The beneficial effect of oxidation on the lubricating properties of oil

Journal bearing friction experiments have been made generally at relatively low temperatures and otherwise in conditions tending to prevent oxidation of the lubricating oil. Thus Beauchamp Tower’s experiments led Reynolds to the conclusion that fluid friction alone prevails in an oil film maintained by continuous rotation of the journal and that boundary conditions do not become sensible. The more recent experiments by Stanton, undertaken after the Physical Society discussion of 1919, were made to verify the conclusion, and confirmed that especially for mineral oils, “the conditions were in all cases those of perfect lubrication ( i. e ., complete fluid lubrication), no approximation to the hypothetical ones of boundary lubrication being observed,” “the conditions of lubrication of a cylindrical journal being of the Reynolds’ type right up to the seizing pressure. Stanton’s experimental conditions were such that oxidation effects were not obtained. The feed to the journal bearing was always by fresh, not circulated, oil and the temperature of the oil film was maintained at 51·6° C., i. e ., at least 50° lower than required to induce oxidation in a mineral oil particularly susceptible to the effect. The possibility that oxidation might lead to boundary conditions becoming a factor in the measurements was not considered. Oxidation of the oil used to lubricate internalcombustion engines cannot be avoided in the usual conditions of operation, and an investigation of the effect on lubricating value was begun, in connection with experiments made in association with Professor Callendar, on the oxidation of the lighter oils used as engine fuel. The results of lubrication experiments made directly on engines were difficult to interpret. The friction measured is mainly that due to the reciprocating motion of the pistons in the cylinders and oxidation being uncontrolled, the resulting accumulation of semisolid products leads to secondary friction effects greater in magnitude than the primary effect attributable to the fluid alone. The conditions of journal bearing lubrication, on the other hand, can be controlled and friction measured with fair accuracy and it appeared therefore that the investigation could be continued most effectively by using journal bearing testing machines. Machines adapted to be run at the relatively high temperature required for the oxidation of mineral oils had been designed at the N. P. L. by Mr. C. Jakeman in association with whom the experiments were continued, by permission of the authorities concerned.