Dear Jason
I hope you are fine
Please i met a problem when i wanted to work with aeroDyn v15
In fact I have indicated the path in the test.18 to the file of the aeroDyn V15 rather that aeroDyn for v14, and i have changed compaero to 2 but the simulation doesn’t run
I want to perform the time simulation in order to obtain an approachable value for TSR of the 5MW, elsewher how i can obtain the TSR with the v14
I wil appreciate your help
Dear Ali,
It’s hard to know what you are doing wrong without seeing your input files, but from the error message you are receiving, I suspect you having set CompAero to 2 in the FAST primary input file.
It appears that you are using a version of FAST v8 that is not quite the newest. FAST v8.16 was released in July. Unless you have a specific reason for using an older version, I suggest that you upgrade to the newest version, available from here: nwtc.nrel.gov/FAST8. In FAST v8.16, Test18 is already set up to use AeroDyn v15.
Best regards,
Dear Mr janson
I want to extrait the rotating force at the main bearing LSSGagFza and LSSGagFya in Simulink (u(strmatch(‘LSSGagFza’,OutList))). I have listed the parameters in the ElastoDyn output list but an error occured:
The expression: u(strmatch(‘LSSGagFza’,OutList))
in ‘NREL5MW/Fcn6’
has a syntax error
Please a help
Dear Ali,
From your description, I don’t see anything wrong with what you’ve done. But I don’t run Simulink often enough to know what would trigger this error. Hopefully another forum user can comment as to what could be the cause.
Best regards,
Dear jason
Tell me if I am right
To calculate the PI parameters of the pitch controller the 5MW WT, we can use the relation rather than the lineraization:
Kp=kp(theta=0)*GK(theta)
Ki=ki(theta=0)*GK(theta), where GK(theta)=1/(1+theta/theta_k) (theta_k=6.30°)
For example at the wind speed v=15 m/s, Ω=12.1 rpm, theta=10.45°, then we replace in the previous equation and we can obtain the PI parameters at the defined point: Kp= 0.0071, Ki= 0.0030 (wind =15 m/s)
Dear Ali,
Yes, your understanding is correct. Of course, the gains at theta=0 and theta_k depend on the turbine design, derivable from through FAST linearization.
Best regards,
Thank you very much Jonson for your reply
When I have implemented the Pi pitch controller with the calculated parameters( Kp= 0.0071, Ki= 0.0030), there is a large divergence from the gain scheduling controller. My simulation have been done by a turbulence wind speed of 15 m/s mean value, i.e around the point of linearization where the controller is acted on the generator error. I expected that the result don’t coincide with that of the baseline controller due the changes of the wind speed and therefore the operating point, but in my simulation there s a large divergence. Can you please tell me what causes be to that divergence.
Dear Ali,
I don’t know enough about the details of what you have done to know why your controller yields different results than the baseline controller, but of course, the gain scheduling in the baseline controller will certainly have some influence–more so for higher levels of turbulence. Are you using the torque-speed curve in your controller that the baseline controller is using?
Best regards,
I don’t understood well what you mean by “the torque-speed curve in your controller that the baseline controller”. But I writ you some details about the simulation
My simulation was done in the region 3 of wind turbine operation, so the torque is kept fixed at the optimal value, and the turbine is controlled to regulate its produced by means of the pitch controller. I have simulated the gain scheduling controller, that i have called baseline controller, for comparison purposes and the controller worked well. My aim now is to design a linear controller (PI), so a linearization of the wind turbine must be given to derive the system parameters. I am using the relations to calculate the parameters PI around wind=15 based on the report of the 5 MW WT:
Kp=kp(theta=0)*GK(theta)
Ki=ki(theta=0)*GK(theta)
So implemented the PI so my problem is that the results are farm from that of the gain scheduling
Dear Ali,
The baseline controller for the NREL 5-MW turbine does not keep torque fixed in Region 3; instead, power is kept fixed in Region 3 by deriving the torque as inversely proportional to generator speed. Using constant power will yield additional variation of the rotor speed (and resulting blade pitch) than using constant torque would.
Best regards,
Thank you the problem is solved
Tell me please what is the name of drive train shaft twist angle output from FAST.
Dear Ali,
The low-speed shaft twist angle is actually not an available output of the ElastoDyn module of FAST v8. But you can calculate the shaft twist by subtracting the azimuth angle on the gearbox-side of the LSS (LSSGagPxa) from the rotor azimuth angle (LSSTipPxa) (just watch out when the angles shift from just below 360 degrees to 0 degrees).
Best regards,
Thank you
Please what about the output; low shaft speed torque (the name of the output in FASt)
Dear Ali,
RotTorq (or equivalently, LSShftMxa etc.) is the name of the low-speed shaft torque output from the ElastoDyn module of FAST v8. Please see the OutListParameters.xlsx spreadsheet provided with the FAST archive for a complete list of output channels available from the ElastoDyn module.
Best regards,
Mr Jonson
Tell me please what about the twist angle (the difference between the High shaft speed and the low shaft speed angle) in FAST, is that different from the twist angle that you have indicated me in the post Fri Mar 17, 2017 9:15 pm
Tell me also if it is equivalent to the drive train shaft twist angle
Thanks
Dear Ali,
The twist angle I described above is the only drivetrain twist DOF in the ElastoDyn module of FAST v8.
Best regards,
Dear Jonson
Thank you very soon response
Please I am not familiar with Mlife. I foud it a reliable tool to calculate the loads.
Can you please summarize me how I can obtain the flapwise moment and LSS torque DEL in some steps. I have confused with the multiples documentations
Thanks
Dear Ali,
When you say that you want to obtain the flapwise moment and LSS torque DEL, what type of DEL do you want? MLife has many options for computing DELs e.g. short-term (i.e. a separate DEL for each FAST output file) or lifetime (weighted based on the wind speed probability distribution); based on a fixed mean, zero mean, or without the Goodman correction; etc.
I suggest looking at Test01.mlif in the CertTest folder of the MLife archive to see an example of how short-term DELs are calculated.
Best regards,
Dear MR Jonson
Thank you very much for the suggestion
I have read the Test but I dont revealed my study case. Realy I dont know what type of DEL correspond to my state. In fact my aim is to test the effect of my controller on the fatigue damage of the drive train and the flapwise moment
Can you please guide me how to conduct this simulation
Sincerely yours
Dear Ali,
I can’t tell you what type of DEL is suitable for your analysis. You should pick an approach and set the related inputs appropriately. If you don’t know what type of DEL you want and/or don’t know how to set the related inputs, I suggest not computing DELs. Instead, you could use the standard deviation of the loads as a proxy for fatigue.
Best regards,
