I have analysed the unsteady forces calculated by FAST v7 to get the periodical unsteady forces to check the fatigue due to them.
I got fn and ft for 30 station on the HAW blade of 7m length.
The time length was 151 seconds and step was 0.004 s.
My procedure was to get the max fn in each time step and assuming it is acting on the blade at this time neglecting the position and the other components to simplify my analysis.
I made fft analysis and got the periodical ones and I assumed that for the next 151 seconds these forces will repeat themselves.
The results of fft (Fn vs frequency) and fn in time domain is shown below
Can you tell me your opinion on this procedure?
I would suggest to check as a minimum the first and second order statistics in adidtion to the power spectrum. These parameters can in turn be checked against expected values.
You should also check if your forces are statistical steady or not. This can simply be done using an STFT (Short Time Fourier Transform). If so you should consider to switch to time-frequency domain analysis and drop the assumption that the forces in the following 151s will repeat themselves. In not your procedure seem fine, just ensure to divide your whole measurement in shorter overlapping segments before passed them the FFT routine. (I usually divide the signals in 50% overlapping segments and apply an Hanning window before do the FFT)
- Badges
- Science topic
- Similar topics
- Analytical Chemistry
- Extraction