly and I'll have > >a look when time permits. > > > >-- > >John > > Hi, > > The files can be seen at: > > http://depositfiles.com/files/ldfmte89w > > Svil > > > > --------------------------------------------------------------------------- >--- Stay on top of everything new and different, both inside and > around Java (TM) technology - register by April 22, and save > $200 on the JavaOne (SM) conference, June 2-5, 2009, San Francisco. > 300 plus technical and hands-on sessions. Register today. > Use priority code J9JMT32. http://p.sf.net/sfu/p > _______________________________________________ > Libmesh-users mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/libmesh-users ------------------------------------------------------------------------------ Stay on top of everything new and different, both inside and around Java (TM) technology - register by April 22, and save $200 on the JavaOne (SM) conference, June 2-5, 2009, San Francisco. 300 plus technical and hands-on sessions. Register today. Use priority code J9JMT32. http://p.sf.net/sfu/p _______________________________________________ Libmesh-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/libmesh-users
