When I create the tetra mesh using the solidembeded node, it treats the two object as one object:
So instead I use a for each loop to create separate tetra meshes:
Then these meshes are animated and I want the solid object to follow the animation as closely as it can while colliding with it's environment. To do this, I pointdeform each tetra by it's animated geometry inside a for each loop again and I use that result as the Target Deformation in the Solid Object DOP.
That works ratter well as each tetra object are simulated like separate objects :
The problem is when the dopimport_embedded deforms the actual mesh, it stretches the parts that were overlapping:
It looks like all tetras will deform all meshes as if they were one object instead of treating them separately. Is there a way to split them up so each tetra deforms only it's associated mesh instead?
Now these are two simple meshes to illustrate the problem, but my actual scene is full of intertwined meshes. I thought of doing a pre-roll process where we make sure each object is never overlapping and they slowly settle back into the start position, but because of the nature of the geometry being intertwined (imagine a ton of electrical wires crisscrossing), it's very difficult to animate a pre-roll that would settle to the desired shape. That's why I'm hoping I can avoid this by figuring out a way to deform each mesh one by one using it's designated tetra mesh instead.
Also a side question: Assuming the above scenario can be done, is there a way to make the finite element simulation push out overlapping tetra so they find a stable state with no overlapping at the beginning of the simulation? Right now, the simulation doesn't seem to care that they are overlapping until they separate and collide again later on.
