Dear deal.II users,
I have a general question regarding constraints on degrees of freedom between
two subdomains.
Suppose I have a domain Omega, and partition it such that Omega_1 U Omega_2 =
Omega with Omega_{1,2} having null intersection. I would like to solve the
elasticity equations on each domain separately, with the constraint of
continuous displacements at the interface between Omega_1 and Omega_2. The
motivation for this is an ALE formulation of FSI, where Omega_2 is the fluid
and mesh motion domain.
If I implement my constraints in the typical ConstraintMatrix manner that
hanging nodes or periodic boundaries use, the solution of Omega_1 and Omega_2
mutually influence each other. I do not want that!
I am looking for a method to strongly enforce continuity on the interface
whereby the displacement solution in Omega_1 is NOT constrained by Omega_2, but
Omega_2 IS constrained by the displacement solution of Omega_1.
Due to the nature of the rest of my system, I do not see a way to do a
block-by-block linear solve and insert non-homogeneous constraints at the
proper step. I would prefer to insert these "one-way" constraints directly
into a single system matrix and solve for all unknowns monolithically.
Has anyone done this in deal.II? Is there functionality for doing this that I
am unaware of? Or, maybe I am just misusing/abusing the ConstaintMatrix such
that it is not doing what I intend?
Thanks,
-Scott
_______________________________________________
dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
