The mechanism of sliding on ice and snow

Abstract The preceding paper has shown that the sliding friction bet ween metals decreases in a regular manner as the temperature is raised, provided that the chemical nature of the surface layers remains unchanged. The behaviour of the friction at temperatures near the melting-point was not investigated. Most metals become very soft at these high temperatures, and the irregularities which normally separate the two surfaces now collapse, so that a comparatively large area of contact is developed. Since the frictional force is the resultant of the minute adhesions acting across the real area of contact, we should expect that the friction between two such surfaces would be very great. The friction between two gold surfaces sliding together has been observed to increase to very high values where the metal becomes soft. With solids of different physical properties, however, there is the further possibility that a surface layer of solid may be melted, either by pressure melting or by frictional heating, when the mass of the solid is still frozen and hard. If the melted fluid is present in sufficient quantity and possesses a low viscosity, the conditions may become those of lubricated sliding and lead to a large reduction in the friction. This effect would naturally be most marked for those substances which retain their rigid character at temperatures close to the melting-point, and which give a liquid of low viscosity on melting. Ice is a solid which has these properties, and it is well known that the friction of solids on ice and snow is very low. The suggestion has often been made that, in skating or ski-ing, the surfaces are lubricated by a layer of water formed by pressure melting (e.g. Reynolds 1901, p. 737), but few measurements of any sort have been made either to support or to disprove this suggestion.