fabric material with large deformations

Hello,

I want to simulate the fluid-structure-interaction between a paraglider and the air. The wing is made of polyester or nylon fabric, so there is a very large deformation/displacement compared to the thickness of the structure.
I made a very simple test structure to find the appropriate element type and settings in code_aster to approach this task (chaining or coupling with code_saturne will follow later on):
It is an empty cube, except that one side is open(missing face). There is pressure applied on the inner walls and the structure is only held at the edge of the opening (no movement in x, y or z direction). So the structure should sort of inflate and reach a stable condition.

As far as I understand the documentation and previous posts I found, COQUE_3D should be the best approach. I tried some (non-linear) static and dynamic simulations with different deformation hypothesis, but as a result, only the (initial) plane faces are inflating. The edges remain the same and are not moving at all. Is it due to the fact that COQUE_3D only supports large displacements and rotations, but no(?) deformations at all or did I miss something?

Should I approach this case with another element type?(Maybe '3D'?) Any suggestions how to get close to the behavior of fabric in code_aster?

Thank you in advance,
Adrian

Hello Adrian,

first of all, this seems like an ambitious and interesting project! I did some time ago a FSI coupling between Code-Aster and OpenFOAM,
so I will be watching closely how this will evolve...

COQUE_3D supports large displacements, but only small strains, which should be ok for your situation. BUT the nylon fabric behaves more like a MEMBRANE than a COQUE_3D, since it has (almost) no bending stiffness. This effect is not possible to achieve with COQUE_3D as far as I know and MEMBRANE has no large displacement formulation available.

Perhaps there is some way around:
- use COQUE_3D modelisation and create A LOT of small patches where the translational DOFs are coupleded between the patches, but the rotationals aren't. Then you might get a behavior that is similar to a membrane (more like a chain mail, but still better than a plate). There will be a lot of python programming involved to achieve this efficiently, but it might work.

Best regards,
Richard

Hello Richard,

thank you for your reply and suggestion.

I hoped there would be an easier way within code_aster to achieve this. I am not limited to code_aster, so I will take a look at some other open source FEM-Solver. Maybe there are some with integrated fabric-like material behaviors. I will keep you informed about the progress if you want.

Best regards,
Adrian

Dear Adrian,

You could use MBDyn for your analysis.
MBDyn has membrane and shell elements for large deformations. Although only linear elastic constitutive laws are supported.
It is possible to couple MBDyn with CFD solvers like OVERTURE .
I think that paper could be interesting for you.

Best regards,
Lagrange