Period elongation on Response spectra

03-Unexpected behaviour/errors
mindlessflip
Posts: 8
Joined: 27 Apr 2021, 10:51

Period elongation on Response spectra

Hi dear all!

As I am working with SeismoMatch to modify my original time histories I have the fallowing issue:

The original time history gives me good results in obtaining my maximum spectra. For my case I have a period of the frame which corresponds to T=0,45s.
As soon as I modify the original time history to the matched time history it gives me a maximum spectral acceleration Sa which occurs on T=0,6-0,7s. Target Spectra is defined like in the EC8.

Does anyone have an idea why?

Thank you very much in advance!
huffte
Posts: 917
Joined: 22 Jul 2011, 10:19
Location: Cookeville, Tennessee, USA
Contact:

Re: Period elongation on Response spectra

I confess to being a bit confused on your question. Are you saying that the original time history matches the target closely to start with? Then why is spectral matching required? Certainly, spectral matching would not result in a lesser match to the target. What do you mean when you say you have "a period of the frame which corresponds to T = 0.45s."?

What do you mean when you say you "modify the original time history to the matched time history"?

In spectral matching, you determine the desired period range of interest. Perhaps you could elaborate to clarify the issue?
Tim Huff
mindlessflip
Posts: 8
Joined: 27 Apr 2021, 10:51

Re: Period elongation on Response spectra

Maybe I chose a wrong way to modify my time histories to a target spectra:

My idea was to use some original time histories which after using SeismoMatch fit for my target spectrum (EC8, i.e. with PGA=1,6 and max. Sa(T)=4,0). Therefore I use a few original time histories with different PGA and Sa(T) and match it to my target spectra. Then I use the modified time histories from SeismoMatch and apply it as a load on the structure. As my structure has a period of T=0,45s, maximum spectral acceleration should occur also on T=0,45s. However this is not the case when I use SeismoSignal to obtain my storey response spectra with the matched time history.
seismosoft
Posts: 956
Joined: 06 Jul 2007, 04:55

Re: Period elongation on Response spectra

Possibly the eigen-periods of the structure have elongated because of the nonlinear behaviour.
If you run eigenvalue analysis during or after the nonlinear analysis (Program Settings>Analysis>Perform Eigenvalue Analysis during the Nonlinear Analysis), you will be able to see this nonlinear effect.

Seismosoft Support