terruptionFew pieces of apparatus have presented more problems, or

been more difficult to build than circuit breakers. This comesabout to a large extent because of the necessity of stopping theelectron stampede on short notice and very quickly. Of coursethere was a time when speed of interruption of an arc was notheld to be important. A circuit breaker was asked only to putout the arc; the time didn't matter mtich.... If it took:UW:0,Wpart of a second or even more than*geetihit:kcptie:aied except:'::::::.the maintenance men wiles:hid:O.-repair the burkiid contacts and

clean up thi'carbatiiized oil, if oil used.Wheu:syeti,e.ms'grew large and::hifefConnection be-

tween them fitgan, a fault nO quickly isolated re-ulted in instability, meaning. that the tie between

-..k4, rent parts of a7system or 'the interconnection be -stems

eers4,4

cycles. 1/2§- gpand by 1930 ltreaker-tfetatfe. les be-came standard. In 1932 the rat&l time loweredto 8 cyclei' feivprelic.tically all outdoor, high -voltagebreakers. Since 1940 .five-cydelmokers have been.standard at voltages of 115 kv and above. Now wehave the three -cycle breaker for high -voltage sys- ":.tems whose stability conditions require it. On recenttests of Westinghouse 220-kv breakers at GrandCoulee Dam a total of 35 interruptions were made,

was severed, impairing interrupting9g aware of the impor41 Zs:

j.926.the 22 k

all of them in less than three cycles. The heaviest current inter-ruption was achieved in 1.85 cycles, although this was unusual.

Engineers quaintly refer to any visible evidence of breaker op-eration as "demonstration." Of this there used to be plenty. Itdoesn't require overlong memories to recall when a breaker spew-ing out a barrel or so of oil resulted in no more concern than afew cuss words and calling out the mop brigade. Smoke, bulgingtanks, trembling of the foundations-anything short of a fire-was simply evidence that a breaker had successfully performeda tough job. But not today. Anything more than an audiblethud is a black mark against the breaker. On one of the recenttests at Grand Coulee Dam the observers couldn't be surewhether the breaker had performed a high -current interruptionuntil the oscillogram had been developed. It showed an inter-ruption of 7% million kva, a record.

Engineers used circuit -interrupting devices for 40 years beforethey had a clear idea of how they worked. Of theories therewere aplenty, but even up to the late '20s the conceptions of arcextinction were, in short, cockeyed. About that time Dr. JosephSlepian began some fundamental studies of arcs and their habits.Those studies resulted in the classic papers on short and long arcspresented to the AIEE in 1928 and 1930, and which laid thegroundwork upon which such designers as J. B. MacNeill, R. C.Dickinson, B. P. Baker, H. M. Wilcox, L. W. Dyer, and othersfashioned the now large family of De -ion interrupting devices.

Slepian introduced the concept of extinction of an a -c arc asa race at each current zero between recovery of dielectricstrength of the arc path and the rising voltage, with dielectricrecovery having an invaluable head start of about 250 volts(peak) at the instant of current zero.

Another of the new concepts was that in the interruption ofa short a -c arc to a" cold" cathode, most of the dielectric strength

occurs in a thin positive space -charge layer of less than I/33 inchnext to the cathode, this layer being developed as electrons moveout of the space with great speed-in a few microseconds at most.This led Slepian and his associates to suggest that one good wayto interrupt an arc would be to chop it into many short ones andby magnetic action keep the arcs hustling over cold contact sur-faces, i.e., cold cathodes, each one of which provides its quotaof approximately 250 crest volts dielectric strength essentiallyinstantaneously after each current zero. The familiar metallic -plate De -ion air circuit breaker thereupon was born.

The role of oil in Orcuit interruption had not been understood.46:44-0.1 about 1930. it::was generally held that the gasificationa the oilby the hot arc as a nufiii:Ace to be lived with. Slepianproved the arc to be indispensable circuit interruption. Inhis words, " The decomposition of the'ihil mixed:turbulently withthe arc space ... is the principal cause "4: the arc -Interrupting ca-

; achy of the oil circuit breaker. The oiliicircuit breaker is a gas-a.$ switch, the gas blast arising from i.the decomposing oil.

eli.At.:1304.p?int of view leads to condiisithis diametrically opposite

it'

rel. , ':, V't .:j.: v..".prey

the ioni.- rt:.;:., ,

a ir. l'%...Fronk,risc:confinquantiti aThertask . 1 -

the arcing.

ally held as to desirable.'ind undeg4ble features ineakers. From this pointipf view, the: decomposition'het of being entirely unaesiirable is the very4eature

e.oil breaker function. Toimprovelhe oil bitaker,Lion of gas should be increased, not decr:!iiased,

that the gas formed is thoroughly mixed withich is carrying the arc."signers built the De -ion grid in whic*:the arc

ced to meet, tbkoughout its full length;:':iimple7 ,or A ,arc-space deionization.asification but 0 keep

d oil deterioration.

The arc, itsel ng been heschemes had beet tried-such asvacuum-to interrupt Cirrnits withourft;unwise even if it were possible. As he s..the matter: "In the past the imi*.essioiiibeen general that the arc is an oh§taclea nuisance which annoyingly intrudesself and prevents easy control of high-pocircuits. It is believed that if the arc did bo,appear of itself, or could be prevented froappearing by some simple means, then the'N.,4,,problem of circuit interruption would haver0an easy and ideal solution. However, closerscrutiny of the fundamentals involved showsthat this impression is altogether false, andthat far from being all nuisance, the arc playsa very necessary and useful role in circuit inter-ruption and that if the arc did not occur spon-taneously it would have to be invented."

Interrupting capacities have come a long way since the daysof the first Niagara Falls plants, where the first appreciable con-centration of current appeared. The interrupter used there wasan air switch, rated at 5000 horsepower, as was the early custom.Since then interrupting capacities have risen as steadily as thegenerating capacity behind potential faults has mounted. Itrose in reasonable increments to 3 million kva, where it hasstood for several years. The tests at Grand Coulee Dam raisedthis to 7% million.

And the end is not yet. Engineers planning water -power de-velopments of the Pacific Northwest envision tightly connectedgenerating capacities that, if they materialize, will call for tenmillion kva or even higher. Circuit -breaker designers are quitecalm about all this. They anticipate they can meet the occasionwithout requiring any new interrupting principle or a funda-mental change in mechanisms.

1

ipisance. Manynces and

proved this

ck"

t-4

a

3

3

WESTINGHOUSE

TOLEDO Pg5Lir, LIBRARY

Technology Department

VOLUME SEVEN JULY, 1947 NUMBER FOUR

On the Side In This IssueThe Cover-The decision having beenmade to feature the Rototrol on the cov-. er of this issue (as well as on the inside-

CIRCUIT BREAKERS-OIL OR OILLESP 98see p. 121), the problem became not whatto include but what to omit. The Roto- M. H. Hobbstrol has so many regulating functions thatthe choice is difficult. The artist, DickMarsh, has portrayed but three: rolling of D -C MOTORS ON AC 103steel, making of paper, and arc -furnaceproduction of alloy steel.

LANDING AIRPLANES IN ANY WEATHER 105The most powerful fighter plane ever W. A. Pennow

built is the McDonnell "Banshee" (XF2D)recently announced by the Navy. With-out disclosing details of engine or plane STORIES OF RESEARCH 108performance, the Navy simply says thatthe engines provide more thrust than More Light on Fireavailable in the previous most powerful Blows Hot and Coldjet-propelled fighter plane. The output is Super -Fine Fuel Atomizer

more than equivalent to a reciprocating -engine output of 5500 hp. The Bansheeemploys two Westinghouse 24C jet en- FUTURE FORMS OF AVIATION GAS TURBINES 110gines, the 24 representing the diameter A. H. Reddingin inches. Like its predecessors, the 9A,19B, and 19XB (which powered theNavy's first all -jet plane, the Phantom),

EXCITATION SYSTEMS FOR TURBINE GENERATORS 115the 24C consists essentially of an in -linearrangement of an axial -flow compressor, C. Lynna combustor, a gas turbine, and a tail-pipe. Its small diameter for its high thrustpermits the engines to be buried in the ROTOTROL AND ITS APPLICATIONS 121wings, cutting drag to a small figure.

E. Frisch

A sheet of steel 42 inches wide emergesat a speed of a mile a minute from the new WHAT'S NEW! 128tandem continuous cold strip mill of theWeirton Steel Company. The six motors Rural Line Recloser Breakerdriving the five stands and the reel aggre- A More Rugged Mill Motorgate the most power ever applied to such Concentrated -Filament Projection Lamp

a mill -17 550 hp. The last stand is drivenby a 4500 -hp twin motor.

EditorCHARLES A. SCARLOTT

Editorial AdvisorsR. C. BERGVALLT. FORT THE WESTINGHOUSE ENGINEER IS PRINTED IN THE UNITED STATES By THE LAKESIDE PRESS, CHICAGO, ILLINOIS

The Westinghouse ENGINEER is issued six times a year by the Westinghouse Electric Cor-poration. Dates of publication are January, March, May, July, September, and November. Theannual subscription price in the United States and its possessions is $2.00; in Canada, $2.50;and in other countries, $2.25. Price of a single copy is 35c. Address all communications to theWestinghouse ENGINEER, 306 Fourth Ave., P.O. Box 1017, Pittsburgh (30), Pennsylvania.

Circuit BreakersOil or Oilless?

The degree to which the oilless circuit breaker,still a relative newcomer, has muscled in on thefield of the oil breakers, particularly for low andmedium voltages, might suggest the doom of theoil breaker. A close examination of the high -voltage, heavy interrupting duty application,however, does not support that extrapolation.

M. H. HOBBSManager, Suitchgear EngineeringWestinghouse Electric Corporation

Nowis a good time to review the record of the oil and the

oilless circuit breaker to see if a definite pattern is evi-dent as to the field of application of the different types. The oilbreaker has the longest history, but the oilless variety hascome along rapidly since the introduction of the De -ion aircircuit breaker for high -voltage (13.8-kv) service in 1929. Inthe ensuing years a large amount of research and developmentwork has been done on various forms of oilless circuit inter-rupters, both in the United States and abroad. For voltagesup to 600, air breakers have been available for many years.More recently commercial lines of breakers for 2.3- to 34.5-kvindoor service have been produced. A few high -voltagebreakers have been installed outdoors. These have beenlargely of an experimental nature to provide service experi-ence for comparison with that of oil breakers.

Circuit breakers can be considered in four major classes,each of which covers a definite field and presents its own prob-lems of design and application:

a-Low voltage, 600 volts and below, as used for stationauxiliary circuits and industrial plants of all types.

b -Medium voltage and capacity, 2.3 to 15 kv, for distribu-tion substations, small generating stations, and industrialplant primary circuits.

c -Heavy capacity for large generating stations and sub-stations, 15 to 34.5 kv.

d-High voltage for outdoor service, 15 to 287 kv.

Low -Voltage Circuit BreakersFor service voltages below 600, air circuit breakers have a

long, satisfactory service record, and have become the stand-ard for this class of duty. This does not mean that oil break-ers have not been used in considerable quantity, as they con-tinue to be preferred in some refineries and chemical plantswhere atmospheres are corrosive or explosive. However, thelarge majority of low -voltage circuits are served by air break-ers, even when outdoor installation in weatherproof housingsis required.

The favorable position of the air type has been maintainedand improved in recent years by several advances in design.For example, ten years ago the open arc was replaced by mul-

tiple -plate arc chutes, which provided limitation and controlof the arc gases during circuit interruption. This permittedmounting of breakers in metal enclosures of reasonable size,later in metal -enclosed switchgear with the draw -out methodof disconnecting and removal, which has become standard forcentral -station and industrial structures.

At the same time, the carbon -tipped arcing contacts usedwith the completely open -type breaker were superseded bymetal contacts, using arc -resisting alloys such as copper tung-sten. The main contacts were changed to solid silver -to -silverline type that supplanted the multiple leaf elliptical brush,which was difficult to manufacture and maintain.

More recently there has been a change to pole units mountedon individual bases of Moldarta, which incorporate the ter-minal studs and permit supporting the breaker on steel insteadof a slab of slate, ebony asbestos, or other insulation. The de-sign at present is being developed for the smaller ratings, andwill also be extended to the higher capacities.

Particular attention is now being given to the selective trip-ping of low -voltage air breakers. This becomes of great im-portance on generating -station auxiliary circuits and in largeindustrial plants. The problem differs somewhat from thaton high -voltage breakers where protective relays and simpleshunt trip coils are universally used for selectivity. For low -voltage breakers the overcurrent and time -delay trippingcharacteristics are self-contained on the breaker, except in thefew cases where relays and current transformers are used. Inaddition, the low -voltage currents are high. If full selectivityis to be obtained breakers located in the circuits nearest thepower source must have adequate short -time current -carryingcapacity to permit their remaining dosed while breakers far-ther from the source are tripped, thus clearing faults with theleast possible service interruption. While the need for thishas been recognized to a certain extent in the past, it has notreceived the attention merited.

The newer breakers will adequately provide both for selec-tive tripping devices and for the short -time current -carryingcapacity. The air breakers thus will be even better equippedthan before to serve the varied applications of low -voltagecircuits, and with high current -interrupting capacities.

98 WESTINGHOUSE ENGINEER

a

..

Power Circuit Breakers of Medium Voltage and Capacity

The breakers of medium voltage and capacity comprisethose of 5, 7.5, and 15 kv, with interrupting ratings from25 000 to 500 000 kva. For public -utility systems, the appli-cations include 2300 -volt circuits for station auxiliaries, pri-mary distribution substations of both the indoor and outdoortype, and main switchgear for smaller generating stations, of2300 to 13 800 volts. Municipal plants, as well as industrialplants and steel mills, use many breakers in this class, thesteel -mill circuits usually requiring interruption of 500 000kva at 6600 or 13 200 volts.

In this range of capacities, oil and magnetic -blowout oillessbreakers have almost parallel ratings, although all are notavailable in both types. For example, in the smallest -25 000-kva step-there is no oilless breaker, because a 50 000-kva rating is about the lowest that can be manufactured eco-nomically. Putting it another way, an air breaker suitablefor circuits of 5 kv automatically has at least 50 000-kvainterrupting capacity.

In the medium -capacity, 15-kv class the insulation is theso-called 15L, or "low" 15-kv level (36-kv, 60 -cycle, 95-kv,impulse), and normally should be limited in application tosystems having effectively grounded neutrals if the line -to -linevoltage exceeds 7.5 kv. This is recognized in the circuit -

breaker tables of American Standards Association C37,Power Circuit Breaker Standards, applying to both oil andoilless types of interrupters.

To obtain even this insulation level is difficult in small di-

mensions comparable with those of the corresponding oilbreakers of the single -tank type. For air breakers of theseratings, the standard insulation level across the open contactshas been one step lower than the full insulation phase -to -phase and to ground, although present designs of breakersmeet this with some margin to spare. Even so, this has placedsome limitation on application to ungrounded circuits above7.5 kv, where one pole may have to interrupt voltage ap-proaching the full line -to -line value. The number of un-grounded 13.8-kv systems is decreasing, it being recognizedthat grounding of the neutral provides more positive relayoperation and permits more effective application of lightningarresters. Development work is being carried on to enhancethe insulation level of air breakers as their increasingly broadapplication on power systems requires.

The general construction of the oilless breaker for 2.3- to13.8-kv operation somewhat resembles that of the low -volt-age device, with a hinged moving -contact arm, primary,secondary, and arcing contacts, and multiple -plate arc chutesto confine and control the arc. This is brought about bythe magnetic -blowout coil through which the current passes.This action drives the arc into the slotted plates of the arcchute. Copper -tungsten arcing contacts are used and silver-to -silver main contacts. Each breaker is mounted on itsown wheeled truck that withdraws horizontally from itsmetalclad housing, providing convenient disconnection andaccessibility for inspection and maintenance.

For general application, in distribution substations andindustrial plants, the oilless magnetic -blowout breaker defi-

The compressed -air breaker (type CA), less than ten years old, is rapidlysweeping the field in the medium -voltage classes. It is used on a 15-kvcircuit and has interrupting capacity of 1 500 000 kva. It is mountedin a station cubicle with bus above, and disconnecting switches at the rear.

A typical low -voltage solenoid -operated,three -pole circuit breaker mounted ondraw -out truck for metal -enclosed switch -gear. It is rated at 600 volts, 50 000 am-peres interrupting capacity. This type hasbecome a str ndard for most low -voltage uses.

JULY, 1947 99

nitely has become the favored type as compared to oil, andalready comprises the majority of breakers manufactured inthis class. For certain locations in corrosive or explosiveatmospheres, the oil breakers may continue to be preferred,and there may be additions to existing installations for whichthe air breaker, (which is manufactured only in draw -out de-sign for metalclad structures) may not be physically adapt-able. Otherwise the air type seems definitely to have "takenover" the 5- to 15-kv, small- and medium -capacity field.

The situation in the United States is somewhat differentfrom that in England and the Continent. There has beenrelatively little development of magnetic -blowout air breakersfor the higher voltages in Europe. In those countries, the bulkof the breakers for medium -voltage service are oil, with someoil -poor, water, and compressed -air types. However, recentreports from England indicate some development work on themagnetic -blowout air type and a greater interest in it.

Circuit Breakers for Heavy -Capacity Stationsat Generator Voltage

For the generating station and substation that requirelarger breakers, oilless breakers again are favored. The rat-ings here are 15 and 34.5 kv with interrupting capacities of500 000 to 2 500 000 kva. For this duty the oilless design isthe compressed -air or air -blast type with pneumatic mechan-ism. The oil breaker is solenoid operated with a tank perpole. The insulation for this class of breaker for 15 kv is the15H, or "high level," with a 60 -cycle, "one minute withstand"rating of 50 kv, and an impulse level of 110 kv. This providesa more conservative rating for these heavy -capacity installa-tions than the "low level" mentioned previously.

Oil circuit breakers supplied for such applications have beendeveloped over the years into thoroughly reliable devices, andtheir operation record is excellent. The modern breaker with

111,11111.1111.fr - .Oil breakers of this type are in great demand. This one(type GM) is for 115 kv and interrupts 1 500 000 kva.

De -ion grids is entirely a different performer from the oldplain -break type that developed high pressure in the tank,and, when interrupting heavy currents, was prone to throwoil. Many old breakers have been modernized by the additionof De -ion grids, new mechanisms, and condenser bushings.

Prior to the oilless breaker, much study was given to thestructures for mounting heavy -capacity oil breakers, withspecial cell construction even to the extent of locating phaseson separate floors or isolated from the other phases by severalfeet horizontally. The compressed -air breaker is responsiblefor the recent return to group -phase construction, with metalisolation of the phases.

During development of heavy, power -station breakers toreduce or eliminate the oil hazard, several types of inter-rupters were studied, and experimental models of some weretested. Among these were the water breaker, and the "oil -poor" type. However, because the compressed -air or air -blast appeared to offer the most promise, intensive develop-ment work continued along that line.

The compressed -air breaker appeared in 1940. Even in thisshort time it has achieved such popularity that oil breakersare now given serious consideration only for additions to ex-isting installations with structural limitations, and even herethe air blast is often used. The reason for this is simply that,although oil breakers have generally a good record, even asmall oil fire generates a considerable quantity of smoke andsoot deposit, and, for indoor service, elimination of such pos-sibilities with attendant shutdown for cleaning is of first im-portance with most operators.

The compressed -air breaker for 15- and 34.5-kv service isthe cross -blast type. That is, the air is released across theopen -contact gap at right angles to the movement of the con-tact. This is done by means of a blast valve and tube thatdirects the air and is designed to have the shortest directpath between storage tank and contacts. The opening andclosing mechanism is pneumatic, and movement of the breakercontacts is mechanically interlocked with the blast valve toinsure proper release of air. An air relay prevents operationof the mechanism unless adequate pressure is available.

Several hundred compressed -air breakers have been in-stalled in a large variety of stations, not only on public -utilitysystems, but also in large industrial plants such as steel mills.In most, the structures for mounting have been factory -assembled steel cubicles, complete with disconnecting switches,instrument transformers, busses and connections. Operatingexperience has been essentially all that was expected, not onlyfrom the standpoint of circuit interruption and reduction offire hazard by the elimination of oil, but also in low mainte-nance. The compressed -air breaker for indoor, heavy power -station service has largely supplanted the oil type, and thereis every evidence that it will continue to maintain its favoredposition for all but hazardous atmospheres.

Foreign practice resembles ours except for two differences.First, greater use of oil circuit breakers is made, at least inEngland. Second, indoor breakers and bus structures arecarried to higher voltages (up to 69 kv or higher) than in theUnited States, where it is universal practice to install equip-ment outdoors for voltages above 34.5 kv.

High -Voltage Circuit BreakersIt is in the high -voltage service that the outdoor oil circuit

breaker seems to be holding its own against the inroads ofother types, such as compressed -air or the low -oil contenttype, also known as oil poor. This comprises the outdoor fieldfrom 7.5 kv, 50 000 kva, to 287 kv, 3 500 000 kva, which is the

1O0 WESTINGHOUSE ENGINEER

-*

highest commercial rating yet installed, although contractsfor 5 000 000-kva oil breakers for 138 kv have been placed inthe United States and for 230 kv in Canada. In addition,there is every indication that the heavy -capacity systems ofthe Northwest will require 7 500 000 kva, or even higher, at230 kv, more than twice the kva of the present maximum.

The high -voltage oil circuit breaker has been brought to itspresent state of development by intensive research and high -power laboratory tests. Improved interrupters have broughtabout major reductions in tank size and oil content, and nowprovide standard interrupting times of five and three cycles,not only on heavy fault currents but also on charging currentof long lines. Where required, the pneumatic mechanism pro-vides fast reclosing time of 20 cycles or less even on the high-est voltages, limited only by the time required to prevent thefault arc from restriking. The simpler solenoid mechanism isavailable where fast reclosing is not required. Performance ofthe breakers is consistent over the entire range of current andis, for all practical considerations, independent of voltage -recovery rates.

Interrupters for the higher voltages (115 kv and above) arethe "multi -flow De -ion grid" type in which two arcs aredrawn in series. The first arc generates gas pressure thatforces oil across the second gap, quenching the arc in theshortest possible time. Each pole contains two interrupters.With short arcing time the arc energy is low, resulting in lowmaintenance required by either contacts or oil. When this isnecessary, access has been provided by manholes in the sideof each tank. Condenser bushings permit simple, convenientmounting of bushing current transformers to the extent of asmany as 12 per breaker. Potential is readily obtained by thetapped bushing -potential device.

Tests on a circuit breaker of this type recently have beencompleted at Grand Coulee power plant, where, in 35 test in-terruptions, as many as six 108 000-kva generators and sixparallel lines were tied together to produce the maximum kvapossible at 230 kv. Single -pole, standard"duty cycle"tests, each comprising aclose -open, 15 -seconds -de-lay, close -open, were firstmade, reaching the maxi-mum of 7 500 000 kva,equivalent three phase.Later, reclosing tests weremade under similar condi-tions with both one -secondand 20 -cycle reclosingtimes. Every interruptionwas successful, with theinterrupting time consist-ently below three cycles,and as low as 1.85 cycles.A similar number of three-phase operations on charg-

Compressed-air breakers forhigh -voltage, outdoor serv-ice are being given a thor-ough trial. This one hasbeen in experimental serv-ice several years on the 138-kv lines of the West PennPower Company at Kittan-ning (Pa.). The interruptingcapacity is 1 500 000 kva.

ing current of a transmission line had been made previouslyon the same system with the same uniform interruptingtime of three cycles or less.

Much development work also has been done in this countryand abroad on other types of interrupters, particularly com-pressed air and oil poor, and some have been installed inEngland and in Europe. In some cases, as in Germany, de-velopment of these types was insisted upon by the govern-ment because of shortages of metal and oil. Recently a com-pressed -air breaker of German A. E. G. manufacture, broughtto this country by our Federal Government Technical Mis-sion, was tested on the Bonneville system at Vancouver,Washington. The breaker, rated 2 500 000 kva at 230 kv, in-terrupted fault currents equivalent to some 2 300 000 kva,although requiring 5% to 6 cycles interrupting time on acircuit of low voltage -recovery rate (some 200 volts permicrosecond), as compared to 3 cycles for an American oilbreaker. On charging current, with a relatively short trans-mission line, interruption was quite unsatisfactory, the time,due to multiple restrikes, being approximately double that onfault currents. Because of this, only three of the scheduled12 charging -current tests were made at that time.

The A. E. G. design consists of two rotating porcelain col-umns per pole, each carrying an arm which in turn supportshalf of the interrupting element enclosed in a porcelain insula-tor. In the breaker -closed position, cylindrical contacts pro-jecting from these insulators butt against each other. Inter-ruption takes place by withdrawal of the contacts by air pis-tons, and release of the air blast through the resulting orificesin the interrupters. Following interruption, the contacts areswung apart by the two supporting columns which are rotatedby the mechanism in the base, to provide the necessary airspace between contacts. A considerable amount of porcelainis required, plus that for separate instrument transformers.The space occupied is much greater than for an equivalent oilbreaker of American design.

JULY, 1947

The principle of operation of other compressed -air high -voltage breakers is somewhat similar, except for variations inthe manner in which the blast is applied and the number ofinterrupters in series. However, contacts after interruptionare usually isolated by a separate switch, either a blade typeor a rod type withdrawn by a piston, so that the main porce-lain columns are stationary.

In the United States several manufacturers have built com-pressed -air breakers experimentally and have installed themon operating systems for field experience. Such a programwas agreed upon with the operators' engineers through theAEIC-EEI-NEMA Joint Committee on Circuit Breakersseveral years ago, as a basis for possible commercial develop-ment. The voltages range from 34.5 to 138 kv. The field ex-perience has been quite completely covered in papers beforethe AIEE and has been somewhat varied. However, severalpoints seem evident. A large amount of porcelain is required,possibly reinforced by other insulation, resulting in a struc-ture not conducive to strength and ruggedness. Weather-proofing of joints and their maintenance present an import-ant problem. The operating mechanism is relatively compli-cated, and made more so by the necessity for isolator switches.This has an adverse effect on satisfactory reclosing duty.Separate current and potential transformers are required, in-creasing both space and cost. All installations have requiredconsiderable field service and frequent inspection.

In Canada, the Quebec Hydro Electric Power Commissionhas several compressed -air breakers and appears to have hadsatisfactory service from them. On the other hand, the On-tario Hydro Electric Commission has installed oil breakersexclusively, and has indicated its intention to continue to do so.

In England opinion is divided between oil and compressedair, although it is recognized that the cost comparison favorsoil, particularly below 138 kv. Until recently the trend wastoward the air blast breaker for high voltage. However, thepicture is clouded to the extent that at least two manufac-turers are now offering new oil breakers similar to the im-proved American design, and the Scottish Electricity Boardhas recently purchased breakers of this type.

Designs of oil -poor or low -oil -content breakers have beenmade in the United States for high -voltage outdoor service,as well as some commercial installations. However, thesebreakers, since they involve large porcelain insulator enclo-sures arranged either horizontally or vertically, have similardisadvantages to those of the compressed -air designs, with theadditional problem of oil leakage through gasketed joints.

Operable breakers can be developed along the lines ofeither the compressed -air or the low -oil -content form of con-struction. That this may be done competitively in cost withthe dead -tank oil circuit breaker is questionable, particularlywhen provision is made for instrument transformers. Forhigh voltage, an oil circuit breaker, which has high permanentinsulation, fast, consistent interruption, short -time reclosing,and provision for instrument transformers, is a relativelysimple device. It occupies generally less space than a porce-lain -clad breaker and has the ruggedness and reliability, with-out excessive maintenance, held of great importance by Amer-ican power systems.

Development work, of course, will continue on compressed -air, oil -poor, and on any other principle of arc interruptionthat offers promise of a better commercial design of circuitbreaker than those on the market. Meanwhile, the highlydeveloped, dead -tank oil breaker is widely accepted and pres-ent evidence is that some time will elapse before it is sup-planted.

ENGINEERING TRENDSTo relieve the strain on tired equipment after the strenuous

war loads and to permit some extension of service, 250 new sub-way cars are being built for New York City's Transit System.For these cars Westinghouse is providing new types of motorsand control. Each car will have four 100 -hp motors (two 200 -hpmotors are common) with ball and roller bearings. They will bespring suspended and the gears will be connected through flexiblecouplings. An important new feature is dynamic braking, whichwill provide stops that are more rapid yet more comfortable andwith vastly reduced brake -shoe wear. Cars will be brought from30 mph to a stop in 10 seconds instead of 15, as with presentairbrakes. The controllers provide 18 steps of acceleration anddeceleration in contrast with the nine for acceleration only onexisting cars.

Tomorrow's subway train is in the making. It is to be a sleek,10 -car train, which because of its light -metal construction, light-weight electrical equipment, and other pounds -saving features,

will weigh 75 tons less than a similar trainof present-day cars. Which means 75 tonsless to accelerate and decelerate every fewminutes. Among the many new technicalfeatures of this train will be a new Westing-house unit -switch type control that providesdynamic braking to give high-speed deceler-

ation, three mph per second as against the present two. Acceler-ation rates have been increased by 45 percent; from 1.75 to 254mph per second. An entirely new -type ventilating system in-cludes dust removal, via Precipitrons. Fluorescent lamps willprovide high illumination levels. A motor -generator set, withflywheel, will provide 60 -cycle power on each car, so that thelamps do not go out whenever a car passes over a third -rail gap.The train is being built for New York City's Transit System.

An application of dry -type air-cooled network transformers inunder -the -sidewalk vaults will be made by the Rochester Gas &Electric Corporation. This installation will consist of 500-kva,three-phase, 11 500 -volt submersible network units containing noliquids and cooled only by the natural circulation of air throughthe vault.

Motors to be nitrogen -cooled are being built for a large oil re-finery. These are big motors -2250 hp at 1800 rpm-the largestexplosion -proof squirrel -cage motors to date. They are so largethat reliance cannot be placed on the abilityof the motor walls to resist bursting shouldan explosion occur inside, as is done withsmaller motors where the ratio of outer sur-face to motor volume is much greater.Hence explosive gases must positively beexcluded. The motors will be connected to asupply of nitrogen or other inert gas and an internal pressuremaintained slightly above atmospheric. A separate motor -drivenoil pump will maintain the seal around the shafts when a motor isshut down.

Distribution systems are going underground. The winds fre-quently experienced in southern Florida and the general trendtoward elimination of pole structures have prompted the utilitiesto a large-scale underground distribution system in business andresidential areas. Transformers of the regular subway type, pro-vided with CSP distribution -transformer circuit breakers, will beused. The self-protection against short circuits and overloadsand the loading of the transformers on the basis of copper tem-perature, furnished by this scheme, make these units particularlysuitable for this service.

102 WESTINGHOUSE ENGINEER

D -C Motors on A -C

The Buffalo plant where Westinghousebuilds motors has many d -c motors inservice, but it has no direct -current dis-tribution system. Where needed, directcurrent is manufactured on the spot fromthe alternating -current supply lines bya metallic rectifier of the selenium type.

THE selenium rectifier offers a simple, static means of pro-viding on -the -spot direct current for adjustable -speed,

field -control motors from the plant a -c system. Over therange of 1 to 15 hp, for which the rectifier unit has been de-veloped and for which it appears most suitable, the rectifierand d -c motor offers several advantages over other methods ofobtaining adjustable speed within a four -to -one range. Ascompared to the motor -generator set, for example, the effi-ciency is higher, regulation essentially the same, commutationnot adversely affected, mounting is less of a problem, andmaintenance is simplified.

The complete assembly consisting of autotransformer, rec-tifier, controls, and (in the larger sizes) motor -driven fan, iscontained in a small metal cabinet that may be placed in anyconvenient location such as on the floor beside the d -c poweredmachine, or on a light, overhead platform. It is in the use ofselenium -cell rectifiers that the saving in size and weight isaccomplished, as compared to copper -oxide rectifiers or tomotor -generator sets. Control can be reduced to the bare es-sentials-a standard a -c line starter and a field rheostat. Thecircuit connections of the rectifier and motor, Figs. 1 and 2,indicate the simplicity of the system.

Six sizes have been developed having 1-, 3-, 5-, 7%2-,10- and15 -hp ratings. These are available for 230 -volt d -c output,the source of power being any of the conventional a -c, three-phase system voltages normally found in industrial plants.The small -size power packs (1, 3, and 5 hp) are of the self -cooled type. These consist of an autotransformer and a recti-fier. The autotransformer makes it possible to use any a -csystem voltage; by suitable choice of transformer, the d -c out-put will be 230 volts at full load.

I

a

Applications and PerformanceWhen using the metallic rectifier, the conventional d -c con-

trol can be supplied, although a simplification takes advantageof the natural internal resistance of the rectifier. It is possible,on the 1- to 7 sizes, to dispense with the d -c controland retain only the field rheostat. A safety switch or mag-netic breaker can be installed in the a -c line for starting andstopping, as it is possible to start the motor across the linethrough the rectifier, even when the shunt -field rheostat is setto give a speed of four times the base or full -field speed.Where dynamic braking or reversing is required, a similarsimplification is possible. This consists of omitting the flut-tering relay on the shunt field, the starting resistance and allits relays and contactors.

The voltage regulation of the rectifier, Fig. 3, is about the

From a paper presented to the Machine Tool Forum at Buffalo, New York, April 22 and23, 1947, by E. C. Watson, Section Manager, and F. L. Reed, Engineer, both of the D -CEngineering Dept., Motor Division, Westinghouse Electric Corporation, Buffalo, N.Y.

same as would be obtained from a motor -generator set. Regu-lation of the fan -cooled unit is slightly higher than for theself -cooled unit. This results from working the rectifier cellsat a higher load and causing a larger drop in voltage. Thevoltage from 50 -percent load to 100 -percent load varies from240 to 230 volts. The high regulation at the low end of thecurve, Fig. 3, is not of too much concern in the average d -cmotor application as this has but little effect on the motor -speed regulation.

The speed regulation, Fig. 4, of an m -g set and the rectifieris essentially the same. This results because the motor andrectifier voltage regulations are about equal.

Commutation is a major consideration in d -c motor opera-tion. Rectifier voltage pulsates at a frequency six times thesupply frequency -360 cycles for a 60 -cycle system because ofthe six -phase rectifier connection-with a magnitude of ap-

The floor -mounted metallic rectifier power pack used todrive the 15 -hp motor d a machine tool. It is cmpled by amotor -driven fan located above the stacks of selenium discs.

JULY, 1947 103

for any airport lights used for navigation or other purposes.The system requires no additional equipment on the plane.

It works equally well with any type of instrument approachsystem: Instrument Landing System (ILS), Ground Control-led Approach (GCA), or microwave. Thus the all-weather ap-proach lighting system does not conflict with any radio orradar control. In fact it supplements the existing and pro-posed electronic systems. An airplane must be brought, attimes of poor visibility, to the approach portal by radio orother instrument -control system. The proposed new lightingsystem, in effect, takes over where instrument control leavesoff, to guide the pilot through that difficult period of bringinghis plane those last few feet before making physical contactwith the ground.

The complete all-weather approach lighting system con-sists of three parts: the approach line, the runway designa-tors, and the high -intensity runway lights.

Approach LightingA single row of two entirely new types of lights, called

flash and blaze units, extends out from the runway for adistance of 3225 feet. This is the approach line and pointsthe way to the runway available for use. This row of lightsconsists of 36 each of the flash and blaze units, as they aredesignated, making 72 in all. Alternate flash and blaze unitsare spaced close together at the outermost end (where they aremost needed) and farther apart near the runway.

Mr. Pennow squats before one of the blaze units mounted ina weatherproof housing similar to the one housing a kryp-ton unit. It consists of six special neon lamps and reflectors.The capacitors and controls are behind the reflector panel.

The flash unit employs a krypton -filled quartz tube aboutthe size of a cigarette. It is equipped with tungsten electrodes,which have high heat -resisting properties, fused into enlargedend chambers. When about 2000 volts are impressed on thetube and the gas ionized, a discharge of capacitor -stored en-ergy occurs through the quartz tube. The discharge producesa white light of astounding brightness. By control of thestored energy discharged through the tube the brightnesscan be controlled from a relatively low value to as high asnine million candlepower per square inch. At the high valuethe lamp is nine times brighter than the sun, which has at theearth's surface a brightness of one million candlepower persquare inch. The energy absorption is equally "astronomical."At the peak of the discharge about 3000 kw are being deliveredto the gas within the quartz tube, only two inches long andabout a quarter inch in diameter. This is why the krypton en-velope must be made of quartz; any other material wouldmelt, even though the duration of the discharge is short, about17 millionths of a second. A polished -parabolic reflector con-centrates the output of the krypton lamp, when burning atmaximum brightness, into a beam of 3.3 billion candlepower.

The second member of the pair of new lights is the so-calledblaze unit. This is made up of six tubular lamps, each abouttwo feet long and a half inch in diameter, fitted with a specialelectrode and filled with neon. Neon lamps have not hereto-fore been considered as susceptible to intensity variation.However, by variable electronic control of the power supply,these blaze units can be operated at different brightnesslevels. These lamps resemble the familiar neon tubes andgive off, at the lower intensities the characteristic red light.At higher intensities the color shifts toward white, althoughred is still dominant.

Unlike the krypton units, which are operated only as flash-ing units, the blaze units can be either flashed intermittentlyor burned continuously. On steady burning service, they pro-duce either 100, 1000, or 10 000 candlepower, as selected,and on flashing service they operate at about 100 000 or10 000 000 candlepower, as needed.

The 72 krypton and blaze units, alternately spaced in a lineextending two thirds of a mile from the runway, provide anapproach -line lighting with the flexibility required to suit anyvisibility condition. Their controls are synchronized so that,when in flashing service, the units are tripped successively,beginning at the outer unit, each one firing before the oneahead appears to have been extinguished. This flashing is re-peated 40 times per minute, so that to an incoming pilot theapproach line appears as a streak of light of three -hundredthssecond apparent duration running for two-thirds of a miletoward the runway and recurring every 1% seconds.

At times of worst visibility, such as a ground fog in the day-time, the approach line is operated with the krypton andblaze units flashing at their maximum intensity, the kryptonlamps emitting flashes of 3.3 billion beam candlepower ofwhite light and the blaze units flashing with a reddish light often million beam candlepower. At the other extreme of thevisibility scale, i.e., on clear nights when only a relativelyfaint approach line is desired so as not to produce glare, onlythe blaze units are turned on. They are burned continuouslyat their minimum brightness level of 100 candlepower. Be-tween these two conditions the krypton and blaze units canbe used in such combinations of intensities as to suit the oc-casion. At all times the object is to provide sufficient light toassure positive visual identification but not enough to blindthe pilot when landing. The degree of reflective density pres-ent, i.e., fog, controls the intensity used.

106 WESTINGHOUSE ENGINEER

The runway is unmistak-ably outlined by rows ofunits of this type. It con-sists of two powerful in-candescent, sealed -beamlamps, for use when visi-bility is poor, and, on top,one incandescent lampemployed when low -in-tensity only is required.

Runway DesignatorThe second part of the system, the runway designator, con-

sists of an assembly of neon (red) and zeon (green) luminoustube elements forming either a green arrow or a flashing redcross. The green arrow indicates that the runway is clear forlanding while the flashing red cross shows that it is closed. Atall other ends of runways, the designator is a large flashingred cross, showing it is closed for landings. As an airplanepasses over the green arrow, the control officer changes thearrow to a flashing red cross, closing the runway until the air-plane has landed and been cleared. It is then changed backto the green arrow, instructing the next airplane in the land-ing -sequence pattern to come in.

During instrument weather the designator is a last -momentsafety signal to the approaching airplane. The green arrow isassurance to the pilot that the runway is clear and that he isat the proper position to make a landing. If an emergencyarises that makes the runway unsafe for a landing, the flash-ing red cross is a warning to pull up and resume position in thetraffic pattern until the emergency is cleared.

High -Intensity Threshold and Runway LightsThe third part of the system consists of the high -intensity

threshold and runway lights spaced at 200 -foot intervals, 10feet out from both sides of the runway borders and producingcandlepowers up to 100 000. This is one hundred times thecandlepower of the semi -flush runway lights in common use,and nearly three times that of the high -intensity runway lightused by the Air Force. The newhigh -intensity runway light usesthree lamps and can be operatedas a high -intensity or a low -in-tensity unit, as visibility condi-tions require. For high -intensityservice, two sealed -beam lampsare used, one for each operatingdirection on the runway. Theselamps, only one of which is usedat any time, are operated at fivedegrees of intensity: 100 000,30 000, 10 000, 3000, or 1000candlepower. For low -brightnessservice the high -intensity lampis turned off and a single lamp,

a

JULY, 1947 107

700 FEET

IIIIIIIIIIM111111111141.41ilii

,NNER RADIO

mounted in a two -directional optical system at the top of theunit, is used. Like the larger lamp, it has five steps of intensityand can be operated to give 500, 150, 50, 15, or 5 candlepower.These ten steps match the ten steps of brightness available inthe approach line.

During weather requiring only the top element, the emittedlight is in two directions, along the runway, but with somescattered light. In thick weather, when higher candlepowersare used, all emitted light is severely restricted to the direc-tion of approach. This prohibits background lighting and pre-vents lighted haze or fog curtains behind each light thatwould reduce perception of the next light in the line.

The color of the lens for the runway lights is chosen by theposition of the unit along the runway. Clear light is usedalong the outermost section of the runway, and yellow indi-cates to the pilot the last 1500 feet. When used as range lightsat the very end of the runway, the lenses are colored green.

Visual Contact with Ground AssuredThe all-weather approach lighting system offers pilots posi-

tive visual contact with the ground at the critical momentwhen they must leave instruments and descend to a landingin all conditions of visibility from unlimited to zero -zero. Thepilot is assured of positive identification of the runway loca-tion through the approach portal until he taxies down the run-way. At the end of the runway,while he still has flying speed, theall-weather system runway designator gives the pilot a last-minute traffic signal by telling him to set his plane down, or tocircle and come in again when the runway is cleared. Whenthe new system is used, the runway itself is outlined with thepowerful lights built to guide him as he lands.

It is expected that air crashes resulting from poor landingvisibility will be cut by as much as 90 percenttion of the all-weather approach lighting system.

TO OUTER RADIOMARKER -2 MILES

FROM AIRPORT

HIGH INTENSITYRUNWAY LIGHTS

APPROACH ANGLE INDICATOR

RUNWAY DESIGNATOR

...-2.

C.

/ / 9/,

.

14

s

0C2

////9099IIIII

'0f9tAZ

/0911111

0

I

347

1101

The plane is brought by radio orother means to an approach por-tal which is an imaginary frame400 feet high, 700 feet wide.Reaching outward 3225 feet fromthe end of the runway is theapproach line comprised of 72krypton and blaze units, closelyspaced at the portal entrance.The runway designator is at theend of the approach line, whilehigh -intensity runway lights out-line therunway path on both sides.

STORIES OF RESEARCHMore Light on Fire

MANhas known fire for thousands of

years. Only recently, however, hashe got around to an actual study of itsfundamental properties and behavior.And when it comes to an investigation offlame as it comports itself in the gas tur-bine, the science is really embryonic.However, fundamental research in thisfield is proceeding rapidly. Dr. W. C.Johnston of the Westinghouse ResearchLaboratories, for example, is attackingthe problem of flame velocities and theirrelation to the general combustion effi-ciency of the gas turbine. Such velocitiesmeasure the rate at which a flame, mov-ing through a mixture of combustiblegases, converts the fuel into heat energy.

For a still flame in a mixture of non-moving gases, this rate may be aroundfive feet a second, resulting in the conver-

sion of 50 to 100 pounds of fuel per hourper square foot of combustion chamber.This seems low in comparison to the totalconversion efficiency of present-day jetengines, which burn around 1200 poundsof fuel per hour per square foot. The dif-ference is due mainly to turbulence in thefuel -air stream that breaks up the smoothflame front into a jagged, stair -step ar-rangement, which presents a much greaterburning surface to the fuel. Researchmen consider flame velocities so funda-mental to the problem of total combus-tion efficiency that, conceivably, a dou-bling of velocity would result in a doublingof the fuel consumption per hour.

Johnston is seeking the answer to anumber of fundamental questions. Howdoes this velocity vary with different fuel -air ratios? What fuels are more likely togive high flame velocities? Here one clueis that fuels with low carbon -to -hydrogen

Dr. W. C. Johnston of :he Westing-house Research Labaratories uses thisequipment to study lane velocitiesand combustion rates of jet engines.

The "hot -cold pipe" Foes through itsmysterious act for Narita Witzig, atthe Research Laboratories. The tem-perature of the air cowing out theopposite ends of the pine is measuredby the thermocoupL-s is glass tubes.The temperature difference shows onthe voltmeter at tht pipe center.

->

ratios (such as the " saturated" acetylene,C2H2) give comparatively high velocities.How does the velocity vary with inlettemperature, with pressure, with blendsof gases, with pure types of fuels? Howdoes ionization affect flame velocity?

In digging for his facts, Johnston usesan apparatus employing the so-called" schlieren mirror" technique. The flameis produced by burning the test gas in aBunsen burner. A 1000 -watt mercury-vapor light, situated ahead of the flame,is reflected to a large eight -inch mirrorbehind the flame. As the light rays areprojected forward through the flame, theyare refracted upward or downward, de-pending upon the density or temperatureof the hot gases. The resulting image,focussed by a lens on a screen, shows theinverted cone of flame surrounded by lightand dark shadows indicative of the dif-ferent temperatures of the gases above thecone. By measuring the angle of the apexof the cone, Johnston is able to determinethe rate at which the fuel is being con-verted into heat energy.

The fundamental information thatJohnston is collecting should fill a gap inthe existing knowledge of the behavior offlame in the presence of a moving streamof air. In addition, it should provide thedesign and development engineer withthat basic fund of data he requires tomeet a wide range of problems.

Blows Hot and Cold

SCIENTISTScurrently are intrigued by

the performance of a narrow piece ofpipe that receives a blast of compressed

108 WESTINGHOUSE ENGINEER

air through a nozzle at its center andneatly divides it into two jets: one as hotas 154 degrees F, the other as cold as 10degrees F. Aside from its practical appli-cations-at the moment not too promis-ing-the tube produces a startling ther-modynamic effect that physicists are eagerto explore and understand.

Called the "Hilsch tube," after a Ger-man scientist who did experimental workon it, the device recently was brought toAmerica for further study. G. W. Penneyof the Research Laboratories has built arevised version of the tube for fundamen-tal research into the thermodynamic prin-ciples involved. His model-made sev-- eral times larger in diameter for easierstudy-consists of a 15 -inch length ofbrass pipe about an inch in diameter.Compressed air up to 50 or 60 pounds ispumped into a nozzle at one end of thetube to produce a vortex of rapidly spin-ning gases.

-* The cooler air concentrated in the cen-ter of the whirlpool is drawn out throughan opening in the middle of the tube, orat either end if desired. The warm air isdriven out through an opening near thewalls of the cylinder. Thermocouples in-serted in the air stream measure the tem-perature differences. Because he is moreinterested in understanding the nature ofthis separation of heat from cold, Penney'smodel is not constructed to produce thewide variations observed in the original.

What really happens is not clear. Onetheory deduces that the various layers ofwhirling gas have different velocities, ris-ing from virtually zero at the center of thevortex to a peak at some intermediatepoint and then falling away at the pe-riphery. This may cause the gases of highervelocity to exert a frictional or "shear"effect on those of lower velocity, thustransmitting energy to the latter. Thenet effect is, then, a considerable warmingof the outer layers of gas and a corres-ponding cooling of the inner layers. Thelaw of conservation is binding here. What-ever heat is extracted to produce the coldstream of air must be added to the warmerend of the flow.

A natural application of this ultra -simple device would seem to lie in thefield of refrigeration. But its efficiency isonly a small fraction of that attained inhousehold refrigerators and larger coolingunits. However, it may be useful as alaboratory tool.

A Super -Fine Fuel Atomizeraz shift from basic research to prac-r-rtical

application is neatly illustratedin the development of a unique "fuelatomizer" by Keith V. Smith, Westing-house Research Engineer. Smith was en-gaged in research studies for suitable fuelnozzles to use in aviation gas turbines,when he conceived the idea of using air toproduce a much finer atomization and dis-tribution of fuel in the combustion cham-ber. He designed a nozzle that provides a

-

-

The difference between the old and the new is dramatically evident in this picture.The nozzle at the left sprays fuel that will blc.w aga nst the firing chamber walls,resulting in caking and the formation of carbon. The nozzle on the right spraystie less than 10 micron particles, assuring complete firing without residue.

spray of particles so fine that superior dis-tribution and greater combustion efficien-cy obtains over a range of speeds.

The new nozzle seems ideal for the oil -burning gas turbine, now under develop-ment by Westinghouse for locomotive ap-plications. Here one of the major prob-lems is to utilize the heavy Bunker C fueloil to the best advantage. At peak opera-tion of the turbine, the conventional fuelnozzle without air pressure does ex-tremely well, about 96 percent of the fuelbeing converted into heat. However atlow fuel rates, i.e., when the turbine isidling, its performance is poor. Consider-able caking of carbon deposits due to im-perfect combustion takes place around thenozzle and on the walls of the chamber,with the result that sometimes less thanhalf the fuel is converted into usable heat.

The nozzle developed by Smith appearsto solve these problems. It provides aneven distribution of fuel throughout thecombustion chamber rather than allowing

it to blow against the walls. The fine de-gree of atomization results in completecombustion even when the turbine is id-ling. Up to 98 percent efficiency over thewhole range of fuel rates is obtained.

The key to the new nozzle's efficiencyis found in the manner in which high vor-tex air is created when the streams underpressure enter six pin -hole angled slots, toimpinge on a spiral -shaped metal whorl.The incoming fuel, itself whirled around,collides with this high circular velocity airand is ripped into particles smaller than10 microns in diameter.

The unique feature is the fact that thissuper -atomization can be obtained at fuelpressures down to 10 pounds per squareinch for the heavy oil and lower for thelighter grades. With the conventionalnozzle, pressures of at least 50 pounds persquare inch are required for satisfactorystarting. With the new nozzle maximumfuel pressures required for top -speed oper-ation of the turbine are much reduced.

JULY, 1947 109

Fig. 1-A turbo -propeller combines the merits ofjet propulsion and propeller drive and is es-pecially suitable for rates of 250 to 500 mph.

ARCRAFT propulsion is not the relativelystraightforward application of reciprocat-

ing engines and propellers it once seemed des-tined to be. Airplanes have been powered bypiston engines driving propellers through gearsever since the Wright Brothers flew at Kitty -hawk in 1903, although all three componentshave been vastly improved both in perform-ance and capacity. The application of the gasturbine to aircraft, on the other hand, is only beginning.Nevertheless, concurrent with the rapid progress in compres-sors, combustors, gas turbines, and other components, nu-merous ways of using the new propulsion machinery havebeen proposed and are undergoing engineering analysis. Itis apparent that the gas turbine will be used in different ways,depending on such factors as flight speed, desired range, typeof load, and economics.

All present known forms of aircraft propulsion-jet, rocket,or engine -driven propeller-employ one common principle:the acceleration rearward of a large weight of gas causing areaction on the airplane to result in a forward thrust. Thegas can be free air or combustion products, or both. Thesethree propulsion forms differ in the relative amounts of matterhandled and the acceleration given it. The rocket carrieswithin it the entire substance (fuel and oxygen) to be acceler-ated, the propulsion being accomplished by discharging theproducts of combustion with an extremely high velocity. Asindicated in Fig. 2, in a turbo -jet engine a considerably largermass of free air is mixed with the products of combustion, thetotal weight being discharged at a lower relative velocity,giving a reaction thrust. The propeller drive obtains its for-ward motion by giving relatively small acceleration to a farlarger mass of free air by the mechanical action of the propel-ler. In the conventional aircraft some forward thrust is ob-tained by expulsion of the engine exhaust gases to the rear,

Future Forms of AviationGas Turbines

The reciprocating engine, itself exceedingly complex, is used in air-craft in a single, simple way: it drives a propeller through a gear.The gas turbine, on the other hand, while it is the essence of sim-plicity, can be used in a variety of ways to propel aircraft. Air-craft gas -turbine versatility is indicated by the numerous combina-tions of jet and propeller drives and different turbine cycles beingstudied. They suggest a pending revolution in aircraft propulsion.

A. H. REDDING

Development Engineer, Aviation Gas Turbine Division,Westinghouse Electric Corporation

but this is only a small portion of the total propulsive effect.Two important considerations arise in evaluating any pro-

pulsion system:1-How much useful thrust power is developed for a given

size and weight of engine and its essential adjuncts?2-What is the effectiveness with which the energy of the

fuel is utilized as useful propulsion energy, as measured by theweight of fuel used per hour or the thrust per horsepower?

The net thrust on the aircraft is proportional to the productof weight of gases accelerated and the change in velocity.Either an increase in mass or a greater change in velocity pro-duces greater propulsive thrust. Power, as always, is forcemultiplied by velocity, hence useful thrust power is propor-tional to the net thrust force times flight speed.

Analysis of the performance of a given jet -propulsion sys-tem requires consideration of the different components ofboth power and efficiency. These include thrust, jet, and in-put powers, and wake and thermal efficiencies. Thrust poweris the power that actually propels the vehicle. Jet power isthe kinetic energy added to the mass of gases accelerated forpropulsion and is, therefore, proportional to the weight of gashandled multiplied by the difference of the squares of the rela-tive velocities before and after acceleration. The input poweris the product of the fuel flow times the heating value of thefuel. Wake efficiency is that percentage of the jet power thatis utilized to give useful thrust power.

110 WESTINGHOUSE ENGINEER

Three numerical examples wilt clarify these points. Inthese the thrust and flight speed remain constant, but theweight of gas accelerated in each. case is reduced by a factorof 2. The wake efficiency is twice the flight speed divided bytwice the flight speed plus the difference between the jet andflight velocity. This difference between the jet and flightvelocity is, of course, proportional to the thrust per pound ofair accelerated and, if total thrust is kept constant, is in-versely proportional to the amount of gas accelerated. Forthe first condition

1 ++

1

1/2E..12= - 80 percent

and for the second condition1 + 1

E.2 = 2 + 1 - 66 2/3 percent

While if the weight of accelerated gas is cut again in half, wehave for the third case

E,,31 + 1

2=

2- 50 percent

-I-

The jet power in each case is proportional to the mass ofgas multiplied by the difference between the squares of thejet and flight velocities. For the first condition

= 1 (1.52 12%) = 1.25and for the second condition

Pte = 1/2 (22 - 12) = 1.5Thus for a given thrust and flight speed, doubling the amountof gases accelerated reduces the amount of power in the jet inthis case by a ratio of 6 to 5. The thrust power remains thesame

= 1 (1 1/2 - 1) = 1/2= 1/2 (2 - 1) = 1/2

Thermal efficiency is the one familiar to the power or engineman. It is the efficiency with which chemical energy con-tained in the fuel is transferred into kinetic energy of the pro-pelling jet. Its determination is complicated because itdepends upon many variables such as cycle pressure ratio,cycle temperature, component efficiency, as well as the wholemakeup of the thermal cycle modified by the use of inter -coolers, gears, propellers, blowers, reheaters, heat exchangers,or accessories.

Gas -Turbine Propulsion SystemsAnalysis of any propulsion system must necessarily be

made upon the basis of the efficiencies. Several different pro-pulsion systems based on the gas turbine can be compared.

The Simple Turbo -Jet Engine, Fig. 3, consists of four mainelements between air intake and exhaust jet: (a) a compressorin which the air used for both the thermal cycle and propulsionis compressed to several atmospheres, (b) a combustion cham-ber after the compressor where heat of the fuel is added tothis compressed gas thereby greatly increasing its volume, (c)a turbine for expanding the gas through a sufficient pressure

I-u.t

2......=171

AIR FLOW -10 000 POUNDSPER SECOND

VELOCITY CHANGE9 FEET PER SECONI

-r VIDEAL WAKE EFFICIENCY:99%

AIR FLOW -2500 POUNDSt PER SECOND

IDEAL WAKE EFFICIENCY.96%

-AIR FLOW -65 POUNDS PER SECOND

V311K101.1171.144t1,114Riant:14:WW

IDEAL WAKE EFFICIENCY:45%

Fig. 2-Propulsive thrust is the product of mass acceleratedand the velocity change. The 16 -foot and 8 -foot propellers,and the small turbo -jet here shown all provide the samethrust, but by widely different masses and velocity changes.

ratio to give power to drive the compressor, and (d) an ex-haust nozzle through which the hot gases complete their ex-pansion to atmospheric pressure and therefore leave with thehigh -velocity jet reaction necessary for propulsion.

In the simple turbo -jet engine the air ejected to the rear toprovide the reaction for propulsion is the same air used in thethermal cycle. Because the turbine exhaust and the jet arethe same, the wake efficiency, thermal efficiency, and flightspeed are interrelated. In present-day turbo -jet engines oper-ating with a compressor ratio of 4 to 1, jet velocities for air-craft moving at 500 mph are about 2200 feet per second (1500mph) which gives a wake efficiency of 50 percent. Withthermal efficiencies, at high altitude, of 28 percent, this resultsin an overall propulsion efficiency of about 14 percent. Ifoverall efficiency is to be improved for a given flight speed twoalternatives are possible: (1) reduce the jet velocity, whichimproves the wake efficiency, or (2) improve the thermal effi-ciency of the machine.

A reduction of the jet velocity can be achieved by reducingthe turbine inlet temperature. This improves the wake effi-ciency but reduces the thermal efficiency. Thermal efficiencycan be improved in some cases by selecting more favorablecompression ratios, usually higher ones. However, higherpressure ratios present serious problems in coordinating thedesign of the compressor and turbine for maximum combinedperformance, particularly at conditions other than those bestsuited to the specific design. Any method that reduces thethrust per pound of air handled by the main elements of the

011 Cooler

Inlet and FrontBearing Support

Compressor

......mow

Accessory Otte

Combustion Chamber

es,sk- - re~.-+.-

N.-

I

Exhaust Nozzle

Turbine

Fig. 3-In the simple turbo-jet, air is compressed, burned,expanded in a turbine that ex-tracts enough energy for com-pression, and is finally ejectedto the rear at high velocity.

JULY, 1947 111

engine-that is, the compressor, com-bustion chamber, and turbine-tendsto increase the weight of the enginefor a given output.

The Simple Turbo -Prop Engine ofFig. 1 is the basic turbo -jet enginebuilt with a larger turbine (i.e., morestages) to allow more complete expan-sion of the hot gases, plus a reductiongear and a propeller, to utilize theextra power obtained on the turbineshaft. In this arrangement more of the energy in the combustion gases is taken outas mechanical energy to drive the propeller, as well as the compressor, leaving lessenergy in the jet.

An advantage of the simple turbine -driven propeller is that the thermal enginecan be best designed to deliver shaft power while the propeller can be designed tohandle large masses of air at high wake efficiency. Unfortunately, the thermal effi-ciency of the propeller -drive turbine is basically considerably lower than that of theturbo -jet, because, instead of converting pressure energy into jet energy in the highlyefficient exhaust nozzle, additional turbine stages must be used with lower expansionefficiency. In addition, gear losses and aerodynamic losses of the propeller must beincluded with all engine losses.

To compensate for this lower thermal efficiency, however, propeller wake effi-ciency is extremely high, reaching 98 percent at speeds above 250 mph for presentdesigns. The modem turbo -prop engine has high static take -off thrust for a givensize of power plant. It is, in general, superior to the simple turbo -jet for speeds upto about 600 mph, which covers the range for all commercial flight in the foreseeablefuture. For still higher speeds, which at present, are only of military interest, theturbo -jet is more efficient. Also the weight of the turbo -prop engine per pound ofthrust may be two to three times larger because of the additional turbine stage,gears, and propeller. For commercial aircraft where the speeds will not exceed about550 mph the simple turbo -propeller shows great promise because it is only two thirdsas heavy as the reciprocating engine. Also because of the much smaller frontal areaof the gas -turbine engine the nacelle drag is significantly less, which should leadeventually to much improved aircraft for both long and short range.

Where speeds above 450 mph are required, low weight per pound and high specificthrust are of paramount importance. While fuel consumption is important at pres-ent, it must be sacrificed for low specific weight. Where highest speeds are sought, asimple jet engine operating at the highest possible turbine inlet temperature is indi-cated. Wake efficiency becomes less important, while extremely high jet velocitiesor high thrust per unit of mass handled become important. Mechanical limitationsof the turbine prohibit the use of excessive turbine inlet temperatures to achievethis aim. Higher temperatures in pure jet engines result in higher thermal efficien-cies, also jet velocities are increased, signifying lower wake efficiency. Thus, sinceoverall efficiency is the product of the two, an increase in turbine inlet temperaturewith present-day jet engines may actually lower the efficiency. With turbo -propengines, however, increasing the turbine temperature may not mean a boost in jetvelocity, and hence overall performance is improved.

Simple Turbo -Jet with Afterburner- A possible solution to the demand for higherpowers at flight speeds near and beyond sonic is a simple turbo -jet engine in whichfuel is additionally burned in the gases after they leave the turbine exhaust. Such anafterburner or combustor has been added in the tailpipe ahead of the exhaust nozzleof the simple turbo -jet engine, as shown in Fig. 4. This system provides an increasein thermal energy to the gases issuing from the turbine before they have been com-pletely expanded in the exhaust nozzle. Increases of 35 percent in thrust at staticsea -level conditions to 60 percent at 500 mph can be realized by using the after-burner. At still higher speeds the percentage thrust increase is proportionallygreater. At speeds above that of sound, (742 mph in air at 32°F) thrusts in flight

Gears

C

Afterburner

Fig. 4-Burning additional fuel inthe tailpipe of a turbo -jet enginegreatly augments thrust, at a loweredefficiency acceptable for short periods.

Fig. 6-A regenerator orhea: exchanger can also be

+ applied to a turbo-prop.

Turbine

112

PropellerCompressor

Regenerator

WESTINGHOUSE ENGINEER

CombustorExhaust Nozzle

Fig. 8-Still better performance ov.two cross -compound turbines, with a

-4611-1 trittliwouiwri

MOINIngiumwo!Gears

//Low -Pressor,

Compres:-

Intercooler--

elkillangiill

1 I 1111M011111_111111

Hugh -PressureCompressor 7

Fig. 9 -An engine with separate turbine fc

Gears

-4_01111Witi

- 001111Minimmil

Propeller Compressor

Fig. 10-A unit with propellers located in the ait

Accessories

-Wpiltom

AAR "V

Ducted Props Gears Co

Fig. 5-With a turbo -jet unit, recovery ofsome of the exhaust energy in a heat ex-

* changer improves thermal efficiency ofthe cycle but reduces available thrust.

Fig. 7-To obtain better perform -'Lae over a wide speed range two

rbines can be used in tandem.

"-High-Pressure Compressor

%ACl4atam ;----,r1;;xifZ:ZIZrt::*-

ReductionGears

Propeller

Low -Pressure Compressor

weide ranges in load is secured byaddition to ducting complications.

l,)2

h;)1

Low -PressureTurbine

Reheater

- propeller gives greater design flexibility.

po oo o0000000o O O-°op

00

[Combustor

O 0 0 0 0 00 0 0 0 0 00

PropellerDrive Turbine

ExhaustCompressor Nozzle

Drive Turbine

J the compressor permits elimination of the gears.

Turbine

Regenerator Combustor

Low -PressureTurbine

Exhaust Nozzle

can be increased manyfold by this means.In the subsonic speed range, the after-

burner results in considerably increasedfuel consumption for a given thrust.However, a turbo -jet engine so equippednormally would cruise at a better specificfuel consumption with the afterburner

High -Pressure TurbineCombustor turned off. The afterburner would beused for short bursts where extremelyhigh powers were required. In the sub-

sonic speed range, the afterburner, in general, results in a considerably lower wakeefficiency as well as a somewhat lower thermal -efficiency cycle; this gives a consider-ably poorer overall efficiency. The additional weight of the afterburner is a furtherdisadvantage to overall performance.

A Turbo -Jet with Heat Exchanger, which in gas -turbine parlance is called a regen-erator, as shown in Fig. 5, is one that combines improved thermal efficiency withlowered jet velocity. In this case, the hot gases leaving the turbine are piped for-ward to a heat exchanger between the compressor and combustion chamber. Heresome of the heat of the turbine exhaust gas is removed and is added to the air leavingthe compressor. This reduces the heat to be added in the combustion chamber andthe amount of fuel burned and reduces the temperature of the gas flowing throughthe exhaust nozzle. As a consequence total thrust is reduced.

The reduction in fuel heat input per pound of air results in reduced jet velocityand therefore a lowered total thrust. Reduction in jet velocity improves wake effi-ciency at the same time thermal efficiency is raised. Thus while the jet engine witha heat exchanger is a heavier engine for a given thrust, fuel economy is bettered.With a moderate amount of regeneration, fuel consumption in a given case is de-creased by 20 percent, accompanied by a reduction of thrust of the same percentage.This means for 80 percent of the thrust, the fuel rate is reduced to 64 percent.

A Composite Turbo -Jet Engine with Regeneration and Afterburning provides agood fuel economy at low thrusts with the afterburner turned off. It would becapable, with afterburner functioning, of extremely high thrust for short periods oftime with large fuel consumptions. Such an engine would be attractive for speedsof 400 mph and upwards.

A Turbo -Propeller Engine with Regenerator-The scheme suggested in Fig. 6 usesa turbo -propeller engine with regenerator. With the ideal case of a regenerator with-out pressure drop, the temperature entering the turbine is the same as in a simpleturbo -propeller. With a full pressure drop across the turbine, the output of theengine is the same as from the simple turbo -propeller engine. Actually, flow lossesin the regenerator result in a small drop in available power. The regenerator haslittle or no influence on the wake efficiency of the propeller in a turbo -propellercombination. The reduction in fuel consumption results in an attractive gain inoverall efficiency.

A Tandem -Compound, Turbo -Propeller Engine-As efficiencies of the gas -turbinecomponent parts are improved by development, the most favorable pressure ratiofor the simple turbo -propeller engine is increased. However, a major problem exists;a single -cylinder, high-pressure ratio, multi -stage, axial -flow compressor does notperform well at speeds and pressure ratios much below those for which it is designed.Under rated conditions, a compressor suitable for a high-pressure ratio operateswith the axial air velocities correct for the peripheral speeds of all stages. However,at lower speeds, the flow of air at the exhaust end of the machine is high comparedto peripheral velocity of the blading, while at the low-pressure end of the compres-sor, gas flows are small compared to the peripheral velocity. A serious loss in com-pressor efficiency then obtains. This reduction in efficiency is a function of the ratedpressure ratio of the machine. The higher the pressure ratio, for which the compres-sor was intended the lower the partial -speed efficiency.

One solution is to use a compound engine, such as shown in Fig. 7, using two corn -

JULY, 1947 113

pressor Corn bustor

Jet engines of this new type (24C) power the Navy's recentlyannonnced "Banshee."

pressors and two turbines. A low-pressure compressor isdriven by a low-pressure turbine geared to a propeller. On aconcentric hollow shaft a high-pressure compressor is inde-pendently driven by a high-pressure turbine. The two shaftsare independent and rotate at different speeds. The speedratio depends on the load or the average speed of the two.With this system, at starting, the low-pressure compressorrotates at relatively low speed while the high-pressure com-pressor turns much faster. As the speed of the high-pressureset approaches normal, the low-pressure combination in-creases its speed much more rapidly.

Table I- Fundamental Relationships for Populsiorin Air at Sub -Sonic Velocities

Net Thrust -(WA + Wx)C- WAVOR

WA(C- N.( )

(Moss x Change of Velocity)

Thrust Power [(WA + WF)C-WAV]OR

VIA(C-V)V- 550 g 550 g

(Thrust x Flight Speed)

(WA + WF)C2- WAV2 Wx(C2- V2)Jet Power - OR2g550 2g550(Moss x Velocity Sguored)

Fuel Input Power = WF x HF X550

( Fuel Flow x Fuel Heating Value)

Woke Efficiency= 2 [((WAW + ) CV- vVAV21A

WF)C2V2 CR

(Thrust Power +Jet Power)

Thermo( Efficiency [(WA *WF)CV- WAV22 WFHF9J

(Jet Power +Fuel Input Power)

+ )Overall Ef ficiency- [(WAg

WJF WFHF

CV-WAV21...R

(Thrust Power 4 Fuel Input Power)Symbols

WA =

WF =

9 =

OR

2V

WA (C2- V2)2xHFWF9J

wAv(C-v)

Free Air Lsed in Propulsion -Pounds Per Second

Fuel and Moss Corned 0 Airframe -Pounds Pe> Second

Acceleration of Grovi ty -Feet Per Second 2

C = Jet Velocity Relative to Airplane -Feet Per Second Jet VelocityRelative to Ground Flight Velocity

V = Flight Ve-ocity-Feet Per Second

HF = Heating Value of Fuel (Lower Value Usuoity Used)- British ThermalUnits Per Pound

J = Mechonical E0AiirOlen, of Heat

Such a system would allow better partial -loadoperation. It would actually permit use of aturbo -propeller engine designed for pressure ra-tios of between 16 and 20 to operate satisfacto-rily from starting speed up to full speed.

A Cross -Compound, Turbo -Prop System, Fig.8, results in a more complicated ducting prob-lem but a simpler mechanical problem. Be-tween the two compressors is an intercooler to re-move heat from the air leaving the low-pressurecompressor. This reduces the mechanical workin the high-pressure compressor, leaving a largerexcess of power for the propeller. A reheatingcombustion chamber, shown after the high-pres-sure turbine, raises the exhaust temperature ofthe high-pressure turbine to a point correspond-ing to the mechanical maximum allowable. Thisenables the low-pressure turbine to deliver morepower. The intercooler and regenerator are bothaimed at increasing output of the engine. De-pending on the effectiveness of these two ele-ments and the cycle in which they are used, again or loss in efficiency is realized. More sig-nificant is the increase in power.

A Turbo -Propeller with Separate Drive and Compressor Tur-bines, Fig. 9, is attractive because it simplifies control of theengine. Also it enables the designer to create a turbo -jetengine and then use it as a propeller drive.

A Ducted -Propeller Gas Turbine, Fig. 10, provides a double -rotation multibladed propeller driven from the tips of rotatinglow -speed turbine elements in such a manner that the reduc-tion gear is eliminated. Such an arrangement illustrates pos-sible variations from the simple turbo -propeller engine.

Ducted -Propeller Turbo-Jet-The simple turbo -jet enginesdiscussed all have low wake efficiency at present-day flightspeeds. On the other hand simple turbo -propeller engineshave extremely high wake efficiency at such speeds. An inter-mediate between these two extremes is a ducted -propellerturbo -jet. A two -to -one ratio compressor handles a largemass of air, part of which is sent into the inlet of a compressorof higher pressure ratio, through a combustion chamber, andthrough a turbine that drives both compressors. The high-pressure -ratio compressor is connected directly to the turbinewhile the low -pressure -ratio compressor must be driventhrough a gear. The resulting thrust is obtained from thereaction of the gas used in the combustion cycle as well asfrom the air handled by the low -pressure -ratio compressor.Thus a thrust at static conditions considerably better thanwith the straight turbo -propeller is obtained. At mediumflight speed, the specific weight is better than with either theturbo -jet or turbo -propeller, while the specific fuel consump-tion at the high range of intermediate flight speed surpassesthat of the turbo -propeller. The possible use of a combustionchamber in the outer flow pass can increase the thrust obtain-able by this engine as much as 80 percent at take -off and wellover 200 percent at 500 mph.

A great variety of propulsion systems can be developedaround the gas turbine. The combinations are almost limit-less. The gas turbine, built up of a number of relativelysimple components and each susceptible to individual devel-opment, allows great latitude to the design engineer in creat-ing an engine for optimum performance for a given applica-tion. In the future more successful models of gas -turbineengines can be expected than there were types of reciprocatingengines before the last war.

114 WESTINGHOUSE ENGINEER

Figs. 1 to 3 show the development stages of thepresent fully enclosed Westinghouse direct -con-nected exciter. Fig. 1 (below) is the early open -type design, Fig. 2 (right) the first encloseddesign, and Fig. 3 (lower right) an intermediatedevelopment stage, showing unit covers inuse before the "roll -away" cover was devised.

Until the first World War the shaft -driven exciter was easyto build because the generators were small and ran at 1800rpm or less. By 1917, the largest 3600 -rpm machine was only6250 kva, requiring a direct -connected exciter of 50 kw. By1927, the appearance of the first superposed turbine -generatorunit had raised the exciter requirements to 62 kw. To date,the largest 1800 -rpm generator built is 180 000 kva, using a375 -kw, 250 -volt, direct -connected exciter; the largest 3600 -rpm unit is 111 800 kva, and the largest direct -connected ex-citer is 280 kw.

Direct -connected exciters have been built by various manu-facturers in sizes up to approximately 150 kw. Beyond thisrating some manufacturers have built exciters geared to theturbine shaft and operating at about one-half turbine speed.Westinghouse, however, has built only direct -connected ex-

- Excitation Systems for Turbine GeneratorsI.

Excitation facilities for a -c generators demand the utmost in reliability over long periods. Experience over manyyears with the direct -connected exciter proves its ability to meet the need for reliability in any capacity required.That excellent record, however, does not preclude consideration of other forms of excitation now being tried suchas the electronic and the Rototrol pilot exciter. For the largest machines the trend is to higher excitation voltages.

C. LYNN, Manager D -C Engineering Department Transportation er Generator Division Westinghouse Electric Corporation

THE shaft -driven exciter has been the traditional methodof furnishing excitation for a turbo -generator and con-

tinues to be the most popular because of its reliability and theability of turbine, generator, and exciter to function as a unit.Any source of excitation other than mechanical drive direct

a from the turbine is subject to failure of the drive or other linksor units introduced into the system. The one disadvantage ofthe shaft -driven exciter is that, in case of serious trouble in

4 that unit, the entire turbine generator must be shut down.Alternate methods of excitation, until quite recently, al-

ways used rotating d -c machines, usually motor driven,although separate small turbines have been used. With thisseparate system of excitation, either individual exciter unitsfor each a -c machine, or a common exciter bus can be used.More recently the electronic exciter has made its appearance.Meanwhile improvements and modifications havebeen made to most excitation schemes so that to-day the choice of excitation systems is quite varied.

a

JULY, 1947 115

Shrink Rings

Balance Holes

Mica Micas

1

Commutator Bar

Fig. 6-An early type of high-speed, shrink -ring com-mutator. Note the positioning of the balance holes.

Fig. 7-The shrink -rings of this modern com-mutator are made of nickel -molybdenum steel.

erator. Various expedients to eliminate the pilot exciter havebeen tried from time to time, such as notching the poles of themain exciter. The weakness of such schemes has always beenthe possibility of instability when the main exciter operates atlow voltages, self-excited.

The same ease of replacing brushes on the main exciter ob-tains for pilot exciters, but in a different manner. For relia-bility on this small, four -pole machine two brushes per armare used to eliminate any possible loss of pilot -exciter voltage,such as might occur if the operator changed brushes with onlyone brush per arm.

The current per brush being low-about 2 amperes-itcan be carried directly into the brush from the tip of the brushfinger. Standard brushes for the pilot exciter are used, butwith the shunts cut off to within about one inch of the brushesto act as a means of removing a worn brush. Flexible shuntsare used on the fingers, so that spring and finger bearingsare not required to carry any current.

Separately Driven ExcitersExciters driven by motors or separate turbines have nor-

mally been of the open, conventional construction. As animprovement, where motor -driven exciters still are consid-ered, a drive consisting of an oversize induction motor and adirect -coupled flywheel is recommended. This allows the ex-citer to ride through system disturbances where the motor

Voltage RegulatorAutomatic Control

Control Unit- Switch

Y

RototrolControl

Field

PotentialTransformer A

VoltageRegulatorPotential

Unit

Voltage Separately Exci edAdjusting Stability Field

Unit

Voltmeter

410

Main ExciterShunt Fields

InCurrentTrans-former

A -CMachine

Stator

A -C Ma-chine Field

E

Pi o Field t Y. Rheostat. Motor or Manually OperatedExciter / Breaker Swi chbo;r"a---Rototrol Equipment Mechanically Connected

Self -Energizing Field

Fig. 8-A schematic drawing showing the compon-ents of "buck -boost" Rototrol controlled excitation.

receives its power from the main generator being served. Thea -c voltage can drop to as low as 70 percent of normal for arelatively long period or to zero for 0.3 second without loss ofstability. This provision for maintenance of excitation is am-ple for normal fault contingencies.

Electronic ExcitersElectronic exciters of the ignitron type have been applied

as an excitation source for large generators.* When the igni-tron receives its power from the generator it serves, specialprecautions are necessary against line disturbances becausethe d -c voltage output of an ignitron bears a definite relationto the a -c voltage. With some electronic exciters a separate,small a -c generator coupled to the shaft of the main generatorsupplies an independent source of exciter power.

The separately driven exciter and the electronic exciterboth are more costly and occupy considerably more spacethan the direct -connected exciter, although the rectifier unititself can be located anywhere convenient. Electronic unitsare subject to normal hazards of rectifier operation. Accord-ingly they are designed oversize, so that they can operate withpart of the unit out of service.

Rototrol Excitation ControlRotating regulators, usually built to replace the conven-

tional pilot exciter, constitute a recent improvement appli-cable to rotating exciters. Several schemes employing Roto-trols have been or are being applied. In one type, the mainexciter is a conventional unit and receives its full excitationfrom a Rototrol pilot exciter. In a modified scheme the mainexciter is conventional but has three shunt fields, as shown inFig. 8. (Rototrols are additionally described on pp. 121-127.)

In this system shunt field 1 provides most of the main -ex-citer field excitation. It receives energy from the main -exciterarmature as usual. Field 3 is a small, separately excited shuntfield fed from the station battery or any other source of ap-proximately constant d -c voltage. This provides stability atlow voltages, during starting, or when the unit is under handcontrol. The arms of the rheostats in fields 1 and 3 aremechanically connected for motor or hand operation, as de-sired. As the arms are turned, resistance is cut into one fieldcircuit and out of the other. Also, the rheostat for separatelyexcited field 3 is built so that at some point (approximatelyhalf voltage on the main exciter), the rheostat is open -cir-cuited and remains so as it is turned further in a direction to

*See "Construction and Tests of an Ignitron Exciter," by R. F. Lawrence and C. R. Mar-cum, Westinghouse ENGINEER, Vol. 6, May, 1946, p. 86 and "Why the Electronic Ex-citer," by C. M. Laffoon, in the same issue, p. 85.

118 WESTINGHOUSE ENGINEER

increase the voltage. This is done because field 3, the stabilityfield, is not needed at the higher exciting voltages. Thus thedrain on the station battery or other source for this field isreduced. Shunt field 2 is excited from the Rototrol pilot ex-citer. This scheme uses a "buck -boost" Rototrol pilot exciter.

In the scheme of Fig. 8, the rheostat on the main exciter isadjusted to provide a base value of excitation for the a -cgenerator. This can be set to maintain steady-state stabilityat any load desired on the a -c generator. The Rototrol pilotexciter operates so that field 2 of the main exciter either addsto or subtracts from the excitation of field 1. This requiresthe pilot exciter to operate at positive or negative polarity andof a magnitude to produce the necessary excitation. A gen-erator static voltage regulator and the "buck -boost" Rototrolpilot exciter produce excitation of correct polarity and magni-tude to maintain the a -c generator voltage constant in spiteof changing loads.

This system is superior to the type where full excitation forthe main exciter is furnished from the Rototrol pilot exciterin that, should anything happen to the Rototrol pilot or toany part of the voltage regulator, complete excitation wouldnot be lost, or load dropped. In this case, the main exciter,operating as a self-excited exciter, provides base value excita-tion for the a -c generator corresponding to an excitation of the

a -c machine when set to carry, for example, 75 percent ofrated load at the predetermined voltage. If such a failureoccurs and the load is other than 75 percent, the generatorcontinues to carry whatever kilowatt loads are placed on it,within its capacity. However, the power factor then changes,because the a -c generator is under -excited or over -excited de-pending on whether the load is more or less than that of theinitial setting. The equipment continues to function in thismanner, even should a short circuit or open circuit develop inthe Rototrol pilot exciter. Such would not be the case if fields1 and 2 of the main exciter were combined, and if the "buck-boost" Rototrol pilot was connected in series with the circuitof sell -excited field 1. If a failure does occur, the excitationsystem can be operated manually.

Standby sources of excitation, too, can be made in the formof three -field -winding main exciters and Rototrol pilot ex-titers, if desired. Where such machines operate at 1200 rpm,the Rototrol pilot can be mounted on the same shaft as themain exciter. Separate motor -driven Rototrol pilot excitersoperating at 1200 or 1800 rpm can be used if desired.

In the scheme of Fig. 8, only two fields are used on the

Rototrol pilot exciter: a self -energizing series field, and a con-trol field that combines a differential shunt field and a patternfield. Its excitation is secured from the turbine generatorthrough a voltage -regulating potential unit and a voltage -regulating, automatic -control unit. These two devices are ad-justable, but have no moving parts in the sense of an ordinaryvoltage regulator of the contact -making or rheostatic type.

The various component parts of the voltage -regulator po-tential unit and voltage -regulating automatic control unitfunction as follows:

Current transformers are located bn conductors A and Cwhile a potential transformer is connected to A and B. Cur-rent transformer A supplies a filter reactor. Current trans-former C is connected through a resistor to half of the sameprimary winding of the filter reactor. The output voltage ofthe two current transformers produced in the secondary wind-ing of the filter reactor is combined with the voltage output ofthe secondary of the potential transformer to make the volt-age regulator responsive to three-phase voltage conditions.The final a -c output voltage is a measure of the voltage de-livered by the turbine generator compensated for the loadbeing carried. This output voltage is fed through the two con-trol switches and a voltage -adjusting resistor into two imped-

Figs. 9 (left) and 10 (below) show the facility with whichbrush replacement can be made. The brush shunt is seento terminate in a jack. A pull on the spring tension keyholding plug and jack together, and a lift on the brush -holder finger, allows the brush to be lifted out. Locatingone pressure finger on each side of a brushholder (Fig. 9)does away with the awkwardness resulting when pres-sure -finger centers are located on the same holder side.

JULY, 1947 119

Fig. 11-The curve of acapacitor -reactance im-pedance for a static vol-tage regulator. Selectionof proper capacitance andreactance values creates Ean intersection corres-ponding approximatelyto the normal a -c voltageon the turbine generator.

Normal A -C Volts

ance circuits, one capacitive and the other inductive, eachterminating in a Rectox.

The curve of capacitor -reactance impedances for a staticvoltage regulator is shown in Fig. 11. The relation betweenthe applied a -c voltage E and the current I through a capaci-tor is a straight line OPC. The curve OPX is non-linear, andshows the relation of the a -c applied voltage and the currentin a saturated reactor. With the proper capacitance and re-actance these two curves intersect as indicated. This point ischosen purposely to correspond approximately to the normala -c voltage on the turbine generator. The voltage -adjustingresistance unit permits reasonable variations in this voltage asdesired by station operation.

The two Rectox units are connected so that their d -c outputvoltages, which are equal when the a -c generator voltage isnormal, are in series through a mid -tapped resistor. Underthis condition there is no voltage output at YY, (Fig. 8).

If the generator voltage drops, the voltage applied to thecapacitor and reactor is reduced so that their delivered volt-ages change in accordance with the curve of Fig. 11. Thisproduces a different d -c output voltage from each of the twoRectox units. When this occurs, the voltage that appearsacross points YY is fed into the Rototrol control field. Thepolarity is such as to cause the "buck -boost" Rototrol to in-crease its voltage in magnitude depending upon the drop inthe a -c voltage. The Rototrol pilot exciter thereupon in-creases its voltage and adds excitation into field 2 of the mainexciter. This raises the excitation on the main exciter and inturn the excitation of the a -c machine, until its voltage is re-turned to normal. There is then no difference in the outputfrom the two rectifiers and thus no voltage is produced at YYand the Rototrol control field operates at zero excitation.

If the voltage of the a -c generator increases, say due to re-duction in load, the same action takes place, but of oppositepolarity, as can be seen in Fig. 8. In this case, the polarityof the voltage from YY is reversed and, when delivered to thepilot Rototrol control field, puts a voltage of opposite polarityon field 2 of the main exciter. This bucks its voltage down-ward and decreases the excitation on the main exciter and onthe field of the a -c machine, bringing the a -c voltage back toits normal value.

At very low voltages the currents in the capacitance andthe reactance circuits differ appreciably. This means that ifconductors A and B of the turbine generator unit are shortcircuited, in which case the potential transformer voltage isextremely low, the regulating equipment still provides an ad-ditive voltage to cause the Rototrol pilot to boost the excita-tion on the main exciter in an attempt to return the a -cterminal voltage to normal.

A new Rototrol excitation scheme is under development, inwhich the functions of the main exciter, the pilot exciter, andthe contact -making type of voltage regulator are all incor-porated in one two -stage main exciter without a pilot exciter.The automatic voltage -regulating potential and control unitshave no moving parts. With such a scheme the same stability,speed of operation, and smoothness of control are obtainedwith only one rotating exciter, as was secured in the past witha main exciter, a pilot exciter, and a voltage regulator.

Higher Excitation VoltagesDirect -connected exciters can readily be built in capacities

up to 300 kw at 250 volts. Above these ratings a higher volt-age is preferable to simplify exciter current collection andreduce the size of generator collector rings, thus simplifyingoperation and reducing maintenance. The largest 3600 -rpmgenerator contemplated is 150 000 kw, which requires in theneighborhood of 400 -kw excitation.

Adoption of 375 volts as a new standard excitation voltageintroduces no problems for either exciter or a -c generator, asit is in the same insulation class as the present lower voltages.The excitation voltage for machines of 12 500 kw and smalleris 125 volts; for larger machines 250 volts has been standard.At present, Westinghouse is building two generators employ-ing 375 -volt excitation. One particular advantage of this volt-age is that, when used in a station with 250 -volt exciters anda 250 -volt standby unit, a new 125 -volt standby exciter canbe provided whose capacity need be only one-third that of theshaft -driven unit. The new 125 -volt exciter, connected inseries with the existing 250 -volt standby unit or bus, providesthe desired 375 -volt excitation.

Pictorial Wall Chart EasesElectron Tube Studies

FOR secondary schools, both junior and seniorhigh, a wall chart depicting the theory, basic

forms and applications of electron tubes has beenmade available by the School Service Division ofWestinghouse.

The chart gives a complete pictorial story,which is unique inasmuch as one chart never be-fore has contained such a wealth of detail in aform guaranteed to hold the attention and in-terest of students. It graphically outlines theatomic structure, and gives in a series of illus-trations, the manner in which electrons are freedin gas -discharge and photoelectric tubes, field,secondary, and thermionic emissions. It showsthe essential parts of tubes, together with thebasic structures of four types, the diode, triode,tetrode, and pentode. Six of the important appli-cations are pictorialized from artists' drawings:rectification, amplification, generation, changinglight to electricity (sound motion pictures andtelevision), control (welding), changing electricityto radiant energy (x-ray).

Finished in eight colors on heavy linen paper,the 25 -inch by 36 -inch chart has been priced at anominal $2.00. It will find many uses in physicsclasses, electrical theory lectures, radio communi-cation, and in general science classrooms.

>

120 WESTINGHOUSE ENGINEER

110

EALING FRISCHControl Engineering

Westinghouse Electric Corporation

Rototrol and Its Applications

What a difference a few fields make. A direct -current generator equippedwith one or more extra field windings becomes a rotating regulator of elec-trical or mechanical quantities, requiring for the job only a feeble indica-tion. It is capable of enormous amplification, high speed of response, goodaccuracy, and versatility. Virtually every industry has found use for it.

THE Rototrol is a rotating amplifier suitable for regulatingany quantity that can be measured electrically, such as

voltage, current, speed, torque, tension, or position. Funda-mentally it is similar in design and theory of operation to astandard d -c generator, but has a number of field windings,connected in various ways depending on the regulating prob-lems. The Rototrol performs its regulating functions entirelythrough interactions of these fields.

The Rototrol FieldsAny regulating device functions on the principle that the

regulated quantity is measured and compared to a fixedstandard or pattern. If deviation occurs, the regulator pro-duces corrective action to restore the quantity involved. Thevalue of the regulated quantity, as well as of the pattern,must be converted into proportional voltages that are im-pressed on the Rototrol field windings. The fields are arrangedto oppose each other, and are related in strength so that theresultant magnetomotive force is zero when the regulatedquantity is correct. Any deviation produces a definite cor-rective magnetomotive force that alters the output voltage ofthe Rototrol in a direction to restore the quantity to normal.

In its simplest form the Rototrol, Fig. 1, has the followingfield windings:

A self -energizing field, connected in the Rototrol armaturecircuit, and described below.

A pattern field, commonly energized from a constant -voltagesource. It sets the pattern for the regulated quantity.

A pilot field measures the regulated quantity and comparesit with the pattern. Sometimes the voltages that provide themeasurement for the pattern and regulated quantity are com-pared externally to the Rototrol, and their voltage differenceimpressed on a single Rototrol control field.

The Self -Energizing Field

To obtain accurate regulation the net ampere -turns of pat-tern and pilot fields must be zero, or nearly so, when theregulated quantity is as desired. The ampere -turns requiredfor Rototrol excitation, therefore, must be produced by othermeans. This is the function of the self -energizing field, theheart of the Rototrol, and it is responsible for the sensitivity.

The effect of the self -energizing field is illustrated in Fig. 2.Assume first that the total excitation is supplied by the netampere -turns AT of pattern and pilot fields, the Rototroldevelops voltage VI. With the effect of the sell -energizing fieldincluded, the total ampere -turns of all fields equals AT

(1, X T.) and the voltage rises to V2. If the Rototrol cur-rent is now increased by adjustment of tuning resistor R1, theampere -turns AT required to develop V2 volts decrease untilthey become zero when the resistance line of the self -energiz-ing field circuit coincides with the straight part or air -gapportion of the saturation curve. For this condition the

steady-state amplification of the Rototrol is infinite. To ob-tain satisfactory regulation over the required range theoperation, therefore, is usually limited to the air -gap portionof the saturation curve.

Rototrol IntelligenceThe principles can be applied to the regulation of any

quantity that can be converted into a proportional d -c voltagefor supplying intelligence to the Rototrol fields. For example,the delivered torque of a d -c motor with constant excitationis proportional to the armature current and can be measuredby the voltage drop across a fixed resistance in the motor -armature circuit. Speed can be measured by the voltage of apilot generator coupled to the motor.

Two or more fields are sometimes required to measure agiven quantity. For example, the speed of a d -c motor withconstant excitation is proportional to the counter emf of themotor. This quantity is equal to the generator voltage minusthe IR-drop in the armature circuit and can be measured bythe combined effect of two properly adjusted Rototrol fieldsconnected differentially across the motor terminals and themotor commutating field respectively.

Other fields on the Rototrol may serve to modify the regu-lated quantity when conditions change. It may, for example,be desired to reduce the voltage of a regulated generator as afunction of load current. This is accomplished by a secondaryfield connected across the generator commutating field, whichmodifies the effect of the pattern field when the voltage dropacross this field, and consequently load current changes.

Rototrol Amplification and Forcing ActionThe regulating ability of the Rototrol depends on amplifi-

cation factors, which fall into two classifications:1-The steady-state system amplification, which deter-

mines the sensitivity and accuracy of regulation.2-The dynamic system amplification, which determines

the speed at which the regulated quantity is restored after adisturbance. The effect of the dynamic amplification is re-ferred to as Rototrol forcing action.

Amplification on a voltage basis should not be confused withpower amplification. It is the change in the generator voltagecaused by a one -volt change in pilot -field voltage. If theamplification factor is 10, a change of one volt in field voltage,therefore, produces a 10 -volt change at the load.

Dynamic amplification is the system amplification obtainedwithout the benefit of the self -energizing field.

Steady-state amplification is the system amplification ob-tained with the assistance of the self -energizing field. For aproperly tuned Rototrol armature circuit the steady-stateamplification is infinite.

The effect of high dynamic amplification on the speed ofresponse for typical values of field -time constants for a simple

JULY, 1947I

121

Rototrol voltage regulator is illustrated in Fig. 3. With anamplification of 16, the time of recovery to within one percentof the initial voltage is only a small percentage of the cor-responding time for unity amplification. With infinite steady-state amplification, the voltage eventually reaches 100 per-cent regardless of the dynamic amplification.

Although high dynamic amplification greatly increases theaccuracy during transients it also tends to increase overshoot-ing and hunting, as Fig. 3 shows. If the dynamic amplificationis so high that these effects become objectionable, specialstabilizing and anti -hunting circuits or devices must be used.

The forcing effect is illustrated in the lower part of Fig. 3.For unity dynamic amplification the net ampere -turns ormagnetomotive force of pattern and pilot fields resulting fromthe drop in generator voltage are just sufficient to raise theRototrol voltage by enough to restore generator voltage to100 percent. For higher amplification, however, the resultingnet magnetomotive force equals that actually required, multi-plied by the amplification factor. As a consequence, the Roto-trol voltage builds up rapidly to a magnitude much higherthan that ultimately required for generator excitation. Therate of build-up is limited only by the time delay of theRototrol fields, which lies within the control of the designer.An excess in Rototrol voltage is thus available to force addi-tional current through the sluggish generator field and hastenvoltage build-up. As the generator voltage approaches thecorrect value the net magnetomotive force of pattern andpilot fields approaches zero, and the additional currentthrough the self -energizing field becomes just sufficient tomaintain the generator excitation at the new value.

Two -Stage Rototrol

The dynamic amplification of the standard Rototrol, whileit is considerably higher than is required by the large majorityof applications, cannot be increased beyond a certain limitwithout excessive power input to the pattern and pilot fields.Very high amplification with a low power input can be ob-tained when necessary by two stages of amplification. Thiscan be accomplished by the use of two separate machines, onea conventional exciter and the other a Rototrol. A preferredmethod, however, is to use a special two -stage Rototrol,where both stages of amplification are obtained with a singlearmature. Sensitivity and speed of response of such construc-tion are extremely high. Tests have shown that a 10 -kw, two-stage Rototrol can be controlled over its entire range with amaximum power input of less than 0.01 watt.

Current -Limit RototrolThe current -limit Rototrol is often used in regulating sys-

tems requiring current -limiting features. It is similar to anormal d -c generator except for the specially designed polepieces and the magnetic shunts between poles as shown inFig. 4. Coils B on the magnetic shunts and coils A on the maincoils are connected in series and energized with a current pro-portional to the generator or motor -armature current.

With correctly adjusted magnetic -shunt air gaps the totalmain -pole flux passes through the magnetic shunts, and theRototrol voltage becomes zero. With increasing field currentand flux the magnetic shunts become saturated, and furthercurrent increase beyond this point produces a flux in thearmature air gaps. The unusual saturation curve obtained inthis manner is shown in Fig. 5. This characteristic is utilizedto limit to a safe value the current of any Rototrol-controlled,d -c machine, when the current -limiting features cannot con-veniently be incorporated in the Rototrol.

Fundamental Applications of the RototrolVoltage Regulation

A schematic diagram of the simplest form of voltage regula-tion of a d -c generator by Rototrol is shown in Fig. 1. Thepattern field PF is energized from a constant -voltage referencebus, and the differentially connected pilot field VF is ener-gized by the generator. When the generator voltage is correct,the ampere -turns produced by the two fields are equal, andthe actual Rototrol excitation is supplied by the self -energiz-ing field SF, assuming that the field circuit is tuned perfectly.

The regulated voltage is a direct function of the pattern -field current. Accuracy of regulation, therefore, depends onvariations in reference voltage. If the reference bus is ener-gized from an exciter, an error in regulation of at least two tothree percent must be expected. When higher accuracy isrequired special provisions must be made to obtain a moreconstant reference voltage, as indicated in Fig. 6.

Current Regulation

Regulation of current is used principally as a means ofmaintaining strip tension in wind-up and unwind operationsutilizing core -type reels in the steel, textile, paper, film, plas-tic, and rubber industries. Other applications are rapid accel-eration and deceleration of high -inertia loads and variouselectrochemical processes. For practically all of these appli-cations, the Rototrol regulators, which supply excitation forthe regulated motor or generator, have a pattern field con-nected to a constant -voltage, d -c source and a differentialpilot field connected across the commutating field of themachines for current measurement. The pilot -field currentchanges with the temperature of the commutating field,making accurate regulation impossible with this method.

For some electrochemical work, where high direct currentsmust be regulated within very narrow limits, an interestingmethod of obtaining accurate intelligence is available as illus-trated in Fig. 7. This method requires a saturating three-legged two -coil reactor, with one coil energized by a connec-tion made to a constant -voltage a -c source, while the otheris energized by a bus carrying the main d -c generator currentand acting as a one -turn coil. Below the saturation point, thereactance of the a -c coil remains nearly constant, independentof the generator current; the coil current does not change.The coil current is rectified in a Rectox rectifier and ener-gizes the Rototrol pilot field. When the generator currentexceeds a certain amount the reactor becomes saturated andthe reactance of the a -c coil decreases rapidly. As a resultthe current in the pilot field rises sharply with increasinggenerator current as illustrated by the Rototrol field curves.The generator current at which saturation occurs can beadjusted by varying the reactor air gap.

The reactor circuit is energized by the secondary voltage ofa transformer with two secondary windings. After rectifica-tion the output voltage is impressed on the Rototrol patternfield PF and the other on the current field CF. Ampere -turns of the pattern field are nearly constant and independentof d -c generator current. The Rototrol current and patternfields are connected to oppose each other, and the generatorcurrent is regulated for a value corresponding to the pointof intersection of the two field -current curves.

Changes in the a -c bus voltage within reasonable limits donot affect the regulation because the ampere -turns of the pat-tern and current fields change in the same proportion and thecurrent value at the point of intersection of the field curvesremains unchanged. As a result, a given direct current always

122 WESTINGHOUSE ENGINEER

produces the same intelligence; therefore, accurate regulation is possible.

Tension Regulation

Wind-up and unwind operations of core -type reels, such as those inuse on the new Sendzimir sheet mills, usually are tension -controlled byregulation of the reel -motor current. This type of tension regulationpresents several problems in that it is necessary to compensate for coilbuild-up and for reel and coil inertia during speed changes. The com-plications resulting from attempts at compensation can be avoided iftensiometers are used to supply the information required by the Roto-trol regulators.

The tensiometers may be spring loaded or, as shown in Fig. 8, com-pressed -air loaded. The tension in the rolled strip is actually developedby the tensiometer action, and desired tension is obtained by variationof the air pressure. Because of the special lever arrangement, the tensionvaries with the position of the tensiometer roller above the pass line asindicated by the tension curves. To hold tension accurately the ten-siometer must be kept in range as close to the middle position as pos-sible, by regulation of the relative speed of the reel motor. This positionshould correspond to a point on the flat portion of the tension curves toobtain the highest accuracy.

Tensiometers have two contact units to obtain a measure of thedeparture from the middle position. The contact units consist of flatphosphor bronze springs with silver contact tips connected to resistors.If the position changes from the middle, the contacts of one or the otherclose gradually to short circuit the associated resistance.

A schematic diagram of a control scheme for a tension regulator isshown in Fig. 9 where power for the reel motor is supplied by a separategenerator. A motor -operated rheostat driven by motor RHM controlsthe reel -motor field to obtain the necessary range in motor speed tocompensate for coil build-up. The tension is regulated by control of thegenerator excitation supplied by Rototrol R. To reduce the physicaldimensions of reel motor and generator to a minimum the motor fieldmust be controlled during the winding operation so that the generatorvoltage is maintained at a value approximately proportional to thestrip speed as measured by the voltage of pilot generator PG.

Before starting the mill, line contactor M is open and the rheostatmotor is connected to the d -c supply bus by auxiliary relay M., whichcauses the rheostat to be moved to the minimum -field position corre-sponding to empty -reel speed. With no metal in the mill, the tensiometeris in the extreme low-tension position and limit switch TLS is open, andrelay A is de -energized. When the reel motor is started by closing ofcontactor M, the Rototrol pattern field PF is connected to the pilotgenerator and voltage field VF is connected to the reel generator bus,while the tension field TF is disconnected. The Rototrol therefore func-tions as a voltage regulator and regulates the generator voltage to avalue proportional to the mill speed. The fields are so designed that theresulting reel -motor "threading" speed is slightly higher than the corre-sponding operating speed, to obtain a tight initial wrap after the reel

PF IV

D -CGenerato

Load

Tuning Resistor

Voltage -AdjustingRheostat

PF-Pattern FieldVF-Voltage Pilot FieldSF-Self-Energizing Field >

Constant -Voltage Bus

Fig. 1-Schematic of Rototrolwhen used as a voltage regulator.

Fig. 2 -Saturationcurve for Rototrol.

Is x Ts -n -TAT

VI

V2

Resistance Line forSelf -Energizing Field

Saturation Curve

PatternField

Self -Energizing Field

Generator

AT= Net Ampere -Turns of Pattern and Pilot FieldsIS a TS -Ampere -Turns of the Self -Energizing Field

Rototrol Field-Ampere-Turns

JULY, 1947 123

104A = 16

102

100

A =4

98

11111".-

96

94

ro0111111 A=1

92

90

200

180

160

LIde, 140ea

120

CcG 100

80

13 05 10Time-Seconds

I 5 20

A - 16

El t l

a.

E3no

A -4

A-3A -2

A = 1

0 10Time-Seconds

Fig. 3-Schematic view of Rototrol used as voltage regulator.Curves show the effect of dynamic amplification on the speedof response of generator voltage and on Rototrol forcing.

Brass Block'

0s 15

Fig. 4 --Pole piece of current -limit Rototrol.

Fig. 5-Saturation curve obtained with spe-cial pole piece used on current -limit Rototrols.

S

00

C4

0

20

Man PolesSaturate Here

titin

///N,--Si - Normal Saturation

CurvePI//1/:

/// .

S2r

-Saturation CurveResulting fromcreased MagneticShunt Air Gaps

-Saturation CurveResulting fromcreased MagneticShunt Air Gaps

t Saturates1

In-

//n/

S3De-

Shur___1---- --17ow AO

////j/Magnetic Here

Field Current

has been threaded by the automatic belt wrapper. The speedcan be adjusted by a threading -speed rheostat.

When the coil is properly started, tension develops andforces the tensiometer toward the high-tension position andthe tensiometer limit switch TLS closes. Relay A becomesenergized and connects the Rototrol tension field TF to thetensiometer bridge circuit while disconnecting fields PF andVF. The bridge circuit consists of the two tensiometer resistorunits LT and HT and the two fixed resistors RT. For correcttensiometer position the current in field TF is zero, but thefield is energized when the position changes and resistor LTor HT is gradually short circuited as a result of the reel -motor speed being too high or too low. The Rototrol thenmaintains correct tensiometer position by regulating the gen-erator voltage and, consequently, the reel -motor speed.

As the coil builds up, the generator voltage decreasesgradually unless the motor field current is increased sufficient-ly to compensate for the build-up. For this purpose the driv-ing motor RHM for the reel -motor rheostat is connected togenerator RHG through relay contacts Ma, when the reelmotor is started by closing of the line contactor M. Thegenerator is actually a Rototrol with a slightly under -tunedshunt self -energizing field SF in addition to the pattern fieldPF, connected to the mill -motor pilot generator to obtain ameasure of strip speed, and pilot field VF which measures thevoltage of the reel generator. When the voltage is of the cor-rect value the two latter fields balance and the Rototrol volt-age is zero. When the voltage decreases as a result of coilbuild-up the Rototrol becomes energized and supplies powerto the rheostat motor in a direction to increase the reel -motorfield. The movement stops as soon as the reel generator volt-age again rises to the correct value as a result of tensiometeraction. A Rectox block is connected in the rheostat motor cir-cuit to prevent mis-operation or hunting caused by temporaryreversal of the Rototrol. This slow step-by-step movement ofthe rheostat continues until the coil is completed.

The tension regulator maintains tension also when the millis stopped at any time during the coil build-up. Because thereel motor cannot carry rated current without damage whenat standstill, the value of the stalled tension and, conse-quently, of the motor current, is reduced by lowering thetensiometer air pressure when the mill stops.

Speed Regulation with Current LimitWhere the speed of a motor in a variable -voltage system

must be changed or reversed frequently, the motor must bebrought to the desired speed as quickly as possible withoutdangerous overloading of the equipment. This must be ac-complished regardless of the speed at which the master con-troller is moved to the selected speed position. A typicalexample is a mine -hoist drive, which requires good speedregulation in addition to positive limitation of load currentunder all conditions including motor stalling and plugging.

A standard Rototrol in combination with a current -limitRototrol, as described, is well suited for this application andmakes possible a control system that, for simplicity and reli-ability, cannot be matched by other types of control. Aschematic diagram of a hoist drive with Rototrol speed andcurrent -limit regulation is shown in Fig. 18.

The generator field is energized by a Rototrol, which oper-ates entirely as a speed regulator for steady-state operation.The speed of a motor with constant excitation is proportionalto the counter emf of the motor, which in turn equals thegenerator voltages less the IR drop in the armature circuit.The Rototrol intelligence for motor speed, therefore, is ob-tained by two differentially connected fields VF and IRFdesigned to measure voltage and current respectively. Thecombined effect of the fields is compared to the Rototrol pat-tern field PF, and the Rototrol functions to reduce to zero anydifference existing between the ampere -turns of the speed -measuring fields and the pattern field. The pattern field isenergized by a constant voltage, and the field current is ad-justed and reversed by contacts on the master -speed con-troller. Any change in the pattern -field current causes asimilar change in motor speed.

While motor speed is being changed as a result of a rapidchange of the pattern -field current, motor accelerating cur-rent is safely limited by the current -limit Rototrol. Thismachine has an unusual saturation curve. The Rototrol volt-age remains close to zero as long as the net field ampere -turnsare below a certain amount A corresponding to the maximumpermissible motor current. For currents in excess of this valuethe voltage rises rapidly and energizes field CLF on the mainRototrol. The ampere -turns of this field oppose the netdifference in ampere -turns of the speed -measuring fields and

3

a

Exciter A -C Generator

Self -Energizing Field

Saturating

Roi

ReactorCNFControl

Field

RegulatedVoltage

otrol

Rectifiers

BalancingPoint

-.ReactorCurrent

IXCapacitorCurrent

IC

II

II

Voltage -Adjusting -=Rheostat

Capacitor

Rototrol Field Current

Fig. 6-Schematic of Ro-totrol when used as avoltage regulator forlarge synchronous gen-erators. Also shown isRototrol field current asa function of the con-trolled generator voltage.

Fig. 7- Schematic viewof Rototrol and a saturat-ing reactor where highdirect currents must beregulated within narrowlimits for certain work.

Atte nating-CurrentSupply

Transformer

Saturating Reactor

Vernier Rheos at

Rectifiers

Air -Gap Adjustment

Load

Generator

TuringRes storrrrh

MinimumAir Gap

Field CFField PF-Maximum

MaximumAir Gap

VernierAdjustment

Generator Current

the pattern field that causes the speed change, and will notpermit the motor current to exceed the preset magnitude.

As the motor speed approaches the selected value the netdifference in the ampere -turns of the fields decreases gradual-ly, and less voltage is required to produce the required oppos-ing ampere -turns of the current -limiting field. If the current -limit Rototrol were excited by current field CF alone, themotor accelerating current would vary considerably through-out the speed range.

To maintain accelerating current constant, independent ofspeed, the current -limit Rototrol CLR has two additionalfields PF and VF, which are connected in series with the pat-tern and voltage fields on the main Rototrol. For steady-statecondition the combined effect of the two fields is zero. Fortransient conditions, however, the net ampere -turns of thefields produce the necessary increase in Rototrol voltage forenergizing the current -limit field CLF without any change inaccelerating current, as illustrated in Fig. 18. As the motoraccelerates to the value set by the pattern field, the voltagefield ampere -turns gradually increase until they equal thepattern field ampere -turns, and the voltage of the current -limit Rototrol changes from B to A.

Power -Factor Regulation

When synchronous motors with constant excitation areused for drives with variable loads, the load -current powerfactor is subject to wide variations. If the motor is large anddepended on for system p -f correction, it becomes necessaryto provide automatic p -f regulation by control of the motorfield current. Depending on the conditions, it may be necessaryto maintain either the power factor or the reactive kva at aconstant value.

provides a simple and efficient solution for thisimportant regulating problem. As shown in the schematicdiagram, Fig. 10, the Rototrol is provided with two differen-tially connected voltage fields VFI and VF2, which are inbalance when the power factor or the reactive kva is at thepreset value. Any deviation causes an increase or a decreasein the motor excitation to produce the necessary correctionin the regulated value. The two fields are energized by recti-fied voltages from a system of potential transformers, currenttransformers, and rheostats as shown. Vector diagrams have

been drawn to illustrate the theory of operation for regulationof power factor and reactive kva.

Letters A to F are used for corresponding potentials in theschematic diagram and the vector diagrams. Vectors OA andEF represent the motor current and phase voltage for phaseA. The two Rototrol fields VFI and VF2 are energized withrectified voltages CF and DF, respectively, and the Rototrolfunctions to raise or lower the motor -field current to restorethe voltages to equal values whenever a disturbance occurs.Current vector EFI produces excess ampere -turns in fieldVF2 (vector DFI is longer than CF1) thereby causing a re-duction of the motor excitation until the phase -current vectoragain coincides with vector EF.

Power -Factor Regulator

The reactive-kva adjusting rheostat is placed in the middleposition. By means of the p -f adjusting rheostat potential Dis shifted along vector AB and changes the regulated value ofthe power factor from 50 percent (leading) at point A to 100percent at point B.

Reactive-Kva Regulator

The p -f adjusting rheostat is placed in the 100 -percentposition. (Potential D coincides with potential B.) By movingthe reactive-kva adjusting rheostat away from the middleposition potential E is shifted on vector BC away from themiddle point El. Since vectors DF and CF are regulated toequal values by means of the Rototrol, it follows that thereactive -current component EEI is held at a constant value.

Rototrol Electronic RegulationFor many applications where extreme accuracy and speed

of response are required, the signal for the Rototrol controlfields is supplied through electronic amplifiers. The partner-ship of Rototrols and electronics has proved very successful inservice. Because of the tremendous amplification obtainedwith this combination, large machines can be controlled witha signal input of less than one ten -thousandth of a watt,thereby simplifying the introduction of the necessary antici-patory and anti -hunt intelligence. Typical applications aresectionalized paper machines, rod mills, side -register windercontrol, and flying shears.

Oil Dash Pot

V` III

1.1F14111

111

-1

S. inl tvaecr

Flat BronzeSprings

Low -TensionContacts

Resistor iii k 'N,-.`1111:tesisto9

SilverstatContact

I Units

High -Tension

ContactsHT

RegulatedAir Pressure

L Insulated Contact Supports

High AirPressure

I RegulatedTension

D - DistanceAbove Pass Line

HTContacts-, I I I LT

Close I ...J. -ContactsClose

Low AirPressure I 'Regulated Tension

Fig. 8 (left)-Schethatic view and ten-sion curves of compressed -air loaded ten-siometer, and Fig. 9 (below) tensioncontrol where power for reel motor issupplied by a separate generator.

Limit Switch Opens InExtreme Low -Tension Position

TensiometerWinding Reel

TLSLT HT

Mill RT RTMotor

RHG

LS RHM,_

LSD

VI-

and Coil

Threading- I ReelSpeed Motor

Rheostat

31 rill-. Reel -MotorRheostat

Fig. 10-Simplified version of the Roto-trol circuit as used for a -c generatorpower -factor regulation and vector dia-grams for power -factor and reactive kva.

SynchronousMotorIncoming 3 -Phase Line

B

Phase An :le AF

F l. -

PhaseCurreit

FieldVFI

PhaseVoltage

FieldVF2

Power -Factor

AdjustingRheostat

Reactive-Kva

AdjustingRheostat

ReactiveCurrent

Com-ponent F 1

FieldVFI

B C El E B(D)Power -Factor Regulation Reactive-Kva Regulation

711 VF2Lower)

PhaseVoltage

PhaseCurrent

FieldVF2

Distance "D Above Pass Line D -C Supply Bus

TABLE I-REPRESENTATIVE USES OF

Fig. 11

A

Variable Speed/A -C Generator

Reactance

Constant Potential Transformer/

Fig. 13D -C Suppl

Speed -Adjusting Rheostat

Fig. 15

NN Ma terSwi ch

Contacts

To A -CSupplySource

F -Combined Pattern anoSelf -Energizing Fields

CombinatiorSeries andShunt Self -Energizing

Field D -C Supply Bus

GeneratorDifferential

Field

igVF

FairRototrol Ampere -Turns

Fig. 17

D -C Supply Bus

These Lines RepresentTotal Ampere -Turns for Relative

Rheostat Displacement as Indicated

VFSF r-,CNF

4.4Ampere -Turns

Rototrol Saturation Curve

Motor Current

Rudder MotorRudder

Stock

°-)0)c

/ Rudder//Follow -UpRheostat

0 0

Rheos at Displacement-Percent

Motor Speed Curve

Fig. 12

] Speed-

Rheostat

Fig. 14

D -C Supply Bus

Winding Reel

SI

IncreasedSpeed

DecreasedSpeed-.,

Fig. 16

Fig. 18

Tension -AdjustingRheostat

D -C Su Bus

Tension Regulator

GeneratorVoltage

Proportionalto Str,p Speed

DriveMotor

ThreadingSpeed

F haostat

Proportional toArc Curren

Field CFElec rode

0 -LowersZ Balance-----

-Poin

Elec rodeRaises-

Propor ional toArc Voltage Field VF

Electrode Current..,Rectifiers

Adiustino4.

irrwr,Master Switch Contacts

dCF

ElectrodeMotor O

nstant Voltage Bus

a.

o

r.. -s

50t Hoist ";\

Red dation Acceleratior.0gi

0c' 0

10

c,..,''

Acceleration50

Retardation

t 1 Lower f___V' 10,),. _../

Current-Percent

Fielc A -T A

CFI

V Fr -1

QuantityRegulated

Rototrol FieldIntelligence

Volt age

A -c generator voltage.

D -c generator voltage.

Vo'ts-per-

Cycle

A -c generator voltage pro-portional to frequency.

Speed

Voltage of pilot generatorcoupled to the regulated

motor.

Motor counter emf meas-ured by the combinedeffect of motor current

and voltage.

Tension

Motor armature current. ,

Tensiometer position asmeasured by operatic'', ofSilverstat contact units.

Current D -c generator armature'current.

SpeedwithCurrentLimit

Motor counter emf forspeed regulation. Cumnt

for current limitation.

Speed andTorque

Motor armature current.Motor voltage.

Power Load current and voltage.

PositionDepending on the particu-

lar application.

126 WESTINGHOUSE ENGINEER

ROTOTROLS AND THEIR ADVANTAGES

Typical ApplicationsTypicalDiagramReference

How Rototrol Performs the Assigned Function Why Rototrol Is Selected

]Central -station generators.Illkh-frequency generators forinduction heating. Ships'

service generators.

Fig. 6

Generator voltage is impressed on two circuits with a saturatingreactor and a capacitor, and supplied through rectifier to theRototrol control fields with opposing polarities. Capacitor cur-rent is a linear function of voltage; reactor current is non-linear.Generator is regulated for voltage corresponding to intersection

of field ampere -turn curves.

The control -field current and, consequently, theregulated voltage, is independent of temperature

changes, which permits accurate regulation.

Generators to supply variablevoltage for speed control ref-

erence bus.Fig. 1

Pattern field PF, energized by d -c supply bus, is connected dif-ferentially to voltage field VF, energized by generator voltage.Generator is regulated for voltage at which the two field ampere-

turns are equal.

Great simplicity and stability. Because the regu-lated voltage depends on temperature variationsand supply -bus voltage, accuracy of regulation is

limited.

Veiable-speed generators forpower supply to: wind -tunnelmodel motors; ships' cargo

pump motors.

F 1118.

A pattern field of fixed magnitude is established by the constant-

P°tential transformer. The differentially connected voltage fieldis energized by a transformer through an air -gap reactor and arectifier. Hence, the field current is a function of generator volts -

per -cycle, which is regulated for equal field ampere -turns.

The control scheme is simple and gives good ac-curacy when the frequency range and the volts -per -cycle range are not excessive. Accuracy

decreases at low frequencies.

Single -motor paper machinedrives; wind -tunnel motors;flying shears;. turbo-compres-so& testing rigs; rubber andplastic calender drives; dy-

namometers.

fl -12

Approximate speed adjustment is obtained by rheostat controlof generator main field MF and the motor field. The Rototrol,connected to the auxiliary field AF, supplies required correctionfor accurate speed regulation. Rototrol field SPF measures themotor speed and compares it to a set pattern as measured byfield PF. Any deviation in motor speed upsets the field balance

and starts corrective action.

Series connection of Rototrol pattern field and-pilot generator compensates for changes in sup

Div voltage and field temperature. Accurate- - .regulation of speed over a wide range is possible.

I

Teed drives on adjustable-voltage planers; high-speedelevators; skip hoists; paper

for textile winders; supercal-enders.

i

-

Winding and un-wind drivesinoteel, textile, paper, plastic,and rubber plants where ten-sion of the processed material

I must be accurately held.

Fig. 13

Motor current is measured by Rototrol field CF. Motor voltageis measured by the differentially connected field VF. The result-ing total ampere -turns provide a measure of motor emf and, con-sequently, of motor speed. Good regulation over a wide speedrange is obtained by connecting the two Rototrol pattern fieldsPF, which also serve as self -energizing fields, in a bridge circuit

with the generator shunt fields.

Because a pilot generator is avoided, this methodof speed control is preferred except when thehighest accuracy is required. It provides goodspeed regulation over wide ranges. Drives withspeed ranges up to 1: 120 are in operation. Highrunning and starting torques are available down

to the lowest speed.

Fig. 14

Reel -motor field MF is adjusted to obtain correct threading orempty reel speed. Auxiliary field AF is energized by the Roto-trol. The Rototrol controls motor flux, to hold the motor cur-rent, measured by Rototrol field CF, at a fixed preset value.Consequently, strip tension can be adjusted by a rheostat in thepattern -field, PF, circuit. Rototrol field IF is energized onlyduring acceleration and deceleration and serves to modify the

regulated current for inertia compensation.

. .TensionTension is automatically held within close limitswhile the reel speed gradually increases to com-Pe nsate for coil build-up. The strip speed can beincreased without danger to the machinery and

without sacrificing the quality of the coils.

Figs. 8and 9

The necessary tension in the strip is developed by compressed -air loaded tensiometer, and control of tension is reduced to repo-lation of the tensiometer position. Changes in position affect theresistance of LT and HT, connected in a bridge circuit with theRototrol field TF. The Rototrol controls generator excitationand, consequently, reel motor speed to hold the tensiometer inthe middle position. Motor field rheostat automatically corn-

pensates for coil build-up. For detail description, see text.

The complications of inertia compensation areavoided by use of tensiometers. Tension can bemaintained accurately during speed changes aswell as for steady-state operation. The methodis recommended where tension must be variedover a wide range. A tandem cold mill with a top

speed in excess of 55 mph is in operation.

!Electrochemical processes re-rquiring accurate current regu-

lation.Fig. 7

Rototrol pattern and current fields are energized through recti-fiers by a transformer energized by any constant a -c voltage.A three-legged saturating reactor, with one coil energized by thegenerator current is connected in the current field circuits. The

rises sharply as a function of generator current afterreactor saturation, providing an amplified current indication.

nRototrol current signal is independent of tern-perature. Accurate regulation (within 3,6 per-

cent) of high direct currents is possible.

Mine hoists; reversing bloom-ing mills; balancing machines;

car dumpers.Fig. 18

Rototrol acts as a regular counter-emf speed regulator throughinteraction of fields PF, VF, and IRF. Field CLF, energized bythe current -limit Rototrol CLR, modifies generator excitation toprevent excessive current peaks during speed changes or when

the motor torque exceeds a safe value.

The control provides: 1-maximum accelerationor deceleration with safe motor current; 2-goodspeed regulation at all speeds; 3-full protection

during periods of slugging and stalling.

kctric shovels; electric drag-

es'lines; steel -mill table drivreversing -mill screwdownincrives; crop shear drives.

Fig. 15

Unlike most applications, the Rototrol operates beyond thesaturation point at no load. When the drive -motor load in-creases, the differential current field CF gradually reduces theRototrol and the generator voltage until it reaches near zerovalue, when the current reaches the maximum permissible stallvalue. The differential voltage field VF provides the desired

slope of the load curve at low speeds.

Because of the high forcing effect of the Rototrol,motor acceleration, and deceleration are ex -tremely fast, thereby producing 20% reduction. the time for a swing cycle of a shovel usingconventional magnetic control. The stall andplugging current peaks are reduced nearly 50%.

Electric -arc furnace. Fig. 16

Because of the relationship existing between arc -furnace powerand electrode position it is possible to control the power input bypositioning of the electrodes. Rectified voltages proportional toelectrode current and arc voltage are impressed on the differ-entially connected Rototrol fields CF and VF. The net ampere-turns at the field are zero for the desired value of power input.A change in this value produces a voltage in the Rototrol andcauses the electrode motor to restore balance in the field ampere-

turns.

Accurate control of the furnace power results ina drastic reduction in electrode breakage andgeneral maintenance. There is an appreciableincrease in furnace output and the power con -

sumption per ton of steel is reduced.

Positioning of the cutters onship-propeller milling ma-ch&es in response to a tracerfollowing a master pattern.Side -register control for accu-rate coil winding. Position-

ing of ship rudders.

Fig. 17

Duplicate potentiometer rheostats are coupled to the pilot housewheel and the rudder stack. The Rototrol control field CNFmeasures the relative displacement between the two rheostatcontact arms. The Rototrol energizes the generator and turnsthe rudder motor to reduce the displacement nearly to zero. TheRototrol operates beyond the point of saturation to obtainreasonably constant motor speed for large displacements. Thepurpose of field is to slow down the motor shortly before reach-ing the final position, thereby preventing overtravel and hunting.

Rototrol steering -gear control, while extremelysimple, provides better than 54° accuracy in rud-der positioning. For propeller -milling machines,where the Rototrol operates in partnership withan electronic amplifier, extreme accuracy andspeed of response is obtained. The milling cutterfollows the master pattern with an accuracy of

0.002 inch.

JULY, 1947 127

What's New!A More Rugged Mill Motor

MILLmotors must be able to take it,

with a margin of reserve strengthconsidered unnecessary for other applica-tions. The grueling work they perform-driving bloom rolls in a steel mill; supply-ing power to massive shears; poweringhigh-speed, hot -cut saws, ladle -hoistcranes, traveling cranes-any of the hun-dred and one jobs that must be done ingrimy, muck -laden atmospheres-de-mands a construction robustness far ex-ceeding the normal.

Present-day mill motors have doneand continue to do an eminently satisfac-tory job-steel production at today's as-tronomical figures would be impossiblewithout them. However, advances inmotor -manufacturing techniques, devel-opment of new insulations, modificationssuggested by operating practices, haveresulted in a new mill motor in the 5- to200 -hp range that, while keeping the ex-cellent characteristics of the earlier de-sign, embodies new features in armature,field, brush and body construction.

Instead of holding armature coils in theslots with core bands they are now heldin place by wedges made of class B mate-rial, designed to give the armature as higha safe maximum speed as if bands wereused; the wedges reduce armature lossesand increase efficiency. The commutatorcontinues to be of bolted construction,with a fit directly on the shaft. Liberalwearing depth of bars and an extraordi-nary length of neck, provide for a greatnumber of refacings, and still allow ade-

quate contact area between coil ends andthe bars.

The new field -coil design has the en-tire coil wrapped with tape. The coils arefirst wound on a mould, each turn sepa-rated by asbestos paper. Mica and glasstape are then wound on, but before im-pregnating, the coil is assembled over asteel shell. A steel washer is then placedon top and tack -welded at several places.Thus the coil becomes a solid mass withthe edges of the conductors lined up evenbefore impregnation. Vacuum and pres-sure impregnating in Thermoset varnishfills all crevices and voids.

The brushholders are now of cast brassand are machined accurately for closebrush fit. The triple -paired brushes areheld in place by pressure fingers, essenti-ally a part of the spring which winds andunwinds as the brushes ride in the box.Heavy porcelain insulating bushingsmount over a Micarta tube fitted on sup-porting pins.

The old thrust washers and collars onbearing housings have been removed, andthe center moved about two inches to-ward the shaft extension, reducing shaftstress greatly. The new cylindrical rollerbearing is narrower than the old. Greaseis introduced on the outboard side whereit is stored in an annular recess; from hereit works through the bearing to be storedin a similar recess on the inboard side.

The bearing housings are so construct-ed that, in normal maintenance, the en-tire housing with outer race and rollerscan be pulled away from the armaturewithout dismantling. The cast steel frame

A Concentrated -Filament Projection Lamp

This five -inch 300 -watt lamp doesa man-size job in the slide -film pro-jection field. The high -efficiencytungsten" biplane" filament for homemotion pictures has been adapted forthis new use. Composed of twin ser-ies of three tungsten coils on a float-ing -bridge suspension, a 25 -hourlamp life at high brightness is as-sured because the filament is free tomove under heat -expansion pres-sure. The extreme light concentra-tion-twice that of the former 300 -watt lamp-requires natural convec-tion cooling only, making unneces-sary expensive blowers and ducts.The new lamp promises bigger andbrighter screening to meet the de-mand for a longer projection " throw"in schools, at public lectures, and forsales -training classes.

The automatic rural line recloserbreaker mounted one. pole cross -arm.

is split, which allows the upper frame to beswung back on its hinge, an importantconvenience in accessibility for normalmaintenance.

Rural Line Recloserr-rniE old adage, "Out of sight; out of1 mind," is particularly fitting when

viewing the new Westinghouse single -poleautomatic reclosing breaker, which can behung on a cross arm or directly on thepole to protect circuits operating at volt-ages up to 15 000-and then forgotten.Fully enclosed in a metallic tank andweighing only 65 pounds, in many re-spects it is quite similar to its big brothersthe heavy-duty power breakers, in its useof De -ion grid arc interrupters, and its fastoperation. It is capable of reclosing thecircuit three times before locking out.

No auxiliary power source is needed,the force of the overload current openingthe breaker and at the same time storingenergy for reclosing. Either four time-lagtripping operations before lockout or twotime delays followed by two slow opera-tions and a lockout, can be selected bychanging a pipe plug, an adjustment thatcan be done in the field in less than a min-ute. When fast tripping is chosen, the firstcycle, from parting of contacts until theyclose again, is accomplished in the ultra -short time of twelve cycles or less; closecoordination with fuses is thus, possible.

The breaker is completely self -protect-ed against burnouts and damage bylightning. The trip coil can carry mini-mum trip current continuously and isshunted by a De -ion coil for protectionagainst lightning surges. Oil and contactburning are kept to a minimum by theuse of the De -ion grid arc interrupters.

z

a

WESTINGHOUSE ENGINEER3

1

is

1

PERSONALITY PROFILES-40111111111111111111111111111111111111

Kansas University has two of its alumniwriting for this issue: M. H. Hobbs and C.Lynn. Both studied electrical engineer-ing. They arrived at their present postsas managers of large and important West-inghouse departments by different routes.

191

\\ /

I Hobbs, after the dean handed him his"sheepskin" in 1913, set out for Montana.For nine years he worked for the MontanaPower Company, most of that time in con-struction and operation of power plants.Curiously enough, he helped build many aheavy masonry switchgear cell, a con-struction his later work at Westinghousewas to make obsolete. When Hobbsjoined Westinghouse in 1922, it was as adesigner of central -station switchgear. Hehas since received major promotions atalmost exactly seven-year intervals. In1929 he was placed in charge of metalcladswitchgear design; in 1937 he was madehead of the Switchboard Engineering De-partment; and then in 1944, promoted tothe managership of all Switchgear Engin-eering. Hobbs has been associated withswitching equipment for most of the bigpower installations in the country. Mostglamorous was the Manhattan Project.However, Hobbs is best known for hismany contributions to metalclad switch -gear construction, with which he was as-sociated almost from the birth of the idea.

C. Lynn saw Europe in both wars, butunder very different circumstances. Onthe first occasion it was as a Lieutenantnot long after he left Kansas Universityin 1918. The second time it was as amember of the Navy's Technical Com-mission sent into Europe on the heels ofthe retreating Nazis to scout their tech-nical accomplishments in the field inwhich Lynn is preeminent -d -c machin-ery. Lynn was in the first wave of suchobservers sent into Nazi -overrun Europe.As he tells it: "We flew to Kiel before allthe shooting was over, and found thetown so bomb -wrecked that there was noplace for us to sleep. We spent the firstnight aboard a German ship that hadgrounded on top a sunken submarine."

Between these two wars lies a career indesign and application of d -c machinery.Particularly a -c generator exciters. Wher-ever exciters are discussed in the UnitedStates, Lynn is known. He has cam-paigned aggressively for the direct -con-nected exciter, and backed it up with ma-chines that are by far the largest high-

4

speed exciters built, and which have setan almost perfect service record. But hispersonal contributions in the d -c field donot stop there. He has had much to dowith the really big d -c stuff, such as the7000 -hp steel -mill motors and with d -cmachinery for marine service.

Lynn can best be described as a dynam-

ic engineer. He walks fast, thinks fast;his ideas come fast. He believes thattime, like machines, should be used effi-ciently. He is noted for his home -moviework, entering this field in 1930 whenhome -movies were all but unknown. Hehas 18 000 feet of 16 -mm film and hassince graduated to sound -film recordingand reproduction. At present, he is build-ing his own recording system. His pic-tures combine artistry with engineeringingenuity. When he wanted to show mov-ies to friends at a cabin, the lack of elec-tric power deterred him not at all. Hemounted a d -c generator under the hoodof his automobile, belted it to the engine,and led a cable from the generator througha cabin window. Presto, 110 -volt power!

Arnold H. Redding was born in Zambo-anga, Philippine Islands, of American par-ents, and came to the States in 1925. Hereceived his mechanical -engineering train-ing at Columbia University, graduating in1938, and immediately moved to Phila-delphia, to work in the plant where West-inghouse builds steam turbines. Being ofa particularly analytical turn of mind,Redding was well adapted to develop-ment work concerned primarily with im-proving the efficiency and flow character-istics of steam -turbine blades. Then camePearl Harbor day-and with it the needfor a new type of jet -propulsion unit.Redding was given the large job of creat-ing a practical, high-pressure axial -flowcompressor for the first American -designedjet -propulsion power plant.

Much of the success of these aircraftgas turbines is due to Redding's ability todevelop satisfactory compressor designsin a short time from sketchy experiments.

He was early interested in developingways to improve gas turbines to makethem more suitable for aircraft propul-sion. Such is his major concern as SectionEngineer in charge of Preliminary Designand Flow Research.

'11611.6%.**140....1..11

W. A. Pennow's experience in the firstwar-he was seventeen months with thefamous 364th Pursuit Group which had216 replacements for an original comple-ment of 64; only 19 came back-reads likethe saga it is, adventure in the days whenaviation was a risky business with itswire -and -wood airplanes. Although helaughs off any belief that he was bornlucky, the fact that a German machine-gun bullet early in 1918 did nothing morethan put a crease in his shin bone, carom-ing off to aid its lethal brethren in smash-ing his plane to kindling, and the fact thathe literally walked away from anothercrash in a treetop, must be taken as a fairmeasure of his good fortune.

After the war he entered college and in1922 graduated from Milwaukee Schoolof Engineering with a B. S. in electricalengineering. He took his B. A. from Uni-versity of Wisconsin in 1924, and in 1926won his master's degree from Yale. Hejoined Westinghouse at the South BendLighting Division on January 1, 1930.Pennow has been associated with flood-lights, searchlights, both lamp and arc,beacons, and a wide variety of special-purpose lights for marine, aviation, andairport use. He has a superlative knowl-edge of optics, which he combines with agood understanding of materials, mechan-ics, and light sources, to produce unbe-lievable results. He was able, for example,to raise the crushing strength of runwaycontact" lights from 35 000 pounds to

200 000 pounds without increasing theirsize. Engineers know that a 28 -volt arc isnot stable; nevertheless, he made an en-tirely practical arc searchlight for opera-tion at that voltage. He was primarilyresponsible for the lighting of the giganticairport built in the woods of Newfound-land, that in the war proved invaluablein ferrying planes, men, and materials toEurope. He now offers a new solution,startling in concept, to the problem oflanding airplanes when visibility is bad.

His home is an orderly maze of elec-trical labor-saving devices. What withair conditioning and Sterilamps installedthroughout the house, he insists he hashimself so healthy that the germs giveup in disgust.

FOG PIERCERSThe worst enemy of com-mercial flying-fogged-in air-ports-may be substantiallyovercome by a new All-Weathei Approach LightingSystem. Standing beside oneof these powerful kryptonlamps is W. A. Pennow whoconceived the system andwho describes it in this issue.