Generating Campbell with matlab-toolbox-main has an error, and some issues.

[1335064763fansir]

I'm currently using matlab-toolbox-main to generate Campbell, and I'd like to ask if the openfast version has any impact on using this toolkit, why do I get an error when I use the latest version of openfast to run the examples in it, and the examples can't be run directly? If I want to learn more about this toolkit, I would appreciate some advice!

[jjonkman]

Dear @1335064763fansir,

The MATLAB Campbell diagram generation scripts should work for different versions of OpenFAST. But in your case, it looks like your input file is not compatible with the version of OpenFAST you are using. Input file changes with each release of OpenFAST are documented here: https://openfast.readthedocs.io/en/main/source/user/api_change.html. And you can always enable the Echo option to debug issues associated with input file format.

FYI: The new Automated Campbell Diagram Code (ACDC) is a vast improvement over the old MATLAB scripts: https://github.com/OpenFAST/acdc.

Best regards,

[1335064763fansir]

Dear Sir,
Also, my 7th and 9th order frequencies have the same mode shape, is it caused by asymmetric progressive (top) and asymmetric regressive (bottom) as you said?

[1335064763fansir]

I also found that as shown in this graph, it is logical that the lower order modes should be for lower Damping Ratio, but it shows the opposite pattern, is this a problem due to the ACDC itself is not mature enough to process the data incorrectly?

[jjonkman]

Dear @1335064763fansir,

I was able to reproduce your issue and learned that is an artifact of how you set up the analysis. You're operating point for blade pitch is zero for all wind speeds, which is not correct when aerodynamics are disabled. And with the generator degree of freedom disabled (GenDOF = FALSE), the internal controller use to trim pitch above rated is also disabled. Without blade pitching, the aerodynamic loads get very large at higher wind speeds, resulting in unrealistically large deflection. The aerodynamic damping also drops as wind speed increases (as you are showing), until the model is aero-elastically unstable and unable to find a steady-state solution above 20 m/s. I would suggest enabling GenDOF and setting the pitch schedule correctly for above-rated wind speeds.

For a given rotor mode (e.g., first out-of-plane bending), I would expect the progressive mode (with high frequency) to have a lower damping than the regressive mode (with lower frequency); ACDC is picking this up correctly.

Best regards,

[1335064763fansir]

Thank you for your answer!
This does not mean that in this case, the phenomenon of the blade is
aero-elastically unstable。 In other words, the wind speed above 20m/s, the blades are unstable and will be destroyed under the operation I set?

This result is the result of my running with the MATLAB toolbox, and it is also run with the same blade pitch, which can also explain my doubts above.

[jjonkman]

Dear @1335064763fansir,

When you run with the MATLAB toolbox, are you computing a steady-state solution before linearizing and does that solution remain stable? Do you have the same DOFs enabled/disabled as in ACDC?

Your results look a bit odd because damping is dropping with increasing wind speed until about 20 m/s, in which case the response changes; I'm not sure what is leading to this change, but it seems like you've changed something in the model for the higher wind speeds.

Best regards,