## Abstract

The friction coefficient $\mu$ and the mass loss M' per centimetre, of Al, Ag, Cu, Pt, Fe, Mo, U and W specimens $\sim$ 1 to 2$\cdot$5 cm long are investigated as a function of the mean particle diameter D of abrading emery-paper surfaces (dry) on which they slide at a few cm/s and $\sim$ 1 Kg load. On any grade of abrasive, $\mu$ and M' vary with distance of sliding, M' decreasing and $\mu$ increasing for initially wet-abraded metals, due to progressive pick-up of emery by the metal surface. After 200 to 500 cm, depending on the metal and the grade of abrasive, they become constant (M' then is wear of metal only, M) and independent of the initial surface preparation. $\mu$ does not increase linearly with D as suggested by previous authors (who used small specimens, smaller range of D, and neglected the effect of pick-up of abrasive). For any given metal, $\mu$ (initially and at equilibrium) is practically constant, $\mu_{\max.}$, at D from 70 to > 150 microns, and decreases with increasing hardness of the metal, at equilibrium being 0$\cdot$88 for Al, to 0$\cdot$53 for W. At D < 70, $\mu$ decreases-in general to $\sim$ 0$\cdot$3 to 0$\cdot$4 for specimens 1 to 2$\cdot$5 cm diam.-due to partial clogging of the abrasive by worn-off metal. The high and constant $\mu_{\max.}$ on emery- and glass-papers is due to the angular shape of the abrasive particles, since $\mu$ is only 0$\cdot$15 for Ag and Cu on an array of glass spheres of D $\sim$ 125 microns. For the metals initially abraded on wet emery, then slid on dry emery-papers, M at equilibrium also increases with D and is practically constant at D > 70 microns; and M $\bumpeq$ kW($\mu - \mu_0$), where W is the load, and $\mu_0$ is a constant, in general > $\mu_{mo.mo}$ of metal-oxide on metal-oxide because of pick-up of abrasive particles by the metal. 1/(M/$\rho$), the reciprocal of the volume wear per cm, is approximately proportional to the Vickers hardness H$_D$ of the abraded metal surface. Most of the indenting particles cause wear in the form of strips of metal analogous to those caused by machine tools, but there is also much plastic flow (and lateral pile-up) as electron diffraction also indicates. A theory defining friction and wear of metals in abrasion in terms of the shape and orientation of the abrasive particles, is given elsewhere, together with quantitative analysis of the present data. For emery-papers, only about 10% of the groove volume (on the average) appears to be removed as wear.