Physicists in the US have developed a new imaging technique that has allowed them to observe the splashing process in great detail. They used the method to look for an air gap that was believed to form when a liquid drop strikes a solid surface, causing tiny droplets to splash out. While they did see such a gap form, they were able to conclude that it is not responsible for splashing – at least in the water–glycerol drops studied.

For more than a century, physicists have used high-speed photography to capture often-beautiful images of the splashes that occur when a liquid drop strikes a solid surface and produces a ring of smaller droplets. However, this beauty belies the complex physics underlying splashing and successive generations of physicists have struggled to understand the process – which is important in practical pursuits as diverse as spray-painting and pesticide application. In 2005 Sidney Nagel and colleagues at the University of Chicago added to the mystery by discovering that a reduction in the ambient air pressure reduces the amount of splashing. This seemed counterintuitive because it had been thought that greater air pressures would tend to hold the drop together, while lower pressures would allow it to break up into a splash.

One proposal put forth by Michael Brenner and colleagues at Harvard University is that air gets trapped under the drop when it nears the surface – and it is this thin layer of air that causes the splash. The lower the pressure, the less trapped air and therefore the splash should be smaller. Others, including Nagel, believe that splashing is driven by interactions between liquid and air at the edge of the drop as it flattens and spreads out on the surface. Now, however, Nagel and Michelle Driscoll have come up with a new technique to monitor how much air is trapped under a drop. Their measurements suggest that trapped air has nothing to do with making the splash, at least for the liquids they studied.

Mind the gap

Their technique is based on an established method for measuring the thickness of very thin films. Monochromatic light from an LED is fired at the drop as it flattens on the surface. Some of this light reflects from the surface and some from the interface between the liquid and the trapped air. The result is an interference pattern from which the thickness of the air gap can be deduced. The challenge for Driscoll and Nagel was to capture this pattern in real time – with particular emphasis on the first few hundred microseconds, which proved to be crucial to understanding the role of trapped air. To do this, they used a high-speed camera that was able to capture the diffraction pattern at 67,000 frames per second. Nagel and Driscoll also had to make sure that they measured the thickness of the air gap and not the thickness of the liquid – which they did by dying the liquid black.

The liquid used by the pair was a mixture of glycerol and water that was chosen because its higher viscosity means that splashing occurs a relatively long time after impact. This makes the process easier to study than the splashing of lower-viscosity fluids such as water. The measurements revealed that a bubble is present under the fluid after about 50 μs after impact. "The behaviour of this air bubble is, to the extent that we are able to investigate in our experiments, consistent with what was predicted in the theory and simulations of Michael Brenner and his collaborators," Nagel says.

However, as the liquid spreads out, the bubble appears to flatten out slightly, but nowhere near as rapidly as the liquid itself. After about 600 μs a thin sheet of liquid lifts off from the edge of the expanding liquid. It is this sheet that physicists believe will eventually break up to form a splash. However, after 2 ms this thin sheet is still expanding but there is no sign of any splash droplets forming.

Indeed, the main finding of the experiment is that the bubble is completely formed by about 150 μs after impact, whereas the splashing occurs much later and far away from the bubble. As a result, Driscoll and Nagel believe that the splashing is likely to be caused by interactions between the uplifting sheet at the edge of the flattening drop – not by air trapped under the drop.

Brenner told that the experiments do not rule out the air-layer theory completely because the viscosity of the fluids used are much higher than the fluids considered in the theory. "The air layer was predicted for drops with low viscosity, such as water," he explained. "It is abundantly clear that when the viscosity increases too much, the assumptions of the calculations break down – and then we have no idea what should happen."

So it seems that Nagel and Brenner have more experimental and theoretical work to do, respectively, before the mystery of the splashing drops is solved.

These latest results are described in Phys. Rev. Lett.107 154502.