The idea of a self-contained engine which should need no boiler is an old one. It has several advantages over the steam engine. This chapter describes the evolution of the modern gas engine
SECTION THROUGH A GAS ENGINE manufactured by Messrs. Crossley Brothers, Limited.
THE idea of an engine which should be self-contained, an engine which should need no boiler, is a very old one. Denis Papin and others tried, at the end of the seventeenth century, to produce power by exploding gunpowder inside a cylinder and arranging for the piston to do work as it fell by its own weight. At odd times in the eighteenth century, and very frequently in the nineteenth century, men worked at the problem of the internal combustion engine, trying now gases and again highly inflammable volatile liquids. The Patent Office records bear witness to the ingenuity and persistence of these early inventors, and the history of applied science indicates their failures by naming Étienne Lenoir as the first to achieve any degree of success, in 1860.
Lenoir’s engine was intended to use gas from the town supply, and its manner of working shows, by comparison with present-day engines, how feeble and uncertain was man’s control over the forces which he sought to harness in his service. The cylinder, like that of the earlier steam engines, was set vertically, with the shaft above the upper end of the cylinder, and the piston was very heavy, lest by the explosion of a mixture of gas and air underneath it should be blown out of the top. This piston, moreover, was not connected to the shaft by a connecting rod and crank in the ordinary way. There was a connecting rod, but it was provided with teeth all the way along the upper half, forming a rack, and the shaft was provided with a toothed wheel into which the teeth of the rack fitted. When an explosion occurred the piston rose rapidly, and the rack, acting on the toothed wheel, caused the shaft to spin round. Having arrived at the top of its stroke, the rack disengaged with the wheel, fell away from it, and allowed the piston to return to the bottom of the cylinder without affecting the rotating shaft. During this return stroke the exploded gases were swept out through the exhaust valve, and in the interval between the strokes a heavy flywheel kept the shaft in motion.
Crude in design, cumbrous in action, and irregular in speed as the Lenoir engine was, a great many were made before, in 1876, Dr. Nicolaus Otto invented a far more perfect form — the parent of the gas engine of to-day. Its mode of operation is shown diagrammatically below.
DIAGRAM SHOWING how the gas engine works.
The first outward stroke of the piston draws in a mixture of air and gas, the valves being opened just long enough for the right proportion of each to enter. As the shaft continues to rotate the piston makes its return stroke, and as the gas and air valves are now closed the mixture is compressed. The efficiency of the engine increases with the degree of compression, but it must not be too high or the explosion will take place prematurely — that is, before the piston has reached the end of its stroke. Generally speaking, it must not reach 125 lb. on the square inch. At this point the mixture is ignited, and an explosion occurs which drives the piston towards the open end of the cylinder, and communicates energy to the flywheel. Finally, on the next return stroke, the spent gases are swept out through the exhaust valve, which is opened just before the former stroke is completed.
In this sequence of operations, which is called a cycle, the crank shaft receives an impulse every two revolutions or every four strokes. For that reason an engine working on this plan is often called a four-stroke engine. It is very simple in construction, as will be seen from the illustration at the top of this page, which represents the type made by Messrs. Crossley Bros., who were the first firm in this country to manufacture engines under Dr. Otto’s patent.
The piston, it will be noticed, is a cylinder open at one end, the type being known as a “trunk” piston. This form renders a piston rod unnecessary, and reduces the length, and therefore the weight, of the bed. Moreover, it serves also as a cross-head in taking up the thrust in the connecting rod, and the pressure produced in this way is distributed over a large area of the piston and cylinder, so that the wear is uniform.
The valves of gas engines are invariably of the mushroom type, held down on their seats by springs, and opened at the right moment and for the right period by cams and levers. A cam is like a small eccentric without the strap. One for each valve — air, gas, and exhaust — is fixed on a shaft parallel to the engine bed, and driven from the main shaft by bevel wheels or by worm gearing at half the speed. This ensures that each valve shall be opened once in every two revolutions of the main shaft. In small engines the cams may be in direct contact with the valve spindles, but the usual plan is for the motion to be communicated through levers. Other kinds of valve have been tried, but none are so satisfactory as these. Slide valves do not work well under the high temperatures which occur during explosion, and rotary valves are not easily adjusted for wear.
Ignition of the mixed gases in the original Otto engine was effected by means of a hot tube. The tube was closed at one end, the open end communicating with the interior of the cylinder. It was kept hot by a flame which played upon it all the time the engine was working. At each compression stroke the mixture of gases was forced into the tube and became ignited. Sometimes the ignition occurred prematurely, but not as a rule, because there always remained some of the spent gases from a previous explosion. For many years it was the only plan which worked satisfactorily, but in recent years it has been replaced entirely by the electric spark.
For producing the electric spark there are two methods — an induction coil and a magneto machine or small dynamo. Both are good, but the magneto is cheaper than a really good coil and less liable to get out of order than a cheap one. Moreover, for the
coil accumulators are necessary, and they must be kept charged or a duplicate set must be held in readiness. For large engines, where the cost of accessories is less important, the arrangement devised by Sir Oliver Lodge is frequently used. This gives a very powerful spark, which is certain and effective in its action. By whatever method the electricity for the spark is generated, it is produced inside the cylinder between the ends of two wires fixed in. a sparking plug. This is easily removable for cleaning or renewal.
The gas engine needs to be governed just as a steam engine does, or it would “race” when the load was decreased or removed. The governor itself is practically the same as a steam engine governor, but the method by which it controls the speed of the engine may differ. Many of the early gas engines were fitted with a “hit and miss” governor. Between the cam and the gas valve spindle was a small rod which when the engine ran too fast was lifted out of the way. When this occurred the gas valve did not open, no charge entered, and no explosion occurred.
Missing a whole explosion is rather a drastic method of reducing the speed, especially when there is only one explosion in two revolutions; and the modern practice is to employ a “throttling” governor, which controls very delicately the quantity of gas drawn into the cylinder on each occasion. One of the neatest devices for effecting this is that adopted by Messrs. Crossley Bros., and shown in the accompanying diagram.
THE GOVERNING ARRANGEMENT of a Crossley gas engine.
It will be noticed that only one valve is used to admit both gas and air, that no additional throttle valve is used, and that the amount of gas and air entering is controlled at the moment of admission to the cylinder. The curved bar pinned to the top of the valve spindle acts as a lever, the fulcrum of which is a “radius rod”, actuated by the governor. The amount of movement given to one end of the curved bar by the cam is constant; but as the speed of the engine varies the radius rod swings to right or left, alters the length of the lever arms, and varies the extent of opening of the admission valve. The same cam also serves to open the exhaust valve, which is situated immediately below that by which air and gas are admitted.
Another very interesting method of governing is that adopted on the National gas engine and illustrated in the accompanying diagram. The cam operates one bent lever, and the valve is opened by another. Between the ends of the two levers is a metal plate suspended by a rod to a lever which rises or falls as the engine increases or decreases in speed. Now the movement of this metal plate alters the length of one arm of the lever which operates the valve and thus controls the amount of opening of the valve. Both are very pretty pieces of mechanism.
GOVERNING ARRANGEMENT of National gas engine.
While the cylinder of the steam engine had to be kept warm by means of a steam jacket, that of the gas engine must be kept cool by means of a water jacket. The walls of the cylinder are double, and cold water is maintained in constant circulation —sometimes naturally and sometimes by the aid of a pump. If this were not done the valves would soon wear out, and the piston would expand and jam owing to the heat produced by the explosions. Moreover, the combustion chamber — the small space behind the piston into which the mixed gases are compressed — would become so hot that the mixture would explode immediately on admission, with the certainty of damage to the engine, and probably also injury of the man in charge. With the steam engine the absence of steam in the jackets means merely a loss in efficiency; with a gas engine even an interruption to the flow of water means a smash. And even when the cooling system is working effectively the gases which escape from the exhaust are hot enough to render very useful service. In some cases they are employed to produce hot water. Thus about 2 gallons of boiling water can be obtained for each horsepower per hour, and the waste heat from a 50-horsepower engine will produce 100 gallons per hour.
Having now learnt how a gas engine works, let us see what advantage it possesses over the steam engine. In the first place, it occupies less space. No boiler is required, and the heat is produced in the same chamber in which it is to be utilised. There is no loss of heat by radiation from the surface of boiler or steam pipes, and the cost of these, together with stop valves and other accessories, is saved. It is cleaner, and though coal and ashes have to be handled at the gas works, this can be done on a large scale which justifies employment of labour-saving machinery. The gas engine utilises from 25 per cent, to 30 per cent, of the heat obtainable from the fuel, while the steam engine converts less than 15 per cent, into useful work. But town gas is rather an expensive fuel, and for many years the engines were small, so that their use was in consequence limited. How it has developed during the last thirty years deserves further attention.
The Food of the Gas Engine
The disadvantages of the early gas engines were their dependence upon a supply of town gas, the cost of this form of fuel, and the fact that the piston received only one impulse for two revolutions of the shaft — that is, for one-quarter of the time the piston drove the shaft, and for three-quarters of the time the shaft drove the piston. Consider the fuel question first.
Coal gas is made by heating coal out of contact with air in closed retorts. The result is :
(a) Gas.
(b) Ammoniacal liquor.
(c) Tar.
(d) Coke.
In order to obtain gas of the highest illuminating value the process of distillation cannot be carried out in the cheapest way. When the gas engine was first introduced — and, indeed, for many years afterwards — the value of coal gas lay in its illuminating value, so that it was a dear fuel. But in 1878 Joseph Emerson Dowson showed not only how to produce a cheaper gas, but also how this gas could be produced wherever it was required, so that a town supply was no longer necessary. The principle was this: that if coal burns in an ample supply of air it forms carbon dioxide which is no longer inflammable, while if the air supply is limited, carbon monoxide is formed, and this gas will burn with a further supply of air, producing carbon dioxide. Thus, in a fairly deep fire, red hot throughout, a lambent blue flame will frequently be seen playing over the top. The oxygen in the air entering the lower part of the grate produces carbon dioxide, and this, passing through the red-hot carbon in the upper part, takes up carbon and forms carbon monoxide. Those who have learnt a little chemistry will recognise the equations corresponding to the two processes:
C + O₂ = C O₂
C O₂ + C = 2 CO
The nitrogen, which forms four-fifths by weight of the atmosphere, passes out unchanged with the carbon monoxide, and the mixture is known as producer gas.
The apparatus consisted of a deep cylindrical furnace charged with coke fed by a hopper from the top. It was of sheet iron, lined with firebrick, and there were dampers at the bottom to regulate the supply of air, and poker holes in the top to enable the red-hot fuel to be stirred occasionally. At the meeting of the British Association for the Advancement of Science, at York, in 1881, it was shown driving a 3-horse-power gas engine, and created a great deal of interest. It occupied no more space than a boiler and produced no smoke, burning the coke completely away to fine ashes, which were raked out at the bottom from time to time.
A gas with a greater calorific, or heating value, free from nitrogen, can be obtained by passing steam through red-hot coke. The result in this case is a mixture of carbon monoxide and hydrogen. Unfortunately, the steam would soon put the fire out, so the process has to be stopped every few minutes while air is blown in to raise the temperature again. In spite of the fact that the process is intermittent, water gas, as it is called, has been made in gasworks for many years for mixing with coal gas. The carbon-dioxide gas, formed when air is blown through, is allowed to escape into the atmosphere, and the water gas alone is employed for this purpose. To the student of chemistry the equation C + H₂ O = C O + H₂ will again be familiar.
A tremendous advance was made in 1889 when Dr. Ludwig Mond found that if air and steam were used together a mixture of producer gas and water gas was obtained, which was better than producer gas, while the process could be worked continuously. He found, moreover, that small coal, which is cheaper than coke, could be used, and that the ammoniacal liquor from which a valuable fertiliser is made, could be recovered. The sale of this by-product reduced the cost of fuel in some cases to 3s. 6d. a ton, and it was evident that here was a source of power which rendered gas engines independent of the town supply. Engineers began at once to study the possibilities of large engines, and whereas engines of 100 horse-power had been the largest before they soon began to be made of 500 and 600 horse-power.
But improvements in gas producers did not cease with Mond’s invention. In the earlier ones the air and steam were forced in, and the engine had to take what came, whether the load and its appetite were large or small. But in the modern producer the engine sucks gas just as fast as it requires it, like a baby with a bottle, and there is no fear of choking.
SECTION THROUGH a suction gas producer.
In the sectional illustration of one of Messrs. Crossley Bros.’ suction gas producers (see illustration above) it will be seen to consist of three parts: the producer proper or gas generator (A), the vaporiser or steam raiser (B), and the scrubber or gas cleaner (C). The gas generator is a steel casing lined with fire brick and having a hopper on the top through which the fuel is admitted. The vaporiser consists of a metal box containing a number of tubes which have “gills” in order to encourage the transfer of heat. The hot gases from the producer pass through the box, and as water flows through the tubes the steam which is necessary for the process is generated. The gases then pass up through a tower filled with lumps of coke over which water is continually flowing. In this way they are freed from dust, which, if it entered the engine, would choke up the valves and lead to undesirable wear. As the coke in the lower portion of the scrubber comes first into contact with the gases it needs to be renewed more frequently than that in the upper portion. Each portion, therefore, rests on a separate tray, so that one can be renewed without interfering with the other.
One great advantage of the suction gas producer, as compared with the earlier form, lies in the fact that the pressure inside is less than that of the atmosphere. Carbon monoxide is an extremely poisonous gas, and is exceptionally dangerous because it has no smell. There is no simple chemical test by which its presence in the air can be detected, and it is a common plan to use a bird or a mouse as a sentinel. These little creatures are much more sensitive than human beings, and fall into a stupor long before a man would be affected. But with a suction gas producer there is practically no danger.
The fuel employed may be anything that will burn. Good results are obtained with coke, but the best are given by anthracite. From this fuel at 30s. a ton, gas equal in heating value to the best town gas can be produced at 11d. per thousand cubic feet. The stand-by losses are also small. When the engine is not working, just sufficient air is allowed to pass through to keep the fire in, and the coal consumed in a 100-horse-power plant during an all-night stoppage of twelve hours amounts to less than 40 lb. The plant can be started again in from ten to fifteen minutes.
When bituminous coal is used the plant must be fitted with a tar extractor, otherwise the scrubber would soon get choked up, and the tar would get into the engine and cause trouble. Provided this precaution is taken, then with fuel at 10s. a ton the cost of power amounts to only about one-sixteenth of a penny per horse-power per hour. It is not necessary, however, to use coal at all. Wood waste, sawdust, bark, spent tan, coir dust, coconut shells, mealie cobs, rice husks, sugar-cane refuse, cotton seed, olive refuse, and other material that tends to accumulate and prove a nuisance may be burnt in a gas producer. In this way many industries, especially those concerned with the preliminary processes of manufacture from the products of the soil, can obtain power practically at no cost for fuel, and at the same time get rid of stuff which can only be destroyed by burning. From being tied to the town the gas engine has spread to the outskirts of civilisation, lightening the labour of the pioneer, increasing his output, and providing more cheaply those things for which people in the Old Country have need.
During the last twenty years a new source of food for gas engines has been discovered in which fuel is not specially burnt for the purpose at all. The process of obtaining iron from its ore consists essentially in heating the oxide of iron with carbon in the form of coke, using a blast of air to raise the temperature. The gas, which was formerly allowed to escape freely from the top of the furnace, contains rather more than 30 per cent. — chiefly carbon monoxide — which is combustible, and at night the blaze could be seen for miles around. It was suggested by the late Benjamin Howarth Thwaite in 1892 that this waste should be prevented, and the gas used in gas engines. For every ton of coal charged into the furnace more than 120,000 cubic feet of gas is produced, and, taking the whole country, there is sufficient energy to yield 750,000 horse-power continuously all the year round. Three years later, in 1895, the Glasgow Iron Company adopted the suggestion, and their example was followed by many other firms, especially in Germany and the United States.
The Growth of the Gas Engine
When, after 1889, cheaper gas became available the problem of the large engine arose. There were two principal difficulties. One lay in constructing the cylinder, and the other in keeping the piston cool. The gas engine cylinder is rather a complicated casting, and owing to explosions inside and cooling water outside great strains are set up, so that it is liable to crack. The difference of temperature of two points an inch apart may be 50°, and even though iron has not a high rate of expansion, this difference between points so near together sets up very severe strain. Engines giving 1,000 horse-power for one cylinder have been made, but they are not numerous, and for anything over 200 horse-power engineers usually prefer to employ two or more cylinders. English makers for a long while arranged them side by side, while Continental firms preferred the horizontal tandem arrangement with one cylinder behind the other.
All large gas engines are started by means of compressed air, which Is produced by a separate engine of relatively small size. The flywheel is barred round until the piston and valves are in the correct position for starting, then compressed air is admitted, and as soon as the engine has fairly started the compressed air is cut off and the gas supply turned on. The absolute necessity of some such aid as this for large engines will be realised by anyone who has seen even a 20-horse-power engine started by hand. It is due both to the fact that with only one explosion in four strokes a very heavy flywheel is necessary to equalise the motion, and to the force required to compress the charge before an explosion can take place.
When very large cylinders are necessary they are often cast in four or more parts and then bolted together before being bored. All castings, and especially large ones of complicated form, are subject to strains which are set up during the process of cooling in the mould. This strain would disappear in time, especially if the whole of the “skin” could be removed. But the shape of a gas engine cylinder is such that it cannot be machined all over, and while the strain is there it is a source of weakness.
Again, in engines of small size the water jacket is effective in keeping the piston cool, but not in a large one. The central portion of the end, continually exposed to explosions, is too far removed from the cooling influence of the jacket to be affected. It is not only liable to excessive stresses, but the temperature may rise so high that it jams in the cylinder. It is necessary, therefore, to make the head of the piston hollow and to keep it supplied with water by means of jointed pipes which follow the motion of the piston. This all adds to the initial cost and possibility of breakdown. Up to a certain point the rise of the big gas engine was rapid, and then unexpected difficulties such as these held it back, so that it has failed to compete with the steam engine for higher powers.
Increased power is not obtained, however, merely by an increase in size, but also by devices which secure that there shall be more than one explosion every two revolutions. In 1881, when the gas engine was yet in its infancy, Dugald Clerk patented a method of obtaining an explosion for each revolution; but the Otto cycle had too strong a following, and no one would take the matter up. Years afterwards Koerting and others in Germany revived the two-stroke engine and met the demand for large powers without a corresponding increase in size. To understand how this is done, recall for a moment the operations in the Otto cycle. In four successive strokes there are (1) charge drawn in, (2) compression, (3) explosion, (4) exhaust. In order to pro-
vide for an explosion every revolution, two of these operations must be carried out in such a way that they do not occupy a stroke, and this is achieved by arranging ports in the cylinder walls which shall be uncovered by the piston on its outward stroke. Through these ports a blast of compressed air is driven, which sweeps out the waste gases and leaves sufficient air for the next explosion. Towards the end of the stroke gas is admitted, so that by the time the piston is ready to return the new charge is there to be compressed.
From the two-stroke to the double-acting engine is only a step. Using a piston of the ordinary type, like that of a steam engine, with a piston rod and cross head, and putting a front cover on the cylinder, the operations of a two-stroke engine can then be performed both in front of and behind the piston. The engine then looks — and acts — more like a steam engine, but explosions replace the steady, persistent force of expanding steam.
The large gas engine has been an attractive field for inventors, and numerous attempts have been made to overcome the disadvantages produced by high temperatures, suddenly applied pressures, and the great weight of the parts. In several engines, for example, the explosion does not take place between the fixed end of a cylinder and a moving piston, but between two pistons, free to move in opposite directions and each communicating its energy to the same or a neighbouring crank. Attempts have also been made to produce a gas turbine, but so far the problem has defied solution. There is, however, one invention which deserves description by reason of its simplicity, its originality, and its success. And that is the explosion pump.
The Explosion Pump
The explosion pump was invented by Herbert Alfred Humphrey, who was one of the pioneers of the large gas engine in Great Britain. Suppose a large U-tube open at both ends, as in the accompanying diagram, to be filled with water or any other liquid, and suppose pressure to be suddenly applied to the surface of the water in the left-hand limb. The liquid will fall in that limb and rise in the other, and as soon as the pressure is released it will flow back past its original level until it rises nearly as high in the left-hand limb as it reached, under pressure, in the right. But it will not stop there. It will swing again in the direction in which it was originally driven, and then back, several times, the swing gradually decreasing in extent until the liquid comes to rest. The internal friction of the liquid, the friction on the walls of the tube, and the work done each time in pushing back the atmosphere gradually destroy the motion.
DIAGRAM TO EXPLAIN the principle of the Humphrey pump.
Now this swing or vibration, or oscillation, is exactly like the swing of a pendulum. So long as the swings are not very large their magnitude does not matter: they each take the same time. But for our purpose the time of swing is not of much consequence. The essential fact is that if you give water in a tube of this form a push, it will always come back again. And that is just what a gas engine piston linked up to a crank will do. So that this column of water, once it is set in vibration, will draw in, or compress, or expel gases just like the solid piston of an ordinary gas engine.
DIAGRAMMATIC SECTION of the Humphrey pump.
At the Chingford Reservoir of the London Water Board there are five pumps constructed on this plan and represented diagrammatically as shown above.. The horizontal portion, called the play pipe, is of cast iron, 6 feet in diameter and 60 feet long. The left-hand limb, forming the pump proper, is 7 feet in diameter and 10 feet long. It is provided (a) with gas and air inlet valves, and (b) with exhaust valves, in the upper end, and with a water inlet valve in the side to admit water from the well into which it is built, and which is supplied with water from the River Lea. The right-hand limb is open and has a discharge pipe through which water flows into the reservoir.
The mode of operation is as follows: A mixture of gas and air is exploded above the surface of the water in the cylinder, the water being driven forward along the play pipe and up into the right-hand limb. When once a large body of water is set in motion it does not stop readily, so it does not cease when the pressure in the cylinder has fallen below that of the atmosphere. Water pours out of the discharge pipe into the reservoir, and fresh water enters from the well through the valve in the side of the cylinder. Gradually the forward motion in the play pipe ceases, and the water begins to return. Swinging back with increasing and then decreasing velocity, it sweeps out through the exhaust valve the waste gases from the explosion, and then begins to flow forward again towards the reservoir. During this stroke gas and air are admitted to the cylinder, the water flows back and compresses them, an explosion occurs, and the whole cycle of operations is repeated.
While the pump is at work no attention is required. There are no rubbing surfaces to be lubricated, the valves are self-acting, held down on their seats by springs, and locked when not in use by the action of a small water motor. Unlike the ordinary gas engine, all the strokes in a cycle are not of the same length. They have different duties to perform and are made under different conditions. So long as the supply of gas does not fail and the ignition device is in order the pump will go on working week after week, month after month, year after year, delivering with unfailing regularity from 12 to 14 tons of water per stroke. Four of the Chingford pumps are capable of lifting 40,000,000 gallons through a height of from 20 to 25 feet every twenty-four hours.
Sidney Smith describes an old woman who took a cottage on the west coast that was liable to be invaded by the sea; and when the waves came rolling in she stood at the door, broom in hand, resolutely prepared to sweep back the Atlantic Ocean! She would have been safer had she lived to-day, for she might have dug a deep ditch and emptied it between tides by the aid of an explosion pump.
But there is a more important role for this pump if tradition and vested interests do not stand in the way. Suppose a town with a hill close at hand and a lake or river curling round the lower slopes. With a Humphrey pump, working under more ideal conditions of constant load than is possible with any other form of prime mover, the water could be pumped up into a reservoir on the top of the hill, and thence it could flow down, through pipes, to drive water turbines. And these turbines, constituting the most perfect form of drive for generating electricity, would be coupled to dynamos, providing all the light and heat and power that the town required. Such is an ideal arrangement which the Humphrey pump has brought within the range of practicability.
But improvements in gas producers did not cease with Mond’s invention. In the earlier ones the air and steam were forced in, and the engine had to take what came, whether the load and its appetite were large or small. But in the modern producer the engine sucks gas just as fast as it requires it, like a baby with a bottle, and there is no fear of choking.