Tutorial for Structure/Structure coupling with FEniCS

[BenjaminRodenberg]

What do we want in the end?

• we want a tutorial similar to the already existing heat-heat coupling with FEniCS. Here, we want to split a beam into two partitions to solve the linear elasticity problem (more information in the FEniCS tutorial book, p.50ff)
• this tutorial can be used to test the exchange of vector values quantities via the fenics adapter. See Handling of vector-valued coupling data fenics-adapter#26

Current State

• Implement (pseudo-)time loop to solve static problem. See Comment below
• Use advance in (pseudo-)time loop.
• Resolve Support multiple Neumann boundary conditions per domain and provide a tutorial on this feature fenics-adapter#31, we need this for applying zero traction on the outer boundary while applying non-zero traction on the coupling boundary.

Pseudo Time Loop

Linear elasticity is a static problem, which is generally not handled by preCICE, but can be circumvented by reformulating the problem as laplacian(u) - f = du/dt. When du/dt = 0 (= convergence), the static problem laplacian(u) = f is solved

[uekerman]

As discussed with @BenjaminRueth on the phone:

I think FSI would be the better choice here than SSI:

• FSI is much more relevant to our users.
• SSI has still some numerical issues that we have not yet fully resolved / understood. See the master thesis of @Alex-tru (not online afaik, I can send an offline copy if somebody needs it). My guess is that it is a similar issue than the flux computation for heat-heat coupling.
• For FSI, you only have "half of the work" as you "only" need to read forces and write displacements, but not the other way around as well.

So, I would recommend to approach FSI first and maybe SSI afterwards if needed.

[BenjaminRodenberg]

I added the wontfix label and we can leave this PR open to indicate that this PR is postponed.

[BenjaminRodenberg]

This PR is very outdated. Rewriting the setup based on https://github.com/precice/tutorials/tree/master/partitioned-elastic-beam and adapting https://github.com/precice/tutorials/tree/master/perpendicular-flap/solid-fenics makes more sense.