I’ve observed this happening in coke but I’m sure it happens in all other carbonated drinks too. As the bubbles rise they zig-zag rather than just rising in a straight line. Why is this?
I assume it has something to do with air pressure but not being terribly knowledgable about physics/chemistry I can’t figure out why. Could someone please enlighten me?
I’m guessing that as the bubbles rise they encounter resistance from the liquid. This resistance deforms the bubbles so that they are larger on one side than on the other, so they veer off in that direction. As they rise in the new direction, they encounter resistance there and deform to the other side.
IANA Physicist.
As an object (like a more-or-less spherical bubble) moves through a fluid (the Coke), after a certain point the flow of the Coke around the bubble is no longer “steady”.
At extremely low speeds, the Coke will neatly move out of the way of the bubble, and will “close back up” behind it - all in an axially symmetric fashion. However, above this (pretty low) threshold, such a flow pattern becomes unstable, and that doesn’t happen any more.
The bubble will essentially begin shedding vortices, and it will settle into something of an oscillation - vortex off one side, followed by one off the other. Doing this creates some lateral forces, and since there’s really nothing holding the bubble to straight-up path, it’ll oscillate somewhat in the process.
This a is really simplified explanation, and I could be quite a bit more detailed if anybody likes. It’s a very well-studied problem in fluid mechanics, with many practical implications. A bubble is more complicated than, say, a cylinder in a freestream for a couple of reasons:[ul][li]The bubble can (and does) deform in response to the motions/forces of the Coke around it3-d objects are more complicated to visualize than 2-d ones (although flow over a cylinder breaks down into 3-d structures eventually, as well).[/ul][/li]
A few years ago, somebody in my adviser’s group did a small project detailing statistics of bubbles rising in (I think) champagne. They certainly didn’t discover the phenomena, and nor were they the first to study it in detail. I believe what they were doing was testing a new technique for tracking small objects in 3-d, and this was a convenient “test bed” for that.
And that single “phenomena” I referred to would be more properly called a “phenomenon.” I hate committing my own pet peeves.
That’s plenty complicated enough for me. brad_d. Thank you 
Is the bubble moving through the Coke, or is the Coke falling to the bottom of the bubble?
Both. The gas that is the bubble is moving, and the Coke is falling. But I suppose the latter is a reaction to the former, since it is the weight of the Coke that is pushing the gas out of the way.
I would imagine that the surface tension around the inside of the bubble keeps the coke from falling through, otherwise the bubble would fragment into many smaller bubbles as a pillar of coke pored through the middle. A better analogy is that the coke-pressure below the bubble is forcing it upwards with enough force to displace the coke above it.
Here is a site with a short description (and a cool animation) of cylindrical objects shedding vortices in their wakes. You can probably imagine just from looking at the animation that there might be oscillating lateral forces on the cylinder, which would cause it to move side-to-side if not constrained. In fact the main thrust of that webpage is “flow induced vibration” of things like cables held in a breeze.
This (cylinders free to move laterally) is a more interesting and complicated physical problem than it might first appear, with many different “modes” of vortex shedding possible depending on how various parameters interact (cylinder diameter, freestream speed, fluid density and viscosity) - the seminal work was done by Williamson & Roshko sometime in the 1980’s.
This is related to the flow over the gas bubble. The obvious difference is that the bubble is three-dimensional, while a cylinder is (effectively) two-dimensional. The flows are not identical, but the general idea is the same.
Mirage, I believe you’re pretty much spot on about why the bubble doesn’t break down. If the bubble is big enough, the hydrodynamic forces from the Coke prove too much even for surface tension, and it falls apart (notice how huge air bubbles in water tend to break up?). However, if the bubbles are small enough the relative balance of all the forces changes and they retain their integrity.
Generally speaking, they do still deform from a spherical shape due to the forces involved. Which changes the flow. Which changes the forces. Which… This is a cheap type of what’s known as “aeroelasticity.”