Time domain based flutter prediction

Wind & Water Structural Analysis
1

Hello everyone,
I am trying to simulate flutter behavior of wind turbines in the time domain analysis. For that, I want to fix the blade pitch angle to zero and accelerate the rotor in still air (wind velocity is zero) from 0 rpm by specifying constant Generator torque. With increasing rotor speed, aerodynamic loads increase and lead to flutter above a specific rotor speed. I have considered onshore NREL 5 MW case for this study and changed inputs in the Test26 case to the following.

0 CompInflow (In “Test26.fst”)
0 RotSpeed (In “NRELOffshrBsline5MW_Onshore_ElastoDyn_BDoutputs.dat”)

(In “NRELOffshrBsline5MW_Onshore_ServoDyn.dat”)
0 PCMode
1 VSContrl
9999.9E-9 VS_RtGnSp
1.0E+4 VS_RtTq
9999.9E-9 VS_Rgn2K
9999.9E-9 VS_SlPc

I have following questions and issues while running simulations:

  1. Are these changes in inputs are sufficient for what I am trying to simulate?

  2. After making the above changes in the inputs, simulations fail due to the following error.

FAST_Solution:FAST_AdvanceStates:AD_UpdateStates:BEMT_UnCoupledSolve:DeterminePh
iBounds:There is
no valid value of phi for these operating conditions! psi = -2.6203, Vx = -1.29
296E-03, Vy =
9.95915E-03, rlocal = 11.743, theta = 0.23227

FAST encountered an error at simulation time 1.00000E-03 of 5 seconds.
Simulation error level: FATAL ERROR
Aborting FAST.

  1. I have changed the “WakeMod” to zero in “NRELOffshrBsline5MW_Onshore_AeroDyn15.dat” which disables the BEMT. The simulation now runs but the rotor speed, generator torque are close to zero.

Can anyone help me in resolving above issues and clarify what inputs I should use to run the simulations.

Thanks in advance
Sudhakar Gantasala

2

Dear Sudhakar,

I’m not surprised the BEMT solution of AeroDyn fails when the inflow wind speed is zero, so, I agree with setting WakeMod=0.

It sounds like you want to motor the rotor to speed up, but the simple variable-speed controller (VSContrl=1) only functions when generating power. To motor the rotor to speed up with constant torque, you could set:
VSContrl = 0 (use a generator model instead of an active torque controller)
GenModel = 1 (enable the simple-induction generator)
SIG_SlPc = 9999.9E-9 (make the torque-speed curve very steep)
SIG_SySp = 9999.9 (or a generator speed far beyond what you want to simulate)
SIG_RtTq = 1.E+4 (or whatever constant generator torque you want)
SIG_PORt = 1.0 (don’t scale that constant torque)

I hope that helps.

Best regards,

3

Dear Jason,
Thanks for your quick response. The rotor is accelerating as intended with the inputs you have specified.

I am running another simulation with turbulent wind profile, Test26 case. When I activate “AFAeroMod” with the input 2, which considers Beddoes-Leishman unsteady model, I get the following error.

FAST_InitializeAll:AD_Init:ValidateInputData:Unsteady aerodynamics module is not enabled in this version of AeroDyn. Set AFAeroMod to 1.
FAST encountered an error during module initialization.
Simulation error level: FATAL ERROR
Aborting FAST.

Can you help me how can I run this simulation with Beddoes-Leishman unsteady model.

Thanks & Regards
Sudhakar

4

Dear Sudhaker,

I’m glad that worked for you.

The unsteady airfoil aerodynamic (UA) routines were not functional in FAST v8.12 and FAST would trigger that error if enabled. The UA routines are now functional in FAST v8.15 released in April, 2016, so, you should upgrade if you wish to use UA.

Best regards,

5

Dear Jason,

Thanks for your help. I can now run the simulations with unsteady aerodynamics model in FAST_V8.15.

I have few more doubts in the Test26 input files.

  1. I read the NREL 5 MW baseline report (nrel.gov/docs/fy09osti/38060.pdf), where it is stated that “The FAST code assumes that the blade pitch axis passes through each airfoil section at 25% chord”. But when I see the input file “NRELOffshrBsline5MW_Blade.dat”, pitch axis varies from 0.25 to 0.375 times of the chord length. Can you comment on this?

  2. In the file “NRELOffshrBsline5MW_BeamDyn_Blade.dat”, the 6x6 section mass matrix is having only diagonal elements. I expect cross-coupling terms due to the offset between pitch axis and center of gravity as defined in BeamDyn theory manual (wind.nrel.gov/nwtc/docs/BeamDyn_Manual.pdf). Are these terms don’t influence the vibration response? If we have to consider these cross-coupled terms, then the reference for Xcm and Ycm values used in the 6x6 mass matrix are with respect to the pitch axis defined above?

Regards
Sudhakar

6

Dear Sudhaker,

Here are my answers to your questions:

  1. Actually, ElastoDyn input PitchAxis is not used in Test26, or any FAST v8 model that uses AeroDyn v15. ElastoDyn input PitchAxis is used in conjunction with AeroDyn v14, based on the old definition of input AeroCent from FAST v7 and earlier. See the FAST v8 ReadMe file for more information: wind.nrel.gov/nwtc/docs/README_FAST8.pdf.

  2. The absence of the sectional center-of-mass offsets from the NREL 5-MW turbine in Test26 is discussed briefly in my post dated Feb 23, 2016 in the following forum topic: 6x6 Stiffness Matrix of NREL5 MW Turbine.

I hope that helps.

Best regards,

7

Dear Jason,
Thank you very much for your answers. I understood better now.

Regards
Sudhakar