Bug in the Wing Editor

  • David

    David - 2012-03-05


    there is a problem with the airfoils during the design of a wing. Whenever the
    twist of a profile section is changed the profiles chord length gets smaller.
    A twist of 90° seems to result in a chord length of zero...
    See the screenshots:

    this is with zero twist

    now 70° twist...

    Is this only a problem with the GL representation or does this even affect the
    size of panels etc?
    Where would I have to look for a fix?



  • André

    André - 2012-03-05

    Unfortunately, this is a bug, not only a display thing.
    I've already committed the fix to the repository.
    The impact is minor for usual small twist angles, which is why I failed to
    spot the problem.

    Thanks for reporting the bug.


  • David

    David - 2012-03-06

    Wow - that was fast! Unfortunately now the thickness seems to be affected by
    twist, it goes to zero at 90° twist.
    Also the panels are not consistent with the surfaces anymore. See the

    I adapt your panel method for unsteady wind turbine simulations. Wind turbine
    geometries require quite high twist angles near the blade root - thats why I
    spot these bugs...



  • André

    André - 2012-03-06

    Right, I need to take a little time to figure out how to reconstruct and join
    properly 3D surfaces including dihedral and twist.


  • Bubo Maximus

    Bubo Maximus - 2012-07-13

    Firstly thank you very much - I'm new in XFLR5, but it's a great tool...

    I tried to model the effect by the Horten brothers skew hinge, but I can't
    define an asymmetric wing - any change in the left other in the right side
    makes a symmetric change in the other side despite of the disabled "Symmetric"


  • Robin

    Robin - 2012-07-13

    The only asymmetric wing feature which can be implemented is the wing profile.
    Unfortunately geometry is alway symmetric, i hope this could be a new feature
    in the near future.

    David, can you tell me something more about your adaption to the program?
    I have also made an adaption to the program such that it uses a viscous
    correction in 3D calculations and therefore is now able to better predict the
    nonlinear area and stall point.

  • David

    David - 2012-07-31

    I have changed the Boundary Element Method to be able to compute unsteady
    simulations over a number of timesteps. The geometry of the rotor is rotated
    with time. I also update the position of the wake panels being influenced by
    all other elements (wake rollup). However I didnt finish a releasable product
    yet. It is planned as an extension to qblade (also here on sourceforge), based
    on XFLR. I would be really interested in your viscous correction - I might use
    a similar approach...

    This is the result of one simulation with the panel code...

  • Robin

    Robin - 2012-07-31

    It sounds really interesting what you are doing. At the moment i am finishing
    my thesis report, but still have to do some small investigations on the
    viscous correction implementation in XFLR5. I implemented two different
    theories (for which each theory i used several variants) and validated the
    results with straight wings, swept wings and a nonplanar wing. From validation
    a best method can be chosen.
    At the moment, unfortunately, i cannot give you my code, since it is
    confidential within the TU Delft. However, i will discuss with my supervisor
    whether e.g. collaboration is possible.
    The used theories can be found in these papers:
    author = {B. J. C. Horsten and L. L. M. Veldhuis},
    title = {A New Hybrid Method to Correct for Wind Tunnel Wall- and Support
    Interference On-line},
    year = {2009},
    journal = {World Academy of Science, Engineering and Technology},
    volume = {vol. 58}
    author = {M. Gaunaa and P. F. P. Carqueija and P. E. R$\'e$thor$\'e$ and N. N.
    title = {A computationally efficient method for determining the aerodynamic
    performance of kites for wind energy applications},
    year = {2010},
    journal = {Ris$\o$ National Laboratory, Denmark}


Get latest updates about Open Source Projects, Conferences and News.

Sign up for the SourceForge newsletter:

No, thanks