When I run Hywind concept with no floating platform ( PtfmModel=0) with wind speeds greater than 17 m/s I get the following message.
(I am using constant wind speed)
" Small angle assumption violated in SUBROUTINE SmllRoTrans<> due to a large blade deflection.
The solution may be in accurate.
Future warnings suppressed. Simulation continuing … "
I am wondering how I can simulate the turbine for wind speeds higher than 17 without violating the blade deflection assumptions?
You can reduce the Integration time step DT to see if this violation occurs. Or it only happens at the beginning of the simulation, the rest of simulation should be good. See the explanation of small angle assumption in this post: http://forums.nrel.gov/t/nrel-5mw-controls-dll-interface/183/1
The post Yinping Yang pointed you to is very useful for your question (thanks, Yinping!)
Unless you’ve changed something such as the system masses or stiffnesses, the default time step supplied with the OC3-Hywind model is sufficient to achieve a numerically stable solution in FAST. So, problem (1) from the aforementioned post is not the isssue. I suspect that you’ll be able to eliminate the warning by setting the initial blade pitch angle and blade deflections suitably for the given simulation (i.e., based on their expected values when operating at steady state conditions above 17 m/s).
Thanks to both of you for the posts.
I changed the initial blade pitch angles and I no longer receive the message.
I ran the simulation for different constant wind speeds (8, 11, 18 m/s). I expect to observe larger out of plane blade tip deflection (OoPDefl1) for higher wind speeds.
OoPDefl1 gets larger by altering wind speed from 8 to 11m/s. However, contrary to my expectation, the deflection predicted for wind speed 18 m/s was smaller than the ones of wind speeds 8 and 11.
(for wind speed 8, OoPDefl1 oscillates around 3 m, for 11 m/s it oscillates around 5.2 m but for wind speed 18 it drops to 1.7 m !)
Do you think anything is going wrong with my simulations?
The behaviour you describe is correct and does not indicate a problem with the simulation.
The NREL 5-MW baseline turbine is a blade-pitch-to-feather-controlled turbine. In pitch-to-feather turbines, the thrust peaks at rated wind speed (11.4 m/s for the NREL 5-MW turbine). So, the the blade deflection should increase with wind speed up to rated, then drop with increasing wind speed above rated. This is the same behavoir that is shown in Figure 9-1 of the NREL 5-MW specifications report: nrel.gov/docs/fy09osti/38060.pdf.
Thanks for your helpful explanation.
In that report maximum wind speed is 25 m/s. I did simulation for wind speed 35 m/s and I am getting negative OoPDefl1!
What does negative OoPDefl1 mean here?
Negative OoPDefl means that the blade is deflected upwind of the undeflected cone of rotation.
35 m/s is far beyond the normal operational wind speed of the NREL 5-MW turbine. At very high wind speeds, the blade-pitch angles will become very large so as to keep the power and rotor speed at rated values (5 MW and 12.1 rpm, respectively). While the overall torque and thrust will remain positive in this case, the thrust will be very small and the outer-most portions of the blade can see negative angles of attack–and thus negative lift forces. It is these forces that can deflect the blade upwind. (The innermost portions of the blade will still maintain positive angles of attack–and thus positive lift forces–but while these forces will keep the overall torque and thrust positive, they have less contribution to the blade deflection because the blade is far stiffer inboard than it is outboard.)
Do you know of any tools or publications that discuss how to predict the limit states of a wind turbine under extreme wind conditions?
Limit states can include instability and fatigue failure of blades or tower.
Volume 11, Issue 6 of the Wind Energy journal–the special issue on design load definition–would be a good first place to start: www3.interscience.wiley.com/jour … 8239/issue.
I am trying to develop the loads associated with extreme events for the design of tripod substructure and tower; however, the generator continues to show output above the cut-out wind speed, I’m not sure how to adjust the fast input file so that the turbine uses its rated speeds. I’m new to FAST so any help would greatly be appreciated.
The turbine response (power, speed, etc.) is typically dictated by the controller. It is hard to offer any advice without knowing what type of controller you’re using. Are you using one of the simple controllers built into FAST, a Simulink controller, a DLL controller, or some user-defined routine?
I’m using the NREL 5MW turbine, but altered slightly with the RePower 5MW wind speeds for offshore cut-in, rated, and cut-out. I’m using the FAST_v6.02 so that i didn’t have to recompile FAST to use the BladedDLLInterface, which avoids the deflection error for most wind cases.
I’m not sure what you mean when you say, “I’m using the FAST_v6.02 so that i didn’t have to recompile FAST to use the BladedDLLInterface.” The NREL 5-MW turbine uses a torque and pitch controller in GH Bladed-style DLL format. These controllers require FAST’s interface to GH Bladed-style DLLs (via BladedDLLInterface.f90), which is not a default option in the current version of FAST (v6.01 or v6.02). You can find a version of FAST with this interface here: wind.nrel.gov/public/jjonkman/FA … erface.exe. Or if you’re adventurous, you can recompile FAST using the guidance found in the FAST User’s Guide.
The control system DLL we developed for the NREL 5-MW turbine, as described in its report (nrel.gov/docs/fy09osti/38060.pdf), is only for normal operation. We have not specified logic for start-up and shut-down scenarios. So–without modification–the turbine will continue to operate if you simulate with wind speeds higher than cut-out. What we do when we simulate this turbine using above cut-out wind speeds is that we modify the model so it considers the turbine in idling mode.
As you may remember, I am working on thunderstorm downburst winds which have very high wind speeds in some part (say, 30-60 m/s). You mentioned in previous post that you model the wind turbine in idling mode (or parked mode). How do you (or how can I) do that? Is there any chance that I could test your modified model(s) with thunderstorm downburst winds?
To model an idling rotor in FAST with AeroDyn, specify the following inputs:
In FAST’s primary input file:
PCMode = 0 (disable pitch control)
GenTiStr = True
TimGenOn = 9999.9 (a large don’t care > TMax)
BlPitch1 = BlPitch2 = BlPitch3 = 90.0 (fully feathered)
RotSpeed = 0.0 (or close to zero)
In AeroDyn’s primary input file:
StallMod = STEADY (disable dynamic stall, which isn’t valid at large angles of attack)
IndModel = NONE (disable aerodynamic induction)
More advice is given in the “Simulating Special Events” Section of the “Controls” Chapter of the FAST User’s Guide.
I am trying to simulate an idling rotor under an extreme HH wind speed of 50 m/s. I am using TurbSim to generate the wind files.
I am using the older version of FAST (i.e. v.7) to simulate a monopile supported OWT in a water depth of 20 m. In addition to the input mentioned in your previous post, I am assuming still water conditions. On plotting the base shear at the mudline, I am getting a very lengthy transient period, and i believe the values are too low (my mean is around 95kN, while a published study using HAWC2 had yielded 350 kN under the same conditions). Am i missing something?
I don’t really know how you’ve set up your simulation, so, it is difficult for me to comment on the results, but it appears that it is the tower natural frequency that has been excited (about 3-s period, or 1/3 Hz) and is damping out very slowly. I don’t really see much effect of the turbulence.
Regarding the comparison to HAWC2, one problem with FAST v7 was the inability to model direct wind (drag) loads on the tower, which likely dominate the tower loads when running simulations in high winds with the rotor parked or idling with the blades fully feathered and the shaft aligned with the mean wind direction. FAST v8 enables you to model the tower wind loads; it may be appropriate to upgrade. The NREL 5-MW turbine atop the OC3-monopile (in 20-m water depth) is available as Test19 in the FAST v8 archive.
I am trying to model a “parked” turbine in OpenFAST and I have a question about it. I have read the suggested FAST modifications that were posted before and applied those in my model. The modifications that I made:
I applied 8 mps steady wind only for the model (no wave, WaveMod=0)
in ElastoDyn: GenDOF = false; BlPitch(1),(2),(3) = 90; RotSpeed = 0
in ServoDyn: PCMode = 0 ;
in AeroDyn: WakeMod = 0 ;
OpenFAST run successfully but I am getting a warning like:
Warning: Turning off Unsteady Aerodynamics due to high angle-of-attack. BladeNode = 19, Blade = 1
Warning: Turning off Unsteady Aerodynamics due to high angle-of-attack. BladeNode = 18, Blade = 1
Warning: Turning off Unsteady Aerodynamics due to high angle-of-attack. BladeNode = 16, Blade = 3
Is the warning acceptable if the objective is looking at structure movements (6 DOF) due to environmental forcing (wind)?
Normally you’d also disable unsteady airfoil aerodynamics during a parked/idling simulations with high angles of attack (AFAeroMod=1); this will eliminate the warning about “to high angle-of-attack”.
You’re probably aware, but 8 m/s is low for an extreme event. Also, extreme events are often modeled with an idling rotor (GenDOF=True; GenTiStr=True; TimGenOn>TMax) instead of a parked rotor (GenDOF=False).
Dear Dr. Jonkman,
Thank you for your suggestion, now the model run without warning.
For the wind, yes, I am aware that 8 m/s is not related to Extreme Event. For now, I am trying to set up a “parked” model correctly and later will try to explore more about the environmental forcing and loading on the structure in a “parked” condition. Again, thank you very much for the suggestions.