Rope Dynamics Test

OK, here's a rope challenge. In order to work out the support structure for the organic shapes in his buildings, Antonio Gaudi would suspend a string model of whatever building he was working on from his ceiling, upside down. The natural curve the string found between two fixed points helped him estimate how much reinforcement his rafters would need: if the naturally dangling curve suited him, it could be a lightweight roofing structure, if he wanted a different shape for the curve, he could create it by adding weights to strings, and the position of the weights helped him calculate the angles and support he would need . The models he built are gorgeous:

Gaudi string model.jpg

I haven't worked with Cheetah's rope much: is there a way to hang a second rope to middle of an existing rope? My early tests make it seem anchors can be targeted to entire rope objects, but pointing to specific points on a rope seems to be a no-go. Would I need to use a lightweight rigid body shape as the intersection between them, with both ropes having a target point tied to it?

(I saw a reference Frank made to a couple of Cat's Cradle discussion threads that I hope to dig up soon.)
 
I've done some testing of the interaction between Rope Tag and Rigid Body dynamics and
although there is some interaction, the rigid body ends up passing through the "rope".

So using a rigid body as a weight did not work.

He was working outside the box, but this is a whole different box.
 
Nice Frank, I didn't try to anchor the Rigid Body to the "rope" yet, great work. :oops:

I did discover something interesting using a Bezier Spline with a
Rope Tag and and a Rigid Body to catch the spline points.

bezier_rope_trick.gif


Bezier rope trick setup.jpg
 
That´s the reason to avoid Béziers in conjunction with rope-tags - but who said you can´t do other cool thinks nonetheless?
Haywire.gif

Haywire2.gif

Cheers
Frank
 
Last edited:
I saw the hanging lights example, which is what gave me the idea of the "intersection" object. My experiments with intersecting ropes didn't work well, but I'm new to this part of C3D and could very well have set them up wrong.

I notice that in the two examples Frank posted, the join point between the lines appears to be fixed: I'm hoping to attach the "T" segment to a point in the original line which dangles. I suspect the math involved would be horrendous.
 
I'm hoping to attach the "T" segment to a point in the original line which dangles.
As I said this works just well with linear splines because the anchored points will stay in place while all other types will be interpolated to some stage and "leave" their "constructive" points.
Good luck.

Cheers
Frank
 
When you get right down to it, I'm not really interested in animating these forms - animation is only necessary to let the ropes find their natural dangling shape. I'm chasing the ability to create a computer generated curve that has more in common with a parabola than a semi-circle (or a stretched semi-circle). these things can be built by hand, but then I'd be relying on my eyes and my instincts, instead of what physics would provide.

---And after some experimentation and careful planning, the intersection theory seems to work!

Gaudi strings A.jpg
 
:unsure: @MonkeyT: I assume that a single jump of the bounce Java script shows a parabola.
PS There is also a difference to a catenary, but the 1 jump bounce is a spline with 3 points, so a miniscute tweak is no problem if you get a template.

Screenshot 2020-05-22 at 19.04.12.png
 
Back
Top