Methods of Measuring the Antiknock Value of Fuels 270003
THIS paper was prepared as part of the Cooperative Fuel Research program and presents mainly a review of published data on methods of measuring the anti-detonating qualities of motor fuels. Although detonation as a factor in gaseous explosive reactions is not a new subject, the general recognition of anti-knock value as an important quality of automotive engine-fuels is comparatively recent. Reference is made to bibliographies covering earlier work in this field, and an outline is given of work now in progress at various laboratories and universities.
Ricardo's two test-engines and his methods of rating fuels in terms of highest useful compression-ratio and toluene value are described, and the applicability of his results to other engines is discussed. The British Air Ministry Laboratory, using a Ricardo variable-compression engine, compares fuels on the basis of percentage of increase in highest useful compression-ratio.
The method developed in the General Motors Corporation Research Laboratories employs the Midgley bouncing-pin indicator and a Delco engine to rate fuels in terms of a system of reference fuels. This same indicator is used by the Sun Oil Co. to compare the knock intensity of different fuels under like operating-conditions in a given engine, the precautions necessary to assure such conditions being enumerated.
Work at Ohio State University is centered on the value of various knock-suppressers, a proposed method of estimating detonation intensity by measuring gaseous ionization during combustion being described.
At Armour Institute, fuels are tested in an automobile engine and rated in terms of the degrees of spark-advance required to produce certain specified knock-intensities.
A bomb method used at the University of Michigan bases fuel ratings on the quantity of nitrogen that must be added to each gasoline-oxygen mixture to eliminate detonation. Peaks on the pressure-time curves, obtained with an optical indicator, are taken as evidence of detonation. The Delco engine and Midgley bouncing-pin indicator are also used to obtain benzol equivalents of blended gasolines.
The Kansas City Testing-Laboratory, using a single-cylinder stationary engine, rates fuels by two methods, namely, (a) comparing the maximum brake horse-power, and (b) finding the benzol equivalents.
Egloff and Morrell, proposing to compute the aromatic equivalents of gasolines from chemical analyses and from Ricardo's data on the toluene value of pure hydrocarbons, verified the results by comparing them with those obtained from benzol-gasoline blends in an automobile engine, the criterion being the spark-advance giving a definite knock-intensity.
Investigators at the University of Iowa, using a two-cylinder motorcycle engine, rate fuels in terms of the brake mean effective pressure at which knocking begins under identical operating-conditions.
The methods in use at the Bureau of Standards are described, with particular reference to the routine engine-test devised in 1924. In this test, a single-cylinder high-compression engine is used, the throttle being opened for each fuel until a certain minimum intensity of detonation is present. Fuels are compared with a reference fuel under like operating-conditions, the antiknock value being expressed as the percentage of increase in maximum permissible horsepower. Data are presented with regard to a detonation bomb being developed for the Bureau of Aeronautics. This bomb has a diaphragm in the head. A steel ball placed on this diaphragm is thrown upward when detonation occurs, the square root of the height of travel being taken as an indication of the detonation intensity.
The conclusions reached are that (a) nearly all the methods in use consist of engine tests or depend on engine tests for their interpretation; (b) knock intensity is measured in various ways and with differing degrees of definiteness; (c) the antiknock value of a fuel may be expressed in a variety of terms according to the particular method of test; and (d) the rating of fuels by existing methods is usually not independent of test conditions.