Dear all,
I am recently working on a project using Bladed, and I want to compare its results with OpenFAST, so I tried to write the .fst and .dat files based on the data from Bladed. However, there are still some parameters I could not match. If you have any ideas, please give me some help. Thank you!
- “PitchAxis” column from _Blade.dat: I could only find Pitching Moment Centre in Blade, do they refer to the same parameter?
- “BLADE MODE SHAPES” also from _Blade.dat: I am not sure why we need to input these values – how did we know before we do modal analysis?
- “DAMPING COEFFICIENT“ from BeamDyn_Blade.dat: Are these some empirical parameters or calculation results?
Maybe I did not understand the whole process clearly, do we need to do any pre-processing before we run OpenFAST? Thank you again:)
All best wishes,
Xingyu
Dear @Xingyu.Liu,
Here are my responses to your specific points:
- The
PitchAxis column in the ElastoDyn blade input file is no longer used by ElastoDyn and will be removed in the next release.
- ElastoDyn requires the flapwise and edgewise mode shapes of the blade in addition to the distributed mass and stiffness. The former can be derived from the latter (plus other inputs such as the rotor speed and hub radius) in a preprocessing step, e.g., using BModes or another tool.
- BeamDyn uses a stiffness-proportional structural damping as defined here: 4.6.3. Input Files — OpenFAST v4.1.2 documentation. Typically the damping coefficients are specified to match the proper modal damping of the first modes, e.g., as discussed here: Need suggestion regarding the calculation of BeamDyn damping · Issue #2938 · OpenFAST/openfast · GitHub.
Best regards,
1 Like
Dear @Jason.Jonkman ,
Thank you for your reply. With your help, I think now I could run some self-defined case. 
But I am still a little bit confused while dealing with the modal analysis. Now I understand how OpenFAST get natural frequencies using Bmodes, but how does OpenFAST calculate damping ratio? I read the BeamDyn User’s Guide and Theory Manual but it talked about Damping Forces and Linearization. It seems that BeamDyn assert that user already knows Damping Coefficient, which could be calculated if we know damping ratio. Do we calculate the Damping ratio using BEM theory? Just like Petersen’s work in backend.orbit.dtu.dk/ws/portalfiles/portal/7751222/ris_r_1045.pdf
All best wishes,
Xingyu
Dear @Xingyu.Liu,
I’m not sure I fully understand your question. If you are referring to the structural damping coefficients of the blades, these are inputs to the blade structural modules of OpenFAST (ElastoDyn or BeamDyn), so, are assumed to be known or estimated by the user. If you are referring to the damping ratio of a given full-system mode, including structural and aerodynamic effects, these can be calculated through an OpenFAST linearization analysis based on the damping terms inherent in the model.
Best regards,
Dear @Jason.Jonkman ,
I am sorry for my poor description. Actually, what I mean is that I want to calculate the damping ratio of a given full-system mode, including structural and aerodynamic effects. According to the literature, once the mode shape and frequency are known, we can compute it using
based on BEM theory. I would like to know if OpenFAST follows a similar process. It seems that BeamDyn does not focus on this step but instead directly discusses the damping force, and I am somewhat confused about the relationship between the damping force and the damping ratio mentioned above.
Thank you again for your reply.
All best wishes,
Xingyu
Dear @Xingyu.Liu,
OpenFAST directly captures the aero-structural coupling (aero-elastics) in its time-domain solve. When OpenFAST is linearized, these aero-elastic effects are captured within the linearized state-space matrices output from OpenFAST. From the linear model, you can calculate the full-system modes and their natural frequencies and damping ratios through subsequent eigenanalysis of the state matrix “A” (that relates the state derivatives to the states).
Best regards,
1 Like
Dear @Jason.Jonkman ,
Thank you for your quick reply. I think now I understand the whole process. I may need to refer to the corresponding literature/document to learn the details of how OpenFAST solves aero-structural coupling (aero-elastics) in its time-domain solve.
All best wishes,
Xingyu
Dear @Jason.Jonkman ,
I hope this mail finds you well. As we talked above, recently I am trying to use OpenFAST run my own case (the major difference between NREL 5MW case is that I used a new blade), and it works well with ElastoDyn+InflowWind+AeroDyn. But when I switched it to BeamDyn, the OpenFAST aborted with some warning. I tried to run BeamDyn alone, but it still didn’t work. Follows are the error information:
Running ElastoDyn.
Nodal outputs section of ElastoDyn input file not found or improperly formatted.
Running BeamDyn.
Running BeamDyn.
Running BeamDyn.
Running InflowWind.
Running AeroDyn.
AD15 Nodal Outputs: Nodal output section of AeroDyn input file not found or improperly formatted.
Skipping nodal outputs.
AeroDyn: projMod: 1
Projection: legacy (NoSweepPitchTwist), BEM: legacy (2D)
FAST_InitializeAll:FAST_Init:ValidateInputData:LinOutJac can be used only when
LinInputs=LinOutputs=2.
FAST_InitializeAll:BD_Init:BD_QuasiStatic:BeamDyn could not find a quasi-static solution to
initialize with. Proceeding with no initial solve.
FAST_InitializeAll:BD_Init:BD_QuasiStatic:BeamDyn could not find a quasi-static solution to
initialize with. Proceeding with no initial solve.
FAST_InitializeAll:BD_Init:BD_QuasiStatic:BeamDyn could not find a quasi-static solution to
initialize with. Proceeding with no initial solve.
Time: 0 of 1.0015 seconds.
FAST_Solution:FAST_UpdateStates:FAST_AdvanceStates:B1::BD_GA2:BD_DynamicSolutionGA2:Solution does
not converge after the maximum number of iterations
FAST_AdvanceStates:B2::BD_GA2:BD_DynamicSolutionGA2:Solution does not converge after the maximum
number of iterations
FAST_AdvanceStates:B3::BD_GA2:BD_DynamicSolutionGA2:Solution does not converge after the maximum
number of iterations
OpenFAST encountered an error at simulation time 1.50000E-02 of 1.0015 seconds.
Simulation error level: FATAL ERROR
Aborting OpenFAST.
Dear @Xingyu.Liu,
Well, the warnings you are receiving imply that your BeamDyn model is not set up properly to obtain a reasonable result with the solver settings you are using. Do your key points form a smooth reference axis? Are the elements of the mass and stiffness matrices you are using based on physically correct values?
Best regards,
Dear @Jason.Jonkman ,
After a series of checks, I found two main issues: First, the mu value I set was too large, approximately around 0.01, while referencing the NREL 5MW case setting it to 0.001 resolved the problem; second, the last node in the AeroDyn_blade.dat file needed to be slightly smaller than the actual blade length. After addressing these issues, OpenFAST can now run smoothly, and I am preparing to study the impact of mu on the final results.
All best wishes,
Xingyu
1 Like