then no

any other solver that comes to mind?

as far as I know, there is no FSI solver in 2 main OpenFOAM development branches. There is one solver for that in foam-extend, and you can also couple OpenFOAM with other software using the preCICE library (I have no experience with that, I just know it exists)

yes so I am coupling calculix and dealii with openfoam for the thermoaeroelasticity part, I need a solver for fluid part that can do transient pressure temperature and viscosity conditions, which will in turn affect the solid part that will be computed using precice

Alright, sounds like a job for ``rhoPimpleFoam``

I see, sorry for the late reply

thank you, I'll try it asap!

Hi everyone, I am working on a Fluid-structure interaction (3D modelling) using fenics, precice and openFoam. Needed some help with the coupling adapter (precice) setup. The documentation and tutorials are made for simple geometry(written within the code itself), but mine is a complex imported mesh, thats why the hassle. I am unable to define a subdomain(a sub-entity in the geometry) for coupling domain, and was seeking help for doing that. Already posted my query on precice discourse (https://precice.discourse.group/t/setting-coupling-interface-for-precice-in-an-imported-complex-geometry/2119) but no help or reads so far. Thanks for your time.

It's a generic divergence happening in OpenFOAM. try to understand the effect of different things here, I can see that the divergence occurred in the solve() function of sixDoFRigidBodyMotion, try to increase accelerationRelaxation or damping and try to see what happens. Also, try to change mooring line parameters; they might be too taut, and sudden acceleration of the attachment point causes this divergence.

You may be mixing things up. Divergence would be when you would begin to see ridiculous numbers before crash. These would go away with better setup of conditions, or better mesh.. etc Floating point errors occur when some illogical math occurs like divide by zero... These typically mean some bad number in your setup which ends up creating a divide by zero somewhere. Another easy one to get is segmentation faults, which occurs due to memory access issues, eg something was saved in RAM but got deleted before you tried to access it so now that memory location does not give a valid results. All these three are quite common but stem from different reasons during the calculation process.

I was wondering, Has anyone tried to use AMR with sixDoFRigidBodyMotion as motion solver?

Hello everyone Has anyone worked with foam extend 5.0 for FSI related simulations . I need some help . Any sort of help is appriciated . Thanks in advance

Which solvers are you using?

Its a custom solver

Well, kinda difficult to help you there buddy

Nah anyways I've figured out that

Currently I'm working in preCICE ...if anyone knows how to establish a connection can anyone help

I've written my python code and openfoam case is ready the .XML is also ready but there are some issue

good day eneryone. I need to simulate a rod tumbling in a simple shear flow as a test case. I suppose the best way would be to use a rigidBodyMotion like in floatingObject tutorial, but with pimpleFoam since I'm not interested in multiphase simulation at the moment. The problem is, I can't really get it how to set a cylindrical rod. Does anyone knows about any related tutorials about how to write a dynamicMeshDict for arbitrary rigid bodies? I'm using OF 12.

What do you mean by "how to set a cylindrical rod"? Create the mesh?

The dynamicMeshDict itself should be pretty similar to the floatingObject tutorial if this is the kind of motion you want to do

I mean, in the floatingObject rigid body consists of a single cuboid. I would like to know how to change it from cuboid to cylindrical rod.

yeah, that's the meshing part. You just need to mesh the shape you want. In the tutorial the mesh is created entirely with blockMesh. You can mesh a cylinder with blockMesh even if this isn't the most pleasant experience or you can use snappyHexMesh to create your mesh.

Don’t I need to set an inertia tensor, mass, center of mass in a different way from how it is done in a tutorial?

Oh ok, yes of course you need to adjust the values/setup to reflect the properties of your object.

if you want to know what other choices you have, instead of the ``cuboid`` type, you can just replace it by ``banana`` in the tutorial, and run it again. You will get an error listing the valid types. This is the good old banana trick (https://openfoamwiki.net/index.php/OpenFOAM_guide/Use_bananas)

In the .org branch, if you don't like bananas a more elegant way could be to use the ``foamInfo`` and ``foamToC`` tools to find information about it. For instance : ``foamInfo "cuboid"``

Thank you! I will try to figure it out that way

more details about these tools: https://cfd.direct/openfoam/v11-useful-tools/

shouldnt be too hard with curved edges and enough mesh resolution tbh if they have rotational symmetry, could even use a wedge, which would be equivalent in that sense