A couple fluid whirlpool tests

Whoops, I think I posted these in the wrong forum the other day.

Here they are again:

Second try:

I’m working on a 3rd where there is an invisible tube to contain the water. I think that the square domain is killing the sweet whirlpool action. I seem to be having issues with it though. Hopefully I’ll have it all rendered up to post tonight or tomorrow.


Hey thats coooool. I seem to remember that the cheat for this in Pirates 3 was to make a turbulent surface and then deform the plane into a whirlpool. Avoiding the volumetric effect. but I really like what you are trying to achieve here.

looks good! how long did it take your pc to bake that?

it’s funny because i was in the middle of these same tests, hah, i’ve got quite a few different setups with different sized spinning paddles at different speeds, about the most i’ve been able to figure out is how hard it is to get just the right balance between the “mass” of the spinning object and it’s speed

good work thus far

Assuming your hardware can bake it, here’s a .blend that will give you a vortex guaranteed. link


3rd try didn’t go so well. These bakes are taking 4 - 8 hours each. I cheated a little this time and turned the resolution down to like 90. I think you need a higher resolution at this size. Also, I angled the blades of the propeller, and I think that was kind of a mistake. It seems to just throw all of the fluid to the top of the domain and keep it there, especially when the prop goes into super mega turbo nitrous at the end.

Hahahahaha! I just had a funny idea, we should try one with counter-rotating propellers. Pretty much guarantee our computers would explode, there has GOT to be a divide-by-zero in there somewhere if we tried that.

I’ve got another one baking right now with straightened blades and higher resolution fluid.

Very interesting stuff. Creating vortexes like that hadn’t even occurred to me!

I don’t have any experience with the Blender fluids engine so I can’t really comment on the simulation settings, it seems to me that you want to leave some space to the side of the paddles so that the water has more room to move around (make the paddles smaller in comparison to the diameter of the container cylinder.) Has anyone tried tilting the paddles a little bit to make them into props? Seems like you’d get better results that way, since it would move the fluid downwards from the center. That may create the vortex faster.

I don’t know how much actual physics can be applied to the fluid simulator, but it looks accurate enough from what I’ve seen so far. Might be worth a shot.

Hmm well you could always change the viscosity settings see what settings work…

try toilet bowl style: no moving parts, just a flush.

You should put the motivating object at the top and have its diameter be half the diameter of the water. Set its origin to be at the center of the fluid. Rotate it around the z-axis, simulating a spoon stirring a liquid. I think that might be better than having it at the bottom where its influence has to translate all the way to the top before the vortex begins.

inverting this whirlpool would give a very cool Funnel or Tornado effect with some halo material swishing around it. hmmmm