Grasshopper

algorithmic modeling for Rhino

Hi, 

I not sure I understand the function "Make Beam-set" very well. 

Can I use it for the following problem? 

I have a main group of beam elements to analyse. 

For some of these beams I would like to add some hinge points at the starting node (set A), for some at the ending node (set B). 

But there are beams of the main group which aren't in either set A or set B, and there are beams which are in both set A and B. 

This makes that I can't link both set A and set B beams in the Elem input of the Assembly component, because then there are some beams double in the list and some not even in the assembly list.

But if I just link the beams of the main group into the Elem input of the Assembly, how do I take into account that there are two sets of beam for which I want to add beam joints? In a way that the systems knows that the two beams sets are part of the main group, but that you only have to analyse it for the clean group of beams of the main group? 

Thanks!

Views: 940

Replies to This Discussion

ok, i think i somehow got what you mean.

probably it is better to start off with one straight list of beams that you feed into the assemble component. you can then assign beam joints by their indices in this list: at the beam-joint component, you feed in the indices you want to be worked, not the element-ids.

those indices you have to distinguish first into three groups: those with joints at the start, the end, and on both sides .. maybe just make your lists for start and end first, then filter out those indices that appear in both lists.

i dont know if it is possible to assign joints to beam-sets - clemens just thought it was possible, you have to try! but anyway, the strategy remains similar!

best

Robert

Okay thanks, this is really helpful. 

Just this: 

For the beam indices, can't I make just to groups? 

One with joints at the start and one with joints at the end. Is it a problem that some indices are in both groups? 

Or is it just no good to release all rotations at both ends of a beam? 

Thanks!

in principle you can release all rotations on both ends, except the x-rotation - this you can do only once, since the beam then for sure could rotate around its x-axis (the longitudinal one). the others depend on how the structure looks like.

for testing these kinds of things i suggest supplying a vector for the stiffness of the joints into the 'Beam Joint' component. you can alter the stiffnesses then, and see how your structure works when they become weaker. the advantage is that karamba then calculates at all, instead when having 'real' joints so the structure is kinematic. so even if it was kinematic with full joints, you cheat a bit for debugging and see the tendencies.

as for the technical part of supplying the beam indices - you can assign just once, karamba does not do that for you (for now). if you specify an index twice, the later appearance will be taken for the joint-assignment.

Thanks a lot for this clear information!! This is very helpfull!!

Hi both of you,

I've been reading the discussion but I still have a problem. I had trouble making "real" hinges in the joints of my beams. So I did has afviced by Robert and put an "artifical" rotational stiffness really low. However the bending moment in curves do not converge to zero (as their theoretically should do).

Is it a bug of Karamba ?

Thanks in advance

Thomas

PS: It's a bit messy but here's the gh definition

Hi Thomas, 

I didn't look at your script in detail, but from the picture above it seems that you are calculating a truss-like system. Is this correct? 

If so, I think you want to introduce these real hinges at your beams because you want them to act purely axial (no bending), as we know truss-systems. No?

If so, there is a better way to to this then introducing hinges at the end/start of the beams.
There is a component "Modify element". This element as an input "bending". If you put this boolean to false, your system is modified to a truss..

Does this help?

Best

Lara 

Oops forgot the file.

Here it is.

Thomas

Attachments:

RSS

About

Translate

Search

Photos

  • Add Photos
  • View All

Videos

  • Add Videos
  • View All

© 2024   Created by Scott Davidson.   Powered by

Badges  |  Report an Issue  |  Terms of Service