Re: [Libmesh-users] Solving a system with 2 subdomains

[Viviana P. B. <vpa...@uc...>]

Hi John,

Thank you for the clarification, it's really helpful.

My assembly function is separated, one for the matrix and another one for
the rhs. I loop over the elements of each subdomain to avoid the if
statement.

Best,
Viviana.

Viviana Palacio Betancur
PhD Student | 2016 Cohort
de Pablo Group
Institute for Molecular Engineering
University of Chicago

On Thu, Jan 4, 2018 at 12:10 PM, John Peterson <jwp...@gm...> wrote:

>
>
> On Thu, Jan 4, 2018 at 9:16 AM, Viviana Palacio Betancur <
> vpa...@uc...> wrote:
>
>> Hello,
>>
>> I am solving a system on a mesh with two subdomains: bulk with elements
>> TET10, and surface with elements TRI6. So far, I've sucessfully asembled
>> the system's matrix and rhs, and initialized the system.
>>
>> My problem is when I execute system.solve(). Apparently, the solution is
>> only done for the bulk subdomain and thus the surface solution is almost
>> 0.
>> Is the default setting to solve only for the first subdomain? Should I be
>> using the system.restrict_solve_to?
>>
>
>
> All variables (even subdomain restricted variables, if you are using
> those) in the System are solved for simultaneously.
>
> Do you have an if-statement in your assembly function that assembles
> different PDE terms depending on what subdomain you are currently on? How
> is your boundary subdomain actually coupled to the volume?
>
> --
> John
>