Strange deformation of membrane

1

Hello everyone,

I simulate the deformation of membrane under impact of wave using openfoam and calculix.
As we can see in file “precice-exports”, the force mapping from openfoam to calculix is normal and right. And the displacement on nodes in Calculix is also right.

fluid-mesh-force
fluid-mesh-force1920×1242 38.1 KB

solid-mesh-displacement
solid-mesh-displacement1920×1242 39.1 KB

solid-mesh-force
solid-mesh-force1920×1242 39.5 KB

I check the frd file in ProPoMax and find the deformation is strange, which is shown here.
deformation
The outer arc interface deformed toward the inner direction and the inner arc interface deformed toward the outer direction. But this is a solid. So I think that this deformation is strange.
I think the problem occurs on the setup of Calculix. My setup files are presented here. GitHub - NUGUY2022/Solid. I use the flap.inp file to run Calculix.
I give a simple description about this case. I define outer and inner arc interface as the exchange interface to get the force from openfoam. The low boundaries are set as fixed boundary. And I hope the front and Back interface have no effect for simulation because I conduct a 2D simulation.

Please feel free to giving me comments and advices.

2

I get zero displacement when I solve it. Also a lot of warnings like

 *WARNING in calinput. Card image cannot be interpreted:
 *WARNING reading the input file. Card image:
          860,8,153.5729,-666.7229,-153.945,-1771.911,-6.76E-12,1.24E-10

I had to move flap.inp up one level in the directory tree. Is there anything else needed to get the solution you have? Which version of CCX?

I uncommented **INCLUDE,INPUT=Solid/boundary.inp in flap.inp and got a solution. Maybe your solution is without the displacement boundary conditions?

Untitled
Untitled864×642 15.4 KB

3

Hi vicmw,
I am doing the coupling simulation using openfoam and Calculix. Calculix get the force from Openfoam. If you just run Calculix, you will get nothing. Because there is no force data input from openfoam to Calculix. In my situation, the membrane is under the hydrostatic pressure. The water depth is 0.5m and density of water is 1000kg/m3. Do you think my deformation is right?

4

I’m confused. Is the line I uncommented the forces from Openfoam? I don’t know if that deformation is right or not but it’s different from what you got so I might be looking at the wrong problem.

Are you looking for a way to verify the results?

5

¿Where is the surface?. ¿Is the membrane submerged or just one side is under pressure? ¿Which side?.

1 Like
6

Membrane is submerged. So, both sides suffer from the pressure. Outer-side suffers from pressure from the Fluid-outer. Inner-side suffers from pressure from the Fluid-inner. I think my deformation is right. I find that the deformation is scaled by a parameter 4.6e7, which looks like so large but small in reality. Sorry, this is my visual illusion.

7

Your result is right. But you just simulate the deformation of one time step in my FSI simulation. Because what you uncommented is the initial displacement. OpenFOAM will transfer force to Calculix every time step. My results include several time steps. I think my result is right because both sides suffer from hydrostatic pressures. The deformation scale makes me think that the deformation is strange. It is just a visual illusion.

8

maybe, i need to look into details, since force magnitude is gradually changed (singularity) at the support from lowest to highest.

9

Did you try simulating this using the imported pressure load in PrePoMax first ? This feature allows you to simply convert the pressure from OpenFOAM analysis to structural load without going full FSI.

Also, try in 3D for now.

10

I have solved my problem. The problem is due to the boundary condition of OpenFOAM. Now I am succeeded to run FSI simulation of a pre-stretched membrane.

11

From your files it seems you are using C3D8 elements, that is first order bricks.
These will probably be too stiff.
Try switching to C3D20R. Those will give a more realistic behavior.

See the “Golden rules” section of the “Calculix Crunchix user’s manual”.

12

Thanks for your advice. I will try to use C3D20R to get a more realistic behavior and learn more about Calculix.

13

Hi COCO,

I am wondering how the pre-tension is applied in the pre-stretch membrane. There is a boundary condition applied using boundary.inp file. How do I get this boundary condition file? It involves many nodes and I cannot enter them one by one manually. In addition, the displacement that needs to be applied to the specific nodes is also unknown to obtain the desired pretension. Look forward to your rely. Thank you.