Here, I attached simple case that I encounter same error message.
Most of code is just same at that of step-5, but I am using mesh-file that
is made from other software. (I attach my mesh file too)
At constructor, there is a issue.
template
Step5::Step5 () :
triangulation (Triangula
On 03/31/2017 01:30 PM, Jaekwang Kim wrote:
I found that what was the problem.
I was using 'maximum' Meshsmoothing which does not allow
So, It was not a bug of deal.ii but I was using wrong smoothing method
that does not go along with anisotropy,
Well, it is still a bug that the library did
Thanks Bruno ! and Dr. Wolfgang!
I found that what was the problem.
I was using 'maximum' Meshsmoothing which does not allow
So, It was not a bug of deal.ii but I was using wrong smoothing method that
does not go along with anisotropy,
I just changed my Meshsmoothing Method, that allow anis
Jaekwang,
2017-03-30 17:39 GMT-04:00 Jaekwang Kim :
> //This one is problematic
> triangulation.prepare_coarsening_and_refinement ();
It could be a bug in the library, most people don't use anisotropic
refinement. This function is not always necessary (and it is not used
in step-30
Always thanks for you.
I have read step-30, and now i understand how it is done,
I was trying to implement on my problem as like
Frist, refine_mesh function , I have designated which cell to be refined,
template
void
*StokesProblem::refine_mesh (**const unsigned int
refinem
On 03/29/2017 06:37 PM, Jaekwang Kim wrote:
I am sorry but I am not that good at c++, so I was bit confused how to use
this function.
I have just set flag cell like
*cell->set_refine_flag ();*
Reading the document pages, you linked, I can understand that I should put
some argument inside par
thanks,
I would need it and I am still running serially
I am sorry but I am not that good at c++, so I was bit confused how to use
this function.
I have just set flag cell like
*cell->set_refine_flag ();*
Reading the document pages, you linked, I can understand that I should put
some argum
To add to Bruno's answer:
If you don't want to use anisotropic refinement you have two other options:
1. Just accept that you get additional refinement in areas/directions you
don't particularly care about. If you have a realistic computation, you
likely need to refine many more times and thus the
Jaekwang,
you can do anisotropic refinement
(http://dealii.org/developer/doxygen/deal.II/classCellAccessor.html#afb6cc537720a5b6381c237abe0887de2)
but only in serial. So depending if your code is parallel or not this may
be what you are looking for.
Best,
Bruno
On Wednesday, March 29, 2017
