Unable to apply the next load step in dynamic time-history

[roshen.rj]

Dear Developers,

I am running an 8-story building retrofitted with a bracing system under dynamic time-history analysis. I have thoroughly checked the model under eigenvalue analysis and all 10 models have converged. I have also checked the model under static pushover analysis and it is running till the completion of all steps. However, under the dynamic time-history analysis, the model does completely run under a selected earthquake record. I tried several approaches of increasing the time step, increasing the tolerance and convergence criteria as per recommendations from the Seismostruct user manual. Could you please suggest some steps which I follow to complete the dynamic analysis run?

Regards
Roshen

[huffte]

Have you estimated the displacement demand on the structure due to the selected earthquake record? It could be that the loading history is too severe for the model to withstand. Can you try applying a factor less than 1.0 to the specified acceleration history to lessen the load and see if the model runs to completion then? Also, compare the maximum displacement form the history analysis to the displacement capacity from the pushover analysis. This should give an indication of whether the structure fails prior to the end of the loading history.

[roshen.rj]

Have you estimated the displacement demand on the structure due to the selected earthquake record? It could be that the loading history is too severe for the model to withstand. Can you try applying a factor less than 1.0 to the specified acceleration history to lessen the load and see if the model runs to completion then? Also, compare the maximum displacement form the history analysis to the displacement capacity from the pushover analysis. This should give an indication of whether the structure fails prior to the end of the loading history.

Thank you for the reply. However, that is not the case as the unstrengthened structure runs at the particular intensity while the retrofitted structure model is unable to run at the specified intensity. Moreover, the discrepancy was not noticed in the pushover analysis as the retrofitted structure had higher strength in the pushover analysis.

[seismosoft]

Possibly there is a connectivity error in the strengthened model. Have you checked the shapes of all the modes in eigenvalue analysis?
Also, make sure that you are applying the correct ground motion in dynamic analysis; it is a typical error in dynamic analysis to confuse the units (e.g. g against m/sec2) and adopt very large or vey small scaling factors.
Seismosoft Support

[seismosoft]

Note that typically pushover analysis is more demanding in terms of numerical stability, thus it is a surprise that you get a full capacity curve in pushover analysis, and you get divergence in dynamic analysis.
Seismosoft Support

[roshen.rj]

Possibly there is a connectivity error in the strengthened model. Have you checked the shapes of all the modes in eigenvalue analysis?
Also, make sure that you are applying the correct ground motion in dynamic analysis; it is a typical error in dynamic analysis to confuse the units (e.g. g against m/sec2) and adopt very large or vey small scaling factors.
Seismosoft Support

I am a PhD researcher with more than 3 years of experience in Seismostruct. I have modelled and analysed complex structures under static and dynamic analyses but it is the first time I am getting such an observation in dynamic history. I have checked thoroughly all mode shapes of the model and it converges to get 10 modal periods! It is really in the dynamic time-history analysis that I am struggling. All those fundamental things I have considered properly.

[seismosoft]

Hi roshen,

This is rather strange indeed. Can you send your model to support@seismosoft.com, so that we have a look at it?
Seismosoft Support

[roshen.rj]

thank you, I was able to solve it.