Commander W. T. Sampson, Vice-President, in the chair.—The subject of the paper to be presented to the Institute this evening is " Electric Motors," and the lecturer, Mr. Frank J. Sprague, formerly an officer of the Navy and a graduate from the Naval Academy in 187S.
It is but a short time since Mr. Sprague was a student in this laboratory, and I recall with much satisfaction the interest he always manifested in the subjects taught in the department of Physics and Chemistry. He was always among the leading men in his class in all branches ; it was here that he passed his otherwise leisure hours and often his Saturday afternoons. Although his success in the scientific work which he has adopted as his profession is due to his ability and untiring energy, yet, no doubt, his first taste for it and his first successes in its pursuit were acquired here.
The same may, doubtless, be said of a number of other graduates of the Naval Academy who have already attained places of distinction in the scientific world.
I can imagine that it must be a source of gratification to Mr. Sprague to stand in this lecture hall and instruct those who were once his teachers. Though it is not among the substantial rewards of his success, it is well calculated to satisfy a commendable ambition.
Prof. N. M. Terry, head of department of Physics and Chemistry, will preside during the lecture and discussion.
Mr. Chairman and Members of the Institute:—In extending my thanks to you for assembling to hear some remarks on a subject of such new and growing interest as that of electric motors, I feel somewhat at a loss how to proceed, because the subject admits of such a variety of treatment. Feeling thus, I have come to the conclusion that perhaps I had better speak of the subject from the standpoint of my own personal experience and practice. I will try to set forth, at first, some of the simpler laws governing the construction of motors and operation of motors, then touch upon the general question of the transmission of power, and, finally, speak of the present and future development of this work.
If some of the truths seem too familiar, I beg you will bear in mind the absolute necessity of fully understanding their fundamental character, and the importance of always referring to them as a basis, if we would correctly explain the operation of a motor, or forecast its operation with any certainty.
In any part of an electric circuit the expenditure of energy is measured by the product of the current and the extreme difference of potential, or W= CE.
That is, the electrical efficiency of the circuit is the ratio of the motor electromotive force to the generator electromotive force, or of the counter to the initial.
All the energy which is delivered to the motor is expended in two ways—heat and work. Our object should be to have as little as possible in heat ; as much as possible in useful work. The current which goes into the field magnets serves to excite them, but the current thus used in maintaining the lines of force in the field magnets may be expressed simply as waste heat, and is measured by the quotient of the square of the electromotive force by the resistance of the field coils. The current that flows in the armature is expended partly in heat. This is measured by the product of the square of the current and the resistance of the armature ; hence the resistance of the armature should be as low as possible. All the rest of the energy of the current is expended in work. This work appears in three forms. One is driving the armature in the magnetic field and forming currents in the body of the armature, and will be greater or less according to the formation of that body. Another expenditure of this work is the overcoming of friction of the armature bearings. The third is useful work.
Two distinct theories have been advanced for the proper construction of motors. The one is advocated by Messrs. Ayrton and Perry, and is to the effect that different general principles should govern the construction of dynamos and motors : that in the former the field moment should greatly predominate over that of the armature, while in motors the moment of the armature should greatly predominate over that of the field.
On the other hand, Professor Rowland, Dr. Hopkinson, Messrs. Mordey and Kapp, and I, hold the opinion that the best apparatus for converting mechanical power into electricity must necessarily be the best apparatus for converting electricity into mechanical power, and that the principles governing the electrical construction of one must be also the best for constructing the other. In short, in motors as well as in dynamos, we should aim to have the most intense magnetic field possible, and the greatest ratio between the magnetic moment of the field and armature which the construction of the machine will permit. It is true that the great varieties of work demanded of motors make their construction often different from dynamos, but this does not affect the truth of the principle involved. So many practical as well as theoretical reasons exist for this view that I cannot see how any doubt can long exist of its soundness, and the fact exists that the most extensive and numerous developments of the transmission of power are made on these lines. Some of the reasons for adopting this view will appear as I proceed.
A motor when running may be considered as a dynamo machine propelled by a current : it has a field magnet like any other dynamo ; it has an armature situated in that field, which, either because of the attraction and repulsion of the lines of force, or of the double attraction and repulsion of the poles which are set up in the armature acting on the poles of the field magnet, is caused to rotate. This armature rotating in the magnetic field has an electromotive force developed in it which is precisely of the same kind as would be developed were the motor driven by a belt instead of by a current. The strength of this electromotive force depends upon the resulting strength of field and the speed of the armature. This electromotive force, which may be termed a motor electromotive force, is ordinarily called the counter electromotive force, because it is opposed to that of the line current which is flowing into the motor. The difference between this line electromotive force and that of the motor is termed the effective electromotive force, and determines, in combination with the resistance of the circuit, the strength of current which will flow in the circuit.
As I have stated, the province of the field magnet is to produce lines of force which shall pass through the armature. Since this is the fact, it is apparent that we wish to create as many lines of force as possible in the smallest space, with the least weight of metal and with the least expenditure of energy.
In considering, however, the way in which the lines of force can be created, we must take into account a great many elements. One is the character, mass, shape, and distribution of the iron which is used in the cores of the field magnets and in the pole pieces. Another is the mass, space, and distribution of the wire around the magnets. Not less important is the character of the armature body, the space taken up by the wire, and clearance outside of the wire. This magnetizing force is represented by the product of the turns of wire and the number of amperes flowing in these turns—in other words, by the ampere-turns ; and it is absolutely immaterial, so far as the magnetic result is concerned in the creation of the lines of force, how this product is made up—that is, whether there are a few turns and a great many amperes, or a great many turns and few amperes. For instance, if it takes ten thousand ampere-turns to magnetize a field magnet to a proper degree, we can magnetize it with one turn of wire and ten thousand amperes, or one ampere and ten thousand turns.
When we consider the economy, however, then we wish to have the greatest number of turns and the least number of amperes. In the case cited, if we used ten thousand amperes and one turn, we would be using ten thousand times the energy to magnetize the field magnets which would be expended were we to use ten thousand turns and one ampere.
That form of iron which gives the least length of wire for one turn is of course the circular, and that is the form which I have adopted in the field magnets of most of my machines. When exciting a field magnet from lines of given difference of potential, the ampere-turns which will flow is dependent upon the cross section of the wire alone, provided the mean length of turn is not changed. In other words, if the field magnet is wound with a certain gauge of wire it does not make any difference, so far as the strength of the field is concerned, whether there be simply one turn of the length of say two feet, or whether there be ten thousand turns each of the length of two feet, because just as we increase the number of turns we increase the resistance and we decrease the current, but we leave the product ampereturns, which we will call Z, the same. Economy, however, as I have said, determines that we shall use the largest number of turns which we can place practically in a given space on our field.
Having, then, determined the magnetizing force, we have next to consider what opposes the creation of the lines of force. In some very beautiful experiments made by Dr. John Hopkinson and published in a communication to the Royal Society of London, some elaborate tables were given, illustrated by a number of curves, which showed the magnetic property of about thirty-five specimens of iron and steel, and the strength of field magnets for different magnetizing forces. It was there shown that the best material for commercial purposes to be used in field magnets was pure wrought iron annealed, because it offered less opposition to the creation of lines of force than any other material. This had been practically accepted as a fact before the publication of these experiments, but I know of none which are entitled to more consideration. The introduction of manganese, sulphur, silicon, phosphorus, or carbon means very radical changes in the qualities of the iron, more so than has generally been realized.
We may make some analogies between magnetic and electric circuits. The magnetizing force or the ampere-turns may be likened to the electromotive force which urges a current over a circuit. As in an electric circuit the current generated depends upon the resistance of the circuit as well as the electromotive force, so too in the magnetic circuit may we very properly consider that the formation of lines of force actually created depends also upon the resistance of the path. The resistance of this path depends upon not only the cross section of any space over which the lines of force have to pass, whether in the metal or in the air, but also upon the specific resistance per unit cross section of the material over which they pass, this being different for cast or wrought iron, and being very much higher than either for air. It also depends upon the length of the space. We would then suppose that the number of lines of force which are created would be the quotient of the magnetizing force by the resistance; but when we come to magnetize a mass of iron, we find that a totally different element comes into play, and this seems to be a force which we may liken, if you please, to a counter electromotive force in an electric circuit—that is, it is an opposition set up to the formation of lines of force, which increases with the degree of magnetization, and finally reaches such a point that it effectually bars any further increase of this magnetization. In other words, if we plot out, as was first done by Mr. Hopkinson, the characteristic of a field magnet, laying off as ordinates the number of lines of force created, and as abscissae the magnetizing forces, we have a curve which starts almost as a straight line from the origin at an angle to both the axes of x and y, but soon begins to curve and finally becomes almost parallel to the axis of X, or so lightly inclined to it that it becomes a straight line which would intercept the axis of j; at a long distance to the left of the origin. This has been explained by considering the curve to be formed of two components, one expressing the lines of force due to the presence of the iron, and the other additional lines offered set up by the coils around the iron ; that the first part of the characteristic curve is formed by the sum of the ordinates of both, and the latter part of it simply by a slowly increasing ordinate expressing magnetization due to the coils around the iron.
I do not feel prepared here to say whether this theory is right or not. If the iron can be brought up to a certain saturation and not be increased any more, then it seems that the air surrounding the conductor may be brought up to a degree of saturation and no further saturation possible. If this is so, it follows that we would finally reach a point in the characteristic curve of either a mass of iron or coils of wire, or these combined, when it would become tangent to the axis of x, and no further increase of magnetization would be possible under any circumstances. Physical conditions, however, prevent our reaching this point, because before we would get to that we would burn the coils off our field magnet because of the great heat developed. It follows that with every different shape of field magnet, quality of iron, and distribution of mass, that an empirical formula must be derived to truly express the conditions of magnetization.
To express a value for the number of lines of force, we must in addition consider the inter-polar space. This is occupied by the mass of iron in the body of the armature, which ought to have at least the cross section of iron that is present in the core pieces of the magnet ; and it is further important that the space taken by the coils of wire on the field magnet should be just as small as possible. In other words, the gap or space between the pole faces and the body of iron of the armature should be as short as possible because of the very high specific resistance of air.
Mr. Siebert Kapp, in a very excellent book on the electric transmission of energy, has given some formulas for dynamos and motors of wrought and cast iron, and of different forms. But, as he has pointed out, they apply only to low degrees of magnetization, and of course depend upon the quality of iron used.
Although the quality and dimensions of the iron masses in a motor are important, it will be found in practical experience that the resistance of the gap is a no less important feature, and I will cite an illustration showing the important bearing which it has on the construction of a machine. As I have stated, I differ radically from Messrs. Ayrton and Perry in the theory which should govern the construction of a motor. I have said the field magnet ought to be as intense as possible; that is, the greatest number of lines of force should be created that it is in our power economically to produce. The resistance of the armature should be as low as possible, and the magnetization of the armature should also be as low relatively to that of the field magnet as possible. Suppose that a machine has wound on its armature four layers of wire, and that it has a further space for clearance equal in depth to the thickness of one of these wires. The total gap then may be said to be 5a, a being the thickness of one wire. The magnetic moment of this armature with any given radiation of heat per square inch of surface is fixed, no matter how it may be wound, and no matter what its power. This magnetic moment is, of course, the product of the amperes by the turns of wire. Suppose, for example, this armature be stripped and then wound with one-half the turns of wire, the wire being of double the cross section of that first used. We will assume that the insulation always bears the same ratio to the cross section. The armature would now have one-fourth the resistance, one-half the number of turns, it would carry double the current, the heat radiated per square inch of surface would be the same, the speed would be doubled, the power would be doubled, but the ampere turns would be a constant ; and so on in other proportions. Suppose, on the other hand, that the armature be intended for a certain horse power, and have a certain radiating surface, it would, with a given efficiency, have a certain resistance. With this fixed resistance, the cross section of the wire would be in direct proportion to the number of turns. Instead of using eight turns of a certain cross section, laid four deep, suppose we use four turns, keeping the resistance the same. The cross section of the wire would be one-half. The weight of the wire would be one-fourth. The room occupied by the wire would be one-fourth. The capacity of the machine would be the same. The heat radiated with any given current would be the same. The gap would be reduced in the proportion of 5 to 2, assuming that we leave the same clearance. Since, however, we use a lighter wire, we can use a smaller gap, because, having less weight, the wires will have less centrifugal force and a finer binding wire can be used. We have then this new condition of affairs : the resistance of the armature has remained constant; the heat-radiating surface constant; the capacity of the armature constant ; the weight of the copper in the armature has been reduced 75 per cent; the number of turns has been halved ; the gap has been reduced 60 per cent. If, now, our field magnet was not nearly saturated, the number of lines of force which appear will be about doubled, because of the reduction of the gap; and since the lines of force have been doubled, the counter electromotive force which is set up by the armature at the same speed will be the same, although there are only one-half the turns of wire that we had in the first place. The magnetic moment of the armature will be one-half what it was. The strength of the field being doubled, it follows that the ratio between the magnetic moment of the field and of the armature will have been quadrupled. There will be less distortion of the lines of force, there will be less sparking at the commutator brushes, there will be less change of lead at the commutator brushes. In addition to this, we have double the strength of field without increasing the expenditure of energy in the field magnet ; in short, we see that one of our main objects should be to make with the most intense magnet field the smallest possible gap, taking care, however, that the increase of lines of force which we desire will be obtained by the proper proportion of the gap and the energy expended in the field magnet coils.
One of the most instructive experiments, showing the value and importance of masses of iron in both the field and armature of a motor, is the following : Arrange the motor with a balanced lever of known length attached to the armature shaft, with accurate means of reading the pull in a direction strictly tangent to the circumference of the circle which would be described by the end of the lever. With fixed strengths of current in the field, vary the current in the armature by any of the ordinary controllable means, and take simultaneous readings of the pull. This will give a curve showing the characteristic of the armature, which is just as important as that of the field. Then, with known currents in the armature, vary the current in the field, and likewise note the current and torsion readings. These results will give a saturation curve of the field, and an investigation of the data will give some interesting facts about your machine.
Several years ago Professor Moses G. Farmer suggested that I would find it an interesting investigation to determine a characteristic of the effective strength of field magnets by determining the number of feet per second a conductor one foot long would have to move to develop an electromotive force of one volt at its terminals when moving in the active part of the field.
Motors may be described as belonging to one of three different systems. First, those in which the field magnet is excited by a coil in parallel circuit with the armature; second, those in which the field magnet is in series with the armature circuit ; and third, those in which there is a combination of these two circuits. There are in addition a very large variety of each of these classes, different conditions demanding different performance. Furthermore, similar machines may be placed upon three different kinds of circuits, their performances varying widely in each case. These three conditions are: First, the case of special transmission with varying potential and current; second, constant current circuits; and third, constant potential circuits. I will now briefly consider the action of these different kinds of machines on two classes of circuits : First, on the constant current circuit. If a series-wound machine be placed upon such a circuit, the same current passing through the field magnet, it will develop a constant torque, which torque is directly proportional to the strength of the field magnet and the armature. If the masses of iron are large, this torque will be directly proportional to the effective ampere-turns in the field magnet, and the work done will be direcdy proportional to the speed. If the machine be at rest there will exist a difference of potential at the terminals of the machine equal to the product of the current and the resistance of the machine. When running, however, an electromotive force will be developed in the machine, and the potential at the terminals of the machine will rise by this increment. The work done may be expressed by the product of this counter electromotive force and the current, or eC, and is independent of the resistance of the machine. The resistance, however, determines, in combination with the other elements, the total efficiency of the motor.
When running at any particular speed, the work will be increased directly as the field magnet strength is increased. So also will be the economy. The heat waste with any given resistance in a machine under these conditions is constant. The direction of rotation of such a machine can be reversed by reversing either the armature circuit or the field circuit ; if both circuits are reversed, then the machine will run in the same direction. For many classes of work this kind of machine is exceedingly useful, because it admits of a great range of hand control. If such a machine, however, be put on ordinary work, and this work be lightened up, the machine will run faster and faster, and unless the field is weakened or the brushes are shifted to check it, the speed will theoretically increase without limit. Every change of speed and every change of load is accompanied by a corresponding change in the potential which exists at the terminals of the machine. Moreover, on a constant current, the motors being in series with each other and with lamps, this continual variation of potential is apt to cause trouble on the circuit, especially if the machines are not automatic, since, as already stated, with any fixed field the torque is constant, the work done is directly proportional to the speed. The machine has the highest efficiency when running at the highest speed.
With shunt machines, however, the action on the constant current circuit is much different. Here the current is divided in two circuits, such division, when the motor is at rest, being inversely proportional to the resistances of the two parts of the circuit. With such a motor, the field is weakest when the machine is at rest, and its torque or rotary effort is also very weak. As the speed of the machine increases a counter electromotive force is set up, the potential at the terminals of the armature and field magnets rises, the current in the armature diminishes, and that in the field magnet increases.
There are two ways by means of which a constant current motor can be governed. One consists in automatically changing the counter electromotive force by changing the position of the brushes on the commutator to positions more or less removed from their normal one. To this objection is offered because the proper position for the brushes of any machine is at the points of least sparking. The other method consists in varying the counter electromotive force by automatically weakening the field as the load is diminished, or strengthening it as the load is increased. Several methods have been proposed for doing this, generally by the action of a centrifugal governor, but I am now engaged on another system which promises better results.
On constant potential circuits the different classes of motors present other peculiarities. A plain series-wound motor, when there is sufficient iron in the field, has a torque proportional to the square of the current flowing through it. It is capable of exerting a great rotary effort and doing a large amount of work at a slow speed. The range of speed for different loads is, however, great, and the motor is entirely unfitted for ordinary work where steadiness of speed is an object; as the load is diminished the speed increases, and, if thrown off entirely, the motor will run faster and faster, the field continually growing weaker, and the armature all the time accelerating its speed in a vain attempt to generate an electromotive force equal to the initial.
A series motor has some excellent characteristics for work where great changes in speed and torque are desired automatically, these changes being varied inversely.
Since with any given strength of field the torque varies with the increase in the strength of the armature field, and with any given strength of armature the torque varies with the increase in field strength, it follows that with large masses of iron in the armature and field, the torque will vary with the square of the current, and the ratio between the field and armature moment will remain constant.
The following will show the law of variation of speed for a series motor within the limits of the straight line saturation on a constant potential circuit.
On the other hand, the shunt-wound machine will run fairly well on a constant potential circuit. The field, being excited independently of the armature, is constant, and since the load varies with the motor electromotive force, and the field is constant, it follows that the speed must vary with e. The torque is proportional to the current in the armature, and the speed will be slowest with the greatest load and fastest with the lightest, that is, when e=iE. The lower the resistance of the armature, the less the variation in speed, and if sufficiently low with a large ratio of magnetic moments, the motor will practically run at a constant speed.
It is with the third class of motors, when used on constant potential circuits, that the difficulties which are involved in the governing of a motor mostly disappear, and, without the use of any such apparatus as centrifugal governors or movable contacts, it becomes possible to satisfy the most exacting conditions, both as regards efficiency, steadiness of running, power to start under very heavy loads, and freedom from sparking.
I must now particularly request your attention to a seemingly paradoxical statement. In a motor with the armature and field magnet independently supplied, the work which the motor will do in a given time, its economy and efficiency, are all independent of the strength of the field magnet, provided the translating devices intermediate between the motor and whatever is the recipient of its motion are not limited as to the rate of transmission of the motor speed; and that in all cases where a motor is working on a constant potential circuit and not up to its maximum capacity, but still above fifty per cent armature efficiency, in order to increase the mechanical effect either of speed or power, or both, or to compensate for any falling off of the potential on a line, it is necessary to weaken the field magnets, instead of strengthening them, and vice versa.
The strength of the field determines the speed at which a motor must run to get a required efficiency. With a given initial potential at the armature terminals, no matter how the load varies from the maximum allowed, the speed may be maintained constant by changing the strength of the field ; such strength being diminished as the load is increased, and, vice versa, increased as the load is diminished.
These facts may be demonstrated as follows:
Let us consider the motor current as derived from mains having a fixed difference of potential, and the motor with its field and armature in shunt relation. In this case the armature runs with a velocity dependent upon the strength of field, the initial potential, the number of turns, resistance, etc., of the armature, and the load, and a counter electromotive force is set up which regulates the armature current. The higher the speed the greater this counter electromotive force. Let E be the initial and e the counter electromotive force, and r the resistance of the armature.
To maintain speed or power constant under varying initial potential, if the potential at the motor terminals increases, these mechanical effects increase or tend to increase. By strengthening the field an increased counter electromotive force is produced, so that the increased power or speed, or the tendency thereto, is counteracted, and this counteraction may evidently be itself considered a decrease in mechanical effect, whether the regulation is performed simultaneously with the increase of potential or before or after such increase. If the regulation is performed simultaneously, with a gradual change of potential, there may be less change in counter electromotive force or armature current ; but there is still the counteracting of the tendency to increased mechanical effect, which counteracting is itself a decrease of mechanical effect. For a decreased or decreasing initial potential, the field is weakened to counteract the decrease in mechanical effect which would otherwise occur, and therefore to produce an increased mechanical effect.
Hence, to change the speed or power of a motor on a circuit of constant potential, the speed or power is increased by weakening the field, which produces a decreased counter electromotive force and an increased armature current, and consequently the increased mechanical effect desired ; and such mechanical effect is decreased by strengthening the field, and thus increasing the counter electromotive force.
In brief, then, this method of regulating shunt-wound motors consists in strengthening the magnetizing effect of the field magnet coils of the motor to decrease the mechanical effects, such as speed or power, or both, and vice versa, weakening such magnetizing effect to increase the mechanical effects, and under varying loads the speed is maintained constant by an inverse varying of the strength of the field magnet strength.
This may be accomplished in several ways, one by varying the field circuits by a mechanical governor which responds to any variation in the speed of the motor, or in hand-controlled devices. For automatic work, however, I prefer to make use of certain coils in series with the armature and dependent upon it, which coils have a resultant magnet action which is opposed to that of the main coils of the machine. While the main principle is the same, I have a number of different ways of applying it which may be classed as the long shunt, the short shunt, the distorted differential with long and short and compound shunts, and the distorted cumulative-differential with long and short and compound shunts. The first of these was originally proposed by Professor Ayrton, but the proportion of his winding was not the same as my own, because determined experimentally and with field magnets whose magnetic moment was less than that of the armature, while in my motors the converse is true.
In the following demonstration, then, it is to be assumed that the field moment greatly predominates over that of the armature, and also that the characteristic is practically a straight line.
Let f denote the resistance of the main or shunt field coils ; m the number of turns therein ; r the resistance of the differential or series field coils, and n the number of turns ; E the difference of potential at the shunt terminals ; e the counter electromotive force set up in the armature ; and R the resistance of the armature.
In starting, then, it becomes necessary to cut out this governing coil, making the motor a pure shunt motor, or to reverse it and make it a cumulative motor. I use both methods.
What has already been pointed out may be again stated, that the motor will regulate itself perfectly for all potentials so long as we work with a straight line characteristic, but it must be with a theoretical armature efficiency of not less than fifty per cent, for if we go below this, the governing coil works in the wrong direction.
A better way of arriving at this same law is as follows:
When running absolutely without load the motor e. m. f. is E, and the field is that due to the shunt coil. When running with a load the motor e. m. f is e, and the magnetic field that due to the difference of the shunt and series field moments.
From the foregoing demonstrations it follows that a victor of either class depending for its regulation on this differential winding will regulate with a constant current only when working at less than fifty per cent armature efficiency ; and that the same machine with the same winding will regulate on a constant potential circuit only when working at over fifty per cent armature efficiency.
The laws above set forth are for pure electro- dynamic motors ; if there is any permanent magnetism, as in hard cast iron, or where permanent steel magnets are used, the law of winding is modified in so far as the residual or permanent magnetism is the equivalent of an electro-magnetic moment ; but in this case, too, there should exist a zero field if the governing coil is normally closed when the motor is at rest.
It should be remarked here that in practice the quality of commercial iron makes some departure from theoretical laws necessary.
As an instance of the effect of reversing the governing coil in starting, if a constant potential motor has the series coil reversed when the full circuit is dosed, there being margin enough on the field characteristic, we would have a field twice as strong as the strongest normal field, four times the strength when the motor is doing its maximum work per unit of time, and a momentary rotary effort eight times that existing when the maximum work is on. As soon as the speed comes up, if the governing coil is short circuited and then reversed, the motor will be self-regulating.
On some machines I take another step in overcoming, or rather counteracting, the distortion set up by the armature, by producing a distortion in the field magnets which is dependent on precisely the same current that flowed through the armature. I will describe one method only.
Main field-magnet coils are employed in shunt relation to the armature, differential field-magnet coils in series with the armature, and additional cumulative field-magnet coils, also in series with the armature. The main field-coils may be shunted upon the armature alone, or upon the armature and both the cumulative and differential series coils, or upon the armature or either of the series coils, the other series coil remaining outside the terminal of the main field shunt.
The object sought is to maintain the non-sparking points of the commutator cylinder constant by opposing the distortion of the magnetic field due to variations in the armature current by a counter distortion dependent upon such variations, whereby the magnetic resultant due to the armature and field magnet is unchanged, and the line of parallel cutting of the lines of force or point of least sparking is maintained in the same position.
In accomplishing the counter distortion of the field, the motor used is one in which the field-magnet cores extend in different directions from the field of force in which the armature revolves. The differential series coils are wound or arranged so that their greatest effect is produced on diagonally opposite parts of the magnetic field, and the cumulative series coils, so that their greatest effect is produced on the other diagonally opposite parts. The differential coils are arranged to have a greater magnetizing effect than the cumulative coils. A decrease of load, causing a decreased armature current, tends to shift the magnetic resultant of the armature and field magnet ; but this also decreases the magnetizing effect of all the series coils, and, therefore, the parts of the field principally affected by the cumulative coils are weakened, and those principally affected by the differential coils are strengthened, whereby a distortion of field is produced opposed to that produced by the decrease of armature current, and hence the magnetic resultant—the line of parallel cutting and the points of least sparking—remains unchanged. Thus no shifting of the commutator brushes is ever required, except on account of wear.
I will now consider, from both the technical and commercial standpoints, the different classes of circuits on which motors may be used.
First. Arc light, or constant current circuits, in which the current supplied to the motor is kept constant at a certain number of amperes, ranging from six to nineteen amperes in different systems. The electromotive force at the terminals of the motor varies with the load.
Second. Constant potential, or incandescent light circuits, in which the difference of potential at the terminals of the machines is kept practically constant while the current varies with the load.
Third. Circuits in which the current and the potential both vary, as is the case where there is an appreciable drop or fall of potential on connecting lines somewhat removed from the source of power of a constant potential system, and in cases of special transmission.
I am now operating on all classes of these circuits, but since, because of the small ampere capacity of the current on arc light circuits, any large power must require a great difference of potential at the motor terminals, and variations of power will cause sudden and great changes of potential, the arc light circuits are useful in conjunction with arc lights for transmission of small powers only, or for constant work.
In considering the transmission of power from a general central station as an industry, that is, in a broad and comprehensive way, and not as an adjunct to some system of lighting, and leaving out of consideration for the present special cases of transmission, which forms a class of work by itself, practical and theoretical considerations make it imperative that the constant potential method of distribution is the only safe and feasible one.
It must be borne in mind that I do not agree with those writers who hold that central station distribution in limited areas is confined to " small domestic industries." We have then to deal with large and small powers, varying and constant loads, and the necessity of absolute independence both in operation and regulation of each motor. Furthermore, the motors may be designed for different classes of work, some constant in speed, others variable, and so on.
The only existing constant current circuits are used primarily for the operation of arc lights. At present, as I have said, these range in capacity from nineteen to six amperes. The unit of light which is required for general purposes necessitates an expenditure of about half a horse power of electrical energy. A greater expenditure would be extravagant. In order to keep the size of the conductors as small as possible, and to allow long extended circuits, the tendency is to reduce the amperes to the smallest number.
The conditions of an arc light probably will not allow this to go below about six amperes ; the more ordinary unit is about nine and a half The commercial conditions unquestionably will not permit in the future a much higher ampere capacity, because the size of the wire varies as the square of the current used. With a nine and a half ampere current, a motor to develop one horse power—supposing it to have an efficiency of eighty per cent—must have supplied to its terminals about 933 watts of electrical energy. In other words, there would be at the terminals of the machine an electromotive force of about 98 volts for each horse power developed. Suppose, now, we want to transmit one hundred horse power; it would then be necessary (if the motor be of the same efficiency) to supply to it 93,250 watts of electrical energy ; and if it were on this gk ampere circuit we would have an electromotive force of over 9800 volts.
In practice, arc light circuits vary from, say, 1000 to 2500 volts. On a 1000 volt circuit about ten horse power could be recovered ; on a 2500 volt circuit about twenty-five horse power. In other words, existing arc light machines, if devoted entirely to the transmission of power, are limited to the actual development, on even the largest machines in ordinary use, of twenty-five horse power; and if used in combination with light, there would be available on any particular circuit only a fraction of this. If we are going to deal with the transmission of the tenth of a horse power, or one-half, or even two or three horse power, and are willing to have the element of danger as well enter into these small transmissions, then we can work with an arc light circuit ; but if we are going to transmit units of five, ten, fifteen, twenty, or twenty-five horse power, we cannot deal with the arc light circuit ; it is utterly impossible.
In Boston we have recently put upon certain lines nearly two hundred horse power. Suppose this were supplied on gi ampere circuits. No less than eight circuits of 2500 volts each would be required, and probably more, because with units as high as fifteen horse power and this division of circuits, the law of general average could come into play in but a very limited manner.
These eight 2500-voh circuits represent the capacity of a 460 arc lamp station, and to deliver the power not a single lamp could be used at the same time. The constant current method of distributing power is the limited and unnatural method ; the constant potential, the comprehensive and natural method. This is a fact entirely independent of the question of relative electrical potentials, because on the constant potential circuit we can work at 100, 1000, or 5000 volts, if we please. In short, electrical distribution follows much the same laws that hold in gas, water, and steam distribution.
One of the most important applications of the principle of reversibility which I pointed out as characterizing dynamos and motors is its possible use in braking trains. The electromotive force depends upon the strength of the field and the velocity of the armature, and is independent of everything else. It is evident, therefore, that if the strength of the field magnet is increased, the motor electromotive force is also increased ; and if the increase of field strength is continued, the initial and counter electromotive forces will become equal, and then the counter or motor electromotive force will predominate. The propelling motor will now become a generator and give current to the line, and its mechanical effects are reversed, so that it brakes the train instead of propelling it, and the current generated by it and the braking power or reversed mechanical effect are now controllable by further increasing or by rediminishing the strength of field.
It will be seen that mechanical energy is received by the reversed motor according to the mass of the train and its velocity. If a train should start on a down grade unprovided with a brake, the energy of falling would tend to increase its speed ; but when this method of braking is used, this mechanical energy is transformed in the machine into electrical energy delivered to the line, and augmenting that supplied from the generating stations to the other trains, which may be moving upon up-grades or on levels.
When it is desired to slow down a train on a level grade, the field is increased, as before, until the counter electromotive force predominates over the initial, and the energy stored up in the moving train is exerted to run the machine as a braking dynamo. As the train slows, however, the diminution of speed of the armature will tend to diminish the counter electromotive force, and the increase of field strength must therefore be continued, so as to still maintain the counter electromotive force above the initial and keep the machine running as a generator as long as practicable, when other methods of braking, as, for example, closing the armature circuit on a local regulating resistance or brake circuit, may be used.
I will give an instance to show how effectively this method may be employed, premising that when large masses of iron are used in the field magnets, the strength of the field can be varied within effective limits two or three hundred per cent, and also that a well constructed armature can carry for a short time fifty, seventy-five, or perhaps even one hundred per cent more current than it can stand for any long run. Suppose the armature of a motor to have a resistance of three-tenths of an ohm, with field coils in shunt relation, and provided with suitable means for varying their strength. Suppose the initial electromotive force to be five hundred volts, and forty horse power to be required from the motor when running at its maximum. Allowing for losses in conversion, this forty horse power would be about thirty-two thousand watts. The counter electromotive force would be four hundred and eighty volts, the effective electromotive force twenty volts, and the current sixty-seven amperes. The electrical efficiency of the armature would be ninety-six per cent. Suppose the strength of the field to be increased about four per cent, the speed remaining the same, the motor running on a down grade; the counter electromotive will be increased to five hundred volts, and the motor armature will then be perfectly passive electrically, neither taking from nor giving to the line. Let the field strength be increased again one per cent, and let the increase be continued in the same ratio. The result is shown in the following table :
Total field increase. Current 10 line. Approximate energy required from train, allowing for losses.
5 per cent. 13.3 amperes. 9.5 horse power.
6 29.3 20.9
7 45.3 32.7
8 61.3 44.7
9 77.3 56.9
10 93.3 69.3
From the above it will be seen that by simply increasing the field strength one twenty-fifth part, the machine is converted from a motor driving a train with forty horse power of effective work, to a perfectly passive machine, allowing the train to run absolutely free. Then by increasing the field one one-hundredth part the motor at once exerts a positive braking force, and on an increase of about eight and one-half per cent above its original strength it will give back to the line current equal to that which was originally taken from it, sufficient, evidently, to run some other motor of the system which may at that time require that amount of current ; and by increasing the original field ten per cent the machine acts as a dynamo, requiring more than fifty per cent more energy than is demanded to run it as a motor developing forty horse power.
This method of braking, it will be seen, is under perfect control, and it is the most economical system possible for an electric railway, since whenever a train descends a grade, and whenever a train stops, the energy stored up in the moving train is delivered in the form of electrical energy upon the line.
Consider for a moment what occurs on a steam railway train making frequent stops at short intervals. Suppose it to be ascending a grade, then the engine exerts more than its average amount of power. When it reaches the top and enters upon a down grade, steam is shut off, and when the train begins to run faster than is desired, the brakes are put on, and the energy of the train is then converted into heat on the rims of the wheels and on the brakeshoes. In other words, all the energy in excess of that necessary to run the train on a level which has been required to climb the grade is now thrown away in going down the grade, instead of being utilized, as in this system ; and, furthermore, additional steam is actually required to check the tendency to augmented speed. Then when the train approaches a stopping place, or whenever it is necessary to slow the train down quickly, steam has to be employed to actuate the brakes, and all this additional power is simply thrown away. By the electrical method of working, the greater part of this loss may be entirely obviated.
It is true that not all of the energy will be converted into electricity, but a large proportion of it will be.
On a double-track road, with both tracks supplied from the same main circuit, the energy given to one track is also communicated to the other. The up-grades on one track being always balanced by the down-grades on the other track, it is evident that the total up-grade of the whole system is equal to the total down-grade thereof. Therefore, energy being expended on the up-grades, and given out to nearly the same extent on the down-grades, the energy required in the system is that sufficient to move a train upon a level with a slight percentage added ; but on a steam railway the energy required is not only that sufficient to run a train on a level, but, in addition, that necessary to raise it from the lower to the higher grades on both tracks, no matter how many of such grades there may be.
Suppose the Third Avenue " L" road of New York were equipped with electric motors, let us consider what the application of this method of braking would mean. There were not long ago, at about six o'clock in the evening, sixty-three trains in operation at one time on about eight miles of double track. The maximum capacity of the engines is about one hundred and eighty horse power, giving an aggregate of 1 1,340 horses. The work, however, is continually varying, because the trains have to make stations about one-third of a mile apart in eighty seconds, and the average work done is only about seventy-one horses, or 4473 horses total.
Yet the wasteful character of the expenditures shown by the fact that fifty-nine per cent is expended in overcoming the inertia of the train in getting underway, twenty-four per cent in lifting it on grades, and seventeen per cent only in traction. If the 4473 horse power which represents the average were supplied by electricity in the ordinarily proposed methods, with a recovery of say sixty per cent, then there would be required not less than 7455 horse power at the central stations. By using the system of braking I have described on down grades and in stopping, enough of the energy of the train can be reconverted into electricity to entirely make up all the losses of conversion, transmission and reconversion, so that at the central station only about 4473 horse power would be required, a saving in investment, and coal consumption of 7455— 4473 = 2982 horse power.
Instead of the current being all supplied by the main generating station at one or more points, it would be supplied from nearly as many additional moving stations distributed along the whole line, as there are trains slowing down or running on down grades. With any given size of conductor, the loss would be much less, and, with the same losses, the sizes of conductors could be very much reduced.
I have thus shown how, in the application of motors to both stationary and railroad work, the control of the motor electromotive force is the key to the control of the power and speed. I will give one or two more illustrations of the practical character of this theory and how it may be utilized. In our factory one of the methods that we have of testing motors is the following. The power which we have is limited. We sometimes wish to test a number of motors together, and in doing so to develop an amount of power which it is impossible for us to spare from our shafting. Suppose, for example, that we wish to test two 20 horse power and two 7 ½ horse power motors, an aggregate of about 55 horse power. One of these 7 ½ horse power motors is driven as a dynamo. This is electrically connected to the second one, which is driven from the first as a motor. Overhead, we have a line of counter-shafting, on which are three pulleys. Our two 20 horse power machines are belted up to these pulleys, and likewise the 7 ½ horse power motor. The like terminals of the large motors are then connected together, and an ampere meter put into one branch of the circuit. One terminal of each field is likewise connected to its proper line, and the other terminals of these fields are brought to the movable levers of a three-part circuit changing switch. One contact is carried to the other line, and between it and the other two is inserted a variable resistance, which, in the middle position of the switch, is short circuited. We have, then, the two large motors connected in an electro-mechanical couple, and to the same mechanical shafting is connected a third motor. The dynamo being started, the switch set in the middle position, the motor is speeded up, which sets the countershafting in operation, and drives both the 20 horse power motors as dynamos, each exciting its own field. If the machines are symmetrical, no current whatsoever will pass over the branch connecting the two they are simply in the position of two dynamos in parallel circuit with each other, with no external circuit and no path over which the current can flow, except that through their field magnets; consequently, very little power, save that of friction, is taken. The switch being moved in one direction, the resistance is thrown into the field of one machine. The electromotive force which it develops at this particular speed is now reduced; it becomes a motor, and current will flow over the connecting mains from the other machine, which is now a dynamo, which current is expressed by the quotient of the difference of the electromotive force of the two machines by the resistance of the two armatures and the connecting mains. By varying this resistance, this current can be made of any value up to the limit. We have here, then, one machine acting as a motor, and driving on to the countershafting with a certain number of horse power, this countershafting driving the other machine as a dynamo with a certain greater amount of horse power, this second machine furnishing the current which operates the first as a motor. The deficit, or loss of efficiency between the two machines, is made up by the third 7 1/2 horse power motor ; in other words, one machine is driving the other electrically as a motor, and the motor through the countershafting is driving the first one mechanically as a dynamo, the deficiency being made up as I have just said. By reversing the switch, the resistance is first cut out of the field, and then thrown into the field of the oppositemachine. This machine now becomes a motor, and the other machine becomes a dynamo. This reversal is not instantaneous, because it takes time for the field magnets to charge and discharge. The ampere meter will drop to zero, and will then rise again progressively.
This method of testing can be used for two purposes, one for testing the actual horse power developed and the couple efficiency, which can be done by measuring the current, the electromotive force between the machines and the horse power delivered to the shafting by the third motor, and for the other purpose of testing simply the heating capacity of the armature coils with a given number of amperes. For this latter purpose it does not matter practically whether the machines are run at their normal speed and generate their normal electromotive force, or whether some lower electromotive force is used. If a lower electromotive force is present, it simply means that there must be a greater ratio of difference between the field magnet strengths and a larger resistance used with the reversing switch.
This method of testing, where there can be an instantaneous and controllable reversibility of the dynamo and motor by the mere touch of the finger, is probably one of the most beautiful, as it is one of the most useful, applications of this characteristic of machines. By it we can test motors of a capacity of, say, fifty-five horse power, with a belt expenditure of about eight or ten horse power.
A similar method of testing can be used in testing the dynamos employed on alternating circuits. It is now customary in testing large machines, say one thousand lights capacity, to use banks of lamps or artificial resistances to test their capacities. Mr. Hopkinson pointed out, some five or six years ago, that two alternating-current machines could be operated in multiple circuit, but could not be operated in series, and one of the practical things which have been developed is this : that if two alternating machines are run in multiple circuit, before they can be thrown together in safety that they have to be brought into unison. It must be remembered that in an alternating-current machine the variation of electromotive force instead of being two or three, or four or five per cent, as with a constant-current machine, is of the greatest degree. With a one thousand-volt machine, the electromotive force is first as one thousand volts positive and then one thousand volts negative, and this may occur a hundred times a second. If two machines which are not in unison are thrown together on the same circuit, one is retarded and the other is accelerated, the one acting as a dynamo and the other as a motor, and in the twinkling of an eye they are brought into unison with a report and a shock so sharp that it has for the uninitiated a startling effect, and is, of course, a very severe wrench upon the machines. It is common to excite the field magnets of alternating current machines by separate exciters. If they run at the same speed, have the same characteristics and have field magnets of the same strength, then when they are in unison both machines will act as dynamos, and neither will take or give current to or from the other. If they are out of unison, then because of the rapidity of the pulsations of these phases and the great change of electromotive forces in the machines they will necessarily come together. One cannot be run as a motor and the other as a dynamo practically by maintaining them out of unison.
Let the two machines be, however, started together with zero fields, and when running at their normal speed let the field magnets be gradually strengthened. The machines will come to unison without trouble. If now, when thus running in unison, the exciting circuit be arranged, as is very easy, so that the field magnets of either can be weakened, then the phase curve of one may be made lower than that of the other by any required degree. The pulsations will be the same in number, they will change in character at the same instant and in the same way, but since there is a difference of electromotive force, the one developing, say, five per cent more than the other, then one will be running as a motor and the other as a dynamo. Both can be driven off the same line of countershafting or engine, and two one thousand-light machines can be tested to their full current-carrying capacity with a very small fractional part of the power which they each separately require, and their dynamo and motor phases can be reversed with the same rapidity and with the same ease and safety as I have described with the constant current machines.
You have been kind enough to express a wish to hear something of the commercial development that has taken place, and I will give you briefly some facts showing how entirely beyond the visionary realm the transmission of power has gone. Some three years ago, the first electric motors which I put into practical use in territory outside of my own jurisdiction were sent to Lawrence, Massachusetts, and one of these was used in an isolated plant for operating a cotton elevator in the Pemberton Mills. Since that time the work has gone forward steadily, at first slowly, and of late with a rapidity which has exceeded my most sanguine expectations.
The first companies on whose circuits what are called the constant speed motors were first introduced were the circuits of the Edison Electric Illuminating Companies, and now a very large number of these companies, as well as many others, are coming to realize that one of the most important matters to which they can give their attention is the development of the use of motors for all industrial purposes within the range of the territory covered by their conductors. The result has been that these machines are now being introduced in the United States in the cities of New York, Chicago, Boston, Des Moines, Elgin, Oskaloosa, Pittsburgh, Chester, Williamsport, Lancaster, Shamokin, York, Detroit, Topeka, Hutchinson, New Orleans, Cleveland, Cincinnati, Springfield, New Brunswick, Fall River, New Bedford, Milford, Taunton, Lawrence, Woonsocket, Fort Meyer, Waterbury, Annapolis, St. Louis, Abilene, Pawtucket, Syracuse, and also into Canada, the Argentine Republic, Austria, Germany, Italy, and Japan, and they are being applied to every possible use. In Boston there have been introduced within the past few months nearly sixty of these machines on one circuit, varying from one-half to fifteen horse power in capacity. In New York there are about sixty machines in operation, and the number is now rapidly increasing.
The position I early took in this matter was briefly this : that electricity could fill the field in a far more complete and satisfactory manner than, but in very much the same general way, as the distribution of power from a central station by means of gas, steam, water, or compressed air. The use of any of these elements involves engineering problems. To be successfully solved, not merely as a scientific fact, but in a commercial sense, every part of such system should be in thorough accord with every other part. It is not sufficient that there should be a good dynamo, or a good system of distribution, or a good motor, but that they should be definitely related, the units, the distribution, the area of the distribution, the sizes of the conductors, and the electromotive forces being determined with reference to each other, and to the ultimate demands to be made upon the central station. Nor could any elements of such a system be in doubt. It goes almost without saying that large engines can beoperated and large powers developed under as little general supervision as smaller ones, and under far less supervision than several units of small powers. The cost of the generation of power is much less in large engines, both in the amount of coal and water used and the matter of attendance, and also in the general expense. It follows, then, that if the power of a manufacturing district can be centralized, it can be developed at a much higher efficiency and at a very much lower cost than where it is being generated in a large number of units. If, after such centralization, it can be distributed to a large number of users in the district with only a moderate loss, then supposing only that the same amount of power is sold as is generated in the central station, this plan of generating power would still have the advantage, in point of economy ; but when we take into account, as shown by our records, and as is being borne out by practical experience every day, that under the circumstances a central station can take advantage of a low average percentage in use of the capacity of machines, it then becomes apparent that if the system is properly constructed and operated there is not only not a loss, but a great gain in the centralization of steam power ; and this will hold true whether dealing with a central station of one hundred or ten thousand horse power, and, if operating under proper electromotive forces, whether the district is a square mile or ten square miles. In fact, electricity is the most convenient, tractable, yet powerful means of carrying and distributing energy or motion from one point to another. As a means of transmission, it has greater advantage over steam, water, gas, or compressed air, in that the system of conductors is far more flexible in its arrangements, and capable of much greater ramifications than any of its competitors. It can be operated under much higher relative pressures or potentials than the other methods of conducting power, and hence it can be distributed with very much smaller conductors and much smaller investments and losses. Of course the truth of this depends upon the efficiency of the different parts of the system.
There are three ways in which power is lost after it is generated in a steam engine and delivered to the dynamo pulleys. The first is in the dynamo. This has now been brought to such a high state of efficiency that there are several good types of machines which have a commercial efficiency as high as ninety per cent. The second loss is that which occurs in the distributing wires or conductors, and this depends upon the electromotive force used, the distance over which the current is distributed, the distribution and arrangement of the conductors, and the amount of current carried. This loss should not exceed in a general district over ten per cent. The third loss is in the motors, where the energy of the electricity is reconverted into mechanical power. Motors vary in their efficiency. To truly answer the commercial conditions, the current used should be almost directly proportional to the work done. The higher the efficiency of the motor, the more nearly will this law hold. Small machines are not as efficient as large ones, nor is every motor of equal efficiency under all loads; but in large machines the efficiency under full loads is now about ninety-one per cent.
As between two motors of differing efficiency, with that of the lesser efficiency not only would the size of the conductors which would be needed in a particular area to distribute a given amount of mechanical work be much increased, and the area over which any given amount of power could be distributed with a given weight of conductor be very much smaller, but also the capacity of both the dynamos and engines in a central station would have to be very materially increased to get a given output of mechanical work. Increased area, lessened weight of copper, greater recovery of power, and a larger amount of mechanical output, as well as a greater excess capacity of motors, and a less proportional investment in central station appliances and equipments—all these are in favor of the motor having the highest efficiency. The effect upon the lights where they are supplied from the same circuit is also very much less marked with motors of higher than with motors of lower efficiency. In transmitting power over long distances, even in single units, any increase whatsoever in the efficency of the motor or dynamo becomes exceedingly important, because the investments in conductors are very large.
Theoretical conclusions, however, were not of sufficient weight, nor were they generally understood, and one of the greatest difficulties we met with in the introduction of motors was that of determining a method of charging for power, because, on account of the lack of knowledge of the character of the machines and the variable work which they have to do, a contract basis seemed at first inadvisable, and the ordinary current rates for lamps seemed exorbitantly high. Few managers could rid themselves of the notion that, because there was a loss, first in the dynamo, then in the conductors, and finally again in the motors, covering an aggregate of about thirty-five per cent, that it would be impossible to compete with steam when supplied directly. In order to determine these facts, and to prove to people who were selling power, not only that there was not a loss, but actually an apparent gain in the transmission from a central station ; that motors were not to be treated as toys, but as important factors in the industrial arts, and that they would be justified in making a charge for electric power service based upon the sensible method of a contract—records were obtained and calculations made something after the following manner: After several motors had been put into use, we sent out a blank circular requesting, among other things, the number of hours' use which the motors averaged every day, the class of duty, and also the monthly payments, then mosdy made on a meter basis. The reports were varied. Some of the records extended over considerable periods of time, and they established the fact that motors, as a rule, do not average over thirtythree per cent of their capacity for the ten hours of a working day. This, in fact, is a large average. This may to a great extent be understood from the following statements : A motor is absolutely automatic. The total current used in machines of high efficiency is almost directly proportional to the work done at any instant. The result is that if the work slacks for a moment, so also does the current, and in nearly the same ratio. Almost all work is spasmodic in character. Motors have to be put in for their maximum capacity, but not one class of work out of a hundred offers continuous duty. Some motors are stopped, and in others which are running the work is constandy varying from the maximum to the minimum. An illustration of this may be cited in the case of elevators. If twenty elevators, each requiring a maximum of five horse power, were driven from the same line of countershafting, they would not require an engine of over forty horse power, because some would be going up, some going down, others standing still, and in not two cases out of the twenty would simultaneous trips be made, while in not one trip in five hundred would an elevator be hoisting its maximum load from the basement to the loft. Comparison also may be made with existing incandescent light plants. Practical experience shows that not over fifty-five or sixty per cent of the total number of lights connected on a station are used at any one time, although it is possible to use them all. But incandescent lamps when they are in use take their full amount of current. Not so with motors. A certain percentage of the motors, just as with the lights, would not be in use at all, and in most of those in use the current would be continually varying, just as if the lights of an incandescent light station were being turned up and down from one to sixteen candle power, as well as many of them being turned out. The recovery in an ordinary district with eighty-eight to ninety per cent of the power delivered to the dynamo pulleys converted into electricity on the line, and ten per cent loss in distribution, is about sixty-five per cent. This sixty-five per cent, however, where the work is distributed, is only thirty-three per cent of the capacity of the motors for which power can be contracted. In other words, for every one hundred horse power in a central station, sixty-five horse power at any one instant can be delivered under an ordinary distribution, and this sixty-five per cent is only one-third of the power which can be contracted for, provided there are a large number of motors in use. Of course, where only one motor can be used, then sixty-five horse power only would be obtained, but the object of a central station is to take advantage of the intermittent character of the work, and that it can be done is shown by the records which we have of a large number of motors in use.
The future, then, of motors in connection with central stations is assured, but the application of motors to stationary work, or more properly in combination with centres of supply such as the ordinary stations for electric lighting and power, is but one field, although a large one. There are many special fields of work, each of which demands large capital, and offers a wide range of application. Among these may be mentioned the application of the transmission of power to mining works. At present, in single plants, both in the mining of the common and of the precious metals, immense sums are invested in elaborate systems for the transmission of power by compressed air, single plants sometimes costing as much as a half a million dollars. Such a system of transmission must necessarily be costly. Not one case exists in which electricity could not be better used and with a very much reduced investment.
The transmission of power in single units for a variety of purposes too numerous to mention, such as, for instance, the operation of the transfer tables of the C, B., and Q. Railroad at Aurora, and the operation of overhead cranes, is itself a field requiring a great deal of thought. Some idea of the progress made may be gathered from the fact that we have now about twenty-five or thirty different types and classes of machines, and I know that the near future will require fifty or seventy-five.
Again, the substitution of motors for horses in street car work, using in some cases the conduit, in others the overhead wire, and Others the storage battery, is now rapidly coming into favor. On the two latter classes of work we are now actively engaged, and are just at present much interested in the most extensive electric railway system which has ever been contracted for, that of Richmond, which comprises forty cars and eleven miles of track. The great margin between steam and horse power gives ample room for increased efficiency and lessened cost of operation, especially where a number of cars are in operation, because here, as in regular central station work, the apparent output of power will be greater than the actual power of the central station, and, unlike horses, the motors do not require to be fed for twenty-four hours in the day independent of whether they are being used or not.
In one sense, that of size, the most important immediate experiment in which I am engaged is the development of motors for application to the movement of heavy trains on the elevated railroad. I was engaged last year for a considerable period in carrying on experiments on one of the branches of the elevated railroad, during the progress of which one of the standard elevated railroad cars, on a heavy grade and with a current of about six hundred volts potential, was handled under all speeds up to twenty-one miles an hour entirely by the motors, and without the use of hand-brakes, the practice of the recovery of the energy of the car, of which I have already spoken, being there very thoroughly developed. We have now stopped this experiment and have under way the largest apparatus of the kind ever undertaken. This consists of a motor car of special construction, which is designed partially for passengers, and is intended to haul itself and five or six other cars ; in other words, a hundred and forty ton train, and this at speeds as high as twenty-eight miles an hour. The middle portion of this car is to be used for passengers, but each end is reserved for the engineer and for duplicate controlling apparatus. Under each end of the car is a heavy iron truck fitted with two pairs of forty-two inch Krupp wheels and five-inch Midvale steel axles. On each axle is to be placed a duplex motor, the normal capacity of which is seventy-five horse power, but which can be worked up to a capacity considerably higher than this. Each motor is centered on the driving axle and drives from each end of the armature. Each motor weighs four tons. The armatures, of which there will be eight altogether, are about eighteen inches long and seventeen inches in diameter. Each of the eight field magnets with its wire will weigh at least a ton. The total weight of the car with passengers will be about thirty-two tons, but it is distributed over the entire distance between two columns. With a tangential strain on each armature of only about twelve hundred pounds, there will be a draw-bar strain of nearly 13,000 pounds; and this is not by any means the full capacity of the motors. The total power of the motors on this car will be double that of the heaviest locomotive on the road, and the traction coefficient more than twice as much. It will be some months before this is ready, because the work is being elaborated in great detail.
One of the most important applications to which I may call the attention of the gentlemen present is that of the use of electric motors on board ship. The uses to which motors can be applied on a man-of-war are numerous. It will probably be argued by many, when the proposals are made to carry on certain operations by means of electricity, that there is some doubt as to their practicability in so complex an organization and one subject to so many peculiar conditions not met with elsewhere as is a man-of-war. Sitjiilar objections were made a few years ago when it was suggested that the incandescent electric light ought to be introduced in the place of the then existing system of lighting. In 1881 I made some efforts to introduce the incandescent electric light on board our ships, and afterwards in 1882 I attended an exhibition at the Crystal Palace, Sydenham, England, and being required to report upon the progress of electric lighting, made as strong a report as I could in favor of this innovation. Whether this report had any weight or not, of course I do not know, but it seems that most of the objections which were raised against electric light proved unfounded, for various systems are being introduced and are giving satisfaction, and so thoroughly satisfied are officers that the proper method of lighting ships is by electricity that few of them would to-day think, for one moment, of seeing the old days return with their antiquated, unpleasant and unhealthy methods of illumination. So may we look upon the application of electricity as a motive power to a great variety of uses on board ship, and that with far more certainty of operation, more ease in manipulation, than in many machines now used. Few people not acquainted with the actual conditions met with on board ship realize the number of small engines in use and the variety of purposes to which they are applied. Steam has superseded hand steering apparatus ; steam ventilation takes the place of wind-chutes ; blowers, exhaust fans, and pumping machinery are driven by steam ; ashes are hoisted by its agency. Shell hoisting and gun training, which is now performed sometimes by hand and again by pneumatic, hydraulic, or steam apparatus, will certainly, sooner or later, be better done by means of electro-motors. In fact we have been asked already by the Department to submit designs for operating the new 8-inch rifle gun for the Chicago, and are now working on these ; the motor will be carried beneath the gun inside the carriage and the training will be under perfect control. Of course I do not say that the first experiment of this kind will be entirely successful, or that it will be devoid of objections. A perfect apparatus in this, as in any other line, must be developed by the process of evolution. It is not always possible to foretell every condition which will be met ; but failures are simply the keys to success. The hoisting of shell and other ammunition, now done laboriously and uncertainly by hand, can be better done by positive acting electro-motors which will be under perfect control. Centrifugal pumps, with their rotary movement and high speed, and likewise fans and exhausters, both for ventilation and for forced draft in the furnaces, will be run by motors. It may be urged that there is a liability of accidents occurring with apparatus of this nature. In answer we can say that duplicate parts are easily carried and injured portions quickly replaced—if anything, more quickly than in any steam apparatus. One of the great advantages of the use of electricity for these purposes will be the avoidance of a great deal of heat now caused by carrying steam pipes all over the ships. Steam steering gear is now operated in a clumsy, uncertain method from the forward and after pilot houses by long lines of countershafting and gearing which, because of the general structure of the ship, cannot be other than objectionable. The electro-motor will replace this gearing in the near future, and so it will go on and eventually replace the steam apparatus itself. I do not think I can be rightly accused of speaking merely as an enthusiast, but may fairly claim to understand some of the conditions which obtain in ship life, and I feel warranted in expressing my own confidence that the future will solve every difficulty which the application of electricity to classes of motive power I have mentioned on board ship may demand. One of the great advantages in the use of electricity on board ship is the ease with which the conductors may be handled and run as compared with steam pipes, and also the facility with which they can be replaced in case of accident. The generating apparatus can be placed well out of the range of shot and shell, and easily protected. Another field for electric motors on board ship is in making repairs. To-day, if a drill is to be operated, a new bolt hole made, or any other work requiring tools of this character, they must be operated by a slow, tedious handgear ; whereas a portable motor, with its flexible shaft and brace, can be introduced in the most out of the way part of boilers and engines, and work there performed which, were it done in the ordinary method, would require extensive preparations involving, possibly, the removal of heavy parts of machinery. Torpedoes can be operated as well as controlled by electricity. Diving operations and repairs to the ship's skin under water and in dock, repairs to armor, turrets, and to gun carriages can all be facilitated by the use of electric motors.
The methods of making these various applications of electric power may well receive the careful thought and criticism of all officers who desire the best good of the service, and I have no doubt but that many most valuable suggestions will in the future come from those who to-day look with conservative doubt upon the practicability of these proposals.
DISCUSSION.
The Chairman.—Gentlemen :—There is no practical application of electricity that is now attracting more attention or is of greater importance than the transmission of power by electricity. The telephone may be greatly improved, but it has taken its place as a successful invention and an indispensable instrument in conducting the ordinary affairs of life. It is not improbable that in a few years the electric motor will be made more useful than the telephone.
The science of electricity is so far advanced that no ignorant man will by accident hit upon a successful motor. Such a motor must be constructed in strict accordance with scientific principles. To be successful in these days of close competition it must be more than almost correct—it must be exactly right. It will be the result of the thoughtful work and experiment of an ingenious man well trained in modern science.
We are very fortunate in having had the opportunity to listen to one so thoroughly competent, so ingenious, so enthusiastic, and so successful as the lecturer of this evening.
The Naval Academy is to be congratulated that one of its graduates is in a position to return and, in this building, where he received his first theoretical and practical training in physical science, speak with authority as one of the foremost inventors and investigators in this branch of electrical science.
The subject is now before the Institute for general discussion. Mr. Sprague will, I doubt not, be happy to answer any questions that members of the Institute may wish to ask.
After some discussion of the feasibility of transmitting power over great distances, and of methods for testing the efficiency of motors, a vote of thanks, on the motion of Commander P. F. Harrington, was extended to Mr. Sprague for his valuable and instructive paper. The meeting then adjourned.