Dear STKO Team,
Before creating my actual model for a journal paper, I am trying to create a simple 3D model with pile similar to "sfsi model" from one of your advance lessons.
I did not want to change many parameters and features before I was confident for constructing the model.
But, I am having errors stating that the analysis did not converge. It is I think mainly because of mumps solver.
Even if the original example of yours uses mumps,
Could you please help me with this issue? Attached file is my model.
Thank you very much.
Best Regards,
About a convergence problem
About a convergence problem
- Attachments
-
- SimpleModel.rar
- (709.04 KiB) Downloaded 41 times
Re: About a convergence problem
The problem was a singular model.
We found these issues:
We found these issues:
- The selection set for Soil Domain was missing some volumes
- The selection set for Contact elements was empty
- The local axes -Z must be assigned to the beam-end contact interaction, not to the beam-end node
Re: About a convergence problem
Thank you very much for your prompt reply and for providing a working version of the model!
I followed the issues you've pointed out. I think I have corrected the issues regarding contact element being empty and the local axis.
But Could you tell me in the selection set which part of the soil domain was missing? If it is possible I just wanted to know.
I basically choose only "select solid" from the subselection mode.
Select "hole domain" in the selection sets and hide it.
Then selected the remaining soil domain to assign it as "soil domain". However, my old model still did not work.
By the way I want to use this topic, if I have similar issues for my actual model, if you don't mind.
Thank you again!
Best Regards,
I followed the issues you've pointed out. I think I have corrected the issues regarding contact element being empty and the local axis.
But Could you tell me in the selection set which part of the soil domain was missing? If it is possible I just wanted to know.
I basically choose only "select solid" from the subselection mode.
Select "hole domain" in the selection sets and hide it.
Then selected the remaining soil domain to assign it as "soil domain". However, my old model still did not work.
By the way I want to use this topic, if I have similar issues for my actual model, if you don't mind.
Thank you again!
Best Regards,
Re: About a convergence problem
That's the issue. If you hide some solids (in this case the hole domain) you will also hide the face they share with the surrounding domains, and they will not be selectable.I basically choose only "select solid" from the subselection mode.
Select "hole domain" in the selection sets and hide it.
Then selected the remaining soil domain to assign it as "soil domain". However, my old model still did not work.
What I did instead was to select all solids of the soil domain, unselect the solids of the hole, and with this selection I created the selection set for the soil-domain.
sureBy the way I want to use this topic, if I have similar issues for my actual model, if you don't mind.
Re: About a convergence problem
Thank you very much! Now it worked.
I really appreciate your support!
Best Regards,
I really appreciate your support!
Best Regards,
Re: About a convergence problem
You're welcome