Model verification

This forum if for discussing controls. Questions about how to implement controls in FAST are more appropriate to the CAE Tools forum.

Moderator: Bonnie.Jonkman

Jason.Jonkman
Posts: 4539
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: Model verification

Postby Jason.Jonkman » Mon Jan 13, 2020 1:40 pm

Dear Sina,

I see that you've enabled tower potential flow (TwrPotent=1). You should get constant results after disabling this feature (TwrPotent=0).

Best regards,
Jason Jonkman, Ph.D.
Senior Engineer | National Wind Technology Center (NWTC)

National Renewable Energy Laboratory (NREL)
15013 Denver West Parkway | Golden, CO 80401
+1 (303) 384 – 7026 | Fax: +1 (303) 384 – 6901
nwtc.nrel.gov

Sina.Ameli
Posts: 68
Joined: Mon Oct 15, 2018 11:23 am
Organization: FSU
Location: Florida

Re: Model verification

Postby Sina.Ameli » Mon Jan 13, 2020 7:37 pm

Dear Dr. Jonkman,
Thank you so much. Now, I can get a fixed and clean signal. But there is a big problem! When I apply different inputs (pitch angles) and different step wind inflows, I get the same rotor speed that is always equal to the initial condition! It seems that it plays as a gain of one that directly send out its input (rotor speed as its initial condition) directly to the output. It seems that the control signal (pitch angel) and the wind speed do not any affect on the rotor speed!

In addition, when I apply the input of 10.45 deg and the wind speed of 15 m/s I get this power coefficient:

Power Coefficient.PNG
CP
Power Coefficient.PNG (63.91 KiB) Viewed 722 times


, which is very weird because it should not be negative!



I have another question. When I create my own folder, I get this error:

Error reported by S-function 'FAST_SFunc' in 'OpenLoop_ModelVerfication/FAST
Nonlinear Wind Turbine/S-Function':
FAST_InitializeAll:FAST_Init:FAST_ReadPrimaryFile:OpenFInpFile:The input file,
"..\..\CertTest\Test24.fst", was not found.


I can only run Simulink files located exactly in the "Samples" folder. It seems that there might be some files located in this folder that needed to be located to other folders of mine. However, I have already added the "ReadFASTbinary", but still gets the same error.

Thanks
Best regards,
Sina
Last edited by Sina.Ameli on Tue Jan 14, 2020 7:55 am, edited 1 time in total.

Jason.Jonkman
Posts: 4539
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: Model verification

Postby Jason.Jonkman » Tue Jan 14, 2020 7:55 am

Dear Sina,

When you disable the generator DOF, the generator will rotate at a fixed speed (the initial speed) regardless of the torques applied. If you want time-varying generator speed, you must enable the generator DOF in ElastoDyn (GenDOF = True).

The error you are getting is not related to the location of ReadFASTBinary.m, but is related to the location of your FAST primary input file relative to the location of the FAST-SFunction. If the path and filename "..\..\CertTest\Test24.fst" is incorrect, fix it.

Best regards,
Jason Jonkman, Ph.D.
Senior Engineer | National Wind Technology Center (NWTC)

National Renewable Energy Laboratory (NREL)
15013 Denver West Parkway | Golden, CO 80401
+1 (303) 384 – 7026 | Fax: +1 (303) 384 – 6901
nwtc.nrel.gov

Sina.Ameli
Posts: 68
Joined: Mon Oct 15, 2018 11:23 am
Organization: FSU
Location: Florida

Re: Model verification

Postby Sina.Ameli » Tue Jan 14, 2020 8:42 am

Dear Dr. Jonkman,

Thanks for your prompt response. I return the DOF of the generator. But there is a problem. When I apply equilibrium points for the pitch and wind speed and with the initial condition in rotor speed at its rated value, then I expect that it stays in its rated value for the rotor speed, but it increases the rotor speed and track the value above the rated value making the CP (power coefficient) is zero. It seems strange because a nonlinear plant at its equilibrium points along with its initial conditions should stay near it.

Thanks,
Best regards,
Sina

Jason.Jonkman
Posts: 4539
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: Model verification

Postby Jason.Jonkman » Tue Jan 14, 2020 8:50 am

Dear Sina,

It sounds like you've enabled the generator DOF, but haven't activated any generator torque within ServoDyn, thus, the rotor is accelerating until the aerodynamic power drops to zero.

Best regards,
Jason Jonkman, Ph.D.
Senior Engineer | National Wind Technology Center (NWTC)

National Renewable Energy Laboratory (NREL)
15013 Denver West Parkway | Golden, CO 80401
+1 (303) 384 – 7026 | Fax: +1 (303) 384 – 6901
nwtc.nrel.gov

Sina.Ameli
Posts: 68
Joined: Mon Oct 15, 2018 11:23 am
Organization: FSU
Location: Florida

Re: Model verification

Postby Sina.Ameli » Tue Jan 14, 2020 12:33 pm

Dear Dr. Jonkman,

Thanks for your fast response. Yes, it makes sense. I corrected it. But there is a problem. Actually, the rotor speed does not stay at its initial value when I apply equilibrium points in it. It has a little steady-steady state error. It finally goes to 1.24 rad/s instead of the rated value that is 1.267 rad/s! When I apply different operating points, I get a different rotor speed. For instance, for the if the wind speed is 22m/s and pitch angle is 19.94 deg, the rotor speed goes from 1.267 to 1.24. And when I apply a wind speed of 15m/s and the pitch angle of 10.45 deg, it goes to 1.18! What is the underlying cause of this problem?

Thanks,
Best regards,
Sina

Jason.Jonkman
Posts: 4539
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: Model verification

Postby Jason.Jonkman » Tue Jan 14, 2020 1:04 pm

Dear Sina,

I gather based on your description that the aerodynamic torque and generator torque are not balanced for the given wind speed, blade-pitch angle, and generator torque, so, the rotor speed will accelerate/decelerate until they are balanced.

Best regards,
Jason Jonkman, Ph.D.
Senior Engineer | National Wind Technology Center (NWTC)

National Renewable Energy Laboratory (NREL)
15013 Denver West Parkway | Golden, CO 80401
+1 (303) 384 – 7026 | Fax: +1 (303) 384 – 6901
nwtc.nrel.gov

Sina.Ameli
Posts: 68
Joined: Mon Oct 15, 2018 11:23 am
Organization: FSU
Location: Florida

Re: Model verification

Postby Sina.Ameli » Tue Jan 14, 2020 1:26 pm

Dear Dr. Jonkman,

Thanks for responding me with celerity. So, how can I balance them? Do you mean because I deactivated structural degree of freedom it is unbalanced? Do you mean now the equilibrium points has changed due to changes of the dynamics?

Thanks,
Best regards,
Sina

Jason.Jonkman
Posts: 4539
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: Model verification

Postby Jason.Jonkman » Tue Jan 14, 2020 1:58 pm

Dear Sina,

Yes, you've disabled the structural DOFs and eliminated shaft tilt, so, the aerodynamic loads differ, and the blade-pitch angles appropriate to each mean wind speed will change. You'll have to calculate the blade-pitch angle at each wind speed necessary to obtain the desired rotor speed, given the desired torque. Or simply enable the blade-pitch controller and it will do that for you.

Best regards,
Jason Jonkman, Ph.D.
Senior Engineer | National Wind Technology Center (NWTC)

National Renewable Energy Laboratory (NREL)
15013 Denver West Parkway | Golden, CO 80401
+1 (303) 384 – 7026 | Fax: +1 (303) 384 – 6901
nwtc.nrel.gov

Sina.Ameli
Posts: 68
Joined: Mon Oct 15, 2018 11:23 am
Organization: FSU
Location: Florida

Re: Model verification

Postby Sina.Ameli » Tue Jan 14, 2020 2:04 pm

Dear Dr. Jonkman,

Thank you so much. I got it.

Best regards,
Sina

Sina.Ameli
Posts: 68
Joined: Mon Oct 15, 2018 11:23 am
Organization: FSU
Location: Florida

Re: Model verification

Postby Sina.Ameli » Fri Jan 17, 2020 3:05 pm

Dear Dr. Jonkman,

When I apply a zero input (pitch angel is zero) fpr test 24, I will get this error message:

An error occurred while running the simulation and the simulation was terminated
Caused by:
Error reported by S-function 'FAST_SFunc' in 'PI2_FAST/FAST Nonlinear Wind Turbine/S-Function':
FAST_Solution:FAST_AdvanceStates:AD_UpdateStates:BEMT_UpdateStates(node 6, blade 1):BEMT_UnCoupledSolve:DeterminePhiBounds:There is no valid value of phi for these operating conditions! Vx = -1.295, Vy = 34.21, rlocal = 15.852, theta = 0.20019

What is the cause of this error?
I also get an error of this :

an error occurred while running the simulation and the simulation was terminated
Caused by:
Error reported by S-function 'FAST_SFunc' in 'PI2_FAST/FAST Nonlinear Wind Turbine/S-Function':
FAST_Solution:FAST_AdvanceStates:AD_UpdateStates:BEMT_UpdateStates(node 7, blade 2):BEMT_UnCoupledSolve:DeterminePhiBounds:There is no valid value of phi for these operating conditions! Vx = 1.5908, Vy = -24.552, rlocal = 20.113, theta = 0.1497

When I apply a nominal pitch angel for test 18!

Besides, I implement a PI control and the input of the system is equilibrium point, but I get a very oscillatory response for Test 24 like this:
Capture.JPG
Rotor Speed
Capture.JPG (132.78 KiB) Viewed 557 times

while I remember that Test 18 didn't give me these much oscillations with a PI control in the operating point! I don't understand why it is too much oscillatory with only a step wind?!

Thanks,
Best regards,
Sina

Jason.Jonkman
Posts: 4539
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: Model verification

Postby Jason.Jonkman » Sat Jan 18, 2020 1:41 am

Dear Sina,

The error regarding "no valid value of phi" has been discussed many times on this forum was fixed in versions of AeroDyn v15 newer than FAST v8.16. I suggest that you upgrade from FAST v8.16 to OpenFAST: https://github.com/OpenFAST/openfast.

Regarding the large oscillations in response, I can't really comment more without understanding more about your simulation settings.

Best regards,
Jason Jonkman, Ph.D.
Senior Engineer | National Wind Technology Center (NWTC)

National Renewable Energy Laboratory (NREL)
15013 Denver West Parkway | Golden, CO 80401
+1 (303) 384 – 7026 | Fax: +1 (303) 384 – 6901
nwtc.nrel.gov

Sina.Ameli
Posts: 68
Joined: Mon Oct 15, 2018 11:23 am
Organization: FSU
Location: Florida

Re: Model verification

Postby Sina.Ameli » Sat Jan 18, 2020 5:11 pm

Dear Dr. Jonkman,
Thank for your reply. Actually, I have several questions.

1- Regarding the oscillations, you're right. You cannot tell anything about the those oscillations without having detailed information. Let me ask my question in other word. Suppose I apply the control input (pitch) in an operating point, say 19.94 deg, and the exogenous input, say a step wind speed of Vx=22m/s, Vy=0; Vz=0, starting at the initial condition for the pitch angel exactly at the equilibrium point, i.e 19.94 deg, and the generator speed of 12.1 rpm, then the generator speed and the pitch angel for Test24 compared with the PI control (the one you exactly used in your documents for offshore WT) are as follows:
Generator Speed.PNG
Generator Speed
Generator Speed.PNG (90.21 KiB) Viewed 512 times

Pitch.PNG
Pitch
Pitch.PNG (78.77 KiB) Viewed 512 times

Generator Speed_PI.PNG
PI Control for Test 18 and Test 24
Generator Speed_PI.PNG (82.39 KiB) Viewed 512 times

Both open loop and closed loop have the same initial conditions, and step wind speed.
As in the onshore WT the response settles, is these oscillations normal in the offshore WT with an step wind speed?
I did a little research and found a paper published recently. It shows the results of the pitch control (rotor speed tracking problem) for a 5 MW offshore WT in FAST. The responses settled very well.

2- I compared the primary input files of test 18 with 24. The flags "CompHydro", and "CompMooring" are one in test 24, while for test 18 they are zero. Could you explain what are these physically and what are their effects on WT?

3- Is it possible to get output from blade pitch 2 and 3 in test 18? If so, how?

4- Regarding the OpenFAST, where are the "bin" and "source" files to set path in Matlab? I set path to D:\openfast-master\glue-codes\simulink\src, where the file "FAST_SFunc.c" is located in, and I downloaded the "windows_openfast_binaries" file separately and pasted it to the D:\openfast-master then set path to D:\openfast-master\windows_openfast_binaries\openfast_v2.2.0_binaries\DISCON_DLLS\64bit. When I run the OpenLoop Simulink file, Matlab gives still gives me the following error:
Error in S-function 'OpenLoop/FAST Nonlinear Wind Turbine/S-Function': S-Function 'FAST_SFunc' does not exist\

Besides, where are the primary and other input files for the 5MW offshore WT? Is there any organized documentation to be able to run the Simulink quickly?

Thanks,
Best regards,
Sina

Jason.Jonkman
Posts: 4539
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: Model verification

Postby Jason.Jonkman » Sun Jan 19, 2020 3:00 pm

Dear Sina,

Here are my answers to your questions:

1) In your plots, it doesn't look like the rotor speed is initialized to 12.1 rpm because (1) the generator speed (=rotor speed*97 for the NREL 5-MW turbine) starts well below 1173.7 rpm in your plots and (2) the blade-pitch angle drops suddenly at time zero when control is enabled. Other than that, I'm not sure I can comment more because I'm not sure what controller you've enabled.

2) Test18 is for a land-based NREL 5-MW baseline turbine, which does not have hydrodynamic loads on the substructure nor does it have a mooring system, so HydroDyn and a mooring module are disabled for this simulation. Test24 is a model of the OC3-Hywind spar, which has HydroDyn and MAP++ enabled.

3) Yes, but without modification, the pitch controller in Test18 uses collective pitch, so, the pitch angle of all blades should be the same. Regardless, pitch angles may be output from the ElastoDyn module (that is, include BldPitch1, BldPitch2, and BldPitch3 in the OutList of the ElastoDyn input file).

4) The compiled OpenFAST S-Function does not appear to be included in the "windows_openfast_binaries.zip" file. You'll have to compile the OpenFAST S-Function before using the Simulink-OpenFAST interface. This interface is very similar to the one for FAST v8, which is documented in the FAST v8 ReadMe file: https://wind.nrel.gov/nwtc/docs/README_FAST8.pdf. What was previously the CertTest in earlier versions of FAST (which contained sample input files) has been converted into the regression-test (r-test) in OpenFAST: https://github.com/OpenFAST/r-test.

Best regards,
Jason Jonkman, Ph.D.
Senior Engineer | National Wind Technology Center (NWTC)

National Renewable Energy Laboratory (NREL)
15013 Denver West Parkway | Golden, CO 80401
+1 (303) 384 – 7026 | Fax: +1 (303) 384 – 6901
nwtc.nrel.gov

Sina.Ameli
Posts: 68
Joined: Mon Oct 15, 2018 11:23 am
Organization: FSU
Location: Florida

Re: Model verification

Postby Sina.Ameli » Tue Jan 21, 2020 12:48 pm

Dear Dr. Jonkman,

Thanks for replying to all of my question. Please see the following results for the rotor speed and control signal:
RotorSpeed.JPG
Rotor Speed
RotorSpeed.JPG (109.46 KiB) Viewed 374 times

Pitch.JPG
Pitch
Pitch.JPG (65.14 KiB) Viewed 374 times

Pitch_Beginning.JPG
Pitch_beginning
Pitch_Beginning.JPG (55.91 KiB) Viewed 374 times


The initial conditions are the operating condition (wind speed=22m/s, pitch=19.94 deg, rotor speed=1.267rad/s, Test 24). I used a simple PI control and a just add a bias term to the PI, which is exactly 19.94 deg. I wonder why the rotor speed has a lot of oscillations? (No matter what are the PI gains are, although it tracks near the rated rotor speed value, it always has a lot of oscillations and never settled while I apply rated values in the steady state situations). The problem is that in the open loop, the rotor speed cannot stay in the equilibrium point and has oscillations! As you mentioned that I don't know what is happening at the beginning that pitch angle is suddenly becomes zero then goes to its steady state situation.

Thanks,
Best regards,
Sina


Return to “Controls”

Who is online

Users browsing this forum: No registered users and 1 guest