Elementary theories regarding the evaporation characteristics of pure substances and mixed liquids are discussed briefly and the difficulties likely to be encountered in attempts to calculate the volatilities of motor fuels from data relating to pure substances or in the extrapolation of volatility data corresponding to the atmospheric boiling-range of the fuel to the range of temperatures encountered in utilization of the fuel are pointed out. A brief review of previous methods of arriving at fuel volatility is also presented.
Volatility, as applied to motor fuels, is defined as being measured by the percentage of a given quantity of the fuel which can be evaporated under equilibrium conditions into a specified volume. The weight of air under known pressures is taken as a convenient measure of the volume.
The new method described is an equilibrium distillation of the fuel in the presence of a known weight of air. The fuel is supplied at a predetermined rate by displacement from a reservoir by the fall of a clock-controlled cylinder, and flows into a long metal helix immersed in a bath at the temperature of the test. Air is also delivered to this helix at a predetermined rate, as measured by a small-orifice meter. Evaporation takes place to equilibrium, and the uncondensed fuel is drained from the lower end of the helix and measured. By such means the distillation curve for the fuel in any desired air-fuel mixture can be determined accurately.
Data are presented for five fuels of varied characteristics which has also been used in engine tests of starting volatility. Such volatility data will be useful in connection with studies of engine performance, carburetion, the blending of fuels, and the production of fuels for specified performance.
The importance of the very low end of the distillation curve for the fuel that is distributed generally throughout the United States is emphasized in the discussion.