Each would be a spectacular improvement:
Rigid bodies interacting with fluid (two ways)
Fire hose bowling;) . (well, not in real time).
Surfing. Blowing sand. paper airplanes. Soooo many things to simulate…
Multi fluid. Bubbles in jacuzzi.
Those little toys that have clear oil and colored water.
Combining those two: surfing with bubbles!
for air the viscosity would have to be set so low. Would that cause problems?
Also, fluids would have a density setting.
Do you have a 68 Ghz quadruple CPU with 4 TB ram? And oh, which team of scientist will have the honor to perform this task.
AFIK even in Madagascar the water sim (like when Marty came in surfing) was choreographed. In the special feature it looked like they modeled the rough shape of the waves and superimposed some kind of fluid effect on top of it (using particles perhaps?)
And as far as rigid bodies interacting with the fluid goes: It is already in Blender 2.42. You can’t simulate a wild pool party with it though… just keep it simple.
You can’t simulate a wild pool party with it though… just keep it simple.
You could you’d just need a
68 Ghz quadruple CPU with 4 TB ram
to do it on.
Any of the things you mentioned would either choke alot of computers or take days to simulate a few frames, maybe a supercomputer could do it though.
you want to see some good fluids?
check this out: made by a german specialFX company:
http://www.flowlines.info.nyud.net:8090/rndreel01.html?coral-no-redirect
I have 3 gigahertz (two processors each 1.5X1^9 hertz). Following moores law, one double per 2 years, 68 Ghz will be accomplished in a little over 9 years.
But that won’t be enough. 300 Ghz is enough.
I have 3 gigahertz (two processors each 1.5X1^9 hertz). Following moores law, one double per 2 years, 68 Ghz will be accomplished in a little over 9 years.
But that won’t be enough. 300 Ghz is enough.
It ain’t just the Hz that are important, and moores law says nothing of the power of computers. In fact, it was just an observation that has happened to stay true so far.
Ian
No it isn’t. Animated obstacals are supported, but those just are ipo animations that do not react to the fluid. I mean rigid bodies that both react ti the fluid and react with the fluid.
If i’m wrong, post a .blend.
here i made a quick test with a rigid-body suzanna and a fluid, so point A is half solved… (-:
With the monkey you can tell that it doesn’t react to the fluid. You should make it look like Iridium (the dencest stable element), and the fluid look like the lightest liquid known:) .
I don’t understand why you say the monkey isn’t reacting to the fluid??
Holy wow…
Which actually has also the computer-power to compute this
Not more than implementing all that stuff you are proposing here …
… as for the paperplane …
… well I am working at a facility which houses Germanys experts in flow-dynmic-simulaion for aerodynamic research … so I can tell you, what you want is totally over the top … that means that it is sure possible ( we do it for example … but with years of scientific work and programming … calculated on supercomputers), but it would take up so much time to design and implement these things, that we could completely recode blender at least a dozen times with this effort … don’t you think this manpower should be used for more important problem than simulating airplanes and bubbles, like …
N-GONs
More moddeling Tools based on N-Gons (like better bridging, etc …)
fully funktional NURBS
REYES
SSS
fully funktional Normal Maps
Micro Displacment
GI
a powerfull history system
a powerfull preset and libary system (for materials, post-pro effects …)
a recode of the GUI and event-System
etc…
In all probability: realising all this things together whouldn’t need so much effort than these, anything but trivial, simulation problems you mentioned.
actually Blender’s fluid simulator in its current state is already capable of some astonishing results, such as:
http://mpan3.homeip.net/gallery/My%20CG%20Portfolio/slides/rain.html
and even animations:
http://mpan3.homeip.net/sub.php?mid=vids
mind you that these are created on a very moderate PC… with lots of patience and skillz 
He probably means that only the fluid is reacting to the monkey head, but the monkey head is not ‘aware’ of the fluid. IE: it’s movement should be damped when inside the fluid (larger fluid resistance), the waves that it creates in the water should rock it a bit, etc.
But yeah, that would take ages both to implement, and to actually use, there are far more important things to do first.
I know these, but simulation flow dynamics with the addition of aerodynamic effects and TWO WAYS of interaction between obstace and fluid/gas ( to simulate a flowing paper plane, as suggested) is more complicated than solving fluid.
And who said Moores Law stil works?
The performance improvements today come from better compilers, better prediction, using multi-processors and multi-computers.
There are physical limitations to the current technology. A transistor gate can’t be made much smaller, and to make faster, you have to make smaller. Besides, each time you make the gates smaller, you have more and more energy leak, and that is not good.
Thats right, and don’t forget quantum effects, and the problem of wireing all units in the cpu together … the cpu layout really is a problem today, and it will stop the minituralisation long before physikal problems come into account
What about those technologies such as quantum computing and chips using diamond as a base. It’s those types of technologies beyond silicon that will keep Moore’s law going.
I think i’ve read the limit of silicon will be reached sometime in the next decade. Which is why they are finding ways of moving beyond traditional silicon.
If you start using Diamond wafers in a chip you’ll be able to run a computer at speeds at around 60-80 Ghz. They can also run at temperatures where current computers would melt.
quantum computers, if one day they succeed in making one, and researchers themselves are not sure the will be able, won’t be available before decades…for the diamond, the cost will be enormous no ?
