CFD Online Discussion Forums - [solids4Foam] Interface mismatch in the FSI simulation

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)

- OpenFOAM CC Toolkits for Fluid-Structure Interaction (https://www.cfd-online.com/Forums/openfoam-cc-toolkits-fluid-structure-interaction/)

- - [solids4Foam] Interface mismatch in the FSI simulation (https://www.cfd-online.com/Forums/openfoam-cc-toolkits-fluid-structure-interaction/226465-interface-mismatch-fsi-simulation.html)

Interface mismatch in the FSI simulation

Dear Foamers

I want to ask if someone has already deal with interface mismatch in the solids4Foam toolkit.
I have applied the "Warp By Vector" filter in ParaView on the solid region to show its deformed position, and "poindD" is selected as the vector by a scale factor of 1.
Please find the attachments, the red cylinder is the solid domain and another part is the fluid domain. Moreover, the first picture shows the Geometry and Grid in the initial time, the second reveals the deformed geometry in the steady-state solution. While there is no gap at the beginning, the gap is widening over time.

configuration of this test case are as follow:

Code:

FSI fluidSolidInterface    IQNILS;

interfaceTransferMethod GGI;

solidModel     unsLinearGeometry;

The FSI loop is converged very well with residual of 1e-6.

Regarding the numerical results and experimental data, I think the deformation of the solid domain is approximately correct but the fluid boundary has not moved correctly.
Any comments or suggestions will be greatly appreciated.

Best Regards,
Peyman

Warp by the vector D

Quote:

Originally Posted by ManuelOliveira (Post 767941)

Warp by the vector D

I check it again, I can warp by poindD, pointDD, and U (of solid domain). I cannot find D for warping.

Also, as it has been mentioned in another thread, pointD could be more accurate than D.

Hi. I did a bunch of test on the FSI with the solids4Foam toolbox and I noticed that the other factor has to be set is the number of outerloops. and relaxation factor

The weird thing is that if you set the number to high number the fsi loop is struggling to converge. Here is the example of my residuals output

You can onvesitigate the fsi residuals by putting

writeResidualsToFile yes;

The. residuals folder in processor0 will be created.