Help with convergance of difficult contact problem

[Fr1day]

Hello,

I am working on my PhD project of simulating bolt connection in angle profile.

Model as in this picture

Material:
-Plasticity with 4 points on the curve
-Bolt has stiffer material

Contact:
-Tangential behaviour: penalty 0.3, normal behaviour: Hard.
-surface-to-surface, finite sliding

Example of contact pairs:

Bolt is prestressed - I tried different prestressing options: Bolt load and Interference fit (in contact option)

Abaqus starts to have problems when the bolt edge passes the edge of the hole (the hole was deformed before during simulation). It ends up with convergence problem and stops the simulation giving Negative Eigenvalue error and THE STRAIN INCREMENT HAS EXCEEDED FIFTY TIMES THE STRAIN TO CAUSE FIRST YIELD AT ... POINTS.

gif:

https://image.ibb.co/hPk3Yb/ezgif_com_optimize.gif

I am guessing it may be problem with realising the force when edge of the bold starts to overhang. Could you suggest any changes to this model to solve it and help converging behaviour? Also could you suggest how to diagnose the problematic area of the model?
I am a beginner Abaqus user. I quite often visit this forum and I find usually very useful information. Thank you for any help!

Fr1day

 

[JBlack68]

No washer?
is the animation showing the bolt head about to "snap" in the clearance hole?

 

[Fr1day]

Well, that is good point to include the washers - it may spread stresses more evenly. Even if it is not 100% accurate to the reality (head of the bolt usually does not have washer under it), but the important part in this simulation is behaviour of the profile, not the bolt. I will definitely try it and post results later, but still I am open to other options.
Yes, it actually "snaps" just at the end. You can observe how this concentration of stresses move along the edge of the bolt - but its not fluent, it jumps from element to element rapidly.