340 picked up from this chamber by means of the injection air. With a light. fluid fuel this difficulty is not noticed. As the oil becomes more heavy and sluggish this difficulty increases. The sheet referred to ex- plains the situation. In the first two figures the heavy black line shows the duration of the delivery stroke of the fuel pump. This would be for full stroke or the full-power condi- tion. top shows the duration of the open- ing of the fuel valve into the cylin- der. From this it becomes at once evident. that the fuel is delivered to the fuel valve and there comes to rest. From this position of rest it must be again put in motion and torn apart by the action of the injection air. There is a certain inertia to the oil and this inertia is greatly augmented in the case where the fuel has a high viscosity. This type of fuel valve is that proposed by Geo. B. Brayton and patented by him. In these figures the time of a, complete cycle of the engine is shown by the full circle of 360 degrees. The vari- ous events of the cycle are shown in the figures. the method which is proposed by the author. As will be seen in the consideration of the fuel pump, the two strokes of the pump are of short duration and the rest of the time the pump is _ idle. this method of fuel injection is to co- ordinate the operation of the fuel valve and the fuel pump so that the fuel valve will open, starting a stream of injection air into the cylinder, and at the same time the discharge from the fuel pump will take place, deliv- ering the fuel charge directly into the stream of moving air where it will be broken up and delivered into the eylinder in the form: of .a_ suitable spray. By this method one of the obstacles to heavy fuel will be over- come. From the time the pump. be- gins to deliver its charge the fuel will not be allowed to come to rest. In this way the fuel valve can be simplified in its design and the ques- tion of a delicate proportioning of parts to suit a special type of fuel will be done away with. It will be possible with a little experience to vary the pressure of the injection air, giving an additional saving. This, how- ever, will be a matter of adjustment possible while the engine is in op- eration. Even if this be ignored, the gain through the use of heavy fuel, not practical in the present types of fuel-feeding arrangements, will be made without any counteracting loss. The fuel valve of this arrangement consists simply of two passages, each controled by a separate needle valve. The short, light space at the . The next figure shows. The theory of THE MARINE REVIEW These valves are worked from one tappet rod and the air valve must be opened first, with the opening of the fuel valve being timed to the de- livery of the fuel by the pump. In this way a stream of air will be started and the oil delivered into this air, and then the air will continue to flow for a short time after the deliv- ery of the fuel in order to clean. the passage to the cylinder so that the valve may not be clogged. The fourth figure is an enlarge- ment of the fuel and air delivery events. The duration of the opening of the valves and of the pump strokes can be controled as well as the tim- ing of those events. At first thought this seems to be quite a complication but on closer observation this com- plication is shown to be nothing but nv So S an So S 4 S oO /dea/ a S So & S S § N << S - s g 8 aS ec August, 1918 to receive consideration. The reason is that the time under the Starting cycle is very short, and a matter of efficiency is not worth considering, By efficiency is meant the thermal efficiency. It is a matter of efficiency whether the engine will start or not. It becomes a matter of efficiency if jt be shown how the engine can be run for an appreciable time on the en- ergy which can be stored-in a com- pact form available for instantaneous use. This is a matter of practical . utility rather than of thermal eff- ciency. That the. matter of thermal efficiency \does come into this consid- eration is only incidental. The steam engine starts with an expansion stroke and has steam ayvail- able for this stroke, so the starting feature is. not a factor, In. the in- ara Nal Frtarged /93_Diese! Card of Low Fressure CT. - 40 50 Perceriiage Volur7e FIG. 7--IDEAL CARD FOR LOW PRESSURE TYPE ENGINE what exists in every automobile en- gine. Certainly, if it is advisable to vary the timing of the spark on an automobile engine for the sake of ~speed control, it is not out of place to contemplate a similar modification in an expensive plant where the effi- ciency gain will be a great factor. Again we have several elements here. It would be impractical to have the elements all under individual control. It will be possible, however, to have a connection between the fuel-pump control and the fuel-valve control so that one lever will regulate all. This would be similar in idea to the present method of changing the cut- off on all cylinders of a compound steam engine from one lever, and at the same time with a provision for varying the cut-off of the separate cylinders at will. This is the first thing to be done with the engine and has been the last ternal-combustion engine the starting: problem is a most difficult one. Not alone does the engine have to be ac- celerated but the acceleration has to impart sufficient energy to overcome inertia, do the work of the engine and at the same time produce the compression stroke. Up to the present compressed air has been universally used for this purpose. The oil engine is still in its swaddling clothes and many devel- opments must be worked out before the engine can take its place in the world's work as a tried and reliable element. For the generation of heat on shipboard the use of steam is practically essential. Then it is rec- ognized that air starting is not abso- lutely reliable, for air must be gen- erated. This may be done by a small oil engine. The auxiliary plant in turn must be started by air. This con- dition has brought about the use of 4