How does water reduce friction?

Ice skating: why is ice so slippery?

There appears to have been little or no curiosity about the physical meaning of the unique property of ice. When I think of myself it's easy to explain: ice was smooth when I was born, I didn't know it any other way, and in short, it was smooth because it was ice.

- Reynolds (1898)

Gliding along the narrow blades of an ice skate is made possible by the extremely low friction of solids on ice. The question of the cause of this low friction was put on the agenda of the scientific world by Reynolds in 1898. Motivated by his earlier fundamental work on the effects of oil in gears, Reynolds suggested that a thin layer of water between the ice and the runners is responsible for the low friction.

Where does this layer of water come from that enables sliding?

Pressure doesn't help much

Ice skater

Before Reynolds, Joly had already suspected a thin layer of water between the ice and the runners in 1886: he blamed the melting of the ice due to the high pressure of the runners on the ice for its existence. This explanation has so much charm that it haunted the textbooks for a long time - but unfortunately it cannot be correct on closer inspection; at least not without benevolent support from other more substantial causes.

Of course, it is true that a local increase in pressure on the ice leads to a local decrease in the melting temperature of the ice. In quantitative terms, this pressure-induced lowering of the melting point only leads to a very poor result: a skater who weighs around 70 kg and glides on skates with a blade length of 30 cm and a blade width of 0.5 mm exerts a pressure of about 23 atmospheres on the ice - about as much as the pressure of a fully packed moving truck on the street. This very considerable pressure lowers the melting point of the ice by just a fifth of a degree! Even at ice temperatures just a few degrees below freezing point, no water film would form.

Another argument against pressure-induced melting: This would result in more and more water under the runners when standing on ice skates, and the skater would sink into the ice. Obviously, this contradicts our experience.

Friction does a lot more

Studies by Bowden and Hughes from 1939, by Evans and colleagues from 1976 and von Colbeck from 1995 point to a completely different, much more effective mechanism as the cause of the formation of the water film: the frictional heat generated when the skate blades move across the ice results in a significant melting of the ice on the surface and thus creates the thin film of water that can be observed experimentally.

This mechanism is supported by the fact that it explains the dramatic difference between the frictional force that is required to set an ice skate in motion from a standstill (static friction) and the friction that acts on a skate that has been moved (sliding friction). The sliding friction on ice is reduced to a hundredth of the static friction as soon as you slide. The difference is based on the fact that the liquid film is only generated by the heat generated during sliding. In contrast, the ice skates of a stationary ice skater experience static friction that is practically the same as on other smooth solid surfaces. In contrast to this, the tiny film of water induced by the contact pressure is of the same thickness when standing and sliding, so that there is no significant difference between the strength of static and sliding friction. The sliding friction on ice is reduced to a hundredth of the static friction as soon as you slide.

Beyond the qualitative insights, this explanation allows quantitative predictions that agree well with the observations. With the amount of heat from the sliding friction of the 70 kg skater from the above example, up to 12 mm can be achieved, depending on the ice temperature3 The ice melts, which would result in a water film with a maximum thickness of 40 micrometers. In fact, the film is then thinner because the water is pushed out to the side by the weight of the skater on it and because not all of the frictional heat generated can be used to melt the ice. Some of the heat is dissipated upwards through the runners and downwards through the ice.

However, the heat loss through dissipation decreases with increasing speed, since the heat can only flow away through a piece of ice while the ice skate is running over it. At lower speeds, the crossing time is longer, so a larger proportion of the heat diffuses into the depths of the ice. When running fast, less heat dissipates and is used almost entirely to melt ice. As a result, the thickness of the water film increases with increasing speed and the friction decreases sharply, as Evans and his colleagues observed experimentally in 1976.

With even faster gliding, however, the friction then increases again because, on the one hand, the water film no longer changes significantly and, on the other hand, according to Stokes' law, the frictional force increases linearly with the speed. So there is a speed range of minimal sliding friction. With a suitable choice of runner material, runner size and shape, for a given ice temperature and a known weight of the skater, this range of minimal sliding friction can be achieved at speeds of a few meters per second, i.e. at the speeds relevant for ice skating.

Ski and toboggan good!

The exceptionally low sliding friction of ice is caused by a film of water on the surface. Pressure-induced melting and "surface melting" (see box below) can support the formation of this water film, but they are not dominant or in themselves sufficient contributions. Rather, the decisive factor is the frictional heat generated during sliding, which melts the ice.

The considerations summarized in this article are directly applicable to skiing and sledding. As the contact area increases, the proportions of frictional heat and pressure-induced melting in the formation of the water film decrease compared to surface melting, but the frictional heat is still clearly dominant even for the contact area of ​​a ski.