search

OpenFAST / openfast

Main repository for the NREL-supported OpenFAST whole-turbine and FAST.Farm wind farm simulation codes.
http://openfast.readthedocs.io
Apache License 2.0
664 stars 451 forks source link

The Tower Natural Frequency of the Tower of the NREL 5 MW-OC4-DeepCwind SemiSubmersible wind turbine. #1980

Open 126qoq opened 7 months ago

126qoq commented 7 months ago

Dear Dr. Jonkman,

I am now working om the Test 25 of the NREL 5 MW-OC4-DeepCwind SemiSubmersible wind turbine. I designed two models, one is a normal operation floating wind turbine model, and for the other one all six platform degrees was disabled. The wind speed was set to be the rated wind speed in turbulent wind, and the waves were irregular waves of 6 m, 10 s. FFT analysis of the tower top acceleration of the two models was carried out. I found that the first-order natural frequency of the tower appeared was 0.463 Hz for the normal model and 0.405 Hz for the constraints, as shown in Fig. 1. The FFT analysis of the acceleration in the SS direction shows 0.442 Hz for the normal model and 0.376 Hz for the constraints, as shown in Fig. 2. Here are my questions.

  1. I would like to know why there is a large difference in the first order natural frequency of the tower obtained by analyzing the accelerations of the top of the tower in the FA direction and the SS direction using the normal model.
  2. I would like to figure out the actual first order natural frequency of the tower for these two models in the FA and SS direction.
  3. I would like to know why for the normal model with all six platform degrees disabled, the first-order tower natural frequency obtained I get through FFT differ from the results from the linearization of the wind turbine explored, which can be 0.43 Hz in the FA SS direction

002 Figure 1 003 Figure 2

Best regards, Yao Ge

jjonkman commented 7 months ago

Dear @126qoq,

Here are my responses:

  1. The tower-top boundary conditions are quite different in the fore-aft and side-side directions, e.g, different rotor-nacelle assembly (RNA) inertias, different blade flexibility flapwise and edgewise, as well as the impact of the free generator rotation on the side-side mode.
  2. Computing FFTs of time series from simulations with stochastic excitation is one way to identify full-system natural frequencies. Full system linearization of the OpenFAST model, followed by eigenanalysis is another way.
  3. I'm not sure. Can you clarify the approach you used to compute the tower natural frequencies through linearization? Where the same degrees of freedom in the RNA enabled in both models (FFT versus linearization)?

Best regards,

126qoq commented 7 months ago

Dear Dr. Jonkman,

Thank you for your answer. I didn't actually do the linearization.The tower natural frequencies through linearization are my quotes from other people's calculations: https://forums.nrel.gov/t/the-frequency-component-of-flapwise/1730/4 Actually, I'm more interested in figuring out if my calculations via FFT make sense or not.

Best regards,

jjonkman commented 7 months ago

Dear @126qoq,

The preferred way to calculate the full-system natural frequencies is through an OpenFAST linearization analysis. I suggest using that to verify your estimation of natural frequencies through FFT.

Best regards,