On 6/20/22 05:00, Мария Бронзова wrote:
>
> I am investigating interface phenomena, taking place between porous absorbers
> and metal plates. I am trying to create a model that would account for
> adhesive, slip and roughness effects on such interfaces.
>
> In this respect I was wondering, whether some models with interface finite
> elements have already been developed with the help of Deal II, so that I could
> take a look and understand the implementation logic.
Mariia:
are you asking whether it is possible to solve a problem where some of the
variables live in d-dimensional domains, and some others on d-1 dimensional
interfaces?
Yes, that's possible. I've implemented mortar methods this way many years ago.
The challenge is to find a mapping between points on the domain(s) and on the
interface. If the meshes for the two are generated in specific ways (e.g., one
corresponds to the faces of the other) then that mapping is easy to compute
and everything becomes easy. If the two meshes are completely unrelated, and
if the interface is curved, then things are difficult.
Best
W.
--
------------------------------------------------------------------------
Wolfgang Bangerth email:
bang...@colostate.edu
www:
http://www.math.colostate.edu/~bangerth/