Non-uniform particle density

It seems in the GenCase, only uniform particle density can be used. Will the solver accept non-uniform particle densities?

If the solver accept, i think maybe i could manually generate the floating object (import complicate geometries from STL) using higher particle density while the fluid can use smaller density. This may be helpful in providing more accurate results while saving calculating efforts i hope.


  • particles of floating objects can have different density than particles of the fluid... so I am not sure what is your question... please can you explain the problem again?

  • edited January 2019
    @Alex Sorry that I didn't make myself clear. I plan to analysis the motion of a floating object which enters the water with a initial speed. The floating object is imported from STL with relatively complicate geometries. So i would like to have different dp for the floating object and the water. But in the _def.xml file, only one dp can be input, so my question is how to use sperate dp values for the floating object and the water respectively?

    See attached the _def.xml file for reference.
  • Ahhhhhhhhhh ok ok
    With the current online version we can only create particles with fixed dp for fluid, bound and floating, sorry!

  • @Alex If I use my own code instead of GenCase to generate particles with different dp for floating object/water, can the current DualSPH solver take this kind of inputs and run analysis? Or is there any modificatiins needed from the solver as well?
  • There are more modifications to do.
    When you use different "dp", then we will have different "h", so that different cell sizes for the neighbour list....... In addition the governing equations... when computing interaction between two particles of different "h" need to use kernel functions defined for different "h" lengths.....

    More information in R. Vacondio, B.D. Rogers, P.K. Stansby, P. Mignosa, J. Feldman (2013a), Variable resolution for SPH: A dynamic particle coalescing and splitting scheme, Computer Methods in Applied Mechanics and Engineering, 256, April, pp. 132–148,
  • edited January 11

    @Alex Can you suggest how to simulate or modify the source code, if it was another non-uniform property of the floating body such as thermal conductivity, that doesn't affect the particle spacing?

Sign In or Register to comment.