Skip to main content

Acceleration peak in FFT analysis

Submitted by adanoop on
education

Dear All,

I have collected time-domain acceleration signals from the test rig through an accelerometer. Voltage signals are scaled to acceleration signals based on voltage sensitivity of accelerometer. Please refer the following PDF attachment to see the time domain acceleration signals. Acceleration varies from 4 to -13 m/sec2. Later FFT analysis is done and FFT magnitudes (absolute of FFT) are calculated in Matlab software. Please refer to the attached PDF file to see the Frequency-domain acceleration signals. But here I got the acceleration peaks very high (of the order 70-100 m/sec2). Kindly suggest me whether the acceleration peaks convey the meaning ful result? Whether some kind of normalization is required for Y axis (acceleration peak) of the frequecy domain signal? Please guide me how to present the acceleration peaks in a meaning ful way?

PS: Time-domain signals are also attached if you want to perform the FFT analysis.

Thanks in advance!

Waiting for ur valuable suggestions...

Yours,

Anoop A D

Attachment Size
Doubt on FFT analysis.pdf 42.13 KB
Time domain signal.xls 425 KB
Ajit R. Jadhav

Dear Anoop,

I will supply only a few hints, not answers. (Don't wish to take away the fun you would have from your learning process.)

1. Can't make out much from your .PDF file. The reason is: the time-domain data is, I guess, 7200 samples long, whereas the FFT has only 1000 data points. Can't tell to which part of the time-domain data the FFT data refers. Therefore, also can't apply any informal judgments regarding the obtained data---can't anticipate how strong a particular frequency might be expected to be.

2. Are you sure that you have labelled correct physical quantities (in correct units) in each of the two graphs you have supplied?

3. What will be the nature of the FFT output? Will it be a real-valued quantity? complex-valued (esp. if the input is only real-valued)? If the latter (complex-valued), you should get two sets of output data, one each for the real and the imaginary part---wouldn't you? If so, what would be the physical significance of each part? And, what is the quantity you have chosen for plotting the FFT part in that PDF document? Also, what does it imply for the physics of it when you take absolute values of some data series? the square (possibly with some multiplying factors)? if you apply some other operation?

Think, and have fun.

Best,

--Ajit

- - - - -

[E&OE]

Mon, 11/01/2010 - 08:51 Permalink
jianghengqiang

My  program for your problem:

data=xlsread('Time domain signal.xls'); 

Fs=1e4;

T=1/Fs;

L=size(data,1);

NFFT=2^nextpow2(L);

Y=fft(data(:,2),NFFT)/L;

f=Fs/2*linspace(0,1,NFFT/2+1);

 

Yf=2*abs(Y(1:NFFT/2+1));

h=figure(1);

f1=f(f<1000);

Yf1=Yf(f<1000);

plot(f1,Yf1),xlabel('Frequency(Hz)'),ylabel('|Y(f)');

hgsave(h,'AccSpec.fig') 

 

 

Mon, 11/01/2010 - 09:41 Permalink