FAST Yaw moment calculation

Hi All,

I am new to this forum and relatively new to FAST and I hope this is the right place to post this question. My company is trying to incorporate a Yaw drive into our 50kW turbine and I am trying to use FAST to estimate what kind of forces/moments the yaw drive will see. Although the rest of the turbine is pretty well defined the only data I have about the yaw drive at this point in time is its weight. I am not sure what the right approach would be to estimate the loads on the yaw mechanism. I have tried to use the TYawManS, TYawManE, NacYaw and NacYawF variables to force the turbine to yaw out of the wind at the yaw rate we plan to use. I have also set the YawDOF to FALSE. And I am outputting YawBrMxn, YawBrMyn and YawBrMzn. All these moments seem to “ring” once the turbine reaches the final yaw angle. I would appreciate it if someone could tell me if this is the right way to estimate these loads. I read in the manual that setting the YawDOF to false turns of the yaw acceleration. Does this mean that the loads calculated are going to be erroneous? I do not have values for YawSpr and YawDamp, which I think are needed if I turn on the YawDOF. When I used hypothetical values for these two variables from a Cert Test case, the amplitude of variation of YawBrMxn, YawBrMyn and YawBrMzn changes dramatically. I am not sure I understand how FAST is dealing with yaw. I would appreciate any help in this matter. Thank you in advance.

Best Regards,

Subin Sethuram

Dear Subin,

Correct, disabling the YawDOF disables yaw acceleration in the FAST model. If the yaw motions are “sizeable,” this will not lead to accurate calculations of the yaw loads. And without more information, it is hard to know what is causing the “ringing” in your model.

A recent topic on this forum addresses a similar question: http://forums.nrel.gov/t/structural-loads-at-pitch-and-yaw-actuator/252/1. In my responses to that topic, I describe how YawSpr and YawDamp can be based on a desired frequency and damping of the yaw actuator. See the topic linked above and the FAST User’s Guide for more information.

Best regards,

Dear Jason,

Thank you for the prompt response. I am sorry I didnt get back to you earlier since I was at the Windpower 2010 conference. I had read the discussion that you refered to in your response and that is why I decided to post my question since it was along similar lines. In your reply you mentioned “desired” frequency for the actuator. Is that a function of the control system for the actuator? Am I supposed to pick a value and then design the actuator accordingly? Also with the YAwDOF set to false, if I ramp the turbine to a fixed final Yaw angle, and assuming everything is set up correctly with my model, should the Yaw moment settle down to a constant value or will it vary due to the aerodynamic loads acting on the turbine. I am quite new to this and do not quite understand all the workings of either FAST or turbines. Thank you.

Regards,

Subin Sethuram

Dear Subin,

When the nacelle-yaw DOF is enabled (YawDOF = True), FAST applies a built-in actuator between the commanded yaw motion (from the controller) and the actual yaw response. This actuator is 2nd order, which can be characterized by stiffness and damping constants–or equivalently, by a natural frequency and damping ratio. The stiffness and damping constants are what are specified in the FAST input files; these can be derived from the natural frequency and damping ratio or vice versa using the relationships I described in the forum topic linked above. The values can be tuned to match the actuotor available in the turbine you are trying to model or chosen based on what type of response you desire. For example, if you want to have the actual yaw response follow nearly identically to the commanded yaw motion, you should choose the damping ratio to be between 0.6 and 0.7 and the natural frequency to be large relative to the other natural frequencies prominent in the FAST model of the turbine system.

When the nacelle-yaw DOF is disabled (YawDOF = False), the actual yaw response will follow the commanded yaw motion exactly, but all yaw accelerations will be zero within the model, which may not be realistic.

Regardless of how the nacelle-yaw DOF is set, the yaw moment in FAST is computed as the difference between the applied moments (from aerodynamic excitation) and the inertia moments (from turbine dynamics). If the yaw angle is fixed, the yaw moment will only be constant if the model is in some sort of equilibrium–e.g., a balanced 3-bladed rotor excited by uniform inflow after all transients have died out.

I hope that helps.

Best regards,

Dear Jason,

Thank you for taking time from your busy schedule to answer my question. That certainly helped clear up how the yaw moments are being calculated. I was curios about the comment you made about the model being in equilibrium so I decided to run a case to see if I can get the moment that the bearing will see once the turbine is yawed to a certain angle and in equilibrium. I did the following:

  1. Turned off all degrees of freedom except CompAero.
  2. Set an initial yaw to 30 degrees and the time to start Yaw manuevers to some high value so that Final Nacelle Yaw wouldnt come into play.
  3. Set Ycmode to 0.
  4. Yawspr and Yawdamp and YawNeut, I left at 0 (I assume that with Ycmode 0, this wouldnt make a difference)
  5. Used the same blade data file for all three blades to simulate balanced rotor.
  6. Rotor rpm is 68.

After I ran the test, I still see a ringing in the moments. I have attached the file showing the moments.Test06_yaw.pdf (111 KB)

I ran the simulation for longer time period to see if it were some sort of transient but it stayed the same. I also tried to remove the precone of the blades and that changed the magnitute but not the frequency. The frequency seems to be half that of the rotor. I am not sure what to make of this. Is this something you would expect to see or do you think something is wrong with my model? As usual any light you can shed on this would be greatly appreciated. Thanks.

Regards,

Subin.

Dear Subin,

I mispoke in my prior post. If the nacelle-yaw error is fixed to a nonzero value, then the yaw moment will not remain constant even if the rotor is mass balanced. Instead, the response will be periodic due to the skewed wake. The periodic frequency should be 3 times the rotor frequency (3P). This is because the axial induction will be periodic with the azimuth angle of the rotor, but the aerodynamic loads are not a linear function of induction, so the aerodynamic loads, while periodic with the rotor frequency, will not be sinusoidal. Thus, the aerodynamic loads on a mass-balanced 3-bladed rotor will not balance. See the AeroDyn Theory Manual for a description of how AeroDyn handles skewed wake.

What I don’t understand in your results is that the frequency is ½P instead of 3P. Are you outputting the data at a high enough sampling rate to capture a 3P response? Have you verified that the rotor is in fact spinning at 68 RPM in your model?

Please note, it is not YCMode that determines whether or not YawSpr, YawDamp, and YawNeut are used. YawSpr, YawDamp, and YawNeut are not used only when YawDOF = False. See Figure 24 from the FAST User’s Guide for more information.

Best regards,

Dear Jason,

I had suspected that the yaw moment would be periodic if the nacelle is yawed. What threw me off was the frequency. But you nailed it, I was decimating the output. I had FAST output every time time step and bingo - frequency is exactly 3p :slight_smile: ! That clears it all up. Thanks a ton for your help.

I misspoke too regarding YCMode. I meant YawDOF. Thanks for pointing that out. It might have confused other people who read this post.

I have another question regarding Dynamic stall but I think I will start a new thread so that it will easier for people to find. Thank you once again.

Subin.

Hi all,

My model is a 3-bladed upwind turbine completely rigid (FAST fst: All DOF false, except for CompAero). I want to find the yaw moment with a fixed yaw angle.
My purpose is to find which forces mainly contribute to the yaw moment, so I would like to know how YawBrMzn is calculated. Is it like described at p.7 in nrel.gov/docs/legosti/old/4822.pdf ?
According to that I calculated the yam moment in the following way(psi= azimuth angle):

Yaw Moment = RootFxc1HubRadsin(psi)-RootFyc1*|OverHang|*cos(psi)+RootFzc1|OverHang|sin(psi)+RootMOoP1sin(psi)+……+…

But it doesn’t correspond to YawBrMzn. For ex., with Yaw angle=0 I obtain YawBrMzn~e-10 kNm, but the calculated Yaw Moment~e-04 kNm, so sizable.
N.B.The rotor radius is R = 0.45 m with RotorSpeed = 636.6 rpm.
Thank you!

Tania

Dear Tania,

I agree that the yaw moment is related to blade-root loads, and that one should be able to calculate the former from the later. I haven’t checked your equation, but it is possible all six blade-root load components could influence the yaw moment if the blade-precone angle and shaft-tilt angle are nonzero. I don’t see precone or shaft tilt in your equation. I don’t have time right now to develop the expression, but you should be able to derive it from vector math. Disabling all DOFs certainly simplifies things; enabled DOFs would introduce a number of inertia terms too.

I hope that helps.

Best regards,

Hi Jason,

Thank you for your fast reply! Yes, it helped because you confirmed that the yaw moment depends on the blade-root loads and I am not missing something important.
Correct. I haven’t got tilt or cone angles in my model. Never mind about developing the eq.n :wink:

Regards,
Tania