0
$\begingroup$

I have the following torque signal picked up with a 10.240Hz sampling rate from a testbench.

signal over time and its fft

I am studying its fft which I create on Octave with the following code:

ftpar=fft(torque);

nfft=length(ftpar); 

freqtpar=abs(ftpar)/nfft;

freqt2par=freqtpar(1:nfft/2); 

mediapar=freqt2par(1); 

freqt2par=2*freqt2par(2:end); 

deltafreq=Fs/nfft; 

xfft=[deltafreq:deltafreq:(Fs/2-Fs/nfft)];

plot (xfft.',freqt2par);

When looking at the 10 frequencies with the biggest amplitudes the one at the lowest frequency is at 24 Hz all of the others are at higher frequencies. However, by looking at the time domain graph there seems to be a big oscillation with around a 0.5 Hz frequency.

Where is that oscillation coming from? Am I missing something when doing the fft? The lowest frequency I get from the fft is 0.08 Hz so I should be seeing that kind of frecuency.

Thanks for your help

$\endgroup$
0
$\begingroup$

What is the expected signal? I assume that you expect the variation with respect to some periodic variation. Over this variation, you might expect modulation which might be caused by the first mode of the structure which can be at the (1 Hz - not 0.5 Hz). This modulation should appear as sidebands around the main peak and its harmonics.

I suggest that you will use the Welch method...

$\endgroup$
  • $\begingroup$ Thank you for your reply. This low frequency signal is not expected according to rotating speed or of any component, so it could be related to the structure. I'm new to the field, and I cannot find a good source for applying the Welch method. Could you point me towards some bibliography or guides? Thanks again. $\endgroup$ – asier Feb 13 at 15:54
  • $\begingroup$ Also, what would I obtain from the Welch method? I read what it does is divide the signal in pieces and average it. Could that make the low frequency appear? $\endgroup$ – asier Feb 13 at 16:01
  • $\begingroup$ I think that you might miss some basic knowledge. Look for Mcfadns bearing article $\endgroup$ – Gideon Genadi Kogan Feb 13 at 19:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.