GEBT for wind turbine simulation

  1. Peter Greaves


    I'd like to use GEBT to model a wind turbine. To do this, I will need to do a few things which I am not sure if GEBT does:
    • Create spherical joints between bodies (eg. rotor end of the main shaft and the nacelle, and generator end of main shaft and the nacelle to create a revolute joint). To do this, two nodes are created that are coincident in space and their translations are kept the same by using Lagrange multipliers. - It looks like it doesn't do this, but it would be easy to implement I think.
    • Create rigid joints between bodies, eg. between the main shaft and the rotor blades. - It think it does do this.
    • Use the information about position and velocity of the nodes in an external subroutine which calculates aerodynamic loading for the next time step, then calculate the next time step using Analysis.dll - This one I'm really not sure about, can you run one step at a time?
    • Model structural damping as a linear combination of the mass and stiffness matrices (Rayleigh damping) - Definitely doesn't do this as far as I can tell, quite easy to implement.
    • Run more or less in real time - As far as I can tell, the time integration method does not introduce numerical damping so you need to use a very small time step. Would it be possible to implement the time stepping part of the code with a Newmark-beta type algorithm with Hilber-Hughes Taylor damping as is used in commercial FE codes? On a 'per time step' basis it seems the code is fast, so I think this objective could be satisfied.
    I've tried to figure some of this stuff out from the source code, but I'm not really a Fortran guy! Is there any plan to add this kind of functionality?


  2. Qi Wang

    Hi Pete,

    Here are my responses to your comments and questions:

    • There is no joint element available at this moment in GEBT. GEBT is a beam solver; while a "multibody dynamic code" will connect the components, including beam, plate, shell, and lumped mass, by joints.
    • I think GEBT can be used as a "structural module" in solving fluid-structure interacting problems.
    • Currently, there is no physical damping effect in GEBT. We are working on it and hopefully a viscous damping option will be available soon.
    • I agree with you that numerical damping is critical for realistic problems where huge high-frequency modes can be found and needed to be damped out to increase the numerical stability. The generalized-alpha or HHT are time integrators for second-order governing equations while GEBT are first-order (in time domain). We will look into the details to see if we can develop a first-order time integrator featuring controllable numerical damping.

    Thanks for your comments!


  3. Li Zed Shawl

    Hi, I got a question about GEBT, hope some one can help:

    In the GEBT input data files, if the "analysis_flag" is 1, it means to carry out a steady state analysis. This "steady state analysis" bothers me. What kind of problems it can solve? And what kind of loads and boundary conditions can be classified as "steady state analysis"? I checked the samples and found they were all static problems even with the "analysis_flag" being 1.