E-text prepared by Juliet Sutherland, Harry Lamé,
and the Online Distributed Proofreading Team
(http://www.pgdp.net)

PRINTED IN U. S. A.

COMPOSITION, ELECTROTYPING AND PRESSWORK
BY THE PUBLISHERS PRINTING CO., NEW YORK

[9]

In presenting this treatise on “Aviation Engines,” the writer realizes that the rapidly developing art makes it difficult to outline all latest forms or describe all current engineering practice. This exposition has been prepared primarily for instruction purposes and is adapted for men in the Aviation Section, Signal Corps, and students who wish to become aviators or aviation mechanicians. Every effort has been made to have the engineering information accurate, but owing to the diversity of authorities consulted and use of data translated from foreign language periodicals, it is expected that some slight errors will be present. The writer wishes to acknowledge his indebtedness to such firms as the Curtiss Aeroplane and Motor Co., Hall-Scott Company, Thomas-Morse Aircraft Corporation and General Vehicle Company for photographs and helpful descriptive matter. Special attention has been paid to instructions on tool equipment, use of tools, trouble “shooting” and engine repairs, as it is on these points that the average aviation student is weakest. Only such theoretical consideration of thermo-dynamics as was deemed absolutely necessary to secure a proper understanding of engine action after consulting several instructors is included, the writer’s efforts having been confined to the preparation of a practical series of instructions that would be of the greatest value to those who need a diversified knowledge of internal-combustion engine operation and repair, and[10] who must acquire it quickly. The engines described and illustrated are all practical forms that have been fitted to airplanes capable of making flights and may be considered fairly representative of the present state of the art.

Victor W. Pagé,
1st Lieut. A. S. S. C., U. S. R.

Mineola, L. I.,
October, 1917.

[11]

[17]

AVIATION ENGINES
DESIGN—CONSTRUCTION—REPAIR

The conquest of the air is one of the most stupendous achievements of the ages. Human flight opens the sky to man as a new road, and because it is a road free of all obstructions and leads everywhere, affording the shortest distance to any place, it offers to man the prospect of unlimited freedom. The aircraft promises to span continents like railroads, to bridge seas like ships, to go over mountains and forests like birds, and to quicken and simplify the problems of transportation. While the actual conquest of the air is an accomplishment just being realized in our days, the idea and yearning to conquer the air are old, possibly as old as intellect itself. The myths of different races tell of winged gods and flying men, and show that for ages to fly was the highest conception of the sublime. No other agent is more responsible for sustained flight than the internal combustion motor, and it was only when this form of prime mover had been fully developed that it was possible for man to leave the ground and alight at will, not depending upon the caprices of the winds or lifting power of gases as with the balloon. It is safe to say that the solution of the problem of flight would have been attained many years ago if the proper source of power had been available as all the essential[18] elements of the modern aeroplane and dirigible balloon, other than the power plant, were known to early philosophers and scientists.

Aeronautics is divided into two fundamentally different branches—aviatics and aerostatics. The first comprises all types of aeroplanes and heavier than air flying machines such as the helicopters, kites, etc.; the second includes dirigible balloons, passive balloons and all craft which rise in the air by utilizing the lifting force of gases. Aeroplanes are the only practical form of heavier-than-air machines, as the helicopters (machines intended to be lifted directly into the air by propellers, without the sustaining effect of planes), and ornithopters, or flapping wing types, have not been thoroughly developed, and in fact, there are so many serious mechanical problems to be solved before either of these types of air craft will function properly that experts express grave doubts regarding the practicability of either. Aeroplanes are divided into two main types—monoplanes or single surface forms, and bi-planes or machines having two sets of lifting surfaces, one suspended over the other. A third type, the triplane, is not very widely used.

Dirigible balloons are divided into three classes: the rigid, the semi-rigid, and the non-rigid. The rigid has a frame or skeleton of either wood or metal inside of the bag, to stiffen it; the semi-rigid is reinforced by a wire net and metal attachments; while the non-rigid is just a bag filled with gas. The aeroplane, more than the dirigible and balloon, stands as the emblem of the conquest of the air. Two reasons for this are that power flight is a real conquest of the air, a real victory over the battling elements; secondly, because the aeroplane, or any flying machine that may follow, brings air travel within the reach of everybody. In practical development, the dirigible may be the steamship of the air, which will render invaluable services of a certain kind, and the aeroplane will be the automobile of the air, to be used by the multitude, perhaps for as many purposes as the automobile is now being used.

[19]

One of the marked features of aircraft development has been the effect it has had upon the refinement and perfection of the internal combustion motor. Without question gasoline-motors intended for aircraft are the nearest to perfection of any other type yet evolved. Because of the peculiar demands imposed upon the aeronautical motor it must possess all the features of reliability, economy and efficiency now present with automobile or marine engines and then must have distinctive points of its own. Owing to the unstable nature of the medium through which it is operated and the fact that heavier-than-air machines can maintain flight only as long as the power plant is functioning properly, an airship motor must be more reliable than any used on either land or water. While a few pounds of metal more or less makes practically no difference in a marine motor and has very little effect upon the speed or hill-climbing ability of an automobile, an airship motor must be as light as it is possible to make it because every pound counts, whether the motor is to be fitted into an aeroplane or in a dirigible balloon.

Airship motors, as a rule, must operate constantly at high speeds in order to obtain a maximum power delivery with a minimum piston displacement. In automobiles, or motor boats, motors are not required to run constantly at their maximum speed. Most aircraft motors must function for extended periods at speed as nearly the maximum as possible. Another thing that militates against the aircraft motor is the more or less unsteady foundation to which it is attached. The necessarily light framework of the aeroplane makes it hard for a motor to perform at maximum efficiency on account of the vibration of its foundation while the craft is in flight. Marine and motor car engines, while not placed on foundations as firm as those provided for stationary power plants, are installed on bases of much more stability than the light structure of an aeroplane. The aircraft motor, therefore, must be balanced to a nicety[20] and must run steadily under the most unfavorable conditions.

The capacity of light motors designed for aerial work per unit of mass is surprising to those not fully conversant with the possibilities that a thorough knowledge of proportions of parts and the use of special metals developed by the automobile industry make possible. Activity in the development of light motors has been more pronounced in France than in any other country. Some of these motors have been complicated types made light by the skillful proportioning of parts, others are of the refined simpler form modified from current automobile practice. There is a tendency to depart from the freakish or unconventional construction and to adhere more closely to standard forms because it is necessary to have the parts of such size that every quality making for reliability, efficiency and endurance are incorporated in the design. Aeroplane motors range from two cylinders to forms having fourteen and sixteen cylinders and the arrangement of these members varies from the conventional vertical tandem and opposed placing to the V form or the more unusual radial motors having either fixed or rotary cylinders. The weight has been reduced so it is possible to obtain a complete power plant of the revolving cylinder air-cooled type that will not weigh more than three pounds per actual horse-power and in some cases less than this.

If we give brief consideration to the requirements of the aviator it will be evident that one of the most important is securing maximum power with minimum mass, and it is desirable to conserve all of the good qualities existing in standard automobile motors. These are certainty of operation, good mechanical balance and uniform delivery of power—fundamental conditions which must be attained before a power plant can be considered practical. There are in addition, secondary considerations, none the less desirable, if not absolutely essential. These are minimum[21] consumption of fuel and lubricating oil, which is really a factor of import, for upon the economy depends the capacity and flying radius. As the amount of liquid fuel must be limited the most suitable motor will be that which is powerful and at the same time economical. Another important feature is to secure accessibility of components in order to make easy repair or adjustment of parts possible. It is possible to obtain sufficiently light-weight motors without radical departure from established practice. Water-cooled power plants have been designed that will weigh but four or five pounds per horse-power and in these forms we have a practical power plant capable of extended operation.

Work is performed whenever an object is moved against a resistance, and the amount of work performed depends not only on the amount of resistance overcome but also upon the amount of time utilized in accomplishing a given task. Work is measured in horse-power for convenience. It will take one horse-power to move 33,000 pounds one foot in one minute or 550 pounds one foot in one second. The same work would be done if 330 pounds were moved 100 feet in one minute. It requires a definite amount of power to move a vehicle over the ground at a certain speed, so it must take power to overcome resistance of an airplane in the air. Disregarding the factor of air density, it will take more power as the speed increases if the weight or resistance remains constant, or more power if the speed remains constant and the resistance increases. The airplane is supported by air reaction under the planes or lifting surfaces and the value of this reaction depends upon the shape of the aerofoil, the amount it is tilted and the speed at which it is drawn through the air. The angle of incidence or degree of wing tilt regulates the power required to a certain degree as this affects the speed of horizontal flight as well as the resistance. Resistance may be of two kinds, one that is[22] necessary and the other that it is desirable to reduce to the lowest point possible. There is the wing resistance and the sum of the resistances of the rest of the machine such as fuselage, struts, wires, landing gear, etc. If we assume that a certain airplane offered a total resistance of 300 pounds and we wished to drive it through the air at a speed of sixty miles per hour, we can find the horse-power needed by a very simple computation as follows:

The result is the horse-power needed, or

Just as it takes more power to climb a hill than it does to run a car on the level, it takes more power to climb in the air with an airplane than it does to fly on the level. The more rapid the climb, the more power it will take. If the resistance remains 300 pounds and it is necessary to drive the plane at 90 miles per hour, we merely substitute proper values in the above formula and we have

The same results can be obtained by dividing the product of the resistance in pounds times speed in feet per second by 550, which is the foot-pounds of work done in one second to equal one horse-power. Naturally, the amount of propeller thrust measured in pounds necessary to drive an airplane must be greater than the resistance by a substantial margin if the plane is to fly and climb as well.[23] The following formulæ were given in “The Aeroplane” of London and can be used to advantage by those desiring to make computations to ascertain power requirements:

The thrust of the propeller depends on the power of[24] the motor, and on the diameter and pitch of the propeller. If the required thrust to a certain machine is known, the calculation for the horse-power of the motor should be an easy matter.

The required thrust is the sum of three different “resistances.” The first is the “drift” (dynamical head resistance of the aerofoils), i.e., tan α × lift (L), lift being equal to the total weight of machine (W) for horizontal flight and α equal to the angle of incidence. Certainly we must take the tan α at the maximum K_y value for minimum speed, as then the drift is the greatest (Fig. 1, A).

Another method for finding the drift is D = K × AV², when we take the drift again so as to be greatest.

The second “resistance” is the total head resistance of the machine, at its maximum velocity. And the third is the thrust for climbing. The horse-power for climbing can be found out in two different ways. I first propose to deal with the method, where we find out the actual horse-power wanted for a certain climbing speed to our machine, where

In this case we know already the horse-power for climbing, and we can proceed with our calculation.

With the other method we shall find out the “thrust” in pounds or kilograms wanted for climbing and add it to drift and total head resistance, and we shall have the total “thrust” of our machine and we shall denote it with T, while thrust for climbing shall be T_c.

The following calculation is at our service to find out this thrust for climbing

thence

[25]

then from (1)

thence,

Whether T means drifts, head resistance and thrust for climbing, or drift and head resistance only, the following calculation is the same, only in the latter case, of course, we must add the horse-power required for climbing to the result to obtain the total horse-power.

Now, when we know the total thrust, we shall find the horse-power in the following manner:

We know that the

in kilograms, or in English measure,

where

When P × r = M, then

thence,

or in English system

Now the power on the circumference of the propeller will be reduced by its radius, so it will be M/r = p. A part[26] of p will be used for counteracting the air and bearing friction, so that the total power on the circumference of the propeller will be (M/r) × η = p where η is the mechanical efficiency of the propeller. Now η/tan α = T, where α is taken on the tip of the propeller.

I take α at the tip, but it can be taken, of course, at any point, but then in equation p = M/r, r must be taken only up to this point, and not the whole radius; but it is more comfortable to take it at the tip, as tan α = Pitch/r2π (Fig. 1, C).

Now we can write up the equation of the thrust:

thence

The computations and formulæ given are of most value to the student engineer rather than matters of general interest, but are given so that a general idea may be secured of how airplane design influences power needed to secure sustained flight. It will be apparent that the resistance of an airplane depends upon numerous considerations of design which require considerable research in aerodynamics to determine accurately. It is obvious that the more resistance there is, the more power needed to fly at a given speed. Light monoplanes have been flown with as little as 15 horse-power for short distances,[27] but most planes now built use engines of 100 horse-power or more. Giant airplanes have been constructed having 2,000 horse-power distributed in four power units. The amount of power provided for an airplane of given design varies widely as many conditions govern this, but it will range from approximately one horse-power to each 8 pounds weight in the case of very light, fast machines to one horse-power to 15 or 18 pounds of the total weight in the case of medium speed machines. The development in airplane and power plant design is so rapid, however, that the figures given can be considered only in the light of general averages rather than being typical of current practice.

Internal combustion engines are best for airplanes and all types of aircraft for the same reasons that they are universally used as a source of power for automobiles. The gasoline engine is the lightest known form of prime mover and a more efficient one than a steam engine, especially in the small powers used for airplane propulsion. It has been stated that by very careful designing a steam plant an engine could be made that would be practical for airplane propulsion, but even with the latest development it is doubtful if steam power can be utilized in aircraft to as good advantage as modern gasoline-engines are. While the steam-engine is considered very much simpler than a gas-motor, the latter is much more easily mastered by the non-technical aviator and certainly requires less attention. A weight of 10 pounds per horse-power is possible in a condensing steam plant but this figure is nearly double or triple what is easily secured with a gas-motor which may weigh but 5 pounds per horse-power in the water cooled forms and but 2 or 3 pounds in the air-cooled types. The fuel consumption is twice as great in a steam-power plant (owing to heat losses) as would be the case in a gasoline engine of equal power and much less weight.

[28]The internal-combustion engine has come seemingly like an avalanche of a decade; but it has come to stay, to take its well-deserved position among the powers for aiding labor. Its ready adaptation to road, aerial and marine service has made it a wonder of the age in the development of speed not before dreamed of as a possibility; yet in so short a time, its power for speed has taken rank on the common road against the locomotive on the rail with its century’s progress. It has made aerial navigation possible and practical, it furnishes power for all marine craft from the light canoe to the transatlantic liner. It operates the machine tools of the mechanic, tills the soil for the farmer and provides healthful recreation for thousands by furnishing an economical means of transport by land and sea. It has been a universal mechanical education for the masses, and in its present forms represents the great refinement and development made possible by the concentration of the world’s master minds on the problems incidental to internal combustion engineering.

Although the ideal principle of explosive power was conceived some two hundred years ago, at which time experiments were made with gunpowder as the explosive element, it was not until the last years of the eighteenth century that the idea took a patentable shape, and not until about 1826 (Brown’s gas-vacuum engine) that a further progress was made in England by condensing the products of combustion by a jet of water, thus creating a partial vacuum.

Brown’s was probably the first explosive engine that did real work. It was clumsy and unwieldy and was soon relegated to its place among the failures of previous experiments. No approach to active explosive effect in a cylinder was reached in practice, although many ingenious designs were described, until about 1838 and the following years. Barnett’s engine in England was the first attempt to compress the charge before exploding. From this time[29] on to about 1860 many patents were issued in Europe and a few in the United States for gas-engines, but the progress was slow, and its practical introduction for power came with spasmodic effect and low efficiency. From 1860 on, practical improvement seems to have been made, and the Lenoir motor was produced in France and brought to the United States. It failed to meet expectations, and was soon followed by further improvements in the Hugon motor in France (1862), followed by Beau de Rocha’s four-cycle idea, which has been slowly developed through a long series of experimental trials by different inventors. In the hands of Otto and Langdon a further progress was made, and numerous patents were issued in England, France, and Germany, and followed up by an increasing interest in the United States, with a few patents.

From 1870 improvements seem to have advanced at a steady rate, and largely in the valve-gear and precision of governing for variable load. The early idea of the necessity of slow combustion was a great drawback in the advancement of efficiency, and the suggestion of de Rocha in 1862 did not take root as a prophetic truth until many failures and years of experience had taught the fundamental axiom that rapidity of action in both combustion and expansion was the basis of success in explosive motors.

With this truth and the demand for small and safe prime movers, the manufacture of gas-engines increased in Europe and America at a more rapid rate, and improvements in perfecting the details of this cheap and efficient prime mover have finally raised it to the dignity of a standard motor and a dangerous rival of the steam-engine for small and intermediate powers, with a prospect of largely increasing its individual units to many hundred, if not to the thousand horse-power in a single cylinder. The unit size in a single cylinder has now reached to about 700 horse-power and by combining cylinders in the same machine, powers of from 1,500 to 2,000 horse-power are now available for large power-plants.

[30]

This form of prime mover has been built in so many different types, all of which have operated with some degree of success that the diversity in form will not be generally appreciated unless some attempt is made to classify the various designs that have received practical application. Obviously the same type of engine is not universally applicable, because each class of work has individual peculiarities which can best be met by an engine designed with the peculiar conditions present in view. The following tabular synopsis will enable the reader to judge the extent of the development of what is now the most popular prime mover for all purposes.

[33]

Fig. 2.—Plate Showing Heavy, Slow Speed Internal Combustion Engines Used Only for Stationary Power in Large Installations Giving Weight to Horse-Power Ratio.

[34]

Fig. 3.—Various Forms of Internal Combustion Engines Showing Decrease in Weight to Horse-Power Ratio with Augmenting Speed of Rotation.

[35]

Fig. 4.—Internal Combustion Engine Types of Extremely Fine Construction and Refined Design, Showing Great Power Outputs for Very Small Weight, a Feature Very Much Desired in Airplane Power Plants.

[36]Of all the types enumerated above engines having less than eight cylinders are the most popular in everything but aircraft work. The four-cylinder vertical is without doubt the most widely used of all types owing to the large number employed as automobile power plants. Stationary engines in small and medium powers are invariably of the single or double form. Three-cylinder engines are seldom used at the present time, except in marine work and in some stationary forms. Eight- and twelve-cylinder motors have received but limited application and practically always in automobiles, racing motor boats or in aircraft. The only example of a fourteen-cylinder motor to be used to any extent is incorporated in aeroplane construction. This is also true of the sixteen- and eighteen-cylinder forms and of twenty-four-cylinder engines now in process of development.

The duty an engine is designed for determines the weight per horse-power. High powered engines intended for steady service are always of the slow speed type and consequently are of very massive construction. Various forms of heavy duty type stationary engines are shown at Fig. 2. Some of these engines may weigh as much as 600 pounds per horse-power. A further study is possible by consulting data given on Figs. 3 and 4. As the crank-shaft speed increases and cylinders are multiplied the engines become lighter. While the big stationary power plants may run for years without attention, airplane engines require rebuilding after about 60 to 80 hours air service for the fixed cylinder types and 40 hours or less for the rotary cylinder air-cooled forms. There is evidently a decrease in durability and reliability as the weight is lessened. These illustrations also permit of obtaining a good idea of the variety of forms internal combustion engines are made in.

[37]

Before discussing the construction of the various forms of internal combustion engines it may be well to describe the operating cycle of the types most generally used. The two-cycle engine is the simplest because there are no valves in connection with the cylinder, as the gas is introduced into that member and expelled from it through ports cored into the cylinder walls. These are covered by the piston at a certain portion of its travel and uncovered at other parts of its stroke. In the four-cycle engine the explosive gas is admitted to the cylinder through a port at the head end closed by a valve, while the exhaust gas is expelled through another port controlled in a similar manner. These valves are operated by mechanism distinct from the piston.

The action of the four-cycle type may be easily understood if one refers to illustrations at Figs. 5 and 6. It is called the “four-stroke engine” because the piston must make four strokes in the cylinder for each explosion or power impulse obtained. The principle of the gas-engine of the internal combustion type is similar to that of a gun, i.e., power is obtained by the rapid combustion of some explosive or other quick burning substance. The bullet is driven out of the gun barrel by the pressure of the gas evolved when the charge of powder is ignited. The piston or movable element of the gas-engine is driven[38] from the closed or head end to the crank end of the cylinder by a similar expansion of gases resulting from combustion. The first operation in firing a gun or securing an explosion in the cylinder of the gas-engine is to fill the combustion space with combustible material. This is done by a down stroke of the piston during which time the inlet valve opens to admit the gaseous charge to the cylinder interior. This operation is shown at Fig. 5, A. The second operation is to compress this gas which is done by an upward stroke of the piston as shown at[39] Fig. 5, B. When the top of the compression stroke is reached, the gas is ignited and the piston is driven down toward the open end of the cylinder, as indicated at Fig. 6, C. The fourth operation or exhaust stroke is performed by the return upward movement of the piston as shown at Fig. 6, D during which time the exhaust valve is opened to permit the burnt gases to leave the cylinder. As soon as the piston reaches the top of its exhaust stroke, the energy stored in the fly-wheel rim during the power stroke causes that member to continue revolving and as the piston[41] again travels on its down stroke the inlet valve opens and admits a charge of fresh gas and the cycle of operations is repeated.

The illustrations at Fig. 7 show how the various cycle functions take place in an L head type water cooled cylinder engine. The sections at A and C are taken through the inlet valve, those at B and D are taken through the exhaust valve.

The two-cycle engine works on a different principle, as while only the combustion chamber end of the piston is employed to do useful work in the four-cycle engine, both upper and lower portions are called upon to perform the functions necessary to two-cycle engine operation. Instead of the gas being admitted into the cylinder as is the case with the four-stroke engine, it is first drawn into the engine base where it receives a preliminary compression prior to its transfer to the working end of the cylinder. The views at Fig. 8 should indicate clearly the operation of the two-port two-cycle engine. At A the piston is seen reaching the top of its stroke and the gas above the piston is being compressed ready for ignition, while the suction in the engine base causes the automatic valve to open and admits mixture from the carburetor to the crank case. When the piston reaches the top of its stroke, the compressed gas is ignited and the piston is driven down on the power stroke, compressing the gas in the engine base.

When the top of the piston uncovers the exhaust port the flaming gas escapes because of its pressure. A downward movement of the piston uncovers the inlet port opposite the exhaust and permits the fresh gas to bypass through the transfer passage from the engine base to the cylinder. The conditions with the intake and exhaust port fully opened are clearly shown at Fig. 8, C. The deflector plate on the top of the piston directs the entering fresh gas to the top of the cylinder and prevents the main portion of the gas stream from flowing out through the open exhaust port. On the next upstroke of the piston[44] the gas in the cylinder is compressed and the inlet valve opened, as shown at A to permit a fresh charge to enter the engine base.

The operating principle of the three-port, two-cycle engine is practically the same as that previously described with the exception that the gas is admitted to the crank-case through a third port in the cylinder wall, which is uncovered by the piston when that member reaches the end of its upstroke. The action of the three-port form can be readily ascertained by studying the diagrams given at Fig. 9. Combination two- and three-port engines have been evolved and other modifications made to improve the action.

In the earlier years of explosive-motor progress was evolved the two types of motors in regard to the cycles of their operation. The early attempts to perfect the two-cycle principle were for many years held in abeyance from the pressure of interests in the four-cycle type, until its simplicity and power possibilities were demonstrated by Mr. Dugald Clerk in England, who gave the principles of the two-cycle motor a broad bearing leading to immediate improvements in design, which has made further progress in the United States, until at the present time it has an equal standard value as a motor-power in some applications as its ancient rival the four-cycle or Otto type, as demonstrated by Beau de Rocha in 1862.

Thermodynamically, the methods of the two types are equal as far as combustion is concerned, and compression may favor in a small degree the four-cycle type as well as the purity of the charge. The cylinder volume of the two-cycle motor is much smaller per unit of power, and the enveloping cylinder surface is therefore greater per unit of volume. Hence more heat is carried off by the jacket water during compression, and the higher compression available from this tends to increase the economy during compression which is lost during expansion.

[45]From the above considerations it may be safely stated that a lower temperature and higher pressure of charge at the beginning of compression is obtained in the two-cycle motor, greater weight of charge and greater specific power of higher compression resulting in higher thermal efficiency. The smaller cylinder for the same power of the two-cycle motor gives less friction surface per impulse than of the other type; although the crank-chamber pressure may, in a measure, balance the friction of the four-cycle type. Probably the strongest points in favor of the two-cycle type are the lighter fly-wheel and the absence of valves and valve gear, making this type the most simple in construction and the lightest in weight for its developed power. Yet, for the larger power units, the four-cycle type will no doubt always maintain the standard for efficiency and durability of action.

The distribution of the charge and its degree of mixture with the remains of the previous explosion in the clearance space, has been a matter of discussion for both types of explosive motors, with doubtful results. In Fig. 10, A we illustrate what theory suggests as to the distribution of the fresh charge in a two-cycle motor, and in Fig. 10, B what is the probable distribution of the mixture when the piston starts on its compressive stroke. The arrows show the probable direction of flow of the fresh charge and burnt gases at the crucial moment.

In Fig. 10, C is shown the complete out-sweep of the products of combustion for the full extent of the piston stroke of a four-cycle motor, leaving only the volume of the clearance to mix with the new charge and at D the manner by which the new charge sweeps by the ignition device, keeping it cool and avoiding possibilities of pre-ignition by undue heating of the terminals of the sparking device. Thus, by enveloping the sparking device with the pure mixture, ignition spreads through the charge with its greatest possible velocity, a most desirable condition in high-speed motors with side-valve chambers and igniters within the valve chamber.

[46]

[47]

The laws controlling the elements that create a power by their expansion by heat due to combustion, when properly understood, become a matter of computation in regard to their value as an agent for generating power in the various kinds of explosive engines. The method of heating the elements of power in explosive engines greatly widens the limits of temperature as available in other types of heat-engines. It disposes of many of the practical troubles of hot-air, and even of steam-engines, in the simplicity and directness of application of the elements of power. In the explosive engine the difficulty of conveying heat for producing expansive effect by convection is displaced by the generation of the required heat within the expansive element and at the instant of its useful work. The low conductivity of heat to and from air has been the great obstacle in the practical development of the hot-air engine; while, on the contrary, it has become the source of economy and practicability in the development of the internal-combustion engine.

The action of air, gas, and the vapors of gasoline and petroleum oil, whether singly or mixed, is affected by changes of temperature practically in nearly the same ratio; but when the elements that produce combustion are interchanged in confined spaces, there is a marked difference of effect. The oxygen of the air, the hydrogen and carbon of a gas, or vapor of gasoline or petroleum oil are the elements that by combustion produce heat to expand the nitrogen of the air and the watery vapor produced by the union of the oxygen in the air and the hydrogen in the gas, as well as also the monoxide and carbonic-acid gas that may be formed by the union of the carbon of gas or vapor with part of the oxygen of the air. The various mixtures as between air and gas, or air and vapor, with the proportion of the products of combustion left in the cylinder from a previous combustion, form the elements to be considered in estimating the amount of[48] pressure that may be obtained by their combustion and expansive force.

The working process of the explosive motor may be divided into three principal types: 1. Motors with charges igniting at constant volume without compression, such as the Lenoir, Hugon, and other similar types now abandoned as wasteful in fuel and effect. 2. Motors with charges igniting at constant pressure with compression, in which a receiver is charged by a pump and the gases burned while being admitted to the motor cylinder, such as types of the Simon and Brayton engine. 3. Motors with charges igniting at constant volume with variable compression, such as the later two- and four-cycle motors with compression of the indrawn charge; limited in the two-cycle type and variable in the four-cycle type with the ratios of the clearance space in the cylinder. This principle produces the explosive motor of greatest efficiency.

The phenomena of the brilliant light and its accompanying heat at the moment of explosion have been witnessed in the experiments of Dugald Clerk in England, the illumination lasting throughout the stroke; but in regard to time in a four-cycle engine, the incandescent state exists only one-quarter of the running time. Thus the time interval, together with the non-conductibility of the gases, makes the phenomena of a high-temperature combustion within the comparatively cool walls of a cylinder a practical possibility.

The natural laws, long since promulgated by Boyle, Gay Lussac, and others, on the subject of the expansion and compression of gases by force and by heat, and their variable pressures and temperatures when confined, are conceded to be practically true and applicable to all gases, whether single, mixed, or combined.

[49]The law formulated by Boyle only relates to the compression and expansion of gases without a change of temperature, and is stated in these words:

If the temperature of a gas be kept constant, its pressure or elastic force will vary inversely as the volume it occupies.

It is expressed in the formula P × V = C, or pressure × volume = constant. Hence, C/P = V and C/V = P.

Thus the curve formed by increments of pressure during the expansion or compression of a given volume of gas without change of temperature is designated as the isothermal curve in which the volume multiplied by the pressure is a constant value in expansion, and inversely the pressure divided by the volume is a constant value in compressing a gas.

But as compression and expansion of gases require force for their accomplishment mechanically, or by the application or abstraction of heat chemically, or by convection, a second condition becomes involved, which was formulated into a law of thermodynamics by Gay Lussac under the following conditions: A given volume of gas under a free piston expands by heat and contracts by the loss of heat, its volume causing a proportional movement of a free piston equal to ¹⁄₂₇₃ part of the cylinder volume for each degree Centigrade difference in temperature, or ¹⁄₄₉₂ part of its volume for each degree Fahrenheit. With a fixed piston (constant volume), the pressure is increased or decreased by an increase or decrease of heat in the same proportion of ¹⁄₂₇₃ part of its pressure for each degree Centigrade, or ¹⁄₄₉₂ part of its pressure for each degree Fahrenheit change in temperature. This is the natural sequence of the law of mechanical equivalent, which is a necessary deduction from the principle that[50] nothing in nature can be lost or wasted, for all the heat that is imparted to or abstracted from a gaseous body must be accounted for, either as heat or its equivalent transformed into some other form of energy. In the case of a piston moving in a cylinder by the expansive force of heat in a gaseous body, all the heat expended in expansion of the gas is turned into work; the balance must be accounted for in absorption by the cylinder or radiation.

This theory is equally applicable to the cooling of gases by abstraction of heat or by cooling due to expansion by the motion of a piston. The denominators of these heat fractions of expansion or contraction represent the absolute zero of cold below the freezing-point of water, and read -273° C. or -492.66° = -460.66° F. below zero; and these are the starting-points of reference in computing the heat expansion in gas-engines. According to Boyle’s law, called the first law of gases, there are but two characteristics of a gas and their variations to be considered, viz., volume and pressure: while by the law of Gay Lussac, called the second law of gases, a third is added, consisting of the value of the absolute temperature, counting from absolute zero to the temperatures at which the operations take place. This is the Adiabatic law.

The ratio of the variation of the three conditions—volume, pressure, and heat—from the absolute zero temperature has a certain rate, in which the volume multiplied by the pressure and the product divided by the absolute temperature equals the ratio of expansion for each degree. If a volume of air is contained in a cylinder having a piston and fitted with an indicator, the piston, if moved to and fro slowly, will alternately compress and expand the air, and the indicator pencil will trace a line or lines upon the card, which lines register the change of pressure and volume occurring in the cylinder. If the piston is perfectly free from leakage, and it be supposed[51] that the temperature of the air is kept quite constant, then the line so traced is called an Isothermal line, and the pressure at any point when multiplied by the volume is a constant, according to Boyle’s law,

pv = a constant.

If, however, the piston is moved very rapidly, the air will not remain at constant temperature, but the temperature will increase because work has been done upon the air, and the heat has no time to escape by conduction. If no heat whatever is lost by any cause, the line will be traced over and over again by the indicator pencil, the cooling by expansion doing work precisely equalling the heating by compression. This is the line of no transmission of heat, therefore known as Adiabatic.

The expansion of a gas ¹⁄₂₇₃ of its volume for every degree Centigrade, added to its temperature, is equal to the decimal .00366, the coefficient of expansion for Centigrade units. To any given volume of a gas, its expansion may be computed by multiplying the coefficient by the[52] number of degrees, and by reversing the process the degree of acquired heat may be obtained approximately. These methods are not strictly in conformity with the absolute mathematical formula, because there is a small increase in the increment of expansion of a dry gas, and there is also a slight difference in the increment of expansion due to moisture in the atmosphere and to the vapor of water formed by the union of the hydrogen and oxygen in the combustion chamber of explosive engines.

The ratio of expansion on the Fahrenheit scale is derived from the absolute temperature below the freezing-point of water (32°) to correspond with the Centigrade scale; therefore ¹⁄_492.66 = .0020297, the ratio of expansion from 32° for each degree rise in temperature on the Fahrenheit scale. As an example, if the temperature of any volume of air or gas at constant volume is raised, say from 60° to 2000° F., the increase in temperature will be 1940°. The ratio will be ¹⁄_520.66 = .0019206. Then by the formula:

Ratio × acquired temp. × initial pressure = the gauge pressure; and .0019206 × 1940° × 14.7 = 54.77 lbs.

By another formula, a convenient ratio is obtained by (absolute pressure)/(absolute temp.) or ^14.7⁄_520.66 = .028233; then, using the difference of temperature as before, .028233 × 1940° = 54.77 lbs. pressure.

By another formula, leaving out a small increment due to specific heat at high temperatures:

[53]

absolute pressure due to the acquired temperature, from which the atmospheric pressure is deducted for the gauge pressure. Using the foregoing example, we have

460.66 being the absolute temperature for zero Fahrenheit.

For obtaining the volume of expansion of a gas from a given increment of heat, we have the approximate formula:

In applying this formula to the foregoing example, the figures become:

From this last term the gauge pressure may be obtained as follows:

III. 4.72604 × 14.7 = 69.47 lbs. absolute - 14.7 lbs. atmospheric pressure = 54.77 lbs. gauge pressure; which is the theoretical pressure due to heating air in a confined space, or at constant volume from 60° to 2000° F.

By inversion of the heat formula for absolute pressure we have the formula for the acquired heat, derived from combustion at constant volume from atmospheric pressure to gauge pressure plus atmospheric pressure as derived from Example I., by which the expression

[54]= absolute temperature + temperature of combustion, from which the acquired temperature is obtained by subtracting the absolute temperature.

Then, for example,

the theoretical heat of combustion. The dropping of terminal decimals makes a small decimal difference in the result in the different formulas.

By Joule’s law of the mechanical equivalent of heat, whenever heat is imparted to an elastic body, as air or gas, energy is generated and mechanical work produced by the expansion of the air or gas. When the heat is imparted by combustion within a cylinder containing a movable piston, the mechanical work becomes an amount measurable by the observed pressure and movement of the piston. The heat generated by the explosive elements and the expansion of the non-combining elements of nitrogen and water vapor that may have been injected into the cylinder as moisture in the air, and the water vapor formed by the union of the oxygen of the air with the hydrogen of the gas, all add to the energy of the work from their expansion by the heat of internal combustion. As against this, the absorption of heat by the walls of the cylinder, the piston, and cylinder-head or clearance walls, becomes a modifying condition in the force imparted to the moving piston.

It is found that when any explosive mixture of air and gas or hydrocarbon vapor is fired, the pressure falls far short of the pressure computed from the theoretical effect of the heat produced, and from gauging the expansion of the contents of a cylinder. It is now well known that in practice the high efficiency which is promised by theoretical calculation is never realized; but it must always be[55] remembered that the heat of combustion is the real agent, and that the gases and vapors are but the medium for the conversion of inert elements of power into the activity of energy by their chemical union. The theory of combustion has been the leading stimulus to large expectations with inventors and constructors of explosive motors; its entanglement with the modifying elements in practice has delayed the best development in construction, and as yet no really positive design of best form or action seems to have been accomplished, although great progress has been made during the past decade in the development of speed, reliability, economy, and power output of the individual units of this comparatively new power.

One of the most serious difficulties in the practical development of pressure, due to the theoretical computations of the pressure value of the full heat, is probably caused by imparting the heat of the fresh charge to the balance of the previous charge that has been cooled by expansion from the maximum pressure to near the atmospheric pressure of the exhaust. The retardation in the velocity of combustion of perfectly mixed elements is now well known from experimental trials with measured quantities; but the principal difficulty in applying these conditions to the practical work of an explosive engine where a necessity for a large clearance space cannot be obviated, is in the inability to obtain a maximum effect from the imperfect mixture and the mingling of the products of the last explosion with the new mixture, which produces a clouded condition that makes the ignition of the mass irregular or chattering, as observed in the expansion lines of indicator cards; but this must not be confounded with the reaction of the spring in the indicator.

Stratification of the mixture has been claimed as taking place in the clearance chamber of the cylinder; but this is not a satisfactory explanation in view of the vortical effect of the violent injection of the air and gas or vapor mixture. It certainly cannot become a perfect mixture in the time of a stroke of a high-speed motor of the two-cycle[56] class. In a four-cycle engine, making 1,500 revolutions per minute, the injection and compression in any one cylinder take place in one twenty-fifth of a second—formerly considered far too short a time for a perfect infusion of the elements of combustion but now very easily taken care of despite the extremely high speed of numerous aviation and automobile power-plants.

In an examination of the times of explosion and the corresponding pressures in both tables, it will be seen that a mixture of 1 part gas to 6 parts air is the most effective and will give the highest mean pressure in a gas-engine. There is a limit to the relative proportions of illuminating gas and air mixture that is explosive, somewhat variable, depending upon the proportion of hydrogen in the gas. With ordinary coal-gas, 1 of gas to 15 parts of air; and on the lower end of the scale, 1 volume of gas to 2 parts air, are non-explosive. With gasoline vapor the explosive effect ceases at 1 to 16, and a saturated mixture of equal volumes of vapor and air will not explode, while the most intense explosive effect is from a mixture of 1 part vapor to 9 parts air. In the use of gasoline and air mixtures from a carburetor, the best effect is from 1 part saturated air to 8 parts free air.

[57]

The weight of a cubic foot of gas and air mixture as given in Col. 2 is found by adding the number of volumes of air multiplied by its weight, .0807, to one volume of gas of weight .035 pound per cubic foot and dividing by the total number of volumes; for example, as in the table, 6 × .0807 = .⁵¹⁹²⁄₇ = .074195 as in the first line, and so on for any mixture or for other gases of different specific weight per cubic foot. The heat units evolved by combustion of the mixture (Col. 6) are obtained by dividing the total heat units in a cubic foot of gas by the total proportion of the mixture, ⁶⁶⁰⁄₇ = 94.28 as in the first line of the table. Col. 5 is obtained by multiplying the weight of a cubic foot of the mixture in Col. 2 by the specific heat at a constant volume (Col. 4), Col. 6/Col. 5 = Col. 7 the total heat[58] ratio, of which Col. 8 gives the usual combustion efficiency—Col. 7 × Col. 8 gives the absolute rise in temperature of a pure mixture, as given in Col. 9.

The many recorded experiments made to solve the discrepancy between the theoretical and the actual heat development and resulting pressures in the cylinder of an explosive motor, to which much discussion has been given as to the possibilities of dissociation and the increased specific heat of the elements of combustion and non-combustion, as well, also, of absorption and radiation of heat, have as yet furnished no satisfactory conclusion as to what really takes place within the cylinder walls. There seems to be very little known about dissociation, and somewhat vague theories have been advanced to explain the phenomenon. The fact is, nevertheless, apparent as shown in the production of water and other producer gases by the use of steam in contact with highly incandescent fuel. It is known that a maximum explosive mixture of pure gases, as hydrogen and oxygen or carbonic oxide and oxygen, suffers a contraction of one-third their volume by combustion to their compounds, steam or carbonic acid. In the explosive mixtures in the cylinder of a motor, however, the combining elements form so small a proportion of the contents of the cylinder that the shrinkage of their volume amounts to no more than 3 per cent. of the cylinder volume. This by no means accounts for the great heat and pressure differences between the theoretical and actual effects.

The utilization of heat in any heat-engine has long been a theme of inquiry and experiment with scientists and engineers, for the purpose of obtaining the best practical conditions and construction of heat-engines that would represent the highest efficiency or the nearest approach to the theoretical value of heat, as measured by empirical laws that have been derived from experimental researches relating to its ultimate volume. It is well known that the[59] steam-engine returns only from 12 to 18 per cent. of the power due to the heat generated by the fuel, about 25 per cent. of the total heat being lost in the chimney, the only use of which is to create a draught for the fire; the balance, some 60 per cent., is lost in the exhaust and by radiation. The problem of utmost utilization of force in steam has nearly reached its limit.

The internal-combustion system of creating power is comparatively new in practice, and is but just settling into definite shape by repeated trials and modification of details, so as to give somewhat reliable data as to what may be expected from the rival of the steam-engine as a prime mover. For small powers, the gas, gasoline, and petroleum-oil engines are forging ahead at a rapid rate, filling the thousand wants of manufacture and business for a power that does not require expensive care, that is perfectly safe at all times, that can be used in any place in the wide world to which its concentrated fuel can be conveyed, and that has eliminated the constant handling of crude fuel and water.

The utilization of heat in a gas-engine is mainly due to the manner in which the products entering into combustion are distributed in relation to the movement of the piston. The investigation of the foremost exponent of the theory of the explosive motor was prophetic in consideration of the later realization of the best conditions under which these motors can be made to meet the requirements of economy and practicability. As early as 1862, Beau de Rocha announced, in regard to the coming power, that four requisites were the basis of operation for economy and best effect. 1. The greatest possible cylinder volume with the least possible cooling surface. 2. The greatest possible rapidity of expansion. Hence, high speed. 3. The greatest possible expansion. Long stroke. 4. The greatest possible pressure at the commencement of expansion. High compression.

[60]

Efficiencies are worked out through intricate formulas for a variety of theoretical and unknown conditions of combustion in the cylinder: ratios of clearance and cylinder volume, and the uncertain condition of the products of combustion left from the last impulse and the wall temperature. But they are of but little value, except as a mathematical inquiry as to possibilities. The real commercial efficiency of a gas or gasoline-engine depends upon the volume of gas or liquid at some assigned cost, required per actual brake horse-power per hour, in which an indicator card should show that the mechanical action of the valve gear and ignition was as perfect as practicable, and that the ratio of clearance, space, and cylinder volume gave a satisfactory terminal pressure and compression: i.e., the difference between the power figured from the indicator card and the brake power being the friction loss of the engine.

In four-cycle motors of the compression type, the efficiencies are greatly advanced by compression, producing a more complete infusion of the mixture of gas or vapor and air, quicker firing, and far greater pressure than is possible with the two-cycle type previously described. In the practical operation of the gas-engine during the past twenty years, the gas-consumption efficiencies per indicated horse-power have gradually risen from 17 per cent. to a maximum of 40 per cent. of the theoretical heat, and[61] this has been done chiefly through a decreased combustion chamber and increased compression—the compression having gradually increased in practice from 30 lbs. per square inch to above 100; but there seems to be a limit to compression, as the efficiency ratio decreases with greater increase in compression. It has been shown that an ideal efficiency of 33 per cent. for 38 lbs., compression will increase to 40 per cent. for 66 lbs., and 43 per cent. for 88 lbs. compression. On the other hand, greater compression means greater explosive pressure and greater strain on the engine structure, which will probably retain in future practice the compression between the limits of 40 and 90 lbs. except in super-compression engines intended for high altitude work where compression pressures as high as 125 pounds have been used.

In experiments made by Dugald Clerk, in England, with a combustion chamber equal to 0.6 of the space swept by the piston, with a compression of 38 lbs., the consumption of gas was 24 cubic feet per indicated horse-power per hour. With 0.4 compression space and 61 lbs. compression, the consumption of gas was 20 cubic feet per indicated horse-power per hour; and with 0.34 compression space and 87 lbs. compression, the consumption of gas fell to 14.8 cubic feet per indicated horse-power per hour—the actual efficiencies being respectively 17, 21, and 25 per cent. This was with a Crossley four-cycle engine.

The efficiencies in regard to power in a heat-engine may be divided into four kinds, as follows: I. The first is known as the maximum theoretical efficiency of a perfect engine (represented by the lines in the indicator diagram). It is expressed by the formula

and shows the work of a perfect cycle in an engine working between the received temperature + absolute temperature (T₁) and[62] the initial atmospheric temperature + absolute temperature (T₀). II. The second is the actual heat efficiency, or the ratio of the heat turned into work to the total heat received by the engine. It expresses the indicated horse-power. III. The third is the ratio between the second or actual heat efficiency and the first or maximum theoretical efficiency of a perfect cycle. It represents the greatest possible utilization of the power of heat in an internal-combustion engine. IV. The fourth is the mechanical efficiency. This is the ratio between the actual horse-power delivered by the engine through a dynamometer or measured by a brake (brake horse-power), and the indicated horse-power. The difference between the two is the power lost by engine friction. In regard to the general heat efficiency of the materials of power in explosive engines, we find that with good illuminating gas the practical efficiency varies from 25 to 40 per cent.; kerosene-motors, 20 to 30; gasoline-motors, 20 to 32; acetylene, 25 to 35; alcohol, 20 to 30 per cent. of their heat value. The great variation is no doubt due to imperfect mixtures and variable conditions of the old and new charge in the cylinder; uncertainty as to leakage and the perfection[63] of combustion. In the Diesel motors operating under high pressure, up to nearly 500 pounds, an efficiency of 36 per cent. is claimed.

The graphic diagram at Fig. 12 is of special value as it shows clearly how the heat produced by charge combustion is expended in an engine of average design.

On general principles the greater difference between the heat of combustion and the heat at exhaust is the relative measure of the heat turned into work, which represents the degree of efficiency without loss during expansion. The mathematical formulas appertaining to the computation of the element of heat and its work in an explosive engine are in a large measure dependent upon assumed values, as the conditions of the heat of combustion are made uncertain by the mixing of the fresh charge with the products of a previous combustion, and by absorption, radiation, and leakage. The computation of the temperature from the observed pressure may be made as before explained, but for compression-engines the needed starting-points for computation are very uncertain, and can only be approximated from the exact measure and value of the elements of combustion in a cylinder charge.

Owing to the decrease from atmospheric pressure in the indrawing charge of the cylinder, caused by valve and frictional obstruction, the compression seldom starts above 13 lbs. absolute, especially in high-speed engines. Col. 3 in the following table represents the approximate absolute compression pressure for the clearance percentage and ratio in Cols. 1 and 2, while Col. 4 indicates the gauge pressure from the atmospheric line. The temperatures in Col. 5 are due to the compression in Col. 3 from an assumed temperature of 560° F. in the mixture of the fresh charge of 6 air to 1 gas with the products of combustion left in the clearance chamber from the exhaust stroke of a medium-speed motor. This temperature is subject to[64] considerable variation from the difference in the heat-unit power of the gases and vapors used for explosive power, as also of the cylinder-cooling effect. In Col. 6 is given the approximate temperatures of explosion for a mixture of air 6 to gas 1 of 660 heat units per cubic foot, for the relative values of the clearance ratio in Col. 2 at constant volume.

In view of the experiments in this direction, it clearly shows that in practical work, to obtain the greatest economy per effective brake horse-power, it is necessary: 1st. To transform the heat into work with the greatest rapidity mechanically allowable. This means high piston speed. 2d. To have high initial compression. 3d. To reduce the duration of contact between the hot gases and the cylinder walls to the smallest amount possible; which means short stroke and quick speed, with a spherical cylinder head. 4th. To adjust the temperature of the jacket water to[65] obtain the most economical output of actual power. This means water-tanks or water-coils, with air-cooling surfaces suitable and adjustable to the most economical requirement of the engine, which by late trials requires the jacket water to be discharged at about 200° F. 5th. To reduce the wall surface of the clearance space or combustion chamber to the smallest possible area, in proportion to its required volume. This lessens the loss of the heat of combustion by exposure to a large surface, and allows of a higher mean wall temperature to facilitate the heat of compression.

In an experimental investigation of the efficiency of a gas-engine under variable piston speeds made in France, it was found that the useful effect increases with the velocity of the piston—that is, with the rate of expansion of the burning gases with mixtures of uniform volumes: so that the variations of time of complete combustion at constant pressure, and the variations due to speed, in a way compensate in their efficiencies. The dilute mixture, being slow burning, will have its time and pressure quickened by increasing the speed.

Careful trials give unmistakable evidence that the useful effect increases with the velocity of the piston—that is, with the rate of expansion of the burning gases. The time necessary for the explosion to become complete and to attain its maximum pressure depends not only on the composition of the mixture, but also upon the rate of expansion. This has been verified in experiments with a high-speed motor, at speeds from 500 to 2,000 revolutions per minute, or piston speeds of from 16 to 64 feet per second. The increased speed of combustion due to increased piston speed is a matter of great importance to builders of gas-engines, as well as to the users, as indicating the mechanical direction of improvements to lessen the wearing strain due to high speed and to lighten the vibrating parts with increased strength, in order that the[66] balancing of high-speed engines may be accomplished with the least weight.

From many experiments made in Europe and in the United States, it has been conclusively proved that excessive cylinder cooling by the water-jacket results in a marked loss of efficiency. In a series of experiments with a simplex engine in France, it was found that a saving of 7 per cent. in gas consumption per brake horse-power was made by raising the temperature of the jacket water from 141° to 165° F. A still greater saving was made in a trial with an Otto engine by raising the temperature of the jacket water from 61° to 140° F.—it being 9.5 per cent. less gas per brake horse-power.

It has been stated that volumes of similar cylinders increase as the cube of their diameters, while the surface of their cold walls varies as the square of their diameters; so that for large cylinders the ratio of surface to volume is less than for small ones. This points to greater economy in the larger engines. The study of many experiments goes to prove that combustion takes place gradually in the gas-engine cylinder, and that the rate of increase of pressure or rapidity of firing is controlled by dilution and compression of the mixture, as well as by the rate of expansion or piston speed. The rate of combustion also depends on the size and shape of the explosion chamber, and is increased by the mechanical agitation of the mixture during combustion, and still more by the mode of firing.

To the uninitiated, indicator cards are considerable of a mystery; to those capable of reading them they form an index relative to the action of any engine. An indicator card, such as shown at Fig. 13, is merely a graphical representation of the various pressures existing in the cylinder for different positions of the piston. The length is to some scale that represents the stroke of the piston. During the intake stroke, the pressure falls below the[67] atmospheric line. During compression, the curve gradually becomes higher owing to increasing pressure as the volume is reduced. After ignition the pressure line moves upward almost straight, then as the piston goes down on the explosion stroke, the pressure falls gradually to the point of exhaust valve, opening when the sudden release of the imprisoned gas causes a reduction in pressure to nearly atmospheric. An indicator card, or a series of them, will always show by its lines the normal or defective condition of the inlet valve and passages; the actual line of compression; the firing moment; the pressure of explosion; the velocity of combustion; the normal or defective line of expansion, as measured by the adiabatic curve,[68] and the normal or defective operation of the exhaust valve, exhaust passages, and exhaust pipe. In fact, all the cycles of an explosive motor may be made a practical study from a close investigation of the lines of an indicator card.

A most unique card is that of the Diesel motor (Fig. 14), which involves a distinct principle in the design and operation of internal-combustion motors, in that instead of taking a mixed charge for instantaneous explosion, its charge primarily is of air and its compression to a pressure at which a temperature is attained above the igniting point of the fuel, then injecting the fuel under a still higher pressure by which spontaneous combustion takes place gradually with increasing volume over the compression for part of the stroke or until the fuel charge is consumed. The motor thus operating between the pressures of 500 and 35 lbs. per square inch, with a clearance of about 7 per cent., has given an efficiency of 36 per cent. of the total heat value of kerosene oil.

That the compression in a gas, gasoline, or oil-engine has a direct relation to the power obtained, has been long known to experienced builders, having been suggested by M. Beau de Rocha, in 1862, and afterward brought into practical use in the four-cycle or Otto type about 1880. The degree of compression has had a growth from zero, in the early engines, to the highest available due to the varying ignition temperatures of the different gases and vapors used for explosive fuel, in order to avoid premature explosion from the heat of compression. Much of the increased power for equal-cylinder capacity is due to compression of the charge from the fact that the most powerful explosion of gases, or of any form of explosive material, takes place when the particles are in the closest contact or cohesion with one another, less energy in this form being consumed by the ingredients themselves to bring about their chemical combination, and consequently[69] more energy is given out in useful or available work. This is best shown by the ignition of gunpowder, which, when ignited in the open air, burns rapidly, but without explosion, an explosion only taking place if the powder be confined or compressed into a small space.

In a gas or gasoline-motor with a small clearance or compression space—with high compression—the surface with which the burning gases come into contact is much smaller in comparison with the compression space in a low-compression motor. Another advantage of a high-compression motor is that on account of the smaller clearance of combustion space less cooling water is required than with a low-compression motor, as the temperature,[70] and consequently the pressure, falls more rapidly. The loss of heat through the water-jacket is thus less in the case of a high-compression than in that of a low-compression motor. In the non-compression type of motor the best results were obtained with a charge of 16 to 18 parts of gas and 100 parts of air, while in the compression type the best results are obtained with an explosive mixture of 7 to 10 parts of gas and 100 parts of air, thus showing that by the utilization of compression a weaker charge with a greater thermal efficiency is permissible.

It has been found that the explosive pressure resulting from the ignition of the charge of gas or gasoline-vapor and air in the gas-engine cylinder is about 4¹⁄₂ times the pressure prior to ignition. The difficulty about getting high compression is that if the pressure is too high the charge is likely to ignite prematurely, as compression always results in increased temperature. The cylinder may become too hot, a deposit of carbon, a projecting electrode or plug body in the cylinder may become incandescent and ignite the charge which has been excessively heated by the high compression and mixture of the hot gases of the previous explosion.

With gasoline-vapor and air the compression should not be raised above about 90 to 95 pounds to the square inch, many manufacturers not going above 65 or 70 pounds. For natural gas the compression pressure may easily be raised to from 85 to 100 pounds per square inch. For gases of low calorific value, such as blast-furnace or producer-gas, the compression may be increased to from 140 to 190 pounds. In fact the ability to raise the compression to a high point with these gases is one of the principal reasons for their successful adoption for gas-engine use. In kerosene injection engines the compression of 250 pounds per square inch has been used with marked economy. Many troubles in regard to loss of power and increase of fuel have occurred and will no doubt continue,[71] owing to the wear of valves, piston, and cylinder, which produces a loss in compression and explosive pressure and a waste of fuel by leakage. Faulty adjustment of valve movement is also a cause of loss of power; which may be from tardy closing of the inlet-valve or a too early opening of the exhaust-valve.

The explosive pressure varies to a considerable amount in proportion to the compression pressure by the difference in fuel value and the proportions of air mixtures, so that for good illuminating gas the explosive pressure may be from 2.5 to 4 times the compression pressure. For natural gas 3 to 4.5, for gasoline 3 to 5, for producer-gas 2 to 3, and for kerosene by injection 3 to 6.

The compression temperatures, although well known and easily computed from a known normal temperature of the explosive mixture, are subject to the effect of the uncertain temperature of the gases of the previous explosion remaining in the cylinder, the temperature of its walls, and the relative volume of the charge, whether full or scant; which are terms too variable to make any computations reliable or available.

For the theoretical compression temperatures from a known normal temperature, we append a table of the rise in temperature for the compression pressures in the following table:

A very useful chart (Fig. 16) for determining compression pressures in gasoline-engine cylinders for various ratios of compression space to total cylinder volume is given by P. S. Tice, and described in the Chilton Automobile Directory by the originator as follows:

[72]It is many times desirable to have at hand a convenient means for at once determining with accuracy what the compression pressure will be in a gasoline-engine cylinder, the relationship between the volume of the compression space and the total cylinder volume or that swept by the piston being known. The curve at Fig. 16 is offered as such a means. It is based on empirical data gathered from upward of two dozen modern automobile engines and represents what may be taken to be the results as found in practice. It is usual for the designer to find compression pressure values, knowing the volumes from the equation

which is for adiabatic compression of air. Equation (1) is right enough in general form but gives results which[73] are entirely too high, as almost all designers know from experience. The trouble lies in the interchange of heat between the compressed gases and the cylinder walls, in the diminution of the exponent (1.4 in the above) due to the lesser ratio of specific heat of gasoline vapor and in the transfer of heat from the gases which are being compressed to whatever fuel may enter the cylinder in an unvaporized condition. Also, there is always some piston leakage, and, if the form of the equation (1) is to be retained, this also tends to lower the value of the exponent. From experience with many engines, it appears that compression reaches its highest value in the cylinder for but a short range of motor speeds, usually during the mid-range. Also, it appears that, at those speeds at which compression shows its highest values, the initial pressure at the start of the compression stroke is from .5 to .9 lb. below atmospheric. Taking this latter loss value, which shows more often than those of lesser value, the compression is seen to start from an initial pressure of 13.9 lbs. per sq. in. absolute.

Also, experiment shows that if the exponent be given the value 1.26, instead of 1.4, the equation will embrace all heat losses in the compressed gas, and compensate for the changed ratio of specific heats for the mixture and also for all piston leakage, in the average engine with rings in good condition and tight. In the light of the foregoing, and in view of results obtained from its use, the above curve is offered—values of P₂ being found from the equation

In using this curve it must be remembered that pressures are absolute. Thus: suppose it is desired to know the volumetric relationships of the cylinder for a compression pressure of 75 lbs. gauge. Add atmospheric pressure to the desired gauge pressure 14.7 + 75 = 89.7 lbs. absolute. Locate this pressure on the scale of ordinates[74] and follow horizontally across to the curve and then vertically downward to the scale of abscissas, where the ratio of the combustion chamber volume to the total cylinder volume is given, which latter is equal to the sum of the combustion chamber volume and that of the piston sweep. In the above case it is found that the combustion space for a compression pressure of 75 lbs. gauge will be .225 of the total cylinder volume, or .225 ÷ .775 = .2905 of the piston sweep volume. Conversely, knowing the volumetric ratios, compression pressure can be read directly by proceeding from the scale of abscissas vertically to the curve and thence horizontally to the scale of ordinates.

The difference realized in the practical operation of an internal combustion heat engine from the computed effect derived from the values of the explosive elements is probably the most serious difficulty that engineers have encountered in their endeavors to arrive at a rational conclusion as to where the losses were located, and the ways and means of design that would eliminate the causes of loss and raise the efficiency step by step to a reasonable percentage of the total efficiency of a perfect cycle.

An authority on the relative condition of the chemical elements under combustion in closed cylinders attributes the variation of temperature shown in the fall of the expansion curve, and the suppression or retarded evolution of heat, entirely to the cooling action of the cylinder walls, and to this nearly all the phenomena hitherto obscure in the cylinder of a gas-engine. Others attribute the great difference between the theoretical temperature of combustion and the actual temperature realized in the practical operation of the gas-engine, a loss of more than one-half of the total heat energy of the combustibles, partly to the dissociation of the elements of combustion at extremely high temperatures and their reassociation by expansion in the cylinder, to account for the supposed continued[75] combustion and extra adiabatic curve of the expansion line on the indicator card.

The loss of heat to the walls of the cylinder, piston, and clearance space, as regards the proportion of wall surface to the volume, has gradually brought this point to its smallest ratio in the concave piston-head and globular cylinder-head, with the smallest possible space in the inlet and exhaust passage. The wall surface of a cylindrical clearance space or combustion chamber of one-half its unit diameter in length is equal to 3.1416 square units, its volume but 0.3927 of a cubic unit; while the same wall[76] surface in a spherical form has a volume of 0.5236 of a cubic unit. It will be readily seen that the volume is increased 33¹⁄₃ per cent. in a spherical over a cylindrical form for equal wall surfaces at the moment of explosion, when it is desirable that the greatest amount of heat is generated, and carrying with it the greatest possible pressure from which the expansion takes place by the movement of the piston.

The spherical form cannot continue during the stroke for mechanical reasons; therefore some proportion of piston stroke of cylinder volume must be found to correspond with a spherical form of the combustion chamber to produce the least loss of heat through the walls during the combustion and expansion part of the stroke. This idea is illustrated in Figs. 18 and 19, showing how the relative volumes of cylinder stroke and combustion chamber may be varied to suit the requirements due to the quality of the elements of combustion.

Although the concave piston-head shows economy in regard to the relation of the clearance volume to the wall area at the moment of explosive combustion, it may be clearly seen that its concavity increases its surface area and its capacity for absorbing heat, for which there is no provision for cooling the piston, save its contact with the walls of the cylinder and the slight air cooling of its back by its reciprocal motion. For this reason the concave piston-head has not been generally adopted and the concave cylinder-head, as shown in Fig. 19, with a flat[78] piston-head is the latest and best practice in airplane engine construction.

The practical application of the principle just outlined to one of the most efficient airplane motors ever designed, the Mercedes, is clearly outlined at Fig. 20.

The mean temperature of the wall surface of the combustion chamber and cylinder, as indicated by the temperatures of the circulating water, has been found to be an important item in the economy of the gas-engine. Dugald Clerk, in England, a high authority in practical work with the gas-engine, found that 10 per cent. of the gas for a stated amount of power was saved by using water at a temperature in which the ejected water from the cylinder-jacket was near the boiling-point, and ventures the opinion that a still higher temperature for the circulating water may be used as a source of economy. This could be made practical in the case of aviation engines by adjusting the air-cooling surface of the radiator so as to maintain the inlet water at just below the boiling point, and by the rapid circulation induced by the pump pressure, to return the water from the cylinder-jacket a few degrees above the boiling point. The thermal displacement systems of cooling employed in automobiles are working under more favorable temperature conditions than those engines in which cooling is more energetic.

For a given amount of heat taken from the cylinder by the largest volume of circulating water, the difference in temperature between inlet and outlet of the water-jacket should be the least possible, and this condition of the water circulation gives a more even temperature to all parts of the cylinder; while, on the contrary, a cold-water supply, say at 60° F., so slow as to allow the ejected water to flow off at a temperature near the boiling-point, must make a great difference in temperature between the bottom and top of the cylinder, with a loss in economy[79] in gas and other fuels, as well as in water, if it is obtained by measurement.

From the foregoing considerations of losses and inefficiencies, we find that the practice in motor design and construction has not yet reached the desired perfection in its cycular operation. Step by step improvements have been made with many changes in design though many have been without merit as an improvement, farther than to gratify the longings of designers for something different from the other thing, and to establish a special construction of their own. These efforts may in time produce a motor of normal or standard design for each kind of fuel that will give the highest possible efficiency for all conditions of service.

[80]

The principal elements of a gas engine are not difficult to understand and their functions are easily defined. In place of the barrel of the gun one has a smoothly machined cylinder in which a small cylindrical or barrel-shaped element fitting the bore closely may be likened to a bullet or cannon ball. It differs in this important respect, however, as while the shot is discharged from the mouth of the cannon the piston member sliding inside of the main cylinder cannot leave it, as its movements back and forth from the open to the closed end and back again are limited by simple mechanical connection or linkage which comprises crank and connection rod. It is by this means that the reciprocating movement of the piston is transformed into a rotary motion of the crank-shaft.

The fly-wheel is a heavy member attached to the crank-shaft of an automobile engine which has energy stored in its rim as the member revolves, and the momentum of this revolving mass tends to equalize the intermittent pushes on the piston head produced by the explosion of the gas in the cylinder. In aviation engines, the weight of the propeller or that of rotating cylinders themselves performs the duty of a fly-wheel, so no separate member is needed. If some explosive is placed in the chamber formed by the piston and closed end of the cylinder and exploded, the piston would be the only part that would yield to the pressure which would produce a downward movement. As this is forced down the crank-shaft is[82] turned by the connecting rod, and as this part is hinged at both ends it is free to oscillate as the crank turns, and thus the piston may slide back and forth while the crank-shaft is rotating or describing a curvilinear path.

In addition to the simple elements described it is evident that a gasoline engine must have other parts. The most important of these are the valves, of which there are generally two to each cylinder. One closes the passage connecting to the gas supply and opens during one stroke of the piston in order to let the explosive gas into the combustion chamber. The other member, or exhaust valve, serves as a cover for the opening through which the burned gases can leave the cylinder after their work is done. The spark plug is a simple device which may be compared to the fuse or percussion cap of the cannon. It permits one to produce an electric spark in the cylinder when the piston is at the best point to utilize the pressure which obtains when the compressed gas is fired. The valves are open one at a time, the inlet valve being lifted from its seat while the cylinder is filling and the exhaust valve is opened when the cylinder is being cleared. They are normally kept seated by means of compression springs. In the simple motor shown at Fig. 5, the exhaust valve is operated by means of a pivoted bell crank rocked by a cam which turns at half the speed of the crank-shaft. The inlet valve operates automatically, as will be explained in proper sequence.

In order to obtain a perfectly tight combustion chamber, both intake and exhaust valves are closed before the gas is ignited, because all of the pressure produced by the exploding gas is to be directed against the top of the movable piston. When the piston reaches the bottom of its power stroke, the exhaust valve is lifted by means of the bell crank which is rocked because of the point or lift on the cam. The cam-shaft is driven by positive gearing and revolves at half the engine speed. The exhaust valve remains open during the whole of the return stroke of the piston, and as this member moves toward[83] the closed end of the cylinder it forces out burned gases ahead of it, through the passage controlled by the exhaust valve. The cam-shaft is revolved at half the engine speed because the exhaust valve is raised from its seat during only one stroke out of four, or only once every two revolutions. Obviously, if the cam was turned at the same speed as the crank-shaft it would remain open once every revolution, whereas the burned gases are expelled from the individual cylinders only once in two turns of the crank-shaft.

Owing to the vibration which obtains from the heavy explosion in the large single-cylinder engines used for stationary power other forms were evolved in which the cylinder was smaller and power obtained by running the engine faster, but these are suitable only for very low powers.

When a single-cylinder engine is employed a very heavy fly-wheel is needed to carry the moving parts through idle strokes necessary to obtain a power impulse. For this reason automobile and aircraft designers must use more than one cylinder, and the tendency is to produce power by frequently occurring light impulses rather than by a smaller number of explosions having greater force. When a single-cylinder motor is employed the construction is heavier than is needed with a multiple-cylinder form. Using two or more cylinders conduces to steady power generation and a lessening of vibration. Most modern motor cars employ four-cylinder engines because a power impulse may be secured twice every revolution of the crank-shaft, or a total of four power strokes during two revolutions. The parts are so arranged that while the charge of gas in one cylinder is exploding, those which come next in firing order are compressing, discharging the inert gases and drawing in a fresh charge respectively. When the power stroke is completed in one cylinder, the piston in that member in[84] which a charge of gas has just been compressed has reached the top of its stroke and when the gas is exploded the piston is reciprocated and keeps the crank-shaft turning. When a multiple-cylinder engine is used the fly-wheel can be made much lighter than that of the simpler form and eliminated altogether in some designs. In fact, many modern multiple-cylinder engines developing 300 horse-power weigh less than the early single- and double-cylinder forms which developed but one-tenth or one-twentieth that amount of energy.

Referring to Fig. 22, A, the sequence of operation in a single-cylinder motor can be easily understood. Assuming that the crank-shaft is turning in the direction of the arrow, it will be seen that the intake stroke comes first, then the compression, which is followed by the power impulse, and lastly the exhaust stroke. If two cylinders are used, it is possible to balance the explosions in such a way that one will occur each revolution. This is true with either one of two forms of four-cycle motors. At B, a two-cylinder vertical engine using a crank-shaft in which the crank-pins are on the same plane is shown. The two pistons move up and down simultaneously. Referring to the diagram describing the strokes, and assuming that the outer circle represents the cycle of operations in one cylinder while the inner circle represents the sequence of events in the other cylinder, while cylinder No. 1 is taking in a fresh charge of gas, cylinder No. 2 is exploding. When cylinder No. 1 is compressing, cylinder No. 2 is exhausting. During the time that the charge in cylinder No. 1 is exploded, cylinder No. 2 is being filled with fresh gas. While the exhaust gases are being discharged from cylinder No. 1, cylinder No. 2 is compressing the gas previously taken.

[85]

The same condition obtains when the crank-pins are arranged at one hundred and eighty degrees and the cylinders are opposed, as shown at C. The reason that the[86] two-cylinder opposed motor is more popular than that having two vertical cylinders is that it is difficult to balance the construction shown at B, so that the vibration will not be excessive. The two-cylinder opposed motor has much less vibration than the other form, and as the explosions occur evenly and the motor is a simple one to construct, it has been very popular in the past on light cars and has received limited application on some early, light airplanes.

[87]

To demonstrate very clearly the advantages of multiple-cylinder engines the diagrams at Fig. 23 have been prepared. At A, a three-cylinder motor, having crank-pins at one hundred and twenty degrees, which means that they are spaced at thirds of the circle, we have a form of construction that gives a more even turning than that possible with a two-cylinder engine. Instead of one explosion per revolution of the crank-shaft, one will obtain three explosions in two revolutions. The manner in which the explosion strokes occur and the manner they overlap strokes in the other cylinder is shown at A. Assuming that the cylinders fire in the following order, first No. 1, then No. 2, and last No. 3, we will see that while cylinder No. 1, represented by the outer circle, is on the power stroke, cylinder No. 3 has completed the last two-thirds of its exhaust stroke and has started on its intake stroke. Cylinder No. 2, represented by the middle circle, during this same period has completed its intake stroke and two-thirds of its compression stroke. A study of the diagram will show that there is an appreciable lapse of time between each explosion.

Three-cylinder engines are not used on aircraft at the present time, though Bleriot’s flight across the British Channel was made with a three-cylinder Anzani motor. It was not a conventional form, however. The three-cylinder engine is practically obsolete at this time for any purpose except “penguins” or school machines that are incapable of flight and which are used in some French training schools for aviators.

[88]

In the four-cylinder engine operation which is shown at Fig. 23, B, it will be seen that the power strokes follow each other without loss of time, and one cylinder begins to fire and the piston moves down just as soon as the member ahead of it has completed its power stroke. In a four-cylinder motor, the crank-pins are placed at one hundred and eighty degrees, or on the halves of the crank circle. The crank-pins for cylinders No. 1 and No. 4 are on the same plane, while those for cylinders No. 2 and No. 3 also move in unison. The diagram describing sequence of operations in each cylinder is based on a firing order of one, two, four, three. The outer circle, as in previous instances, represents the cycle of operations in cylinder one. The next one toward the center, cylinder No. 2, the third circle represents the sequence of events in cylinder No. 3, while the inner circle outlines the strokes in cylinder four. The various cylinders are working as follows:

It will be obvious that regardless of the method of construction, or the number of cylinders employed, exactly the same number of parts must be used in each cylinder assembly and one can conveniently compare any multiple-cylinder power plant as a series of single-cylinder engines joined one behind the other and so coupled that one will deliver power and produce useful energy at the crank-shaft where the other leaves off. The same fundamental laws governing the action of a single cylinder obtain when a number are employed, and the sequence of operation is the same in all members, except that the necessary functions take place at different[89] times. If, for instance, all the cylinders of a four-cylinder motor were fired at the same time, one would obtain the same effect as though a one-piston engine was used, which had a piston displacement equal to that of the four smaller members. As is the case with a single-cylinder engine, the motor would be out of correct mechanical balance because all the connecting rods would be placed on crank-pins that lie in the same plane. A very large fly-wheel would be necessary to carry the piston through the idle strokes, and large balance weights would be fitted to the crank-shaft in an effort to compensate for the weight of the four pistons, and thus reduce vibratory stresses which obtain when parts are not in correct balance.

There would be no advantage gained by using four cylinders in this manner, and there would be more loss of heat and more power consumed in friction than in a one-piston motor of the same capacity. This is the reason that when four cylinders are used the arrangement of crank-pins is always as shown at Fig. 23, B—i.e., two pistons are up, while the other two are at the bottom of the stroke. With this construction, we have seen that it is possible to string out the explosions so that there will always be one cylinder applying power to the crank-shaft. The explosions are spaced equally. The parts are in correct mechanical balance because two pistons are on the upstroke while the other two are descending. Care is taken to have one set of moving members weigh exactly the same as the other. With a four-cylinder engine one has correct balance and continuous application of energy. This insures a smoother running motor which has greater efficiency than the simpler one-, two-, and three-cylinder forms previously described. Eliminating the stresses which would obtain if we had an unbalanced mechanism and irregular power application makes for longer life. Obviously a large number of relatively light explosions will produce less wear and strain than would a lesser number of powerful ones. As the parts can be built lighter if the explosions are not heavy, the engine can be operated[90] at higher rotative speeds than when large and cumbersome members are utilized. Four-cylinder engines intended for aviation work have been built according to the designs shown at Fig. 24, but these forms are unconventional and seldom if ever used.

The six-cylinder type of motor, the action of which is shown at Fig. 23, C, is superior to the four-cylinder, inasmuch as the power strokes overlap, and instead of having two explosions each revolution we have three explosions. The conventional crank-shaft arrangement in a six-cylinder engine is just the same as though one used two three-cylinder shafts fastened together, so pistons 1 and 6 are on the same plane as are pistons 2 and 5. Pistons 3 and 4 also travel together. With the cranks arranged as outlined at Fig. 23, C, the firing order is one, five, three, six, two, four. The manner in which the power strokes overlap is clearly shown in the diagram. An interesting comparison[91] is also made in the diagrams at Fig. 25 and in the upper corner of Fig. 23, C.

A rectangle is divided into four columns; each of these corresponds to one hundred and eighty degrees, or half a revolution. Thus the first revolution of the crank-shaft is represented by the first two columns, while the second revolution is represented by the last two. Taking the portion of the diagram which shows the power impulse in a one-cylinder engine, we see that during the first revolution there has been no power impulse. During the first half of the second revolution, however, an explosion takes place and a power impulse is obtained. The last portion of the second revolution is devoted to exhausting the burned gases, so that there are three idle strokes and but one power stroke. The effect when two cylinders are employed is shown immediately below.

[92]Here we have one explosion during the first half of the first revolution in one cylinder and another during the first half of the second revolution in the other cylinder. With a four-cylinder engine there is an explosion each half revolution, while in a six-cylinder engine there is one and one-half explosions during each half revolution. When six cylinders are used there is no lapse of time between power impulses, as these overlap and a continuous and smooth-turning movement is imparted to the crank shaft. The diagram shown at Fig. 26, prepared by E. P. Pulley, can be studied to advantage in securing an idea of the coordination of effort that takes place in an engine of the six-cylinder type.

[93]

In the diagrams previously presented the writer has assumed, for the sake of simplicity, that each stroke takes place during half of one revolution of the crank-shaft, which corresponds to a crank-pin travel of one hundred and eighty degrees. The actual duration of these strokes is somewhat different. For example, the inlet stroke is usually a trifle more than a half revolution, and the exhaust is always considerably more. The diagram showing the comparative duration of the strokes is shown at Fig. 27.[94] The inlet valve opens ten degrees after the piston starts to go down and remains open thirty degrees after the piston has reached the bottom of its stroke. This means that the suction stroke corresponds to a crank-pin travel of two hundred degrees, while the compression stroke is measured by a movement of but one hundred and fifty degrees. It is common practice to open the exhaust valve before the piston reaches the end of the power stroke so that the actual duration of the power stroke is about one hundred and forty degrees, while the exhaust stroke corresponds to a crank-pin travel of two hundred and twenty-five degrees. In this diagram, which represents proper time for the valves to open and close, the dimensions in inches given are measured on the fly-wheel and apply only to a certain automobile motor. If the fly-wheel were smaller ten degrees would take up less than the dimensions given, while if the fly-wheel was larger a greater space on its circumference would represent the same crank-pin travel. Aviation engines are timed by using a timing disc attached to the crank-shaft as they are not provided with fly-wheels. Obviously, the distance measured in inches will depend upon the diameter of the disc, though the number of degrees interval would not change.

Those who have followed the development of the gasoline engine will recall the arguments that were made when the six-cylinder motor was introduced at a time that the[95] four-cylinder type was considered standard. The arrival of the eight-cylinder has created similar futile discussion of its practicability as this is so clearly established as to be accepted without question. It has been a standard power plant for aeroplanes for many years, early exponents having been the Antoinette, the Woolsley, the Renault, the E. N. V. in Europe and the Curtiss in the United States.

The reason the V type shown at Fig. 29, A is favored is that the “all-in-line form” which is shown at Fig. 29, B is not practical for aircraft because of its length. Compared to the standard four-cylinder engine it is nearly twice as long and it required a much stronger and longer crank-shaft. It will be evident that it could not be located to advantage in the airplane fuselage. These undesirable factors are eliminated in the V type eight-cylinder motor, as it consists of two blocks of four cylinders each, so arranged that one set or block is at an angle of forty-five degrees from the vertical center line of the motor, or at an angle of ninety degrees with the other set. This arrangement of cylinders produces a motor that is no[96] longer than a four-cylinder engine of half the power would be.

Apparently there is considerable misconception as to the advantage of the two extra cylinders of the eight as compared with the six-cylinder. It should be borne in mind that the multiplication in the number of cylinders noticed since the early days of automobile development has not been for solely increasing the power of the engine, but to secure a more even turning movement, greater flexibility and to eliminate destructive vibration. The ideal internal combustion motor, is the one having the most uniform turning movement with the least mechanical friction loss. Study of the torque outlines or plotted graphics shown at Figs. 25 and 30 will show how multiplication of cylinders will produce steady power delivery due to overlapping impulses. The most practical form would be that which more nearly conforms to the steady running produced by a steam turbine or electric motor. The advocates of the eight-cylinder engine bring up the item of uniform torque[97] as one of the most important advantages of the eight-cylinder design. A number of torque diagrams are shown at Fig. 30. While these appear to be deeply technical, they may be very easily followed when their purpose is explained. At the top is shown the torque diagram of a single-cylinder motor of the four-cycle type. The high point in the line represents the period of greatest torque or power generation, and it will be evident that this occurs early in the first revolution of the crank-shaft. Below this diagram is shown a similar curve except that it is produced by a four-cylinder engine. Inspection will show that the turning-moment is much more uniform than in the[98] single cylinder; similarly, the six-cylinder diagram is an improvement over the four, and the eight-cylinder diagram is an improvement over the six-cylinder.

The reason that practically continuous torque is obtained in an eight-cylinder engine is that one cylinder fires every ninety degrees of crank-shaft rotation, and as each impulse lasts nearly seventy-five per cent. of the stroke, one can easily appreciate that an engine that will give four explosions per revolution of the crank-shaft will run more uniformly than one that gives but three explosions per revolution, as the six-cylinder does, and will be twice as smooth running as a four-cylinder, in which but two explosions occur per revolution of the crank-shaft. The comparison is so clearly shown in graphical diagrams and in Fig. 31 that further description is unnecessary.

Any eight-cylinder engine may be considered a “twin-four,” twelve-cylinder engines may be considered “twin sixes.”

[101]

The only points in which an eight-cylinder motor differs from a four-cylinder is in the arrangement of the connecting rod, as in many designs it is necessary to have two rods working from the same crank-pin. This difficulty is easily overcome in some designs by staggering the cylinders and having the two connecting rod big ends of conventional form side by side on a common crank-pin. In other designs one rod is a forked form and works on the outside of a rod of the regular pattern. Still another method is to have a boss just above the main bearing on one connecting rod to which the lower portion of the connecting rod in the opposite cylinder is hinged. As the eight-cylinder engine may actually be made lighter than the six-cylinder of equal power, it is possible to use smaller reciprocating parts, such as pistons, connecting rods and valve gear, and obtain higher engine speed with practically no vibration. The firing order in nearly every case is the same as in a four-cylinder except that the explosions occur alternately in each set of cylinders. The firing order of an eight-cylinder motor is apt to be confusing to the[100] motorist, especially if one considers that there are eight possible sequences. The majority of engineers favor the alternate firing from side to side. Firing orders will be considered in proper sequence.

The demand of aircraft designers for more power has stimulated designers to work out twelve-cylinder motors. These are high-speed motors incorporating all recent features of design in securing light reciprocating parts, large valve openings, etc. The twelve-cylinder motor incorporates the best features of high-speed motor design and there is no need at this time to discuss further the pros and cons of the twelve-cylinder versus the eight or six, because it is conceded by all that there is the same degree of steady power application in the twelve over the eight as there would be in the eight over the six. The question resolves[102] itself into having a motor of high power that will run with minimum vibration and that produces smooth action. This is well shown by diagrams at Fig. 31. It should be remembered that if an eight-cylinder engine will give four explosions per revolution of the fly-wheel, a twelve-cylinder type will give six explosions per revolution, and instead of the impulses coming 90 degrees crank travel apart, as in the case of the eight-cylinder, these will come but 60 degrees of crank travel apart in the case of the twelve-cylinder. For this reason, the cylinders of a twelve are usually separated by 60 degrees while the eight has the blocks spaced 90 degrees apart. The comparison can be easily made by comparing the sectional views of Vee engines at Fig. 32. When one realizes that the actual duration of the power stroke is considerably greater than 120 degrees crank travel, it will be apparent that the overlapping of explosions must deliver a very uniform application of power. Vee engines have been devised[103] having the cylinders spaced but 45 degrees apart, but the explosions cannot be timed at equal intervals as when 90 degrees separate the cylinder center lines.

While the fixed cylinder forms of engines, having the cylinders in tandem in the four- and six-cylinder models as shown at Figs. 33 to 35 inclusive and the eight-cylinder V types as outlined at Figs. 36 and 37 have been generally used and are most in favor at the present time, other forms of motors having unconventional cylinder arrangements have been devised, though most of these are practically[104] obsolete. While many methods of decreasing weight and increasing mechanical efficiency of a motor are known to designers, one of the first to be applied to the construction of aeronautical power plants was an endeavor to group the components, which in themselves were not extremely light, into a form that would be considerably lighter than the conventional design. As an example, we may consider those multiple-cylinder forms in which the cylinders are disposed around a short crank-case, either radiating from a common center as at Fig. 38 or of the fan shape shown at Fig. 39. This makes it possible to use a crank-case but slightly larger than that needed for one or two cylinders and it also permits of a corresponding decrease in length of the crank-shaft. The weight of the engine is lessened because of the reduction in crank-shaft and crank-case weight and the elimination of a number of intermediate bearings and their supporting webs which would be necessary with the usual tandem construction. While there are six power impulses to every two revolutions of the crank-shaft,[105] in the six-cylinder engine, they are not evenly spaced as is possible with the conventional arrangement.

In the Anzani form, which is shown at Fig. 38, the crank-case is stationary and a revolving crank-shaft is employed as in conventional construction. The cylinders are five in number and the engine develops 40 to 50 H.P. with a weight of 72 kilograms or 158.4 lbs. The cylinders are of the usual air-cooled form having cooling flanges only part of the way down the cylinder. By using five cylinders it is possible to have the power impulses come regularly, they coming 145° crank-shaft travel apart, the crank-shaft making two turns to every five explosions. The balance is good and power output regular. The valves are[106] placed directly in the cylinder head and are operated by a common pushrod. Attention is directed to the novel method of installing the carburetor which supplies the mixture to the engine base from which inlet pipes radiate to the various cylinders. This engine is used on French school machines.

In the form shown at Fig. 39 six cylinders are used, all being placed above the crank-shaft center line. This engine is also of the air-cooled form and develops 50 H. P. and weighs 105 kilograms, or 231 lbs. The carburetor is connected to a manifold casting attached to the engine base from which the induction pipes radiate to the various cylinders. The propeller design and size relative to the engine is clearly shown in this view. While flights have been made with both of the engines described, this method of construction is not generally followed and has been almost entirely displaced abroad by the revolving motors or by the more conventional eight-cylinder V engines. Both of the engines shown were designed about eight years[107] ago and would be entirely too small and weak for use in modern airplanes intended for active duty.

Rotary engines such as shown at Fig. 40 are generally associated with the idea of light construction and it is rather an interesting point that is often overlooked in connection with the application of this idea to flight motors, that the reason why rotary engines are popularly supposed to be lighter than the others is because they form their own fly-wheel, yet on aeroplanes, engines are seldom fitted with a fly-wheel at all. As a matter of fact the[108] Gnome engine is not so light because it is a rotary motor, and it is a rotary motor because the design that has been adopted as that most conducive to lightness is also most suited to an engine working in this way. The cylinders could be fixed and crank-shaft revolve without increasing the weight to any extent. There are two prime factors governing the lightness of an engine, one being the initial design, and the other the quality of the materials employed. The consideration of reducing weight by cutting away metal is a subsidiary method that ought not to play a part in standard practice, however useful it may be in special cases. In the Gnome rotary engine the lightness is entirely due to the initial design and to the materials employed in manufacture. Thus, in the first case, the engine is a radial engine, and has its seven or nine cylinders spaced equally around a crank-chamber that is no wider or rather longer than would be required for any one of the cylinders. This shortening of the crank-chamber not only effects a considerable saving of weight on its own account, but there is a corresponding saving in the shafts and other members, the dimensions of which are governed by the size of the crank-chamber. With regard to materials, nothing but steel is used throughout, and most of the metal is forged chrome nickel steel. The beautifully steady running of the engine is largely due to the fact that there are literally no reciprocating parts in the absolute sense, the apparent reciprocation between the pistons and cylinders being solely a relative reciprocation since both travel in circular paths, that of the pistons, however, being electric by one-half of the stroke length to that of the cylinder.

While the Gnome engine has many advantages, on the other hand the head resistance offered by a motor of this type is considerable; there is a large waste of lubricating oil due to the centrifugal force which tends to throw the oil away from the cylinders; the gyroscopic effect of the rotary motor is detrimental to the best working of the[109] aeroplane, and moreover it requires about seven per cent. of the total power developed by the motor to drive the revolving cylinders around the shaft. Of necessity, the compression of this type of motor is rather low, and an additional disadvantage manifests itself in the fact that there is as yet no satisfactory way of muffling the rotary type of motor. The modern Gnome engine has been widely copied in various European countries, but its design was originated in America, the early Adams-Farwell engine being the pioneer form. It has been made in seven- and nine-cylinder types and forms of double these numbers. The engine illustrated at Fig. 40 is a fourteen-cylinder form. The simple engines have an odd number of cylinders in order to secure evenly spaced explosions. In the seven-cylinder, the impulses come 102.8° apart. In the nine-cylinder form, the power strokes are spaced 80° apart. The fourteen-cylinder engine is virtually two seven-cylinder types mounted together, the cranks being just the same as in a double cylinder opposed motor, the explosions coming 51.4° apart; while in the eighteen-cylinder model the power impulses come every 40° cylinder travel. Other rotary motors have been devised, such as the Le Rhone and the Clerget in France and several German copies of these various types. The mechanical features of these motors will be fully considered later.

[110]

There is no appliance that has more material value upon the efficiency of the internal combustion motor than the carburetor or vaporizer which supplies the explosive gas to the cylinders. It is only in recent years that engineers have realized the importance of using carburetors that are efficient and that are so strongly and simply made that there will be little liability of derangement. As the power obtained from the gas-engine depends upon the combustion of fuel in the cylinders, it is evident that if the gas supplied does not have the proper proportions of elements to insure rapid combustion the efficiency of the engine will be low. When a gas engine is used as a stationary installation it is possible to use ordinary illuminating or natural gas for fuel, but when this prime mover is applied to automobiles or airplanes it is evident that considerable difficulty would be experienced in carrying enough compressed coal gas to supply the engine for even a very short trip. Fortunately, the development of the internal-combustion motor was not delayed by the lack of suitable fuel.

Engineers were familiar with the properties of certain[111] liquids which gave off vapors that could be mixed with air to form an explosive gas which burned very well in the engine cylinders. A very small quantity of such liquids would suffice for a very satisfactory period of operation. The problem to be solved before these liquids could be applied in a practical manner was to evolve suitable apparatus for vaporizing them without waste. Among the liquids that can be combined with air and burned, gasoline is the most volatile and is the fuel utilized by internal-combustion engines.

The widely increasing scope of usefulness of the internal-combustion motor has made it imperative that other fuels be applied in some instances because the supply of gasoline may in time become inadequate to supply the demand. In fact, abroad this fuel sells for fifty to two hundred per cent. more than it does in America because most of the gasoline used must be imported from this country or Russia. Because of this foreign engineers have experimented widely with other substances, such as alcohol, benzol, and kerosene, but more to determine if they can be used to advantage in motor cars than in airplane engines.

Crude petroleum is found in small quantities in almost all parts of the world, but a large portion of that produced commercially is derived from American wells. The petroleum obtained in this country yields more of the volatile products than those of foreign production, and for that reason the demand for it is greater. The oil fields of this country are found in Pennsylvania, Indiana, and Ohio, and the crude petroleum is usually in association with natural gas. This mineral oil is an agent from which many compounds and products are derived, and the products will vary from heavy sludges, such as asphalt, to the lighter and more volatile components, some of which will evaporate very easily at ordinary temperatures.

The compounds derived from crude petroleum are composed[112] principally of hydrogen and carbon and are termed “Hydrocarbons.” In the crude product one finds many impurities, such as free carbon, sulphur, and various earthy elements. Before the oil can be utilized it must be subjected to a process of purifying which is known as refining, and it is during this process, which is one of destructive distillation, that the various liquids are separated. The oil was formerly broken up into three main groups of products as follows: Highly volatile, naphtha, benzine, gasoline, eight to ten per cent. Light oils, such as kerosene and light lubricating oils seventy to eighty per cent. Heavy oils or residuum five to nine per cent. From the foregoing it will be seen that the available supply of gasoline is determined largely by the demand existing for the light oils forming the larger part of the products derived from crude petroleum. New processes have been recently discovered by which the lighter oils, such as kerosene, are reduced in proportion and that of gasoline increased, though the resulting liquid is neither the high grade, volatile gasoline known in the early days of motoring nor the low grade kerosene.

The process of carburetion is combining the volatile vapors which evaporate from the hydrocarbon liquids with certain proportions of air to form an inflammable gas. The quantities of air needed vary with different liquids and some mixtures burn quicker than do other combinations of air and vapor. Combustion is simply burning and it may be rapid, moderate or slow. Mixtures of gasoline and air burn quickly, in fact the combustion is so rapid that it is almost instantaneous and we obtain what is commonly termed an “explosion.” Therefore the explosion of gas in the automobile engine cylinder which produces the power is really a combination of chemical elements which produce heat and an increase in the volume of the gas because of the increase in temperature.

If the gasoline mixture is not properly proportioned[113] the rate of burning will vary, and if the mixture is either too rich or too weak the power of the explosion is reduced and the amount of power applied to the piston is decreased proportionately. In determining the proper proportions of gasoline and air, one must take the chemical composition of gasoline into account. The ordinary liquid used for fuel is said to contain about eight-four per cent. carbon and sixteen per cent. hydrogen. Air is composed of oxygen and nitrogen and the former has a great affinity, or combining power, with the two constituents of hydro-carbon liquids. Therefore, what we call an explosion is merely an indication that oxygen in the air has combined with the carbon and hydrogen of the gasoline.

In figuring the proper volume of air to mix with a given quantity of fuel, one takes into account the fact that one pound of hydrogen requires eight pounds of oxygen to burn it, and one pound of carbon needs two and one-third pounds of oxygen to insure its combustion. Air is composed of one part of oxygen to three and one-half portions of nitrogen by weight. Therefore for each pound of oxygen one needs to burn hydrogen or carbon four and one-half pounds of air must be allowed. To insure combustion of one pound of gasoline which is composed of hydrogen and carbon we must furnish about ten pounds of air to burn the carbon and about six pounds of air to insure combustion of hydrogen, the other component of gasoline. This means that to burn one pound of gasoline one must provide about sixteen pounds of air.

While one does not usually consider air as having much weight, at a temperature of sixty-two degrees Fahrenheit about fourteen cubic feet of air will weigh a pound, and to burn a pound of gasoline one would require about two hundred cubic feet of air. This amount will provide for combustion theoretically, but it is common practice to allow twice this amount because the element nitrogen, which is the main constituent of air, is an inert gas and[114] instead of aiding combustion it acts as a deterrent of burning. In order to be explosive, gasoline vapor must be combined with definite quantities of air. Mixtures that are rich in gasoline ignite quicker than those which have more air, but these are only suitable when starting or when running slowly, as a rich mixture ignites much quicker than a weak mixture. The richer mixture of gasoline and air not only burns quicker but produces the most heat and the most effective pressure in pounds per square inch of piston top area.

The amount of compression of the charge before ignition also has material bearing on the force of the explosion. The higher the degree of compression the greater the force exerted by the rapid combustion of the gas. It may be stated that as a general thing the maximum explosive pressure is somewhat more than four times the compression pressure prior to ignition. A charge compressed to sixty pounds will have a maximum of approximately two hundred and forty pounds; compacted to eighty pounds it will produce a pressure of about three hundred pounds on each square inch of piston area at the beginning of the power stroke. Mixtures varying from one part of gasoline vapor to four of air to others having one part of gasoline vapor to thirteen of air can be ignited, but the best results are obtained when the proportions are one to five or one to seven, as this mixture is said to be the one that will produce the highest temperature, the quickest explosion, and the most pressure.

While it is apparent that the chief function of a carbureting device is to mix hydrocarbon vapors with air to secure mixtures that will burn, there are a number of factors which must be considered before describing the principles of vaporizing devices. Almost any device which permits a current of air to pass over or through a volatile liquid will produce a gas which will explode when[116][115] compressed and ignited in the motor cylinder. Modern carburetors are not only called upon to supply certain quantities of gas, but these must deliver a mixture to the cylinders that is accurately proportioned and which will be of proper composition at all engine speeds.

Flexible control of the engine is sought by varying the engine speed by regulating the supply of gas to the cylinders. The power plant should run from its lowest to its highest speed without any irregularity in torque, i.e., the acceleration should be gradual rather than spasmodic. As the degree of compression will vary in value with the amount of throttle opening, the conditions necessary to obtain maximum power differ with varying engine speeds. When the throttle is barely opened the engine speed is low and the gas must be richer in fuel than when the throttle is wide open and the engine speed high.

When an engine is turning over slowly the compression has low value and the conditions are not so favorable to rapid combustion as when the compression is high. At high engine speeds the gas velocity through the intake piping is higher than at low speeds, and regular engine action is not so apt to be disturbed by condensation of liquid fuel in the manifold due to excessively rich mixture or a superabundance of liquid in the stream of carbureted air.

The problem of gasoline storage and method of supplying the carburetor is one that is determined solely by design of the airplane. While the object of designers should be to supply the fuel to the carburetor by as simple means as possible the fuel supply system of some airplanes is quite complex. The first point to consider is the location of the gasoline tank. This depends upon the amount of fuel needed and the space available in the fuselage.

A very simple and compact fuel supply system is shown at Fig. 41. In this instance the fuel container is placed immediately back of the engine cylinder. The carburetor[117] which is carried as indicated is joined to the tank by a short piece of copper or flexible rubber tubing. This is the simplest possible form of fuel supply system and one used on a number of excellent airplanes.

As the sizes of engines increase and the power plant fuel consumption augments it is necessary to use more fuel, and to obtain a satisfactory flying radius without frequent landings for filling the fuel tank it is necessary to supply large containers.

When a very powerful power plant is fitted, as on battle planes of high capacity, it is necessary to carry large quantities of gasoline. In order to use a tank of sufficiently large capacity it may be necessary to carry it lower than the carburetor. When installed in this manner it is necessary to force fuel out of the tank by air pressure or to pump it with a vacuum tank because the gasoline tank is lower than the carburetor it supplies and the gasoline cannot flow by gravity as in the simpler systems. While the pressure and gravity feed systems are generally used in airplanes, it may be well to describe the vacuum lift system which has been widely applied to motor cars and which may have some use in connection with airplanes as these machines are developed.

One of the marked tendencies has been the adoption of a vacuum fuel feed system to draw the gasoline from tanks placed lower than the carburetor instead of using either exhaust gas or air pressure to achieve this end. The device generally fitted is the Stewart vacuum feed tank which is clearly shown in section at Fig. 42. In this system the suction of a motor is employed to draw gasoline from the main fuel tank to the auxiliary tank incorporated in the device and from this tank the liquid flows to the carburetor. It is claimed that all the advantages of the pressure system are obtained with very little more complication than is found on the ordinary gravity feed. The mechanism is all contained in the cylindrical tank shown,[118] which may be mounted either on the front of the dash or on the side of the engine as shown.

The tank is divided into two chambers, the upper one being the filling chamber and the lower one the emptying chamber. The former, which is at the top of the device, contains the float valve, as well as the pipes running to the main fuel container and to the intake manifold. The lower chamber is used to supply the carburetor with gasoline and is under atmospheric pressure at all times, so the flow of fuel from it is by means of gravity only. Since[119] this chamber is located somewhat above the carburetor, there must always be free flow of fuel. Atmospheric pressure is maintained by the pipes A and B, the latter opening into the air. In order that the fuel will be sucked from a main tank to the upper chamber, the suction valve must be opened and the atmospheric valve closed. Under these conditions the float is at the bottom and the suction at the intake manifold produces a vacuum in the tank which draws the gasoline from the main tank to the upper chamber. When the upper chamber is filled at the proper height the float rises to the top, this closing the suction valve and opening the atmospheric valve. As the suction is now cut off, the lower chamber is filled by gravity owing to there being atmospheric pressure in both upper and lower chambers. A flap valve is provided between the two chambers to prevent the gasoline in the lower one from being sucked back into the upper one. The atmospheric and suction valves are controlled by the levers C and D, both of which are pivoted at E, their outer ends being connected by two coil springs. It is seen that the arrangement of these two springs is such that the float must be held at the extremity of its movement, and that it cannot assume an intermediate position.

This intermittent action is required to insure that the upper part of the tank may be under atmospheric pressure part of the time for the gasoline to flow to the lower chamber. When the level of gasoline drops to a certain point, the float falls, thus opening the suction valve and closing the atmospheric valve. The suction of the motor then causes a flow of fuel from the main container. As soon as the level rises to the proper height the float returns to its upper position. It takes about two seconds for the chamber to become full enough to raise the float, as but .05 gallon is transferred at a time. The pipe running from the bottom of the lower chamber to the carburetor extends up a ways, so that there is but little chance of dirt or water being carried to the float chamber.

If the engine is allowed to stand long enough so that the[120] tank becomes empty, it will be replenished after the motor has been cranked over four or five times with the throttle closed. The installation of the Stewart Vacuum-Gravity System is very simple. The suction pipe is tapped into the manifold at a point as near the cylinders as possible, while the fuel pipe is inserted into the gasoline tank and runs to the bottom of that member. There is a screen at the end of the fuel pipe to prevent any trouble due to deposits of sediment in the main container. As the fuel is sucked from the gasoline tank a small vent must be made in the tank filler cap so that the pressure in the main tank will always be that of the atmosphere.

The early types of carbureting devices were very crude and cumbersome, and the mixture of gasoline vapor and air was accomplished in three ways. The air stream was passed over the surface of the liquid itself, through loosely placed absorbent material saturated with liquid, or directly through the fuel. The first type is known as the surface carburetor and is now practically obsolete. The second form is called the “wick” carburetor because the air stream was passed over or through saturated wicking. The third form was known as a “bubbling” carburetor. While these primitive forms gave fairly good results with the early slow-speed engines and the high grade, or very volatile, gasoline which was first used for fuel, they would be entirely unsuitable for present forms of engines because they would not carburate the lower grades of gasoline which are used to-day, and would not supply the modern high-speed engines with gas of the proper consistency fast enough even if they did not have to use very volatile gasoline. The form of carburetor used at the present time operates on a different principle. These devices are known as “spraying carburetors.” The fuel is reduced to a spray by the suction effect of the entering air stream drawing it through a fine opening.

The advantage of this construction is that a more[121] thorough amalgamation of the gasoline and air particles is obtained. With the earlier types previously considered the air would combine with only the more volatile elements, leaving the heavier constituents in the tank. As the fuel became stale it was difficult to vaporize it, and it had to be drained off and fresh fuel provided before the proper mixture would be produced. It will be evident that when the fuel is sprayed into the air stream, all the fuel will be used up and the heavier portions of the gasoline will be taken into the cylinder and vaporized just as well as the more volatile vapors.

The simplest form of spray carburetor is that shown at Fig. 43. In this the gasoline opening through which[122] the fuel is sprayed into the entering air stream is closed by the spring-controlled mushroom valve which regulates the main air opening as well. When the engine draws in a charge of air it unseats the valve and at the same time the air flowing around it is saturated with gasoline particles through the gasoline opening. The mixture thus formed goes to the engine through the mixture passage. Two methods of varying the fuel proportions are provided. One of these consists of a needle valve to regulate the amount of gasoline, the other is a knurled screw which controls the amount of air by limiting the lift of the jump valve.

The modern form of spraying carburetor is provided with two chambers, one a mixing chamber through which the air stream passes and mixes with a gasoline spray, the other a float chamber in which a constant level of fuel is maintained by simple mechanism. A jet or standpipe is used in the mixing chamber to spray the fuel through and the object of the float is to maintain the fuel level to such a point that it will not overflow the jet when the motor is not drawing in a charge of gas. With the simple forms of generator valve in which the gasoline opening is controlled by the air valve, a leak anywhere in either valve or valve seat will allow the gasoline to flow continuously whether the engine is drawing in a charge or not. The liquid fuel collects around the air opening, and when the engine inspires a charge it is saturated with gasoline globules and is excessively rich. With a float-feed construction, which maintains a constant level of gasoline at the right height in the standpipe, liquid fuel will only be supplied when drawn out of the jet by the suction effect of the entering air stream.

The first form of spraying carburetor ever applied successfully was evolved by Maybach for use on one of the[124] earliest Daimler engines. The general principles of operation of this pioneer float-feed carburetor are shown at Fig. 44, A. The mixing chamber and valve chamber were one and the standpipe or jet protruded into the mixing chamber. It was connected to the float compartment by a pipe. The fuel from the tank entered the top of the float compartment and the opening was closed by a needle valve carried on top of a hollow metal float. When the level of gasoline in the float chamber was lowered the float would fall and the needle valve uncover the opening. This would permit the gasoline from the tank to flow into the float chamber, and as the chamber filled the float would rise until the proper level had been reached, under which conditions the float would shut off the gasoline opening. On every suction stroke of the engine the inlet valve, which was an automatic type, would leave its seat and a stream of air would be drawn through the air opening and around the standpipe or jet. This would cause the gasoline to spray out of the tube and mix with the entering air stream.

The form shown at B was a modification of Maybach’s simple device and was first used on the Phœnix-Daimler engines. Several improvements are noted in this device. First, the carburetor was made one unit by casting the float and mixing chambers together instead of making them separate and joining them by a pipe, as shown at A. The float construction was improved and the gasoline shut-off valve was operated through leverage instead of being directly fastened to the float. The spray nozzle was surrounded by a choke tube which concentrated the air stream around it and made for more rapid air flow at low engine speeds. A conical piece was placed over the jet to break up the entering spray into a mist and insure more intimate admixture of air and gasoline. The air opening was provided with an air cone which had a shutter controlling the opening so that the amount of air entering could be regulated and thus vary the mixture proportions within certain limits.

[125]

The form shown at B has been further improved, and the type shown at C is representative of modern single jet practice. In this the float chamber and mixing chamber are concentric. A balanced float mechanism which insures steadiness of feed is used, the gasoline jet or standpipe is provided with a needle valve to vary the amount of gasoline supplied the mixture and two air openings are provided. The main air port is at the bottom of the vaporizer, while an auxiliary air inlet is provided at the side of the mixing chamber. There are two methods of controlling the mixture proportions in this form of carburetor. One may regulate the gasoline needle or adjust the auxiliary air valve.

A Schebler carburetor, which has been used on some airplane engines, is shown in Fig. 45. It will be noticed that a metering pin or needle valve opens the jet when the air valve opens. The long arm of a leverage is connected to the air valve, while the short arm is connected to the needle, the reduction in leverage being such that the needle valve is made to travel much less than the air valve. For setting the amount of fuel passed or the size of the jet orifice when running with the air valve closed, there is a screw which raises or lowers the fulcrum of the lever and there is also a dash control having the same effect by pushing down the fulcrum against a small spring. A long extension is given to the venturi tube which is very narrow around the jet orifices, which are horizontal and shown at A in the drawing. Fuel enters the float chamber through the union M, and the spring P holds the metering pin upward against the restraining action of the lever. The air valve may be set by an easily adjustable knurled screw shown in the drawing, and fluttering of the valve is prevented by the piston dash pot carried in a chamber above the valve into which the valve stem projects. The[127] primary air enters beneath the jet passage and there is a small throttle in the intake to increase the speed of air flow for starting purposes. The carburetor is adapted for the use of a hot-air connection to the stove around the exhaust pipe and it is recommended that such a fitting be supplied. The lever which controls the supply of air through the primary air intake is so arranged that if desired it can be connected with a linkage on the dash or control column by means of a flexible wire.

This carburetor is of extremely simple construction, because it has no supplementary or auxiliary air valve and no moving parts except the throttle controlling the gas flow. The construction is already shown in Fig. 46.[128] The spray jet is eccentric with a surrounding sleeve or tube in which there are two series of small orifices, one at the top and the other near the bottom. The former are about level with the spray jet opening. The sleeve surrounding the nozzle is closed at the top. The air, passing the upper holes in the sleeve, produces a vacuum in the sleeve, thereby drawing air in through the bottom holes. It is this moving interior column of air that controls the flow of gasoline from the nozzle. Owing to the friction of the small passages, the speed of air flow through the sleeve does not increase as fast as the speed of air flow outside the sleeve, hence there is a tendency for the mixture to remain constant. The throttle of this carburetor is of the barrel type, and the top of the spray nozzle and its surrounding sleeve are located inside the throttle.

The carburetor shown at Fig. 47 is a metering type in which the vacuum at the jet is controlled by the weight of the metering valve surrounding the upright metering pin. The only moving part is the metering valve, which rises and falls with the changes in vacuum. The air chamber surrounds the metering valve, and there is a mixing chamber above. As the valve is drawn up the gasoline passage is enlarged on account of the predetermined taper on the metering pin, and the air passage also is increased proportionately, giving the correct mixture. A dashpot at the bottom of the valve checks flutter. In idling the valve rests on its seat, practically closing the air and giving the necessary idling mixture. A passage through the valve acts as an aspirating tube. When the valve is closed altogether the primary air passes through ducts in the valve itself, giving the proper amount for idling. The one adjustment consists in raising or lowering the tapered metering pin, increasing or decreasing the supply of gasoline. Dash control is supplied. This pulls down the metering pin, increasing the gasoline flow. The duplex type for eight- and twelve-cylinder motors is the same in[129] principle as model 25, but it is a double carburetor synchronized as to throttle movements, adjustments, etc. The duplex for aeronautical motors is made of cast aluminum alloy.

To secure properly proportioned mixtures some carburetor designers have evolved forms in which two or more nozzles are used in a common mixing chamber. The usual construction is to use two, one having a small opening and placed in a small air tube and used only for low[130] speeds, the other being placed in a larger air tube and having a slightly augmented bore so that it is employed on intermediate speeds. At high speeds both jets would be used in series. Some multiple jet carburetors could be considered as a series of these instruments, each one being designed for certain conditions of engine action. They would vary from small size just sufficient to run the engine at low speed to others having sufficient capacity to furnish gas for the highest possible engine speed when used in conjunction with the smaller members which have been brought into service progressively as the engine speed has been augmented. The multiple nozzle carburetor differs from that in which a single spray tube is used only in the construction of the mixing chamber, as a common float bowl can be used to supply all spray pipes. It is common practice to bring the jets into action progressively by some form of mechanical connection with the throttle or by automatic valves.

The object of any multiple nozzle carburetor is to secure greater flexibility and endeavor to supply mixtures of proper proportions at all speeds of the engine. It should be stated, however, that while devices of this nature lend themselves readily to practical application it is more difficult to adjust them than the simpler forms having but one nozzle. When a number of jets are used the liability of clogging up the carburetor is increased, and if one or more of the nozzles is choked by a particle of dirt or water the resulting mixture trouble is difficult to detect. One of the nozzles may supply enough gasoline to permit the engine to run well at certain speeds and yet not be adequate to supply the proper amount of gas under other conditions. In adjusting a multiple jet carburetor in which the jets are provided with gasoline regulating needles, it is customary to consider each nozzle as a distinct carburetor and to regulate it to secure the best motor action at that throttle position which corresponds to the conditions under which the jet is brought into service. For instance, that supplied the primary mixing chamber[131] should be regulated with the throttle partly closed, while the auxiliary jet should be adjusted with the throttle fully opened.

This is a two-stage vaporizing device, hot air being used in the primary or initial stage of vaporization and cold air in the supplementary stage. Referring to the sectional illustration at Fig. 48, it will be seen that there is a hot-air passage with a choke-valve; the primary venturi appears at B; J is its gasoline jet, and V is a spring-loaded idling valve in a fixed air opening. These parts constitute the primary system. In the secondary system A is a cold-air passage, T a butterfly valve and J a gasoline jet discharging into the cold-air passage. This system is brought into operation by opening the butterfly T. A connection between the butterfly T and the throttle, not shown, throws the butterfly wide open when the throttle is not quite wide open; at all other times the butterfly[132] is held closed by a spring. The cylindrical chamber at the right of the mixing chamber has an extension E of reduced diameter connecting it with the intake manifold through a passage D. A restricted opening connects the float chamber with the cylindrical chamber so that the gasoline level is the same in both. A loosely fitting plunger P in the cylindrical chamber has an upward extension into the small part of the chamber. O is a small air opening and M is a passage from the cylindrical chamber to the mixing chamber. Air constantly passes through this when the carburetor is in operation. The carburetor is really two in one. The primary carburetor is made up of a central jet in a venturi passage. The float chamber is eccentric. In the air passage there is a fixed opening, and additional air is taken in by the opening through suction of a spring-opposed air valve. The second stage, which comes into play as soon as the carburetor is called upon for additional mixture above low medium speeds, is made up of an independent air passage containing another air valve. As the valve is opened this jet is uncovered, and air is led past it. For easy starting an extra passage leads from the float bowl passage to a point above the throttle. All the suction falls upon this passage when the throttle is closed. The passage contains a plunger and acts as a pick-up device. When the vacuum increases the plunger rises and shuts off the flow of gasoline from the intake passage. As the throttle is opened the vacuum in the intake passage is broken, and the plunger falls, causing gasoline to gather above it. This is immediately drawn through the pick-up passage and gives the desired mixture for acceleration.

This carburetor, shown in detail in Figs. 49 and 50, has been very popular in racing cars and aviation engines because of exceptionally good pick-up qualities and its thorough atomization of fuel. Its principle of operation is the breaking up of the fuel by a series of jets, which[133] vary in number from fourteen to twenty-one, according to the size of the carburetor. These are uncovered by opening the throttle, which is curved—a patented feature—to secure the correct progression of jets. The carburetor has an eccentric float chamber, from which the gasoline is led to the jet piece from which the jets stand up in a row. The tops of these jets are closed until the throttle is opened far enough to pass them, which it does progressively. The air opening is at the bottom, and the throttle opening is such that a modified venturi is formed.[134] The throttle is carried in a cylindrical barrel with the jets placed below it, and the passage from the barrel to the intake is arranged so that there is no interruption in the flow. For easy starting a dash-controlled shutter closes off the air, throwing the suction on the jets, thus giving a rich mixture.

The only adjustment is for idling, and once that is fixed it need never be touched. This is in the form of a screw and regulates the position of the throttle when at idling position. The dash control has high-speed, normal and rich-starting positions. In installing the Master carburetor the float chamber may be turned either toward the radiator or driver’s seat. If the float is turned toward the radiator, however, a forward lug plate should be ordered; otherwise it will be difficult to install the control. The throttle lever must go all the way to the stop lug[135] or maximum power will not be secured. In adjusting the idle screw it is turned in for rich and out for lean.

The Zenith carburetor, shown at Fig. 51, has become very popular for airplane engine use because of its simplicity, as mixture compensation is secured by a compensating compound nozzle principle that works very well in practice. To illustrate this principle briefly, let us consider the elementary type of carburetor or mixing valve, as shown in Fig. 52, A. It consists of a single jet or spraying nozzle placed in the path of the incoming air and fed from the usual float chamber. It is a natural[137] inference to suppose that as the speed of the motor increases, both the flow of air and of gasoline will increase in the same proportion. Unhappily, such is not the case. There is a law of liquid bodies which states that the flow of gasoline from the jet increases under suction faster than the flow of air, giving a mixture which grows richer and richer—a mixture containing a much higher percentage of gasoline at high suction than at low. The tendency is shown by the accompanying curve (Fig. 52, B), which gives the ratio of gasoline to air at varying speeds from this type of jet. The mixture is practically constant only between narrow limits and at very high speed. The most common method of correcting this defect is by putting various auxiliary air valves which, adding air, tends to dilute this mixture as it gets too rich. It is difficult with makeshift devices to gauge this dilution accurately for every motor speed.

Now, if we have a jet which grows richer as the suction increases, the opposite type of jet is one which would grow leaner under similar conditions. Baverey, the inventor of the Zenith, discovered the principle of the constant flow device which is shown in Fig. 52, C. Here a certain fixed amount of gasoline determined by the opening I is permitted to flow by gravity into the well J open to the air. The suction at jet H has no effect upon the gravity compensator I because the suction is destroyed by the open well J. The compensator, then, delivers a steady rate of flow per unit of time, and as the motor suction increases more air is drawn up, while the amount of gasoline remains the same and the mixture grows poorer and poorer. Fig. 52, D, shows this curve.

By combining these two types of rich and poor mixture carburetors the Zenith compound nozzle was evolved. In Fig. 52, E, we have both the direct suction or richer type leading through pipe E and nozzle G and the “constant flow” device of Baverey shown at J, I, K and nozzle H. One counteracts the defects of the other, so that from the cranking of the motor to its highest speed there is[138] a constant ratio of air and gasoline to supply efficient combustion.

In addition to the compound nozzle the Zenith is equipped with a starting and idling well, shown in the cut of Model L carburetor at P and J. This terminates in a priming hole at the edge of the butterfly valve, where the suction is greatest when this valve is slightly open. The gasoline is drawn up by the suction at the priming hole and, mixed with the air rushing by the butterfly, gives an ideal slow speed mixture. At higher speeds with the butterfly valve opened further the priming well ceases to operate and the compound nozzle drains the well and compensates correctly for any motor speed.

With the coming of the double motor containing eight or twelve cylinders arranged in two V blocks, the question of good carburetion has been a problem requiring much study. The single carburetor has given only indifferent results due to the strong cross suction in the inlet manifold from one set of cylinders to the other. This naturally led to the adoption of two carburetors in which each set of cylinders was independently fed by a separate carburetor.[139] Results from this system were very good when the two carburetors were working exactly in unison, but as it was extremely difficult to accomplish this co-operation, especially where the adjustable type was employed, this system never gained in favor. The next logical step was the Zenith Duplex, shown at Fig. 53. This consists of two separate and distinct carburetors joined together so that a common gasoline float chamber and air inlet could be used by both. It does away with cross suction in the manifold because each set of cylinders has a separate[140] intake of its own. It does away with two carburetors and makes for simplicity. The practical application of the Zenith carburetor to the Curtiss 90 horse-power OX-2 motor used on the JN-4 standard training machine is shown at Fig. 54, which outlines a rear view of the engine in question. The carburetor is carried low to permit of fuel supply from a gravity tank carried back of the motor.

Many carburetors include a filtering screen at the point where the liquid enters the float chamber in order to keep dirt or any other foreign matter which may be present in the fuel from entering the float chamber. This is not general practice, however, and the majority of vaporizers do not include a filter in their construction. It is very desirable that the dirt should be kept out of the carburetor because it may get under the float control fuel valve and cause flooding by keeping it raised from its seat. If it finds its way into the spray nozzle it may block the opening so that no gasoline will issue or may so constrict the passage that only very small quantities of fuel will be supplied the mixture. Where the carburetor itself is not provided with a filtering screen a simple filter is usually installed in the pipe line between the gasoline tank and the float chamber.

Some simple forms of filters and separators are shown at Fig. 55. That at A consists of a simple brass casting having a readily detachable gauze screen and a settling chamber of sufficient capacity to allow the foreign matter to settle to the bottom, from which it is drained out by a pet cock. Any water or dirt in the gasoline will settle to the bottom of the chamber, and as all fuel delivered to the carburetor must pass through the wire gauze screen it is not likely to contain impurities when it reaches the float chamber. The heavier particles, such as scale from the tank or dirt and even water, all of which have greater weight than the gasoline, will sink to the bottom of the[141] chamber, whereas light particles, such as lint, will be prevented from flowing into the carburetor by the filtering screen.

The filtering device shown at B is a larger appliance than that shown at A, and should be more efficient as a separator because the gasoline is forced to pass through three filtering screens before it reaches the carburetor. The gasoline enters the device shown at C through a bent pipe which leads directly to the settling chamber and from thence through a wire gauze screen to the upper compartment which leads to the carburetor. The device shown at D is a combination strainer, drain, and sediment[142] cup. The filtering screen is held in place by a spring and both are removed by taking out a plug at the bottom of the device. The shut-off valve at the top of the device is interposed between the sediment cup and the carburetor. This separating device is incorporated with the gasoline tank and forms an integral part of the gasoline supply system. The other types shown are designed to be interposed between the gasoline tank and the carburetor at any point in the pipe line where they may be conveniently placed.

On four- and six-cylinder engines and in fact on all multiple-cylinder forms, it is important that the piping leading from the carburetor to the cylinders be made in such a way that the various cylinders will receive their full quota of gas and that each cylinder will receive its charge at about the same point in the cycle of operations. In order to make the passages direct the bends should be as few as possible, and when curves are necessary they should be of large radius because an abrupt corner will not only impede gas flow but will tend to promote condensation of the fuel. Every precaution should be taken with four- and six-cylinder engines to insure equitable gas distribution to the valve chambers if regular action of the power plant is desired. If the gas pipe has many turns and angles it will be difficult to charge all cylinders properly. On some six-cylinder aviation engines, two carburetors are used because of trouble experienced with manifolds designed for one carburetor. Duplex carburetors are necessary to secure the best results from eight- and twelve-cylinder V engines.

The problem of intake piping is simplified to some extent on block motors where the intake passage is cored in the cylinder casting and where but one short pipe is needed to join this passage to the carburetor. If the cylinders are cast in pairs a simple pipe of T or Y form can be used with success. When the engine is of a type[143] using individual cylinder castings, especially in the six-cylinder power plants, the proper application and installation of suitable piping is a difficult problem. The reader is referred to the various engine designs outlined to ascertain how the inlet piping has been arranged on representative aviation engines. Intake piping is constructed in two ways, the most common method being to cast the manifold of brass or aluminum. The other method, which is more costly, is to use a built-up construction of copper or brass tubing with cast metal elbows and Y pieces. One of the disadvantages advanced against the cast manifold is that blowholes may exist which produce imperfect castings and which will cause mixture troubles because the entering gas from the carburetor, which may be of proper proportions, is diluted by the excess air which leaks in through the porous casting. Another factor of some moment is that the roughness of the walls has a certain amount of friction which tends to reduce the velocity of the gases, and when projecting pieces are present, such as core wire or other points of metal, these tend to collect the drops of liquid fuel and thus promote condensation. The advantage of the built-up construction is that the walls of the tubing are very smooth, and as the castings are small it is not difficult to clean them out thoroughly before they are incorporated in the manifold. The tubing and castings are joined together by hard soldering, brazing or autogenous welding.

The low-grade gasoline used at the present time makes it necessary to use vaporizers that are more susceptible to atmospheric variations than when higher grade and more volatile liquids are vaporized. Sudden temperature changes, sometimes being as much as forty degrees rise or fall in twelve hours, affect the mixture proportions to some extent, and not only changes in temperature but variations in altitude also have a bearing on mixture proportions by affecting both gasoline and air. As the temperature[144] falls the specific gravity of the gasoline increases and it becomes heavier, this producing difficulty in vaporizing. The tendency of very cold air is to condense gasoline instead of vaporizing it and therefore it is necessary to supply heated air to some carburetors to obtain proper mixtures during cold weather. In order that the gas mixtures will ignite properly the fuel must be vaporized and thoroughly mixed with the entering air either by heat or high velocity of the gases. The application of air stoves to the Curtiss OX-2 motor is clearly shown at Fig. 54. It will be seen that flexible metal pipes are used to convey the heated air to the air intakes of the duplex mixing chamber.

Any internal combustion engine will show less power at high altitudes than it will deliver at sea level, and this has caused a great deal of questioning. “There is a good[145] reason for this,” says a writer in “Motor Age,” “and it is a physical impossibility for the engine to do otherwise. The difference is due to the lower atmospheric pressure the higher up we get. That is, at sea level the atmosphere has a pressure of 14.7 pounds per square inch; at 5,000 feet above sea level the pressure is approximately 12.13 pounds per square inch, and at 10,000 feet it is 10 pounds per square inch. From this it will be seen that the final pressure attained after the piston has driven the gas into compressed condition ready for firing is lower as the atmospheric pressure drops. This means that there is not so much power in the compressed charge of gas the higher up you get above sea level.

“For example, suppose the compression ratio to be 4¹⁄₂ to 1; in other words, suppose the air space above the piston to have 4¹⁄₂ times the volume when the piston is at the bottom of its stroke that it has when the piston is at the top of the stroke. That is a common compression ratio for an average motor, and is chosen because it is considered to be the best for maximum horse-power and in order that the compression pressure will not be so high as to cause pre-ignition. Knowing the compression ratio, we can determine the final pressure immediately before ignition by substituting in the standard formula:

in which P is the atmospheric pressure; P¹ is the final pressure, and V/V¹ is the compression ratio, therefore P¹ = 14.7 (4.5)^1.3 = 104 pounds per square inch, absolute.

“That is, 104 pounds per square inch is the most efficient final compression pressure to have for this engine at sea level, since it comes directly from the compression ratio.

“Now supposing we consider that the altitude is 7,000[146] feet above sea level. At this height the atmospheric pressure is 11.25 pounds per square inch, approximately. In this case we can again substitute in the formula, using the new atmospheric pressure figure. The equation becomes:

P¹ = 11.25 (4.5)^1.3—79.4 pounds per square inch, absolute.

“Therefore we now have a final compression pressure of only 79.4 pounds per square inch, which is considerably below the pressure we have just found to be the most efficient for the motor. The resulting power drop is evident.

“It should be borne in mind that these final compression pressures are absolute pressures—that is, they include the atmospheric pressure. In the first case, to get the pressure above atmospheric you would subtract 14.7 and in the latter 11.25 would have to be deducted. In other words, where the sea level compression is 89.3 pounds per square inch above the atmosphere, the same motor will have only a compression pressure of 68.15 pounds per square inch above the atmosphere at 7,000 feet elevation.

“From the above it is evident that in order to bring the final compression pressure up to the efficient figure we have determined, a different compression ratio would have to be used. That is, the final volume would have to be less, and as it is impossible to vary this to meet the conditions of altitude, the loss of power cannot be helped except by the replacing of the standard pistons with some that are longer above the wrist-pin so as to reduce the space above the pistons when on top center. Then if the ratio is thereby raised to some such figures as 5 to 1, the engine will again have its proper final pressure, but it will still not have as much power as it would have at sea level, since the horse-power varies directly with the atmospheric pressure, final compression being kept constant. That is, at 7,000 feet the horse-power of[147] an engine that had 40 horse-power at sea level would be equal to

“If the original compression ratio of 4.5 were retained, the drop in horse-power would be even greater than this. These computations and remarks will make it clear that the designer who contemplates building an airplane for high altitude use should see to it that it is of sufficient power to compensate for the drop that is inevitable when it is up in the air. This is often illustrated in stationary gas-engine installations. An engine that had a sea-level rating amply sufficient for the work required, might not be powerful enough when brought up several thousand feet.” When one considers that airplanes attain heights of over 18,000 feet, it will be evident that an ample margin of engine power is necessary.

A system of fuel supply developed by the late Dr. Diesel, a German chemist and engineer, is attracting considerable attention at the present time on account of the ability of the Diesel engine to burn low-grade fuels, such as crude petroleum. In this system the engines are built so that very high compressions are used, and only pure air is taken into the cylinder on the induction stroke. This is compressed to a pressure of about 500 pounds per square inch, and sufficient heat is produced by this compression to explode a hydrocarbon mixture. As the air which is compressed to this high point cannot burn, the fuel is introduced into the cylinder combustion chamber under still higher compression than that of the compressed air, and as it is injected in a fine stream it is immediately vaporized because of the heat. Just as soon as the compressed air becomes thoroughly saturated with the liquid fuel, it will explode on account of the degree of[148] heat present in the combustion chamber. Such motors have been used in marine and stationary applications, but are not practical for airplanes or motor cars because of lack of flexibility and great weight in proportion to power developed. The Diesel engine is the standard power plant used in submarine boats and motor ships, as its efficiency renders it particularly well adapted for large units.

A writer in “The Aeroplane,” an English publication, discourses on some features of carburetor installation that may be of interest to the aviation student, so portions of the dissertation are reproduced herewith.

The modern float feed carburetor is a delicate and nicely balanced appliance that requires a certain amount of attention and care in order to obtain the best results. The adjustments can only be made by one possessing an intelligent knowledge of carburetor construction and must never be made unless the reason for changing the old adjustment is understood. Before altering the adjustment of the leading forms of carburetors, a few hints regarding the quality to be obtained in the mixture should be given some consideration, as if these are properly understood this knowledge will prove of great assistance in adjusting the vaporizer to give a good working proportion of fuel and air. There is some question regarding the best mixture proportions and it is estimated that gas will be explosive in which the proportions of fuel vapor and air will vary from one part of the former to a wide range included between four and eighteen parts of the latter. A one to four mixture is much too rich, while the one in eighteen is much too lean to provide positive ignition.

A rich mixture should be avoided because the excessive fuel used will deposit carbon and will soot the cylinder walls, combustion chamber interior, piston top and valves and also tend to overheat the motor. A rich mixture will[152] also seriously interfere with flexible control of the engine, as it will choke up on low throttle and run well on open throttle when the full amount of gas is needed. A rich mixture may be quickly discovered by black smoke issuing from the muffler, the exhaust gas having a very pungent odor. If the mixture contains a surplus of air there will be popping sounds in the carburetor, which is commonly termed “blowing back.” To adjust a carburetor is not a difficult matter when the purpose of the various control members is understood. The first thing to do in adjusting a carburetor is to start the motor and to retard the sparking lever so the motor will run slowly leaving the throttle about half open. In order to ascertain if the mixture is too rich cut down the gasoline flow gradually by screwing down the needle valve until the motor commences to run irregularly or misfire. Close the needle valves as far as possible without having the engine come to a stop, and after having found the minimum amount of fuel gradually unscrew the adjusting valve until you arrive at the point where the engine develops its highest speed. When this adjustment is secured the lock nut is screwed in place so the needle valve will keep the adjustment. The next point to look out for is regulation of the auxiliary air supply on those types of carburetors where an adjustable air valve is provided. This is done by advancing the spark lever and opening the throttle. The air valve is first opened or the spring tension reduced to a point where the engine misfires or pops back in the carburetor. When the point of maximum air supply the engine will run on is thus determined, the air valve spring may be tightened by screwing in on the regulating screw until the point is reached where an appreciable speeding up of the engine is noticed. If both fuel and air valves are set right, it will be possible to accelerate the engine speed uniformly without interfering with regularity of engine operation by moving the throttle lever or accelerator pedal from its closed to its wide open position, this being done with the spark lever advanced. All types of carburetors do not have the same[153] means of adjustment; in fact, some adjust only with the gasoline regulating needle; others must have a complete change of spray nozzles; while in others the mixture proportions may be varied only by adjustment of the quantity of entering air. Changing the float level is effective in some carburetors, but this should never be done unless it is certain that the level is not correct. Full instructions for locating carburetion troubles will be given in proper sequence.

It is a fact well known to experienced repairmen and motorists that atmospheric conditions have much to do with carburetor action. It is often observed that a motor seems to develop more power at night than during the day, a circumstance which is attributed to the presence of more moisture in the cooler night air. Likewise, taking a motor from sea level to an altitude of 10,000 feet involves using rarefied air in the engine cylinders and atmospheric pressures ranging from 14.7 pounds at sea level to 10.1 pounds per square inch at the high altitude. All carburetors will require some adjustment in the course of any material change from one level to another. Great changes of altitude also have a marked effect on the cooling system of an airplane. Water boils at 212 degrees F. only at sea level. At an altitude of 10,000 feet it will boil at a temperature nineteen degrees lower, or 193 degrees F.

In high altitudes the reduced atmospheric pressure, for 5,000 feet or higher than sea level, results in not enough air reaching the mixture, so that either the auxiliary air opening has to be increased, or the gasoline in the mixture cut down. If the user is to be continually at high altitudes he should immediately purchase either a larger dome or a smaller strangling tube, mentioning the size carburetor that is at present in use and the type of motor that it is on, including details as to the bore and stroke. The smaller strangling tube makes an increased suction at the spray nozzle; the air will have to be readjusted to meet it and you can use more auxiliary[154] air, which is necessary. The effect on the motor without a smaller strangling tube is a perceptible sluggishness and failure to speed up to its normal crank-shaft revolutions, as well as failure to give power. It means that about one-third of the regular speed is cut out. The reduced atmospheric pressure reduces the power of the explosion, in that there is not the same quantity of oxygen in the combustion chamber as at sea level; to increase the amount taken in, you must also increase the gasoline speed, which is done by an increased suction through the smaller strangling aperture. Some forms of carburetors are affected more than others by changes of altitude, which explains why the Zenith is so widely employed for airplane engine use. The compensating nozzle construction is not influenced as much by changes of altitude as the simpler nozzle types are.

[155]

One of the most important auxiliary groups of the gasoline engine comprising the airplane power plant and one absolutely necessary to insure engine action is the ignition system or the method employed of kindling the compressed gas in the cylinder to produce an explosion and useful power. The ignition system has been fully as well developed as other parts of the engine, and at the present time practically all ignition systems follow principles which have become standard through wide acceptance.

During the early stages of development of the gasoline engine various methods of exploding the charge of combustible gas in the cylinder were employed. On some of the earliest engines a flame burned close to the cylinder head, and at the proper time for ignition a slide or valve moved to provide an opening which permitted the flame to ignite the gas back of the piston. This system was practical only on the primitive form of gas engines in which the charge was not compressed before ignition. Later, when it was found desirable to compress the gas a certain degree before exploding it, an incandescent platinum tube in the combustion chamber, which was kept in a heated condition by a flame burning in it, exploded the gas. The naked flame was not suitable in this application[156] because when the slide was opened to provide communication between the flame and the gas the compressed charge escaped from the cylinder with enough pressure to blow out the flame at times and thus cause irregular ignition. When the flame was housed in a platinum tube it was protected from the direct action of the gas, and as long as the tube was maintained at the proper point of incandescence regular ignition was obtained.

Some engineers utilized the property of gases firing themselves if compressed to a sufficient degree, while others depended upon the heat stored in the cylinder-head to fire the highly compressed gas. None of these methods were practical in their application to motor car engines because they did not permit flexible engine action which is so desirable. At the present time, electrical ignition systems in which the compressed gas is exploded by the heating value of the minute electric arc or spark in the cylinder are standard, and the general practice seems to be toward the use of mechanical producers of electricity rather than chemical batteries.

Two general forms of electrical ignition systems may be used, the most popular being that in which a current of electricity under high tension is made to leap a gap or air space between the points of the sparking plug screwed into the cylinder. The other form, which has been almost entirely abandoned in automobile and which was never used with airplane engine practice, but which is still used to some extent on marine engines, is called the low-tension system because current of low voltage is used and the spark is produced by moving electrodes in the combustion chamber.

The essential elements of any electrical ignition system, either high or low tension, are: First, a simple and practical method of current production; second, suitable timing apparatus to cause the spark to occur at the right point in the cycle of engine action; third, suitable wiring[157] and other apparatus to convey the current produced by the generator to the sparking member in the cylinder.

The various appliances necessary to secure prompt ignition of the compressed gases should be described in some detail because of the importance of the ignition system. It is patent that the scope of a work of this character does not permit one to go fully into the theory and principles of operation of all appliances which may be used in connection with gasoline motor ignition, but at the same time it is important that the elementary principles be considered to some extent in order that the reader should have a proper understanding of the very essential ignition apparatus. The first point considered will be the common methods of generating the electricity, then the appliances to utilize it and produce the required spark in the cylinder. Inasmuch as magneto ignition is universally used in connection with airplane engine ignition it will not be necessary to consider battery ignition systems.

To properly understand the phenomena and forces involved in the generation of electrical energy by mechanical means it is necessary to become familiar with some of the elementary principles of magnetism and its relation to electricity. The following matter can be read with profit by those who are not familiar with the subject. Most persons know that magnetism exists in certain substances, but many are not able to grasp the terms used in describing the operation of various electrical devices because of not possessing a knowledge of the basic facts upon which the action of such apparatus is based.

Magnetism is a property possessed by certain substances and is manifested by the ability to attract and repel other materials susceptible to its effects. When this phenomenon is manifested by a conductor or wire through which a current of electricity is flowing it is termed “electro-magnetism.” Magnetism and electricity are closely related, each being capable of producing the other. Practically[158] all of the phenomena manifested by materials which possess magnetic qualities naturally can be easily reproduced by passing a current of electricity through a body which, when not under electrical influence, is not a magnetic substance. Only certain substances show magnetic properties, these being iron, nickel, cobalt and their alloys.

The earliest known substance possessing magnetic properties was a stone first found in Asia Minor. It was called the lodestone or leading stone, because of its tendency, if arranged so it could be moved freely, of pointing one particular portion toward the north. The compass of the ancient Chinese mariners was a piece of this material, now known to be iron ore, suspended by a light thread or floated on a cork in some liquid so one end would point toward the north magnetic pole of the earth. The reason that this stone was magnetic was hard to define for a time, until it was learned that the earth was one huge magnet and that the iron ore, being particularly susceptible, absorbed and retained some of this magnetism.

Most of us are familiar with some of the properties of the magnet because of the extensive sale and use of small horseshoe magnets as toys. As they only cost a few pennies every one has owned one at some time or other and has experimented with various materials to see if they would be attracted. Small pieces of iron or steel were quickly attracted to the magnet and adhered to the pole pieces when brought within the zone of magnetic influence. It was soon learned that brass, copper, tin or zinc were not affected by the magnet. A simple experiment that serves to illustrate magnetic attraction of several substances is shown at A, Fig. 57. In this, several balls are hung from a standard or support, one of these being of iron, another of steel. When a magnet is brought near either of these they will be attracted toward it, while the others will remain indifferent to the magnetic force. Experimenters soon learned that of the common metals only iron or steel were magnetic.

If the ordinary bar or horseshoe magnet be carefully[159] examined, one end will be found to be marked N. This indicates the north pole, while the other end is not usually marked and is the south pole. If the north pole of one magnet is brought near the south pole of another, a strong attraction will exist between them, this depending upon[160] the size of the magnets used and the air gap separating the poles. If the south pole of one magnet is brought close to the end of the same polarity of the other there will be a pronounced repulsion of like force. These facts are easily proved by the simple experiment outlined at B, Fig. 57. A magnet will only attract or influence a substance having similar qualities. The like poles of magnets will repel each other because of the obvious impossibility of uniting two influences or forces of practically equal strength but flowing in opposite directions. The unlike poles of magnets attract each other because the force is flowing in the same direction. The flow of magnetism is through the magnet from south to north and the circuit is completed by the flow of magnetic influence through the air gap or metal armature bridging it from the north to the south pole.

Magnets are commonly made in two forms, either in the shape of a bar or horseshoe. These two forms are made in two types, simple or compound. The latter are composed of a number of magnets of the same form united so the ends of like polarity are laced together, and such a construction will be more efficient and have more strength than a simple magnet of the same weight. The two common forms of simple and compound magnets are shown at C, Fig. 57. The zone in which a magnetic influence occurs is called the magnetic field, and this force can be graphically shown by means of imaginary lines, which are termed “lines of force.” As will be seen from the diagram at D, Fig. 57, the lines show the direction of action of the magnetic force and also show its strength, as they are closer together and more numerous when the intensity of the magnetic field is at its maximum. A simple method of demonstrating the presence of the force is to lay a piece of thin paper over the pole pieces of either a bar or horseshoe magnet and sprinkle fine iron filings[161] on it. The particles of metal arrange themselves in very much the manner shown in the illustrations and prove that the magnetic field actually exists.

The form of magnet used will materially affect the size and area of the magnetic field. It will be noted that the field will be concentrated to a greater extent with the horseshoe form because of the proximity of the poles. It should be understood that these lines have no actual existence, but are imaginary and assumed to exist only to show the way the magnetic field is distributed. The magnetic influence is always greater at the poles than at the center, and that is why a horseshoe or U-form magnet is used in practically all magnetos or dynamos. This greater attraction at the poles can be clearly demonstrated by sprinkling iron filings on bar and U magnets, as outlined at E, Fig. 57. A large mass gathers at the pole pieces, gradually tapering down toward the point where the attraction is least.

From the diagrams it will be seen that the flow of magnetism is from one pole to the other by means of curved paths between them. This circuit is completed by the magnetism flowing from one pole to the other through the magnet, and as this flow is continued as long as the body remains magnetic it constitutes a magnetic circuit. If this flow were temporarily interrupted by means of a conductor of electricity moving through the field there would be a current of electricity induced in the conductor every time it cut the lines of force. There are three kinds of magnetic circuits. A non-magnetic circuit is one in which the magnetic influence completes its circuit through some substance not susceptible to the force. A closed magnetic circuit is one in which the influence completes its circuit through some magnetic material which bridges the gap between the poles. A compound circuit is that in which the magnetic influence passes through magnetic substances and non-magnetic substances in order to complete its circuit.

[162]

Magnetism may be produced in two ways, by contact or induction. If a piece of steel is rubbed on a magnet it will be found a magnet when removed, having a north and south pole and all of the properties found in the energizing magnet. This is magnetizing by contact. A piece of steel will retain the magnetism imparted to it for a considerable length of time, and the influence that remains is known as residual magnetism. This property may be increased by alloying the steel with tungsten and hardening it before it is magnetized. Any material that will retain its magnetic influence after removal from the source of magnetism is known as a permanent magnet. If a piece of iron or steel is brought into the magnetic field of a powerful magnet it becomes a magnet without actual contact with the energizer. This is magnetizing by magnetic induction. If a powerful electric current flows through an insulated conductor wound around a piece of iron or steel it will make a magnet of it. This is magnetizing by electro-magnetic induction. A magnet made in this manner is termed an electro-magnet and usually the metal is of such a nature that it will not retain its magnetism when the current ceases to flow around it. Steel is used in all cases where permanent magnets are required, while soft iron is employed in all cases where an intermittent magnetic action is desired. Magneto field magnets are always made of tungsten steel alloy, so treated that it will retain its magnetism for lengthy periods.

There are many points in which magnetism and electricity are alike. For instance, air is a medium that offers considerable resistance to the passage of both magnetic influence and electric energy, although it offers more resistance to the passage of the latter. Minerals like iron or steel are very easily influenced by magnetism and[163] easily penetrated by it. When one of these is present in the magnetic circuit the magnetism will flow through the metal. Any metal is a good conductor for the passage of the electric current, but few metals are good conductors of magnetic energy. A body of the proper metal will become a magnet due to induction if placed in the magnetic field, having a south pole where the lines of force enter it and a north pole where they pass out.

We have seen that a magnet is constantly surrounded by a magnetic field and that an electrical conductor when carrying a current is also surrounded by a field of magnetic influence. Now if the conductor carrying a current of electricity will induce magnetism in a bar of iron or steel, by a reversal of this process, a magnetized iron or steel bar will produce a current of electricity in a conductor. It is upon this principle that the modern dynamo or magneto is constructed. If an electro-motive force is induced in a conductor by moving it across a field of magnetic influence, or by passing a magnetic field near a conductor, electricity is said to be generated by magneto-electric induction. All mechanical generators of the electric current using permanent steel magnets to produce a field of magnetic influence are of this type.

The accompanying diagram, Fig. 58, will show these principles very clearly. As stated on an earlier page, if the lines of force in the magnetic field are cut by a suitable conductor an electrical impulse will be produced in that conductor. In this simple machine the lines of force exist between the poles of a horseshoe magnet. The conductor, which in this case is a loop of copper wire, is mounted upon a spindle in order that it may be rotated in the magnetic field to cut the lines of magnetic influence present between the pole pieces. Both of the ends of this loop are connected, one with the insulated drum shown upon the shaft, the other to the shaft. Two metal brushes are employed to collect the current and cause it[164] to flow through the external circuit. It can be seen that when the shaft is turned in the direction of the arrow the loop will cut through the lines of magnetic influence and a current will be generated therein.

The pressure of the current and the amount produced vary in accordance to the rapidity with which the lines of magnetic influence are cut. The armature of a practical magneto, therefore, differs materially from that shown in the diagram. A large number of loops of wire would be mounted upon this shaft in order that the lines of magnetic influence would be cut a greater number of times in a given period and a core of iron used as a backing[165] for the wire. This would give a more rapid alternating current and a higher electro-motive force than would be the case with a smaller number of loops of wire.

The illustrations at Fig. 59 show a conventional double[166] winding armature and field magnetic of a practical magneto in part section and will serve to more fully emphasize the points previously made. If the armature or spindle were removed from between the pole pieces there would exist a field of magnetic influence as shown at Fig. 57, but the introduction of this component provides a conductor (the iron core) for the magnetic energy, regardless of its position, though the facility with which the influence will be transmitted depends entirely upon the position of the core. As shown at A, the magnetic flow is through the main body in a straight line, while at B, which position the armature has attained after one-eighth revolution, or 45 degrees travel in the direction of the arrow, the magnetism must pass through in the manner indicated. At C, which position is attained every half revolution, the magnetic energy abandons the longer path through the body of the core for the shorter passage offered by the side pieces, and the field thrown out by the cross bar disappears. On further rotation of the armature, as at D, the body of the core again becomes energized as the magnetic influence resumes its flow through it. These changes in the strength of the magnetic field when distorted by the armature core, as well as the intensity of the energy existing in the field, affect the windings, and the electrical energy induced therein corresponds in strength to the rapidity with which these changes in magnetic flow occur. The most pronounced changes in the strength of the field will occur as the armature passes from position B to D, because the magnetic field existing around the core will be destroyed and again re-established.

During the most of the armature rotation the changes in strength will be slight and the currents induced in the wire correspondingly small; but at the instant the core becomes remagnetized, as the armature leaves position C, the current produced will be at its maximum, and it is necessary to so time the rotation of the armature that at this instant one of the cylinders is in condition to be fired. It[167] is imperative that the armature be driven in such relation to the crank-shaft that each production of maximum current coincides with the ignition point, this condition existing twice during each revolution of the armature, or at every 180 degrees travel. Each position shown corresponds to 45 degrees travel of the armature, or one-eighth of a turn, and it takes just three-eighths revolution to change the position from A to that shown at D.

The magnets which produce the influence that in turn induces the electrical energy in the winding or loops of wire on the armature, and which may have any even number of opposed poles, are called field magnets. The loops of wire which are mounted upon a suitable drum and rotate in the field of magnetic influence in order to cut the lines of force is called an armature winding, while the core is the metal portion. The entire assembly is called the armature. The exposed ends of the magnets are called pole pieces and the arrangement used to collect the current is either a commutator or a collector. The stationary pieces which bear against the collector or commutator and act as terminals for the outside circuit are called brushes. These brushes are often of copper, or some of its alloys, because copper has a greater electrical conductivity than any other metal.

These brushes are nearly always of carbon, which is sometimes electroplated with copper to increase its electrical conductivity, though cylinders of copper wire gauze impregnated with graphite are utilized at times. Carbon is used because it is not so liable to cut the metal of the commutator as might be the case if the contact was of the metal to metal type. The reason for this is that carbon has the peculiar property in that it materially assists in the lubrication of the commutator, and being of soft, unctuous composition, will wear and conform to any irregularities on the surface of the metal collector rings.

The magneto in common use consists of a number of[168] horseshoe magnets which are compound in form and attached to suitable cast-iron pole pieces used to collect and concentrate the magnetic influence of the various magnets. Between these pole pieces an armature rotates. This is usually shaped like a shuttle, around which are wound coils of insulated wire. These are composed of a large number of turns and the current produced depends in great measure upon the size of the wire and the number of turns per coil. An armature winding of large wire will deliver a current of great amperage, but of small voltage. An armature wound with very fine wire will deliver a current of high voltage but of low amperage. In the ordinary form of magneto, such as used for ignition, the current is alternating in character and the break in the circuit should be timed to occur when the armature is at the point of its greatest potential or pressure. Where such a generator is designed for direct current production the ends of the winding are attached to the segments of a commutator, but where the instrument is designed to deliver an alternating current one end of the winding is fastened to an insulator ring on one end of the armature shaft and the other end is grounded on the frame of the machine.

The quantity of the current depends upon the strength of the magnetic field and the number of lines of magnetic influence acting through the armature. The electro-motive force varies as to the length of the armature winding and the number of revolutions at which the armature is rotated.

The magneto in the various systems which employ a transformer coil is very similar to a low-tension generator in general construction, and the current delivered at the terminals seldom exceeds 100 volts. As it requires many times that potential or pressure to leap the gap which exists between the points of the conventional spark plug, a separate coil is placed in circuit to intensify the current to one of greater capacity. The essential parts[169] of such a system and their relation to each other are shown in diagrammatic form at Fig. 60 and as a complete system at Fig. 61. As is true of other systems the magnetic influence is produced by permanent steel magnets clamped to the cast-iron pole pieces between which the armature rotates. At the point of greatest potential in the armature winding the current is broken by the contact breaker, which is actuated by a cam, and a current of higher value is induced in the secondary winding of the transformer coil when the low voltage current is passed through the primary winding.

It will be noted that the points of the contact breaker are together except for the brief instant when separated by the action of the point of the cam upon the lever. It is obvious that the armature winding is short-circuited[171] upon itself except when the contact points are separated. While the armature winding is thus short-circuited there will be practically no generation of current. When the points are separated there is a sudden flow of current through the primary winding of the transformer coil, inducing a secondary current in the other winding, which can be varied in strength by certain considerations in the preliminary design of the apparatus. This current of higher potential or voltage is conducted directly to the plug if the device is fitted to a single-cylinder engine, or to the distributor arm if fitted to a multiple-cylinder motor. The distributor consists of an insulator in which is placed a number of segments, one for each cylinder to be fired, and so spaced that the number of degrees between them correspond to the ignition points of the motor. A two-cylinder motor would have two segments, a three-cylinder, three segments, and so on within the capacity of the instrument. In the illustration a four-cylinder distributor is fitted, and the distributing arm is in contact[172] with the segment corresponding to the cylinder about to be fired.

The true high-tension magneto differs from the preceding inasmuch as the current of high voltage is produced in the armature winding direct, without the use of the separate coil. Instead of but one coil, the armature carries two, one of comparatively coarse wire, the other of many turns of finer wire. The arrangement of these windings can be readily ascertained by reference to the diagram B, Fig. 60, which shows the principle of operation very clearly. The simplicity of the ignition system is evident by inspection of Fig. 62. One end of the primary winding (coarse wire) is coupled or grounded to the armature core, and the other passes to the insulated part of the interrupter. While in some forms the interrupter or contact breaker mechanism does not revolve, the desired motion being imparted to the contact lever to separate the points of a revolving cam, in this the cam or tripping mechanism is stationary and the contact breaker revolves. This arrangement makes it possible to conduct the current from the revolving primary coil to the interrupter by a direct connection, eliminating[173] the use of brushes, which would otherwise be necessary. In other forms of this appliance where the winding is stationary, the interrupter may be operated by a revolving cam, though, if desired, the used of a brush at this point will permit this construction with a revolving winding.

During the revolution of the armature the grounded lever makes and breaks contact with the insulated point, short-circuiting the primary winding upon itself until the armature reaches the proper position of maximum intensity of current production, at which time the circuit is broken, as in the former instance. One end of the secondary winding (fine wire) is grounded on the live end of the primary, the other end being attached to the revolving arm of the distributor mechanism. So long as a closed circuit is maintained feeble currents will pass through the primary winding, and so long as the contact points are together this condition will exist. When the current reaches its maximum value, because of the armature being in the best position, the cam operates the interrupter and the points are separated, breaking the short circuit which has existed in the primary winding.

The secondary circuit has been open while the distributor arm has moved from one contact to another and there has been no flow of energy through this winding. While the electrical pressure will rise in this, even if the distributor arm contacted with one of the segments, there would be no spark at the plug until the contact points separated, because the current in the secondary winding would not be of sufficient strength. When the interrupter operates, however, the maximum primary current will be diverted from its short circuit and can flow to the ground only through the secondary winding and spark-plug circuit. The high pressure now existing in the secondary winding will be greatly increased by the sudden flow of primary current, and energy of high enough potential to successfully bridge the gap at the plug is thereby produced in the winding.

[174]

The Berling magneto is a true high tension type delivering two impulses per revolution, but it is made in a variety of forms, both single and double spark. Its principle of action does not differ in essentials from the high tension type previously described. This magneto is used on Curtiss aviation engines and will deliver sparks in a positive manner sufficient to insure ignition of engines up to 200 horse-power and at rotative speeds of the magneto armature up to 4,000 r. p. m. which is sufficient to take care of an eight-cylinder V engine running up to 2,000[175] r. p. m. The magneto is driven at crank-shaft speed on four-cylinder engines, at 1¹⁄₂ times crank-shaft speed on six-cylinder engines and at twice crank-shaft speed on eight-cylinder V types. The types “D” and “DD” BERLING Magnetos are interchangeable with corresponding magnetos of other standard makes. The dimensions of the four-, six- and eight-cylinder types “D” and “DD” are all the same.

The ideal method of driving the magneto is by means of flexible direct connecting coupling to a shaft intended for the purpose of driving the magneto. As the magneto must be driven at a high speed, a coupling of some flexibility is preferable. The employment of such a coupling will facilitate the mounting of the magneto, because a small inaccuracy in the lining up of the magneto with the driving shaft will be taken care of by the flexible coupling, whereas with a perfectly rigid coupling the line-up of the magneto must be absolutely accurate. Another advantage of the flexible coupling is that the vibration of the motor will not be as fully transmitted to the armature shaft on the magneto as in case a rigid coupling is used. This means prolonged life for the magneto.

The next best method of driving the magneto is by means of a gear keyed to the armature shaft. When this method of driving is employed, great care must be exercised in providing sufficient clearance between the gear on the magneto and the driving gear. If there should be a tight spot between these two gears it will react disadvantageously on the magneto. The third available method is to drive the magneto by means of a chain. This is the least desirable of the three methods and should be resorted to only in case of absolute necessity. It is difficult to provide sufficient clearance when using a chain without rendering the timing less accurate and positive.

Fig. 64, A shows diagrammatically the circuit of the “D” type two-spark independent magneto and the switch used with it. In position OFF the primary winding[176] of the magneto is short-circuited and in this position the switch serves as an ordinary cut-out or grounding switch. In position “1” the switch connects the magneto in such a way that it operates as an ordinary single-spark magneto. In this position one end of the secondary winding is grounded to the body of the motor. This is the starting position. In this position of the switch the entire voltage generated in the magneto is concentrated at one spark-plug instead of being divided in half. With the motor turning over very slowly, as is the case in starting, the full voltage generated by the[177] magneto will not in all cases be sufficient to bridge simultaneously two spark gaps, but is amply sufficient to bridge one. Also, this position of the switch tends to retard the ignition and should be used in starting to prevent back-firing. With the switch in position “2” the magneto applies ignition to both plugs in each cylinder simultaneously. This is the normal running position.

Fig. 64, B shows diagrammatically the circuit of the type “DD” BERLING high-tension two-spark dual magneto. This type is recommended for certain types of heavy-duty airplane motors, which it is impossible to turn over fast enough to give the magneto sufficient speed to generate even a single spark of volume great enough to ignite the gas in the cylinder. The dual feature consists of the addition to the magneto of a battery interrupter. The equipment consists of the magneto, coil and special high-tension switch. The coil is intended to operate on six volts. Either a storage battery or dry cells may be used.

With the switch in the OFF position, the magneto is grounded, and the battery circuit is open. With the switch in the second or battery position marked “BAT,” one end of the secondary winding of the magneto is grounded, and the magneto operates as a single-spark magneto delivering high-tension current to the inside distributor, and the battery circuit being closed the high-tension current from the coil is delivered to the outside distributor. In this position the battery current is supplied to one set of spark plugs, no matter how slowly the motor is turned over, but as soon as the motor starts, the magneto supplies current as a single-spark magneto to the other set of the spark-plugs. After the engine is running, the switch should be thrown to the position marked “MAG.” The battery and coil are then disconnected, and the magneto furnishes ignition to both plugs in each cylinder. This is the normal running position. Either a non-vibrating coil type “N-1” is[178] furnished or a combined vibrating and non-vibrating coil type “VN-1.”

The magneto may be set according to one of two different methods, the selection of which is, to some extent, governed by the characteristics of the engine, but largely due to the personal preference on the part of the user. In the first method described below, the most advantageous position of the piston for fully advanced ignition is determined in relation to the extreme advanced position of the magneto. In this case, the fully retarded ignition will not be a matter of selection, but the timing range of the magneto is wide enough to bring the fully retarded ignition after top-center position of the piston. The second method for the setting of the magneto fixes the fully retarded position of the magneto in relation to that position of the piston where fully retarded ignition is desired. In this case, the extreme advance position of the magneto will not always correspond with the best position of the piston for fully advanced ignition, and the amount of advance the magneto should have to meet ideal requirements in this respect must be determined by experiment.

1. Designate one cylinder as cylinder No. 1.

2. Turn the crank-shaft until the piston in cylinder No. 1 is in the position where the fully advanced spark is desired to occur.

3. Remove the cover from the distributor block and turn the armature shaft in the direction of rotation of the magneto until the distributor finger-brush comes into such a position that this brush makes contact with the segment which is connected to the cable terminal marked “1.” This is either one of the two bottom segments, depending upon the direction of rotation.

4. Place the cam housing in extreme advance, i.e.,[179] turn the cam housing until it stops, in the direction opposite to the direction of rotation of the armature. With the cam housing in this position, open the cover.

5. With the armature in the approximate position as described in “3,” turn the armature slightly in either direction to such a point that the platinum points of the magneto interrupter will just begin to open at the end of the cam, adjacent to the fibre lever on the interrupter.

6. With this exact position of the armature, fix the magneto to the driving member of the engine.

1. Designate one cylinder as cylinder No. 1.

2. Turn the crank-shaft until the piston in cylinder No. 1 is in the position at which the fully retarded spark is desired to occur.

3. Same as No. 3 under First Method.

4. Place the cam housing in extreme retard, i.e., turn the cam housing until it stops, in the same direction as the direction of rotation of the armature. With the cam housing in this position, open the cover.

5. Same as No. 5 under First Method.

6. Same as No. 6 under First Method.

The wiring of the magneto is clearly shown by wiring diagram.

First determine the sequence of firing for the cylinders and then connect the cables to the spark plug in the cylinders in proper sequence, beginning with cylinder No. 1 marked on the distributor block.

The switch used with the independent type must be mounted in such a manner that there will be a metallic connection between the frame of the magneto and the metal portion of the switch.

It is advisable to use a separate battery, either storage or dry cells, as a source of current for the dual equipment.[180] Connecting to the same battery that is used with the generator and other electrical equipment may cause trouble, as a “ground” in this battery causes the coil to overheat.

Use only the very best of oil for the oil cups.

Put five drops of oil in the oil cup at the driving end of the magneto for every fifty hours of actual running.

Put five drops of oil in the oil cup at the interrupter end of the magneto, located at one side of the cam housing, for every hundred hours of actual running.

Lubricate the embossed cams in the cam housing with a thin film of vaseline every fifty hours of actual running. Wipe off all superfluous vaseline. Never use oil in the interrupter. Do not lubricate any other part of the interrupter.

With the fibre lever in the center of one of the embossed cams, as at Fig. 65, the opening between the platinum contacts should be not less than .016′′ and not more than .020′′. The gauge riveted to the adjusting wrench should barely be able to pass between the contacts when fully open. The platinum contacts must be smoothed off with a very fine file. When in closed position, the platinum contacts should make contact with each other over their entire surfaces.

When inspecting the interrupter, make sure that the ground brush in the back of the interrupter base is making good contact with the surface on which it rubs.

The distributor block cover should be removed for inspection every twenty-five hours of actual running and the carbon deposit from the distributor finger-brush wiped off the distributor block by rubbing with a rag[181] or piece of waste dipped in gasoline or kerosene. The high-tension terminal brush on the side of the magneto should also be carefully inspected for proper tension.

Trouble in the ignition system is indicated by the motor “missing,” stopping entirely, or by inability to start.

It is safe to assume that the trouble is not in the magneto, and the carburetor, gasoline supply and spark-plugs should first be investigated.

If the magneto is suspected, the first thing to do is to determine if it will deliver a spark. To determine this, disconnect one of the high-tension leads from the spark-plug in one of the cylinders and place it so that there is approximately ¹⁄₁₆′′ between the terminal and the cylinder frame.

Open the pet cocks on the other cylinders to prevent the engine from firing and turn over the engine until the piston is approaching the end of the compression[182] stroke in the cylinder from which the cable has been removed. Set the magneto in the advance position and rapidly rock the engine over the top-center position, observing closely if a spark occurs between the end of the high-tension cable and the frame.

If the magneto is of the dual type, the trouble may be either in the magneto or in the battery or coil system, therefore disconnect the battery and place the switch in the position marked “MAG.” The magneto will then operate as an independent magneto and should spark in the proper manner. After this the battery system should be investigated. To test the operation of the battery and coil, examine all connections, making sure that they are clean and tight, and then with the switch, in the “BAT,” rock the piston slowly back and forth. If a type “VN-1” coil is used, a shower of sparks should jump between the high-tension cable terminal and the cylinder frame when the piston is in the correct position for firing. If no spark occurs, remove the cover from the coil and see that the vibrating tongue is free. If a type “N-1” coil is used, a single spark will occur. The battery should furnish six volts when connected to the coil, and this should also be verified.

If the coil still refuses to give a spark and all connections are correct, the coil should be replaced and the defective coil returned to the manufacturer.

If both magneto and coil give a spark when tested as just described, the spark-plugs should be investigated. To do this, disconnect the cables and remove the spark-plugs. Then reconnect the cables to the plugs and place them so that the frame portions of the plugs are in metallic connection with the frame of the motor. Then turn over the motor, thus revolving the magneto armature, and see if a spark is produced at the spark gaps of the plugs.

The most common defects in spark-plugs are breaking down of the insulation, fouling due to carbon, or too large or small a spark gap. To clean the plugs a stiff brush[183] and gasoline should be used. The spark gap should be about ¹⁄₃₂′′ and never less than ¹⁄₆₄′′. Too small a gap may have been caused by beads of metal forming due to the heat of the spark. Too long a gap may have been caused by the points burning off.

If the magneto and spark plugs are in good condition and the engine does not run satisfactorily, the setting should be verified according to instructions previously given, and, if necessary, readjusted.

Be careful to observe that both the type “VN-1” and type “N-1” coils are so arranged that the spark occurs on the opening of the contacts of the timer. As this is just the reverse of the usual operation, it should be carefully noted when any change in the setting of the timer is made. The timer on the dual type magneto is adjusted so that the battery spark occurs about 5° later[184] than the magneto spark. This provides an automatic advance as soon as the switch is thrown to the magneto position “MAG.” This relative timing can be easily adjusted by removing the interrupter and shifting the cam in the direction desired.

The Dixie magneto, shown at Fig. 66, operates on a different principle than the rotary armature type. It is used on the Hall-Scott and other aviation engines. In this magneto the rotating member consists of two pieces of magnetic material separated by a non-magnetic center piece. This member constitutes true rotating poles for the magnet and rotates in a field structure, composed of two laminated field pieces, riveted between two non-magnetic rings. The bearings for the rotating poles are[185] mounted in steel plates, which lie against the poles of the magnets. When the magnet poles rotate, the magnetic lines of force from each magnet pole are carried directly to the field pieces and through the windings, without reversal through the mass of the rotating member and with only a single air gap. There are no losses by flux reversal in the rotating part, such as take place in other[186] machines, and this is said to account for the high efficiency of the instrument.

And this “Mason Principle” involved in the operation of the Dixie is simplified by a glance at the field structure, consisting of the non-magnetic rings, assembled to which are the field pieces between which the rotating poles revolve (see Fig. 68). Rotating between the limbs of the magnets, these two pieces of magnetic material form true extensions to the poles of the magnets, and are, in consequence, always of the same polarity. It will be seen there is no reversal of the magnetism through them, and consequently no eddy current or hysteresis losses which are present in the usual rotor or inductor types. The simplicity features of construction stand out prominently here, in that there are no revolving windings, a detail entirely differing from the orthodox high-tension instrument. This simplicity becomes instantly apparent when it is found that the circuit breaker, instead of revolving as it does in other types, is stationary and that the whole breaker mechanism is exposed by simply turning the cover spring aside and removing cover. This makes inspection and adjustment particularly simple, and the fact that no special tool is necessary for adjustment of the platinum points—an ordinary small screw-driver is the whole “kit of tools” needed in the work of disassembling or assembling—is a feature of some value.

With dust- and water-protecting casing removed, and one of the magnets withdrawn, as in Fig. 69, the winding can be seen with its core resting on the field pole pieces and the primary lead attached to its side. An important feature of the high-tension winding is that the heads are of insulating material, and there is not the tendency for the high-tension current to jump to the side as in the ordinary armature type magneto. The high-tension current is carried to the distributor by means of an insulated block with a spindle, at one end of which is a spring brush bearing directly on the winding, thus shortening[187] the path of the high-tension current and eliminating the use of rubber spools and insulating parts. The moving parts of the magneto need never be disturbed if the high-tension winding is to be removed. This winding constitutes all of the magneto windings, no external spark coil being necessary. The condenser is placed directly above the winding and is easily removable by taking out two screws, instead of being placed in an armature where it is inaccessible except to an expert, and where it cannot be replaced except at the factory whence it emanated.

[188]

The bearings of the magneto are provided with oil cups and a few drops of light oil every 1,000 miles are sufficient. The breaker lever should be lubricated every 1,000 miles with a drop of light oil, applied with a tooth-pick. The proper distance between the platinum points when separated should not exceed .020 or one-fiftieth of an inch. A gauge of the proper size is attached to the screwdriver furnished with the magneto. The platinum contacts should be kept clean and properly adjusted. Should the contacts become pitted, a fine file should be used to smooth them in order to permit them to come into perfect contact. The distributor block should be removed occasionally and inspected for an accumulation of carbon dust. The inside of the distributor block should be cleaned with a cloth moistened with gasoline and then wiped dry with a clean cloth. When replacing the block, care must be exercised in pushing the carbon brush into the socket. Do not pull out the carbon brushes in the distributor because you think there is not enough tension on the small brass springs. In order to obtain the most efficient results, the normal setting of the spark-plug points should not exceed .025 of an inch, and it is advisable to have the gap just right before a spark-plug is inserted.

The spark-plug electrodes may be easily set by means of the gauge attached to the screwdriver. The setting of the spark-plug points is an important function which is usually overlooked, with the result that the magneto is blamed when it is not at fault.

In order to obtain the utmost efficiency from the engine, the magneto must be correctly timed to it. This operation is usually performed when the magneto is fitted to the engine at the factory. The correct setting may vary according to individuality of the engine, and some[190] engines may require an earlier setting in order to obtain the best results. However, should the occasion arise to retime the magneto, the procedure is as follows: Rotate the crank-shaft of the engine until one of the pistons, preferably that of cylinder No. 1, is ¹⁄₁₆ of an inch ahead of the end of the compression stroke. With the timing lever in full retard position, the driving shaft of the magneto should be rotated in the direction in which it will be driven. The circuit breaker should be closely observed and when the platinum contact points are about to separate, the drive gear or coupling should be secured to the drive shaft of the magneto. Care should be taken not to alter the position of the magneto shaft when tightening the nut to secure the gear or coupling, after which the magneto should be secured to its base. Remove the distributor block and determine which terminal of the block is in contact with the carbon brush of the distributor finger and connect with plug wire leading to No. 1 cylinder to this terminal. Connect the remaining plug wires in turn according to the proper sequence of firing of the cylinders. (See the wiring diagram for a typical six-cylinder engine at Fig. 70.) A terminal on the end of the cover spring of the magneto is provided for the purpose of connecting the wire leading to a ground switch for stopping the engine.

A special model or type of magneto is made for V engines which use a compound distributor construction instead of the simple type on the model illustrated and a different interior arrangement permits the production of four sparks per revolution of the rotors. This makes it possible to run the magneto slower than would be possible with the two-spark form. The application of two compound distributor magnetos of this type to a Thomas-Morse 135 horse-power motor of the eight-cylinder V pattern is clearly shown at Fig. 71.

[191]

[192]

With the high-tension system of ignition the spark is produced by a current of high voltage jumping between two points which break the complete circuit, which would exist otherwise in the secondary coil and its external connections. The spark-plug is a simple device which[193] consists of two terminal electrodes carried in a suitable shell member, which is screwed into the cylinder. Typical spark-plugs are shown in section at Fig. 72 and the construction can be easily understood. The secondary wire from the coil is attached to a terminal at the top of a central electrode member, which is supported in a bushing of some form of insulating material. The type shown at A employs a molded porcelain as an insulator, while that depicted at B uses a bushing of mica. The insulating bushing and electrode are housed in a steel body, which is provided with a screw thread at the bottom, by which means it is screwed into the combustion chamber.

When porcelain is used as an insulating material it is kept from direct contact with the metal portion by some form of yielding packing, usually asbestos. This is necessary because the steel and porcelain have different coefficients of expansion and some flexibility must be provided at the joints to permit the materials to expand differently when heated. The steel body of the plug which is screwed into the cylinder is in metallic contact with it and carries sparking points which form one of the terminals of the air gap over which the spark occurs. The[194] current entering at the top of the plug cannot reach the ground, which is represented by the metal portion of the engine, until it has traversed the full length of the central electrode and overcome the resistance of the gap between it and the terminal point on the shell. The porcelain bushing is firmly seated against the asbestos packing by means of a brass screw gland which sets against a flange formed on the porcelain, and which screws into a thread at the upper portion of the plug body.

The mica plug shown at B is somewhat simpler in construction than that shown at A. The mica core which keeps the central electrode separated from the steel body is composed of several layers of pure sheet mica wound around the steel rod longitudinally, and hundreds of stamped steel washers which are forced over this member and compacted under high pressure with some form of a binding material between them. Porcelain insulators are usually molded from high-grade clay and are approximately of the shapes desired by the designers of the plug. The central electrode may be held in place by mechanical means such as nuts, packings, and a shoulder on the rod, as shown at A. Another method sometimes used is to cement the electrode in place by means of some form of fire-clay cement. Whatever method of fastening is used, it is imperative that the joints be absolutely tight so that no gas can escape at the time of explosion. Porcelain is the material most widely used because it can be glazed so that it will not absorb oil, and it is subjected to such high temperature in baking that it is not liable to crack when heated.

The spark-plugs may be screwed into any convenient part of the combustion chamber, the general practice being to install them in the caps over the inlet valves, or in the side of the combustion chamber, so the points will be directly in the path of the entering fresh gases from the carburetor.

Other insulating materials sometimes used are glass,[195] steatite (which is a form of soapstone) and lava. Mica and porcelain are the two common materials used because they give the best results. Glass is liable to crack, while lava or the soapstone insulating bushings absorb oil. The spark gap of the average plug is equal to about ¹⁄₃₂ of an inch for coil ignition and ¹⁄₄₀ of an inch when used in magneto circuits. A simple gauge for determining the gap setting is the thickness of an ordinary visiting card for magneto plugs, or a space equal to the thickness of a worn dime for a coil plug. The insulating bushings are made in a number of different ways, and while details of construction vary, spark-plugs do not differ essentially in design. The dimensions of the standardized plug recommended by the S. A. E. are shown at Fig. 73.

It is often desirable to have a water-tight joint between the high-tension cable and the terminal screw on top of the insulating bushing of the spark-plug, especially in marine applications. The plug shown at C, Fig. 72,[196] is provided with an insulating member or hood of porcelain, which is secured by a clip in such a manner that it makes a water-tight connection. Should the porcelain of a conventional form of plug become covered with water or dirty oil, the high-tension current is apt to run down this conducting material on the porcelain and reach the ground without having to complete its circuit by jumping the air gap and producing a spark. It will be evident that wherever a plug is exposed to the elements, which is often the case in airplane service, that it should be protected by an insulating hood which will keep the insulator dry and prevent short circuiting of the spark. The same end can be attained by slipping an ordinary rubber nipple over the porcelain insulator of any conventional plug and bringing up one end over the cable.

On most aviation engines, especially those having large cylinders, it is sometimes difficult to secure complete combustion by using a single-spark plug. If the combustion is not rapid the efficiency of the engine will be reduced proportionately. The compressed charge in the cylinder does not ignite all at once or instantaneously, as many assume, but it is the strata of gas nearest the plug which is ignited first. This in turn sets fire to consecutive layers of the charge until the entire mass is aflame. One may compare the combustion of gas in the gas-engine cylinder to the phenomenon which obtains when a heavy object is thrown into a pool of still water. First a small circle is seen at the point where the object has passed into the water, this circle in turn inducing other and larger circles until the whole surface of the pool has been agitated from the one central point. The method of igniting the gas is very similar, as the spark ignites the circle of gas immediately adjacent to the sparking point, and this circle in turn ignites a little larger one concentric with it. The second circle of flame[197] sets fire to more of the gas, and finally the entire contents of the combustion chamber are burning.

While ordinarily combustion is sufficiently rapid with a single plug so that the proper explosion is obtained at moderate engine speeds, if the engine is working fast and the cylinders are of large capacity more power may be obtained by setting fire to the mixture at two different points instead of but one. This may be accomplished by using two sparking-plugs in the cylinder instead of one, and experiments have shown that it is possible to gain from twenty-five to thirty per cent. in motor power at high speed with two-spark plugs, because the combustion of gas is accelerated by igniting the gas simultaneously in two places. The double-plug system on airplane engines is also a safeguard, as in event of failure of one plug in the cylinder the other would continue to fire the gas, and the engine will continue to function properly.

In using magneto ignition some precautions are necessary relating to wiring and also the character of the spark-plugs employed. The conductor should be of good quality, have ample insulation, and be well protected from accumulations of oil, which would tend to decompose rubber insulation. It is customary to protect the wiring by running it through the conduits of fiber or metal tubing lined with insulating material. Multiple strand cables should be used for both primary and secondary wiring, and the insulation should be of rubber at least ³⁄₁₆ inch thick.

The spark-plugs commonly used for battery and coil ignition cannot always be employed when a magneto is fitted. The current produced by the mechanical generator has a greater amperage and more heat value than that obtained from transformer coils excited by battery current. The greater heat may burn or fuse the slender points used on some battery plugs and heavier electrodes are needed to resist the heating effect of the more intense arc. While the current has greater amperage it is not of as high potential or voltage as that commonly produced by the secondary winding of an induction coil, and it[198] cannot overcome as much of a gap. Manufacturers of magneto plugs usually set the spark points about ¹⁄₆₄ of an inch apart. The most efficient magneto plug has a plurality of points so that when the distance between one set becomes too great the spark will take place between one of the other pairs of electrodes which are not separated by so great an air space.

Airplane work calls for special construction of spark-plugs, owing to the high compression used in the engines and the fact that they are operated on open throttle practically all the time, thus causing a great deal of heat to[199] be developed. The plug shown at Fig. 74 was recently described in “The Automobile,” and has been devised especially for airplane engines and automobile racing power plants. The core C is built up of mica washers, and has square shoulders. As mica washers of different sizes may be used, and accurate machining, such as is necessary with conical clamping surfaces, is not required, the plug can be produced economically. The square shoulders of the core afford two gasket seats, and when the core is clamped in the shell by means of check nut E, it is accurately centered and a tight joint is formed. This construction also makes a shorter plug than where conical fits are used, thus improving the heat radiation through the stem. The lower end of the shell is provided with a baffle plate O, which tends to keep the oil away from the mica. There are perforations L in this baffle plate to prevent burnt gases being pocketed behind the baffle plate and pre-igniting the new charge. This construction also brings the firing point out into the firing chamber of the engine, and has all the other advantages of a closed-end plug. The stem P is made of brass or copper, on account of their superior heat conductivity, and the electrode J is swedged into the bottom of the stem, as shown at K, in a secure manner.

The shell is finned, as shown at G, to provide greater heat radiating surface. There is also a fin F at the top of the stem, to increase the radiation of heat from the stem and electrode. The top of this finned portion is slightly countersunk, and the stem is riveted into same, thereby reducing the possibility of leakage past the threads on the stem. This finned portion is necked at A to take a slip terminal.

In building up the core a small section of washers, I, is built up before the mica insulating tube D is placed on. This construction gives a better support to section I. Baffle plate O is bored out to allow the electrode J to pass through, and the clearance between baffle plate and electrode is made larger than the width of the gap between[200] the firing points, so that there is no danger of the spark jumping from the electrode to the baffle plate.

This plug will be furnished either with or without the finned portion, to meet individual requirements. The manufacturers lay special stress upon the simplicity of construction and upon the method of clamping, which is claimed to make the plug absolutely gas-tight.

[201]

The importance of minimizing friction at the various bearing surfaces of machines to secure mechanical efficiency is fully recognized by all mechanics, and proper lubricity of all parts of the mechanism is a very essential factor upon which the durability and successful operation of the motor car power plant depends. All of the moving members of the engine which are in contact with other portions, whether the motion is continuous or intermittent, of high or low velocity, or of rectilinear or continued rotary nature, should be provided with an adequate supply of oil. No other assemblage of mechanism is operated under conditions which are so much to its disadvantage as the motor car, and the tendency is toward a simplification of oiling methods so that the supply will be ample and automatically applied to the points needing it.

In all machinery in motion the members which are in contact have a tendency to stick to each other, and the very minute projections which exist on even the smoothest of surfaces would have a tendency to cling or adhere to each other if the surfaces were not kept apart by some elastic and unctuous substance. This will flow or spread out over the surfaces and smooth out the inequalities existing[202] which tend to produce heat and retard motion of the pieces relative to each other.

A general impression which obtains is that well machined surfaces are smooth, but while they are apparently free from roughness, and no projections are visible to the naked eye, any smooth bearing surface, even if very carefully ground, will have a rough appearance if examined with a magnifying glass. An exaggerated condition to illustrate this point is shown at Fig. 75. The amount of friction will vary in proportion to the pressure on the surfaces in contact and will augment as the loads increase; the rougher surfaces will have more friction than smoother ones and soft bodies will produce more friction than hard substances.

Friction is always present in any mechanism as a resisting force that tends to retard motion and bring all moving parts to a state of rest. The absorption of power by friction may be gauged by the amount of heat which exists at the bearing points. Friction of solids may be divided into two classes: sliding friction, such as exists between the piston and cylinder, or the bearings of a gas-engine, and rolling friction, which is that present when the load is supported by ball or roller bearings, or that which exists between the tires or the driving wheels and the road. Engineers endeavor to keep friction losses as low as possible, and much care is taken in all modern airplane engines to provide adequate methods of lubrication, or anti-friction bearings at all points where considerable friction exists.

The reason a lubricant is supplied to bearing points will be easily understood if one considers that these elastic substances flow between the close fitting surfaces, and by filling up the minute depressions in the surfaces and covering the high spots act as a cushion which[203] absorbs the heat generated and takes the wear instead of the metallic bearing surface. The closer the parts fit together the more fluid the lubricant must be to pass between their surfaces, and at the same time it must possess sufficient body so that it will not be entirely forced out by the pressure existing between the parts.

Oils should have good adhesive, as well as cohesive, qualities. The former are necessary so that the oil film will cling well to the surfaces of the bearings; the latter, so the oil particles will cling together and resist the tendency to separation which exists all the time the bearings are in operation. When used for gas-engine lubrication the oil should be capable of withstanding considerable heat in order that it will not be vaporized by the hot portions of the cylinder. It should have sufficient cold test so that it will remain fluid and flow readily at low temperature. Lubricants should be free from acid, or alkalies,[204] which tend to produce a chemical action with metals and result in corrosion of the parts to which they are applied. It is imperative that the oil be exactly the proper quality and nature for the purpose intended and that it be applied in a positive manner. The requirements may be briefly summarized as follows:

First—It must have sufficient body to prevent seizing of the parts to which it is applied and between which it is depended upon to maintain an elastic film, and yet it must not have too much viscosity, in order to minimize the internal or fluid friction which exists between the particles of the lubricant itself.

Second—The lubricant must not coagulate or gum; must not injure the parts to which it is applied, either by chemical action or by producing injurious deposits, and it should not evaporate readily.

Third—The character of the work will demand that the oil should not vaporize when heated or thicken to such a point that it will not flow readily when cold.

Fourth—The oil must be free from acid, alkalies, animal or vegetable fillers, or other injurious agencies.

Fifth—It must be carefully selected for the work required and should be a good conductor of heat.

The first oils which were used for lubricating machinery were obtained from animal and vegetable sources, though at the present time most unguents are of mineral derivation. Lubricants may exist as fluids, semifluids, or solids. The viscosity will vary from light spindle or dynamo oils, which have but little more body than kerosene, to the heaviest greases and tallows. The most common solid employed as a lubricant is graphite, sometimes termed “plumbago” or “black lead.” This substance is of mineral derivation.

The disadvantage of oils of organic origin, such as those obtained from animal fats or vegetable substances, is that they will absorb oxygen from the atmosphere,[205] which causes them to thicken or become rancid. Such oils have a very poor cold test, as they solidify at comparatively high temperatures, and their flashing point is so low that they cannot be used at points where much heat exists. In most animal oils various acids are present in greater or less quantities, and for this reason they are not well adapted for lubricating metallic surfaces which may be raised high enough in temperature to cause decomposition of the oils.

Lubricants derived from the crude petroleum are called “Oleonaphthas” and they are a product of the process of refining petroleum through which gasoline and kerosene are obtained. They are of lower cost than vegetable or animal oil, and as they are of non-organic origin, they do not become rancid or gummy by constant exposure to the air, and they will have no corrosive action on metals because they contain no deleterious substances in chemical composition. By the process of fractional distillation mineral oils of all grades can be obtained. They have a lower cold and higher flash test and there is not the liability of spontaneous combustion that exists with animal oils.

The organic oils are derived from fatty substances, which are present in the bodies of all animals and in some portions of plants. The general method of extracting oil from animal bodies is by a rendering process, which consists of applying sufficient heat to liquefy the oil and then separating it from the tissue with which it is combined by compression. The only oil which is used to any extent in gas-engine lubrication that is not of mineral derivation is castor oil. This substance has been used on high-speed racing automobile engines and on airplane power plants. It is obtained from the seeds of the castor plant, which contain a large percentage of oil.

Among the solid substances which may be used for lubricating purposes may be mentioned tallow, which is obtained from the fat of animals, and graphite and soapstone, which are of mineral derivation. Tallow is never[206] used at points where it will be exposed to much heat, though it is often employed as a filler for greases used in transmission gearing of autos. Graphite is sometimes mixed with oil and applied to cylinder lubrication, though it is most often used in connection with greases in the landing gear parts and for coating wires and cables of the airplane. Graphite is not affected by heat, cold, acids, or alkalies, and has a strong attraction for metal surfaces. It mixes readily with oils and greases and increases their efficiency in many applications. It is sometimes used where it would not be possible to use other lubricants because of extremes of temperature.

The oils used for cylinder lubrication are obtained almost exclusively from crude petroleum derived from American wells. Special care must be taken in the selection of crude material, as every variety will not yield oil of the proper quality to be used as a cylinder lubricant. The crude petroleum is distilled as rapidly as possible with fire heat to vaporize off the naphthas and the burning oils. After these vapors have been given off superheated steam is provided to assist in distilling. When enough of the light elements have been eliminated the residue is drawn off, passed through a strainer to free it from grit and earthy matters, and is afterwards cooled to separate the wax from it. This is the dark cylinder oil and is the grade usually used for steam-engine cylinders.

The oil that is to be used in the gasoline engine must be of high quality, and for that reason the best grades are distilled in a vacuum that the light distillates may be separated at much lower temperatures than ordinary conditions of distilling permit. If the degree of heat is not high the product is not so apt to decompose and deposit carbon. If it is desired to remove the color of the oil which is caused by free carbon and other impurities it can be accomplished by filtering the oil through charcoal. The greater the number of times the oil is filtered,[207] the lighter it will become in color. The best cylinder oils have flash points usually in excess of 500 degrees F., and while they have a high degree of viscosity at 100 degrees F. they become more fluid as the temperature increases.

The lubricating oils obtained by refining crude petroleum may be divided into three classes:

First—The natural oils of great body which are prepared for use by allowing the crude material to settle in tanks at high temperature and from which the impurities are removed by natural filtration. These oils are given the necessary body and are free from the volatile substances they contain by means of superheated steam which provides a source of heat.

Second—Another grade of these natural oils which are filtered again at high temperatures and under pressure through beds of animal charcoal to improve their color.

Third—Pale, limpid oils, obtained by distillation and subsequent chemical treatment from the residuum produced in refining petroleum to obtain the fuel oils.

Authorities agree that any form of mixed oil in which animal and mineral lubricants are combined should never be used in the cylinder of a gas engine as the admixture of the lubricants does not prevent the decomposition of the organic oil into the glycerides and fatty acids peculiar to the fat used. In a gas-engine cylinder the flame tends to produce more or less charring. The deposits of carbon will be much greater with animal oils than with those derived from the petroleum base because the constituents of a fat or tallow are not of the same volatile character as those which comprise the hydro-carbon oils which will evaporate or volatilize before they char in most instances.

The suitability of oil for the proper and efficient lubrication of all internal combustion engines is determined chiefly by the following factors:

[208]1. Type of cooling system (operating temperatures).

2. Type of lubricating system (method of applying oil to the moving parts).

3. Rubbing speeds of contact surfaces.

Were the operating temperatures, bearing surface speeds and lubrication systems identical, a single oil could be used in all engines with equal satisfaction. The only change then necessary in viscosity would be that due to climatic conditions. As engines are now designed, only three grades of oil are necessary for the lubrication of all types with the exception of Knight, air-cooled and some engines which run continuously at full load. In the specification of engine lubricants the feature of load carried by the engine should be carefully considered.

Full Load Engines.

Variable Load Engines.

Of the forms outlined, the only one we have any immediate concern about is the airplane power plant. The Platt & Washburn Refining Company, who have made a careful study of the lubrication problem as applied to all types of engines, have found a peculiar set of conditions to apply to oiling high-speed constant-duty or “full-load” engines. Modern airplane engines are designed to operate continuously at a fairly uniform high rotative speed and at full load over long periods of time. As a sequence to this heavy duty the operating[209] temperatures are elevated. For the sake of extreme lightness in weight of all parts, very thin alloy steel aluminum or cast iron pistons are fitted and the temperature of the thin piston heads at the center reaches anywhere between 600° and 1,400° Fahr., as in automobile racing engines. Freely exposed to such intense heat hydro-carbon oils are partially “cracked” into light and heavy products or polymerized into solid hydro-carbons. From these facts it follows that only heavy mineral oils of low carbon residue and of the greatest chemical purity and stability should be used to secure good lubrication. In all cases the oil should be sufficiently heavy to assure the highest horse-power and fuel and oil economy compatible with perfect lubrication, avoiding, at the same time, carbonization and ignition failure. When aluminum pistons are used their superior heat-conducting properties aid materially in reducing the rate of oil destruction.

The extraordinary evolutions described by airplanes in flight make it a matter of vital necessity to operate engines inclined at all angles to the vertical as well as in an upside-down position. To meet this situation lubricating systems have been elaborated so as to deliver an abundance of oil where needed and to eliminate possible flooding of cylinders. This is done by applying a full force feed system, distributing oil under considerable pressure to all working parts. Discharged through the bearings, the oil drains down to the suction side of a second pump located in the bottom of the base chamber. This pump being of greater capacity than the first prevents the accumulation of oil in the crank-case, and forces it to a separate oil reservoir-cooler, whence it flows back in rapid circulation to the pump feeding the bearings. With this arrangement positive lubrication is entirely independent of engine position. The lubricating system of the Thomas-Morse aviation engines, which is shown at Fig. 76, is typical of current practice.

[210]

[211]

The construction and operation of rotative radial cylinder engines introduce additional difficulties of lubrication to those already referred to and merit especial attention. Owing to the peculiar alimentation systems of Gnome type engines, atomized gasoline mixed with air is drawn through the hollow stationary crank-shaft directly into the crank-case which it fills on the way to the cylinders. Therein lies the trouble. Hydrocarbon oils are soon dissolved by the gasoline and washed off, leaving the bearing surfaces without adequate protection and exposed to instant wear and destruction. So castor oil is resorted to as an indispensable but unfortunate compromise. Of vegetable origin, it leaves a much more bulky carbon deposit in the explosion chambers than does mineral oil and its great affinity for oxygen causes the formation of voluminous gummy deposit in the crank-case. Engines employing it need to be dismounted and thoroughly scraped out at frequent intervals. It is advisable to use only unblended chemically pure castor oil in rotative engines, first by virtue of its insolubility in gasoline and second because its extra heavy body can resist the high temperature of air-cooled cylinders.

The oiling system of the Hall-Scott type A-5 125 horse-power engine is clearly shown at Fig. 77. It is completely described in the instruction book issued by the company from which the following extracts are reproduced by permission. Crank-shaft, connecting rods and all other parts within the crank-case and cylinders are lubricated directly or indirectly by a force-feed oiling system. The cylinder walls and wrist pins are lubricated by oil spray thrown from the lower end of connecting rod bearings. This system is used only upon A-5 engines. Upon A-7a and A-5a engines a small tube supplies oil[212] from connecting rod bearing directly upon the wrist pin. The oil is drawn from the strainer located at the lowest portion of the lower crank-case, forced around the main intake manifold oil jacket. From here it is circulated to the main distributing pipe located along the lower left hand side of upper crank-case. The oil is then forced directly to the lower side of crank-shaft, through holes drilled in each main bearing cup. Leakage from these main bearings is caught in scuppers placed upon the cheeks of the crank-shafts furnishing oil under pressure to the connecting rod bearings. A-7a and A-5a engines have small tubes leading from these bearings which convey the oil under pressure to the wrist pins.

[213]

A bi-pass located at the front end of the distributing oil pipe can be regulated to lessen or raise the pressure. By screwing the valve in, the pressure will raise and more oil will be forced to the bearings. By unscrewing, pressure is reduced and less oil is fed. A-7a and A-5a engines have oil relief valves located just off of the main oil pump in the lower crank-case. This regulates the pressure at all times so that in cold weather there will be no danger of bursting oil pipes due to excessive pressure. If it is found the oil pressure is not maintained at a high enough level, inspect this valve. A stronger spring will not allow the oil to bi-pass so freely, and consequently the pressure will be raised; a weaker spring will bi-pass more oil and reduce the oil pressure materially. Independent of the above-mentioned system, a small, directly driven rotary oiler feeds oil to the base of each individual cylinder. The supply of oil is furnished by the main oil pump located in the lower crank-case. A small sight-feed regulator is furnished to control the supply of oil from this oiler. This instrument should be placed higher than the auxiliary oil distributor itself to enable the oil to drain by gravity feed to the oiler. If there is no available place with the necessary height in the front seat of plane, connect it directly to the intake L fitting on the oiler in an upright position. It should[214] be regulated with full open throttle to maintain an oil level in the glass, approximately half way.

An oil pressure gauge is provided. This should be run to the pilot’s instrument board. The gauge registers the oil pressure upon the bearings, also determining its circulation. Strict watch should be maintained of this instrument by pilot, and if for any reason its hand should drop to 0 the motor should be immediately stopped and the trouble found before restarting engine. Care should be taken that the oil does not work up into the gauge, as it will prevent the correct gauge registering of oil pressure. The oil pressure will vary according to weather conditions and viscosity of oil used. In normal weather, with the engine properly warmed up, the pressure will register on the oil gauge from 5 to 10 pounds when the engine is turning from 1,275 to 1,300 r. p. m. This does not apply to all aviation engines, however, as the proper pressure advised for the Curtiss OX-2 motor is from 40 to 55 pounds at the gauge.

The oil sump plug is located at the lowest point of the lower crank-case. This is a combination dirt, water and sediment trap. It is easily removed by unscrewing. Oil is furnished mechanically to the cam-shaft housing under pressure through a small tube leading from the main distributing pipe at the propeller end of engine directly into the end of cam-shaft housing. The opposite end of this housing is amply relieved to allow the oil to rapidly flow down upon cam-shaft, magneto, pinion-shaft, and crank-shaft gears, after which it returns to lower crank-case. An outside overflow pipe is also provided to carry away the surplus oil.

The oil strainer is placed at the lowest point of the lower crank-case. This strainer should be removed after every five to eight hours running of the engine and cleaned thoroughly with gasoline. It is also advisable to squirt distillate up into the case through the opening[215] where the strainer has been removed. Allow this distillate to drain out thoroughly before replacing the plug with strainer attached. Be sure gasket is in place on plug before replacing. Pour new oil in through either of the two breather pipes on exhaust side of motor. Be sure to replace strainer screens if removed. If, through oversight, the engine does not receive sufficient lubrication and begins to heat or pound, it should be stopped immediately. After allowing engine to cool pour at least three gallons of oil into oil sump. Fill radiator with water after engine has cooled. Should there be apparent damage, the engine should be thoroughly inspected immediately without further running. If no obvious damage has been done, the engine should be given a careful examination at the earliest opportunity to see that the running without oil has not burned the bearings or caused other trouble.

Oils best adapted for Hall-Scott engines have the following properties: A flash test of not less than 400° F.; viscosity of not less than 75 to 85 taken at 21° F. with Saybolt’s Universal Viscosimeter.

Zeroline heavy duty oil, manufactured by the Standard Oil Company of California; also,

Gargoyle mobile B oil, manufactured by the Vacuum Oil Company, both fulfill the above specifications. One or the other of these oils can be obtained all over the world.

Monogram extra heavy is also recommended.

The splash system of lubrication that depends on the connecting rod to distribute the lubricant is one of the most successful and simplest forms for simple four- and six-cylinder vertical automobile engines, but is not as well adapted to the oiling of airplane power plants for reasons previously stated. If too much oil is supplied the surplus will work past the piston rings and into the combustion chamber, where it will burn and cause carbon[216] deposits. Too much oil will also cause an engine to smoke and an excess of lubricating oil is usually manifested by a bluish-white smoke issuing from the exhaust.

A good method of maintaining a constant level of oil for the successful application of the splash system is shown at Fig. 78. The engine base casting includes a separate chamber which serves as an oil container and which is below the level of oil in the crank-case. The lubricant is drawn from the sump or oil container by means of a positive oil pump which discharges directly into the engine case. The level is maintained by an overflow pipe which allows all excess lubricant to flow back into the oil container at the bottom of the cylinder. Before passing into the pump again the oil is strained or filtered by a screen of wire gauze and all foreign matter removed. Owing to the rapid circulation of the oil it may be used over and over again for quite a period of time. The oil is introduced directly into the crank-case by a breather pipe and the level is indicated by a rod carried by a float which rises when the container is replenished and falls when the available supply diminishes. It will be noted that with such system the only apparatus required besides the oil tank which is cast integral with the bottom of the crank-case is a suitable pump to maintain circulation of oil. This member is always positively driven, either by means of shaft and universal coupling or direct gearing. As the system is entirely automatic in action, it will furnish a positive supply of oil at all desired points, and it cannot be tampered with by the inexpert because no adjustments are provided or needed.

In most airplane power plants it is considered desirable to supply the oil directly to the parts needing it by suitable leads instead of depending solely upon the distributing action of scoops on the connecting rod big ends. A system of this nature is shown at Fig. 77. The oil[217] is carried in the crank-case, as is common practice, but the normal oil level is below the point where it will be reached by the connecting rod. It is drawn from the crank-case by a plunger pump which directs it to a manifold leading directly to conductors which supply the main[218] journals. After the oil has been used on these points it drains back into the bottom of the crank-case. An excess is provided which is supplied to the connecting rod ends by passages drilled into the webs of the crank-shaft and part way into the crank-pins as shown by the dotted lines. The oil which is present at the connecting rod crank-pins is thrown off by centrifugal force and lubricates the cylinder walls and other internal parts. Regulating screws are provided so that the amount of oil supplied the different points may be regulated at will. A relief check valve is installed to take care of excess lubricant and to allow any oil that does not pass back into the pipe line to overflow or bi-pass into the main container.

A simple system of this nature is shown graphically in a phantom view of the crank-case at Fig. 79, in which[219] the oil passages are made specially prominent. The oil is taken from a reservoir at the bottom of the engine base by the usual form of gear oil pump and is supplied to a main feed manifold which extends the length of the crank-case. Individual conductors lead to the five main bearings, which in turn supply the crank-pins by passages drilled through the crank-shaft web. In this power plant the connecting rods are hollow section bronze castings and the passage through the center of the connecting rod serves to convey the lubricant from the crank-pins to the wrist-pins. The cylinder walls are oiled by the spray of lubricant thrown off the revolving crank-shaft by centrifugal force. Oil projection by the dippers on the connecting rod ends from constant level troughs is unequal upon the cylinder walls of the two-cylinder blocks of an eight- or twelve-cylinder V engine. This gives rise, on one side of the engine, to under-lubrication, and, on the other side, to over-lubrication, as shown at Fig. 80, A. This applies to all modifications of splash lubricating systems.

[220]

When a force-feed lubricating system is used, the oil, escaping past the cheeks of both ends of the crank-pin bearings, is thrown off at a tangent to the crank-pin circle in all directions, supplying the cylinders on both sides with an equal quantity of oil, as at Fig. 80, B.

The reader should understand from preceding chapters that the power of an internal-combustion motor is obtained by the rapid combustion and consequent expansion of some inflammable gas. The operation in brief is that when air or any other gas or vapor is heated, it will expand and that if this gas is confined in a space which will not permit expansion, pressure will be exerted against all sides of the containing chamber. The more a gas is heated, the more pressure it will exert upon the walls of the combustion chamber it[221] confines. Pressure in a gas may be created by increasing its temperature and inversely heat may be created by pressure. When a gas is compressed its total volume is reduced and the temperature is augmented.

The efficiency of any form of heat engine is determined by the power obtained from a certain fuel consumption. A definite amount of energy will be liberated in the form of heat when a pound of any fuel is burned. The efficiency of any heat engine is proportional to the power developed from a definite quantity of fuel with the least loss of thermal units. If the greater proportion of the heat units derived by burning the explosive mixture could be utilized in doing useful work, the efficiency of the gasoline engine would be greater than that of any other form of energizing power. There is a great loss of heat from various causes, among which can be cited the reduction of pressure through cooling the motor and the loss of heat through the exhaust valves when the burned gases are expelled from the cylinder.

The loss through the water jacket of the average automobile power plant is over 50 per cent. of the total fuel efficiency. This means that more than half of the heat units available for power are absorbed and dissipated by the cooling water. Another 16 per cent. is lost through the exhaust valve, and but 33¹⁄₃ per cent. of the heat units do useful work. The great loss of heat through the cooling systems cannot be avoided, as some method must be provided to keep the temperature of the engine within proper bounds. It is apparent that the rapid combustion and continued series of explosions would soon heat the metal portions of the engine to a red heat if some means were not taken to conduct much of this heat away. The high temperature of the parts would burn the lubricating oil, even that of the best quality, and the piston and rings would expand to such a degree, especially when deprived of oil, that they would seize in the cylinder. This would score the walls, and the friction which ensued would tend to bind the parts so tightly[222] that the piston would stick, bearings would be burned out, the valves would warp, and the engine would soon become inoperative.

The best temperature to secure efficient operation is one on which considerable difference of opinion exists among engineers. The fact that the efficiency of an engine is dependent upon the ratio of heat converted into useful work compared to that generated by the explosion of the gas is an accepted fact. It is very important that the engine should not get too hot, and on the other hand it is equally vital that the cylinders be not robbed of too much heat. The object of cylinder cooling is to keep the temperature of the cylinder below the danger point, but at the same time to have it as high as possible to secure maximum power from the gas burned. The usual operating temperatures of an[223] automobile engine are shown at Fig. 81, and this can be taken as an approximation of the temperatures apt to exist in an airplane engine of conventional design as well when at ground level or not very high in the air. The newer very high compression airplane engines in which compressions of eight or nine atmospheres are used, or about 125 pounds per square inch, will run considerably hotter than the temperatures indicated.

There are two general systems of engine cooling in common use, that in which water is heated by the absorption of heat from the engine and then cooled by air, and the other method in which the air is directed onto the cylinder and absorbs the heat directly instead of through the medium of water. When the liquid is employed in cooling it is circulated through jackets which surround the cylinder casting and the water may be kept in motion by two methods. The one generally favored is to use a positive circulating pump of some form which is driven by the engine to keep the water in motion. The other system is to utilize a natural principle that heated water is lighter than cold liquid and that it will tend to rise to the top of the cylinder when it becomes heated to the proper temperature and cooled water takes its place at the bottom of the water jacket.

Air-cooling methods may be by radiation or convection. In the former case the effective outer surface of the cylinder is increased by the addition of flanges machined or cast thereon, and the air is depended on to rise from the cylinder as heated and be replaced by cooler air. This, of course, is found only on stationary engines. When a positive air draught is directed against the cylinder by means of the propeller slip stream in an airplane, cooling is by convection and radiation both. Sometimes the air draught may be directed against the[224] cylinder walls by some form of jacket which confines it to the heated portions of the cylinder.

A typical water-cooling system in which a pump is depended upon to promote circulation of the cooling liquid is shown at Figs. 82 and 83. The radiator is carried at the front end of the fuselage in most cases, and serves as a combined water tank and cooler, but in some cases it is carried at the side of the engine, as in Fig. 84, or attached to the central portion of the aerofoil or wing structure. It is composed of an upper and lower portion joined together by a series of pipes which may be round and provided with a series of fins to radiate the heat, or which may be flat in order to have the water pass through in thin sheets and cool it more easily. Cellular or honeycomb coolers are composed of a large number of bent tubes which will expose a large area of surface to the cooling influence of the air draught forced through the radiator either by the forward movement of the vehicle or by some type of fan. The cellular and[225] flat tube types have almost entirely displaced the flange tube radiators which were formerly popular because they cool the water more effectively, and may be made lighter than the tubular radiator could be for engines of the same capacity.

The water is drawn from the lower header of the radiator by the pump and is forced through a manifold to the lower portion of the water jackets of the cylinder. It becomes heated as it passes around the cylinder walls and combustion chambers and the hot water passes out of the top of the water jacket to the upper portion of the radiator. Here it is divided in thin streams and directed against comparatively cool metal which abstracts the heat from the water. As it becomes cooler it falls to the bottom of the radiator because its weight increases as the temperature becomes lower. By the time it reaches[226] the lower tank of the radiator it has been cooled sufficiently so that it may be again passed around the cylinders of the motor. The popular form of circulating pump is known as the “centrifugal type” because a rotary impeller of paddle-wheel form throws water which it receives at a central point toward the outside and thus causes it to maintain a definite rate of circulation. The pump is always a separate appliance attached to the engine and driven by positive gearing or direct-shaft connection. The centrifugal pump is not as positive as the gear form, and some manufacturers prefer the latter because of the positive pumping features. They are very simple in form, consisting of a suitable cast body in which a pair of spur pinions having large teeth are carried. One of these gears is driven by suitable means, and as it turns the other member they maintain a flow of water around the pump body. The pump should always be installed in series with the water pipe which[227] conveys the cool liquid from the lower compartment of the radiator to the coolest portion of the water jacket.

Some automobile engineers contend that the rapid water circulation obtained by using a pump may cool the cylinders too much, and that the temperature of the engine may be reduced so much that the efficiency will be lessened. For this reason there is a growing tendency to use the natural method of water circulation as the cooling liquid is supplied to the cylinder jackets just below the boiling point and the water issues from the jacket at the top of the cylinder after it has absorbed sufficient heat to raise it just about to the boiling point.

As the water becomes heated by contact with the hot cylinder and combustion-chamber walls it rises to the top of the water jacket, flows to the cooler, where enough of the heat is absorbed to cause it to become sensibly greater in weight. As the water becomes cooler, it falls to the bottom of the radiator and it is again supplied to the water jacket. The circulation is entirely automatic and continues as long as there is a difference in temperature between the liquid in the water spaces of the engine and that in the cooler. The circulation becomes brisker as the engine becomes hotter and thus the temperature of the cylinders is kept more nearly to a fixed point. With the thermosyphon system the cooling liquid is nearly always at its boiling point, whereas if the circulation is maintained by a pump the engine will become cooler at high speed and will heat up more at low speed.

With the thermosyphon, or natural system of cooling, more water must be carried than with the pump-maintained circulation methods. The water spaces around the cylinders should be larger, the inlet and discharge water manifolds should have greater capacity, and be free from sharp corners which might impede the flow. The radiator must also carry more water than the form used in connection with the pump because of the brisker[228] pump circulation which maintains the engine temperature at a lower point. Consideration of the above will show why the pump system is almost universally used in connection with airplane power plant cooling.

The earliest known method of cooling the cylinder of gas-engines was by means of a current of air passed through a jacket which confined it close to the cylinder walls and was used by Daimler on his first gas-engine. The gasoline engine of that time was not as efficient as the later form, and other conditions which materialized made it desirable to cool the engine by water. Even as gasoline engines became more and more perfected there has always existed a prejudice against air cooling, though many forms of engines have been used, both in automobile and aircraft applications where the air-cooling method has proven to be very practical.

The simplest system of air cooling is that in which the cylinders are provided with a series of flanges which increase the effective radiating surface of the cylinder and directing an air-current from a fan against the flanges to absorb the heat. This increase in the available radiating surface of an air-cooled cylinder is necessary because air does not absorb heat as readily as water and therefore more surface must be provided that the excess heat be absorbed sufficiently fast to prevent distortion of the cylinders. Air-cooling systems are based on a law formulated by Newton, which is: “The rate for cooling for a body in a uniform current of air is directly proportional to the speed of the air current and the amount of radiating surface exposed to the cooling effect.”

There are certain considerations which must be taken into account in designing an air-cooled engine, which are often overlooked in those forms cooled by water. Large[229] valves must be provided to insure rapid expulsion of the flaming exhaust gas and also to admit promptly the fresh cool mixture from the carburetor. The valves of air-cooled engines are usually placed in the cylinder-head, in order to eliminate any pockets or sharp passages which would impede the flow of gas or retain some of the products of combustion and their heat. When high power is desired multiple-cylinder engines should be used, as there is a certain limit to the size of a successful[230] air-cooled cylinder. Much better results are secured from those having small cubical contents because the heat from small quantities of gas will be more quickly carried off than from greater amounts. All successful engines of the aviation type which have been air-cooled have been of the multiple-cylinder type.

An air-cooled engine must be placed in the fuselage, as at Fig. 85, in such a way that there will be a positive circulation of air around it all the time that it is in operation. The air current may be produced by the tractor screw at the front end of the motor, or by a suction or blower fan attached to the crank-shaft as in the Renault engine or by rotating the cylinders as in the Le Rhone and Gnome motors. Greater care is required in lubrication of the air-cooled cylinders and only the best quality of oil should be used to insure satisfactory oiling.

The combustion chambers must be proportioned so that distribution of metal is as uniform as possible in order to prevent uneven expansion during increase in temperature and uneven contraction when the cylinder is cooled. It is essential that the inside walls of the combustion chamber be as smooth as possible because any sharp angle or projection may absorb sufficient heat to remain incandescent and cause trouble by igniting the mixture before the proper time. The best grades of cast iron or steel should be used in the cylinder and piston and the machine work must be done very accurately so the piston will operate with minimum friction in the cylinder. The cylinder bore should not exceed 4¹⁄₂ or 5 inches and the compression pressure should never exceed 75 pounds absolute, or about five atmospheres, or serious overheating will result.

As an example of the care taken in disposing of the exhaust gases in order to obtain practical air-cooling, some cylinders are provided with a series of auxiliary exhaust ports uncovered by the piston when it reaches the end of its power stroke. The auxiliary exhaust ports open just as soon as the full force of the explosion has[231] been spent and a portion of the flaming gases is discharged through the ports in the bottom of the cylinder. Less of the exhaust gases remains to be discharged through the regular exhaust member in the cylinder-head and this will not heat the walls of the cylinder nearly as much as the larger quantity of hot gas would. That the auxiliary exhaust port is of considerable value is conceded by many designers of fixed and fan-shaped air-cooled motors for airplanes.

Among the advantages stated for direct air cooling, the greatest is the elimination of cooling water and its cooling auxiliaries, which is a factor of some moment, as it permits considerable reduction in horse-power-weight ratio of the engine, something very much to be desired. In the temperate zone, where the majority of airplanes are used, the weather conditions change in a very few months from the warm summer to the extreme cold winter, and when water-cooled systems are employed it is necessary to add some chemical substance to the water to prevent it from freezing. The substances commonly employed are glycerine, wood alcohol, or a saturated solution of calcium chloride. Alcohol has the disadvantage in that it vaporizes readily and must be often renewed. Glycerine affects the rubber hose, while the calcium chloride solution crystallizes and deposits salt in the radiator and water pipes.

One of the disadvantages of an air-cooling method, as stated by those who do not favor this system, is that engines cooled by air cannot be operated for extended periods under constant load or at very high speed without heating up to such a point that premature ignition of the charge may result. The water-cooling systems, at the other hand, maintain the temperature of the engine more nearly constant than is possible with an air-cooled motor, and an engine cooled by water can be operated under conditions of inferior lubrication or poor mixture adjustment that would seriously interfere with proper and efficient cooling by air.

[232]Air-cooled motors, as a rule, use less fuel than water-cooled engines, because the higher temperature of the cylinder does not permit of a full charge of gas being inspired on the intake stroke. As special care is needed in operating an air-cooled engine to obtain satisfactory results and because of the greater difficulty which obtains in providing proper lubrication and fuel mixtures which will not produce undue heating, the air-cooled system has but few adherents at the present time, and practically all airplanes, with but very few exceptions, are provided with water-cooled power plants. Those fitted with air-cooled engines are usually short-flight types where maximum lightness is desired in order to obtain high speed and quick climb. The water-cooled engines are best suited for airplanes intended for long flights. The Gnome, Le Rhone and Clerget engines are thoroughly practical and have been widely used in France and England. These are rotary radial cylinder types. The Anzani is a fixed cylinder engine used on training machines, while the Renault is a V-type engine made in eight- and twelve-cylinder V forms that has been used on reconnaissance and bombing airplanes with success. These types will be fully considered in proper sequence.

[233]

Methods of Cylinder Construction—Block Castings—Influence on Crank-Shaft Design—Combustion Chamber Design—Bore and Stroke Ratio—Meaning of Piston Speed—Advantage of Off-Set Cylinders—Valve Location of Vital Import—Valve Installation Practice—Valve Design and Construction—Valve Operation—Methods of Driving Cam-Shaft—Valve Springs—Valve Timing—Blowing Back—Lead Given Exhaust Valve—Exhaust Closing, Inlet Opening—Closing the Inlet Valve—Time of Ignition—How an Engine Is Timed—Gnome “Monosoupape” Valve Timing—Springless Valves—Four Valves per Cylinder.

The improvements noted in the modern internal combustion motors have been due to many conditions. The continual experimenting by leading mechanical minds could have but one ultimate result. The parts of the engines have been lightened and strengthened, and greater power has been obtained without increasing piston displacement. A careful study has been made of the many conditions which make for efficient motor action, and that the main principles are well recognized by all engineers is well shown by the standardization of design noted in modern power plants. There are many different methods of applying the same principle, and it will be the purpose of this chapter to define the ways in which the construction may be changed and still achieve the same results. The various components may exist in many different forms, and all have their advantages and disadvantages. That all methods are practical is best shown by the large number of successful engines which use radically different designs.

One of the most important parts of the gasoline engine and one that has material bearing upon its efficiency is the cylinder unit. The cylinders may be cast[234] individually, or in pairs, and it is possible to make all cylinders a unit or block casting. Some typical methods of cylinder construction are shown in accompanying illustrations. The appearance of individual cylinder castings may be ascertained by examination of the Hall-Scott airplane engine. Air-cooled engine cylinders are always of the individual pattern.

Considered from a purely theoretical point of view, the individual cylinder casting has much in its favor. It is advanced that more uniform cooling is possible than where the cylinders are cast either in pairs or three or four in one casting. More uniform cooling insures that the expansion or change of form due to heating will be more equal. This is an important condition because the cylinder bore must remain true under all conditions of operation. If the heating effect is not uniform, which condition is liable to obtain if metal is not evenly distributed, the cylinder may become distorted by heat and the bore be out of truth. When separate cylinders are used it is possible to make a uniform water space and have the cooling liquid evenly distributed around the cylinder. In multiple cylinder castings this is not always the rule, as in many instances, especially in four-cylinder block motors where compactness is the main feature, there is but little space between the cylinders for the passage of water. Under such circumstances the cooling effect is not even, and the stresses which obtain because of unequal expansion may distort the cylinder to some extent. When steel cylinders are made from forgings, the water jackets are usually of copper or sheet steel attached to the forging by autogenous welding; in the case of the latter and, in some cases, the former may be electro-deposited on the cylinders.

The advantage of casting the cylinders in blocks is that a motor may be much shorter than it would be if individual castings were used. It is admitted that when[235] the cylinders are cast together a more compact, rigid, and stronger power plant is obtained than when cast separately. There is a disadvantage, however, in that if one cylinder becomes damaged it will be necessary to replace the entire unit, which means scrapping three good cylinders because one of the four has failed. When the cylinders are cast separately one need only replace the one that has become damaged. The casting of four cylinders in one unit is made possible by improved[236] foundry methods, and when proper provision is made for holding the cores when the metal is poured and the cylinder casts are good, the construction is one of distinct merit. It is sometimes the case that the proportion of sound castings is less when cylinders are cast in block, but if the proper precautions are observed in molding and the proper mixtures of cast iron used, the ratio of defective castings is no more than when cylinders are molded individually. As an example of the courage of engineers in departing from old-established rules, the cylinder casting shown at Fig. 86 may be considered typical. This is used on the Duesenberg four-cylinder sixteen-valve 4³⁄₄′′ × 7′′ engine which has a piston displacement of 496 cu. in. At a speed of 2,000 r.p.m., corresponding to a piston speed of 2,325 ft. per min., the engine is guaranteed to develop 125 horse-power. The weight of the model engine without gear reduction is 436 lbs., but a number of refinements have been made in the design whereby it is expected to get the weight down to 390 lbs. The four cylinders are cast from semi-steel in a single block, with integral heads. The cylinder construction is the same as that which has always been used by Mr. Duesenberg, inlet and exhaust valves being arranged horizontally opposite each other in the head. There are large openings in the water jacket at both sides and at the ends, which are closed by means of aluminum covers, water-tightness being secured by the use of gaskets. This results in a saving in weight because the aluminum covers can be made considerably lighter than it would be possible to cast the jacket walls, and, besides, it permits of obtaining a more nearly uniform thickness of cylinder wall, as the cores can be much better supported. The cooling water passes completely around each cylinder, and there is a very considerable space between the two central cylinders, this being made necessary in order to get the large bearing area desirable for the central bearing.

It is common practice to cast the water jackets integral[237] with the cylinders, if cast iron or aluminum is used, and this is also the most economical method of applying it because it gives good results in practice. An important detail is that the water spaces must be proportioned so that they are equal around the cylinders whether these members are cast individually, in pairs, threes or fours. When cylinders are cast in block form it is good practice to leave a large opening in the jacket wall which will assist in supporting the core and make for uniform water space. It will be noticed that the casting shown at Fig. 86 has a large opening in the side of the cylinder block. These openings are closed after the interior of the casting is thoroughly cleaned of all sand, core wire, etc., by brass, cast iron or aluminum plates. These also have particular value in that they may be removed after the motor has been in use, thus permitting one to clean out the interior of the water jacket and dispose of the rust, sediment, and incrustation which are always present after the engine has been in active service for a time.

Among the advantages claimed for the practice of casting cylinders in blocks may be mentioned compactness, lightness, rigidity, simplicity of water piping, as well as permitting the use of simple forms of inlet and exhaust manifolds. The light weight is not only due to the reduction of the cylinder mass but because the block construction permits one to lighten the entire motor. The fact that all cylinders are cast together decreases vibration, and as the construction is very rigid, disalignment of working parts is practically eliminated. When inlet and exhaust manifolds are cored in the block casting, as is sometimes the case, but one joint is needed on each of these instead of the multiplicity of joints which obtain when the cylinders are individual castings. The water piping is also simplified. In the case of a four-cylinder block motor but two pipes are used; one for the water to enter the cylinder jacket, the other for the cooling liquid to discharge through.

[238]

The method of casting the cylinders has a material influence on the design of the crank-shaft as will be shown in proper sequence. When four cylinders are combined in one block it is possible to use a two-bearing crank-shaft. Where cylinders are cast in pairs a three-bearing crank-shaft is commonly supplied, and when cylinders are cast as individual units it is thought necessary to supply a five-bearing crank-shaft, though sometimes shafts having but three journals are used successfully. Obviously the shafts must be stronger and stiffer to withstand the stresses imposed if two supporting bearings are used than if a larger number are employed. In this connection it may be stated that there is less difficulty in securing alignment with a lesser number of bearings and there is also less friction. On the other hand, the greater the number of points of support a crank-shaft has the lighter the webs can be made and still have requisite strength.

[239]

Another point of importance in the design of the cylinder, and one which has considerable influence upon the power developed, is the shape of the combustion chamber. The endeavor of designers is to obtain maximum power from a cylinder of certain proportions, and the greater energy obtained without increasing piston displacement or fuel consumption the higher the efficiency of the motor. To prevent troubles due to pre-ignition it is necessary that the combustion chamber be made so that there will be no roughness, sharp corners, or edges of metal which may remain incandescent when heated or which will serve to collect carbon deposits by providing a point of anchorage. With the object of providing an absolutely clean combustion chamber some makers use a separable head unit to their twin cylinder castings, such as shown at Fig. 87 and Fig. 88. These permit one to machine the entire interior of the cylinder and combustion chamber. The relation of valve location and combustion chamber design will be considered in proper sequence. These cylinders are cast of aluminum, instead of cast iron, as[240] is customary, and are provided with steel or cast iron cylinder liners forced in the soft metal casting bores.

A question that has been a vexed one and which has been the subject of considerable controversy is the proper proportion of the bore to the stroke. The early gas engines had a certain well-defined bore to stroke ratio, as it was usual at that time to make the stroke twice as long as the bore was wide, but this cannot be done when high speed is desired. With the development of the present-day motor the stroke or piston travel has been gradually shortened so that the relative proportions of bore and stroke have become nearly equal. Of late there seems to be a tendency among designers to return to the proportions which formerly obtained, and the stroke is sometimes one and a half or one and three-quarter times the bore.

Engines designed for high speed should have the stroke not much longer than the diameter of the bore. The disadvantage of short-stroke engines is that they will not pull well at low speeds, though they run with great regularity and smoothness at high velocity. The long-stroke engine is much superior for slow speed work, and it will pull steadily and with increasing power at low speed. It was formerly thought that such engines should never turn more than a moderate number of revolutions, in order not to exceed the safe piston speed of 1,000 feet per minute. This old theory or rule of practice has been discarded in designing high efficiency automobile racing and aviation engines, and piston speeds from 2,500 to 3,000 feet per minute are sometimes used, though the average is around 2,000 feet per minute. While both short- and long-stroke motors have their advantages, it would seem desirable to average between the two. That is why a proportion of four to five or six seems to be more general than that of four to seven or eight, which would be a long-stroke ratio. Careful analysis of a number[241] of foreign aviation motors shows that the average stroke is about 1.2 times the bore dimensions, though some instances were noted where it was as high as 1.7 times the bore.

The factor which limits the stroke and makes the speed of rotation so dependent upon the travel of the piston is piston speed. Lubrication is the main factor which determines piston speed, and the higher the rate of piston travel the greater care must be taken to insure proper oiling. Let us fully consider what is meant by piston speed.

Assume that a motor has a piston travel or stroke of six inches, for the sake of illustration. It would take two strokes of the piston to cover one foot, or twelve inches, and as there are two strokes to a revolution it will be seen that this permits of a normal speed of 1,000 revolutions per minute for an engine with a six-inch stroke, if one does not exceed 1,000 feet per minute. If the stroke was only four inches, a normal speed of 1,500 revolutions per minute would be possible without exceeding the prescribed limit. The crank-shaft of a small engine, having three-inch stroke, could turn at a speed of 2,000 revolutions per minute without danger of exceeding the safe speed limit. It will be seen that the longer the stroke the slower the speed of the engine, if one desires to keep the piston speed within the bounds as recommended, but modern practice allows of greatly exceeding the speeds formerly thought best.

Another point upon which considerable difference of opinion exists relates to the method of placing the cylinder upon the crank-case—i.e., whether its center line should be placed directly over the center of the crank-shaft, or to one side of center. The motor shown at Fig. 90 is an off-set type, in that the center line of the[242] cylinder is a little to one side of the center of the crank-shaft. Diagrams are presented at Fig. 91 which show the advantages of off-set crank-shaft construction. The view at A is a section through a simple motor with the conventional cylinder placing, the center line of both crank-shaft and cylinder coinciding. The view at B shows[243] the cylinder placed to one side of center so that its center line is distinct from that of the crank-shaft and at some distance from it. The amount of off-set allowed is a point of contention, the usual amount being from fifteen to twenty-five per cent. of the stroke. The advantages of the off-set are shown at Fig. 91, C. If the crank turns in direction of the arrow there is a certain resistance to motion which is proportional to the amount of energy exerted by the engine and the resistance offered by the load. There are two thrusts acting against the cylinder wall to be considered, that due to explosion or expansion of the gas, and that which resists the motion of the piston. These thrusts may be represented by arrows, one which acts directly in a vertical direction on the piston top, the[244] other along a straight line through the center of the connecting rod. Between these two thrusts one can draw a line representing a resultant force which serves to bring the piston in forcible contact with one side of the cylinder wall, this being known as side thrust. As shown at C, the crank-shaft is at 90 degrees, or about one-half stroke, and the connecting rod is at 20 degrees angle. The shorter connecting rod would increase the diagonal resultant and side thrusts, while a longer one would reduce the angle of the connecting rod and the side thrust of the piston would be less. With the off-set construction, as shown at D, it will be noticed that with the same connecting-rod length as shown at C and with the crank-shaft at 90 degrees of the circle that the connecting-rod angle is 14 degrees and the side thrust is reduced proportionately.

Another important advantage is that greater efficiency is obtained from the explosion with an off-set crank-shaft, because the crank is already inclined when the piston is at top center, and all the energy imparted to the piston by the burning mixture can be exerted directly into producing a useful turning effort. When a cylinder is placed directly on a line with the crank-shaft, as shown at A, it will be evident that some of the force produced by the expansion of the gas will be exerted in a direct line and until the crank moves the crank throw and connecting rod are practically a solid member. The pressure which might be employed in obtaining useful turning effort is wasted by causing a direct pressure upon the lower half of the main bearing and the upper half of the crank-pin bushing.

Very good and easily understood illustrations showing advantages of the off-set construction are shown at E and F. This is a bicycle crank-hanger. It is advanced that the effort of the rider is not as well applied when the crank is at position E as when it is at position F. Position E corresponds to the position of the parts when the cylinder is placed directly over the crank-shaft center.[245] Position F may be compared to the condition which is present when the off-set cylinder construction is used.

It has often been said that a chain is no stronger than its weakest link, and this is as true of the explosive motor as it is of any other piece of mechanism. Many motors which appeared to be excellently designed and which were well constructed did not prove satisfactory because some minor detail or part had not been properly considered by the designer. A factor having material bearing upon the efficiency of the internal combustion motor is the location of the valves and the shape of the combustion chamber which is largely influenced by their placing. The fundamental consideration of valve design is that the gases be admitted and discharged from the cylinder as quickly as possible in order that the speed of gas flow will not be impeded and produce back pressure. This is imperative in obtaining satisfactory operation in any form of motor. If the inlet passages are constricted the cylinder will not fill with explosive mixture promptly, whereas if the exhaust gases are not fully expelled the parts of the inert products of combustion retained dilute the fresh charge, making it slow burning and causing lost power and overheating. When an engine employs water as a cooling medium this substance will absorb the surplus heat readily, and the effects of overheating are not noticed as quickly as when air-cooled cylinders are employed. Valve sizes have a decided bearing upon the speed of motors and some valve locations permit the use of larger members than do other positions.

While piston velocity is an important factor in determinations of power output, it must be considered from the aspect of the wear produced upon the various parts of the motor. It is evident that engines which run very fast, especially of high power, must be under a greater strain than those operating at lower speeds. The valve-operating mechanism is especially susceptible to the influence[246] of rapid movement, and the slower the engine the longer the parts will wear and the more reliable the valve action.

As will be seen by reference to the accompanying illustration, Fig. 92, there are many ways in which valves may be placed in the cylinder. Each method outlined possesses some point of advantage, because all of the types[247] illustrated are used by reputable automobile manufacturers. The method outlined at Fig. 92, A, is widely used, and because of its shape the cylinder is known as the “T” form. It is approved for automobile use for several reasons, the most important being that large valves can be employed and a well-balanced and symmetrical cylinder casting obtained. Two independent cam-shafts are needed, one operating the inlet valves, the other the exhaust members. The valve-operating mechanism can be very simple in form, consisting of a plunger actuated by the cam which transmits the cam motion to the valve-stem, raising the valve as the cam follower rides on the point of the cam. Piping may be placed without crowding, and larger manifolds can be fitted than in some other constructions. This has special value, as it permits the use of an adequate discharge pipe on the exhaust side with its obvious advantages. This method of cylinder construction is never found on airplane engines because it does not permit of maximum power output.

On the other hand, if considered from a viewpoint of actual heat efficiency, it is theoretically the worst form of combustion chamber. This disadvantage is probably compensated for by uniformity of expansion of the cylinder because of balanced design. The ignition spark-plug may be located directly over the inlet valve in the path of the incoming fresh gases, and both valves may be easily removed and inspected by unscrewing the valve caps without taking off the manifolds.

The valve installation shown at C is somewhat unusual, though it provides for the use of valves of large diameter. Easy charging is insured because of the large inlet valve directly in the top of the cylinder. Conditions may be reversed if necessary, and the gases discharged through this large valve. Both methods are used, though it would seem that the free exhaust provided by allowing the gases to escape directly from the combustion chamber through the overhead valve to the exhaust manifold[248] would make for more power. The method outlined at Fig. 92, F and at Fig. 90 is one that has been widely employed on large automobile racing motors where extreme power is required, as well as in engines constructed for aviation service. The inclination of the valves permits the use of large valves, and these open directly into the combustion chamber. There are no pockets to retain heat or dead gas, and free intake and outlet of gas is obtained. This form is quite satisfactory from a theoretical point of view because of the almost ideal combustion chamber form. Some difficulty is experienced, however, in properly water-jacketing the valve chamber which experience has shown to be necessary if the engine is to have high power.

The motor shown at Fig. 92, B and Fig. 88 employs cylinders of the “L” type. Both valves are placed in a common extension from the combustion chamber, and being located side by side both are actuated from a common cam-shaft. The inlet and exhaust pipes may be placed on the same side of the engine and a very compact assemblage is obtained, though this is optional if passages are cored in the cylinder pairs to lead the gases to opposite sides. The valves may be easily removed if desired, and the construction is fairly good from the viewpoint of both foundry man and machinist. The chief disadvantage is the limited area of the valves and the loss of heat efficiency due to the pocket. This form of combustion chamber, however, is more efficient than the “T” head construction, though with the latter the use of larger valves probably compensates for the greater heat loss. It has been stated as an advantage of this construction that both manifolds can be placed at the same side of the engine and a compact assembly secured. On the other hand, the disadvantage may be cited that in order to put both pipes on the same side they must be of smaller size than can be used when the valves are oppositely placed. The “L” form cylinder is sometimes made more efficient if but one valve is placed in the pocket[249] while the other is placed over it. This construction is well shown at Fig. 92, D and is found on Anzani motors.

The method of valve application shown at Fig. 87 is an ingenious method of overcoming some of the disadvantages inherent with valve-in-the-head motors. In the first place it is possible to water-jacket the valves thoroughly, which is difficult to accomplish when they are mounted in cages. The water circulates directly around the walls of the valve chambers, which is superior to a construction where separate cages are used, as there are two thicknesses of metal with the latter, that of the valve-cage proper and the wall of the cylinder. The cooling medium is in contact only with the outer wall, and as there is always a loss of heat conductivity at a joint it[250] is practically impossible to keep the exhaust valves and their seats at a uniform temperature. The valves may be of larger size without the use of pockets when seating directly in the head. In fact, they could be equal in diameter to almost half the bore of the cylinder, which provides an ideal condition of charge placement and exhaust. When valve grinding is necessary the entire head is easily removed by taking off six nuts and loosening inlet manifold connections, which operation would be necessary even if cages were employed, as in the engine shown at Fig. 93.

[251]At Fig. 94, A and B, a section through a typical “L”-shaped cylinder is depicted. It will be evident that where a pocket construction is employed, in addition to its faculty for absorbing heat, the passage of gas would be impeded. For example, the inlet gas rushing in through the open valve would impinge sharply upon the valve-cap or combustion head directly over the valve and then must turn at a sharp angle to enter the combustion chamber[252] and then at another sharp angle to fill the cylinders. The same conditions apply to the exhaust gases, though they are reversed. When the valve-in-the-head type of cylinder is employed, as at C, the only resistance offered the gas is in the manifold. As far as the passage of the gases in and out of the cylinder is concerned, ideal conditions obtain. It is claimed that valve-in-the-head motors are more flexible and responsive than other forms, but the construction has the disadvantage in that the valves must be opened through a rather complicated system of push rods and rocker arms instead of the simpler and direct plunger which can be used with either the “T” or “L” head cylinders. This is clearly outlined in the illustrations at Fig. 95, where A shows the valve in the head-operating mechanism necessary if the cam-shaft is carried at the cylinder base, while B shows the most direct push-rod action obtained with “T” or “L” head cylinder placing.

[253] Fig. 97.

CENSORED

Fig. 98.

CENSORED

The objection can be easily met by carrying the cam-shaft above the cylinders and driving it by means of gearing. The types of engine cylinders using this construction are shown at Fig. 96, and it will be evident that a positive and direct valve action is possible by following the construction originated by the Mercedes (German)[254] aviation engine designers and outlined at A. The other forms at B and C are very clearly adaptations of this design. The Hall-Scott engine at Fig. 97 is depicted in part section and no trouble will be experienced in understanding the bevel pinion and gear drive from the crank-shaft[255] to the overhead cam-shaft through a vertical counter-shaft. A very direct valve action is used in the Duesenberg engines, one of which is shown in part section at Fig. 98. The valves are parallel with the piston top and are actuated by rocker arms, one end of which bears against the valve stem, and the other rides the cam-shaft.

The form shown at Fig. 99 shows an ingenious application of the valve-in-the-head idea which permits one to obtain large valves. It has been used on some of the Panhard aviation engines and on the American Aeromarine power plants. The inlet passage is controlled by the sliding sleeve which is hollow and slotted so as to permit the inlet gases to enter the cylinder through the regular type poppet valve which seats in the exhaust sleeve. When the exhaust valve is operated by the tappet rod and rocker arm the intake valve is also carried[256] down with it. The intake gas passage is closed, however, and the burned gases are discharged through the large annular passage surrounding the sleeve. When the inlet valve leaves its seat in the sleeve the passage of cool gas around the sleeve keeps the temperature of both valves to a low point and the danger of warping is minimized. A dome-shaped combustion chamber may be used, which is an ideal form in conserving heat efficiency, and as large valves may be installed the flow of both fresh and exhaust gases may be obtained with minimum resistance. The intake valve is opened by a small auxiliary rocker arm which is lifted when the cam follower rides into the depression in the cam by the action of the strong spring around the push rod. When the cam follower rides on the high point the exhaust sleeve is depressed from its seat against the cylinder. By using a cam having both positive and negative profiles, a single rod suffices for both valves because of its push and pull action.

Valve dimensions are an important detail to be considered and can be determined by several conditions, among which may be cited method of installation, operating mechanism, material employed, engine speed desired, manner of cylinder cooling and degree of lift desired. A review of various methods of valve location has shown that when the valves are placed directly in the head we can obtain the ideal cylinder form, though larger valves may be used if housed in a separate pocket, as afforded by the “T” head construction. The method of operation has much to do with the size of the valves. For example, if an automatic inlet valve is employed it is good practice to limit the lift and obtain the required area of port opening by augmenting the diameter. Because of this a valve of the automatic type is usually made twenty per cent. larger than one mechanically operated. When both are actuated by cam mechanism, as is now common practice, they are usually made the same[257] size and are interchangeable, which greatly simplifies manufacture. The relation of valve diameter to cylinder bore is one that has been discussed for some time by engineers. The writer’s experience would indicate that they should be at least half the bore, if possible. While the mushroom type or poppet valve has become standard and is the most widely used form at the present time, there is some difference of opinion among designers as to the materials employed and the angle of the seat. Most valves have a bevel seat, though some have a flat seating. The flat seat valve has the distinctive advantage of providing a clear opening with lesser lift, this conducing to free gas flow. It also has value because it is silent in operation, but the disadvantage is present that best material and workmanship must be used in their construction to obtain satisfactory results. As it can be made very light it is particularly well adapted for use as an automatic inlet valve. Among other disadvantages cited is the claim that it is more susceptible to derangement, owing to the particles of foreign matter getting under the seat. With a bevel seat it is argued that the foreign matter would be more easily dislodged by the gas flow, and that the valve would close tighter because it is drawn positively against the bevel seat.

Several methods of valve construction are the vogue, the most popular form being the one-piece type; those which are composed of a head of one material and stem of another are seldom used in airplane engines because they are not reliable. In the built-up construction the head is usually of high nickel steel or cast iron, which metals possess good heat-resisting qualities. Heads made of these materials are not likely to warp, scale, or pit, as is sometimes the case when ordinary grades of machinery steel are used. The cast-iron head construction is not popular because it is often difficult to keep the head tight on the stem. There is a slight difference in expansion ratio between the head and the stem, and as the stem is either screwed or riveted to the cast-iron head[258] the constant hammering of the valve against its seat may loosen the joint. As soon as the head is loose on the stem the action of the valve becomes erratic. The best practice is to machine the valves from tungsten steel forgings. This material has splendid heat-resisting qualities and will not pit or become scored easily. Even the electrically welded head to stem types which are used in automobile engines are not looked upon with favor in the aviation engine. Valve stem guides and valve stems must be machined very accurately to insure correct action. The usual practice in automobile engines is shown at Fig. 100.

The methods of valve operation commonly used vary according to the type of cylinder construction employed. In all cases the valves are lifted from their seats by cam-actuated mechanism. Various forms of valve-lifting cams are shown at Fig. 101. As will be seen, a cam consists[259] of a circle to which a raised, approximately triangular member has been added at one point. When the cam follower rides on the circle, as shown at Fig. 102, there is no difference in height between the cam center and its periphery and there is no movement of the plunger. As soon as the raised portion of the cam strikes the plunger it will lift it, and this reciprocating movement is transmitted to the valve stem by suitable mechanical connections.

The cam forms outlined at Fig. 101 are those commonly used. That at A is used on engines where it is desired to obtain a quick lift and to keep the valve fully opened as long as possible. It is a noisy form, however, and is not very widely employed. That at B is utilized more often as an inlet cam while the profile shown at C is generally depended on to operate exhaust valves. The cam shown at D is a composite form which has some of the features of the other three types. It will give the quick opening of form A, the gradual closing of form B, and the time of maximum valve opening provided by cam profile C.

The various types of valve plungers used are shown at Fig. 102. That shown at A is the simplest form, consisting of a simple cylindrical member having a rounded end which follows the cam profile. These are sometimes[260] made of square stock or kept from rotating by means of a key or pin. A line contact is possible when the plunger is kept from turning, whereas but a single point bearing is obtained when the plunger is cylindrical and free to revolve. The plunger shown at A will follow only cam profiles which have gradual lifts. The plunger shown at B is left free to revolve in the guide bushing and is provided with a flat mushroom head which serves as a cam follower. The type shown at C carries a roller at its lower end and may follow very irregular cam profiles if abrupt lifts are desired. While forms A and B are the simplest, that outlined at C in its various forms is more widely used. Compound plungers are used on the Curtiss OX-2 motors, one inside the other. The small or inner one works on a cam of conventional design, the outer plunger follows a profile having a flat spot to permit of a pull rod action instead of a push rod action. All the methods in which levers are used to operate valves are more or less noisy because clearance must be left between the valve stem and the stop of the plunger. The space must be taken up before the valve will leave its seat, and when[261] the engine is operated at high speeds the forcible contact between the plunger and valve stem produces a rattling sound until the valves become heated and expand and the stems lengthen out. Clearance must be left between the valve stems and actuating means. This clearance is clearly shown in Fig. 103 and should be .020′′ (twenty thousandths) when engine is cold. The amount of clearance allowed depends entirely upon the design of the engine and length of valve stem. On the Curtiss OX-2 engines the clearance is but .010′′ (ten thousandths) because the valve stems are shorter. Too little clearance will result in loss of power or misfiring when engine is hot. Too much clearance will not allow the valve to open its full amount and will disturb the timing.

Two systems of cam-shaft operation are used. The most common of these is by means of gearing of some form. If the cam-shaft is at right angles to the crank-shaft it may be driven by worm, spiral, or bevel gearing.[262] If the cam-shaft is parallel to the crank-shaft, simple spur gear or chain connection may be used to turn it. A typical cam-shaft for an eight-cylinder V engine is shown at Fig. 104. It will be seen that the sixteen cams are forged integrally with the shaft and that it is spur-gear driven. The cam-shaft drive of the Hall-Scott motor is shown at Fig. 97.

While gearing is more commonly used, considerable attention has been directed of late to silent chains for cam-shaft operation. The ordinary forms of block or roller chain have not proven successful in this application, but the silent chain, which is in reality a link belt operating over toothed pulleys, has demonstrated its worth. The tendency to its use is more noted on foreign motors than those of American design. It first came to public notice when employed on the Daimler-Knight engine for driving the small auxiliary crank-shafts which reciprocated the sleeve valves. The advantages cited for the application of chains are, first, silent operation, which obtains even after the chains have worn considerably; second, in designing it is not necessary to figure on maintaining certain absolute center distances between the crank-shaft and cam-shaft sprockets, as would be the case if conventional forms of gearing were used. On some forms of motor employing gears, three and even four[263] members are needed to turn the cam-shaft. With a chain drive but two sprockets are necessary, the chain forming a flexible connection which permits the driving and driven members to be placed at any distance apart that the exigencies of the design demand. When chains are used it is advised that some means for compensating chain slack be provided, or the valve timing will lag when chains are worn. Many combination drives may be worked out with chains that would not be possible with other forms of gearing. Direct gear drive is favored at the present time by airplane engine designers because they are the most certain and positive means, even when a number of gears must be used as intermediate drive members. With overhead cam-shafts, bevel gears work out very well in practice, as in the Hall-Scott motors and others of that type.

Another consideration of importance is the use of proper valve-springs, and particular care should be taken with those, of automatic valves. The spring must be weak enough to allow the valve to open when the suction is light, and must be of sufficient strength to close it in time at high speeds. It should be made as large as possible in diameter and with a large number of convolutions, in order that fatigue of the metal be obviated, and it is imperative that all springs be of the same strength when used on a multiple-cylinder engine. Practically all valves used to control the gas flow in airplane engines are mechanically operated. On the exhaust valve the spring must be strong enough so that the valve will not be sucked in on the inlet stroke. It should be borne in mind that if the spring is too strong a strain will be imposed on the valve-operating mechanism, and a hammering action produced which may cause deformation of the valve-seat. Only pressure enough to insure that the operating mechanism will follow the cam is required. It is common practice to make the inlet and exhaust valve springs of[264] the same tension when the valves are of the same size and both mechanically operated. This is done merely to simplify manufacture and not because it is necessary for the inlet valve-spring to be as strong as the other. Valve springs of the helical coil type are generally used, though torsion or “scissors” springs and laminated or single-leaf springs are also utilized in special applications. Two[265] springs are used on each valve in some valve-in-the-head types; a spring of small pitch diameter inside the regular valve-spring and concentric with it. Its function is to keep the valve from falling into the cylinder in event of breakage of the main spring in some cases, and to provide a stronger return action in others.

[266]

The sectional view through the cylinder at Fig. 105 shows the Knight sliding sleeves and their actuating means very clearly. The diagrams at Fig. 106 show graphically the sleeve movements and their relation to the crank-shaft and piston travel. The action may be summed up as follows: The inlet port begins to open when the lower edge of the opening of the outside sleeve which is moving down passes the top of the slot in the inner member also moving downwardly. The inlet port is closed when the lower edge of the slot in the inner sleeve which is moving up passes the top edge of the port in the outer sleeve which is also moving toward the top of the cylinder. The inlet opening extends over two hundred degrees of crank motion. The exhaust port is uncovered slightly when the lower edge of the port in the inner sleeve which is moving down passes the lower edge of the portion of the cylinder head which protrudes in the cylinder. When the top of the port in the outer sleeve traveling toward the bottom of the cylinder passes the lower edge of the slot in the cylinder wall the exhaust passage is closed. The exhaust opening extends over a period corresponding to about two hundred and forty degrees of crank motion. The Knight motor has not been applied to aircraft to the writer’s knowledge, but an eight-cylinder Vee design that might be useful in that connection if lightened is shown at Fig. 107. The main object is to show that the Knight valve action is the only other besides the mushroom or poppet valve that has been applied successfully to high speed gasoline engines.

It is in valve timing that the greatest difference of opinion prevails among engineers, and it is rare that one will see the same formula in different motors. It is true that the same timing could not be used with motors of[267] different construction, as there are many factors which determine the amount of lead to be given to the valves. The most important of these is the relative size of the valve to the cylinder bore, the speed of rotation it is desired to obtain, the fuel efficiency, the location of the valves, and other factors too numerous to mention.

Most of the readers should be familiar with the cycle of operation of the internal combustion motor of the four-stroke type, and it seems unnecessary to go into detail except to present a review. The first stroke of the piston is one in which a charge of gas is taken into the[268] motor; the second stroke, which is in reverse direction to the first, is a compression stroke, at the end of which the spark takes place, exploding the charge and driving the piston down on the third or expansion stroke, which is in the same direction as the intake stroke, and finally, after the piston has nearly reached the end of this stroke, another valve opens to allow the burned gases to escape, and remains open until the piston has reached the end of the fourth stroke and is in a position to begin the series over again. The ends of the strokes are reached when the piston comes to a stop at either top or bottom of the cylinder and reverses its motion. That point is known as a center, and there are two for each cylinder, top and bottom centers, respectively.

All circles may be divided into 360 parts, each of which is known as a degree, and, in turn, each of these degrees may be again divided into minutes and seconds, though we need not concern ourselves with anything less than the degree. Each stroke of the piston represents 180 degrees travel of the crank, because two strokes represent one complete revolution of three hundred and sixty degrees. The top and bottom centers are therefore separated by 180 degrees. Theoretically each phase of a four-cycle engine begins and ends at a center, though in actual practice the inertia or movement of the gases makes it necessary to allow a lead or lag to the valve, as the case may be. If a valve opens before a center, the distance is called “lead”; if it closes after a center, this distance is known as “lag.” The profile of the cams ordinarily used to open or close the valves represents a considerable time in relation to the 180 degrees of the crank-shaft travel, and the area of the passages through which the gases are admitted or exhausted is quite small owing to the necessity of having to open or close the valves at stated times; therefore, to open an adequately large passage for the gases it is necessary to open the valves earlier and close them later than at centers.

That advancing the opening of the exhaust valve was[269] of value was discovered on the early motors and is explained by the necessity of releasing a large amount of gas, the volume of which has been greatly raised by the heat of combustion. When the inlet valves were mechanically operated it was found that allowing them to lag at closing enabled the inspiration of a greater volume of gas. Disregarding the inertia or flow of the gases, opening the exhaust at center would enable one to obtain full value of the expanding gases the entire length of the piston stroke, and it would not be necessary to keep the valve open after the top center, as the reverse stroke would produce a suction effect which might draw some of the inert charge back into the cylinder. On the other hand, giving full consideration to the inertia of the gas, opening the valve before center is reached will provide for quick expulsion of the gases, which have sufficient velocity at the end of the stroke, so that if the valve is allowed to remain open a little longer, the amount of lag varying with the opinions of the designer, the cylinder is cleared in a more thorough manner.

When the factor of retarded opening is considered without reckoning the inertia of the gases, it would appear that if the valve were allowed to remain open after center had passed, say, on the closing of the inlet, the piston, having reversed its motion, would have the effect of expelling part of the fresh charge through the still open valve as it passed inward at its compression stroke. This effect is called blowing back, and is often noted with motors where the valve settings are not absolutely correct, or where the valve-springs or seats are defective and prevent proper closing.

This factor is not of as much import as might appear, as on closer consideration it will be seen that the movement of the piston as the crank reaches either end of the stroke is less per degree of angular movement than it is when the angle of the connecting rod is greater. Then,[270] again, a certain length of time is required for the reversal of motion of the piston, during which time the crank is in motion but the piston practically at a standstill. If the valves are allowed to remain open during this period, the passage of the gas in or out of the cylinder will be by its own momentum.

The faster a motor turns, all other things being equal, the greater the amount of lead or advance it is necessary to give the opening of the exhaust valve. It is self-evident truth that if the speed of a motor is doubled it travels twice as many degrees in the time necessary to lower the pressure. As most designers are cognizant of this fact, the valves are proportioned accordingly. It is well to consider in this respect that the cam profile has much to do with the manner in which the valve is opened; that is, the lift may be abrupt and the gas allowed to escape in a body, or the opening may be gradual, the gas issuing from the cylinder in thin streams. An analogy may be made with the opening of any bottle which contains liquid highly carbonated. If the cork is removed suddenly the gas escapes with a loud pop, but, on the other hand, if the bottle is uncorked gradually, the gas escapes from the receptacle in thin streams around the cork, and passage of the gases to the air is accomplished without noise. While the second plan is not harsh, it is slower than the former, as must be evident.

A point which has been much discussed by engineers is the proper relation of the closing of the exhaust valve and the opening of the inlet. Theoretically they should succeed each other, the exhaust closing at upper dead center and the inlet opening immediately afterward. The reason why a certain amount of lag is given the exhaust closing in practice is that the piston cannot drive the[271] gases out of the cylinder unless they are compressed to a degree in excess of that existing in the manifold or passages, and while toward the end of the stroke this pressure may be feeble, it is nevertheless indispensable. At the end of the piston’s stroke, as marked by the upper dead center, this compression still exists, no matter how little it may be, so that if the exhaust valve is closed and the inlet opened immediately afterward, the pressure which exists in the cylinder may retard the entrance of the fresh gas and a certain portion of the inert gas may penetrate into the manifold. As the piston immediately begins to aspirate, this may not be serious, but as these gases are drawn back into the cylinder the fresh charge will be diluted and weakened in value. If the spark-plug is in a pocket, the points may be surrounded by this weak gas, and the explosion will not be nearly as energetic as when the ignition spark takes place in pure mixture.

It is a well-known fact that the exhaust valve should close after dead center and that a certain amount of lag should be given to opening of the inlet. The lag given the closing of the exhaust valve should not be as great as that given the closing of the inlet valve. Assuming that the excess pressure of the exhaust will equal the depression during aspiration, the time necessary to complete the emptying of the cylinder will be proportional to the volume of the gas within it. At the end of the suction stroke the volume of gas contained in the cylinder is equal to the cylindrical volume plus the space of the combustion chamber. At the end of the exhaust stroke the volume is but that of the dead space, and from one-third to one-fifth its volume before compression. While it is natural to assume that this excess of burned gas will escape faster than the fresh gas will enter the cylinder, it will be seen that if the inlet valve were allowed to lag twenty degrees, the exhaust valve lag need not be more than five degrees, providing that the capacity of the combustion chamber was such that the gases occupied one-quarter of their former volume.

[272]It is evident that no absolute rule can be given, as back pressure will vary with the design of the valve passages, the manifolds, and the construction of the muffler. The more direct the opening, the sooner the valve can be closed and the better the cylinder cleared. Ten degrees represent an appreciable angle of the crank, and the time required for the crank to cover this angular motion is not inconsiderable and an important quantity of the exhaust may escape, but the piston is very close to the dead center after the distance has been covered.

Before the inlet valve opens there should be a certain depression in the cylinder, and considerable lag may be allowed before the depression is appreciable. So far as the volume of fresh gas introduced during the admission stroke is concerned, this is determined by the displacement of the piston between the point where the inlet valve opens and the point of closing, assuming that sufficient gas has been inspired so that an equilibrium of pressure has been established between the interior of the cylinder and the outer air. The point of inlet opening varies with different motors. It would appear that a fair amount of lag would be fifteen degrees past top center for the inlet opening, as a certain depression will exist in the cylinder, assuming that the exhaust valve has closed five or ten degrees after center, and at the same time the piston has not gone down far enough on its stroke to materially decrease the amount of gas which will be taken into the cylinder.

As in the case with the other points of opening and closing, there is a wide diversity of practice as relates to closing the inlet valve. Some of the designers close this exactly at bottom center, but this practice cannot be commended, as there is a considerable portion of time, at least ten or fifteen degrees angular motion of the crank, before the piston will commence to travel to any extent on its compression stroke. The gases rushing into the[273] cylinder have considerable velocity, and unless an equilibrium is obtained between the pressure inside and that of the atmosphere outside, they will continue to rush into the cylinder even after the piston ceases to exert any suction effect.

For this reason, if the valve is closed exactly on center, a full charge may not be inspired into the cylinder, though if the time of closing is delayed, this momentum or inertia of the gas will be enough to insure that a maximum charge is taken into the cylinder. The writer considers that nothing will be gained if the valve is allowed to remain open longer than twenty degrees, and an analysis of practice in this respect would seem to confirm this opinion. From that point in the crank movement the piston travel increases and the compressive effect is appreciable, and it would appear that a considerable proportion of the charge might be exhausted into the manifold and carburetor if the valve were allowed to remain open beyond a point corresponding to twenty degrees angular movement of the crank.

In this country engineers unite in providing a variable time of ignition, though abroad some difference of opinion is noted on this point. The practice of advancing the time of ignition, when affected electrically, was severely condemned by early makers, these maintaining that it was necessary because of insufficient heat and volume of the spark, and it was thought that advancing ignition was injurious. The engineers of to-day appreciate the fact that the heat of the electric spark, especially when from a mechanical generator of electrical energy, is the only means by which we can obtain practically instantaneous explosion, as required by the operation of motors at high speeds, and for the combustion of large volumes of gas.

It is apparent that a motor with a fixed point of[274] ignition is not as desirable, in every way, as one in which the ignition can be advanced to best meet different requirements, and the writer does not readily perceive any advantage outside of simplicity of control in establishing a fixed point of ignition. In fact, there seems to be some difference of opinion among those designers who favor[275] fixed ignition, and in one case this is located forty-three degrees ahead of center, and in another motor the point is fixed at twenty degrees, so that it may be said that this will vary as much as one hundred per cent. in various forms. This point will vary with different methods of ignition, as well as the location of the spark-plug or igniter. For the sake of simplicity, most airplane engines use set spark; if an advancing and retarding mechanism is fitted, it is only to facilitate starting, as the spark is kept advanced while in flight, and control is by throttle alone.

It is obvious by consideration of the foregoing that there can be no arbitrary rules established for timing,[276] because of the many conditions which determine the best times for opening and closing the valves. It is customary to try various settings when a new motor is designed until the most satisfactory points are determined, and the setting which will be very suitable for one motor is not always right for one of different design. The timing diagram shown at Fig. 108 applies to the Hall-Scott engine, and may be considered typical. It should be easily followed in view of the very complete explanation given in preceding pages. Another six-cylinder engine diagram is shown at Fig. 109, and an eight-cylinder timing diagram is shown at Fig. 110. In timing automobile engines no trouble is experienced, because timing marks[277] are always indicated on the engine fly-wheel register with an indicating trammel on the crank-case. To time an airplane engine accurately, as is necessary to test for a suspected cam-shaft defect, a timing disc of aluminum is attached to the crank-shaft which has the timing marks indicated thereon. If the disc is made 10 or 12 inches in diameter, it may be divided into degrees without difficulty.

In timing a motor from the marks on the timing disc rim it is necessary to regulate the valves of but one cylinder at a time. Assuming that the disc is revolving in the direction of engine rotation, and that the firing order of the cylinders is 1-3-4-2, the operation of timing would be carried on as follows: The crank-shaft would be revolved until the line marked “Exhaust opens 1 and 4” registered with the trammel on the motor bed. At this point the exhaust-valve of either cylinder No. 1 or No. 4 should begin to open. This can be easily determined by noting which of these cylinders holds the compressed charge ready for ignition. Assuming that the spark has occurred in cylinder No. 1, then when the fly-wheel is turned from the position to that in which the line marked “Exhaust opens 1 and 4” coincides with the trammel point, the valve-plunger under the exhaust-valve of cylinder No. 1 should be adjusted in such a way that there is no clearance between it and the valve stem. Further movement of the wheel in the same direction should produce a lift of the exhaust valve. The disc is turned about two hundred and twenty-five degrees, or a little less than three-quarters of a revolution; then the line marked “Exhaust closes 1 and 4” will register with the trammel point. At this period the valve-plunger and the valve-stem should separate and a certain amount of clearance obtain between them. The next cylinder to time would be No. 3. The crank-shaft is rotated until mark “Exhaust opens 2 and 3” comes in line with the trammel. At this[278] point the exhaust valve of cylinder No. 3 should be just about opening. The closing is determined by rotating the shaft until the line “Exhaust closes 2 and 3” comes under the trammel.

This operation is carried on with all the cylinders, it being well to remember that but one cylinder is working at a time and that a half-revolution of the fly-wheel corresponds to a full working stroke of all the cylinders, and that while one is exhausting the others are respectively taking in a new charge, compressing and exploding. For instance, if cylinder No. 1 has just completed its power-stroke, the piston in cylinder No. 3 has reached the point where the gas may be ignited to advantage. The piston of cylinder No. 4, which is next to fire, is at the bottom of its stroke and will have inspired a charge, while cylinder No. 2, which is the last to fire, will have just finished expelling a charge of burned gas, and will be starting the intake stroke. This timing relates to a four-cylinder engine in order to simplify the explanation. The timing instructions given apply only to the conventional motor types. Rotary cylinder engines, especially the Gnome “monosoupape,” have a distinctive valve timing on account of the peculiarities of design.

In the present design of the Gnome motor, a cycle of operations somewhat different from that employed in the ordinary four-cycle engine is made use of, says a writer in “The Automobile,” in describing the action of this power-plant. This cycle does away with the need for the usual inlet valve and makes the engine operable with only a single valve, hence the name monosoupape, or “single-valve.” The cycle is as follows: A charge being compressed in the outer end of the cylinder or combustion chamber, it is ignited by a spark produced by the spark-plug located in the side of this chamber, and the burning charge expands as the piston moves down in the cylinder while the latter revolves around the crank-shaft. When[279] the piston is about half-way down on the power stroke, the exhaust valve, which is located in the center of the cylinder-head, is mechanically opened, and during the following upstroke of the piston the burnt gases are expelled from the cylinder through the exhaust valve directly into the atmosphere.

Instead of closing at the end of the exhaust stroke, or a few degrees thereafter, the exhaust valve is held open for about two-thirds of the following inlet stroke of the piston, with the result that fresh air is drawn through the exhaust valve into the cylinder. When the cylinder is still 65 degrees from the end of the inlet half-revolution, the exhaust valve closes. As no more air can get into the cylinder, and as the piston continues to move inwardly, it is obvious that a partial vacuum is formed.

When the cylinder approaches within 20 degrees of the end of the inlet half-revolution a series of small inlet ports all around the circumference of the cylinder wall is uncovered by the top edge of the piston, whereby the combustion chamber is placed in communication with the crank chamber. As the pressure in the crank chamber is substantially atmospheric and that in the combustion chamber is below atmospheric, there results a suction effect which causes the air from the crank chamber to flow into the combustion chamber. The air in the crank chamber is heavily charged with gasoline vapor, which is due to the fact that a spray nozzle connected with the gasoline supply tank is located inside the chamber. The proportion of gasoline vapor in the air in the crank chamber is several times as great as in the ordinary combustible mixture drawn from a carburetor into the cylinder. This extra-rich mixture is diluted in the combustion chamber with the air which entered it through the exhaust valve during the first part of the inlet stroke, thus forming a mixture of the proper proportion for complete combustion.

The inlet ports in the cylinder wall remain open until[280] 20 degrees of the compression half-revolution has been completed, and from that moment to near the end of the compression stroke the gases are compressed in the cylinder. Near the end of the stroke ignition takes place and this completes the cycle.

The exact timing of the different phases of the cycle is shown in the diagram at Fig. 111. It will be seen that ignition occurs substantially 20 degrees ahead of the outer dead center, and expansion of the burning gases continues until 85 degrees past the outer dead center, when the piston is a little past half-stroke. Then the exhaust-valve opens and remains open for somewhat more than a complete revolution of the cylinders, or, to be exact, for 390 degrees of cylinder travel, until 115 degrees past the top dead center on the second revolution. Then for 45 degrees of travel the charge within the cylinder is expanded, whereupon the inlet ports are uncovered and remain open for 40 degrees of cylinder travel, 20 degrees on each side of the inward dead center position.

Springless valves are the latest development on French racing car engines, and it is possible that the positively-operated types will be introduced on aviation engines also. Two makes of positively-actuated valves are shown at Fig. 112. The positive-valve motor differs from the conventional form by having no necessity for valve-springs, as a cam not only assures the opening of the valve, but also causes it to return to the valve-seat. In this respect it is much like the sleeve-valve motor, where the uncovering of the ports is absolutely positive. The cars equipped with these valves were a success in long-distance auto races. Claims made for this type of valve mechanism include the possibility of a higher number of revolutions and consequently greater engine power. With the spring-controlled, single-cam operated valve a point is reached where the spring is not capable of returning the valve[281] to its seat before the cam has again begun its opening movement. It is possible to extend the limits considerably by using a light valve on a strong spring, but the valve still remains a limiting factor in the speed of the motor.

A part sectional view through a cylinder of an engine designed by G. Michaux is shown at Fig. 112, A. There are two valves per cylinder, inclined at about ten degrees from the vertical. The valve-stems are of large diameter, as owing to positive control, there is no necessity of lightening this part in an unusual degree. A single overhead[282] cam-shaft has eight pairs of cams, which are shown in detail at B. For each valve there is a three-armed rocker, one arm of which is connected to the stem of the valve and the two others are in contact respectively with the opening and closing cams. The connection to the end of the valve-stem is made by a short connecting link, which is screwed on to the end of the valve-stem and locked in position. This allows some adjustment to be made between the valves and the actuating rocker. It will be evident that one cam and one rocker arm produce the opening of the valve and that the corresponding rocker arm and cam result in the closing of the valve. If the opening cam has the usual convex profile, the closing cam has a correspondingly concave profile. It will be noticed that a light valve-spring is shown in drawing. This is provided to give a final seating to its valve after[283] it has been closed by the cam. This is not absolutely necessary, as an engine has been run successfully without these springs. The whole mechanism is contained within an overhead aluminum cover.

The positive-valve system used on the De Lage motor is shown at D. In this the valves are actuated as shown in sectional views D and E. The valve system is unique in that four valves are provided per cylinder, two for exhaust and two for intake. The valves are mounted side by side, as shown at E, so the double actuator member may be operated by a single set of cams. The valve-operating member consists of a yoke having guide bars at the top and bottom. The actuating cam works inside of this yoke. The usual form of cam acts on the lower portion of the yoke to open the valve, while the concave cam acts on the upper part to close the valves. In this design provision is made for expansion of the valve-stems due to heat, and these are not positively connected to the actuating member. As shown at E, the valves are held against the seat by short coil springs at the upper end of the stem. These are very stiff and are only intended to provide for expansion. A slight space is left between the top of the valve-stem and the portion of the operating member that bears against them when the regular profile cam exerts its pressure on the bottom of the valve-operating mechanism. Another novelty in this motor design is that the cam-shafts and the valve-operating members are carried in casing attached above the motor by housing supports in the form of small steel pillars. The overhead cam-shafts are operated by means of bevel gearing.

Mention has been previously made of the sixteen-valve four-cylinder Duesenberg motor and its great power output for the piston displacement. This is made possible by the superior volumetric efficiency of a motor provided with four valves in each cylinder instead of[284] but two. This principle was thoroughly tried out in racing automobile motors, and is especially valuable in permitting of greater speed and power output from simple four- and six-cylinder engines. On eight- and twelve-cylinder types, it is doubtful if the resulting complication due to using a very large number of valves would be worth while. When extremely large valves are used, as shown in diagram at Fig. 113, it is difficult to have them open directly into the cylinder, and pockets are sometimes necessary. A large valve would weigh more than two smaller valves having an area slightly larger in the aggregate; it would require a stiffer valve spring on account of its greater weight. A certain amount of metal in the valve-head is necessary to prevent warping; therefore, the inertia forces will be greater in the large valve than in the two smaller valves. As a greater port[286] area is obtained by the use of two valves, the gases will be drawn into the cylinder or expelled faster than with a lesser area. Even if the areas are practically the same as in the diagram at Fig. 113, the smaller valves may have a greater lift without imposing greater stresses on the valve-operating mechanism and quicker gas intake and exhaust obtained. The smaller valves are not affected by heat as much as larger ones are. The quicker gas movements made possible, as well as reduction of[287] inertia forces, permits of higher rotative speed, and, consequently, greater power output for a given piston displacement. The drawings at Fig. 114 show a sixteen-valve motor of the four-cylinder type that has been designed for automobile racing purposes, and it is apparent that very slight modifications would make it suitable for aviation purposes. Part of the efficiency is due to the reduction of bearing friction by the use of ball bearings, but the multiple-valve feature is primarily responsible for the excellent performance.

[288]

The piston is one of the most important parts of the gasoline motor inasmuch as it is the reciprocating member that receives the impact of the explosion and which transforms the power obtained by the combustion of gas to mechanical motion by means of the connecting rod to which it is attached. The piston is one of the simplest elements of the motor, and it is one component which does not vary much in form in different types of motors. The piston is a cylindrical member provided with a series of grooves in which packing rings are placed on the outside and two bosses which serve to hold the wrist pin in its interior. It is usually made of cast iron or aluminum, though in some motors where extreme lightness is desired, such as those used for aëronautic work, it may be made of steel. The use of the more resisting material enables the engineer to use lighter sections where it is important that the weight of this member be kept as low as possible consistent with strength.

A number of piston types are shown at Fig. 116. That at A has a round top and is provided with four split packing rings and two oil grooves. A piston of this type is generally employed in motors where the combustion chamber is large and where it is desired to obtain a higher degree of compression than would be possible with a flat top piston. This construction is also stronger because of the arched piston top. The most common form[289] of piston is that shown at B, and it differs from that previously described only in that it has a flat top. The piston outlined in section at C is a type used on some of the sleeve-valve motors of the Knight pattern, and has a concave head instead of the convex form shown at A. The design shown at D in side and plan views is the conventional form employed in two-cycle engines. The deflector plate on the top of the cylinder is cast integral and is utilized to prevent the incoming fresh gases from flowing directly over the piston top and out of the exhaust port, which is usually opposite the inlet opening. On these types of two-cycle engines where a two-diameter cylinder is employed, the piston shown at E is[291] used. This is known as a “differential piston,” and has an enlarged portion at its lower end which fits the pumping cylinder. The usual form of deflector plate is provided at the top of the piston and one may consider it as two pistons in one.

One of the important conditions in piston design is the method of securing the wrist pin which is used to[292] connect the piston to the upper end of the connecting rod. Various methods have been devised to keep the pin in place, the most common of these being shown at Fig. 117. The wrist pin should be retained by some positive means which is not liable to become loose under the vibratory stresses which obtain at this point. If the[293] wrist pin was free to move it would work out of the bosses enough so that the end would bear against the cylinder wall. As it is usually made of steel, which is a harder material than cast iron used in cylinder construction, the rubbing action would tend to cut a groove in the cylinder wall which would make for loss of power because it would permit escape of gas. The wrist pin member is a simple cylindrical element that fits the bosses closely, and it may be either hollow or solid stock. A typical piston and connecting rod assembly which shows a piston in section also is given at Fig. 118. The piston of the Sturtevant aëronautical motor is shown at Fig. 119, the aluminum piston of the Thomas airplane motor with piston rings in place is shown at Fig. 120. A good view of the wrist pin and connecting rod are also given. The iron piston of the Gnome “Monosoupape” airplane engine and the unconventional connecting rod assembly are clearly depicted at Fig 121.

The method of retention shown at A is the simplest and consists of a set screw having a projecting portion[294] passing into the wrist pin and holding it in place. The screw is kept from turning or loosening by means of a check nut. The method outlined at B is similar to that shown at A, except that the wrist pin is solid and the point of the set screw engages an annular groove turned in the pin for its reception. A very positive method is shown at C. Here the retention screws pass into the wrist pin and are then locked by a piece of steel wire which passes through suitable holes in the ends. The method outlined at D is sometimes employed, and it varies from that shown at C only in that the locking wire, which is made of spring steel, is passed through the heads of the locking screws. Some designers machine a large groove around the piston at such a point that when the wrist pin is put in place a large packing ring may be sprung in the groove and utilized to hold the wrist pin in place.

The system shown at F is not so widely used as the simpler methods, because it is more costly and does not offer any greater security when the parts are new than the simple lock shown at A. In this a hollow wrist pin is used, having a tapered thread cut at each end. The wrist pin is slotted at three or four points, for a distance equal to the length of the boss, and when taper expansion plugs[295] are screwed in place the ends of the wrist pin are expanded against the bosses. This method has the advantage of providing a certain degree of adjustment if the wrist pin should loosen up after it has been in use for some time. The taper plugs would be screwed in deeper and the ends of the wrist pin expanded proportionately to take up the loss motion. The method shown at G is an ingenious one. One of the piston bosses is provided with a projection which is drilled out to receive a plunger. The wrist pin is provided with a hole of sufficient size to receive the plunger, which is kept in place by means of a spring in back of it. This makes a very positive lock and one that can be easily loosened when it is desired to remove the wrist pin. To unlock, a piece of fine rod is thrust into the hole at the bottom of the boss which pushes the plunger back against the spring until the wrist pin can be pushed out of the piston.

Some engineers think it advisable to oscillate the wrist pin in the piston bosses, instead of in the connecting rod small end. It is argued that this construction gives more bearing surface at the wrist pin and also provides for more strength because of the longer bosses that can be used. When this system is followed the piston pin is held in place by locking it to the connecting rod by some means. At H the simplest method is outlined. This consisted of driving a taper pin through both rod and wrist pin and then preventing it from backing out by putting a split cotter through the small end of the tapered locking pin. Another method, which is depicted at I, consists of clamping the wrist pin by means of a suitable bolt which brings the slit connecting rod end together as shown.

Aluminum pistons outlined at Fig. 122, have replaced cast iron members in many airplane engines, as these weigh about one-third as much as the cast iron forms of the same size, while the reduction in the inertia forces[296] has made it possible to increase the engine speed without correspondingly stressing the connecting rods, crank-shaft and engine bearings.

Aluminum has not only been used for pistons, but a number of motors will be built for the coming season that will use aluminum cylinder block castings as well. Of course, the aluminum alloy is too soft to be used as a bearing for the piston, and it will not withstand the hammering action of the valve. This makes the use of cast iron or steel imperative in all motors. When used in connection with an aluminum cylinder block the cast iron pieces are placed in the mould so that they act as cylinder liners and valve seats, and the molten metal is poured around them when the cylinder is cast. It is said that this construction results in an intimate bond between the cast iron and the surrounding aluminum metal. Steel liners may also be pressed into the aluminum cylinders after these are bored out to receive them. Aluminum has for a number of years been used in many motor[297] car parts. Alloys have been developed that have greater strength than cast iron and that are not so brittle. Its use for manifolds and engine crank and gear cases has been general for a number of years.

At first thought it would seem as though aluminum would be entirely unsuited for use in those portions of internal combustion engines exposed to the heat of the explosion, on account of the low melting point of that metal and its disadvantageous quality of suddenly “wilting” when a critical point in the temperature is reached. Those who hesitated to use aluminum on account of this defect lost sight of the great heat conductivity of that metal, which is considerably more than that of cast iron. It was found in early experiments with aluminum pistons that this quality of quick radiation meant that aluminum pistons remained considerably cooler than cast iron ones in service, which was attested to by the reduced formation of carbon deposit thereon. The use of aluminum makes possible a marked reduction in power plant weight. A small four-cylinder engine which was not particularly heavy even with cast iron cylinders was found to weigh 100 pounds less when the cylinder block, pistons, and upper half of the crank-case had been made of aluminum instead of cast iron. Aluminum motors are no longer an experiment, as a considerable number of these have been in use on cars during the past year without the owners of the cars being apprised of the fact. Absolutely no complaint was made in any case of the aluminum motor and it was demonstrated, in addition to the saving in weight, that the motors cost no more to assemble and cooled much more efficiently than the cast iron form. One of the drawbacks to the use of aluminum is its growing scarcity, which results in making it a “near precious” metal.

As all pistons must be free to move up and down in the cylinder with minimum friction, they must be less in[298] diameter than the bore of the cylinder. The amount of freedom or clearance provided varies with the construction of the engine and the material the piston is made of, as well as its size, but it is usual to provide from .005 to .010 of an inch to compensate for the expansion of the piston due to heat and also to leave sufficient clearance for the introduction of lubricant between the working surfaces. Obviously, if the piston were not provided with packing rings, this amount of clearance would enable a portion of the gases evolved when the charge is exploded to escape by it into the engine crank-case. The packing members or piston rings, as they are called, are split rings of cast iron, which are sprung into suitable grooves machined on the exterior of the piston, three or four of these being the usual number supplied. These have sufficient elasticity so that they bear tightly against the cylinder wall and thus make a gas-tight joint. Owing to the limited amount of surface in contact with the cylinder wall and the elasticity of the split rings the amount of friction resulting from the contact of properly fitted rings and the cylinder is not of enough moment to cause any damage and the piston is free to slide up and down in the cylinder bore.

These rings are made in two forms, as outlined at Fig. 123. The design shown at A is termed a “concentric[299] ring,” because the inner circle is concentric with the outer one and the ring is of uniform thickness at all points. The ring shown at B is called an “eccentric ring,” and it is thicker at one part than at others. It has theoretical advantages in that it will make a tighter joint than the other form, as it is claimed its expansion due to heat is more uniform. The piston rings must be split in order that they may be sprung in place in the grooves, and also to insure that they will have sufficient elasticity to take the form of the cylinder at the different points in their travel. If the cylinder bore varies by small amounts the rings will spring out at the points where the bore is larger than standard, and spring in at those portions where it is smaller than standard.

It is important that the joint should be as nearly gas-tight as possible, because if it were not a portion of the gases would escape through the slots in the piston rings. The joint shown at C is termed a “lap joint,” because the ends of the ring are cut in such a manner that they overlap. This is the approved joint. The butt joint shown at D is seldom used and is a very poor form, the only advantage being its cheapness. The diagonal cut shown at E is a compromise between the very good form shown at C and the poor joint depicted at D. It is also widely used, though most constructors prefer the lap joint, because it does not permit the leakage of gas as much as the other two types.

There seems to be some difference of opinion relative to the best piston ring type—some favoring the eccentric pattern, others the concentric form. The concentric ring has advantages from the lubricating engineer’s point of view; as stated by the Platt & Washburn Company in their text-book on engine lubrication, the smaller clearance behind the ring possible with the ring of uniform section is advantageous.

Fig. 124, A, shows a concentric piston ring in its groove. Since the ring itself is concentric with the groove, very small clearance between the back of the ring[300] and the bottom of its groove may be allowed. Small clearance leaves less space for the accumulation of oil and carbon deposits. The gasket effect of this ring is uniform throughout the entire length of its edges, which is its marked advantage over the eccentric ring. This type of piston ring rarely burns fast in its groove. There are a large number of different concentric rings manufactured of different designs and of different efficiency.

Figs. 124, B and 124, C show eccentric rings assembled in the ring groove. It will be noted that there is a large space between the thin ends of this ring and the bottom of the groove. This empty space fills up with oil which in the case of the upper ring frequently is carbonized, restricting the action of the ring and nullifying its usefulness. The edges of the thin ends are not sufficiently wide to prevent rapid escape of gases past them. In a practical way this leakage means loss of compression and noticeable drop in power. When new and properly fitted, very little difference can be noted between the tightness of eccentric and concentric rings. Nevertheless, after several months’ use, a more rapid leakage will always occur past the eccentric than past the concentric. If continuous trouble with the carbonization of cylinders, smoking and sooting of spark-plugs is experienced, it is[301] a sure indication that mechanical defects exist in the engine, assuming of course, that a suitable oil has been used. Such trouble can be greatly lessened, if not entirely eliminated, by the application of concentric rings (lap joint), of any good make, properly fitted into the grooves of the piston. Too much emphasis cannot be put upon this point. If the oil used in the engine is of the correct viscosity, and serious carbon deposit, smoking, etc., still result, the only certain remedy then is to have the cylinders rebored and fitted with properly designed, oversized pistons and piston rings.

In order to reduce the compression loss and leakage of gas by the ordinary simple form of diagonal or lap joint one-piece piston ring a number of compound rings have been devised and are offered by their makers to use in making replacements. The leading forms are shown at Fig. 125. That shown at A is known as the “Statite” and consists of three rings, one carried inside while the other two are carried on the outside. The ring shown at B is a double ring and is known as the McCadden. This is composed of two thin concentric lap joint rings so disposed relative to each other that the opening in the inner ring comes opposite to the opening in the outer ring.

The form shown at C is known as the “Leektite,” and is a single ring provided with a peculiar form of lap and dove tail joint. The ring shown at D is known as the “Dunham” and is of the double concentric type being composed of two rings with lap joints which are welded together at a point opposite the joint so that there is no passage by which the gas can escape. The Burd high compression ring is shown at E. The joints of these rings are sealed by means of an H-shaped coupler of bronze which closes the opening. The ring ends are made with tongues which interlock with the coupling. The[302] ring shown at F is called the “Evertite” and is a three-piece ring composed of three members as shown in the sectional view below the ring. The main part or inner ring has a circumferential channel in which the two outer rings lock, the resulting cross-section being rectangular just the same as that of a regular pattern ring. All three rings are diagonally split and the joints are spaced equally and the distances maintained by small pins. This results in each joint being sealed by the solid portion of the other rings.

The use of a number of light steel rings instead of one wide ring in the groove is found on a number of automobile power plants, but as far as known, this construction is not used in airplane power plants. It is contended that where a number of light rings is employed a more flexible packing means is obtained and the possibility of leakage is reduced. Rings of this design are made of square section steel wire and are given a spring temper. Owing to the limited width the diagonal[303] cut joint is generally employed instead of the lap joint which is so popular on wider rings.

An examination of the engine design that is economical in oil consumption discloses the use of tight piston rings, large centrifugal rings on the crank-shaft where it passes through the case, ample cooling fins in the pistons, vents between the crank-case chamber and the valve enclosures, etc. Briefly put, cooling of the oil in this engine has been properly cared for and leakage reduced to a minimum. To be specific regarding details of design: Oil surplus can be kept out of the explosion chambers by leaving the lower edge of the piston skirt sharp and by the use of a shallow groove (C), Fig. 126, just below the lower piston ring. Small holes are bored through the piston walls at the base of this groove and communicate with the crank-case. The similarity of the sharp edges of piston skirt (D) and piston ring to a carpenter’s plane bit, makes their operation plain.

The cooling of oil in the sump (A) can be accomplished most effectively by radiating fins on its outer surface. The lower crank-case should be fully exposed to the outer air. A settling basin for sediment (B) should be provided having a cubic content not less than one-tenth of the total oil capacity as outlined at Fig. 126. The depth of this basin should be at least 2¹⁄₂ inches, and its walls vertical, as shown, to reduce the mixing of sediment with the oil in circulation. The inlet opening to the oil pump should be near the top of the sediment basin in order to prevent the entrance into the pump with the oil of any solid matter or water condensed from the products of combustion. This sediment basin should be drained after every five to seven hours air service of an airplane engine. Concerning filtering screens there is little to be said, save that their areas should be ample and the mesh coarse enough (one-sixteenth of an inch) to[304] offer no serious resistance to the free flow of cold or heavy oil through them; otherwise the oil in the crank-case may build up above them to an undesirable level. The necessary frequency of draining and flushing out the oil sump differs greatly with the age (condition) of the engine and the suitability of the oil used. In broad terms, the oil sump of a new engine should be thoroughly drained and flushed with kerosene at the end of the first 200[305] miles, next at the end of 500 miles and thereafter every 1,000 miles. While these instructions apply specifically to automobile motors, it is very good practice to change the oil in airplane engines frequently. In many cases, the best results have been secured when the oil supply is completely replenished every five hours that the engine is in operation.

The connecting rod is the simple member that joins the piston to the crank-shaft and which transmits the power imparted to the piston by the explosion so that it may be usefully applied. It transforms the reciprocating movement of the piston to a rotary motion at the crank-shaft. A typical connecting rod and its wrist pin are shown at Fig. 120. It will be seen that it has two bearings, one at either end. The small end is bored out to receive the wrist pin which joins it to the piston, while the large end has a hole of sufficient size to go on the crank-pin. The airplane and automobile engine connecting rod is invariably a steel forging, though in marine engines it is sometimes made a steel or high tensile strength bronze casting. In all cases it is desirable to have softer metals than the crank-shaft and wrist pin at the bearing point, and for this reason the connecting rod is usually provided with bushings of anti-friction or white metal at the lower end, and bronze at the upper. The upper end of the connecting rod may be one piece, because the wrist pin can be introduced after it is in place between the bosses of the piston. The lower bearing must be made in two parts in most cases, because the crank-shaft cannot be passed through the bearing owing to its irregular form. The rods of the Gnome engine are all one piece types, as shown at Fig. 127, owing to the construction of the “mother” rod which receives the crank-pins. The complete connecting rod assembly is shown in Fig. 121, also at A, Fig. 127. The “mother[306]” rod, with one of the other rods in place and one about to be inserted, is shown at Fig. 127, B. The built-up crank-shaft which makes this construction feasible is shown at Fig. 127, C.

Some of the various designs of connecting rods that have been used are shown at Fig. 128. That at A is a simple form often employed in single-cylinder motors, having built-up crank-shafts. Both ends of the connecting rod are bushed with a one-piece bearing, as it can be assembled in place before the crank-shaft assembly is built up. A built-up crank-shaft such as this type of connecting rod would be used with is shown at Fig. 106. The pattern shown at B is one that has been used to some extent on heavy work, and is known as the “marine type.” It is made in three pieces, the main portion being a steel forging having a flanged lower end to which the bronze boxes are secured by bolts. The modified marine type depicted at C is the form that has received the widest application in automobile and aviation engine construction.[308] It consists of two pieces, the main member being a steel drop forging having the wrist-pin bearing and the upper crank-pin bearing formed integral, while the lower crank-pin bearing member is a separate forging secured to the connecting rod by bolts. In this construction bushings of anti-friction metal are used at the lower end, and a bronze bushing is forced into the upper- or wrist-pin end. The rod shown at D has also been widely used. It is similar in construction to the form shown at C, except that the upper end is split in order to permit of a degree of adjustment of the wrist-pin bushing, and the lower bearing cap is a hinged member which is retained by one bolt instead of two. When it is desired to assemble it on the crank-shaft the lower cap is swung to one side and brought back into place when the connecting rod has been properly located. Sometimes the lower bearing member is split diagonally instead of horizontally, such a construction being outlined at E.

In a number of instances, instead of plain bushed bearings anti-friction forms using ball or rollers have been used at the lower end. A ball-bearing connecting rod is shown at F. The big end may be made in one piece, because if it is possible to get the ball bearing on the crank-pins it will be easy to put the connecting rod in place. Ball bearings are not used very often on connecting rod big ends because of difficulty of installation, though when applied properly they give satisfactory service and reduce friction to a minimum. One of the advantages of the ball bearing is that it requires no adjustment, whereas the plain bushings depicted in the other connecting rods must be taken up from time to time to compensate for wear.

This can be done in forms shown at B, C, D, and E by bringing the lower bearing caps closer to the upper one and scraping out the brasses to fit the shaft. A number of liners or shims of thin brass or copper stock, varying from .002 inch to .005 inch, are sometimes interposed between the halves of the bearings when it is first[309] fitted to the crank-pin. As the brasses wear the shims may be removed and the portions of the bearings brought close enough together to take up any lost motion that may exist, though in some motors no shims are provided and depreciation can be remedied only by installing new brasses and scraping to fit.

The various structural shapes in which connecting rods are formed are shown in section at G. Of these the I section is most widely used in airplane engines, because it is strong and a very easy shape to form by the drop-forging process or to machine out of the solid bar when extra good steel is used. Where extreme lightness is desired, as in small high-speed motors used for cycle propulsion, the section shown at the extreme left is often used. If the rod is a cast member as in some marine engines, the cross, hollow cylinder, or U sections are sometimes used. If the sections shown at the right are employed,[310] advantage is often taken of the opportunity for passing lubricant through the center of the hollow round section on vertical motors or at the bottom of the U section, which would be used on a horizontal cylinder power plant.

Connecting rods of Vee engines are made in two distinct styles. The forked or “scissors” joint rod assembly is employed when the cylinders are placed directly opposite each other. The “blade” rod, as shown at Fig. 129, fits between the lower ends of the forked rod, which oscillate on the bearing which encircles the crank-pin. The lower end of the “blade” rod is usually attached to the bearing brasses, the ends of the “forked” rod move on the outer surfaces of the brasses. Another form of rod devised for use under these conditions is shown at Fig. 130 and installed in an aviation engine at Fig. 132. In this construction the shorter rod is attached to a boss on the master rod by a short pin to form a hinge and to permit the short rod to oscillate as the conditions dictate.[312] This form of rod can be easily adjusted when the bearing depreciates, a procedure that is difficult with the forked type rod. The best practice, in the writer’s opinion, is to stagger the cylinders and use side-by-side rods as is done in the Curtiss engine. Each rod may be fitted independently of the other and perfect compensation for wear of the big ends is possible.

[313]

Before going extensively into the subject of crank-shaft construction it will be well to consider cam-shaft design, which is properly a part of the valve system and which has been considered in connection with the other elements which have to do directly with cylinder construction to some extent. Cam-shafts are usually simple members carried at the base of the cylinder in the engine case of Vee type motors by suitable bearings and having the cams employed to lift the valves attached at intervals. A typical cam-shaft design is shown at Fig. 133. Two main methods of cam-shaft construction are followed—that in which the cams are separate members, keyed and pinned to the shaft, and the other where the cams are formed integral, the latter being the most suitable for airplane engine requirements.

The cam-shafts shown at Figs. 133 and 134, B, are of the latter type, as the cams are machined integrally. In this case not only the cams but also the gears used in driving the auxiliary shafts are forged integral. This is a more expensive construction, because of the high initial cost of forging dies as well as the greater expense of machining. It has the advantage over the other form in which the cams are keyed in place in that it is stronger, and as the cams are a part of the shaft they can never become loose, as might be possible where they are separately formed and assembled on a simple shaft.

The importance of the crank-shaft has been previously[314] considered, and some of its forms have been shown in views of the motors presented in earlier portions of this work. The crank-shaft is one of the parts subjected to the greatest strain and extreme care is needed in its construction and design, because practically the entire duty of transmitting the power generated by the motor to the gearset devolves upon it. Crank-shafts are usually made of high tensile strength steel of special composition. They may be made in four ways, the most common being from[315] a drop or machine forging which is formed approximately to the shape of the finished shaft and in rare instances (experimental motors only) they may be steel castings. Sometimes they are made from machine forgings, where considerably more machine work is necessary than would be the case where the shaft is formed between dies. Some engineers favor blocking the shaft out of a solid slab of metal and then machining this rough blank to form. In some radial-cylinder motors of the Gnome and Le Rhone type the crank-shafts are built up of two pieces, held together by taper fastenings or bolts.

The form of the shaft depends on the number of cylinders and the form has material influence on the method of construction. For instance, a four-cylinder crank-shaft could be made by either of the methods outlined. On the other hand, a three- or six-cylinder shaft is best made by the machine forging process, because if drop forged or cut from the blank it will have to be heated and the crank throws bent around so that the pins will lie in three planes one hundred and twenty degrees apart, while the other types described need no further attention, as the crank-pins lie in planes one hundred and eighty degrees apart. This can be better understood by referring to Fig. 135, which shows a six-cylinder shaft in the rough and finished stages. At A the appearance[316] of the machine forging before any of the material is removed is shown, while at B the appearance of the finished crank-shaft is clearly depicted. The built-up crank-shaft is seldom used on multiple-cylinder motors, except in some cases where the crank-shafts revolve on ball bearings as in some automobile racing engines.