Negative jacobian (shell elements & Tri3)

[axel3969]

Hello,

I'm new to FEBio and FEBio Studio. I am trying to run a static simulation of a rectangular shell, fixed on one side and tracted on the opposite.
The problem is that I get a Negative jacobian error.
I tried to reduce the thickness, increase elements size and modifying the nodal force magnitude but nothing correct this.

I read many topics about negative jacobian issues but nothing helpful for my case. Please can you explain me the problem here ?

Thanks,
Best regards,

Axel

[maas]

Hi Axel,

I would like to suggest a few changes to your model:

1. The "zero rotation" boundary condition has no effect on the shells. (This is a left-over from an older shell formulation.) For the default shell formulation in FEBio, you need to fix the "shell displacement" degrees of freedom. (In addition to the regular displacement degrees of freedom.)
2. Your nodal force does not have a load curve assigned. It might be because you are trying to achieve the load in one step, but the load you are applying is very big for the given material parameters and shell thickness. (Keep in mind that a nodal load applies that load to each of the nodes of the assigned selection.) I recommend using a load curve for this load.
3. I also recommend using multiple time steps.

Even with all these recommendations, the model doesn't run for the given load. I was able to get it to run for a lower load and have attached my modified model file. Please take a look and let me know if you have any questions.

Best,

Steve

[axel3969]

Hi Steve,

Thanks a lot for your answer.

I looked at your file and increased the shell thickness and the nodal force. Then, I observed that the shell not only elongate but also bend. When I check "shell_bottom" I observe a bending in the opposite direction. Therefore, I assess that it is due to the application of the Nodal force. Any way to apply the force on the "neutral line" and not on the "top line" or "bottom line" ?

Best regards,

Axel

[maas]

That's an interesting observation; I had not thought of that. I don't think there is a way to apply a force to the "neutral" line of the shell. You could consider applying two loads with half the value, one with and one without the "shell bottom" option checked. I think that's the closest you will be able to get.

Best,

Steve

[axel3969]

Hi Steve,

Thanks for your answer, I obtain the expected behavior.

Best regards,
Axel