Fire extinguishers, such as sprinkler systems, typically deliver a fine, pressurized spray of water droplets. Some of the smaller water droplets from a sprinkler may be deflected by the plume of the fire and so do little to extinguish it directly. But hot surfaces around the fire can be cooled by the evaporation of droplets that land on them, and the resulting vapour may also help to stifle the flames.

Because of environmental concerns, the widely used fire-quenching agent Halon 1301 was banned by international treaty in 1987. An alternative is to use water with some additive, such as sodium acetate trihydrate, that enhances its fire-quenching ability. How does that affect droplet behaviour?

Samuel L. Manzello and Jiann C. Yang (Proc. R. Soc. Lond. A 458, 2417–2444; 2002) have investigated the collision dynamics of pure water droplets and droplets containing 30% (by mass) sodium acetate trihydrate as they impact on a hot metal surface. Whether each droplet spreads, splashes or rebounds depends on its impact energy as well as on the temperature and roughness of the surface.

Manzello and Yang recorded the collisions of 2.7-mm-diameter drops on a stainless-steel surface for a range of surface temperatures and impact energies — the latter quantified by the 'Weber number', related to the liquid's density and surface tension, and the velocity and diameter of the droplets. The images shown here were taken for Weber number = 181, at times (rows top to bottom) of 0, 1, 2 and 7 ms after collision and at surface temperatures (columns left to right) of 20, 104, 230 and 340 °C.

Compared with the behaviour of droplets of pure, distilled water, Manzello and Yang found that the presence of sodium acetate trihydrate made quite a difference to the collision dynamics at low impact (low Weber number), but less so for high-velocity impacts. In most fire-extinguishing systems the droplets are delivered from a pressure nozzle at high velocity, so the authors conclude that knowing how pure water droplets behave is a reasonable approximation to the behaviour of water with an additive.

They also note that, for high-impact collisions, a droplet forms a liquid film of greater diameter than in low-impact collisions, which increases its capacity for surface cooling. Although the relation of liquid-film diameter to surface temperature was clear for droplets of pure water, Manzello and Yang struggled to find such a correlation for water droplets containing the additive. More work, both theoretical and experimental, is needed here.