Non-convergence of displacement control cyclic analysis in ABAQUS

[Winhe]

Hi all,

I'm new to the site and this is my first post.

I am currently modeling a column base plate connection. All my elements are C3D8R. General contact is used. Friction property is defined between baseplate and grout as well as between anchor plates and base plate. Other surfaces are all frictionless. Tie constraints are used between column and baseplate also between grout and concrete.

I did a monotonic push (displacement control) on the column to 10% drift. The model converged and results look good comparing to experimental data.

However, when I applied a cyclic displacement to the column. The model was having convergence difficulty and the model eventually stoped at a column drift of about 0.5%. I have tried to increase the number of cutbacks and palyed around with the mesh size but the model still wouldn't work.

At the moment, I don't have any clue on what's causing the problem. Any help would be much appreciated.

 

[g.alshamsi]

What solver are you using(i.e.: standard or explicit)? You can check the job diagnostics tool in the visualization module for clues on where the problems are occurring (tools > job diagnostics). Some extra info would also help, for example; what material model are you using?n

 

[FEA way]

Did you try using *Direct cyclic or *Fatigue analysis step ? Those are special procedures meant for cyclic loads.

 

[Winhe]

Hi g.alshamsi,

I am using standard at the moment. Looking at the last increment (in job diagnostics), force equilibrim is not achieved at column nodes (more specifically nodes at the top part of the column) for consecutive iterations and attemps. I initially thought this could be a hourglassing issue so I refined the column mesh couple times but it still wouldn't work.

For steel component, I'm using elastic + plastic material. For grout, I use concrete damage plasticity where tensile properties are defined using stress-displacement (crack width) relationship. Concrete remains elastic.

I didn't define any deterioration as I want to see how the results look first without deteriortion.

 

[Winhe]

Hi FEM way,

No, I have not as I thought these steps are more for fatigue analysis where I am not looking at fatigue. I am using multiple general steps and applying the desired displacments with "Amplitudes" assgined accordingly.

But since you have mentioned it, I should try both of these steps as well.

 

[FEA way]

Direct cyclic approach is often used for fatigue analyses but not necessarily. It’s optional to include low cycle fatigue when setting this step. However, direct cyclic procedure assumes linear geometry and fixed contact conditions. Fatigue step may include these sources of nonlinearity but it’s meant specifically for fatigue crack growth analyses.