Tips for Transient Analysis Convergence

Forum for OpenSees users to post questions, comments, etc. on the use of the OpenSees interpreter, OpenSees.exe

Moderators: silvia, selimgunay, Moderators

Post Reply
Pablongo
Posts: 15
Joined: Tue Oct 11, 2011 7:42 pm
Location: Georgia Tech

Tips for Transient Analysis Convergence

Post by Pablongo »

Hi all,
Analyzing a 2-story, 2-bay RC frame. Pushover ran successfully to 5% drift (about 14 in displacement of 2nd story). However, I keep getting convergence issues for some Ground Motion records (4 so far). I've ran 4 ground motions successfully.

I'm using nonlinear Force based beam-columns. I have played around with element convergence tolerance (from 1e-1 to 1e-8), different integrators and different parameters, different time steps (have used VariableTransient analysis) and I'm running a loop with several solution algorithms, iteration steps, AND tolerances (from 1e-6 to 1e-8 global) but I still can't get the solution to converge for some GM records.

What else can I do?? I'm getting convergence problems at drift ratios that are far less than what I ran in the pushover analysis. Any ideas?
Pablongo
Posts: 15
Joined: Tue Oct 11, 2011 7:42 pm
Location: Georgia Tech

Re: Tips for Transient Analysis Convergence

Post by Pablongo »

Update... after running some of these same GM records, but scaled to anywhere from 25% to 75% of the intensity, I can get some to converge.
Is this an indication that my structure can't handle the forces from these accelerations even though the displacements are "small"? Or could it be something else?
vesna
Posts: 3033
Joined: Tue May 23, 2006 11:23 am
Location: UC Berkeley

Re: Tips for Transient Analysis Convergence

Post by vesna »

There may be several different issues that may cause the convergence problems. If you have convergence problems at small level of displacements I would think that convergence problems are related to your model. Check your elements, your sections and your materials. If everything is to your liking try changing force based elements with several displacement based elements.
Pablongo
Posts: 15
Joined: Tue Oct 11, 2011 7:42 pm
Location: Georgia Tech

Re: Tips for Transient Analysis Convergence

Post by Pablongo »

Vesna,
Thank you for the reply. I was able to solve this issue by using finer section meshing and tweaking damping parameters. I am now having a different issue after introducing the beamColumnJoint element in my model. I have posted a question about it on a separate post. If you read this reply first, please search for "beamcolumnjoint element and dynamic analysis" to see if you may be able to help. Thank you once again...
vesna
Posts: 3033
Joined: Tue May 23, 2006 11:23 am
Location: UC Berkeley

Re: Tips for Transient Analysis Convergence

Post by vesna »

That is a contributed element that I never used and do not have experience with it. I'm sorry but can not help you on this issue.
Post Reply