Comparing OpenFAST and steady-state performance from ccblade for the IEA 3.4 MW onshore wind turbine

Wind & Water Computer-Aided Engineering Software Tools
1

Dear wind turbine enthusiasts,

I am working on the IEA 3.4 MW onshore wind turbine.

I decided to compare between CCblade and OpenFAST.

The steady-state rotor performance from CCblade could be found here.

Regarding OpenFAST:

  • i have disabled all DoFs

  • I have set the Wake induction model in AeroDyn to steady BEM

  • The inputs used for the OpenFAST model are the wind speed, the rotor speed and the blade pitch. All three parameters were obtained from here. I have automated the computations using MATLAB.

  • Regarding the outputs from OpenFAST, i take the mean value of the time series after ignoring the first 30 seconds. However, i found discrepancies which seems to me odd.

Did anyone perform the comparison and obtain good agreement ?

Thank you in advance,

Best Regards,

Riad

2

Dear @Riad.Elhamoud,

Can you share the results that you’ve generated?

Best regards,

3

Dear @Jason.Jonkman,

Thank you for your reply.

After i have posted on the form, i re-checked the input files and i found that i forgot to disable rotor speed and the driverttain DoFs. I re-run the 50 wind speeds presented here.

I found one inacceptable discrepancy for the blade root flapwise moment which is called in OpenFAST oulist “RootMyb”. Also the generator power called “GenPwr” was negative (inacceptable!). For the other outputs the discrepancy is lower than 5%. In the following, kindly find attached a table showing the comparison between OpenFAST and CCBlade.

Capture
Capture1163Ă—609 27.1 KB

An other question that comes to my mind: Why the results from OpenFAST are periodic but they are not constant while those from CCBlade are constant. I assme that CCBlade gives the value at a time. If you agree what is this time ?

Thanks a lot in advance.

Best Regards,

Riad

4

Dear @Riad.Elhamoud,

Just a few comments:

  • Regarding the flapwise root moment, I suspect this difference is caused by the weight of the blade, which is captured in OpenFAST, but not in CCBlade. You can output the aerodynamic-only blade root loads through AeroDyn.
  • Regarding the electrical power, this is an output of ServoDyn and will depend on the generator or torque controller you’ve enabled. When you disable GenDOF in ElastoDyn, the generator torque coming from ServoDyn won’t have any impact on the simulation results, but the ServoDyn output will still compute electrical power and my not be physically correct if the ServoDyn settings are not appropriate.
  • I’m not an expert on CCBlade, but my understanding is that CCBlade computes its quantities at user-specified number of azimuth angles, and can azimuth-average its results.

Best regards,

1 Like
5

Dear @Jason.Jonkman ,

I re-made a comparison between AeroDyn driver (AeroDyn standalone) and CCBlade.

Although some discrepancies (~10%) in some outputs, however, the discrepancy of blade root moment has been drastically reduced (from 47% to 3%).

This confirms your statement that OpenFAST considers gravity even if all DoFs are disabled.

In the following, kindly find attached a table showing the comparison between OpenFAST and CCBlade as well as the comparison between AeroDyn driver (AeroDyn standalone) and CCBlade.

Capture_1
Capture_1799Ă—746 26.2 KB

Thanks a lot,

Finally, i would add that AeroDyn driver outputs till simulation time - dt. In other word, if i set the simlation length to 630 seconds, it outputs 630-dt. I dont know whether i am doing something wrong.

Best Regards,

Riad

1 Like
6

Riad,

the implementation of BEM in AD is more rigorous than the one in CCBlade. Significantly more efforts have gone into the improvement of AD, and far less to CCBlade. I would trust AD, maybe with the exception of non dimensional coefficients. It is my understanding that OpenFAST defines those accounting for the instantaneous area of the rotor, whereas traditional aeroelastic solvers define rotor diameter as two times the sum of hub radius and blade length along the z direction times the cosine of the rotor precone. In formula, this is 2 x (Rhub + blade length along z) * cos(precone). You don’t care about this as long as the elastic DOFs are kept off.

Regards

1 Like