Hi Miguel,

Try troubleshooting according to priority as follows:

1- Check contact Normals;

2- Control increment size to prevent sudden Stifness change:

*STATIC

0.01, 1.0 ,1.0000E-08, 0.1

3- For highly curved surfaces of deformable-body where contact take place use unsymmetric matrix storage

*STEP, UNSYMM=YES

4-Chek your boundary conditions;

5- In case all the above is done and yet no result, then:

*CONTACT CONTROLS, AUTOMATIC TOLERANCES

Kind regards,

Alireza

There could be many reasons for this non-convergence. Please share your input file.

Ahad Javanmardi I simply defined the tangencial and normal behavior with the default values

Johannes A. J. Huber I attach the input file.

Thanks for your help

The issue seems to be that you are not dealing with a quasi-static problem in your model. Since you have no tangential friction, part A will accelerate while sliding on part B. This could also happen if you use a friction coefficient, depending on the load. Using a static solver gives then reported issues. When I removed the pressure in your model and instead added a displacement along U2 downwards, the model converged without problems.

Also, given the setup of your model, I would consider accounting for non-linear geometric changes, since the bar (part B) can potentially deflect so much that it violates the assumption of small displacements.

Note, that I also removed your definition of friction coefficient to be 0 in the interaction property since it is unnecessary, the default value is zero if you don't define it. Also, I wonder whether the definition of the geometric property in the interaction property is necessary.

Thank you very much for your help Johannes A. J. Huber

I was unaware of the impossibility of applying loads such as concentrated forces or pressures if there was no tangential friction between the parts.