Ellipsoidal grid and boundary descriptor

[Markus Klinsmann]

Hello,

I have added a hyper_ellipsoid() function and a HyperEllipsoidBoundary class to my own deal.II installation (in grid_generator.* and tria_boundary_lib.* ) to generate  geometries that have an ellipse-like shape. They are closely related to the hyper_ball() and HyperBallBoundary function and class.
I am wondering if it would be nice to add them to deal.II. However, I have no experience in how to add such contributions and I am also not using subversion (there is no internet connection and a lot of firewalls on the company cluster that I am working on). So, if there is interest, please let me know how to proceed with that. Alternatively, I could just provide the modified files.

Thanks,
Markus

[Luca Heltai]

Dear Markus,

could you send your files to me? I'm working on a branch which will modify internally the tria_boundary* files and classes. I'll add your contribution to the development version of deal.II as soon as we push out the new release.

Luca.

--
Luca Heltai <luca....@gmail.com>
http://people.sissa.it/~heltai/
Scuola Internazionale Superiore di Studi Avanzati
Phone: +39 040 3787 449, Office: 622
--
There are no answers, only cross references

> --
> 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.
> For more options, visit https://groups.google.com/groups/opt_out.

[Wolfgang Bangerth]

> I have added a hyper_ellipsoid() function and a HyperEllipsoidBoundary class
> to my own deal.II installation (in grid_generator.* and tria_boundary_lib.* )
> to generate geometries that have an ellipse-like shape. They are closely
> related to the hyper_ball() and HyperBallBoundary function and class.
> I am wondering if it would be nice to add them to deal.II.

Always.

> However, I have no
> experience in how to add such contributions and I am also not using subversion
> (there is no internet connection and a lot of firewalls on the company cluster
> that I am working on). So, if there is interest, please let me know how to
> proceed with that. Alternatively, I could just provide the modified files.

Since Luca has already volunteered I'll leave it up to him what form is most
convenient to him.

In general, it is easiest if you can send patches, for example between the
unmodified file and the one that has your changes. Since you work for a
company, let me also point out that it is generally understood that if you
send someone a modified file that carries a statement of open source license
at the top, that the intention is to place your modifications under the same
license. I assume you and your company are ok with that.

Best
Wolfgang


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

[Michał Wichrowski]

I was dealing with similar problem - I had to create cylinder with walls by function in cylindrical coordinates. I can post my code too.

BTW: did anyone considered using CGAL in deal.II ?

Michał

[Wolfgang Bangerth]

On 12/20/2013 04:17 AM, Michał Wichrowski wrote:
> I was dealing with similar problem - I had to create cylinder with walls by
> function in cylindrical coordinates. I can post my code too.

More code people post = more better :-)

> BTW: did anyone considered using CGAL in deal.II ?

For triangulations, our own data structures are probably better suited to
support finite element discretizations. But there may be other contexts where
CGAL could help. What did you have in mind?

Best
W.

[Michał Wichrowski]

The included file makes cylindrical grid, refines it and outputs. I think that it can be added to step-49.

> BTW: did anyone considered using CGAL in deal.II ?

For triangulations, our own data structures are probably better suited to
support finite element discretizations. But there may be other contexts where
CGAL could help. What did you have in mind?

Best
  W.

                                 www: http://www.math.tamu.edu/~bangerth/

I never used cgal so I have no experience with it, but some parts of it seem very useful.
There are a lot of thing that are not implemented/implemented in inefficient way in grid classes. For example finding a closest vertex/cell to point, computing distance between meshes (for contact problems) etc.. Generating complex mesh in 3D is also a problem.
I don't know if cgal supports hexaheral mesh.

[Wolfgang Bangerth]

On 12/20/2013 08:07 AM, Michał Wichrowski wrote:
> The included file makes cylindrical grid, refines it and outputs. I think that
> it can be added to step-49.

I'm not quite clear why. What does it provide in terms of explaining how to do
things that step-49 doesn't show yet?

> For triangulations, our own data structures are probably better suited to
> support finite element discretizations. But there may be other contexts where
> CGAL could help. What did you have in mind?
>

> I never used cgal so I have no experience with it, but some parts of it seem
> very useful.
> There are a lot of thing that are not implemented/implemented in inefficient
> way in grid classes. For example finding a closest vertex/cell to point,
> computing distance between meshes (for contact problems) etc.. Generating
> complex mesh in 3D is also a problem.

Yes, that may be correct. If someone came up with an interface to CGAL and a
good use case, we may consider including it.

[thomas stephens]

Markus,  I don't think this was ever added to dealii, do you still have an implementation of the ellipsoid handy?  I'm trying to make an ellipsoid (just the surface) out of a ChartManifold and it's a bit of a pain, I sure wouldn't mind looking over someone's shoulder for some help here ;)

Thanks,

Tom

[Markus Klinsmann]

Hi Tom, I have attached the files. They were working with deal.ii 8.1, but that is already quite some time ago. Hope it helps you, Markus

[thomas stephens]

Markus, thank you