The paper represents a study of analyses obtained from 656 samples of contaminated crankcase-oil and states the results of cooperative research, the sponsors being the Society, the American Petroleum Institute, the National Automobile Chamber of Commerce, and the Bureau of Standards. Reliable information was sought regarding existent conditions throughout the Country and, since analyses of a large number of samples were a requisite, arrangements were made with service stations located at points representative of the Country's atmospheric and geographical conditions for the collection of samples of contaminated crankcase-oil, a uniform procedure calculated to assure accuracy of the results being enjoined. Each participating service station was requested to select 10 cars, all of the same make, to drain the oil and to refill with new oil. After the cars so prepared had run a distance equal to that commonly allowed for one filling of oil, the oil was to be withdrawn into a clean container immediately after stopping the engine. From this well-mixed crankcase-draining an 8-oz. sample was to be sent in a glass container to the Bureau of Standards for analysis. The trade names and the grade of new oil used were to be recorded in all cases.
Some of the influential factors that could not be controlled in a survey of this kind were the make of engine, the age and the condition of the engine, the service that a car was rendering, the care and attention that was given to the engine, and the quality of the lubricant. The statement is made that, in general, the number of samples so far analyzed is not large enough to permit the drawing of clearly definite conclusions and that, therefore, definite answers to some of the questions propounded by the Steering Committee cannot be given. But since much valuable information can be secured from the analyses made thus far, the data are presented so that this information may become available.
The method of analysis used by the Bureau of Standards is described and the data obtained from the 1925 and the 1926 surveys are explained. Subjects such as the pathology of the internal-combustion engine, tracing causes of contamination by ingredients found in the ash, and the influence of various factors on contamination are discussed. Other parts of the paper have to do with the grouping of samples according to iron oxide, the relation between silica and iron oxide, the effect of mileage on the silica-iron oxide relation, qualifications of the influence of mileage on contamination, and the effect of viscosity on the accumulation of iron oxide.
After making tabular diagnosis of some of the 1926 samples, the author outlines the effects of temperatures and explains the data showing that the silica content of the 1926 samples indicates little cause for wear. He discusses the effect of a high percentage of sulphur in the fuel, the significance of ash and the effect of air-cleaners and oil-filters. A comparison is made between the results of the 1925 and the 1926 surveys and the differences of dilution for the various groups of samples are analyzed. Factors influencing the accumulation of water are enumerated and the tabular data pertaining thereto are explained. In conclusion, the suggestion is made that the interpretation of the data obtained in this extensive survey can be elaborated, and that such elaboration be undertaken by universities.