Model a beam using Grasshoopper+Kangaroo

Is it possible to model a line/cuboid to behave like a beam?

I want to introduce the load as applied on a real beam from a load-cell and feed it to grasshopper using firefly.

Or in place of load, I can also input the displacement at a point of the real beam using an LVDT and feed it to grasshopper and my model beam in grasshopper should deflect as a whole.

For this I need the ability to define atleast two points as supports and define the degree of fixity at these supports.

I read the kangaroo manual but wasnt able to do it based on one reading. Will post my work, if I get something done.

Replies to This Discussion

Permalink Reply by Chris Tietjen on December 24, 2012 at 8:31pm

This gives an approximation of bending in a simple beam. It could be modified to allow for the shortening between support points that occurs when a beam deflects. It could also be used to show real time deflection with a suitable sensory input but the accuracy of any values would have to be calibrated beyond what's being done here.

Attachments:

9_KangaSimpleBeamApprox.gh, 14 KB

Permalink Reply by Chintan Pathak on December 28, 2012 at 12:01am

Thanks Chris,

This is exactly what I needed.

I extended it to now simulate for two points on a rectangular beam. The simulation looks fine.

Permalink Reply by Chintan Pathak on December 28, 2012 at 12:05am

Next, I tried to input geometry defined value of stiffness for the spring i.e. E*A/l, for different finite elements in this case can have different lengths and hence different stiffnesses.

The simulation crashes, or doesnt output at all. What can be done to resolve this ?

Attachments:

9_KangaSimpleBeamApprox2.gh, 11 KB
9_KangaSimpleBeamApprox2i.gh, 12 KB

Permalink Reply by Chintan Pathak on December 28, 2012 at 12:15am

What I want to try next is to verify the simulation results with the exact results.

So for the 3-point bending problem, for the real steel beam I have, I will apply some fixed value of load (say 5kgs) and read the displacement value (at say the mid-point) with an LVDT, and compare this displacement with what the grasshopper model.

Can I view the physics/code behind Kangaroo, to see if it resembles anything like the theory of elasticity we read in Timoshenko ?

Permalink Reply by Robert Vier on December 28, 2012 at 3:21am

have you heard about karamba for gh? it is parametric finite elements and a.o. gives you the deformed shape / the deformed mesh of the cross section, with an incremental approach (no theory of 2nd order so far) also for larger deflections.

RSS