FFT plot of velocity

11 views (last 30 days)
Melvin Koshy
Melvin Koshy on 5 Nov 2015
Edited: Rick Rosson on 5 Nov 2015
I have a plot of velocity (mm/s) against time for 15 sec. I want to observe the same in frequency domain. I have done the following code in MATLAB also for the same.  
Whether the FFT results also have mm/s as the unit? If I want to get the result in terms of mm/s, what should I do?
How can I get the one-third octave band plots from the FFT above?
Am I approaching this correctly?
[v,T,vT]=xlsread('vib.xls');
t=v(:,1);
y=v(:,2);
fs=750;
t=0:1/fs:(length(y)-1)/fs;
figure(1);
plot(t,y);
title('PLOT OF VELOCITY');
ylabel('Velocity Amplitude');
xlabel('Time (in seconds)');
grid on;
nfft = length(y);
K = fft(y,nfft);
K = K(1:nfft/2);
mx = abs(K);
f = (0:nfft/2-1)*fs/nfft;
figure(2);
plot(f,mx);
grid on;

Answers (3)

Rick Rosson
Rick Rosson on 5 Nov 2015
mx = abs(K)/nfft;
  1 Comment
Melvin Koshy
Melvin Koshy on 5 Nov 2015
Edited: Melvin Koshy on 5 Nov 2015
Could you please elaborate on the reply? Should I make the line in the program as mx = abs(K)/nfft At present what is the unit in the y-axis of FFT plot. What if I change the program line to mx = abs(K)/nfft Do you mean to say that if I change it to mx = abs(K)/nfft I can believe that the units on Y-axis of FFT is mm/s. Please clarify. I have made the change in program line. And now I get the value in the order of 10^(-3)

Sign in to comment.


Rick Rosson
Rick Rosson on 5 Nov 2015
Edited: Rick Rosson on 5 Nov 2015
Then, in your code, try both Option 1 and Option 2, and compare the results. What does the comparison show?
  1 Comment
Melvin Koshy
Melvin Koshy on 5 Nov 2015
The maximum value on the y-axis the plot when i use mx = abs(K) is 20 and when i use mx = abs(K)/nfft the magnituede is 0.018. Could you please clarify further?

Sign in to comment.


Rick Rosson
Rick Rosson on 5 Nov 2015
Edited: Rick Rosson on 5 Nov 2015
Please try the following experiment:
Fs = 48000;
dt = 1/Fs;
t = (0:dt:0.25-dt)';
A = 5;
x = A*ones(size(t));
N = size(x,1);
df = Fs/N;
f = -Fs/2:df:Fs/2-df;
X = fftshift(fft(x));
figure;
plot(t,x);
grid on;
xlabel('Time (in seconds)');
ylabel('Velocity (in mm/sec)');
title('Time Domain');
figure;
ax(1) = subplot(2,1,1);
plot(f/1000,abs(X)/N);
grid on;
xlabel('Frequency (in kilohertz)');
ylabel('Amplitude (in mm/sec)');
title('Amplitude Spectrum');
ax(2) = subplot(2,1,2);
plot(f/1000,abs(X)/Fs);
grid on;
xlabel('Frequency (in kilohertz)');
ylabel('Amplitude Density (in mm/sec per hertz)');
title('Amplitude Spectral Density');
linkaxes(ax,'x');
zoom xon;
Now try a second experiment:
Fc = 5000;
A = 12;
y = A*cos(2*pi*Fc*t);
Y = fftshift(fft(y));
figure;
plot(t,y);
grid on;
xlabel('Time (in seconds)');
ylabel('Velocity (in mm/sec)');
title('Time Domain');
figure;
ax(1) = subplot(2,1,1);
plot(f/1000,abs(Y)/N);
grid on;
xlabel('Frequency (in kilohertz)');
ylabel('Amplitude (in mm/sec)');
title('Amplitude Spectrum');
ax(2) = subplot(2,1,2);
plot(f/1000,abs(Y)/Fs);
grid on;
xlabel('Frequency (in kilohertz)');
ylabel('Amplitude Density (in mm/sec per hertz)');
title('Amplitude Spectral Density');
linkaxes(ax,'x');
zoom xon;

Categories

Find more on Fourier Analysis and Filtering in Help Center and File Exchange

Tags

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!