Does errors in simulations with poor initial conditions indicates a faulty model?

Provide feedback, request enhancements, and get help with wind-turbine computer-aided engineering tools.

Moderators: Bonnie.Jonkman, Jason.Jonkman

Pedro.Pozzobon
Posts: 7
Joined: Tue Jul 07, 2020 3:08 pm
Organization: University of Sao Paulo
Location: Brazil

Does errors in simulations with poor initial conditions indicates a faulty model?

Postby Pedro.Pozzobon » Wed Mar 24, 2021 2:37 pm

Hi,

I'm currently facing some strange behaviour when simulating my model at high wind speeds (close to cut-out).

- if I disable FlapDOF1 and FlapDOF2, the simulations runs fine with any initial conditions (even poor ones)
- if all DOFs are enabled, the slightest deviation in pitch initial conditions leads to an aborted simulation due to tower strike (excessive blade deflection), or an irregular pitch motion. For example, a 15°pitch ( which is the expected pitch value for the mean wind speed) initial condition makes the simulation run fine, but a 20° initial pitch fails. This happens even when the pitch controller is deactivated.
- If I set a more "realistic" initial condition of pitch = 90°and RotSpeed = 0 rpm, the simulations runs fine.

This does not happen for lower wind speeds (simulations run fine even with poor initial conditions)

That said,
1 - The sensitivity to initial conditions is a numerical problem or does it have some physical meaning?
2 - Is this a sign that something is 'wrong' with my model (perhaps the flap vibration mode is coupling with another mode?)

Best regards,

Jason.Jonkman
Posts: 5738
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: Does errors in simulations with poor initial conditions indicates a faulty model?

Postby Jason.Jonkman » Sat Mar 27, 2021 7:23 am

Dear Pedro,

I'm not sure I understand what you are saying. You seem to say that a deviation in pitch initial condition results in irregular pitch motion while the pitch controller is deactivated, which does not make sense to me. Can you clarify?

It is generally recommended to set the initial rotor speed and initial blade-pitch angles to their respective mean values associated with the mean wind speed being simulated. Otherwise, the speed and pitch may be outside a region the controller was designed for and result in unstable control response and never recover (worst-case scenario), or a start-up transient that is unrealistic and lasts a long time.

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

Pedro.Pozzobon
Posts: 7
Joined: Tue Jul 07, 2020 3:08 pm
Organization: University of Sao Paulo
Location: Brazil

Re: Does errors in simulations with poor initial conditions indicates a faulty model?

Postby Pedro.Pozzobon » Mon Mar 29, 2021 6:29 am

Dear Jason,

The deviation in the pitch initial value causes the simulations to fail even when I deactivate the pitch controller (ServoDyn, PCMode = 0). I simulated this condition to test if it the controller was causing the large blade deflections.

Re-writing my observations in a more clear way:
- if I disable FlapDOF1 and FlapDOF2, the simulations runs fine with any initial conditions (even poor ones) and any controller configuration (activated/deactivated)
- if all DOFs are enabled, the slightest deviation in pitch initial conditions leads to an aborted simulation, for any controller configuration. On some simulations with the controller activated there is an irregular pitch motion (depending on the seed)
- If I set a more "realistic" initial condition of pitch = 90°, RotSpeed = 0 rpm, and pitch controller activated, all simulations runs fine.

Thanks for your time.

Best regards,

Jason.Jonkman
Posts: 5738
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: Does errors in simulations with poor initial conditions indicates a faulty model?

Postby Jason.Jonkman » Mon Mar 29, 2021 6:55 am

Dear Pedro,

Are you running a FAST / OpenFAST model provided by NREL or are these your own input files?

I'm not sure I have more to add at this time, but it sounds like you at least have a workaround.

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

Pedro.Pozzobon
Posts: 7
Joined: Tue Jul 07, 2020 3:08 pm
Organization: University of Sao Paulo
Location: Brazil

Re: Does errors in simulations with poor initial conditions indicates a faulty model?

Postby Pedro.Pozzobon » Mon Mar 29, 2021 7:19 am

Dear Jason,

I'm running OpenFAST with my own input files. These files were made to model a commercial wind turbine.

The workaround works. I just worry that these issues may indicate some problem with my model.

Based on your experience, these kind of issues may happen even with a model based on a commercial wind turbine?

Thanks for your prompt response.

Best regards,

Jason.Jonkman
Posts: 5738
Joined: Thu Nov 03, 2005 4:38 pm
Location: Boulder, CO
Contact:

Re: Does errors in simulations with poor initial conditions indicates a faulty model?

Postby Jason.Jonkman » Mon Mar 29, 2021 7:53 am

Dear Pedro,

This is not a symptom of a commercial wind turbine model. I would guess that the controller you are using is highly sensitive to the initial conditions being set.

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


Return to “Computer-Aided Engineering Software Tools”

Who is online

Users browsing this forum: No registered users and 1 guest