SimulationData - scottcarr/wind GitHub Wiki
For the wind speed data source I used the output of the FAST simulation. See the file Test13.out. It's just a CSV file with a bunch of columns. It looks like:
Time WindVxi WindVyi WindVzi HorWndDir VerWndDir BldPitch2 ...
(s) (m/s) (m/s) (m/s) (deg) (deg) (deg) ...
10.000 1.313E+01 -5.167E-01 4.897E-01 -2.254E+00 2.134E+00 1.308E+01 ...
For the audio data, I hacked together a Python script to generate obviously good and obviously bad sample data. Try running generate_audio_data.py to see how it works.
Below is the plotted FFT of the obviously good data:
Here we see spikes at the pumping frequency (10Hz) and the excitation frequency (1000 Hz)
Below is the plotted FFT of the obviously bad data:
Here we see the spikes for the excitation frequency are spread out. This is the suppposed behavior when the blade has a crack.