Grasshopper

algorithmic modeling for Rhino

Structural output shows errors

Hello, 

If gravity (defined by the material and cross section) is not included as a load case (or it has a value of zero), I thought that the values of the section forces were only depending on the external loads acting on the structure.
But if gravity is zero and I change the material or the cross section (while the external loads are unchanged), than the values of the section forces change as well.
The difference is not much, but it is there. How is this explained?

 

I made two files to address the gravity problem.
The first test file is based on a simple supported beam. Here, whether the gravity is on or not, you don't see any change in the values of the section forces of the load case (LC1) with the external load (when you are looking at the lists which are the output of the section force component), if the material or the cross section is changed.
And this is how it should be because these list list the values for each loadcase, and the values of the loadcase with the external load don't take into account the gravity (hence material and cross section). 
 
The second file is a smaller definition of the main file I have been working on. 
It is with beams and springs, corresponding to a deployable scissor structure. 
To use this file, you have to use the corresponding rhino file. You have to set the curves (in the grasshopper file, indicated with the arrows). In the first set of curves (GH) you set the curves of layer 1 (RH), the red ones, and please do so by selecting them clockwise. In the second set of curves (GH) you set the curves of layer 2 (RH), the purple ones, again selecting them clockwise. 
Normally the Karamba output should appear now. 
In this file the values of the section forces of the loadcase with the external loads (LC1) (output list of the section force component) change when you change the material (steel or aluminium) and the cross section. 
And this is not correct, because these values should be independent from gravity, whether gravity is on or not, right? 
Is it because I am using springs (and their related cross section and material)? 

 

Thanks for the help!

Best

Lara

Views: 551

Attachments:

Replies to This Discussion

Permalink Reply by Robert Vier on October 16, 2012 at 6:32am

Hey Lara!

Thanks for the well formulated question :)

Clemens, the main developer of Karamba, is not here for the next four weeks but I hope I also can help you with this quite technical issue.

In general - statically determinate structures are not dependent of material properties, but so are statically indeterminate. Meaning not only an applied gravity force derived from the material density, but also material stiffnesses influences internal forces and displacement of a statically indeterminate model. I think your scissor-structure is statically indeterminate, and therefore material has an effect even if no gravity is applied - opposed to the simple bending beam.

In karamba, springs do not use a specific material - even if you could supply one like you did, the stiffnesses are given explicitly when you create the cross sections. springs in karamba are partially abstract models that can represent more sophisticated parts. Hence, there is no gravity force applied to them also.

Meaning, that in your scissor example the varying sectional forces come from beam stiffnesses that vary with the change of material.

Do you agree?

Best,

Robert

Permalink Reply by Lara Alegria Mira on October 16, 2012 at 7:04am

Hey Robert! 

Thank you for your quick response. 

I don't really understand the thing about the statically (in)determinate structures you explained. 

In my opinion you can make the example of the simple bending beam also statically indeterminate (adding more support constraints than statically necessary), and in that case the section forces of the external load load case still stay the same when cross sections and material are changed (which is not in the scissor example). Only the section forces under the gravity load case change (if not zero), which seems right.

Or am I seeing it wrong? 

Best

Lara

Permalink Reply by Robert Vier on October 16, 2012 at 7:54am

I could imagine you cannot see the changes for an example simple as this one (the beam). linear FE-methods rely on assumptions for small displacements, so I think the material-influence for a beam is rather small.

Further, FE on a normal computer issues numerical precision problems, respective the 'conditioning' of a stiffness matrix that is inverted with FE (meaning the range between lowest and highest values and their distribution within the matrix). this, besides the typological option from above, could also be a reason why you dont see changes for the simply supported beam with increased constraints.

maybe a cantilever supported with a column would be better suited..?

we tested karamba quite extensively, it should not give results that wrong from the core. anyway, you never know..

Best

Permalink Reply by Lara Alegria Mira on October 16, 2012 at 8:20am

Okay Robert, thank you..

It is more clear to me now.

Best

LAra

RSS

About

Scott Davidson created this Ning Network.

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