Explosions in Maya fluids

[modezty-blaze]

THE CLIP:

http://www.youtube.com/watch?v=7eR2yp4lBPg

I have been struggeling quite some time with maya fluids. I want to make fat-ass exposions but i feel kind of constrained by all the amount of time it takes to simulate all the new tweaks. I'm getting faster with producing these sims but it's hard to get to a real nice end result.

When i am eventually done with my sim, it can look really different rendered since it's improssible to see the texture properly in the viewport. The other issue is resolution. If i increase the resolution of the container, the simulation will change, and might look completely different. when i go into higher resolution than 60*100*60 (like this explosion) every frame takes like 2 minutes to sim, which makes it untweakable.

It would be nice if there was some way of making the sim in one resolution, than just hightening resolution before render, knowing it's going to look sort of like the other simulation.

In this simulation theres no texture time anymation, but can't see stuff not moving. Do you? Is there stuff that i can look after in the playblast that reveals stuff that will look crap in render, or some kind of effective workflow?

I know it's kind of fuzzy what i'm exactly asking here. Some rule of thumb for above mentioned, or tips för wheree i can get some good info on explosions. Fluids are mean, and i'm all alone

[joojaa]

It would be nice if there was some way of making the sim in one resolution,
than just hightening resolution before render, knowing it's going to look sort of like the other simulation.

Well yes however, that is inddeed the nature of how the Navier-Stokes equations work, especially how turbulence works. See turbulence is a sort of feature that exists because of the grain in the solution, ita a really hard problem. Lets just say that if you could make a fluid simulator thet would in actuallity converge with a real turbulence result then you would actually use that. But its not really a thing that scientists actually know how to solve properly.

In fact IF you can solve the navier stokes equation mathematically to solve thsi problem you WILL be awarder 1 million dollars immideately on proof. This should give you a idea what your up against.

So you should probably just ignore the results of the turbulence whan simulating as the turbulence look WILL change, no question about this. The other alternative is not to use fluids at all, because you will not be able to manibulate teh result very well with fluids directly, after all your not looking for a realistic simulation but a artistic expression of the result.

[modezty-blaze]

Thank you for telling me about the nature of fluids. My end goal is to be able to do a "realistic explosion" with enough production value to suit a film.

I have had the impression that the nicest looking explosios are made with fluids/flow. Are you suggesting any specific alternatives?

[joojaa]

I have had the impression that the nicest looking explosios are made with fluids/flow.

Perhaps, but theres no reason why you can not mix and match techniques as you see fit. Theres nothing that says you cant use particles that render looking exactly teh same as fluids because your using the same particle renderng techinque on a particle sim with no fluid sim at all.

See the problem is that the kinds of flashy film studio renders you may see are often done over LONG period of times, they include many layers, many effects are combined, and alot of computation time is wasted, and a lot of compositioning is involved. So while thay may use a fluid, perhaps for some part of the effect its allmost certainly true that theres a LOT of work on top of the thing.

See simulation is not allways the easeiest thing in the world. Its often a prcess of what you see is what you got. That is unless you have the mathematical skill to build up the simulation of even more detailed pieces. Then you might get to say what you actually need.

For many tasks doing the thing by hand is better than actually simulating it.