In my experience with these grippers, I would say that it is better to start off with the right material from the beginning because you may be chasing an issue that is not there in reality. However, that is just my experience.
If you are only looking for deformation and filling ratios, maybe you could take a shortcut and simply load the grippers normally instead of moving the flange tangentially. That would not give you a meaningful insertion/extraction force, but you could then check your filling ratio.
It would be nice to better understand the real setup to have the model built with the correct 2D assumption. I was thinking it was a long extrusion, hence, plane strain.
I did some changes and I got the first complete running simulation!
What helped to get it running after the geometry change:
Switch from 3D to 2Dax
Contact Pair: reduce contact surface
surface interaction: friction from 0.01 to 0.5
Step: change from Automatic to Direct (is not running with automatic, dont know why)
Material: change material model to Ogden
BC: change parameters of the movement distance in the calculation by try and error. For example it failed with 4mm but was successful with 4.5mm. Very sensitive
Now I will return to the old design version and check it again.
But I would say material and change to 2Dax was the biggest advantage.
This is important assumption. You could also try pseudo-axisymmetric 3D model (wedge/segment), maybe even with cyclic symmetry.
This is definitely a good practice, but to reduce the model complexity rather than improve convergence.
Usually, larger friction is worse for convergence unless friction is needed to eliminate some rigid body motion.
Strange, usually automatic incrementation is sufficient and better (allows larger increments if convergence goes well) unless you use very low increment size for direct incrementation.
This is probably the key change here. With limited or no test data, you should use the simplest hyperelastic models. But of course, their constants are the key.
Position of the pusher with respect to the lip should be more important. Normally, larger range of motion should be worse unless it jumps through some difficult regions when using direct incrementation with different BC magnitude.
But in general yes - material properties and geometry are the key aspects. This case is very difficult for the mesh and such modifications may reduce its distortions.
