The steady development of marine dynamo machinery and the number of applications opening for its further use aboard men-of-war, render the subject of direct importance to every student of naval construction or equipment. Recognizing the interest felt throughout the service in electrical matters, it affords me pleasure to comply with the invitation of the Board of Control to contribute a paper on the subject.
An inspection of the systems and electrical apparatus in use in the different European navies in 1891, followed by duty at the New York Navy Yard, where the electrical plants of nearly all the vessels of the Navy have come under my inspection, has demonstrated to my satisfaction the two facts that our electrical plants are in most respects superior to those in use in other navies, but deteriorate more rapidly.
It has been the practice of the Bureau of Navigation, and later of the Bureau of Equipment, to publish in the Naval Intelligence Annuals general descriptions of the apparatus introduced in our ships, and the articles therein contained cover the subject very fully. As no article of the kind appears in the last Annual, the attempt will be made to partly cover the subject in this paper.
The Bureau of Equipment now issues standard specifications for nearly all of the electrical material used aboard ship.
The specifications for generating sets were referred to in the Intelligence Annual for 1892, and an illustration given of an eight-kilowatt set, of the new type, as constructed by the General Electric Co. The principal points of the specifications which determined the construction of this type were the continuous operation of the engines without oil in the steam spaces, the provision limiting the heating of the dynamo, and the absence of external magnetic field. Two eight-kilowatt sets of this type have been placed aboard the Concord, and two others aboard the Bennington. Sixteen-kilowatt sets are in use on the Dolphin, Detroit and New York, while thirty-two-kilowatt sets, shown in Fig. 1, are also installed aboard the latter vessel. Similar sets of various sizes are constructed and in process of installation aboard the Maine, Cincinnati, Raleigh, Montgomery, Marblehead, Columbia, Puritan, Terror and Amphitrite. Under these circumstances a few details may be of interest.
The finished weights and dimensions are as follows:
Watts
Weight. Length. Width. Height. per lb.
400 amperes, 11,100 lbs. 9ft. 2in. 4ft. 1 in. 5ft. 9 ¾ in. 2.89
300 amperes, 9,100 lbs. 8ft. ¾ in. 3ft. 10 1/8 in. 5ft. 5 ½ in. 2.64
200 amperes, 6,160 lbs. 6ft. 9 ½ in. 3ft. 4 ½ in. 5ft. ¼ in. 2.60
100 amperes, 3,750 lbs. 5ft. 6 ¼ in. 2ft. 10in. 4ft. 9in. 2.13
The construction of the engines is much heavier than was formerly the case, the number of breakages of those in use in the service having apparently shown that strength had been unduly sacrificed to decrease weight. The new engines have more clearance in the cylinders, and although this is uneconomical so far as steam is concerned, it is of value when working with wet steam or when threatened by water. As the presence of water in the cylinder has always been a cause of damage to dynamo plants, special attention is given to drainage, and the cylinders are fitted with both hand and the automatic relief valves. The governor is of the common automatic type, the weights being placed in a wheel midway between the cranks. The space allowed is barely sufficient in the smaller engines, and they do not govern as closely as the larger sizes. The closest regulation has been obtained from the 24 K. W. sets which, on test for acceptance, averaged only five revolutions change of speed from full load to no load, the momentary increase, when all load was thrown off suddenly, being only eight revolutions. The requirement of the specifications, that the variation of speed should not exceed eight revolutions when the steam pressure changed from sixty to one hundred pounds, necessitates considerable reserve power at ordinary working pressures, and even more was furnished, the engines having carried full load on the dynamos at thirty-five pounds pressure, although intended for eighty, and at the latter pressure having under test carried double load. The pistons are without packing rings, and all four sizes of the engines have the same stroke, six inches, the cylinder diameters being respectively 13, 10 3/4, 8 1/2 and 6 1/2 inches. All bearings are of gun-metal except in the 32 K. W. sets, where the crank-pin boxes are partially babbitted. The cranks are set at 90°, causing a poor balance and consequent noisy working, probably the most serious defect of the engine.
The dynamos are four-pole, compound wound. The frame and field cores are of cast steel in one piece, the pole pieces being bolted on. The field coils are cylindrical and wound on vulcabeston spools, the spools on the small dynamos being coned, on account of the lack of room. The series coils are next the core, the ends being brought out directly opposite each other, and all four spools are coupled in series. The shunt winding is outside the series, and the spools are also in series. This arrangement greatly facilitates repairs, as a defective spool can be removed and rewound without touching the others.*
*The arrangement of the field magnets has been modified in dynamos for the Boston, Atlanta, Texas and the battleships. In the new type the frame is divided horizontally, the upper half lifting off, while the cores and pole pieces are cast together, and each core bolted to the frame, so that the axes of the cores are inclined 45° to the vertical and horizontal instead of being in those planes. The armature has its own shaft, which is rigidly coupled to that of the engine, but is in other respects unchanged. The battleships will have a new type of engine, which is somewhat smaller but better balanced than the one described in the text.
The shape of the field frame and the position of the coils are such as to minimize all external magnetic field. No magnetism can be detected on the outer part of the frames except in one or two positions, and the external field is imperceptible on a horizontal force instrument at a distance of fifteen feet. While it may be urged that the influence of a dynamo placed in an iron compartment is imperceptible on a compass at a distance, there are instances on record in which peculiar arrangements of steel bulkheads have resulted in a deflection of the compass whenever the dynamos were in operation. This magnetic effect could be easily counteracted, but the present policy of entirely preventing it is a most effective safeguard. The multipolar type of dynamo allows of reduction of weight, although necessitating somewhat greater floor space.
The armatures of the new type dynamos are Gramme wound, and slide off and on a shaft common to both engine and dynamo, being secured by a feather and slot and kept in place by set-screws. The armature is built up of soft iron discs, clamped together, there being no through bolts. In the first armature constructed for the 16 K. W. (200 ampere) dynamos, it was found on making efficiency tests that the total core loss due to eddy and foucault currents in the mass of the armature itself was 1207 watts, and the armature, when run on open circuit, the field being separately excited, heated to 22° F. above the air in two and three-quarter hours. As the limit of heating in a full load run of four hours is 50°, this allowed but little margin for the heating due to the current, and armatures tested under these conditions heated from 60° to 80° above the air. A new type of construction was therefore adopted. The core discs were held together by clamps, all through bolts being removed and the discs themselves being carefully insulated. The armature winding was also changed, being made of stranded conductors, each wire being separately insulated before stranding. The effect of these changes was most marked, the total core losses falling below 700 watts, and heating on the full load run of four hours to about 45° F.
The commutator is of bronze, the number of segments varying from 120 to 144 in the different sizes. The commutator has cross connections inside for use with two sets of brushes.
As a result of the range of steam pressure called for under the engine specifications, and of the low heating limit allowed for the dynamo, the generating set possesses great reserve power. In one of the tests, while operating two 200-ampere dynamos in multiple with about 80 pounds of steam, the whole load of 400 amperes was inadvertently thrown on one dynamo. No noticeable diminution of speed of the engine occurred, nor was there any sparking on the dynamo, and the overloading was manifest only on the ammeter. When the double load of 400 amperes was thrown off it was not necessary to adjust the brushes, the sparking being insignificant. This capacity for carrying an overload is of the greatest value in ship work, as, in case of damage to one set, all the work could be done by another. There is no question of the capacity of the dynamo to carry one hundred and fifty per cent, of its rated load for long runs without material damage, nor is there any lack of power in the engine. In any particular case, the amount of safe overloading would probably depend upon the heating of engine bearings and crank pins.
All dynamos as yet used in our service, have been fitted with rheostats in the shunt field, to admit of variation of the voltage. This design has found but little favor in other countries, either for commercial or marine work, but it is universally adopted in the United States, and the ease of adjustment thus obtained is so great that no other system is considered. Commercial practice also includes an adjustable resistance in parallel to the series field, and this is authorized on board ship when the dynamos are required to operate in parallel.
WIRING.
The greatest care has always been taken in wiring vessels of the United States Navy, and the standards of insulation and of work are of the very highest order. The student of electrical engineering can nowhere find a better example of a thoroughly high grade installation than on one of our ships. This excellence of work is, under the prevailing high wages and short working hours, obtainable only at a great expense as considered with the cost of foreign installations, but economy should not be sought in any changes involving the efficiency of the installation or the lowering of the standard of work.
The general system of wiring is to run two sets of circuits. One set called the "battle circuits" includes all lamps necessary in action. These circuits generally include all lights in fire and engine-rooms, coal bunkers and store-rooms below the protective decks, lights not needed in action being turned off separately. On upper decks the battle lanterns are so placed as not to show from outboard. All exposed lights and all used for ordinary illumination are put on "the lighting circuits." On going to quarters all these lights are cut out at the switchboard in the dynamo-room, and there is no necessity for an inspection of the whole ship to determine if unauthorized lights are in operation.
To further insure proper lighting in action, the ends of the battle mains are interlocked at many places in the ship, and the circuits are kept as much as possible below the protective deck. This system becomes more important as the necessity for electrical energy in time of action increases, as it is doing daily by the extended use of electric motors.
The number of circuits at the switchboard has been decreased of late by the use of "feeders." These are heavy conductors leading from the dynamo-room to the general location to be lighted, where, by means of a "feeder junction box," they branch right and left into "mains." Each pair of mains may have a feeder box placed in it, allowing its division into two other pairs of mains. These feeder boxes contain no fuses, and the mains leading from them must therefore have a safe carrying capacity equal to that of the feeder fuses at the main switchboard. Switches may be interposed in the mains as desired, and in case the arrangement of the ship is such as to necessitate the use of many switches on the lighting circuits', it would be advisable to place the feeder boxes and switches together in sub-stations.
Subdivisions known as branch mains are taken off the mains by means of ordinary junction boxes, each of which contains a fuse. Branches may be taken from the branch mains in the same way. Ordinarily it is not intended that more than four lights shall be taken off one junction box, and the specifications call for an unbroken wire from the box to the lamp. All wires are enclosed in pine mouldings, there being at least three-quarters of an inch of wood between wires, and three-eighths between either wire and any metal connected with the ship. As a rule all fixtures are secured on wood, avoiding a heavy ground in case of low insulation resistance on the sockets or other metal parts. Of late, commercial ceiling fixtures, ceiling cut-outs, and non-watertight junction boxes have been installed in quarters where their use in nowise impairs the insulation efficiency of the plant, with the idea of avoiding the expense and loss of light attendant on use of the standard water-tight fixtures. There is probably no case recorded of any report having been made complaining of too much light in any of the living spaces of a ship. The tendency is always to seek a brilliant illumination, and it is very poor economy to obtain light and then interpose heavy ground shades for the purpose of subduing it.
The search-light leads extend from the switchboard, directly to the projectors or to the control stands, being broken at intervals of about every fifty feet for convenience in installation, the ends being joined in search-light junction boxes which contain no side holes and no fuses, being used simply to avoid splices.
The standard wire is now without any lead covering. The latter was used as a waterproof covering to the insulation and as protection from mechanical injury. In neither of these two respects did it prove satisfactory. The spar-deck and engine-room circuits of several ships, wired with lead-covered wire, have had to be entirely renewed at the end of three years. On removing the old wire, it is found that the handling involved in getting it into place is very liable to break or dent the lead, admitting water to the insulation and retaining it. Any defect in the insulation will then cause a bad ground. The lead covering practically puts all grounds in parallel, and is, moreover, objectionable from its extra cost and weight. The new standard wire depends for its durability on the insulation being to a greater extent of vulcanized rubber, and on a heavy braid covering which is saturated with a waterproof compound. Sufficient mechanical protection is given by the moulding. The new wire is, of course, to a certain extent experimental, and if experience should show that it is inferior to wire of the same grade covered with lead, the latter might be again used.
In order to obtain the high insulation resistance called for in naval plants, great care is necessary in installation. The standard called for, of at least one megohm on each circuit, and 500,000 ohms on all in parallel, is intended to secure care in installation, and would be unobtainable in anything but dry weather, the surface conduction over socket shells and over exposed metal parts of the circuit being sufficient in ordinary weather to bring the insulation far below that figure. The general precautions taken, in order to reach the limit, are the use of high insulation wire and the separation of all parts of the circuit from the metal of the ship by means of wood moulding, or by hard rubber tubing when passing through bulkheads and beams. Grounds frequently occur on the lamp sockets, which would be dead grounds if it were not for the interposition of wood.
The high insulation called for is not only a guarantee of good material and careful work, but also insures against accident.
With telephones, fire and water alarms, torpedo and gun circuits, helm and revolution indicators, and various kinds of call signals, all operated by electricity, high insulation of the dynamo circuits becomes a necessity, a cross between two wires involving many possibilities of trouble. An electric light system would probably operate perfectly well on an insulation of five hundred ohms, although there would be a greater risk of sudden break down aboard ship than would exist in shore plants. The specifications call for the best possible work in installation, as a guarantee of good operation and durability. Every care should be taken to localize and eliminate even light grounds. Nearly all the trouble is experienced on the fire-room and spar-deck circuits, and arises very largely from the water-tight fittings not being properly closed and screwed up, allowing damp air or water to get inside.
In laying out the wiring, the size of the wires is determined under three provisions of the specifications. One stipulates that no single conductor shall be smaller than number 16 B. W. G. The second calls for a cross section of one thousand circular mils (one square inch equals 1273236 circular mils) for each ampere of current at full load, authorizing a double current in search-light leads. The third calls for such a cross section that the fall in potential between the dynamo and the farthest lamp shall not be more than three per cent on full load.
In laying out a ship installation the calculations are made on the area of the conductors in circular mils. As stranded cables are used for all sizes above No. 14 B. W. G., results cannot be expressed in terms of diameter. The exact area required cannot always be obtained in stranded cables, and the nearest size admitting of convenient manufacture is therefore taken. The following are the sizes of conductors commonly used in the service:
No. 16 B. W. G. 4,225 circ. mils.
No. 14 “ 6,889 “
20-30 B. W. G. flexible cord 2,880 “
7-22 B. & S. flexible cord 4,498 “
7-20 B. W. G. 8,575 “
7-18 B. & S. 11,368 “
7-19 B. W. G. 12,348 “
7-16 B. & S. 18,080 “
7-17 B. W. G. 23,548 “
7-16 “ 29,575 “
7-15 “ 36,288 “
19-18 “ 45,619 “
19-17 “ 63,916 “
19-16 “ 80,275 “
37-18 “ 88,837 “
37-17 “ 124,468 “
37-16 “ 156,325 “
The ruling condition of manufacture of stranded cables is that the cable must consist of 7, 19, 37 or 51 wires, if all are of equal size. By varying the size of each of the component wires, stranded cables of any desired cross section may be made.
The determination of the cross section of the mains is as follows: The maximum drop allowed is three per cent, on eighty volts or two and four-tenths volts:
... 2.4 = current X resistance of mains.
Although navy specifications call for pure copper wire, the twisting involved in making stranded conductors shortens the cable from three to five per cent, and practically diminishes its conductivity to that extent.
Substituting in the first formula and reducing we have
Area in circular mils = 4.5 X LC,
L being the total length of wire in circuit in feet (both conductors) and C the current in amperes.
Making C unity and substituting 1000 circular mils, the corresponding area called for by the navy specifications, we obtain a value of L of 222 feet. For shorter lengths than this, the area calculated from drop alone would be less than allowed by the specifications for carrying capacity. We have, therefore, the simple rule: for mains extending to shorter distances than 111 feet from the dynamo, the area is determined by allowing one thousand circular mils per ampere; for greater distances, the area must be determined from the fall of potential alone by the above formula.
A useful modification of the formula is to determine the area of conductor for any given number of 16 candle-power lamps, each averaging about eight-tenths of an ampere.
These formulae apply, of course, only to navy standard lamps and to naval wiring specifications.
Most of the fittings in use aboard ship are figured in General Information Series, No. XI. A few others are shown in Figures 2 and 3. Figure 2 is a receptacle for search-light leads, being placed near the projector. Its use is not universal, as the leads may, in many cases, lead unbroken to the switch in the control stand or pedestal. Figure 3 shows the stuffing boxes used in carrying wires through water-tight bulkheads or decks. The small one is the ordinary bulkhead tube. The one on the left is for ordinary decks, the right hand one being for protective or other curved decks. The length is necessary to pass through the planking. All these tubes are of brass, lined with hard rubber tubing.
SEARCH-LIGHTS.
Several kinds of search-lights are now in use in the service. The Chicago, Atlanta, Boston, San Francisco and some others are fitted with apparatus of French manufacture; later cruisers received projectors made by the Thomson-Houston Co., while those finished within the last year are fitted with Thomson-Houston projectors of still another type.
The Mangin projector, as originally manufactured by Messrs. Sautter and Lemonnier, of Paris, was admirably designed and constructed for the conditions of use on shipboard, and but few improvements in the general plan have since been made. It is desirable to have the projectors used in our service of American manufacture, and the Thomson-Houston Electric Co. have, for several years, supplied good apparatus. The principal trouble has been with the mirrors, great difficulty having been experienced in making the silver permanently adherent to the glass. The manufacturers are now willing to guarantee the silvering, and greater durability is hoped for.
The latest type of projectors, illustrated in Fig. 4, are much larger and heavier than their predecessors, and in those respects are not improvements. They include, however, the new principle of electrical control, and the experience obtained with them will be of value. The apparatus is quite thoroughly explained in General Information Series, No. XI, June, 1892, and no further description is necessary.
It has been the practice for several years in the service to connect all search-lights in parallel, controlling each circuit from the switchboard, where they may all be thrown on a separate dynamo, or, if desired, may be worked from the same dynamo which is feeding the incandescent circuits. The former practice is generally followed, but is by no means necessary, as the damage done to the incandescent lamps by the sudden momentary changes of voltage, caused by the search-lights being thrown on or off, is more theoretical than real. The sudden jumps are of course unpleasant to the eye, and for this reason the use of a separate dynamo is preferable. In action, it is intended with plants of the latest type to operate all dynamos in parallel, and the flickering of the incandescent lights in such circumstances is of no importance.
Each search-light has its own dead resistance. It has been argued, from economical considerations, that this resistance should be avoided by lowering the potential of the dynamo to that of the lamp, but it has been found that the resistance is of so much practical benefit in the operation of the lamp as to make economical considerations less important than those of good working.
In the operation of arc lights carrying the heavy currents now used, the simplicity of Ohm's law is widely departed from, and the sudden variations of resistance, due to the change of position of the arc or to impurities in the carbons, defy all permanent adjustments. Under these circumstances, the dead resistance acts somewhat like a buffer, diminishing the effect of the change and affording a working margin of potential. Any arc lamp operates to better advantage with dead resistance in circuit, and, this being the case, it is true economy to waste sufficient energy to secure good operation.
The best results with search-lights will always be obtained with the hand lamp. It requires but little attention from an operator to keep the crater in an inclined lamp well shaped and placed, but that little attention is necessary, and no automatic mechanism can take its place. The latest type of automatic lamp furnished our ships, shown in Fig. 5, works as well as any inclined lamp in use in other navies, but when the best results are desired must be controlled by hand. The two great difficulties that automatic action cannot control are the shifting of the crater into unfavorable positions and the "mushrooming" of the arc or building up of an obtuse point on the end of the negative carbon, which masks the crater. The operator adjusts the carbons to keep the crater turned full towards the mirror, and gets rid of the mushrooms by bringing the carbons together and breaking it off. In an automatic lamp the mushroom grows slowly, cutting off much of the crater radiation and causing the crater to form too near the end of the carbon. The arc adjusts itself to the potential for which the lamp is set, and may operate steadily under the conditions stated until the mushroom suddenly falls off, causing an immediate lengthening of the arc, which is generally attended with the extinction of the light.
In spite of these difficulties there are certain advantages in the automatic lamp which recommend it, although no type should be used that cannot also be worked by hand. There are frequently cases where the search-light may be needed suddenly, as with a man overboard, vessels sighted, or danger reported by the lookout. Under these conditions it is a great convenience to have a lamp ready for operation by turning a switch.
The difficulties experienced with the inclined automatic lamp are largely overcome by placing the carbons in line horizontally in the axis of the projector, the crater of the positive being in the focus of the mirror. Left to itself in this position, the upper part of the crater is consumed by the flame of the arc which rises vertically, causing the crater to become inclined upward. This can be avoided in several ways, the simplest being to place a semi-circular bar of soft iron of about two inches radius below and at right angles to the arc and concentric with it. This iron becomes magnetized by the lines of force of the current, and reacts on the arc, keeping it down in place and preventing the deformation of the crater. The objection to the horizontal lamp is that the middle of the mirror is inoperative, the crater radiation being cut off by the negative carbon. The diameter of the latter cannot be reduced beyond a certain point, nor can the length of the arc be increased, so that considerable loss of reflecting surface is inevitable. Experiments have been made on horizontal lamps, resulting very favorably when working automatically, and it is probable that apparatus of this type may be issued to the service for trial.
The control of projectors from a distance by means of electricity has been tried in foreign services, and several of our ships will have this system installed. The New York has four thirty-inch projectors, all operated from the forward bridge. It has been claimed that a great advantage is gained by having the observer at a distance from the projector, as his eyes are not dazzled by the glare, and he may thus be capable of seeing faint objects which he might miss if nearer the light. For navigation purposes also, it might be a great advantage for an officer on the bridge to be able to throw the beam wherever he wishes without the delay or uncertainty involved in transmitting orders. The objections to the system are, however, serious. The first is the necessity for having two commanding positions for each projector, as unless the beam of light can be thrown to any quarter and followed everywhere by the observer no advantage is gained. It is very difficult on most of our ships to obtain good positions for the projectors alone. The control stand must have some protection afforded from weather, and yet, if housed in, the view of the operator becomes restricted. If it is necessary for him to operate by commands received from another, the advantage of electrical control is lost, as the orders could be given with equal facility by word of mouth, voice tube, or telephone to an operator at the projector itself. The second objection to electrical control is in the complexity and weight of the mechanism. The weight of each projector and its accessories aboard the New York is about 1980 pounds, and although this will be reduced in future manufacture, the weight would probably be as great as that of two hand projectors occupying the positions of search-light and control stand. The control mechanism is complicated, liable to injury and difficult to repair with the facilities aboard ship, involving also considerable expense in installation.
It will have to be determined by experience whether the advantages of electrical control from a distance are sufficiently great to compensate for the increased expense and complexity, the extra weight involved, and the double space occupied.
In working the new type automatic lamps but few precautions are necessary. When used automatically, the lamp should always be left in good order after use, with a well formed crater, the carbons properly placed and in contact, and the spring adjusting the potential at proper tension. A good light will then be obtained in a few seconds after switching on the current. In working automatically, much better results will be obtained by having an operator in position to inspect the lamps and correct any fault that may cause a serious diminution of light. When working entirely by hand, the carbons can be brought into contact without risk of any injury to the dynamo, as the series magnet of the lamp immediately separates the carbons striking the arc. The carbons should not, of course, be held together when the current is on.
The Mangin projector was for many years unequalled in the power of the beam it produced. Many other kinds had been tried, but none could compete with it successfully. The Mangin has the advantage of having both its surfaces sections of spheres, and the spherical curves greatly facilitate the process of grinding and polishing. The mirrors have considerable thickness, supplying the strength necessary in service. The principal disadvantage is the fact that the mirror subtends a small solid angle at the focus, and that much light is lost by falling on the projector barrel instead of on the glass. Recent mirrors have a shorter focal distance.
Parabolic mirrors furnish theoretically the simplest type for search-light use, but glass parabolas are difficult to construct, and metal reflectors are worthless from the rapid deterioration of the reflecting surfaces. Recently Schuckert, of Nuremburg, has succeeded in making parabolic glass mirrors with a sufficient degree of accuracy for search-light work, and has for several years supplied all used in the German Navy, besides furnishing a large number to other governments. The Schuckert mirror is a parabola, with its two surfaces parallel and silvered on the back, the thickness of the glass being only that necessary to provide strength, about four-tenths of an inch. It is easy to construct a parabola with any desired focal distance, while preserving the relative lightness of the mirror. The makers favor the use of a large mirror with a short projector, the standard size in the German Navy being ninety centimetres, the focal distance being either 35, 42 or 45 centimetres. The Schuckert lamp is of the horizontal type, the carbons being in line. The current in the navy 90-centimetre projector is either 120 or 150 amperes, with a difference of potential of 60 volts at the arc. The positive carbon is 38 millimetres in diameter, the negative 26, and the length of the arc about 18. The projector is fitted with double diverging lenses, by which the divergence of the beam may be varied at will, and is generally arranged for electrical control from a distance. The lamp can be operated either automatically or by hand.
It is to be hoped that the Schuckert projector may soon be given a practical trial in our navy, to determine its relative efficiency under service conditions when compared with the Mangin. Theoretically the parabolic mirror has decided advantages, but it is probable that these may be greatly modified by considerations of manufacture and use. The parabola is more difficult to grind and polish, and consequently costs more, while perhaps more liable to mechanical defects than the Mangin form. The mirror is most satisfactory when large, but must then be thin to reduce the weight, and is thus rendered more liable to injury. These points could, however, be waived if it were conclusively shown that the Schuckert projector was more powerful and equally well constructed.
Avoiding any theoretical discussion, it is necessary here to only illustrate the difference between the mirrors, leaving all questions of utility to be determined by trial. Figure 6 shows two mirrors of the same diameter, one Mangin and the other Schuckert, the focal distance of the former being twice that of the latter. A horizontal lamp is used and the radii of the curves show the intensity of the light on each angle. As drawn, the parabolic mirror P practically receives all the light radiated from the arc, while much less falls on the Mangin mirror M. In order to make the latter effective it must be brought nearer to the focus, but as there is always a fixed relation between the radii of curvatures of the Mangin mirror, the refractive index of its glass, and the focal distance, the latter can be varied only within limits without sacrificing the desired concentration of light. The Schuckert mirror, on the contrary, can be made with any desired focal length, the concentration becoming only slightly less as the focal distance is decreased.
The limit is found in the practical considerations of leaving room for the negative carbon of the lamp, and of not overheating the mirror. In practice the 90-centimetre Schuckert projector, using a current of 120 to 150 amperes, has about the same focal distance as the 60-centimetre Mangin, using 75 to 90 amperes. The solid angle, subtended by the mirror at the arc, is always greater with the parabolic form, and other things being equal, the amount of light reflected is greater.
Another point already alluded to, on which practical tests are very desirable, is the relative efficiencies of inclined and horizontal lamps under service conditions. The latter seem to offer great advantages for automatic use, as the action of the lamp is better and the movement of the crater on the positive carbon less when the carbons are in line than when they are displaced, as in the inclined lamp. The artificiality of the latter arrangement almost precludes any automatic working. If the horizontal lamp is used it gives the parabolic mirror another advantage, as the focal distance of the Mangin mirror is comparatively so great that much of the surface is in the shade of the negative carbon and is useless as a reflector.
Figure 7 shows the type of electrically controlled 90-centimetre projector in common use in the German Navy. In the latest type there is less gearing, and most of it is concealed in the base.
There is probably no question of naval outfit in a more unsettled condition than that of the utility of search-lights. It may safely be said that most officers think they are more trouble than they are worth, and yet when a decision must be made, few, if any, would advocate their abolition. The fact that they are retained in every principal navy in the world is proof that they are thought to be a necessity. The uncertainty as to their advantages is probably very largely due to the fact that officers obtain but little experience in their use as military weapons. The weekly practice is generally confined more to a test of the actual condition of the apparatus than to solving any of the tactical questions involved.
Lieutenant Bacon, an experienced torpedo officer of the British Navy, in a short article published after the manoeuvres of 1892, stated the conditions underlying search-light work with exceptional clearness, and no better presentation of the subject can be made than in his own words:
"Objects on the water or against the horizon are visible solely by the contrast of their color or shade with that of their surroundings. Were the object exactly similar to its surroundings, naturally it would appear but as part of them and be invisible. In bright sunshine it is impossible to imitate water so exactly in its varying color and shades as to render a boat absolutely invisible upon it; so much light is always present ready to be reflected from all parts of an object that slight variations are easily detected, but at night this is different. Smooth water at night, when lit up by a search-light, appears very dark indeed, and naturally so, since to an observer near a projector but few rays are reflected back from the small ripples; the water, therefore, appears almost black. Now place on this an object which also reflects but little light to the observer, such as a 'dead' black boat (one painted with black lead), and the contrast between the two is almost nil; in fact, so nearly may the two sets of reflected rays match one another as to leave the boat, though in the centre of the beam, practically invisible from the ship. It is not sufficient to merely let the beam fall on the boat, but to see it, it is necessary to produce a distinct difference between the light reflected by the boat and that reflected by the water. This at once explains a very common occurrence to all who have taken part in night attacks—that is, to find your boat brilliantly lighted by a search-light beam, and yet for the beam to pass by harmlessly without the boat being observed by the ship. Nor is this all. Observers on board the ship are, to a certain extent, dazzled by the brilliancy of the beam even when at some considerable lateral distance, since the particles of moisture or dust in the air reflect and diffuse light to a considerable extent, and an effect is produced similar, but, of course, to a less extent, to looking out of a light room into a dark night; the exact extent of the effect is determined by the angular position of the observer from the projector and object. There is yet another effect that limits the distinct vision of an observer near a projector, and that is the fact that when near a projector the rays reflected from an object have to travel back for some distance through the beam—that is, they suffer from considerable interference. The result is that the object appears indistinct. It will be easily seen that all these effects greatly reduce the efficient use of a search-light when used to pick up an object if the observer is placed near the projector, as is usually the case on board ship, especially when the object is traveling towards the ship at a high speed. The practical proof of this is shown by the fact that a boat is rarely itself recognized in the first instance at any great distance, but originally attention is attracted to her by her bow wave—that is, the wave of foam, which reflect a comparatively large amount of light and shows up distinctively against the surrounding dark-colored water.
"Were the above case, viz., that of detecting boats from a ship by means of a search-light in the ship the sole use of that weapon, it might, indeed, be an open question whether the electric beam is any material good in naval warfare, for there is one point which must never be forgotten, and that is that the fact of burning such a light reveals the position of the ship to an attacking boat, when, by an absence of all lights, the darkness of the night might be a more efficient protection. It is this particular case which is the one that correspondents in manoeuvres are most frequently brought into contact with, and on it opinion is so much divided that in open water it is probable that no captain would promiscuously use his search-light as a method of defense. Happily, however, other occasions arise in war time when the searchlight is of great value. These may be divided roughly under two heads: (1) Defense of a fleet in close waters, where the position of the anchorage water of itself reveals the probable position of the fleet; (2) Fleet action at night, or action between single ships.
"Regarding (1) the case may be briefly stated as follows: Ships are at anchor in a close anchorage for some particular reason, either for repairing or refitting purposes, or else the close presence of a superior force. For it is inconceivable, considering the deadly nature of a torpedo-boat attack, that an officer would keep his fleet at anchor at night in a harbor that afforded by its position a possibility of an attack from torpedo-boats if he could possibly avoid it. Supposing such the case, the search-light may be a most efficient help against such an attack. Light may be burned at a distance from the ship on shore to light up the entrance or different sections of the passages through which the torpedo-boats will have to pass, or the ships may illuminate the passages and use observers in their guard-boats. In either case the effect is the same, for no longer has the light reflected from the attacking boats to be reflected at a small angle to the observer, but the observer is placed in a far better position as regards reflection, and far more minute differences in shade can be observed, since the reflected light is much greater. In other words, a greater contrast is produced between non-reflecting bodies, and the water gun-fire can, therefore, be more effectively used and greater havoc effected on the attacking force. In such a case one thing is certain—for detecting boats, all beams must be fixed so that observers may get their eyes accustomed to a constant intensity of light, and not a variable one, as produced by 'sweeping' beams. This is the most effective use of an electric search-light. A few beams may, of course, be kept in reserve for following up and illuminating boats after they have been detected, but these should be quite independent of the fixed beams. The guard-boats, some two thousand yards away, can see the approach of others several hundred yards off, since the surrounding water is well illuminated by the beams of light playing on it. When the boats are discovered, it is necessary to light them up sufficiently for the guns' crews in the ships or boats to use their guns. This may be done by the ships' projectors with divergent lenses in conjunction with parallel beams, the divergent lenses being used to light the boats up, and from their large divergence prevent the chance of losing sight of them again, and the parallel beams to more brightly illuminate them and make them more clear to the guns' crews.
"Again, in single ship or fleet action at night search-lights must be used to light up the enemy, and now the search-light is no longer an objection as showing the position of the ship, since its height prevents an accurate estimation of the position of the waterline (the most vulnerable position in the ship to fire at), but in reality it is absolutely safeguard, both since it is misleading as to the distance of the ship and also is blinding to the opposing guns' crews. From the foregoing remarks it will be seen that the function of the search-light may be either to discover or to light up an object for guns to fire at. These two different uses are important, since they lead to the height of the light being varied to some considerable extent. It has long been the opinion that for discovering boats the beam should be low, since its direction is parallel to the surface of the water, and therefore lights up a larger area than would be the case if thrown down on the water from a height, when the plane of the water would cut the beam in an ellipse, whose area depends on the inclination of the beam. But a little thought will show that if the beam be parallel to the water the water is but feebly illuminated, whereas a body vertical to the water cuts the beam at right angles and receives a large amount of light. Were the body absolutely non-reflecting, it would be invisible; but if reflecting, a contrast will be obtained. Now that boats can be painted so as to be practically non-reflecting, the low beam is viewed with more disfavor, especially as detection is not the chief function of a light in a ship. The real function of the light on board is to light up an object whose approach has been detected, so that the guns' crews can accurately lay their guns. To do this, it is necessary to keep the light as far from the guns as possible, and also to keep the beam off the water near the ship, so as to prevent the men being dazzled. The best position in the case would appear to be, to have the light above the guns, so that the crews practically do not see the beam, except where it strikes the object, and also should the opposing ship use the light to aim at, the shot will be more likely to pass over the ship. Another great point is to have the projectors so placed as not to illuminate any portion of the side or superstructure of the ship; this, again, is best obtained by a fairly high position of the light.
"For attacking forts or in action in moderately smooth water, a light in the military tops would seem to answer these requirements; but, of course, with even a moderate motion of the ship a beam from a projector so placed would sweep through a far larger arc than one in a lower position, and would, therefore, be far harder to keep steadily illuminating an object. For navigating purposes, where the water close to and for some distance ahead of the ship has to be illuminated, a low position of the light is best."
A brief consideration of the preceding is enough to show that much of the uncertainty existing in the minds of our officers as to the value of search-lights is due to the fact that no experiments are made to test the conditions of their efficiency or to develop proper methods of utilizing them. The search-light is a weapon of offense as well as of defense, and requires more judgment and skill in its control than any other weapon entering into the military equipment of a ship. If used at the wrong time it may lead to the destruction of the ship by betraying the exact position to a powerful enemy; if used properly it may save the vessel by depriving an attack of the darkness necessary for its success. Nothing but practice under conditions as nearly as possible simulating those of warfare can give officers clear opinions of how the search-light enters into the fighting power of the vessel. Weekly practices may develop interesting electrical facts as to the light obtainable from heavy currents, but these can be carried on as well from a wharf as from a ship's deck. The many tactical questions, such as how to use the search-lights under different conditions of the atmosphere or of the sea; how to utilize them in case of a vessel at anchor in a harbor; how when lying in an open roadstead; the advantages obtainable by dazzling an enemy and preventing his sighting his gun; the best way of avoiding detection by a search-light; the best method of stationing observers; the practical advantages or disadvantages of electrical and hand control for projectors; the utility of diverging lenses; and, above all, the conditions under which search-lights should not be used, are all matters of the greatest importance, concerning which but few officers have opinions based on experience. A practical investigation of these mooted points is possible aboard almost any ship in commission, and a course of experimental drills would be most valuable in the practical information it would develop.
STANDARD SWITCHBOARD.
The specifications for the system of switchboard connections have gradually been growing more complex from year to year. In the specifications for the Columbia and Olympia they read, "the design to be such that the dynamos may be run either singly or in multiple on the incandescent circuits, either singly or in multiple on the arc light circuits, and either singly or in multiple on the arc and incandescent circuits connected in parallel." The standard switchboard was devised to meet these specifications without taking up valuable space.
The switchboard has been briefly described in General Information Series, No. XI., but further details may be advisable. The system is the same, whatever may be the number of circuits or dynamos installed, so that a person familiar with it can operate the switchboard at first sight.
Figure 8 is the switchboard of the Bancroft arranged for two dynamos and ten branch circuits, and shows the most symmetrical and simplest arrangement. Figure 9 illustrates the switchboard for three dynamos and eighteen branch circuits. The board has on each of its edges a vertical bus bar fitted with fuse clamps and terminals at the lower ends, the two bars being connected by a cable on the back of the board. The bus bar thus formed is common to all the dynamos, and in practice is connected to the positive terminals. Inside of this bar are the fuses on the positive ends of the branch circuits. On each side of the centre of the board there is a set of vertical bars, each set containing as many bars as there are dynamos. The similarly situated bars of each set, counting from the left as number one, are connected together by a cable on the back of the board, forming the negative bus bar of the same dynamo. Between the two sets of negative bars are the negative branch circuit fuses.
One-half of the branch circuits are placed on each side of the vertical center line. The positive terminal of each is in the clip near the common bus bar, in which the switch shown in the figure engages. The negative end of the circuit is in the socket, in which fits the shifting plug switch near the center of the board. One of these sockets is placed between each two of the negative bus bars. The plug switch has two small bosses on each side which serve to clamp it in the socket in the open position at right angle to the face of the board. By withdrawing the plug slightly the bosses are disengaged, and the switch may then be thrown to right or left, connecting its circuit to either one of the dynamo bus bars between which it is placed. In a board designed for three or four dynamos the plug switch may be shifted from one socket to another so as to be placed in position to connect its circuit to any bus bars.
The switches for connecting the dynamos in parallel are at the bottom of the board. They are double pole, the upper half connecting the negative bus bars, while the lower half connects the two equalizing bars. The board for two dynamos has only one multiple switch, but those for three or four dynamos have a switch for each dynamo. In the latter case, the upper half of the switch connects the negative bus bar of its dynamo to a common negative bus bar, while the lower connects the equalizing bar of the same dynamo to a common equalizer. It is then necessary to close two switches to connect two dynamos in multiple, the connection not being complete until the last switch is closed. When two or more dynamos are operating singly the multiple switches are open, and hard rubber guards are placed in the clips to prevent any accidental closing.
As one pole of each dynamo is connected to a common bus bar it can be entirely cut out of circuit only by a double pole switch on the headboard of the dynamo. This switch is therefore a necessity when the standard switchboard is used, and should always be open when the dynamo is not in operation.
All connections are made on the back of the board. Each conductor is soldered to a conical copper plug. The end of this plug has a screw thread tapped in it, and fits into a corresponding conical socket in the bus bar or fuse terminal, being secured in place by a screw inserted from the front of the board. When this is screwed up, it makes a good and reliable contact.
The design of the switchboard is very simple. All vertical lines are dynamo circuits, all horizontal ones branch circuits, the plug switches being at the intersections. The following directions are from the official instructions:
1. No metal, such as screw-drivers, monkey-wrenches or other tools or watch-chains, should be allowed near the front of the board. Any neglect of this kind may cause dead short-circuits.
2. Putting more than one plug switch on any one section is positively forbidden, as the use of two might put dynamos in parallel without any equalizing bar.
3. If any section is not wired up, it is advisable to remove its plug switch entirely.
4. In throwing a section in circuit it is advisable to first place the plug switch in proper position for putting the section on the dynamo desired, and then close the switch on the common bus bar. In cutting out a section, first open the common bus bar switch and then the plug switch. In changing a section from one dynamo to another, it is best to first break the circuit with the common switch, then throw the plug switch over into proper position and close the common switch. The plug switches are not intended for throw-over switches.
5. When any section is wired up, but not in circuit, its common switch should be open, and its plug switch open and locked.
6. When the dynamos are operating singly the multiple switches should be open and the spring clips covered by hard-rubber covers. These covers to be removed before attempting to connect in parallel.
The standard arrangement of switch and instrument boards is shown in Figure 9. The left hand instrument board carries at the top a ground detector fitted with a multi-throw switch, having as many contacts as there are dynamos. By connecting this with one of the negative bus bars, starting the corresponding dynamo and throwing any plug switches on the same bus bar, each and all branch circuits can be tested for grounds, or an existing ground may be speedily localized. On the middle of the board is a vertical reading Weston voltmeter. One of its terminals is connected to the common bus bar, the other to a multi-throw switch by which it can be put in connection with any one of the negative bus bars. By moving this switch, the potential of any one of the dynamos may be measured and the polarity also tested, the voltmeter needle showing the direction of the current in the instrument. This voltmeter is always kept in circuit and can be read from a distance. The lower voltmeter is connected so as to give the difference of potential at the terminals of the search-lights, having a multi-throw switch with as many contacts as there are projectors. It is identical with the other voltmeter, and is in reserve in case of accident. As its use on the search-lights is unimportant in case of electrical control projectors, it may be used for other purposes.
Underneath the switchboard are the shunt field rheostats, the left hand being in circuit of the dynamo connected to the left hand bus bar, which in turn is connected to the left terminal of the voltmeter multi-throw switch. The instrument board on the right of the switchboard carries the ammeters, one being in circuit of each dynamo. It is highly desirable that the same system of connections should be followed on all switchboards, to enable persons familiar with one plant to take charge of another at sight. The system shown in Plate I. has therefore been approved by the Bureau of Equipment. It is, that the same sequence should always be followed from left to right on the negative bus bars, voltmeter terminals and shunt field rheostats, and from top to bottom on the ammeter board. Thus number one dynamo would have the left hand bus bar, left hand voltmeter terminal, left hand rheostat and upper ammeter; number two dynamo would have the second bus bar, voltmeter terminal and rheostat, counting from the left and second ammeter from the top. A stranger, then, on entering a dynamo-room could by a glance at the ammeters see which dynamos were carrying load; the switches on the common bus bar indicate what branch circuits are in operation, while the position of the corresponding plug switches shows to which dynamo each branch circuit is connected. Finally, the multiple switches would show whether the dynamos were operating singly or in parallel, and if the dynamos were numbered, the whole situation could be grasped in a few seconds, and the stranger could operate the plant. At present many ships have switchboards which can hardly be understood with the assistance of plans, and no two are identical.
An examination of Fig. 8 shows that the condition of the Bancroft's circuits is as follows, neglecting the fact of the absence of fuses. Commencing at the top and numbering the left hand circuits from 1 to 5 and the right hand from 6 to 10, we have circuits 3, 4, 5, 9 and 10 on number one dynamo; circuits 6 and 7 on number two dynamo; circuits 1, 2 and 8 open; both dynamos operating in parallel.
The operation of compound dynamos in parallel seems to be practically confined to the United States, where it is almost universal in all large power stations. The difficulty encountered in operating in parallel is, that any inequality between the electromotive force of the two dynamos results in the one of higher potential reversing the series field of the other, and driving it as a motor. This is evident, from Fig. 10, any current from the upper dynamo to the lower passing through the series coil of the latter in the direction opposite to that of normal excitation. The series field would therefore be reversed. This trouble is easily removed by connecting the brushes of the two dynamos by a conductor of very low resistance called an equalizer, as shown by the heavy dotted line. If, then, the electromotive force of one dynamo becomes higher than that of the other, current flows through the equalizer and through the series field of the weaker dynamo in the right direction, strengthening it and raising the potential. The vital point of safety in the system is to have the resistance of the equalizer so small that the greater part of the current may flow through it from brush to brush, instead of in the parallel circuit formed by the series fields and the mains. The operation is most successful when the dynamos are similar, but this is by no means essential, dynamos having been operated in parallel whose normal outputs were in the ratio of thirty-eight to one. The presence of the rheostat in the shunt field is of the greatest assistance. The ammeters are connected on that side of the circuit opposite the equalizing bar, so as to show the total current passing in the external circuit. If, then, one shows very much higher than the others, it is probable that it is feeding them through the equalizer. A slight adjustment of its rheostat weakens the shunt field, and by reducing the voltage of the stronger dynamo brings all nearer equality. The attendant can, therefore, by an inspection of the ammeters and adjustment of shunt rheostats make each one of the dynamos take its proper proportion of the load, although they differ widely in size, speed, type and output.
The following directions are given for connecting dynamos in parallel on the standard switchboard:
"In throwing dynamos in parallel care will be taken to see that each machine is poled right and kept on open circuit until the normal potential is reached before being thrown into circuit. The operation of throwing dynamos in parallel with the naval standard switchboard is as follows: Each dynamo must first be tested separately. It is brought to the proper potential of 80 volts in light load by adjusting the shunt field rheostat. Full load is then thrown on and the voltage should remain the same. If not, it should be adjusted to give the proper voltage by the shunt to the series field. This adjustment should be once made and occasionally verified; too much care cannot be given to it, and once adjusted should not be tampered with. With loads on each dynamo practically the same, and the difference of their potentials less than one volt, the multiple switch may safely be closed and the dynamos will work together. Any inequality of load between them may then be rectified by the shunt field rheostats. The polarity of the dynamos will be shown by these measurements, and, as has been said, should be the same. When one dynamo is to be thrown in parallel with another actually working, it may be advisable to divide the load about equally between them, repeating the observations as to potential before throwing them in parallel."
In explanation of these directions, the polarity of the dynamos is shown by putting the voltmeter first on one and then on the other. The needle should indicate the same voltage on each. If it moves in opposite directions, one dynamo is reversed. Great care should be taken in connecting up the leads and equalizers the first time, to get everything in the right place. Once permanently connected right, no error could ensue except from a reversal of one of the dynamos, and this is shown on the voltmeter. In ships in which the dynamos are frequently operated in parallel, it might be well to connect the lower voltmeter on the instrument board to the equalizers instead of to the search-lights, especially if the latter have individual control stands. In this case, one terminal of the voltmeter would be connected to the common equalizer, and the contacts on the multi-throw switch to the individual dynamo equalizers. The switch should not be closed if there is more than one or two volts difference between the two equalizer sides.
In order to secure good working with the standard switchboards the equalizers from the dynamos must be large. A safe working rule would probably be to make them of the same cross-section as the dynamo leads.
The standard switchboard for three dynamos is shown in Plate I. with all its connections, made as authorized by the Bureau of Equipment.
MOTORS.
In General Information Series, No. VII., June, 1888, the writer dwelt at some length upon the advantages attendant on the use of the electric motor on shipboard. Since that time they have been largely introduced in different navies for hoisting ammunition or for training light guns, as well as for ventilation purposes. It is unquestionable that they will come into more extended use for these purposes as well as for others, but a wise policy will limit their use to position, in which they are actually superior to any other type of machine. Thus, it would probably be inadvisable to place electric motors anywhere in the engine or fire-rooms, as the combined heat and moisture would cause speedy deterioration, and in such locations, moreover, the presence of steam-pipes is no objection. One of the most promising immediate applications of the electric motor is in the operation of ventilating fans, and this problem is now under consideration. The presence of steam-pipes always carrying live steam is a nuisance on a berth-deck, directly affecting the health and comfort of the whole crew. In some of the very best and latest ships of the Navy, however, the ventilating fans are placed on the berth-deck and are driven by steam engines supplied with steam through long leads of pipe. This method heats the living space of the crew more than it cools it. From a military point of view it is vitally weak, the whole or a greater part of the ventilating piping, on which the fighting parts of the ship must depend for air in action, being wholly unprotected against machinegun fire which would riddle the pipes, utterly destroying the ventilation system. In some ships, moreover, immense air ducts are carried through the coal bunkers, displacing a hundred or more tons of coal and cutting up the remaining space so as to greatly interfere with coal passing. Nor can it be considered good policy to cut holes two or three feet in diameter in water-tight bulkheads, even if they are protected by automatic valves. The system thus outlined seems to interfere with both the military efficiency and the sanitary well-being of the ship. If it were necessary it could be tolerated, but every objection noted can be overcome by the use of electric motors, with a probable gain also in efficiency and weight. The general method will be outlined only.
The system would be to ventilate each compartment between main water-tight bulkheads independently, placing in each two small electric ventilating blowers. These could be used as either exhaust or supply at will. The advantage of placing the blower in the compartment to be ventilated is that no injury to the piping above can cut off the supply of air, as air now enters through any holes in the pipes instead of escaping. A general feed pipe might run fore and aft on each side of the berth-deck, from which all vertical pipes leading to lower deck compartments could lead, the feed pipe being connected to the outer air by ventilators placed where convenient. This method prevents heating of the ship by steam fans, saves water-tight bulkheads and coal bunkers for their legitimate uses, secures ventilation of magazines and fighting spaces in action, and would probably be fully as efficient both in actual ventilation and in the energy required to produce it. It cannot be carried out with steam fans, the necessary net work of hot pipes heating the ship beyond endurance, but the electric motor will work in any place in which it can be set up, and neither in itself nor in its connections gives off any perceptible amount of heat.
Fig. 11 shows two ventilating fans operated by electric motors, supplied by the General Electric Co. for the Oregon. The motors are ironclad, the coils being wholly enclosed. A few slight changes were recommended in construction, but the general operation was very satisfactory.
One motor was connected to a Sturtevant exhauster No. 3, the whole set weighing 254 pounds and supplied 1320 cubic feet of air per minute at the nozzle, at a speed of 1520 revolutions. The energy consumed was 10 1/2 amperes at 80 volts, or 840 watts.
The other motor was connected to a Sturtevant exhauster No. 4, the whole set weighing 446 pounds. It supplied 2050 cubic feet of air per minute at the nozzle, at a speed of 1540 revolutions per minute, taking 21 amperes at 80 volts, or 1680 watts.
Both motors were series wound and started without any starting rheostat. The speed of the motors would, of course, be higher and the delivery of air smaller if piping was connected to the exhauster.
If problems presenting themselves in naval construction are impartially considered, it will be found that the electric motor has an unquestionable field of usefulness on shipboard. Much conservatism has been felt in relation to the use of untried or experimental machinery, but the electric motor can no longer be thus classed. Its widely extended use during the last few years, and the durability and general good working it shows under most unfavorable conditions in shore practice, are guarantees that it can be made reliable on shipboard.
It may not be amiss to consider the subject in the light of our necessities.
All motors on shipboard will naturally fall under the head of constant potential motors, supplied from the ship's mains, and either series, shunt or compound motors may be used. The series motor is valuable for its great starting torque, but has the disadvantage of an irregular speed. In many cases, as in training guns or in other heavy work, steady speed is of no consequence, while the ability to start with a heavy load is all essential. For work of this kind the series motor is the best. If thrown in circuit directly across the mains, it might burn out, the resistance being so low as to allow of an enormous current passing when the motor is at rest. As soon as it commences to move, the current is reduced to safe limits by the counter electromotive force. It is advisable, however, to interpose some resistance to save the motor, this being generally in a starting rheostat, the speed of rotation when once in motion being similarly controlled by resistances. An objection to the series motor is the fact that the speed rises dangerously high if the load is thrown off, the excitation being diminished by the rise of the counter-electromotive force of the motor. The motor makes the vain attempt to attain a speed at which its counter-electromotive force equals the difference of potential of the mains, each increase of speed diminishing the field current and counter-electromotive force and calling for higher speed to attain the voltage of the mains. Theoretically the speed of the series motor under such circumstances is infinite, but the practical limit is found in the fact that friction and internal losses always make something of a load, so that no motor ever runs absolutely light.
The shunt motor, when made with low armature resistance, preserves a practically constant speed with wide variations of load. It possesses, however, the great defect of having no starting torque. The field coils and armature being in parallel, the former are short-circuited by the latter, and if the motor is suddenly thrown into circuit, the field would not receive enough current to magnetize it, while the armature would probably be burned out by the excess of current passing. Some special starting device is therefore necessary, and a starting rheostat is commonly provided, by which a current is passed through the field coils before the armature circuit is closed at all. The next step is to send the current through the armature in series with a resistance, the latter being gradually cut out as the counter electromotive force rises from the speed increasing. The special purpose for which the shunt motor is adapted is, therefore, starting with a light load and afterwards preserving an approximately constant speed. The starting rheostat is necessary and is objectionable as an extra fitting. There have been many cases in the service where these rheostats have been used in continuous running to diminish the speed of the motor, but not being designed for this purpose they overheat and finally burn out.
The apparent paradox, that the speed of a motor is increased by weakening the field and consequently the counter electromotive force, must always be borne in mind in considering the speed of motors. The fact that the shunt field remains excited when all load is thrown off, keeps the speed down to near its normal value. If the speed could increase, the field being kept constant, the counter-electromotive force of the armature would rise until it reached the potential of the mains. It could, of course, go no higher without running the generator as a motor. The actual limit attained is less than that of the mains, as there is always internal loss and mechanical friction in the motor, energy to overcome which must be derived from the mains, and the speed settles at such a point as to allow the motor to receive just this amount of energy.
As the compound dynamo preserves a constant voltage at its terminals with constant speed but varying output, the reversibility of the dynamo and motor would lead us to expect that a motor wound with both series and shunt coils on its field magnets would preserve a constant speed at constant potential with varying load, and such, within limits, is found to be the case. An important difference must, however, be considered. In the compound dynamo, the magnetic effect of the series coil is to reinforce that of the shunt, but if the compound dynamo is used as a motor the series coil will be found to act in opposition to the shunt. While this assists materially in preserving the speed constant when once established, difficulty is found in starting the series coil, which is practically the only one operative when the motor is suddenly thrown into circuit, and tends to start the rotation in the wrong way. As the shunt field "builds" it acts in opposition to the series, and this tendency to rotation in the wrong direction is overcome. It will readily be understood that under such conditions the motor can have no starting torque, and if heavily loaded will burn out. This difficulty may be overcome by reversing the series coil, making it co-operate with the shunt coil. This arrangement gives powerful starting torque, and prevents the speed from rising too high when the load is thrown off, but the constant speed is sacrificed, although the variation may be so small as to be unimportant.
The compound wound motor, with both its coils acting in the same direction, possesses therefore the best points of both series and shunt motors, and is advantageous where starting torque and fairly uniform speed are desired, and especially where there is a probability that all load may be thrown off suddenly. If constant working speed is essential, the series and shunt coils must act differentially, the motor being either started light and load thrown on after speed had been attained, or the series coil reversed, as already explained, to gain starting torque, and thrown in differentially as soon as the motor reaches normal speed. This system is complex, requiring an external control board, but answers admirably in such cases as in operating lathes, where constant speed is essential.
The foregoing is sufficient to enable a conclusion to be reached as to the use of motors on shipboard. If starting torque is essential and constant speed is unnecessary, the series motor is the best, an adjustable resistance being inserted as a protection on starting. For purposes calling for constant speed the shunt motor is best adapted, having a rheostat for use only in starting. In special cases the double wound motor, possessing in a lesser degree the advantages of each of the others, may be best, but it is noticeable that its complexity is preventing its wide use in commercial practice.
A writer in General Information Series, No. VIII, in referring to the use of motors on shipboard, dwells at some length on the disadvantages arising from the increase of dynamo plant, and also on the danger resulting from having all the auxiliary motive power of the ship dependent on the dynamo-room. These objections must be considered in any scheme of electrical power transmission on shipboard. The following general plan is suggested for consideration. Electric motors will be advantageous,—
1. In all places where auxiliary engines operate continuously for extended periods, the steam and exhaust pipes necessarily passing through store-rooms or living spaces.
2. For ordnance purposes, where the demands for power are not excessive.
3. In isolated places such as military tops where power is necessary.
Electric motors are inadvisable:
1. In engine or fire-rooms, where steam is always available, and the conditions are unfavorable for the durability of motors.
2. In general, where the demands for power are excessive and steam is available, as in the windlass engine.
3. Where rapid and frequent reversals are necessary, as in the steering engine.
The question of safety of the dynamo plant can be met only by taking all precautions for the protection of the dynamos and wiring, and by operating the dynamos in parallel with a large reserve of power. If the development of ordnance applications of electric motors should become of great importance, a further guarantee of safety should be obtained by dividing the dynamos between two widely removed dynamo rooms.
COMPARISON OF SPECIFICATIONS.
Not only are different types of dynamo used in different navies, but the methods of operating the plant also vary. In nearly every service the voltage is now standardized at eighty volts, this being a compromise between the high voltage desirable for incandescent lamps, and the heavy demands for current made by the search-lights. In any case, the type of dynamo and engine actually adopted depends entirely on the specifications to be met. Although those of European navies are not widely published like our own, they are all in print, and no confidences will be betrayed by referring to them.
The British specifications dwell largely on the heating of the dynamo. The common provision is that “at the end of a six hours’ trial at full load, and one minute after slopping, no accessible part of the armature or magnets shall have a temperature more than 30° Fahrenheit above the temperature of the dynamo-room taken on the side of the dynamo remote from the engine and three feet distant from it, and that the maximum rise of the temperature of the armature at the end of the trial shall not exceed the temperature of the dynamo-room more than 70° Fahrenheit." The wording of these specifications is somewhat peculiar, and the first proviso seems to be utterly inconsequential, as the vital question is certainly not what the temperature is "one minute after stopping," but rather what it is while the dynamo is in operation. The temperature shown by a thermometer in one minute depends as much on the thermometer as on the actual temperature of the hot body, and with the latter remaining the same, the thermometer indications vary with the instruments used. The second proviso limiting the maximum temperature of the armature to 70° above the air is definite and intelligible. It is urged that the armature is hotter just after stopping than when in operation on account of the absence of cooling by the fanning of the air, but as the generation of heat ceases when the dynamo stops, it is difficult to imagine any rise of temperature. Most of the British dynamos have large flat field coils, and in case any curiosity exists as to the temperature of the field coils while in operation, instead of "one minute after stopping," it can be determined with considerable accuracy by having thermometers strapped on the coils throughout the test, covering the bulbs with waste or flannel. The armature temperature must be determined in the same way after stopping, no other method being available. The following temperatures are from six hours' tests at New York Navy Yard, made on Siemens dynamo built under Admiralty specifications.
Heating of Shunt field. Heating of Heating of
In 1 minute. Max. Series Field. Armature.
200 amperes, 21° 40° 35° 46°
300 amperes, 16° 33° 38° 43°
400 amperes, -- 39° 39° 35°
The British specifications contain many minor provisions, but leave the design and construction of the dynamo almost entirely to the maker. The prevailing type has been for several years the bipolar vertical magnet.
The engine specifications call for vertical, inverted, direct, double acting engines. These are open, the closed in types having been abandoned. The most important proviso is that the governor shall be so arranged as to allow the speed to be varied while in operation. This necessitates a throttle governor and causes an absolute sacrifice of what, in the estimation of the American electrician, is the most important consideration, viz., close regulation. Another consequence is the advisability or even necessity of changing the setting of the valves if the working steam pressure is largely varied, or if the exhaust is changed from atmospheric to vacuum. The throttle governor, while able to reproduce a speed after a sudden change of load, can act only when a large momentary increase has taken place. This is, of course, attended with a corresponding rise in the voltage of the dynamo, injuring any incandescent lamps left in circuit. The Admiralty specifications allow a temporary increase of speed or "jump" of 25 per cent.
As this system is almost universal in Europe, it may be of advantage to contrast it with that used in the United States Navy. Our specifications allow of a variation of speed of only five per cent, when the whole load is thrown off the dynamo, and although this strictly applies to the difference between the steady speeds of the engine loaded and light, it is seldom exceeded even in the "jump." The latter is directly controlled by another proviso that the dynamo voltage as indicated on a dead-beat voltmeter shall not show a variation of more than ten volts when full load is thrown on or off. This closeness of regulation is obtained by the use of what is commonly known as the "automatic governor," acting directly on the eccentric and changing the throw of the valves, altering the point of cut off, and also adjusting the compression to suit the change of load. No commercial electric light engine with a governor operating on any other principle could be sold in the United States, while only two naval dynamo engines were seen by the writer in Europe in 1891 that used it. The indicator cards shown in Figure 12 illustrate the capacity of this variety of governor to adjust the working of the engine to variations of steam pressure and exhaust. The reason for the retention of the throttle governor is found in the fact that European dynamos are self-contained, the field circuits being entirely on the frames. As the dynamo heats the voltage can be adjusted only by an increase of speed. American dynamos, on the contrary, always have an adjustable exterior resistance in the shunt field circuit, by varying which the voltage may be changed as desired. With a dynamo of this type constant speed is desirable, and automatic engines are therefore invariably used. It is difficult for an electrician accustomed to the facility of adjustment thus obtainable, to conceive of any compensating advantages which justify the abolition of the adjustable rheostat. To him nothing is cruder and practically more inconvenient than the method of adjusting the voltage of the machine by setting up with a monkey-wrench on a nut regulating the tension of the governor spring of the engine. Messrs. G. E. Belliss & Co., the principal contractors for Admiralty dynamo engines, have recently constructed automatic engines for commercial marine use, but so far as is known they are not yet used by the Admiralty.
Returning to the Admirality specifications, another most important proviso is one promoting economy. The is based on water consumption. The contractor, in tendering a bid, guarantees a certain water consumption per hour per electrical horse power, under penalty for each pound of water in excess. The consumption is generally determined by measuring the feedwater. This method is apparently superior to the one called for in our specifications of determining efficiency from indicator cards, as it is simple, practical and not liable to great error. The measurement of feedwater checks up any condensation in pipes or cylinders against the engine, and the consumption is therefore sometimes found by a surface condenser, this process, of course, favoring the engine.
The French Navy seem to have no standard specifications, the requirements varying for different vessels. The heating proviso, which dominates the whole dynamo design in our own specifications as well as in the Admirality, is very elastic, the only demand being that it shell be “tres failble.” This is said to be construed as 35° C. above the air in a six hours’ run, a much greater limit than our own. The controlling feature in the French specifications is that of economy. As in England, this is determined by water consumption per electrical horse power at terminals. The specifications fix consumption which must be obtained under penalty for every decilitre in excess, a limit being also named, the exceeding of which justifies total rejection. A variation of two per cent. is generally allowed in the steady speed of the engine between light and full load. The governor must be adjustable while in operation, and this proviso calls for the throttle governor, with regulation of speed by adjustment of the governor. As a result of the above requirements we find the French dynamo to be small and compact and generally operated by compound engines. Many multipolar dynamos are in use, but in recent installation it is understood that bipolar are preferred.
The tests for acceptance are of two kinds. One is made before the installation of the dynamos on shipboard, generally including two six hour tests at full load, one on atmospheric exhaust and the other on vacuum, each with specified steam pressures. In each case a certain economy must be attained, penalties being exacted for any excess as above stated. It is understood that the tests of the sets for the Marceau gave an economy of 12.2 kilos of water per electrical horse power at terminals, while those of the Neptune gave a consumption as low as 11.7 kilos. At these tests all measurements of insulation and copper resistance are made. If the first tests are successful, the sets are installed on shipboard, and when in place are subjected to another continuous test of twenty-four hours at full load.
Although the French dynamos are not made large by the imposition of a low heating limit, it is noticeable that, in spite of their apparent compactness, they occupy nearly as much floor space as the later American types and are but little, if any, lighter. They are invariably of a high degree of mechanical excellence.
The dynamo apparatus of the German Navy is made either by Siemens Halske of Berlin, Schuckert of Nuremberg, or Kummer of Dresden. Each manufacturer has his own type, most of the recent apparatus having been made by Kummer. One of his types is somewhat like that in use in our service aboard the Philadelphia and Baltimore, but a later type is a ring frame having six interior poles. Incandescent lighting is subordinated, in the German Navy, to the search-light. As the Schuckert projectors in use each take 150 amperes, the practice is to install a 180-ampere dynamo for each projector, leaving but a small margin for simultaneous operation of incandescent lights. The dynamos are never coupled in parallel, and the incandescent lighting is, therefore, comparatively ineffective. Although the general type of apparatus is much like the French, Kummer has made dynamos having adjustable rheostats in the shunt fields and driven by automatic engines.
The foregoing is merely an outline of the prevailing characteristics of the dynamo machinery in the principal foreign navies. The one point which remains for us to notice is that of economy. More stress is laid on this abroad than in our own service, although our specifications call for a percentage of efficiency taken as the ratio of indicated horse power to electrical horse power at the terminals. It is certainly most desirable to have economical dynamo apparatus, but the question of economy presupposes certain working conditions which have to be met. No engine or dynamo is economical when operating at a fraction of its normal output, or with one-half its normal steam pressure, and yet these are exactly the conditions under which our naval dynamos operate two-thirds of the time. Moreover, economy of working depends mainly on the boiler, and as long as the main boilers are used for generation of steam, the consumption of coal must be excessive for the amount of work done. While these considerations will not justify an uneconomical engine or inefficient dynamo, they indicate that economy of actual working depends on many circumstances, and that it may be better practice to design our machinery with special reference to durability and good working under the conditions holding on shipboard. This is the policy that has been followed, and although our dynamo engines are probably not as economical as others, they have a margin of power which, in connection with their automatic governing, makes them largely independent of all variations of steam pressure, enabling them to carry their load on half pressure, and, so far as is possible with a high speed engine, to work with wet steam. Compound engines have been but little used in our navy although not forbidden by specifications, the increased complexity and the ordinary low working steam pressure rendering them of practically no advantage.
American specifications call for the operation of dynamos in multiple, while in other services it is unknown. The advantage of parallel working is in the fact that it minimizes the effect of injury to any one dynamo. In a ship supplied with gun motors and search-lights, the demands made for electrical energy in time of action would vary suddenly and largely, so that it would be impossible to distribute the load on separate dynamos without having great excess of power. In such a system, each dynamo must have a reserve power sufficient to cover all demands made upon it, although nine-tenths of the time it may be working light. This would hold of each dynamo of the plant. By yoking all together, however, we can take advantage of the fact that it is improbable that all motors will simultaneously be working at their maximum, and if not, there is a safe margin of power. The dynamo plant should, however, have an output equal to all demands made simultaneously, and the system of parallel connection would then prevent a total break-down of any motor mechanism from the fact that one or more dynamos may be injured. The advantage of the system is so evident as hardly to admit of question, and the fact that parallel working of dynamos is hardly thought of in any other service than our own must be explained by the difficulty of carrying it out without the automatic engine and the rheostat in the shunt field of the dynamo. Its practicability is abundantly proved by the habitual operation of compound dynamos in parallel in almost every power station in the United States; its military advantages have already been set forth, and justify the use of the method on shipboard.
CARE OF PLANT.
The conditions existing aboard men-of-war are specially trying to dynamo apparatus, and even with the best of care deterioration is more rapid than is usual ashore. The continuous operation of the dynamos, the high temperature of the dynamo-rooms, the presence of salt water and salt air, and the inevitable wear and tear attendant upon the crowded condition of the ships, are all difficulties to be encountered, but intelligent care will materially assist in overcoming them and in minimizing deteriorations.
A great part of the trouble experienced on shipboard arises from the engine. A high speed engine, especially if working with a small clearance, as in the Armington & Sims type, is specially sensitive to water in the cylinders, and but very little is required to cause a break-down. This danger can be much reduced by careful arrangement of steam and drain pipes. The dynamos are now supplied by independent steam pipes, and these are as straight as possible, and without any drop bends in which water may accumulate. Separators are generally placed in the dynamo-room, but cannot be relied upon to prevent all access of water. The engine drains are found to work well on a vacuum if piped directly into the exhaust, but when operating on the atmosphere, must be led into traps, special attention being paid to avoiding back pressure. The engine cylinders are fitted with automatic relief valves, but a valve is frequently put in the drain pipes back of these, and if shut, prevents their action. The dynamo attendants must thoroughly understand the drain piping, and see that the drains are always clear. Care must be taken on starting the engine to drain the pipes and work all the water through at low speed, and as this precaution is obvious, accidents do not often occur then, most of the damage by water being done while the plant is in full operation, and is due to foaming or priming in the boilers, causing a sudden lifting of masses of water into the pipes. Shifting from one boiler to another sometimes causes trouble from water having accumulated in the pipes. Implicit reliance cannot be placed in such cases on either automatic drains or separators, and safety requires the slowing of the engines if the noise of water in the cylinders is serious. The Armington & Sims' engines in service seem to break the crossheads from the effects of water oftener than any other part. The only accident of this kind, within the knowledge of the writer, happening to one of the new engines, resulted in breaking the piston.
Heating of crank pins and bearings may, of course, be due to the shaft being out of line, or to bad fit of the brasses. The latter is easily remedied; the former is not liable to occur, as every engine is subjected to a run of at least two days before being put in place. If heating occurs in engines which have previously given satisfaction, it is probably due to dirt and grit in the oil cups, or to the brasses being set up too tight. One cause of injury to the plant is frequently found in the lack of mechanical skill of the force. Dynamo engines have been found cut and scarred by the marks of cold chisel and hammer used to start nuts or back out pins, while many parts are upset by hammering. No criticism of such treatment is needed.
The dynamo also requires reasonable care and intelligent treatment. The field coils are liable to have their cotton insulation rotted by moisture, or charred by overheating. The former would probably occur only when the dynamos are laid up for a long time. The overheating is, however, an effect of use. One of the most common causes of overheating the field coils is operating at too low a speed, with the mistaken idea of saving wear and tear on the engine. If the latter is properly designed it will be as durable when worked at its normal speed as at any lower. If the dynamo is below speed a stronger current is necessary in the shunt field coils, in order to maintain normal voltage, and as the heating varies as the square of the current, the variation of temperature is much greater than that of speed. If the field coils be found to overheat, the defect may be diminished by increasing the speed and lowering the voltage if necessary.
Armature bearings give little or no trouble, as there is ample surface and no side thrust on the shaft. The armature itself requires no special attention beyond mechanical protection. This should not be such as to prevent good ventilation of the coils.
The point of most rapid deterioration of the dynamo is the commutator, and nowhere is care more required. Damage to the extent of hundreds of dollars may be done in a short time by allowing the brushes to spark heavily. The commutator segments are now made of hard copper or bronze, and are one and a half inches deep. No material can be used, however, which will stand heavy sparking. The ordinary injuries to the commutator are of two kinds, due to scoring and sparking. If a hard copper brush is used and the springs are set up with considerable tension, the commutator segments are scored and cut by the friction, and everything in the vicinity becomes covered with fine copper dust. The tension should always be as light as possible and regulated after the brushes have been placed at the neutral points. Sparking may be due to a great many causes; in fact, almost any defect in construction or design of the dynamo manifests itself in this way. As all dynamos are tested, however, before being placed on shipboard, and excessive sparking is sufficient reason for rejection in these tests, anything of the kind noted after installation is probably due to some defect which has been developed and necessitates careful inspection to determine the cause. It will, in general, be found to arise from defective insulation on the armature causing either a heavy ground or a cross. If the defect is on the outside, it may be repaired by covering it with shellacked tape, but if on an inner turn, the coil must be rewound. The Gramme armature, almost exclusively used in the Navy, has the great advantages of not being liable to short-circuits, and of being easy to rewind if they do occur. In naval practice, armatures have burned out only from actual mechanical injury ruining the wire, or from general deterioration of insulation due to continuous overheating. The first can be guarded against by careful handling and treatment, the latter by avoiding excessive overheating.
There is a great difference in dynamos as to sparking, depending mainly on the strength of the field and on the distribution of the lines of force. Every coil is short-circuited as it passes under the brush, and if it is at that moment generating electromotive force, there will be heavy sparking. It is possible to so arrange the field of the dynamo as to have an electromotive force generated in the coil opposing that due to the short-circuiting, and in this case, the short-circuited coil will be practically dead and no spark occurs. In any dynamo in operation, the standard rule is to set the brushes at the position of minimum sparking. If it occurs on one set of brushes and not on the other, it shows that they are not at the right distance apart, and one set must be moved to bring both at the neutral points. Sometimes the brushes may be too narrow, and it will be found advantageous to spread those on the brush holder, so as to cause them to make good contact on two commutator segments.
Gauze brushes have recently been tried and are found very satisfactory, as, while admitting of sufficient tension for good contact, they do not score the commutator. Carbon brushes are largely used in shore practice in cases where large and sudden variations of load take place. They do not obviate the sparking, but render it less injurious, the brush wearing away instead of the copper of the commutator. They should have a cross section of at least one square inch for every fifty amperes to avoid heating.
Once in place, the fixed wiring of the ship should cause but little trouble. The insulation on the wire in the fire-room is apt to deteriorate from heat, becoming harder and more brittle, and therefore more liable to injury. Most of the deterioration in the fixed wiring occurs, however, in the receptacles, switches and junction boxes. Unless care is taken to always have these screwed up tight, moisture or water is liable to find its way inside, where it is held and slowly corrodes the copper of the conductors. Steam-tight globes also are frequently not set up firmly on their rubber gaskets. The standard fittings are water-tight only when care is taken to keep them so; if neglected, water will work in and remain. Portable double conductor wire is sometimes treated by the crew with about the same consideration that they give a hawser and is, of course, rapidly worn out. The latest type is stronger and more durable than that formerly issued, but should have intelligent handling.
Rubber covered wire should not be painted. Many pieces have been taken out of ships in which the rubber had been rotted by the oil of the paint. If different colors are desired, they should be made with colored braid.
The automatic search-light lamps should be occasionally overhauled, a little jeweler's oil on pivots materially assisting their working. The tension of the springs should be occasionally looked to, that the lamps may feed at the proper potential, which varies from 45 volts when using 50 amperes to 52 with 100 amperes. The lamp should always be left with a good crater and with the carbons properly placed. Sufficient experience has not yet been gained with the electrical control apparatus to predict its special weakness in service, although it is safe to say that its complexity is such that reliability can be obtained only by constant care. The search-light mirrors should be kept clean, and when working with heavy currents in very cold weather it is probably advisable to keep them protected from sudden changes of temperature. The mirrors are sometimes broken by the blast of the guns, but this will seldom occur if the projectors are placed at a proper distance from the latter. The plane glass doors are more liable to injury and are made in standard strips, any one or more of which can be replaced at little expense.
The indicating electrical apparatus thus far installed in our navy has not given satisfaction in service, although working well enough in tests ashore. Other systems will be installed for trial, and with care it is hoped will give good results.
The present type of fixtures and many of the fittings are finished in dark bronze. The durability of this finish is doubtful. As it is lacquered it requires no cleaning except an occasional wiping off with an oily rag. Any scrubbing or polishing is liable to rub off the film, exposing the metal underneath.
The care necessary for the proper preservation of a complex electrical plant can be given only by thoroughly trained men. Whatever electricity may do in the future, it will always depend on human intelligence for the full development of its power, and all complex electrical apparatus requires attention to keep it in working order, as well as technical skill for its operation.