Friction as a Dissipative Process

Abstract

This paper consists of two interwoven strands. One strand describes selected experimental studies of friction carried out by the author over the last 55 years and indicates some of the rather more general ideas that emerged from that work.

The second strand deals with the question: how is energy dissipated when friction occurs? It is clear that if friction involves permanent damage to the sliding surfaces, say plastic flow, fracture, tearing or fragmentation the energy expended should be explicable in terms of the strength properties of the materials. If it involves viscous flow or inelastic deformation (hysteresis) there may be little visible damage and one is tempted to treat the energy loss as an example of internal friction (this is another problem). There are, however, some apparently puzzling cases where none of these explanations are plausible. The paper will discuss these and other borderline cases.

As a starting point it takes up the ideas of plastic shear in a crystalline solid where it is known that atoms in the shear plane are displaced from their equilibrium position until they reach an unstable configuration: at this point they flick back to another equilibrium position and slip occurs by a single atomic spacing. The strain energy is lost by vibrations in the lattice and these in tum are degraded into heat. (This also applies to crystals containing dislocations).

These ideas are certainly applicable to frictional processes involving plastic deformation. It is suggested that a similar mechanism applies, in many cases, to elastic or near-elastic conditions, particularly if sliding occurs truly in the contact interface. Thus it would explain the friction observed when oriented fatty acid molecules on a solid substrate slide over a similar surface without producing any permanent damage. It is probable that this also applies to experiments with the Atomic Force Microscope where it has been found that riders sliding over a surface experience a finite frictional force without damaging the surface (like passing a finger nail over the teeth of a comb).

In this paper this idea is applied to the sliding of oxide-coated metals, to polymers and tentatively to ceramics. (The behaviour of rubber is left to Dr Savkoor). The distortion produced by the frictional force and hence the energy dissipated depends on the strength of the bonds between the surfaces and may be expressed in terms of surface forces or in terms of surface energies.

In those cases where the friction is speed and/or temperature dependent the behaviour may be understood in terms of the effect of these variables on the elastic constants of the bodies. It may be that this is equivalent to certain rate-theories developed by Eyring and his school many years ago: but the Eyring approach gives little guidance as to the modes of energy dissipation.

The basic idea developed in this paper is not original: it is to be found in a 1929 paper by Tomlinson though, for some reason, he does not emphasise the important role of atomic vibrations as a means of dissipating energy.