Karamba - Sandwich beam

Grasshopper Karamba3D
1

Hello,

I am quite new on Karamba. I was trying to model a sandwich beam.
I have created a top and bottom beam element, and to simulate the core, I have made a shell element with the same thickness of the base of my beam.

My model seems to work fine, I don’t get any error. But the result in the displacement differs by 16% with my analytical calculations. I have also tried to make a model with another FEM software, and the result confirms that the analytical calculations are fine, while the Karamba model is not.

What is wrong in the Karamba model? I guess I am missing something…
04.Sandwich Beam Model.gh (51.4 KB)

Thank you in advance,
Marina

2

Hello Marina,
Did you try to refine the shell mesh (e.g. 40 sections in longitudinal direction, 10 over the height)?
Best,
Clemens

3

Hi Clemens,

Thank you for the suggestion… now the result is way more accurate!

Thank you a lot,
Marina

4

Hello,

I have a similar problem. I’m trying to simulate a sandwich beam using Karamba. The sandwich structure consists of 3 layers, each layer should have a different height and material. Currently, I use a shell for each layer since the layers are, compared to the dimensions of the beam, relatively thin.

Unfortunately, I have no idea how to join the stacked layers / shells together. So far, only the core layer (on which the force is applied) shows a displacement.

If anybody has a hint for how to solve my problem, please let me know.

In the end, I would like the core layer to be a shell with a varying height, while the top and bottom layers have consistent heights but follow the shape of the core layer. Does anyone have a hint/idea for how to realize this once the other problem is solved?

Thank you in advance,
Robert
sandwich-beam~0qj.gh (82.7 KB)

5

Hello @robert.frank,

you could try to tie together the nodes along vertical lines using short pieces of beams which have a dummy material of zero weight and act like dowels.
You would have to experiment around with the dowel shear stiffness which is controlled bxy the shear areas Ay and Az and the material stiffness.
–Clemens

6

Is there another possibility of joining these layers together? Or would that necessitate volumetric FEA (on the roadmap?)? I am currently looking into a similar challenge (especially with the varying height).

Thanks,
Rudi

7

Dear @rudolf.neumerkel,
I added this to the roadmap.
– Clemens

8

Very cool, thank you!

How would such a sandwich element work? Would it then also be possible to define variable cross sections?

Cheers

9

Hello @Rudi,
there are two possibilities for implementation:

  • addition of rigid connectors
  • layered cross sections.

For variable cross sections you could divide a beam into small segments and attach changing cross sections to them.
– Clemens

1 Like
10

Did you find a solution ?

11

Hey, it seems to be on the roadmap of Karamba - but unclear, when or if it will be available. I also found Alpaca4D, which supports sandwich cross sections.

I did need the sandwich cross section to do some quick analysis for some basic ideas regarding my thesis, but switched the strategy anyway, so I can’t yet really speak from much experience regarding sandwich elements.

Hope this is helpful, all the best!
Rudi

12

Thank you very much Rudi