Hello,
I am trying to model a steel bracket fastened with steel nails to two pieces of wood (see pictures below, please). To deal with the nail withdrawing, I am using cohesive elements connecting the wood panels to the nails. I am also using hard contact to connect the nails to the bracket (see the input file, if wanted).
However, the model is not converging during analysis. I have tried to suppress the vertical wood part (along with its nails, cohesive elements, interactions, etc) and run only a bracket load with the remaining elements. In this situation, the model converges just fine. I did the same experiment suppressing the floor wood part (along with its nails, cohesive elements, interactions, etc) and the model also converges just fine. No strange or weird results are present from any of these two tests. However, when I resume the entire model, the not-converging situation happens again.
The loading I want to apply is an upward movement of the vertical wood panel (which should interact with the cohesive elements, which should interact with the nails, which should finally interact with the bracket), but this loading condition does not converge. If I fix the vertical wood panel and apply a vertical load on the bracket itself, then the model converges just fine. It seems like moving the wood, somehow, causes the model not to converge.
So I am wondering, is ABAQUS not converging due to the high amount of interactions present (33 constraints and 36 interactions defined in the model)?
Is there anything else I can try to make this run?
Thanks for the help.
Hi Mr. Salgado,
Try meshing the slave surface bearing parts (nails, etc) with a finer mesh compared to the master surface bearing parts (wood block, etc) and also define a general contact interaction for the two blocks (where they are in contact with each other) in initial step.
Let me know if this does not work...
Syed Faizan Husain, Thank you for the suggestion. I have tried the finer mesh before, and the vertical wood block goes up (away from the horizontal block), so I figured no contact interaction would be necessary there. But I'll try your suggestions and let you know.
Assogba Ogoubi Cyriaque, I updated the question with an image of the warnings. It is not all the warnings, but the ones about nodes adjustments repeat several times due to the number of tie constraints that the model has.
Hey Mr. Salgado
Hope you are doing well.
I see that your analysis has severe discontinuous iterations, that might be because of too many contacts continuously changing their state... In such cases, I believe, explicit analysis is the only viable option...
See this
https://info.simuleon.com/blog/6-tips-solving-non-convergence-with-abaqus-fea
And this for your reference
http://blog.myigetit.com/2018/07/12/where-abaqus-explicit-shines-over-abaqus-standard/
Hope it helps!
Good luck...
Syed Faizan Husain, thanks for the suggestion. I will definitely try ABAQUS explicit in this situation. Thanks for the references, it will be of great value.
- Badges
- Science topic