asdeasoft.net

I am new to STKO and OpenSees. I followed a tutorial on YouTube regarding RC Frame Structure by Asdea Software https://www.youtube.com/watch?v=vPB9ZwTZbNw .

In this webinar a custom command of adaptive time step was used. But since the version of STKO I am using has adaptive time stepping built in, so I used that open instead of the custom command. However when the Analysis finished I was not able to see model stage 2 which is related to "PUSH" in the post processor even though the all the analysis got completed when I did the analysis. After searching online about the issue I am across this post 2022 viewtopic.php?f=46&t=2198 .

In this post, an updated SCD file is provided. I downloaded it and compared that file with my own file and I tried to match almost all the things so that I can pin point exactly where I was making some mistake. However I was not able to figure out the mistake.

Another issue that I am facing is that while defining the analysis command for the vertical load, I can not find the option for "pFlag".

I have attached some screenshots highlighting my issues and I also attached my model file. Please help me in this regard as I am unable to find a solution

Hi, Did your model converge at the first step in Model Stage-2. I am able to get the "Mode Stage-2" for the SCD that you attached. Please refer to the picture attached. You can also try running on different machines.

For your second query, I assume that you asked about $printFlag in NormDispIncr. Yes, the current version of STKO does not include that feature, however you can specify the $printFlag in TCL input file (analysis_steps.tcl) generated by STKO.

Please let me know if you need further assistance.

Thanks for your reply.

Both of my analysis were completing but I was getting a lot of "CTestNormDispIncr::test() - failed to converge.....Reducing increment factor due to non convergence" warnings in model stage 2. I have not tested the model on another machine yet, but on my current machine I tried increasing the mesh size and then performed the analysis again. I got relatively less convergence warnings this way and then "Model Stage 2" was showing up in the post processor.

I want to ask that is this normal that a if model has a very fine mesh then it may not work properly on some machines and may work properly on another machine. Was the mesh size actually the problem, or have I still not figured out the mistake yet?

Meshing and step size will influence the FE analyses' results. The issue of machine dependence is not a problem in your case - I asked this aspect because I was able to run your model on my machine. However, I identified some inconsistencies in analysis settings, such as the solver, constraint handlers, tol, and the number of increments in cyclic analysis. Consider always large number of increments for the cyclic analysis (about 4000 to 5000 here). Attached is the hysteretic response from the model as well as the revised SCD model. Please let me know if it is helpful.