The announcement in the newspapers that one of the recently authorized battleships is to be propelled by electricity marks a new step in ship propulsion and will add another achievement to the credit of the United States as a leader in progressive naval engineering. The story of the development of electricity for the propulsion of large ships is an interesting one and, now that it is here, the wonder is that it has not been accomplished before.
Many officers there are in the service to-day whose ideas of electricity were obtained when Professor M. G. Farmer at the Newport Torpedo Station was a man ahead of his time. Many more graduated from the Naval Academy before the useful applications of electrical engineering formed a part of the course. With battleships lighted, ventilated and steered by electricity; with turrets and boat cranes completely electric; with electric anchor windlasses, electric galleys and the wonderful radio, electricity is to-day one of the most vital engineering subjects in the naval profession.
Now we read that the propellers of a battleship are to be driven by electricity. No more main engines! There are only a couple of completely inclosed, turbine-driven alternators in the main engine rooms, and a motor on each shaft. With this advance comes the necessary step to alternating-current electricity, now so generally used in power stations throughout the land, but as yet so seldom used afloat.
This paper was suggested through the realization that few naval officers have had opportunity to keep abreast of the applications of alternating-current power production and transmission. In it an attempt is made to answer, in as elementary a manner as possible, two questions:
1. Why electric propulsion?
2. What is electric propulsion?
The answer to the first question is probably well known in the service, while the answer to the second, at least as to details, is not so generally known. The first will be answered briefly by outlining certain necessary and certain other highly desirable factors in battleship propulsion and in showing how they may be obtained. In the answer to the second an attempt will be made to outline the electric system proposed; followed by an elementary discussion of the alternating-current apparatus.
The time is not long passed when a well-known commander in the navy remarked to his navigator, " Don't talk volts and vampires [amperes] to me; get the dynamo fixed and give us some light! " Such a remark in these days would hardly be possible. To-day the service is familiar with direct current, and it is hoped the term " alternating current " will not, at the outset, suggest deep and mysterious complications, because it is primarily for people not familiar with its use that the latter part of this description is undertaken.
WHY ELECTRIC PROPULSION?
Since the introduction of steam, ship propulsion in general has offered to inventors a field almost unparalleled in breadth. Jet propulsion, stern wheels, side wheels, and various freak forms of screws and propellers are all available as relics of the development.
For battleships the screw propeller and reciprocating engine gradually became standardized. The marine reciprocating engine has reached a stage of development in recent years beyond which it is difficult to go. Consequently in this field the up-to-date four cylinder, triple-expansion, high-speed engine, with forced lubrication and superheated steam at 300 lbs. pressure, is about the last word.
On shore the steam turbine was the next step in gaining a uniform turning moment and in utilizing the energy in the steam at high vacuum. Wonderful results in high-speed turbines have been accomplished. The one phrase "high-speed " is partly the key to the situation. In fact the single-stage impulse turbine was first in the field. The necessity of lower rotative speeds led to compounding; and the turbines, according to the method of compounding, are familiarly known by the names of the men who developed them, as De Laval, Parsons, Curtis and Rateau.
For driving ships two things are essential, viz.: low rotative speed and reversibility. The low speed is necessary on account of the poor results with high-speed propellers, and reversibility in some form is absolutely necessary. Low rotative speed at full power has been obtained by putting in turbines of Many stages, and reversibility has been obtained only by putting a second turbine on the shaft built for operation in the opposite direction from the ahead turbine. Excessive multi-staging means a falling off in efficiency due to increased thermal and frictional (steam) losses, a further loss being the driving idle of a reversing turbine, when the ahead machine is in use. It is customary to run the reversing turbine in a high vacuum when not in use, but even then this idle machine running counter to its own ahead direction requires much more power than is commonly supposed. By .such means the turbine, speed has been reduced to, say, 300 R. P. M., with obvious sacrifices in economy, and at the same time a propeller has been developed for this speed, which is roughly twice the speed for a propeller of maximum desirability. We have therefore lost at both ends in over-all driving power per pound of coal.
This is bad enough at full power, but what happens at cruising speed of, say, half power? At this power the main turbines show no sort of economy and additional turbines have to be put in to better utilize the energy of the steam. All this is done with added turbine weight, added steam piping, with the heat losses incident thereto, and numerous large valves to open and close. The engine room contains truly a complicated-looking array.
For backing, valves must be shut, others opened, and then the relatively small backing turbines acting on the small propellers give a comparatively weak backing and maneuvering power, with much churning of water about the propellers.
It is no wonder the Navy Department was at one time forced to forsake turbines to return to reciprocating engines. It was at that time feared that the problem of economical ship propulsion was not to be solved by turbines, The following figures, based on the reported trials of the ships listed, give at least approximate ideas of coal burned for the production of power:
|
| Tons coal per 24 hrs. at 21 knots | Tons coal per 24 hrs. at 12 knots |
Arkansas | Direct-connected turbines | 465 | 156 |
Wyoming | |||
Texas | Reciprocating engines | 525 | 100 |
New York |
Thanks to the advances in mechanical engineering, another promising solution was offered. This consisted in designing a propeller for its best speed (in a four-shaft ship, say, 160 R. P. M.), then designing a highly efficient turbine at its best speed of, say, 2000 R. P. M. This propeller and turbine form highly satisfactory individual units, and particularly does the turbine respond to an increase in speed by a large increase in economy and greatly reduced weight and size. All that remains to be done is to couple these two units together by a reliable mechanical reducing gear. Although difficulties have to be overcome when the reduction ratio is great, large gears can be built and work at 98 per cent efficiency. The problem at full power ahead is thus perfectly solved.
Unfortunately, when it comes to cruising speed, in order to keep the water rate and the coal rate down we again have to introduce more turbines at the lower rotative speed associated with half power. Also all the while reversing turbines must be driven idle when going ahead.
A single-ratio gear with a non-reversible turbine is still not the ultimate solution.
We then turn to the automobile engineer; he has been up against the same problem. He has solved it for his little gas engines by a multi-ratio gear with a reversing combination. Unfortunately the time has not arrived when such a rig can be built for the 7000 horse-power necessary to be transmitted to each propeller.
The hydraulic engineer became interested and in Germany developed a single-ratio reversible gear with water as the working element. This, too, presents problems in 7000 horse-power sizes and has but a single gear ratio.
Such was the state of the propulsion situation when the electrical engineers were permitted to state their case. On shore the high-speed turbine-driven electric generator had made won derful advances. So, too, had the development of huge electric motors progressed. It seemed possible that electricity, with its single turbo-generators of 40,000 horse-power and 6500 horsepower motors, might offer a solution.
Fortunately for the navy, Mr. W. L. R. Emmet, a Naval Academy graduate of the class of 1881, and now one of the
principal consulting engineers of the General Electric Company, had for years been studying the general subject of ship propulsion. As a result of his researches, Mr. Emmet had come to the conclusion that for battleships the most satisfactory means of propulsion could be obtained with what amounted to an electric reduction gear. By this means a high-speed, non-reversing turbine could be electrically " geared " to one or more low-speed propellers by a two-ratio reversing installation.
In a 30,000 horse-power four-shaft battleship, a 165 R. P. M. 7500 horse-power motor with a two-speed ratio would be put on each shaft, and would be efficiently driven in either direction with a single 30,000 horse-power 2000 R. P. M. generator. With alternating current this was easy, although many new electrical problems were presented. One generator in a single compartment was, however, a bad idea, both from a safety standpoint and from the fact that it is undesirable to concentrate all power in a single unit. Furthermore, one generator at half power would not be working at the desired high efficiency. The obvious cure was to use two 15,000 horse-power generators. With this arrangement full power is obtained by driving four shafts from two generators, and half power is produced by driving all four shafts from one generator only. This single generator will then be under full load and, by a simple pole-changing switch, will also be running at full speed at its best water rate. Power under one generator will be available for 19 knots in a 21-knot ship; and the other generator, with all of its auxiliaries, will be standing idle, ready for any emergency. Backing with either one or two generators in use simply means throwing a switch for each pair of motors, the generators with their turbines continuing to run in the same direction. When the motors are stopped by opening one switch the turbines run on at no load. Unlike other installations, as much power is available for backing as for going ahead.
Apparently this arrangement completely solved the battleship problem, and it only remained for Mr. Emmet and his engineers to prove their figures in an actual ship and later tell the ship designers exactly how much weight and space would be saved in a battleship. Accordingly the installation was made in the collier Jupiter, whose building and subsequent trials and voyages fully upheld the figures of her designers and convinced even the skeptical of the possibilities of the form of propulsion used. It may be remarked that collier propulsion and battleship propulsion are vastly different matters, and that while electricity may and has demonstrated its superiority in driving a ship which habitually travels at full power, its exclusive advantages apply to battleship and not to collier work. At this point it is interesting to compare with the actual trial results on the Arkansas and Texas, above quoted, the coal consumption of an "electric Texas" based on the guaranteed figures of the designers. In these figures dry saturated steam at the turbine throttle is assumed.
Under the same conditions as for the Arkansas and Texas trials the figures for an electric Texas" would be as follows:
Tons of coal per 24 hours at 21 knots 400
Tons of coal per 24 hours at 12 knots 80
Again comparing this coal consumption with the Arkansas, the " electric Texas" at 12 knots would have almost twice the steaming radius of the turbine-driven Arkansas, to say nothing of the increased backing and maneuvering power.
OPERATION
As to operation, consider the " electric battleship " under way at sea, full power ahead. The port turbo-alternator at 2600 R. P. M. is electrically geared "to the two port motors in their low-gear connection and the starboard alternator is Similarly " locked " to the two starboard motors. The speed of the turbo-alternators, hence of the motors, is held constant by an adjustable, automatic governor on each turbine; so the machinist's mate doesn't have to worry about the revolutions. Nor do racing propellers have any dangers, as the motors cannot appreciably speed up.
"Half speed" is signalled from the bridge. The governor is screwed down to give the desired turbine revolutions. No switches are touched. The motors being electrically locked to their alternator, they, too, slow down correspondingly. In formation, by this variable frequency system, a couple of turns faster or slower are similar obtained.
'Stop!" The machinist's mate (or perhaps the electrician) on watch opens a small switch in the alternator field circuit and the power is instantly off the motors. As the load comes off, the turbine and alternator keep running just as they were, the speed being held down automatically by the governor. All power is now off the alternator, and the main motor switches may be opened without sparking, if 'desired.
“Full speed astern!" The reversing switch for each pair of motors is electrically thrown over to the astern position, the alternator field switch is closed, and the governor is opened up for full speed. The time taken to complete the connections was less than three seconds. Driven by the already running turbo-alternators, the large propellers instantly reverse under the powerful starting torque of the motors, and 100 per cent ahead " power is instantly available for backing, and the ship quickly comes to a stop.
Once more on signal to stop, the field switch is opened and power is instantly off the propellers and they soon cease to revolve. Meanwhile the turbine runs on at full speed, instantly ready to take full load within five seconds. There being no commutators nor slip rings between alternator and motors, there is no sparking or visible show of the work going on within.
Having finished the full-power maneuvers, the cruising rig is put in in a few seconds. Both alternators are disconnected from their motors by opening two field and two main switches, the four motors are connected for the high-gear ratio. by throwing over four switches; the port and starboard bus bars are paralleled by closing one switch, the main motor switches and field switch of one alternator are then closed, and we have the ship going at, half power with four shafts " locked " together and all four motors driven by one turbine running at 2000 R. P. M. and at its best efficiency. The efficiency of the motors remains, above 95 per cent under both conditions, and the ship can cruise at not over a 11-lb. (per S. 'H. P.) water rate under one generator between 13 and 19 knots. The advantages of the system can thus be appreciated. The electric system lends itself readily to contractor control direct from the bridge, should this ever become a desirable feature, and in; the near future we may find the bridge handling the main engines as easily as electric steering is now accomplished.
As to warming up, the turbine and generator can be warmed up without worrying the officer of the deck, and, finally, each motor can be given a kick ahead and astern to see all clear. If the turbine is of the pressure-velocity compound type, there are no small clearances to worry about and actually no warming up required.
Such high-speed turbo-alternators are of ten used in power plants ashore as " stand-by " units when, under usual conditions, power comes in over a transmission line. Sometimes the power line "goes out" for some cause. Steam power must be substituted at once. A cold turbo-alternator is started up as follows:
The throttle is opened all the way as fast as possible and when up to speed and in synchronism with the other alternators to which it is to be connected, the switches are closed to the line and the machine takes its load. The warming up of a cold turbo-alternator is, therefore, in itself, not much of a ceremony.
During the maneuvering above described the alternating voltage in the leads goes up to about 3000 volts at times. Not a single one of these leads moves, however, and from the insulated bushings where the leads come out of the stationary part of the alternator to the same sort of bushings where the leads go into the stationary part of the motors not a " live " bit of conductor is in sight. Between machines and switches the leads are completely covered with an insulating material unaffected by moisture, so that if required the leads themselves could be handled while " alive " with impunity. As a matter of fact the leads are carried well out of reach in order to avoid the possibility of damage. The main switches themselves work in tanks of oil and are enclosed in wiremesh cages with only an insulated handle coming out. This type of oil switch is the one used in all power stations ashore to-day, and the oil is used simply to help suppress the arc when the switch is opened under load. Except the two small 230-volt slip rings to the alternator field there are no brushes on the alternator, and as for the motors there are no leads to the rotors, no commutators, of course, and not even slip rings. In fact, in flooded mine levels, small motors of this, type, driving centrifugal pumps, are at times called upon to run under water. This they do with no damage to themselves except perhaps subsequent corrosion of the steel parts. Similarly, the ship motor could be run in a flooded compartment without any fear of damage (except corrosion), although, due to to the friction of the water, little power could be produced.
There is no commutator on an alternator, consequently no brushes. Due to this and to its inherent characteristics, the alternator can be dead short-circuited and the current will rise only to a value which will do no damage to the machine or conductors. Another valuable trait of this alternating-current machine! There is only one steam pipe, and that to the main turbine, which is located near the engine-room bulkhead.
There will be no steam auxiliary machinery in the engine room, except the boiler feed pump, because all of the engine auxiliaries will be driven by electricity on the " electric battleship." The heated air from the enclosed alternator and motors goes directly out ventilators on the main deck through individual ducts, or this hot air, as in the Jupiter; can be sent under the boiler grates.
Many other details of interest could be mentioned, but after all might only serve to cloud the bigger points. In fact, with a steam power plant the whole raison d'être of electric propulsion may be summed up in a "left-handed " manner by saying that when mechanical engineers succeed in producing a reliable, quick reversing, two-speed gear capable of handling 7000 horsepower at 160 R. P. M., then the day of electric propulsion may be over. Until that time and as long as the experimental Jupiter continues to steam from coast to coast and to weather gales such as that of December, 1914, we may consider ourselves at the beginning of an era of electric propulsion for battleships.
WHAT IS ELECTRIC PROPULSION?
The answer to " Why electric propulsion?" indicates pretty clearly " What is electric propulsion?"
Electric propulsion in the proposed battleship consists essentially of a port and a starboard unit, each consisting of a turbine-driven, two-pole, quarter-phase alternator and two double squirrel-cage induction motors, and it is largely a description of these machines that forms the remainder of this paper. Direct-current field excitation, at 230 volts, for the alternator, is supplied from a turbine driven generator in the engine room or, if desired, is obtained from the ship's mains. Conversely, the two "exciters " may be used in port to supply power to the ship's mains and thus form two valuable additions to the vessel's power plant. The motors have two possible arrangements of poles, the change from 36 poles to 24 poles being accomplished by simply throwing a switch. There is one motor in a separate compartment for each of the four shafts. For 21 knots both turbine units are run and the four motors are used in their 24-pole rig (low gear). This reduces the full turbine speed of 2000 R. P. M. to 165 R. P. M. of motor or propeller. At cruising speeds one turbine is connected to the four motors in their 36-pole rig (high gear). Thus either turbine at 2000 R. P. M. drives four motors at 1 10 R. P. M. The other turbine is meanwhile not in use. When the motors are not connected, the turbine runs at no load and the motors stand idle. If connected to one generator, all the motors turn (if in the same pole setting) at the same speed, although, if it is desired to back the port and at the same time stop or go ahead on the starboard motors, this is perfectly possible. In other words, when in cruising rig the ship can be quickly started, stopped, or turned by various combinations of the four motors driven by one turbine, and power for 19 knots is available. Speed control with either one or two alternators running is entirely by turbine throttle or on the so-called " variable-frequency " principle. It is at once evident that the electricity between turbine shaft and propeller shaft simply performs the same function that a clutch and gear-box do on an automobile. The electric machinery simply makes it possible to reverse the propeller shafts and keep the turbine running in the same direction, and also gives two gear ratios between turbine shaft and propeller shafts. The electric rig does not have a "direct drive," as does a motor car, because this is just what is not wanted in a ship. On the other hand, the electric rig gives two speed ratios ahead and two backing, while the motor car gives usually two ahead (besides " direct ") and one backing. Being in any gear setting with either ship or motor car, to go faster we speed up the engine; to slow, we slow the engine. Without shifting "gear," speed changes can be made in no other way. In the motor car we can at any time throw out the clutch, and leave the engine running; with the ship, we can instantly disconnect the motors electrically and leave the turbine running. In fact, the analogy between electric propulsion and automobile drive is almost exact, and the feature of speed control by the turbine governor and by no other means cannot be too strongly emphasized.
Between the exciter and the alternator there are, besides the usual protective circuit breakers, one switch and one rheostat. As has been said, the ship's circuit can be used for supplying excitation for the alternators, or in port the exciters, which are of about 300 KW. capacity, can be used for supplying power to the ship's mains. The four main leads from the stator of, say, the port alternator go through a four-pole switch to the port bus bars. As this switch is opened, only when no power is on the system, it is of the usual knife type. The two port motors are connected in parallel and then through either one of two electrically actuated main oil switches to the bus bars. When closed, one of these oil switches connects the pair of motors for " ahead " operation; and when closed, the other switch connects the motors for " astern" operation. Only one of these switches can be closed at a time, and the closed switch is locked so that it cannot be opened until after the alternator field switch is Open. This last provision makes it impossible to break the main circuit until the current in it practically ceases to flow.
In each motor circuit there is a four-pole, double-throw switch. One closed position of this switch connects the motor for 36 poles, the other closed position for 24 poles. As the pole-changing switches are never used with power on the circuit, they are simply knife switches.
When in cruising rig with all four motors driven by one turbine unit, the port and starboard bus bars are connected together. At all other times the port and starboard sides of the ship are not electrically connected.
As the alternator field switch, the main switches, and even the turbine governors will in all probability be electrically operated, two master controllers, one in front of the instrument board in each engine room, will suffice for handling the entire main propelling plant.
Alternating current machinery of the type used in ship propulsion is essential rugged and "fool-proof," and by interlocking the main switches no danger of wrong connections is possible. An engineer force experienced in electric work is not at all essential, as is shown by the very satisfactory results attained on the Jupiter, where, aside from one or two officers, there were no experienced power-plant men on board. Confidence comes quickly with experience, and the bogey of alternating current soon disappears.
To one familiar with the applications of alternating current there are only one or two new features connected with the whole design. None of the apparatus is remarkable on account of size, and voltage is low compared to most modern power plants. In fact, the standard navy-yard voltage for generators is 2200 volts, and motors of 50 horse-power and above in navy yards are run at a line voltage of 2200 volts. The standard voltage for generators and motors in commercial power plants is 11,000 volts at the machine. To obtain this low voltage the potential is stepped down by transformers from, say, 110,000 volts of transmission line to 1,000 volts in the motor windings. In other words, electrical engineers class 3000 volts on board ship as low voltage, and, from any point of view, the use of such voltage is entirely reasonable.
* * *
As an aid to those not accustomed to alternating currents, a brief description of the various machines is undertaken. While it is always difficult to explain a machine to one rusty in theory, and while the field is a broad one, it is believed that at least the principles of operation can be readily understood. In so doing a knowledge of the fundamentals of direct-current electricity must be assumed.
THE ALTERNATOR
Practically all electric generators, direct-current or alternating current, depend upon the production of voltage in copper conductors by mechanically cutting magnetic lines of force. Whether the cutting conductors move through a fixed magnetic field, as in the familiar direct-current dynamo, or whether the conductors are stationary, as the magnetic field moves by them, makes no difference. As being the ship type, the latter variety of machine will be described.
Voltage is really an " electro-motive force " and expresses electrical pressure, just as " pounds per square inch expresses steam pressure. The size of the electrical current, direct or alternating, is expressed in amperes. As just said, electrical pressure or voltage is generated in a machine by causing magnetic lines or a so-called magnetic flux to cut a conductor. Once a pressure is produced between terminals, a current Will flow in a wire if connected between the terminals. The dynamo or direct-current generator is essentially an alternator to which is: added a commutator for rectifying the alternating current generated. Without the commutator the machine becomes much simpler, and particularly when the field is made to rotate and the armature is fixed. The fixed part is called the stator and the rotating element the rotor. Fig. 1 shows the principle of a single-phase, two-pole alternator. F is the rotating magnetic field, usually built of sheet-steel discs and wrapped with wire (not shown) carrying direct current for producing the field. Every revolution
{figure}
FIG. 1
of this field causes its lines of force to cut the fixed armature coil, of one or many turns, which is imbedded in the steel armature or stator (Fig. 2)., The steel of the stator forms an easy return path for the magnetism from the north back to the south pole of the field, and prevents its " leaking " into surrounding space.
{figure}
FIG. 2
The cutting of the armature conductor by the rotating lines of force induces an electro-motive force or voltage in the coil which appears across the terminals 1-2, Fig. 1.
C1 and C2 in Fig. 2 show the coil imbedded in the stator. When the rotating field NS of Fig. 2 is horizontal, no lines are being cut in coil C1-C2; hence the voltage in the coil is zero. When the field is vertical, with N at the top (as shown), the coil is most rapidly cutting the rotating magnetic flux and a 'maximum voltage is produced in one direction. When S is at the top the voltage is again t maximum, but in the other direction. An alternating voltage or pressure with instantaneous values following a sine wave about as shown in Fig. 3 is therefore produced between terminals. This voltage wave makes one complete cycle per revolution of the field. If c represents the instantaneous value of the varying voltage between terminals and E the maximum value, then at any instant ?=E sin ?, where ? is a function of time. At 1500 R. P. M. this means 1500/60=25 complete alternations or cycles per second. Such a machine is therefore known as a single-
{figure}
FIG. 3
phase, two-pole, 25-cycle alternator. The number of cycles per second is called the frequency.
If another coil exactly like the first is added to the stator, with the plane of the second coil at 90° to the first, a wave of voltage with characteristics just like the one in Fig. 3 is also produced in the second coil. Due to the location of the second coil in the stator, however, this wave C3-C4 differs 90° in time or in phase with respect to the first wave. Thus when C1-2 is a maximum C3-4 is zero, and conversely. The two waves in their proper relative positions are shown in Fig. 4 and are said to be in time-quadrature. With the addition of this second coil we have the equivalent of two separate machines in which corresponding instantaneous voltages occur 90° or a "quarter phase" apart. The two coils have no electrical connection whatever. This machine will then have four terminals and, at 1500 R. P. M., will be called a two-pole, quarter-phase, 25-cycle alternator. In the design the voltage is governed by the number of turns put in series in each coil; and when completed, the voltage is regulated, at constant speed, by varying a rheostat in the direct-current field circuit. In other words, voltage depends upon the rate of cutting magnetic flux and the voltage may be varied directly by varying the rate of cutting in any way convenient. The most convenient way to do this is by a rheostat in the field circuit. More field current means a denser field, more flux cut, hence higher terminal voltage of the alternator. The frequency (cycles per second) is varied by varying the turbine speed, and by this means only. Such an alternator, conversely, may be driven by supplying quarter-phase power to its four terminals, and when so driven is called a synchronous motor.
{figure}
FIG. 4
A synchronous motor driven by alternating current depends for its speed on the frequency of the alternator supplying the current, and not at all on the voltage of the alternator; hence the electrical bond between alternator speed and motor speed is absolute. Also the synchronous motor must be speeded up without load until it is in step with the applied frequency, and the load be then put on. This form of starting under no load is impossible in most cases; hence the synchronous type of motor is seldom used.
Synchronous motors are therefore not suitable for ship propulsion, as they give no torque at starting, and because, if overloaded, they get out of step and have to be started up without load (like a gas engine) all over again. Many of the principles of the synchronous motor are, however, found also in the induction motor, which is the type used for ship propulsion.
An alternator with a two-pole field running at 1500 R. P. M. (25 cycles) will thus drive another two-pole alternator as a synchronous motor at 1500 R. P. M., and at no other speed. This same alternator at woo R. P. M. will drive the same motor at 1000 R. P. M. only, and so on. If the driving alternator had a four-pole field (Fig. 5) running at 1500 R. P. M., 50-cycle current would be produced and the synchronous motor having two stator poles (Fig. 2) would run at 3000 R. P. M. only, and at moo R. P. M. of alternator the motor would run at 2000 R. P. M. only, etc. For ship propulsion the two-pole alternator is run at 2000 R. P. M., and is *electrically interlocked to one or more induction motors which can be connected for either 24 or 36 poles at will. With turbine and alternator running at 2000 R. P. M. and motor in 24-pole setting, the propeller revolves at 165 R. P. M. for a speed of 21 knots. With turbine and alternator still at 2000 R. P. M. and motor in the 36-pole setting, the propeller revolves at 110
{figure}
FIG. 5
R. P. M., say, for 13 knots. Remembering that 2000 R. P. M. is the turbine's most economical speed and that one turbine drives all four motors at 13 knots, the reason for low cruising water rate is apparent. We thus get the desired two-ratio reduction gear with electricity as the working medium.
For many purposes alternators are built with three coils 1200 apart instead of two coils 90° apart, and when thus constructed they are known as three-phase machines. The quarter-phase type is adopted for description, largely because of greater simplicity of connections, and, in fact, a quarter-phase system may be put in the first battleship, as giving better possibilities for, obtaining the desired number of motor poles in the full-power and half-power rigs.
It is likewise, possible to build alternators and certain types of motors to run single phase, but these machines are not generally satisfactory, and a single-phase induction motor has no starting torque, although when up to speed it will run and produce mechanical power. As will be explained, polyphase induction motors do have a starting torque and are a very satisfactory type of machine. For analogy, take a single-cylinder engine, a cross-compound engine, and a three-crank (120°) triple-expansion engine. Here we have, so far as the crank effort is concerned, one, two- and three-phase machines, and here, too, the one-phase engine is apt to be " on the center " and fail to start.
ALTERNATING-CURRENT MOTORS
Electric signs frequently have a couple of "squirrels" running around the rim. The " squirrels " consist of a group of successively lighted incandescent lamps mounted at the edges of the
{figure}
FIG. 6
sign. The lamps of course do not move, although the "squirrels" actually do travel around the edge of the sign. To produce a " squirrel" about nine adjacent lamps are lighted at once, the contact being moved on the actuating drum so as to cause successive illumination of the border lamps. If, instead of nine or ten lamps, only about six are on at -one time, instead of a "squirrel” a "rat" chases around the sign. Similarly 'four lamps produce a "mouse." The number of " animals " chasing around the edge is limited only by the size of the sign and the size of the lamps.
The point is, we do get rotation of the, illuminated section, although the lamps themselves do not move. Another familiar case is found in water waves. Here a wave may progress at considerable velocity across the ocean, yet the water particles actually move only vertically up and down or in restricted oval paths. Take now the alternator, or motor stator, of Fig. 6, omitting any rotor, and consider terminals C1, C2, C3 and C4 to be connected with corresponding terminals of the alternator, Fig. 2. Fig. 4 shows that the voltages impressed across the ends of the motor conductors will be sine waves. Current, according to
{figure}
FIG. 7
Ohm's law, being voltage divided by a constant, the current in each phase will also be a sine wave of similar characteristics.
The current waves (Fig. 7), to a suitable scale, will therefore be similar to the voltage waves already shown in Fig. 4. When the current is a maximum in one direction in coil C1-2, there is no current in coil C3-4. This is shown in position 1, Fig. 7, at which
{figure}
FIG. 8
point one current wave is zero and the other is at a positive maximum. The current running round and round the turns of C1-2 makes a magnet out of the stator, with temporary poles induced in the iron of the stator as shown in Fig. 8. One coil is seen to be working with maximum current, while the other, carrying no current at this instant, has no effect. For illustration, the current is here assumed to flow down C, (normal to the plane of the paper) and up C2. Current flowing down is shown as x; current flowing up, as ¤, and no current flowing, as ¢. When the alternator of Fig. 2 has turned 45°, the current wave in the two motor windings will be as in position 2, Fig. 7. C1-2 is plus a certain value and C3-4, is minus an equal value. Therefore both coils now have current in them, and in such a direction that the resultant magnetism of the stator is as shown in Fig. 9, and the fields produced by the two coils adding together, as they do, the total flux is of the same strength as before. Another 45° rotation of the alternator, and the stator magnetism in the motor has
{figure}
FIG. 9
moved around as shown in Fig. 10. Coil 1-2 now carries no current, and C3-4, is at a negative maximum.
For convenience we have been going by jumps' of 45°. In the actual stator the induced poles actually revolve at uniform speed and the strength of the rotating stator field resulting from the two phase currents is constant. Thus the cycle of events goes on indefinitely, and the induced magnetic poles marked N and S rotate just as truly as though the whole frame of the stator were permanently magnetized like the field of a direct-current dynamo and revolved at uniform speed in a lathe. The illuminated sign is now a stator, and the pair of " squirrels " are the induced magnetic poles chasing each other around. The pair of induced poles on the inside of the stator thus revolve once per one revolution of the alternator field, and this again brings out the fact that motor field and alternator field (when of the same number of poles) must revolve at the same speed. Upon the rotation of the stator poles depends the whole action of synchronous and induction motors. In the case of the quarter-phase currents of Fig. 7, the rotation
{figure}
FIG. 10
of the resultant flux produced by the two windings can be easily shown in another, way. Let C1C2 (Fig. 11) represent the plane of one stator coil and C3C4 that of the other coil. These two coils are,
{figure}
FIG. 11
of course, 90° apart. At any instant the flux due to the current in one coil may be represented by ON1, and at the same instant the flux due to the other coil will equal 0M1 both in magnitude and in direction. The resultant flux is OP1. At another instant the component fluxes will be ON2 and OM2 respectively. The resultant of these two is OP2, etc. OP1 equals OP2; hence the resultant stator flux is of a constant value and actually rotates at uniform speed. All the way through, for the sake of simplicity a two-pole stator winding has been considered. Actually the stator can be wound to produce practically as many poles as desired. Fig. 12 shows a stator wound for 14 poles. In such a stator one complete cycle of current, instead of making a complete revolution of the rotating flux, as in Figs. 8, 9, and 10, would revolve it uniformly only from N to the next N in Fig. 12. Assuming the alternator to have two poles, the flux in the 4-pole motor rotates only one-seventh of a revolution per revolution of the driving alternator.
{figure}
FIG. 12
A 7-to-1 reduction is thereby obtained. By throwing a switch the windings of the 14-pole motor can instantly be so connected as to produce, say, eight poles, and a 4-to-1 reduction results. Here, by throwing a switch, is our two-speed gear!
All motors, for either direct or alternating current, depend for operation on the magnetic attraction of unlike poles and the repulsion of like poles. Having demonstrated the fact that the stator of an alternating-current motor possesses a rotating magnetic field, we will now insert a rotor (Fig. 13) having one pair of poles just like the rotating field of the alternator. If this field is excited by direct current through slip rings, we shall have a two-pole synchronous motor, previously described. The magnetically rotating stator-field will be looked to a mechanically revolving rotor-field. The lock is the magnetism across the air gap. We have all along assumed C1 and C2 of the alternator to be connected to C1 and C2 of the motor, and C3 and C4 of the alternator to C3 and C4 of the motor. With this arrangement the motor revolved in the same direction as the alternator. If now C1 and C2 remain as before and C3 and C4, of the alternator are connected to C4 and C3 of the motor, respectively, conditions will be identical, except that the rotating stator field (hence the rotor itself) will revolve in the opposite direction. This can be readily demonstrated, as was done in Figs. 8, 9 and to.
{figure}
FIG. 13
Hence, to reverse the direction of rotation of a quarter-phase motor, interchange the two connections to either one of the phases. Here, by throwing a switch, is instant reversibility!
From what has been said, it is evident that a two-pole alternator at 1500 R. P. M. will drive a four-pole motor at 750 R. P.M. We have already called the alternator at 1500 R. P. M. a 25-cycle machine; therefore the speed of the motor in revolutions per minute is 25x60/2=750; 25 is the frequency and 2 represents pairs of stator poles. The rule thus becomes: The synchronous R. P. M. of an alternating-current motor equals the frequency multiplied by 60 divided by the number of pairs of stator poles.
In general, therefore, the alternating-current motor can be made to fulfill the functions of the gear-box on the automobile, as also certain of the desired features of ship propulsion.
The foregoing general description of the alternating-current motor was introduced largely to demonstrate more easily the fact that there is produced in the stator an actual rotating magnetic field. Among other disadvantages, the synchronous type of motor will not start under load, and of course it requires direct-current field excitation. The induction motor will exert a torque at starting and requires (in the type finally adopted for battleship propulsion) no leads whatever to the rotor. Whatever differences exist, one important feature is the same in both types. The stator
{figure}
FIG. 14
of a synchronous motor and the stator of an induction motor are identical in every respect. Pole changing and reversing are therefore identical in the two types, and the rotating magnetic field is present in each. Referring to Fig. 14, consider the stator to be wound for quarter-phase current as before described, and that the rotating field has been established. Insert a solid iron rotor with one coil of wire on it, as R1R2 (Fig. 15). The coil may be of one or many turns, with the ends brought out. Fig. 15 shows an elementary axial section. Consider for a moment the rotor to be held stationary, also remembering that a current in a wire wrapped around iron produces magnetic poles on the iron whether the current be direct or alternating, and no matter how the current is produced. In the synchronous motor the current in the coil R1R2 would be direct current through slip rings. In induction motor the rotor current is induced.
Start with the rotor stationary, and with coil R and rotating field NS in the relative positions shown in Fig. 14. The rotating two-pole field does not actually cut the stationary conductors of coil R in this position, as all of the stator flux threads through the
{figure}
FIG. 15
coil. The number of lines through the coil in this position is a maximum, but the rate of cutting is zero. No rate of cutting flux, no induced voltage. Therefore, no voltage is induced in
{figure}
FIG. 16
the coil, and if the ends were short-circuited no current would flow around the rotor. Keeping the ends R1 and R2 (Fig. 15) short-circuited on each other, assume the rotating stator flux proceeding clockwise towards NS (Fig. 16). This moving flux is now cutting coil R at an increasing rate; hence a sine wave of current is induced in the rotor toil just as in a transformer. A maximum rotor current occurs when NS is vertical, as shown in Fig. 16. This current magnetizes the rotor iron in the direction N1S1. S1 is naturally attracted by N1 and N2 by S, and the rotor tends to revolve in a clockwise direction. Hence we have an action similar to, a synchronous Motor; only the poles on the rotor are not made by direct current from an outside source; but by an induced alternating current in the rotor. Furthermore, even when the rotor is stationary, as at starting, mechanical torque is produced. If followed out through a complete revolution of the stator flux, it will be found that the rotor will tend to turn (or be dragged around) through a complete revolution also, and if mechanically released will so turn, in a clockwise direction. As actually built, a symmetrical winding of many turns is put on the rotor to get even balancing and constant torque. The elementary theory is equally true, however, with many coils. Reversing the direction of rotation of the stator flux by interchanging the two leads to either phase will reverse the rotor just as in a synchronous motor, and by winding the stator for many poles low rotative speeds are also possible. The following difference in the two types of motor is, however, paramount. In the synchronous motor the rotor receives its magnetizing current through slip rings from an outside source, while in this motor the rotor current is produced directly by having its windings cut by the rotating stator flux.
Hence the name induction motor. Right here there conies a hitch. Suppose, after releasing the rotor, it should speed up until the coil R is rotating just as fast as the stator field? At this point, there being no relative motion between stator field and rotor winding, there will be no cutting of flux, no rotor voltage, no current induced in the rotor winding, hence no rotor magnetism. There will, therefore, be no torque.
The speed of the rotating stator flux in R. P. M. is called the synchronous speed, and is the exact and only speed at which a synchronous motor will run. But according to our theory of the induction motor, the rotor ceases to get its induced current and magnetism at synchronous speed and will fail to work. Such is the case. Therefore, what happens?
Instead Of running at synchronous speed the induction motor slows down until its rotor windings do have a Motion relative to the rotating stator field, and to a point where this relative motion is sufficient to generate the required low-frequency current in the rotor windings. The difference between synchronous speed and actual rotor speed is called the slip of the motor. Slip in R. P. M. is usually divided by synchronous R. P. M. and expressed as a percentage.
As assumed all the way through this description, the frequency in the stator windings has been 25 cycles. When the rotor of Fig. 14 is stationary, the induced frequency in the rotor coil or coils is, of course, 25 cycles, too. If the rotor could get up to synchronous speed, the rotor frequency would be zero; therefore no current would flow. This case has just been dealt with. At a slip 5 per cent, what is the rotor frequency? It is obviously S/100 X 25 cycles per second. We have said voltage is rate of cutting flux. The rate at which the rotor conductors cut the rotating stator flux is zero at synchronous speed, is full frequency at no speed (rotor stationary), and between these two is proportional to the slip. As the rate of cutting varies as the slip, the induced rotor voltage also varies as the slip. But current=voltage/resistance.
For given current a small induced voltage or slip implies low rotor resistance. Therefore, the lower the rotor resistance the nearer the rotor can approach synchronous speed. This low resistance is obtained, often, by making the rotor winding simply of copper bars and short-circulting the ends of these bars with heavy copper rings. It is not even necessary to insulate the copper bars from the iron core. See Fig. 17, which is a photograph of such a rotor. This is called a "squirrel-cage" rotor. The stator used with this rotor, as has been said, consists only of insulated copper wire imbedded in slots and, of course, does not move. This, then, is essentially the type of motor which may run under water without possibility of injury, as mentioned at one place under " Why electric propulsion?'" The slip of a squirrel-cage motor at full load may be as high as 5 per cent. The slip decreases as the load is taken off, and at no load may be a fraction of 1 per cent.
The squirrel-cage induction motor is therefore essentially a constant-speed motor, and its characteristics are comparable to a direct-current shunt motor.
If the load on a squirrel-cage motor is increased, the slip increases, and at a certain point the magnetic bond between the rotor and stator becomes overstrained and the motor stops, or breaks down. There being no commutator to burn up, no damage is done, although the current in the windings may reach a very high value. This breakdown torque is at least double the rated full-load torque. From this it follows that in starting a squirrel-cage motor, at full voltage, under, say, full-load torque, a very
{figure}
FIG. 17
heavy current will flow ; perhaps six or seven times normal full-load current. This in many motors is prevented by introducing starting compensators. A compensator is a small transformer which reduces the stator voltage until the motor is up to speed. Holding down the voltage applied to the motor in starting has the desired effect in preventing the flow of excessive currents, but at the same time this limiting of the current at starting cuts down the starting torque considerably. It can be proved that the maximum torque which an induction motor can exert depends directly on the square of the applied voltage, and that the slip at which this maximum torque occurs can be regulated at will by adjusting the rotor resistance. This is shown in Fig. 18. Curve A shows a squirrel-cage torque curve. Here maximum torque is at 10 per cent slip, and full-load torque at 3 per cent slip. Curve B is a higher-resistance rotor with maximum torque at 28 per cent slip. The line MN represents full-load torque. By making the rotor resistance sufficiently high, maximum torque can be given at 100 per cent slip or, in other words, at starting (curve C). Note that at constant voltage the maximum value of torque is the same in all cases, and only the position with respect to slip is changed by
{figure}
FIG. 18
varying the rotor resistance. Broadly speaking, then, in a squirrel-cage induction motor the speed at any given load can be varied by varying the applied frequency, and the maximum to can be varied by varying the applied voltage. This type, with its poor starting features, is therefore not altogether suitable for ship propulsion.
THE PHASE-WOUND INDUCTION MOTOR
This brings us to a second type of motor called the phase-wound induction motor. The stator is the same as before. Here the rotor windings are not short-circuited at R, and R., (Fig. 15), but are brought out to slip rings. The motor then becomes a phase wound motor.
If these slip rings are short-circuited, we have the equivalent of a low-resistance or squirrel-cage winding just described, and in this rig the motor is suitable for constant-speed work. If, instead of short-circuiting slip rings R, and 122, we insert resistance between them (B, Fig. 15), we thereby demand more rotor voltage to force the current through the coil. We therefore have a temporary high-resistance rotor. This means we can get the same torque that we can in a squirrel-cage motor, but the slip will be greater to get it. By suitably proportioning this resistance, any starting torque (up to the maximum possible) can be obtained with only normal full-load current, and by using a controller the
{figure}
FIG. 19
resistance can be varied at will. Curves of torque plotted on slip for the squirrel-cage and phase-wound types of motor are shown in FIG. 18. Curve A represents a squirrel-cage motor of low resistance and shows maximum torque at it) per cent slip. Curves B and C show a phase-wound rotor with resistance enough added across the slip rings to give maximum torque at 28 per cent and too per cent slip at will. This then makes a good variable-speed motor, as does the direct-current series motor with rheostatic control. Furthermore, the heat losses in the resistance B (Fig. 15) are strictly comparable to the heat losses in the starting rheostats of, say, a direct-current series street-car motor. Fig. 19 shows a three-phase-wound rotor.
A phase-wound induction motor, with slip rings which can be either short-circuited at the higher running speeds or connected through a resistance for starting or maneuvering, demands no excessive starting currents and must, then, be a good solution for ship propulsion. Motors of this type are in satisfactory use on the Jupiter.
It was only after years of work and research that resistances could be built for handling the tremendous rotor currents in a ship motor. This was solved on the Jupiter by putting the special alloy resistances in tubes of sea water. The water is circulated through the tubes when necessary by a connection to the main circulating pump.
While a phase-wound motor with variable rotor resistance makes a possible solution for either battleship or collier, the elimination of the resistance and slip rings would in general make a much better solution, provided the desired torque slip characteristics could be obtained.
Electrical engineers think that the solution is obtained in the double squirrel-cage induction motor, and there is every reason to believe they are right. Motors of this type have been built in the smaller sizes and thoroughly tested, although for general power purposes ashore they are little used.1
THE DOUBLE SQUIRREL-CAGE INDUCTION MOTOR
It has been pointed out that the squirrel-cage induction motor, with its low-resistance rotor windings, makes an ideal, highly efficient constant-speed motor, but, due to these same low-resist-
1 From an electrical point of view the induction motor is most readily understood by comparison to a transformer. Practically all calculations on induction motors are made on this assumption.
The stator is the transformer primary, the rotor the secondary. Both machines inherently take a lagging current, hence form an inductive load on a line. The output of the transformer secondary is electrical energy; the output of the induction motor appears partly as heat energy in the rotor conductors and resistances and partly as mechanical power.
The mechanical power of the rotor is obtained as a result of magnetic stress between rotor and stator. This same magnetic stress exists in transformers and in. the early days of transformer construction wrecked many a suddenly overloaded machine. Steinmetz evolved the " equivalent " transformer circuit as the readiest means of calculating transformer results. This same device applies equally well to the induction motor and is readily understood.
ance windings, gives low starting torque, when only normal current is consumed, and, if full-load starting torque is demanded with a starting compensator, about 250 per cent full-load current is obtained. If full voltage is applied to a loaded squirrel-cage motor, approximately 700 per Cent full-load current is required for starting.
The phase-wound motor, by temporarily increasing the rotor resistance, outside of the machine, gets around the starting difficulties of the squirrel-cage type. It does this, however, at the sacrifice of efficiency and with the mechanical difficulties incident to slip rings and cumbersome resistances. In large motors the resistances become very sizable, and, indeed, their construction was for years the main hindrance to the adoption of phase-wound motors for ship propulsion.
Designers, scanning the numberless schemes that have been evolved to perfect the starting of induction motors, finally eliminated all but the Boucherot method. M. Boucherot succeeded in building a squirrel-cage induction motor with a rotor so designed electrically that at low speeds (high slip) it offered high "resistance" to the rotor currents, while at high speeds the effect of the windings was to give a low "resistance.”1 This means a rugged construction, with the advantages of a squirrel-cage and also the starting characteristics of the phase-wound type. All this is obtained with no additional mechanical difficulties, with negligible loss in efficiency, and without any exterior resistances or connections. Furthermore, the two characteristics, that of low "resistance" at high speed and high " resistance " at low speed, are adjusted automatically as required by the machine itself. All this is obtained by what is known as the double squirrel-cage construction.2 Just above, it was stated that the rotor of M. Boucherot's motor had a high " resistance " under certain conditions, and, under other conditions of slip, a low "resistance." In so saying, certain liberties were taken with the good word "resistance," and it is here necessary to more accurately describe what actually happens. A short digression into theory is involved.
1 See " Electric Motors," by Henry M. Hobart.
2 The full-load power factor of a double squirrel-cage motor may be 10 per cent lower than in an equal-size single squirrel-cage machine.
Ohm's law and Kirchoff's law give all that is needed for direct-current work, and the former is often expressed, simply as E=RI, where E= electro-motive force or voltage, /= current in amperes, and R= resistance in ohms.
A current, direct or alternating, in a wire always creates a closed magnetic field or flux around that wire (Fig. 22). When the current is direct, once the current is established, the field stays around the wire and its presence is forgotten (unless, perhaps, there is a compass near). Every time an alternating current pulsates, however, the field around the wire is set up in one direction, brought back to zero, then set up in the other direction.
In the present case this happens 25 times a second. This magnetic field, waxing and waning, and every time cutting the wire, induces an electro-motive force in the wire just as the rotating field of the stator induces a voltage in the rotor conductors. No, longer does Ohm's law (as usually stated) apply. The new condition is expressed by an equation ?=Ri+K d?/dt.
Here again ?=the electro-motive force in volts and i=the current in amperes. Small letters are used to denote instantaneous values. According to the calculus d?/dt=rate cutting lines, where ? represents the number of encircling magnetic lines (the flux) and t represents time. Rate of cutting flux represents an electro-motive force in the wire, and the constant K is introduced to reduce this to volts. Therefore the law becomes ?=Ri+K d?/dt, where the term K d?/dt must be added to the Ri voltage of Ohm's law on account of the varying magnetic field always present with an alternating current.
As the flux in the air around the wire is caused by the current in the wire, it is found that the size of the flux ? varies directly with the current i, and, if we wish, i may be put in our equation in place of ?; i can be easily measured, while ? cannot. The only change will be in the constant, which must be of a different size to still express volts. K d?/dt therefore becomes L di/dt, and the equation may be written ?=Ri+L di/dt. It is now in convenient form, and R and L being constants, we have ? expressed simply as a function of i.
Look at Fig. 20, where the solid line represents the varying values of i, any point on the curve being reached 25 times a second. How can we draw a curve expressing L di/dt? From the calculus it is evident that when i is a plus or minus maximum di/dt is zero; also the rate di/dt is a maximum when the curve of i is steepest. Examination of the current curve show i that L di/dt must therefore look like ?1 (Fig. 20).
{figure}
FIG. 20
?1 is a sine wave of the same period as i and occurs 90° ahead of i; is still amperes and by hypothesis we have used L of such a value as to make ?1 volts. We can now say ?=Ri+?1. Unfortunately, c, cannot be added algebraically to Ri because we have just said ?1 is 90° ahead of i, while Ri is in phase with i. They can be added, however, just as forces are added in mechanics. This is called vector addition and is necessary where quantities have direction as well as magnitude. The clots under the letters in the last equation represent vector addition. The vector addition is correctly shown in Fig. 21.
In the preceding paragraph we shifted from K d?/dt to L di/dt, saying i (amperes) was easier to measure than ? (flux). L we
{figure}
FIG. 21
called an arbitrary constant. A straight wire in air carrying a given current will have a definite flux around it as shown roughly
{figure}
FIG. 22
in Fig. 22. if this wire were bent into a coil of two turns (Fig. 23), the augmented flux caused by the current in the parallel
{figure}
FIG. 23
turns would encircle or link more turns of wire; hence the induced voltage would be much greater. A bigger value of L is therefore required for a coiled wire carrying current than for the same length of straight wire with an equal current. L is therefore dependent upon the shape of the circuit, and is called the coefficient of self-induction or often the inductance of the circuit. A piece of iron in the coil will increase L, etc., and in general L is difficult to determine. Once given a coil permanently wound, L can be found, if desired, and is really some sort of a constant depending on the shape and surroundings (iron or no iron) of the coil.
Doubling the frequency of i doubles the rate di/dt, hence doubles L di/dt or ?1. To provide for this possibility we measure L at one dt cycle per second and use fL, where f= frequency in cycles per second.
Ri, according to Ohm's law, is an electro-motive force in phase with the current and equal to current (amperes) multiplied by a constant, and when the constant R is in ohms the electro-motive force is in volts, irrespective of frequency. ?1 is also an electromotive force equal to current times fL, and if the current is multiplied by 2?fL, ?1 will also be in volts. For convenience, then 2?fL will be written X, and is a constant expressed in ohms.
For any coil, at any given frequency, X is called the inductive reactance and is a constant with which to multiply the current in the coil to get a voltage across the coil occurring 90° in time ahead of the current. ?1=XI. Fig. 21 represents such a voltage.
As indicated by the arrow, counter-clockwise rotation is assumed, and XI is shown to be 90° ahead of RI. While ?, ?1, and i have been written in small letters to indicate the instantaneous values of the alternating voltage and current at any moment, the equation is equally true for the effective values which are read on all voltmeters and ammeters. The equation will hereafter be written E=RI+XI. From the figure it is evident that E2=(RI)2+ (X1)2, or E = I√R2+X2. If for √R22+X2 we use Z, we have E= ZI. Z is a constant expressed in ohms by which to multiply an alternating current (amperes) in a circuit to get the total voltage across the circuit, and is called impedance. In a circuit where the frequency varies (as in the rotor of an induction motor starting up), the resistance (R) is a constant and the impedance (Z) varies as the reactance (X). We thus have the equivalent of a rotor of variable " resistance " previously mentioned. Z now takes the place of R in Ohm's law for direct currents, and we will hereafter use it where applicable. The apology to the word' resistance" is complete, and we accordingly say that the Boucherot rotor is so designed that at low speeds (high slip) the impedance to the rotor currents is high and at high speeds it is low. It follows that low rotor speed=high slip=high rotor frequency=high rotor reactance (2?fL) =high rotor impedance. The converse is likewise true.
Therefore at low speed we have a high impedance rotor like a Phase-wound motor in starting rig, and at high speed the impedance is low and we approach the squirrel-cage characteristics. The ideal characteristics for ship propulsion are thus obtained,
{figure}
FIG. 24
and all the while with no rotor rheostats nor even slip rings—simply d rugged iron and copper squirrel cage.
How, then, are these conditions realized? Fig. 24 represents a section through the rotor showing an outer and inner bar of the double squirrel-cage winding. Thinking of these copper bars simply as two parallel wires, we would like to have the pair of them (acting always together in parallel) at low speeds offer high impedance and at high speeds low impedance.
Notice the annular air gap between stator and rotor, also the axial air gap between inner and outer bars. In designing, these air gaps can be made any size at will.
In developing the idea of reactance it was stated that XI was the voltage produced in a wire by the alternating magnetic flux around it. The magnitude of the reactance, therefore, depends upon the amount of this flux. It takes 2000 to 3000 times as much force to drive a magnetic flux through an air gap as through the same thickness of steel. Therefore, by introducing an air gap near a wire imbedded in iron we can reduce the encircling flux, hence the reactance of the wire. Refer now to Fig. 24 and note the two air gaps and stator slots near B2 which the flux around that bar must cross when a current flows in the bar. At the same time B, has only one narrow gap to hinder its flux. The reactance of B2, therefore, can be made as much larger than the reactance of B1 as the designer wishes, and by burying B2 deep in the rotor iron its reactance may become very high.
We have, therefore, a bar of high reactance, B2, and a bar of low reactance, B1. Reactance being a function of frequency, it follows that at starting (high rotor frequency) the bar B2 will have high reactance; hence the parallel circuit of B1 and B2 will have high impedance (the ohmic resistance of both being assumed not to vary), and we have produced the high-"resistance" rotor originally mentioned as so desirable in starting. With the very small slip at running speeds the reactance of B2, hence the impedance of the two bars is low and the squirrel-cage characteristics appear.
So far as the reactance of B2 is concerned, the radial air gap could be omitted, but if there were no air gap between B1 and B2 the main flux through the rotor would go between B1 and B2, and thus much of its use on B2 would bp lost. Figs. 25 and 26 show the real reason why the radial gap extending between B1 and B2 is used. The fact that its presence decreases the reactance of B2 is incidental and actually makes little difference. In other words, with the gap (Fig. 26) it is easier for the rotating main flux from N to S to go around B2 (as is desired) than to short-circuit or "leak " between the bars (Fig. 25).
* * *
Summing up the machines, we may say that a quarter-phase alternator delivers separate alternating sine waves of current to two windings 90° apart. The four stationary terminals of the alternator are connected to four similar terminals of a motor stator. Whether the stator be that of a synchronous motor or that of an induction motor of any type makes no difference, as the stators are identical in any case. The stator may be wound for any even number of poles within reason, more poles meaning simply a larger speed reduction with respect to the driving alternator. With a given number of poles, speed changes can be made only by varying the frequency.
{figure}
FIG. 25
The synchronous motor, once speeded up by outside means, and its field excited by direct current through two slip rings, will deliver power at synchronous speed.
{figure}
FIG. 26
The squirrel-cage induction motor having a low-resistance, indestructible rotor, operates slightly below synchronous speed by an, amount called its " slip." This type of motor takes excessive starting currents and is suitable only for constant-speed work.
The phase-wound induction motor has slightly poorer full-speed characteristics than the squirrel-cage type; it is more difficult to build, but serves very well, when connected through slip rings to external resistances, for variable-speed work. In large sizes the rotor resistances become excessively large and waste much energy as heat.
The double squirrel-cage induction motor combines many of the good qualities of the two preceding types, and for ship propulsion is by far the best solution. In this type the rotor is of variable impedance, depending upon the slip, so that at starting when slip is 100 per cent it offers high impedance to rotor currents, and when up to speed (say, 5 per cent slip) its rotor conductors offer an easy path to the then low frequency currents. The power factor of the double squirrel-cage motor is low, but for ship work with no other loads on the alternators than the main motors this fact matters little.
CONCLUSION
There are, of course, as many advocates of the different means of battleship propulsion as there are types. As in correspondence filing systems, there are good points in all and bad points in all, and no one is perfect. There is usually one best method for any particular case, and trying to get all the good points and none of the bad is hopeless. So with the power for propelling battleships. Many means have been tried many more proposed, several of these, too, involving electricity in one form or another. Taken broadly, however, from a point of view of reliability, economy, minimum weight and space, ease of upkeep, and general ruggedness, it is believed the California will be in all respects a pacemaker among the battleships of the world.