Each ground motion records have different response spectrum depending upon their origin type of earthquakes,local site responses.Target response spectrum is designed in codes considering the large number  of Crustal earthquakes ground motion from the existing database.Scaling to the Target spectrum will cover the pseudo acceleration of maximum earthquake events in database.Yes,we can use the the ground motion data direct for analysis depending upon the response spectrum shape different from the target spectrum and the results must be interpreted accordingly for specific cases.But in general Scaling is recommended for normal case.

Making the explanation short, if you are doing a site specific analysis then it is compulsory to perform matching of the spectra of the recorded motion with the spectra of the site. For example, you have a site located in Kathmandu and you want to use any earthquake record located away from the site.  The site in Kathmandu will have a particular spectrum, while the spectra of the recorded ground motions will not be similar to the site spectrum. In this case, if you use the recorded ground motion without any matching, the response may be small or large than expected depending on the spectrum of the ground motion.  So we need to match the spectrum of the recorded ground motion to the site spectrum. 

It is generally done for a range of period based on the fundamental period of the building in order to incorporate the elongation of the period due to nonlinearity as well as the higher mode periods. 

You can refer to ASCE 7 codes for the range of scaling and the procedure of scaling can be referred to any journal papers related to this field.

Satish Bhagat, Thank you for explanation!

and here is another query that should i scale  peak ground acceleration of recorded time history to some specific value of 'g'. i.e. i got  time history data with peak acceleration of 0.15g, than how can i scale it to 0.2g or 0.3g?  should i multiply whole time history data by two to get  0.3g  as PGA or is there any specific  procedure to make 0.3g or some other value from the available time history data.

Suraj Thapa, Thank you for your explanation.

Hi Mahesh,

Actually, different country prefers different ways for scaling. If you are scaling to a specific PGA then what you said is the way to do.  Just use a scale factor of 2 for 0.15g to get 0.3 PGA.

However, if you want to match to the spectrum of the site then you need to write MATLAB codes or use any ground motion matching tools.

Dear Satish Bhagat,

thank you for your reply. i scaled ground motion time history data to 0.3g by multiplying directly. and i have matched its response to the target response spectrum as IS 1893:2002 code design spectrum in  time domain matching in ETABS directly not manually. BUT  interesting is that the base shear and displacement caused by time history analysis are lesser than that of equivalent static method as per indian standard as well  lesser than values obtained by response spectrum analysis using IS 1893:2002, design response spectrum.

what should we use as scale factor in time history load case in ETABS if we match time history data with target response spectrum of IS 1893:2002? Here i have used Scale factor=(I.g/2R), is this right for scale factor in time history load case in ETAB ?

Thank you!

Hi Mahesh, 

In my case, I personally do not prefer matching the time history records to a certain value of PGA.  I prefer matching to the target spectrum considering a period of range.  However, sometimes it is possible to get base shear in time history analysis that is lesser than the results of the equivalent static method.   Could you draw the response spectrum of the site that you considered and the response spectrum of the ground motion record in a comparative manner and post it here or send it to me, I can tell you the reason after looking at it.  Please post your building detail+ Type, no. of bays, height, time period etc.

Satish

Dear  Satish Bhagat, here is target response spectrum and time history response spectrum not matched and matched both are plotted. the building height is 3.5@12=42.0m, three bay in each direction i.e. 5@3=15.0m,  first mode time period is 1.35 sec. HERE you can go through the spectrum plot.

Thank you!

A bit late, but this could help you! Along with this article you can find the ground motion data that compatible with the Gorkha earthquake!

Article SIMULATION OF EARTHQUAKE DESIGN GROUND MOTION COMPATIBLE TO ...

Thank you Parajuli Sir!

Good question.. Please share me the best answer might you trust...

Regards…

I think we should use scale factor of (IG/R) while performing time history analysis in ETABS as the 2 in the denominator from codal formula is for converting MCE to DBE. If I am wrong and anybody knows better please let me know.

Can you Please share a time history data of Gorkha Earthquake? I am doing Final year Project on Time History analysis.I would be thankful

Hi everyone

I am doing the like same with Mahesh Raj Bhatt. I am interesting what did Satish Bhagat say. Because I am very curious how does to scale acceleration to target, e.g., from 0.19 to 0.21 g. Do you have any provisions that say to scale using the equation of scale is PGA'/PGAo, where PGA' is PGA target and PGAo is PGA existing?

Could we say so? Please give me suggestions.

Thank you.

How can i generate response spectrum curve in Microsoft excel based on ASCE 7-05 Code ?

The Spectral Matching is done to match the spectra to the code based design spectra.

Hello all dears

To do time history analysis in ETABS

I downloaded 7 acceleration records from the Peer site

and then I made the spectrum according to the rules and got the scale factors

I introduced it to Etabs software and did a time history analysis

but does not occur nonlinear hinges in beams and columns

I'm likely to make a mistake in scaling

when I Multiplied the acceleration in an extra g(=9.81) I find my goal. But I have to have a reason for my work

If I get a scaled acceleration that is correct and ready to use in Etabs

I can compare and find the problem.

please send me the scaled acceleration record that you used for time history analysis in Etbas.

Sincere thanks

It should be noted that EQ is random in nature. The acceleration travel from ‘a Number of Known Faults’ to the area we consider is also random.

Each EQ event create a different Time history. Even it is equal Peak value but the frequency content inside the record might be totally different.

Structure responses depends also on the frequency content inside of the record not only the peak value.

Then, it means that you need to analysis with a million of TH record to cover all events created by faults around your area.

Response Spectrum Concept try to avoid those tedious analysis by capture all events and possibilities into one curve which is Envelope of the Event.

As a result, we match the TH to Define Spectrum, we can avoid a million TH analysis. However, you still see that, to match TH to define Spectrum, you can also can get a number of TH. All THs can also create the same Spectrum but it is different frequency content. Nonetheless, with TH matching, you can narrow down the possibility from million TH events into 7-10 TH.