Testing Tcl and Python

[mhscott]

Thank you for providing Tcl and Python scripts.

I verified that the material parameters are the same and that updateMaterialStage is called in both scripts.

I get the same results for the response history analysis (norms the same between the Tcl and Python versions)

I'm not sure what the issue could be with the original model.

[mhscott]

What version of OpenSeesPy are you using?

python3 -m pip show openseespy

or

pip show openseespy

[fjjimenez1]

Doctor, thank you very much for your help, since I am in this problem for a project that is being carried out. The results are identical until reaching the time (dt = 18s), from that time the results vary as shown in the following graph.

[image: image.png]

I don't know if the problem is true, the version of OpenseesPy that works is the following:

[image: image.png] [image: image.png]

El mié., 19 de ago. de 2020 a la(s) 09:14, Michael H. Scott ( notifications@github.com) escribió:

What version of OpenSeesPy are you using?

python3 -m pip show openseespy

or

pip show openseespy

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/fjjimenez1/Effect-site-column-3D/issues/1#issuecomment-676427641, or unsubscribe https://github.com/notifications/unsubscribe-auth/AIE6RA5ZJTYPFNZNFPAT4ITSBPM5PANCNFSM4QERHPYA .

[mhscott]

The images didn't come through on the GitHub page or through the email generated by GitHub.

It's likely there's a memory leak somewhere if the results are the same up to 18s, then become different. Is there any significance to 18s, or is it random?

[fjjimenez1]

The results are identical until reaching the time (dt = 18s), from that time the results vary as shown in the following graph. I don't know if the problem is true, the version of OpenseesPy that works is the following:

At time 18 is when the frequency of the earthquake begins and there are variations, prior to 18 there is no record of increased accelerations.

[mhscott]

Gracias! There is a new version of openseespy (3.2.2.3). Can you upgrade then run the analysis again to see if you get the same results?

pip install openseespy –upgrade

or

python3 -m pip install openseespy –upgrade

[fjjimenez1]

Thank you very much for your contribution, update the OpenseesPy (3.2.2.3) and I continue with the same problem since time 19, which is where the acceleration of the registry increases.

Without updating the materials from elastic to plastic if the results match.

[mhscott]

It looks like you're getting very large accelerations (10-15g), that are not physically possible. Even at t=19s, the acceleration is about 5g. Can you reduce the intensity of the input ground motion by a factor of 10 and see if the models give the same results with updating the materials?