Hey guys, soon I am going to be shooting a vfx video for our YouTube channel, and it will be using Blender and After Effects. The biggest job for me in Blender is to make a realistic car explosion, so I thought I’d get some explosion practice. It’s a lot of fun, but my computer sucks so it takes forever, but I think it will turn out well.
I decided to start small and blow up Suzanne:
Please feel free to comment, critique, or add your own explosion tests! Sadly I haven’t seen too many good ones out of Blender, but hopefully that will change!
I am working on an actual car right now, but it might be a day or two before I post it because of baking and rendering.
So the file I was working on crashes like crazy for some reason, but I know it’s my computers fault and not Blenders. Anyway, I had a cool car crash going where the car crumples via shape keys, the windows explode, and it gets flipped over. But it’s gone.
The one big problem is still the explosion. Ones like the one above don’t look quite right for a car. I need something more rounded, like this:
I will get to work on this again later today. I don’t need this to be super realistic, but more exaggerated and stylized.
Any suggestions?
Thanks,
Jonathan
Yet another try at a nice explosion. I think the materials are all right, but it just doesn’t seem explosive. It just kinda plops out, which is not quite right. I guess more tests will be on the way seeing as I definitely haven’t found the secret yet.
i think the scaling of fire/smoke compared to monkey and "ground"doesnt fit well.
Good point, I will keep that in mind for later.
Thanks! I am putting this project on hold because we decided to release the video that needed an explosion a little later, and push up a vido less heavy on the VFX side. I will come back to this though, because I really want to push the smoke sim into making something close to professional. If you guys have any tips or examples that would be much appreciated for later.
I think that you really need to look at durations of effect. The initial expansion is a concussive wave through the air followed by gas. Then debri and fire as ignition of air takes place. I would time some explosion from Mythbusters, they are shooting the real thing with high speed cameras all the time. Try graphing the different expansion stages and brightness. BTW I love both those tests, they look amazing.
^^ Thanks, I actually read that before I started my tests. The steps of an explosion is pretty clear (flash, fireball, smoke overtaking fire, etc… , but for some reason it is really really hard to set up in the smoke sim.
Thanks though.
I would suggest watching andrew price’s tutorial on explosins, it will show you how to perfect those smoke tyails, also, try and work on your color ramp, the explosion is a little bit too yellow.
I would suggest watching andrew price’s tutorial on explosins, it will show you how to perfect those smoke tyails
I have, but to be honest, I don’t like the result. I learned a few tips for sure, but overall the explosion doesn’t look that great. Sorry, but it’s true.
also, try and work on your color ramp, the explosion is a little bit too yellow.
I had a bit of time today to make another explosion. I think this one worked pretty well except the very beginning looks bad. I also had a problem with the lights immediately dying because I forgot to key frame their intensity so they just die and don’t fade. It’s easy to fix, but I don’t really want to re-render for just a test.
It doesn’t look as cool, but I think the feel is a lot better.
Your third video is look a lot better. My issue would be the smoke it kind of pumes out instead of explosing out. I havn’t played with smoke yet so I don’t even know of a way to do this.
I think what needs to happen is the smoke needs to expand/explode in all directions, then slow down and pume upwards. Other then that I think you have nailed it.
I was watching the making of Apollo 13, they said the key to the success of large fire simulation was changing direction of air flow. That is, when the rocket burned a large area of oxygen and created a low pressure zone, air from outside the flame rushed in. This created a reversing (or inwards flow) of the flow of smoke and flame.
So I noticed that there was a rapid explosive expansion, followed by a sucking inthen a re-ignition of oxygen. Also Mythbusters explosions tend to have a concussive wave that pushes the smoke and flame out in a sphere immediately after the first strike, then the ball of flame and heat rises seperately. Perhaps you have the slower heat ball to close to the initial explosion?
All this is just speculation, sorry i can’t be more helpful.
I was watching the making of Apollo 13, they said the key to the success of large fire simulation was changing direction of air flow. That is, when the rocket burned a large area of oxygen and created a low pressure zone, air from outside the flame rushed in. This created a reversing (or inwards flow) of the flow of smoke and flame.
Very interesting, cool tip!
Have you seen this page? The Anatomy of an Explosion
Yep, I have been using it as a rough guideling, but for some reason it is really tricky. Thanks for the “cute” breakdown.
Here is another simple test of a fireball hitting some explosive barrels. As usual I forgot something (my computer is so slow I pretty much have to change settings from memory and bake). This time the fire disappears all of a sudden because I forgot to reduce the flame amount over time, but it just happens to be the exact second of that camera shake. The shake is supposed to be from the initial explosion, but from far away - hence the wait time. Sorry about the domain being so obvious, but it’s a test and I couldn’t spare the memory.
Hey guys, long time no update on the explosions because I’m still awaiting my beast computer, but in the meantime I thought I would try an explosion in cycles.
Probably the worst looking explosion in VFX history, but I had very little time and was more concerned about getting it working in cycles than it looking pro. Perhaps the next one will be better.
Haha I know it sucks badly, the point was to try cycles. I had like an hour total, so I didn’t have time to do much else other than the extreme basics. Don’t worry I haven’t forgotten what I’ve learned.