Hello, everyone,
I have a contact mechanics problem to solve. The problem is a thin rod (flexible rod) needs to navigate a rigid 3D tube (e. g, endovascular vessel). A large deformation and small strain was assumed for this problem. An incremental method was used to find the solution. Newton method was applied to do iteration. As far as boundary, a penalty force (F=Kx) was used to make sure the configuration of rod is inside the boundary. The specialty of this problem are:
- the incremental load is not Applied force, instead, it is a known incremental displacement (either an incremental displacement or a rotation angle) is applied at the end of thin rod. So there is a difficulty to compute the imbalanced force, since the applied (external) force is deformation-dependent force and computed by penalty force, which obvious has big error compared to the real contact force.
- I used the force imbalance as a convergence criteria, however, by using a absolute normal value of imbalanced force seems no sense. If I use a ratio, it is very hard to get converged
- Since the penalty force is deformation-dependent, the number of contact points changes back and forth (jagging) between iterations, especially when multiple contact points occurs, so it is very hard to get converged in a few iterations.
My question is:
- Is there any better convergence criteria than imbalanced force?
- Any suggestions for dealing with jagging phenomenon?
Your help is really appreciated.....
Helen