nemesis wrote:
> But here's a question that may be or not relevant to this:  did you parent the
> fluid object to its container? 

That's not how it works. (Indeed, I'm not even sure what you mean by 
"container"? The pitcher, or the domain?)

The problem I had was I let the pitcher (an obstacle) touch the boundary of 
the domain, which apparently crashed out the fluid simulation without 
actually giving any sort of error message. It does look funky, for sure.

> Looks way better, 

Probably because he spent the hours of tweaking to make it look nice, while 
I'm still at getting it to look at all. :-)  Once you have the fluid 
simulated in a mesh, you can start adding all the materials and lights and 
stuff you need.

> though it doesn't attempt to move the liquid along with the
> container as you do. 

That was basically what I was testing - whether it would work with moving 
obstacles. Otherwise, Sunday breakfast might be a PITA to animate.

> For that, you need parenting.

No, it's nothing to do with parenting. Otherwise, it wouldn't pour from 
container to container, for example. Parenting is good for armatures and IPO 
curves, and if you try to do the physics engine with bits inside other bits 
without parenting them ( http://www.youtube.com/watch?v=hjbBcqiK3FU ) then 
the physics engine gets rather unhappy, but it's not to do with the fluid 
sim.  (Indeed, I think the fluid sim was probably done as a completely 
separate script kind of thing and later incorporated, since it works rather 
differently than all the other physics kinds of stuff out there, in terms of 
parameters and such.)

> BTW, this is mighty cool:
> http://www.youtube.com/watch?v=WruTNnF6Ztg

Yes it is.

-- 
   Darren New, San Diego CA, USA (PST)
   My fortune cookie said, "You will soon be
   unable to read this, even at arm's length."

Post a reply to this message