Fig. I.—Section through cylinder of Knight sleeve-valve engine.
depends upon two factors, the mean effective pressure upon the piston head and the linear speed of the piston. It was not possible to increase greatly the mean effective pressure; in fact, changes in the character of the fuel used for motor-cars, by which some very much less volatile fractions were included than were found in motor fuel of the earlier period, made it necessary to operate with lower compression, which resulted in a lower mean effective pressure. Improvements in combustion chamber design and other changes more than balanced this loss however. A great gain was made by increasing the speed of operation. In 1905 the average engine speed corresponding to maximum engine output was about 1,000 ft. p.m.; in 1920 it was above 2,000 ft. p.m. for passenger-car engines and over 3,000 ft. p.m. for -racing engines. The first step in endeavouring to increase engine speed was the enlargement of valve ports and passages, to enable the engine to draw in a normal amount of charge at a higher speed. The valve timing was also changed, the exhaust valve being given a greater lead and the inlet valve a greater lag. Next, the reciprocating parts (piston and connecting-rod) were lightened, so as to reduce the inertia forces on them and the bearing pressures resulting therefrom. This led eventually to the adoption of aluminium alloy pistons. One difficulty with aluminium pistons is that owing to the fact that aluminium has a higher coefficient of heat expansion than cast iron, the aluminium piston must be given a greater clearance in the cylinder, which tends to result in unpleasant piston slap when the engine is cold, and also in "oil pumping," that is, transfer of lubricating oil from the crankchamber to the combustion chamber, with consequent smokiness of the exhaust. For this reason some makers who had adopted aluminium pistons gave them up. Special means were resorted to in attempts to build ultra high-speed engines, as for racing and similar purposes. These included the use of two inlet and two exhaust valves per cylinder, the use of two simultaneous ignition sparks in each cylinder, and the use of crankshafts in which each individual throw was counterbalanced.
It was recognized early in the history of motor-car development that by increasing the number of cylinders above the one or two employed in the first machines, not only could the engines be reduced in weight, but the objectionable vibration could be minimized. In 1920 the four-cylinder engine was the foremost type for use on vehicles of a strictly utilitarian character. The six-cylinder engine was, however, widely used for the larger and more powerful types of passenger car, particularly in the United States, and the eight- and twelve-cylinder engines also had a run of popularity. In a petrol engine the torque impressed upon the crankshaft is always nonuniform, no matter how many cylinders there are, but the fluctuations decrease with an increase in the number of cylinders. In a four-cylinder engine there is, as with one or two cylinders, a reversal of the torque; that is, just before the end of each stroke the flywheel not only supplies all the power delivered by the engine, but also some of the power necessary for keeping the crankshaft and pistons in motion. Six cylinders are the smallest number delivering continuous torque at the crankshaft; but while continuous the torque is still far from being uniform; with eight cylinders the torque fluctuations are reduced, and with twelve they are still smaller. Four- and six-cylinder engines were always arranged vertically, with all cylinders in a row; eight-cylinder engines were generally of the V type with an angle of 90º between cylinders. The Lincoln eight-cylinder car had a 60º V engine, and several all-in-a-row eights were built; twelve-cylinder motor-car engines were always built in V form, with a 60º angle. In deciding upon the form of the crankshaft of a multi-cylinder engine and the angle between cylinders in a V engine two objects are aimed at, namely, to ensure uniform spacing of explosions and inherent balance of reciprocating parts. Both objects can be attained in six- and eight-cylinder vertical engines and in twelve-cylinder V engines; in a four-cylinder vertical engine there is an unbalanced reciprocating force in a vertical plane, causing vibration of the engine. In an eight-cylinder 90º V engine there is an unbalanced reciprocating force in a horizontal plane.
When multi-cylinder engines were first used the cylinders were generally either cast separately or in pairs; later it became the practice to cast all cylinders in one row in one block. This greatly simplified the outward form of the engine, as with such a cylinder block only one pipe connexion each need be made for the cooling-water inlet, the cooling- water outlet, the combustible charge from the carburetter and the exhaust. Some manufacturers even cast the top part of the crankcase integral with the cylinder block and made the lower part a steel pressing. This construction lent itself well to quantity production. Most makers of the higher-priced cars, produced in smaller numbers, cast all parts of the crankcase of aluminium. In American practice the cylinder heads were generally cast separate from the cylinder block and fitted to the block with a gasket of sheet copper and asbestos between. This construction facilitated manufacturing operations, and when the engine was in service permitted decarbonizing the combustion chamber by scraping without removing the cylinder. It also made it possible to machine completely the combustion chambers, and thus to get all the chambers in one engine of exactly equal volume. European engineers up to 1920 adhered largely to the integral cylinder head.
One thing that caused both manufacturers and users of motorcars a great deal of trouble between 1910 and 1920 was the continual change in the volatility of the fuel used. When motor-cars were first used the fuel sold consisted of a comparatively narrow range of highly volatile hydro-carbons. When sprayed into air at atmospheric temperature in the required proportion of about one part by weight of petrol to 15 parts of air it vaporized readily. The fuel supplied in the United States in 1920 had an end point of close to 500º F., that is, the least volatile constituents, when under atmospheric pressure, boiled only at that temperature. Hence, in order to vaporize this fuel completely it was necessary to supply heat to the mixture or to the components before they were mixed. When trouble from incomplete vaporization was first experienced the carburetters were provided with a jacket through which hot water from the engine jacket was circulated. When this no longer sufficed the air for the carburetter was drawn through a muff surrounding a part of the exhaust manifold, and to prevent recondensation after the mixture was formed the inlet manifold was so arranged that it was completely surrounded by hot water. Still later this also proved inadequate, and then what is known as the hot spot or exhaustheated manifold was introduced. When the fuel is incompletely vaporized the liquid particles tend to separate out of the mixture at the bends in the manifold, and it is very difficult to insure that all cylinders get mixture of the same composition. Those portions of the manifold wall where the liquid particles tend to accumulate are then made to form parts of the exhaust manifold wall also, so that they are constantly kept at a high temperature, and the liquid particles upon striking them flash into vapour. The change in the character of motor fuel between 1916 and 1919 is strikingly illustrated by the diagram (fig. 2) of distillation curves of fuels purchased in Detroit, Mich., at various times during 1916-19.
One difficulty encountered in the use of exhaust heat for vaporizing the fuel is that the heat supplied does not vary in accordance with the needs when the load on the engine is varied. When the engine is heavily throttled and runs under light load at low speed, the suction on the spray nozzle is small, and consequently the fuel is not finely sprayed. Relatively more heat is needed to ensure the vaporization of the larger globules of fuel, but under these conditions of operation the exhaust does not supply a great amount of heat. A device designed to overcome this difficulty was developed by the Packard Motor Car Co., and is known as the Fuelizer. With this (see fig. 3) a variable fraction of the mixture prepared in the carburetter mixingchamber is shunted around the throttle valve and through a heating jacket of the carburetter, where it is kept burning by a constant stream of sparks delivered by a sparking plug. The products of combustion are combined with the main stream of combustible charge and pass on into the cylinder. When the throttle valve is fully open
