Smoothed Particle Hydrodynamics

joshuawildejoshuawilde Member Posts: 96
edited February 2012 in Working with GS (Mac)
Has anyone ever thought of or tried to use Smoothed Particle Hydrodynamics (SPH). I would really like to use it but I want to see what the GS Community has to offer first. Thanks.

Comments

  • joshuawildejoshuawilde Member Posts: 96
    If you don't know what SPH is, it is a way to create liquid physics without the built in physics. It could also potentially run faster.
  • CloudsClouds Member Posts: 1,599
    Has anyone ever thought of or tried to use Smoothed Particle Hydrodynamics (SPH). I would really like to use it but I want to see what the GS Community has to offer first. Thanks.
    It's not entirly clear what you are asking ? GS doesn't support SPH or anything similar as it's not a 3D games engine.
  • beefy_clyrobeefy_clyro Member Posts: 5,394
    No i do not think you can do any liquid physics properly in gamesalad .. hopefully 1 day, i have some cool ideas if that was to ever get implemented
    Warning There was an error rendering this rich post.
  • RThurmanRThurman Member, Sous Chef, PRO Posts: 2,881
    The challenge is in the current physics engine that GameSalad uses. As far as I know, Box2D does not support SPH. I might be wrong on this but I think Box2D uses Euler integration, but SPH is largely based upon Vertlet integration.

    Without Box2D support, SPH would need to be 'coded' using GameSalad behaviors. It's doable, but optimizing the algorithms seem to be beyond the reach of this tiny brain. (Mostly due to trying to use constrain behaviors to perform Vertlet integration.)

    If you have an idea on how to do this -- (translation: If you feel comfortable coding SPH in another language/system) -- I'd be happy to explore integrating SPH concepts into GameSalad.

    RThurman
  • MotherHooseMotherHoose Member Posts: 2,456
    have you tried the free trial?

    http://www.algodoo.com/wiki/Download

    lots of fun! … Thanx @joshuawilde!

    @};- MH
  • RThurmanRThurman Member, Sous Chef, PRO Posts: 2,881
    @MH
    The videos look like this is a great physics simulation tool (er... playground?!?).

    Looking forward to the day when GS can do more of this kind of stuff! (Or when we can figure out how GS can do this kind of stuff.)

    RThurman
  • joshuawildejoshuawilde Member Posts: 96
    All SPH is is a "huge" equation. If we can just figure that out then we can get it working.
  • joshuawildejoshuawilde Member Posts: 96
    the only thing i can't get is how to get the x and y position of liquid around another liquid actor without creating tons of global attributes
  • RThurmanRThurman Member, Sous Chef, PRO Posts: 2,881
    Yes that "huge" equation is only the beginning! Lots of variables to keep track of for each particle.

    It will be easier to do when GS can use arrays (Read/write tables). I just hope table lookup is not too slow.

    You will need significant amounts of information for each particle. And each particle will need to be able to access all the attributes of its closest neighbors (at least two particle-width depth away).

    Right now, I would estimate that with 20 particles or more, speed would drop to under 10 frames per second. (Based upon experiments with Vertlet integration for rope simulations.)

    RThurman
  • JohnPapiomitisJohnPapiomitis Member Posts: 6,256
    Particle colliders would be nice too :)
  • RThurmanRThurman Member, Sous Chef, PRO Posts: 2,881
    @JohnPapiomitis -- yes, I'm hopeful that GS will have a chance to continue to implement new functionality. Particle systems like in [unnamed 3D program] would be great!

    I am hoping that GS developers will eventually open up Box2D and add in these kinds of 'plugins'. (Maybe they already are!) Meanwhile, we can try to figure it out with the behaviors we are given. (Much is possible -- but many calculations currently brings FPS to a standstill.)

    RThurman
Sign In or Register to comment.

© Vanilla Keystone Theme 2025