Cantilevered Slab Optimization - Branching Pattern


I am looking to optimize the structure of a cantilevered slab.

I have found a way to do this using Millipede, particularly referencing the “004_leaf” example included by the developers.

However, I’d like to use Karamba instead given that the software seems to have a fresh and vibrant support community online.

Is there a way that I can achieve the same branching pattern which emerges through the Millipede example by using Karamba? I have included an attempt at setting up the problem, but would appreciate any and all guidance. It seems that I cannot “Step” in the same way I can in Millipede to allow for the branching pattern to emerge. Is there analogous functionality?

Thank you.


Setup in Millipede and the output (after pressing the Step button on the FE Solver several times)

Mimicking (or at least attempting to) the setup within Karamba.

I have also included my .gh file for reference: (41.5 KB)

Hi @jptroise,
you can use the cross section optimisation to achieve something similar. As I do not have milipede, I am not sure of the loading that you used. However you can refer to the example to see the process of this optimisation. (57.0 KB)

Hello @matttam ,

Thank you for your response and help. Is it possible to show a screen shot of your results, as I do not currently have the full version of Karamba and am thus limited by number of shell elements?

In the meantime, I will work with a low-resolution version of my model in order to understand the optimization process you have laid out. Thanks again.