Donald G. Flom

Theory of Rolling Friction for Spheres

A theory of rolling friction featuring the importance of elastic hysteresis losses is presented. A simple model of retarded elasticity is chosen to represent the physical properties of the material. A prediction resulting from the theory is that the coefficient of friction for a relatively hard sphere rolling on a softer base material should vary with speed so as to go through a maximum. This relationship resembles closely the variation of mechanical loss with frequency.The results are not restricted to rolling but also apply to well‐lubricated sliding where shearing forces have been minimized. Although the theory is developed for a material with idealized physical properties, it nevertheless affords a basis for comparing real materials and for predicting their frictional properties in cases where deformation losses are predominant.

Friction of Teflon Sliding on Teflon

The low coefficient of friction for Teflon sliding on Teflon, as observed previously by others, has been found in this study to hold only if low sliding speeds and newly prepared surfaces are used. At high sliding speed, the nature of the surface is irreversibly changed such that subsequent sliding at low speed reveals a two‐ to threefold increase in the friction coefficient. In addition to the effect of high speed sliding, there are reversible changes in the friction of Teflon on Teflon dependent in part on temperature. As the temperature of Teflon is increased from below room temperature, a sharp and pronounced increase in the coefficient of friction is observed in the vicinity of 20°C. The existence of a phase transition in Teflon at this temperature indicates a close correlation between the frictional and the structural properties of this material.

Rolling Friction of Polymeric Materials. I. Elastomers

Coefficients of friction for the rolling of steel balls on butyl, silicone and Neoprene elastomers have been measured in the temperature range of 25 to 100°C. For equivalent amounts of deformation rolling friction is directly proportional to dynamic mechanical losses measured by a rebound method. In addition, the coefficients of friction vary directly with (load)⅓ and inversely with (ball radius)⅔. These results are in agreement with recent theoretical predictions.The high mechanical losses and rolling friction for the butyl elastomer at 25°C drop sharply on increasing the temperature to 100°C. For the silicone and the Neoprene the losses and the friction decrease only slightly with increase in temperature.

Rolling Friction of Polymeric Materials. II. Thermoplastics

The results of rolling friction studies of several thermoplastics provide additional evidence for the correlation of such friction with dynamic mechanical losses in polymeric materials. Among the polymers discussed are polymethyl methacrylate, polytetrafluoroethylene, polyethylene, nylon, polyvinyl chloride, polyvinyl acetate, and polystyrene. The effects of spin and other deviations from pure rolling are demonstrated for polymethyl methacrylate, polytetrafluoroethylene, and nylon by varing the experimental parameters. In essentially pure rolling, the dependence of friction on temperature illustrates the importance of the extent of crystallinity in polytetrafluoroethylene, the amount of branching in polyethylene, and the concentration of plasticizer in polyvinyl chloride. In the latter effect, it is found that the temperature at which the rolling friction goes through a maximum varies linearly with plasticizer content.

Metal Transfer in Sliding Contacts

Transfer of radioactive silver has been studied quantitatively in sliding contacts containing graphite and metal. The rate of transfer from silver‐graphite riders to graphite cylinders is initially high but soon approaches a limiting equilibrium state. The amount of silver transfer increases markedly with surface roughness. Further, the transfer is enhanced by the addition of liquid water to the rider track. Back transfer to the rider has been found to consist mostly of loosely clinging dust in the rider face.Silver transfer to a copper cylinder is much greater than to a graphite cylinder, the difference being roughly twenty‐fold. Within the experimental conditions, the transfer/unit time to both graphite and copper cylinders is greater at 1 cm/sec than at 390 cm/sec.