Last week I posted Rolling, Rolling, Rolling, with the video about the moment-of-inertia problem with the liquid-filled and ice-filled soda cans. Nick submitted the post to reddit (Thanks!), and I got a spike in traffic about 25x normal, for a day or so. That makes it (by far) my most popular post to date, and also the most popular home-grown post (I had gotten a spike or two from some interesting links I had posted).

One of the questions asked at reddit was what would happen if the ice was fee to move inside the can, so I tried that experiment — and there was no noticeable change. Even though the ice was riding on a layer of water it still spun up, and to the resolution of the video, the lag between the ice can and liquid can is the same. Water has pretty strong adhesion to a surface, and the entire interior surface of the can and exterior surface of the ice are interacting, so this shouldn’t be that surprising. Liquids, on the other hand, can be thought of as a very large number of layers (laminar flow), which gives much weaker coupling.

There are one or two other experiments I could run that require a sealable container, and so aren’t convenient with soda cans. I should be able to acquire some transparent sports bottles that should do the trick. If I can think of any comparisons that give non-intuitive results, I’ll post them.