problem measure() for quads

[Andreas Kyritsakis]

Hi,

I have a problem with using the measure() function for quads in 3D. I am currently using version 9.1.1 and the function seems to check whether the face is planar and if not, it returns nan. The problem is that the tolerance for considering it planar is extremely low (1.e-24), which causes problems. If for example the mesh is read in from a standard .msh file, written by the GridOut class, we cannot expect to get such a precision in the vertices. I saw that in the 9.2 version this is partially fixed by handling the non-planar case by numerical intergration, but the tolerance has remained the same extremely low, which might result in unnecessary computational effort. 

I wonder whether the tolerance can be increased to something more reasonable (e.g. 1.e-10) or there is something I am missing that suggests such a low tolerance.

Best regards,

Andreas Kyritsakis

[Wolfgang Bangerth]

It's a good question. I checked that the expression in question is
dimensionless and scaled correctly so that the mesh size h does not appear in
the size of the expression. I think it would be quite ok if you changed the
tolerance to 1e-10.

Want to propose a patch for this?

Best
W.


--
------------------------------------------------------------------------
Wolfgang Bangerth email: bang...@colostate.edu
www: http://www.math.colostate.edu/~bangerth/

[Andreas Kyritsakis]

Thank you very much Wolfgang. Yes I would like to propose a patch. How is it done? I am not familiar with the process.

[Jean-Paul Pelteret]

Hi Andreas,

Thanks for being willing to make a pull request of this. The basic instructions for how to do so are listed in our FAQ, 

https://github.com/dealii/dealii/wiki/Frequently-Asked-Questions#i-found-a-typo-or-a-bug-and-fixed-it-on-my-machine-how-do-i-get-it-included-in-dealii

but of course if you have any troubles just reach out to us and we’d be happy to help further.

Best,

Jean-Paul

-- 
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see https://groups.google.com/d/forum/dealii?hl=en
--- 
You received this message because you are subscribed to the Google Groups "deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dealii+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dealii/231d1ecd-03a2-4cce-9590-ed1470fd5278%40googlegroups.com.