Currently I have to solve a system of several time-dependent equations, one 
set being a set of different heat equations, and one set is a set of 
schrödinger/maxwell (depending on the requirement) equations. All equations 
are coupled. The domain of the latter set is bigger than the domain of the 
former set, but both domains overlap. My current approach is to solve the 
first set using deal.II and the CG-method, and the latter set on a custom 
grid with a split-step method. According to my knowledge the interpolation 
between both domains (which has to happen after every step, using 
FEFieldFunction) is by far the most expensive part of the problem.  
Thus, would it be easier to follow (for example) example 46 and use a 
single DoFHandler, while calculating both sets of equations on two 
different meshes (which overlap, unlike in example 46), or if I should stay 
with my current approach? I do not want to have all sets of equations at 
the same domain, after I am only interested into a small part of the domain 
of the second set of equations with respect to the first set of equations. 
That allows me to have a higher grid density for the first set of 
equations, without increasing the amount of calculation time and used 
memory. Unfortunately I do not have the necessary experience for deciding 
that (yet), thus I am asking here.

Thanks!

-- 
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 [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to