Definite knowledge as to the behavior of gases and liquids in the manifold of an internal-combustion engine being lacking, an attempt is made to answer the questions: (a) How bad is the distribution, (b) how do the different types of manifold compare, (c) why is the liquid distribution in some manifolds poor and (d) how shall we proceed to correct the trouble? The solution of the problem is affected by the facts that, in extremely cold weather, nearly all fuel is delivered to the engine, at the time of starting, as a liquid; that all cars perform poorly under such conditions, some engines, when cold, “hitting” on only one or two cylinders; and that, because of inferior distribution, many multi-cylinder engines are outperformed by single-cylinder engines of similar design.
In earlier days when the fuel was vaporized to a greater extent the distribution of it was easy, but as fuels have continually become heavier and portions of liquid are carried over in increasing amounts, the problem has become one of distributing liquid. Two solutions presented themselves: (a) the fuel might be vaporized before a division of the mixture stream occurred, or (b) means might be provided for dividing the mixture stream properly. It was found that fuels vaporized before distribution are more sensitive to knock, owing to the high temperature necessary to produce vaporization, whereas cold mixtures allow higher compression-pressures and greater economy of operation.
Heretofore, exhaust-gas analysis has been commonly employed to determine accurately the distribution of fuel, but this method requires great skill, and leaks may completely change the result; such analyses repeated a large number of times become very laborious and consequently are not widely used. The method usually employed is to observe the color of the flame given off from the exhaust-port.
By means of a specially constructed dummy engine, from which the crankshaft, the connecting-rods and the pistons had been removed, and having separators connected to the bottom of each cylinder, results extremely close to normal were obtained. A large number of tests could be performed in a short time and the cold fuel tests were interpreted under “hot-engine” conditions.
The three most important causes of improper distribution are said to be (a) the stratification of the liquid with respect to air, (b) puddles of liquid forming in the manifold in dead-air pockets and (c) the effect of the firing order. Stratification is produced by the difference in the weight of the liquid and the air and the tendency of the liquid, after having velocity imparted to it by the air, to continue in the same direction and impinge on the wall of the manifold at the turn. Photographs of the formation of dead-air pockets were obtained by depositing ammonium chloride smoke on black paper. Under conditions of steady or uniformly pulsing air-flow the puddles in the air-pockets remain practically constant in size, but these conditions seldom occur.
The tests made covered many combinations of stratification and puddle formation, a disturbing factor being the variability of the flow of air due to the varying degrees of turbulence set up by uneven surfaces, sharp corners, the position and the shape of the throttle, tortuous bends and other causes difficult to analyze.
The manifolds tested included one of the hot-spot type, a four-cylinder square transparent manifold, an inverted ramshorn-type manifold, another one similar to it made of glass and a six-cylinder manifold.