FumeFX

#5221

So, I am trying to create a directional explosion, and I could use some advice. I know how to create general explosions pretty well, but I need it to VERY directional, like a battleship firing it’s deck guns. I need it to travel upward at a 60ish degree angle.

How would you guys approach this? I tried directional sources, but it reads as such, and begins to “mushroom” as the velocity reaches too much. I think I might try putting the emitter in cylinder geometry and try to get the pressure to “push” it like a real cannon would.

Any suggestions would be appreciated. :slight_smile:

#5222

Joker I recently did a scene with multiple candle flames using FumeFX, I found issues when using instance so I used copy instead. I had 30+ candle flames that all functioned and rendered without an issue.
I’m not a pro with FumeFX but this was what I ran into.

Rich

#5223

Thank you guys for your suggestions.

I’m rendering dozens of sim boxes but I’ve broken them up into sections of 12 at a time.

I was doing tests on 2 sim boxes which were around 5 minutes a frame down to 1:30 m/s a frame.

Noe the longest part of my render is, what the render dialog calls “Updating FumeFXXX”. What exactly is that doing? Loading in the cache I’m assume?

Does anyone know of anyways to speed that process up? I’m narrowing in on the problem and I believe that is it. The render of the scene is simple. Just 2 lights and 12 sim boxes.

#5224

Have you tried post processing your cache? You can strip out channels you don’t need for rendering and minimize your grid.

#5225

rendered 500 sim boxes with each containing a cache of around 800Mb each using the new fileformat f3d. When using the default fumeFX cache…no render at all.

90 sim boxes took around 5 minutes. The 600 took a few hours.

Thanks guys!

#5226

There are some issues surrounding instanced grids and/or source object, it it recommended to make copies instead. It is being looked into, I haven’t had time to check in it in the final 3.5 build to see if it has been fixed.

#5227

Exactly what I said, Instanced FumeFX would not work for me but making Copies did. It does take a bit of time and the 3dsMax file is quite large but it renders beautifully. I have not had the opportunity to check the update but I hope they have fixed it as well.

#5228

I find it very useful to have a very high turbulence for a few frames after it fires, breaks up the mushrooms, then drop the timescale for the swirling after.

#5229

And/Or localize it with an Effectorizor :slight_smile:

#5230

Yeah with effectors you can localize the turb to the high velocities or temperatures. That is sooooooooo useful!

#5231

Speaking of effectors and vorticity/turbulence…in an explosion, wouldn’t the fire be more vortic (is that a word?) and turbulating(word?) than the smoke? Since the fire is hotter, in my mind it would be swirling more than the smoke.

So I was thinking about having an effector on the fire channel make it a vorticity of 1, and just have the regular grid on the smoke have a vorticity of about 0.5-0.6. This makes sense in my head at least. :slight_smile:

#5232

It is my experience (firefighter) that the smoke will be more turbulent as the fire typically creates it’s own wind. Fire will follow the Oxygen, and the Fuel as long as the heat source is present. Smoke tends to move all over the place. Usually where ever I sit at the campfire is where the smoke goes …LOL

#5233

But is it not true that hotter air (i.e. fire) will create more turbulence than cooler (i.e. smoke)? I could be mistaken, but this was my assumption. Whenever I see very turbulent smoke, it seems to be the result of coming directly out of a very turbulent flame.

#5234

Smoke is typically considered the byproduct of combustion, so fire will effect the smoke by the winds or turbulence it creates. Fire can be turbulent as well but you normally see the smoke as the most turbulent. That’s not to say that fire can’t be turbulent especially in the case of a fire tornado, and fire can be turbulent in a rollover where as you see fire rolling across a ceiling etc. Flashover is a smoke explosion so that’s not fire turbulence its hot gases exposed to oxygen rapidly. Smoke can be greatly affected by heat and the turbulence created by the fire itself. Let’s not forget that some fires can burn very hot so much so there is no smoke, alcohol, Sterno cans are all examples of a fuel that burns very hot and typically no smoke. Many factors affect fire and smoke mostly the type of fuel, the type of heat, and the amount of oxygen. Some would also say that the Chain reaction involved to create fire may also effect the Fire and smoke. In my observation of fire I think it moves according to the oxygen and it gets turbulent when searching for it.

#5235

afaik there is only ONE velocity field under the ffx’s hood. Vorticity is computed using this field. So there is no fire or smoke turbulence. If fire is in the voxel - then it gets this computed vorticity, if smoke - smoke gets. The only difference MAY be in shader, where for smoke and fire there are slightly different mappings of the temp and velocity values.

#5236

True, but I was talking about in reality. If you watch a standard hollywood movie explosion (a real one) it seems the fire is much more turbulent, and after it burns off, you have a nice soft rolling black smoke which doesn’t seem as turbulent and violent as the fireball it came from.

I’m running explosion effector tests actually right now which one has a lower vorticity on fire (0.3) and the smoke is at 0.6. On the other test, the fire vorticity is 0.8, and the smoke is again at 0.6.

I will post my findings. :slight_smile:

#5237

I can’t wait to see the results Daniel-B. Hollywood Fire is a control based fire using several forms of pyrotechnics. They use different chemicals and fuels to either get more fire or more smoke. Again in real fire smoke is determined on fuel and oxygen fire is based on heat fuel and oxygen. With FumeFx we can actually create better realistic fire than Hollywood’s real (pyrotechnic) fire, its a matter of hitting that right setting. I wished I had a helmet cam on when I entered fires to share real life fires for everyone for reference. When someone tells you fire is alive it is true, and it has a mind of it’s own.

Rich

#5238

So I’ve simmed a cloud with the desired settings and whatnot. I’m pleased with the look and feel of the cloud as well. The only problem is that when I scale the cloud both larger and smaller than it’s original 100% size, I get odd looking results. Now is there a way or settings I need to be aware of when change the size of the fumeFX box to better max the original cloud which is represented at 100%?

The image helps make more sense. I want the scaled cloud to look exactly like the cloud at 100%, but just scaled larger in the scene. I’m not understanding clearly why it changes so much in appearance?

#5239

It looks like the density is changing with the scale. Try increasing the render opacity in the smoke the same proportion that you are scaling the fume grid down, or vice verca.

Be careful when you scale as your global step size gets affected too, so you may see render slowdowns/ render artifacts depending on which way/how much you scale.

#5240

Yeah I ended up changing the opacity.
I scaled the sim box from 100 to 1000
so i changed the opacity of the smoke from .75 to .075 and it fixed it exactly.
Only problem is the 1000 sized sim box takes 8 minutes to render where as the 100 takes 54 seconds.