Get k1, k2,k3 from GEBT output result

  1. Daniel Kharlamov

    Dear all,


    I would like to ask how to get the twist and curvatures, k1, k2,k3 from gebt output in b frame of reference. I am doing currently aeroelasticity and need to set initial k1,k2,3 for the next iteration.

    However, I am a bit confused how to compute them, as (F1,F2,F3,M1,M2,M3) in the output file of GEBT are in B-frame of reference. So, if I multiply the inverted stiffness matrix with (F1,F2,F3,M1,M2,M3) in B-frame of reference, I get stretches and strains also in B-frame of reference. But, as I've understand the initial twist and curvatures are in b-frame of reference.

    I am coming normally from the aerodynamic corner and need to fix the issue in my code. Please, help me here to find a way how to obtain k1,k2 and k3 in b-frame of reference.

    Thanks for help!


    Kind regards,


  2. Qi Wang

    Hi Daniel,

    You may need to use the output rotation parameters \theta_i to construct a rotation matrix; then use this matrix rotate the pseudovector [k1,k2,k3] back to the initial frame b.

    The rotation parameters, \theta_i, in GEBT are Miener-Milenkovic parameters. You can find the details in the paper attached, see equation 12 and 14.


  3. Daniel Kharlamov

    Dear Qi Wang,


    thank you for your answer.


    I am asking because I am using this tool in an aeroelasticity framework. When I put gravitational loads only onto the beam, I am getting the correct displacement with respect to experimental data. However, when I put gravitational loads and aero loads and moments, my displacement is slightly off and the discrepancy increases with increasing flow velocity. Till now I was running an initially non curved and twisted configuration for any iteration and that's why I am wondering if I need to pre-set the initial curvature and twist for a particular iteration by using values from the previous iteration. Do you have some experience here?


    I've followed your advice and transformed the [k1,k2,k3] vector resulting from (S^-1)*[M1_B,M2_B,M3_B] back to b frame by multiplying it with CbB rotation matrix. However, now I am getting a non-physical result and my aeroelastic solution does not converge anymore.


    I am using gebt-26 (name of the folder when I unpack the tar container), which I've downloaded from this homepage. However, your paper is from 2017 and this homepage states that most recent version of the code had been published in 2014. Is this the most recent version of the code published on this website?


    Kind regards,