Adds functionality to allow FSI to work with solids4foam as the solid solver

[solids4foam]

This pull request adds functionality to allow FSI to work with solids4foam as the solid solver.

TODO list:

• I updated the documentation in docs/: I don't think I need to change anything here, do I?
• I added a changelog entry in changelog-entries/ (create directory if missing): apologies, I am not sure what I meant to add here.

[davidscn]

Looks nice, thanks a lot. Have you tested the implementation using a simple tutorial? Maybe our perpendicular-flap tutorial?

[MakisH]

Hi @solids4foam (who's that? 😅),

thanks a lot for the contribution! I left some first comments.

We will also need a tutorial to try this out. You can add one here: https://github.com/precice/tutorials Ideally, you could recreate the perpendicular-flap example.

Regarding the changelog, I can also handle that. You just need a file 236.md in the changelog-entries/ directory.

[solids4foam]

Looks nice, thanks a lot. Have you tested the implementation using a simple tutorial? Maybe our perpendicular-flap tutorial?

I initially tested it with a simple beamInCrossFlow test case from solids4foam [1,2], but I agree it would be good to set up the perpendicular flap test cases. I will start on this.

1. Richter, T., Goal-oriented error estimation for fluid-structure interaction problems, Computer Methods in Applied Mechanics and Engineering 223-224 (2012), pp. 28–42, https://doi.org/10.1016/j.cma.2012.02.014.
2. Tuković Ž, Karač A, Cardiff P, Jasak H, Ivanković A (2018) OpenFOAM finite volume solver for fluid–solid interaction. Transactions of FAMENA, 42(3):1–31. 10.21278/TOF.42301.

[solids4foam]

Hi @solids4foam (who's that? 😅),

Ha, it is I, Philip Cardiff!

thanks a lot for the contribution! I left some first comments.

Sure, of course.

We will also need a tutorial to try this out. You can add one here: https://github.com/precice/tutorials Ideally, you could recreate the perpendicular-flap example.

Yep, I will set up the perpendicular-flap example, and then I could create a pull request for that too.

Regarding the changelog, I can also handle that. You just need a file 236.md in the changelog-entries/ directory.

I will add this.

[solids4foam]

FYI, I have created a pull request to the tutorials repository at precice/tutorials#286, where I have added the perpendicular-flap/solid-solids4foam.

I have checked this with OpenFOAM-v2012 and the nextRelease branch of solids4foam.

I think it will not be difficult for me to also add a solid-openfoam case which uses the builtin OpenFOAM solidDisplacementFoam solver and the same openfoam-adapter. It should just require a new boundary condition. I will think about this tomorrow...

[MakisH]

Thanks for the updates! A few more comments.

I have not yet managed to build solids4foam to test this, I guess I will bother you on the side regarding this.

[MakisH]

I think it will not be difficult for me to also add a solid-openfoam case which uses the builtin OpenFOAM solidDisplacementFoam solver and the same openfoam-adapter. It should just require a new boundary condition. I will think about this tomorrow...

That would also be a nice idea! I have never used that solver, but could be a nice intro and hint to solid simulations with OpenFOAM. One could just run the solidDisplacementFoam and we could link from that to the solids4foam tutorial for something more elaborate.

[solids4foam]

I think it will not be difficult for me to also add a solid-openfoam case which uses the builtin OpenFOAM solidDisplacementFoam solver and the same openfoam-adapter. It should just require a new boundary condition. I will think about this tomorrow...

That would also be a nice idea! I have never used that solver but could be a nice intro and hint to solid simulations with OpenFOAM. One could just run the solidDisplacementFoam, and we could link from that to the solids4foam tutorial for something more elaborate.

OK, good news, I got the case working with the default solidDisplacementFoam solver. I added this to this tutorials pull request (1db82dc); you will also see a minor change (will probably require some discussion) required to the adapter in 4b49300.

[MakisH]

We are getting there! I tried to resolve all conversations that are now resolved. Still open:

• Add a warning: Adds functionality to allow FSI to work with solids4foam as the solid solver #236 (comment)
• Internal field vs boundary field (displacement reading): Adds functionality to allow FSI to work with solids4foam as the solid solver #236 (comment)
• Pointer or other approach for the force field (probably concluded already): Adds functionality to allow FSI to work with solids4foam as the solid solver #236 (comment)

We then have a couple of points open in the related tutorial(s): precice/tutorials#286

I triggered the CI already and formatting and building are fine. After the state of both PRs converges, I will also run a few tests (unfortunately we don't have automated system tests at the moment, this is on my todo list for a while).

[MakisH]

• documentation in docs/: I don't think I need to change anything here, do I?

Two Three small changes are needed:

• docs/config.md: add displacements to the writers and forces to the readers

  openfoam-adapter/docs/config.md

  Line 123 in 14c5ca1

  For fluid-structure interaction, `writeData` can be `Force` or `Stress`, where `Stress` is essentially a force vector scaled by the cell face in spatial coordinates (with any postfix), thus, a conservative quantity as well.`readData` can be `Displacement` and `DisplacementDelta` (with any postfix). `DisplacementDelta` refers to the last coupling time step, which needs to considered in the case of subcycling.

• docs/README.md: same https://github.com/precice/openfoam-adapter/blob/develop/docs/README.md?plain=1#L21-L23
• Also, we can add the two solvers we now added in the tutorials as examples of solvers we couple, here: https://github.com/precice/openfoam-adapter/blob/develop/docs/openfoam-support.md

[MakisH]

I updated a couple of things in the documentation and it now finally looks ready! 😄 Thanks a lot for the contribution! I am sure that this is going to be very nice both for preCICE and for solids4foam users.