Re: [deal.II] Best approach for multi-variable system with multiple grids

[Daniel Arndt]

Maxi,

I know that both components will behave quite differently, thus I assumed 
> that I have to use two grids. 
>
Having to deal with two grids always creates a lot of overhead for the 
coupling terms. So you normally want to avoid this if you can.
 

> In addition I was interested in how both parameters evolve over time. If I 
> need only one grid, then I still have to store them separately, but it 
> would be closer to example 33, wouldn't it?
>
In step-33 you only have one DoFHandler in combination with a FESystem that 
stores all components for all variables. In particular, these variables are 
stored in one common Vector. This is the approach I would suggest 
unless you have special requirements.
Of course, you can also use BlockSparseMatrix objects and BlockVector 
objects instead if you want to be able to access the individual components 
more easily. This approach is shown in step-22 [1].

Best,
Daniel

[1] https://www.dealii.org/8.5.0/doxygen/deal.II/step_22.html

-- 
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+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [deal.II] Best approach for multi-variable system with multiple grids

['Maxi Miller' via deal.II User Group]

I know that both components will behave quite differently, thus I assumed 
that I have to use two grids. In addition I was interested in how both 
parameters evolve over time. If I need only one grid, then I still have to 
store them separately, but it would be closer to example 33, wouldn't it?

Am Dienstag, 11. Juli 2017 22:22:39 UTC+2 schrieb Wolfgang Bangerth:
>
> On 07/11/2017 08:29 AM, 'Maxi Miller' via deal.II User Group wrote: 
> > Which of those two approaches is more beneficial for me (especcially 
> after I 
> > would like to extend it afterwards to two additional equations? Or is 
> there 
> > another approach I did not consider yet? 
>
> I don't think I quite understand your two options, but in any case: What 
> is 
> the reason why you need two different meshes? Is the qualitative behavior 
> of 
> the two variables vastly different? Unless the diffusion constants for the 
> two 
> variables is vastly different, there is usually no particularly good 
> reason to 
> use different meshes. 
>
> Best 
>   W> 
>
> -- 
>  
> Wolfgang Bangerth  email: bang...@colostate.edu 
>  
> www: http://www.math.colostate.edu/~bangerth/ 
>
>

-- 
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+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [deal.II] Best approach for multi-variable system with multiple grids

[Wolfgang Bangerth]

On 07/11/2017 08:29 AM, 'Maxi Miller' via deal.II User Group wrote:
Which of those two approaches is more beneficial for me (especcially after I 
would like to extend it afterwards to two additional equations? Or is there 
another approach I did not consider yet?


I don't think I quite understand your two options, but in any case: What is 
the reason why you need two different meshes? Is the qualitative behavior of 
the two variables vastly different? Unless the diffusion constants for the two 
variables is vastly different, there is usually no particularly good reason to 
use different meshes.


Best
 W>

--

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

--
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+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[deal.II] Best approach for multi-variable system with multiple grids

['Maxi Miller' via deal.II User Group]

I have a coupled system of two equations:








which I want to solve. Here k_L, k_E, f_1 and f_2 are nonlinear functions 
depending on T_E and T_L, while Q_E is a nonlinear function depending on 
the time. Now I can either reformulate it as seen in example 33, but then 
have to extend the grids to two grids, not only one (for TE and TL):

>
>


or I can do the same approach as in example 28 (where I already have two 
grids), but there the variables L and F are set to zero (only X is 
nonzero). Which of those two approaches is more beneficial for me 
(especcially after I would like to extend it afterwards to two additional 
equations? Or is there another approach I did not consider yet?

The target application is in 3d, and isotrop.

-- 
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+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.