LIBRARY OF THE UNIVERSITY OF CALIFORNIA. Class ELECTRIC RAILWAYS A Series of Papers and Discus- sions Presented at the Interna- ** tional Electrical Congress in St. Louis, 1904 REPUBLISHETBY THE McGRAW PUBLISHING COMPANY NEW YORK By Special Arrangement w : th the International Electrical Congress 1907 A &. COPYRIGHTED, 1906 by the McGRAW PUBLISHING COMPANY New York PREFACE. The papers and discussions here reprinted define the state of the art of electric traction at the epoch of the St. Louis Exposition. The Electrical Congress of 1904 was in all respects valuable, but the papers on electric traction are particularly important, as marking the beginning of a new era that of heavy railway work, with the introduction of single phase alternating motors. The valuable papers of Arnold, Dawson, Steinmetz, Deri, and others represent the new point of view which entirely changes the aspect of electric railway operation. For the previous decade the methods of electric traction had remained essentially unchanged. They had been stretched be- yond the elastic limit, so to speak, in the endeavor to reach expanding conditions, and the logic of events demanded a change. The realization of this was strongly in evidence at the Congress of 1904, and the developments of the two years since passed have been directly along the lines then plainly foreseen. In fact, with few exceptions, the papers here presented bear the prophetic impress, and while the older methods found some vigorous support, the handwriting on the wall was plain for all to read. Achievement follows foresight, and commercial adapt- ation trails in the rear, so that one need not wonder at the ap- parently gradual progress in the actual equipment of roads. Such has been the history of other improvements, of the appli- cation of alternating currents to lighting and of polyphase cur- 1B1539 iv PREFACE. I rents to the electrical transmission of power. But methods in- evitably change with the times, and one reading these papers half a dozen years hence will marvel, not so much at the insight of the engineers who wrote them, as that the work had already been so long delayed. Progress seems easy when viewed from a sufficient distance. This volume records the field notes of the advance guard. LOUIS BELL. BOSTON, MASS. CONTENTS. Page, The history and development of electric railways 1 Some early work in polyphase and single-phase electric traction 21 Electric traction on British railways 52 The Monorail railway 76 The electrification of steam railroads 83 Alternating vs. Direct-current traction Ill Notes on equipment of the Wilkesbarre & Hazelton railway 189 Transmission and distributing problems peculiar to the single-phase railway 230 Protection and control of large high-tension alternating-current dis- tribution systems 238 Rotary converters and motor-generator sets 252 The Booster machine in traction service and its proper regulation... 262 Storage batteries in electric railway service 275 Electrolysis of underground conductors 288 Braking high-speed trains 315 Alternating-current motors 323 Single-phase motors 376 Alternating-current machines with gramme commutators 396 Single-phases railway motors 402 Theory and operation of the repulsion motor , 410 Theory of the compensated repulsion motor 429 ELECTRIC POWER TRANSMISSION COMMITTEE 1904 Honorary Chairman, M. PAUL JANET AND ING. A. MAFFEZINI. Chairman, MR. CHAS. F. SCOTT. Vice-President, ING. E. JONA. Secretary, DR. LOUIS BELL. OF THE A UNIVERSITY ) F THE HISTORY AND DEVELOPMENT OF ELEC- TRIC RAILWAYS. BY FRANK J. SPRAGUE. Although the earliest recorded experiments date back three-quar- ters of a century, the electric railway is essentially of modern de- velopment, for it achieved a recognized position less than twenty years ago, long after the telephone, the arc and incandescent lamp, and the -stationary electric motor had been thoroughly established. This is but natural, for it is the logical outcome of the establish- ment of certain cardinal principles and practices in the kindred arts. The first roads to carry passengers commercially were built in Europe, but the first railway experiments and the modern com- mercial impetus, as well as most of the essential and distinctive features of the art as it stands today, an example of almost unpre- cedented industrial development, are distinctively American. Brandon, Vt., birthplace, and Thomas Davenport, blacksmith, father, are the names first on the genealogical tree of the electric railway, in the year 1834. A toy motor mounted on wheels, pro- pelled on a few feet of circular railway by a primary battery, ex- hibited a year later at Springfield, and again at Boston, in the infant's photograph. This was only three years after Henry's in- vention of the motor, following Faraday's discovery ten yearn earlier that electricity could be used to produce continuous motion. The records of Davenport's career, unearthed by the late Franklir* Leonard Pope, show this early inventor a man of genius deserving a high place in the niche of fame, for in a period of six years he built more than a hundred operative electric motors of various NOTE The writer having been requested to prepare a paper on the sub- ject of electric railways has done so with considerable reluctance because of his own connection with the art, and the difficulty under such circumstances in presenting events in a true perspective, unbiased by personal experiences. That such must be spoken of is, while embarassing, somewhat necessary, and due allowances should be made in his estimate of their importance. ELEC. BYS. 1. [1] Z 8PRAGUE: ELECTRIC RAILWAYS. designs, many of which were put into actual service, an achieve- ment, taking into account the times, well nigh incredible. For nearly two score years various inventors, handicapped with the limitations of the primary battery, and in utter ignorance of the principles of modern dynamo and motor construction, labored with small result. About 1838, a Scotchman, Robert Davidson of Aberdeen, began the construction of a locomotive driven by a motor similar to that used by Jacobi in his experiments on the river Keva, which was tried upon the Edinboro-Glasgow Railway, and attained a speed of about four miles an hour. In an English patent issued to Henry Pinkus in 1840, the use of the rails for currents was indicated; also in a United States patent issued to Lilley and Colton of Pittsburg in 1847. In 1847, Prof. Moses G. Farmer, late United States govern- ment electrician at the Newport Station, one of the most learned and able of the early electric experimenters, operated an experi- mental model car at Dover, N. H. ; and about three years later one Thomas Hall exhibited in Boston an automatically reversing car mounted on rails through which current was supplied from M battery. These are said to be the first instances in which rails were: actually used as carriers of the current, as well as the first time where there was a reduction by gear from the higher speed on the motor to the lower speed of the driven axle. About the same time Prof. Paga of the Smithsonian In- stitute, aided by a special grant from Congress, constructed a locomotive in which he used a double solenoid motor with recipro- cating plunger and fly-wheel, as well as some other forms. This locomotive, driven by a battery of 100 Grove elements, was tried the 29th of April, 1851, upon a railroad running from Washington to Bladensburg, and attained a fair rate of speed. Patents issued in 1855 to an Englishman named Swear and a Piemontais named Bessolo indicated the possibility of collecting current from a conductor suspended above the ground, and in 1864 a Frenchman named Cazal patented the application of an electric motor to the axle of the vehicle. From the experiments of Farmer and Hall a decade elapsed be- fore the invention by Pacinotti in 1861 of the continuous current dynamo, from which may properly be said to date all modern elec- tric machines. These were developed in their earliest forms by Gramme and Siemens, Wheatstone and Varley, Farmer and Row- 8PRACUE: ELECTRIC RAILWAYS. land, Hefner- Alteneck and others, and brought into existence the elements essential to any possible commercial success. Yet not- withstanding that the principle of the reversibility of the dynamo- electric machine, and the transmission of energy to a distance by the use of two similar machines, said to have been discovered and described by Pacinotti in 1867 the same year in which Prof. Farmer described the principle of the modern dynamo in a letter to Henry Wilde and demonstrated independently at the Vienna Exposition by Fontaine and Gramme in 1873, many years more passed before the importance and availability of this principle were generally recognized. From 1850 to 1875, is a long period, relatively, and yet there seemed to have been practically an entire cessation of experimental electric railway work, until in the latter year George F. Greene, a poor mechanic of Kalamazoo, Mich., built a small model motor which was supplied from a battery through an overhead line, with track return, and three years later he constructed another model on a larger scale. Greene seemed to have realized that a dynamo was essential to success, but he did not know how to make one, and did not have the means to buy it. Shortly afterward, in 1879, at the Berlin Exposition, Messrs. Siemens and Halske constructed a short line about a third of a mile in length, which^ was the beginning of much active work by this firm. The dynamo and motor were of the now well-known Siemens type, and the current was supplied through a central rail, with the running rails as a return, to a small locomotive on which the motor was carried longitudinally, motion being transmitted through spur and beveled gears to a central shaft from which connection was made to the wheels. The locomotive drew three small cars having a capacity of about 20 people, and attained the speed of about eight miles an hour. In the same year important experiments were carried on by Messrs. Felix and Chretien at the little village of Sermaize in France to demonstrate the possibilities of the transmission of energy. At Vienna in the following year, Egger exhibited a model of an electric railway, the current to be supplied through the running rails. About the same time Messrs. Bontemps and Desprez made a study of a scheme for replacing pneumatic transmission of dis- patches by miniature electric locomotives in Paris. 4 SPRAQUE: ELECTRIC RAILWAYS. The Siemens and Halske demonstration in Berlin was followed by others for exhibition purposes at Brussels, Dusseldorf and Frank- fort, but no regular line was established until a short one with one motor car at Lichterfelde, near Berlin, the first in Europe, or in fact in the world. This road was 1 1/2 miles in length, used all rail conductors, and was opened for traffic in May, 1881. The motor was carried on a frame underneath the car between the wheels, and current transmitted from the armature to drums on the axles by steel cables. The car was of fair size, having a capacity of 36 passengers, and attained a maximum speed of about 30 miles. The e.m.f. used was about 100 volts. This line was con- tinued in regular service, but 12 years later the rail method of dis- tribution was replaced by two conductors carried on top of poles, upon which ran a small carriage connected to the gear by a flexible cable. Shortly afterward the same firm installed at the Paris Electrical Exposition of 1881, a small tramway about a third of a mile long, and used for the first time overhead distribution. In this case the conductors consisted of two tubes slotted on the under side, and sup- ported by wooden insulators. In the tubes slid shoes which were held in good contact by an underrunning wheel pressed up by springs carried on a frame-work supported by the conductors, and connected to the car by flexible conductors. The motor was placed between the wheels, and the power was transmitted by a chain. About the same time Siemens constructed an experimental road near Meran in the Tyrol with a view of demonstrating the possibili- ties of electric traction for the San Gothard tunnel, and later other small lines at Frankfort, Molding and elsewhere. These were fol- lowed by a comprehensive scheme for a combined elevated and underground road submitted to the city authorities at Vienna. The invention about this time of accumulators directed attention to the possibilities of the self-contained car, and in 1880 a loco- motive with accumulators was used at the establishment of Du- chesne-Fournet at Breuil, and in the following year Kaffard with a large battery of Faure accumulators made experiments on the tramway at Vincennes. In 1881, Dr. John Hopkinson, in describing the application of motors to hoists, proposed both for them and for tramways the simple series-parallel control for speed, a principle which combined vdth resistance variation later became universal. SPRAGUE: ELECTRIC RAILWAYS. 5 Meanwhile in the United States two inventors, Stephen D. Field and Thomas A. Edison, began electric experiments almost simul- taneously. Edison was perhaps nearer than any other on the verge of great possibilities had it not been that he was intensely absorbed in the development of the electric light, for he had in the face of much adverse criticism developed the essentials of the low internal resistance dynamo with high-resistance field, and many of the essential features of the multiple arc system of distribution. In fact, in 1880 he built a small road at his laboratory at Menlo Park, on which he ran a car operated by one of his earliest dynamos from which the power was transmftted to the axle by a belt. One set of wheels was insulated, and the two rails were used for current. But beyond taking out a few patents, and for a while acting in conjunction with Field, Edison did little in this particular field, and soon ceased to be a factor. Perhaps more than to any other the credit for the first serious proposal in the United States should be awarded to Field. Curi- ously enough, patent papers were filed by Field, Siemens and Edison, all within three months of each other in the spring and summer of 1880. Priority of invention was finally awarded to Field, he having filed a caveat a year before. He had been actively interested in electric telegraphs, and in an account of his work pub- lished some 20 years ago, it is stated that he early constructed two electric motors, and had in mind the operation of street cars in San Francisco, but had not been able to do anything in the matter because 1 of a realization that a dynamo must be used instead of a battery. In 1877 while in Europe he saw some Gramme machines, and on his return two of them were ordered but not delivered. Later a dynamo was ordered from Siemens Brothers in London which was lost, and this was replaced by another which arrived in the fall of 1878. Meanwhile two Gramme machines were placed at his disposal, and shortly afterward an electric elevator was operated. In February, 1879, he made plans for an electric rail- way, the current to be delivered from a stationary source of power through a wire enclosed in a conduit, with rail return, and in 1880-81, he constructed and put in operation an experimental electric locomotive in Stockbridge, Mass. Pending the settlement of patent interferences between Edison and Field (the Siemens application being late was rejected), the two interests were combined in a corporation known as " The Elec- tric Railway Company of the United States," and the first work of 6 8PRAGUE: ELECTRIC RAILWAYS. the company was the operation of an electric locomotive at the Chicago Eailway Exposition in 1883. This locomotive called "The Judge, 57 after the late Chief Justice Field, ran around the gallery of the main exposition building on a track of about one-third of a mile in length. The motor used was a Weston dynamo mounted on the car and connected by beveled gear to a shaft from which power was transmitted by belts to one of the wheels. The current was taken from a center rail, with track return. A lever operated clutches on the driving shaft, and the speed was varied by re- sistance. The reversing mechanism consisted of two movable brushholder arms geared to a cfisk operated by a lever, each arm carrying a pair of brushes one of which only could be thown into circuit at a time, to give the proper direction of movement. Meanwhile several other inventors were getting actively into the field of transmission of power and electric railways. In the sum- mer of 1882, Dr. Joseph E. Finney operated in Allegheny, Pa., a car for which current was supplied through an overhead wire on which traveled a small trolley connected to the car with a flexible cable, and about the same time in England Dr. Fleming Jenkin, following a paper by Messrs. Ayrton and Perry before the Royal Institution on an automatic railway, proposed a scheme of telpher- age which was developed by those gentlemen. In the early part of the same year, the writer, then a midship- man in the United States Navy, who had in 1879 and 1880 begun the designing of motors, was ordered on duty at the Crystal Palace Electrical Exhibition, then being held at Sydenham, England. While in London he became impressed with a belief in the pos- sibility of operating the underground railway electrically. He first considered the use of main and working conductors, the latter being carried between the tracks, with rail return, but noting the complication of switches on certain sections of the road, conceived the idea of a car moving between two planes, traveling on one and making upper pressure contact with the other, those planes being the terminals of a constant potential system. For practical application the lower of the two planes was to be replaced by the running track and all switches and sidings, and the upper plane by rigid conductors supported by the roof of the tunnel, and following the center lines of all tracks and switches, contact to be made there- with by a self-adjusting device carried on the car roof over the center of the truck and pressed upward by springs. SPRAGUE: ELECTRIC RAILWAYS. 7 In 1882 he applied for a patent on the first idea, which was but a variation from that shown in other patents, but the second laid dormant for nearly three years because of central station work and the development of the application of stationary motors. The storage battery still attracted attention, and in 1883 experi- ments were carried on at Kew Bridge, London. In the latter part cf 1884 the Electrical Power & Storage Company of London, under the direction of Anthony Keckenzaun, began a number of trials. The same engineer repeated his work at Mill Wall, and later in Berlin. The car body in his last experiment was carried by two trucks, each of which was equipped with a motor driving one axle through a worm gear. Keversal was accomplished by using two sets of brushes, and speed was varied by using one or both motors, also by using the motors in series or parallel with a resistance to cut down sparking when making the change over. Reckenzaun subsequently had charge of the experiments conducted by Wm. Wharton of Philadelphia, in which both a Reckenzaun and a Sprague motor were used in 1886. Here series parallel grouping of both batteries and motor circuits were used on the Sprague car, and a series parallel and resistance variation of motors on the car operated by Reckenzaun and Condi ct. Meanwhile, in the United States, Charles J. Van Depoele, a Bel- gian by birth and a sculptor by original trade, and an indefatigable worker, had become interested in electric manufacturing, and soon energetically attacked the railway problem. His first railway was a small experimental line constructed in Chicago in the winter of 1882-83, the current being supplied from an overhead wire. In the fall of 1883, a car was also run at the Industrial Exposition at Chicago. A year later a train pulled by a locomotive car, and taking cur- rent from an underground conduit, was successfully operated at the Toronto Exhibition to carry passengers from the street car sys- tem, and again in the year following Van Depoele operated another train at the same place, using on this occasion an overhead wire and a weighted arm pressing a contact up against it. Experiments were also carried on by him on the South Bend Railway in the fall of 1885, where several cars were equipped with small motors, and also in Minneapolis, where an electric car took the place of a steam locomotive. Other equipments were operated a i the New Orleans Exhibition, and at Montgomery, Ala., where the 8 8PRAGUE: ELECTRIC RAILWAYS. current was at first taken from a single-overhead wire which carried a traveling trolley connected to the car by a flexible con- ductor. Other equipments were put in operation at Windsor, Ont., Detroit, Mich., Appleton, Wis., and Scranton, Pa. In these several equipments the motors were placed on the front platforms of the cars, and connected to the wheels by belts or chains. The cars were headed in one direction, and operated from one end only. In 1888, the Van Depoele Company was absorbed by the Thom- son-Houston, which had recently entered the railway field, and Van Depoele continued in its active development until his death in 1892. Among the early American workers of this period, none was for a time more prominent than Leo Daft, who after considerable development in motors for stationary work took up their applica- tion to electric railways, making the first experiments toward the close of 1883 at his company's works at Greenville, N. J., these being sufficiently successful to be repeated in November of that year on the Saratoga and Mt. McGregor road. The locomotive used there was called " The Ampere," and pulled a full sized car. The motor was mounted on a platform, and connected by belts to an intermediate shaft carried between the wheels, from which another set of belts lead to pulleys on the driving axles. A center rail and the running rails formed the working conductors. Varia- tion of speed was accomplished by variation of field resistance, this being accentuated by the use of iron instead of copper in some of the coils. In the following year Daft equipped a small car on one of the piers at a New York seaside resort, and a little later another one at the Mechanic's Fair in Boston, the motor for this last being subsequently put on duty at the New Orleans Exposition. In 1885 work was begun by the Daft Company on the Hampton Branch of the Baltimore Union Passenger Eailway Company, where in August of that year operations were begun, at first with two and a year later with two more small electric locomotives which did not carry passengers themselves, but pulled regular street cars. A center and the running rail were used for the normal distribution, but at crossings an overhead conductor was installed, and connection made to it by an arm carried on the car and pressed up against it. The SPRAGUE: ELECTRIC RAILWAYS. * driving was by a pinion operating on an internal gear on one of the axles. Daft's most ambitious work followed when a section of the Ninth Avenue Elevated Eoad was equipped for a distance of 2 miles, on which a series of experiments were carried on during the latter part of 1885, with a locomotive called " The Benjamin Franklin."' The motor was mounted on a platform pivoted at one end, and motion was communicated from the armature to the driving wheel through grooved friction gears held in close contact partly by the weight of the machine and partly by an adjustable screw device. This locomotive, pulling a train of cars, made several trips, but the experiments were soon suspended. This work was followed by street railway equipments at Los Angeles and elsewhere, using double overhead wires carrying a trolley carriage. Meanwhile Bentley and Knight, after some experiments in the yards of the Brush Electric Company at Cleveland in the fall of 1883, installed a conduit system in August, 1884, on the tracks- of the East Cleveland Horse Eailway Company. The equipped sec- tion of the road was 2 miles long, the conduits were of wood laid between the tracks, and two cars were employed which were each equipped with a motor carried under the car body and transmitting power to the axle by wire cables. These equipments were operated with varying degrees of success during the winter of 1884-85, but were abandoned later. This work was followed by a double overhead trolley road at Woon- socket, the motors being supplied by the Thomson-Houston Com- pany, and later by a combined double trolley and conduit road at Allegheny, Pa. In 1884, Dr. Wellington Adams of St. Louis proposed a de- parture in motor mounting which recognized the necessity of re- moving the motor from the car body and directly gearing it to the axle. In his plan the field magnets were carried by the pedestals, and inclosed the axle on which the armature was to revolve, its motion to be transmitted by gearing. The method was impractica- ble, and found no application. In 1884-85, J. C. Henry installed and operated in Kansas City a railway supplied by two overhead conductors on each of which traveled a small trolley connected to the car by a flexible cable. The motor was mounted on a frame supported on the car axle, and the power was transmitted through a clutch and a nest of gears giving five speeds. In the following year a portion of another 10 8PRAOUE: ELECTRIC RAILWAYS. road was equipped. A number of experiments seem to have been conducted there, and on some the rails were used as a return. The collectors were of different types, and it is said that among others there was one carried on the car. The final selection was a trolley having four wheels disposed in pairs in a horizontal plane, carried by and gripping the sides of the wires; this feature, but using one wire and rail return, characterized a road installed by Henry in San Diego, Cal. ? opened in November, 1887. In the early part of 1885, Sidney H. Short began a series of experiments on a short piece of track in Denver which was fol- lowed by the construction, in conjunction with J. W. Nesmith, of a section of road for operation on the series system. These experi- ments were continued through 1885 and 1886, and were repeated at Columbus, but were doomed to ultimate failure because of the principle involved. Subsequently Short adopted the multiple system of distribution, and for a time essayed the use of gearless motors for tramway work, but reverted later to the geared type. Meanwhile work had begun in Great Britain, where the first regular road to be put in operation was that known as the Portrush Electric Eailway, in Ireland, installed in 1883 by Siemens Brothers of London. Power was generated by turbines, and the current was transmitted by a third rail supported on wooden posts alongside of the track, the running rails constituting the return. The pres- sure used was about 250 volts. This was followed in the same year by a successful short road at Brighton, installed by Magnus Volk, the current being trans- mitted through the running rails. Then came the railway in- stalled at Bessbrook, Newry, in 1885, under the direction of the Messrs. Hopkinson, and at Eyde, in 1886, in which latter year was also installed the Blackpool road by Holroyd Smith. In this latter case the conduit system was used with complete metallic circuit. The motor was carried underneath the car between the axles, and connected by chain gearing. Fixed brushes with end contact were used for both directions of running. Eeverting to work in the United States, Sprague again took up the electric railway problem, and in 1885, before the Society of Arts, Boston, advocated the equipment of the New York Elevated Eailway with motors carried on the trucks of the regular cars, and work was actually begun on the construction of experimental motors. Shortly afterward a regular truck was equipped, and a long series of tests made on a private track in New York city. In May, 8PRAGUE: ELECTRIC RAILWAYS. 11 1886, an elevated car was equipped with these motors, and a series of tests begun on the Thirty-fourth Street branch of the road. These motors may be considered the parent models of the modern railway motor. They were centered through the brackets on the driving axles, connected to them by single reduction gears, and the free end of the motor was carried by springs from the transom, the truck elliptics being interposed between this support and the car body. The truck had two motors, they were run open, had one set of brushes, and were used not only for propelling the car but for braking it. The motors were at first shunt wound, but later had a correcting coil in series with the armature at right angles to the normal field to prevent shifting of the neutral point. The car was operated from each end by similar switches, current at 600 volts were used, and increase of speed was effected by cutting out resistance in the armature circuit and then by reducing the field strength. This enabled energy to be returned to the line when decreasing from high speed. It being impossible to interest the railway management, the experiments were finally suspended. Soon afterward a locomotive designed by Field had a short trial on the same section of the Elevated. Sprague then turned his attention to building a locomotive car of 300-hp capacity, each truck to be equipped with two motors, each having a pair of armatures geared to the axle, but this evidently being ahead of the times, and the possibilities of street tramway traction becoming evident, these equipments were abandoned, and he began the development of the type of motor finally used in Richmond, one crude form of which was first used in storage battery experiments in Philadelphia, and others in New York and Boston, in 1886. One of the Elevated motors was put into ser- vice at the East Boston Sugar Refinery, and continued so for some- time. Reviewing the conditions at the beginning of 1887, statistics compiled by Mr. T. Commerford Martin show that, including every kind of equipment, even those a fraction of a mile long and operated in mines, there were but nine installations in Europe, aggregating about 20 miles of track, with a total equipment of 52 motors and motor cars, none operated with the present overhead line or conduit, and seven cars operated by storage batteries, while in the United States there were only ten installations, with an ag- gregate of less than 40 miles of track and 50 motors and motor cars, operated mostly from overhead lines with traveling trolleys 12 SPRAGUE: ELECTRIC RAILWAYS. flexibly connected to the cars. These were partly Daft, but prin- cipally Van Depoele roads. Almost every inventor who had taken part in active work was still alive. The roads, however, were limited in character, varied in equipment, and presented nothing sufficient to overcome the prejudices of those interested in transportation, and command the confidence of capital. The whole electric railway art may fairly be termed, and was in fact for sometime afterward,, in an experimental condition, and some radical step was necessary to overcome the inertia which existed, and inaugurate that develop- ment which has been so remarkable. This came in the spring of 1887, when the Sprague Electric Railway & Motor Company took contracts for roads at St. Joseph,. Mo., and Richmond, Va., the latter covering a road not then built,. and including a complete generating station, erection of overhead lines, and the equipment of 40 cars each with two 7 1/2-hp motors,, on plans largely new and untried. The price, terms, and guaran- tees were such as to impose upon the company extreme hazards,, both electrical and financial. The history of the Richmond road has been too often written to dwell upon it at any length here. Suffice it to say that after experimental runs in the latter part of 1887 it was put into commercial operation in the beginning of February, 1888, and for a year there followed an experimental period of development which taxed the technical and financial resources of the company to the limit. But it won out, and Richmond, by com- mon consent of history, now stands as that pioneer road which more than any other was effective in the creation of the electric- railway as it stands today. The general features characterizing it may be briefly summarized as follows: A system of distribution by an overhead line carried over the center of the track, reinforced by a continuous main con- ductor, in turn supplied at central distributing points by feeders from a constant potential plant operated at about 450 volts, with reinforced track return. The current was taken from the over- head line at first by fixed upper pressure contracts, and subsequently by a wheel carried on a pole supported over the center of the car and having free up and down reversible movement, exposed motors, one to each, were centered on the axles, and geared to them at first by single, and then by double reduction gears, the outer ends being spring supported from the car body so that the motors were individually free to follow every variation of axle movement, and yet maintain at all times a yielding touch upon the gears an abso- SPRAGUE: ELECTRIC RAILWAYS. 13 lute parallelism. All the weight of the car was available for trac- tion, and the cars could be operated in either direction from either end of the platform. The controlling system was at first by graded resistances affected by variation of the field coils from series to multiple relations, and series-parallel control of armatures by a separate switch. Motors were run in both directions with fixed brushes, at first laminated ones placed at an angle, and later solid metallic ones with radial bearing. The well-nigh heart-breaking experiences and the alternation of good and bad performances are largely matters of personal history, but the results accomplished soon commanded the attention of those interested in the street transportation, most prominent among whom at that time was Henry M. Whitney, President of the West End Railway of Boston, who was considering the adoption of the cable. He consented to come to Richmond, and accompanied by his associates stopped also at Allegheny City to see the underground conduit of the Bentley-Knight Company. The demonstrations made for his benefit were conclusive, the cable was abandoned, and orders given for trial installations on both the overhead and under- ground systems to run from the Providence depot in Boston to the suburb of Allston. A winter's run resulted in the abandon- ment of the conduit and the adoption of the overhead trolley sys- tem, the principal orders for equipment going to the Thomson- Houston Company which, having absorbed the Van Depoele Com- pany, was now pushing work' energetically. Mr. Whitney's decision had a vital bearing upon the commercial development of electric railways, and from that time there followed a period of extraor- dinary activity, in which for a time two companies, the Sprague Electric Railway & Motor Company and the Thomson-Houston Electric Company, were the principle competitors. There was a continuous improvement and increase in the size of apparatus. Form wound armatures, proposed by Eickemeyer, replaced irregular windings, and metallic brushes gave way to carbon, this single change, initiated by Van Depoele in 1888-9, going a long way to- ward making the art a success. Cast and wrought iron yielded to steel, two-pole motors to four-pole, double reduction gears to single, and open motors to closed, protected only by their own casings. In 1892 combined series parallel and resistance control was adopted, when the Thomson magnet blow-out was successfully applied to controllers by Mr. Potter, and this was a most effective agent in reducing the troubles of operation. 14 8PRAGUE: ELECTRIC RAILWAYS. The progress of the electric railway, however, was not unimpeded, for no sooner had the Kichmond road started than there was em- phasized a series of disturbances on the telephone lines which threatened the use of the rails for return, and brought on a con- flict with the Bell Telephone Company, far reaching in its char- acter and involving new legal questions. At that time it was almost universal practice for the telephone to be installed with single circuits and earth return. Already the service had become most unsatisfactory because of the multiplicity of electric install- ations of various kinds, with consequent leakages, troubles from induction and variations in earth potential. To the hissing and frying incident to the system as installed was now added the hum of the motor and exaggerated differences of potential at the ground connections. The first attempt to meet this was made in Eichmond by the superintendent of the exchange, who disconnected from the ground and joined all return wires to a common circuit. This obviated most leakage troubles, but did not get rid of the troubles of in- duction. Numerous law suits followed in nearly half the States of the Union, the telephone companies attempting to force the railways to use double overhead circuits, and the railway companies demanding their share of the heritage of the earth. The trolley contentions were in the main successful, and individual metallic circuits, vital to successful operation, and without which long dis- tance telephone is impracticable, were adopted, for which condition of affairs the electric railway may be thanked. The work accomplished at Eichmond, the widespread advertising of the equipment and the rapid spread of electric railways in the United States commanded the attention of the Old World, and work was begun in Italy, Germany and elsewhere along the same lines, but it was not until a number of years later that there was any general adoption of the electric railway in the more conserv- ative countries. Meanwhile the Sprague Electric Eailway & Motor Company was absorbed in 1890 by the Edison General Electric, which later combined with the Thomson-Houston Company and others in the General Electric. For the next six years the record of the electric railway is that of industrial development, practically as indicated in the improve- ment of apparatus, the replacement of horse and cable power on existing lines, and the creation of new ones. Electric operation SPRAGUE: ELECTRIC RAILWAYS. 15 on tramways having become established, there naturally followed more ambitious attempts in limited applications of electricity to heavier work. In November, 1890, a line on South London road, which was originally designed for cable, was opened, the trains being pulled by electric locomotives equipped with a pair of gearless motors having armatures mounted on the axles of the drivers. In June, 1891, Sprague offered to install on the New York Elevated road a train to be operated by a locomotive car, and also one with motors distributed under the cars, and to make an express speed of 40 miles an hour. Two years later the Liverpool overhead railway was put in operation. Here the trains were composed of two-car units, each car having one motor, the two being operated by hand control. In the spring of the same year, 1893, the Intramural Kailway was constructed at the World's Fair, the equipment being supplied by the General Electric Company. Four motor cars with hand control were used to pull three trail cars, and a third-rail supply with running-rail return was adopted. Two years later the Met- ropolitan West Side Elevated road in the same city was equipped on the same general plan except using two motors instead of four. In May, 1896, the Nantasket Beach road, a branch of the New York & New Haven Kailway, was put in operation, and in Sep- tember the Lake Street Elevated of Chicago began electrical oper- ations. In November of the same year, electric service was in- stituted on the Brooklyn Bridge, the motor cars being used to handle the trains at first at the terminals but later across the bridge. There were few attempts, however, to replace steam on regular roads, and only occasionally were electric locomotives adopted for special reasons. Among the earlier ones built were one of 1000 horsepower, 1892-94, designed by Sprague, Duncan and Hutchin- son for Mr. Henry Villard for experimental operation on lines out of Chicago, which was never undertaken, and the still larger loco- motives built by the General Electric Company, which began operation of the trains in the Baltimore & Ohio tunnel in 1895. For a long time the conduit system, after its abandonment at Allegheny and Boston, remained quiescent, and all work was practi- cally with the overhead trolley. In 1893 a short line was tried in Washington on the Love system, but it was not until the following year that work was begun in New York on the Lenox Avenue 16 8PRAOUE: ELECTRIC RAILWAYS. line, and carried to that successful conclusion which warranted its widespread adoption in that city, under the auspices of Wm. Whitney and Henry Vreeland, and in Washington under Con- nett, although a line had been in operation at Budapest for some time. All this of course was largely because of the necessary cost of the heavy construction, and because street railway man- agers would not and could not undertake any such investment except under most favorable traffic conditions, and then with the additional restriction of a prohibition of the use of overhead wires. About this period there began that rapid introduction of inter- urban railways, soon aided by the developments in transformers by Stanley, in polyphase transmission by Tesla and Ferraris, and in rotary transformers by Bradley and others, which has had sucE an influence upon steam railway operation and been so instrumental in knitting together urban and rural communities. The first practical proposal for a railway using high-tension alternating-current transmission, seems to have been made in 1896 by Bion J. Arnold in plans for a road to run from Chicago to the Lake region, and although this road was never built the general plans were utilized for a line actually put into operation about two years later, which was the forerunner of the standard practice of today by means of which the limitations of distance have been so effectively reduced. In 1896 Sprague again sought the opportunity to make a demonstration on the Elevated Eailway in the form of a proposition to the management to equip a section of the line, and operate a train of cars on a new principle, the " Multiple Unit." Although the advantages of the system, such as higher schedules, reduced weights, variable train lengths, more frequent trains, dis- tributive motive equipment and increased economy were presented, .and supplemented by an offer to equip the whole system, no re- sponse whatever was made. A similar proposal repeated seven months later met with like fate, but in the spring of 1897 he made a contract with the South Side Elevated Kailfoad, in Chicago, to equip the line on this plan in lieu of the locomotive car plan then under consideration. This system has now become so widely known that any detailed description of it is unnecessary. Generally speaking, however, it is essentially the control of controllers, by means of which cars equipped with motors and controllers for them are operated from SPRAGUE: ELECTRIC RAILWAYS. 17 master switches through a secondary line, with provision for so coupling up cars that from any master switch all cars can be oper- ated irrespective of number, order or end relation, or whether all or only part of the cars, are equipped with motors. The first equipment was for 120 cars, and the first public demonstration was made in July, 1887, at Schenectady, on a full train of cars which had been sent from Chicago for that purpose. A regular train was put into operation before the close of the year, and within a few months steam operation was entirely replaced, As originally equipped, the main controller consisted of a mag- net-operated reverser and pilot-motor driven cylinder, operated semi-automatically and with throttle restraint through a secondary line and relays from master switches on the platforms. A number of variations have since been developed, such as operating the reverser and cylinder by air pistons electrically controlled, or breaking the main controller up into several magnetically operated parts, and all forms of equipment are now in operation. The es- sential principle of the system, however, has not been changed, and it has become standard wherever required to operate electric trains at high schedules. Equipments have grown from 100 horse- power per car to 2200 horse-power per locomotive, for in the largest work under way, that of the New York Central, the locomotives are to be controlled on this plan. The necessities of tunnel traffic on the one hand and a grave accident on the other have curiously enough centered in New York the largest two electric transportation problems, namely, that of the operation of the Pennsylvania tunnel and terminals, and more extensive still, that of the New York & Hudson Eiver Kailroad for 35 miles out from its terminals. The general requirements are so exacting, and the installation of the latter under such difficult continuous working conditions that they will prove of historic in- terest, and be influential in determining the disposition of many terminal problems. Up to comparatively recent times most of the electric railways, including those just mentioned, have been planned for operation with continuous current motors at moderate potentials, but this has often required the conversion of alternating current trans- mitted at high potential into continuous current at a lower one through the medium of transformers and rotary converters. While this bids fair to be the practice for some time, there are of course certain objections which are apparent, and the best energies of many KI.F.C. RYS. 2. 18 SPRAGUE: ELECTRIC RAILWAYS. of the ablest electrical engineers have for some time been bent upon solving the problem of operating directly with alternating cur- rents. Among the most active and successful of these have been the Ganz Company, whose Valtellina line, equipped on the polyphase plan for Italian Government, is of special interest. Among note- worthy experimental installations is that conducted under the auspices of the German Government on the Zossen military line, where the highest record for speed of a car carrying passengers, about 126 miles per hour, has been made during the past year, the current being collected frpm the three overhead wires by sliding contacts. The multiplicity of conductors, however, distinctly militates against this as any general solution of the larger railway problems, quite independently of other limitations affecting trunk-line trans- portation, and hence single-phase operation, using one overhead conductor with track return, is being energetically prosecuted. Among the workers who have sought solution and been active in invention along this line, as well as one of the earliest and most persistent advocates of single-phase railway operation, is Mr. Arnold, who has developed an electro-pneumatic plan in which is combined on a locomotive a constant speed single-phase alternat- ing-current motor with reversible air pumps and a storage tank, by which starting and running can be controlled by compressed air with a more even demand upon the capacity of the station. Arnold's experiments, a long time delayed from various causes, are now being subjected to the actual tests which will demonstrate the practicability of this scheme. Meanwhile, becoming alive to the limitations of past practices and the increasing demands of the art, the engineers of the various manufacturing companies in the United States and Europe, among whom must be especially men- tioned Finzi, Lamme, Latour, Winter, Eichberg and Steinmetz, are developing the single-phase alternating-current motor along two general lines. One is by using a series motor of special construc- tion, plain or compensated current being supplied from the second- ary of a transformer carried on the car and operated at moderate frequency. Another form is that originally proposed by Thomson, and known as the " repulsion " type, in which the field is supplied directly at high potential, and the armature is short-circuited upon itself and operates at low potential. An alternative of this form is that developed by European engineers, in which a variable potential is delivered to the armature from a transformer, the field being SPRAGUE: ELECTRIC RAILWAYS. ID supplied direct from the line. One desideratum is of course to be able to operate both from alternating and continuous currents, and this has been done, but the best results may possibly be gotten by ignoring this limitation. It is unnecessary to go into the many variations or details of these various schemes. Suffice it to say that all are being sub- mitted to the crucial test of commercial operation, and the over- coming of difficulties of the early days of electric railroading war- rant expectation that a great measure of success will likewise be attained on these new lines, and that another bar to the wider spread of electric railway operation may be speedily removed. This paper will not be burdened with detail statistics, but to illustrate in a general way the growth of the electric railway it should be noted that three years after the inauguration of the Kich- mond road there were in operation or under contract in the United States, England, Germany, Italy and Japan, not less than 325 roads, representing an equipment of about 4000 cars and 7000 motors, with 2600 miles of track, on which there was made a daily mileage of not less than 400,000 miles, and three-quarters of a billion of passengers were carried annually. By the end of 1903, in the United States alone, there was a total of over 29,000 miles equipped, 60,000 motors and 12,000 trail and service cars in service, and the passengers carried ran into billions. What the electric railway has done may only briefly be referred to here, but the writer may be permitted to repeat the substance of remarks written some nine years ago, for it has become a most potent factor in our modern life, and left its imprint in the indelible stamp of commercial supremacy. It has given us better paved streets, greater cleanliness, more perfect tracks, and luxurious, well-lighted and well-ventilated cars. With the higher speeds it has made possible the extension of the taxable and habitable areas of towns and cities in a much greater ratio than is repre- sented by the increase of speed. It has released from -drudgery tens of thousands of animals, and increased the morale of transportation employees. It has given employment to an army of men, and hundreds of millions of capi- tal. It has improved and extended the telephone service by forc- ing the abandonment of ground circuits. It has built up com- munities, shortened the time between home and business, made 20 8PRAGUE: ELECTRIC RAILWAYS. neighbors of rural communities, and welded together cities and their suburbs. Will it replace the steam locomotive? Perhaps the best answer is that " its future is not in the whole- sale destruction of existing great systems. It is in the development of a field of its own, with recognized limitations but of vast possibil- ities. It will fill that field to the practical exclusion of all other methods of transmitting energy; it will operate all street railway systems, and elevated and underground roads ; it will prove a valu- able auxiliary to trunk systems; but it has not yet sounded the death-knell of the locomotive any more than the dynamo has that of the stationary steam engine. Each has its own legitimate field." SOME EAELY WORK IN POLYPHASE AND SINGLE-PHASE ELECTRIC TRACTION. BY BION J. ARNOLD. In 1896 I became interested in a proposed road projected to run west and north from Chicago into the lake regions of Wiscon- sin, and to be known as the Wisconsin Inland Lakes & Chicago Electric Eailway. The rotary converter was then just beginning to be commercially exploited, and had, I believe, been used in some instances for power transmission, but so far as I know it had not been used for railway work. Desiring to construct the road, some 75 miles in length, as economically as practicable, and seeing no. reason why rotary converters would -not operate on railway work, I decided to adopt a three-phase high-tension transmission system with sub-stations, using rotary converters and storage batteries a radical depart- ure from the then standard 500-volt direct-current system. Complete detailed specifications for the road and its equipment were prepared, calling for three-phase generators capable of sup- plying current at 1040 volts, the necessary step-up and step-down transformers, switchboard apparatus, rotary converters, etc., re- quired to generate alternating-current energy at 1040 volts, trans- mit it at 5000 volts, and convert it into direct current at 700 volts to supply the overhead conductor, from which standard direct-current railway motors were to be operated, using storage batteries as equalizers in sub-stations distributed along the line. Fig. 1, which shows the arrangement proposed, is a reproduction of one of the original drawings attached to the specifications sub- mitted to the railway company at the time the final specifications were delivered. It happened, unfortunately, that the promoters of the road were unable to secure the necessary franchises for its construction, and it remains unbuilt today, while the specifications and plans repose among the archives of my office as evidence that an engineer, [21] 22 ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION eager to see his ideas executed, is apt sometimes to do much work for no pay and stand the preliminary expenses himself. The ter- ritory has since been partially occupied by the Aurora, Elgin & Chicago Electric Railway, and the Chicago & Milwaukee Electric Railroad. However, while this experience was somewhat disappointing financially, the time and study put upon it were not lost. A few months later the promoters of another road, now a part of the Chicago & Milwaukee Electric Railroad, came and stated that they must build 15 miles of new road in order to connect two small roads, each about a mile long, and that out of the total money Distance* JbtropoUtan Road to Elgin -< Power House 12 MU .. H " Wilmot 54 MU Elgin Branch 20 MU Total length of road 74 MU FIG. 1. MAP SHOWING LOCATION OF POWER-HOUSE; SUBSTATION AND DIS- TRIBUTION SYSTEM OF THE WISCONSIN INLAND LAKES & CHI- CAGO ELECTRIC RAILWAY AS PLANNED IN 1896. available to build this road they had provided but $10,000 to put into copper. After carefully calculating the cost of the road and finding it prohibitive, if built under the then standard 500- volt direct-current system of distribution, the plans of the Inland Lakes Road were resurrected. To have built the new road under the 500-volt direct-current system would have necessitated in- vesting almost as much money for copper alone as the parties had at their disposal for building the complete electrical and* mechani- cal equipment. ARNOLD: POLYPHASE AND SINGLE-PHASE TRAGTION. 9.3 After explaining the alternating-current plan and showing its adaptability to the case, and the impossibility of constructing under the standard 500-volt system, a remark was made by one of the owners to the effect that : " If the engineer was willing to take the professional risk the owner would take the financial risk." Au- thority was secured to build in accordance with the rotary-con- verter plan I had submitted, on condition that the road must be in operation within 90 days, in order to save the franchises under which it was authorized. One of the leading manufacturing companies had on hand at this time (March, 1898) three 120-kw rotary converters, which had been built for experimental purposes mainly, and by contracting with this company for the new electrical machinery required for the road, the use of these rotaries, provided with temporary trans- formers and switchboard apparatus, was secured. A new power-house was built, a transmission line eight miles long, consisting of three No. 8 bare copper wires carried upon or- dinary Western Union single-petticoat glass insulators, was con- structed, and the temporary apparatus installed. It was necessary to belt two of -the rotaries in tandem from the fly-wheel of the engine, and use them as generators, one supplying direct current to the section of the line nearest the power-house, while the other supplied three-phase current to the third rotary placed in the sub-station eight miles away. The alternating cur- rent was stepped up at the power-house and transmitted at 5000 volts. The road was opened for traffic July 1, 1898, and ran with fair success with the temporary apparatus until the following spring. In the meantime the ownership had changed hands, and the new owners, owing to their unfamiliarity with electric railways and the trouble due to the. temporary character of the plant (the new machinery not yet having been received from the manufacturers), desired to change the road into a standard direct-current system, and in this position they were supported by several engineers whom they consulted, and who reported adversely to the new system. It was also intended to extend the road southward 10 miles to Evanston, the road previous to this time having extended only from Waukegan to Highland Park, a distance of about 15 miles. In order to prevent the abandonment of my plans and of the alternating system it became necessary for me to assume the entire risk, and a contract was entered into whereby 24 ARNOLD: POLYPHASE AND SINGLE-PEASE TRACTION. I undertook to complete and extend the road in accordance with the original designs and guarantee, under a bonus and forfeiture contract, a certain efficiency between the steam-engine cylinders and the car motors under working conditions, and the successful operation of the system as a whole. The contract was dated March 21, 1899, and as an example of how rapidly engineering and construction work can be done when necessary, I will state that the conditions of the contract were successfully met on time, and when the work called for by it was completed the road stood, on June 20, 1899, equipped with a central power station, and two sub-stations, each eight miles from the DIAGRAM OP FEEDER SYSTEM 1 . CHICAGO AND MILWAUKEE ELECTRIC R.JR. FlQ. 2. MAP SHOWING LOCATION OF POWER-HOUSE, SUB-STATION AND DIS- TRIBUTION SYSTEM OF THE CHICAGO & MILWAUKEE ELECTRIC RAILWAY COMPANY, AS PLANNED IN 1898 AND COMPLETED IN 1899. FIRST ROTARY CONVEETEB SUB-STATION ROAD. power-house, all equipped with new machinery, regulating bat- teries, together with all necessary high-tension transmission lines and direct-current feeders for operating 16 40-ton trains be- tween Evanston and Waukegan, a distance of 27 miles, at an average speed of 20 miles per hour, with stops averaging one per mile. The energy was generated and transmitted at 5500 volts, as this was the highest pressure that the manufacturers, whom the con- ditions made it desirable to contract with for the electrical machin- ery on account of their experience and ability to make prompt deliveries, were prepared to furnish machinery for at that time. ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. 23 The success of the road was immediate, and its traffic has grown so rapidly that its capacity has been increased to three times its original capacity, during the past year, under the direction of my office. While there was an instance of a one-car road at Concord, N. H., taking its power through a rotary converter, located about four miles from a water-power generating station, the road I describe, I believe, was the first road to be put in operation designed to MILWAUKEE I RACINE KEN OS HA WAUKEGAN NORTH CHICAGO LAKE BLUFF LAKE FOREST fORT SHERIDAN i tilGHWOOD tMGHLANO PARK RAVINIA GLENCOE IAK6SIDE WINNETKA KCNIUWORTH WlLMETTt .WRTh EVANSTON EVANSTON" CHIC FlO. 3. MAP SHOWING GEOGRAPHICAL LOCATION OF THE CHICAGO & MIL- WAUKEE ELECTRIC RAILWAY. run from a central alternating-current power station, using high- tension transmission lines, rotary converters and sub-stations. It was thus probably the prototype of the system that rapidly became standard, and upon which almost all suburban lines have been built since. Fig. 2 is a map of the road, drawn to scale, giving the relative locations of the power-house and sub-stations, and is a reproduc- tion of one of the original sketches attached to the contract en- tered into on March 21, 1899. The portion of the line, north of the power-house at Highwood, was installed during the previous year and equipped with the temporary machines. 26 ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTIOX. Fig. 3 shows the relative location of the road to the surrounding territory. While this system was a marked step in advance in electric rail- roading, effecting as it did a great reduction in first cost and oper- ation, it did not seem to me to be the final solution of the electric railway problem on account of the losses due to the many conver- sions of the current and the excessive investment in sub-station machinery, with the attendant operating expenses. In 1899, while still engaged upon this work, I, therefore, com- menced to develop a system which should utilize the alternating current directly in the motor and employ but one overhead con- ductor, and thus eliminate the sub-station completely, together with the disadvantages of the complicated overhead work made necessary by the use of three-phase motors as then applied to alter- nating-current railway work in Europe. Realizing the advantages that storage batteries offered for equalizing the load in direct- current work I planned to retain a similar advantage for the alternating-current system by utilizing some form of a storage sys- tem to be carried upon the car. As the single-phase motor was not at that time capable of self -starting under load, some supplemental means must be provided for starting it. Air was the medium chosen, for by its use in combination with a high-tension single- phase motor I saw a possibility of requiring not only a single overhead working conductor, but of maintaining a constant load upon the power-house, thus enabling the investment in machinery and transmission lines for any given case to be much less than would be possible with the heavy fluctuating loads common to all electric- railway systems. The essentials decided upon were: (1) A motor which would use single-phase alternating current without conversion. (2) Single overhead working conductor. (3) Steady load upon the power-house. (4) Independent unit for switching purposes. The principles underlying the system which I developed to accomplish these results were: (a) A single-phase motor mounted directly upon the car axles, designed for the average power required by the car, running at a constant speed and a constant load, and, therefore, at maximum efficiency. ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. 27 (ft) Instead of stopping and starting this motor and dissipat- ing the energy through resistance, as was then common to all rail- way systems, the speed of the car was controlled by accelerating or retarding the parts usually known as the rotor and the stator, by means of compressed air in such a manner as not only to regu- late the speed of the car but also to store the kinetic energy of the car when stopping and utilize it in starting. Draughtsmen were put at work preparing the Patent Office drawings for different methods of applying the above principles, and late in 1899 an opportunity for trying the system was offered in the case of a road designed to extend about 60 miles northward from Lansing, Mich., and to be known as the Lansing, St. Johns & St. Louis Electric Railway. In January, 1900, I rode over the pro- posed right of way with a party of gentlemen interested in the road, and as a result of the negotiations that ensued a contract for its Wirt Maple Rapid* I?IG. 4. MAP SHOWING LOCATION OF THE LANSING, ST. JOHNS & ST. Louis ELECTBIC RAILWAY. FIBST SINGLE-PHASE BOAD. construction was entered into on April 23, 1900, wherein I under- took to build the road, assuming part of the financial risk. Fig. 4 is a reproduction of one of the original sketches attached to the contract, and Fig. 5 is a map showing the relative location of this road to the other roads in the State of Michigan. Locating engineers were at once placed in the field, and the construction proceeded systematically until 20 miles of the road (extending from Lansing to St. Johns) were completed to such an extent that it was opened for operation with steam locomotives about Nov. 15, 1901. For financial reasons the construction work was delayed but in the meantime the development of the electrical system was going on in different offices and shops. The overhead work of the 20-mile section of the road was com- pleted and ready for operation about Dec. 15, 1902, and the power installed, so that experiments with the electropneumatic system began in March, 1903. During these and all subsequent experi- 28 ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. ments the power was supplied from a 300-kw rotary converter, generating at 25 cycles and located in a combined water and steam- FlQ. 5. MAP SHOWING RELATIVE GEOGRAPHICAL LOCATION OF THE LANSING, ST. JOHNS & ST. Louis ROAD. power plant about two miles from the Lansing end of the line. The energy was carried to the motor over two No. 3 bare copper ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. 2!) wires, one of which was attached to the rails of the track and the other to the No. .00 trolley wire. Much experimental work had been done at the shops where the machine was constructed during the preceding year. On June 15, 1903, two trips were made, each about three miles long, with the first experimental machine, which is illustrated in Fig. 6. On the first trip eight persons 1 were carried and on the second trip 13 2 persons were aboard, and I give the names, as I believe this was the first public demonstration of a single-phase railway built for commercial use. At this time the voltage on the over- head conductor was carried at 2400 volts. The locomotive was a crude affair made hastily from a truck of one of the cars (Fig. 7) upon which was placed the motor, some rough timber for supporting the transformers, and the air tanks and controlling devices originally planned to be placed on a large car as shown in Figs. 8 and 9, but which a single motor was unable FIG. 8. DRAWING OP CAB OF LANSING, ST. JOHNS & ST. Louis ELECTRIC RAILWAY. to drive, thus necessitating the temporary construction shown in Fig. 6. The above tests demonstrated that the motor would work, and as the first machine was necessarily a makeshift and had been con- siderably damaged during its preliminary trials, it was thought bost not to attempt -further tests until a complete equipment could be built. 1. A. S. Courtright, G. A. Damon, W. A. Blanck, J. F. Scott, T. M. Keeley, Fred Rider, M. P. Otis and B. J. Arnold. 2. Mr. and Mrs. A. S. Courtright, Paul Courtright, Mr. and Mrs. T. M. Keeley, Leroy Keeley, Mr. and Mrs. Fred Rider, Mrs. T. E. Hamilton, Mrs. A. N. Hamilton, Miss Isabel Hamilton, H. B. Quick and M. P. Otis. 30 ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. A new double-motor equipment in the form of a locomotive, illustrated in Figs. 10 and 11, was completed and made ready I FlG. 10. LONQITUDIONAL SECTION OF LOCOMOTIVE NO. 2. for operation early in December, 1903, but on the morning of Dec. 18, a few days prior to the date set for public tests, the carhouse in which it was stored was completely destroyed by fire ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. 31 and with it went the locomotive, two new cars built for the system, and a steam locomotive used on the line. FIG. 11. TBANSVEBSE SECTION OF LOCOMOTIVE wo. 2. 32 ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. Unfortunately no photographs were secured of the complete machine before it was destroyed, but Fig. 12 shows the wreck the morning after the fire, and Fig. 13 shows the character of the weather and the conditions of the road at the time. No insurance was carried upon the machine, but the work of rebuilding was at once commenced. All of the electrical machin- ery and other electrical parts were returned to the manufacturers to be rewound or rebuilt, and all parts of the air machinery that could not be repaired on the ground were ordered new, except the main cylinder castings, which though cracked were in such a con- dition as to warrant attempting their repair by pumping a strong solution of sal ammoniac and water into them under pressure and thus attempting to close the cracks by oxidization. This was partially successful, and a new locomotive, Figs. 14 and 15, christened "Phoenix" was completely and recently made ready for trial. In the meantime, as it became necessary to place the road in operation electrically in order to operate in conjunction with the local street railway system in the city of Lansing, which had been acquired by the owners of the Lansing, St. Johns & St. Louis line, provision for operating the direct-current motor cars of the city line was made, under my direction, by adding additional copper and the installation of a rotary sub-station. It is interesting to know that the rotaries and sub-station appara- tus now operating this road are the same ones installed on the Chicago & Milwaukee Electric Eailway in 1899, they having served their purpose well and been removed to make room for larger ones recently installed to take care of the increased demands of that road. The Lansing, St. Johns & St. Louis road is now so equipped that by throwing suitable switches in the sub-station, either direct cur- rent at 600 volts, or alternating current at 6000 volts, can be turned on the trolley-wire at will, thus making it practicable for the road to run direct-current cars a large part of the time, and allow the operation of my experimental locomotive at such times as may be desired. On the evening of Aug. 3, 1904, the Phoenix made its trial run from Lansing to Dewitt, a distance of eight miles, carry- ing the superintendent of the road, two newspaper men, the writer and three assistants. ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. 33 FIG. 16. DETAILS OF OVERHEAD WORK, USED ON LANSING. ST. JOHNS & ST. ELEC. BYS. 3. LOUIS RAILWAY. 34 ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. Trouble in the power-house, due to the breaking of an engine prior to the trial, made it impossible to maintain the current on the line continually, on account of the blowing of the circuit breaker; otherwise the run would have been made over the entire 20 miles to St. Johns. The run was made with 6000 volts on the trolley- wire, and on the whole was satisfactory, as it demonstrated the ability of the machine to run smoothly at all speeds from zero to synchronous speed, and maintain a constant load on the power- house. The control of the speed of the car seemed perfect. Owing to the cracks in the cylinder castings not having been fully stopped, and loss of current from the absence of several in- sulators on the line, no attempt to determine efficiency of opera- tion was made, but as these defects can be remedied additional FlQ. 17. VIEWS OF SPECIAL INSULATOR USED FOB BUPPOBTINQ THE WOBK- INQ CONDUCTOR. run& will be made to determine the efficiency of the system. This is, I believe, the longest run yet made upon a road built for single- phase operation. Having thus described the conditions surrounding the develop- ment and application of the system, a more detailed description of it may be of interest. The track of the road does not differ from standard steam or electric railroad construction, except that but one line of rails was bonded, as it was thought that at the high-working voltage the amount of current would be so small that the bonding of the other rail would be unnecessary. "Wood was used for both pole and bracket, as illustrated in Fig. 16, which also shows the details of construction of the over- head work. A special trolley insulator was designed, Fig. 17, as ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. 33 it was intended to experiment with pressures as high as 15,000 volts on the working conductor. The insulators were made of annealed glass and tested up to 30,000 volts. Had a bow or some form of sliding contact been used as originally intended, these insulators would probably have proven satisfactory ; but with the running of short four-wheeled direct-current cars over the line came the frequent jumping off of the trolley wheels, resulting in the breaking of many of the insulators. Such con- struction should, therefore, not be used with anything but a slid- ing contact or bow trolley. One of the most difficult problems in the development of the electropneumatic system was to design an air compressor which would not only work efficiently as a compressor but could also be made to work efficiently as an engine. Much time was spent upon the development of various valve mechanisms and many types of engines were designed. The objects to attain were first, quick- opening and quick-closing valves; and second, valves so driven that when the machine was not running as an engine they would not be mechanically moved. They should also be capable of oper- ating automatically when the machine is running as a compressor. By the development of electropneumatically operated valves, described later, these objects were accomplished, and the inequality of the point of cut-off, due to what is technically known as "the angularity of the connecting rod " was eliminated, thus making it possible for each compressor when running as an engine to open its inlet and outlet valves at exactly the right point of cut-off for each end of the cylinder under all conditions of operation, regardless of the direction in which the engine runs. This was accomplished by the use of valves which operate pneumatically without loss of air, the time of opening and closing being electrically piloted by means of collector rings mounted upon or driven by the main shaft of the engine. These collector rings consist of several insulated seg- ments so placed with reference to the crank that they operate the valves instantaneously at such times as an eccentric would if it were placed directly in line with or directly opposite the crank pin. Primarily a car-motor equipment consists of a single-phase motor having both its rotor and its stator free to revolve (Figs. 18 and 19), each of which is attached to an air compressor in such a manner that when it revolves its compressor will be driven or either air compressors may at times become an air engine and drive the part H6 ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. of the electric motor to which it is attached. Fig. 20 shows the bottom view and Fig. 21 the top view of the combined electro- pneumatic motor standing on end in the shop prior to being placed upon the truck, and Figs. 22 and 23 show the two motors com- plete mounted upon a truck. The following description will make clear the application of the principles and the operation of the different parts of the system. Perhaps I cannot describe the theory and working of the machine better than by employing language which I have previously used, so amplified as to conform to the additional figures given in this paper showing more clearly the interior mechanism of the machine. Fig. 24 represents diagrammatically the working parts of the Fio. 4. DIAGRAMMATIC ARRANGEMENT OF ELECTRO-PNEUMATIC MOTOR. system when a reciprocating type of air compressor is used. Fig. 25 shows a transverse section through the air cylinders, the regulating valves and the individual cylinder valves of the machine shown in Figs. 22 and 23. The rotor R, Fig. 24, is geared to the axle of the car, and by means of crank pin <7', secured in pinion P f also drives the com- pressor cylinder R C, while the stator 8 is free to revolve around the rotor and drive by means of crank-pin C the compressor cylinder 8 C. Both cylinders are piped to air reservoirs located under the car, and are also provided with suitable valves, A, B, C and C' , shown in Fig. 25, which in connection with the pneumatically operated cylinder valves previously mentioned, are manipulated ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. 37 3S ARNOLD: POLYPHASE AND SINGLE-PHASE TRACT10X. from the controller in such a manner as to make them perform their various functions. Thus the entire regulation of the speed of the car is controlled by the air cylinders. For the purpose of making clear the different operations of the system, Fig. 26, showing a speed diagram, has heen prepared, in which on the axis of abscissae D L are represented the different car speeds in per cent of the synchronous motor speed, and the co- ordinate axis A B represents the rotor and stator speeds cor- responding to the car speeds shown on axis D L. The operation of the car may be divided into the following periods : 1. Standing in the Station. In Fig. 24, the rotor R being rigidly geared to the car axle is now standing still, while the stator 8 runs with full synchronous speed, and is thus transferring the full energy of the electric motor through crank C to the compresser cylinder S C, which energy is being delivered in the form of compressed air into the air reservoir. Since the relative velocity between the stator and the rotor is con- stant under all conditions of operation, the speed curves of stator and rotor may be represented by two parallel lines, OCR and A D S, shown in Fig. 26. The origin of the given co-ordinate system represents the period of rest of the car, and, therefore, indi- cates zero rotor speed and full stator speed in a negative or down- ward direction, as the stator is now revolving in an opposite direc- tion from that which the rotor must revolve to drive the car for- ward. If it is assumed that A equals the active torque of the stator, then B, which equals A, will represent the reactive torque of the rotor exerted on the car axle, so that if the car is free to move the reactive torque can be used for starting and accelerating the car. When the car is standing in the station it is held at rest by placing valve B (Fig. 25) by means of the controller, in the position shown ir. full lines, thus allowing air from the storage tanks to enter through opening Q in the direction of the arrow R to passage ZT, which is in communication with the high-pressure valves of the rotor cylinder. The pressure may be thus increased behind the rotor piston to such an extent that it overcomes the effort of the rotor to revolve, thus tending to cause the stator to revolve, while at the same time it holds the car at rest without the use of wheel- brake?. When the car is standing, the stator is running at full ARNOLD: POLYPHASE AND SINGLE.-PHASE TRACTION. 39 synchronous speed and the stator cylinder is drawing in cold air through opening D in the direction of arrow 0, which enters the stator cylinder through the inlet valves shown at the top of the cylinder. The air is delivered from the stator cylinder through the outlet valves into passage H, and may be delivered in the direction of arrow R into opening Q and thence to the storage tanks or into the passage H' for the purpose of holding the rotor cylinder still or supplying it with air in starting. 2 Starting and Accelerating. To start the car the air cushion behind the piston of rotor cylinder R C, Fig. 24, is removed by so manipulating the controller that the exhaust valves shown at the top of Fig. 25 are opened; Fio. 26. DIAGRAMMATIC REPRESENTATION OF OPERATION OF ELECTRO-PNEU- MATIC MOTOR. the air which is being compressed by the stator cylinder is then delivered from passage H into H' , as indicated by the arrow R, supplemented by the stored air from the tanks. The controller is now set at the position of maximum cut-off for the inlet valves of the rotor cylinder, shown at the bottom of Fig. 25. The rotor then begins to revolve and as it accelerates the stator slows down by exactly the same amount that the rotor has increased its speed; as the rotor and car speed increase the controller is gradually moved so that the inlet valves of the rotor cylinder give a smaller percentage of cut-off until the car speed corresponds to the full synchronous speed of the motor, at which time the stator 40 ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. comes to rest. During this period of acceleration the air compressed by the stator cylinder, instead of being delivered to the tanks to lose its heat, is delivered, hot, directly to the rotor cylinder through the passages H and H' , either directly, as indicated by arrow R f in case the valve A is placed as shown in full lines, or through the automatic valve (7, as indicated by arrow S, thence through a pas- sage (not shown) communicating with opening Q. In the latter case the valve A is placed in position A'. The valve G, known as the stator automatic valve, is provided with a spring so set that it maintains a constant pressure in passage H and hence a constant load upon the electric motor. After the air thus delivered from the stator cylinder has done its work behind the rotor piston, it is exhausted cold, owing to the rapid expansion, into the passage L' f and thence in the direction of the arrow N into the passage L leading to the inlet valves of the stator cylinder. Thus a complete cycle is established and the same air may be used repeatedly if the rate of acceleration is such that the rotor cylinder uses all of the air supplied by the stator cylinder and under these conditions no exhaust to the atmosphere from the rotor cylinder will take place. Since all of the air passages and both cylinders are enclosed in a water-jacket, the heat generated while compressing is delivered to the water and extracted by the rotor cylinder when working as an engine, the water performing the double function of cooling the air during compression and reheat- ing it during the process of expansion, thus increasing the efficiency of the combination. Tests already made indicate that this jacket- ing water will remain at a fairly constant and comparatively low temperature. Opening D is known as the cold-air inlet and the exhaust out- let. It is provided with a valve acting against a spring which nor- mally keeps opening D closed to the outside air. In case the volume of air required by the stator cylinder is greater than the amount exhausted from the rotor cylinder, this valve automatically opens and permits the outside air to enter the passage L through the open- ing D, as indicated by the arrow 0. This valve also opens auto- matically to admit air to the rotor cylinder in the direction of the arrow P at such times, hereinafter described, as it may be compress- ing air. The valve is also electrically controlled in such a manner that it can be opened by the motorman when it is desired to operate the car as an independent unit with air alone by means of the rotor cylinders acting as engines. ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. 41 Kef erring to Fig. 26, which graphically represents the period of acceleration, since the electric motor always runs at a constant speed and constant load, it has a constant torque, and, therefore, the vertical distance A between ADS and OCR may be considered as representing the energy delivered by the electric motor. The length of any ordinate extending from D to C represents the proportionate amount of energy derived from the electric motor which is applied directly through pinion P and gear G f Fig. 24, to the propulsion of the car wheel. The corresponding ordinate extending below D to S D represents the proportionate amount of the energy of the electric motor which is absorbed in compressing air through the cylinder S C, which energy, in the form of air, is immediately transferred to cylinder, the R (7, and is utilized in accel- erating the car. In practice, however, since there will be some loss IL transferring the energy from electrical energy to energy in the form of compressed air and back again into mechanical energy, the energy thus lost, whatever it may be, must be drawn from the stor- age tanks and the requisite amount of air from these tanks sup- plied to the rotor cylinder R C in order to maintain the full power of the electric motor upon the car axle during the period of accelera- tion. Should it be desired to accelerate at a greater rate than the full power the electric motor is capable of giving to the car, the ad- ditional energy may be supplied in the form of air from the stor- age tanks through the rotor cylinder, thus increasing the total energy given to the car during acceleration, in which case this total power would be represented for any given instant by a point above line B C. The air thus drawn from the tanks enters through the opening Q and flows in the direction of arrow R into the passage H', and thence to the rotor cylinder. 8 Running Speeds. Assuming that during the accelerating period valve A has been in position A' f the air from the stator cylinder has been delivered through the stator automatic C, and a constant load has been main- tained upon the motor. As soon as the car by the previous processes reached a speed corresponding to the synchronous speed of the motor, the exhaust valves of rotor cylinder R C are held open by setting the controller at a suitable position and the piston of the rotor cylinder now runs free. The electric motor now gives its 42 ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. full power to the car axle and the stator and its air mechanism remain at rest as long as the car runs at the speed corresponding to the synchronous speed of the motor. Since the pressure behind the piston of the stator cylinder is maintained constant by the valve C f the stator will remain at rest only so long as the resistance offered by the car is exactly equal to the power of the electric motor. In case this resistance is less than the capacity of the electric motor, the stator cylinder will automatically reverse and begin to rotate ir the same direction as the rotor is running, and slowly compress air and deliver it to the storage reservoir. In case the resistance of the car is greater than the capacity of the motor, the speed will decrease and the stator automatically reverse and run in an opposite direction from that of the rotor, and will then be operating in the same manner as during the accelerating period. It will thus be seen that no attention need be paid to the stator during the running period, for it automatically takes care of itself. When the resistance of the car is greater than the capacity of the electric motor, speeds above synchronism can be maintained only by supplying the rotor cylinders with stored air from the tanks, and can only be maintained for short distances, or until the storage capacity of the air reservoirs is exhausted. The distance from the line D L to that portion of the line ADS above D L in Fig. 26 represents, at any given speed, the proportionate amount of energy which must come from the tanks and be supplied through cylinder S C. The distance from D L to C R represents the total energy given to the car by the combined action of the electric motor and stator cylinder. 4 Retardation. To bring the car to rest, instead of applying mechanical brakes to the wheels in the ordinary manner, thereby dissipating the entire stored energy of the car in the form of heat, this energy is saved in the form of compressed air to assist in starting the car, by set- ting the controller in such a position that the rotor cylinder com- presses air and delivers it into the storage tanks. Any desired rate of retardation can be secured by throttling the delivery pas- sages from the rotor cylinder by means of valve B, Fig. 25, by mov- ing it toward the direction indicated at B'. When the valve is in the position B' f the passage H' is brought into communication with the automatic valve C' , so set that it will release just before the slipping point of the wheels is reached. The kinetic energy of the ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. 43 car can thus be all absorbed by means of the rotor cylinder and the car brought to rest without wheel brakes, although such brakes are supplied for emergency, but need not be often used. 5. Reversing. When it is desired to run the car backward for short distances the electric motor is not disturbed, and the power is furnished from the rotor cylinders acting as engines; but if it is desired to run backward for any great distance, the current is thrown off the motor, the stator engine is reversed and the stator is brought to speed with the air, when the current is again thrown on to the motor, and the cycle of operation is the same as when running forward. A detailed description of the valves may now be of interest: DESCRIPTION OF VALVES. Eef erring to Fig. 25, the lower valves are termed the high-pres- sure valves and act as inlet valves when the machine is running us an engine and as outlet valves when the machine is running as ji compressor. The upper valves are the outlet or exhaust valves when the ma- chine is running as an engine and the inlet or admission valves vvhen the machine is running as a compressor. Both valves are shown in detail drawn to a larger scale in Fig. 27. In Fig. 27 (bottom valve) part 15-79 is the valve proper and is of steel; it is carried in a brass guiding case, 15-464, screwed solidly in the retaining walls of the cylinder. Into this seat 15-464, is screwed a brass guiding piece, 15-463, which serves the double purpose of guiding the solenoid plunger, 15-78, and as a chamber for the solenoid coil X. In the center of the valve 15-479 is bored a round, true socket or port chamber, into which fits a round plunger or piston, this being an integral part of the solenoid core, 15-78. This core carries a flange, also integral with it, against which the spring 15-80 rests, the other end of the spring resting against 15-463. Surrounding the solenoid core 15-78, is placed * solenoid coil X which, when energized, draws 15-78 downward and with it the piston which fits into the port chamber, 15-79. Valve 15-79 is provided with one or more ports, a, drilled into its face and terminating in the central port; chamber. It is also provided with radial ports, I, terminating in the port chamber. The portion of the solenoid core, 15-78, which enters the port Chamber is also provided with channels, c, drilled longitudinally, 44 ARNOLD: POLYPE ASE AND SINGLE-PHASE TRACTION. r ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. 45 which are connected with radial openings d and e. Under normal conditions of operation the space between walls / and g is filled with air. Valve 15-78 is round, and the portion h is slightly less in diameter than portion j t the latter sliding air tight in 15-464, so that if pressure is admitted through ports b f d, c and e, into the chamber behind 15-79, the pressure will act upon the portion ; of the piston or that portion which has the largest diameter and consequently the greatest area, and the valve will be held tight against its seat. The operation of the valve is then as follows : When working as an admission valve for the engine, current is sent through the solenoid coil X, which causes the solenoid core 15-78 to be pulled downward, thus withdrawing its upper portion which fits into the port chamber, causing port a, normally closed by 15-78, to be opened, thus allowing the air to flow from the in- terior of portion / out through ports e f c and a into the cylinder. While this air is thus permitted to escape into a larger opening, it is not lost for it must act upon the piston before escaping to the at- mosphere. Since portion h is smaller in diameter, and, therefore, of less area than j, the high-pressure air surrounding the valve will force 15-79 downward, thus opening the main port previously closed by 15-79, allowing the high-pressure air to flow from the high-pressure air chamber into the cylinder. Port 15-79 will re- main open as long as current is held upon the solenoid coil X; but as soon as current is turned off from the solenoid coil, spring 15-80 forces 15-78 upward, thus closing port a, and allowing air to again enter through ports, ~b, d, c and e into the chamber behind 15-79, which forces it upward to its seat on account of the larger diameter and consequently larger area of portion /. By sending current through the solenoid coil at suitable intervals by means of the col- lector rings previously referred to, the valve can be made to open and close and act as an admission valve when the machine is operating as an engine, using air for its driving power and utilizing the air to be used in the cylinder of the engine afterward. The solenoid feature of the valve, therefore, acts only as a pilot and requires but little energy, which can be supplied from the line or from any secondary source, such as a small motor-generator or a storage battery. When acting as an outlet valve for the compressor, no current is sent through the solenoid coil, and 15-78 is held in its upward position by the spring 15-80, thus, as before, admitting high-pres- 46 ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. sure air through ports 5, d, o and e f behind portion j of 15-79, the air thus supplementing spring 15-80 to hold 15-79 against its seat. The valve will thus operate automatically like the outlet valve of an ordinary air compressor whenever the pressure in the cylinder is sufficiently great to overcome the combined action of spring 15- 80 and the air pressure behind 15-79. Kef err ing now to the upper or low pressure valve, Fig. 27, part 15-466 is a brass seat normally screwed into the casting of the cylinder. In the drawing these valves on the stator cylinder are shown screwed at their bases into brass bushes which have nothing to do with the valves, but were used on the stator side on account of a mechanical defect in the stator cylinder casting. As in the case of the high-pressure valves, part 15-467 is a brass seat screwed into the cylinder casting, and screwed on it for mechanical protection of the solenoid coil is a cast-iron part 15-460. On the interior of 15-467 fits piston 15-81, which is screwed on to valve seat 15-83, thus making parts 15-81 and 15-83 practically integral so far as operation is concerned, they having been made in different parts only for convenience in assembling. Part 15-83 is provided with a round port chamber into which ports ra and n enter in such a manner that they can be closed or opened by plunger o. Plunger o is made of steel and is firmly secured to plunger rod p and provided with ports q extending 'com- pletely through it. To the upper portion of rod p is attached the solenoid core r. Solenoid r and with it rod p and plunger o are normally held in their upward position by means of spring 15-87 resting against part s which is screwed into part i, the latter form- ing the path for the lower part of the magnetic circuit created by the solenoid coil. The chamber between walls u and v is the ex- haust or low-pressure chamber. The action of the valve when in operation as an exhaust valve when the stator cylinder is operating as an engine is as follows: Spring 15-87 normally holds plunger r and with it rod p and plunger o in their upward position, thus causing plunger o to close the ports ra. When it is desired to operate the valve, and thus exhaust air from the cylinder, current is sent through the solenoid coil, which causes plunger r to be drawn into the solenoid coil and downward against the resistance *of the spring 15-87, thus carry- ing the stem p and plunger o to the downward position and open- ing the ports ra so that the air behind the piston of the stator cylinder can flow freely through ports ra up through the interior ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. 47 of the port chamber inside of 15-83 and enter the space above piston 15-81. As the piston 15-81 is larger in area than the valve FIG. 28. WIRING DIAGRAM OF LOCOMOTIVE " PHCENIX." 15-83, the air thus admitted above the piston causes it to press downward, thus carrying with it and opening valve 15-83, which 48 ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. will remain open and allow the air to exhaust from the cylinder into the exhaust or low-pressure space so long as current remains upon the solenoid coil. When the valve is used as an inlet valve for the compressor, no current is sent through the solenoid coil and Fte. 20. LONGITUDINAL SECTION OP LOCOMOTIVE "PHCENIX." the valve works mechanically, due to the suction of the piston in the cylinder, which draws valve 15-83 and piston 15-81 downward against spring 15-84, the latter heing only of sufficient strength to normally hold valve 15-83 against its seat. The valves when used ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. 49 for the purpose of operating the air cylinders as engines are con- trolled by means of revolving commutators and suitable circuits FIG. 30. TRANSVERSE SECTION OF LOCOMOTIVE " PHCENIX." in combination with the controller, all as shown diag: ainmatically in Fig. 28. ELEC. RYS. 4. 50 ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. Since it was impracticable for me to get the manufacturer to build a single-phase motor for my first machine at the time the order was placed (January, 1901), I was compelled to utilize the parts of a three-phase motor and have it built as such in order to get it at all. For this reason the machines were built as three- phase machines, and provision was made in the locomotive for run- ning them three-phase when it was desired to do so during the preliminary tests in the carhouse. The diagram, Fig. 28, therefore, shows the connections necessary for running three-phase, but all tests on the line were made run- ning single-phase. Figs. 29 and 30 show longitudinal and transverse sections of the locomotive "Phoenix/' outside views of which are shown in Figs. 14 and 15. FIG. 31. ELEVATION OF ELECTRO-PNEUMATIC TEUCK WHERE ROTARY AIR MOTORS ARE USED. This machine was similar to " Number 2," the one destroyed by fire, both being equipped with the truck and motors shown in Figs. 22 and 23, the only difference being in the fdrm of the cab, the type of the transformer and the location of the auxiliaries in the cab. In both cases the current came from the working con- ductor directly into the terminal on one side of the stationary trans- formers while the terminal of the other side was grounded. The secondaries of the transformers led to the collector rings of the stator parts of the motor and supplied current at about 250 volts. In order to permit the machines to operate as independent units by using air, each was supplied with a motor generator and a stor- '!re battery to supply energy for operating the valves of the engines. ARNOLD: POLYPHASE AND SINGLE-PHASE TRACTION. r,l While the development of this system has proven to be a most interesting and fascinating field of work, I regard the machine in its present form as somewhat complicated for coninierciali 'applica- tion, for like most all new mechanical problems the first designs are much more complicated than subsequent experience finds necessary. By the development of suitable rotating air machinery the system is capable of great simplification, as by this means all of the above mentioned reciprocating parts, valves with their revolving collector rings and connections, together with the motor generator and battery disappear. The machine would then take the form shown in Fig. 31 and be controlled entirely by two valves similar to those shown at the top of Fig. 25. If the motors then be designed for the working pressure of the line, the transformer will also disappear from the car; and as the current is not manipulated in controlling the speed of the car, the use of high-pressure motors becomes practicable. What the commercial value of the system is will depend upon the results shown by future tests, and on the relative merits of the various single-phase systems that have been developed since the announcement of the principles of this system were made public at the Great Barrington Convention of the American Institute of Electrical Engineers in June, 1902. Whatever its value may be commercially, I believe its influence in stimulating others to greater effort along new lines cannot be denied, and that the art of electric railroading is one step nearer its final solution than it would be today had my efforts not been exerted in this attractive field of achievement in which I have publicly, 8 and often unsupported, proclaimed my faith in the ultimate supremacy of the alternating-current motor for railway work. 3. See Transaction* American Institute of Electrical Engineers as fol- lows: Joint meeting with the British Institution of Electrical Engineers, Paris, August 16, 1900; Niagara Ealls Convention, August 24, 1P01; Great Barrington Convention, June 19, 1902 : New York Meeting, Sept. 26, 1902 ELECTRIC TRACTION ON BRITISH RAILWAYS, BY PHILIP DAWSON. INTRODUCTORY. The introduction of electric traction on British railways is a subject of great interest, but can only be discussed very briefly in this paper. The position of our railways is one which is beginning to make all those connected with these interests fully alive to the necessity of improvement, both as regards increasing their freight and passenger traffic and reducing the working expenses. Owing to the stringent regulations imposed by the Government, and the very densely-populated districts which the railways traverse, the capitalization of English lines is exceedingly heavy, as the follow- ing figures clearly show: CAPITALIZATION AND MILEAGE OF EAILWAYS IN THE UNITED KINGDOM IN 1901. Debenture stock 304,577,862 Preferential share capital 310,819,740 Guaranteed share capital 114,293,436 Ordinary share capital 454,379,107 Total capitalization 1,184,070,145 Double or more lines, length of route 12,272 miles. Single line 9,806 " Total length of route 22,078 miles. The ever-increasing taxation, as well as of the competition which the railway companies are beginning to feel in consequence of the rapid, introduction of electric traction on tramways in and around [52] f OF THE ( UNIVERSITY jl V Ns^ OF ELECTRIC TRACTION ON BRITISH RAILWAYS. all the large cities of Great Britain, are some of the many reasons which, notwithstanding the fact that the total number of passen- gers carried, as well as the total merchandise and goods conveyed, has been more or less steadily rising, has contributed, as will be seen by the following table, to reducing the percentage of net re- ceipts to total paid-up capital: SUMMARY OF RAILWAY RESULTS OF THE UNITED KINGDOM FROM 1850 TO 1901. Year. Total number of passengers carried (exclusive of season ticket-holders.) Weight of goods and minerals conveyed, Tons. Percentage of net receipts to total paid-up capital. Percentage of working expen- diture to gross receipts. 1850 72 854 422 1860. 1870 163,435,678 836 545 397 89,857,719 4.19 4 41 47 48 1880. 1885. 1890. 1895. 1899. 1900. 1901. 603,885,025 697,213,031 817,744.046 929,770,909 1,106,691,991 1,142,276,696 1,172,895,900 235,305,629 257,288,454 808,119,427 834.230,991 413,623,025 424,929,518 415,593,441 4.38 4.02 4.10 8.80 8.61 8.41 8.27 51 53 54 56 59 62 68 The question, therefore, arises as to what our railways can do in order to increase the ratio of net receipts to the total paid-up capital. In my mind the answer is, that their salvation lies in the judicious adoption of electric traction. Railways have to deal with three classes of traffic. 1). The short-distance, suburban, and interburban traffic in the neighbor- hood of our large towns and between the large centers which, in many parts of England, lie so close together; as for instance, such cities as Bradford, Leeds, Halifax, Blackburn, and the numerous towns on the borders of Lancashire and Yorkshire. 2). The long- distance, main-line traffic. 3.) The goods traffic. As regards the suburban and short-distance interurban traffic, there is no doubt that electric traction will be a great benefit to the railways, and owing to the dense population of this country which makes the building of new roads expensive and difficult, and the very extensive network of railways which already exist, the steam railways in Great Britain are in exceptionally favorable conditions to benefit by electrification. As regards long-distance, main-line traffic, there may be indi- vidual isolated cases where, after electric traction has been intro- duced on suburban lines it may be found advisable to extend it to the main lines. 54 DAWSON: ELECTRIC TRACTION ON BRITISH RAILWAYS. English railway companies have progressed very considerably of late; the track construction is as good as any to be found in the world, and the locomotives, signaling apparatus, and rolling stock are, as far as steam traction is concerned, beyond criticism. At the same ; tirrie, as already pointed out, the competition of electric tramways, and the demands of the public for increased facilities of locomotion, call for a development on entirely new lines. As trustee to many millions of the public for money, it is evident that no railway company can take any action or adopt any novel system involving considerable expenditure, except with the greatest care and after most thorough investigation. Necessarily, the railway companies would prefer to be perfectly certain of the results before taking any important steps, and to know what has been achieved financially by electrification of other lines. The figures of the cost of operating, and receipts on the tube lines and on the Liverpool Overhead, are very instructive, but at the same time, owing to the different conditions under which they are con- structed and operated, their results do not necessarily apply to all cases. It will be some years yet before reliable information is available of the results obtained on the Lancashire and Yorkshire, and North Eastern, and under these circumstances the experience of the Mersey railway is most useful. The conversion of the Mersey line was taken in hand and was in progress during 1902 and was completed in May of last year, and, therefore, it could hardly be said to be in full working order during the latter half of last year, for which accounts were available. Under these circumstances it is not necessary to go into these accounts in detail, but it will be interesting to note that the three minutes' service adopted has caused their train mileage to be in- creased from some 155,000 miles to over 401,000 miles in six months. That this result was justified is shown by the fact that the number of passengers has increased from 2,844,708 to 4,153,777, and the results show that the traffic was almost entirely made up of first and third-class passengers, the second class having to be greatly diminished, as might have been anticipated. The results consequent on the electrification of the Milan-Varese line of the Mediterranean Eailway Company are no less surprising. For six months ended June, 1903, the total number of passengers carried was 2,977,812. During the whole year 1900, when the line was entirely operated by steam, the total number of passengers carried was 2,768,541. DAWSON: ELECTRIC TRACTION ON BRITISH RAILWAYS. 55 SUBURBAN AND SHORT-DISTANCE INTERURBAN TRAFFIC CON- DITIONS. The position of railways with respect to suburban traffic varies considerably with their location. In some cases there has been a decrease both in the number of passengers carried and in the gross receipts, due in a large measure to the competition of paral- leling electric tramways. In other cases there has been little or no change, whilst in others again, particularly those serving the London suburbs, the requirements of the traveling public are so great that the steam railways have never been able to cope with them, and consequently the presence of competing tramways has not as yet been seriously felt. The electrification of the tramways in all the big towns, as well as the construction of a large number of so-called light rail- ways connecting the various towns, and the activity shown, both by the local authorities and private companies in promoting new lines, is rapidly bringing matters to a crisis. The speeds allowed on tramways are consequently being increased and there seems but little doubt that on a large portion of the electric lines the average speeds of from 12 to 15 miles an hour may be allowed at no very distant date. This increase in speed, as well as the fre- quent service given by electric tramways, will make them most serious competitors to steam railways unless their local time-tables are considerably modified and improved, both as regards frequency and average speed, and this will only be rendered possible by the in- troduction of electric traction. Thus there is an urgent need for a revision of the mode of transport adopted on railways, in one case to turn the ebbing tide to traffic, and in the other to satisfy the claims of a public anxious to travel but unable to do so because of the congested state of the lines. If the railways allow competing lines to proceed unmolested, the problem will be solved in a manner extremely detrimental to the former. The congestion will steadily diminish by reason of the traffic being diverted from the railways by the opposing interests, which will give the facilities so urgently needed at the present day. The suburban traffic of the railways has been growing rapidly with the suburbs, and as it is largely concentrated at certain sta- tions instead of being uniformly distributed, it is naturally very congested. The state of affairs is further complicated by the inter- 56 DAW SON: ELECTRIC TRACTION ON BRITISH RAILWAYS. mingling of main line and suburban traffic in the termini, owing to lack of space which prevents their being kept entirely separated as they should be; and by the delay due to the impossibility at present of getting in and out of the terminal stations expeditiously. The electric tramways, though strong and healthy, are at present a young growth, but they are extending with amazing rapidity in all directions, with the result that they are pressing hard upon the railways, even in the matter of comparatively long-distance suburban traffic; this is a branch which is essentially a province of the railways, and one in which they should easily maintain their supremacy if they are properly equipped to satisfy the require- ments of the situation. The great need of the railway companies in respect of their local traffic is both to increase their service and to improve their methods generally, and there is only one practical way of doing this, viz., electrification. Under the present conditions it is impracticable to increase the frequency of suburban trains, first, because of the cost of handling such an increase by steam locomotives, and secondly, because of the mutual interference of the main line and suburban traffic. In other words, the lines are at present being worked very near to the limit of their capacity as far as steam is concerned. In the case of suburban trains the amount of time occupied in starting and stopping forms a very large proportion of the whole time spent on the journey, so that a very considerable saving would be effected if this waste of time could be reduced. At the same time it is essential that this should be supplemented by more efficient methods of taking up and detraining passengers, which process, under the present circumstances, entails an unnecessary waste of time at stations. Attempts have been made in various cases to avoid the electrifica- tion of steam lines by the adoption of special locomotives giving exceptionally high rates of acceleration, but the failure of these has only served to emphasize the necessity for electric traction, for by that means alone is it possible to obtain really high rates of accelera- tion. Experience has conclusively shown that the only really satis- factory method of handling suburban traffic is by means of electric traction, and it is a great pity that hitherto the question has not been faced with greater boldness by those concerned. There are such certain benefits to be derived from the electrifi- cation of steam suburban lines that the only wonder is that more DAWSOy: ELECTRIC TRACTION ON BRITISH RAILWAYS. 57 progress has not been made. As already stated, the main advantage is that the use of electric power by permitting very high rates of acceleration enables the frequency and consequently the carrying capacity of the service to be largely increased, while at the same time the working expenses per train mile are decreased. Also with electric traction the delays due to signals are of less importance, owing to the quickness with which an electric train can get under way. Over and above the foregoing, the cleanliness of electricity and the consequent enhanced comfort of the passengers is a considerable factor in increasing the popularity of the line. Not a few railway companies have expressed their opinion that they have no objection to the tramways taking their suburban traffic. According to them this branch of the service is both costly to maintain and difficult to manage, and at the same time it is not a profitable source of revenue, and they appear to be quite content to drop it altogether and fall back upon the lucrative main-line traffic. There are, however, two objections to this. In the first place, it is extremely doubtful whether any railway would be allowed by the Government to drop its suburban traffic com- pletely. In the second place, there is too much capital tied up in this branch to render it possible to dispense with it entirely. Few companies could afford to let such a large amount of capital lie idle, and there is no reason why they should. In my opinion, there is not a single railway company that could not operate its suburban traffic in the neighborhood of most of our large manufacturing towns at a substantial profit if it were to be electrified, and in most cases the profit resulting therefrom would be more than sufficient to pay the interest on the necessary capital outlay called for by the change of motive power. As soon as the railways electrify their suburban lines, they will hold a very strong position against the attacks of competing tram- ways and light railways, since in the matter of speed they will have all the good points of the tramways, without the disadvantage of having to operate in crowded thoroughfares; the greater distance between the stops will naturally permit a far higher schedule speed to be maintained, and the higher the speed the railways are able to offer to the public, the shorter will be the distance of the journeys for which the tramways will prove more convenient. That there is room for great improvement in the railway service, and that there is a larger amount of latent traffic to be secured pro- 58 DAW SON: ELECTRIC TRACTION ON BRITISH RAILWAYS. vided the railway companies go to work in the proper way, is clearly shown from the statistics giving the number of times the population of the large cities in Great Britain are carried annually. The evidence given hefore the Koyal Commission for London Traffic by Mr. Edgar Harper, the statistician of the London County Council, shows that whereas, in 1867, the population of London was carried 22.7 times, in 1901 it was carried 128.7 times. These figures only deal with the traffic in the London area, and do not in- clude the passengers brought in by suburban trains. It must be noted that these figures do not include all the omnibus lines. It is interesting to note that the number of journeys per head of population in London is at present small compared to that in many other large cities, as will be seen by the following figures : London, 1901 129 journeys. Glasgow, 1901 174 " Liverpool, 1901 187 " London (Mr. Harper's estimate), 1903 200 " Berlin 223 " Greater New York 320 " Facilities for traffic always create traffic, and as facilities are improved traffic will not only be actually but also relatively greater. This is shown by the following figures. INCREASE IN NUMBER OF JOURNEYS PER HEAD OF POPULATION. Greater London. Greater New York. 1867 23 1860 47 1870 27 1870 118 1880 55 1880 182 1890 92 1890 283 1900 , 126 1900 320 1901 129 1903 (estimated) 415 1903 (estimated) 200 From these figures it will be seen that the population of London is carried only half the number of times that the population of New York is. DAW SON: ELECTRIC TRACTION ON BRITISH RAILWAYS. 59 CONDITIONS TO BE FULFILLED FOR A SYSTEM OP ELECTRIC TRAC- TION TO BE SATISFACTORY. There are certain conditions which require to be fulfilled before any system can be considered capable of giving satisfactory results, and these conditions are briefly set forth below. 1). Should moving machinery be found necessary in the trans- forming of sub-stations, these stations must be as few in number as possible. The apparatus used in sub-stations should be such as to require but little attendance and should be efficient at all loads and capable of dealing for short periods with very heavy overloads. 2). The number of conductors required to supply current to trains should be as few as possible, and should be capable of un- limited extension, and must not interfere with the tracks; hence the use of a third rail is not possible. 3). It must be possible to collect from a single conductor sufficient power to haul one or more fast trains in service on the lines between the feeding points of the conductor. 4). It is very desirable that the system should be applicable to main line as well as suburban traction, and that it should be possible to utilize at least two working pressures a low pressure where found necessary in or near the station, and a higher pressure outside. 5). The system should be such that the trains can be operated at any speed required, and thus be capable of making up lost time. 6). All controlling apparatus must be of the simplest character, and such that no skilled labor is necessary to operate the trains; also there must be no dangerous high pressure anywhere accessible to either railway officials or passengers. 7). If alternating currents are used it is essential that the power factor be high and that the motor be capable of giving an acceleration equal to that obtained with the best series-wound direct- current motors at present in use. 8). In certain cases it might be advantageous for the motor to be constructed to return current to the line, but in any case it must be constructed to reverse and to be used for braking purposes. OVERHEAD CONDUCTORS. The doubts that have been expressed as to the feasibility of adopt- ing overhead wires on the lines where steam locomotives are run- ning, and the objections which have been urged against their use QQDAW80N: ELECTRIC TRACTION ON BRITISH RAILWAYS. on this account, are, in my opinion, quite groundless, and there is no reason to anticipate any trouble from this cause. The engineers of the Valtelina railway informed me on the occasion of my last visit that they had never experienced the slightest difficulty in re- spect to the two overhead 3000-volt conductors which have been in use for over two years on that line, in spite of the fact that steam locomotives burning soft coal are continually passing over the line, and that the aerial conductors in some places have to pass through tunnels from the roof of which large quantities of water are always descending. The conditions I have mentioned as being those with which a traction system has to comply appear to be exceedingly difficult to fulfil, and the only system which could possibly comply with the conditions is a single-phase one. As long as electric traction was applied only to tramways or lines with few or no complicated junc- tions, and on which only electric trains operated and there was no steam service, the continuous-current railway motor has given per- fect satisfaction. But this type of motor has its limitations, and the necessity for dispensing with third rails and using a single high-tension over- head conductor, has recently induced manufacturers and directors to investigate the question thoroughly and experiment upon the possibility of constructing a really reliable single-phase motor. As might have been expected, as soon as there was a real demand it was not long before an article was produced to supply it. Aided by the experience obtained in the design of all types of electric machinery, consequent on the enormous extension of the applications of electricity that has taken place during the last few years, a satis- factory alternating-current single-phase motor has now been de- veloped. The single-phase motor at present developed may be divided into two classes, the " series " type which has been investigated and brought out in Europe by Dr. Finzi, and in America by Mr. Lamme, and the " repulsion " type, both in the original form as investigated many years ago by Prof. Elihu Thomson, and the " compensated repulsion " form as theoretically studied and discussed by Mr. Latour in France, and practically investigated by Messrs. Eichberg and Winter. The restriction as to the pressure at which it is feasible to operate a continuous-current motor is a great drawback to its employment on electrified railways, for it means that the sub-stations must be DAW SON: ELECTRIC TRACTION ON BRITISH RAILWAYS. 61 placed close together. The use of an alternating-current motor Introduces a considerable saving in the cost of distributing mains and conductors on account of the high voltage which can be utilized, and not only are the sub-stations fewer in number, but they are smaller and cheaper in first cost, maintenance, and attendance, owing to the absence of rotating machinery. The great advantage of the single-phase motor in dispensing with the necessity for a third rail and enabling a single small high-ten- sion overhead conductor to be used instead is further enhanced by the fact that in its operation the rheostatic losses involved in the control of continuous-current motors are avoided. An additional gain in efficiency also results from the better distributions of the sub-stations and the decreased losses at these points of distribution, whilst in some cases a line voltage can be employed which is suffi- cient to dispense entirely with transformer sub-stations. Also owing to the increased efficiency of the whole system the amount of plant required at the power station is less than would be the case for a similar direct-current system. A very important point about the single-phase motor is the fact that it can easily be adapted to operate upon direct-current cir- cuits, a simple switching device being all that is necessary to make the change. Besides the solutions mentioned above, there have been various more or less unpractical solutions suggested, such as that proposed by the Oerlikon company and now being tried by them. CONDITIONS TO BE FULFILLED BY CONDUCTORS BRINGING CURRENT TO TRAIN. The conditions governing the type and position of the conductor from which the motor cars or locomotives obtain their supply of power in the case of most of the steam suburban railway systems of Great Britain are very different to those which apply to ordinary tramways, newly-built electric urban or suburban systems. On most suburban systems the traffic is very dense and either local long-distance passenger, or goods' trains are operating over the lines for the greater portion of the 24 hours for six days a week. At many junctions the traffic is largely increased by the numerous other companies who use that station and there is but little time available for keeping in proper repair the existing track rails and points and crossings, which are very congested. As things stand at 62 DAWSON: ELECTRIC TRACTION ON BRITISH RAILWAYS. present the tracklayers have the greatest trouble in finding time to keep, the permanent way in proper condition, and they are greatly hindered in their work owing to the very frequent service of trains. Under these conditions, the introduction of a third " live " rail is practically impossible. Even if it was guarded it would consti- tute an additional and constant source of danger to the permanent way men who, besides having to avoid the passing trains, would also have to keep clear of the " live " third rail. Furthermore, it is highly probable that a " fourth " or return rail, such as has been adopted on the Metropolitan district and the Lancashire and Yorkshire, would be found necessary in order to keep within the 7-volt drop in the return circuit required by the Board of Trade. It might be possible to sectionize the third rail, and arrange so that no portion of it was alive except while a train was actually passing over it ; but the necessary automatic switches would intro- duce most undesirable additional complications, whilst there would always be a possibility of their failing to work so that it would not do for the men to treat the rail as quite harmless. In any case, with the complicated track work existing at many large junctions it is probable that there would be no space available to place the third and fourth rail, owing to the numerous signal wires and the rods used to operate the points. The consequences following even a slight derailment would be most serious and the danger of fire due to short-circuits thus in- curred, would be very great, not to mention the danger of electric shocks to passengers and the entire stoppage of the service for a considerable time, while the damage to the third rail was being made good. These considerations have led a large number of railway man- agers and engineers in the United Singdom, on the Continent, and in America, to the conclusion that the idea of using any " live " rail conductor installed at or near the level of the track rail must be discarded. The onty other alternative is to employ an overhead conductor. From a careful study of the conditions which have to be fulfilled, I have come to the conclusions embodied in the following: 1). The conductor must be overhead. 2). The conductor must be as far as possible at a uniform height above the track rails and have no sag. DAW SON: ELECTRIC TRACTION ON BRITISH RAILWAYS. G3 3). The conductor must be supported in such a way that it is practically impossible for it to fall down on the track and get in the way of the train, even in the event of its breaking. 4). The current collector must be light and require little or no attention. Any wear must take place mainly on the collector and not on the overhead conductor. 5). The collector must be such that it cannot slip or slide out of contact with the conductor. 6). In the event of the collector fouling any portion of the over- head work, the collector should give way and not the overhead work. 7). The collector must be cheaply and easily replaceable. 8). The overhead conductor must be connected to safety devices that will automatically cut it out of circuit the instant any breakage occurs. 9). The insulation of the conductor and collector must permit the use of very high pressures, say, up to 10,000 volts. I have designed a form of construction which I think will meet all requirements and which will obviate any interruption of service taking place. Furthermore, I would propose, as far as possible, not to use steel and iron except for poles or brackets and to avoid the employment of galvanized wire or hooks in any form or shape whatever. The supporting wires would be stranded wire, composed of either steel covered with an outer layer of copper rolled onto it, or else composed of phosphor or silicon bronze wire; the main con- ductor from which current would be collected would be of hard- drawn copper and of a diameter of at least one-half inch ; the sup- ports of this wire should not be more than four feet apart and, there- fore, it would be possible to hang this wire in such a way that to all intents and purposes it would be absolutely parallel to the track rails. CURRENT COLLECTOR. The question of the form of current collector or trolleys to be used is one which will have to be most carefully considered. The Oerlikon company's type of trolley loses most of its special ad- vantages when the conductor is suspended from above, over the tracks. With the wire in this position it acts almost exactly like an ordinary sliding bow, and is in no way superior to that type of collector. The chief merit of the Oerlikon trolley lies in its wide range of movement, and in the fact that it can be arranged to make 04 DAW SON: ELECTRIC TRACTION ON BRITISH RAILWAYS. contact with the conductor either on top, underneath, or at the side. But necessarily the. first and last positions require a special form of construction of the overhead work, and are quite impossible where the conductor is suspended from span wires, whether longitudinal or transverse. For main line work where there are no complicated crossings or sidings, it is quite possible that the form suggested by the Oerlikon company may be adopted with the greatest advantage. Bow trolleys may be divided into two classes, one of which is the ordinary scraping type, as used by Messrs. Siemens & Halske, and which is the more common. With the operation and construction of this trolley I am fully acquainted. Such bows have been running for many years, the soft metal on the top of the bow which make the contact preventing wear of the trolley wire. The contact piece is easily replaceable when it wears out. The other type of trolley is that designed and constructed by Messrs. Ganz & Company and used by them on the Valtelina line; this trolley instead of a scraping bow has a roller mounted on ball bearings. This type is considerably more expensive than the scrap- ing trolley and I do not see any necessity for the additional com- plications introduced by the use of a revolving roller. In connection with trolleys the question may arise as to whether any difficulty is likely to be encountered from the high speed at which the trolleys will run along over wire, but there is no reason to anticipate any trouble on that account. In the experiments carried out on the high-speed experimental electric railway between Berlin and Zossen, the bow was only pressed against the wire by a pressure not exceeding from 3 to 4 kgs, whereas the ordinary trolley has to be pressed against the wire with a pressure between four and five times as great: this smaller pressure is of course ad- vantageous as it reduces the wear and tear on the trolley wire and makes it possible to have a much lighter trolley construction than would otherwise be the case. In my opinion a trolley of the " scissors " type would present many advantages. The contact bar could be made at least as long as the whole width of the carriage and this would allow considerable latitude in the position of the overhead wire. In situations where there was not sufficient room for the conductor to be suspended over the center of tracks, it could be diverted to one side, and increased head room thus be obtained by reason of the curvature of the top of the carriage. In this way it would be quite possible DAWSON: ELECTRIC TRACTION ON BRITISH RAILWAYS. 6,T to place the conductor at an altitude not greater than the highest point of the carriage roof, and thus obtain the necessary clearance. In places where it might be considered inadvisable to place a bare conductor under very low bridges, this portion might be made " dead/' electrical continuity being obtained by insulated cables. With this arrangement a dummy trolley wire would be provided for the trolley to run on whilst passing under the bridge. There is no reason to fear that a " scraping " contact would not be satisfactory, since there is ample evidence to the contrary. Many years' experience with third-rail working has demonstrated that very heavy currents can be collected in this way, and with the single-phase high-tension system the current per trolley would be very small; in fact, it would probably be considerably less than the amount which is frequently collected by small trolley wheels in ordinary practice. ADVANTAGES OF ELECTRIC TRACTION. There are several further advantages possessed by the modem method of traction which render it greatly superior to steam haul- age, quite apart from the fact that the high acceleration demanded for the proper operation of suburban traffic can only be obtained from the use of electric power. In the first place steam trains have to carry their own power; that is to say, a locomotive must not only be able to haul a certain weight of train, but it must carry coal, and machinery to consume that coal and convert its heat energy into tractive energy "en route," and a steam locomotive is a most uneconomical instrument for transforming heat into work for traction purposes. All this adds to the weight of the train and greatly increases the weight to be hauled per passenger. In the case of some of the trains on the suburban systems serving London, the locomotives weigh over one-third of the total useful weight of passenger coaches hauled. In the modern electric system the heavy locomotive is replaced by a comparatively light motor car, and energy is generated under the most economical conditions at a certain power station from which it can be transmitted many miles with but slight loss. The wear and tear of the permanent way, particularly at junc- tions and crossings, would be considerably less with electric trac- tion than with steam traction, as in the former there is not that tendency to roll or pitch which exists in the case of steam loco- EIEC. RYS. 5. (>6 DAW SON: ELECTRIC TRACTION ON BRITISH RAILWAYS. motives and which is due to the movement of their reciprocating parts. Electric trains with two or more motor cars on the multiple-unit system have the great advantage of distributing the weight of the train more evenly over the track and also of permitting a smaller weight on each driving axle than would be the case with a loco- motive, owing to the larger number of driving axles. A steam loco- motive has to be heavy enough to give sufficient weight on the driving wheels to haul the heaviest train up the steepest gradient. at the highest required speed. This concentration of the weight is very detrimental to the track, and the bad effects are accentuated by the pounding action caused by the reciprocating motion of the engine. For the same weight on the driving wheels an electric motor can exert a much greater tractive effort than a steam engine, because the electric motor exerts a constant torque upon the driving wheels, whilst the steam engine does not. In the case of steam locomotives, the ratio of the maximum tractive effort to the weight on the driving wheels is not much above 16 per cent, whereas experience has shown that with electric traction this is increased to from 25 per cent to 30 per cent. The cost of operating electric trains will also be reduced by the fact that only one man is required, that is to say, only a driver in- stead of both a driver and a fireman. The Board of Trade should take no exception to this, as should the driver be incapacitated in any way, the method of control employed is such that it automatic- ally brings the train to a standstill. The men for operating these trains need not be mechanics, and the work will be cleaner and nicer, and, therefore, sought after, as in the case of electric trains, the driving cabin is entirely inclosed and perfectly clean. Electric traction is much more flexible than steam, and trains can either be split up into units of one or two cars or joined up into trains, the length of which is only limited by the length of plat- form available. There are other advantages, but the crowning one is certainly the much higher average speed due to rapid acceleration and the econ- omy of power and labor, as well as the reduced cost of production which is everywhere effected. An incidental advantage in favor of the electric motor as com- pared to the steam engine is that the former can stand an amount DAWSON: ELECTRIC TRACTION OH BRITISH RAILWAYS. G7 of continuous service and hard usage which would be impossible with the latter, besides having far less internal friction. A most important benefit resulting from the use of electric trac- tion is the diminution of the present difficulties due to the lack of accommodation in termini. Mr. Aspinall, the general manager of the Lancashire & Yorkshire railway, stated to me that the recent electrification of the line from Liverpool to Southport will not only double the carrying capacity of the line but will also practically double the terminal accommodation. How this is brought about is easily seen when we consider the time wasted at present in getting a steam train out of a station after it has once entered. First the line has to be cleared to allow another locomotive to back on to the train in readiness to take it out, which it does. Then before another train can be brought in the line has to be cleared again so that the original locomotive which brought in the first train can run out. These various manceuvers occupy a considerable amount of time, besides necessitating a considerable amount of siding accommodation, not to mention possible blocking of other lines by the steam locomotives constantly either running out or else backing on to the trains. CONCLUSIONS. Comparatively little has so far been done toward the introduction of electric traction on main line railways in Great Britain. This is not" surprising, and as far as that is concerned, neither in the States nor on the Continent of Europe are main line railways at present operating anything like very long stretches of line by means of electric traction. The country which has the longest stretches of line is undoubtedly Italy, with its three-phase 3000-volt line work- ing with overhead trolley between Lecco and Sondrio, and its third- rail system between Milan, Gallarate, and Varese, both of which lines have been exhaustively described in the technical press of the world. As regards this country, the Lancashire & Yorkshire railway has equipped and is operating 23 miles of route, and the North Eastern is operating 40 miles of route, both on the third-rail system. These lines have only recently been put into regular service, and no figures, either as regards increase of traffic or cost of operation, are as yet available; the only results so far obtained go to show the excessive danger of third rail. In this country a large number of accidents to third parties have taken place, some of OS DAWSON: ELECTRIC TRACTION ON BRITISH RAILWAYS. them fatal, and in one instance a train has been set on fire and seriously damaged, fortunately with no loss of life. These results, as far as they go, amply confirm the conclusions at which I have arrived and are strong evidence that the adoption of third rails, at any rate on main line railways, is not at all desirable. There are several lines in this country operating at high speeds with fairly heavy loads and many others are being constructed. There is, for instance, the Mersey railway, the City & South Lon- don, the Central London, the Great Northern & City, and the Liverpool Overhead, all of which have been working most satis- factorily for a considerable number of years, as well as the Metro- politan, the Metropolitan District, and other tube lines now being constructed by the Underground Electric Eailway Company of Lon- don and which will commence working next year. The power sta- tion which will supply energy for the last-mentioned railways is. so far, the largest that has ever been built. To give some idea of the large amount of railways which already exist in London, the following tabulated statement may not be with- out interest. I have taken them from figures published by that eminent American, the Hon. Eobert P. Porter: RAILWAYS RUNNING INTO LONDON. North Side. t Railway. Un - > i Great Central e"#th of Number of es within stations county. wiU in [miles). county. 2.37 1 16.79 27 4.31 4 4.75 4 9.64 12 .62 1 12.25 18 10.42 16 2.12 8 7.27 6 1.92 6 Great Eastern Great Northern Great Western ' London & North Western London Tilbury & Southend Metropolitan Metropolitan District Metropolitan & Metropolitan District (joint) . . . Midland Totten & Hampstead Junction Total . 72.46 103 DAWSON: ELECTRIC TRACTION ON BRITISH RAILWAYS. 09 South Side. Tergthof Number of T?5!rrax/ UD68 Within Stations "* county within (miles). county. London, Brighton & South Coast 31 . 14 29 London, Chatham & Dover 26 . 22 33 London & South Western 14.05 12 South Eastern 37.86 27 London & South Western and London & South Coast (joined) .60 Totals 109.87 101 Wholly in London. City of London Electric 6 . 50 13 City & South London Electric 6.65 14 East London 7.22 7 Hammersmith & City 3 . 00 5 North London 11.19 18 Waterloo & City Electric 1.50 2 West London 2 . 30 2 West London Extensions 4. 76 4 Whitechapel & Bow 3.00 4 Totals 46.12 69 Making a grand total of 228.45 273 In this connection it must be borne in mind that the mileage here given are miles of route and not miles of single track, and that they only represent the miles of route actually inside the county of London. In order to represent the actual mileage which is only, or to a large extent, devoted to suburban service, the total would have to be more than doubled; in the case of the London, Brighton & South Coast railway only just over 31 miles of route are given, whereas the suburban system comprises 75 miles, and on this sys- tem, the average distance between stations does not exceed one mile. A glance at a railway map of London and its environs clearly shows the enormous network of railways which converge into the center of the city, and a careful examination of such a map on which the existing electric tramways and light railways have been drawn will 70 D4WSON: ELECTRIC TRACTION ON BRITISH RAILWAYS. show how, in the next few years, the electric tramways and light railways will enable passengers to travel from any portion of Lon- don to places from 20 to 30 miles outside the center of the city. Under the existing conditions it would in many cases be advan- tageous for passengers to travel as far as 15 miles by tramway in- stead of taking the steam railway, owing both to the low average speed and, in many cases, the long interval between the trains. What has been stated as regards London, applies quite as well to other large towns of the United Kingdom, and I am firmly con- vinced that there is no country in which electrification will be a greater benefit to the railways than the United Kingdom. I believe that the British railway companies are rapidly realizing that a move will become necessary, and when once the movement begins, the transformation as regards our railways will be quite as great as that which has taken place during the last few years with tramway construction. DISCUSSION. CHAIRMAN DUNCAN: The paper abstracted by Mr. Armstrong is now open for discussion. Mr. H. WARD LEONARD: There is one figure that drew my attention in the eaily part of the paper, and that is the statement of the very high capitalization of the English roads, as an average. I can only speak from memory, but I believe that the most efficient railway as regards earnings we have the Pennsylvania railroad has a capitalization of about $370,000 per mile, which is about 50 per cent in excess of the figure named. So that very high capitalization, while of course it is of tremendous importance, is not necessarily an indication of poor earning capacity. The average figure in New York, Pennsylvania, and New Jersey which is the part of the United States most fairly comparable with England is about $120,000 a mile. Mr. F. J. SPBAGUE: You refer now to single or double tracks? And does the paper refer to single or double tracks? SECRETARY ARMSTRONG: It is per mile of road. It is partly single and partly double. Mr. LEONABD: The figures I have named are all per mile of road. The figures I have stated are from the statistics of the United States Railway Commission Reports. They are per mile of road, that is, per mile of line. I think it is quite proper to emphasize the statement the author has made in the paper as to the steam locomotive having to carry around with it continuously a very large number of tons that are entirely idle. I think it would probably be conservative to say that the net cost repre- sented by the ton miles of a locomotive due to the non-tractive part of a heavy freight locomotive in this country would be not far from $50 p*.r day. And that brings up another point, namely, that in discussions on this subject, a great deal of attention is usually spent on the cost of DAW SON: ELECTRIC TRACTION ON BRITISH RAILWAYS. 71 fuel, and the question of whether or not it would be possible to save the wages of the fireman. Personally, I consider all those matters as ex- tremely trivial as compared with the really important matters. Con- sidering $100 of earnings by railways of this country only about $7 of that amount is spent for fuel. So that it does not seem advisable to confine the discussion so much to fuel consumption. I think that the very greatest importance should be laid upon the re- quirement of endeavoring to utilize to the highest degree the very large investments and fixed charges that are represented by the equipment and maintenance of a mile of road. Something like 86 per cent of the total cost of moving a ton-mile is represented in this way, and is totally independent of the coal and wages on the train. And it seems to me con- spicuous that the problem narrows itself down to the question of getting from every mile of track per hour the maximum possible ton miles which means again the maximum possible number of tons moved at the highest possible rate of speed. If we go back, for example, in the statistics of the Pennsylvania Rail- way for about thirty years we find that the cost of moving a ton one mile used to be at that time about a cent and six-tenths; and in 1902 the cost of moving a ton one mile by the same railway was thirty-six hundredths of one cent. Now, this very striking reduction in the cost of moving freight has not been in any way due to any reduction in cost of wages or cost of coal. On the contrary, those have increased. And per- sonally I am strongly of the belief that to-day the cost of moving freight is inversely proportionate to the power that is employed in moving the train, and with that thought in mind it seems to me that the electric moving of freight has possibilities that are not at all to be expected from any steam operation. There are probably not more than 10 per cent of the locomotives that are used in this country that are capable of developing over 1000 horse- power. Those large locomotives are the most economical ones we have as regards moving freight. There are some 40,000 locomotives in the United States, and less than 4000 of them are of modern efficient size. The boiler is the principal limitation to the power of the locomotive, and it seems unlikely that there will be very much growth in the power of the boiler used on steam locomotives; whereas, theoretically speaking, there is no limit to the amount of power that could be applied to the move- ment of a freight train by electricity. The draw-bar pull of the freight locomotive, in the best types, reaches sometimes as high as 50,000 pounds, but that draw-bar pull is only obtained when steam is taken at full stroke, which means at an ex- tremely slow rate of speed, and by the time such a rate of speed is reached as would represent the average speed desired the draw-bar pull is less than half of that figure. The mountain sections of our principal railways are the places where the requirements for power are most keenly felt to-day. There is always a great congestion of freight at such places. If we employ electric loco- motives we have, fortunately, coincident with the grades of those nioun- 72 DAWSON: ELECTRIC TRACTION ON BRITISH RAILWAYS. tain sections, as a usual thing, power in cheap form repre?ented by the water power of the mountain section. It seems to me that what is wanted is a locomotive which will produce about 50,000 pounds draw-bar pull and maintain that at about thirty miles per hour and that means 4000 horse-power at the draw-bar. We already are subjecting our draw-bars to that strain, and they are strong enough to stand it, provided that we have some form of operation which is not going to subject those draw-bars to intermittent large strains due to irregular methods of control, or to the bucking of the various units that are employed under multiple control. It seems to me that the principal cause of the poor showing of the British railroads in the cost of handling freight, as compared with the United States, can be found in the fact that as an average the horse-power of their locomotive is very small compared with the best practice in this country. Mr. SPRAGUE: I have only a few words, Mr. Chairman, on this paper. Not having read it, I am not prepared to discuss it at length. Probably in what little I say I may disagree with Mr. Dawson, and to some extent with what Mr. Leonard has said. In one form or another I have for many years advocated electric traction. It has already naturally supplanted a method of traction which at the best was poor that is, animal traction ; and in supplanting it, it has achieved results in transportation greater than its most ardent advocates had hoped, in cheapening operating cost, increasing schedule speeds and opening up new fields. But we must not forget that one of the chief reasons for the great success of the electric railway has been the fact that it has been what may be called a house-to-house railway, one making frequent stops convenient to the passenger. As a result we have seen here in the United States practically every horse-car disappear, almost all cables abandoned, exist- ing lines consolidated, and new lines link together towns and cities and wipe out the divisions between urban and rural communities. But when approaching the steam railway problem I have always done so with a good deal of deference to existing conditions. Electricity, after all, is merely a convenient method of transmitting power. We do not create anything by it, we do not establish any new laws by its use. By concentrating at central stations the power used on a railroad and dis- tributing it in the best possible manner we hope to utilize that power more economically. But in order to do so successfully from a power standpoint there is one essential; the load factor must be high, which brings me back to an assertion which I have made again and again for the last fifteen years; namely, that leaving out for the moment the influence of competing lines, diversion of traffic and what not, there H a point on any railroad where the adoption of electricity may be justified, and that point is primarily determined by one essential, density of traffic. And I do not mean by density of traffic concentration of loads at one point, but multi- plicity of units well distributed. So long as the operation of any road means the sending out of high powered units at long and irregular in- tervals over great distances we might as well be frank with ourselves and say that there is not the field for electric transportation. DAW SON: ELECTRIC TRACTION ON BRITISH RAILWAYS. 73 As the traffic increases we approach a point where the number of units between terminal points warrants the consideration of a change of motive power, and I think that condition has arisen on a number of steam rail- roads. Beyond that point there can, I think, be very little question as to whether electricity can be economically adopted. The problems in Great Britain, the Continent and the United States are somewhat different. Here it will be many years before we can in even the most hopeful attitude look for many of our main roads to be operated other than by steam. There are some roads and certain sections of rail- roads which will undoubtedly be operated by electricity but oftentimes for reasons not determined by economy of operation. I may, perhaps, cite the most important two instances in this country, if not in the world, at present the operation of the Pennsylvania Rail- road tunnels and terminals in Jersey City, New York and Long Island, and then those of the New York Central Railroad in New York and a part of its main line. I have the honor to be a member of the Commission on the latter road, which has to do with the electrification of the equipment. It is, perhaps, too soon to go into details in connection with it, except this : One of the requirements a legal one which determined the use of electricity on this road was that no steam-operated train should be used below the Harlem River. The Harlem River is well within the city limits of New York, and only a comparatively short distance from the terminal at 42nd Street. The movement of trains within that district is enormously congested. According to a report made by Mr. Arnold some- time ago, at certain periods there are over 700 daily train movements, and the trains vary anywhere from 150 to 700 tons in weight. The law said we should abandon the use of steam. Of course, we were permitted to use anything else in the tunnel, but that was practically the same as saying that electricity should be used. But within even the district determined for electrical operation extending out some 35 miles, I do not think any calculation made, taking into account the interest on investment, shows any real economy in operation over steam, all things considered. The determining considerations may be stated as first, the law which practically required the use, part of the way, which part was to a point where there were no terminal facilities whatever, and second, that it was advisable from a transportation standpoint to operate suburban trains electrically certainly within a distance of perhaps an hour's run to New York. Having determined that, then it was common sense to operate all trains located within that zone by electricity, instead of having a duplicate system. The result is that for some distance from New York city we will have what may be considered a great terminal, within which there are suburban trains operated by motors under the cars on the multiple-unit plan, and other trains dropping off their steam locomotives at the termini, and taking on electric locomotives likewise so operated. It is, perhaps, unwise -to attempt any limitation as to this particular development. Certainly I would not be rash enough to hazard it, but it is a special problem, and I do not think has yet any great bearing upon the 74 DAWSON: ELECTRIC TRACTION ON BRITISH RAILWAYS. broad question whether trunk railways will be operated by electricity. There are many other problems to be taken up when that subject is con- sidered. The British railways, especially those terminating in London, will mostly adopt electricity only when they are compelled to; few of them will do it of their own volition. The competition which exists in this country with electric railways will not be quite so forcibly felt on English roads, because there is not that same freedom in granting of franchises tor parallel lines of railways that we have here. But there is a special reason why electric equipment should be considered. The traffic in Lon- don, for example, is enormously congested at certain times of the day. The result is that the facilities are entirely inadequate, and new con- struction, whether of an overhead line over existing tracks, or tunnels beneath or tracks parallel to them, is almost prohibitive in cost. So that the natural, and so far as I can see about the only way, in which they can increase their capacity is by electric equipment, and I think that the British roads in time will see that fact. Perhaps one of the most important means by which the changes could be brought about would be in the reduction of the age limit in their directorates. You gentlemen know the English and American practice is somewhat different in this regard. When a man becomes a director on an English railway the position practically terminates only at his death or permanent disability, or when for some good practical reason he gives away to another. He rarely represents, in that conservative method which governs English railways, the progressive element of the stock- holders. Being a man of mature years, and often having reached that age when most men are ultra conservative, he will hesitate to abandon an existing system and adopt another. It usually takes younger men to do that, and to believe what can be accomplished by such a change. Here in the United States the directors have not that hold upon the administration that they have abroad. A change of ownership in the road, a change in the holdings of the stock, may result in a very prompt and radical change in its management from the president down, but such a thing is almost impossible in the British Isles. I do not know that there is anything further, Mr. Chairman, that I want to add. In fact, I did not intend to say anything. I am as hopeful, perhaps, as any one can be that electricity will be used on steam railroads, but I do not want to shut my eyes to the fact that there are a good many difficulties inherent to trunk-line railway service for which elec- tricity is no cure-all. Mr. LEONARD: Mr. Chairman, if I may be allowed to add a word, I should like to speak of one point I have heard frequently raised and it seems a fitting opportunity to mention it. In comparing the cost of operation of English railways with American railways, I find that the Englishmen are very apt to retreat behind the argument that in England the mail and express business is classified in the freight figures, whereas we in this country have those classified in the passenger business, and receive compensation for those, which make so important a factor as to distort the figures materially; and that, therefore, deductions cannot be DAWSON: ELECTRIC TRACTION ON BRITISH RAILWAYS. 75 fairly drawn by comparison in the cost per ton-mile of the English rail- \vays and the American railways. The express is 2 per cent of the receipts in this country, and the mail also about 2 per cent of the total receipts, so that those figures are not sufficiently influential to in any way influence the very striking difference in the costs. About nine-tenths of our present steam locomotives in our country seem to me a liability, rather than an asset, to the railways that operate them; and since there seems to me to be a necessity for "scrap- ping" about nine-tenths of the inefficient small locomotives and the re- placement of them by the larger efficient ones, we are not confronted with the same condition of affairs as we would be if the steam railways had already made a comprehensive equipment of the highest class of steam locomotives. THE MONORAIL RAILWAY. BY F. B. BEHR. For many years eminent engineers, including the great Telford in 1828, have taken great interest in designing single-rail rail- ways, and many patents, covering a variety of forms and combina- tions, have been taken out, but none of these attempts have been carried to a practical issue until recently. Within a comparatively recent period, however, two systems have been so far perfected as to offer real practical value as means of transportation for passengers and goods. Of these two systems, one is identified with the author, and the other is known under the name of the Langen system, and is used to connect the towns of Barmen and Elberfelde, Germany. This paper will more especially describe the development and ap- plication of the former system. No claim is made by the author of this paper as the originator of the fundamental principle of the system of monorail railway. It is difficult to trace who first suggested it, as there were several almost simultaneous attempts in that direction between the years of 1875 and 1884. His only claim is to having taken up the original idea in 1884, when it was still in its simplest and most primitive shape; to having developed the general ideas and principles of others in designing the prac- tical details; and to having constructed for the first time, in 1886 for steam power and in 1896 for electricity, monorails which have been worked successfully for the carrying of passengers and goods. The Behr monorail is applicable to three distinct purposes, namely, to light railways in sparsely populated districts and in hilly countries where they would serve as feeders to existing rail- ways, as elevated railways in towns, and as supplementary to ex- isting systems of mail lines all over the world for carrying express passenger traffic, mails and parcels at much higher speeds and with much greater frequency of trains. The advantages over ordinary lines in the first of these appli- cations, result principally from the possibility of using very sharp [76] BEER: MONORAIL RAILWAY. 77 curves, avoiding the expense of earthworks and tunnels, and also from the smaller cost of bridges, etc. The result of this economy would be about 50 per cent in hilly countries, in comparison with an ordinary meter gauge railway. The special principle of the Behr monorail was in 1883 applied by Mr. Charles Lartigue in the construction of some primitive FIG. 1. PROPOSED METROPOLITAN RAILWAY FOB CHELSEA EMBANKMENT, LONDON. lines in Algeria and Tunis for carrying esparto grass and similar produce, the tractive power being by animals in all cases. An experimental line was built by the author in London, in the rear of Victoria street, Westminster, in the year 1886, where for the first time locomotives and carriages were run on a monorail. On the section was a gradient of 1 in 10, and for about a year the engine took up this incline, without a rack, one light carriage besides its own weight, showing that the adhesion on this form of railway is considerably greater than on an ordinary two-rail 78 BEER: MONORAIL RAILWAY. railway, on which on such an incline an engine is hardly able to pull up its own weight. An act of Parliament for a railway from Listowel to Bally- bunion, in Ireland, for regular passenger and goods traffic, was obtained in July, 1887, and the line was passed by the Board of Trade and opened to the public March 1, 1888. It has been working ever since without any difficulty or accidents, and in over 16 years has not been subject to a single claim for compensation of any kind. The line is especially remarkable for its very sharp curves, the smallest having a radius of 54 feet. The second application, as an elevated railway in towns, is of very great importance, especially in the United States, where such FIG. 2. CAB FOB PROPOSED MANCHESTER AND LIVERPOOL ELECTRIC EXPRESS RAILWAY. railways are in common use. Elevated railways of this system in towns could be built at a very much smaller cost, with much less obstruction of the ordinary road traffic, requiring much less room on the roads, with much less obstruction of both light and air. Such a line has been proposed for the Chelsea Embankment, in London, to Putney Bridge. The cost of the line (double track) would be under $200,000 per mile. The system has also been pro- posed for another metropolitan line in London, about 17 miles long, starting in the west end and going to the city and docks, and nowhere along any street, merely crossing streets. The total cost of this line, double track, is estimated at about $500,000 per mile. Passing next to the question of high speeds, the great merit of this system for high-speed service is that the cars are absolutely underailable; that it possesses important economic advantages for BEER: MONORAIL RAILWAY. 79 working at very high speeds; and that the rise in grade in ap- proaching stations greatly helps the acceleration of the trains when starting and is of equally great assistance in stopping the trains when approaching a station. The cost of construction of such lines is generally slightly less for a speed of 100 to 110 miles an hour than the cost of an ordinary two-rail railway for speeds of 50 to 60 miles an hour. There are many causes which contribute to the absolute safety of the system which can only be understood by carefully examin- ing the detailed construction of the carriage as it fits to the track, when it will be observed, among other things, that whereas an ordinary railway carriage is held on the rails by a flange of about three-fourths of an inch in depth, the arrangement of the mono- rail carriage is really equivalent to a continuous flange of over three feet in depth. A feature of great importance to the pas- senger is that it is not only a safe way of traveling, but it looks also very safe and produces on the mind of the traveler a feeling of absolute security. On an experimental elevated high-speed monorail built in 1897, with a carriage weighing about 72 tons, a speed of 84 miles per hour was obtained over curves of 1500 ft. radius, and a speed of 70 miles per hour on an ascent of 1 in 90. It was a much greater feat to attain 84 miles an Hour on such a line and on such curves, with straight sections so short that it was impossible to construct a proper parabola between them and the curve, than to attain a speed of 110 miles on a properly constructed monorail, under such conditions as would arise in ordinary railway practice. This line consisted only of embankments about 25 ft. high and cuttings 20 ft. deep, with a total fall of 130 ft in 1% miles. The road was built during a very wet winter in a few months, and worked immediately afterwards during a very wet summer, when considerable portions of the embankments had prac- tically been washed away, and many of the sleepers were really suspended in mid-air. Notwithstanding these conditions, experi- ments were carried out during a period of over twelve month? without a single accident. The line was three miles in length, and formed by two short straight lines joined by two curves at each end, so that continuous runs of any length could be made. But as there was a fall of 130 ft., it was necessary to rise to the same level, so that to develop the speed there was practically only a length of about one and one-half miles, and the highest speed 80 BEER: MONORAIL RAILWAY. of 84 miles an hour always occurred at the bottom of the incline, at the center of a curve of 1500 ft. radius. The British Parliament has authorized and the Board of Trade has approved the construction of a monorail between Manchester and Liverpool, on condition that the speed shall not exceed 110 miles an hour. The sharpest curve to which this speed applies has a radius of 1800 ft. The whole of the materials proposed to be used in its construction, above the level of the sleepers, will be of steel. The maintenance will be similar to that on an ordi- nary railway, as there will be practically no difference in the manner of packing the sleepers or of inspecting the various parts. For the greater security, however, of the workmen employed on the line, the clear space left between two trains passing will be 3 ft., as against 1 ft. 8 in., the space provided between Pullman cars. All trains will consist of only one car, for reasons of safety, economy in working and construction, and for the convenience of the public. There are three classes of cars designed and approved for this line. The smallest car will carry 40 passengers, the second size 52 and the largest 80. It is proposed to begin the service between Manchester and Liverpool with cars carrying 40 passengers each and running every 10 minutes. The working expenses of this service at 110 miles an hour, including maintenance repairs, management and everything else, are estimated at less than 15 cents per train mile. The center of gravity of this carriage is at least 12 inches below the top surface of the monorail, as required by the Act of Parlia- ment. The whole working of this line, which will carry, if neces- sary, 48,000 passengers a day at a speed of 110 miles an hour, doing the whole distance in 20 minutes, is very simple. Collisions are impossible, there are no level crossings, no switches, and not- withstanding the number of passengers carried, there are never more than two carriages on the whole line from end to end. With regard to the electrical working, full details cannot be given, as the author does not consider that he is especially quali- fied for that purpose. The joint electrical engineers for the Man- chester and Liverpool Kailway are Lord Kelvin and Sir W. H. Preece. Following are given, however, some general data: The distance to be traversed is 34% miles, without a stop, in 20 minutes. The acceleration at starting is to be 2 ft. per second per BEER: MONORAIL RAILWAY. 81 second, diminishing to 9 in. per second per second or an average of iy 2 ft. per second per second, attaining a speed of 110 miles in 1 minute 47 seconds and in a distance of under 2 miles. The resistance due to friction and air pressure is taken at 45 Ibs. per ton at full speed. The coefficient of adhesion is taken at about one- sixth, say 400 Ibs. per ton for the worst weather. The total weight of the car is over 40 tons and the weight of the driving wheels is 20 tons; hence the limit of adhesion that can be calculated on these driving wheels under all circumstances is over 200 Ibs. per ton plus 15 Ibs. per ton weight for air resistance, giving a total of 215 Ibs. per ton weight, being more than the weight required, as 140 Ibs. per ton is all that is necessary for an acceleration of 2 ft. per second per second. For braking purposes, a high-speed Westinghouse brake will be able to retard the train at the rate of 3 ft. per second per second, which will absorb 210 Ibs. per ton weight of the car distributed over the four wheels, or 52% Ibs. per ton per wheel. This will stop the car in about 1380 yds. If, in addition to this, the motors are short-circuited, the remaining adhesion can be utilized on the two driving wheels, which amounts to another 52% Ibs. per ton per wheel, and is sufficient for an additional retardation of 1 ft. 6 in. per second per second. This will give a total retardation of 4 ft. 6 in. per second per second, and would stop the car in 768 yds. In this arrangement, the retardation produced by the motors will be at exactly the ratio of the average acceleration to attain the full speed. If the shortrcircuiting of the motors was used alone without the Westinghouse brake for stopping the train, there would be an available adhesion on the driving wheels of 215 Ibs. per ton weight of the car, amply sufficient for a retardation of 3 ft. per second per second, and also for stopping the car in 1380 yds. Therefore, either of the brakes used alone will stop the car in that distance, whereas both combined will stop the car in 768 yds. This does not take into account the steep up grades at the stations. The power required during acceleration is about 1100 hp, and during the run about 515 hp, or 129 hp per motor, there being four motors to a car. The generating station is situated exactly half way, at Warring- ton. Three-phase currents will be generated at 15,000 volts and converted in five sub-stations placed along the line into continuous current at 650 volts. The motors are wound for 600 volts and ELEC. RYS. 6. 82 BEER: MONORAIL RAILWAY. weigh each about 2% tons. The system used is three-wire con- tinuous current. Each car will be fitted with four continuous current traction motors arranged in pairs. Each motor will have a normal capacity of 160 hp at the full speed of 720 revolutions per minute, but will be capable of giving at least 320 hp for short periods during accel- eration. The driving wheels have a diameter of 4 ft. 4 in., the speed at 720 revolutions per minute corresponding to 110 miles an hour. The whole line will be fenced with an unclimbable fence from end to end, preventing all possibility of trespassing, as there are no level crossings and no means of access of any kind. By an arrangement on the axles of the guide wheels, which are freely suspended in slots fixed on the bogie or truck frames, the guide wheels on both sides of the car remain always horizontal and in fair contact with the guide rails, whatever may be the inclina- tion of the bogie frames and the car itself, which can swing freely on the top or bearing rail under the influence of centrifugal force in the curves, or from any other causes. The main rail itself remains always perfectly horizontal, even on the sharpest curves. The result is that the pressure on the guide rails need never be increased or the inclination of the car, and this pressure can be limited in such a manner as to combine the greatest comfort of the passengers with the greatest economy in electrical energy through the diminution of friction, THE ELECTRIFICATION OF STEAM RAILROADS. BY BION J. ARNOLD. Eleven years ago this summer it was our privilege to meet under the auspices of a great Exposition, located upon the shores of Lake Michigan, organized not only to commemorate the 400th anni- versary of the discovery of this country, but also to direct atten- tion to the advancement made in the various fields of the world's activities, and especially in those arts in which we, as workers, were most interested. To-day we meet under the auspice* of another great Exposition, brought into being to commemorate the 100th anniversary of the peaceful acquirement by the Government of the United States of a large portion of the territory now contained within its bor- ders, to have our attention directed to the development of the various industries of this and other countries that have taken place during the intervening years. For a few years preceding the former Exposition, engineers and others engaged in electrical pursuits had had their energies ab- sorbed in attempting to show the owners of street railways that operation by electricity was cheaper and better than by means of the horse or the cable. We, at that time, had seen the horse prac- tically disappear from street railway service and the cable sup- planted in some instances. The more ambitious engineers were then advocating the use of electricity on elevated railways, and making figures to prove to the owners of such railways that electricity was cheaper in opera- tion and more desirable for such conditions than steam locomo- tives, then universally used for such work. At that Exposition was placed in operation an elevated electric road, known as the Columbian Intramural Kailway, which, though the city and South London Underground, a road of light equip- ment, was started some time before, and the Liverpool Overhead Eoad soon after, was the first practical commercial application on a large scale of electricity for the propulsion of heavy railway trains. [831 84 ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. The success of these roads gave the electric railway industry an impetus which has since resulted in the abandonment of steam and the adoption of electricity on every elevated railway now in operation, and practically on all of the underground roads, thus effectually proving the soundness of the theories of those engineers who pinned their faith to the correctness of the conclusions which their figures showed, and who staked their reputations upon the future to prove them true. The interval between these Expositions has also been one of great activity and development in the field of interurban railways, which has brought into being the extensive use of the alternating- current, rotary-converter, sub-station system of operating direct- current roads, resulting in the interlinking of thousands of cities with each other and intervening points, thus not only affording a new field for the investment of capital but bringing, to most of the inhabitants of the territory through which these roads pass greater facilities for the prosecution of business and the widening of their social life. With the introduction of the suburban railway came an increased volume of passenger travel, induced by the increased facilities, which may well be noted by the managers of great steam railway- properties as an example of what may be expected in increased revenue when frequent and pleasant service is available to the public. The energies of those engaged in electrical industries have thus far been absorbed in fields which now seem to have been naturally theirs, and their success has been such that they now aspire to enter the field occupied by the steam locomotive as a legitimate field of conquest. The questien now is whether this field is one in which the ad- vantages of electricity will be sufficient to overcome -the obstacles which seem almost unsurmountable, and enable it to win as it has in the cases cited. Those who have given the subject little thought or who are unable to analyze it carefully on account of the lack of the tech- nical knowledge necessary to appreciate the difficulties to be overcome, are most apt to predict the early supremacy of the electrically driven train over the steam locomotive. That the fields referred to have been apparently formidable yet quickly overcome is not necessarily proof, or even good evi- ARXOLD: KLBCTRIFIGATION OF STEAM RAILROADS. 85 donee, that the legitimate field of the steam locomotive can be entered and successfully achieved. Those most familiar with the subject are now prepared to admit that our great steam railway terminals, where many switching locomotives are shunting back and forth continuously, and those portions of the steam roads entering our great cities, where sub- urban trains are numerous, frequent and comparatively light, can be more economically operated by electricity than by steam. This is evident to most of those engaged in the work, for the reason that it simply means duplicating, on a large scale, the systems which have proven successful in our street railways, operat- ing, as they do, numerous units running at frequent intervals. Proof that this field is recognized as a legitimate one for elec- tricity is furnished in the examples of steam railway terminals that are now being equipped electrically, such as the lines of the New York Central and Pennsylvania Eailroad Companies in the vicinity of New York, involving an expenditure of something over $70,000,000, where not only suburban service will be operated electrically, but where in the case of the New York Central, the main line trains will be brought into the city from points 30 to 40 miles distant. While these are great examples of electrical operation on steam railroads, and heroic instances of faith on the part of the railway managers in the ability of electricity to successfully meet the con- ditions of steam railroad work, where the trains are sufficiently frequent, they are by no means conclusive evidence that electrically propelled trains can be made to successfully meet the conditions of trunk line passengers and freight service, the field now so suc- cessfully held by the steam locomotive. The best conditions for electrical success are a great number of units moving at a practically uniform schedule, at equal intervals, within a limited distance. The legitimate field of the steam locomotive is now one in which there are few but heavy units moving at uneven speeds over long distances at unequal intervals and at high maximum speeds. The amount of energy transmitted to any great distance and used by electric cars that have been put in use until recently has been small when compared with the amount of energy that it takes to propel a steam railroad train of five or six hundred tons weight at the speeds ordinarily made by such trains. It may be taken as axiomatic that when investment is taken 80 ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. into consideration, power cannot be produced in a steam central station, under conditions that exist to-day, and transmitted any great distance to a single electrically propelled train, requiring from 1000 to 2000 hp to keep it in motion, as cheaply as a steam locomotive, hitched directly in front of the train will pro- duce the power necessary for its propulsion. Therefore, there must be other reasons than the expected economy in power prp- duction to warrant the adoption of electricity on a trunk line rail- road unless it can be shown that the trains are frequent enough to make the saving in the cost of producing power greater than the increased fixed charges made necessary by the increased invest- ment due to the adoption of electricity. There are undoubtedly in existence to-day conditions where water power in abundance is available along the right of way of existing roads, in which the substitution of electricity for steam could be made a paying one, with apparatus now available, even on roads having a comparatively infrequent service, but these are special cases and only tend to prove the correctness of the position, for in these special cases the cost of power would be but little over half the present cost of producing it by means of a central steam- driven station. The ideal conditions for any trunk line railroad having a traffic heavy enough to warrant the investment in a sufficient number of tracks to properly handle this traffic in such a manner as to get the most efficient service out of its rolling stock, would be to have four or more tracks between terminal points, upon which, in pairs, could be run the different classes of service at uniform rates of speed. Thus, if six tracks were used, the through line, passenger, and express service would be run on one pair of tracks ; the local passenger, local express, and local freight service upon another pair of tracks; while the through freight service would be run upon a third pair of tracks, and all the trains upon any pair of tracks would run at the same average speed and stop practically at the same places. If these conditions could prevail and the traffic were sufficient to warrant this investment in tracks, such a service could be operated more economically and more satisfactorily electrically than by steam. The difficulty is that few roads in existence have sufficient traffic to warrant such an investment in a permanent way, and the result is that all of their traffic must be handled over one or two tracks. ARK OLD: ELECTRIFICATION OF STEAM RAILROADS. 87 thus necessitating trains of all weights and all speeds running upon the same rails. This results in a tendency to bunch the cars into as few trains as practicable, in order not only to reduce the cost of train service to a minimum but to give the fast-running trains greater headway to allow them to make their time safely. Such an arrangement of trains necessitates the concentration of large amounts of power in single units, which is leading away from the ideal conditions for the application of electricity to the pro- pulsion of trains; and it is this element, combined with the fact that the traffic on most roads is not great enough to warrant the investment necessary in electrical machinery to produce and trans- mit the power to the distances necessary to keep a few heavy trains in motion, that makes the trunk line railway problem so difficult, as it is more economical to propel these heavy trains by steam- driven locomotives, which are practically portable power-houses. It being admitted that electricity becomes most economical when a sufficient number of trains are available, and that the steam locomotive is most economical when the trains have become few and heavy, the problem then resolves itself into one of the density of traffic and the question then is: where is the dividing line? It was my intention to attempt such an analysis of this subject as to be able to formulate some general law which could be readily applied to any given case, and thus enable one to decide whether electrical operation would be more economical than steam in any concrete case. After carefully analyzing the subject I have become convinced that no general law or formula can be laid down which will apply to all cases, for the reason that the elements entering into different cases vary so greatly that any formula would contain too many variables, dependent upon local conditions, to admit of a general application. I shall, therefore, only attempt to point out a way in which the dividing line between steam and electricity can be determined after the elements of each case are known. It will readily be seen that with steam locomotive operation the fixed charges, and cost of fuel and engine labor increase almost directly proportional as the train miles increase, for in this case an additional locomotive means simply a given amount of increased investment, a given amount of increased fuel and labor, and this total investment is least when the number of locomotives is small. On the other hand, with electricity it is necessary to invest at 88 ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. once a large amount of capital in the power houses and trans- mission systems, which amount must be great enough to provide for handling the maximum number of trains required upon the line, and unless this number of trains is great enough so that the economy effected in the different method of producing and applying the power is sufficient to offset the increased fixed charges, due to the additional invested capital, it will not pay to equip and operate electrically. Any problem, therefore, must be analyzed for the relative cost in operation. In case this does not show a saving the advisability of equipping electrically will depend entirely upon the probable increased traffic to be derived from the adoption and operation of electrically propelled trains. That electricity will be generally used on our main railway terminals, and ultimately on our main through lines for passenger and freight service, I am convinced, but I do not anticipate that it will always be adopted on the grounds of economy in operation ; neither do I anticipate that it will come rapidly or through the voluntary acts of the owners of steam railroads, except in special instances. At first the terminals will be equipped for special reasons, due either to the voluntary act on the part of the terminal companies to effect economy in operation, or to public pressure brought to bear upon the owners through an increased demand on the part of the public for better service, on the grounds that the use of the steam locomotive is objectionable in our great cities. Those roads which run through populous countries will either build new roads, or acquire, for their own protection, those electric railways already built and operating in competition with them, and utilize them as feeders to their through line steam trains. Thus the steam railroad companies will gradually become interested in electric railways and eventually become practically the real owners of them. With these roads operating as feeders to the main line system and with the terminals thus equipped and the public educated to the advantages of riding in electrically equipped cars, the next step will logically be the electrical equip- ment of the trunk lines between the cities already having electrical terminals. Thus some favorably located trunk line having a sufficient density of population will feel warranted in equipping electrically, and when this is once done the other roads running between the ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. 89 same competing points must, sooner or later, follow in order to hold their passenger traffic. This may result in temporarily relegating some roads to freight service, so long as they operate exclusively by steam, but with the increased demand on the part of the public for better and cleaner service will come a corresponding increase in passenger revenue to the roads equipped for handling it until one road after another finds it advantageous to furnish an electric passenger service. With the terminals and main lines equipped electrically, and the desire on the part of the public for more prompt and effective freight service resembling that which is given by the steam roads in England and on the Continent, due to the great density of population, there will be developed a great high-class freight serv- ice conducted in light, swiftly moving electric trains which can be quickly divided and distributed over the surface tracks of our smaller cities, or through underground systems similar to that which is now being built in Chicago. Such a system would soon prove indispensable to the public and a source of great profit to the roads as it is now getting to be to many suburban railways. This class of freight service would soon prove so large a part of the freight traffic of a road that the operation of the through freight traffic by steam locomotives, though at present cheaper, would in time, as the cost of coal increases, grow less, until those roads operating an electric passenger service would ultimately use electricity exclusively. It has not seemed advisable to me in an address of this char- acter to attempt to furnish detailed figures to support my theories for the subject is of such general interest that many able men are presenting papers upon it at the International Electrical Con- gress now in session here, in which papers will be found informa- tion of much value to those interested, and from which I believe the correctness of some of my assumptions can be proved. The principal problem before the electric railway engineer to-day is how to make the most effective use of the high-pressure trans- mission, and high-tension working conductor and maintain safety of operation. Experiments conducted during the past year by engineers in this country and abroad have made this problem simpler than it seemed before and to-day we seem reasonably certain of the solution. f)0 ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. Until recently the cost of electrically equipping a trunk line under the standard direct-current, rotary-converter system, has been such as to practically prohibit its adoption, but recent devel- opments in the single-phase alternating-current motor field have made it possible to eliminate a large part of the investment here- tofore necessary and the prospects for the application of electricity to long-distance running are better than ever before. When it is recalled that the rotary converter, which was the means of reducing the cost of long-distance roads, was introduced in 1898, and that within the six years from the time of its adop- tion through the development of the single-phase motor it has been practically rendered obsolete for heavy railroad work, it will be seen that the dividing line between the steam locomotive and the electrically propelled train has moved several points in favor of the latter, due to the reduction which can now be made in first cost and the saving in operating expenses. With the single-phase motor and the steam-turbine a reality, the transmission problem almost solved, and with the rapid devel- opment of the internal combustion engine now taking place, are we, as engineers, not warranted in believing that we can so com- bine them into a system which will ultimately supplant the steam locomotive in trunk line, passenger and freight service? I do not anticipate that all roads will soon adopt electricity, for the steam locomotive will hold its field in this country for many years to come, but I do expect, judging somewhat from "positive knowledge," a remarkable development to soon begin in the electrical equipment of favorably located steam roads. From Richmond, where the first commercial electric road was built, to the present is but 17 years, yet within that time the horse has been relegated to the past as a serious factor in transportation, the cable has served its usefulness and awaits its end, and the suburban railway has been developed and is now rapidly encroach- ing upon the field of the steam railroad. With the terminals of the two greatest roads in the United States now being equipped electrically and with an investment of something more than $4,000,000,000 in electrical industries made within a quarter of a century, we have reason to feel satisfied with the past. With several of the leading roads in this country, of England, of Sweden, of Switzerland, of Italy, and Australia electrically equipping branch lines and seriously considering changing large ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. 91 portions of their present systems from steam to electricity, we, as personal factors in this great industrial advancement, have every reason to be hopeful for the future. DISCUSSION. PRESIDENT EGBERT KAYE GRAY : I do not know whether I am perfectly in order under the American procedure or not, but our habit on the other side, when we receive an address from our President, is to tender him our thanks. As President Arnold has said, during the Paris Exposition we had a joint meeting of the two Institutions, and I am very glad indeed to say that we have in this hall to-day the two gentlemen who presided on that occasion, namely, Mr. Carl Hering and Professor Perry. I do not think that any one could even have wished to criticise, in any way, the address which has been so ably given by your President, because if any man, either on this side or on the other side of the Atlantic, is pre- eminent in connection with the subject he has treated, I think it is President Arnold. His name is exceedingly well known to us on the other side, and I think I am not giving away any secret in telling you that the evidence of his work which he has been tendering to us has received a very warm reception there, and the evidence is considered to be the best that can be obtained in relation to the matters with which it deals. I therefore wish, in the name of the Institution of Electrical Engineers of Great Britain, to tender to my colleague, President Arnold, our very sincere thanks for hit exceedingly able address ; and, with your permission, I will ask the senior Past-president of the Institution of Electrical Engineers of Great Britain to second the motion Colonel Crompton. COLONEL R. E. B. CROMPTON : It is with the most heartfelt pleasure that I rise to second the motion of President Gray, that the thanks of the Ameri- can Institute of Electrical Engineers, as well as our own Institution of England, be given to President Arnold for his address, which I personally feel is worthy of this great occasion the meeting of the two Institutions. PRESIDENT GBAY: I presume it is unnecessary to put this motion to the meeting, and I shall put it by acclamation if it meets your approval. PRESIDENT BION J. ARNOLD: I assure you that your expression of ap- proval is very much appreciated indeed. We have for our discussion this morning a subject similar to that which I have treated in my address ; in fact the address was written as a sort of introduction to the discussion of the subject entitled " Different Methods and Systems of Using Alternating Current in Electric Railway Motors." This subject has received the attention of engineers interested in electric railways for the past three or four years. During the past two years it has received very energetic attention on the part of leading engineers of Europe and this country, and it bids fair to be one of far greater importance as we : get more thoroughly into heavy railway work. Since I have talked to you quite a while, notwithstanding the fact that my name appears first on the program to discuss the question, I am going to ask a gentleman to open the discussion who is one of the most distinguished engineers in this country, .and one of the most distinguished living authorities in electrical matters. I have the pleasure of introducing Dr. C. P. Steinmetz, of the General 92 ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. Electric Co. and Past-president of the American Institute of Electrical Engineers. DR. C. P. STEINMETZ: The problem which we have before us here for discussion the problem of the direct application of alternating currents to electric railways is not a new one, but it has become of primary importance and interest in the last few years. The early pioneers in electric rail roading, 10 or 15 years ago, started the development of the alternating- current railway motor, and prominently among them I may mention Mr. R. Eickemeyer and Mr. Vandepoele, who designed alternating motors for railway purposes and investigated their characteristics. However, very little progress was made in this field for many years, for a number of reasons; one being that in those early days frequencies of 125 to 130 cycles were customary, far higher than we are using now and the difficulties of the problem were thereby increased so formidably that advance was neces- sarily very slow. In addition the very rapid development of the direct- current railway motor fully occupied the attention of all electrical engineers, and therefore the less urgent field of the alternating-current motor was necessarily somewhat sidetracked. Then the alternating-current poly- phase induction motor came into the foreground, showed its superiority over other types of motors for stationary work, and impressed the engineers to such an extent that for a long time it overshadowed the work done by the early investigators on the variable-speed alternating- current motor, that is, on motors with series characteristic. Attempts then were made to intro- duce this very successful polyphase induction motor into electric railway work, attempts which have not been successful to any great extent. In the meantime, in the United States the synchronous converter was developed and became a standard piece of apparatus familiar to everybody standard as much as the direct-current generator and the alternating-current generator, and experience with such synchronous converters shows that for electric railway work, for the violently fluctuating loads on the railway system, the synchronous converter is superior even to the direct-current generator: the absence of armature reaction, the phase control of pressure feasible in the converter, and corresponding close pressure regulation makes it specially able to withstand and take care of very violent fluctuations of load and to carry overloads which no direct-current generator can carry. This apparatus became standard, and with its introduction the field of the direct-current railway motor the distances which could be covered by the direct-current railway was extended practically without limit, and a field opened which has been exploited in the last years, which was the field dreamed of by the early pioneers ; the difficulties, however, being over- come, not by the development of the alternating-current motor, but by the development of methods of transmitting alternating currents and transform- ing them into direct currents along the routes, in synchronous converter sub- stations. Now, however, in the last year or two, with the still further development of the electric railway we have approached and in many instances reached the limits of applicability of this synchronous converter. The synchronous converter is a piece of machinery which requires sub-stations, requires some attendance, and as a necessary result has a high economical efncieney only ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. 93 where the traffic is sufficiently condensed to warrant the maintenance of sub-stations within relatively short distances from each other. Where the number of trains is less or the power per train greater than can be supplied at 500 volts from sub-stations, without excessive expenditure in line con- ductors, and too excessive fluctuations of load, pressures are required higher than can be utilized efficiently in direct-current motors, and there we strike the limit of the synchronous converter, and the alternating-current motor has to come in. Personally I do not believe that the alternating-current motor will make very serious inroads in the field now occupied by the direct-current railway motor. I do not believe that direct-current railway systems will be changed into alternating-current railway systems; but what I expect of the alter- nating-current railway motor is that it will find a field of its own, a new field; just as when the alternating-current method of distribution was developed in this country, it did not displace the direct-current method of distribution which occupied the centers of our large cities, but it found a field of its own, a field which has gradually developed so as to be equal in importance if not superior to the field occupied by the direct current. Hence, to conclude these remarks, what I expect of the alternating-current railway motor is that it will find and develop a field of its own, that field which the direct-current railway motor cannot reach suburban and inter- urban service, long-distance service, secondary railway service. When considering the technical aspect of the subject before us for dis- cussion to-day, the relative advantages and disadvantages of the direct- and alternating-current railway motors, we have to consider, first, the character of the problem we have to meet in electric propulsion; secondly, the character of the apparatus which we have available to solve these problems ; thirdly, the additional features imposed upon the problem, or conditions more or less outside of the problem, as, for instance, the condition of the electrical industry at present, the existing investment in direct-current and in steam railroads, which have to be taken into consideration when dis- cussing any new system of railway propulsion. Regarding the characteristics of the different types of motors the direct- current series motor now in universal use for railroad work, the polyphase induction motor proposed, and, to a certain extent, tried in recent years for railway work, a motor eminently successful in stationary work and the alternating-current single-phase railway motor with commutator, I have in a paper before the International Electrical Congress given the results of a theoretical investigation and discussion of these different motors and shown the speed-torque curves, or characteristic curves of these motors in relation to each other. In Fig. 1 is given a comparison of the typical speed-torque curves of the different types of motors. In general, the characteristic of the polyphase induction motor is essen- tially that of a constant-speed motor, with shunt-motor characteristics; that is, it can efficiently operate over a certain limited range of speed only, cannot exceed the synchronous speed, and when operating below its normal speed, it operates less efficiently; that is, when operating at a lower speed than normal, or approximately synchronous as can be done by a rheostat in the secondary circuit, the polyphase induction motor merely wastes that 94 ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. part of the power corresponding to the difference between its actual speed and synchronous speed. Or, in other words, at low speed the induction motor consumes the same power which it consumes with the same torque at full speed, though its power output is reduced in proportion to the speed, and its efficiency correspondingly. Speed Torque Characteristics of Railway Motors FIG. In the direct-current series motor the torque developed by the motor de- creases with increase of speed, and inversely with increasing load the speed of the motor decreases. The maximum torque of such a motor occurs in ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. 95 starting. All variable-speed commutator motors, alternating and direct, more or less differ from each other in the rate at which the torque varies with the speed, and that brings us to a consideration of the requirements of electric propulsion. Important classes of work to which an electric motor may be put in loco- motion are: first, city railway or tram car work; secondly, rapid transit service as on elevated and underground roads ; thirdly, suburban and inter- urban service ; fourthly, trunk line passenger service ; fifthly, long distance freight service, and sixthly, elevator service. Now, discussing these briefly in succession. The city tram car service is characterized by its frequent stops of irregular duration, at irregular intervals. To maintain good average speed it is therefore essential that the motor should get under way after the stop as rapidly as possible; that is, have a very high starting torque and accelerating torque, and carry this high torque up to a considerable speed. Beyond this speed, then, the torque of the motor should decrease fairly rapidly down to the torque required to run on a level track, which we may assume roughly to be at twice the speed to which the high torque of acceleration should be maintained. In addition thereto, it is necessary that the motor should accelerate efficiently and that it should be able to operate efficiently at low speeds in those city districts where the general traffic is dense and where it is not possible to run at high speeds. The characteristics of this type of motor are pre- eminently given by the direct-current series motor. If we assume the torque required to run on a level track as 1, probably the starting or accelerating torque may be something like six times as high. At that torque we start and run up to considerable speed, and then strike what is called the motor curve and after cutting out all regulating devices, accelerate with decreas- ing velocity up to the free-running speed. Such a curve, that of a typical direct-current railway motor, is given in Fig. 1, marked " direct-current parallel " and " direct-current series." The induction motor, although it may accelerate with a high torque, at the end 'of acceleration, the speed is limited. It accelerates up to or near synchronous speed, and there the torque falls off to zero ; and hence that part of the torque curve which is so essential to city tramway work, the curve of running with decreasing torque, from the limit of acceleration to the free-running speed, does not exist in the induction motor. We can indeed reach the free-running speed of a direct-current motor with a polyphase induction motor by gearing it to twice the speed, making synchronism the free-running speed, but this means that the available torque of acceleration, and therefore the rate of getting under way, is reduced by one-half, or, if we make it the same, the motor capacity is twice as great, requiring a motor twice as large. Considering that in this service a very considerable part of the running time is occupied running on approximately level track with torque very small compared with the accelerating torque, we see that the highest possible efficiency of the motor at light load is essential. Here, however, is the place where the induc- tion motor falls down. A polyphase induction motor running at, say, one- tenth of its maximum output runs very uneconomically and with very poor power-factor. So in the polyphase induction motor, when used for railway service, you cannot combine very high acceleration with high efficiency in 06 ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. free running, and with the ability of running efficiently at low speeds, aa you can in the direct-current series motor with series- parallel control. Therefore, the induction motor is not suitable to the class of work which we call city service or tram-car work. The alternating-current commutator motor of which two sets of curves are shown in Fig. 1, marked " alternating-current parallel " and " alter- nating-current series " has characteristics very closely similar to the direct- current railway motor, except that possibly the variation of torque with the speed is less. That means, with the same decrease of speed the torque does not increase at the same rate as with the direct-current motor; if we assume again the same free-running torque as 1, and the torque of accelera- tion six times the free-running torque, the direct-current series motor will carry the acceleration torque up to half speed ; the alternating-current motor not quite as high. This means with the same maximum acceleration you will strike the motor curve at a lower speed, accelerating on the motor curve, you get under way, then, slightly slower, or to get the same average acceleration you have to start with a higher maximum acceleration. Now, this is an advantage in some cases in so far as you run for a longer period of time and over a wider range of speed on the motor curve, that is without controlling devices, hence in the most efficient manner possible, and thereby make up to a considerable extent for that power which the alternating- current motor inherently loses by its slightly lower efficiency due to the alternating character of the magnetic field, and the losses by magnetic- hysteresis in the motor field cf the alternating-current motor which do not exist in the direct-current motor. This difference in the speed-torque curve of the alternating-current series motor, compared with the direct-current series motor, is due to the lower magnetic density used in the alternating- motor field, and, at low speeds, also to the e.m.f. of self-induction. The first phenomenon, therefore, also occurs in an unsaturated direct-current series motor ( Fig. 1 ) , and such a direct-current series motor therefore has at high and medium speeds, the same characteristics as the alternating-current motor. It is undoubtedly true that alternating-current motors can be designed to give very closely the same characteristics as the standard direct- current series motor. However, the motor as it is before us at present reaches the motor curve at a lower speed, therefore with the same maximum acceleration, gives a lower average acceleration up to full speed, or with the same average acceleration requires slightly higher maximum acceleration. Coming now to the second class of service, rapid transit service, here the problem and the conditions of operation are almost identically the same as in city service, except that the units are larger, the speeds are higher, the stops not as frequent, absolutely, but about just as frequent relatively in comparison to the maximum speed of the motor, so that we can directly apply our considerations to rapid transit service regarding a comparison of polyphase induction motors of alternating-current commutator motors, and of direct-current motors. In interurban and suburban work, that is, in railroads running out from the cities far across the country into the suburbs or into other cities, we have a much lesser frequency of stops. That means that rapidity of acceleration is of lesser importance, and we can well get along with a lesser average ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. 97 torque of acceleration, but we must have the same surplus torque as on city service work, or rather a greater surplus torque, because, while in city service and in rapid transit service, where the distances are relatively short, we can count on maintaining fairly constant pressure in the supply system, we cannot to the same extent count on this in interurban and suburban service where we are far away across the country, except by investing much greater sums in line conductors and feeders than is commonly economically desirable or feasible. Hence, in this service the motor should have a greater surplus torque than in city service, so as to get a sufficient margin to start the train or the car under the most severe conditions on an up- grade or an overload, even if the pressure in the system is low. The motor which is most sensitive to pressure variation is the polyphase induction motor. The maximum torque which this motor can give necessarily cannot very much exceed the acceleration torque without badly spoiling the characteristics of the motor either electrically or mechanically; but the maximum torque varies with the square of the pressure and hence rapidly decreases if the pressure of the system is low. In the motors with series characteristics, however, like the single-phase commutator motor, the direct- current series motor, the torque does not depend on the pressure, or rather, while the maximum torque so depends, the theoretical maximum torque which you get from the motor when standing still is so far in excess of the torque of self-destruction, or rather of slipping the car wheels, that it is not reached, and the effect of variation in the supply pressure is merely a variation in the motor speed. That is, if the pressure is low in the system, the direct-current motor and the alternating-current commutator motor run at lower speeds, but still are able to give the same torque, while the polyphase induction motor runs at the same speed, but is not able to give the same margin of torque, and at a certain load falls down or does not start. That means that in designing a system of transmission and dis- tribution for alternating-current commutator motors or direct-current motors we are permitted to design the system for the average drop of pres- sure in the system while in designing it for induction motor service, we have to take into consideration the maximum drop of pressure in the system which is very much greater that the average. For interurban and suburban service we require an excess overload in torque, but do not require an acceleration up to high speed. The alter- nating-current commutator motor appears to be preeminently satisfactory in this work, and there is where I believe it will be used extensively, and where the advantage of a high-pressure trolley and of the absence of sub- stations is specially important. In trunk line passenger service the rate of acceleration as given at present by the steam locomotive is very much less than every- day practice in electric railway service. So here we do not need this excess acceleration torque sustained up to high speeds. Here, again, we find a field where we may apply the alternating-current commutator motors. The polyphase induction motors could be used if the question of pressure supply did not come in, as I discussed above, and if furthermore the limit in speed of the induction motor was not so objectionable in passenger trunk service, where, more than anywhere else, we desire to get the full benefit of the track by running at ELEC. RYS. 7. 118 ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. the highest safe speeds wherever the track is lerel. This the induction motor with its limited speed cannot do. In trunk line freight service, the same considerations come in, except there the speeds are relatively low, the train weights great; and it is more than anywhere essential to have a very large surplus torque available to get under way or to hold the train on a grade. You must, therefore, in this class of service, just as in suburban and interurban service, have a motor running efficiently at light load, but being able to give very high torque, although it does not need to carry this torque up to high speeds. On the contrary, it is not desirable in freight service that the motor should sustain a high torque up to high speeds, because that would mean the con- sumption of very large power. In freight service the highest possible economy is especially necessary, and the highest possible economy means the least fluctuations of power consumption; that means on up-grades you would desire to go slowly and reduce the power consumption and get the high speeds on the level track. In mountain railways and such classes of work, the running torque is of the same magnitude as the starting torque, and so the load on the motor is more nearly constant than in any other class of railway work, and on the down-grade the motor is preferably used for braking, by returning power into the line. Here then the polyphase induction motor appears well suited, and is indeed being used successfully. Such service, however, is in its character more nearly akin to elevator service than to railway service. In the discussion so far I have considered the requirements of the different classes of railway service, irrespective of extraneous conditions. When con- sidering the alternating-current motor and the direct-current motor, we have to take in view what exists at the present time in this country and abroad. There exists the enormous network of steam railroad and of direct- current electric railways. The steam locomotive is a unit of very high efficiency, but a very large unit. It therefore for efficient operation requires the massing of traffic in heavy trains, and results in less frequent but large trains. This has practically rearranged and reorganized the whole system of locomotion by collecting it into a small number of very large units. That is not the most efficient manner of operating electrically propelled vehicles, but rather the contrary. Furthermore, you have to consider that every city and almost every village has a direct-current railway system. Now, the main and most important features by which the electric railway motor and electric propulsion has gained and is gaining rapidly in competition with the steam locomotive, appears to me to be the frequency of headway and the absence of passenger stations, not the speed, which frequently in electric lines is lower than that on steam railroads paralleled by them. The electric railway picks up its passengers anywhere in the city and deposits them anywhere and it does not require them to consult time tables, but runs its cars so frequently that the passenger can always find a car within a few minutes at any point; on the other hand, the steam locomotive re- quires you to consult a time table and go to a depot. As soon as the electric railway gives up this advantage which I have just mentioned, I believe one of the main advantages of the electric railway over the steam railroad will be lost, and this, therefore, is the feature which has to be kept ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. 99 in view. It means that whatever type of motor may be adopted in inter- urban or suburban service, etc., it must be able to carry the passengers through the cities over existing railways. The existing railways are direct-current railways, and I believe will remain so. That means that the long-distance motor, at least the sub- urban and interurban motor, must be able to run over the direct-current system. Hence, it must be a type of motor equally applicable and capable of operation on a high-pressure alternating-current or on the 500 volt direct- current system. Taking this for granted the methods of control must also be as simple as possible; that is, the same control for alternating as for direct current. Even if the motor could be used on direct current and alternating current, if we would have to carry a double system of control, one for city service and direct current, the other for long-distance service and alternating cur- rent, this would be a very serious handicap. It means that really to solve the problem before us, of extending the electric railway into interurban and suburban service, and into the field now occupied by the steam railroad systems, and into new fields not yet developed, to a large extent not even dreamed of, that we must have a motor which with the same controlling appliances and the same characteristics, can run either on the high-pressure alternating circuits or on the existing direct-current circuits. Furthermore, the enormous investment in electric railway systems existing at present has all been made, in the large systems, on 25-cycle, three-phase apparatus. That means that we shall have to continue to operate at 25 cycles. It may be preferable, possibly, to run at lower fre- quencies, or it may be preferable to run at higher frequency in this instance or that instance, but regardless of whether it is preferable or not, if it can be done on 25 cycles, it will have to be done on 25 cycles, and if another frequency had to be used, it would be a very severe handicap to the new system. I am glad to say that there is no doubt that 25 cycles is the frequency best suited to the alternating-current single-phase railway motor. PRESIDENT GRAY: Dr. Steinmetz' remarks have been so clearly stated and so closely reasoned out that they do not give us much chance for dis- cussion, but I am glad to refer to my English colleague, Professor John Perry, upon whom I call to take part in this discussion. PROFESSOR JOHN PERRY: I have to confess that 1 am not prepared to take part in the discussion. We have had the address of President Arnold and this excellent address of Professor Steinmetz, and two such addresses in one morning I think we have never had before. Clearly, they are men who have thoroughly studied the subject, and in view of what they have said, I think what it comes to is this that everything seems to depend to a very great extent as to what is to occur in connection with the electrification of steam railroads in the next ten years, on the success of the single-phase alternating- current motor. I knew of the progress that had been made by the General Electric Company and the Westinghouse Company, I had heard a great deal about it before leaving the other side, and it is one of the things that I promised myself to learn something about during my visit here. I have not yet been able to do much in the way of getting accurate knowledge on the subject. I have been on a tram-car at Schenectady, the motor of which, 100 ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. I was informed, was driven by direct current and the car ran well; and then I would get on another car, and I was told that the motor was driven by alternating current, and it seemed to run just as well, so that I was not able to acquire any knowledge. 1 had no means of experimenting or ascertaining what the efficiency of the various arrangements were. Some 10 or 12 years ago I was tremendously interested in the single-phase alter- nating-current motor, perhaps for a selfish interest, as I had invented a sys- tem of traction which required the use of that system. I suppose we are all tremendously interested in this thing, and are all anxious to learn what we can about the alternating-current motor. I wanted to go to the section in which Mr. Steinmetz was giving an account of the work yesterday, but I was told it was my duty to attend a discussion upon the subject of electro- magnetic units in another section, and as a man cannot be in two places at once, I had to attend to my duty as it was pointed out to me. In these circumstances, I can only say that I should like to hear the discussion of this subject proceed further before I shall feel able to take any part in it. PBESIDENT AENOLD: It has been said that the fame of a scientific man is a quiet fame, but that is the most satisfactory after all. It does not attract the multitude. A man is able to walk in a crowd without being pointed out, which by the way, is a very satisfactory thing to do; but he finds that when he reaches different parts of the world his name has pre- ceded him in the circles in which he moves, so that he after all enjoys in the most satisfactory way the results of his efforts in the particular line of work which he has been following. We have many such men present to-day, and among them is one who has done excellent work in the special line we are discussing this morning. I shall now call upon one of our dis- tinguished engineers and colleagues, Mr. B. G. Lamme, of the Westinghouse Electric & Manufacturing Co., of Pittsburg, to discuss the question further. MB. B. G. LAMME: Away back in the dark ages of electric traction, about 15 years ago, there was great confusion in the types of apparatus used. There were all kinds of motors and all kinds of apparatus on the car. They only had one property in common they were all direct-current. After putting a number of these systems into commercial use it was dis- covered that certain types of apparatus were superior to others, and those particularly interested in the manufacture of such apparatus followed up this matter to ascertain what properties were of the greatest value. It was gradually discovered that one type of motor was taking precedence of all others, namely, the series motor. Practically all development for a certain time was in the direction of the direct-current series motor. The reasons which led to this were partly based on theory and partly on practice. The series motor gave the effect of a cushion on a car. The motor is inherently a variable-speed machine and automatically varies its speed with the condition of the load. That was discovered to be a matter of first importance in the smooth operation of electric cars. Also the motor auto- matically increases its torque in a greater proportion than the current, which is of great importance in regard to starting and acceleration. These points were possibly not as well understood at that time as at present, but experience showed that certain equipments were superior to others and development was along that line. ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. 101 .After a few years, when the motors had reached standard proportions and practically but one type was used, a second limitation was discovered ; namely, in the transmission conditions. It was found that in the extension of the railway system, the ordinary 550- or 600-volt direct-current system was becoming cumbersome, and it was evident that some method of trans- mitting power at higher pressure and transforming to lower pressure for utilization would be necessary. The most evident method was naturally to transmit by alternating current and convert to direct current, in order to use existing car equipment. This led to the use of motor-generators, and later to synchronous converters. The motor-generator was found to fit the existing alternating system fairly well, but in the development of the synchronous converter the manu- facturers discovered a great difficulty in existing systems. The frequencies of 125 and 133, which were the standards for many plants, were entirely unsuited for synchronous converters and also not well adapted for synchro- nous motors. Another frequency, coming into general use, namely, 60 cycles, was found to be possible for use with synchronous converters, but the difficulties of design were very great in that case, and the synchronous converters were rather heavy and cumbersome. At that time there was fortunately a new frequency adopted which was of prime importance in the development of the synchronous converter, namely, 25 cycles. So far as I know, the origin of that on a large scale, was as follows : in the Niagara Falls power plant, when it was first laid out, the engineers for the power company had arranged for a frequency of 2000 alternations per minute, or 16% cycles per second. They wished to use 8-pole machines, running at 250 revolutions. The company which I repre- sent, which was one of the prominent bidders on the contract, objected seriously to the proposed frequency, as it was considered entirely uncom- mercial and also not suited for the best design of machine. The engineers of this company recommended 4000 alternations per minute or 33% cycles per second. That was considered extremely low compared with anything then in use. As we could not come to any agreement to use that frequency, we finally compromised on 3000 alternations per minute, or 25 cycles per second, and the first Niagara machines were built in that way. There were various reasons for the adoption of a low frequency, one of which was that commutator type of motors might possibly come into use. Another reason was that it was better adapted to synchronous converters, but it was admitted that 33% cycles would also be satisfactory. After the Niagara Falls plant was installed, there was then a precedent for the adoption of this frequency for large units, and the manufacturers began to build apparatus of this frequency for the Niagara Falls plant and also adopted it for other plants. This opened quite a field for the synchro- nous converter and it soon began to be extensively used for railway work, as it was recognized that this was the link needed for extending the direct- current system. Even at the early date of 1893 and 1894 it was believed by many engineers that the synchronous converter was simply a machine to meet an emergency condition, that it would not last, that the time would oome when synchronous converters would be dropped from the railway service, but as the most convenient and apparently the best solution of the 102 ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. problem, it was adopted extensively. About that time electric railway service began to be greatly extended and synchronous converters have thus come into very general use. By the use of synchronous converters, the advantages of the alternating-current system in transmission are obtained and the advantages of the direct-current system with the series motor are retained. Distances could be extended indefinitely by increasing the number of synchronous converter stations and raising the pressure of the alter- nating-current lines. Shortly after this system came into general use it was recognized that a purely alternating-current system, in which purely alternating current was supplied to the motors, would be advantageous and considerable work was done along that line. The polyphase motor apparently had' the field, and naturally the manufacturing companies took up the question of the application of the polyphase motor to traction work. The company which I represent, the Westinghouse Electric & Manufacturing Company, took up this question in an active way about 1895, and built two motors of 75 hp each for traction work. These motors were equipped with collector rings and rheostatic control and tests were made in regard to performance, both with straight rheostatic control and with the new well-known " tandem " control, in which the secondary of one motor is connected to the primary of the other to obtain half-speed conditions. Even with this latter arrange- ment it was found that the motors would not compare at all favorably with the direct-current motor or the system with the direct-current system using rotary converters, and this work was abandoned. It was recognized that the polyphase motor did not possess the proper series characteristics which long experience had shown to be so necessary for railway work. Other experiments along this line were made, using polyphase motors wound for two or more speeds, and two 100-hp motors were built which were wound for several speeds. While this was better than the other arrangements, it still appeared that this was not a solution of the problem. Previous to this time the company had done some work in the direction of using single phase, but not as a solution of the problem which presented itself in 1895 and later. In 1892 the question of the use of the commutator type alternating-cur- rent motor for railway work was taken up. Two motors of nominally 10 hp each were designed and built. These were built for a frequency of 2000 alternations per minute, or 16% cycles per second. They were mounted on a car and were operated for awhile, but the system was not a success. In the first place the pressure used 400 volts as compared with 550 in the direct-current motor was rather low. It was considered that as 550 volts was the limit in the direct-current motor, 400 volts would be the limit with alternating current. The motors were tested on a track of iron rails with practically no bonding. The track drops were excessive and the pressure fluctuations were great. The generator used of about 20 kw capacity was entirely too small for this work and it was not adapted to handle the inductive loads which were found with alternating-current motors. A series of tests was run and it was finally decided that for city work, for which the system was then laid out, the motor could not compete with the direct-current motor. It was decided, however, that such a type of motor ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. 103 would probably furnish the solution of the heavy railroad problem, but as there was no such heavy railroad problem at that time, the work was dropped for awhile. But in 1897 the question of the use of the commutator type of alternating- current motor was again taken up this time on a some- what larger scale. Motors of 50 hp were built for variable-speed work, and given a long series of tests. Then after sufficient experience had been obtained, the work was gradually carried to the larger sizes. In 1900 and 1901, when the question of the polyphase traction in Europe was so extensively advertised, it became evident that there was actually a demand for an alternating- current railway system. It was therefore decided to continue the previous work with large motors of the commutator type, and two motors of 100 hp were designed and built. For these also, the frequency adopted was 2000 alternations per minute, or 16% cycles per second. This fractional figure was primarily adopted on account of certain steam-engine conditions. It was recognized that an even frequency of 16 or 18 would have been practically as good. In the earlier work, with the 10-hp motors at the low frequency, it was recognized that it would be absurd to put such a system on the market, as at that time even 25 cycles had not been adopted. The frequencies in common use were 50 or 60 and a drop to 16 cycles was considered pro- hibitive. In the latter work, as 25 cycles had come into general use, and 15 or 20 cycles had been talked of and proposed by certain companies, it was considered that in view of their advantages for railway work such fre- quencies should be adopted. The motors were hence built for the above frequency. The results obtained with these large motors were so satis- factory that a contract was taken for a rather large road and the apparatus prepared. Knowing that news of this would soon be abroad, it was decided that the matter should be brought before the American Institute of Elec- trical Engineers, and a paper was presented on the 26th of September two years ago which I believe was the first announcement of the application of the single-phase alternating current to railway motors. There was con- siderable discussion mostly criticism and it was generally considered by the engineering public that the weak point of the system was the commu- tation. At the present time, however, I believe this is no longer considered as a serious point. Previous to building the 100-hp motors we had had considerable experience with the commutation of such motors. Besides a long series of tests, we had run 40-hp motors at practically full load on a 60-cycle sys- tem for nine months, day and night. At the end of the nine months the commutators were in practically as good condition as in the beginning, showing that the commutator on such machines could be made to have a long life. The conditions of the 60-cycle machines were much worse than on the lower frequency, and the nine months of operation under the con- dition of steady service probably equalled two or three years of traction service; but the commutator stood up so well that we decided definitely that there was no difficulty on that point. The principal reasons which led to the adoption of the single-phase motor were stated in the paper above referred to, and were that but one trolley wire would be required and that the motors had the series characteristics. 104 ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. It was considered that no motor, except one of the commutator type would give suitable characteristics for the service, and it was stated that there were several types of motors, with commutators, which had the proper characteristics. All of these may be classed as series motors, although some of them are combined with transformers and may be considered as transformer series motors, or, under another name, as repulsive motors, and others are pure series motors. The pure series motor is one which can operate on direct current as well as alternating current. The repulsion motor can be modified so as to operate on direct current, but as ordinarily arranged it is not as well adapted for this as the other type. It was recognized in the first undertakings with this system that the motor would probably be required to operate on direct current at times, and the fact that the pure series motor was primarily a direct-current motor of a first-class design was one of the reasons which led us toward the adoption of that type. As both theory and experience indicated that such motors would probably be wound for 200 or 250 volts, it was recognized that the motors would probably have to be operated in series for direct current, and either in series or in parallel for alternating current as might be desired. The arrangement required for permitting operation on direct current as well as alternating are rather complicated, due to the fact that it is necessary to switch from one system to the other in passing from the alternating to the direct current. We did not suppose that the electrical public would con- sent to such a combination, but since that time we have found that in some instances they do not object seriously to the increased complication. At the time that the alternating-current system was brought out it was considered that the principal field would be in heavy railway work, because this motor furnished what was considered a general solution of the railway problem; as the railways would have their own terminals and their own rights of way, the system would be an alternating-current system through- out. At the present time, however, roads are being installed which operate primarily on alternating current, but at the terminals and where they pass through intervening towns they operate on direct current. The direct-current motor has never been considered as entirely suitable for the heavy railway problem, as usually but two speeds, and at most but three speeds can be obtained with four motors, the third speed increas- ing the complication considerably. With the alternating-current motor of the commutator type any speed can be obtained for locomotive work, because any pressure can be applied to the terminals of the motor. As soon as alternating current is used for motors, we at once have a ready means of pressure transformation. As on locomotives for large capacity the diffi- culty of handling the current is considered a very prominent one, it was considered that some form of pressure control which varied the pressure without opening the circuit would probably be the best one. One form of pressure control permissible is what is called the induction regulator. This regulator varies the pressure without opening the circuit. The relation of the primary and secondary windings with respect to each other is varied. This gives a means of varying the pressure to the motors and varying the speed of very large motors with no tendency to sparking at the controller. The only time the circuit is opened is at the end of the operation when ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. 105 cutting it off. Therefore it was considered as an important feature in the solution of the general railway problem. The single-phase system is the one means presented at the present time as the solution of the heavy railway problem. It has all the advantages of the direct-current motor in the variable-speed characteristic, and has also the advantage possessed by alternating current in the ability to use any line pressure desired, and to vary the pressure applied to the motor and thus vary the speed over any range desired. It also has the advantage of permitting a system of control that can be obtained without sparking. In the adaptation of the alternating- cur rent motor to direct-current serv- ice, two 250-volt motors can be connected in series for 500 volts; also in operating on alternating current the motors can be connected in series, if desired, or in parallel. There is a possibility of danger in operating two motors in series in this way on alternating current, or even on ordinary direct current. In ordinary direct-current practice the use of two motors in series for part of the service is common practice, but there is this differ- ence between the direct-current equipment and the alternating-current equip- ment. In the direct current we have motors wound normally for 500 or 600 volts. When operating in series the motors are connected, two in series, each one receiving 250 volts. Therefore, if one motor should slip its wheels and take the full pressure of the pair, it would still be operating at its normal pressure. But with two 250-volt motors connected across a 500-volt circuit, we have a different condition. In case one motor should take the entire pressure, we should have 500 volts across a 250-volt motor. That condition was considered early, and in the Washington, Baltimore, Annapolis project, a description of which was given in the American Insti- ture paper read two years ago, we showed an arrangement by which this could be avoided. We had balancing transformers connected across the two motors in series. The balancing transformer was across the outside terminals, and a tap from the middle of the transformer was connected between the two motors. In this way equal pressure was supplied to the two motors in series, and the danger of a runaway was thus avoided. It is not yet determined how important this is, but I believe that something like this will be found advisable for the operation of motors in series, especially where high-power motors are used on medium weight cars for high-speed service. Possibly with comparatively low speed, and with very heavy cars, there may not be the same tendency to slip. On the direct- current part of the road, of course, the balancing transformer could not have any effect; but as the direct current is usually a very small part of the service, this danger would be lessened, due to the proportionate time in service. In the application of the motor to use on both alternating and direct current, we have found some special conditions which affect the arrange- ment of control. Take, for instance, a large road being installed between Cincinnati and Indianapolis, where it is intended to run on direct current at the terminals and alternating current on the rest of the line. The normal speed on the alternating current part of the line is so great that it would be prohibited in the towns, and it is found that to get the speed down to the desired rate in the city service on the direct-current portion of the road, 106 ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. it is necessary to connect the four motors all in series, and thus no series- parallel arrangement can be used. Pure rheostatic control is therefore necessary in the city. On the suburban part, a switch is used to throw the current from direct to alternating, simply throwing the four motors in parallel, and taps are used on the lowering transformers to get a number of pressures. In that way we get the effect of series-parallel control and even better, by having more than two steps. On a long line it is possibly of no great advantage to have many steps, but as a rule the more steps there are, the easier is the service on the controlling apparatus, and the more running speeds are available. With regard to the application of the system to locomotives, on the steam roads where the systems are not tied up with existing electric plants, it is probable that in time the railroads will adopt their own pressures, and possibly their own frequency. This may not be 25 cycles but may be some- what lower. I believe that the electrification of the steam road may be a controlling factor in the change from direct to alternating current in city service. If the large railroads with their own large power plants adopt alternating current throughout, then the towns lying along the roads will in time probably adopt the same power system, and even the large cities will sooner or later adopt the same system. At the present time the rail- roads, as far as they have gone, have adopted direct current because the cities through which they pass or enter are using direct current. When the railroads make the big end of the project, however, then the cities will adopt what the railroads are using. When this comes about the direct- current railway systems in the cities will be superseded by the alternating. ME. C. V. DEYSDALE : At this late hour in the discussion, I do not pro- pose to take up your time very much, especially as I am afraid that very few of us over in England have had much experience on this important subject. I should like, in the first place, to take this opportunity of con- gratulating you on this side of the water on having carried this important problem to such an extremely successful issue as has been recently shown in Ballston and in other places. I think this subject has been worked on in several places, yet to America belongs the honor of having constructed the first line of any considerable length working on the single-phase system. We must still further admire the way in which it has been done when we re- member that the result has been achieved by getting over the great diffi- culties that stood in the way of the series motor, and that in so doing it has been found practicable to use the same motive plants on direct- and alter- nating-current lines. That, in itself, is an enormous advantage over and above that of being able to use the single-phase alternating current. It would be impossible for anyone to criticize any of the statements that have been made this morning, because they come from gentlemen who have had such exceedingly minute experience in the special branch of the subject, that their remarks must be taken as gospel, at any rate for the present. My object in taking part in the discussion is rather to bring the matter back to first principles. This subject has been worked upon in many differ- ent ways, and although the laminated series motor, which seems to have been the first to give us results, will probably explain and solve the problem, yet AJiXOLD: ELECTRIFICATION OF STEAM RAILROADS. 107 there are some interesting questions as to whether there are any other ways of fulfilling the problem which may have other advantages. There is one thing that does not seem always to be kept in view in traction matters, in the starting of the cars, and that is the very simple matter that in the starting of the car you do not require power, you require force; if you wish to get anything into motion, what you require in the first instance is purely force, and until the body moves, it does not require power at all. One of the great advantages which the steam-engine has over any electrical system up to the present time, is the fact that when you first turn the steam into the locomotive you get the pressure on the back of the cylinder and get the starting force without taking any power from the steam. If it were not for the other disadvantages of the locomotive, there is no ques- tion that that one point would give it a strong pull over anything we have electrical, because if we turn to the ordinary direct-current motor, we find that we have to use half, or with one motor, the whole, of the full-load power merely to secure a starting torque. This has several objections. Not only is this uneconomical and wasteful of power, but it throws a sudden strain on the general plant, and furthermore has to be wasted in resistances, and these resistances sometimes attain a considerable magnitude. With alternating-current motors these matters are worse, as we have in addition low power-factors and consequently difficulties in regulation. The time is too short to refer to many other systems, but I will mention one, that known as the Ward-Leonard system, which at first sight appears to be an unworkable one. In the Ward-Leonard system, as I understand it, the system is to use a single-phase motor coupled to a direct-current generator which runs direct current on the locomotive or cars. Of course, the indirectness of the method seems to put it at fault, but on the continent that method has been developed with considerable hope of success, in fact with considerable practical success ; and it has this great advantage that by the use of this arrangement you can start get your starting effort with very small power taken from your station. In the other system it is too well known for me to describe it here you have your single-phase motor continuously running, and you can do the whole of the regulation of your speed, etc., by merely regulating the excitation of the generator. The result is that it is possible to get the full starting effort with only something like one-third or one-quarter of the full-load current on the motors. That is so important a matter, especially in view of the huge trains liable to be thrown on the plant in the large schemes which we are hoping to see realized in the future, that I think we should give that method the consideration which it deserves, although it at first sight appears to be roundabout. In addition to that, we have the magnificent system invented by your President, Mr. Arnold, and I hope we shall hear more of that in the future. My only object in rising was to ask that we should hear as much aboul these systems as possible. PRESIDENT ABNOLD: I am pleased to be put down as one of the speakers on this subject, but Messrs. Steinmetz and Lamme have so thoroughly covered the subject, and Dr. Drysdale has so kindly referred to the other systems known to most of you, that it is not necessary for me to say much more, particularly as the time is growing short. 108 ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. I will correct one statement by Mr. Lamme, which rather puts me on the defensive. I understood him to state that his announcement of the single- phase motor made in September, 1902, was the first announcement of a single-phase system. I beg to state that in the month of June preceding, I read a paper on a single-phase railway, known as the Lansing, St. Johns & St. Louis Railway, which was built at that time and which I have since put in operation. I do not think it is just for the statement to be placed on record just in the manner in which it was made. I think Mr. Lamme meant to say that his paper was the first formal paper on the subject, but my road was built and almost ready to operate at the time that he made his announcement. Now, without further discussing the question, I am going to call upon a gentleman whose name is known to all of you, and introducing him, I am reminded of an anecdote about a little negro boy who sat on a log chopping away with a hatchet. A man coming along the road asked him how old he was and the boy answered : " If you goes by what mother says, I'se six, but if you goes by de fun I'se had, I'se 'most a hundred." If you judge the man who is to address you by his looks, he is a young man, but if you judge him by his experience he is " most a hundred," and is the father of the commercial electric railway. I have pleasure in presenting Lieut. Frank J. Sprague. MB. F. J. SPRAGUE: I feel quite embarrased by this pleasant introduc- tion by our worthy President and the reception which you kindly give me. The subject under discussion is one which I will not enter into at any length to-day, for I see by the hungry and thirsty look on the faces of some of the gentlemen present that one o'clock is near at hand, and that they would probably rather adjourn for luncheon than to listen to any discussions whatever. The subject on the card is how best to use the alternating current in railway motors. It is largely a technical question. The alternating-cur- rent motor is like a somewhat brilliant boy, who being exposed to various diseases has contracted a number of them ; he has had a moderate experience in mumps and measles, and a touch of typhoid fever, and the various doctors, many able ones here and elsewhere, have administered, sometimes in homeo- pathic but oftentimes in allopathic doses, large measures of quinine and other drugs. Whether, as the child grows and we are all hopeful of that child and he is subjected to the various climatic conditions of commercial introduction and use, those undercurrents of disease common to all fevers will recur, or whether the child will outlive them and become strong and robust is a matter which must be left to future developments. There is a larger problem, and I will not take over two minutes to speak of it. It is perhaps a more popular one, but of vital interest to us as engineers who are called upon to advise managers and others as to their financial expenditures, and that is : will electricity be used on trunk lines ? Our worthy President, with whom I have the honor to be associated on some important work in that line, is very hopeful, and so am I. But what are the reasons which may dictate the adoption of electricity on trunk lines ? Will it be because an economical service cannot be gotten by steam ? No. Will it be because there cannot be obtained to-day an efficient service ? ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. 100 Again, no. Will it be because of aesthetic reasons? Distinctly not. If electricity be adopted on any trunk line service it will be because of the hard and fast rule of financial necessity, not because we engineers urge it. It will be because the men who raise the money, run the road and have to provide dividends find that it is the best way to do it, and the reasons which will apply to one road are not necessarily those which will apply to another. It is my belief that some of the largest expenditures, and those most fruitful of return to those who own the steam railroads of the country to-day will be for the purchase and control of competing electric railways which, having in the past acquired franchises of undoubted value which cannot be duplicated, have built up a profitable business which they can hold and which will increase. Many a steam railroad will be better off financially and get bigger returns if it gathers in these franchises and systems, and operates its whole property with proper regard to the needs and capacities of each division than by electrification of its main lines, at least for a long time to come. I know there are one or two gentlemen back of me who feared that I would make some break on the subject, so I will close my remarks. I thank you for your kind attention. PRESIDENT ARNOLD: We now desire to hear from a gentleman whose early work is known in many fields, especially in the electric lighting field. His name was carried by one of the leading electrical manufacturing com- panies for many years and it stands to-day on much of the material that was manufactured in the early days. He is a man who has done much research work, and also considerable experimental work on the repulsion motor, a gentleman whom you all know and whom you recently honored by electing him President of the International Electrical Congress. I have much pleasure in introducing Professor Elihu Thomson. PRESIDENT ELIHU THOMSON: It is certainly a pleasure to me to listen to a discussion of this kind in a joint meeting of the Institution of Electrical Engineers of Great Britain, the American Institute of Electrical Engineers, and a Section of the International Electrical Congress. It is gratifying to find that there is so little dissent from the statements which have been made as to the future of alternating-current traction. Many of you will recall, no doubt, that at one time the electrical profession might have been said to have been divided into two camps, the alternating-current camp and the direct-current camp. The gentleman who preceded me was probably at that time more to be found in the direct-current camp than any other. The other gentlemen who have preceded me were to be found in the alternating-current camp. It is a fact, however, and those who have visited the power stations on the circular tour have noticed, that the direct- current men have called in the alternating current to help them out, and combine, therefore, the virtues of the alternating current with the virtues of the direct current. I was connected in the early days, and am still connected, with an organization which had not many prejudices of one kind or another. We had direct current, we had constant current series arc lights, constant- potential direct-current systems, and when the alternating current came we were ready to take that up without prejudice, and find out what there was in it. 110 ARNOLD: ELECTRIFICATION OF STEAM RAILROADS. In 1886 we put out our original alternating-current apparatus, and find- ing that the necessity might perhaps arise for motors on the system, it was at that time I undertook to get a motor for that system, a self-starting alternating-current motor, and the first motor of the repulsion type was made in 1886 and finished in the fall of that year. It was a little affair and was found not to operate very well on the higher fre- quencies, but by connecting it to a machine, which 1 was using for electric welding, giving 30 cycles, I found it operated very well and satisfied me as to the general features of the machine. That machine, unfortunately, was sent to an exposition and lost I could never trace it, and it never came back. The Paris Exposition of 1889 had a couple of examples of machines on a little different basis. One of them, I believe, is in England, at the Royal Institution, and another we have at Lynn. It was a machine which was started as a series alternating-current motor, and as soon as it reached a certain speed the commutator was short-circuited and it became an induc- tion motor. It combined, therefore, the elements of both, but 1 will admit that the design of such machines in those days was poor. We did not have even the distributed winding; we did not have the arrangements and the proportioning which we have to-day ; nevertheless those little motors would give a half-horse power for a moderate-sized motor on 125 cycles, which was the highest frequency used. 1 merely mention these items as matters of history touching on the discussion. They have nothing to do with the discussion as to the different methods and systems of using alternating current in electric railway motors, but I am a strong believer in the field being open for such work. I believe that not only will the direct-current motor maintain its place, but that certain lines of service which the direct- current motor cannot easily take will undoubtedly be taken by the alternat- ing-current motor for railway service, and the exhibition of a system, which you have been able to see in use, and which adapts itself to the use of both currents, is certainly a very instructive one. PBESIDENT ARNOLD: It occurs to me that I may not have put my explanation in regard to Mr. Lamme's statement in just the way it should be put. I think what he meant was that his announcement was the first of a purely single-phase commutator motor system. I think with this cor- rection he will accept my statement. He has not sent me any word, but this additional statement is due him. I think my work was first, but he got in with his announcement in September regarding the single-phase commutator motor. ALTERNATING vs. DIRECT-CURRENT TRACTION.* BY PROF. DR. F. NIETHAMMER. Even if one considers only serious proposals, there will be found available quite a considerable number of electric railway systems which might be used. None of the known systems possess, how- ever, advantages or features of such a kind as to render it able to replace all others. This fact becomes specially conspicuous, if an electric railway is desired which is adapted equally well to all the various services occurring in railway practice, viz., short and long lines, high and low speeds, short and long distances between sta- tions, heavy and light traffic. At the present time the following electric systems may be considered : 1. DIRECT-CURRENT RAILWAYS. a. Two-wire systems. Constant pressure from 500 to 1000 volts on the train, the return circuit being the rails. Motors with large inputs and moderate speeds may, however, safely be built for pres- sures up to 2000 volts, by using double commutators if necessary, or by grouping several motors in series. ml FIG. 1. THREE- WIRE DISTRIBUTION SYSTEM (KEIZIK, PBAG). b. Three-wire systems. Constant pressure up to 2X1200=2400 volts (Line built by Thury, Geneva; City & South London Eail- * As I did not think it advisable to change the paper as it originally ran, I add an appendix containing the most recent progress. The author treats the above subject in a much more complete way in a German book " Die elektrischen Bahnsysteme der Gegenwart." As to single-phase rail- way motors see also Electrical Magazine, October and November, 1904. run 112 NIETHAMMER: ELECTRIC TRACTION. way with 4X500 volts). 2X2000 volts seem to be a safe limit for this system. Each train may either be fed by one branch of the three-wire net-work, in which case the pressure per train is half of the whole voltage, and only one trolley is necessary; or each train may take the whole voltage by two trolleys the connection of two motor groups being grounded and the rails being the neutral conductor. (Fig. 1.) By installing boosters or direct-current transformers in the cen- tral station or along the line, the distance which may be covered by the simple direct-current system may be somewhat enlarged; the same result may be arrived at by using storage batteries at some feeding points of the line. c. Direct-current railways with transmission by three-phase cur- rents. The direct-current pressures on the trains remain the same as under (a) and (&) ; the three-phase voltages go as high as 60,000, transformation being by rotaries or motor generators. The transmission may also be effected by the direct-current system with constant current and variable voltage up to 30,000 volts (Thury). In all the cases mentioned, the simple series motor is almost ex- clusively used, with the exception of mountain railways with low constant speed. For the latter case the shunt motor is preferable, as this type is able to return energy to the line. 2. THREE-PHASE EAILWAYS. Induction motors on the train. Safe trolley wire voltages are 500 to 3000 (5000) volts, the pressure of 10,000 volts of the Berlin-Zossen high speed line being only experimental. The trans- mission may use pressures up to 60,000 volts which are reduced by stationary transformers preferably of the oil type. Of some inter- est for three-phase lines is the compensated polyphase induction motor with commutator, which has been proposed for this work. 3. SINGLE-PHASE COMMUTATOR MOTORS. The highest trolley wire voltage advisable at present is about 6000 volts, though the Maschinenfabrik Oerlikon proposes and is using 15,000 volts on an experimental line, which involves less risk for a system with one trolley wire than for a system with two or three. There are available for traction three types of single-phase motors: The series motor (Westinghouse Company, General Electric Company), the repulsion motor (Brown, Boveri & Cie) NIETHAMMER: ELECTRIC TRACTION. 113 and the compensated or series-repulsion type (Winter-Eichberg, Union E. G.). The following systems are of sufficient practical interest to be mentioned, but they cannot be considered serious competitors to the foregoing. 4. CURRENT CHANGERS. Converters or electric generators on the train. a. Ward-Leonard system and shunting locomotive of the Mas- chinenfabrik Oerlikon. The train takes single-phase current at high voltage from the trolley line and transforms it into direct current by a high-speed motor-generator set. By changing the small exciting current of the direct-current generator, speed con- trol of the train may be accomplished, and during retardation energy may be returned to the line. The main drawback of this excellent system is the excessive weight and price of the motor generator, which weighs more than all car motors together. b. Combination of steam engines, steam turbines or oil (petro- leum) engines with direct-connected direct-current generators. Regulating is done as in case a. Of this type is the old Heilmann locomotive and a new car of the North Eastern Eailway in Eng- land, which latter contains a horizontal oil engine directly con- nected to a compound generator of 55 kilowatts 300 to 550 volte and 420 to 480 revolutions per minute. 5. SINGLE-PHASE INDUCTION MOTORS. First proposed by C. E. L. Brown but having little chances of being applied practically. a. Stator and rotor both revolve; the stator is brought up to speed without load. By gradually retarding the stator by means of a brake, the rotor which is connected to the car wheels is put in motion. b. The induction motor is started empty and connected to the car wheel through a flexible friction clutch. c. Stator and rotor are connected with a device capable of stor- ing up energy, which device absorbs or delivers energy at will, i. e., an air compressor (B. J. Arnold) or oil pumps with variable stroke (Swinburne) or water pumps (Siemens & Halske). This system possesses the advantage of ability to operate for short dis- tances without connection to the electric supply circuits. ELEC. BYS. 8. 114 NIETHAMMER: ELECTRIC TRACTION. 6. CONSTANT DIRECT-CURRENT SYSTEM. Various trains are switched in series by a double trolley wire system. The total voltage is variable, and it is proposed to go as high as 30,000 volts. The constant-current system is, however, too complicated and unreliable for distributing purposes, though it is excellent for transmission lines. Speed variation and starting would be very economical and returning energy to the line very simple. The continuous losses in the line would, however, be considerable. 7. MOTOR CARS WITH STORAGE BATTERIES. Independent of every outside source of current and always ready for service. Such locomotives are heavy 1 and expensive and serve only for factory or shunting purposes or for short lines with light and constant traffic and with low acceleration. The mixed service using partly storage batteries and partly overhead line has entirely failed. There were also proposed railway plants combining storage batteries with single-phase induction motors. Starting is done b) the battery, free running and charging by the transformed single- phase current. The above-mentioned systems should be compared with regard to first cost, operating expenses, reliability and safety in service. This comparison is to be made for lines of few kilometers in length (street railways) and for hundreds and thousands of kilometers (main lines), for speeds of 10 km per hour up to 100 and 150 Ion; for accelerations of 0.1 m per sec. 2 to more than 1 m per sec. 2 , the retardation being even higher by 20 to 40 per cent. The train weights vary from 5 tons to about 2000, and the number of horse-power per train from 10 to 4000. There are lines with only 10 trains a day, on others the trains follow at intervals of 3 min- utes. In one case the stations are separated by a distance of only 200 meters, on others more than 100 km. In the first case the whole service is starting, coasting and braking; in the other case free running is of most importance. Either motor cars with mul- tiple unit control for motors distributed over the whole train or lo- comotives may be used. Electric traction is specially qualified for motor car service of passenger as well as of freight trains which service requires frequent short trains of variable length. Electric 1. A storage battery locomotive for a whole train weight of 100 tons for 16 km an hour weighs 26.8 tons; the storage battery absorbs 10 tons and the remaining electric equipment 4.3 tons. NIETHAUUER: ELECTRIC TRACTION. 115 motors are also able to haul heavier trains than steam engines and to exceed the latter in speed. (Baltimore-Ohio 1600 tons trains. New York Central trains of 2X2200 horse-power, max. 2X2800 horse-power.) Single motor cars take more watts per ton-km than long trains. A very hard problem is offered by motor cars which must be used for short and long distances between stations, for low and high speeds, and for short and long lines and for light and heavy service at the same time. For this service, however, elec- tricy is better adapted than steam. The main data for usual railway traffic may be taken from the following table, which gives the limiting values corresponding to light and heavy traffic: 116 NIETHAMMER: ELECTRIC TRACTION. I 1 | CO CM M 8 S o S2S ,0 jl .2t i! o 3 3 a d 3 CM 3 3 3 333 coccus S o c 5 2 J^ eo -a t- CM CO to a *3 i d a T-l O o 3 2 1 3 oa 3 3 3 3 3 w- - 1 J2 P ,_| ^4 en 10 ,H (N >, > 8 ^ O O o d 1 Is -4 O 3 o 3 3 I o 1 s s 3 g f i S 3 CO s, 1 35 s o ** 1H CO 3 CO 3 CM 3 1 I I^THtH iH I 53 1 fc " 3 8 1 1 1 1 11 1 3 3 o 3 3 3 CO 3 3 eo 3 3 &H H 8 S 8 1 V 8 S ? 1 1 8 9g| S Motors SI 1 3 o 3 3 3 8 3 3 3 8 3 1 CM 3 l-H 3 00 3 00 3 CM CO 3 i 2 ^"^ CM 3 rl a o "^ s 1 fl 55 03 M * 8 8 8 a g s 883 & 3 o 3 00 3 3 3 o 3 3 3 IO I" P 8 g 4 1 i s 1 Street railways, with radial lengths of 1 to 30 km trorr cen er and gauges of 600 to 1 .43< mm . . Industrial and mine locon olives, gauge 800 to 1,436 mm Elevated and underground, 10 to 40 km length, multiple unit Suburban and interurban ser- vice, 10 to 200km radial length, multiple unit oa 1 I I V- .i 1 1 II passenger trains, locomotives x 1 (Steep) Gothard line, 26* grade freight express | Mountain Ry -NIETHAMMER: ELECTRIC TRACTION. 117 II J f l Id 1 fc c I i EH Q i K 1 1 X V sg S 8 s So II ^S r irect curr Baltimore New York MS NIETHAMMER: ELECTRIC TRACTION. The energy input, the watts per ton-kin grow with increasing speed, i. e., an increase from 32 to 128 km per hour means an in- crease of input of 45 per cent per ton-km. High acceleration cer- tainly ^requires lowest total watt consumption for starting, but it causes excessively high starting currents, necessitating large and expensive motors and drawing excessive loads from the central sta- tion. For short distances between stations it is most economical to run very fast up to a high speed, if admissible, and to coast as long as possible without current. The maximum acceleration de- pends upon the allowable shock to passengers when starting; much more than 1 m (per sec.) 2 will not be permissible. The best method is to increase acceleration gradually when starting and to let it die out finally without any shock. Undoubtedly a reasonable electric railway service can be offered in economical competition with steam. On the Italian line from Milan to Porto 'Cerisio (130 km direct length, 70-ton trains, accel- eration 0.35 m (per sec.) 2 and speed of 90 km), the introduction of electric service has increased the number of travellers 2 times, the train-km 4 times and the trains per day have grown to 120 from about 20 during steam service. The receipts and profits ob- tained render this line the most economical in all Italy, though steam is used for generating electricity. Most favorable to elec- tric traction are most urban and suburban lines, railways with dense traffic or those so located that the traffic could not be increased without an additional line, railway tracks with long tunnels and heavy grades and lines which are in the neighborhood of coal mines and water powers. A comparison of the various electric systems should comprise the whole electric equipment, viz.: a. Motors and gearing; &. regulating and braking devices; c. current-collecting devices; d. central stations and sub-station equipments. MOTORS. The characteristic features of railway motors are, 1.. Mechanical reliability. 2. Maximum pressure possible on motor terminals and maxi- mum pressure at the trolley. 3. Sparking on commutator or collector. 4. Weight per horse-power at a definite speed. 5. Space occupied by motor. JflETHAMMER: ELECTRIC TRACTION. 119 With heavy train loads, high speeds and great accelerations it is often extremely difficult to make sufficient space available on the car truck for the large motors required. The room available increases with increasing diameter of the wheels and with broader gauges. It is a fact, that on a 300-mm gauge only 15-hp motors are possible, on a 700-mm gauge only 90 horse-power, on 1000-mm gauge about 150 horse-power and on normal gauge 250 horse-power. 6. Efficiency at full and partial loads. 7. Starting losses of the motor. 8. Power factor at full load, partial loads and at starting. 9. Heating for normal continuous running and for frequent starting. 10. Starting torque and possibility of produciag high accelera- tion; current consumption for a definite starting torque. 11. Efficiency of acceleration. 12. Speed variation; losses and efficiencies at variable speeds. Steadiness of regulation. Speed characteristic for variable loads. 13. Braking on resistances and return of power into the line, when coasting or on grades. The direct-current series motor has an air-gap of 2.5 to 7 mm for usual armature diameters of 300 to 600 mm, the upper gap being smaller by J to 1 mm than the lower. Experience on hundreds of thousands of such motors prove that this air-gap is absolutely safe and that there is no danger of sticking. The direct-current arma- ture winding, with open slots and carefully wound separate coils, as well as the commutator, may be insulated in .in absolutely reli- able way for voltages up to 2000. The field winding has no high potential between its terminals and is easily protected against the frame, whilst the field winding of shunt motors, being subject to full pressure, is much more liable to break down. The induction motor, if only that type without commutator is considered, must have an air-gap from 1 to 3 mm in depth for usual railway motors, according to size, in order to secure a satisfactory power factor and a sufficient overload capacity. The Valtellina motors with a rotor diameter of 800 mm have an air-gap of 2 mm. Values for other machines may be taken from Table II. According to the long practical experiences of Brown, Boveri & Cie., and of Ganz & Company, this small air-gap has never given rise to trouble, when the bearings are liberally designed. C. E. L. Brown has suc- cessfully used automatic ring lubrification for three-phase motors. Nearly closed slots should preferably be used to get smooth cylin- 120 NIETHAMMER: ELECTRIC TRACTION. drical surfaces along the air-gap, which makes it a necessity to wind the coils by hand. This type of winding with closed mica tubes in the slots and end connections well protected by bronze caps has never caused trouble on the Burgdorf-Thun or the Valtellina line. Special care must be given to the crossings of the end con- nections, but insulation may be obtained to withstand easily pres- sures up to 5000 volts. High voltage motors must, however, be very Liberally dimensioned to keep down heating which deteriorates in- sulation. It may be of advantage to put the stators of two three- phase motors in series to reduce the voltage per motor (Fig. 2). The air-gap of the single-phase commutator motor must also be rather small, though larger than with the three-phase motor, i. e., 3 mm for a rotor diameter of 450 mm. Commutator motors, the rotors of which are not fed directly from the line, are the best machines for high voltages up to 8000 volts, as all crossings of the end-connections can be easily avoided. For equal line voltage tin. IM j i UI-JT |-"S*f *** n Fw. 2. STATORS IN SEBIES. the single-phase motor in comparison with the direct-current motor is at a disadvantage in that there is an active e.m.f. across not only the armature but also the field coils. The trolley voltage of all alternating current equipments of the single or polyphase type may be lowered at will by transformers on the car, if on account of limited space or due to troubles on the commutator, one is bound to use low voltage motors. That means, however, additional weight and expense, though the transformers may be used for regulating purposes at the same time. The frame of single or polyphase motors can hardly be split, as is frequently done with direct-current machines. The joints might give rise to noise. As, however, even for direct-current motors in limited space the splitting of the frame is being abandoned in favor of the box frame, this fact is not of much importance. From Fig. 3 which represents a single-phase commutator motor of the Union Company, Berlin, for 50 horse-power, 800 rev. p. min., 40 mETHAMMER: ELECTRIC TRACTION. 121 periods, 6 poles, 400 volts, it may, however, be seen that the splitting of an alternating current motor is not an entire impossibility. Single-phase winding is more favor- able yet, as no coils have to be cut. The laminated field of alternating-current motors is less rigid than that of the direct- current machine, so that an additional solid frame becomes neces- sary. For direct-current motors which must undergo rapid varia- tions of the magnetic flux and of the speed or which must be quickly braked, a laminated frame would, however, also be of advantage. The greatest drawback of direct-current motors is the difficulty of commutation. Sparking in the neutral zone is due to the react- ance voltage of the short-circuited coils and to the voltage induced by the distorted main field. The distortion may be kept low by using a high number of field ampere turns and high saturations of FlO. 3. 50-HP SINGLE-PHASE MOTOR, UNION COMPANY. teeth and pole shoes. The reactance is small for low speed motors, for short armatures, for small currents per armature circuit, and for commutators with many segments. Flashover is produced by high voltages per segment and by current rushes, when at high speeds the current circuit is suddenly opened and closed again. These are the reasons why direct-current motors have not been built as yet for more than 1000 volts, though larger low-speed types may success- fully be designed for about 2000 volts. To raise the trolley voltage, several motors may be switched in series, but this scheme has the drawback that when some wheels with motors are slipping and others not, one or several motors may get the full voltage at their terminals and be burned out. The series motor is, of course, much less liable to sparking than the shunt motor, as the reactance does not vary much with load and speed, besides that armature and field ampere- J22 NIETHAMMER: ELECTRIC TRACTION. turns increase together. This commutation trouble is the most serious handicap to the direct-current motor, as it limits the exten- sion of its supply lines. The three-phase motor has no commutation problems. The space for the three slip rings with carbon brushes is, however, not smaller \L> if short circuit current T time for commutation. FIGS. 4 AND 5. COMMUTATION CHABACTERISTICS. than that for a commutator. Even the commutator for compensated polyphase motors is easily designed, as it is a mere frequency changer with low voltage. With regard to sparking, single-phase commu- tator motors offer the greatest difficulties. First of all, an alternat- ing -current has to be commutated, a process which changes every moment (Figs. 4 to 6). Sparking is due not only to the reactance voltage (e r ) and the voltage ( a ) induced by the main field during FIG. 6 COMMUTATION CHAKACTERISTICS. rotation, but to a transformer voltage e i which is induced by the oscillations of the main field independent of speed and which pro- duces a high short-circuit energy. By using low commutator volt- ages (smaller than 200 volts), a high number of commutator bare preferably with multiple parallel winding, by selecting thin brushes NIETBAMMER: ELECTRIC TRACTION. 12;] (minimum 6 to 10 mm), by inserting high resistances into the short- circuited coils, by reducing the main field and by building only motors for small outputs and small periodicities, the transformer voltage, el,, may be kept sufficiently low. The reactance voltage, ^ is cut down by the same expedients as used for direct-current motors. Equalizers and auxiliary commutation poles may be of ad- vantage, but there will rarely be space available for them. By a double (horseshoe) pole excited by the main current opposite to the short-circuited coil, one may neutralize the whole transformer effect. The General Electric Company uses a distributed field wind- ing to neutralize the reactance voltage similar to the Eyan winding of direct current machinery. The repulsion motor and the compensated motor (Fig. 7) have this advantage that for synchronism, and in its neighborhood, a regular rotating field is built up, replacing the pulsating alternating fields. Near synchronism the transformer effect in the coils under Fio. 7. COMPENSATED MOTOB. PIG. 8. STARTING COMMUTATOB MOTOR. the brush is, therefore, eliminated and the commutation is similar to that of direct-current machine. Flashing over on the brushes of a repulsion motor seems next to impossible and even for other commutator motors flashover appears less probable, as self-induc- tion damps away sudden current rushes and the laminated stator frame facilitates the rapid building up of magnetic fields. When starting, all commutator motors are equally bad and one of the best schemes besides those already mentioned is to use a series transformer for the armature circuit (Fig. 8) which cuts down the starting field, allowing at the same time any intensity of . the starting current. For repulsion motors, the same effect is possible by shifting the brushes toward the position of complete transformer action (brush axis in line with field axis). The main 124 NIETUAMMER: ELECTRIC TRACTION. field at starting may also be prevented from rising too much by choosing the iron inductions very high. The distortion of the field by armature reaction and the wattless voltage component produced by it may easily be neutralized for the single-phase series motor by a field winding, the axis of which coin- cides with the armature cross-field and which may be short-circuited or in series with the armature current circuit. Figs. 9 and 10 show this arrangement as used by Ganz & Company 15 years ago. Finzi splits the poles for the same purpose and cuts down the polar arc. Blathy of Ganz & Company also used some 15 years ago high tooth inductions 2 and ohmic and inductive resistances between armature winding and commutator, sometimes imbedded in the slots. FIG. 9. STATOB PUNCHING. A stator with definite projecting poles has the advantage of cutting down the reactance voltage in the short-circuited armature coils and gives rise to smaller armature cross-field, which means a better power factor than with a distributed winding imbedded in slots equally spread round the whole circumference. This is the reason why series motors should always have definite poles, while 2. Lamme proposes high pole-shoe induction. NIETHAMMER: ELECTRIC TRACTION. 125 the good operation of repulsion motors depends upon the full devel- opment of the armature cross-field to get a rotating field at syn- chronism. Repulsion motors must, therefore, have a distributed winding. The better leakage factor of the last-mentioned winding FIG. 10. ROTOR PUNCHING. is outbalanced by the better voltage factor or winding efficiency of a concentrated winding. Table II shows weight, outside dimensions, air-gap, efficiency, etc., of a great many railway motors of the direct-current, three- phase and single-phase type, most of which are in actual service : VIETHAMMER: ELECTRIC TRACTION. w I 1 1 (5 a meS unn JB93 S-9 o :!!L * GO GO GO GO 10*0 : 8S5 : 3 S o * co *3 < I 111 TOO oar-i 18 8 :8 8 :l !T t -9 .55 36 .< oq cq OTH oq e to * * * ^H iq ee lOTfl^ <3SO 16 I! oo Sd MO aequmx -i N 00 * 5 D t- X OS O r-i O* * >00t- 00 OS < NIETHAMMER: ELECTRIC TRACTION. 127 unn ja3 rain JCB&S H - a :S8i I ! eon o 10 : to 1 lllllllli CO 03 05 II :i OOOOC5OO O 8 8 c* c3 in 10 GO o = 1 for various speeds. Table III gives an interesting comparison of power factor and efficiency for three-phase motors and various methods of regulating them : TABLE III. Half speed. Full speed. Concatenated motors. Variable number of poles. Rotor resistance. Primary compen- sator. Variable frequency. Efficiency 81 88 - - 43 59 I Same 87 f motor. 86 74 85 75 80 _ _ __ 90 81 86 93 86 89 Power factor 85 60 60 85 75 85 93 77 84 93 85 93 With light loads the power factor of three-phase motors is usu- ally very poor, and the mean value is sometimes as low as 0.5. For starting however the cos is 0.8 to 0.95. The opposite is the case for single-phase motors, the power factor at starting is extremely low, about 0.3, increasing with speed and decreasing with load. Of all motors the direct current shows by far the smallest losses in the motor itself when starting with same torque, mainly because the iron losses are zero at standstill and the starting current is least for a given torque. From this fact it results that a direct-current motor heats least, when frequently started. The following Table IV gives a comparison of the motor losses at starting for various types of motors and starting arrangements. JflETHAMMER: ELECTRIC TRACTION. 133 1 ee 1 o cc ft^i fl3 C? 1- fl o 03 O ill! Hi UL 'IS III -2.2 O CO ll u II V 06 $& flj !! Si sgl ai ill f 3_d p I 1 ?.: si 134 NIETHAMMER: ELECTRIC TRACTION. Since in the direct-current motor most of the losses are produced far away from the motor surface, the capability for radiating heat is better for the alternate-current motor and best for the three- phase machine. For equal losses the difference in favor of the three-phase motor may amount to 25 per cent. The distributed winding is also better for cooling than the mummified concentrated field coils, for which latter copper strips on edge are best In heavy locomotives or motor cars for high acceleration, it may occur that there is not sufficient space for the necessary motor capacity at a predetermined rise of temperature. This limit is much sooner reached by three-phase and single-phase motors than by direct-current motors, and of all motors concatenation is worst in this respect. In extreme cases artificial cooling becomes neces- sary. The air of the running train may be directed by special pipes and chimneys on the surface of the motors and starters. If it is Ifto. 12. G. E. co. MOTOR. possible to nse openings covered by gauze or perforated sheets at both ends of the motor, one may drive an air draught through the motor by the ventilating ducts of the armature, thereby throwing the heated air to the outer surface. The new G. E. motors for the New York Subway are ventilated similar to Fig. 12 by means of air entering near the back bearing and passing through armature ducts of variable breadth over the field coils and escaping through holes in the yoke. The waste air of the universally used air- brake may also serve for cooling purposes; the pressure of the air must, however, be kept very low to avoid the creeping of oil. If there is sufficient space on the shaft, there may be added a fan to the motor. Eeichel proposed to install these fans inside the second- ary motors of a concatenated group, and to cool the main motors from these fans. There are other means for saving space: Siemens & Halske (German patent 131,299) propose to put the commutator outside NIETHAMMER: ELECTRIC TRACTION. the car frame to leave all the space inside available for the armature. In this case, however, the axle must be hollow and many connec- tions through the bearings are necessary. This scheme may, how- ever, be much better realized for the three slip rings of three-phase motors, as may be seen from Fig. 13, which shows the very interest- ing concatenated motors of Ganz and Company for the new Valtel- lina locomotives. (Each motor for 600 horse-power, 225 revs. p. 135 13. CONCATENATED MOTOB, VALTELLINA LOCOMOTIVE. min., concatenated 500 horse-power, 112 revs. p. min., 15 periods, 3000 volts.) To best utilize the given height, the projecting poles of direct current and single-phase motors must be arranged at 45 deg. on an octagonal frame with two sides horizontal; the most favorable design for getting a large armature diameter is to use a bipolar frame putting the poles horizontally and closing the yoke with its bearings outside the frame as is done by the General Elec- 136 NIETHAMMER: ELECTRIC TRACTION. trie Company for the New York Central motors. Kandd cuts off a segment of the cylindrical stator iron of three-phase motor at the lower side. (Fig. 13.) On locomotives there is sometimes suffi- cient space to put the motors above the car truck, the design being much simplified thereby. There are, in fact, cases where the motors may be of the open type, if they are well protected inside the car box (Jungfrau locomotives). As soon as it becomes possible to build reliable ball bearings, these may be used to reduce the space absorbed by the bearings. The last expedient in cases of limited space consists in insulating the motors by heat and fireproof mate- rials, such as mica and asbestos, and allowing a temperature rise of 100 deg. or more, though the commutators will hardly with- stand these temperatures continuously. The starting torque of motors for high accelerations may be three to ten times larger than the torque for free running, whilst for slow speed trains there is no great difference between these two torques. The best motor for accelerations higher than 0.5 m (per sec.) 2 is undoubtedly the direct current motor which starts very economi- cally against any torque taking less than two and one-half times the normal current for three times the normal torque. All three- phase railway motors in actual service have low accelerations, smaller than 0.3 m (per sec.) 2 ; the large locomotives and motor cars on the Valtellina lines have only 0.16, though tests were maclo up to 0.45 m (per sec.) 2 . Three-phase motors can yield specially high starting torques only by adopting complicated switching de- vices (mesh-star connection) or heavy regulating transformers, or by sacrificing the best running conditions (bad power factor for free running). This holds specially good for lines with very variable grades. Concatenated motors give only 50 to 70 per cent more maximum torque than the primary motor alone if one does not increase the motor voltage for concatenated working. It is, therefore, reasonable never to switch concatenated motors in mul- tiple, but to leave the secondary motors idle for full speed. More- over the acceleration up to 50 per cent of synchronism must bo double of the value after 50 per cent of synchronism which is also true for the mesh-star connection. The starting torque of the single-phase motor is for a given voltage the highest possible torque just as for the direct-current motor. On account of sparking difficulties and self-induction, the maximum torque is, however, smaller, but may be three to five NIETHAMMER: ELECTRIC TRACTION. 137 times the normal torque for well-designed types. The starting current is nearly entirely wattless, but is only about two to twj and one-half times normal current for three times the normal torque. When starting very slowly with a large torque by a strong field, the short-circuit effect under the brushes may burn out the motor. The torque of a single-phase motor, which is 30 per cent smaller than that of a corresponding direct-current motor is not constant as is true with direct and three-phase cur- rent machines, but varies between a maximum and zero with double the periodicity of the line current. The mean value of the torque is only half of the maximum, which fact is very important for the limit of slipping of the wheels. The wheels slip when the mean useful torque is only half of the maximum torque which is propor- tional to the adhesion of the wheels. This limit will, however, be reached only in very few practical cases. The starting torque of the direct-current series motor is inde- pendent of the terminal voltage, whilst the torque of the three-phase and single-phase motor is proportional to the square of the line voltage. This fact is specially dangerous for starting several trains at a time on a steep grade. The single-phase commutator motors have such a high starting torque that they may do their service, in emergency cases, with 40 per cent of the full line voltage. A disadvantage with the three-phase motor is due to the fact that its breakdown torque occurs at a slip of about 10 per cent coming to standstill when overloaded and absorbing a high wattless current and developing no torque and thus being liable to be burned out in that way. For frequent starting the watt consumption, or economy of the whole starting period, that means the efficiency of acceleration, is of utmost importance. Direct-current equipments are started by series parallel control, resistances are in circuit only for a short time as the motors accelerate a long time without resistances. In principle the single-phase motor can be started with the smallest losses, as they need no resistances. Starting transformers absorb, however, continuously a certain amount of energy and the ef- ficiency of the motor itself is low. The most economical way of stating consists in brush shifting (Brown, Boveri & Cie). The following Tables V and VI contain a comparison of the starting losses of various systems : 138 , NIETHAMMRR: ELECTLtLV TRACTION. TABLE V. Total starting losses for one entire trip of an elevated train of about 160 tons, distance about 1300 m, mean speed = 30 km p. h. Direct current Three-phase. Single-phase commutator motors (start- 1 Sgj parallel 2 motor- groups. JS A | .a 11 a o IN ing trans- former or brush shift- ing). a 1 OS * 33 i o > Mean kw hours on car .... 1.00 1.85 1.50 1.17 1.10 0.90 to 1.20 Mean kvA hours on car . . 1.00 1.55 1.85 1.67 1.55 1.10 to 1.50 For smaller distance the values become worse for three-phase equipments and better for single-phase motors. VIETHAMMER: ELECTRIC TRAGTIOM. 139 s lUI Isle II 111 g g co ,jj O 9-0 C-g o grtOgg Psr 111! lisfl fsgjS I Sal 1 140 NIETHAMMER: ELECTRIC TRACTION. Changing the direction of rotation is easily done for direct and single-phase current motors by crossing the connections of the armature; for high- voltage single-phase motors this ought to be arranged in the low-voltage secondary of a series transformer (Fig. 9) or a safe reversing oil switch becomes desirable. Three^ phase motors may be reversed by interchanging the primary wires, whilst the repulsion motor, whose armature is only in inductive connection with the line, must be reversed either by shifting the brushes through about a polepitch (Brown, Boveri & Company) or by shifting the line connections to the stator winding by about a polepitch or by using two primary windings. The direct current and the single-phase series motor vary in speed automatically about inversely proportional to the load with the effect that for variable torque the input and current consump- tion does not materially fluctuate, though this property is not used to its full extent in the direct-current motor, as may be seen from the following table: TABLE VII. Current 1.9 1.6 1 0.78 0.58 0.88 of normal Speed 0.75 0.90 1.0 1.15 1.40 1.90 for direct current. Speed 70 1 1 80 1 8 Torque 2 6 1.8 1 60 80 10 for direct current Torque 1.65 1 70 50 85 for single-phase Output 1.9 1.6 1.0 0.73 58 88 for direct current 1.2 1.0 0.92 0.90 0.88 for single-phase. The three-phase motor and direct-current shunt motor have prac- tically constant speeds for all loads and grades. On long lines with constant grades or on mountain railways this quality is no direct disadvantage, as the timetable is independent of the length of the trains and the motorman may quietly leave his regulating switches alone all along the trip. One may even state that the series motor ir. a certain sense is unable to make up for delays which usually occur with overloaded trains, in which latter case the series motors diminishes its speed. But practice proves that the motorman can easily avoid delays by making the best of the variable speed charac- teristic of the series motor according to the variable grades of hi* line. Of course on three-phase lines the main current may bo interrupted for intervals, either to increase speed when descending NIETHAMMER: ELECTRIC TRACTION. 141 or to reduce speed when ascending. Concatenated motors in themselves an additional possibility of varying the schedule time. The inherent constant speed quality, however, means high current consumption on grades and when starting compared with series motor characteristics. Moreover the direct-current motor has very economical means for speed variation through wide ranges and the single-phase motors possess this quality even to a higher degree. Speed variation of the three-phase motor is possible by one of the following methods: 1. Ohmic resistances inserted into the rotor circuit, the regulation of speed depends, however, from the torque used for a given resist- ance and is very uneconomical. Large resistances become neces- sary and small speeds at small torques are hardly possible. The Line Starting Rheostat Fra. 14. CONCATENATED MOTOR CONNECTIONS. higher the resistance, the more the three-phase motors acquire the variable speed quality of the series motor. 2. Concatenation of motors which has been admirably perfected by Ganz & Company. This company uses double motors in one frame (Fig. 13) uniting primary and secondary motors on one shaft need- ing only three slip-rings for both motors (Fig. 14). The secondary motor is never in circuit for full speed and may be specially di- mensioned for concatenation. The well-known reproaches made against concatenation are : Bad power factor and bad efficiency for half speed (see Table III), increase of weight, space and heating. The maximum torque of concatenated motors is rarely more than 50 per cent greater than that of one primary motor. The starting and switching devices are rather complicated. Ganz & Company have de- cidedly reduced these difficulties to a minimum by building the double motors and by using a frequency of 15 periods. Efficiency and power factor are both as high as 93 per cent (without gears) for full 142 NIETHAMMER: ELECTRIC TRACTION. speed; 85 per cent efficiency and 77 per cent power factor for half speed ; including gear loss the efficiency is still 80 per cent for half speed. Such an equipment is certainly not inferior to a single-phase car for full and half speeds. The Ganz motors have very high over- load capacities enabling them to exert high drawbar pulls in tandem connection. The complication of the car wiring and of the starting devices has been avoided by using only three sliprings for two motors and by adopting very simple and safe liquid resistances (Fig. 15). In recent tenders Ganz & Company propose only one secondary motor for three primary motors reducing the dead weight materially. Brown, Boveri & Cie have two heavy three-phase locomotives for the Valtellina line under construction. The two motors of each have FIG. 15. LIQUID RESISTANCE, GANZ oo. 450 horsepower and will be regulated by varying the number of poles from 16 to 8 ; drawbar pull 6000 kg for 37 km an hour and 3500 kg for 74- km, maximum pull for half speed 9000 kg. This scheme promises higher efficiency, higher torque for half speed, and less space than concatenation. These motors need, however, 5 or 6 slip- rings per motor, if resistances have to be in the rotor circuit above and below 50 per cent of synchronism. The resistances will be metallic in this case, not liquid. The type of winding for varying the number of poles must be a multiple parallel loop winding with 2X3 terminals (Fig. 16). The winding pitch is only 60 to 75 per cent of the polepitch at the high speed and 120 to 150 per cent of the polepitch at the low speed. Concatenated motors with a different number of poles or motors with more than two numbers of poles are surely too complicated for railway work. Variable frequencies NIETHAMMER: ELECTRIC TRACTION. 143 would certainly give a very economical speed variation, but the com- plication and the increase of price of the central station or sub- station and of the line are prohibitive. Brown, Boveri & Cie have installed a variable gear ratio on their Burgdorf Thun locomotives, which makes two economical speeds 18 and 36 km possible. Most direct-current equipments possess series-parallel control either with two or four motor groups giving a very efficient speed variation, as the efficiency at half voltage or a quarter voltage is only Zo Zs -12 pole, 600 revs. -E|g 8 pole, 750 revs. 6 pole, 1000 revs. 4 pole, 1500 ret* |2 & 6 pole 8 & 4 pole 1 7 Two separate windings FIG. 16. VAEIABLE POLE MOTOR CONNECTIONS. 1 to 5 per cent lower than for full voltage. Double commutators may fulfill the same purpose. An increase of speed may easily be effectuated by shunting the field, in which case the efficiency is even better than for normal speed. Commutation troubles may, however, prohibit the extensive use of this method. The single-phase commutator motor has in principle the most ideal and economical as well as the most uniform speed variation, by the use of regulating transformers in the primary or secondary circuit of the motor or by brush shifting or by varying the connec- tions between the line and a series of taps on the primary winding. The last two methods are specially suitable for repulsion motors. 144 NIETHAMMER: ELECTRIC TRACTION. All these methods work with good efficiency and good power factor for many speeds. The continuous losses in the regulating trans- formers, however, decrease the total efficiency of the equipment. In fact, the single-phase regulation is not more economical than that with a four motor direct-current equipment. The losses are spe- cially high for straight single-phase series motors using an auto- transformer, a potential regulator and balancing transformers. Series transformers, as used by Winter and Eichberg to supply only the small exciting current of the armature, are decidedly preferable to regulating transformers, and the best scheme seems to be shifting of the brushes or the shifting of the taps on the primary winding as used for repulsion motors. The three-phase motor could be very economically regulated by providing a polyphase commutator on the rotor and a three-phase transformer to change the size and phase of the rotor voltage, but this scheme is somewhat complicated and is not suitable for railway work. If a direct-current series motor whose field connections are re- versed, is separated from the line and short-circuited or switched on resistances, it will act as a brake, the effect depending upon tho speed and the resistance in circuit. The series motor is, however, unable to return energy to the line. By arranging a small exciter which just yields the small exciting voltage of the series winding and the full exciting current, returning of energy could be easily effectuated. The best motor for energy returning is the direct- current shunt motor which acts as a generator without making necessary any switching. The simple field regulation enables the shunt motor to work as generator and motor within a very wide range of speed; without any change the shunt motor works also on resistances or as a short-circuited brake. On mountain railways the braking on resistances is, however, rarely desirable, as the resist- ances on the locomotives become too cumbersome and heavy (i. e. f 2000 kg on an 11-ton engine). If other motor cars are on the line, the downgoing shunt motor feeds the ascending. If there is only one car on the. line, the energy returned will speed up the generators and will be only troublesome. One may provide re- sistance, in parallel with the generators, to absorb the superfluous energy, but by far the best method is to install storage batteries in the sub-stations which are charged by the descending cars. The three-phase motor has braking qualities similar to those of the direct-current shunt motor, but throughout a very restricted range. The three-phase motor returns energy only for speeds above NIETHAMMER: ELECTRIC TRACTION. 145 synchronism, that means, within a very narrow range and the energy cannot be stored up. Braking on resistances independently from the line is only possible by an additional exciter. The range of returning energy may be somewhat increased by applying con- catenated motors, but this advantage must be very expensively paid for, besides the fact exists that a short-circuited concatenated group only acts as generator between 50 and 75 per cent of synchronism and then again above synchronism. By inserting resistances in the rotor of the secondary motor this range may be slightly increased. On level lines as encountered on elevated roads not more than 10 per cent of the stored up energy can be returned by concatenated motors. On lines with many steep grades and dense traffic, the returned energy may be more and become of decided advantage. For mountain railways the three-phase motors have been fre- quently used (Jungfrau, Gornergrat & Engelberg), but it does seem not to have been a complete success, as new mountain lines (Vesuvius, Opcima Triest) are not equipped with three-phase motors but with direct-current shunt motors. The main reasons are that for the three-phase motor the downgoing speed must be higher than the ascending one which is prohibited by most railway regulations and that the energy of the descending car cannot be stored up, neither of which reasons is applicable to the shunt motor. There are very ingenious schemes for perfecting the three-phase motor for steep grades. The Maschinenfabrik Oerlikon switched the motors on their Jungfrau locomotive No. 3 in the upward sense for going downward in such a way that the primary field revolved against the rotor rotation. By inserting resistances into the rotor, in which the frequency is higher than in the line, any speed between stand- still and full speed may be obtained, but the resistances must dissi- pate twice the energy braked and the line has to provide just as much energy for descending as for ascending. The next step was to iise a special direct-current exciter directly connected to the motor shaft for braking, the motor works as a three-phase synchronous generator. The A. E. Gr. had arranged a storage battery for th'e same purpose on its high-speed car. If three-phase currents must be used, the simplest scheme would be to take the compensated three- phase motor with commutator on one side and sliprings on the other, which acts as generator at will (newest Jungfrau locomotive 5 of 5. The brushes on the commutator of these motors are automatically lifted, when the locomotive is connected to the line. The speed may be cut down to 5 per cent of full speed. ELEC. RYS. 10. 140 NIETHAMMER: ELECTRIC TRACTION. Brown, Boveri & Company), though it is inferior to the direct- current shunt motor. Those single-phase commutator motors, the armature and field of which are interconnected directly or through a transformer, may be separated from the line, and caused to work on resistances as single-phase generators of variable frequency. The return of energy to the line is possible onl} r by rather complicated switching devices, such as changing the variable speed feature into a constant speed one or, in other words, by creating a shunt motor or a sepa- rately excited motor. This may be done practically by feeding field and armature from a transformer having a series of taps which are changed according to speed and load. (Fig. 17, Union motor.) If the repulsion motor is driven backwards, it acts as a brake; by varying the brush angle any braking torque may be produced and even at low speeds energy may be returned to the line. Pio. 17. CONNECTIONS or MOTOR USED AS GENERATOR. For returning energy at speeds from the highest down almost to standstill, the most perfect system is a direct-current equipment with two double-commutator motors, with a combined series and shunt-field winding and regulating resistances in series with the shunt field and in parallel with the series field. For the highest speed, the four commutators are in multiple and the field weakest ; for the lowest speed, all commutators are in series, the field strong- est. This scheme is, however, too complicated for practical railway service. Motors which are regularly and frequently used for braking pur- poses must be much more liberally laid out and they are more liable to injuries than those used simply for haulage. The shunt motor which has several very valuable features for braking and speed variation has the great fault which rather excludes it from most railway services in that it is almost unsuitable for parallel running. This adverse criticism must be made concern- ing all motors with constant speed characteristics including the three-phase motor. If by chance the wheel diameters are not NIETHAAIMER; ELECTRIC TRACTION. 147 identical in general, if the slip in speed is not equal or if the mag- netic characteristics of two shunt motors or of two three-phase motors 6 are slightly different (not the same air-gap or not the same permeabilities), one motor takes more of the whole load than the other. It may even happen that one motor acts as generator, deriv- ing its energy from the other which must carry the whole load, causing a break-down and throwing the locomotive from the rails. This has actually occurred on mountain railways. For emergency cases rail tongs must, therefore, be provided which prevent the de- railing of the locomotive. For the shunt motor, there may be used the following remedies: Two shunt regulators may be used, ono for each motor, adjusted in a manner such as to equalize the load. The adjustment is, of course, different for an ascending and a de- scending car. and it must be modified before reversing. A scheme installed by the Austrian Union Company on their locomotives for Fio. 18. EQUALIZING SHUNT MOTORS. Opcima-Triest seems to possess many advantages (Fig. 18). The two armature terminals of same polarity are connected by a small regulating resistance, and the lever of this resistance is grounded, the rail being the return conductor. By adjusting this resistance which takes at most 2 per cent of the whole voltage, the load in any case may be equally distributed. The position of the lever is differ- ent for motor and for generator action. The automatic breaker in the trolley circuit cannot be used for avoiding overloads, as the cur- rent does not flow to the line and as the armature circuit is not allowed to contain a circuit breaker, as it would render emergency braking very dubious. Brown, Boveri & Cie arrange a friction clutch between each motor and the axle, which transmits only a certain maximum torque. The simple remedy of using only motors in series is not to be recommended as on steep lines it happens that 6. For three-phase motors that means different magnetizing and different short-circuit current. 148 NIETHAMMER: ELECTRIC TRACTION. one wheel slips and the other stands still, in which case the former motor is subjected to the whole voltage and may burn out. Cars with direct-current equipments may be run on lines with variable voltage, if the motors are connected only in series on one part of the line and only in multiple on the other, or by adopting double commutator motors. The trolley line voltage of three-phase and single-phase cars may be varied at will, if a stationary trans- former is provided on the car, which transformer, however, in- creases the weight of the equipment considerably. The Austrian Union Company is just completing a suburban single-phase line, starting from Innsbruck, which is fed at 400 volts inside the town and at 2700 volts outside. Single-phase cars may even be run over direct-current tracks, though, a good single-phase motor usually is a pretty bad direct-current machine; for the repulsion motor this is a specially bad case. Moreover the primary and sec- ondary motor voltage rarely agrees with the direct-current line voltage and a special set of starting resistances must be provided, or the single-phase equipment must use rheostatic control, which is very uneconomical. Best is series-parallel control in this case. Motor Gearing. In most cases the motors drive the car axle by 1 a single gear of cylindrical tooth wheels with ratios of 1 : 1 to 1 : 5 which withstand the wear of 8000 to 200,000 train-km. In few cases one finds 2 cogged wheels (Alioth) or double and treble threaded worm gears (Maschinenfabrik Oerlikon), which in some cases allow a better disposal of the available space. For very low speeds double gear becomes necessary, i. e. f Jungfrau and other mountain loco- motives. 3. The direct coupling of motor and axle may either be (a) rigid (Central London, Siemens & Halske high speed car, new locomotives for New York Central) or (b) elastic, by means of a hollow shaft and a flexible coupling (Heilmann Locomotive, A. E. G. high speed car, Valtellina locomo- tives of Ganz & Company). The rigid connection of the armature on the car axle has up to the present not been a complete success, but the method with the hollow shaft and coupling is decidedly compli- cated and entails the waste of much precious space. Siemens & Halske support the frame of their rigidly connected motors from the truck by means of springs, by which the bearings are pressed NIETHAMMER: ELECTRIC TRACTION. 149 against the axle from below, an oil cushion on the upper half of the bearing boxes damping vertical shocks of the frame. From Fig. 19 one may get an idea of the design of the Valtellina motors for 250 FIG. 19. VALTELLINA GEABLESS MOTOB. horse-power, 300 revolutions, 3000 volts, with a hollow shaft and a flexible coupling. The new 550-hp motors of the General Electric Company are rigidly fixed on the axle, but the frame may freely move in the vertical direction, as the motor has only two poles, one at each side, and the pole shoes have plain vertical surfaces. For motors mounted on the car axle, special care is necessary to exclude oil and dirt from the motor windings. 4. Driving by cranks and connecting rods, well known from steam locomotives was probably first proposed for electric loco- motives by Eickemeyer, and first used by Brown, Boveri & Cie. The location of the motors above the axle is decidedly facilitated by this mode of driving. Very disagreeable vertical and other move- ments and shocks, such as are incident to steam driving, can hardly be avoided when this construction is used. Ganz & Company have laid out a special arrangement for their new Valtellina locomotives PIG. 21. LOCOMOTIVE WITH CONNECTING BODS. (Fig. 20), the crank turning point being supported in such a way as to allow vertical movements. The General Electric Company possess a patent on the arrangement (Fig. 21), in which two double 150 NIETHAMHER: ELECTRIC TRACTION. WIETHAMMER: ELECTRIC TRACTION. 151 commutator motors are mounted at the end of the locomotive and four axles are joined by cranks and connecting rods. Starters. For starting are used (a) Metallic or liquid resistances combined with series parallel control ; (b) Transformers or autotransformers with taps or potential regulators, mainly for single-phase motors ; (c) Brush shifting for repulsion motors (Brown, Boveri & Cie). The last method is undoubtedly the cheapest. Regulating transformers for single-phase motors are heavier and more ' expensive than starting resistances, .even if high iron in- ductions (15,000 per cm 2 ) and high copper densities (3 to 5 amp. per mm 2 ) are adopted. They should be submerged in oil or ar- tificially cooled by compressed air. If series transformers (Fig. 8) are used for the armature alone, the size and weight are materially reduced. The heaviest are the potential regulators which must be provided with a short-circuited coil to neutralize the cross-field or self-induction of the armature. They allow, however, a very steady regulation, and avoid all contacts liable to spark; the higher the line current, the smaller is the range of voltage control. The main difficulty in the design of the usual regulating transformers with taps is due to the necessity for a reliable switch dial which works sparklessly. The best scheme seems to involve a solid snap switch which interrupts the current for a moment, when jumping from one tap to the next thus avoiding all auxiliary contacts and the short-circuiting of coils and eliminating the use of resistances or inductances. For various reasons three-phase starting resistances must be 'heavier and more voluminous than direct-current starters. To this is probably due the fact that liquid resistances have been thought more desirable for three-phase than for direct-current railways. Though at first sight a liquid resistance seems to be mechanically much poorer than solid parcels of nickeline strips or eastiron grids, the dimensions of which may be very much reduced "by forced air cooling (Jungfrau line) or by placing them into oil tanks, the designs of Ganz & Company and those of the A. E. Gr. 'high-speed car are of practical interest, and the first mentioned 152 NIETHAMMER: ELECTRIC TRACTION. design 7 (Fig. 15) has operated satisfactorily for years. The prin- ciple involved in their construction consists in using stationary electrodes within a tank into which the liquid is forced either by air pressure or by a rotating pump. The time consumed in starting may be varied by regulating the air pressure or by adjust- ing a throttle valve (Fig. 15). The electrodes consist of solid parcels of iron sheets which may be readily replaced. For frequent starting and shunting purposes, the liquid tanks must be very liberally dimensioned; it is desirable that the motorman control the resistance according to a main current ammeter, to avoid current rushes. The outer surface of the tank is provided with cooling ribs. The overload capacity of liquid starters is very high, as when the water is evaporating, an immense amount of heat may be absorbed. There is, however, the drawback that the water level may oscillate and that the evaporated water must regularly be replaced (2 liters on 500 km for the Valtellina line). On very cold days freezing is possible. To avoid a heavy current rush before short-circuiting the liquid starter, it is necessary that the electrodes have such a shape as to finally reduce the resistance to a value less than that of the armature. The starting switches may be, 1. Cylindrical controllers with contacts for reversing and series- parallel control, and for the control of resistances or transformer coils, sometimes provided with flat dials at the lower end for field regulation. 2. Multiple unit control with a series of single switches actuated by electromagnets or by compressed air pistons. 3. Liquid starters. For small inputs, the cylindrical con- trollers are in almost universal use. For three-phase equip- ments, they become heavier and more voluminous, on account of the increased number of contacts, which number may be somewhat diminished by using two-phase rotors. The multiple- unit system has been developed for direct current by the General Electric Company (electromagnetic switches), by the Westinghouse Company and the Siemens Schuckert Werke (electropneumatic control), for single-phase cars by the Union E. C-. Berlin, the system resembling very much the direct-current con- trol of the G. E. Company. The electromagnets, however, must be 7. Fig. 15 shows the original design of the Ganz rheostat which has been changed somewhat in its details; p is the throttle valve, k the short- circuiting switch of the rheostat. NIETHAMMER: ELECTRIC TRACTION. 153 laminated, and for the primary circuit, high-tension oil switches must be used. As alternate-current electromagnets are known to have various bad qualities, it seems to be a good plan to propose direct-current control from a small storage battery or electropneu- matic control for single-phase cars. For three-phase equipments multiple-unit control has never been used or proposed, 8 on account of its being rather complicated. Moreover the tendency of three- phase railways is toward locomotives and not toward the use of a series of motor cars in a train. On account of the many wires and contacts for three-phase current liquid starters with pneumatic con- trol have come to the front, as already stated. The liquid resistance does away with the great number of contacts and the sparking troubles of switches for heavy currents, allowing a very steady regulation and occupying only a moderate amount of space. Several single-phase or direct-current motors in multiple may be equipped with one common starting resistance if desired, whilst this arrangement is not possible with three-phase motors, unless the relative position of all parallel rotors is continuously identical, which condition seems impossible to be obtained. When this is not the case, the rotors may be partly short-circuited by the cross- connections. In the following Table VIII I have tried to make a comparison of the weight of various starting devices. 8. There are several patents on polyphase multiple control granted to the G. E. Co. four years ago. 154 NIETHAMMER: ELECTRIC TRACT 'ON. TABLE VIII. a. Direct current. Builder. Motors. Weight of starting devices in kg. Volts. H.P. Weight, kg. Krizik Prague.... 2x650 4x30 4x985 Entire electric equipment per car with- out motors: 1,560. Q E Co ... 500 1x27 700 1 controller and resistance: 150. Alioth 500 2x81 2x770 Entire electric equipment without motors: 960. Alioth 500 4x88 4x890 Entire electric equipment without motors: 3,600. Westinghouse.... 500 2x55 2x1,860 2 controllers and resistances: 660. Q. E Co 500 2x65 2x1,600 2 controllers and resistances: 600. Alioth 500 2x65 2x1,850 Entire electric equipment without motors: 4,100. G. E. Co 500 2x80 2x1,800 2 controllers and resistances: 700. 4x600 4x125 4x8,600 Entire electric equipment without motors: 6,600. Westinghouse. . . . 500 2x150 2x2,400 2 controllers and resistances: 750. G E Co .... 600 2x165 2x2,400 Weight of whole control apparatus: 1,000 (multiple unit). Westinghouse. . . . 600 2x150 2x2,400 Weight of whole control apparatus (including small battery): 800 (mul- tiple unit). Brown, Bo veri.... 600 b. Thr 2x25 ee-phase 2x830 currents. Entire electric equipment without motors: 1,500. Brown, Boveri .... 750 2x150 2x4,000 2 controllers and starting resistances: 2,000. Siemens & Halske 10,000 2x200 2x4,100 Entire electric equipment without motors: 8,800 (metallic resistances). Siemen & Halske. 10,000 4x250 4x4,000 Metallic resistances 5,000, controllers 4,800, transformers for 10,000 1 1,000 volts: 12,000. AEG 10,000 4x250 4x8,200 Liquid starters 4.800, transformers 6,400. 8,000 2x600 2x12,500 Entire electric equipment without motors: 7,000 (liquid starters). 500 c. Sine 1x27 jle-phast 1x800 ! current. Transformer 800. Union 2,700 2x50 2x1,240 Transformer 2,700 1 400 volt: 680 kg (oil type), regulating transformers: 2 x 315 kg. Union 6,000 2x100 2x2,860 Regulating transformers 1,100 kg, whole electric equipment without motors 1,800 kg. Oerlikon 14,000 4x145 4x8,000 Transformers 5,600 kg, apparatus and switches 800, trolley 1,200. Oerlikon 14,000 4x200 4x3,400 Transformers 8,400 kg, apparatus and switches 900, trolley 1,200. For the repulsion motor with brush snifting no special starting devices are necessary. NIETHAMMER: ELECTRIC TRACTION. 155 For operating whistles and brakes, electricity is not directly applicable, and, in most cases, compressed air must be used for this purpose. The air brakes and the main controllers should be so interconnected that applying the brakes instantly interrupts the main current. The air compressor should be driven electrically and should run noiseless, which latter condition seems to be most easily obtained when slide valves are used. On steep grades there should be provided electromagnetic rail brakes, or a braking rack should be placed along the rails. To avoid derailing on mountain railways, rail tongs are desirable. In order to eliminate the pos- sibility of racing on steep grades, there should be provided a device which prevents the motorman's leaving his car or train if he has not first put the controller-handle on the short-circuit brak- ing point. Although universally used, the scheme of lighting the trains directly from the trolley line is a bad one. Periodicities below 40 give a flickering light, unless very low voltages and thick fila- ments are used. The question of train lighting is, however, not so important as to render useless a system which, though defective ii: this one respect, is first class in all others, it being possible, in any event, to provide for lighting the train from some source in- dependent of the trolley circuit. Current Collectors and Line. The problem of collecting current from the line is one of the most difficult in electric traction. While direct-current and single- phase equipments using the rails as return require but a single con- ductor, three-phase and certain other three-wire cars necessitate at least two-line wires, which drawback to such equipments in some cases is so serious as to prohibit their use. The two conductors may be installed either above the center of the line beside each other in the same height or in different heights, or beside the line one above the other, the current collector sliding from the side, or one wire may be on each side of the line (Fig. 22). The lateral current collection avoids the oscillation of the current collector, arising from the deflection of the wire, but if the wires hang above each other, short- circuits may easily occur. As far as my experiences go, there seems to be no difficulty in collecting current for single-phase lines for voltages up to 6000. For three-phase lines the limit, as derived from the experiences on the Valtellina line with humid tunnels, 156 NIETHAMMER: ELECTRIC TRACTION. sharp curves and steep grades seems to be 3000 volts. For equal line voltage the voltage drop of the line which influences very much the starting torque of alternate-current motors is much greater for three-phase and single-phase currents than for direct current and on account of phase displacement, the equivalent cur- rent is also higher. The increase of voltage drop is the higher, the higher the frequency. The resistance of the iron rails foi alternating current amounts to between 3 and 15 times the value for direct current on account of the skin effect.- A high voltage drop in the rails causes electrolytic effects for direct current and czir FIG. 22. HIGH POTENTIAL OVERHEAD CONSTRUCTION OF SIEMENS & HALSKE. telephone or telegraph disturbances for direct and alternate cur- rents. For this reason special return wires may become necessary (fourth rail), which must be frequently connected to the main rails, and which, for alternate currents, should be as close to the other trolley wires as possible. Kapp proposed to place boosting dynamos or transformers between two consecutive rails at various spots. The Maschinenfabrik Oerlikon is using a separate return wire along the rails and puts the boosting transformers into this special wire, the primary of the transformer being in the overhead con- ductor. In this way the voltage drop of the return wire which NIETHAMMER: ELECTRIC TRACTION. 157 is regularly connected to the rails is reduced to naught and the drop is transferred to the overhead wire, the drop of which is correspondingly increased. If high trolley voltage or the boosting scheme of Oerlikon is used, railbonds are no longer a necessity; they are, therefore, omitted in a new single-phase line of the Austrian Union Company with 2700 volts. The current collectors used nowadays are: 1. The trolley wheel with overhead conductor consisting of a circular or an 8-shaped profile-wire, suitable for about 200 amp. voltages below 1000, and speeds not exceeding 80 km an hour. The trolley wire may hang just above the line or on the side of it (lateral trolley), as the trolley arm is hinged upon a vertical bolt; the height of the, wire above the line may also vary considerably. A disadvantage is the hammering of the wheel against the wire and the frequent derailing, which may be somewhat reduced by using Section c-d FIG. 23. TROLLEY SHOE. very light and elastic trolley arms, the movements of which may be damped by various springs or air and oil cushions. For cur- rents greater than 200 amp., two trolleys may be adopted. The wheel may be replaced by a sliding shoe on the end of the trolley arm (Fig. 23), the inside of which may be covered with aluminum (Jungfrau railway). The ability to collect current is increased in this way, but the deterioration of the wire is augmented. 2. The sliding bow consists of a tube of brass or aluminum con- taining a V-shaped groove and stands with axis perpendicular to the trolley wire. To get a larger surface of contact, Brown, Boveri & Company have given a triangular cross-section to the bow (Fig. 24), one plain surface of which is continuously on the wire. The inside of the tube may be filled with grease. The bow may carry 100 amp. for voltages up to 10,000 or 200 amp. for low voltages; sliding on two overhead wires 300 amp. may be safely 158 NIETHAMMER: ELECTRIC TRACTION. collected at 1000 volts or less. The line equipment, especially the overhead switches, are much simpler for the bow than for the trolley wheel; there is no derailing. The bow automatically ad- justs itself for forward and backward movements which feature is very important for shunting. The bow is probably the best cur- FIG. 24. SLIDING SHOE OP BBOWN, BOVERI & CO. ^ M rent collector devised for high speeds, in which Case it may be as light and elastic as possible; the pressure on the wire should not exceed 2 to 4 kg, and several springs acting after each other must neutralize shocks and prevent the interruption of the contact by the bow. The wind pressure must be compensated by wings. The satisfactory result of the current collection on the experimental 9 Siemens high speed car at 10,000 volts and 100 amp. per wire is due to the lateral sliding of the bow (Fig. 25 10 ), avoiding thereby the movements due to the deflection of the wire, to the great elas- ticity produced by three consecutive springs and to the small weight of the cross-bar of the bow (650 grammes) and to the very small pressure of only 2J to 3 kg; on three-phase lines one may apply either two separate bows beside or behind each other or one bow the cross-bar of which consists of two insulated pieces (Brown, Boven & Company) (Fig. 26). 3. Most of the good qualities of the bow are also to be found in the new original current collector of the MachinenfdbriJc Oerlikon (Fig. 27) consisting of a curved rod of brass tubing sliding on the lateral overhead wire from above, when running on the free line. At stations and in tunnels, or wherever it is desired, the rod makes contact from below exactly as the bow. The turning of the rod through an angle of nearly 270 deg. is effectuated either by hand or pneumatically. There are 2X2 rods on each loco- motive, which may collect the current from either of the two wires on each side of the line. In this way one wire may be repaired, 9. The treble bow is of course much too cumbersome to suit for regu- lar service. 10. Consisting of a brass tube with aluminum filling. NIETHAMMER: ELECTRIC TRACTION. 159 while the other is working. The repairs of the wire and of the current collector are quickly made and do not necessitate the use of a turret car which blocks up the line. Without any serious sparking, the rod collects full current while the locomotive travels at full speed from a section with 15,000 volts to a line section which is interrupted by the semaphore FIG. 27. OEBLIKON CUBEENT COLLECTOB. 4. The cylindrical roller collector somewhat resembles the one of the bow type, the cross-bar consisting, however, of a rotating roller usually running on ball bearings. On the Valtellina loco- motives this roller for 3000 volts and 200 amp. per wire makes 4000 revolutions a minute and consists of two copper tubes or two steel tubes electrolytically covered with copper, insulated from each other by impregnated . wood. The tubes have to be replaced after a service of about 15,000 train-km. There is one roller collector for forward running and one for backward movement (Fig. 20), 13 24 85 - M ^s&j MV... 45 -< ~ 2 - - 3 4 FIG. 28. GANZ & co. TEOLLEY SUSPENSION. each being controlled by compressed air. The roller is usually heavier and less elastic than the bow. To avoid the hammering effect of the deflection of the wire, Ganz & Company propose to use two trolley wires (Fig. 28) which cross each other, the support of one wire being at the spot where the other has the deepest deflec- tion. 100 NIETHAMMER: ELECTRIC TRACTION. Overhead wires should not be fastened rigidly but In an elastic manner, to avoid break-downs of the wires by the hammering effect of the collector, which effect increases with the speed. The Union Company fastens the trolley wire for their single-phase lines FIG. 29. UNION co. TROLLEY SUSPENSION. to a special suspension wire, at distances of 3 to 4 metres, by vertical wires of variable length, obtaining an almost straight trolley wire with unnoticeable deflection (Figs. 29, 30 and 31). For voltages above 1000 double insulation of the trolley wire is to recommended. 5. From the third rail the sliding shoe which is usually pressed on the rail from above by its own weight, or by springs or pneu- matically from the side or from below, may collect currents of FIG. 30. TEOLLEY SUSPENSION. more than 2000 amp. For heavy currents this is the cheapest and most durable scheme yet proposed, though for alternate currents, the increase of the resistance by skin effect is very objectionable. The main difficulty which makes the third rail prohibitive for NIETHAMMER: ELECTRIC TRACTION. 161 three-phase lines is the necessity for the thorough protection of the live rail, mainly at stations. It is, however, possible to cover the third rail at stations by wcoden boards leaving only a narrow crevice for connection with the shoe (Baltimore & Ohio Ky) . If the shoe projects laterally from the car, the third rail may easily be protected by overhanging boards in such a way as to eliminate danger to operators and officials when crossing the rails. Too much protecting, however, prevents rapid inspection. On the Fribourg-Murten line (Switzerland) the third rail was allowed only for the free line, at the stations two bows and two overhead wires were prescribed, complicating the system materially. During the erection of the third rail, special care must be taken to allow for heat extension and to prevent the movement of the rail. Overhead FIG. 31. TROLLEY SUSPENSION. conductors for heavy currents above 500 amps, would necessitate very expensive framework and a conductor having the shape either of a usual rail or of a U iron or two Z irons. The elevated railway of Elberfeld is using such an overhead rail and the Baltimore & Ohio Ey. formerly used an overhead tube which, however, has been discarded. If the same car has to run on tracks with different voltages, two different kinds of current collectors must be provided. On the line already mentioned with 400 and 2700 volts single-phase, the Austrian Union Company has installed a high bow for high tension and a low bow for low tension. The trolley wire at the end of the low voltage track is gradually raised and the low bow automatically leaves the trolley wire. "ELEC. RYS. 11.. 1G2 NIETHAMMER: ELECTRIC TRACTION. Disagreeable disturbances are caused on trolley wires and third rails by ice and sleet. A mechanical remedy consists in using scrapers and metal brushes which, however, deteriorate the con- ductor; it is also not always sufficiently effective. On heavy third rails may be applied certain chemicals, as calcium chloride, as they readily melt all ice and sleet, the soft mass being easily swept away by brushes on the motor car. Electric heating, though somewhat expensive, has also proved a success, as on the Burgdorf Thun rail- way. The line is short-circuited with low voltage only as long as is necessary to soften the ice ; the sliding bow sweeps it away after- ward. A thin coat of varnish on the trolley wire may prevent the formation of ice without disturbing the collection of current. It is noteworthy that ice not only depends from the lower side of the wire, but it forms on the upper side also. In overhead switches of three-phase lines either both wires or at least one must be entirely omitted and replaced by insulated pieces, to avoid crossings of conductors of different phases (Fig. FIG. 32. OVERHEAD SWITCH. 32, of Brown, Boveri & die). One should, therefore, always pro- vide at least two current collectors on the motor cars, spacing them at a distance somewhat greater than the length of the over- head switch. It is bad practice to be compelled to pass all switches without current, as it necessitates special attention on the part of the motorman and means low acceleration and very inconvenient shunting. It is advantageous to have at least one phase running all through the switches (Fig. 32), as in this case the motors con- tinue to work as single-phase machines; starting is, of course, ex- cluded. On third-rail tracks there occur similar interruptions along street crossings ; at least two sets of shoes are, therefore, neces- sary at each end of the car. The intersecting roads should cross at a small angle, and the third rail should continue on different sides of the line beyond the crossing. The following characteristic features seem desirable in a cur- rent collector for universal use: Running and shunting in both XlETUAUUElt: ELECTRIC TRACTION. 103 directions must be possible without reversing the position of the collector. There must be neither hammering of wire nor breaking of contact at high speed. The current collector must not be able to destroy the line construction and must be incapable of being de- railed during service. Eepairs of the collector and of the line must be quickly made without interrupting the regular service. This ne- cessitates either the simple direct-current system for volt ages less thr- 1000, or for high voltage single-phase railways two separate trolley lino* have to be erected on both sides of the line (Maschinenfabrik Oerlikon). Simplicity of the line and of the switches which must be crossed with full current on dictates the use of only one current- collecting conductor. High voltages from 3000 upwards seem ab- solutely necessary for long lines. Central Stations and Sub-stations. The generators and transformers for single and polyphase railways must be designed for the apparent input of the railway motors, that is, for the kilovoltamperes which are considerably higher than the kilowatts. The low power factor of the current entails a much higher voltage drop than current at unity power factor, which feature is specially bad for alternate-current motors so sensitive to voltage variations. The mean power factor on three- phase lines is sometimes as low as 50 per cent, and during starting it is even lower for single-phase motors. Single-phase generators and transformers are larger and more expensive than those of the polyphase type. It is, of course, possible, or even necessary, to use two-phase generators for single-phase lines, but both phases will always be far from equally loaded. Generators and sub-stations must be able for moments to deliver the maximum output, which in some cases may be more than 10 times the mean value on long lines with light traffic, but which in other cases may fall down to only 50 per cent above the mean value. If in the sub-stations of direct- current railways storage batteries are provided, the converters and generators may be very much reduced in size and price, having to yie7 400 kw or larger. It has been found advantageous to start both generators and rotary converters from the direct-current side with a starting bar operated by hand, thus reducing the amount of cur- rent required to overcome the friction of rest about one-half. This is often important when the sub-station is heavily loaded and the current is drawn from it to start additional machines. It also re- duces the size of the starting resistance ; for example, 400-kw motor- generator sets or rotary converters may require from 400 to 600 amperes to start them from rest, providing the machines have been standing for some time, and thus the oil allowed to squeeze out be- tween the bearing and the shaft. With the assistance of the start- ing bar this current can be cut down to 200, and not exceeding 250 amperes at 230 volts. The operation of these motor-generator sets was all that could be expected of them and was satisfactory in all respects; but, on account of their inefficiency and high first cost, improvements were demanded. A very rational step was the abolishment of step-down trans- formers and the substitution of a high-tension for the low-tension winding on the motors ; the introduction of one 250 to 300-volt gen- erator instead of two 125 to 150-volt generators, and, in some cases, the substitution of a synchronous motor for an induction motor. The use of the high-tension winding on the motor removed the necessity for and cost of the step-down transformers, and more than compensated for the additional floor space required by the motor- generator set. Regulators on the alternating-current side were un- necessary the voltage on direct-current generators being readily controlled by a rheostat on the shunt field and the operation was simplified, allowing the equipment to be handled by the regular class of dynamo operators, this being of great importance in large systems, as it requires less time to train the operators. Disturbances on the alternating-current side of the system have little effect on the motor-generator sets and are of such a character that they are easily provided for by protecting devices. The objection to motor-generators, especially those of induction type, is the low-power factor, which increases the losses in the feeders. The losses in the feeders are usually a small part of the total loss, so that this in many cases is not important. The first cost of motor-generator sets is usually somewhat higher than of rotary converters, particularly in the large sizes. The ELEC. BYS. 17. 258 EGLIN: CONVERTERS AND MOTOR-GENERATOR SETS. actual difference, however, is not so great as is generally supposed. The relative costs of the various sizes are shown in the following table: COMPARATIVE TABLE or COSTS. notaries with trans- Synch. Ind. Synch. Ind. formers and regs. Mt. Gen. Mt. Gen. Mt. Gen. Mt. Gen. 25 Cyc. 60 Cyc. 25 Cyc. 25 Cyc. 60 Cyc. 60 Cyc. Capacity. 1.00 1.05 1.05 1.10 1.03 1.08 1,000 kw 1.00 1.00 1.05 1.05 1.00 1.00 500 kw. 1.00 .95 1.00 .95 .95 .95 250 to 300 kw. This is based on quotations by the same manufacturer of rotary converters and motor-generator sets of 25 and 60 cycles, with a 25- cycle rotary converter with transformers and regulators as a unit. The following table shows the efficiency of a 400-kw, 60-cycle ro- tary converter and a 400-kw 60-cycle motor-generator set with both high and low-voltage motors. These tests were made at the works of the manufacturers, and show that even at high frequencies the rotary converter is more efficient than the motor-generator set : Rotary Converter. Two-phase, 16 poles, 400 kw, 450 r.p.m., 230 to 300 volts rotary. Two-phase, 60 cycles, 5000 volts primary, 210 to 160 volts sec- ondary transformer. Per Cent. Combined Efficiency. Load. at 210 Volts. at 160 Volts. 100 89.6 90.5 75 88.0 88.9 50 84.3 85.5 25 73.0 75.3 Motor-Generator Sets. Direct current, 10 poles, 400 kw, 450 r.p.m., 125 to 150 volts. Alternating current, 16 poles, 560 kw, 450 r.p.m., 380 volts. Per Cent. Combined Efficiency Load. at 150 Volts. 100 82.0 Direct ojurent, 10 poles, 400 kw, 450 r.p.m., 230 to 300 volts. Alternating current, 16 poles, 560 kw, 450 r.p.m., 220 volts. Per Cent. Combined Efficiency Load. at 300 Volts. 100 84.9 EGLL\ .- C-OA r/;y,'7'/-;A'.s AAL> MOTOR-GENERATOR SETS. 259 Direct current, 10 poles, 400 kw, 450 r.p.m., 230 to 300 volts. Alternating current, 16 poles, 560 kw, 450 r.p.m., 5500 volts. Per Cent. Combined Efficiency Load. at 300 Volts. 100 86.9 After these machines were installed in the suh-stations, a series of tests were made, using the same observers and the same instru- ments (the instruments being checked between tests), so as to obtain the all-day efficiencies when operating under commercial conditions; the rotary converters being placed in one sub-station and the motor generators in another, but both supplied from the same generating station. It would appear from these tests that there is no practical difference in the commercial efficiency between the high-voltage motor- generator set and the rotary converter, and, as was to be expected, the low-voltage motor-generator set was the most inefficient. ALL-DAY EFFICIENCIES UNDER COMMERCIAL CONDITIONS. Power Factor. No. 1 No. No. 3 Type of Mach ne. Load. H. P. Eff. A Ph. CPh Av. f Ind. Motor Empty 0.0 .0 45.9 23.0 1 Two-phase 1/4 140 72.9 67 .1 86.4 76.7 j H. P., 560 1/2 280 82.1 83 .6 92.8 88.2 | Volts, 220 3/4 420 85.3 86 .4 90.4 88.4 [ Amp., 1150 Full Ld. 560 85.4 88 .7 92.1 90.4 ' Ind. Motor Empty 0.0 3.9 29.1 16.5 Two-phase 1/4 140 72.7 56.9 69.7 63.3 H. P., 560 1/2 280 81.2 77.4 82.5 80.0 Volts, 6000 3/4 420 84.7 82.5 86.2 84.3 Amp., 47 Full Ld. 560 85.9 85.6 88.9 87.3 Rotary 1/4 Two-phase 1/2 KW, 400 3/4 Volts, 250 Full Ld. 134 70.9 268 77.2 402 80.4 536 84.1 99.2 100.5 97.4 97.6 106.0 102.6 105.1 102.8 102.4 99.9 98.0 97.8 CONCLUSIONS. The type of machine to be installed in sub-stations depends principally upon the frequency of the system and the importance -00 EGLIN: CONVERTERS AND MOTOR-GENERATOR SETS. of reliable and continuous service. The frequency to be used de- pends upon other conditions which are outside the scope of this paper. In cases where the largest percentage of the output of the gener- ating station is to be transformed to low-tension direct current, 25-cycle rotary converters should be installed on account of their higher efficiency and lower first cost. The very large number of these machines which are now in successful operation proves con- clusively their reliability and effectiveness. In mixed systems, and where the percentage of current transformed for low-tension dis- tribution is small, motor-generators are desirable. With higher fre- quency, particularly 60 cycles, it has been shown that motor gener- ators compare favorably in efficiency and are much more reliable and simple in their operation. DISCUSSION. CHAIRMAN FERGUSON: Mr. Eglin's paper is now ready for discussion. You know that in Europe motor generator sets are used very much more extensively than in this country, and we shall be glad to hear from any of our European friends as to their experience. Col. Crompton, we will be glad to hear from you. Col. R. E. B. CROMPTON: I am unable fully to discuss this important subject as I have not studied the paper sufficiently carefully but I can com- municate one figure which appears important that is, that in the large London system with which I have most experience, where we generate and transmit at 5000 volts transformed by motor generator sets to 400 volts, and charge batteries through these sets; the total losses, including those in the high-pressure mains, motor transformers, accumulators, low pressure mains to consumers, amount to 27 per cent as a maximum, but about 25 per cent on the average. If we used rotary transformers, these losses would be greatly reduced. Mr. PHILIP TOECHIO: In comparison with the efficiency obtained from motor-generator sets, I would say that with 25-cycle rotary converters of 500 to 1000-kw capacity, in American cities the all-day efficiency is above 90 per cent. Mr. M. J. E. TILNEY: The writer mentions that he only starts up from the direct-current side, owing to the heavy starting current. There is a large system in London where they start up from the high-tension side, with resistance in the rotors, and they find the maximum current never exceeds the full-load current of the machine, and in many cases is only 60 per cent. Is there any special advantage in starting from the direct- current side? Mr. PHILIP TORCHIO: I want to add another point, and emphasize a matter that Mr. Eglin touched upon in the paper, but in my opinion, did not dwell upon strongly enough ; that is, the advantage oi the greater EGLIN: CONVERTERS AND MOTOR-GENERATOR KETti. 2(51 capacity you get from a rotary converter than from a motor generator set for overload conditions. This is an important factor in laying out the reserve capacity for a sub-station. Mi-. EGLIN: The figure given by Col. Crompton is similar to the figure in this country on motor-generator sets. As to the question of cutting down the starting current, it is not the practice in this country to start from the alternating side. As the motor-generator sets are started from the direct-current side, the starting current would be much smaller than 60 per cent of the full-load current of the motor. It would not exceed 25 per cent. The motor-generator sets are started in the same way as the rotary con- verters are started, using the generator as a motor. THE BOOSTER MACHINE IN TRACTION SER- VICE, AND ITS PROPER REGULATION. BY PROF. DR. GUSTAV RASCH. In power stations for electric railways, especially for those feed- ing a network having a small number of cars running at the same time, there is a demand for a device to steady the power-station service. The unsteady load and current consumption in such power-stations grows worse as fewer cars are running at the same time on the line. It is known that ammeters and voltmeters of small power-stations indicate fluctuations continually, while large power-stations show mostly a steady load with small variations only. The disadvantages of the unsteady service with regard to the efficiency and the life of the steam-engines and the generators of 'the power-stations are obvious. It is desirable also that the cars, especially those cars running on the outer ends of the line, be fed with a constant voltage, which, however, cannot be expected from a power-station having too heavily fluctuating a load. To steady the machine service, a buffer storage battery is often used; that is, a floating battery connected parallel with the power-station genera- tors. It is indisputable that such batteries possess valuable features. For instance, they are of great importance in case of breakdown of the machine service and they give a chance to run some cars just before starting up and after shutting down the regular power- station service. They have a disadvantage, however, in that they do not react upon the fluctuations of the current, but only upon the voltage. Though an absolutely steady voltage on the bus cannot be assumed, it is evident that the battery does less work as a buffer battery the steadier the voltage. It is very likely, however, that even with nearly steady voltage the machine may be subject to heavy fluctuations. The following discussion may explain these phenomena still more cloarly. The generator at the railway power station is shunt- wormd, and, to simplify the discussion, it may be assumed that [262] RASCH: BOOSTERS IN TRACTION 8ERVIC3. 263 the e.m.f.= E, may drop proportionally with rising current /! that is may follow the law E~E Q I^c where E Q is the elec- tromotive force with no load, and c a constant. The constant c, may cover the influence of the armature reaction and the drop of speed of the machine with increasing current 7 1 . It has evidently the character of a resistance. The armature resistance of the machine is r (Fig. 1). The parallel connected buffer battery may F.. t 1 FIG. 1. have an electromotive force A and an internal resistance a. The heavy fluctuating current in the network (the actual line current) is I, while I t and 7 2 are the currents of the machine and the storage battery. It is easy to derive the formula, ^, - A + al a + c + r . _(c ' I- - A) (2) a + c -T- , At the average value 7 m of the actual line current, the current of the storage battery must reach zero value, so that the battery may not be either overloaded or underloaded during daily service. That is, following equation (2) : o = (c + r) l m (E A),OT, That changes the equations (1) and (2) to and, (4) It is necessary to make ample estimate of / m . The equation (3) shows that an absolute steadying of the machine 2(U RASCR: BOOSTERS IN TRACTION SERVIGE. current 7 X is impossible, because it is not independent of the heavy liuctuating actual current /. The larger the value of c + r, and smaller the value of a, the larger will be the dampening effect. The first means high internal resistance and large drop of voltage, features of the generators which cannot be called desirable ones. A small storage battery resistance a, means plates of large surface, that is expensive cells. The buffer machine (booster) is another means of steadying the service. The value of such a machine, especially for hoisting in- stallations, was thoroughly treated in a paper by Mr. Meyersberg,* read at the meeting of the Institution of Gorman Electrical Engi- neers. It is indisputable, however, that these machines are also of great value for all railway central stations, and for all similar services with load fluctuations of shprt duration. A large centrifugal mass is driven by an electric motor (under certain conditions two motors may suitably be used). The arma- ture of the motor is in multiple with the network. The centri- fugal mass naturally accumulates energy with decreasing network current and gives out energy in the network with increasing current consumption. Therefore the buffer machine works at one time as a motor, and at another time as a generator. For the moment we will not consider the character of the field excitation of this machine. It will be the object of this paper, however, to calculate the most favorable device for one special case. It may be assumed that the actual current I, of a 600-volt power station for a small railway is the subject of regular fluctuations from 500 to 200 amperes inside of periods of 12 seconds. The condition regarding the regularity of the sequence of the fluctua- tions is of minor importance; it is important, however, that the utmost data be obtained regarding the fluctuation itself; that is, 500 and 200 amperes be not increased or decreased, because other- wise the buffer machine would be forced to run at a speed which would not agree with the speed which was assumed when designing the machine. The curve ABODE (Fig. 2), shows the actual current fluctuation inside of one period. The average current of the gen- erator ought not to be 200 4- 500 IL = 350 amperes, 1. Meyersberg, Elektrotechnische Zeitschrift, 1903, page 261. RASCH: BOOSTERS IN TRACTION SERVICE. 265 but somewhat higher, about 360 amperes, on account of the efficiency of the buffer machine, which is naturally below 100 per cent. With absolutely equalized service the generator would work continuously with this amount of current. One may be satisfied, however, in practice to limit the fluctuations to 10 per cent above and below 'this amount. We therefore assume that the generator current I g must be dampened by means of the buffer machine to the limits of fluctuations between 400 and 320 amperes (see curve A t 7? x C^ DI EI in Fig. 2). Now then, 7 g -f 7 p = 7, and it follows that the buffer current 7 p =7 I g . The curve A z B 2 C 2 Z> 2 E 2 Fig. 2, shows the work of the buffer current 7 . The positive values mean taking energy (charging), the negative values, giving out energy (discharging) by the buffer machine. For the moment, the e.m.f. may be assumed as constant, at 600 volts, the output of the Ampere D Seconds FIG. 2. OUTPUT OF GENEBATOB AND BUFFEB. buffer machine being a maximum during charging: 600 (320 200)= 72,000 watts. This value may be used as specific for the designing of the buffer machine, though naturally there is the intermittent service to be taken care of in addition. The curve ABODE (Fig. 3), shows the period of loading and unloading the buffer machine. This figure shows a time phase retardation against Fig. 2, inasmuch as the period starts with the beginning of the load. It shows that the loading has a dura- tion of 6.54 seconds, and the unloading a period of duration of 5.46 seconds. The energy consumption is, 72,000 2 The energy output is: 60,000 X 6.54 = 236,000 watt-seconda. X 5.46 = 164,000 watt-seconds. 260 RASGH: BOOSTERS IN TRACTION SERVICE. The total efficiency is therefore assumed as 164 . 00 =0.695 236,000 corresponding to an efficiency of yT695 = .834 for the single conversion of electric energy to mechanical energy, or vice-versa. If a lower average current of the generator were to be assumed, the efficiency would be proportionately higher. It may be practical to reckon with more than 83.5 per cent efficiency; which was purposely not done, however, for the reason to be men- tioned in the latter part of this paper. BO 70 60 60 40 90 20 10 10 20 80 40 60 B X x-" Ci -\ x. 4WW 80000 20000 10000 I / 'V \ // V \ \ DI ^ fr \ S, A! / \ E a / \ / \ / V E A \ i 1 1 / \ / \ / \ / \ / r 2 FJQ. 3. ENERGY AND POWER OF BUFFER. We will not define how much of the energy stored in the centri- fugal masses is to be used, but state only that at the end of an unloading period there has still to be stored an energy L % + mkg This energy storage will be T r ^ < 70 Giip > re 100 80 i i 60 40 20 20 40 60 Buffer current 80 100 120 600 ' t I Jo ' 400 ' 360 ' 820 ' 280 Line current \ I I t l ' 240 ' J IB 880 3GO 840 Generators current FIG. 5. B> the difference of which is the greater as the current is lower. Ob- viously, this condition is not to be reached in practice. However, we may influence the regulation by the current, one value of it will always call for only one definite value of the speed. The same considerations regarding the buffer current holds true also for the actual line current and for the generator current. To Fig. 5 are added values of the actual and generator current, which prove in this case that a suitable regulation cannot be reached, since one of them, decreasing or increasing according to the network current load, cannot produce two different speeds of the buffer machine. It would be possible to influence the field of the buffer machine by the combined buffer current and line current, in ad- dition to a special shunt winding fed from the bus, but even this would not give the correct result. We wish to transmit the fluctua- 272 RASCH: BOOSTERS IN TRACTION SERVICE. tions of the line current to the buffer current, which means that both of them must reach their highest values simultaneously, as shown in Fig. 2. The two currents working in two different windings on the magnet field of the buffer machine would, according to the way they were connected, work either with or against each other. In the first case they would not reach the desired result; in the second case, the result would be equal to that which could be reached with one of the currents alone, which is not suitable, as above mentioned. The question of making use of a change in speed for regulating the buffer effect gives us two possibilities : (1) Using a centrifugal regulator to adjust a rheostat placed in the field winding of the buffer machine. (2) The use of a special exciter machine driven from the buffer machine, and having therefore a speed proportional to that of the buffer machine. 2 The latter seems to be the most favorable, since this regulation is not applied step by step, but gradually. We assume again a steady voltage on the bus, and assume further- more that the e.m.f. of the buffer machine with maximum current varies from 5 per cent above to 5 per cent below this voltage ; in this case the vibration of e.m.f. during the period of 12 seconds is shown in the curve A z B 2 C 2 D 2 E 2 (Fig. '4). The proportion of the ordinates of this curve, and of the curve ABODE that is E E M.F. U ~ ~ speed are proportional to the lines of force which are necessary for the field of the buffer machine to produce the desired regulation. Fig. 6 shows a diagram, the abscissae of which are proportional to the V value =, and the ordinates proportional to the speed. In this case there are also generally two ordinates for one abscissa, but the dif- ference in the most unfavorable case is 8 per cent, while the regula- tion by means of the currents showed differences of 30 per cent (Fig. 5). If both branches of Fig. 6 could be combined which would call for no armature resistance whatever in the buffer ma- chine the regulation would be a perfect one. We remark, there- fore, that a low armature resistance is favorable. 2. Both kinds of regulation are the subject of the German patents No. 129,553, assigned to the German General Electric Co., Berlin. KASCH: BOOSTERS IN TRACTION SERVICE. 273 The connections may be as follows (Fig. 7) : The buffer machine P is provided with two field windings M t and a . M l ia excited from the busses ; M 2 from a small generator e, 100 90 80 TO 5 ^ \ \ \ \ \ \ ^ J> 6.0 7.0 8.0 9. FIG. 6. placed upon the axle of the buffer machine, and excited from the bus also. The windings M t and M 2 are differentially connected, so that with increasing current in Jf 2 , the total field of the buffer machine will be weakened. Another arrangement would be to use a part of the winding M FIG. 7. as the second winding M 2 , as shown in Fig. 8. The latter arrange- ment has the advantage of a smaller winding space necessary for the buffer machine ; it has the disadvantage, however, that the small machine e has eventually to be designed for rather high voltages, while with an arrangement according to Fig. 7, voltage and current of the small generator may be chosen at will. ELEC. RYS. 18. . 274 RASCH: BOOSTERS IN TRACTION SERVICE. The small generator may also be designed as motor, the arma- ture of which is to be connected in series with the magnet winding M of the buffer machine (Fig. 9). In this case increasing the speed of the buffer machine will cause an increase of the e.m.f. of the FIG. 8. small machine, and result in decreasing the exciter current of the buffer machine. In all arrangements, while working, there is a weakening of the field of the buffer machine, and, therefore, the di- mensions of these machines must be made ample, in comparison with other machines of the same average load. We may mention FIG. 9. that in all cases the maximum armature current of the buffer machine does not reach its highest value at the same time with the weakest field, but with an average strong field, which, of course, is favorable. The discussion has shown that the most suitable regulation of the buffer machine is to be effected by means of speed regulation. STORAGE BATTEEIES IN ELECTEIC RAILWAY SERVICE. BY JUSTUS B. ENTZ. The principal applications of batteries to electric railway systems are made at the generating stations, at distributing sub-stations, and directly connected to points on a direct-current distributing line. The objects of such installations are to store electrical energy at efficient and convenient periods and to return it when most useful, generally at periods of increasing or heavy load. A storage bat- tery, which is a reservoir of electrical energy, when connected to the circuit, makes the conditions of generation and transmission up to the point where the battery is connected independent of the load demand of the circuit beyond that point. The storage battery also permits the rate of production of energy to be independent of the rate of demand. The demand for electric current may occur at a time when its production will be inconvenient or inefficient, or both, The demand may call for a very high rate of output for a short time, as in electric railway systems, where the maximum demand lasts only for a period of a few seconds. By the use of a battery the rate of producing energy may be adjusted with reference to other considerations, such as efficiency of generating apparatus. This energy may be produced at low and constant rates and stored in a battery and given out to meet the demands of variable and very high rates. The results thus obtained are improved efficiency of operation and greater reliability of service. In comparing a battery with the generating and transmission ap- paratus, it will be noted that the battery handles most economically those portions of the load which are least economical for generating or transmission apparatus, namely, those which are of extremely short duration and excessive in amount, as well as those which are of considerable duration, but of very small amount. The maximum economy of usefulness is secured by such division of load between the battery and other apparatus that each handles that portion to which it is best adapted. [275J 270 ENTZ: ELECTRIC RAILWAY STORAGE BATTERIES. The reasons for installing storage batteries in railway work are as follows : 1). Keasons affecting investment. 2). Reasons affecting economy of operation. 3). Reasons affecting reliability and public convenience and safety. It is impossible to draw a sharp line between the three classes of reasons. A part of the total investment for railway equipment is made for economy or for reliability. The consideration of a bat- tery in this class of work usually involves the comparison of bat- tery with generating machinery, or in some cases the transmission copper and sub-station equipment, or with all of them on the three headings enumerated above. In such a comparison the following points must be considered : 1). The results of the comparison of investment will depend upon the shape of the load diagram and upon the methods and dis- tances of transmission and upon what portion of the load is as- signed to the battery. In general, however, it may be stated that there is almost invariably some portion of the maximum load which may be carried by a battery at an investment cost not exceeding that for the apparatus it actually displaces to do the same work. It is often sound engineering to increase the proportion of the bat- tery considerably beyond that point to secure more fully the ad- vantages in headings 2 and 3 ; and it must be borne in mind that many of the functions of a battery can be performed by no other class of apparatus, and where these functions are vital, the invest- ment comparison is of minor importance. 2). Under the head of economy of operation must be included both generation and transmission of energy and both labor and fuel economy, as well as cost of maintenance and depreciation. As loads of certain nature are handled more economically by the bat- tery than by generating machinery, the maximum fuel economy is secured by such division of load between the battery and machinery that each handles that portion for which it is best adapted. The question of the size of a power equipment is, of course, confined to a determination of the requirements necessary to meet the maximum load conditions, whereas, considering economy of operation, the average load conditions must be considered. In the sub-station there is an increased efficiency of rotaries and trans- formers due to the operation of batteries which must be compared with the losses in the battery. With the batteries used very con- ENTZ: ELECTRIC RAILWAY STORAGE BATTERIES. -277 eiderably on peak work, their output will amount to from 15 to 20 per cent of the total output of the station. As the hattery under average conditions will not be fully used, but a certain por- tion of it held in reserve to meet abnormal conditions, the efficiency of the battery will be high. Taking this at 85 per cent as a mini- mum, we find that the losses in the batteries, where their output is 20 per cent of the total, is 3 per cent of the total output of the system. It is safe to say that on account of the improved efficiency in transmission and the improved load factors on the rotaries, the efficiency of the sub-station should be increased by considerably more than this amount, and that any improved economy in the generation of power at the main power-house will be net gain. The economies at the power-house from the operation of bat- teries will be such as to produce ideal economy in both boiler and engine-room. The load on the engines and boilers can be adjusted to practically the 24-hour average, and need be varied only when this average is changed. With peaks in the morning and evening to double the height of the average load, this will mean operating through the day at practically one-half the capacity that will be needed at the peaks without the battery. To handle these peaks without the battery, it would be necessary to keep one-half of the total boiler capacity with fires banked from 18 to 20 hours a day for operation during the hours of peaks. The constant loss in these boilers through radiation and the escape of heated gases would probably not be less than 20 per cent of their capacity; and one- half of these losses, or 10 per cent of the boiler capacity required with the battery in service, would be saved. This would mean a saving of 10 per cent of the total fuel. The improved load factor on the engines and generators and the reduction in the number of engine hours of operation would effect an additional economy. There is a considerable loss of steam when every unit is started up, this being the steam consumed from the time the throttle is opened to the time the load is thrown on the generator. As the operation of batteries would reduce the number of times the unit is started up and shut down, there would be a saving on this point. I believe that it would be conservative to expect a saving on this point of from 5 per cent to 10 per cent, making a total saving of fuel in the operation of a power plant with batteries of from 15 to 20 per cent. It has been stated that a storage battery is a good thing to patch 278 UNTZ: ELECTRIC RAILWAY STORAGE BATTERIES. up bad engineering. This is true, and there is a considerable field for its application in this way. It is, however, not limited to such cases, and is often the only means of preventing engineering prov- ing to be bad owing to the impossibility of foretelling what con- ditions are to be met exactly. The extreme flexibility of a battery in meeting conditions varying over a very wide range renders it peculiarly applicable to such cases. Under the question of maintenance and depreciation it may be noted that with a storage battery these two items are combined in one. The renewals of plates which are made from time to time keep the battery up to date, so that at the end of a period of years it is not an obsolete piece of apparatus, but it is up to date in every respect and equal to the batteries then in the market, including all the improvements in plate construction which have been introduced since it was installed. The flexibility of the battery to meet changes in conditions, such as desirability of increased voltage or larger capacity, is also to be noted, such changes in conditions often in- volving the discarding of generating apparatus; whereas in a bat- tery, the simple modification in the number of cells, or the number of plates in each cell, will suffice. 3). The reliability of the storage battery and its absolute freedom from break-down without warning is due in part to the fact that it is composed of a multitude of small units, each unit being a bat- tery plate, any one of which can be put out of service without noticeably affecting the operation of the entire installation ; whereas, in a generating plant, the various parts, such as boilers, engines, generators, switchboard apparatus, transmission lines, transformers, and converters, are all connected in series and the derangement of any one class of these parts instantly interrupts completely the operation of the whole. The deterioration of a battery is in all cases very gradual, and repairs can be made without taking the battery out of service. As an emergency reserve, the battery can be found of immense value in any one of the following ways : a). In case of a total shut-down of the power-house or high-ten- sion lines, the amount of battery which would usually be installed from other considerations would be sufficient to maintain the en- tire service of the road at the time of the peak for three-quarters to one and one-half hours, or for twice as long during the middle of the day, thus permitting temporary repairs to be made. In case of un interruption of longer duration, the battery would at least CD- ENTZ: ELECTRIC RAILWAY STORAGE BATTERIES. 270 able the trains to be run into the station and the passengers dis- charged, instead of leaving them stalled between stations. b). At a sub-station the rotaries could be shut down for an in- definite period of time, the battery being floated on the line at a somewhat reduced voltage. c). The batteries are available instantly to take care of sudden excessive load of short duration, due to any unusual congestion of traffic. d). They will take care of and prevent interruptions from short- circuits on the line which would otherwise fall on the machines, saving overloading them and then throwing out the breakers and interrupting the traffic. e). The batteries would permit the entire machinery of the power- house and sub-station to be shut down at night and the current cut off the alternating-current lines for a period of several hours for repairs and inspection. /) . The batteries would often make it possible to purchase either alternating or direct current from other systems at times when they were not overloaded, and at a constant and controllable rate which would cause no disturbance. This power could be utilized on the system at times of peak load, when it probably could not be pur- chased. The fact that the batteries are available in case of emergency would permit the shutting down of machinery when signs of trouble first appear, thus reducing the extent of the damage which might be caused by continuing to run partially disabled machinery until a substitute could be put in service. The points enumerated above apply to batteries installed at the power-house and those installed on the line. Certain additional ad- vantages arise in many cases from installing a battery at some dis- tance from the source of power, due to the improved conditions of transmission. With such a battery, it becomes necessary to trans- mit only the average power required instead of the maximum. The result will be a saving in the amount of copper required for a given drop in voltage, or an improvement in the voltage with a given amount of copper, or the advantages may be divided between the two methods. An increase in economy will also be secured, since it is a well-known fact that to transmit a given amount of energy over a certain conductor in a given time, with a minimum loss, the rate of transmission should be constant. The installation of a storage battery at a generating station is 280 ENTZ: ELECTRIC RAILWAY STORAGE BATTERIES. to take the peak of the load for its maximum two or three hours, and to regulate or control the rapid fluctuations of load occurring all day. Where the station voltage has not a drooping character- istic it is necessary to add to the voltage of the battery the voltage of an auxiliary generator, in order to cause it to discharge at the time and by the amount necessary. This auxiliary generator, com- monly called a booster, also serves for charging the battery without varying the bus-pressure of the station by adding its voltage to that of the bus, the armature of the booster being in series with the bat- tery and its field strength being automatically controlled where the changes of load are at all rapid. When located at some distance from the power-house, the booster may be dispensed with, as the variation in the line voltage will be sufficient to cause the battery to do its work. Located in this way the battery will maintain the voltage on the line at approximately its average point. If the num- ber of cells in the battery are properly adjusted to float this aver- age voltage, the battery will remain in the same average state of charge. If the average voltage at the point where a line battery is located is found too low for satisfactory results, a booster may be installed at the power-house and sufficient current transmitted over a feeder direct to the battery at a voltage higher than the bus to maintain the battery voltage at the desired point, this latter arrange- ment affording means for adjusting the voltage at the battery to meet changes in local conditions which is usually very desirable. Such installations are very satisfactory and economical, showing a saving in investment over copper and generating machinery, as well as a considerable saving in energy, as not only is the energy transmitted at its average current value to such a point on the line, but the average current consumption is lessened by the increase of voltage at the point of consumption; and this increase and main- tenance of voltage very often brings about an actual reduction in wattage at the point of consumption, because of the higher accelera- tion rates permitted by the cars themselves, resulting, as is well known, in a considerable reduction in energy consumed where stop- ping and starting is at all frequent. A booster for such a purpose is usually an independently excited booster located at the power- house and hand-controlled, so as to have control over the average output over the battery feeder. The automatic control of a battery by its booster when the battery is connected in parallel with generating machinery of a conptant or rising characteristic is accomplished in one of the following ways : ENTZ: ELECTRIC RAILWAY STORAGE BATTERIES. 281 A regulating battery is generally discharged at a rate at least as high as its one-hour rate ; that is to say, the rate at which it would discharge continuously without its voltage drop becoming too great. This does not mean that a battery is totally discharged at this rate in one hour's time, as a reduction in the rate would permit consid- erably greater capacity to still be taken out of the battery without its voltage falling too few. Within the full range of the one-hour capacity of a battery the voltage change for a change of the one-hour rate of current is from 5 to 7 per cent, due to the internal ohmic resistance of the battery, and this change of voltage is simultaneous with the change of current. If the full rate of current be main- tained for 30 seconds, an increased change of voltage of from 4 to 5 per cent will take place in about 30 seconds' time, due to polariza- tion. After 30 seconds the increased change of voltage due to po- larization is comparatively slight, except at the very end of discharge or of a full charge. The booster must, therefore, be pro- vided so as to give a voltage of about 12 per cent of the battery volt- age at the time that the battery is charging or discharging at its maximum rate; and we must further insure that it will give a voltage of 20 per cent of that of the battery at a rate of current of from one-third to one-fifth that of the maximum rate, in order to bring the battery up to a point of full charge. The characteristic of most boosters allows them to give this additional voltage at re- duced current with but comparatively little increase in the size of their field magnets. The automatic excitation of the booster field is accomplished either by including an exciting coil in the working circuit by means of which the full output of the station to be regulated passes through this coil, so that an increased load demand strengthens the booster field and gives added voltage to the battery circuit sufficient to cause it to discharge by an amount equal to the increase, thus keeping the load on the generator constant, or to take any propor- tion of the increased load that is desirable. Such a main exciting coil in the working circuit must be neutralized by a separate excit- ing coil, so that with any predetermined average output of the station the booster shall neither add nor oppose its voltage to that of the battery. For currents below this established load, this opposing coil becomes stronger and reverses the polarity of the booster, caus- ing the battery to charge by the proper amount to maintain the regulation desired. In order to make such a combination as stable as possible, another main-current coil has been included in the gener- 28i> ENTZ: ELECTRIC RAILWAY STORAGE BATTERIES. ator circuit, so that an increase of current falling upon the gener- ator following an increase of outside load would further affect the battery and cause it to discharge. Where the outside main-current coil has been adjusted to exactly meet the state of the battery and so effect absolutely constant current delivered from the generator, this inside coil in the generator circuit, of course, accomplishes no purpose ; but it prevents any lack of exact adjustment affecting the regulation to any great degree, and where very perfect regulation is required, this form of booster is very extensively used and is gen- erally known as a differential booster. Eegulating altogether by variations in the generator load while trying Jx> keep that variation within small limits, calls for some means of magnifying the effect of such variations upon the booster excitation. There are two methods of this kind in general use, in one of which a small generator with a voltage normally equal to that of the station bus has included in its circuit the exciting coil of the battery booster. When the voltage of the small generator and that of the bus are equal, no current flows through this booster ex- citing coil. This small generator is known as a counter e.m.f. gen- erator, and derives its field excitation, and, consequently, its voltage, from a coil placed in the generator circuit, the said coil being so adjusted that the average load that is to be kept upon the generator produces a voltage of the counter e.m.f. generator equal to and op- posed to that of the station voltage, so that under such conditions the battery is neither charging nor discharging. If, now, the gener- ator output increased 10 per cent, the voltage of the counter e.m.f. generator, if it has a perfectly straight characteristic, will increase 10 per cent above the station voltage, and this excess of voltage should be sufficient to excite the booster to an extent necessary to cause the battery to discharge the balance of the load increase which caused the increase upon the generator, part having fallen upon the generator for the purpose of effecting the regulation. The lowering of the generator output following the lowering of the station out- put acts in the same manner, sends a reverse current through the booster field and causes the battery to charge. If, as cited above, regulation of the generator load within 10 I er cent were to be maintained, the output of the counter e.m.f. generator would have to be 10 times that of the energy required for the field excitation of the booster, as but 10 per cent of its voltage is applied for that purpose. If the regulation were to be 5 per cent in either direction, the output of the counter e.m.f. generator ENTZ: ELECTRIC RAILWAY STORAGE BATTERIES. 283 would have to be 20 times that of the energy required for the booster field excitation. The excess output is, of course, not lost, but passes to the line. The maintenance of any fixed load upon the generators in this system is controlled by means of variable shunts around the exciting coil of the counter e.m.f. generator, which carries all the generator output. The other method of regulating by variations in the generator load is by means of an electromechanical regulator. This regulator consists of two or more groups of carbon discs, connected in the manner of the Wheatstone bridge, with the exciting field coil of the booster connected in the position of the galvanometer. A pivoted lever is so mounted that its movement brings pressure to bear upon one set of the groups and releases it upon the other, so as to change their respective resistance and to vary and reverse current through the field of the exciting coil. To one end of the lever an adjustable spring is attached and to the other end a magnet core influenced by the current in the generator circuit. At the average generator load which is to be maintained, the pull of the magnet is balanced by the pull of the spring at the other end of the lever. Under these conditions the pressure upon the two groups of carbons is the same, and no current flows through the booster field coil. A slight in- crease of current in the generator circuit is sufficient to cause addi- tional pressure upon one of the groups of carbons compared with the other, and send current to the field regulating coil of the booster in a direction to cause the battery to discharge, which it does to an amount practically equal to the increase of load in the outside cir- cuit, letting only a small portion of the additional load fall upon the generator to effect the regulation. If the generator load is de- creased following the decrease in the outside load, the spring be- comes stronger than the magnet, and a pressure is put upon the opposite group of carbons, reversing the current through the booster field coils and causing the battery to charge. It has been found that very close regulation can be maintained in this way, even with a load varying almost instantaneously. Regula- tion of less than 2 per cent in either direction has been frequently obtained. Complete control of the output of the generators is -secured by this system, and the generators can be set to run at any average load desired, by simply varying the strength of the spring opposing the magnet. If the pull of the spring is increased, the ienerator current is immediately increased to a corresponding de- cree, as otherwise the battery would charge till the increase of the 284 ENTZ: ELECTRIC RAILWAY STORAGE BATTERIES. generator load would balance the spring pull. The end of the spring carries a pointer, and there is a calibrated scale in amperes by which the generator output can be instantly set at its desired value. This form of regulator is mounted on a switchboard, and occupies not much more space than the ordinary recording wattmeter. The spring and its indicator, as well as the carbons, are on the front of the board, and the lever extends through the board, and in stations of any considerable size carries a simple horseshoe of soft iron which is hung over the bus-bar carrying the total load of the generators. The usual connection for such a regulator is to have, electrically considered, two groups of carbons. These are connected all in series, and by means of a connection made to the storage battery a small current is maintained through them. At the middle point of the carbons, which is the point where pressure is divided, a lead is taken through the field coil of the booster to the middle point of the battery, to which the two ends of the carbons are connected. In this way, when the pressure on the two groups of carbons is equal and the resistance is, therefore, equal, there is no difference of po- tential between the midway point of the carbons and the midway point of the battery. In plants where very large boosters are used it is desirable to magnify the effect of the regulator by means of an exciter con- nected between the regulator and the booster, rather than to increase the size of the regulator. This regulator has some advantages over any other method of battery regulation, in that it is possible to ad- just the sensitiveness of regulation on the charge side of the battery as compared with the discharge, and vice versa. For instance, in a generating station or a sub-station with a fluctuating load, it is not necessary or always desirable to maintain the load on the generating machinery absolutely at its ratings but to allow it to share the in- creased loads to some considerable extent, in order to reduce the discharge rate of the battery. If this is done by any direct means of field-coil regulation it will follow that if the generating apparatus shares a portion of the overload it will also have to share the under- loads, or loads below the average, in the same proportion. With the carbon regulator, on the other hand, by the introduction of a resistance in one of the groups of carbons, the regulation on discharge, for instance, may be made of any degree of sensitiveness, so as to allow the generator machinery to share any portion of the overloads, while on the underloads full sensitiveness of regulation ENTZ: ELECTRIC RAILWAY STORAGE BATTERIES. 285 may be maintained and the load on the generators not allowed to drop off below the average. In this way the battery may be accu- mulating charge; as it receives more charge than discharge, the actual variation of load on the station is considerably lessened and the maximum output of the generating machinery and the battery is not increased. In this manner the overload capacities of en- gines, generators, rotary converters, etc., may be utilized to the fullest advantage, and the battery may be discharging at very high rates; but by taking full advantage of every dropping off of the load below a predetermined point sacrifice as little of its capacity as possible, and may assist the generator on the peak of the load, while losing but a minimum of its capacity. Also with this form of regulator, a zone of non-regulation may be created extending say from 10 per cent above and below the average load, whereas for loads above and below this the regulation may be as perfect as pos- sible. This permits of reducing the total amount of charge and discharge in ampere-hours that a battery may receive by a very great amount, while keeping the variation of the load on any system within non-objectionable limits, and the life of the battery may be materially increased, often without reducing any of its benefits. As to the construction of a battery for railway service, it is pretty well established that the positive plates should be of the " Plante " type and not of the " pasted " type ; while the negative plates are preferably of the " pasted " type. The characteristic trouble of negative plates has been loss of ca- pacity due to shrinkage of the spongy, finely divided, active material into a denser and less porous material. This has particularly been true of " Plante " negative plates, where the active material is rela- tively small in quantity and has been reduced from the peroxide previously formed from the plate itself. A process has been dis- covered of manufacturing a negative active material which always retains its loose, spongy, porous condition ; but, as this has but little mechanical strength, means have to be provided in the plate for retaining it in position. Such plates have proven eminently satis- factory in service and extended tests show a very greatly increased life and the maintenance of low resistance and low polarization factors. In considering the life of a positive " Plante " plate, it should be taken into account that the life in ampere-hours of a pound of lead entering into the construction of a positive plate is not gov- erned by the surface development of the lead, but by the means 286 ENTZ: ELECTRIC RAILWAY STORAGE BATTERIES. which have been provided for retaining the active material formed from the plate itself in proper contact with it, and the prevention of the loss of such active material from being washed away or car- ried away by the gases which rise from the surface of the plate. The extended development of a pound of lead increases the capacity which it would yield on any one discharge, but lessens the total number of discharges available by more than a proportionate amount, a very highly-developed plate yielding less total life in ampere-hours mainly because its mechanical structure and its con- ductivity are affected to a greater extent by the removal or loss of a portion of the substance of the plate. For this reason the de- velopment of the active lead should be made in such a manner that it will provide secure receptacles for the retaining of the active material; and the necessary further corrosion of the active lead for the purpose of replacing active material carried away should not interfere with the mechanical strength or with the conductivity of the plate. No modern battery installed for railway service should be in dan- ger of a break-down at any rate of discharge that could possibly be imposed upon it, and in well-designed batteries there is absolutely no danger of break-down due to any rate of overload. Some years ago the electrical engineer was disposed to look upon a storage battery with more or less misgiving. Even at the present time there may be found occasionally an engineer who, not realiz- ing the progress that has been made in this art and the place that the storage battery has established for itself, is disposed to take this skeptical attitude. If, however, the history of the storage battery business for the past 10 years, which period practically covers its entire commercial history, should be compared with the first 10 years of any other electrical apparatus, we believe that the com- parison will show a series of complete successes and the absence of anything approaching a failure or setback, that will compare favor- ably with the history of any other electrical apparatus. DISCUSSION. CHAIRMAN DUNCAN: The paper is now open for discussion. I would like to ask Mr. Sprague if the New York Central is going to put in bat- teries, or if he can say whether they are or not? Mr. F. J. SPRAGUE: That is a question I cannot answer at present. My experience with storage batteries has been such as to lead me to regard them with favor in some classes of work. On the South Side Ele- ENTZ: ELECTRIC RAILWAY STORAGE BATTERIES. 287 vated road in Chicago, storage batteries were introduced for two reasons, one to help take the sharp fluctuations in load, and the other to provide additional facilities when the demands of the road were growing so rapidly as to run ahead of possible direct equipment. No boosters were used, the batteries responding fairly well automatically to the rise and fall of potential where connected to the line, but varied somewhat in action by cutting in or out an extra feeder. The New York Central presents a problem which is materially different from that of elevated and suburban roads. Usually on those classes of service there are a large number of units, and the load is fairly dis- tributed. The New York Central has about nine sub-stations, the units weigh from 150 to 700 tons, and the sub-stations are a considerable dis- tance apart. It is impossible to avoid a condition which is emphasized on heavy steam railway work, extreme local variations of load. There will probably be at times as many as four trains supplied almost entirely by one sub-station, while at other times there will not be any load whatever on it. Of course that means a pretty large variation, and sometimes a very rapid one. Personally, I am strongly in favor of the use of storage batteries in this instance, not only partly to relieve the sub-station machinery and to reduce its capacity; but also to provide a reserve, in case of any acci- dent to the central power plants or transmission system. The equipment is being laid out with the idea of maintaining train movements from two stations, either of which in emergency can operate the entire service for a reasonable time. I do not think that in the matter of cost, all things considered, there would be much difference be- tween the installation of a plant with or without storage batteries, that is, the saving of central station and sub-stations would be about offset by the cost of batteries and boosters. ELECTROLYSIS OF UNDERGROUND CONDUCTORS. BY PROF. GEORGE F. SEVER, Columbia University. In the spring of 1903, Mr. L. B. Stillwell, Mr. F. N. Water- man and the writer felt that it was desirable to compile and co- ordinate as much information as could be procured on the subject of the electrolysis of underground conductors, due to the operation of electric railways. It was felt that both the opinions regard- ing electrolysis and the practice in remedying the same were so diverse that it would be of value to collect all this information and present it before the International Electrical Congress. Through the efforts of the first-named gentlemen the practice of the electric street railways was secured, and all the world's litera- ture, which was available, was collected and put into the form of a digest. The writer collected information regarding the atti- tude of the municipalities, including such ordinances regarding electrolysis as had been put into effect up to that time. The data was put into tabular form by Mr. Waterman, and through the courtesy of both Mr. Stillwell and Mr. Waterman the writer has been able to present the final results before this Congress. The data is presented in the five tables which are attached hereto. Table I shows the street railway practice in the United States regarding the use of return feeders and the effect of increasing the capacity of these feeders. The reports are shown from 102 electric railways. Table II shows the recommendations which have been made to 29 municipalities by city and other engineers. The results of these recommendations are shown in a few cases. Table III shows the most essential electrical features of the municipal ordinances which are in force in 12 different munici- palities. The inconsistencies in some of these ordinances are remarkable, particularly in the cases of Atlantic City and Altoona. [288J SEVER: UNDERGROUND CONDUCTORS. 289 Table IV presents a summary of the opinions of municipal offi- cers as extracted from the letters received from them. Fifty municipalities, widely distributed, were heard from. Table V presents a summary of expert opinion concerning elec- trolysis. This expert opinion shows many differences in the recom- mendations as to remedy. It is the writer's hope that the discussion on this presentation may be full and that some definite conclusions may be arrived at for the betterment of the conditions which are known to exist in some localities. ELEC. RYS. 19. 200 SEVER: UNDERGROUND CONDUCTORS. TAB SUMMARY OF STKF-ET RAILWAY 1 E: X 1 1 a 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 28 1 03 City. | 1 | (2 Name of electric street railway company. System of operation. Date of electrical in- stallation. Miles of track. Weight of rails. Bonding system . Number and size of 1 bond per joint. Nature of re- turn feeder system. Ala... Ala... Conn. Conn. KJonn. Fla.. Ga... Ga... 111. . . . 111.... 111.... 111.... m.... Ind.. Ind.. Ind.. Iowa. Ky... Me... Me... Me... Md... Md... Birmingham Huntsville 88,400 8,100 9,600 9,600 2,400 28,500 10,200 12,000 22,500 1,700,000 25,000 18,800 15,100 7,800 10,500 18,100 15,000 20J,000 22,000 8,000 8,100 17,100 18,600 Birmingham R y . Lt & Pw Co 1894 1900 1895 99 8 10 13 18 7 27 12 184 14 5 4 7 18 8 142 27 7 2 7 14 60- 88 60 56- 80 60 60 45- 70 40- 60 70 60- 75 95 60- 70 50 40- 60 56 30- 86 60 56 60- 100 56 48- 56 60 40- 73 56- 72 W.&M.. Copper . . Copper.. Copper . . Copper . . Chase S.. Protect.. All wire. 4-0 4-0 2-0 2 4-0 1-0 2-0 2 2-0 2 4-0 2-0 4-0 4-0 2-0 4 4-0 1-0 2-0 4-0 None Huntsville Ry. Lt. & Pw Co None None Bristol B. & P Ry. Co.... Middletown Montville . . M. St. Ry. Co M St Ry Co No. 2 track wire None Jacksonville .... Athens Dahlonga . . 1 ret. feeder. None A. Elec. Ry. Co G. & D. Elec. Ry. 1895 None Alton Chicago A. Ry. Gas & Elec. On all lines. . On all main lines Cast weld Copper .. Wire .... Protect. . Wire Wire A.S.&W. Wire .... A.S.&W. Chase S. protect. Chic ago protect. Brown crown . . CJDI er None F. Ry. Lt. & Pw. Co Ret. feed.... None None.. Jacksonville .... Columbus J. S. Cramp's Elec. None LaFayette .. Keokuk L F St Ry Co... On one line.. None K. Elec. Ry. & Louisville Bangor ......... L. Ry. Co Penobscot Central M to 2 miles from P. S.. None Ret. feeds... None No.2gr.wlre None Calais Kennebunkport. . Cumberland* . . Hagerstown Atlantic Shore Line Ry H Rv Co... Copper . . 2-0 4-0 8L'\'1-:R: CONDUCTORS. 201 LE I. PRACTICE IN THE UNITED STATES. .S 2 | 8 a 6 I i 1 m 1 Q. S* 1- a, 1 of power statioi :imum curre from each. imumlinevolta Nature of soU. Nature of corro- sion. Extent of corro- sion. Any claim against railway company? What remedy applied? Effect of remedy. p E 3 -S X S .2 N'o... No... None . 1 350 Some. ... No. None No No None 1 300 475 Fav . . No No No None 1 500600 No No Yes.. 1 500450 No No... No... None . 1 2,500350 Some Yes Larger bonds. . No more trouble. No No... None . 1 700375 None .... No No No .. None . 1 400 450 Unfav.. No No No None . 1 500 4OI No No... No... None . i 1,200 450 None .... No No... Yes.. 4 12,000400 Some. ... Nothing definite. . None No . No... None . 1 8,600 450 Fav.... Some . . . No None No No... None . 1 400425 No None . No... No... None . I '450 None .... Once Analysis showed rust. Yes Yes.. 1 250480 None No No... No... None . 1. 500 Unfav.. None.... Once Proved earth No No... None '. 1 500450 None No No .. No... None . 1 500450 None .... No Yes.. Yes.. 1 10,000 500 Some .... Yes Ret. feeds. .... Less com- plaint. Mo... No... None . 2 480 None.... No Tee.. Yes.. 1 850 450 Some .... Yes Improving re- Less trou- ble. N.... No... None. 3 1,000 360 None . . , No ) No... No No... No... None. 1 800 350 Fav. at point At one point .. No None 21)2 SEVER: UNDERGROUND CONDUCTORS. TABLE I Number. | 1 City. Population served. Name of electric street railway company. d .0 Date of electrical in- stallation. Miles of track. Weight of rails. Bonding system. Number and size of bond per joint. Nature of re- turn feeder system. 24 25 20 27 28 29 30 31 32 83 84 85 36 87 38 89 40 41 42 48 44 45 46 47 48 49 Mass. Mass. Mass. Mass. Mass. Mass. Mass. Mass. Mass. Mass. Mass. Mich. N. H. N. J. N. J. N. J. N. J. N.Y. N.Y. N.Y. N.Y. N.Y. N.Y. N.Y. N.Y. N.Y. Amherst Athol 5,000 7,100 A. & S. St. Ry. Co. A. &O. St. Ry C. C. & E. Traction Co :::::: '.'.'.'. 15 7 5 B 32 16 23 60 50- 90 50 60 70- 90 45- 60 48- 60 Copper.. S.V'.'&B. Crown. . . 4-0 4-0 1-0 2-0 4-0 4-0 4-0 2 4-0 2-0 2 4-0 Fewgr. wires Ret. feed.... Gr. wire None On all lines. . None 1,600 12,400 10,000 8,000 45,700 45,700 658,000 8,100 9,500 2,500 4,000 76,000 8.100 10,60 C. Elec. St. Ry. Co. F. & L. St. Ry. Co. G.W.&F.St.R.R. Co Fitchburg ...... Gardner ........ Greenfield ...... fG.D.&N.St.Ry. 1 Co Holyoke 1 G. & T. F. St. Ry. [ Co I None Holyoke Lowell H St Ry Co Crown... On one line. . On nearly El- lines None . None Some 2-0 for each track B. & N. U. Ry. Co. C. M. & H. St. Ry. Co 440 18 8 8 24 67 6 12 42 37 8 6 27 7 4 19 5 90 60- 90 45- 60 48 60- 70 70- 90 65- 70 60 80 40 56- 90 50- 90 56 48- 60 56- 80 56 Crown... Escanaba Chester Asbury Park.... Camden KevDort E. Elec. St. Ry. Co. C. & D. Ry. Ass'n.. A. C. Elec. R. R. Co .... C. & Sub. Ry. Co. J. C. Traction Co. M Traction Co .... None Millville .. M. & E.. Copper . . Crown... M. & E.. Chicago plastic wire . . 4-0 4-0 2 4-0 None None Albany A. & H. R. R. Co -j B.Ry. Co Ont. Lt. & Trac. Co Third rail. E: Binghamton Canandaigua .... Corning El mi ra .......... 60,000 6,000 12,500 85,700 8,700 18,500 18.000 12,000 Ret. feed.... Ret. feed.... None C. &P. P.St.Ry.. E. Wat. Lt. & R. R. Co .... Ret. feed.... Ret. feeds... Fishkill Citizen R. R. Lt. & Fredonia Gloversville Hornel Lsville . . . D. &F.R. R. Co.. F. J. & G. R. R... .... M. & E. O.B.Co. Copper . . Copper . . 4-0 2-0 4-0 Ret. feed.... SEVER: UNDERGROUND CONDUCTORS. (Continued). 2 2 s s 43 o 6 0 None . . . No I No... Yes.. 1 150 too None No No... No... None. 1 850 175 Fav. ... None .... No No... No... None. 1 On rails Yes Ret feed Economy Ho... Yes.. 1 Some.... I n d e fl - nitely. . . He... No... None. 1 800 500 Clay, gravel. Some No Better bonding Suppressed. No... , { 400 Fav NTrmo No... 1 470 i.. None None ! 294 SEVER: UNDERGROUND CONDUCTORS. TABLE I 1 ~50 51 52 53 54 55 56 57 58 59 60 61 63 63 64 65 66 67 68 69 70 71 72 73 74 1 City. Population served. Name of electric street railway company. 1 OQ Date of electrical in- stallation. Miles of track. Weight of rails. Bonding system. s i n frnTS E- Nature of re- turn feeder system. N.Y. N.Y. N.Y. N.Y. N.Y. N.Y. N.Y. Ohio. Ohio. Ohio. Ohio. Ohio. Ohio. Ohio. Ohio. Ohio. Ohio. Pa... Pa... Pa... Pa .. Hoosick 30,000 3,000 13,000 28,000 7,500 10,500 114,000 15,500 1,400,000 882,000 126,000 8,800 125,000 22,000 22,000 16,500 12,000 1,600 9,600 53,000 Ben. H. Valley Ry. 17 8 8 21 17 18 88 6 42 ^ 60 56 45- 60 45- 80 80- 100 45- 73 80- 90 50 70 Plastic wire Wire.... 1-fl 2 1-0 2-0 3-0 2 800- 000 2-0 2 250- 000 2-0 None None None Huntington Ithaca Jamestown . H. R. R. Co I. St. Ry. Co J. St. Ry. Co .... No. 2 to No. gr. wire... None Port Chester. . . . Seneca Falls Utica N. Y. & S. Ry. Co. G. W.,S. F.,&C. L. Tr. Co .... O.B.Co. Wire.... 1-0 ret. feed. U. & Mo. Val. Ry. Chillicothe C. Elec. St. Ry. & Pw. Co Wire .... Protect.. None C. L. & A. Elec. St.Ry. Co Cleveland East O. Tr. Co.... Columbus Dennison Toledo C. L. &S.Ry. Co. 50 2 60 12 47 12 5 70- 90 48 60 60- 70 60- 72 56- 82 60 70- 80 65 80 70 60- 90 56 m. 90 90 w!?r- .... None U. Elec. Co T. B. G. & So. Tr. Co .... Copper.. Crown.,. Crown... Copper.. Plastic 4-0 4-0 4-0 2 4-0 None Ret. feed .... None None . . . L. Elec. Ry. &Lt. Lima. East Liverpool.. Marion W. 0. Ry.Co U. Pw.Co M. St. Ry.Co C. E. & I. St. Ry. Co .... Gr. wire at sw M. &B.. 8-0 4-0 4-0 2 4-0 4-0 44) None None . . Carlisle 0. & Mt. H. Ry. Co . 6 81 19 49 20 92 24 Erie ;.. E. Elec. Motor Co. Media,Middlatown, A. & C. El. Ry.. Copper.. Copper . . Pa... Pa... Pa... Pa... Harrisburg Hazleton Lancaster Lebanon .. 60,000 14,000 41,000 18,000 L. Tr.Co C. Tr. Co L. V. St. Ry... .... Plastic wire... Copper.. 4-0 ret. feed. None None... SEVER: UNDERGROUND CONDUCTORS. 200 Yes Area. 1 4frf) Fav.... No No... No... None . 1 375 None .... No No... Yes.. 5 300 None .... No r iOO None .... No No... 1 10,000 .10(1 None .... No... No No... No... None. 9 4-(0 ^one ...; No No... 1 800500 tf one No... No... None . 4 800400 None .... No No... Yes.. 1 1,200 275 None .... No No No None . _ 900 5BB No No... Yes.. 2 1,000 Some .... No No . No None - 500 ^00 No No. No None 400 VAS... No No... MOI- No No... Yes.. 1 2,000 400 In some Alleged Yes No Yes.. No... None. 2 1,800 390 Some .... No Repaired bonds . . No more trouble No... No... None . 2 500850 Fav.... None.... No No... No... No... None . No... None . 1 1 1,800400 5,500... Noth i n g serious . . None.... No 296 BEYER: UNDERGROUND CONDUCTORS. TABLE I I 3 fc 75 76 77 78 79 80 81 82 83 84 S.5 80 87 88 j 89 90 91 92 93 J4 1 City. Population served. Name of electric street railway company. 1 * >> CO Date of electrical in- stallation. S Weight of rails. Bonding system. I" I | || fc Nature of re- turn feeder system. Pa... Pa... Pa... Pa... Pa... Tenn. Tenn. Tenn. Tenn. Vt.... Vt.... Va... Va... Va... W.Va Wis.. Wis.. Wis.. 1 DL.J Mass. Lewistown McKeesport Philadelphia .. 4,500 81,000 L. & R. Elec. Ry. 6 85 60- 70 70- 90 60- 90 40- 80 f>0- 70 35- 60 (50 15- 75 60 60 50- (50 8-0 4-0 4-0 4-0 8 4-0 1-0 4-0 None P. McK. & C. Ry. Co A.S.&W. Protect. . Many.... 4-0 ret. feed. None Scranton ....... 103,000 S. Ry. Co 77 50 T 16 41 8 5 9 4 Ret. feed.... Ret. feed... Ret. feed.... None Philadelphia . Bristol 5,000 82,500 82,500 14,500 19,000 3,500 10,000 16,500 28,000 80,uOO 17,500 18,500 10,3UO 17,000 B. B L. Ry Co.... Copper.. Chatanooga Chatanooga R. T. Co. of C... C. E. Ry. Co J. & Sub. St. R. R. Co .... Copper, V mile from P. S ... Ret. feed.... Gr. wires. . . . None Turlington Springfield ...... M. P. St. Ry Co... Chicago . Crown... .... S Elec Ry Co... Charlottsville.... C. C. & Sub. Ry.Co. 6 16 20 22 2 11 90 45- 100 60 45- 70 : ?0 54- 94 Protect.. Crown, G. E... Protect. . A.S.&W. Crown, G. E... None.... AS &W. 4-0 4-0 4-0 4-0 None Ret, feed.... None jvnchburer L. Tr. & Lt. Co., R. Ry. & Elec. Co . . Parkersburg .... Eau Claire 7 P. M.&I. Ry.Co.. C. V. Elec. Ry. Co. M. AN. Tr. Co.... M.Ry.&Lt.Co.-j Manitowac MerriU D'ble trol. h None Metallic Boston Elevated Protect, steel plug . . 1 4-0 2 4-0 8 4-0 Ret. feeds on a heaw traffic line. SEVER: UNDERGROUND CONDUCTORS. 297 ( Continued). 3 K 3 a 2 o 0) S3 O 1 6 Maximum current from each. Minimum line voltage. | Nature of soil. Nature of corro- sion. Extent of corro- sion. Any claim against railway company ? What remedy applied ? Effect of remedy. No... No... No... None . 1 2 600 1,200 375 450 No Fav. ... None .... No Yes Rebond ret. feed No more trouble. No more trouble. 1 -8 volt P. D. 4- Pipes No... No... Yes.. No... Yes.. No... Yes.. Yes.. No... None . None. 2 1 1 2 2,100 'i',m 1,100 350 550 4SO 400 450 450 ** Some At two places. None.... Some ... None .... No Yes Taps to pipes bonding Rebond 4-0 C. s No No No No Repaired poor bonds Yes.. No... Yes 400 No On rails. . Nr>n No River plates . . . I m p ro v e d power . No N.... No... None. 450 1,600 500 000 No NYn . . No... No... No... No Yes.. No... + None . 2 2 1 1 1 8 600 350 500 80 8,800 450 450 350 500 550 400 'TJVniA . . No None Yes No Doubled bond . No more claims. No Pipes He... No... No... None. None. Far.... Graphitic None .... Some ... No Yes Heavier return Reduced danger 1 areas. 298 SEVER: UNDERGROUND CONDUCTORS. TABLE I- J a 3 i 1 || 03 || City. a o Name of electric street railway company. 1 o Bonding system. Is, Nature of re turn feeder system . h *ri q_l +2 1 OJ t->d J s o ^ O ^ 5 1 Pi I 1 1 3 I" 5 02 CH 02 M K* fc 95 TVTnca Middleboro 6 900 Mid W # R R St. Ry Co BO 70 2 4-0 500,000 C. M. ret % N J. Pub . Serv Corp of 70 107 Cast pro- tect . -. . 4-0 Ret. feed.... 97N Y Buffalo International Ry Co 98 Ohio 125 500 C Ry & Lt Co 1% ^n M C. B. L. & N. Tr. 107 Copper . . .... Ret. feed.... Co 70 4-0 None 100 S. C. Columbia 21,000 C. El. St. Ry. Lt. & Pw Co M 48 101 Term. Knoxville 82,700 K. Tr. Co 24 80 40- Roebling 2-0 3 ret. feed... 100 Copper.. 4-0 102 Wis.. Madison M. Tr. Co no- 1-0 None SEVER: UNDERGROUND CONDUCTORS. 290 {Concluded). 3 5 i 1 49 a & o +J r 03 C8 |C 1 irea is drai power stat um c u r r rom each. 1 g a Nature of soiL Nature of corro- sion. Extent of corro- sion. Any claim against railway company ? What remedy applied ? Effect of remedy. a m g a 1 2 i 1 i|i .5 & No... None. 2 500 400 Gravel, No No... No... None. 6 None .... No 9 8 000 100 Noue .... Some None No... Yes.. 4 1,500 150 None .... No No... No... None . 1 1,400 400 None.... Yes None No... No... None. 2 1,500 400 Fav.... Due to soil.... Some Yes Improved bonds. . Cons idera- ble im- provem't. Pipes Yes.. 1 800 450 Some.... Yes Tap pipes to Station NA tnnre trouble. 300 SEVER: UNDERGROUND CONDUCTORS. TAB SUMMARY OF MUNICIPAL Number. \ 00 1 I S | Was electrolysis alleged? 2 L i! 5 $ Was electric railway company blamed? 1 2 5 e 7 8 9 10 11 5! 14 !S 17 18 19 20 21 22 23 24 25 26 27 28 29 Conn . Conn. Ill .... Ind.... M5..'!.' Mass.. Mass .. Mich . Minn. . Minn. . Minn.. Mo.... Mo.... N. J.. N.Y.. N. Y.. N. Y.. Ohio.. Ohio.. Ohio. . Ohio. . Pa.... Pa.... R.I .. B.I .. Wis .. Wis .. v.... Hartford .... Middletown.. 1901 1902 18% Council Water Commas Yes.. Yes.. Water . Water . Water . Yes.. Yes.. Yes.. Vfts [ndianapolis. Newport Baltimore 1901 1902 1901 Yes City Water Works Com. . . . Yes. . Water . i Yes. . Ch Eng Elec Com Snm VAC Chelsea Worcester 1902 1901 City . . . Water Comm'rs City Engineer Yes.. Water. Yes.. No Detroit 1896 1901 1901 1903 1894 1908 1901 Bd. Water Comm'rs. City Engineer Yes.. Little Yes.. Yes.. Yes.. Yes.. No .. Water . Yes. . Gas i Minneapolis . St. Paul. St. Paul St. Joseph... St. Louis.... Newark .... City Engineer Water Comrn'r City D. B. Maury, G. H. Benzenburg & Co Water Comm'rs . Water. Water Water, gas .. Yes.. Yes.. Yes.. Water Department. City Electrician E. E. Brownell Engineer Water Dept. Supt. Bur. of Water.. Chief Engineer Albany 1901 1 1901 ' 1898. . No .. Water. Yes.. Rochester... Cincinnati . . . Cleveland . . . 1901 Com. Public Works.. 1899 Am. Soc. Mun. Irn.. . 1899 E A Fisher Yes Water. Yes.. Com Waterworks... No .. Supt. Water Works. . . F C Caldwell Yes.. Water . Yes.. PiiVilin Wnrks Dayton Philadelphia. Reading Pawtucket.. Providence . Madison 1899 1899 1900 1900 1H9fl Ch Elec Bureau . . No A. A. Knudson A. A. Knudson A. A. Knudson City Water Works .... Q. H. Benzenburg A. Schoen Yes.. Yes.. Yes.. Yes.. i Yes.. Yes.. Water . Water . Water . Water . Water . Water . Yes. . Yes. Yes. Yes. Yes! City Engineers Com. Public Works. Racine Richmond... 1899 Supt. Water Works. SEVER: UNDERGROUND CONDUCTORS. .301 LE II. REPORTS ON ELECTROLYSIS. $ f! *d**- s-g a Was legal action In- 1 stigated? Plaintiff. Better return No more trouble Double trolley Yes.. Peoria Water Co. Reduced P D 's . Reduced P D 's Yes Efficient return, Connected track and pipes Yes.. 302 SEVER: UNDERGROUND CONDUCTORS. TAB SUMMARY OP MUNICIPAL ORDINANCES 1 fc j City. ' Immediate cause for or- dinance. System required. Are taps from pipes I to rails allowed? Is drainage in -j- area allowed ? ll L| Is railway com- pany made lia- ble for corro- ifl Fj,9 HR Allowable leakage of current. ii Remarks. sion? => e8 g p,T3 ti ^9 O *D S* S ft D 9* eS ft 0^ S a Q Q not less than + feeders. 1 1 7<&-300' 8.8 AH damage to water and gas V( No mor6 than ono 74 amp in any pips i^_200' 8 8 Yes '4 Jl^ VA 200' A A 74 25 Electric code of city Yes. No leakage department. Periodic tests and ex- cavations. i/ ^ ^6-200' 8 I Ye&. ... 304 SEVER: UVDEKGKOUXn CONDUCTORS. w - 1 as H 4 A'lpA'qpenAvo sadid aa^uM. aoy i Now O D CD Q) & fl P4 fl fl fl C gggig ^ fe IISS &ogWS hhfc-| "r H E? fefelb b^g^b bb^Ufe S.2 fe oa Illl islsl" sss^^ 06 oo ^^JQ, OJ3 .-S.-S3 ford 1 1 1 II si ii ai II 5SSS ^I3 &8&Z 3 5^5 5IS 33 3Sls 3S SSS i-oooJO^H o*ecTjooo SEVER: UNDERGROUND CONDUCTORS. 305 s-g : ^JH * K2 5.9 .2 oo S 33 3 ET-EC. RYS. 20. 306 SEVER: UNDERGROUND CONDUCTORS. TAB SUMMARY OF EXPERT OPINION Number. Name of expert. Title of article. Where published. I tssi i| 1 Most suitable remedy. 1 8 i I 11 12 13 14 15 16 17 IS 19 20 21 22 23 s A. V.Abbott... Bavlis. . . Electrol. from Ry. Cur. . Electrolysis 1899 Yfts. . . Canadian Elec. Assn. 1894 No ....IT-,.... Only cure is double trolley.. Wm. Brophy.... H. P. Brown .... Ellicott Prevention of Electrol.. Remedy for Electrol. . . . IfW Neg. booster Report in Chicago . . T H F'arnhftTn Cassiers 1 Mag . . 1895 18 |i II Requirements as to return feeder system. Are taps from rails to pipes recommended? !i a o c II a ll Remarks. --2 OS^' P 1} Ii ^ Good Good . Yes.. Good . . pipe joints. Yes... Good . . Good . . No Yes.. No... . Yes.. Suggests balanced feeder system. Good . . Good . . -}- Area Along entire tracks Yes Advises good bonding for comp. met. cir. Good Good . Yes Cood .. Good . . Good Yes.. Yes Yes.. With precaution Almost no trouble in St. Louis. Good . . Good . . No Yes.. Good . . Good . Yes Yes Yes. . Yes . Good . . Good.. Good Yes Good.. Good.. Good Yes.. 30cS SEVER: UNDERGROUND CONDUCTORS. DISCUSSION. Mr. JOHN HESKETH: Being in the position of having had experience on both sides of the problem I have had reason to give the question very close study. There are certain well-defined lines and conclusions from which I think we cannot escape. To begin with, the onus of protecting underground works from electrolysis or from damage by tramway systems cannot possibly be considered as resting on one or the other party ex- clusively. It must, if it is to be a successful work, be a mutual one. It is impossible for the telephone company, even by the adoption of all known reasonable methods, to protect their works if the tramway company, on their part, neglect well-known methods. Further, it is impossible for the tramway company to so run their system as to avoid damage, if the telephone company or others interested are laying their works in an unnecessarily dangerous manner. As an instance: In one case which I have in mind, a water company laid its lead service pipes within six inches of the rails of a tramway system. They invited electrolysis; they got it; and then they complained. Further, there are conditions which are easily imaginable, where a system of water pipes acts as a feeder from the zone in which danger is existent to a zone which otherwise would not be dangerous. In such cases the water-supply company, or the gas company, should so insulate its pipes as to prevent the feeding of danger from the one zone into the other. Further, it has been the effort in one or two places to prevent damage by laying down hard and fast rules as to the drop in the return circuit. For instance, the Board of Trade of London laid down an arbitrary figure of seven volts as the maximum difference of potential between the ends of the return. But any figure of drop in the return must take into consideration the length of the line. It is not necessarily the drop along the return that does the damage. It is rather the difference of potential between the return and the other metal bodies in the neighborhood; and yet not altogether so. It is not the difference of potential only, but the capacity for current carrying from the return into the pipe. There may be a huge difference of potential and yet no passage of current into the pipe. There may be a very small difference of potential, and yet a very dangerous current. There we strike another main principle the method of testing for pos- sible danger, which ought to be clearly defined. It is not sufficient to measure the difference of potential between the pipe and the return. I rather incline to the belief that the method which has during the past year been suggested in Germany, of measuring the difference of potential between the rail and the earth nearest to the rail, is a more correct method. It takes into account the electrolyte between the two bodies. Recently the Australian Government met in conference the engineers of the telegraph department and engineers representing electric supply indus- tries. In conference, we agreed on certain regulations for the protection of the works of the Postmaster-General of Australia, and the points just mentioned were the salient points brought out in the discussion on the question of electrolysis. When I heard that this Congress was to be held, it appeared to me as rather desirable that an effort be made to SEVER: UNDERGROUND CONDUCTORS. 30C have an expression of opinion from the technical associations of different nations on this most important subjejct, and I mention that now for your consideration, if deemed advisable. It is rather a problem as to how such an expression of opinion could be obtained, but it seems to me, in view of the diversity of regulations throughout the world and the lack of authoritative statements based on a scientific principle, that such a statement prepared by scientific bodies would be invaluable to both sides. Prof. F. C. CALDWELL: In Columbus, Ohio, as has been mentioned by Prof. Sever, we have made some investigation of this matter, and our conditions there are particularly favorable for absence from the trouble. I believe the soil there is not such as to produce much electrolysis, and the lay of the railway system is particularly favorable for freedom from it. It seems to me there are two points upon which definite information is needed in connection with this matter of electrolysis. The first is whether we should look for trouble only where the current leaves to go to other metallic structures, or whether we are to look also to the joints of the pipes. There is much difference of opinion upon this question. It has been claimed that trouble has been found at the joints, but on the other hand we find engineers taking very decidedly the stand that ,all that is necessary is to keep the current from leaving the pipes and going to other conducting material. Information on this subject would certainly be very valuable. The second point is as to how much current can be allowed in the pipes or to leave the pipes. This is especially im- portant if it is true that we are to look for trouble at the joints. If we must keep the current out of the pipes practically altogether, then it becomes an important matter to know how much current can be allowed to flow and still not add an appreciable amount to their disintegration. There has been a little data along this line published in regard to the resistance of pipes. What is needed is data as to the resistance in the case of pipes laid in dry sandy soil. Where a pipe is laid through a street, if we make an attempt to measure its resistance we shall get the joint resistance of the pipe, the surrounding soil, and other conducting material, so that we cannot be sure that the resistance we get would show the current going through the pipe. The other question as to how much damage is to be expected from the current when it leaves the pipe, I believe, depends very much upon the surrounding soil. In some cities much more damage may be anticipated, with the same current flowing, than in others. We have been carrying on, at the Ohio State University, some investigation along this line, ob- taining earth from different cities and using an electrode which was weighed before and after the test. Our results so far have not been sufficient to warrant any conclusions, but they are interesting. We have found in two different tests a considerable difference in the amount of material in different cities. Soil from Dayton, Ohio, where there has been much trouble, gave a large amount of electrolysis, while that from Columbus gave a very small amount. It looks as if this was an important point to be considered. Mr. H. E. HARRISON : It does not matter practically how much current or what current density flows into a pipe. It has been assumed that the \ 310 SEVER: UNDERGROUND CONDUCTORS. current flowing into the pipe would come out more or less uniformly through the whole service; but I do not believe this is so. The pipe may pass through a considerable length of soil which will be a very fair in- sulator, and will then come upon a patch of soil that is conductive to a high degree, with the result that the current density is more visible and the damage greater. Prof. SEVER: The data which we have collected contains many refer- ences to underground conductors other than piping systems, so that I think it is perfectly proper that that phase of the situation should be brought before this meeting. About two years ago, when I became con- nected with the city government of New York, Mr. Jones brought me a cable sheath which he claimed had been destroyed by electrolysis. I know that on some cable sheaths in New York city, both on the telephone and the power circuits, there are large currents coming presumably from the operation of the electric railways. In the Bronx there has been con- siderable difficulty. In the borough of Manhattan there has been diffi- culty which to some extent has been remedied by the co-operation of the officials of the railway company and the telephone companies. I know of one instance where the sheaths were bonded at one point by a heavy copper conductor to a return of the Manhattan " L," approximately 1500 amperes passed over that wire sufficient to heat it so one could not put his hand upon it. The Manhattan elevated road uses its structure com- pletely bonded, its service rails completely bonded, and a large amount of return feeder, something like six or seven million circular mils, to get their current back without causing trouble to their own and other con- ductors. In spite of all their precautions, there are still thousands oi amperes coining back on their cable sheaths as well as those of other companies. It has been drawn to the attention of the city officials for their recommendation, as there are at times a higher potential than twenty-five volts between the end of the line and the nearest sub-station, which is the maximum fixed by the city rules. Mr. P. B. DELANY: It seems to me there is one phase of this subject which has been overlooked, and that is the shunting of water and gas pipes or the cable sheath, by the grounding of telegraph wires in the city. This may, to a certain degree, account for the apparent discrepancies electrolytically, in different cities and towns and through different soils. We all know that there is in many places a very great leakage what we call stray or vagrant currents into the telegraph circuits by way of the ground return. I myself have had experience with wires about a hundred miles in length, and it was rather a disagreeable experience. I tried some synchronous experiments four years ago, and I found there was a voltage varying from three or four volts to seventy-five in that circuit not constantly, but running up and down. If it had been con- stant, we might have been able to do something with it, but as it was fluctuating, it was rather disastrous to the experiments at the time. It has occurred to me that in cities where there are hundreds of ground connections made at different points to the pipes and where considerable electric energy is used in the operation of telegraph lines grounded in cities, some of the electrolysis may be even due to that source, as well as SEVER: UNDERGROUND CONDUCTORS. 311 the protection of the pipes from power leakage by the shunting. I think this suggestion may throw some light on the subject, although I presume that this phase of the case has been taken into consideration by Mr. Sever and his associates. It has not been referred to in the discussion. Mr. BANCROFT GHEBABDI: One of the functions of my department is taking precautions against electrolysis trouble on our cables, and in that connection the bulk of our work has been in Brooklyn, on account of our very large underground plant there and the great extent of the overhead trolley system. It is not unusual for us to have to take care of currents as great as 200 or 300 amperes at a single point on our system. This shows that the aggregate amount of current that our system is carrying back to the power-houses amounts to thousands of amperes. There is a certain expense in connection with this work which is quite appreciable, and there still remains, after everything is done that we can do, a certain amount of trouble which is real trouble. The discussion of the respon- sibility for such trouble and expense is one that it seems to me is beyond the scope of this section and I shall not touch on it here. Prof. SEVER: In connection with the situation on the Virginia Pas- senger & Power Company, at Richmond, Va., Mr. Stillwell went at the matter in an engineering way by laying out very carefully on paper the whole railroad system, placing the cars in accordance with their various schedules, and ascertaining those points to which he could most profitably connect a return conductor. He decided upon four points about the city, almost at the corners of a rectangle, and carried directly back to the power station very heavy return feeders, as well as heavily bonding the tracks. From the results which they are getting, it would seem that that is a very satisfactory way in that particular locality to solve the problem. Chemical analyses were made also of the soils. I learned from him a short time ago that the city, through its engineering staff, ap- proved of this scheme and accepted the efforts on the part of the railroad company as an expression of a desire to reduce the trouble. As stated in one of the tables which is presented, the city of Richmond insists that the railroad company must pay for all damage to pipes. How two men are going to agree as to whether damage is due to electrolysis or to ordinary tubercular action or rust, I do not know, and I do not know any- body who does know definitely. The other day we took up in Brooklyn cast-iron water pipes, which had been down fifty-two years, so filled with tubercular nodules that the area of the pipe was reduced to about one-half of its original area. In other places we took up lead pipe, part of which had entirely disappeared, undoubtedly through electrolytic action. Mr. J. SIGFRID EDSTR'OM: We have had very little trouble in Europe from electrolysis. There has been some, however, in the earliest railroads built in England, but lately we have experienced hardly any trouble. I think this is owing to the very solid construction in bonding and in cables carrying return current to the central station. In Berlin the city officials require that there shall be no larger voltage between any two points of the rails in the city system than two volts that is, between any points in the rail system of the tramway there must be no greater pressure than two volts. This, or a similar stipulation, has been adopted by many other 312 SEVER: UNDERGROUND CONDUCTORS. cities, including cities in Switzerland and Sweden, where I have had the pleasure to be a railway engineer. In these places we bond the rails with two heavy copper wires at each joint. We have double track generally, and consequently we have eight copper wires at each double pair of joints of the rails. We bond the rails between each other and also the tracks at certain distances. At crossing of bridges or water pipes, where the rails get close to iron in the earth, we insulate the rail with asphalt as much as possible. The rails themselves in the street are generally in- sulated through a layer of stones or concrete put under the rails. To take the current from the rail, we put it in an insulated cable of very heavy dimensions. It is very important to have the cable insulated", as a bare copper cable, which I know is often used and which generally is buried deep into the street, invites the current to seek other ways home. The general practice in Europe is that, where a feeding cable is connected to a certain part of the overhead wires, a return insulated cable of the same dimensions as the feeding cable is used. This has also the advantage that in case the positive cable becomes damaged, we can easily exchange it for the return cable until the positive cable has been repaired. All the nega- tive returns are carried into the station through resistances, and these are regulated so that the actual current for which the cable is assigned arrives there; thus the current is split up and no cable is overloaded. In this way every feeding point becomes a " central station." These central sta- tions are planted around in the city, and we have no long flows of current running through the city. Street railways built ten years ago in this waj T have given no trouble whatever. As to disturbances on telephones in cities, where the telephones use the earth as a return, there has been some slight disturbance, as naturally a portion of the street-car current must go through the earth and thus some of it also through the telephone wires. In cities where we have a double-wire telephone system, there is no trouble whatsoever. Mr. HESKETH : Although, as you stated, the regulations define the drop in voltage, I should like to ask what in actual practice is found to be the approximation to the regulation? How closely in Berlin do they comply with the regulations? It would be interesting to know, for the purposes of comparison simply, some of the leading dimensions of the system on which the regulations mentioned are found practicable the mileage of track and the number of amperes output from the station per mile of track. Mr. EosTRbM : I am here not loaded with figures, but I will try to give part of the information. When the plant is laid out, it is laid out accord- ing to a certain schedule, and consequently you know the loads on the several points of the city. According to this the dimensions of the cables are figured out. The track itself has the ordinary two heavy copper wires at each rail and four rails at the side of each other are considered to be sufficient for the two volts drop that should be the maximum in the city. Actual tests have not been taken, so far as I know. I have myself been opposed to the two volt requirement, as I consider this limit very low, and I do not think that on any day of heavy traffic for instance, Sun- days or Easterdays or Whitsundays that the two volts will be the limit. but that you will actually find the drop far larger. SEVER: UNDERGROUND CONDUCTORS. 313 CHAIBMAN JONES: I perhaps might give you a few salient facts of the effect of electrolysis upon the Postal Telegraph. The Postal Telegraph Cable Company would be only too glad to submit any of the data it has upon the subject of destruction of their cables by electrolysis to Prof Sever for the purposes of his paper. I think they would do this in the interest of electrical engineers everywhere, and in the interest of munic- ipalities whose pipes are being eaten up, and also our good neighbors, the telephone people, who are in the same boat with us in that respect. I can only, of course, as intimated, speak in a general way on the sub- ject. The telegraph companies were urged to place their wires under- ground, commencing about the year 1880. Cities got tired of the crow's nests and networks of wires which were in their streets. Some of them were curiosities. Commencing with New York, Philadelphia, and other cities, the agitation became so great that eventually they started to put in their wires underground. As a rule, the cables of the telegraph com- pany are not to be compared with the network of water pipes and gas pipes of cities, nor, except in a few cases, the rails of the tramways. The telegraph companies coming into a city and passing through generally follow a line of pipes, and lately the line of rails of the electric railroads, and we have had a great deal of trouble from electrolysis, in times past, in various cities, commencing with Boston, Hartford, Baltimore, Chicago, Atlanta, New Orleans, and other places. In almost all those places, our cables, that had been laid parallel with or near to the electric railroads, have been attacked, and sections have been eaten up, and our service stopped. We were helpless in the matter, because the cities in some cases had ordered us underground, and after having gone underground, our poles were taken down, and it was not possible to place the poles up again and put the wires on them very oxpeditiously ; so we had to suffer and so the public had to suffer. Its telegrams could not be forwarded until we had made the repairs. We found out, however, that by applying the now universal remedy of bonding, where we could secure a good return wire from the point at which the currents were leaving our cable sheaths to get back to the negative brush of the generating station of the railroad companies, we were rendered entirely immune. We have not had any trouble since we have been properly bonded in any city. Quite recently, in New Orleans, our cable was attacked at one point; but we have since bonded and I think there will be no further trouble. In Hartford we have for some years been bonded, and no trouble has arisen there. In all -other places where we have been properly bonded there has been no trouble. It of course follows that the currents which are carried through the trolley pole and down into the motor of the car and so into the rails, is seeking its way back to the generating station, and if the resistance is very high between the point where the car is resting upon the tracks and the negative brush of the machine at the station, it is going to seek a great many ways to get back. It will go all around and follow every route that is possible. As a matter of fact, we loan the sheaths of our cables to the railroad companies to allow them to get their current back to the station, and we bond our sheath to their return wire so they can have every use of it and get back the easiest way possible. We do that to prevent getting hurt. 314 SEVER: UNDERGROUND CONDUCTORS. It is not where their current comes on and starts in to go back that we suffer, but it is where the current leaves our sheath to go through moist ground or some electrolyte to reach the metallic conductor at the power station; so that we have found it was necessary for us to make that path just as good as possible. Our sheaths are one-eighth of an inch lead with 10 per cent of tin, and we have not yet had a case where the carrying capacity has been exceeded by the amount of current that our friends, the railroad people, want to have us carry back for them. It is lying there, doing us no good at all, and we feel no effect from any induction in that respect, and we are glad enough not to be eaten up in the under- taking. It is pretty difficult to tell whether there is any serious electrolysis generated by telegraph currents or not. Of course, the companies are using much more current now than ever before, on account of their increased business, but prior to the time of electric lights and trolley systems, I have never yet heard of any electrolysis arising from telegraph currents, and do not think they are of sufficient quantity to figure in the case at all. There is another question, in regard to alternating currents being used for transportation or trolley purposes. How are we going to be effected when alternating currents are used? That is an open question which I am not prepared to discuss, but I would like to call it to your attention. BRAKING HIGH-SPEED TRAINS. BY R. A. PARKE. During a hearing upon an application for a charter for the New York & Port Chester Eailroad Company,, before the New York Eailroad. (Commission, some three and one-half years ago, the writer was called upon for expert testimony concerning the dis- tance in which electric trains might be stopped, in regular ser- vice, from a speed of about 60 miles an hour. It was then that expression was first publicly given to the opinion that the special conditions under .which the brakes are applied upon trains of such high speeds warrant a force and promptness of application which could be employed at low speeds only with serious shock and dan- ger of train rupture. For years prior to that time, the uniform teaching and recommendation of The Westinghouse Air Brake Company, as given in instruction cars and by authorized repre- sentatives of the company, had been emphatically opposed to the use of what is commonly known as the "emergency application" of the quick-action air brake in ordinary train service, and. a propo- sition which contemplated the use of the emergency application of the quick-action air brake, and particularly the more powerful high-speed form of air brake, appeared to the average railroad officer as nothing short of heresy. Members of the Railroad Com- mission promptly instituted a line of questioning which made it quite evident that they were similarly impressed. One of the fundamental grounds upon which the New York & Port Chester Railroad sought to justify the granting of a char- ter for a railroad line paralleling a steam railroad already in operation, was the materially improved local express-train ser- vice which it was proposed to attain through the superior rate of acceleration acquired upon electric trains by the use of the system of multiple control of motors operating uppn the axles throughout the train, and the higher rate of retardation to be obtained through the high efficiency of the emergency application of the air brake, in bringing such trains to a station stop. 1315] 316 PARKE: BRAKING HIGH-SPEED TRAINS. In presenting the matter to the Eailroad Commission,, elabor- ately arranged curves, indicating the rates of acceleration and retardation, had been prepared by the able engineer, Mr. C. 0. Mailloux, which appeared to justify the claims for the improved character of train services contemplated by the company. Al- though multiple-control systems of electrical train control were, at the time, in successful operation, it did not appear that the rate of retardation indicated by the stopping curves had been at- tained in regular service, and the high efficiency of the proposed train service was characterized as impracticable and chimerical by those opposed to the granting of the charter. It is a notable fact that, while the effort to attain high accelera- tion in bringing electric trains to the required speed had in- volved costly extension of the application of motors to a number of cars throughout the train being applied, in some cases, to the trucks of all the cars practically no effort had previously been made to realize a higher rate of retardation, in regular ser- vice, than that which had been regularly employed in steam-rail- road service, through the service application of the ordinary auto- matic air brake. Without attempting any discussion of the merits of multiple-control systems of electric locomotion, or the commer- cial limitation of the expense justified in extending the appli- cation of motors to a number of cars throughout the train, it may properly be suggested that commercial economy may not result from an indefinite extension of such systems. The inadequacy of a single motor car for the acceleration of a train of several cars easily justifies the application of motors to the trucks of one or more additional cars, depending upon the number of cars in the train; but it may also be readily understood that a point may be reached, in the increased acceleration due to multiplication of motors, beyond which the addition of other motors is accompanied by too small a measure of increased acceleration to justify the added expense of installation and maintenance. An illustration may be found in the operation of city water- works' systems. It is a well-understood fact that refinement of pumping machinery is justified up to the attainment of a practical duty of somewhere in the neighborhood of 90,000,000 gallons; be- yond this, further refinement, whereby an increased duty is accom- plished, is attended by an increased cost of plant and of repairs and necessitates a higher grade of skilled oversight and attendance which more than compensates for the fuel economy acquired. It PARKE: BRAKING HIGH-SPEED TRAINS. 317 is not fuel economy, but it is commercial economy of operation, which defines the limit of such refinement and establishes a duty which may not be exceeded with commercial advantage. Similarly, the extension of multiple-control systems to the application of motors to more than a certain limited proportion of the axles upon a train may easily be attained by an ultimate cost whereby economy of operation is impaired. The pertinence of the foregoing observation lies merely in the fact that, while the application of multiple-control systems has, in some cases, apparently been pushed to extremes, in an effort to improve electrical train service, by attaining the very highest ac- celeration at the sacrifice of commercial economy, the absence of any effort to attain a fuller measure of the possible rate of retarda- tion, in stopping, is the more noteworthy. That materially in- creased stopping efficiency may with propriety be employed in high-speed train service, it is the purpose of this paper to demon- strate; and, as every start, requiring high acceleration, is neces- sarily attended by a corresponding stop, in which a higher rate of retardation correspondingly improves the character of the train service, it is obvious that, if increased expense is justified in mod- erately increasing the rate of acceleration, materially increased stopping efficiency, at a comparatively small cost, is entitled to careful consideration. To those who are familiar with the results of experiments with the friction of brake-shoes upon car-wheels and the difference in the conditions of brake application at high speed from those at low speed, the proposal to increase the force of application of the brake- shoes upon the wheels at high speeds will excite no comment. The various trustworthy experiments upon brako-shoe friction have uniformly demonstrated a declining ratio of the friction to the pressure of the shoe upon the wheel at increased speeds. For the same brake-shoe pressure the friction excited at a speed of 60 miles an hour is but about one-half that which occurs when the speed is but 20 miles an hour. Other causes result in a reduction of the brake-shoe friction during continued application of the brakes ; and this result combines with the increase of the friction through reduction of speed, during the retardation of the train, to main- tain a comparatively uniform, though slightly increasing, rate of friction throughout the stop, until quite near its close. Thus, the average rate of retardation of the brakes, when applied to the wheels at a speed of 60 miles an hour, is about one-half of that acquired 318 PARKE: BRAKING HIGH-SPEED TRAINS. with the same brake-shoe pressure when the initial speed is but 20 miles an hour. It is evident, therefore, that the same rate of re- tardation which may with entire propriety be employed at all speeds can only be acquired by increased pressure of the brake- shoes upon the wheels, to correspond with the reduced rate of friction occurring at the higher speeds. Moreover, an application of the brakes which will produce a given rate of retardation at one speed, without danger to the rolling- stock or discomfort to the passengers, may also be applied at any other speed with no more danger or discomfort. The high-speed brake was designed more particularly for use upon high-speed trains, and it employs a considerably greater brake-shoe pressure in emergency applications than that of the ordinary quick-action brake, to more nearly realize the rate of retardation obtained in the emergency application of the quick-action brake upon trains of lower speeds. At such a high speed as 60 miles an hour, however, even the emergency application does not develop greater brake-shoe friction than does a full service application of the quick-action brake at a speed of 20 miles an hour. It is true that the service ap- plication is attended by a comparatively gradual application of the brake-shoe pressure, while the emergency application develops the greater brake-shoe pressure very quickly; but experience and ob- servation seemed fully to justify the conclusion that the reduced rate of friction at the higher speeds would permit the use of even the high-speed brake without noticeable shock or disagreeable sensation. Though the conviction thus expressed three and one-half year? ago was based upon observation, experience and knowledge of the results of experiments upon brake-shoe friction, it was, neverthe- less, so far as practical employment in train service was concerned, a theoretical conclusion. Since that time experiments in the use of the high-speed brake upon passenger trains have amply confirmed the writer's views upon this subject and demonstrated the absence of disagreeable effect as well as the highly increased rate of re- tardation in employing the emergency application of the high-speed brake for stops in high-speed train service. The time and distance saved in such stops permit the employment of the maximum speed up to a comparatively short distance from the stopping point and cause the train to be brought to a quick, smooth stop in much less than half the time and distance required for an ordinary service stop. PARKS: BRAKING HIGH-SPEED TRAINS. 319 That the shortened running time and increased efficiency of high- speed train service particularly local express-train service by the employment of such higher rate of retardation, may be attained at a small fraction of the expense at which a lesser improvement in such efficiency can be obtained through the increased acceleration resulting from extending the multiple-control system from the use of motors upon one-half the cars in the train to their application to all of them, seems hardly open to doubt. The neglect to take ad- vantage of this higher rate of retardation would seem to be attrib- utable chiefly to the long-established doctrine that emergency ap- plications must not be employed for service stops, under far dif- ferent conditions. It is to be understood that such a doctrine still applies, with all its force, to the operation of passenger trains at moderate speed, as well as to freight-train service. It is only under the special conditions of uniform operation at high speeds not less than 50 miles an hour that the recommendation of a most powerful application of a most powerful brake, in all stops, properly applies. In addition to the advantage of effecting a reduction of from 50 to 75 per cent in the time and distance required by a service ap- plication of the brakes, a collateral advantage of material im- portance is the much greater accuracy of the stop. In a stop by a service application of the brakes, the application is affected by the personal judgment of the operator, whereby an element of uncer- tainty is introduced which almost invariably requires a subsequent release and a second application of the brakes, in order to bring the train to a stop within the range of the station platform. This frequently involves more or less " drifting " of the train, at greatly reduced speed, to avoid stopping short of the station and not in- frequently involves backing of the train because of inaccurate judgment in the application of the brakes, whereby the train runs beyond the stopping point. In the use of the emergency application, not only is the individual application of every brake very much more prompt and powerful, but the rate of serial application from car to ear is almost instantaneous and is automatically established to the exclusion of any influence of the operator's judgment. Grade and alignment of the roadway, of course, influence the stopping dis- tance ; but such influences are readily determined for each stopping point, and the point at which the motive power should be shut off and the emergency application of the brakes should occur, may be 320 PARKE: BRAKING HIGH-SPEED TRAINS. designated by a post or other permanent signal, whereby the train will be brought to a stop at the desired stopping point. In comparison with the rate of acceleration, in starting steam- railroad trains, the rate of retardation in ordinary service stops has been so high that it is not unnatural that increased efficiency of train service has suggested higher rates of acceleration in starting, rather than improved retardation in stopping; but it should now be clear that a really efficient high-speed train service may be ob- tained only by also employing the maximum practical rate of re- tardation, by which so large a reduction of the time and distance of stopping is counted. Electric train service furnishes exceptional conditions for attaining the maximum retardation, as well as the maximum rate of acceleration though for different reasons. Where trains are drawn by steam locomotives the conditions exist- ing at the locomotive and the variable load carried in the tender involve limiting the braking power so that the retarding force is considerably inferior to that realized upon the cars. Where elec- tricity is employed the motive power is applied directly to the cars themselves in such a manner that the maximum braking efficiency may be obtained as well upon motor as upon other cars, and the whole train is thus subject to the maximum rate of retardation. While the special conditions of high-speed train service permit realization of the maximum obtainable retardation in ordinary sta- tion stops, it will be understood, of course, that all the ordinary- means of general brake efficiency are contemplated in connection with the brake apparatus. In a paper presented to the American In- stitute of Electrical Engineers, and published in the January, 1903, volume of proceedings, the writer pointed out the more important features of the brake apparatus for attaining such high efficiency. They included efficient foundation brake-gear automatic slack ad- juster, to maintain the minimum piston stroke in the brake cylin- ders, and brake beams hung between the wheels and adapted to regulate the brake-shoe pressure so as to compensate for the transfer of weight from the rear to the forward pair of wheels of each truck during the application of the brakes. In addition to such general considerations an exceedingly im- portant element of braking efficiency is the character of the brake- shoes applied to the wheels. Extensive experiments have demon- strated a very wide variation in the frictional quality of brake-shoes of different materials, and, further, a marked difference in the friction of the same brake-shoe upon wheels of different materials. PARKS: BRAKING HIGH-SPEED TRAINS. 321 It is, in general, found that the maximum frictional resistance occurs in the application of soft cast-iron shoes to chilled cast-iron wheels, and the friction-producing quality generally declines a8 harder brake-shoe materials are employed. It should not be con- cluded, however, from this general relation of the hardness of the brake-shoe materials to the frictional quality, that soft material only should be employed in brake-shoes. Beside the cost of soft brake-shoes, which wear rapidly, the trouble and expense of replace- ment, together with the complications arising from rapid wear, are highly persuasive elements in favor of the use of harder materials. If the inferior frictional quality of the harder brake-shoes is com- pensated by correspondingly increased pressure of the brake-shoes upon the wheels, the operative objection to the hard brake-shoe practically disappears. The question is, to a large extent, a com- mercial one. Increased pressure upon the harder shoes involves, of course, somewhat increased wear ; but when, in each case, the brake- shoe pressure is so adapted to its frictional quality that the maxi- mum retarding friction is acquired, the practical question resolves itself into the relative cost of initial installation and of subsequent maintenance to which must be added due consideration of trouble and annoyance arising from the necessity of frequent attention. Within the past two or three years, two different series of ex- periments with the high-speed brake have furnished most inter- esting and important information bearing upon this subject. In one series, soft cast-iron brake-shoes were employed with chilled cast-iron wheels. In the other the " Diamond S " form of brake- shoe (of hard cast-iron, with steel inserts) was used with steel-tired wheels. Otherwise, the conditions were fairly comparable, the tests being conducted in the same general locality. In the case where soft cast-iron shoes were employed, the initial air-pressure in the brake-cylinder was about 85 J Ibs., which became reduced, toward the end of the stop, to 60 Ibs. In the tests with the Dia- mond S brake-shoe, the initial air-pressure on the brake-cylinder was also about 85J Ibs., which, by the use of special high-speed reducing valve?, became reduced to a final minimum of from about 69 Ibs., from a speed of 80 miles an hour, to about 78 Ibs., in stopping a six-oar train from a speed of 50 miles an hour. Moreover, in some instances, a brake-cylinder pressure of 75 Ibs. or more occurred in applications of the brakes at speeds of SO miles per hour (and even less), without producing wheel-sliding of an injurious character or exceeding that which occurred with ELEC. RYS. 21. 322 PARKE: BRAKING HIGH-SPEED TRAINS. the use of the soft cast-iron brake-shoe, when the final minimum air-pressure in the brake-cylinder was but 60 Ibs. The stopping distances were phenomenally short in the tests with the Dia- mond S brake-shoe, averaging 602 ft. from a speed of 50 miles an hour, 982 ft. at 60 miles an hour, and 1334 ft. at 70 miles an hour the shortest authentic stops on record. It is true that these tests were made in dry weather, and that the rails were more or less affected by sand, in which the soil of the country abounded, and particles of which were carried about by wind. It is very doubtful whether such high terminal brake- cylinder pressure might be safely employed even with such hard brake-shoes, under the varying rail conditions of regular service the corresponding total brake-slioe pressures, as customarily cal- culated, being from 104 to 117 per cent of the weight of the braked cars; but these experiments clearly illustrate both the fact that wide difference in the friction al qualities of brake-shoes should be given proper consideration in determining the brake-shoe pressure, and also the fact that, with a properly determined pressure, cer- tain forms of hard brake-shoes may yield as good, or perhaps even better, average retarding influence than soft cast-iron brake-shoes. It is worthy of note that the material in the brake-shoes employed in these experiments was so hard that a number of the shoes were broken during the tests but without apparently affecting their utility, inasmuch as the form of the shoes remained unchanged, the parts being held in place by a steel plate cast in the outer sur- face of the shoe. The foregoing considerations assume the use of the automatic air brake. Inasmuch as high-speed trains have been under con- sideration, no other form than an automatic brake could properly be considered. In the case of a service employing single cars, the advantage of an automatic brake practically disappears and more simple forms of apparatus may be employed to advantage; but, where two OT more cars are assembled in trains, and particularly in high-speed trains, the necessity of providing for the contingency of train partings permits but the one prudent and safe course of employing an automatic brake, and, thus far, the automatic air brake alone has become safely established as meeting all the requirements of service. The necessity of the most efficient high- speed train service requires, in addition, the most forcible appli- cation of the most efficient form of automatic air brake the emergency application of the " high-speed " brake. ALTERNATING-CUBEENT MOTOKS. BY CHARLES PROTEUS STEINMETZ. I. In recent years a number of types of alternating-current motors have become of interest, which, while not new in their general principles, but antedating even the polyphase induction motor. have been for some time overshadowed by the latter, due to its greater simplicity, resulting from the absence of the commutator, and its constancy of speed. With the rapid extension of the applications of electricity, alternating-current motors were demanded for railway and similar classes of work, which give high-starting torque efficiency and high efficiency over a wide range of speed; that is a speed-torque characteristic similar to that of the direct-current series motor. The characteristic of the alternating-current induction motor, however, is that of a constant-speed motor, and indeed the poly- phase induction motor can theoretically be considered as an adapta- tion of the direct-current shunt motor to alternating current, as I have shown elsewhere. By the introduction of the commutator almost any speed-torque characteristic can be produced. A number of types of such commutator motors have been produced and more or less developed, but thus far practical experience has not yet advanced so far as to weed out the less desirable types. To enable a, critical judgment of their relative advantages a'nd disadvantages, I shall endeavor in the following to give a general theory of the alternating-current motor, applicable alike to the induction and commutator motors. The starting point of the theory of the polyphase and single- phase induction motor usually is the general alternating-current transformer, and from the equations of the general alternating- current transformer the induction motor equations can be de- veloped. 1 Coming, however, to the commutator motors, this method becomes less suitable. 1. Transactions A. I. E. E., 1895. [323] 224 8TEINMETZ: ALTERNATING-CURRENT MOTORS. In its general form the alternating-current motor consists of one or more stationary electric circuits magnetically related to one or more rotating electric circuits. These circuits can be excited by alternating currents, or some by alternating, others by direct current, or closed upon themselves, etc., and connection can be made to the rotating member either by collector rings that is, to fixed points of the windings or by commutator that is to fixed points in space. The alternating-current motors can be subdivided into two classes those in which the electric and magnetic relations be- tween stationary and moving members do not vary with their relative positions, and those in which they vary with the relative positions of stator and rotor. In the latter a cycle of rotation exists, and therefrom the tendency of the motor results to lock at a speed giving a definite ratio between the frequency of rotation and the frequency of impressed e.m.f. Such motors, therefore, are synchronous motors. The main types of synchronous motors are as follows: (1) One member supplied with alternating and the other with direct current polyphase or single-phase synchronous motors. (2) One member excited by alternating current, the other con- taining a single circuit closed upon itself synchronous induction motors. (3) One member excited by alternating current, the other of different magnetic reluctance in different directions (as polar con- struction) reaction motors. (4) One member excited by alternating current, the other by alternating current of different frequency or different direction of rotation general alternating-current transformer or frequency converter. No. 1 is the synchronous motor of the electrical industry. Nos. 2 and 3 are used occasionally to produce synchronous rotation with- out direct-current excitation, and of very great steadiness of the rate of rotation, where weight efficiency and power factor are of secondary importance. No. 4 is used to some extent as frequency converter. In the following I shall discuss only that type of motor in which the electric and magnetic relations between the stator and rotor do not vary with their relative positions, and the torque is, there- STEIN METZ: ALTERNATING-CURRENT MOTORS. 32-> fore, not limited to a definite synchronous speed. This requires that the rotor when connected to the outside circuit is connected through a commutator, and when closed upon itself several closed circuits exist, displaced in position from each other so as to offer a resultant closed circuit in any direction. In the theoretical in- vestigation I shall use the method of complex quantities, the ap- plication of which to alternating-current phenomena I outlined in a paper before a previous congress. 2 The extension of this method to vector products as torque and power is given in the appendix.* II. An alternating current / flowing through an electric circuit produces a magnetic flux $ interlinked with this circuit. Consider- ing equivalent sine waves of 7 and 0, lags behind I by the angle of hysteretic lag a. This magnetic flux induces an e.m.f. E=s 2 TT Nn&, where N = f requenc}', n = number of turns of electric circuit. This induced e.m.f. E lags 90 deg. behind the magnetic flux $, hence consumes an e.m.f. 90 deg. ahead of 0, or 90 deg. ahead of 7. This may be resolved in a wattless component: E = 2 -IT Nn $ cos a =2 IT N L I = x I, the e.m.f . consumed by self- induction, and an energy component : E" = 2 TT Nn & sin a= 2 tr N H I = r" I = e.m.f. consumed by hysteresis (eddy currents, etc.), and is, therefore, in vector representation denoted by E' = ~jxl and E"=r" I where x = 2 TT N L reactance, L = inductance, r" = effective hysteretic resistance. The ohmic resistance of the circuit, r', consumes an e.m.f. // in phase with the current, and the total or effective resistance of the circuit is, therefore, r = r' + r", and the total e.m.f. consumed by the circuit, or the impressed e.m.f. is E=(r jx) I = Z1 where Z = r jx = impedance, in vector notation, z = V?* 2 + z 2 = impedance, in absolute terms. If an electric circuit is in inductive relation to another electric circuit, it is advisable to separate the inductance L of the circuit into two parts the self-inductance S, which refers to that part of 2. Chicago, 1903, Proceedings Int. Elec. Cong., 1894. 3. See also Transactions A. I. E. E., 1899 326 STEIN METZ: ALTERNATING-CURRENT MOTORS. the magnetic flux produced by the current in one circuit which is interlinked only with this circuit but not with the other circuit, and the mutual inductance, M., which refers to that part of the magnetic flux interlinked also with the second circuit. The desirability of this separation results from the different character of the two components: The self -inductance induces a wattless e.m.f. and thereby causes a lag of the current, while the mutual inductance transfers power into the second circuit, hence generally does the useful work of the apparatus. This leads to the distinc- tion between the self-inductive impedance Z = r j X Q and the mutual inductive impedance Z = r j x. r is the coefficient of power consumption by ohmic resistance, hysteresis and eddy currents of the self -inductive flux effective resistance. x is the coefficient of e.m.f. consumed by the self-inductive flux self-inductive reactance. r is the coefficient of power consumption by hysteresis and eddy currents due to the mutual magnetic flux (hence contains no ohmic resistance component). x is the coefficient of e.m.f. consumed by the mutual magnetic flux. The e.m.f. consumed by the circuit is then E = Z I + Z I If one of the circuits rotates relatively to the other, then in addition to the e.m.f. of self -inductive impedance: Z Q I and the e.m.f . of mutual-inductive impedance or e.m.f. of alternation : Z I, an e.m.f. is consumed by rotation. This e.m.f. is in phase with the flux through which the coil rotates that is the flux parallel to the plane of the coil and proportional to the speed that is the frequency of rotation while the e.m.f. of alternation is 90 deg. ahead of the flux alternating through the coil that is the flux parallel to the axis of the coil and proportional to the fre- quency. If, therefore, Z' is the impedance corresponding to the former flux, the e.m.f. of rotation is / a Z r I, where a is the ratio of frequency of rotation to frequency of alternation, or the speed expressed as a fraction of synchronous speed. The total e.m.f. con- sumed in the circuit is thus : E=Z I-}-ZI-}-jaZ'L Applying now these considerations to the alternating-current motor, we assume all circuits reduced to the same number of turns - that is, selecting one circuit, of n effective turns, as starting STEIN METZ: ALTERNATING-CURRENT MOTORS. 327 point, if HI = number of effective turns of any other circuit, all the e.m.f s. of the latter circuit are divided, the currents multi- plied with the ratio njn, the impedances divided, the admittances multiplied with (n^/n) 2 This reduction of the constants of all circuits to the same number of effective turns is convenient by eliminating constant factors from the equations, and so permitting a direct comparison. When speaking, therefore, in the following of the impedance, etc., of the different circuits, we always refer to their reduced values (as it is customary in induction motor design- ing practice) . Let then, in Fig. 1, E 0f 7 , Z = impressed e.m.f., current and self-inductive impedance resp. of a stationary circuit, E^ 1^ Z^= FIG. i. impressed e.m.f., current and inductive impedance respectively of a rotating circuit, w = angle between the axis of the two circuits, Z = mutual-inductive or " exciting " impedance in the direction of the axis of the stationary coil, Z r = exciting impedance in the direction of the axis of the rotating coil, Z" = exciting impedance at right angles to the latter axis, and a = speed, as fraction of syn- chronism. It is then: In the stationary coil: E.m.f. consumed by self-inductive impedance: Z 7 E.m.f. consumed by mutual-inductive impedance: Z (/ + /! cos w) since the m.m.f. acting in the direction of the axis of the stationary coil is the resultant of both currents. Hence : Eo = Z I + Z (Zo+7, cos*) In the rotating circuit, it is : E.m.f. consumed by self-inductive impedance: Z l J 4 E.m.f. consumed by mutual-inductive impedance or "e.m.f. of alteration:" Z' (I I + 7 cos >) 328 STEIN METZ: ALTERNATING-CURRENT MOTORS. > E.m.f. of rotation: jaZ" 7 sin > Hence the impressed e.rn.f. : E 1 =Z 1 7 + Z' (7 1 + 7 cos ) + ; aZ" 7 sin In a structure with uniformly distributed winding, as used in induction motors, repulsion motors, etc., Z 1 = Z" = -Z t that is, the exciting impedance is the same in all directions. Z is the reciprocal of the " exciting admittance," Y of the in- duction motor theory. In the most general case, of a motor containing n circuits, of which some are revolving, some stationary, if: miu /kj Z k = impressed e.m.f., current and self -inductive im- pedance respectively of any circuit Tc Z 1 , and Z li = exciting impedance parallel and at right angles respectively to the axis of a circuit i f >i = angle between the axes of coils k and i f and fl = speed, as fraction of synchronism, or " frequency of rota- tion." It is then, in a coil i: EI = Zi /i + Z ! > ^k cos i i where : Zi Ii = e.m.f. of self -inductive impedance. n_ Z 1 yK 7k cos oi =^. e.m.f. of alternation E[=jaZ" >F 7 k sin w*, == e.m.f. of rotation which latter = o in a stationary coil, in which a = o. The power output of the motor is the sum of the powers of all the e.m.f s. of rotation, hence, in vector denotation 4 : = X / jZ il XL /k Sin a;,* , /| / and, therefore, the torque, in synchronous watts 8 : 4. See appendix. Also Transactions A. I. E. E., 1899. 5. See Transactions A. I. E. E., 1897, 1898, 1900. STEINMETZ: ALTERNATING-CURRENT MOTORS. 329 The power input, in vector denotation, is: = jr/ jsi. and therefore : P J = true power input P * = wattless voltampere input Qo = V(P l ) z + (P 3 ) 2 = apparent or voltampere input P 7* p-y = efficiency ; -p-f torque efficiency ; *o * o P T 7 - = apparent efficiency; -75 = apparent torque efficiency Vo Vo p 1 -~- = power factor ^o From the n circuits : i = 1, 2, . . . thus result n linear equa- tions, with 2n complex variables: I i and E { . Hence n further conditions must be given to determine th-3 variables. These obviously are the conditions of operation of the n circuits. Impressed e.m.fs. E i may be given. Or circuits closed upon themselves: E i =o. Or circuits connected in parallel : GI E\ = c k E^ f where c i and c k are the reduction factors of the circuits to equal number of ef- fective turns, as discussed before. Or circuits connected in series: c\ 1\ = c k / k> etc. When a rotating circuit is connected through a commutator, the frequency of the current in this circuit obviously is the same a? the impressed frequency. Where, however, a rotating circuit is permanently closed upon itself, its frequency may differ from the impressed frequency, as, for instance, in the polyphase induction motor it is the frequency of slip s = l a. and the self -inductive reactance of the circuit, therefore, is: 5 x, though in its reaction upon the stationary system the rotating system necessarily is al- ways of full frequency. After this introduction we come now to the discussion of a few motor types. We shall, however, consider only such types as have been more or less developed commercially or at least seriously con- sidered. 330 STEIN METZ: ALTERNATING-CURRENT MOTORS. III. (1) Polyphase Induction Motor. In the polyphase induction motor a number of primary circuits, displaced in position from each other, are excited by polyphase e.m.f's. displaced in phase from each other by a phase angle equal to the position angle of the coils. A number of secondary circuits are closed upon themselves. The primary usually is ttw stator, the secondary the rotor. In this case the secondary system always offers a resultant closed circuit in the direction of the axis of each primary coil, ir- respective of its position. Let us assume two primary circuits in quadrature as simplest form, and the secondary system reduced to the same number of phases and the same number of turns per phase as the primary system. With three or more primary phases the method of pro- cedure and the resultant equations are essentially the same. k FIG. 2. POLYPHASE INDUCTION MOTOB. Let in the motor shown diagrammatically in Fig. 2 : E ' and jE I and // . Z = impressed e.m.f's., currents and "OJO O O self-inductive impedance respectively of the primary system. o, Zj and jl lf Z^ = impressed e.m.f ., currents and self-inductive impedance respectively of secondary system, reduced to primary. Z = mutual-inductive impedance between primary and sec- ondary. a = speed ; s = 1 a = slip, as fraction of synchronism. The equation of the primary circuit is then : (/-',) (i) OF UNIVERSITY x/ STEIN METZ: ALTERNATING-CURRENT MOTORS. 331 The equation of the secondary circuit: o = ZJi+Z (I- 1, ) +jaZ (jlf- ]'I ) (2) from (2) follows: Z (l-a) _ Zs 7l o - /0 ZtZ and, substituted in (1) : Primary current: T Zs+Zj _ ^ ... 2 o- ^o ZZ s+ZZ l +Z Z l v ; Secondary current : Exciting current: 9 I = I 7 i = E o E.m.f. of rotation: jl )=aZ C zz, = aE. It is, at synchronism: s = o: T E T .r-T-w- /0 = ^FC ; " ' ~ At standstill : s = 1 ; " Introducing as parameter the counter e.m.f., or e.m.f. of mutual induction: or: it is, substituted: Counter e.m.f . : hence : Primary impressed e.m.f.: Y O \ ^ + ZZi ~Zi + Z l } 332 STEIN HETZ: ALTERNATING-CURRENT MOTORS. E.m.f. of rotation: E^=Ea = E (18). (10) Secondary current: Primary current: Exciting current: These are the equations from which the transformer theory of the polyphase induction motor starts. Since the frequency of the secondary induced currents is the frequency of slip, hence varies with the speed a = 1 5. the sec- ondary self -inductive reactance also varies with the speed, and so the impedance: Z 1 = r 1 jsx 1 (14) The power output of the motor, per circuit, is: e 2 z 2 (l 3) /,.*_ /i-x (fl+}SXl) where the brackets [] denote the absolute value of the term in- cluded by it, and the small letters e , z, etc., the absolute values of the vectors E o , Z, etc. Since the imaginary term of power seems to have no physical meaning, it is: Mechanical power output: P- ~ This is the power output at the armature conductors, hence includes friction and windage. The torque of the motor is: x, __ ( V ; i ] 2 The imaginary component of torque seems to represent the ra- dial force or trust acting between stator and motor. Omitting it, it is: STEIN METZ: ALTERNATING-CURRENT MOTORS. 333 The power input of the motor per circuit is: P /E I / . y7 77 ZZ fi -{- ZZ V -f- Zo% 1 = PJ+jPo* where: PJ = true power, P\= reactive or "wattless power," Q Q = \/ p? 4- pi 2 = apparent power, or voltampere input o o Herefrom follows power factor, efficiency, etc. Introducing the parameter: E, or absolute: e t it is: Power output: (80) O Power input: 2 ,ZZ .9 + ZZ, p ^-_i zz, ' 9/0 , !> ~ And since: - 1 = a + * r, = 2Q + r,, and ifari = P, it is: ^o = ft, ^o + i 2 n + C 2 r + P) + j(i5 "T" ^1 7 = __' E Z. 777 1 77T ^i = ^o , Hence, at synchronism, the secondary current of the single-phase induction motor does not become zero, as in the polyphase motor, but both components of secondary current become equaL At standstill : a = o, s = 1 it is : 2 jg Z + Z\ ZZo ~\~ ZZi + Zo Zi Z,Zo ~h ZZi -\- Zo Zi That is, primary and secondaiy current corresponding thereto have the same values as in the polyphase induction motor, page 8. This was to be expected. Introducing as parameter the counter e.m.f. or e.m.f. of mutual induction : T/T TjJ F7 ~T & =^o ^o 1 o and substituting for I from (6), it is: Primary impressed e.m.f. : *7 ( *7^ o I o ^7 ?7 i ?7 2\ i ?7 17 Primary current: Secondary energy circuit: --P o _ o a'E = ' ~~ ~ Secondary magnetizing circuit: n T?. r 06) ELEC. RYS. 22. 33S 8TEIXMETZ: ALTERNATING-CURRENT MOTORS. E 2 * = jaE (17) And: /o-A=f (18) These equations differ from the equations of the polyphase in- duction motor by containing the term: s = (1 a 2 ) instead of: s=(l a), and by the appearance of the terms:!? - and: 2 Tji - . , , of frequency (1 + a), in the secondary circuit. Z ~r Z\ The power output of the motor is : _o ' a? r t (. z 2 - z, ) (19) [Up and the torque, in synchronous watts : -o (20) From these equations it follows that at sjrnchronism torque and power of the single-phase induction motor are already negative. Torque and power become zero for: hence: : that is, very slightly below synchronism : Let z = 10, z = A it is: a = .9995. In the single-phase induction motor, the torque contains the speed a as factor, and thus becomes zero at standstill. Neglecting quantities of secondary order, it is, approximately: T - F Z*o+*Zi |M x 2 = L Q ^, ^. (22) /i = ^ ^(^^^ + 2^ ^ (23) 7 2 = -> ^ z (Z o*>+ Z Zi)+*ZoZi (24) js', 1 = 2 ^; z(Zo*o-\^z Z i) + *ZoZi (25) ^ => ^ V . V . i ^' L o ^ V < 26 > STEIN METZ: ALTERNATING-CURRENT MOTORS. 339 This theory of the single-phase induction motor differs from that previously communicated (see note 1, ante), in that it repre- sents more exactly the phenomena at intermediate speeds, which are only approximated in the transformer theory of the single- phase induction motor. As instance are shown, in Fig. 5, with the speed as abscissae, the curves of a single-phase induction motor, of the constants: e = 400 volts Z = 1 10; ohms Zo ^^ Z 1 =.'L .3; ohms hence : / = 400 -p- amps. N = (s +.2) ;(10s +.6 .60) T ' = 1616 as 120 JIG 100 90 80 TO 60 50 40 80 20 10 SINGLE PHASE INDUCTION MOT08 400 VOLTS ~850 -250- -200- -50- t 100 90 80 TO 60 60 40 80 20 10 Pro. 5. (3) Single-phase Condenser Motor. The single-phase induction motor is not self -starting, as seen from the equations and diagram, Fig. 5. To secure 340 STEIN METZ: ALTERNATING-CURRENT MOTORS. starting torque, either a commutator has to be used that is, the motor started as repulsion motor or series motor, etc. or a quadrature magnetic flux impressed upon the motor, that is the motor converted into a more or less unsymmetrical, poly- phase motor. To a considerable extent used in practice are only the starting as repulsion motor, which will be discussed later, and the starting by a condenser in the tertiary circuit, both methods giving good starting efficiencies. The use of a condenser also per- mits to greatly increase the power factor in running, by retaining the condenser in circuit. This is usually carried out by employing a three-phase winding on the motor primary, of which two ter- minals are connected to the single-phase supply, two terminals permanently connected to a condenser, either directly or by step- "t JJL Fro. 6. INDUCTION SINGLE-PHASE CONDENSEB MOTOR. up transformer. This condenser so closes a circuit displaced by 60 deg. in position from the primary circuit, as shown diagrammati- cally in Fig. 6. Let, in the diagram Fig. 6, of such a single-phase condenser motor : EO, 7 , Z impressed e.m.f., current and self -inductive im- pedance respectively of the primary circuit, /!, Z = current and self -inductive impedance of the secondary energy circuit, 7 2 , Z x = current and self -inductive impedance of the secondary magnetizing circuit, 7 3 = current in the condenser circuit, or tertiary circuit, Z s =r 3 -\- j x & total effective impedance (leading) of the condenser circuit, Z = mutual-inductive impedance, ut = position angle between the axes of primary and tertiary circuit, STEIN AIETZ: ALTERNATING-CURRENT MOTORS. 341 a = speed. The equations of the motor then are: Primary circuit: E = Z I + Z (I 7 X 7 3 cos ,)_ (1) Secondary energy circuit : o = Z 1 I 1 + Z (/! / + 7 8 coBa,) +jaZ (7 2 I 8 8in") (2) Secondary magnetizing circuit: o = Z l I 2 + Z (7 2 7 3 sin0 +jaZ (I, 7 + 7 3 cos ,) (3) Tertiary or condenser circuit: o = Z z I s -{- Z (I 3 7 cos w -f- 7 1 cos -*3 and thereby the e.m.f ? s. of rotation: E\ = jaZ (7 2 7 3 sin ) (5) 'S t =jaZ (A / O + /,CQB) (6) end therefrom the torque, power output, input, etc. Usually a* is made 60 deg. in this type of motor. (4) Polyphase Shunt Motor. Since the characteristics of the polyphase motor do not depend upon the number of phases, here, as in the preceding, a two-phase system may be assumed : that is, a two-phase stator winding acting upon a two-phase rotor winding, that is a closed coil rotor wind- ing connected to the commutator in the same manner as in direct- current machines, but with two sets of brushes in quadrature posi- tion excited by a two-phase system of the same frequency. Me- chanically the three-phase system here has the advantage to re- quire three sets of brushes only instead of four with the two-phase system, but otherwise the general form of the equations and con- clusions are not different. Let E and / E = e.m.f's. impressed upon the stator, E t and ; E! e.m.f J s. impressed upon the rotor, to = phase angle between e.m.f. E Q and E^ and ">!= position angle between the stator and rotor circuits. The e.m.f's. E and ; E produce the same rotating m.m.f. as two e.m.f's. of equal intensity, but displaced in phase and in position by angle o> from E and / E , and instead of con- sidering a displacement of phase < and a displacement of position >i between stator and rotor circuits, we can, therefore, assume zero-phase displacement and displacement in position by angle <"<> -f w i = ">. Phase displacement between stator and rotor e.m.f s. is, therefore, equivalent to a shift of brushes. 8TEINMETZ: ALTERNATING-CURRENT MOTORS. Without losing in generality of the problem, we can, therefore, assume the stator e.m.f's. in phase with the rotor e.m.fs., and the polyphase shunt motor can thus be represented diagrammatically by Fig. 7. Let, in the polyphase shunt motor, shown two-phase in diagram Fig. 7: E Q and jE , I and ;7 , Z = impressed e.m.f s., currents and self-inductive impedance respectively of the stator circuits, Jit FIG. 7. POLYPHASE SHUNT MOTOR. cE and jcE , 1^ and ;7 1? Z^= impressed e.m.f's., currents and self-inductive impedance resp. of the rotor circuits, reduced to the stator circuits by the ratio of effective turns c, Z = mutual-inductive impedance, a = speed, hence : s = 1 a = slip, w= position angle between stator and rotor circuits, or " brush angle." It is then: Stator: E = Z 7 + Z (7 7 t cos a, + /7 sin ,) Rotor : cEf=ZJi+Z (/!/ COB jig sin )+jaZ 7 sin > ;7 cos ^) Substituting : (l) ff = cos d = cos it ifl- ff6r and: a> -fjF sin =W ai (15) That is, if the brush angle < is complimentary to the phase angle of the self -inductive rotor impedance a^ the motor tends toward approximate synchronism at no load. The rotor current: ' sZZ +ZZi+Z Zi becomes zero, if: or, since Z is small compared with Z, approximately: c= ff 8= s (cos ) hence, resolved: c = s cos of o = s sin o hence : That is, the rotor current can become zero only if the brushes are set in line with the stator circuit or without shift, and in this case the rotor current, and therewith the output of the motor, becomes zero at the slip s = c. STEIN METZ: ALTERNATING-CURRENT MOTORS. 345 Hence such a motor gives a characteristic curve very similar to that of the polyphase induction motor, except that the stator tends not toward synchronism but toward a definite speed equal to (1 + c) times synchronism. The speed of such a polyphase motor with commutator can, therefore, be varied from synchronism by the insertion of an e.m.f . in the rotor circuit, and the percentage of variation is the same as the ratio of the impressed motor e.m.f. to the impressed stator e.m.f. A rotor e.m.f. in opposition to the stator e.m.f. reduces, in phase with the stator e.m.f. increases the free running speed of the motor. In the former case the rotor impressed e.m.f. 'is in opposition to the rotor current, that is the rotor returns power into the system in the proportion in which the speed is reduced, and the speed variation, therefore, occurs without loss of efficiency, and is similar in its character to the speed control of a direct-cur- rent shunt motor by varying the ratio between the e.m.f. im- pressed upon the armature and that impressed upon the field. Substituting in the equations: it is: r IT* iZ ~\~ Zi /i o\ ^ fl = i. ~l*ZZo+ZZi+ZoZilp These equations are very similar to the polyphase induction motor equations. The stator current: can be resolved into a component: ri __ jjr _ s Z ~r Zi _ /oi\ ^ which does not contain c, and is the same value as the primary cur- rent of the polyphase induction motor, and a component: (22) 340 STEINMETZ: ALTERNATING-CURRENT MOTORS. Resolving /o 11 , it assumes the form: /j*fcsJ(>44i^/t;) = c^Ai cos u>-\- A z sin at) j (At sin at A 2 cos o>) i Hence, by choosing: AI cos 01 + 4 2 sin = o or: tan* 4 1 (23) ^2 it is: Hence this component can, by choosing *<*>) + aZ (* a ia ( r cs w + a? sin > r) For : c = 1, or equal number of effective turns in stator and rotor, it is: TJ1 . . ~[Zo + Zi + *Z (l cos w) + oZ^ i)] 1 The characteristics of this motor entirely vary with a change of ae 2 r (c 1) . the brush angle *. It is, for : <*> hence very considerable. email, while for * =90: P= r 2 , hence Some brush angles give positive P: motor, others negative P; generator. Substituting in (7) for Z, etc., it is : j-'L. _ % _ jc 2 r + TI + r (1 -f c 2 2 c cos w) + a (c(r cos w + a; sin o) _ r ) J J| c 2 x +Xi + a; (1 -- c 2 2 c cos a>) + a (c( cos r sin a>) a) ^ STEIN METZ: ALTERNATING-CURRENT MOTORS. 349 hence the angle of lag of the current input behind the impressed e.m.f. is given by: _c 2 x + x l -f x (1 -f-c 2 2c cos o>) -f a (c a; cos a* r ~ ' ) -f a (c(r cos w + x sin w) aj) sin r) (14) In such a motor, by choosing o and c, appropriately unity power factor or leading current as well as lagging current can be pro- duced. The limits of this paper, however, do not permit a further discussion of the very interesting characteristics derived by choosing different values of c and > in polyphase as well as single-phase shunt and series motors, and an investigation of the effect of the short- circuit current under the commutator brushes. I: POLYPHASE SERIES MOTOR 40 VOLTS 30 40 20 20 10 As instance as shown in Fig. 9, with the speed as abscissae, and values from standstill to over double synchronous speed, the char- acteristic curves of a polyphase series motor of the constants: e = 640 volts Z 1 10 j ohms Z = Z 1 = .I .3; ohms 01= 37 (sin to = .6 ; cos w = .8) 350 STEIN METZ: ALTERNATING-CURRENT MOTORS. hence: 640 n-^r * m P** 5.8a) 4673a kw. (.6 + 5.8a, 2 + (4.6 2.6a) 2 As seen, the motor characteristics are similar to those of the direct-current series motor : very high torque in starting and at low speed, and a speed, which increases indefinitely with the decrease of load. That is the curves are entirely different from those of the induction motors shown in the preceding. The power factor is very high, much higher than in induction motors, and becomes unity at the speed: a = 1.7 7, or about one three-quarter syn- chronous speed. IV. SINGLE-PHASE COMMUTATOR MOTORS. In polyphase motors and motors of similar type a distributed rotor and stator winding is used, that is a structure having uniform magnetic reluctance and thus exciting impedance in all directions, and a polar construction of the stator winding results in lower power factor, and thus is permissible only in very small motors as fan motors, etc. In direct-current motors a polar construc- tion of the stator is almost exclusively used, that is a construction in which the reluctance in the direction of the magnetic field, which produces the e.m.f. of rotation, is very much smaller than in the direction at right angles thereto. In single-phase alternating com- mutator motors (as series motors, repulsion motors, etc.) both stator constructions may be used, and in the most general case we must, therefore, assume the magnetic reluctance and so the exciting impedance in the direction of the axis of the rotor cir- cuits Z' as different from the exciting impedance Z at right angles to this axis. When different, the latter Z is usually far larger than the former Z' ', since Z is in the direction of the magnetic flux which produces the e.m.f. of rotation, that is corresponds to the field excitation, while in the direction of Z' energy transfer between stator and rotor, or compensation of rotor reaction takes place, but magnetic flux in the direction Z' does not produce e.m.f. and thereby power by the rotation of the motor. The stator winding can, therefore, be considered as consisting 8TEINMETZ: ALTERNATING-CURRENT MOTORS. of two components, or may be constructed of two separate circuits, in the directions in line and at right angles to the rotor winding, which circuits may be connected in series or energized in any other manner, as, for instance, by exciting one by the impressed e.m.f., short-circuiting the other upon itself, etc. With a com- pletely distributed winding and an angle "> between the axes of the stator and the rotor circuits (the angle of brush position), the exciting or magnetizing component of the stator winding is 7 sin at , the compensating or power transferring component 7 cos > if 7 = stator current, as shown in diagram Fig. 10. When using separate circuits for the two stator components, they can even magnetically be arranged differently, as, for instance, a unitooth or polar arrangement chosen for the field exciting circuit, a dis- 'o< * , FIG. 10. FIG. 11. SINGLE-PHASE SERIES MOTOB. tributed winding for the compensating circuit. In this case ob- viously, when reducing all circuits to each other by the ratio of effective turns, the resultant vector of the distributed winding has to be used. As limit case, with zero compensating winding, appears the plain uncompensated series motor, consisting of a polar field ex- citing circuit and an armature with brushes at the neutral or at right angles to the field, as shown in Fig. 11; as a further limit case, a motor with zero field exciting winding on the stator and excitation of the rotor by a second system of brushes at right angles to the main or power brushes, as shown diagrammatically in Fig. 12. In alternating-current commutator motors, especially of the single-phase type, the short-circuit current in the coils under the brushes during commutation has to be taken into consideration. While with numerous commutator segments, carbon brushes and possibly an additional resistance in the commutator leads, as 352 STEIN METZ: ALTERNATING-CURRENT MOTORS. occasionally used in such motors, these short-circuit currents may be moderate, they still are sufficient to noticeably affect the con- stants of the motor, especially at high speeds, where the mail] current is small, and at standstill, where the main magnetic flux is very large. Furthermore, the character of the commutation of J^ 12. WlNTER-ElCHBERG-LATOUB MOTOR. the motor, and, therefore, its operativeness, depends upon these currents. An excessive short-circuit current gives destructive spark- ing, while zero short-circuit current would be conducive to perfect commutation. In comparing different types of such motors, the investigation of the short-circuit current under the brushes is, therefore, of fundamental practical importance. In its most general form, the single-phase commutator motor can thus be represented diagrammatically by Fig. 13. FIG. 13. Let: E 0f I , Z = impressed e.m.f. current and self-inductive impedance of magnetizing or exciter circuit of stator (field coils), reduced to the rotor energy circuit by the ratio of effective turns c, EI, /!, Zi = impressed e.m.f., current and self-inductive imped- ance of rotor energy circuit (or circuit at right angles to 7 ), 77 2 , 7 2 , Z z = impressed e.m.f., current and self-inductive imped- STEIN METZ: ALTERNATING-CURRENT MOTORS. 353 ance of stator compensating circuit (or circuit parallel to J t ; the ' cross-coil" of the Eickemeyer motor), reduced to the rotor cir- cuit by the ratio of effective turns 6, v I&> Z impressed e.m.f., current and self -inductive im- pedance of the exciting circuit of the rotor, or circuit parallel to 7 , A> + Z 1 (/.+ If- I,) + jaZ (/.+ If-I t ). (6) Substituting : Z*/Z = A t where 4 = 1 with a motor of uniform reluct- ance, (7) Z/Z 6 =l s where >* 4 and ; 8 are small quantities, and suppressing terms of secondary order, equations (5) and (6) give: / 4 = W (/o + IB) +Ja,A (!B - Ii) \ () J 8 ) f (10) Substituting (9) and (10) into (1), (2), (3), (4), gives four equations containing the eight quantities: E 09 E lt E 2 , E S) 7 , I v 7 2 . 7 8 , requiring four further conditions to be given, which are the conditions of operation of the four circuits, and distinguish the different types or modifications of such single-phase alternating- current motors. Some of the types under practical considerations at present are : (1.) Series Motor: E = cE Q + EI; I = c7 t ; 7 2 = o; 7 8 = o. ELEC. RYS. 23. 354 STEINMETZ: ALTERNATING-CURRENT MOTORS. (2.) Compensated Series Motor (Eickemeyer Motor), (a.) direct compensation: E = cE + EI+ bE 2 ; I = cl i; I 2 = 6J i; 7,= r o. (b.) inductive compensation: E = cE + E I; E 2 = o; I = c/ 1 ; I & = o. (3.) Eepulsion Motor (Thomson Motor) : E = cE + bE 2 ; tf t = o ; cI = bI 2 ; I s = o. (4.) Compensated Repulsion Motor (Winter-Eichberg-Latour Motor) : E=bE 2 + fE 3 ; E,= o; I = o; bI 2 = fI 3 . (5.) Inverted Series Motor: E = E,+ bE 2 + fE s ; 7 e = o; 1^=11,; If=fl v (6.) Inverted Repulsion Motor: E = E I; cE + bE 2 = o; cI = II 2 ; I & =o. (7.) Induced Series Motor: E = E 2 ; EI+ cE =o; cI =l^ I s =o. Types (4.) and (5.) have two sets of brushes on the rotor. In types (3.) and (7.), the rotor is not connected to the external or supply circuit, and its voltage can, therefore, be chosen inde- pendent of the supply voltage; in type (4.), by feeding circuit E 3 through transformer, the same may be secured. Frequently in motors of uniform reluctance : Z*= Z, as the two stator circuits / and I 2 the two parts of the same uniformly dis- tributed circuit are used, and then c/&=tan ,, where a>= angle of brush shift. Only a few of the more important types can be discussed in the following : (1.) Single-phase Series Motor. The plain or uncompensated single-phase series motor is usually designed with definite field poles, similar to the direct-current series motor (only that the field is laminated also). The object of the polar construction is to secure as low a value of Z 1 and as high a value of Z as possible, so as to reduce the armature self- induction which is not compensated, and secure a fair power factor. Let then, in the motor shown diagrammatically in Fig. 14 : E = impressed e.m.f., 1 = current, c = ratio of effective field turns to effective armature turns ; E ' I Q , Z Q , Z = impressed e.m.f., current, self-inductive and mutual-inductive impedance of field circuit, reduced to armature circuit ; STEIN METZ: ALTERNATING-CURRENT MOTORS. 355 ^u Ii> %\> %*= impressed e.m.f., current, self-inductive and mutual-inductive impedance of armature circuit; 7 4 , Z= current and self-inductive impedance of the short-cir- cuit under the brush, reduced to the armature circuit. FIG. 14. SINGLE-PHASE SERIES MOTOB. a = speed. Z/Zt=l = ^ It is then: E = c I =cl Hence: And: ^1=^1 h+Z*It+jaZ (7 -7 4 ) = Z. 7 4 + Z (7 4 7 ) E I, = 1(1 e ja (1) (2) (3) (4) (5) (6) (7) c 2 (Z+ Z ) + (eljai. 1 ) Or, denoting: Z, + Z' + Z (o +ja) (c (9) (10) The e.m.f. of rotation of the main circuit is D (12) 350 STEIN METZ: ALTERNATING-CURRENT MOTORS. of the short-circuit under the brush: The power output of the motor is the algebraic sum of the power of the main rotor circuit, and that of the short-circuit under the brush, hence is: . 1 ,'--.^ 1 / 1 ( and since: . /JZI, l/*= it is: _ and the torque : )( (15) In the equation of the current (8), c- (Z + Z) is the total impedance of the field, Z t + Z 1 is the total impedance of the armature, hence : c 2 (#0+ ^) + (^i+ ^) is the ttal impedance of the motor, cor- responding to the e.m.f. consumed by the effective resistance and the self-induction of field and armature, jacZ corresponds to the e.m.f. of rotation, or the mechanical work done by the motor, and Z (c + ja) (d jaP) is the effect of the short-circuit current under the commutator brush. STEIN METZ: ALTERNATING-CURRENT MOTORS. 357 Neglecting the short-circuit current of commutation, as of secondary order, it is : /= 1 -ocr (16) hence, the angle of lag of the current I behind the impressed e.m.f. E is given by: With increasing speed a, the numerator decreases, the denomina- tor increases, hence the angle of lag decreases and the power factor cos increases. The power factor of the motor becomes unity, or < = 0, at the speed: __*(*+*. )+(* +.) cr That is at some very high speed the power factor of the single- phase alternating-current series motor, even if not compensated, would become unity, if there were no commutation losses. On first sight this is unexpected, since even assuming the arma- ture as entirely non-inductive, in addition to the e.m.f. induced in the armature by the rotation through the alternating magnetic field, and in phase thereto, in the field coils a quadrature e.m.f. must be induced by the same magnetic flux, and while the former increases relatively to the latter with the speed, the quadrature e.m.f. obviously never can become zero. The explanation is found in the following: In equation (17) the denominator contains the effective exciting resistances r as factor, which represents the hysteretic loss in the motor, and if r = o, or no hysteresis loss, unity power factor would be reached only at infinite speed. Due to the hysteresis loss in the alternat- ing magnetic field, when considering equivalent sine waves, the magnetic flux lags behind the magnetizing current by the angle of hysteretic lag , and the e.m.f. of rotation, which is in phase with the magnetic flux, therefore, lags behind the current, that is the current leads the e.m.f. of rotation, and so at a certain definite speed compensation for the lag due to the e.m.f. of self-induction in the motor takes place by the lead of the e.m.f. of rotation ahead of the magnetizing current, which in this case is the main current 358 STEIN METZ: ALTERNATING-CURRENT MOTORS. of the motor. This feature is found in nearly all types of single- phase commutator motors, that is at a certain high speed, when neglecting commutation losses, the current is in phase with the impressed e.m.f. (and at still higher speed leading), and when considering equivalent sine waves the power factor is unity. Con- sidering the actual wave shape, however, there remains a wattless component which represents the wave-shape distortion caused by the hysteretic cycle of the magnetic field. It also follows that in all such single-phase commutator motors a certain wave-shape dis- tortion must take place, since the e.m.f. of rotation is of the same wave shape as the magnetic field flux, but the magnetic field flux and the current differ in wave shape by the wave-shape distortion represented in the hysteretic cycle of the magnetic structure. At given speed a, the power factor is a maximum for that value of c, where: substituting (17), and suppressing quantities of higher order, this gives: rx 1 +xr l . / yx l -\- rr l l . / yx l or approximately, for higher speeds a : (20) Since Z*! E 19 /u Z^== impressed e.m.f., current and self-inductive im- pedance of armature circuit, E 2 , 7 2 , Z 2 ,= impressed e.m.f., current and self-inductive imped- ance of (stationary) compensating circuit, reduced to the arma- ture circuit by the ratio of effective turns &, 7 4 , Zf= current and self-inductive impedance of the short-cir- cuit under the commutator brush, reduced to the armature circuit, FlG. 16. ElCKEMETEB MOTOB DIBECT COMPENSATION. Z = mutual-inductive impedance, constant in all directions, (I) (2) (3) (4) (6) (6) (1) (8) (9) (10) a = speed. It is then : I 2 =bl Field circuit: Compensating circuit: E^ZJi+ZV^- Armature circuit: Brush short-circuit: o=ZJ l +Z (/,-/ Herefrom follows: >el approximately 362 8TEINMETZ: ALTERNATING-CURRENT MOTORS. Main current: pi Zi + V s Zz + Z (I- b? + o Z(c where : D=\c*Z,+Z 1 +VZ 2 +Z(l-ir\+cZ(c+}a).(l-i) (13) Short-circuit current under brushes : Ti==s lE\c-ja(\-b) \ IcE /n K x = fi- approx. c.m.f. of rotation of main armature circuit: e.m.f. of rotation of brush short-circuit: Power output: In the equation of the current, (11), c 2 (Z Q +Z) is the total impedance of the field, b 2 Z 2 is the total impedance of the compen- sating circuit, Z-\- Z (\ &) 2 is the total impedance of the arma- ture, the component Z(l &) 2 being due to incomplete compen- sation. In the uncompensated motor on its place stands Z 1 . Neglecting the effect of the short-circuit under the brush in equation (11), and substituting for Z, etc., it is: 7= + 0V, + n + ( I b^r 4- acx }> j^ c\x + x) 4- b* a^+Ki -^ ( I bfx acr } (20J STEINMETZ: ALTERNATING-CURRENT MOTORS. 3G3 hence the angle of lag of the motor : _c^g + g) + 6 2 ^-t-a;H-(l bYxacr ^(r +r)+aS+ri+(l 6)V+oc* 1. Hence, by overcompensation a reverse e.m.f. can be inserted into the coil short-circuited under the brushes, and thereby the commutation controlled, that is, sparkless commutation secured, at Hie expense, however, of some decrease of the power factor. 364 STEIN METZ: ALTERNATING-CURRENT MOTORS. As instance may be considered a motor of the constants: e = 500 volts, Z = I 10; ohms, Z 1 = .1 .3; ohms, Z 2 = .13 Aj ohms, Z Q = A 1.2; ohms, 4 = 30 30; ohms, A = .18 .15; c=.25 500 (.4+ 2.01a) j (1.28 .58a) 502.5g hence : hence: _ -f (1.28 .58a) 2 Since the curves of this motor are almost identical with those of the inductively compensated motor, they are not given. (&.) Inductively Compensated Motor. Let, in the motor shown diagrammatically in Fig. 17 the denota- tions be the same as in (a.), the directly compensated motor, except FlG. 17. ElCKEMEYEE MOTOB INDUCTIVE COMPENSATION. that now 7 2 is a separate, secondary current, and not = 67, and E 2 = o. It is then : E = Ei + cE, (1) Field circuit: (3) (4) STEIN METZ: ALTERNATING-CURRENT MOTORS. 365 Armature circuit: /.-/) (5) Compensating circuit: o = ZJ 2 +Z (!, !,) Short-circuit under brush: o = ZJ t + Z (!<-!) + jaZ (/-/,) (7) From (6) follows: from (7) : hence substituted into equations (1) to (5) Main current: Zi xwl ' ?? i /4 = -. where : D = c>Z + Z, + ~ z - + eZ (c+ja) (1-4) (112) Short-circuit current under commutator brush: (13) = j approx. E.m.f. of rotation of main circuit: EI = ~: \**v E.m.f. of rotation of armature short-circuited coil: hence very small. Power output, suppressing terms of secondary magnitude: Torque : ca;^ 2 / r \ /1 , r^iv '""^ v 3GG STEIN METZ: ALTERNATING-CURRENT MOTORS. zz> As seen, these equations contain: instead of: b 2 Zz + (1 bf Z of the directly compensated motor, which latter, for 7 7 6 = 1, gives Z 2 . Since Z 2 is small compared with Z, - . L- is almost identical with Z 2 , inductive compensation gives almost iden- tically the same results as complete direct compensation, and all conclusions derived under (a.) for the case of complete compensa- tion: 6 = 1, apply to the case of inductive compensation. 800 275 250 200 irs 150 125 100 'IL u_ T\ \ \ EICKEMEYER MOTOR 500 VOLTS T< 280 260 240 220 200 180 160 140 120 100 1.0 1.6 1.8 2.0 FIG. 18. As instance are shown, in Fig. 18, with the speed a as abscissae, the curves of an inductively compensated motor of the constants: e = 500 volts, Z = 1 10; ohms, Z 1 = .l .3; ohms, Z 2 = .l3 Aj ohms, Z = A 1.2 j ohms, Z s = 30 30; ohms, hence : * = .18 .15; Hence: 7 = 500 (.394-2.01a) j ^1.27 58,;) (.39+2.0 1 a)* (1.27 .58) STEIN METZ: ALTERNATING-CURRENT MOTORS. 367 Interesting is the very high power factor reached already at low speed: 80 per cent below half synchronism. At speed: a = 2.19 unity power factor is reached. The starting torque is very large, and with increase of speed the . torque falls rapidly, very similar as in a direct-current series motor. (3) Repulsion Motor (Thomson Motor). In Prof. E. Thomson's single-phase repulsion motor the stator is supplied with the main current, the rotor short-circuited upon itself through the commutator brushes under an angle with the axis of the stator circuit. Amongst the single-phase commutator motors this repulsion motor takes a separate and distinctive position by its magnetic characteristics and their effect on commutation, so that single- phase commutator motors may be divided into series motors and repulsion motors. While both types of motors have similar speed characteristics, the magnetic flux of the repulsion motor is an elliptically rotating flux, while that of the series motor is essentially an alternating flux. In the series motors treated in the preceding, the magnetic flux in the axis of the rotor circuit is either negli- gible, in the compensated motor, or as magnetic flux of armature reaction in phase with the main magnetic flux. The e.m.f. in- duced in the armature coil short-circuited under the brush, by its rotation, is, therefore, either negligible or in phase with the main flux, while that induced by the alternation of the flux enclosed by the short-circuited coil is in quadrature with the main flux, and so with the e.m.f. of rotation, and the short-circuited coil is the seat of an active e.m.f. at all speeds. In the repulsion motor, the magnetic flux in the direction of the axis of the armature circuit is in quadrature with the current and thereby the flux at right angles with the armature circuit, but the former is constant, the latter varying inversely with the speed. The e.m.f. induced by rotation in the coil short-circuited under the commutator brush is in phase with the quadrature field of the motor, while the e.m.f. of alternation is in quadrature with the main field, and since the two fields are in quadrature with each other, the two e.m.f s. in- duced in the short-circuited coil are in opposition to each other, that is neutralize each other more or less completely. At synchron- ism the two e.m.f s. are equal and opposite, the neutralization com- plete and commutation, therefore, theoretically perfect. 308 8TEINMETZ: ALTERNATING-CURRENT MOTORS. The repulsion motor can be constructed with distributed or with polar stator winding. Since, however, compensation takes place of the armature reaction by the primary current and the secondary current flowing in opposite direction, and the rotating !u.m.f. of the motor can produce a uniformly revolving (circular or elliptic) magnetic field only in a structure of uniform reluctance, polar winding gives decidedly inferior characteristics and a dis- tributed stator winding is, therefore, assumed in the following. With polar construction, different exciting impedances Z and Z l have to be introduced in the two quadrature directions. FIG. 19. THOMSON MOTOR, Let, in a repulsion motor : E , / , Z = impressed e.m.f., current and self-inductive im- pedance of primary or stator circuit, I lf Z 1 = current and self -inductive impedance of secondary or rotcr circuit, reduced to primary by the ratio of effective turns, 7 4 , Z 4 current and self -inductive impedance of short-circuit under brush, reduced to primary circuit, Z = mutual-inductive impedance, a = speed, as fraction of synchronism, to = angle between axis of primary and secondary circuit, or angle of brush shift. It is then, in the motor shown diagrammatically in Fig. 19. Stator: E = ZJ + Z (7 *i cos <* 7 4 sin w) (1) ttotor: o == ZJ, + Zi (7 7 cos ") + jaZ (1 sin > J 4 ) (2) STEIN METZ: ALTERNATING-CURRENT MOTORS. ,309 Short-circuit under brush : o = ZJ 4 -{-Z 4 (J 4 J sin <*>)-}- jaZ (J x J cos CM) (3) hence : J 4 = A | TO (sin ut +ja cos CM) ja Ji f (4) and, substituting (4) in (2) : 7 T (cos a- ja sin CM) a A (a cos CM ./sin /M) (5) *i = lo* - -f 7. a A Z substituting (4) and (5) in (1) : Z Zo -h Zi ~h Z in w (iu w +/^ cos w) (l A) (G) or, denoting: D = Z" + Z^ -{- Z" sin CM (sin CM 4- /# cos <) (1 ^) (7) Primary or main current: /o= ~TJ>~~ (8) Secondary current: / _ jg \ ( cos CM - ja sin co> a A (a cos CM j sin CM) } (9) ~~D~ Short-circuit current under brush : sin CM E.m.f. of rotation of main armature circuit: E 1 1 = jaZ (7 sin w J 4 ) E.m.f. of rotation of short-circuit under brush: ES = jaZ (A J cos ) = a ^ ] a ^( 1 A ) sin j" 1 ^ V Cll) = ^^f intt> appro*. j The power output is : P=/SA/ 1 /' + /# 4 M 4 /i _ a ^ 2 sin ut i , j ~TZ)j 1 / - ; ]Zi+ Z ( 1 A) j- , (cos w 7^ a sin w ) a A (a cos w Jsin ai) 7 1 + a (1 a 2 ) sin a> /Z, ^ (1 ELEC. RYS. - 24. 370 8TEINMETZ: ALTERNATING-CURRENT MOTORS. cos _a (r+r t ) sin * ^ (x cos w=ra (Z a 2 ) sin ca ) >1 2 (1 a 2 ) (r cos & a# sin o>) ! (12) As seen, in the repulsion motor, sin > takes the place of c, the ratio of field exciting turns to armature turns, of the series motor. I z sin at is the field exciting or magnetizing, J cos ^ the compensat- ing circuit. , At synchronism : a = 1, it is : I t = 0, that is, at synchronous speed, the short-circuit current under the commutator brushes of the repulsion motor is zero, and the com- mutation perfect. It is then : Z Zo +Zi+ Zsin w (sin (8in at -\-j cos to) and, absolute: *o = and the power factor: p=C08 (a-|- o -90) where : tan a = r The secondary current is: j 8 * n (sinw -f jcosw) or absolute: ^ ' - (17) * 8in hence, at synchronism the secondary current equals the primary current, and leads it by angle at. The power is, at synchronism, approximately: cos* -(r + rOsina. (18) that is, the effect of the short-circuit under the brushes disappeared. Since the repulsion motor contains the factor (1 a 2 ) in the short-circuit current under the brush, I 4 , which does not appear in the series motor, within the range where this factor is small, STEINMETZ: ALTERNATING-CURRENT MOTORS. 371 that is, near synchronism and below synchronism, the short-circuit current is less, and the commutation, other things being equal, better in the repulsion than in the series motor. Considerably above synchronism, however, where [1 & 2 ]>1, the short-circuit cur- rent of the repulsion motor becomes large, and the commutation inferior to that of the series motor. Thus the repulsion motor is specially suited for the range of speed from standstill up to some- what above synchronism, where the plain series motor is unsuitable by its lower power factor. Neglecting the effect of commutation, it is: I - E Zo + Zi + Z sin at (sin at -\-ja cos to) (19) ~ Z Zi + Z j E (cos at -j- fa sin ">) (20) ZQ-\- Z\ -\-Z sin a* ^sin at ~K/a 8 w ) or approximately: j o= E (21) Z + Z\ + Z sin at (sin at -\-ja cos a>) K *o + r i + r 8 i n2w -h a sin a* cos a) j (x -j- i + x sin 2 ^ ra sin w cos at) ( 22 ) or, absolute: = r^j/cos 2 u; + a 3 sin 2 > = - - |/i (1 a* ) sin 2 (24) hence, up to synchronism: a 1, the secondary current is greater than the primary and does the magnetizing of the motor field. The secondary current leads the primary current by the angle: tan 8 ==a tan at (25) The phase angle of the motor is, approximately, and neglecting the effect of commutation: tan = X ~^~ Xl ~^~ x 8 * p2 *" ar 8 * n w cos w r -|- r\ -\- r sin 2 Xv \ / \ ^ \ / ^ ^ [ A A l.O FIG. 20. 1.2 As instance are given, in Fig. 20, with the speed a as abscissas, the characteristic curves of a repulsion motor of the constants: e = 500 volts, Z = 1 10; ohms, Z Q = Z 1 =.l .3 ohms, Zi= 30 30; ohms, hence : =.18 .15;, m = 14 deg., or: sin & =.25; cos to =.97. 8TE1NMETZ: ALTERNATING-CURRENT MOTORS. 373 Hence : 478+13.8 in the case of primary and cos $ in the case of the secondary motor can only approach the value unity in the case of an ideal no-load operation. As regards the process of commutation, we shall investigate in addition to the so-called reactance voltage, also the e.m.f s. in- duced in the short-circuited turns by the fluxes F and F 2 . These e.m.f's. are usually of a higher order, quantitatively, than the re- actance voltage produced by the variation of contact between brush and commutator for the latter, as in the case of direct-cur- rent machines, is determined only by the stray field which is inter- 1 inked with the short-circuited turns, while the former are due to the total flux. The coil, short-circuited during commutation, is interlinked with the flux F 2 and at the same time cuts the field DERI: SINGLE-PHASE MOTORS. 381 flux FI in its densest zone with a speed corresponding to n. There are, therefore, induced in the coil two different e.m.f's. both of the same frequency: i. e., e\\ F 2 independent of the speed, proportional to F 2 , with the phase of which it is in quadrature ; and the e c : nF lt proportional to the speed and to F and in phase with F t . The diagram of these e.m.f s. can be derived directly from the e.m.f. triangle Fig. 3, as shown in Fig. 6. Let ab represent the measure of the e.m.f. e^ corresponding to F 2 maximum and for a certain speed n : cotg S. Draw the triangle abc. The phase direction of F 2 is for, ^ is at right angles to it and proportional to F 2 and hence determined by ^T in dimension and phase. The magnitude of e in the direction of F v is ^#, at the speed n for instance. 5 For the primary series motor ad l : -==- for the secondary ac ad 2 = = . The resultant e of the e.m.f s. is cd l or cd 2 re- ac spectively. As far as the reactance voltage e r is concerned, its phase co- incides with the current phase, being in the direction be, and quantitatively r : nl : E & (on a scale approximately 0.1 0.2 bo ) . From the diagram it appears that in the case of the primary motor, e r and e Q are always at right angles to each other, whereas ir. the case of the secondary motor the obtuse phase angle is variable and only at practicable speeds e r and e are opposed to one another. In the case of the secondary commutation we, there- fore, arrive at an advantageous compensation of the reactance. The result of all three e.m.f s. in the short-circuited winding, e measured by ^/ x or^ 2 respectively must be considered in the com- mutation as the cause of sparking. In Fig. 6 curves of e are shown for both methods of connecting the series motor. One sees that e in the case of the primary series motor (e^) deviates but little from its initial value, and increases at greater speeds. In the case of the secondary series motor, on the contrary e 2 falls considerably below its initial value and is a minimum at a speed which lies within the limits of the usual operating speeds. The latter, therefore, commutates considerably better than the primary motor in which good commutation is out of the question. We shall investigate in a similar way the repulsion motor in 382 DERI: SINGLE-PHASE MOTORS. Fig. 7. The stator winding is connected to the line and the brushes, inclined to the axis Y at an angle a, are short-circuited. The brush axis is indicated as X. We shall not proceed with the in- vestigation of the operation of this motor according to the method usually followed of resolving the fields, e.m.f's. and ampere-turns into components as functions of sine and cosine a, but we shall pro- coed in other ways which are simpler and unobjectionable. We may proceed in two ways: According to Fig. 8, we can divide the stator winding, which we may assume as uniformly distributed over the circumference of a closed stator, into two groups connected in series with the same current flowing through them and which exert and are subjected to the same effect as the com- bined system. The four groups of windings on the stator, as shown in the figure, are so connected as to produce, by the current flowing in all of them, the flux in the direction of the arrow, and to gmerate the field with the axis Y. In the distribution of the lines of force and the amount of magnetizing current, there must be taken into account, in addition to the magnetic resistance of the entire flux, the reaction of that part of the armature circuit which is short-circuited between the brushes. None of the effects suffer any change if we consider, as connected in the order shown, the ampere-turns / whose axis coincides with X, and the ampere-turns II, whose axis is perpendicular to X. Only the sequence of the single elements in the series has been changed, which is without any importance on the result. We have, therefore, two stator windings 7 and 11, in general with different number of turns and also of different magnetic resistance, and the arc covered by. the windings and the polar arc are unequal. The axis of the two windings are perpendicular to each other. We can, therefore, construct the diagram Fig. 9. DERI: SINGLE-PHASE MOTORS. 383 We can proceed in another way in accordance with Fig. 10. in accordance therewith, the armature winding is divided into four groups which are connected in pairs 7 and /, then II and II in parallel, and the two pairs connected in series are closed on Fig. 10 Fig. 9 themselves. The division is such that the groups connected to- gether when traversed by the current produce the same magnetic field in the direction of the arrows and corresponding with the brush axis X, as would the entire armature winding when short- circuited through the brushes. It appears, therefore, that the armature phenomena are as follows: In winding I an e.m.f is Fig. 11 induced by the coaxial stator winding, which e.m.f. is equal to that of the entire armature; winding I forms a circuit closed through winding II, hence excites a flux in the latter which is free from any armature reaction. The diagram of this arrange- ment, in accordance with the foregoing analysis, is shown in Fig. 11. This diagram differs from the arrangement shown in Fig. 2 in that the winding exciting the field is not on the stator but on the 3S4 DERI: SINGLE-PHASE MOTORS. lotor. WhetHer the armature rotates in a field excited by fixed stator windings or by rotor windings amounts to the same inducing effect, provided only that in the latter case the axis of the exciting windings is held fixed by brushes. According to the first arrangement the coils I are the inducing v;inding, and according to the second they are the induced winding ; the coils 11, however, on account of being at right angles tc X or Y, respectively, can neither exert nor receive any induction. The ratio of the number of turns is - : ^ ; the ratio of the Za a poles faces the reciprocal ___^i__. This alone sufficiently indi- 90 _ a cates the importance of the brush position, represented by the angle and its effect on the flux, torque, etc. It appears also that by a variation of this angle by turning the brushes, all of the secondary quantities and their functions can be varied. In Fig. 12 the armature current I in amount and phase is de- termined graphically by the following considerations: abc' is the diagram of the primary pressures, ^ the terminal pressure, ^' the e.m.f. of the armature and also the pressure of the stator winding I, and ac 1 that of the stator winding II. F^ is proportional to ^c 7 and in the direction JJJ 1 ; the same is true of i. The component of the armature e.m.f. which, induced by the com- ponents F 2 , belonging to the exciting current, i t is proportional to gJS F% and to n : and can be measured by C( /. Then fr c is the etc component of the armature e.m.f. belonging to the exciting current / and, therefore, also the direction of 7. The angle

. The phase shifting which takes place in the transformer 7 is to a certain extent balanced by the compensation. Cos will be about the same as in the case of the primary series motor and will, therefore, approach closer to the maximum value than in the case of the secondary series motor. The commutation phenomena in the case of the repulsion motor can be represented by consideration similar to the foregoing, taking into account, however, that the e.m.fs. ^ and e o are induced by those fluxes, which, according to the analysis of Fig. 9, correspond to the number of turns Z^ and Z 2 . Therefore Fig. 14 shows also on the proper scale the diagram of e.m.fs. in the short-circuited winding. In amount and phase direction e^ ^ e c : ad an ^ e o : 7w; ao i g > therefore, the e.m.f., e, resultant of all three. ad is proportional to the speed. We can, therefore, project the values of c as ordinates upon n. The curve shows the dependence of e upon n. e is, therefore, a maximum at starting (just as in the case of the series motor e:ab), diminishes rapidly, however, with increasing speed and reaches at a certain speed a minimum, which is less than the reactance voltage. The compensating e.m.f. of the repulsion motor depends upon the ratio --, hence upon O a _. On the other hand, the torque and energy of this motor is inversely proportional to the number of turns, i. e., inversely proportional to Z 2 or to higher powers of Z. 2 . Herein lies the weakness of the ordinary repulsion motor. It is not at all sufficient, as was originally believed, to shift the brushes by 45 deg. (i. e., one quarter of the polar distance), which would 17 correspond to the ratio 1=1. The e.m.f. induced while at rest DERI: UINGLti-PHAtiE MOTORS. 387 in the winding / would be barely sufficient to excite the field F 2 =Fi with the current I = i. The maximum power current would be t, hence the starting power too small and the output insufficient. Consequently it is necessary to make the angle a much less than 45 deg. in order to obtain sufficient torque. In re- ality the angle is chosen at about 25 deg. to 20 deg., corresponding yr to a value =0.40 to 0.30. In order, therefore, to obtain a Zi greater output one sacrifices a part of the compensating effect. The repulsion motor is, nevertheless, a very useful machine, particularly if the windings are carried out in two parts, as shown in the diagram, Fig. 9. WJth the aid of a switching arrangement, reversal of direction and control is easily made by the inversion and variation of the field of force. Another and more convenient method for the reversal and control within the widest limits con- g sists in varying the ratio -? by shifting the brushes. In order to Zi obtain the maximum output, it would be necessary to make so small that it would embrace only two to three commutator bars, which would, however, make the motor unreliable in operation and commutation. On this account the output of the repulsion motor is limited; in other words, dependent upon the number of the commutator bars and the size of the commutator. We will now refer to an arrangement devised by the author ac- cording to Fig. 15, in which all of the characteristics of the re- pulsion motor are left substantially undisturbed, permitting, how- ever, the angle a to be made twice as large as with the usual arrange- ment. One pair of brushes is placed in the Y axis and another pair at the angle in the V axis. The two pairs of brushes are con- nected as shown, so that they embrace the obtuse / angle (180 ). The effect of this arrangement can be judged if one imagines an armature in the Y axis connected, as shown in series and in closed circuit with another armature in the V axis. The ratio off the & number of turns ? in accordance with the diagrammatic analysis is in this case 7^-^ If this ratio and the resulting output are loO * to be of the same v The brushes have to carry the same load in this arrangement as 388 DERI: SINGLE-PHASE MOTORS. in the case of the ordinary repulsion motor with the same power current and cross-section of all brushes. The number of brushes need not be increased if with the proper winding of the armature fewer brushes are used than the number of poles, .for instance, for eight poles, two positive and two negative poles have two brushes each. This arrangement is particularly adapted to controlling by brush shifting, perhaps by shifting the V brushes alone. Accord- ing to the above presentation a compensating effect is obtained either by exciting the field by primary current, one component of which is the magnetizing current for the flux F 9 or by placing the windings which excite the field and wljich are a part of the main circuit on the armature and subjecting them to induction by F r Both of these causes of the compensating effect are contained in the arrangement of Fig. 16, which shows the so-called compen- Flg. 16 Fig. 17 sated motor of the Union Elektrizitats-Gesellschaft. The brushes in the axis / are short-circuited. The diagonal brushes in this axis II are traversed, on account of the series connection with the stator, by the main current, either directly or at a transformed potential through the insertion of a transformer. The armature carries the power current in short-circuit and rotates in the field of force excited by the primary current in the armature betAveen the diag- onal brushes. . The actions in the two axes of the armature do not interfere with one another notwithstanding that they occur in a common winding, because the axes lines are in neutral positions relatively to each other. Considering that the axes are held fast in the armature by the brushes, the arrangement can also be represented by the diagram of Fig. 17. The diagram of the working current is similar to that of the repulsion motor. DERI: SINGLE-PHASE MOTORS. 380 Fig. 18 shows the polygon of e.m.f s. abcda with relation to the working current. The e.m.f. of the armature coil I, which is re- flected in the stator winding, is made up of E & :bc" and E :ak. These e.m.f's. are generated by the rotation of the coil in the two components of the field F 2 , one of which is excited by I, the other by i. The e.m.f., E n : kd, is induced in the armature coil II on account of its rotation in the field F^ ; on the other hand, E m : ac" is consumed in exciting the field of force and the stray field. E and E n have the direction ad; db and bd form the angle ?. In order to construct the diagram, with relation to the working /%\ current, ad is drawn at right angles to ab (ad: tan r :n \~7~) <*&) On bd as the total useful e.m.f., the work polygon is constructed for n : cot d , in which all quantities ad, be and cd are referred to the phase and amount of the working current I. Not only is the phase displacement and the drop in voltage caused by the trans- formation, compensated for in this way, but the primary diagram also receives a favorable displacement. On the assumption of the direct connection in series between the stator winding and exciting winding, ok and led are equal in amount, ad : tan y is, therefore, twice as large, and the compen- sating effect more powerful in the same proportion than in the repulsion motor, be' shows the direction of the primary current; ac' is the measure of the exciting e.m.f. to be supplied externally. By the insertion of the series transformer the compensation can <7 be varied together with the ratio Still greater is the variation if 390 DERI: SINGLE-PHASE MOTORS. with the aid of a potential regulator the number of turns in both be varied simultaneously in opposite directions. The example in Fig. 19 shows the stronger influence of ad upon cos d, T and P. One can see that cos

, which is greater by (^i-hr) than the angle previously referred to. The power factor will be poor between rest and a speed slightly under synchronism. The motor can, therefore, operate advantageously only at nearly syn- chronous speed, but even then only with limited output, because at this speed a small e.m.f. ~^ e remains, and the working current is, therefore, weak. The triangle a 1 frc 1 shows the work diagram of the induction motor and contains the quantities T, P and cos y^), which is measured between the exciter brushes is practically small and can also become zero. T:(l + n 2 )! 2 ; P: n(i -f- ft 2 )/ 2 and This kind of motor is far superior to all the commutator motors previously described as regards commutation. Brushes carrying main working currents are not used; therefore, all commutator difficulties at starting and at low speeds disappear entirely. For this reason the motor becomes nearly independent of the armature current. Therefore, from this standpoint machines can be built in units of any desired size quite the same as with polyphase motors. The commutation at the exciting brushes is, as previously shown, DERI: SINGLE-PHASE MOTORS. 39o very smooth, particularly in this case, because the turns in closed circuit under the brushes are really parts of closed circuits and because the currents which flow through the brushes have only to furnish a small part of the excitation; they need, therefore, be comparatively small for this reason and especially because the brush e.m.f. ~^ e will be very small at normal speeds. The commutator for this reason can be made much narrower, an advantage which is important in connection with the fitting into car bodies; of equal advantage is the possibility of using a com- paratively small number of brushes. In principle, this motor is an induction motor which transmits external energy by means of transformation to a simple rotor; the torque and speed can never- theless vary, and any variation can be produced which is neces- sary for the control of vehicles or cars. The motor can be started with considerable torque, and its power factor can be made nearly unity. By this combination we have an externally excited, com- pensated induction motor. ALTERNATING-CURRENT MACHINES WITH GRAMME COMMUTATORS. BY MARIUS LATOUR, Delegate of Societe Internationale des Electriciens. If we refer to the technical literature of four or five years ago we shall notice that the problems, the solution of which was sought by engineers applying themselves to the study of alternating-cur- rent machinery, were : 1). The development of alternators in which all difficulties in parallel running should be done away with. 2). The construction of generators of constant voltage. 3). The construction of motors working with a good efficiency at all speeds, and starting under load with single-phase current. 4). The construction of non-synchronous motors working with a power factor equal to unity. It was about that time that I began to take an interest in the ap- plication of the Gramme commutator to alternating-current ma- chines. I have thus been led to a new system of electrical machinery which might take the place, either as generators or motors, of the machines used nowadays and solve, from the technical point of view, all the problems set forth. The description of alternating-current machines comprising a Gramme commutator is comparatively old. Indeed, it is to be traced back to Messrs. Elihu Thomson, Wightman, and Wilson. However, owing partly to the essential phenomena exhibited in direct-current armatures traversed by alternating currents not being very well known, partly to the little interest taken by electricians in these phenomena, partly to the bad opinion that had been formed of the Gramme commutator used in connection with alternating- currents, the arrangements proposed by those inventors were left without much industrial value. I soon realized that the use of the Gramme commutator with alternating currents was full of capabilities and I have been able to realize the machines concerning which I shall say a few words. All these machines have a uniform appearance, due to their comprising a stator with a winding distributed in slots, and a rotor [39GJ LATOUR: ALTERNATING-CURRENT MACHINES. :W7 similar to a direct-current armature with a commutator. These machines are: 1). The panchronous self-exciting polyphase generator. 2). The panchronous self-exciting single-phase generator. 3). The polyphase motor at variable speed with a power factor equal to unity. 4). The compensated single-phase motor. 5). The repulsion motor. 6). The single-phase series motor with perfect commutation. 1. The Self -Exciting Polyphase Generator. This generator is represented in its two most interesting shapes by Figs. 1 and 2, the former showing the shunt connection, and the latter the compound connection. S is the stator, R the rotor with commutator, and t the transformer for supplying the rotor. Such generators are connected to a network like direct-current generators, without any synchronizing operation. The compounded alternator, when well regulated, works at constant voltage whatever may be the inductive or non-inductive load on each phase separately. 2. The Self -Exciting Single-Phase Generator. This generator is represented in its two most interesting shapes by Figs, li and 2 X . In order to allow ' self-excitation two sets of brushes c d are short-circuited on one another. In reality, for the sake of commutation, it is preferable to have several sets of brushes short-circuited on one another, as represented by Fig. 3. The self -exciting single-phase generator has, above the ordinary generator, besides the two advantages regarding the easier parallel running and the perfect compounding, that of admitting a perfect rotary field without any harmonic field likely to weaken the efficiency. 3. The Polyphase Motor at Variable Speed with a Power Factor Equal to Unity. This motor corresponds to Fig. 1 (representing the shunt con- nection of the panchronous generator), the transformation ratio of the transformer t being supposed to be arbitrary and variable. The speed of the motor may be regulated by changing the transforma- tion ratio of the transformer. The power put into play in the trans- former is the larger, the greater the slip of the motor. Such a motor may work with a power factor equal to unity at normal speed. 398 LATOUR: ALTEJtNATlXQ-UURlMNT MACHINES. f!fl.4 Fi fl .6 LATOUR: ALTERNATING-CURRENT MACHINES. 39tt 4. The Compensated Single-Phase Series Motor. This type of motor is represented under its two forms by Figs. 4 and 5. Such a motor works at every speed with good efficiency. The power factor is equal to unity at normal speed, and magnetizing current may be delivered to the network, if desired. 5. The Repulsion Motor, the Stator of Which Has a Distributed Winding. Flo. Fig. 7 This type of motor, represented by Fig. 6, has the general char- acteristics of the compensated series motor, but its power factor is lower and the leakage has in this motor a much worse influence. All the machines of which I have just spoken have a common property regarding the commutation, viz., that if they are properly designed they work with a perfect commutation in the vicinity of synchronism, owing to the existence or the formation of a perfect rotary field. This property, which I have demonstrated for each machine successively, although at first questioned, is now recognized. Let us consider (Fig. 7) a direct-current armature which is re- volving under the action of a rotary field in a stator like that of an induction motor. The rotary field may be excited partly or completely, either from the stator or from the direct-current arma- ture itself if this is traversed by alternating current. Let us consider a section s, which is short-circuited under a brush a. The revolving field may be considered as the resultant of two alternating fields, the first one sin cot in a direction per- pendicular to o a; the second one cos cot in the direction o a itself. 400 LATOUR: ALTERNATING-CURRENT MACHINES. Now the section s is the seat of two e.m.f.s. The first one is pro- duced in a static way by the variation of the field, < sin cot, and is equal to e% = GOI cos GO t, if the armature is revolving at the angular speed GO^ These two e.m.fs. are opposite, and at synchronism (GO I = GO) coun- terbalance one another. The section s not being any longer the seat of any resultant e.m.f., the commutation under the brush a will be perfect, whatever the current under this brush may be. This consideration leads easily to the conception of a device for avoiding sparking in the straight single-phase series motor. I wrote a paper upon this device a few years ago, and Mr. Maurice Milch, working independently on the same line, has reached the same result. 6. The single-phase series motor with perfect commutation. We shall consider at first a single-phase series motor, the field of which is wound like a direct-current armature (Fig. 8). The motor Ba.8 being operated with continuous current, in order to obtain the best commutation the brushes must be located so that the resultant field of the motor sin cot is perpendicular to the line a b. When operated with single-phase current, if the induction is low enough, the power factor of the motor will be pretty high. In order to reverse the direction of rotation of the motor without shifting the LATOUR: ALTERNATING-CURRENT MACHINES. 401 brushes, it will be possible to change the position of the terminals A B on the periphery of the stator. Such a construction represents the best it is possible to obtain with the straight single-phase series motor, as I pointed out two years ago at a time when polar projections were still used. But for the single-phase series motor there is no speed for which a perfect commutation is secured. The variation of the field of the motor induces at any speed an e.m.f. in the short-circuited sec- tions, and owing to this very important reason, I think the series motor is. for larger capacities, inferior to the repulsion motor and to the compensated type. Yet we can improve it in this way : An auxiliary field '/'' cos GO t is produced above the short-circuited sections, which field lags 90 deg. behind the main field, < sin (&t, of the motor. (See Fig. 9.) Conforming with the explanation I have given above, it is easy to see that the new e.m.f. induced in the short-circuited sections in a dynamic way by the movement of these sections under the auxiliary field V f cos GO 1, may counterbalance at a certain speed the e.m.f. induced in a static way by the variation of the main field < sin GO t. The auxiliary field may be excited with special coils c c shunt connected to the motor, these coils encompassing only a few slots. We realize now that four types of motors are possible for single- phase traction : The repulsion type, the straight series motor, the compensated type, and the type with an auxiliary field. The future will decide which is the best. CHAIRMAN DUNCAN: The next paper will be on "The Theory and Operation of Repulsion Motors," by Mr. Bragstad, and will be abstracted by Mr. Steinmetz. ELEC. RYS. 26, SINGLE-PHASE RAILWAY MOTOES. BY FRIEDR1CH E1CHBERG. The standard direct-current railway has probably been developed to its final stage. The combination of alternating current for the transmission of power, rotary converters for the conversion into direct-current, and direct-current car motors, is not, however, an economical solution except in rare cases. Recognizing this fact, Brown & Boveri (Burgdorf-Thun) and Ganz & Company (Valtelina line) took up the direct application of polyphase alternating cur- rents. But even if the polyphase system has achieved practical success in special cases, it has not been proven thereby that the polyphase motor furnishes a universal solution of the electric rail- way problem. It is not necessary here to repeat all the objections that European and American engineers have brought forward in numerous discussions against the polyphase motor. The multiple trolley for the collection of current, which is unavoidable in the polyphase system, leads to complications in the overhead work and sets narrow limits to the line voltage available. For short roads (lines between neighboring cities) the polyphase system, moreover, leads to excessive cost in the installation of the conduct- ing system. Add to this that the polyphase motor, by reason of its characteristic speed-curve, which resembles that of a shunt- wound motor, is almost or quite unfit far railway purposes. It cannot be disputed that it is possible to operate on schedule time upon special lines with a favorable profile but this proves nothing as to the general applicability of the polyphase motor. For two years, as is well known, efforts have been made to apply the single-phase motor to railway purposes. B. J. Arnold, with his electro-pneumatic system and the Oerlikon Company with the Ward-Leonard system, offered only incomplete solutions of the problem of applying single-phase current to railways. The first announcement of the direct application of single-phase motors came from Lamme, of Pittsburg, and was followed soon after by the publication of Finzi in Milan. The former used a frequency of 16, and the latter 18 cycles per second. Both have built series [402] EICHBERG: SINGLE-PHASE RAILWAY MOTORS. 403 motors similar to the direct-current series motor. The former uses, for the compensation of the armature reaction, short-circuited windings, which are applied in the field-magnet coils and whose axis coincides with that of the brushes ; the latter uses slots in the poles for the diminution of the armature reaction. Later the work of G. Winter (see Elektrotechnische Zeitschrift, 1904, No. 4), of Vienna, became known to the writer. This fur- nished the basis of the system worked out by the Union, and especially by the Allgemeine Elektricitats-Gesellschaft. This sys- tem, which forms the subject of this paper, has been put into operation on the Niederschoneweide-Spindlersfeld line under the management of the Koyal Prussian State Kailway, and on the Stubaital line near Innsbruck, which was opened on July 31, 1904. The first line operates with 6000 volts and 25 cycles, and the second with 2350 volts and 42 cycles. In perfecting this single-phase system, the motor of course played the chief part. In a lesser degree the controlling apparatus and those devices which become necessary in the direct application of high tension to the car were also of importance. In regard to the motor of the Winter-Eichberg system, it unites the properties of the ordinary alternating-current series motor with those of the repulsion motor. Its characteristic features are the following: FIG. 1. In the motor, in addition to its own magnetic field (F), there is developed as in the repulsion motor, a cross-field, which, at synchronism, is about as strong as the magnetic field F, from which it differs in phase by 90 deg. This means that when the motor is near synchronism a complete rotary field is established, the field being less developed below or above synchronous speeds. On account of the cross-field developed in the motor, the short-cir- cuited e.m.f. under the brushes diminishes with increasing speed, becomes nearly zero at synchronism and then increases again with increasing speed. 404 EICHBERG: SINGLE-PHASE RAILWAY MOTORS. In regard to armature voltage, these motors are essentially similar to the ordinary series motor. In both the tension per com- mutator segment may not exceed a certain value and, according to the size of the motor, the armature voltage will therefore lie between 100 and 200 volts. In the ordinary series-motor, in which the working voltage appears in the armature, the working voltage would therefore not exceed 200 volts. It is otherwise with our motor. Since the armature is short-circuited along the working axis and the working voltage appears only in the stator field wind- ings, the voltage supplied to the motor may be as great as desired. But even for the case where the excitation is inserted in series with the stator winding (Fig. 2), the entire working voltage (E) is in the same proportion to the armature voltage (e) at rest as the entire volt-ampere input is to the volt-amperes for magnetization at rest. (150 Volts) Ed- (450Volt) Volts FIGS. 2 AND a Let us suppose that the magnetizing current is one-third of the armature current, which is a good practical mean; then the work- ing voltage in the motor of our system, even with the direct intro- duction of the excitation, is three times as high as in the ordinary series motor. Through the insertion of a small transformer (Fig. 3) one can increase at will the proportion of the working voltage to the armature voltage without great expense (Figs. 2 and 3). The excitation by means of the armature in combination with the cross-field yields an e.m.f. which is 90 deg. ahead of the w.ork- mg e.m.f. and directly opposite to the e.m.f. of self-induction. This wattless counter e.m.f. gives the motor the well-known rapidly rising cos< curve. (See EUJctrotechnische Zeitschrift, 1904, No. 4.) Our first 100-hp motor had, with a 3-mm air-space on each side, a power factor of 0.9 even at 70 per cent of synchronism. Even more important is the fact that this good power factor is obtained with a number of ampere-turns per cm almost twice as ElCHBtiltU: SINGLE-PHASE RAILWAY MOTORS. 405 great as in the ordinary alternating-current motors. From this results the possibility of building a very powerful motor for a given armature diameter and external dimensions. Another characteristic property of our system is that the field can be controlled independently of the voltage in the working windings. In every alternating- cur rent commutator motor there are magnetic losses in the coils short-circuited by the brushes. Through the possibility of adjusting this field in proportion to the stator current, one can keep these losses under the brushes within Fuch limits as will permit the commutator and the brush-holders easily to conduct away the resulting heat. By varying the field one can, for a given working voltage, give the motor a variable characteristic. The separate characteristic curves will then be somewhat related as the curves of a 3-, 4-, 5- and 6-winding motor. Control independent of this naturally is possible and also control of the load voltage. The accompanying diagrams give examples of the control as carried out in practical cases. In Fig. 4 the primary voltage is not regulated and only the secondary winding of the exciting transformer is altered. rAAAA/V WW\A j FIGS. 4 AND 5. The absence of any primary regulation in the high-tension cir- cuit offers the special advantage that only low-tension circuits will have to be opened or closed when the car is to start, reverse or alter its speed. A still more complete solution is shown in Fig. o, the stator circuit as well as the exciter circuit being regulated. This method of connection is less advantageously applied to high- tension motors, because the high-tension circuits generally can not be readily altered in operation. The third diagram (Figs. 6a and 6b) shows a method of con- trol which, although not quite so complete as that of Fig. 5, is yet of value for small low-tension cars, and which will shortly bo put into operation on a short Belgian road. The scheme of 406 EICHBERG: SINGLE-PHASE RAILWAY MOTORS. Fig. 5, with the modification represented in Fig. 7, was installed on the Stubaital line near Innsbruck, now in operation, which is at times operated at 2350 volts and at other times at 400 volts. The direct insertion of the excitation in the stator circuit (Fig. 2) in which the control is effected by ohmic or inductive resistances Volts 120 VolW Volts FIGS. 6A AND GB. 120 VolM with the eventual application of series-parallel regulation, is pos- sible for small cars, and hence chiefly applicable to short railway lines. In the latter case the motors can be built simply for 550 volts. Motors for 550 volts connected in this manner are already in operation, and will also run on direct-current lines. (Figs. 7a and 7b.) /WWWNAAAAj V 2850 Volts FIGS. TA AND 7s. The possibility of running an alternating-current motor with direct current is of great importance in practical application. Ordinary series commutator motors, which are built on the com- pensated system of Deri, can of course be run both on direct- current and alternating-current circuits. The voltage for the direct-current motor is l 1 /^ to 2 times higher than the armature EICHBERG: SINGLE-PHASE RAILWAY MOTORS. 407 voltage with alternating current. Since, as we have shown above, the alternating-current voltage, which in our motor system can be directly applied, may be three times greater than the armature voltage, the ratio of direct-current to alternating-current voltage will be, not as in the series motor 3:1^ or 3 :2, but 3:3 to 1% or 2; that is, the direct-current voltage will be about half that of the alternating-current voltage. This allows, for example, the running with direct currents with motors connected in series, and with alternating current with the motors connected in parallel, but in the former case at less speed. This corresponds to the case in common practice where cars which travel over interurban stretches at high speed transfer to the direct-current systems of cities, where lower speed is demanded. There are various ways of running, with a motor connected according to Fig. 1, on direct- current circuits. The method which has proved itself most prac- tical is represented in Fig. 8. In the direction of the diameter of the exciter axis a winding is applied which counteracts and opposes the armature ampere-turns. The stator field windings then produce a magnetic field with direct current, and the exciting windings on the armature represent with direct current the work- ing ampere windings. The field saturation in direct current working is then somewhat greater than with alternating current, while the density in the armature is somewhat less. These prop- erties are extraordinarily favorable for practical operation. The auxiliary winding (h) is inserted only when operating with direct current. In order to make better use of the armature, the field windings of two motors can be connected in parallel in the direct- current circuit, while in order to be able to operate at 500-550 volts with direct currents, the armatures can be connected in series. These conditions are represented in Fig. 9. In the alternating- current system, one can operate according to the method of either Fig. 3 or Fig. G. If operated according to plan 3, then with alter- nating currents the connections 2, 4, 5 are closed and 1 and 3 are open. With direct current, 1 and 3 are closed, and 2, 4 and 5 are open. The motor system which I have above briefly described will not be the only one in the field. I can not, however, undertake to pass an unbiased opinion upon the different systems possible. I can only briefly mention the reasons why, in my opinion, the alternating-current commutator motor, which has been long known in two general types, namely, the ordinary series motor and the 408 EIGHBERG: SINGLE-PHASE RAILWAY MOTORS. repulsion motor, is not to be considered of equal value to the system above described. The ordinary series motor possesses, even if it is compensated, no cross-field; and it has no rotary field. The short-circuit losses under the brushes do not decrease with increasing speed, and the power-factor increases much slower with the speed. The maximum working voltage for which it can be built is 200 volts. When a short circuit takes place in the field winding of the series motor, the motor becomes inoperative. Multipolar machines with series windings on the armature, if pro- vided with the device shown in Fig. 1, can have an entire field coil short-circuited without the motor becoming inoperative. The separate field coils behave like transformers inserted in series. Any one of these can always be short-circuited; the others then receive correspondingly more voltage. FIGS. 8 AND 9. The repulsion motor when contrasted with the arrangement of Fig. 1 has the disadvantage that its reversal is possible only by the application of a second field-winding, or of several sets of brushes, or of reversible brushes. Its power-factor is poorer, and for its control there remains only either the method of primary voltage control, the opening of short circuits, or finally of brush reversal. The disadvantage of the type represented in Fig. 1, as com- pared with the series and repulsion motor, consists in the employ- ment of two exciter brushes, which doubles the number of brushes in multipolar systems. These exciter brushes give rise to no diffi- culties with respect to short-circuit losses; as I have shown (Elektrische Bdknen, Vol. 2, 1904), these short-circuit losses do not occur with exciter brushes. They carry moreover only one- EWE BERG: SINGLE-PHASE RAILWAY MOTORS. 409 third to one-fourth of the entire short-circuit current. On the other hand, the motor of Fig. 1, as compared with the compensated series motor and the repulsion motor with the double field-winding, offers the constructional advantage of only one field phase, which guarantees good economy and great simplicity. In high-tension motors, the increased certainty of operation in consequence of the absence of cross windings must be considered. Motors for either direct-current or alternating-current working provided with the auxiliary winding (h), which plays the part of the compensation winding of the compensated series motor, can therefore only be operated advantageously with low-tension alternating currents. The results of more than a year's operation on the 6000-volt Niederschoneweide-Spindlersfeld line, on which during a great part of the day four 100-hp motors haul a 160- to 170-ton train, and on which daily two motors handle a 100-ton train, prove that the alternating-current motor is adapted to the heaviest traffic. Moreover, the direct application of 6000 volts to the car has been demonstrated to be entirely safe. The Stubaital line, which has been running since July 31, 1904, at 42 cycles and 2350 volts, has introduced an advanced practice for small roads, an advance which exceeds the boldest expectations of the year 1902. At that time it seemed as though only very low frequencies could be used. In the case of many roads running in connection with existing power stations operating with 40-50 cycles, the possibility of using these frequencies limits the availa- bility of alternating-current traction. Moreover, the possibility of operating also with direct current makes the alternating-current commutator motor in a certain sense a universal motor, and places it, as regards its main features, far above the direct-current com- mutator motor, which really represents only a special case of the alternating-current commutator motor. THEORY AND OPERATION OF THE REPULSION MOTOR. BY 0. S. BRAGSTAD. Commutator motors for alternating current have become of great interest within recent years. The main reason for this is the demand for a motor for single-phase alternating current for the operation of electric railways ; but such motors will also find a broad field for other purposes where speed regulation is required. Of special interest in the older forms of alternating commutator motors is the repulsion motor, important in itself as also in that it marks a transition to the different forms of compensated motors. In the following I will develop a general theory of the repulsion motor, and that under the usual assumptions that the magnetic resistance is constant for all magnetic circuits, and that the iron losses are proportional to the square of the induction. We will not consider the processes under the commutator brushes and in the armature coils short-circuited by the brushes. PRINCIPAL EQUATION. Fig. 1 shows the diagram of the motor. W g is the stator and W T the rotor winding. The angle of displacement relative to the shaft, Y Y , of the stator winding is . We further take : I s the stator current according to strength and phase, 1 T the rotor current according to strength and phase, EJ the induced e.m.f. according to strength and phase, in the stator winding, we get the following relation: 1 l )E's = - Za (/s + /r COS ) =/ m Z & . In this we assume the effective number of windings of the stator winding and of the rotor winding to be alike. Z&= r & jx A can be df3signated as the exciting impedance of the motor and is given through the magnetic resistance of the main power current of the same, whereby r a is so determined, that E'l ~2~T 2~ *& ~T~ 1. The period below the letter indicates that the value is a vector of determinable phase. [410| BRAGSTAD: REPULSION MOTOR. 411 equal to the iron effect. 7 m is the resulting current in the Y axis, or the magnetizing current in this axis. In the rotor winding there i*j induced between the brushes through the same power current: 1) . Statically E' r j = E' % cos a = Z a (/ + /r cos a) Za 7m COS a. 2). Dynamically through the rotation: .u r u . u . . T- fj' d as 1 Es 81D a = 9 Za (A + / r COS a) SID a = 7 Z ^ m SID a " c c " c where w is the number of periods of the rotor rotation. c, number of periods of the current, ;= 4/_ i. The prefixed sign depends on the direction of rotation chosen. The rotor current generates a power current in the X axis, perpendicularly to the axis of the stator winding. The e.m.f. in- duced by this power current is between the brushes : 1). Static 3) ^ / rx = -Za^r8m 2 a. 2). Dynamic, through the rotation 4) E'rxA =j ~ E'rx COS a - j Za fr SID a COS O. The prefixed sign must be the reverse in equation 4 to what it is in equation 2. From 2 and 4 follows E'ryd + #'rxd =J ^ Za /. sin O. 412 BRAGSTAD: REPULSION MOTOR. The field generated by the rotor current I r thus produces no dynamic e.m.f., which also follows from the fact that the same runs along the brushes. The entire e.m.f. induced in the rotor must be equal to the rotor current, l t times the impedance of the rotor winding, which we will designate by Z r . We, therefore, have Za 's COS a Z & IT COS 2 a a 7" r Sin 2 a -f -j Z* IB Sin a = I r Z r * C ' IT (Z a + Zr) f s (cos a U ~j sin a) Za l] f r = (cos a U -j sin aj J 8 II) E' 6 = - (Za - 1^ (cos 2 a -j si / . \ c and sn a cos a ) ) 7. / / where Zt= Z&~{- Zr> Let us designate by Z B the impedance of the stator winding; FIG. 2. then the corresponding e.m.f. is equal to IgZ s , and the entire e.m.f. in the stator winding 77T/ ET/ T /7 .o ^Ja s ^s //a Or if we introduce instead of the e.m.f. the terminal pressure, E = E' ) we have III) E= Is (z s + Z & |^ (cos 2 - %j sin cos )) . j^t From this we get for a given constant stator current / g = I s , the pressure diagram shown in Fig. 2. We assume the direction of BltAGSTAD: REPULSION MOTOR. 413 rotation of the vectors to be to the right and carry A = 7 g onto the vertical axis. If we further have we have -T 9 and BF = I n A A * whereby we make the triangle BFD similar to the triangle BCE. We determine a point G on the straight line FD, so that cos 2 a, draw a perpendicular line to DF at (r and make GU = DF sin cos . 7 then represents the terminal pressure E of c the motor for the respective number of revolutions u, and

* 2 a = oo = 45 deg. With an in- creasing speed a is somewhat decreased. At synchronism (u = c) we get 5 a ) tan With = 45 deg. the moment of starting is and the moment at synchronism The energy transformed into mechanical work is found simply by multiplying the moment of rotation by . It is thus V) H m = "ft = (l* y-) 2 (" ajt sin a cos a- (") ' r t sin 2 a) . According to equation III we have for the terminal pressure, if we put 1% =1* III ^=/ 8 (Zg + Za Z (cos 2 a ^; sin a c OB a) ) sn a cos = 7. (r g r oos" a + x sin a cos a ^ f j; g # cos* a ' u r flu a cos a c 410 BRAGSTAD: REPULSION MOTOR. -C ( r k H sin a cos a J lx k -- r sin a COP a\ \ . The following abbreviations have here been introduced: Z& 2 I \ -*-=Z s = s r jx (constant section FD in Fig. 21 Z* + Z & = r s + r & -j (aj g -f a; a ) *=Zg = r g jx g (constant section OZ>in Fig. 2) Z g Z cos 2 a = r g r cos 2 a j (x g x cos 2 a) = Z k =ry i j Xk (section OG in Fief. 2) ^ k is the impedance of the motor at standstill (short-circuit im- pedance), and is dependent on angle a. For a = o we have Z k =Z 3 + ^^- (section OF in Fig. 2.) Witji an increasing angle of displacement of the brushes, the ter- minal point of the vector Z k moves on the straight line FD from F to D. For a = 90 deg. we have Z^=Z 8 -\-Z & = Z s (section OD in Fig. 2). From equation V we can determine the number of revolutions w e for which the output of the motor becomes naught. The same is c r t sin a For this no-load point we get the phase displacement ? e by putting the above value of -1 into equation Ilia. We get Xt T tan Z e = r e jfy can be designated as the no-load impedance of the motor. We can also put 11) =Z K From this we immediately get the following construction of BRAGSTAD: REPULSION MOTOR. 417 (Fig. 3). We make OB = Z e and BE = Z t =r t jx t equal to the geometrical sum of BD and DE, whereby BD = Z & and DE =Z r . Z 2 Thereupon we draw the line DP = Z = -~ and set up at F, a ver- A tical FS. The angle FDS becomes equal to arcta/* and -y-,= y*j W .N cos 2 a. OT is then the vector of the no-load impedance. The termi- nal point of the vector of the short-circuit impedance is G and this, point is the projection of point T onto section Z =DF. For a = FIG. 3. 90 cleg., cos a = o, the no-load point 8 coincides with the short- circuit point G in point D. With the decrease of angle a, the no- load point moves toward S and the short-circuit point toward F. From equation V we find the maximum output of the motor for a constant current, if 12) u Xt sin a cos a 2 r t sirr a = o 1 cos a sin a " c 2 r t The maximum output with constant current, therefore, occurs when the motor has half of the no-load speed (see equation 9). This maximum output would be ELEC. RYS. 27. 418 BRAQSTAD: REPULSION MOTOR. If we introduce the number of revolutions for the maximum out- put into equation III a ), we have I * Xt 2 / I** 9 \ + X cos 2 a j (xk 5 r COS 2 a) 2 r t \ 2 r^ j Or in consideration of equation 10: 14) E= 7 8 ft (r k + r e ) -j (i (* k + x. ) )). This result also follows directly from the fact that the point for maximum output lies in the middle between the short-circuit point and the no-load point. The general expression for the electric power supplied to the motor is 15) W 9 = Jg 2 (r g r cos 2 a -j x sin a cos a j T 9 / , U ' 1 % 2 (r k -{- -x sin a cos a \ c j The degree of efficiency is thus jj 7 2 2 [Xt sin a cos a 5 r t sin 2 a ) 16) w^IlB^ ^ 9 W 2 / , u \ b I r k H * 8m a COS a 1 c / By differentiation we find from this the number of revolutions foi the maximum degree of efficiency , *t 2 _ __ C x sin a cos a x 2 sin 2 a cos 2 a r t X sill 2 a The negative prefix of the root would apply to the operation as generator. We here take the positive. The expression can be easily reduced to the following: 17) U _ ^^e y*k G x sin a cos a This expression for the most favorable number of revolutions in- troduced into equation 16 gives for the maximum degree of ef- ficiency the following value : , Q v 2a 2 (2 r k r t + x t x cos 2 a) Vr k r e 2 r t r k r 1O J Viaax I - ~ / z t 2 x> cos 2 a Vr k r e The values for the most favorable number of revolutions (17), and for the corresponding degree of efficiency (18) are, as is seen. independent of the chosen current strength and of the terminal pressure. If the current strength is assumed, we find the terminal pressure through the introduction of the respective number of BRAGSTAD: REPULSION MOTOR. 419 revolutions into equation III ft . We likewise obtain the output of the motor by introducing the number of revolutions into equa- tion V. OUTPUT DIAGRAM WITH CONSTANT TERMINAL PRESSURE. The mode of operation of the motor, assuming a constant current, can be quite plainly seen in the pressure diagram of Fig. 2. This diagram is useless, however, when it is a question of examining the action of the motor with a constant terminal pressure. We will now also develop a diagram for this case. In equation II I we assume the terminal pressure to be real, in that we put E= E but we allow the current vector to take any arbitrary phase. Equation III a then reads III b y E= 7 8 ^ k -|- _ x sin a co* a j \r r -- r -in a cos j j For the current Z g we will introduce the two components, the real one ^ parallel to e and the imaginary one C, perpendicular to e. We thus put h = *> +? ^t x bin a cos a= - ' - x c c x -\- rj r T If we introduce these two values into equation 19, we get: V a; + TI r) = (? 2 +r y 2 ) (r r k + a; k ) or vi) / 8 2= * 2+ ^ _ . l_^-hA_ r _ r r k + a; x k If we consider ? as abscissa and ^ as ordinate in a rectangular ct^-ordinate system, then this equation represents a circle. The 420 BRAGSTAD: REPULSION MOTOR. same goes through the initial point of the co-ordinates and has the radius The co-ordinates of the central point are x Abscissae w = E- 2 (rr k Ordinates M = With constant terminal pressure we thus have the current vector according to strength and phase as the distance from the initial point to a circle. By means of this circle we can follow the work in the motor. We will first consider the number of revolutions, or what is the same, the relation . c According to equation 19 a we have c sin a cos a ( x -f- f) r) sin a coft a ^ Here we have i? x k r k = Zj = o a straight line through the initial point of the co-ordinate system. According to equation 19, \ = with u = o. The straight line ^==0 thus goes through ^ the short-circuit point of the motor. The short-circuit current is K in Fig. 4. a; -}-^r = L 2 = 0is likewise a straight line, which stands perpendicularly on the connecting line (central line) between the origin of co-ordinates and the central point of the circle. This line is, therefore, a tangent at the initial point. Every vector is a ray in a group of rays, and has the equation u . sin a cos a 1^ L\ mm o. c For u = c the equation of the ray (synchronous line) is Zs = sin a cos a L% Z> } = t) (r sin a cos a a k ) + (a sin a cos a -f- r k ) = o. For any point P of the circle, is the double relation between the four rays and the same is cut off on a transverse line, thus dc BRAVtiTAD: REPULSION MOTOR. 421 If we draw the transverse line parallel to the straight line then the double relation passes over into the single relation and we get c We can, therefore, read on the transverse line parallel to the straight line L. 2 o the speed of rotation for every current. The synchronous line L 3 = o can be constructed as follows : Over the section FD = z as diameter we describe a circle and mark off the brush angle = GDT in point D. The synchronous line then runs through point T. A perpendicular line from T onto sec- tion FD cuts the latter in point G; OG is the short-circuit im- pedance of the motor with the two components r* = r g r cos 2 a and aj k = x g x cos 2 . The short-circuit line = o thus goes through point G. The short-circuit point K on the diagram circle is an inverse point to point G, in that OK is proportional to -Q-Q . A modification of the brush angle a changes the size of the diagram circle and the position of the short-circuit point K as also the posi- tion of the synchronous line 3 =0. If a is made so large, that the synchronous line becomes a tangent to the circle over FD f synchronous speed occurs with the smallest phase displacement. When G moves toward F, if, therefore, a is reduced, the short- circuit point K continually moves higher on the circle; the losses are thus increased. At the same time, however, the diameter of the diagram circle and therewith the maximum output of the motor are increased. The question as to the most favorable brush dis- placement angle can thus be answered in very many different ways according to the special result which is to be obtained. By means of the diagram in Fig. 4, we can easily find the most favorable brush position for a given case. We will now see how we can represent the moment of rotation in the diagram. According to equation IV D = ( ) J 8 2 (x t sin a cos a -- r t sin 2 a) . \ 2% ' * C * In accordance to the circle equation VI, we have however and according to equation 22 u 7 x* * r * _ L\ c sin a cos a ( x -\~ 7 **) sin cos 422 BRAGSTAD: REPULSION MOTOR. 23) Introducing this for the moment of rotation we get D- (Ji)'L* r , -^(eL+ji \ z t ' z * cos a \ \ * r t *) \Xk -- ~ COS 2 a) j \ -'*/ sn - cos a The straight line $ r e y x e = L^= o is the no-load line, the construction of which we have given in Fig. 3. M=C FIG. 4. The moment of rotation is at any point proportional to tHe dis- tance of the same from the no-load line. In short-circuit the cur- rent is (r g r cos 2 a) 2 + ( Xg x cos 2 a) 2 The starting moment is, therefore, / z \ 2 24) D k = ^ -j / k 2 C t sin o cos a (2 a \ 2 2 iC t sin a cos a 'v ov^ r coa 2 a) 2 -f (xg x cos 2 a) r BRAGSTA.D: REPULSION MOTOR. 423 Hereby the measure for the determination of the moment of rotation is given by the distance from the no-load line for every point. We will now introduce the relation between the moment of rotation in watts relative to synchronism and the introduced watts. This is 25) d= --= W e ~ 2 t / 2$ COS a At starting we have sin a cos It is seen that * is the double relation in a group of vectors con- sisting of the abscissae axis (170), the no-load line (L 4 =o) and the short-circuit line (L^=o). Because in this group of FIG. 5. vectors the double relation is given for the short-circuit line as equal d 9 the same can be at once determined for any vector 2 + X (Xe *k + Xk ^ k -f ^*,)-fi ^to+^ We have - x cos 8 Hence / k 2 e 9 and $ are watt and wattless current at maximum output. The output factor with this output is , 2 + (r e Z k + r k Z e ) The number of revolutions with maximum output is found as follows C su a cuS r BRAGSTAD: REPULSION MOTOR. 427 e Zk + r k g e ) - ?* k (Xe 3k sin a cos a [r 3k k r e - sin a cos a L*k (rr e -f- xx e sn a cos a Now we have sin a cos a (rr k -f- #k ) ?*t c where u e is the number of revolutions at no-load (equation 9), Thus we have with maximum output 33) = .- -. C 2 U + 3 e C The moment of rotation with maximum output is D = (^ V -^- r t ^- a ^ ^e 7 *e ) (Equation 23) where for and ^ we must introduce the current component for this maximum output. We then have 2 e ) 8 Here ^ -\-and /i' -^-are the current components when the motor k * Sk2 is at standstill and the torque is /s a \ a 27e sin a / a: k , p r* \ 1}*=* [ ) r t ( r e A = x e & ] \Z t I Z COS a \ 25 k 2 2 k 2 / Consequently the moment of rotation for maximum output is . x~ c - , ~. ,) a + (^e2k + 4) =^k e+^k The output itself is 428 BRAGSTAD: REPULSION MOTOR. We finally find the efficiency for the maximum output W D * 3 We "mmax -^k %* Z e 86) For the use of the formula indicated in the las*t section in con- sidering a motor design, it is only necessary to make no-load measurements at standstill of the impedance Z^ and the torque D k and at no-load, of the impedance Z e and the number of revolu- tions u e . THEORY OF THE COMPENSATED REPULSION MOTOR. BY ERNST DANIELSON. Of late there has been a great deal written about the compen- sated repulsion motor, which fairly may be considered the most modern type of motor in the present electrical industry. Not only analytical but also graphical methods have been given for explain- ing its qualities. It appears, however, that a complete analytical treatment, considering the leakage that is to say, formulas for the calculation of current, torque and lag with known voltage, brush position and speed have not as yet been published. In this paper the author will present the formulas which are used by The Allmanna Svenska Elektriska Aktiebolaget of Westeras, Sweden, for figuring such motors, and which have given results in good agreement with actual tests. It should be mentioned before entering into detail?, that in the following theory the magnetic losses are neglected and that accord- ingly the magnetic vector is considered the same as the resulting ampere-turns. When using the formulas given below, it is there- fore necessary, when aiming at the utmost accuracy, to apply a cor- rection for these losses. Referring to Fig. 1, let 7 k sin at be the current on short-circuit, assuming the direction as the posi- tion when the current is flowing in the winding from c to d; a = ZnN; (N= frequency). This current creates leaking lines of force, viz., lines of force, yiz., lines cutting only the rotor windings: =zn r I k sin at n r Conductors round rotor, = Leakage co-efficient. These lines of force induce on short-circuit an e.m.f. = _ .,,2 j fc jyr cog at 1() -8 yoits. The lines of force in the direction c d which enter from the rotor into the stator we designate, = B. sin (at + ft) [429J 430 DANIELSON: COMPENSATED REPULSION MOTOR. and those which in direction b a cut the rotor winding (sum of leaking and useful lines), A. sin (at -f- a). In these two last formulas, A and B are the lines per pole, and ft and a the difference of phase of these lines and the current I * The lines B induce in the short-circuit an e.m.f., = B Nn t cos (at -f- 0) 10~ 8 volts. The lines A induce in the same circuit on account of the motion (this assumed clockwise), = A, JVi. ftr'sin (at -f- a). 10" 8 volts. NI being the speed (in frequency). In the last two formulas it is Fm. 1. assumed that the rotor winding is either two-pole or series-con- nected, so that each conductor carries a current =_*. If now 2 r = ohmic resistance of short-circuit, then r k r sin at 10 8 = A. A\ ?' P sin (at + a) B n r c * (at-{- fi) f //.J This equation must be valid for any value on t, accordingly, 7 k . r . \(P = A. N v w r cosr-f B Nr, r *iufi. (\) A. N^ sin a =B. N'. cos /3 + .-. n, r . . ^V! ( ) In the stator circuit there flows a curreni, the phase of which DANIELSON: COMPENSATED REPULSION MOTOR. 431 is the same as that of the lines of force A, these lines being caused only by this current. This current we may represent by 7. sin (at + ). The direction is positive when the current in the rotor winding flows from b to a. This current causes following lines of force : 7. Lines of force cutting only the rotor windings: = . n T . I . sin (at -j-a) in direction b a. II. Lines of force cutting both rotor- and stator-winding in direction b a: = Z. I. sin (at -{- a) (n r n B . sin 0) C is a co-efficient depending on the geometrical form of the machine and the permeability of the iron; n g = number of con- ductors on stator, the same assumption being made as for 7/ r> viz., that each conductor carries a current = _ .0 = angle of brush 2 position (see Fig. 1). The sum of these lines of force must equal A. sin (at -{-a). Accordingly A = I. [:K n. . Bin 0) +. n, ] or : A = I. D if : D = C (n r n s . sin 0) -f . n r . The lines cutting both stator and rotor winding in direction c d are caused by the combined influence of ampere-turns in stator and rotor. Hence : B. sin (at + 0)= C [w, . I. sin (at + a), cos -f- Tjj . w, . sin a] . . . (1) This equation is valid for any value of the angle a t, therefore : sin at (B. cos /5 C. w I. cos - C. w r ^ k ) = (2) cos at (B.siap C. w 8 I. sin a. cos 6) = ... (3) and B cos ft C. ?? 8 I. cos a. cos C. r -40 (4) B. sin C. ra s Z sin a. cos = (5) The equations 1, 2, 3, 4 and 5 contain, besides I, five unknown quantities A, B, , and ^k, and hence give value? of these ex- pressed in 7 and known numbers. Now combining equations (1) and (2), then (1) with (4), and afterward eliminating B by means of (5) and A by means of (3), we get : 432 DANIELSON: COMPENSATED REPULSION MOTOR. ~-j. 10 8 N. n r *. n s . cos 9 (C + ) (6) r. w . cos . 10 8 + N^ . n r * . D (l + 1 "~C./*s. cos* yr r . lo s - r . 10 s These equations show that a and y# are independent of the current. From (4) and (5) we obtain C. w s Z sin a. cos . sin/* ~ ................. < 8 ' ,. w g . Z sin a. cos , . Z k = -' - -- (cot /3 cot a) ...... (9) In the stator circuit, the induced e.m.f's. are: /. In the rotor winding, by the lines of force along c d: = [ n r . 7 k . sin at + B. sin (at + ft)}. NI. n^ 10~. //. In the rotor winding, by the lines of force along b a: = A. cos (at +- a). N~. w r . l 8 2 . A.. JV C. N. n B . sin (9 (rc r w g sin 0) -f fr n t J)]. 10~< TF= C. ra s 2 . sin a. cos 2 9. JVI lO" 8 T = C. w s . sin 11 Dimensions: .Millimeter PIG. 9. half-open slots, each containing 12 conductors of 2.5 mm diameter. The winding was divided in two groups which were connected in parallel. The rotor had 49 slots with 6 wires of 2.8 mm diameter. Winding, series drum. Number of poles, 4. Air-gap, 1 mm. Fre- quency, 27. Voltage, 200. Calculation of C. A current in the rotor winding = 7 causes a flow of lines in the air-gap (neglecting iron) per pole : ft. 1.85. -1. 469. 1 42 0.2 2 (459=area of 1 pole). Accordingly : C=180 438 DANIELSON: COMPENSATED REPULSION MOTOR. Calculation of %=\. The constants of Holart (see Elektrotechnische Zeitschrift No. 46, 1903), are used: . I.n r = j. -^-(20 X 0.93 + 0.4 X30).2 20 = length of one conductor in iron in cms. 30 = free length of one conductor in iron in cms. 0.93 = Hobart's constant. 0.4 = Hobart's constant. The constants C and Z obtained from actual measurement (by measuring voltage and primary current with brushes removed en- tirely in one case and complete short-circuit in the other case) are: C = 208; =7.1. Other constants are : # = 27; r = 0.15; B = 0.68; # = 200; n s = 864; n r 294. Calculating from these constants and with = 0, the curves ^50. f*- 1 30> I 18. 17. 16. 13. 14. 13. 12. U 1 io,fl ! 8. 7. 6. _ Cal ula ;ed (pun es \ Bxjlerii len * . \\ s \ S \ \ \\ \ \\ -t- -fl-Q Cos \ Sr- ,^ ^ "**" / Arop 07 ' \ \ \\ \ l\ nn \ \ 3. ' ^ V 1A ^v ^N^^ 10. *o ^r Kgjii 400 1200 600 800 1000 Revolutions per Mln. Pro. 8. in full line (Fig. 3) are obtained. The dotted lines represent the experimental values. DANIELSON: COMPENSATED REPULSION MOTOR. 439 APPENDIX. SUMMARY OF FORMULAS. General Formulas. D= C (n r w g sin 0) -f . n r 111! _ jy nr * Wg cos OOt a F = [n. A. ^V C ^/> 8 n (9 (n r w g si TF = C w s 2 sin or C08 2 ^ 10' 8 CT = C w s ^in cos 6 N[ n r .lO" 8 T =gn t N! n B cos a (cot fl cot a) .lO" 8 P = 72. sin or + T 7 ! cos or + TK cot y5 -f U Q = 72. cos a Fsin a TF + PI cot /? + Tsin a W B J e sin or. cos a (cot fi -- cot or) 7 *- ~^r ^T = 1.625 10- 10 . C 7 e 2 . > cos 0. p [n t + 7*8 sin (sin or. cos a cot /? cos a )] Special Formulas. for <9 = 0. . 10 8 7>. ^i f 1 5. w r 2 . cot a\ Nrt^ cot/?=: T^L^- r. 10' J ~ ~7Tu F = [w. 2 A JV^ + JV"w r J9]. 10- 8 TT = C. rt g 2 sin a N. 1Q- 8 T = C '/'g sin or. JV^ n r . 10~ 8 T = g rt t NI /'s (cot /3 cot a). 10" 8 P = R. bin a + Fcos a + TF cot ft -f IT 440 DANIELSON: COMPENSATED REPULSION MOTOR. Q . JK. cos a Fsin a TF + C^cot /? + Tsin a. /e = j- _ n s I e . sin a (cot ft cot a) K = 1.625. 10~ 10 C. I^n^p n, Notation. D = SL coefficient. n r = number of conductors round rotor, provided the winding is such that each conductor carries half of current on short-circuit. If the rotor has a six-pole parallel winding, then n r = active con- ductors divided by three. n B *= number of conductors on stator, provided that each con- ductor carries half of total current. C= coefficient of magnetization; C n a = number of lines of force per pole at one ampere in stator circuit, with no current in rotor. g = coefficient of leakage; n r = leaking lines per pole at one ampere on short-circuit. A = leaking cofficient for stator. 6 = angle of brush position. a = angle of lag between current in stator circuit and short- circuit. ft = angle of lag between lines of force along b c and short- circuited current. N = frequency. N l = speed expressed in frequency (at synchronous speed *,-.). r = resistance of rotor circuit including short-circuit. R = resistance of short-circuit including rotor winding. V. W, U, T, P and Q = coefficients. / e = stator current (effective value). / ek = short-circuited current (effective value). E e = impressed e.m.f. (effective value). K = torque in kilograms at 1 meter radius. INDEX OF SUBJECTS. ACCELERATION, 80, 136. maximum efficiency with vari- ous methods of control, 139. power required, 81. used by various classes of rail- ways, 116. used on Valtellina railway, 167. Air, compressed, use of, 155. compressors, 155. Arnold system, design of, 35. gap of motors of various makes, 126. resistance, 81. Alternating current motors. (See Motors. ) railways. (See Railways.) Armature reaction, 124. BALANCING transformers, 105. Batteries, storage, 275. a. c. railways, 163. automatic, 280. buffer, parallel connected, 263. carbon regulator, 284. depreciation, 278. distant from power house, 280. efficiency, 163. investment, 276. maintenance, 278. operation economy, 276. plates, construction, 285. plates, life of, 285. plates, troubles, 285. reasons for installing, 276. reliability, 278. reserve, 278. Boosters, 262. calculation of, 264. connections, 273. design of, 265. efficiency, 286. excitation, automatic, 281. excitation, regulation, 282. fly- wheels, 269. slip, 267. Bow trolley. (See Distribution.) Brake, d. c. series motor, used as, 144. equipment, operation, 200. shoes, character of, 320. friction, ratio to pressure, 317. pressures, 317, 322. three-phase motor, used as, 144. Braking, brake shoes, character of, 320. emergency application, 319. experiments, 321. high-speed trains, 315. service application, 319. shortest stops on record, 322. British electric railways, 52. Buffer -battery, 263. machine, 264. CABLES, capacity distribution in, 241. grounds in, 239. insulation, 239. protection from electrolysis, 310. Capacity, distribution in cables, 241. Carbon regulator, 284. Catenary construction, 160. Central stations, 163. Characteristic curves of different types of motors, 373. Circuit breakers, oil, 249. City railways. (See Railways.) Collectors, current. (See Distribu- tion. ) current, 155. Compensated motors. (See Mo- tors.) Compressors, air, motor driven, 155. Conduit system, first, 7. Contact shoe for third rail. (See Collectors. ) Continuous current motors. (See Motors. ) (441) 442 INDEX OF SUBJECTS. Control by brush shifting, 151. of large motors, 217. methods, 151. multiple unit, 152, 228. first, 17. limit of number of units, 317. potential regulators, 151. rheostatic, of induction mo- tors, 133. rheostatic, series parallel, 151. series-parallel, first, 4. of storage batteries, 280. Ward Leonard system, 107. Controllers, cylindrical, 152. d. c., weight of, 154. magnetic blow-out, first, 13. single-phase, weight of, 154. three-phase, weight of, 154. weight of various kinds, 152. Converter car, 198. car operation, 201. Converters, synchronous, 252. efficiency, 253, 258. efficiency all day, 259'. first cost, 257, 258. frequency, 254. hunting, 254. inverted, speed limiting devices, 255. overload capacity, 164. power factor, control of, 253. protection of, 250. starting of, 253. voltage, d. c., 253. Cooling of motors, 134. Cost a. c. railways vs. d. c., 166. Coupling, elastic, 148. rigid, 148. Crank and connecting rod, motor drive, 149. Current collection. (See Distribu- tion.) collectors. (See Distribution.) collectors, 155. Curves, railway, minimum, 78. Deri motor, 387. Dimensions of motors of various makes, 126. Direct current motors. (See Mo- tors.) Distribution, a. c., choice of phase, 174. a. c., polyphase, disadvantages, 174. protection of systems, 238. cables, capacity of, 241. collector current, high speed, 158. ideal current, 162. Distribution Continued: current collectors, 155. drop, total in trolley, 236. fourth rail, 62. overhead conductors, 59. double trolley, first, 9. first, 4. limitation of, 161. lines connected to under- ground lines, 248. work details, 34. rail, return insulation of, 312. losses in, 224, 225. return feeder, practice in United States, 288. third rail, 62, 160. composition of, 192. construction, 192. installation of, 161. pressure of shoe open, 200. protected operation, 197. protection, 161, 193. third rail shoe, 193. construction, 195. current capacity, 160, 200. operation, 199. pressure, 200. three-phase, capacity in, 242. inductive ground, 242. total drop in trolley, 236. trolleys, 155. bow, 64. bow, rolling, 64. bow sliding, 64, 157. conductors, conditions to be fulfilled by, 61. d. c., voltage, 166. drop in line, 236. high speed, "158. insulator, special, 34. Oerlikon, 158. overhead, first, 4. pressure on wire, 64, 158. roller, 159. shoe, 157. suspension and distribu- tion, 161. suspension catenary, 160. switches, 162, voltage, 165, 187. voltage, high, danger of, 210. voltage, safe, 209. wheel, 157. whip, 63, 158. wind pressure compensa- tion, 158. wire, double, first, 9. wire, ice and sleet, 162. INDEX OF SUBJECTS. 443 Distribution Continued: two voltages, 161. underground, protection from electrolysis, 308. EFFICIENCY of motors of various makes, 126. Electrification of steam railways. (See Railways.) Electrolysis, 288. a. c. system, 210. bonding as .a remedy, 313. importance of, 220. in Europe, 311. opinions of experts, 289. ordinances, 288. rail return, insulation of, 312. return feeders used in United States, 288. summary, data, 289-. tables of classified data, 292, 307. E. m. f. of motors of various makes, 126. Electro-pneumatic system of trac- tion, 26. Energy, restoration of, 218. return to line, methods of, 144. Equipment, railway. (See Rail- ways. ) FIELD amp. turns ratio to arma- ture amp. turns in single-phase commutator motor, 13. Fly-wheels, speed limit, 269. Fourth rail, 62. Frequency best adapted to single- phase railways, 99. choice of, 183, 184. GEARING, double, 148. motor, 148. single, 148. Gearless motors first, 15. Gear ratio of motors of various makes, 126. Generators, compounded, 164. grounding of the neutral, 242. high speed fly-wheel, 164. polyphase, self-exciting, 397. vs. single-phase, 231. protection of, 250. railway, capacity of, 163. single-phase vs. polyphase, 231. self-exciting, 397. speed, drop with load, 164. Ground detectors, 247. Grounds, 239. HEILMAN, locomotive, 216. History of electric railways, 1. Hunting of synchronous converters, 254. ICE and sleet, 162. Impedance, mutual inductive, 336. self-inductive, 326. Induction motors. (See Motors.) regulator, 104. Insulation in cables, 239. Insulator, 35. trolley, special, 34. Insurance, 236. Interurban railways. (See Rail- ways. ) LINE construction, 155. Line construction. (See .Distribu- tion.) Liquid, rehostat, 151, 152. construction of, 142. Locomotive, electric, 106. Finzi type, 168. gearless motors, first, 15. heavy, 177. high voltage, 158. Oerlikon system, 217. steam, freight draw-bar pull, 71. power of, 71. weight per h. p. of electrical equipment, 176. London railways, 68. Losses, distribution, losses in rail return, 224, 225. MAGNETIC blow-out, first, 13. Monorail railways. (See Rail- ways. ) Motor-gearing, 148. Motor-generators, 252, 256. efficiency, 258. efficiency all day, 259. first cost, 257, 298. starting of, 256. voltage, 257. Motors, acceleration, 136. a. c., 323. classification of, 60. field of application, 185. on d. c. service, 105. vs. d. c., 98, 205. characteristic curves of differ- ent types, 93, 118, 373. control of. (See Control.) controllers. (See Controllers.) cooling of, 134. crank connection, 149. direct connected, 148. d. c. advantages, 129. data, 126. efficiency of, 129. limitations of, 60, 101. losses in, 133. losses when starting, 138. 444 IXDEX OF SUBJECTS. Motors, d. c. Continued: series, air gap, 119. arranged to return energy to line, 144. characteristics, 94, 140. speed, characteristic, 140. used as brake, 144. shunt, equalization of, 147. speed characteristics, 140. d. c. sparking, 121. starting torque, 137. voltage, 166. gearless, 149. first, 15. induction, air gap, 119. efficiency all day, 259. high voltage, 120. polyphase, 225, 330. advantages, 173. characteristics of, 93. variable speed, unity power factor, 397. induction, single-phase, 113, 335. advantages, 173. condenser, 339. torque, pulsating, 173. induction, three-phase, advan- tages, 213. braking, quality of, 144. concatenated, losses in, 133. concatenation of, 141. data, 127. efficiency of, 129. field of application, 166, 181. losses in, 133. losses when starting, 138. mountain railways, 145. power factor, 132. power factor, high, 131. regulation of, 132. in secondary, 141. rheostatic control, losses in, 133. speed characteristic, 140. speed variation, 141. torque, 170. torque curve, 181. torque starting, 136. variable pole, 143. induction, torque, 172. losses, 133. at starting, 138. N. Y. C. R. R., 131. New York subway, 134. parent, models of, 11. repulsion, 104, 367, 381, 399. Motors Continued: repulsion, compensated, 123, 388. formulas, summary of. 439. Lahmeyer, 168. rotor, 125. slots, design, 437. stator, 124. theory of, 429. torque, calculation, 433. torque, starting, 434, 436. repulsion. Deri, characteristics, 387. repulsion, disadvantages of, 408. performance, 414. principal equation, 410. regulation by brush shift- ing, 168. theory and operation, 410. reversal of, 140. series, polyphase, 347. series, single-phase, 112, 350, 354, 377, 396. advantages of, 103. air gap, 120, 171. as generators, 146. characteristics of, 130. comihutation, perfect, 400. series, single-phase, compen- sated, 359, 399. conductively, 360. disadvantages of, 408. inductively, 364. losses in, 133. winding, 169. series single-phase, data, 128. early experiments, 102. efficiency of, 129. Finzi, 168. frame, 120. frequency, 99, 131. losses when starting, 138. magnetizing current, 171. power factor, 131. power at start, 171, 211. railway, 402. ratio, field amp. turns to armature amp. turns, 131. sparking, 122. speed, characteristic, 104, 140. starting torque, 136. straight, losses in, 133. torque, 170. torque starting, 171. shunt, polyphase, 341. single-phase, 376. S. OF 445 Motor-starters, 151. liquid, 152. starting torque, 130. synchronous, 324. Winter-Eichberg, 403. characteristics, 403. Winter-Eichberg-Latour, 352. Multiple unit control. (See Con- trol.) Neutralj grounding of, 245. N. Y. C. R. R. locomotive motors, 131. New York subway motors used by, 134. OERLIKON locomotive, 217. trolley, 63. Operation of Arnold system of electric railways, 38. Overhead conductor. (See Distri- bution. ) distribution. (See Distribu- tion). work. (See Distribution.) Overload relays, 249. POLYPHASE motors. (See Motors.) Potential regulators, 151. Power, cost factor, 203. developed by motors of various makes, 126. plant equipment, 195. used to operate various classes of railways, 116. Protection of a/c. distribution sys- tems, 238. Pressure against trolley wire, 64. Rails, a. c., resistance, 156. insulation of, 312. Railways, braking, 315. British, capitalization of, 52. mileage of, 52. city, condition of service, 180. requirements of, 95. data concerning various classes, 116. dividing line between steam and electric, 87. electric, acceleration, maxi- mum efficiency, 139. advantages of, 65, 85. a. c.. 92. electric a. c. extension of d. c., 179. extensions of d. c., condi- tions of, 168. rail loss, 224, 225. single-phase, 112. three-phase, 112. vs. d. c., Ill, 206. electric, Arnold system, 26. air-compressor. 35. Railways, electric Arnold system Continued : car motor equipment, 35. detailed description of, 34. operation of, 38. valves, description of, 43. electric, British, 52. expenses, 53. freight carried, 53. passengers carried, 53. electric, conditions in 1887, 11. conduit system, first, 7. control systems. (See Con- trol. ) control systems, data, 154. cost a. c. vs. d. c., 166. current collection. ( See Distribution.) current collection, 155. electric, d. c., 111. a. c. transmission, 112. constant current, 114. limitations of, 101. three- wire, 111. two- wire. 111. voltage, 166. electric, first, 1. first in U. S., 5. heavy service, first, 83. high-tension a. c. trans- mission, first, 16. history of, 1. interurban, first, 16. in various civilized coun- tries, 19. line, construction, 155. power lost in motors of various types, 133. problem principal, 89. requirements of, 59. single-phase, Arnold sys- tem, 21. single-phase, Spindlerfelde, 167. single-phase vs. three- phase, 221. 222. storage battery, 114. suburban. traffic condi- tions, 55. three-phase, field for, 166. three-phase vs. single- phase, 221, 222. traffic, class of, 53. transmission, three-phase, power-factor, 163. trunk line, ideal, 86. vs. steam, 65. Ward, Leonard, 113. wear and tear of, 65. weicrht of a. c. vs. d. e., 186. 446 INDEX OF SUBJECTS. Railways, electric Continued : weight per h. p. of electric equipment on locomotive, 176. with converter substations, first, 23. elevated, data, 116. high-speed, 79. energy, restoration of, 218. entering London, 68. high-speed, 78. industrial, data, 116. journeys per head of popula- tion, 58. monorail, 71. acceleration at start, 80. advantages of 76. center of gravity of car, 80. high-speed, 79. Manchester and Liverpool, 80. minimum curve, 78. safety of, 79. mountain, data, 116. motors for, 145. requirements of, 98. protection of system, 238. steam, British, electrifica- tion of, 67. capitalization of, 70. cost of moving freight, 71. electrification, effect on mileage, 54. electricfication of, 83. electricfication, results of, 54. fixed charges, 71. fuel cost, 71. power of locomotives, . 71. vs. electric, 65, 87. wear and tear of, 65. street, data, 116. suburban, requirements of, 90. suburban, traffic conditions, 55. surface, classification, 180. telephones, disturbances in, 232. tractive effort, 223. trunk, freight, requirements of, 9. trunk line, data, 116. trunk line, ideal, 86. trunk, passenger, requirements of, 97. underground, data, 116. Ward-Leonard system, 214. single-phase, system, 176. weight of a. c. vs. d. c., 186. weight per h. p. of electrical equipment on locomotive, 176. Relays, overload, 249.' Repulsion motors. (See Motors.) Resistance, a. c., of rails, 156. Restoration of energy, 144, 218. Return circuit losses, 224, 225. Rheostat, liquid, 151, 152. construction of, 142. Rotary converters. (See Convert- ers, synchronous. ) SERIES motors. (See Motors.) Series parallel control, 151. Service, classification of electric railways, 180. Shoe, third rail. ( See Distribution. ) Shunt motors. (See Motors.) Single-phase railways. (See Rail- ways. ) Skin effect, 156. Sleet and ice, 162. .Slip relation to centrifugal masses, 267. Speed limiting devices, 255. of motors of various makes, 126. on various classes of railwavs, 116. Spindlerfelde single-phase railway, 167. Starters, 151. Starting torque, 136. Stops, shortest, on record, 322. Storage batteries. (See Batteries.) mounted on motor cars, 114. Substations, 163. distribution of, 165. portable, 197. operation, 201. protection of, 250. Switches, overhead, 162. Synchronous converters. (See Con- verters, synchronous.) TELEPHONE, disturbances in, 232. Third rail. (See Distribution.) Track construction, 192. Traction. (See Railways.) data, 116. electric. (See Railways.) Tractive effort, tio of, to weight on wheels, 66. Traffic, classification of, 53. conditions of surburban rail- ways, 55. journeys per head of popula- tion in larger cities. 58. Train resistance, air pressure, 81. Transformers, balancing, 105. railway, capacity of, 163. regulating, 151. regulating, losses in, 144. IXDEX OF SUBJECTS. 447 Transformers Continued: three-phase-two-phase, balanc- ing of, 175. Transmission, a. c., 196. a. c. wiring formulas, 234. of power, first, 6. protection of system, 238. single-phase, 178. railways, 230. railways, maximum econ- omy, 233. railway, mechanical consid- erations, 236. railway, voltage drop, 235. vs. polyphase, 174. single-phase vs. polyphase, 174. telephone, disturbances in, 232. three-phase, power factor of, 163. voltage, 165. Trolley. (See Distribution.) VALTELLIXA railway gearless motors, 149. acceleration used, 167. Valtellina railway gearless motors Continued: locomotive, 135, 150. WARD-LEONARD system of electric traction, 107, 113, 176, 216. field of application, 216. Weight of motors of various makes, 126. Whip trolley. ( See Distribution. ) Wilkesbarre and Hazelton railway, 189. brake equipment operation, 200. construction, 192. converter car operation, 201. description of road and equip- ment, 191. power plant equipment, 195. profile, 191. third rail, composition of, 192. third rail shoe construction, 195. track construction, 192. Wind pressure on trolley, 158. Wiring formulas, a. c., 234. UNIVERSITY OF CALIFORNIA LIBRARY BERKELEY Return to desk from which borrowed. This book is DUE on the last date stamped below. DE030 1947 f* LIBRARY US APR 2 8 195 1 7 1952 1 6 1952 AU61719S2 LD 21-100m-9,'47(A5702sl6)476 'T7=j 161539