> On what system did you get that compute/render times ?
I did the computation on a P4 3.2 GHz with 1 Go of RAM.

> Very impressive, even if the end of the animation makes the liquid
> surface a little odd (and not only because of color ;-) )
Thank you. I think the end is a little bit odd due to the fact that I didn't
modelize any surface tension.

> Any hope to see such a simulator on-line one of these days ? (If you
> know of one freely available, please tell me).
I plan to release it one day, but I have to correct 2 main problems before
to do so :
- The simulations are to slow to be used in a real project.
- For the moment, the geometry of the objects (for exemple the glass) are
defined by an isosurface hard coded in the simulator. I will write a small
parser (the integration in POVRay would ask for much more work).
I don't know any free liquid simulator suitable for computer graphics. It
exists some commercial ones like RealFlow (which looks very good). You can
find some free simulator of NS equations with source code available (for
instance, you can make a search on sourceforge web page if you are
interested) but still remain the gap to integrate it in a CG package as
POVRay.

Fidos

Post a reply to this message