basement interaction with soil

[spater]

Dear STKO team
Can you help me?

[HC2020]

Dear Spater,

Our Team responds to forum questions during business hours. We are not open on weekends, and our forum is not monitored those days. Furthermore, we do not guarantee a response time. We request that you wait patiently for a response. The team keeps track of the posts we have not responded to. Commenting that you have not received a response will not get you a response any faster. Please note that if you continue to do this, you will be blocked for spamming the forum.

[spater]

Dear STKO team
Thank you
my problem hasn't been solved.

[Clarabella]

We are aware of it, it is not a problem of STKO but of your model, so it is up to you to investigate the solution.

[spater]

Dear STKO team
At first, I made a simple model with this method, which could calculate normally. I asked you for help, and you told me there was no problem. I did now in strict accordance with the method of model, initially unable to partition, you told me to let me to modify, I modified according to your method, but appeared unable to convergence, I also to see the webinar, I myself have been modified, but I find different meshing grid size, I don't know what the problem is. I don't know what went wrong. Please help me.
Thank you very much.

[Horace Horsecollar]

Dear Spater,
we will reply in the next few days, as we already explained to you, we have answering priorities and the free version comes after the paid licenses.

[spater]

Dear STKO team
Ok, looking forward to your reply!
Thank you

[spater]

Dear STKO team
I have waited for 10 days, When will you help me solve this problem?
Thank you

[STKO Team]

Dear spater,

We had a deep look into your model and all you have are convergence issues that do not depend on STKO or OpenSees itself. It's just the nature of nonlinear analysis: You need to understand what issue you are having by looking at the warning messages given by OpenSees and trying to tweak the analysis settings to improve convergence. There is no hard-and-fast rule for this, it's just a matter of experience using the finite element method.

Here is a list of changes that made the model run (we used a coarser mesh to speed up the testing process):

• soil element = bbarBrick (has better accuracy for coarser meshes)
• transition K(penalty) = 1.0e8 (we used a smaller value because high penalty values can create issues with models with lots of DOFs)
• In all the analyses I used -use_nType = True, nType = 0 [when you have lots of DOFs, it can be very hard to achieve a tolerance of 1.0e-5 as the norm of dU. nType = 0 means that your norm will be max(abs(dU)) instead of norm(dU)
• In the 2 dynamic analyses: use TRBDF2 instead of Newmark; use NewtonLineSearch, use 20 iterations or slightly more.
• Move the damping before the first dynamic analysis

Keep in mind that this kind of support is not included in your license. This is not a problem related to STKO. Instead, this is a consultancy on how to model a specific problem, and it can be requested by e-mail.
Have a look here for the support type we give for the different licenses:
https://asdeasoft.net/?stko-purchase

Here is what we got:

running.png (497.34 KiB) Viewed 3908 times

[spater]

Dear STKO team
Thank you for replying!

In the 2 dynamic analyses: use TRBDF2 instead of Newmark; use NewtonLineSearch, use 20 iterations or slightly more.

Do you mean in the all analyses?
I just changed the two dynamic analyses not including the gravatity analyse step.
Is it correct as the screen shots?