Fluid Solid Interaction to Investigate Elastic Bi-Stability

[ABrierley]

Hello,

Firstly, a sincere thanks to the active forum members, moderators and developers. FEBio is an amazing tool and the community around it make it even more useful and accessible.

I am investigating elastic bi-stability. I have a round rubber diaphragm which is moulded flat but is deformed into a cone by having a central shaft of a large diameter stretch its centre. When air passes over the deformed diaphragm it 'flips' from one bi-stable state to the other.
I have modelled the initial deformation by simply using prescribed displacement to pull the diaphragm over the central shaft. I have then added a second, structural mechanics step to apply pressure to the diaphragm to cause it to flip. This is okay however if possible I would rather use FSI to flip the diaphragm, this I believe would be the most accurate representation of the problem and would yield much better results.

I am looking for some advice or an assist on how to make this problem solve. I have attached the intial step of the simulation. I do not have an understanding of how to set up the FSI simulation. Should I be modelling the air around the other 2 components? Any help or guidance on how to set up the FSI simulation from someone with experience would be greatly appreciated.

180320.feb

Alex

[jshim777]

Hi Alex

If you want to make it an FSI simulation then you would have to include a air domain that surrounds the other two parts (you have to decide if you want it to completely surround it or only along the axial direction). There is a short tutorial in the most recent PreView manual that walks you through using both the CFD and FSI solvers in FEBio, which may be helpful to get you started. However, the downside of doing this is that it would be much more computationally expensive and there are a some additional things you have to keep in mind for the boundary conditions (now you also have fluid velocity and pressure/dilatation in addition to just displacement).

Because of this I would recommend that you first make sure that you get desirable results with just structural mechanics by applying a pressure before proceeding with the FSI. As a side note in your example feb file, it would be better if you can use a hex or quadratic tet mesh for your diaphragm (for rigid body mesh any type is okay) since tet4 meshes are susceptible to mesh locking. Also I noticed that the NeoHookean material you are using has a pretty high Poisson ratio of 0.47, it might converge better if you use an uncoupled material such as mooney rivlin instead if your material is almost incompressible. If you do go ahead with FSI the bubble problem from the recent publication on FSI in FEBio might be useful to look at: https://www.ncbi.nlm.nih.gov/pubmed/30835271. It inflates a thin solid sandwiched between two fluid domains where one side is prescribed a pressure, which is probably similar to what you want to do, and should be doable for your problem if you are careful with the boundary conditions and are mindful of any fluid boundary layer (you need to refine your mesh there).

Hopefully this is helpful!

Jay