PREVIOUS papers relating to the cooperative fuel-research investigation of engine-starting have attacked the problem along two distinct lines, namely, (a) by actual tests of the time required to start engines under various conditions, such as temperature, mixture supplied and the like, using different fuels; and (b) by laboratory tests of the volatility of the same fuels under somewhat similar conditions, using the method of equilibrium air-distillation. In this paper the results of these two methods of attack are correlated. For this purpose, it was necessary to extend the air-distillation data to lower temperatures than had been used in the tests. This involved an application of the laws of the perfect gas and the well-established law governing the relation between vapor pressure and temperature.
The results of the engine tests were originally plotted with the rate of fuel supplied, as ordinate, and the time required to start, as abscissa. When the results were plotted in this manner, the time required to start varied with the fuel used, the temperature, the mixture-ratio supplied, the cranking-speed, the type of engine, and the like. By a series of transformations, it was found that, when resultant fuel-air ratios, obtained in equilibrium air-distillations with the same mixture-ratio supplied, and the same temperatures employed in the engine tests were plotted against the number of engine revolutions required to start, all the results could be represented by a single curve. This signifies that the resultant fuel-air ratio alone determines the number of revolutions required to start, the effects of different fuels, different temperatures, and the like appearing only as they affect the resultant fuel-air ratio. The following equation for the curve was found to be representative of all the results:
Since the resultant mixture attained in the cylinders, when the first explosion occurs, is practically always the same (about 20 to 1 air-fuel ratio), the above equation gives a direct relation between the fuel-volatility data and engine-starting performance.
The fuel requirements for engine-starting are defined as follows: (a) possible starting with a 1 to 1 mixture supplied at the lowest temperature ever anticipated and (b) starting in 10 engine-revolutions with a 2 to 1 air-fuel mixture supplied at the average minimum temperature encountered in service. These definitions serve to fix the percentage of the fuel that must be evaporated at each of the temperatures in equilibrium air-distillations, namely, 5 per cent in a 1 to 1 mixture and 15 per cent in a 2 to 1 air-fuel mixture supplied.
It is shown that the ratios of the absolute temperatures corresponding to 5 per cent evaporated in American Society for Testing Materials distillations and in equilibrium air-distillations are practically constant for the nine fuels employed in this investigation. The same result is found in the case of 15 per cent evaporated. Thus, the air-distillation data form the connecting-link that makes it possible to interpret the results of the engine-starting tests directly in terms of the American Society for Testing Materials distillation-curves.
The lowest temperature ever recorded and the average minimum-temperature for any season in any locality are fairly well known from United States Weather Bureau statistics. This information together with the information supplied in this paper furnish a ready means for specifying the fuel to meet the starting requirements in any locality, in summer or in winter.