Re: Newby question - Pb to fix BCs on a mesh from Tetgen
Hi Robert,
I succeeded in applying my Bcs with Tetgen.
Following one example you provided in the Sfepy distribution (acoustic_wg.vtk) I added to my VTK file a section: POINT_DATA number of nodes SCALARS node_groups int 1 LOOKUP_TABLE default
Where I set "number of nodes" according to my mesh Then I copy all node attributes provided by the *.node file generated by Tetgen
Then I use the following syntax in Sfepy:
regions = { 'Omega' : 'all', # or 'cells of group 6' 'Gamma_Left' : ('vertices of group 4', 'edge'), 'Gamma_Right' : ('vertices of group 5', 'edge'), }
'4' and '5' are the number I currently use to identify my BCs.
And surprisingly it works ! Regions definition did not work when I use 'facet' instead or 'edge'.
I wonder why, when I read the *.node file from Sfepy, the node attributes can not be used to set BCs ?
All the very best
Marc
Hi Marc,
On 01/23/2018 10:00 AM, Marc Saudreau wrote:
Hi Robert,
I succeeded in applying my Bcs with Tetgen.
Following one example you provided in the Sfepy distribution (acoustic_wg.vtk) I added to my VTK file a section: POINT_DATA number of nodes SCALARS node_groups int 1 LOOKUP_TABLE default
Where I set "number of nodes" according to my mesh Then I copy all node attributes provided by the *.node file generated by Tetgen
Then I use the following syntax in Sfepy:
regions = { 'Omega' : 'all', # or 'cells of group 6' 'Gamma_Left' : ('vertices of group 4', 'edge'), 'Gamma_Right' : ('vertices of group 5', 'edge'), }
'4' and '5' are the number I currently use to identify my BCs.
And surprisingly it works ! Regions definition did not work when I use 'facet' instead or 'edge'.
Glad to hear that! Compare your solution with my other response in the previous e-mail.
I wonder why, when I read the *.node file from Sfepy, the node attributes can not be used to set BCs ?
It is simply not implemented - the extra data are thrown away. I have created an easy-to-fix issue, so that it is not forgotten, see [1].
r.
[1] https://github.com/sfepy/sfepy/issues/442
All the very best
Marc
SfePy mailing list sfepy@python.org https://mail.python.org/mm3/mailman3/lists/sfepy.python.org/
participants (2)
-
Marc Saudreau
-
Robert Cimrman