Dear Cheng,
I’m not surprised by these results. The decrease in response outside of the 1P peak is not much different; the difference is exaggerated by the log scale on the y-axis. I would suggest plotting with linear scales to better see the true physical differences.
Best regards,
Dear Jason,
I am trying to use an IPC to reduce low frequency loading on the rotor blades. I am interested in the blade root out of plane bending moment, so i plotted the PSD of the out of plane root bending moment using Mcrunch. I have attached the plot below. It is showing peaks at much lower values than the expected 1P(~0.2Hz).
I am running the Fastv8 Test18 Certification case.
Can u clarify why are the peaks lower than expected?
Regards,
Kenneth Dsouza
untitled.tif (352 KB)
Dear Kenneth,
I would expect that you’d see peaks in the PSD of the blade-root moments at 1P, 2P, 3P, etc. I don’t know why you are not seeing this. What does the time series look like; do you see a strong 1P oscillation in that?
Best regards,
Thanks for the quick reply Jason.
There was a slight error in the time series data for which the previous PSD were calculated. So I ran simulations again this time with turbulent wind conditions. The PSDs still don’t show peaks corresponding to 1P or higher frequencies. I have attached the PSD and time series data for out of pane root moment.
Also I am wondering whether there is any error in the Mcrunch Input file that i have used. So I have attached the input of PSD portion of Mcrunch as well. All other modules of Mcrunch are set to false.
Dear Kenneth,
Your MCrunch input data looks OK to me.
I think the biggest problem may be that Test18 from the FAST v8 CertTest is only run for 60 s, which does not provide a lot of resolution at low frequencies. I would suggest rerunning this simulation for at least 600 s
Best regards,
Dear Jason,
I followed your suggestion and ran simulation for 600s using 15m/s uniform wind speed. The PSD now show definitive peaks, but the 1P frequency is still shown as 0.25Hz compared to the 0.2Hz actual value. Can you provide any suggestion as to why this maybe happening.
Regards,
Kenneth
Dear Kenneth,
I would first suggest confirming that the rotor speed output from OpenFAST is what you expect it to be. 0.25 Hz = 15 rpm.
Best regards,
Dear Jason,
The simulations were run at 15m/s uniform wind speed. The rotor speed was 12.1 rpm which should correspond to ~0.2Hz 1P frequency, but it is not the case.
Note:- I am using FAST v8 for running the simulations.
Regards,
Kenneth
Dear Kenneth,
Can you share the times-series output file used to generate this?
Best regards,
Dear Jason,
The time series data File is too large to be uploaded in the forum even after compression(97MB). Is there any other way I can send it to you?
Regards,
Kenneth
Dear Kenneth,
Can you share it through a file-sharing platform like Google drive?
Best regards,
Dear Jason,
I have attached the link for the file below
drive.google.com/file/d/19J_3dk … sp=sharing
Regards
Kenneth
Dear Kenneth,
I plotted your time series and confirm that the rotor is spinning at 12.1 rpm. I used my own MATLAB-based PSD script, which calculates the PSD directly via FFT and get the result I expect–i.e., that RootMyc1 has peaks at the rotor speed (about 0.2 Hz) and its harmonics. See the raw and binned (based on 10 frequency steps per bin) PSDs, as well as my MATLAB script (Jason_PSD.m) attached. I’m not sure why you are not getting the PSD result you expect out of MCrunch (I’m not too familiar with its options), but I would guess it has something to do with the windowing applied.
Dear Jason,
Thank you for the script as well as helping me in sorting the issue.
Regards,
Kenneth
