Dear OpenFAST team,
I hope this mail finds you well. I am recently calculating the Campbell diagram of the IEA 22MW project (using the file from GitHub - IEAWindSystems/IEA-22-280-RWT: Repository for the IEA 22-MW offshore reference wind turbine developed by the IEA Wind Task 55 REFWIND · GitHub).
And everything goes well when I set the UA_mod = 0, which means Quasi-steady. However, when I switched this value to 4, which means B-L HGM 4-states, the Campbell diagram results came like this below.
I think it might because of noisy, as some excepted pattern could still be found. So I then checked the excel file after linearization, and found that the eigenvalue should be plurals, but many positive integers have appeared there.
Do I just need to delete those integers manually? And could you please tell me why this happened?
Thank you so much!
All best wishes,
Xingyu
UPDATED: I just delete those integers manually, but it doesn’t work because of the wrong column number
Dear @Xingyu.Liu,
This happens because with UAMod = 4, many aerodynamic states are included in the linearized model. You can filter these out, leaving only the structural states, by selecting “Filter Non-structural Modes” in the Campbell Diagram section of ACDC.
Best regards,
Dear @Jason.Jonkman ,
Thank you for your explanation. I tried selecting “Filter Non-structural Modes,” but possibly due to too many modes, the ACDC crashed. Therefore, I would like to ask if there is any manual method to do this task? Or rather, how can we identify Non-structural Modes? I originally wanted to try manually deleting some data before feeding it to ACDC to generate the Campbell diagram, but I encountered the problem described earlier — an error because of the wrong number of columns.
Thank you!
All best wishes,
Xingyu
Dear @Xingyu.Liu,
I’m not sure what would cause ACDC to crash. Perhaps try starting the process through ACDC from the beginning?
What the “Filter Non-structural Modes” switch does is to use the location of the maximum value within an eigenvector and remove modes that are not structurally dominated (ElastoDyn, BeamDyn, SubDyn). So, you should be removing an equal number of rows (states) and columns (modes) to keep the system square.
Best regards,
Dear @Jason.Jonkman ,
Thank you for your suggestion. It might be that my previous description wasn’t clear enough—I did initiate the process through ACDC from the very beginning, but the crash still occurred. I also checked the case where UA_mod = 0 and found that there are some positive integers in the modes.csv file as well—I’m curious about what these positive integers, listed after the eigenvalues, mean.
Then I looked at the MaxMod column and noticed that some entries are BD, but most are AD. Perhaps you mean that I should first remove the rows and columns corresponding to these AD parts—hopefully this will work.
All best wishes,
Xingyu
UPDATE: Before it crashed, I captured some error messages, which may help. This message appeared for only a moment, then ACDC crashed and the window also closed.
Dear @Xingyu.Liu,
Yes, that is what I mean about removing rows and columns.
I’ll forward your note about the ACDC crash to the developers.
Best regards,
1 Like
Dear @Jason.Jonkman ,
I think I’m almost certain now that the crash of ACDC is caused by too many rows and columns — I tried using ElastoDyn to calculate IEA15MW with the same setting UA_mod=4 — in this case, “Filter Non-structural Modes” works fine.
All best wishes,
Xingyu
UPDATE: I tried to delete all rows with MaxMod=AD in modes.csv, but it still doesn’t seem to help — ACDC still crashes. I might need to write my own code to complete the connectModes step.
@Xingyu.Liu
I can’t think of a reason that this would cause ACDC to crash. If you haven’t already solved it, could you open an issue at Issues · OpenFAST/acdc · GitHub and post the project’s .json file and list the version of OpenFAST being used. Also, if you could include the full error message from the crash that would be helpful, the source of the error is at the top of the message.
Thanks,
Derek Slaughter
Dear @Derek.Slaughter ,
I have loaded the project’s.json file as required, however, I failed to include the full error message – it disappeared in an instant, so I could only try to take a screenshot.
All best wishes,
Xingyu
