OpenFAST Tower Deflection Under-Estimated

Wind & Water Computer-Aided Engineering Software Tools
1

Hi All,

I am modeling the IEA-10MW turbine on a floating platform. I have modified the tower and run an analysis in BModes to fit out the mode shapes and frequencies. I am confident that the mode shapes are correct.

When I simulate the model in OpenFAST, I get suspicious tower response. I ran a simulation wherein:

  1. Rotor speed and blade pitch were held constant (no control, Gen DOF = False)
  2. Steady wind @ 13.5 m/s
  3. All structural DOFs enabled

Taking an FFT of the Q_TFA1 result shows this:

image
image703×597 9.76 KB

This was after further stiffening the tower, thinking that the tower resonance I was seeing was my first modification attempt still being too close to the 3P operating frequency. Additionally, the actual tower deflection is VERY small:

image
image715×640 15.7 KB

This makes me think that OpenFAST is overestimating the stiffness of the tower. I checked the rotor thrust output and it is ~1200kN, and doing a quick manual calculation the static tower deflection should be about 0.14m, which I did in Matlab like this:

M = 1200000 * (newHt - tower.Elevation);
Curvature = M./tower.TwFAStif;
Theta = cumtrapz(tower.Elevation,Curvature);
Deflection = cumtrapz(tower.Elevation,Theta)

My ElastoDyn tower looks like this:

------- ELASTODYN V1.00.* TOWER INPUT FILE -------------------------------------
Generated with MATLAB tower properties driver 
---------------------- TOWER PARAMETERS ----------------------------------------
11                     NTwInpSt    - Number of input stations to specify tower geometry
1.0                    TwrFADmp(1) - Tower 1st fore-aft mode structural damping ratio (%)
1.0                    TwrFADmp(2) - Tower 2nd fore-aft mode structural damping ratio (%)
1.0                    TwrSSDmp(1) - Tower 1st side-to-side mode structural damping ratio (%)
1.0                    TwrSSDmp(2) - Tower 2nd side-to-side mode structural damping ratio (%)
---------------------- TOWER ADJUSTMENT FACTORS ------------------------------
1.0                    FAStTunr(1) - Tower fore-aft modal stiffness tuner, 1st mode (-)
1.0                    FAStTunr(2) - Tower fore-aft modal stiffness tuner, 2nd mode (-)
1.0                    SSStTunr(1) - Tower side-to-side stiffness tuner, 1st mode (-)
1.0                    SSStTunr(2) - Tower side-to-side stiffness tuner, 2nd mode (-)
1.0                    AdjTwMa     - Factor to adjust tower mass density (-)
1.0                    AdjFASt     - Factor to adjust tower fore-aft stiffness (-)
1.0                    AdjSSSt     - Factor to adjust tower side-to-side stiffness (-)
---------------------- DISTRIBUTED TOWER PROPERTIES ----------------------------
  HtFract       TMassDen         TwFAStif       TwSSStif
   (-)           (kg/m)           (Nm^2)         (Nm^2)
0.000000000000000e+00  1.917165418002903e+04  5.543311365245792e+12  5.543311365245792e+12
9.949824860361640e-02  1.824232866916633e+04  4.775666711761255e+12  4.775666711761255e+12
1.989018271324434e-01  1.610411113499036e+04  3.810408420103280e+12  3.810408420103280e+12
2.984000757360599e-01  1.411100012842773e+04  3.008551066454783e+12  3.008551066454783e+12
3.978036542648869e-01  1.338304444997033e+04  2.566561082972286e+12  2.566561082972286e+12
4.973019028685033e-01  1.160650772637432e+04  1.992355932286332e+12  1.992355932286332e+12
5.967054813973303e-01  9.964050224125884e+03  1.525293366530144e+12  1.525293366530144e+12
6.962037300009468e-01  8.451518159698828e+03  1.149112790713548e+12  1.149112790713548e+12
7.956073085297738e-01  6.186913747460687e+03  7.451822089485516e+11  7.451822089485516e+11
8.951055571333901e-01  4.975599365823232e+03  5.275547630880361e+11  5.275547630880361e+11
1.000000000000000e+00  4.643445033141643e+03  4.287993905217201e+11  4.287993905217201e+11
---------------------- TOWER FORE-AFT MODE SHAPES ------------------------------
     0.5219   TwFAM1Sh(2) - Mode 1, coefficient of x^2 term
     1.3105   TwFAM1Sh(3) -       , coefficient of x^3 term
    -2.1424   TwFAM1Sh(4) -       , coefficient of x^4 term
     2.0489   TwFAM1Sh(5) -       , coefficient of x^5 term
    -0.7389   TwFAM1Sh(6) -       , coefficient of x^6 term
     6.1138   TwFAM2Sh(2) - Mode 2, coefficient of x^2 term
     8.3540   TwFAM2Sh(3) -       , coefficient of x^3 term
   -23.0055   TwFAM2Sh(4) -       , coefficient of x^4 term
    19.2777   TwFAM2Sh(5) -       , coefficient of x^5 term
    -9.7400   TwFAM2Sh(6) -       , coefficient of x^6 term
---------------------- TOWER SIDE-TO-SIDE MODE SHAPES --------------------------
     0.5130   TwSSM1Sh(2) - Mode 1, coefficient of x^2 term
     1.2832   TwSSM1Sh(3) -       , coefficient of x^3 term
    -2.0713   TwSSM1Sh(4) -       , coefficient of x^4 term
     1.9734   TwSSM1Sh(5) -       , coefficient of x^5 term
    -0.6983   TwSSM1Sh(6) -       , coefficient of x^6 term
     3.8012   TwSSM2Sh(2) - Mode 2, coefficient of x^2 term
     5.9813   TwSSM2Sh(3) -       , coefficient of x^3 term
   -15.7293   TwSSM2Sh(4) -       , coefficient of x^4 term
    14.2215   TwSSM2Sh(5) -       , coefficient of x^5 term
    -7.2747   TwSSM2Sh(6) -       , coefficient of x^6 term

The stiffness values are comparable to the IEA-15MW system, and in comparing the input files everything appears in order. I’m not sure where else to look to debug this.

2

Update: I ran the model with the platform DOFs disabled, and the tower deflection went down to zero-mean:

image
image710×632 14.4 KB

Despite the RotThrust output holding steady around 1200kN.

3

Dear @Ian.Ammerman,

I’m not sure I understand what your question is; can you clarify?

Best regards,

4

Dear @Jason.Jonkman,

Apologies for the clarity - I am looking for some guidance on what might cause this result. It seems erroneous enough to me that this is a modelling error rather than design issue.

From the original IEA-10MW model, I modified the ElastoDyn tower file with the updated stiffnesses and mode shapes from BModes. (I also updated the ElastoDyn to include the platform, and added MoorDyn, but in my follow-up above, those DOFs were disabled in ElastoDyn and the .fst).

I ran constant wind with locked blade pitch, rotor speed, and only looking at tower DOFs with a fixed base and got this result:

image
image739×595 29.3 KB

The thrust force plot is pretty constant at 1200kN. Since the main change that I made was to update the tower, is it possible that the mode shapes I calculated from the sheet are incorrect, leading to this performance? I’m not really sure what else could cause this.

For reference, from BModes and the fitting sheet, the first fore-aft mode looks like this (from input page):

image
image660×450 5.7 KB

5

Dear @Ian.Ammerman,

Can you clarify what you are expecting to see that you are not seeing in your result?

Note that in addition to the rotor thrust force, there is likely a large moment applied at the tower-top associated with the weight of the rotor-nacelle assembly offset from the tower centerline.

Best regards,