Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations KootK on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Convergece error

Status
Not open for further replies.

arigato

Civil/Environmental
May 26, 2005
18
Hi folks, Im working with SAP 9.0.3 at the moment and I am running a nonlinear time history for a simple shear 3-story building with hinges located at the ends of the columns. My hinges account for nonlinear effects in the model. My floors are assumed to be rigid diaphragms. I have no gravity loads at all, and no P delta effects and have 5% damping. My hinges have a long, positive post yield slope which basically means my structure cant collapse.

My problem is that when I run the analysis, it stops at about the same time step due to covergence errors (basically the solution never converges so it stops running. Ive played with the number of convergence steps, played with different iteration schemes step sizes and even removed the top 2 stories of my model and still it fails to converge. If I remove the hinges in the structure my solution does converge. I rebuilt the model by hand once more just incase i goofed on the modeling and re-ran the analysis and it still fails. Running the model in Sap 8 gives me the same problem but at a different time step.

Anyone have any idea what could be wrong?

I know that there is a severe drop in the record at that step I get convergence errors, where the acceleration changes by a factor of 100 and then goes back to a more "regular" value, but there are other places in the record where this occurs and I havent had any problems.

 
Replies continue below

Recommended for you

I assume you're no longer stuck on this issue, but I just had one idea (since my main line of work involves lots of time histories - and lots of convergence errors!)

If you stop your analysis and create a video of the structure up to that point, you may see what's going astray with the model. Obviously, you can generate displacement plots or data too, because that means the same thing. I've solved most of my convergence headaches in this way.
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor