[ODE] Ship simulation

[Nikita Buida]

Thank you for the useful link :) It is shame I didn't make a proper
search through mailinglist archives... 

The method described there is very simple but I'm not sure how could it
be used with ODE. I mean it is simple and works just fine if water
surface is approximated with single plane... You just intersect all
spheres with plane and act apropriately on collided spheres. But how can
I make this work with fairly complex water surface, say represented with
heightfield? I'd like to use ODE's own collision detection
functionality. I don't know much about ODE internals but from interface
point of view there's no obvious way to make sphere/heightfield
intersection test. Maybe I'm wrong here?

By the way code for tutorial is not available. From the image in the
article I guess author simulates water with single plane...

Any suggestions?

Nikita


Jean de Largentaye (2006-04-19, 14:26):
> Maybe this buoyancy tutorial could help you:
> http://www.cambrianlabs.com/Mattias/DelphiODE/BuoyancyParticles.asp
> 
> The associated code is in Delphi, but I expect it can help you nevertheless...
> /me takes note to bring it into the wiki
> 
> John
> 
> On 4/19/06, Nikita Buida <nikita.buida at gmail.com> wrote:
> > Hi, ODE community!
> >
> >         I'm planning to make a ship simulation using ODE. There should not be
> > too much precision and main criterion is natural ship movement.
> >
> >         The idea is to approximate body of water surrounding ship with dynamic
> > water 'boxes'. Ship's hull is approximated too with some simple
> > shapes... spheres/boxes. Then all water geoms are placed in separate
> > space so they don't collide with each other. The last step is to assign
> > proper contact joints parameters such that water washes the ship
> > 'softly' but applies contact forces to simulate ship movement. I guess
> > some simple hydrodynamics model could be added with this approach.
> >
> >         The other idea is to perform collision detection of ship's hull/water
> > body and to apply corresponding Archimedean forces... Everything
> > without contact joints at all.
> >
> >         I already have some experience with ODE but I'd surely like to
> > hear your opinions! Maybe somebody already tried something similar? I'd
> > be thankful for any advice :)
> >
> > Cheers!
> >
> > Nikita Buida
> >
> >
> > _______________________________________________
> > ODE mailing list
> > ODE at q12.org
> > http://q12.org/mailman/listinfo/ode
> >