Karamba Cross Section Optimization

Hello everyone.

I'm new to Karamba and I'm wondering if it is possible to optimize the cross-section of a shell, but not by picking from a list of standardized cross-sections.

The picture below is that of the attached example file from Karamba.

I have three questions:

1) How can you visualize the chosen cross-section from the Optimize Cross Section component?

2) Can Karamba perform an optimization on the cross-section, in order to obtain a freeform cross-section?

3) If yes, can Karamba also vary this freeform cross-section along the length of the bridge?

Attachments:

02_ShellCrossSectionOptimization Forum.gh, 40 KB

Replies to This Discussion

Permalink Reply by Karamba3D on August 12, 2016 at 5:24am

Hello Erik,

ad1.) You can visualize the cross section thicknesses by enabling 'Cross section' in the 'BeamView'-component - see attached file.

ad2.) The cross section optimization algorithm makes its choice from a user defined set of cross sections. The larger this set the less constrained the optimized result.

ad3.) Yes, see attached file.

Best,

Clemens

Attachments:

02_ShellCrossSectionOptimizationForum_CP.gh, 41 KB

Permalink Reply by Erik Pelicaen on August 14, 2016 at 4:09pm

Thank you Clemens.

Is it also possible to visualize the cross-section volumetrically, instead of showing colours on the surface?

Permalink Reply by Erik Pelicaen on August 15, 2016 at 11:21am

Actually I still don't see how you know which cross-section was chosen from the Optimize Cross Section component. In the model output of this component I can see 'Cross sections: 4.' How should I interpret this?

Permalink Reply by Karamba3D on August 16, 2016 at 2:44am

Hi Erik,

if you simply bake the mesh when the Cross Section button is selected in the 'ShellView' component is selected, the mesh will bake with with the shell thickness. You can disassemble the model and then extract the cross section heights for all the mesh faces of the shell too.
Please see the attached file.

Matthew

Attachments:

02_ShellCrossSectionOptimizationForum_MT.gh, 45 KB

Permalink Reply by Erik Pelicaen on August 16, 2016 at 3:33am

Ok thank you Matthew.

Sorry I still have some questions. The output says RO: 'RO114.3/4' - Germany.

Is that a type of cross-section? I can't find it anywhere in Google.

It also says isotropicShell: 'defaultShell' height: 1 [cm]

Does this mean an isotropic shell was used for the calculation?

Also, as the mesh is triangulated, do the different heights of the faces correspond with their maximum thickness? Or their average thickness?

Permalink Reply by Karamba3D on August 16, 2016 at 5:25am

RO114.3/4 is a circular hollow cross section with diameter 114.3mm and 4mm wall thickness. It is used as the default cross section for beams and trusses.

The default shell cross section has a thickness of 1cm and is assumed to have the same mechanical properties in all directions (thus isotropic).

The definition of shell cross sections lets you define a different thickness for each triangular face. Therefore there can be multiple shell thicknesses at mesh vertices.
Best,

Clemens

Permalink Reply by Contestables on August 29, 2018 at 10:18am

Hi Clemens,

I'm using the same procedure for my shell. I have the legend with the different thickness values but my shell has just one color and indeed it looks like has one value of thickness throughout the structure or a subtle variation. The values are very low and I don't think they are coherent with all the process.

Moreover, I baked from shell view component and then I compared the values from Rhino with the legend and they are different. (In Rhino they are much lower)

I'd like to understand the reason of this thing.

Thanks a lot,

Marinella

Attachments:

screenshot ottimizzazione karamba.PNG, 2 MB

Permalink Reply by Karamba3D on August 30, 2018 at 1:35pm

Hi Marinella,

without the GH-definition it is hard to say what the reason for the uniform thickness might be.

What happens if you refine the shell mesh? Do you see a variation of utilization throughout the elements? Could it be that the smallest cross section is already sufficient?

Best,

Clemens

Permalink Reply by Contestables on September 5, 2018 at 11:56am

Hi Clemens,

this is the GH definition.

Sorry, but what do you mean for refine the mesh? Do you mean the subdivision?

Thanks a lot.

Marinella

Attachments:

ottimizzazione Karamba.gh, 44 KB

Permalink Reply by Karamba3D on September 5, 2018 at 12:46pm

Hi Marinella,

yes, try to increase the subdivision of the mesh.

I was not able to run your definition, because the initial surface did not load. It would be best, if you internalized the geometry which results from the form-finding, so that the definition gets more compact for testing.

Best,

Clemens

Permalink Reply by Contestables on October 26, 2018 at 9:28am

Hi Clemens,

sorry for my late reply but now I internalized data and I would like to understand it the process is correct and how I can check the variation of thickness.

Your help would be very appreciated.

Thanks,

Marinella

Attachments:

optimization karamba.gh, 64 KB

Permalink Reply by Karamba3D on October 29, 2018 at 9:41pm

Dear Marinelle,

for unknown reasons the Grasshopper-website did not let me respond directly to your post below.

Could you simplify your definition so that it does not contain references to plug-ins other than Karamba3D? That would make it easier for me to check it.

Best,

Clemens