Kangaroo simulation: cloth pulling a rigid body that needs to rotate

Grasshopper Kangaroo
, ,
1

Hi everyone,

I’m trying to simulate a sim

mechanisme.gh (101.0 KB)

ple mechanism in Kangaroo, but I’m having trouble getting the interaction between a cloth and a rigid body to work correctly.

I have 4 elements:

  • A – the base

  • B – a rail/slider moving along the base

  • C – a pivoting piece connected to B and attached to the cloth

  • D – the cloth connecting C to A

The basic movement is:

  • A and B move along the Z direction.

  • The connection between B and C works correctly.

  • The problem is with C-D and D-A.

As B moves, the cloth D becomes detached from C/A instead of pulling C into rotation. It looks almost as if the cloth does not have enough force to attract the point on C, so C does not rotate as I would expect.

I’m wondering if there is a specific way the different goals should be added to the Kangaroo simulation. For example, is there any priority or interaction between cloth/membrane goals and rigid body goals?

I have already tried changing the Strength values of the different components, but I haven’t been able to get a better result.

Could the problem be related to:

  • how the rigid body and cloth points are connected?

  • the way the hinge/pivot is defined?

  • the relative Strength values?

  • the order in which the goals are added to the solver?

  • or something else in the setup?

I’ve attached the Grasshopper/Rhino file so you can inspect the definition directly. You can move the height slider to test the mechanism and see how the simulation behaves.

Any advice or suggestions would be greatly appreciated. Thanks!

2

One of the issues is that you only attach the cloth to the rigid body corners but then you subdivide the cloth. I would attach all 5 points.

About the strength, cannot explain but I increased it for the rigid bodies. From my point of view points which need to be attached to something or anchored I almost always set the strength to 1’000’000.

image
image2563×1472 340 KB

I currently anchored the hinge but I don’t think that is correct.

On another note, I would use Entwine to gather the goals. Using the regular solver, the tree structure stays intact and you can simply extract the branches with the important geometries…

Also I disabled the solver in the saved file…

mechanisme_5-5.gh (115.2 KB)

3

thanks for fast response Martin , its more beter !!!
but i dont arrive to simulate the tisue like a real live , you can see on real video, the curved surface dont lengthen but in the simulation when you push down the curved surface come more long and when you pull up it become very strech

4

Ok I understand now. Maybe I’ll find some time tonight.

5

image
image1887×954 33.4 KB

image
image1906×1261 75.5 KB

image
image1920×581 144 KB

I think you don’t need the concentricity and align faces goals if you attach the end points of the axis to both parts of the hinge.

I rebuilt the surface of the bending tissue so the mesh subdivision is more regular. I attached the first two lines of vertices to the adjacent parts and used the rod goal. This way the tissue has some rigidity.

Use the slider to move the part. The simulation is running a little slow. Not sure why.

mechanisme_5-6.gh (108.9 KB)

6

woooow thanks very much!! You are incredible.
Merci Beaucoup !!