I am trying to perform cloth simulation, but I am struggling with the parameters. Specifically, since it is uses the stiffness, two hanging cloths with the same mass and same stiffness have different deformations based on the number of particles. How can we have consistent deformations independently of the number of particles, like using the Young module? (Obviously, I expect that more particles equal a more accurate simulation)
Hi @giorgio.nicola,
Unfortunately the PBD particle solver we use for the cloth is not accurate in that respect. More particles lead to worse convergence. You can try to fight that to a certain degree using higher iteration counts and smaller time step sizes.
We will introduce a new deformable surface type soon that will be more realistic, featuring Young module as you mention, but even there it will be crucial to tweak iteration counts and time step sizes across different mesh resolutions.
Cheers,
Simon
Hi @SimonPhysX thanks for the quick update. What it concern me most is that in complex meshes the particles are likely to be not equally distributed. In particular density will be defined only based on the geometry complexity. Thus there will be accurate patches and inaccurate patches depending on the local density of particles, inaccurate patches might be where more stresses are applied to the cloth.
I suppose it would be better to manually generate the meshes to tweak the density of particles where the more strain is expected, similarly to FEM analysis.
Have you already planned the release date for the new deformable surface type?
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.