Dear all,
I am using OpenFAST-v3.0.0, and trying to reproduce the free decay in “Definition of the UMaine
VolturnUS-S Reference Platform Developed for the IEA Wind 15-Megawatt Offshore Reference Wind Turbine”. The case is downloaded from GitHub. When I give an initial displacement, time series show that the magnitude and frequency of the displacement are fine. The only problem is that the pitch is not centered at zero. Then I removed the initial displacement, i.e., zero initial displacement, without wind, wave and current. The plot below shows the pitch center is at around -1.3 deg, surge is off as well. What could cause this center shifting? Any suggestions are greatly appreciated!
Best regards,
Xiaoqin
Dear Xiaoqin,
I’m not too familiar with this specific model, but I would guess the nonzero offset in surge and pitch in the absence of wind and wave loading is caused by the overhanging weight of the rotor-nacelle assembly, offset from the tower centerline.
Best regards,
Dear Jason,
I compared the mass of RNA and related parameters, they are indeed different from the paper. I suppose there is offset of CG of RNA from tower axis for the setting in the GitHub package. Thank you for pointing in the right direction.
Regards,
Xiaoqin
Dear Jason
I want to proceed verification for IEA 15MW wind turbine model using VolturnUS-S Platform.(downloaded from Github - IEAWindTASK37/IEA-15-240-RWT)
I saw Free-decay & Steady state result from IEA WindTASK37 Technical Report but it showed only result graph.
I ran free-decay & Steady-state test but my result is not identical to report one.
Maybe one of the reasons could be OpenFAST & 15MW R.W.T with VolturnUS-S platform version has been developed since Report was published in 2020.
I know i can’t ask for OpenFAST initial input data or condition directly according to the Rules here.
At least, Would you please give me a tip for verification?
Best regards,
Sangwon
Dear @Sangwon.Lee,
I was not the one to develop this OpenFAST model or run the simulations documented in the report. But can you clarify how different your results are relative to the report and what you expect are the differences in input file settings?
Best regards,
1 Like
Dear Jason
Here is my free-decay test result and OpenFAST(v3.4.1) input
Comparing my results with Report,
Surge : Only trend but not the value
Sway : Quite similar but not the same
Heave : Almost identical
Pitch : Totally different
Roll : Almost identical
Yaw : Quite different
Left is mine, Right is report
Here’s my set point in OpenFAST
Main module
Hydrodynamic
Elastodyn
ServoDyn
AeroDyn15
I hope these could help you.
Best regards,
Sangwon
Dear @Sangwon.Lee
I agree with your summary. In surge and pitch, I see large differences in frequency and damping, as well as a mean offsets after transients die out. In yaw, I see more damping in your model.
I haven’t reviewed your model in detail, but I do see that you have wind and aerodynamics with rotor rotation and control enabled. I would guess what you are comparing too from the report does not have wind excitation. This would likely explain the differences in mean offset, frequency, and damping levels. So, I would recommend disabling wind and aerodynamics and parking the rotor. Do your results agree better then?
Best regards,
1 Like
Dear @Jason.Jonkman
I followed your advice and there is still some differeneces in pitch
- Turn off InflowWind & AeroDyn15 Module
- Set initial Rotor speed(RotSpeed) 0 in ElastoDyn Module
Comparison with my previous results
- It looks much reasonable than before
Comparsion with IEA Report
- Most of results are closer to Report than before
- But still minor difference in Pitch(marked by red circle)
- I couldn’t find what made this difference
Left is new results, right is Report
I have another question.
I set no wind, no aerydynamic condition, no wave, no current but there is small pitch maneuvering during simulation.
It supposed to be fixed, isn’t it?
One of my simulation processes
Best regards,
Sangwon
Dear @Sangwon.Lee,
I’m glad your updated results match the report much better!
But the nonzero rotation of the rotor may be influencing your platform motions a bit.
It sounds like you’ve kept the operational pitch and torque controller enabled when eliminating wind. When eliminating wind, you should disable the controller either by disabling the generator degree of freedom in ElastoDyn (GenDOF = False) or by disabling the controller within ServoDyn (PCMode = 0; VSContrl = 0; GenTiStr = True with TimGenOn > TMax).
Best regards,
1 Like
Dear @Jason.Jonkman
I followed your advice to set exact parking condition
Also, I confirmed no more pitch maneuvering
But except roll, I found there isn’t much difference between results.
According to IEA Report, Pitch and Roll should be almost same.
But Pitch and Roll are still quite different.
Is it because of updated Model version or OpenFAST version?
Best regards,
Sangwon
Dear @Sangwon.Lee,
In your comparison between pitch and roll free decay, the biggest difference I see appears to be the mean value, which is nonzero in pitch and zero in roll (the natural frequencies and damping appear nearly identical). This offset in pitch is likely the result of the overhanging weight of the rotor-nacelle assembly (RNA) being offset from the centerline of the tower. I’m not sure why the report you are referencing doesn’t show this same offset (again, I’m not the one who performed that analysis), but perhaps the model was simplified (removing the offset) before the free-decay simulations were run.
Best regards,
1 Like
Dear @Jason.Jonkman
I thought I ran wrong simulation or used wrong model.
Now I understood why Pitch Displacement is quite different.
I compared my final result with IEA report much precisely.
Except Pitch and Roll, most of graphs are very similar.
Lastly, I have a one more question.
- Is Roll Displacement difference likely also because of RNA Overhanging weight?
Blue Line : IEA Report / Red Line : My Simulation
Best regards,
Sangwon
Dear @Sangwon.Lee,
The RNA overhanging weight should result in a mean offset in pitch, not roll. Indeed, in your roll results, I don’t see any mean offset.
I do now see, though, a small difference in pitch and roll natural frequency between your results and the IEA Wind report. I’m not sure what is causing this difference.
Best regards,
1 Like
Dear @Jason.Jonkman
I will keep study what might cause this difference.
I sincerely appreciate for your detailed and quick reply so far.
It really helped my reasearch progress.
Thanks again and have a nice day!
Best regards,
Sangwon
Dear @Jason.Jonkman @Sangwon.Lee
I really appreciated the discussion you two had about the results of the Free decay Test.
I am also attempting to validate the 15MW OpenFAST model through the Free decay test and Static equilibrium test.
Therefore, I have some questions regarding the discussion mentioned above and my validation process.
1. In the discussion mentioned above, it was stated that there is a slight difference between the Pitch Free decay test results and the Report results due to “the model was simplified (removing the offset)”. Is there a method in OpenFAST model to perform validation by excluding the offset?
2. I’m also experiencing results similar to the Equilibrium test shown in the figure below. In this case, there appears to be significant pitch occurrence and relatively low damping applied.
Could these results be attributed to the inherent characteristics of the 15MW IEA model itself rather than indicating a flaw in the 15MW OpenFAST model or the interpretation method?
Your discussion has been incredibly helpful. If you could provide answers to the additional questions I’ve asked, it would greatly aid my research efforts. I would greatly appreciate it.
Best regards,
Dear @Jisu.Lim,
Regarding (1), I was suggesting that perhaps the RNA model was simplified by removing the mass offsets. This can be achieved in OpenFAST by changing the geometry in ElastoDyn such that the mass is not offset from the tower centerline (such as by setting PreCone, OverHang, NacCMxn equal to zero).
Regarding (2), are you using the OpenFAST model of the IEA 15-MW reference wind turbine model directly from the GitHub repository (IEA-15-240-RWT/OpenFAST/IEA-15-240-RWT-UMaineSemi at master · IEAWindTask37/IEA-15-240-RWT · GitHub) or did you change some parameters? I agree that the results appear to show little pitch damping.
Best regards,
Dear @Jason.Jonkman
Thank you infinitely for your prompt and clear response.
As you suggested, I have eliminated the mass offset of the RNA model and conducted a free decay test through OpenFAST.
Comparing the results with the report, they appear to be almost identical. However, there seems to be a slight difference in the period as the time step progresses, indicating a need for further investigation.
If you have any insights into why the period might vary, please let me know.
Once again, I extend my gratitude to Jason Jonkman for his continuous efforts in the field of wind energy.
Best regard,
Dear @Jisu.Lim,
Well you have changed the generalized mass matrix, so, I’m not surprised the natural period is affected. But again, I was not the one to generate the results you referencing, so, I’m not sure what the differences are relative to what you are doing.
Best regards,
