ALTHOUGH the injector is not theoretically so efficient as a good pump, practically it has proved itself the best means of feeding water to locomotive boilers that has ever been tried. When a well-made injector is used regularly, it is more reliable than any form of pump, is more easily examined and repaired when it gets out of order, is less liable to freeze or to sustain damage from accidental causes, and it regulates the quantity of water required as well as the ordinary pump, and better than any pump actuated by the machinery of the engine, when the speed of a train is irregular. The injector also possesses the important advantage that it raises the temperature of the feed-water to approach the temperature of the boiler, thereby avoiding shocks and strains to metal that very cold water is likely to impart.

So long as injectors were imperfectly understood, and were used with no regularity, they retained the name of being unreliable; but so soon as they began to be made the sole feeding medium for locomotive boilers, they had to be worked regularly, and kept in order, which quickly made their merits recognized.

The feed injector was invented by Henri Giffard, an eminent French scientist and aeronaut. Its successful action was discovered during a series of experiments, made with the view of devising light machinery that might be used to propel balloons. Although Giffard designed the most perfect balloon that was ever constructed, the injector was not used upon it. and the invention was laid aside and almost forgotten. During the course of a sea-voyage, Giffard happened to meet Stewart of the engineering firm, Sharp, Stewart & Co., of Manchester, England. In the course of a conversation on the feeding of boilers, Giffard remembered his injector, and mentioned its method of action. Stewart was struck with the simplicity of the device, and undertook to bring it out in England, which he shortly afterwards did, representing the interests of the inventor so long as the original patents lasted. By his advice, William Sellers & Co., of Philadelphia, were given control of the American patents. Seldom has an invention caused so much astonishment and wild speculation among mechanics, and even among scientists, as the injector did for the first few years of its use. Scientists were not long in discovering the philosophy of the injector's action, but that knowledge spread more slowly among mechanics. It was regarded as a case of perpetual motion — the means of doing work without power, or, as Americans expressed it, by the same means a man could raise himself by pulling on his bootstraps.

Although the mechanism of the injector is very simple, the philosophy of its action is not so easily understood as the principles on which a pump raises water and forces it into the boiler. On beginning to investigate the action of the injector, it appears a physical paradox, the finding that steam at a given pressure leaves a boiler, passes through several tortuous and contracted passages, raises several check-valves, and then forces water into the boiler against a pressure equal to that which the steam had when it first began the operation. At first acquaintance the operation looks as if it had a strong likeness to perpetual motion, but closer investigation will show that the steam which raises and forces the water by passing through an injector performs mechanical work as truly as the steam that pushes a piston which moves a pump-plunger. A current of any kind, be it steam, air, water, or other matter, has a tendency to induce a movement in the same direction of any body with which it comes in contact. Thus, we are all familiar with the fact that a current of air called wind, passing over the surface of a body of water, sets waves in motion, and dashes the water high up on the shore away above its original level. In the same way a jet of steam moving very rapidly, when injected into a body of water, under favorable conditions, imparts a portion of its motion to the water, and starts it with momentum sufficient to overcome a pressure even higher than the original pressure of the steam. The locomotive blast, blowers, steam siphons, steam jets, jet exhausters, vacuum ejectors, and argand burners, are all common instances of the application of the principle of induced currents.

At a boiler pressure of 140 pounds per square inch steam passes into the atmosphere with a velocity of 1920 feet per second. When steam at this speed strikes like a lightning-flash into the tubes of the injector, it becomes the ram which forces the water towards the boiler; but its power is opposed by the tendency of the water inside the boiler to escape through the check-valve. The velocity with which water will flow from a vessel is known to be equal in feet to the square root of the pressure multiplied by 12.19. Accordingly, in the case under consideration, the water inside of the boiler would tend to escape at a speed of 144 feet per second. This represents the resistance at the check-valve. The mechanical problem, then, to be worked out by the injector is to transform the energy of hot steam moving at a high velocity into the momentum possessed by a heavier and colder mass of water. In the operation the steam yields up a portion of its heat and the greater part of its velocity, but it keeps a current of water flowing fast enough to overcome the static resistance at the check-valve.

A common delivery temperature of the water forced through an injector is 160 degrees Fahr. Taking the feed-water at 55 degrees Fahr., we find that the steam used in operating the injector imparts 105 degrees Fahr. to the feed-water before putting it into the boiler. One pound of steam at 140 pounds boiler-pressure contains 1224 heat units reckoned above zero. When the hot steam speeding at a high velocity strikes the feed-water, part of the heat is converted into the mechanical work required to put the water in motion, but there still is left heat sufficient to raise about 11 pounds of water to the temperature of 160 degrees. One pound of steam, therefore, communicates to 11 pounds of water the motion required for overcoming the resistance encountered at the. check-valve. The steam moving at a speed of 1920 feet per second having imparted motion to a body eleven times its own weight, itself in the mean time having become a portion of the mass, the velocity of the feed-water would be 1920 / 12 = 170 feet per second. When the reduction of speed due to friction of the pipes and other resistances is considered, there still remains momentum enough in the water to raise the check-valve.

Although 160 degrees is about the average heat of the water delivered by lifting injectors, instruments can be designed so that they will heat the water much higher. With non-lifting injectors, the feed-water is nearly always delivered at a higher temperature than with the other kind.

There are numerous forms of injectors in use, but they are all developments of the elementary arrangement of parts shown in the annexed illustration, Fig. 1. Steam at a high velocity passes from the boiler into the tube A, and striking the feed-water at B, is itself condensed, but imparts momentum to the water to send it rushing along into the delivery pipe E with sufficient force to raise the check-valve against the pressure inside and pass into the boiler. As the current of water could not be started into rapid motion against the constant pressure of the check-valve, an overflow opening is provided in the injector, through which the water can flow unchecked till the necessary momentum is obtained, when the overflow valve is closed.

In a lifting injector, the parts are so designed that, in starting, a jet of steam passes through the combining tube at sufficient velocity to create a vacuum in the water-chamber XX, and the water is drawn into this place from the feed-pipe as if by the suction of a pump. The steam-jet then striking the water starts it into motion. If too much steam is admitted for the quantity of water passing, air will be drawn in through the overflow opening, mixing with the water and reducing its compactness, while some uncondensed steam will pass through with the water. This will reduce the force of impact of the feed-water upon the boiler check, and when it becomes so light that the momentum of feed-water is no greater than the resistance inside the boiler, the injector will break. On the other hand, when the quantity of water supplied is too great for the steam to put into high motion, part will escape through the overflow valve. In some forms of injectors, separate appliances are used for raising the water from the forcing chamber to the source of supply.

As the successful operating of the injector is dependent on the feed-water promptly condensing the steam which supplies the power, water of a very high temperature cannot be fed by an injector. A certain amount of live steam must be condensed by the feed-water to impart the momentum necessary to make the latter overcome the resistance at the check-valve. When the feed-water becomes hotter than 100 degrees Fahr. a point is soon reached where it takes such a large body of water to condense the steam that there is not the required velocity generated to force the feed water into the boiler.

All deviations from the elementary form of injector shown are made for the purpose of extending the action of the instrument under varied conditions, for making it work automatically under different pressures of steam, and for increasing its capacity for raising the water to be used above its natural level.

When an engineer finds that an injector refuses to work, his first resort should be the strainer. That gets choked with cinders or other impurities so frequently that no time should be lost in examining it. One day when I was running a round-house, an engineer came in breathless, with the information that his engine was blocked in the yard, and he must dump his fire, as he could not get his injector to work. The thermometer stood at twenty degrees below zero, and an Iowa blizzard was blowing; so the prospect of a dead engine in the yard meant some distressingly cold labor. I asked, the first thing, if he had tried the strainer; and his answer was that the strainer was all right, for the injector primed satisfactorily, but broke every time he put on a head of steam. I went out to the engine, and had the engineer try to work the injector. By watching the overflow stream, I easily perceived that the injector was not getting enough water, although it primed. An examination showed that the strainer was full of cinders, and the injector went to work all right as soon as the obstruction to the water was removed.

Sand and cinders are the most common causes of failure with injectors, as they are indeed with all water feeding apparatus. A very common cause of failure of injectors is leakage of steam through throttle-valve or check-valve, keeping the tubes so hot that no vacuum can be formed to make it prime. A great many injector-checks have been turned out too light for ordinary service, while others are made in a shape that will always leave the valve away from the seat when they stop working. Then the engineer has to run forward, and pound the check with a hammer to keep the steam from blowing back, and that soon ruins the casting. Check-valves set in a horizontal position are worthless with water that contains grit.

To preserve a good working injector, the engineer should see that the pipes and joints are always perfectly tight. Of course it is difficult to keep them tight when they are subjected to the continual jars a locomotive must stand; but injectors cannot be depended on where there is a possibility of air mixing with the water. Leaky joints or pipes are particularly troublesome to lifting injectors; for air passes in, and keeps the steam-jet from forming a vacuum. At first the injector will merely be difficult to start; but as the leaks get worse there will be no starting it at all. Then, the air mixing with the water is detrimental to the working of all injectors, as its tendency is to decrease the speed of the water. The compact molecules of water form a cohesive body, which the steam can strike upon with telling force to keep it in motion. When the water is mixed with air it lacks the element of compactness, and the steam-jet strikes a semi-elastic body which does not receive momentum readily. This mixture of steam and air does not act solidly on the check-valve, but makes the water pass in with a bubbling sound, as if the valve were moving up and down; and the stream of water breaks very readily when it is working in this way.

As maintaining unbroken speed on the water put in motion is the first essential in keeping an injector in good working order, anything that has a tendency to reduce that speed will jeopardize its action. A variety of influences combine to reduce the original efficiency of an injector. Those with fixed nozzles are constructed with the orifices of a certain size, and in the proportion to each other which experiment has demonstrated to be best for feeding with the varied steam-pressures. When these orifices become enlarged by wear the injector will work badly, and nothing will remedy the defect but new tubes. The tubes sometimes get loose inside the shell of the injector, and drop down out of line. The water will then strike against the side of the next tube, or on some point out of the true line, scattering it into spray which contains no energy to force itself into the boiler. A machinist examining a defective injector should always make sure that the tubes are not loose. Injectors suffering from incrusted water-passages will generally work best with the steam low. In districts where the feed-water is heavily charged with lime salts, it is common for injectors to get so incrusted that the passages are almost closed.

Joints about injectors that are kept tight by packing must be closely watched. Many an injector that failed to work satisfactorily has been entirely cured by packing the ram-gland.

During severe frosty weather an injector can be kept in order much easier than a pump; but it needs constant watching and intelligent supervision.

To keep an injector clear of danger from frost, it should be fitted with frost-cocks so that all the pipes can be thoroughly drained. Bends in the pipes, where water could stand, should be avoided as far as possible; and where they cannot be avoided, the lowest point should contain a drain-cock.

To operate an injector successfully, thoughtful care is requisite on the part of the engineer; and where this is given, the injector will prove itself a very economical boiler-feeder.

The injectors principally used in American locomotives are the Sellers, the Nathan, the Mack, and the Rue Little Giant. All are good reliable boiler-feeders, and all are made to wear well under the rough service met with on locomotives.

When the Giffard injector was first introduced into this country by William Sellers & Co., Philadelphia, it was a rather defective boiler-feeder; but that firm effected great improvements and led the way for making the injector the popular boiler-feeder it is to-day. They made the instrument self-adjusting, and improved its design so that it would feed automatically, however much the pressure of the boiler varied, and, finally they perfected it so that, should anything happen to interrupt its working, it would automatically restart itself. The latest development of the injector is shown by a sectional view in Fig. 2.

This instrument will start at the lowest steam pressures with water flowing to it, and will lift the water promptly even when the suction-pipe is hot. At 10 pounds steam pressure it will lift the Water 2 feet; at 30 pounds, 5 feet; and at all ordinary pressures, say 60 pounds and over, it will lift from 12 to 18 feet. It can be used as a heater for the water supply by simply closing the waste-valve and pulling out the steam-lever.

By reference to the cut it will be seen that this injector consists of a case A provided with a steam inlet B, a water-inlet C, an outlet D through which the water is conveyed to the boiler, an overflow opening E, a lever F by which to admit steam, stop and start its working, a hand-wheel G to regulate the supply of water, and an eccentric lever H to close the waste-valve when it is desired to make a heater of the injector.
Its operation is as follows:

The water-inlet C being in communication with water supply, the valve a is open to allow the water to enter the chamber I, Steam is admitted to the chamber B, and the lever F is drawn out to lift the valve b from its seat and permit the steam to enter the annular lifting steam-nozzle c through the holes d d. The steam issuing from this nozzle passes through the annular combining tube e and escapes from the instrument partly through the overflow opening f and partly through the overflow openings provided in the combining tube g g, through the overflow chamber J and passage E E, and produces a strong vacuum in the water chamber I which lifts the water from the source of supply, and the united jet of steam and water is, by reason of its velocity, discharged into the rear of the receiving end of the combining tube g. The further movement of the lever F withdraws the spindle h until the steam-plug i is out of the forcing nozzle K, allowing the steam to pass through the forcing nozzle K and come in contact with the annular jet of water which is flowing into the combining tube around the nozzle K. This jet of water has already a considerable velocity, and the forcing steam jet imparts to it the necessary increment of velocity to enable it to enter the boiler through the delivery tube j and boiler check k.

If from any cause the jet should be broken—say from a failure in the water supply—the steam issuing from the forcing nozzle K into the combining tube g will escape through the overflows m and n and intermediate openings with such freedom that the steam, which will return through the annular space formed between the nozzle K and combining tube g, and escape into the overflow chamber through the opening f, will not have sufficient volume or force to interfere with the free discharge of the steam, issuing from the annular lifting steam-nozzle and escaping through the same overflow F, and hence the lifting steam-jet will always tend to produce a vacuum in the water-chamber I, which will again lift the water when the supply is renewed, and the combined annular jet of steam and water will be forced into the combining tube g against the feeble current of steam returning, when the jet will again be formed and will enter the boiler as before. In actual practice on a locomotive the movement of the lever F in starting the injector is continuous.

One of the most successful and enduring injectors in use is the Monitor, the distinguishing feature of which originally was that the injector is constructed with fixed nozzles, that insure great durability, combined with certainty of action. The injector shown in Fig. 3 is an improvement on the old Monitor, the radical change being that this injector is operated by a single lever. Any one who has studied the operation of the injector already described will have no difficulty in perceiving how the new Monitor works. It will be seen that steam is admitted from the top to the tube that forms the body of the injector, and the water from below. To start the injector, the water-valve W is opened. The main lever S is then pulled out a short distance to lift the water; when the water begins to escape through the overflow the lever S is steadily drawn back, which puts the injector working at its maximum power. The quantity of feed required is graduated by the valve W.

When it is desired to use the injector as a heater, close the valve H and. pull out the lever S all the way. At other times the valve H must be kept open.

With a boiler pressure of 30 pounds this injector will lift the water 5 feet, and at ordinary working pressure the steam will have power to lift the water to a height not likely to arise in locomotive practice.

The engraving gives a sectional view of the well known Mack injector, which is one of the oldest and has long been a favorite on many roads. The parts are strikingly simple, and they are designed in very compact form. The section shows the arrangement very clearly. The cone and tubes can be easily re-moved for cleaning; or should they get cut by the sand in gritty water, or filled with incrustation, they can betaken out and replaced by a new set in a few minutes, the interchangeable parts being kept in stock. There is but one water-passage and it is very large, so that there is very little danger of sand or mineral deposits interfering with the efficiency of the injector.

To work this injector, the steam-valve is opened one quarter of a turn to lift the water; when water begins to escape from the overflow, the steam-valve is opened till the water ceases to pass out of the overflow opening. The supply is regulated by the lazy-cock.

A special claim made for this injector is the wide range of its delivery. The supply can be regulated to the absolute requirements of the boiler, be the train fast or slow, light or heavy. It will start readily at 30 pounds pressure and work up to any pressure required. Below 50 pounds pressure it may be necessary to partly close the supply of water.

This injector, made by the Rue Manufacturing Co. is a highly efficient boiler-feeder, and a very simple apparatus. The construction is clearly seen in the engraving. A unique feature about this injector is the movable combining tube adjusted by a lever, causing the feed to be exactly suited to the service. Moving the lever towards A tends to cut off the feed, and moving towards B increases it.

To work the injector, the combining tube lever is set in position to admit sufficient water to condense the steam from the starting valve. The starting valve is then opened slightly till the water begins to escape from the overflow, when it is opened full. The feed is then regulated by the combining tube lever. To use this injector as a heater, the overflow is closed by the combining tube being moved up against the discharge, and opening the starting valve sufficiently to admit the quantity of steam required.

Table of Contents | Contents Page

Do you have any information you'd like to share on this subject? Please email me!
The Catskill Archive website and all contents, unless otherwise specified,
are 1996-2010 Timothy J. Mallery