Here are the two best things about fluid dynamics: scaling laws and cool videos. A group of researchers has been studying the behavior of water on a hot plate. You are probably all familiar with the effect: you drop some water on a hot plate, the water hisses a bit, and the drop races crazily over the surface. This is called the Leidenfrost effect.
(Its name has nothing to do with frost; rather, the person who first described it was named Johann Gottlob Leidenfrost.)
The drop’s motion was assumed to come predominantly from the shape of the hot plate’s surface. The plate is almost never completely level and typically not very flat, so the drops just follow the local slope. A trace of the drop, presumably, would map out the local contours of the plate. However, this turns out to be wrong. Droplets of water on a hot plate are self-propelling and choose their own destination.
Fast moving water droplets
Leidenfrost explained the water’s behavior in terms of a water drop that flattens but sits on top of a vapor layer rather than contacting the hot plate. The vapor comes from the bottom of the water droplet and escapes from the sides, so the drop becomes smaller and smaller (and crazier and crazier) as time goes on.
The very mobility of the droplets makes them hard to study. Researchers have generally taken two approaches to solve this problem: pin the drops in place using a needle or use a bowl-shaped plate so that the droplets stay in one location.
In the new work, the researchers took a two-camera approach. A side-view camera was used to observe the drop as it was deposited on a level hot plate. Once released, the drop could majestically sail away (video) and become an out-of-focus blur. The side-view camera provided a close-up view of the initial motion.
A top-view camera provided tracking information (video). This allows the researchers to observe the direction, speed, and acceleration of the drop as a whole. The camera shows enough detail to see some features of the drop, but things like internal flows are not visible.
I’m free, free falling
The researchers discovered that small drops just took off in any direction, while large drops consistently traveled toward the same location. Neither small drops nor large drops seemed to change direction very often—once the drop had decided where to go, it went straight there.
The large drops seem to be driven directly by gravity. The hot plate was leveled to the accuracy of the researchers’ spirit level (a precision spirit level, not a household spirit level). The acceleration of the large drops fit that accuracy pretty well. If the hot plate was tilted just slightly less than the amount that the researchers could measure, then the drops would have a consistent acceleration in one direction.
The researchers also found a middle range, where drops follow curved trajectories.
I’m free, free wheeling
That leaves the smaller drops: why don’t they also head in the direction of the slope? Or, why don’t they at least curve in response to the slope?
The researchers’ measurements showed that the small droplet’s acceleration was quite a bit larger than that due to gravity. Any curvature of the path due to gravity is so slight that it is simply not possible to measure it. This tells us that the movement of the drop is dominated by the activity of the water, rather than its environment.
The side-view camera also had an interesting story to tell. It showed that the drop appears to be spinning quite rapidly (video). Critically, the drop typically takes off in the direction of the spin. This doesn’t make a lot of sense when you consider that the drop is basically spinning on a frictionless surface created by water vapor—the spin can’t get any traction to move the drop. So, how is the acceleration generated?
Returning to the top-view camera, the researchers observed that the vapor layer is not evenly distributed (video)—this is visible through the interference fringes in the images, which change shape as the drop moves. Instead, the drops have a tilt in the direction of motion. Observation and calculation showed that the internal motion of the drop causes the tilt and that the tilt is sufficient to explain the observed acceleration.
The potted summary is that the drops are not static, but they have internal flows. The internal flow disrupts the symmetry of the vapor cushion, leading to the drop accelerating off in some direction all on its own. The drop does not need any external driving force to set it in motion; instead, it is self propelling.
And, most importantly, the drop supplies the best movies (there are many more in the supplementary information at the link below, which I think are accessible to everyone).