BEM algorithm implementation according to AeroDyn v15

Wind & Water AeroDyn Development
1

Dear all,

I am coding the BEM using C++ by myself, mainly according to
(https://docs.nrel.gov/docs/fy15osti/63217.pdf). I omitted some parts:

  1. I omitted ‘BlCrvAC’ and ‘BlSwpAC’ from the code, but include all other blade properties.
  2. Also, I omitted the skew angle in my code.
  3. Didn’t include unsteady aerodynamics.

The simulation setting is steady-state wind 8m/s, and the rotor speed is 9.1090 (rpm); others are shown in the first figure, the OpenFAST example I am using ‘5MW_Land_BD_DLL_WTurb.fst’. The second figure below shows the comparison of AeroDyn and my code. I compared the lift and drag coefficients, axial and tangential induction factors, inflow angle, and AoA, elemental lift and drag force.
My implementation strategy:

  1. I upgraded the linear interpolation to cubic interpolation, but the difference is still there.
  2. And I manually set the root and tip axial and tangential induction factors.
  3. I didn’t include the relaxation.

Questions:

1. The NREL 5MW blade property ‘BlCrvAng’ is all zeros, but the output file contains ‘B1N1Curve…’, the data are “-0.04581500, -0.04145112, -0.03705501, -0.03705501, -0.04345870, -0.05162950, -0.05157691, -0.05105104, -0.05089845, -0.05023600,
-0.04941005, -0.04860823, -0.04747296, -0.04658158, -0.04616814,
-0.04576731, -0.04510486, -0.04395841, -0.04330288”. If the input file didn’t include the data, why did the outputs contain it? How does the AeroDyn deal with this data?
2. The main difference between my code and the AeroDyn in mid-span (roughly 0.3 - 0.6), I checked the ‘BlCrvAC’ and ‘BlSwpAC’, they have relatively large data values in mid-span, I am guessing the difference may be caused by this? (I didn’t try to include it yet)
3. I read the source code of AeroDyn, “REAL(ReKi) , DIMENSION(:,:), ALLOCATABLE :: xVelCorr !< projection of velocity when yawed + prebend [m/s]”, the ‘xVelCorr’ is about the yawed + prebend, But how about the tilt angle?

Any help would be greatly appreciated!
Thanks in advance.

Best regards,
Chaozhi Qiu

Screenshot 2025-06-23 at 1.31.27
Screenshot 2025-06-23 at 1.31.271788×1516 266 KB

AeroDynAeroQ
AeroDynAeroQ1920×2692 344 KB

2

Dear @Chaozhi.Qiu,

Just a few comments:

  • Even if you don’t specify blade curvature in the AeroDyn input file, the blade can still have it when it deflects structurally.
  • The effect of blade curvature, whether input or from deflection, on the coordinate systems used in AeroDyn for the aerodynamic calculations and output reporting depends on how you’ve set BEM_Mod in AeroDyn. See AeroDyn’s coordinate system documentation for more information: 4.2.2. Coordinate systems — OpenFAST v4.1.0 documentation.
  • AeroDyn’s skewed wake correction is based on a generic skewed wake angle, expressing the total skew combining the effects of yaw and tilt.

Best regards,

3

Dear @Jason.Jonkman,

Thank you very much for the reply.
I will check the document carefully.

Best regards,
Chaozhi Qiu