spooky fluids!

May 10, 2006 3:14 PM

Order from chaos!Fill a cylindrical bucket with water and make it so the bottom can spin. At certain speeds, stable regular polygonal shapes will spontaneously form at the turbulent surface of the water. See the video. [2.6MB avi] [via last week's PRL]
posted by sergeant sandwich (28 comments total)

Shouldn't the shapes be a reflection of the water molecule structure? I know rocks cleave naturally along planes that refect their atomic structure -- it seems like the shapes might be when the little H₂0s fall into a simple polygon lattice shape, and the macro view reflects that.

(keep in mind I'm talking out of my ass here, ten years after finishing a masters in soil chemistry so I could be way off)
posted by mathowie at 3:19 PM on May 10, 2006

Interesting findings. However, scientists have know for years that water can be made to take almost any shape simply through a reduction in temperature. No spinning required.
posted by redteam at 3:52 PM on May 10, 2006

well, they were apparently able to generate similiar behavior with ethylene glycol, which has a different molecular structure than water (and a much higher viscosity). it sounds like it's got something to do with vortex formation, but i don't think the authors have a definite explanation. i think it's another example of the not-very-well-understood phenomenon of symmetrical structures in chaotic systems.
posted by sergeant sandwich at 3:55 PM on May 10, 2006

Related
posted by gimonca at 3:56 PM on May 10, 2006

From my (very limited) understanding of these types of things, they are not easily explainable. We have no solution for the Navier-Stokes equation, so we can't completely understand these systems.

At least that was the gist of what I got from a physicist I knew that was studying such things (he would put water in a dish and hook it up to a signal generator, which was driving some modified woofers that supported the three struts that held up the dish. He put it on an overhead projector, added food coloring, experimented with fluids, and made movies. It was pretty cool to watch, sometimes. Really surprising and very ordered patterns would result from time to time. There is no complete theory to explain why, when, and how such order would appear).

Cool stuff.
posted by teece at 4:05 PM on May 10, 2006

I find this very interesting. Intuitively, it seems that these shapes are a reflection of somthing, whether it's the physical properties of water molecules as mathowie suggests, or a reflection of the physical forces involved.

My bet is on the angular momentum combined with water molecules finding some sort of equilibrium. Note that all the shapes are radially symmetric.

Of course redteam, you are correct. Sometimes when I look at flowing water, in a stream or river, It's pretty amazing as well, and no polygons involved! :)
posted by kuatto at 4:13 PM on May 10, 2006

Note that all the shapes are radially symmetric.

It's pretty hard to imagine a stable hole that wouldn't be. Whatever forces are in play have to be balanced, an asymmetrical hole seems like it would collapse or migrate.
posted by sonofsamiam at 4:17 PM on May 10, 2006

Wow.
posted by jb at 4:17 PM on May 10, 2006

You're right, kuatto. This "stream" phenomenon is also amazing.

I haven't yet published my findings, but I have found a very simple way to create rectangular solids from water using the temperature reduction method.
posted by redteam at 4:22 PM on May 10, 2006

If you think about the way crystals and things that depend on molecular properties work, you can see why this probably isn't about that. To get that sort of shear, alignment ect, you've got to have those molecular aligning forces be the highest order forces on the object. Here the major force is on the macroscopic order, so we've not got something that admits that sort of subtlety. There's some neat stuff in the physics of this; and its also cool in that its easy to get a picture of it in your head.
posted by apathy0o0 at 4:34 PM on May 10, 2006

Why the heck is my WMP unable to find the codec to watch this?
posted by gigbutt at 4:39 PM on May 10, 2006

Maybe the shape is related to how the experiment's conductors were feeling.
posted by knave at 4:48 PM on May 10, 2006

btw, here's some possibly related phenomena studied by the same crowd.
posted by sergeant sandwich at 4:50 PM on May 10, 2006

Why the heck is my WMP unable to find the codec to watch this?

It's encoded with the Divx codec. Microsoft's "automatic codec download" bullshit is essentially worthless for everything except MS codecs (WMV and WMA.) Install the Divx decoder (or better, skip the adware garbage and install the Xvid.org decoder.)
posted by Rhomboid at 5:10 PM on May 10, 2006

I'm not a physicist, so if I'm wrong please correct me, but isn't it possible that what we're seeing here is an intersection between centrifical force and intermolecular cohesive forces in the water? Centrifical force leads to water pushing against the side of the bucket, moving away from the center. Couldn't the shapes be due to some sort of inaccuracy in the rotation of the bucket which appear in shapes due to symmetry? Obviously, this entire phenomenom is ripe for experimentation.

On an unrelated note, my first post on Mefi! I've been a lurker for a couple years but my desire to comment on this post finally motivated me to get off my butt and pay my $5.
posted by switchsonic at 5:51 PM on May 10, 2006

I haven't yet published my findings, but I have found a very simple way to create rectangular solids from water using the temperature reduction method.

I have personal experience with Eskimo scientists who have been pursuing this technology for years. If there were some way you could sell it to them...
posted by skammer at 6:04 PM on May 10, 2006

These are interference patterns due to the waves in the bucket resonating in tune with the rotation of the bucket.
posted by StarForce5 at 6:15 PM on May 10, 2006

I studied fluid mechanics and materials science (quite a while ago) in college but I haven't a clue how this works - not sure I can even guess. It's excellently, unexpectedly, fantastically cool, however.
posted by normy at 6:16 PM on May 10, 2006

I think its probably the hand of God. Weird that (s)he only has 3 fingers, though...
posted by pkingdesign at 6:23 PM on May 10, 2006

I found the pictures better than the video.
It looks to me like some harmonic situation is reached with the waves radiating out from the vortex, and the reflected waves back from the cylindrical wall. I wonder if a very large cylinder rotating would create the same effect? It would be easy to test the theory, vary the size of the cylinder and see if the speed at which the phenomenon occurs changes.
But then having spent a large part of my adult life studying radio waves, and some of the harmonic aspects of those, it might just be me looking at it through the lense I'm most comfortable with.
Cool stuff though ....
posted by forforf at 6:41 PM on May 10, 2006

Looking at the pics closer, that's exactly what it looks like to me. It appears that there is a harmonic created by the spinning mixer thing that "pushes" the water out of the polygon at the exact frequency that the water naturally would push back in. Ok, that's a terrible explanation of the phenomenon. Think of a pendulum, and there is a frequency that you can tap the pendulum and it will swing further and further out (this is similar to how people swing on swings). Now the spinning mixer thing appears to be the "tapper", with the water acting as the "ball". Throw in the rotational force and this seems like its possible. In fact, I seem to recall that a rotating pendulum describes geometric polygonal arcs at certain frequencies too ... I couldn't find the exact link, but I did find this one of pendulum traces, and to me it looks similar to the formations appearing in the water ...
but again ... pure speculation by me ....
posted by forforf at 6:57 PM on May 10, 2006

I am reminded of the hypercube.
posted by TwelveTwo at 7:33 PM on May 10, 2006

I am reminded of the Timecube.
posted by Toecutter at 7:46 PM on May 10, 2006

Anyway, my theory is that it has to do with mini-vortices being spun off the central one -- you can see them forming in one of the side views. When two, three, four, etc of these mini-vortices occur at once, they migrate outwardly at the same speed from their common experience of centrifigal force (thus producing symmetry), each mini-vortex forming a (rounded in every case!) corner of the polygon. Tornadoes do a similiar thing -- spawning mini-vortices, that is.

Sidetrack warning . . .

Just this very evening my wife and I were talking about how sometimes when you rub your eyes quite regular geometric patterns -- sort of like checkerboards or mosaic pavements -- appear behind your eyelids. Related? Hmmmm . . .
posted by Toecutter at 8:05 PM on May 10, 2006 [1 favorite has favorites]

In fact, in the video, you can quite clearly see the two mini-vortices form and move away from one another at roughly the same rate.
posted by Toecutter at 8:10 PM on May 10, 2006

Its the merkaba appearing in the water. The architect of the universe put it there.
posted by godseyeview at 8:11 PM on May 10, 2006

So, uh, did anyone read the abstract, then? I'm not a physicist, and especially not versed in fluid dynamics, but here's what I got from it:

The authors find that the radial motion of the fluid against the container sets up, for some reason, secondary cycles of flow that move up and down the wall, in addition to the primary motion around the axis (props to switchsonic for getting the radial motion vs. viscosity bit, and, btw, welcome!).

Ordinarily, this sort of thing dissolves into chaos, but, according to the authors' speculations, as the bottom plane of the container wobbles, it drives the system regularly, and for some reason the system's reaction falls into phase with the driving force (props to forforf for the pendulum analogy; sometimes driven pendula are stable, sometimes chaotic). So a whole number of rotations takes place while one cycle of changing pressures and flows plays out (this is, I think, what the authors meant by 'mode-locked'; compare the tidal locking that keeps Io, Europa, and Ganymede orbiting in neat multiples of one another's periods).

I suppose that the ratio of their periods determines how many instances of the system are going on at once, and then that they spread out and push away from each other, for some reason, until the polygon is more or less regular.

The vagueness of 'for some reason' is tied to the mysteries of fluid dynamics, as mentioned by teece.

And pkingdesign, everyone knows that the Hand has five fingers.
posted by eritain at 9:52 PM on May 10, 2006

Thanks sargeant sandwich, that guy's doing a lot of cool stuff
posted by NinjaTadpole at 4:03 AM on May 11, 2006