ABSTRACT
The microscopic atomistic understanding of dissipation at sliding material interfaces is a necessary ingredient in constructing physical models of the frictional interaction between materials at larger length scales. From an engineering point of view, one requires models of the interfacial stresses as functions of the relevant material variables for use in finiteelement and finite-difference macroscopic computer codes. Relevant variables depend on materials, environment, and velocity regimes. Thus, when values of frictional coefficients are quoted for different material pairs, reproducibility requires, in most cases, specification of atmosphere, humidity, surface roughness, chemical composition, and microstructure at and away from the interface. Because a steady sliding state is ultimately achieved, the evolution of these variables toward the steady state is required as well. The character of the steady state depends on several length and time scales. Surface roughness in the form of the asperity distribution function clearly plays a role. The effects of microstructure, dislocation distributions, and their relationship to plastic deformation are important for both ductile metals and brittle materials. Defect production and mobility and crack propagation under strong shear loading become relevant. Surface chemistry (electronic states) is important when mechanical deformation plays a subsidiary role. Then, chemical kinetics, electron-electron, and electron-phonon interactions, as well as surface charge distributions are necessary elements to the description of dissipation. For lubricated interfaces, the rheology of complex fluids or hydrocarbon layers determines the local interfacial motion. If one crudely defines macro-, meso-, and microscale physics by length scales of mm, µm, and nm, and time scales of msec, µsec, and nsec, then the
determination of the interfacial frictional force, measured at a distance from the interface, becomes quintessentially a multiscale problem in materials science.
