We are coupling our inhouse structural solver with Aerodyn.
We want to include the moment due to the offset between FE nodes and AC.
After going thorough several discussions in this forum and the documentation, I have understood that - BlCrvAC and BlSwpAC are the offsets from pitch axis. I want to obtain these coordinates and their motions in terms of global reference frame. Because we already have our FE node coordinates and motions in terms of global reference frame.
As I understand it involves several successive transformations.
Pitched position to Blade local - rotation using pitch angle
Blade local to pitch axis - rotation using BlCrvAng
Pitch axis to Rotor plane - rotation using Precone
Rotor plane to hub axis - rotation using Azimuth and translation using BldSpan and
Hub to Nacelle - rotation using shftTilt and translation using OverHang and Twr2Shft
Nacelle to Global - translation using tower height
Please confirm if these transformation are correct provided my understanding is valid.
I referred to this document
I generally agree with what you are saying; just a few comments:
I would say that your transformations are backwards and that the successive transformations are from the ground (global) to the blade, not the other way around. In terms of rotations, these are just transposes of each other. But the equations you copied from the ElastoDyn Theory Manual are backwards from what you describe.
From the hub to the blade node, the sequence should be pitch angle, followed by BlCrvAng, not the other way around.
Your transformations miss the BlCrvAC and BlSwpAC offsets in step 1, as well as the HubRad in step 4.
Your transformations miss any effect of structural deflection from elasticity.
Regarding the symbols in the ElastoDyn theory manual, q_R, q_P, and q_Y are the roll, pitch, and yaw displacements (rigid-body rotations) of the platform. These and other ElastoDyn structural DOFs are documented in FASTDOFs.xls.
Regardless, I would think your structural FE solver would compute all of these transformations internally, and so, you shouldn’t need to know what ElastoDyn is doing exactly anyway if you goal is to replace ElastoDyn with your own FE solver.
What is the datastructure that contains the deflection of AC in blade coordinates ? Does AeroDyn track it ?
Once this deflection is known - for every time step I can add that to my AC coordinates and then carry on the transformation.
Yes as I understand, I don’t need TransMat. For all the above rotation transformations - Direction cosine matrix (DCM) would be required for each successive ones (of course in the backward order).
For example - for a rotation about a common z axis, the DCM would look like
Structural deflections of the blade in OpenFAST are tracked by the structural solver (ElastoDyn, BeamDyn). The motions (positions/orientations, velocities) are passed to the aerodynamic solver (AeroDyn). AeroDyn expects the motions to be received at the aerodynamic analysis nodes, which on the blade are at the aerodynamic centers. Within OpenFAST, spatial mesh-ot-mesh mapping is used to transfer the data from the structural nodes to the aerodynamic nodes.
I agree that that you can track your own DCMs from your own structural solver and pass the successive DCM to AeroDyn.
