algorithmic modeling for Rhino
Hi everybody, I'm experimenting with WoolyPath algorythm where each line has thickness. It wasn't a problem to do it in Kangaroo using SphereCollide component. In this case all the pipes have similar radius (WoolyPaths_OneRadius.gh). But I would like to have same effect but with different thickness of pipes. I tried to provide SphereCollide component with a list of different radius. It doesn't work (WoolyPaths_MultipleRadius.gh). Can anybody help me?
I think multiple radius is not yet implemented (in July 2015) :
" this sphere-collide component is currently specific to sets of equal radius spheres (as this allows some easy simplification of the collision detection). I do have something working now for accelerated collisions between varying radii spheres, and plan to release this soon."
Probably that means it was implemented for Kangaroo 2? But new components don't work with the old Kangaroo engine.
Great idea, ran into similar trouble trying to vary the force of different forces (power laws and springs) by feeding a list of different values into the kangaroo engine. My guess is that Daniel's interest is in optimization and convergence with his form finding exercises and the solver is written with that interest.
With the old solver running at a slower pace, you can try to make the solver do idiosyncratic things and freeze it at an interesting moment far from equilibrium. That aesthetic appeals rather more than the fully optimized converged solution, which can be elegant doesnt have the same tension as the dynamic moments.
For instance if you were trying to model desires where different types of similar objects have varying preferences of attraction and repulsion to its neighbors and you wanted to throw them all together and see what shakes out in that community of objects, being able to feed in different parameters would be quite helpful. This seems to be what you are trying to do. Please let us know if you figure out anything interesting with this.
I've gotten stalled on that set of design experiments, but there are other implementations of physics models around (vertlet physics for processing, or physics engines in animation software) that might be more amenable to these types of experiments. Kangaroo is so nice to play with since its already embedded in grasshopper and rhino, though.
Thank you for detailed response. I'll let you know if something nice would be achieved.
Which "old solver" do you mean? Kangaroo 0.95?
Does anyone know if this was addressed in K2? I'm trying to use multiple radii with the sphere collide component, and it doesn't seem to be working.