Comparison of Antiknock Ratings Determined in Different Laboratories 300019

TESTS conducted about a year ago in the laboratories of three British oil companies to ascertain the antiknock properties of fuels, with the object of obtaining uniformity in expression of antiknock values, are described and the results compared.

Nine gasolines of qualities normally obtainable in the British market were used in the tests and the results are expressed in terms of percentage by volume of equivalent blends of heptane and pure benzene. Each laboratory followed its own standard method of testing and, to ensure uniformity, the supplies of heptane and benzene were bulked and redivided before use.

The report on the results refers to tests with a Delco, an Armstrong-Whitworth and a Ricardo E-35 engine. Test procedures followed in the several laboratories are described and the results obtained are compared in graphs. Average and maximum deviations of the three laboratories from the mean of the values obtained are tabulated, as are also the average differences, in percentage of benzene, for four fuels of highest antiknock value and for all nine fuels. Maximum deviation of the average of results obtained by each laboratory from the mean of results in all three laboratories range from +2.1 to -2.2 per cent benzene. Agreement among individual results is not so satisfactory for fuels of higher antiknock value as for those of lower value, the average differences for four of the former ranging from 0.85 to 3.10 per cent of benzene in the three laboratories, whereas the difference range for all nine fuels tested is from -0.10 to + 1.60 per cent benzene.

The tests show that, within the very narrow limits investigated, very concordant knock ratings for different fuels can be obtained with engines of widely different design and working conditions if temperature and speed are carefully controlled and settings are adjusted for maximum-knock mixture-strength.