mel08
Bioengineer
- Sep 24, 2008
- 22
I have an issue with contact. I have a screw thread surrounded by highly porous bone, and I am trying to model tightening of the screw.
I have run my model with a fully bonded contact condition & the model solves without convergence issues.
I now am trying to simulate friction with a friction coefficient of 0.1.
The simulation is solved as a static structural. To start with, I am just trying to rotate the screw 10degrees over 10 substeps.
I have set up a local co-ordinate system & the BC's are correct because it behaves properly with the to surfaces bonded.
With friction, it will solve the first substep & then cannot converge on the second.
I have checked the contact status with a zero load (ie before any substeps) & after the first substep it will solve and there doesn't seem to be any standout problems with the contact.
I have set FKN to 1.0 (with lower FKN i have errors stating too much penetration). & pinball region set to 0.02 constant value (element size is 0.06). Stiffness is set to update each iteration, with the augmented lagrange algorithm and contact detection by nodes ( should this be the Guass points?). I have set the initial adjust to exclude everything.
Any suggestions would be much appreciated.
Thank you
Melissa Ryan
PhD Candidate
Biomedical Engineering
Flinders University
South Australia
I have run my model with a fully bonded contact condition & the model solves without convergence issues.
I now am trying to simulate friction with a friction coefficient of 0.1.
The simulation is solved as a static structural. To start with, I am just trying to rotate the screw 10degrees over 10 substeps.
I have set up a local co-ordinate system & the BC's are correct because it behaves properly with the to surfaces bonded.
With friction, it will solve the first substep & then cannot converge on the second.
I have checked the contact status with a zero load (ie before any substeps) & after the first substep it will solve and there doesn't seem to be any standout problems with the contact.
I have set FKN to 1.0 (with lower FKN i have errors stating too much penetration). & pinball region set to 0.02 constant value (element size is 0.06). Stiffness is set to update each iteration, with the augmented lagrange algorithm and contact detection by nodes ( should this be the Guass points?). I have set the initial adjust to exclude everything.
Any suggestions would be much appreciated.
Thank you
Melissa Ryan
PhD Candidate
Biomedical Engineering
Flinders University
South Australia