by zhouxinwei » Thu Aug 04, 2016 1:20 am
We have discussed possible implementations for a more general triangle face - node constraint. The complexity of these implementations depends on what we are trying to accomplish. So we some questions.
For clarity, let's assume that we are dealing with
1. a general surface consisting of triangular facets
2. the node numbering of the facets is such that we can define a positive normal direction for each facet and the normal are consistent
3. if there is material on one side of the surface, the normals are oriented away from the material
4. the 'slave' node is initially bonded to the surface, but that bond can degrade during a simulation.
Here are the questions:
1. Are we dealing with large deformations? In other words, do we want the 'slave' node to move from one facet to an adjacent facet, when it 'gets out' of the facet is on.
2. Are we forcing the 'slave' node to move on the surface with a bilateral constraint, or do we also want to handle the case where the constraint is unilateral (it can never penetrate the surface but it can move away from it once the bond is broken) ?
3. In the second scenario of 2., do we want to address the condition that a node that has moved away from the surface may come into contact with the surface later on in the simulation?
The level of effort for this task depends on how general we want to make this formulation.