[ODE] fluid dynamics question

Stefan Kuzminski pontifor at yahoo.com
Wed Mar 8 13:50:26 MST 2006


I guess what I'm really after is something like RealFlow from Next
Limit.  It would also be nice to tie that with a thermal model such as
when gases get compressed.  I had a fantansy that I could find various
C++ libs and tie them up with Python but it looks like those comercial
codes have many man-years of tuning into them.  I supose that is more
of an engineering or video production tool as realtime is not
important.

thanks for the info.
S

--- jon klein <jk at artificial.com> wrote:

> 
> On Mar 7, 2006, at 4:23 PM, Stefan Kuzminski wrote:
> 
> > Hi,
> >
> > I was wondering if anyone has any experience or thoughts around
> > integrating ODE with any computational fluid dynamics sofware.   
> > Most of
> > that stuff is high-end, but there must be more accessible software
> out
> > there somewhere ( realtime is nice but not necessary ).  This link
> is
> > intriguing, although I have yet to find the C code that goes with
> it..
> 
> I suppose it depends on what kind of integration you're looking for.
> Stam's code (and other similar techniques) simulates only a single
> fluid -- they cannot, for example, simulate the interface between air
> and water.  So while they can simulate underwater currents, they
> cannot simulate waves moving over a body of water, and they could not
> simulate ODE objects interacting with bodies or drops of fluid.
> 
> These fluid simulation methods do, however, allow for obstacles
> submerged in the fluids.  It would therefore be fairly
> straightforward
> to simulate the effects of submerged ODE objects on the fluid
> behaviors.
> It should also be possible (with a lot more work) to simulate the
> effects of the fluid on ODE objects.  But if the effects of wind and
> water resistance are what you're really after, I think this type of
> fluid simulation is probably overkill -- there are easier approaches.
> 
> As a final note, though, any integration with these variety of fluid
> simulations would likely be very impractical because of the
> computation
> required for these kinds of fluid simulations in 3D.  While 2D fluid
> simulations can be pretty fast even at 256x256 or higher, 3D fluid
> simulations are painfully slow.  At 64x64x64 you're lucky to get a
> few frames per second, and that's still probably too small for any
> meaningful interaction with larger scale physical objects.
> 
> - jon klein
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 


More information about the ODE mailing list