Interpretation of State Vector during linearization with AeroDyn

I am trying to interpret the results of a Modal analysis around an Operational Point. To do that i am using the linearization capabilities of OpenFast. I am using the standard NREL 5-MW reference turbine and have enabled the following modules:

  1. ElastoDyn
  2. InflowWind
  3. AeroDyn
  4. ServoDyn

The results after running the python script ex2a_MultiLinFiles_OneOP.py are as follows:

1. Why are aerodynamic wind velocities part of my state vector?

I see that the state vector includes quantities like AeroDyn wind velocities, which I didn’t expect. I had assumed the aerodynamic influence would be captured in the system matrices (e.g., the A matrix), while the state vector would contain only the structural degrees of freedom. Is there a way to exclude these AeroDyn states from the state vector or configure the linearization to focus only on the structural states?

2**.** How should I interpret mode shapes that include wind velocities and structural displacements together?

This is the Main_lin.1.lin output file and as we can see there are 258 continuous states. This many eigenmodes and frequencies make the results less clear and hard to interpret.

These are my input files in case they are needed:

NREL

Thank you very much in advance,

Best regards,

Vasileios Xenodochidis

Dear @Vasilis.Xenodochidis,

Here are my responses:

  1. Aerodynamic states are included in your linearized model because you’ve dynamic wake (DBEMT_Mod = 3) in AeroDyn. While inclusion of aerodynamic states (from dynamic wake or unsteady airfoil aerodynamics) makes interpretation of the eigensolution more difficult, it is a very valuable tool to include the effect of aerodynamics on the stiffness and damping of the eigensolution. See our TORQUE 2022 paper for more information: https://docs.nrel.gov/docs/fy22osti/82343.pdf. To eliminate these aerodynamic states from the linearization, set DBEMT_Mod = -1 to enable frozen wake, which mimics the dynamic inflow effect without directly considering aerodynamic states).
  2. To interpret the eigensolution, I would focus on the the structural state contributions to the modes. Our Automated Campbell Diagram Code (ACDC) is the most convenient way to set-up, analyze, and interpret eigensolutions: GitHub - OpenFAST/acdc: ACDC: Automated Campbell Diagram Code .

Best regards

1 Like