The engine was a single cylinder, 4-stroke-cycle "hot-tube" ignition type. The cylinder, of cast iron quite finely and completely finned for its day, was air-cooled, or rather, air-radiated, as there was no forced circulation of air over it, the atmosphere surrounding the engine simply soaking up the dissipated heat. Although this was possibly a desirable adjunct in winter, inside the small shop in Dayton, the temperature there in summer must have been quite high at times. The operating fuel was city illuminating gas, which was also utilized to heat, by means of a burner, the ignition tube. This part was of copper, with one completely closed end positioned directly in the burner flame; the other end was open and connected the interior of the tube to the combustion chamber. The inlet valve was of the usual automatic type while the exhaust valve was mechanically operated. The fuel gas flow was controlled by a separate valve mechanically connected to the inlet valve so that the opening of the inlet valve also opened the gas valve, and gas and air were carried into the cylinder together.
Figure 16.—Shop engine, 1901, showing governor and exhaust valve cam. (Photo courtesy R. V. Kerley.)
The engine was of normal stationary powerplant design, having a heavy base and two heavy flywheels, one on each side of the crank. These were necessary to ensure reasonably uniform rotational speed, as, in addition to having only one cylinder, the governing was of the hit-and-miss type. It had a 6×7-in. bore and stroke and would develop slightly over 3 hp at what was apparently its normal operating speed of 447 rpm, which gives an MEP of 27 psi.
The engine is noteworthy not only for its very successful operation but also because it incorporated two quite ingenious features. One was the speed-governing mechanism. As in the usual hit-and-miss operation, the engine speed was maintained at a constant value, the output then being determined by the number of power strokes necessary to accomplish this. The governor proper was a cylindrical weight free to slide along its axis on a shaft fastened longitudinally to a spoke of one of the flywheels. A spring forced it toward the center of the wheel, while centrifugal force pulled it toward the rim against the spring pressure. After each opening of the valve the exhaust-valve actuating lever was automatically locked in the valve-open position by a spring-loaded pawl, or catch. The lever had attached to it a small side extension, or bar, which, when properly forced, would release the catch and free the actuating lever. This bar was so positioned as to be contacted by the governor weight when the engine speed was of the desired value or lower, thus maintaining regular valve operation; but an excessive speed would move the governor weight toward the rim and the exhaust valve would then be held in the open position during the inlet stroke, so no cylinder charge would be ingested. Since the ignition was not mechanically timed, the firing of the charge was dependent only on the compression of the inlet charge in the cylinder, so it made no difference whether the governor caused the engine to cease firing for an odd or even number of revolutions, even though the engine was operating on a 4-stroke cycle at all times.
Figure 17.—Shop engine, 1901, showing operation of exhaust valve cam. (Pratt & Whitney drawing.)
The exhaust valve operating cam was even more ingenious. To obtain operation on a 4-stroke cycle and still avoid the addition of a half-speed camshaft, a cam traveling at crankshaft speed was made to operate the exhaust valve every other revolution (see Figure [17]). It consisted of a very slim quarter-moon outline fastened to a disc on the crankshaft by a single bearing bolt through its middle which served as the pivot about which it moved. Just enough clearance was provided between the inside of the quarter-moon and the crankshaft to allow the passage of the cam-follower roller. The quarter-moon, statically balanced and free to move about its pivot, basically had two positions. In one the leading edge was touching the shaft (Figure [17b]), so that when the cam came to the cam follower, the follower was forced to go over the top of the cam, thus opening the exhaust valve. When the cam pivot point had passed the roller, the pressure of the exhaust valve spring forced the following edge of the cam into contact with the shaft and this movement, which separated the leading edge of the cam from the shaft, provided sufficient space between it and the shaft for the roller to enter (Figure [17c]). Thus, when the leading edge of the cam next reached the roller, the roller, being held against the crankshaft by the valve spring pressure (Figure [17d]), entered the space between the cam and the shaft and there was no actuation of the valve. In exiting from the space, it raised the trailing edge of the cam, forcing the leading edge against the shaft (Figure [17a]) so that at the next meeting a normal valve opening would take place. The cam was maintained by friction alone in the position in which it was set by the roller, but since the amount of this could be adjusted to any value, it could be easily maintained sufficient to offset the small centrifugal force tending to put the cam in a neutral position.[20]