Limitation of mass spring model and continuous model
- - Does mass-spring system converge toward a unique solution when sampling increase ?
-
- \(\Rightarrow\) No :(
-
- Depends on the connectivity \(\rightarrow\) bad for physical accuracy
- Corollary
-
- - Mass-springs work well for grid-mesh structure (draping)
- - Less for arbitrary triangular meshes
- 1st improvment: Change toward energy formulation for bending springs (limits locking effect)
-
- \(\displaystyle F=\frac{\partial E}{\partial p}\)
- \(\displaystyle E=\frac{1}{2}K\,L\,\kappa^2\), \(\kappa\): curvature
