Optimizing of cross-sections in Karamba

I want to optimize my truss, as much as I can.

By optimizing I mean, I want to use the best circular (hollow) profile diameter and it's thickness for particular members. For example, I want to different circular (hollow) profiles, for different members (one for compressional, one for tension, and one for null members). I am using Galapagos solver to minimize the displacements as much as it can.

I am using Grasshopper 0.8.0050 and Karamba 0.9.07.

Can you edit my file, in order to show my what is wrong with it?

Karamba 0.9.07 can be downloaded from here:

http://twl.uni-ak.ac.at/karamba/index.php?section=download

And Grasshopper 0.8.0050 from here:

http://www.gamefront.com/files/21702580/grasshopper_wip_20110421_NO...

Thank you for the reply.

Attachments:

proba.3dm, 37 KB
proba.gh, 18 KB

Replies to This Discussion

Permalink Reply by Robert Vier on May 15, 2012 at 1:08am

hey Djordje!

first of all - the twl.uni-ak.ac.at karamba page is outdated- please refer to

www.karamba3d.com

where the latest karamba release 0.9.0.84 is available in different licences.

in the pro version (and also in the 20-member trial version) there is a component doing efficient cross section optimization. for its usage and settings please refer to the manual first.

kind of a workaround is an earlier definition with hoopsnake, doing rough, simple iterative cross section design by moments and normal forces. http://www.grasshopper3d.com/group/karamba/page/example-files

best

Robert

Permalink Reply by Najna Sorin on May 16, 2012 at 4:46am

Djordje your file can not be opened with 0.9.0.84 release.

edit:

sorry I saw that you are using 0.9.0.7

Attachments:

proba.3dm, 37 KB
proba.gh, 18 KB

Permalink Reply by Robert Vier on May 16, 2012 at 10:38am

Hey!

If you want to define cross section types according to axial normal forces you should make sure not have too much bending moments in your elements .. like in a truss system.

To assign the cross section types, you have to analyze your model first, otherwise you of course dont know where the respective forces occur. With the 'resultant section forces' component you can determine in which items of your list of beams the axial force is either below, similar to, or above zero, split the list accordingly and modify the beams separately (how to do this there are several ways , but this is an easy an clear one).

best

robert

Permalink Reply by djordje on May 16, 2012 at 10:53am

Hi Robert.

I did the same thing - I separated the resultant forces based on their values (+ or -).
Still I do not know how to assign a separate cross-sections to each of the members.

Take a look:

Help?

Permalink Reply by Robert Vier on May 16, 2012 at 11:31am

You have to build up a model a second time, after you analyzed it firstly. To do this, use 'disassemble model', and there pick the respective beams that are tension or compression, modify them then assemble a new model.

cheers

Permalink Reply by djordje on May 16, 2012 at 12:23pm

Any chance of making a quick definition for that?
I think in that way it would be much easier, as I did not quite understand you.

RSS