Error boundary particles excluded.

Hello,

For one of my school project I have to simulate a dam break case with 9 solids floating in the initial water column and I decided to do it with DualSPHysics. I saw the example CaseSolids already made in the DEM folder and I decided to start my project using the existing xml file.

I changed the geometry, some of the material properties and I moved the existing boxes at the top of the water column.

When I run the program, an error quickly appears saying that "some boundary particle exceeded the +Y limit" . I already changed the initial position of the solid boxes (in the first error message, the particle exceeded the -X +X -Y and +Y limits) but I don't how to fix this error.

I attached my code (the CaseSolids.xml file modified and the Floatingsmaterials.xml also modified) and screenshots of the error message and the vtk error files visualized in Paraview (I voluntarily increase the radius to spot the errors particles).

Thank you very much for any help.

Clement


Comments

  • Hi,

    No bug here I think. As the file told you, some of the floatings particles are reaching the edge of the simulation.

    On your screenshots, it seems that there is no wall on y+ ?

    Regards

  • Hi,

    Thanks for your help. Indeed I added the wall in y+.

    However the error persist. It seems like there is a problem with the floating particles and the walls, how can I fix the boundary condition to keep floating particles in the tank ?

    I tried with one floating solid and the error don't seem to appear.

    Best regards,


  • The walls need to have a thickness. Give them a thickness, i.e. a few layer of particles to represent the boundary

    Kind regards

  • Hi,

    Thanks for your message.

    I added 3 layers with the command <layers vdp="0.5,1,1.5" /> in the definition of the boundary particles tank and it still doesn't work.

    I runed the case with only one floating solid and it works but there is no collision between the solid and the tank. I guess the error can be either from floatings solids/tank contact or floatings solids contact between them.

    Adding layers to the floating boxes is useless as they are defined as full solids ?

    I am currently running a case where I changed the rigid algorithm value to 1 (SPH interactions are allowed between boundary objects) instead of solving solid interaction with DEM. The code is running and for the moment the error don't appear..so I guess there is something with the DEM resolution..

    Regards

  • I would suggest you to use Chrono collisions instead of DEM since that option is more robust and indepent on the resolution (dp).

    You can find examples in: DualSPHysics_v5.0\examples\chrono\07_DamBreakCubes\DamBreak3Cubes_Def.xml

    And you can read about collisions with CHRONO in: DualSPHysics_v5.0\doc\guides\XML_GUIDE_CHRONO.pdf

    Regards

  • Hi Alex,

    I tried to use Chrono but I have an error saying that Actual VTK are not found.

    Do you know where it can come from ?

    Regards


  • edited April 9

    @Alex I found an old discussion on the forum where you explained that the Case__Actual.vtk needs to be created first, so I add the line and it seems to run.

    Thanks

    Regards

  • perfect

  • Hi @Alex ,

    Is it possible to save the coordinates of the center of gravity of my blocks at each time step ?

    In the output files I don't see where I can extract this data..

    Best regards

  • Sure, you have to use FloatingInfo tool !!!

    Some examples include how to use it. Please check the folders with floating objects and many examples with CHRONO

    Regards

    Alex

Sign In or Register to comment.