How to find the frequency from data measured with a strain gage?

Hi,

I have collected data from a strain gage using a DAQ. I would like to find the frequency from that data.

I set the DAQ sample rate to 1000 samples per second, and I have 43800 samples. I am not sure if the sample rate is adequate and if the code below is correct to find the frequency.

If anyone has worked with this kind of test before and could help me to find the best approach to solve this problem.

Thank you,

signal = load('DataSG.txt');
N=length(signal);
Fs=1000; %Sampling frequency
Ts=1/Fs; %sampling period or time step
dt=0:Ts:43.8-Ts; %signal duration

X_mags = abs(fftshift(fft(signal)));
X_mags = X_mags/max(X_mags);
bin_vals = [0 : N-1];
N_2 = ceil(N/2);
fax_Hz = (bin_vals-N_2)*Fs/N;

subplot(2,1,1);
plot(dt,signal);
xlabel('Time[s]');
ylabel('Amplitude');
title('Time Domain Signal');

subplot(2,1,2);
plot(fax_Hz,X_mags);
xlabel ('Frequency [Hz]');
ylabel ('Normalized Amplitude');
title('frequency Domain Signal');

[B,IX] = sort(2*X_mags); %order the amplitudes
A1=B(end); %amplitude of the first peak
A2=B(end-1); %amplitude of second peak
f1=fax_Hz(IX(end)); %frequency of first peak
f2=fax_Hz(IX(end-1)); %frequency of second peak

BFloor=0.1; %BFloor is the minimum amplitude value (ignore small values)
Amplitudes=B(B>=BFloor) %find all amplitudes above the BFloor
Frequencies=fax_Hz(IX(1+end-numel(Amplitudes):end)) %frequency of the peaks

[Merkhav E]

Excellent. Thanks for your help.

[Star Strider]

As always, my pleasure!

Answer 1

[Star Strider]

Try this:

SG = load('DataSG.txt');
L = numel(SG);
Fs = 1E+3;                                          % Sampling Frequency
Fn = Fs/2;                                          % Nyquist Frequency
Ts = 1/Fs;
t = linspace(0, L, L)*Ts;

figure
plot(t, SG)
grid

SGm = SG - mean (SG);                               % Subtract Mean (D-C Offset) To Make Other Peaks More Visible
FTSG = fft(SGm)/L;                                  % Fourier Transform
Fv = linspace(0, 1, fix(L/2)+1)*Fn;                 % Frequency Vector
Iv = 1:numel(Fv);                                   % Index Vector

[pks,lcs,w,p] = findpeaks(abs(FTSG(Iv))*2, 'MinPeakProminence',3, 'MinPeakDist',10);
[maxamp,idx] = max(abs(FTSG(Iv))*2);

figure
plot(Fv, abs(FTSG(Iv))*2)
hold on
plot(Fv(lcs), abs(FTSG(Iv(lcs)))*2, '^r')
hold off
grid
xlim([0  50])
text(Fv(idx),maxamp, sprintf('\\leftarrowAmplitude = %.3f\n    Frequency = %.3f Hz', maxamp, Fv(idx)), 'HorizontalAlignment','left', 'VerticalAlignment','top')    

It plots the peak frequency and significant harmonics.

Experiment with it to get the result you want.

EDIT — (16 Aug 2020 at 16:38)

Added plot image —

.

[Merkhav E]

@Star Strider I am using the same code with different data, and I am not sure if my plots are correct. The frequency and amplitude seem very low. Do I need to modify the code?

Thank you.

[Star Strider]

The signal has predominantly low frequencies and significant noise (high-frequency, low-amplitude) in a relatively low-amplitude signal. The results appear to be reasonable.

If you used the same code, the only change I would do for the existing code (for this data set only) is to add:

xllim([0 5])

after the Fourier transform plot. This will allow you to see the spectrum in more detail.

[Merkhav E]

Thank you for your help. It looks better now.

[Star Strider]

As always, my pleasure!