where and how should I use the MeshWorker class? how should I do to implement DG without MeshWorker?

[yocz...@gmail.com]

Hi,
I'm reading the step-12, and i get confused that where and how should I use the MeshWorker class?
The MeshWorker class seems just using in the DG examples.

But I do not think it is a good idea to use MeshWorker in my problems, the problems, I will treat,  maybe a little  complicated, it combines
DG and couples different kinds of equations in different parts of the domain. If I don't want to use the MeshWorker, how should I do to implement DG?

Are there any examples using DG without MeshWorker?

[Daniel Arndt]

An old version of step-12[1] has an implementation that doesn't use MeshWorker.

In my opinion, it is covenient to use MeshWorker if you can when you want to consider discontinuous ansatz spaces due to the

different cases that can occur for face contributions. If you want or need to use more than one DoFHandler, this more or less the only case when you can't use MeshWorker directly. You might however, be able to combine different equations in a single DoFHandler.

What exactly is it that you want to do?

Best,

Daniel

[1] https://www.dealii.org/7.1.0/doxygen/deal.II/step_12.html 

[yocz...@gmail.com]

Dear Daniel
Thanks for your answer, the 7.1.0 version of step-12 seems also using the MeshWorker,  i find the 6.2.1 version of step-12 without using MeshWorker: https://www.dealii.org/6.2.1/doxygen/deal.II/step_12.html

I want to repeat the numerical example of the references, which is a stokes-darcy problem in different parts of the domain. The stokes equation may takes DG basis functions.

Actually, I also want to use the MeshWorker,  what worried me is that I can't control MeshWorker by myself, it something likes I know  the MeshWorker can do lots of thing, but I don't know how to use MeshWorker. I am just beginning to use deal.ii, and deal.ii seems a little complex to me.  So I want to find some examples of MeshWorker.  And I am wondering if it is there any more advices?

Best,
Yongchao.



在 2016年8月28日星期日 UTC+8上午3:13:03，Daniel Arndt写道：

[Wolfgang Bangerth]

On 08/27/2016 08:45 PM, yocz...@gmail.com wrote:
>
> Actually, I also want to use the MeshWorker, what worried me is that I can't
> control MeshWorker by myself, it something likes I know the MeshWorker can do
> lots of thing, but I don't know how to use MeshWorker. I am just beginning to
> use deal.ii, and deal.ii seems a little complex to me. So I want to find some
> examples of MeshWorker. And I am wondering if it is there any more advices?

Yongchao -- the issue with MeshWorker is that when it was implemented, we
didn't write enough documentation/tutorial programs/examples that show how to
use it in more complicated situations. Development then also stopped, and so
there are valid cases where it would be nice to use MeshWorker for which
nobody today knows any more whether they can or can not be done with
MeshWorker. The only person who still knows how this all works unfortunately
also doesn't regularly participate in the help forum any more :-(

In other words, as in other threads about MeshWorker on this forum, you can of
course ask questions about how this or that can be done, but you're likely not
going to get a lot of helpful answers because nobody knows any more how this
works. It's really an unfortunate situation we got ourselves into with this
small corner of the library.

Best
Wolfgang

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

[yocz...@gmail.com]

Dear Wolfgang ,
Thanks for your advice, I have read the 6.2.1 version of step-12, and I know how to deal DG without using MeshWorker.
What's more there is the  6.2.1 version of step-12 without using MeshWorker, if someone needs.
https://www.dealii.org/6.2.1/doxygen/deal.II/step_12.html


在 2016年8月29日星期一 UTC+8上午11:38:34，Wolfgang Bangerth写道：

[Daniel Arndt]

Yongchao,

If you are still interested in using MeshWorker, there are some more examples (and in particular Stokes) at
https://bitbucket.org/guidokanschat/amandus

Best,
Daniel

[yocz...@gmail.com]

Dear Daniel,
Thank you, I have looked at it, it may take some time to learn. It's a good news for me.

在 2016年8月29日星期一 UTC+8下午5:23:27，Daniel Arndt写道：