will be discussed later under the heading of "Lubrication," and withdraws a certain heat from the working parts of the engine, especially from the piston. An indicator card and the theory of the scavenging is given herewith (see Fig. 7).

Efficiency of Scavenging.-Scavenging efficiency can be subdivided under two headings, the efficiency of the pump and the efficiency of the scavenging of the working cylinder. As seen from the comparison of the ideal with the working indicator diagram, Fig. 7, there are several losses in the pump: (1) Suction loss due to attenuation of the charge.

(2) Compression loss due to leakage through the main bearings.

(3) Loss of volumetric efficiency due to the heating and re-expansion of the clearance air.

To take the three points in order; suction loss requires no explanation. Loss due to leakages through the bearings is now reduced to a minimum with careful design of the air rings and good workmanship. Fig. 8 shows an arrangement which has proved satisfactory. The clearance volume should be kept a minimum although with this type of engine, this volume is always large even when the crankshaft is fitted with balance weights to give better balance and to minimize this clearance volume, and where the clearance for the crank and bottom end is cut fine, since the air must have access to the piston crown for cooling purposes. It is probable that the greatest loss of efficiency is not in the scavenging pump but concerns the scavenging within the working cylinder.

The effect of the shape, and the size of the scavenging and exhaust passages and ports has not been fully studied, although it can be stated that wide variations in both are possible without any appreciable effect on the performance in practice of the normal semi-Diesel engine, the main considerations being, as would clearly be inferred from the indicator card, Fig. 7, to get rid of the exhaust at a suitable point as rapidly as possible, and that back pressure of the exhaust must be reduced to an absolute minimum. Back pressure has an effect in preventing the entrance of scavenging air to the working cylinder, previously compressed to a pressure dependent in some measure upon the speed of revolution. The scavenging air pressure cannot increase to overcome exhaust back pressure as with separate valve-controlled scavenging pumps. The chief effect of back pressure, however, is to decrease the quantity of air drawn into the crank chamber to be compressed. The pressure from which the clearance air in the crank chamber must expand is the back pressure of the exhaust, and until this clearance air has expanded down to the suction pressure no fresh air will be drawn into the crank-case. The volumetric efficiency is entirely dependent upon this factor, as is clearly shown in Fig. 7. Large exhaust ports, ample passages, the close proximity of a large silencer to the cylinder and the minimum of restriction in the exhaust pipes are necessities.

The two-cycle semi-Diesel engine has almost settled down to a standard design of scavenging and exhaust passages and piston crown without any proof other than that of satisfactory performance, although still at relatively low efficiency. Much experimental work still remains to be done on this subject. This course involves considerable labor and expense due to the large numbers of variables that have a direct influence, of which but a few will be mentioned. Piston speed must have an effect on the efficiency of both the crank chamber scavenging pump and of the cylinder scavenging. Experience has shown that 300 feet per minute approximately is the maximum piston speed, above which with present designs the scavenging efficiency falls off somewhat rapidly, the power output from the engine fails to increase with higher revolutions and increased fuel, and the limiting condition of maximum powers are reached.

With this question is intimately associated the subject of the stroke-bore ratio; it can be said that the higher the stroke-bore ratio, the better the conditions of cooling of the piston, because the smaller the diameter of the

piston for a given power, and so the shorter the path for the heat to travel from the center to the cooled walls; also the lower number of revolutions for the desired output of power gives certain advantages for driving types of machinery which are inherently slow speed machines. Furthermore, the author's experience suggests the larger the stroke-bore ratio within the limits of ratio of 1 to 1 and 1.5 to 1, the less, probably, the escape of scavenging air through the exhaust ports, due to a greater quantity of fresh air being entrapped in the combustion chamber, i. e., with a square engine an engine of approximately equal stroke and bore-a greater percentage of the scavenging air finds its way out through the exhaust ports. The shape and angle of entrance to the cylinder of the air inlet passages, the type of baffle on the piston crown and the location of the bulb in relation to the path of the scavenging air all have an influence.

Injection. To turn to the question of injection, which depends primarily with a solid injection semi-Diesel engine, on the following factors: (1) Turbulence within the cylinder.

(2) Pressure of fuel and rate of injection.

(3) Point of the cycle at which injection occurs.

FIG. 8.-Arrangement of Rings for Securing Air Tightness of Crank Chamber Showing also System of Lubrication of Main and Crank Pin Bearings.

(4) Fineness of the spray.

I. Centrifugal oiler.

J. Airtight rings.

K. Airtight rings, springs.

L. Airtight rings, driving pins.

M. Balance weights.

(5) Distance of injector from the hot igniting surface.

These points are not given necessarily in order of importance. Turbulence, apart from piston speed which is governed as already stated by consideration of scavenging, is determined by the shape of the piston crown and the combustion chamber, and in the immediate vicinity of the spray by the shape, speed and volume of the spray. The question of turbulence is an exact parallel to that of scavenging and as yet has received little attention, excepting for the experiments arising out of the necessity to burn tar oils in Diesel engines.

The pressure of injection is governed by the piston speed of the fuel pump or by the angle of crank revolution allowed for the injection of the fuel and by the size of the orifice or orifices in the injector. Pressure is necessary more to give momentum to the stationary column of oil than to secure fineness of the injection spray. It is an absolute essential that injection shall be as rapid as possible, and the injection devices so designed that no "after drip" takes place.

The angle of revolution allowed for the injection period is governed by the ratio of diameter to stroke of the fuel injection pump which injects the fuel through a non-return valve or valves direct into the hot bulb. Obviously with a large diameter and a small stroke fuel pump the period is short and conversely. Practical considerations of design of the pump and governing mechanism determine the period for full power running to be about 30 degrees (see Fig. 6). As regards fineness of spray, no standard has been fixed, although the spray can certainly be "too fine" for rapid ignition.

The next essential is probably that the whole of the oil should be in the combustion chamber in the form of a spray before the first particle touches the hot bulb, which again is a function of rapidity of injection and of the distance through which the oil is thrown, arguing in favor of a longdistance of throw to give rapid ignition and the maximum degree of turbulence in the vicinity of the spray, although a long throw will militate against flexibility. The efficiency of the scavenging will determine to what extent the hot bulb is charged with burnt gases or with fresh air, and will thus have a direct influence upon the speed of ignition and combustion.

Fuel.-In view of the simplicity of this type of engine in comparison with the usual four-cycle internal-combustion gas or Diesel engine, it might be matter for surprise that the semi-Diesel engine has not made greater headway in the past than has been the case. The one outstanding difference between the semi-Diesel and the Diesel engine has been the small range of working fuels with which it could satisfactorily cope. It is but a few years since the great majority of these engines almost required paraffin or the very lightest of petroleums for their successful operation in practice with reasonable costs for upkeep and maintenance. Recently, however, the advantages of simplicity of this engine have been more generally recognized and have led its producers to experiment on the question of utilizing fuel oils of a heavier nature, which have been more readily procurable within the last few years, thus extending greatly the field of application. This movement has been largely responsible for many attendant improvements such as increasing the compression pressure.

for

The present stage of development of the semi-Diesel engine permits it to use most of the heavy fuel oils ranging from 0.8 to 0.9 in specific gravity and with flash points from 130° to 250° F. Perhaps the most frequently used oils are "Solar" and "Shale," but paraffin at the one end of the scale and Texas at the other may be said to be quite suitable without special adjustments or contrivances. Very thick oils such as "Mexican example, may be used, but require preheating to facilitate pumping, and periodical runs on lighter oils are desirable in such case in order to keep the pipes clear, the pistons clean, and the piston rings free in their grooves. With regard to sulphur, the semi-Diesel is no more sensitive than the Diesel engine. Experiments are being carried out at the present time in order to permit of the use of tar oil.

A note should be made in connection with the subject of burning heavy fuel oils with semi-Diesel engines, that this engine being a solid" injection engine does not run with such a clean exhaust as is customary with air-injection engines, and the amount of overhauling required for cleaning of piston rings, etc., is on that account greater, and, of course, is increased with the heavier oils as compared with shale oil and such lighter oils. The user must, therefore, balance the gain of cheap and readily-obtained fuels with the extra overhauling which may on that account be required, taking

into account the size of the engine as to whether the parts to be handled are of convenient size and weight.

· Lubrication.-On the subject of lubrication, Fig. 8 shows the means provided for the main and the crank-pin bearings, and Fig. 9 illustrates a satisfactory device for the lubrication of the connecting-rod top end bearing, whereby the oil is collected from the cylinder walls and conveyed to this bearing. The lubrication of the cylinder walls is carried out in exactly the same manner as is customary with Diesel engines, a special lead being provided for the top-end bearing. Forced lubrication to the main, crank-pin and top end bearings cannot be used with semi-Diesel engines, so long as crank-case scavenging is utilized in order to avoid excessive impregnation of the crank chamber air with lubricating oil.

The qualities of lubricating oil desirable for semi-Diesel engines differ in no way from those required by the Diesel engine, and, with careful design of the airtight rings shown in Fig. 8, and good fitting piston rings, the

FIG. 9.-Lubricating Oil Collector for Gudgeon Pin.

A. Piston.

B. Gudgeon pin.

C. Oil collector.

D. Oil collector casing.

List of Parts.

E. Oil collector spring.
F. Oil holes.

G. Oil tube from lubricator.

consumption of lubricating oil compares favorably with the figure for twocycle Diesel engines, and is in the neighborhood of 0.02 lb. per brake horsepower per hour.

Starting.-Semi-Diesel engines are started by means of compressed air. Due to the low compression pressure, a comparatively low pressure of starting air is sufficient to ensure reliable starting. The minimum pressure at which the engine will start is from 80 lbs. to 100 lbs. per square inch, and the starting air is generally stored at 200 lbs. per square inch, which, with a suitable volume of storage, gives the requisite number of starts, or a margin for contingencies. Prior to starting the hot bulb is brought to a sufficient

temperature to ignite the fuel, which is accomplished by means of a blow lamp in from 10 minutes to 15 minutes. If the engine is provided with special starting plugs of nickel steel screwed in the hot bulb, very much less time is required. These special plugs quickly attain the necessary temperature for ignition. One impulse is generally sufficient to start the engine, so that operating gear for the starting air valves, other than a hand lever, is not customarily fitted with land engines, excepting occasionally in the case of four-cylinder engines. For the compression of starting air a separate hand or power-driven compressor can be installed. The practice with semi-Diesel engines is, during each working stroke to tap off a portion of the working gases through a combined non-return and screw-down valve on the main cylinder, and to pass them to the starting reservoir or reservoirs until such time as any reduction of contents of these reservoirs has been made good.

For multi-cylinder engines which require to start against a load, as for instance those driving pumps or propellers, the same mechanism can well be fitted as with the Diesel engine, the starting air valves being operated either from a main camshaft or through a distributing box, with a secondary camshaft driven from the crankshaft. The quantity of air storage requisite for starting is generally considerably less than is required by Diesel engines, since the hot bulbs have, of necessity, been heated prior to turning the engine, and one revolution on compressed air is sufficient to start the engine. The quality of reversibility, which is almost exclusively required for marine engines, presents few difficulties where two-cycle semi-Diesel engines are concerned, and some notes in regard to this subject are given in Appendix II.

Reliability and Regularity in Operation.-As would be expected from the extreme simplicity of this prime mover, its reliability and regularity in operation are of a high order. Due to the necessity of restricting the quantity and minimizing the pressure of lubricating oil, as already dealt with under the heading of "Lubrication," earlier designs were subject to bearing troubles attributable to the failure of the lubricating system. With modern designs ample bearing surfaces and carefully-designed means for the provision of the necessarily restricted quantity of lubricating oil, these defects have been entirely eliminated.

The system for fuel injection due to the adoption of solid" or 66 mechanical" injection is considerably simplified in comparison with air injection engines.

Faulty circulation of cooling water and unsatisfactory designs of castings-more particularly those for the cylinder head and hot bulb-have been the cause of a certain amount of trouble with cracked heads and so forth; but modern designs have practically overcome this earlier source of unreliability.

Conclusion. The rapid extension within the last few years not only of the field of application, but also of the size of engine and power developed per cylinder with semi-Diesel engines, foreshadows considerable developments in the near future. In these developments the influence of the design and practice of the pure Diesel engine will probably play a considerable part, and it may be expected that the lines of design of the Diesel and semi-Diesel engines will become more closely merged.

Practical difficulties would seem almost to confine the semi-Diesel engine to the two-stroke cycle. Developments_towards improving the efficiency of scavenging may well be expected. In the United States of America semi-Diesel engines with separate scavenging pumps, crossheads, and so with forced lubrication, are already making their appearance.

The mean effective pressures developed by the semi-Diesel engine under conditions of continuous running are considerably less than those associated with the Diesel engine, due primarily to the questions of scavenging and compression already fully discussed. It is not expected that other than a slight increase in mean effective pressures can be looked for in the near future, since the simplicity of the semi-Diesel engine and its relatively low