Need help with FFT analysis after free vibrational analyis on a beam

Hello, I recently performed a free vibrational analysis and these are the results https://drive.google.com/file/d/0B2F1liITq1icMlk1U05sMjktVWM/view?usp=sharing

When I perform the FFT analysis on it I get results far different than the calculated natural frequency (5.1Hz). I believe there is a mistake in my signal sampling. Can anyone help me with this? Thank you.

Code used for FFT analysis:

Fs=20;
T=1/Fs;
L=20;
ctdata=load('Sravs3.dat');
t=ctdata(:,1);
y=ctdata(:,2);
%plot(Fs*t(1:50),y(1:200))
title('signal in time domain')
xlabel('time (ms)')
ylabel('X{t}')

NFFT=2^nextpow2(L);
Y=fft(y,NFFT);
f=Fs/2*linspace(0,1,NFFT/2);
figure
stem(f,2*abs(Y(1:NFFT/2)))
title('Single-sided Amplitude Spectrum of y(t)')
xlabel('Frequency(Hz)')
ylabel('|Y(f)|')

Answer 1

[Star Strider]

Without your 'Sravs3.dat' file, I cannot be certain what the problem is. However, I always define ‘Fs’, ‘T’, and ‘L’ from the data themselves:

ctdata=load('Sravs3.dat');
t = ctdata(:,1);
y = ctdata(:,2);
L = length(t);
T = mean(diff(t));
Fs = 1/T;

See if those changes produce the result you want. (This assumes a constant sampling interval, ‘T’ in your original data.)

[PriyankaDanda]

Hello, Thank you so much for responding. :) here is the excel file of the data https://drive.google.com/file/d/0B2F1liITq1icMlk1U05sMjktVWM/view?usp=sharing

After trying the above lines of code I got the below response.. but no spikes in the right places. :(

this is the signal that I gave as input.

[Star Strider]

When I looked at the relatively clean version of the signal and counted the peaks between two times, I got a natural frequency of about 24 Hz, so I’m inclined to believe the fft results. There is very little signal energy below about 15 Hz.

I wrote my own (more familiar) code:

t = ctdata(:,1);
y = ctdata(:,2);
L = length(t);
T = mean(diff(t));
Fs = 1/T;
Fn = Fs/2;

y = y - mean(y);                                            % Subtract D-C- Offset

fty = fft(y)/L;                                             % Normalised Fourier Transform
Fv = linspace(0, 1, fix(L/2)+1)*Fn;                         % Frequency Vector
Iv = 1:length(Fv);                                          % Index Vector

figure(1)
plot(Fv, abs(fty(Iv))*2)
grid

[PriyankaDanda]

Thank you so much :)

[Star Strider]

My pleasure!