Point reference of center gravity

edited April 2016 in DualSPHysics_v4.0
Hi all, firstly I like say thanks for this amazing software. I just want ask about coordinate center of gravity in floating case is measure from global coordinate or we measure from floating bodies(local from floating object) If we want for specified for coordinate of center gravity.


  • Center gravity is computed starting from coordinates of the floating object which correspond to the global coordinates of the system. But you can check it first with a simple case as the ones including as testcases.

  • Thanks alex for your explanation and that's true. I try run my previous case in new version but it doesn't work

    GenCase4 v4.0.025 (15-04-2016)
    OmpThreads: 8


    *** Exception (JXml::LoadFile)
    Text: Cannot load the xml file: Failed to open file (row:0 col:0)
    File: CaseFloatingnew_Def.xml

    my batch file

    %boundaryvtk% -loadvtk %dirout%/%name%__Actual.vtk -filexml %dirout%/%name%.xml -motiondata %dirout% -savevtkdata %dirout%/MotionFloating -onlymk:14

    %boundaryvtk% -loadvtk %dirout%/%name%__Actual.vtk -filexml %dirout%/%name%.xml -motiondata %dirout% -savevtkdata %dirout%/MotionPiston -onlymk:21

    my xml
  • You should mimic the new structure of the XML, especially for floatings, and take a look to the XML_GUIDE.pdf
  • edited April 2016
    [floating mkbound="3" relativeweight="1.3"]

    * I have used symbols [ and ] instead of higher and lowe than *
  • Alex still doesn't work,is there mistake on my xml? I have been trying make several modification but still don't work.


  • Can not open ship.vtk file... I need the file
    However GenCase starts to run without problems using your XML
    Are you using the latest version of GenCase?
  • edited April 2016
    I use latest version of GenCase, yes it's work maybe I will try again to solve my case but I give my vtk. file to compare and more understand about my case and Thanks for your help Alex


    Best Regards,

  • Note, that as we mentioned in the guide, floating bodies have been validated when they are created with solid or full options.

    For the "ship" you are using setdrawmode mode="full", however the option drawfilevtk only creates particles at the faces (as you can see checking Case_bound.vtk). Therefore you need to fill the ship with boundary particles. For that you can use the option fillbox.

    With the floating objects full insided of particles you will get properly results and you can check if the mass and density are the real ones.

  • Thank you for your suggestion Alex

    Best Regards
  • Alex for your suggestion is mean I must do like case pumps or just use fillbox to create particle from inside "ship" ?thanks in advance

    Best Regards,
  • You have to fill the interior of the "ship" with boundary particles.
    Then you can also use "fillbox" option to fill the tank with fluid particles.

  • Thanks Alex for your suggestion, I run simulation and is stop with this warning

    Exception (JSphGpuSingle::Interaction_Forces)
    Text: Failed while executing kernels of interaction. (CUDA error: unknown error).
    is there any problem with my CUDA? is I need update my CUDA?

    Best Regards,

  • I have already replied you from dualsphysics@gmail.com:

    "everything looks fine with your GPU
    reboot the machine and execute again the case
    do not execute more than one case on a GPU at the same time"
  • Hi Alex,

    I have also met the same problem while adjusting the dp to lower value.

    Exception (JSphGpuSingle::Interaction_Forces)
    Text: Failed while executing kernels of interaction. (CUDA error: unknown error).

    Is it caused by reaching the limit of my gpu?
Sign In or Register to comment.

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!