Cancel
Showing results for 
Search instead for 
Did you mean: 

Earthquake analysis using FEMAP with NX Nastran

Experimenter
Experimenter

Hello everyone!

 

I'm really new in working with FEMAP, I'm working on seismic analysis of a building.

I have to do a frequency domain analysis with the load being an earthquake (data about the earthquake is accquired for an accelerogram data file in the time domain). How should I represent this earthquake data in the frequency domain? I did an FFT in Matlab, but I got an amplitude and phase spectrum and I have no idea what to do with them.

Which analysis is best suited for this kind of problem?

 

Thanks in advance!

5 REPLIES

Re: Earthquake analysis using FEMAP with NX Nastran

Solution Partner Phenom Solution Partner Phenom
Solution Partner Phenom

Hello!,

In seismic analysis the typical excitation is a Response Spectrum of acceleration vs. frequency, like the following picture, is the envelope of Arizona Power, Duke Power, TVA (Tennessee Valley Authority) and WPPS (Washington Power), see my FEMAP & NX NASTRAN SUPPORT web site in the following address where I explain step-by-step a tutorial to perform a Seismic analysis of RESPONSE SPECTRUM APPLICATION:  
http://www.iberisa.com/soporte/femap/dinamico/response_spectrum_4story.htm

4story_response_spectrum_excitation

If your excitation is in the form of acceleration vs time please note that Time = 1/Frequency, then you can convert easily to the classical excitation of acceleration vs. frequency.

horizontal_ground_design_spectra.png

Best regards,
Blas.

Blas Molero Hidalgo, Ingeniero Industrial, Director
IBERISA • 48004 BILBAO (SPAIN)
WEB: http://www.iberisa.com
Blog Femap-NX Nastran: http://iberisa.wordpress.com/

Re: Earthquake analysis using FEMAP with NX Nastran

Solution Partner Creator Solution Partner Creator
Solution Partner Creator

Hello! The approaches for earthquake analysis described in the Eurocode standard are:

 

1) Response Spectrum analysis (the input data is a SPECTRUM: acceleration vs. frequency)  Blas suggested FE method  use NX NASTRAN SOL103

2) Time Histories analysis (input data is a TIME HISTORY: acceleration vs time or accelerogram)  use NX NASTRAN SOL109 or 112

 

If you would like to use a FFT of the accelerogram (solve the problem in the frequency domain instead of the time domain)  use NX NASTRAN SOL108 or 111. Anyway This approach is not used for earthquake analysis (FFT is valid for periodic input and the earthquake is not)

Re: Earthquake analysis using FEMAP with NX Nastran

Experimenter
Experimenter

Thank you very much for the answers!

I tought that the acceleration spectrum depends on damping and the natural period or natural frequency? Am I right? If I were to divide with increments from the time record I would basically get the sampling frequency, and for frequency response that's not necessary? 

I understood that it is necessary to determine the maximum acceleration of the SDOF system with the earthquake excitation and a speciffic damping. Can I find a .dat file somewhere with the spectrum for the el centro earthquake? If not, how could I generate this file?

Thanks in advance!

 

Re: Earthquake analysis using FEMAP with NX Nastran

Solution Partner Phenom Solution Partner Phenom
Solution Partner Phenom

Dear Flower,

The Response Spectrum excitation of acceleration vs. frequency in the horizontal & vertical direction for different levels of damping is the data of the problem to solve, this information is provided by the contractor.

Best regards,
Blas.

Blas Molero Hidalgo, Ingeniero Industrial, Director
IBERISA • 48004 BILBAO (SPAIN)
WEB: http://www.iberisa.com
Blog Femap-NX Nastran: http://iberisa.wordpress.com/

Re: Earthquake analysis using FEMAP with NX Nastran

Legend
Legend

_flower_ wrote:

Hello everyone!

 

I'm really new in working with FEMAP, I'm working on seismic analysis of a building.

I have to do a frequency domain analysis with the load being an earthquake (data about the earthquake is accquired for an accelerogram data file in the time domain). How should I represent this earthquake data in the frequency domain? I did an FFT in Matlab, but I got an amplitude and phase spectrum and I have no idea what to do with them.

Which analysis is best suited for this kind of problem?

 

Thanks in advance!


That's really more a Nastran question than FEMAP.  FEMAP would let you setup either approach easily.  You could do direct transient, but that's a pain in my opinion.  Or, simply coonvert your time history into SRS  in matlab and do the analysis in the frequency domain...