# Power coefficient/Aerodynamic torque

Dear Dr. Jonkman,
Thanks again for your prompt response. I used FAST 8 and Aerodyn v15.03!
Best regards,
Sina

Dear Jason,

I have performed simulation of WP1.5MW using FAST for the conditions specified below. I observe that Cp is greater than 1 as shown in figures. But it should be below Betz limit. Right? Could you please help me how to interpret this.

Uniform wind speed: 6 m/s.
Pitch controller: No pitch controller. (4 deg constant pitch angle).
Only GenDof is kept on for the gear box.
“WakeMod =0” is chosen.
Simple VScontrol .

From output:
Torque at steady state: 385.4 kN-m (plot not attached)
Rotor speed at steady state:14.8 RPM (plot not attached)

Also, Cp is calculated as the ratio of output power of rotor to the (0.5* air density * Disk Area * wind vel^3). Correct?

Regards,
Kumara

Dear Kumara,

Are you plotting AeroDyn output RtAeroCp?

With WakeMod = 0 in AeroDyn v15, the induced velocity calculation is disabled (meaning the induction is zero and the angle of attack will be determined geometrically), so, you are not using BEM and the Betz limit can be violated. The WakeMod = 0 option should only be used when simulating parked/idling conditions (when not producing power).

I suspect you’ll get a result that makes more sense for an operational wind turbine when enabling BEM by setting WakeMod = 1.

Best regards,

Yes Jason. I am plotting AeroDyn output RtAeroCp.
Yes, Induced velocity calculation is disabled in my simulation. I did not think about the fact that Betz limit is valid only when BEMT is used.
Thanks for pointing that. Now I understand that exceeding Betz limit is sensible, but how is it possible to generate more power than that is available in the wind? Kindly advice me incase i’m missing something fundamental here. Thanks.

Regards,
Kumara

Dear Kumara,

Without induction, energy is not preserved in this simulation, so, energy-related quantities are not accurate. I would set WakeMod = 1.

Best regards,

Dear Jason,

Many thanks for your quick response. I’ll do that.

Regards,
Kumara

Dear all,

Issue: Difference in the aerodynamic power of rotor between Standalone AeroDyn calculation and OpenFAST calculation.

I have calculated Rotor aerodynamic power for WindPact 1.5MW turbine at (pitch=2.2 deg, rotor speed=24RPM, wind speed=12 m/s Uniform) using

1. Standalone version of AeroDyn15.04 and 2) OpenFAST-V2.40 but i see the difference in aerodynamic power of the rotor.

2. RtAeroPwr = 1954.472 kW

3. ‘RtAeroPwr’= 2003.206 kW

In the OpenFAST simulation case some of the settings are GenDOF = 0, CompAero=2, shafttilt=0,precone=0.

I have made sure that the settings are same for both. What could be the reason for this discrepancy?
Also, does the Elastodyn play role in the calculation of the rotor aerodynamic power (“RtAeroPwr’”) in OpenFast?

Regards,
Kumara

Dear Kumara,

If all structural DOFs are disabled in ElastoDyn and if the geometry and wind are set up identically between the OpenFAST and standalone AeroDyn driver simulations, then I would expect the standalone AeroDyn driver to give the same aerodynamic power as OpenFAST. If the power predictions are different, I would expect that structural DOFs are enabled in ElastoDyn or some geometry or wind differs. Enabling structural DOFs in OpenFAST–e.g., blade deflection–can impact the aerodynamic power calculated by AeroDyn.

Best regards,

Dear Jason,

I have made sure that all the parameters are the same between both AeroDyn (stand alone version) and AeroDyn in OpenFAST. But still I see significant difference between the Aerodynamic Power calculated using standalone Aerodyn vs OpenFAST.

Wind speed (m/s) Pitch (deg) Rotor speed (RPM) RtTSR RtArea (m^2) RtAeroPwr (W) RtAeroCp
12 5 15 4.5815 3848.4510 1420687.0 0.3488
12 5 15 4.5815 3848.4510 1417626.0 0.3480

Also, the AeroDyn input file of OpenFAST (V2.40) require some extra inputs like “AA_InputFile”, “DBEMT_Mod”,“tau1_const”, which are not present in the standalone AeroDyn input file. This may be because of the version difference. But I made sure that, the options I chose are agnostic to these differences, unless modifications are made to the way calculations are performed even with the same settings between the two (like “AFAeroMod”=1).

I have attached relevant files. If it is not a big ask, could you please go through the attachment and and let me know if I am making any mistakes? Thanks.

Regards,
Kumara
Test_AeroDYN.zip (60.4 KB)

Dear Kumara,

I did a quick comparison between your OpenFAST model and standalone AeroDyn model and don’t see any obvious differences that would lead to different power.

I’m guessing the problem is that you are using two different versions of AeroDyn between the cases. You mention that you are using OpenFAST v2.4 and standalone AeroDyn v15.04. While we don’t provide precompiled Windows executable of the standalone modules (like AeroDyn) with each OpenFAST release, the source code needed to compile the standalone AeroDyn driver is included in the OpenFAST v2.4 repository. The AeroDyn driver input file has not changed between AeroDyn v15.04 and OpenFAST v2.4. So, all you should have to do is compile the standalone AeroDyn driver for OpenFAST v2.4, use your AeroDyn driver file together with the AeroDyn input file(s) from your OpenFAST v2.4 model, and rerun the driver-level simulation.

Best regards,

Dear Sina Sir,
I am also facing the same error while running the file Run_OpenLoop.m
Error using Run_OpenLoop (line 30)
Error reported by S-function ‘FAST_SFunc’ in ‘OpenLoop/FAST Nonlinear Wind Turbine/S-Function’:

Please guide how to solve this.
Thanks and regards,
Shivaji

Dear Shivaji,

This error means that the input file “Test01.fst” does not exist, i.e., the filename is wrong or the path to it is incorrect.

As I told Sina, "In the transition from FAST v8 to OpenFAST, what used to be the CertTest of FAST v8 has been changed to the regression tests (r-test) of OpenFAST: github.com/OpenFAST/r-test. The primary input file names have also been changed.

Best regards,

Dear Jason,

I wanted to look at the change of the power coefficient Cp of a waked wind turbine. Thus, in FAST.Farm, I considered 2 wind turbines, one 4*D upstream the second one (D is the rotor diameter), and I plot the Tip-Speed Ratio (RtTSR), the rotor speed (RtSpeed), the averaged-disk wind speed (RtVAvgxh), the power coefficient (RtAeroCp), and the output power of the turbines.
I am surprised to have a Cp of the waked wind turbine increasing higher than the Cp_max =0.48, even though I made sure that the controllers are on (pitch & torque), and WakeMod=1. I checked also TSR and RtSpeed, and they seem to be the ones I expected (referring to the plots in the 5MW NREL Definition)
I was expecting a decrease of Cp for the second turbine as the TSR increases (going from Region 2 to Region 1/2). Do you have any idea where this issue may come from?

Kindest regards

Younes

Dear Younes,

Are you running a FAST.Farm simulation using OpenFAST models of the NREL 5-MW baseline wind turbine or something else? What inflow conditions are you simulating in FAST.Farm (presumably steady, uniform inflow based on the smooth curves in your plots)? What structural DOFs and aerodynamic settings are enabled in your OpenFAST models?

I would generally expect the NREL 5-MW baseline wind turbine to have a Cpmax of 0.482 at a TSR of 7.55, but that was computed a while ago and different aerodynamic settings in AeroDyn v15 or enabling/disabling structural DOFs may have some effect on that. The peak of the Cp-TSR surface is also fairly flat for small changes in TSR, so, I could see that small variations in TSR will may have a small effect on Cp. However, a Cp of 0.54 sounds a bit higher than I would expect.

Best regards,

Dear Jason,

Yes, I am using the NREL 5MW baseline. Otherwise, where can I download its files again?

For the structural DOFs, only the Generator is enabled as can be seen here:

[/code]

Here is my AeroDyn file as well:

[/code]

I agree with your statements and I was expecting the same. Do you think the issue comes from one of the files?

Sorry for the inconvenience

Kindest regards

Younes

Dear Younes,

OK, thanks for clarifying. You can find a complete OpenFAST model of the NREL 5-MW baseline wind turbine from the OpenFAST r-test (github.com/OpenFAST/r-test), but I suspect you are using the correct data anyway.

Another thought I have is that the inflow to the second wind turbine is not uniform, but includes a (likely Gaussian-shaped) wake deficit from the upwind turbine. Perhaps the value of the rotor disk-averaged wind speed (RtVAvgxh), which is also used to compute the rotor aerodynamic power coefficient (RtAeroCp), is not computing the disk-averaged wind speed accurately enough for this inflow condition. RtVAvgxh is computed simply by summing the relative velocity at each aerodynamic analysis node along each blade and dividing by the total number of nodes. In reality, the more outboard nodes should have a higher weighting in this summation than the inboard nodes because they sweep a larger area, but this weighting is not included in the spatial averaging. You could look at the inflow to the second turbine from the FAST.Farm outputs and manually compute a more accurate representation of the disk-averaged wind speed and associated rotor aerodynamic power coefficient. I suspect if you do that, you’ll get a TSR and Cp that is more consistent with the upstream turbine.

Best regards,

Dear Jason,

Thank you so much for the reply and guidance. For the maximum power coefficient, I found using the AeroDyn module that it is Cp=0.4862, for a TSR_opt=7.55 and Pitch_opt=0°, as we discussed it in the “NREL 5-MW reference turbine - CP, CQ, CT Coefficients” topic [url]Question #1 for turbulence fields (points outside of field)], which is very close to the 0.482 found a while ago. I think the difference comes from the blade nodes considered while computing Cp, and the different aerodynamic settings in AeroDyn v15, and/or enabling/disabling structural DOFs may have some effect, as you highlighted.

Regarding your suggestion for the computation of RtVAvgxh, I compute it now by double integrating the local wind speeds, in the following nodes, on the swept area divided by (R^2)*pi.

```The nodes: OutRadii = 0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 51 57 60 63 dr=1 (m) ```

The results on the waked wind turbine are:
RtVAvgxh= 6.8052 m/s
TSR_waked= 8.1119
Cp_waked= 0.4829

Which are what I expected when looking at the 5MW characteristics, and also not in the second region anymore, but in region 1/2 (Cp < Cp_max):

Thank you so much again

Kindest regards

Younes

Hi Jason,
A basic question as I haven’t had much experience with openFAST.
I am working with r-tests. But I am struggling to get around the naming convention of r-tests. ReadMe file doesn’t have much information. Also, I didn’t get much info from the “naming scheme from testNN”. E.g what does “5MW_ITIBarge_DLL_WTurb_WavesIrr” mean?
I need to run the simulation for a 5MW, three-bladed,variable-speed offshore(fixed monopile) turbine. Which test should I refer to?
Warm regards

Dear @Ahmed.Hassan,

The r-test named 5MW_ITIBarge_DLL_WTurb_WavesIrr is a simulation of the NREL 5-MW baseline wind turbine atop the floating offshore ITI Energy barge with the controller DISCON DLL excited by wind turbulence and irregular waves.

A description of most of the OpenFAST r-tests is provided in Table 6 of the old FAST v8 ReadMe: https://openfast.readthedocs.io/en/main/_downloads/5f2ddf006568adc9b88d8118dc3f1732/FAST8_README.pdf (although in this table, the tests were numbered from Test01-Test26). The previous CertTest cases from FAST v8 were migrated into what is now the r-test in OpenFAST.

A model of the NREL 5-MW baseline wind turbine atop the fixed offshore OC3-monopile with fixed foundation is provided in the following OpenFAST r-test: r-test/glue-codes/openfast/5MW_OC3Mnpl_DLL_WTurb_WavesIrr at main · OpenFAST/r-test · GitHub.

Best regards,

1 Like