How to model yaw-wind misalignment?

Good everything to everyone :slight_smile:

I’m trying to model a nacelle yaw misalignment with respect to average wind in the x direction. To do that, I impose an initial angle to the variable NacYaw in …ElastoDyn.dat, then keep the yaw control turned off.

In almost all cases I tried so far, FAST stops with an error like this one:

FAST_Solution:FAST_AdvanceStates:AD_UpdateStates:BEMT_UpdateStates(node 7, blade
3):BEMT_UnCoupledSolve:DeterminePhiBounds:There is no valid value of phi for these operating
conditions! Vx = -0.79016, Vy = 25.635, rlocal = 19.937, theta = 0.17734

I’m kinda stuck cause I don’t know what it means :slight_smile: Could anyone help me interpret it? Why can it happen? How can we get around it?

Thank you

Dear Emanuele,

Were you ever able to solve this problem? Which version of AeroDyn v15 are you using? Does upgrading to the latest version of FAST v8 help?

This error means that the blade-element/momentum theory (BEMT) solution algorithm of AeroDyn v15 cannot find a valid solution. The error has been discussed several times on this forum. Please use “Search…” in the upper-right corner to see how other users have solved similar problems in the past.

From the following forum topic, we are aware of numerical problem in the BEMT convergence algorithm of AeroDyn v15 when Vx gets very close to zero, but your Vx is quite a bit lower than zero: AeroDyn v15 convergence issues - #7 by Jason.Jonkman.

A negative Vx is a bit odd. Does the structural solution look satisfactory up until the point of failure, or is there perhaps a numerical instability whereby the structural motions become very large?

Best regards,

Dear Jason,
I have some questions about simulating the parked land-based 5MW wind turbine with the yaw-misalignment. I am now using the Openfast v3.2.1 and run the 5MW_Land_DLL_Wturb.fst. For the parked condition, I think I have made the right settings according to several posts discussed in the Forum (if you think I may make mistakes I can share my settings). Also for Yaw-misalignment, I mainly disabled the YawDOF and set the NacYaw to the value I intended. I have questions regarding the results I obtained as presented below:

A) flap-wise root moments for three blades with NacYaw is 0°

B) flap-wise root moments for blade 1 with NacYaw = 0° , 90°, 180°, -90°

For figure A), I wonder why the moments at the root for different blades are different? I guess it may relate to the azimuth position for the rotor at the initial time I set, which is 0 Azimuth in ElastoDyn. If so, it will mean that the azimuth position will have a great effect on the responses for each blade and where can I find the relevant description for this setting? (I find no info in FAST User’s Manual 2005)

For the figure B), It seems to show that for blade 1 root moment, the response is not exactly symmetrical along 0-180 ° (for example, for NacYaw = 0° and 180°, the response is similar, but the mean of response is slightly different). Except this is related to the azimuth position, what other factor can be?
Thank you for your time for this matter. Look forward to hearing from you.


Dear @Charlie.Sheng,

Presumably in this parked condition the generator DOF is disabled and all three blades are pitch to 90 degrees; is that correct? In this case, the flapwise (yb) moment acts in the in-plane direction, which means the mean value of the moment will be heavily driven by the blade weight, blade center of gravity, and azimuth angle. And the asymmetry associated with winds 180deg from each other is likely the result of the sign changing on the aerodynamic load, but the sign not changing on the gravity load.

I hope that helps.

Best regards

Dear Jason,

Thank you for your prompt reply and pointed out that the gravity load can play a important role for my obtained results. And Yes, In my parked condition the generator DOF is disabled and all three blades are pitch to 90 degrees. As you also mention the azimuth angle, I wonder what does the 0 Azimuth in ElastoDyn mean? because this settings likely affects greatly for individual blade.

I actually want to simulate the DLC 6.1 and 6.2 in IEC, which is stated as below:
In DLC 6.1, for a wind turbine with an active yaw system, a yaw misalignment of up to ±15°
using the steady extreme wind model or a mean yaw misalignment of ±8° using the turbulent
extreme wind model shall be imposed, provided restraint against slippage in the yaw system
can be assured.
In DLC 6.2, a loss of the electrical power network at an early stage in a storm containing the
extreme wind situation shall be assumed. Unless power back-up is provided for the control
and yaw system with a capacity for yaw alignment for a period of at least 6 h, the effect of a
wind direction change of up to ±180° shall be analysed.

I want to consult you that for DLC 6.1 I need to simulate ±8° yaw misalignment only? and for DLC 6.2, I need to simulate for cases with Yaw misalignment ±180°? also, why in the power production, it seems not to require for the Yaw misalignment check?

Thank you for your help.


Dear @Charlie.Sheng,

The Azimuth angle in ElastoDyn is defined relative to ElastoDyn input AzimB1Up. When AzimB1Up = 0, the Azimuth angle is zero when blade 1 is directly vertically upward. See the old FAST v6 User’s Guide for more information:

I agree that DLC 6.1 needs a yaw misalignment of +/-8deg when simulating with turbulence and DLC 6.2, in the absence of sufficient battery backup, needs a yaw misalignment of up to +/-180deg.

In the text discussing DLC 1.1 to 1.5, the IEC 61400-1 standard mentions that, “deviations from theoretical optimum operating situations such as yaw misalignment and control system tracking errors shall be taken into account in the analyses of operational loads.” However, it doesn’t prescribe a specific misalignment requirement.

Best regards,

Dear Jason,

Thank you for sharing the FAST v6 User’s Guide which is helpful. I now have understood the relation between Azimuth and AzimB1Up in ElastoDyn.
I got another further question. In my setting with Azimuth = 0 and AzimB1Up = 0, the blade 1 will point right up. Then, by presuming that Blade 2 and 3 are identical and the angles between blades are all 120°. Thus, the (flap-wise) moments for blade 2 and 3 in my parked scenario should be similar, including the sign, even through with the gravity load. But Figure A) in the my previous post seems not mean it, see How to model yaw-wind misalignment? - #3 by Charlie.Sheng . Do I misunderstand something during this process?


Dear @Charlie.Sheng,

Figure A looks correct to me. The yb axis is directed nominally downwind in this case with 90deg blade pitch and the gravity load will generate a positive moment for one blade and a negative moment for the other.

Best regards,

Dear Jason,

Thank you for your knowledge which enhanced my understanding about FAST system. In our conversion in previous post, you indicated that for DLC 6.2, the ±180° Yaw misalignment shall be considered in IEC provided that there is no battery backup. I wonder how this assumption is true for modern wind turbine? I would believe that in order to protect wind turbine from yaw errors during storms like hurricane, recent developed wind turbine will likely provide sufficient battery backup.


Dear @Charlie.Sheng,

I agree, but I don’t have direct knowledge about how common battery backup is for modern wind turbines.

Best regards,

Dear Sir,

I am using the FAST v7 Seismic module. I intend to run simulations for misaligned seismic motions with respect to wind inflow. Considering that the default in FAST v7 Seismic module assumes concurrent loads; how can run such simulations, I believe the direction of application of earthquakes cannot be changed. Shall I just yaw nacelle using yaw maneuvering inputs in the fst file with wind field generated at different misaligned angles? Will this work?

Need an urgent reply on this. Thanks

Dear @Kashyap.Subham,

In the Seismic version of FAST v7, you can certainly change the yaw angle and wind direction to account for different conditions. But you can also change the direction of the seismic excitation because you can directly change the X and Y components of the seismic excitation.

Best regards,