Electric Railway Handbook

Second Edition

A Reference Book of Practice, Data, Formulas and Tablesfor the Use of

Operators, Engineers and Students

8
-.Flc. 19.-Luzerne car house. Philadelph~a. Pit section an:! flush sectlon of ~nspcctlonbars.

ELECTRIC
A REFERENCE BOOK OF PRACTICE DATA,
FORMIJ1,AS AND TABLES FOR THE USE
OF OI'ICRATORS, ENGINEERS
AND STUDISNTS
BY

ALBERT 8. RICHEY E. E.
Consulting Engineer, Professor of Electric Railway Engineering,
FI'orcratrr Poly&rchnic Instilute. Fellow of !he Amrrican
Inslztute of Blrctrical Engzneers

McGRAW-HILL BOOK COMPANY, INC,

NEW YORK: 370 SEVENTH AVENUE
LONDON 6 & 8 BOUVEKIE ST,E. C . 4

1924

COPYRIGHT,
1915, 1924, B Y THE
MCGRAW-HILLBOOKCOMPANY,
INC.
PRINTED I N T E E UNITED STATES O P AMERICA

THE MAPLE PRESS COMPANY, YORK,

PA.

ELECTRIC RAILWAY HANDBOOK

PUBLISHKRS

OF BOOKS

F o b

Electrical Ubrld v Eng~neenngNews-Record

Poorer v Eng~neenngand Mmn~ngJwmal-WT
Chemical and Metallurgical Engtneertne
Eloctrlc Ratlway Journal v
Coal Agc
American Mach~nlst7 Inficn~enaInfernac~onal
Electrical Merchandlslng7 BusTransporranon
Journal of Electrlclty and Western Indusrry

PREFACE TO SECONI) EDITION

During the decade since the first edit~onof Lhls I I a n d b k was
presented, there have been many changes in electric n ~ l w a ypractlce Urgent necessity for the most economical operation has
resulted in more rapid development than might have been expected
ten years ago. The most str~kingsingle feature of this develop
ment is the light weight car; this has been accompanied by more
scientific design of the car body, both as to structural features and
arrangement of entrances and ex~ts,better design of trucks, especially of the single truck, lighter and more economical motors, and
more e5cient control, brakes and safety appliances. A much
more considerable use of machinery in track construction and
maintenance is another outcome of the past several years.
Such changes in the a r t im lled the present revision, and
necessitated a radical rewriting oeeveral sections. Advantage was
taken of the opportunity to revise and add much valuable material
in other sections. I n those on Train Movement and Motors, for
instance, there is given a very greatly improved treatment of the
general subjects pertaining to the selection of a pro er motor
equipment to meet a required service. I t is felt that tge present
edition is a much better handbook as i t relates to the industry
today than was the first in its presentation of the material available ten years ago.
Costs and prices have had wide fluctuations during this period,
and have not yet reached any stable level. However, where such
figures are given herein, they are accompanied either by the basic
unit costs or the dates a t which they were obtained, so that the
user of the Handbook may convert them to the proper figures as
of the date of use
ALBERTS. RICHEY.
WORCESTFR,
MASSACHL~SETTS,
August, 1924.

FOREWARD
In 1915, Albert S. Richey, E.E., Consulting Engineer, Professor of
Electric Railway Engineering, Worcester Polytechnic Institute, and
Fellow of the American Institute of Electrical Engineers published his
treatise on the current technology of the electric railway industry. Into
this 800 page reference he attempted to present a complete description
of all facets of electric railway practice from the engineer's point of
view. Updated in 1924, the Electric Railway Handbook remains one
of the most complete compendiums of electric railway engineering and
practice.
Richey is still a useful reference for today's museum member.
Topics covered include roadbed and track, motors, trucks and control. Also such useful engineering data as brake rigging design, car
body construction. electric transmission and distribution, signals and
communication. Every chapter contains information valuable for
accurate and safe restoration of antique electric railway equipment.
Each subject is covered in detail including charts and diagrams.
The Association of Railway Museums and the Parts Committee are
proud to present this reprint of the Electric Railway Handbook t o the
museum community. Originally reprinted in 1978 by the Illinois
Railway Museum, the ARM is deeply indebted t o the Illinois Railway
Museum and especially t o Mr. James Johnson for preserving the offset printing plates for use in this edition as well as arranging for
publication. This edition retains the double page format of the 1978
reprint while incorporating double sided printing and a sewed hardcover binding. We expect that each volume will be a useful addition
for bookshelf and shop.
The Electric Railway Handbook is part of a continuing series (7f
publications by the ARM of information for the railway museum
community. We are always on the lookout for new material t o publish
and solicit contributions from our readers.
Rod Fishburn
Chairman, ARM Parts Committee
3 August 1989
Third Printing
April, 2001
This book can be ordered from;
ARM Reprint Books
33 Ashland Street
Manchester, NH 03 104-51 10
603-669-8490

PREFACE TO FIRST EDITION

I n the collection and presentation of the material for the Electric
Railway Handbook, the aim has been to get together in compact
and usable form a large amount of the information which the electric
railway engineer frequently requires, but often finds only after a n
extended search through mechanical, electrical or civil engineers'
handbooks or the files of technical journals or proceedings of engineering societies. The field of the electric railway engineer is so
broad that such information has necessarily been very much scattered, and progress has been so rapid that a great deal of valuable
material has been available only in periodical ublications, where,
unless held by a good index system, i t was soon fost to the operating
engineer. The idea of a reference book for the practical electric
railway man has been kept in mind, and no attempt has been made
to produce a text book, although it is believed that the book may
be a useful aid to the student.
I t has not seemed wise to include detailed material relating to
power plants as such, as this is a field in itself, well covered by existlng books, and requiring much more space than is available here.
Such matters as the influence of amount and character of load on
the capacity and location of power plants and substations are, of
course, given due consideration in connection with such subjects as
power requirements, energy consumption, feeder layout, etc. In
the section on Buildings, the field of the architect has been entered
practically only so far as electric railway building problems may differ from others. Other instances may be cited of intentional omissions or very brief consideration of certain matters, such as the
elementary treatment of the laying out of track curves, the design
of transmission lines, and signaling systems. I n the comparatively
rare cases where the matters in hand require more extended formulas and data than are here included, the problems are so special
and the engineer who is competent to handle them is so well supplied with information that it seems unwise to burden this volume
with the material. I n short, no attempt has been made to covcr
the entire field of electric railway engineering, ramifying, as it
does, into the special provinces of the civil, mechanical, electrical,
chemical engineer and architect. Rather, i t has been the aim to
present data on the subjects which come up in everyday electric
railway practice for constant use by the operating, constructing
or designing engineer; a book which may be used by the nontechnical manager or operator as well as by the engineer; and a
convenient reference book on electric railway practice for those
who may be specializing in other or allied lines.
Much material has been gleaned from standard hooks and publications. Due credit has been given in the text, but special mention

vii

viii

PREFACE T O FIRST EDITION

should be made of such sources as the Manual and the proceeBings

of the American Electric Railway Engineering Association,
Transactions of the American Institute of Electrical Engineers,,onof other engineering societies, the Handbook on Overhead Line rthe
struction of the National Electric Light Association, the files of
Electric Railway Jortrnal, Electric Journal, and Aera.
A special acknowledgment is also due to the following who 1lave
so kindly assisted in furnishing needed information: A. H. Armstrong, M. V. Ayres, J. H. Barnard, E. J. Blair, M. H. ~rons?'
C. L. Cadle, Charles M. Clark, C. L. Crabbs, R. E. Danforth, I! '
Davis, S. R. Dunbar, A. W. French, S. L. Foster, W. G. G~~~~
Charles Rufus Harte, W. J. Harvie, H. C. Ives, A. St. George ~
o
G. H. Kelsay, Norman Litchfield, Bruce Loomis, W. H.
Thomas B. McMath, W. S. Murray, George W. Palmer, Jr., I?'
Phillips, Clarence Renshaw, F. L. Rhodes, Ralph H. Rice, ~
a
Schreiber, Harold B. Smith, W. C. Sparks, Thomas Sproule, Cllter
Squier, R. B. Steams, H. M. Steward, Robert I. Todd.
Jackson has read both the original manuscript and theMr.
finalWa,oof
p
sheets, and has offered many valuable suggestions.
The book is now presented to its user with the request tha/ he
make any criticisms or suggestions for additions which may be h t $ ~
ful in making future editions more valuable in the field. Wxble
great care has been exercised in proof reading, it is inconceivi.ba't no errors e*s'L, anh no'i~bcaT~on
oT any h a t m a y b e b u n 8
be greatly appreciated.
ALBERTS. RICEF?''
WORCESTER,
MASSACHUSETTS,
February, 1915.

:,"

MCAIOPZ

CONTENTS

SECTION I
ROADBED
AND TRACK
Railway Location . . . . . . . . . . . . . . . . . . . .
Right of Way . . . . . . . . . . . . . . . . . . . . . .
Grading . . . . . . . . . . . . . . . . . . . . . . . .
Handling Earthwork . . . . . . . . . . . . . . . . . . .
Power Shovels . .
. . . . . . . . . . . . . . . . .
Transportation of Earth . . . . . . . . . . . . . . . . .
Culverts, Trestles and Bridges . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . .
n~lJnaSL
Ties . . . . . . . . . . . . . . . . . . . . . . . . .
Street Railway ~ o a d b e .d . . . . . . . . . . . . . . .
Rails; Sections and Composition . . . . . . . . . . . . .
Rail Joints . . . . . . . . . . . . . . . . . . . . . . .
Rail Corrugation . . . . . . . . . . . . . . . . . . . .
Track Fastenings . . . . . . . . . . . . . . . . . . . .
Track Curves . . . . . . . . . . . . . . . . . . . . . .
Special Track Work . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . .
Subway and Tunnel Sections
Electric Track Switches . . . . . . . . . . . . . . . . .
Track Machinery and Tools . . . . . . . . . . . . . . .

a
3

7
13

18
20
-2.,

27
29

31
46
53
54
55
65

72
75
76

SECTION I1
'4

CAR HOUSESAND SHOPS

Car House Track Layout . . . . . . . . . . . . . . . . .
Design of Car House Building . . . . . . . . . . . . . .
Fire Protection and Prevention . . . . . . . . . . . . . .
Details of Car House Design . . . . . . . . . . . . . . .
Repair Shop Design . . . . . . . . . . . . . . . . .

Schedules. Headways. Stops . . .

Train Resistance . . . . . . .
Track Curve Resistance . . . . .
ix

81
86
88
92

106

. . . . . . . . . . . . 123

. . . . . . . . . . . 127
. . . . . . . . . . . . 144

1:

CONTENTS

Track Grade Effect . . . . . .

Grades. Actual. Ruling. Virtual .
Acceleration . . . . . . . . .
Coefficient of Adhesion . . . . .
Weight Transfer in Acceleration
Run Curves . . . . . . . .
Traction Power Requirements .
Energy Consumption . . . . .
Regenerative Braking . . . . .
SECTION IV
RAILWAY
MOTORS
A.I.E.E. Standardization Rules on Railway Motors
Preliminary Selection of Motor Rating . . . . . . .
Alternating Current Motors . . . . . . . . . . . .
Characteristic Curves . . . . . . . . . . . . . .
Resistance of Direct Current Motors . . . . . . . .
Lists of Commercial Motors . . . . . . . . . . .
Gear Ratio Selection . . . . . . . . . . . . . . .
Commutating Poles and Field Control . . . . . . .
Ventilation . . . . . . . . . . . . . . . . . . .
Commutator . . . . . . . . . . . . . . . . . .
Brush Holders and Brushes . . . . . . . . . . . .
Armature Maintenance . . . . . . . . . . . . .
Field Coils and Maintenance . . . . . . . . . . .
Insulating Materials . . . . . . . . . . . . . . .
Gears and Pinions . . . . . . . . . . . . . . . .
Bearings and Lubrication . . . . . . . . . . . . .
Motor Suspension . . . . . . . . . . . . . . . .
SECTION V

Types of Controllers . . . . . . . .
Resistance Connections . . . . . .
Cammercial Drum Type Controllers .
Auxiliary Contactors . . . . . . . .
Resistance Calculations . . . . . .
Power Operated Control . . . . . .
Alternating Current Motor Control .
Maintenance of Control Apparatus .
SECTION V I

Trolley Wheels
Trolley Base .

. . . . . . . . . . . . . . . . . . . .

....................

365
370

CONTENTS
Trolley Maintenance .
Trolley Pressure . . .
Third Rail Collector . . . . .
Pantograph and Bow Collector
Roller Trolley . . . .
Slot Plow . . . . .

xi

. . . . . . . . . . . . .

.............

SECTION VII

Classification and Description of T ~ c k s.

Axles . . . . . . . . . . . . . .
Wheels. . . . . . . . . . . . . . .
Wheel Turning. Grinding and Mounting .
Wheel Defects and Inspection . . . . .
Standard Wheel Dimensions . . . . . .
Wheel Base and Track Curves . . . . .
Standard Journal Bearings . . . . . . .
Journal Bearing Lubrication . . . . . .
L

SECTION V I I I

Shoe Pressure. Rate and Time of Stop . . . . . . . . . . . 431

Coefficient of Friction between Shoe and Wheel . . . . . . . 432
Braking Distance . . . . . . . . . . . . . . . . . . . . 436
Brake Rigging E5ciency . . . . . . . . . . . . . . . . 4 4 0
Weight Transfer in Braking . . . . . . . . . . . . . . . 444
Brake Shoe Suspension. . . . . . . . . . . . . . . . . . 447
Brake Rigging Calculations . . . . . . . . . . . . . . . . 448
Variable Load Compensator . . . . . . . . . . . . . . 45'
Relation between Air Pressure. Piston Area and ~ e v e r a ~. e. 454
Brake Cylinders and Levers . . . . . . . . . . . . . . . 455
CIasp Brake . . . . . . . . . . . . . . . . . . . . . 459
Automatic Slack ~ d j u s t e r. . . . . . . . . . . . . . . . 460
Hand Brakes. Arrangement and Maintenance . . . . . . . . 463
Straight Air Brake . . . . . . . . . . . . . . . . . . . 465
Emergency Straight Air Brake . . . . . . . . . . . . . . 467
Automatic Air Brake . . . . . . . . . . . . . . . . . . 469
Electro-pneumatic Brake . . . . . . . . . . . . . . . . . 474
Safety Car Devices . . . . . . . . . . . . . . . . . . . 475
Brake Inspection and Maintenance . . . . . . . . . . . . 477
Air Compressors . . . . . . . . . . . . . . . . . . . . 480
Magnetic Brake . . . . . . . . . . . . . . . . . . . . . 483
Electric Braking. Regeneration. etc . . . . . . . . . . . . 485
Brake Shoes and Shoe Heads . . . . . . . . . . . . . . 486

xii

CONTENTS
SECTION I X
CARS

Present Tendencies in Car Design . . . . . . . . . . . .

Car Weight and Operating Costs . . . . . . . . . . . .
Light Weight Cars . . . . . . . . . . . . . . . . . .
Passenger Capacity . . . . . . . . . . . . . . . . .
Types of Entrances and Exits . . . . . . . . . . . . . .
Characteristics of Representative Cars . . . . . . . . .
Articulated Cars . . . . . . . . . . . . . . . . . . .
Rapid Transit Cars . . . . . . . . . . . . . . . . . .
Interurban Cars . . . . . . . . . . . . . . . . . . . .
Freight and Express Cars . . . . . . . . . . . . . . .
Aluminum in Car Body Construction . . . . . . . . . .
Painting . . . . . . . . . . . . . . . . . . . . .
Couplers and Draft Rigging . . . . . . . . . . .
Standard Dimensions of Cars . . . . . . . . . . . . .
Track Sanders . . . . . . . . . . . . . . . . . . . .
Car Heating . . . . . . . . . . . . . . . . . . . . .
Car Ventilation . . . . . . . . . . . . . . . . . . . .
Car Lighting . . . . . . . . . . . . . . . . . . . . .
SECTION X
TRANSMISSION
AND DISTRIBUTION
A.I.E.E. Standardization Rules . . . . . . . . . . . . .
Overhead Trolley Construction . . . . . . . . . . . . . .
Sag and Tension in Span Wire . . . . . . . . . . . . .
Catenary Trolley Construction . . . . . . . . . . . . . .
IIeavy Electric Traction . . . . . . . . . . . . . . . . .
Trolley Wire Specifications . . . . . . . . . . . . . . . .
Steel Poles . . . . . . . . . . . . . . . . . . . . . .
Wood Poles . . . . . . . . . . . . . . . . . . . . . . .
Concrete Poles . . . . . . . . . . . . . . . . . . .
Transmission Line Construction . . . . . . . . . . . .
National Electric Safety Code . . . . . . . . . . . . . .
Terminology, Electric Wire and Cable . . . . . . . . . .
Rubber Insulated Wire and Cable . . . . . . . . . . . .
Weatherproof Braid . . . . . . . . . . . . . . .
Cable Sheath and Armor . . . .
. . . . . . . . . . .
Paper Insulated Cable . . . . . . . . . . . . . . . . . .
Duct Conduit Construction . . . . . . . . . . . . . . .
Third Rail Construction and Material . . . . . . . . . . .
Conduit (Slot) Contact Conductor . . . . . . . . . . . . .
Track Bonding . . . . . . . . . .
. .
. . .
Electrolysis . . . . . . . . . . . . . . . . . . . . . .
Transmission Line Calculations . . . . . . . . . . . . . .
Positive Feeder System and Substation Location . . . . . .
Feeder Calculations . . . . . . . . . . . . . . . . . .
Negative Return Systems . . . . . . . . . . . . . . . .
Wire Tables . . . . . . . . . . . . . . . . . . . . . .
Wood Preservation

...................

547
550
570
574
584
586
591
598
603
609
609
611
613
619
620
622
625
630
643
645
663
699
706
714
720
727
749

CONTENTS

SECTION X I
SIGNALS
AND COMMUNICATION
Block Signal Delinitions . . . . . . . . . . . . . .
Signaling Schemes for Suburban and Interurban Service .
Signal Indications. Aspects and Clearances . . . . . .
Light Signals in Sunlight . . . . .
. . . . .
Hand Operated Signals . . . . . . . . . . .
Manual Block System . .
. . . . . . . . .
Automatic Block System . . . . . . . . . . . . . . .
Trolley Operated Signals . . . . . . . . . . . . . . . .
Track Circuits . . . . . . . . . . . . . . . . . . . .
Zinc-treated Ties . . . . . . . . . . . . . . . . . .
Interlocking . . . . . . . . . . . . . . . . . . . . .
Crossing Protection . . . . . . . . . . . . . . . . . .
Automatic Train Control . . . . . . . . . . . . . . . .
Telephone Dispatching . . . . . . . . . . . . . . . . .
Telephones . . . . . . . . . . . . . . . . . . . . . .

ELECTRIC
RAILWAY FTANDBOOK
SECTION I

ROADBEDANDTRACK
Railway Location. The following e\tracts from the section on
"l<conomics of Railway Location" of the Manual of the Amer.
Ry. h g . Assn. apply to electric as well as to steam railways.
1.ocating a railway means designing on economical plant for
handling a given tra5c. The economical plant for a given quantity
and class of t r a 5 c may not be the economical plant for a greater
or less quantity of traffic or for traffic of a different class. I t is
considered good practice to discount the future within reasonable
limits, provided the necessary funds are available.
The most general formula for the economic value of a location is:

K-E

Where R
E

=P

Annual revenue (receipts from operation);

Annual expense of operation, including depreciation
and taxes;
C = Capital invested (cost of construction);
p = Percentage of income on investment.
The following equation may be used in certain cases, especially
where the annual revenue, known or unknown, is constant:
=
=

R-(E+I)=P
1Vhere I

Amount of interest on cost of construction;

P Amount of profit (net corporate income).
When the revenue is constant the condition of the second equation is that the sum of operating expenses, plus interest on cost of
construction, shall be a minimum, and is convenient in many cases,
but does not indicate the proportion of profit to investment. Care
should be taken not to use too low a rate of interest. The ratio of
profit to investment should be considered.
In order to make an economical location of a railway, the engineer
must know or make a reasonable assumption of the amount, direction and class of trafic that the railway will be called upon to handle,
class of power and the approximate efficiency and cost of fuel that
will be used, the rate of wages that will be paid to employes, the
cost of maintenance materials, and the rate of interest considered
a proper return for additional expenditures involved in the improvement of the line for the reduction of operating expenses.
=
=

In the construction of a line where the contemplated immediate

tr~fficis small and the future traitic large, sha curvature and steep
ternpmry gradients, so situated as to lie poss%e of reduction when
justified by the tra5c. niay be ad\.antagrously introduced; such a
line will provide lor immediate r~c~uirenients
and can be improved
for future requirements a t a reasonable expense. Before deciding
upon such temporary exjztlients, compare the cost of the work
ultimately to IH: ab;indoneci with the interest sa\.cd on the extra
cost that would have heen necessary to construct a line on the final
location during that period in which the more expensive construction \vould appear uneconomical. In the construction of temporary
lines. due consideration must be given to the location of stations,
and these should not be located on portions of the line where revisions are contemplated, owing to the fact that when a station is
once established, opposition from the public may prevent its removal.
'I'he location of terminal points and ruling gradient having been
decided upon, the eKect of the minor details of location upon operating espenses may be determined approximately by comparing distance, curvature and line resistance; curvature divided into sharp
curvature, necessitating a reduction of speed for trains, and ordlnary curvature, which will again be subdivided into that increasing
line resistance in both directions and that increasing line resistance
in one direction only; and line resistance which is the sum of the
rolhng resistance lor friction resistance), plus the resistance 05
gravity overcoming difference in elevation on upgrades, plus the
resistance due to curvature, minus the energy of gravity on trains
on descending grades, from which has been subtracted the loss of
energy (or velocity head) due to the application of brakes. The
above does not take into account the resistance due to accelerating
trains. This may or may not be a considerable part of the total
resistance, depending on the rate of grades and the distance
between sto s
~ i g hoft &iy.
Crandall and Barnes in "Railroad Construction'
(1913)state that rices a t which land can be purchased for railroad
purposes will be Prom one to two times the market value in cities,
and from one to four times in the country, it being generally recognized that higher prices are justiliable, partly on account of the
greater value of the land for the special purpose and partly on account of the lower value of contiguous lands in consequence of their
prosimity to the road. The latter reason is not so likely to apply
in the case of an electric railway.
Clearing and Grubbing are sometimes lumped together with
grading in a contract. Specifications for clear~ngshould require
the removal from right of way of all trees, brush and other obstruction to grading, escept as reserved; brush and other refuse should
be burned or otherwise disposed of, but timber should be saved in
the form of saw logs, cut into cord wood, or made into ties, as specified; tops of stumps under embankments should be a t least 2 % ft.
below grade; the removal of stumps and roots (grubbing) is usually
required over all places where excavation occurs and between the
slope stakes of embankments less than 2 s ; ft. in height. Payments
for clearing and grubbing are made usually by the acre or square of

roo ft. on a side, or frzction thereof. actually cleared or grubbed,

but the removal of isolated trees or 1)uildings is often separately
contracted for.
Classification of Grading. Too minute classification gives
opportunity for clispute ant1 litigation as the percentages of the
d~fferentclasses are usu;~llyestirnatecl, not bcing susceptible to
exact mcasurement. 'I'lre ri~nericanKnil\\.ay I*;ngineering Association (hfanual, I Q I I j ~ i v e three
s
classes (solid rock, loose rock and
common excavation) for ortlinary use, but recognizes the necessity
of special classes, clearly defined, in some localities. The Manual
defines the three classes mentioned, as follows: "Solid rock shall
comprise rnck in solid Ixds or masses in its original position which
may be best removed by blasting, and Imulders or detached rock
measuring r cu. yd. or over. Loose rock slrall comprise all detached masses of rnck or stone of more than r cu. ft. and less than
1 cu. yd.. ant1 all other rock which can I)e properly removed by
pick and bar and without hhsting, although steam shovel ancl blastIng may be resorted to on favoral)le occasions in order to facilitate
the work. Common rxcavation shall comprise all other materials
of whatsoever nature that do not come under the classification of
solid rock, loose rock or such other classification as may be established before the anarcl of the contract." Common excavation
may be subdivided into loam, strong, heavy soils and stiff clay or
cemented gravel in estimating the cost of excavation with men and
teams, but a steam shovel will handle all of these and even loose
rock a t about the same cost. Hardpan, or earth which cannot be
plowed with a four-horse team, is now generally omitted from classification on account of difficulty of separating it from common
excavation.
Shrinkage must he allowea for both in figuring on the distribution
of the earth and in making the embankments, and the American
Railway Engineering Association (Manual, 1911) recommends the
following: "For green embankments, shrinkage allowance should
be made for both height and width, as follows:
For black tlirt, trestle filling.. . . . . . . . . . . . . . . . . . 15 per cent
For hlack dirt, raising under traffic. . . . . . . . . . . .
5 per cent
For clay, trestle filling.. . . . . . . . . . . . . . . . . . . . . . .
ro per cent
5 per cent
For clay. raising under traffic . . . . . . . . . . . . . .
For sand. trestle filling.. . . . . . . . . . . . . . . . . . . .
For sand. raising under traffic. . . . . . . . . .

6 per cent
5 per cent

Solid rock will swell about 7 0 per cent from cut to fill.
On account of the uncertainty in shrinkage percentages, it is
customary to measure earthwork in excavation, but the method
of mcasurement should be clearly specified.
Pay Quantities. Where the material taken from the cuts is used
in making the fills, the price paid for excavation includes hauling
and placing in the embankment, and this should be done as far as
is practicable. The cost increases, however, until the economic
lead is reached, where the cost is such that it is cheaper to waste
a t the cut and borrow a t the fill. Beyond the economic lead, the
yardage will then become the sum of the remaining cuts and fills
(plus shrinkage).

ELECTRIC RAILWAY HANDBOOK

Overhaul. When a limit is set for " f r ~ ehaul" or the maximum

haul a t the contract yardage rate, a definite additional price per
cubic yard r station for overhaul is allowed. The definition for
overhaul a g p t e d by the American Railway Engineering Association requires the limit of free haul to be h e d so as to include the
grade point and balance the cut on the one side with the fill, plus
proper shrinkage allowance, on the other. All material within
this free-haul limit is omitted from further consideration. The
distance from the center of gravity of the remaining excavation to
the center of gravity of the resulting embankment, less the free
haul, is the overhaul. When the haul is over runways laid out by
the engineer from borrow pits or to waste, the overhaul is one-half
the round trip distance, less the free haul. To calculate the overhaul, find the free-haul limit by adding the yardage each way from
the grade int, allowing shrinkage for the fills and keeping the
quantities Elanced until the free-haul limit is reached. This will
require using plusses with the corresponding yardages for the final
result. Beyond this free point in the cut, multiply the yardage of each volume by the distance of its center of gravity from the
near end of the free-haul limit, add the products and divide by the
corresponding yardage for the overhaul a t the cut end. Find how
far thls material will extend in the fill and determine the distance
of its center of gravity from the near end of the free haul limit as
above. The sum a t the two ends will give the total overhaul. For
heavy work, this method is preferable to a gra hic one as being
is used.
more accurate, unless an inconveniently large
Mass Dingnun. A simple method of determining the most
economical distribution of the material is by means of the mass diagram (or Bruckner's curve). As described by Crandall and Barnes
in "Railroad Construction," the mass diagram is constructed
by starting a t the left or zero end of the profile or other convenient point, say one past which no material will be transported, and
adding the yardage algebraically, station by station, allowing for
shrinkage on fills. The totals are plotted a t the corresponding
stations, or plusses for grade points, and a smooth curve drawn
through the points thus found as in Fig. I. Straight lines joining
the points would mean that the yardage increments were assumed
proportional to the distance
This method of construction gives to the curve the following
properties:
(a) An upward inclination of the curve from left to right indicates
excavation, a downward inclination, embankment, and maximum
and minimum ints, grade points Thus (Fig. I), A'B' indicates
embankment and B' and G', p a d e points.
excavation,
(b) Cuts and fills balance between the ints of intersection of
any horizontal line with the curve, leads E n g to the right where
the curve is above the line and to the left where it is below. Thus,
the cut from D to B will make the hll from B to E, and that from
F to G, the fill from G to E.
(c) Since the cut B D makes the fill to E, the last of the cut a t D
is moved to E, a distance D'E', while the last of the cut from F is
moved a distance F'E'. Equal distances will give the minimum

seal

B'R

MASS DIAGRAM

average lead for the fill, as this allows taking the material for
each part of the fill from the nearest point. Hence in making a
fill from adjacent cuts, or in utilizing the material from a cut in
adjacent fills, the minimum average lead is found by raising or

ELECTRIC RAILWAY HANDBOOK

lowering the horizontal dosing line until the adjacent intercepts

are equal.
(d) The area between the curve and an intersecting ho&?,ontal
line w f i be the product, lead times yardage, for the material between the two points of intersection. For since by (c) the material
,
B to D, will just make the 6U from B to E,
in the cut, e . ~ . from
the material represented by the increment dz to the ordinate a t D
must be moved from D to E, a distance LYE', giving for lead times
yardage the area dz X LYE'. Similarly for the other increments
giving the total area, lead times yardage as stated above. If area
be divided by yardage, the quotient will be the average lead for
the portion taken. The yardage for any cut or fill (not containing
artlal sections) will be the difference of the ordinates a t the ends.
3! ence to find the average lead where the material from a cut is
utilized in making an adjacent fill, or mce versa, find the area between the closing line and the curve by planimeter or by Simpson's
rule and divide by the grade point ordinate or total yardage, In
the case of side-hill work, i.e., both cut and fill between adjacent
sections and no grade point, only the difference of quantities is
used. The balanced quantities in each length are thus ignored.
The lead for this material will depend upon the longitudinal and
transverse slopes; it will generally be small and can be neglected
without serious error. To find the average lead for all the material,
including the side-hill work, add t o the area iound from the mass
diagram the roduct of the balanced yardage or side-hill work by
its lead and &vide by the total yardage.
The total yardage for the cut AB (Fig. I) is 3552, while the maximum or grade point ordinate a t B is only 3506 less by 46 cu. yd.
If D be taken a t do. I 08, D'E' will be 1 0 2 0 above the zero llne
and the area D'E'B' will be 1,868,700 CU. yd. it. Adding 46 X 10
for the side-hill product, 10 being assumed for the lead,
= 1,869,160
Total area
1 0 2 0 = 2532
Total yardage = 3552
Dividing, 1,869,160/2532 = 738 ft.
If the side-hill work be omitted,
1,868,700/2486 = 752 ft., average lead.
If the side-hill lead be omitted,
1,868,7~/2532= 738 ft.
08)]= 1142
The maximum lead, D'E', = stas. [12 50 - (I

.
I

1 L.

If the material from C to F is to be wasted uniformly along the

fill, GB, draw the horizontal C'G" and join G1'B". The area
CIG"B"F divided by the ordinate G"G1" will give the average
lead. If this material is to be wasted a t the grade point, C , divide
the area C'G"G"'F' by the yardage G"G"' for the average
lead. If wasted on the side a t a constant lead, aC', draw the horizontals, aC' and bF' and the curve ab for the yardage lead area.
TOfind the overhaul from the mass diagram, draw the horizontals
giving the greatest lead or haul and the free haul; the distance
between them will give the yardage for overhaul; draw verticals
through the ends of the free-haul line to meet the maximum haul

HANDLING EARTHWORK

line. The sum of the areas between these verticals and the mass
diagram curve will give the overhaul roduct, lead times yardage.
If the overhaul is required, divide $e overhaul product by the
overhaul yardage. Thus in Fig. I , for the cut G I which would
make the fill to K, J'K' gives the greatest haul, 11'2' the free haul,
H'H" or 2'1" the overhaul yardage and the sum of the areas,
HfH"J and III"k?, the overhaul product.
Handling Earthwork. The following figures on cost of handling
earthwork, except where otherwise credited, are taken from CrandaU and Barnes' "Railroad Construction" (1913).
Loosening. All materials excepting possibly sand require loosening for shovels or scrapers and it will often be economical to loosen
even sand, especially for shoveling. Loosening is commonly done
with picks or plows, but explosives may be used to advantage under
some conditions. A man with a pick will loosen about 15 cu. yd.
of s t 8 clay, or cemented gravel, 2 0 of strong, heavy soil, or 30 of
common loam, per 10-hour day. With wages a t 15 cents per hour,
these quantities give the following costs for labor, except foremen,
for loosening with picks:
Stiff clay or cemented gravel. . . . . . . l o o cents per cu. yd.
Strong, heavy soils.. . . . . . . . . . . . . . . . 7 . 5 cents per cu. yd.
5 . o cents per cu. yd.
Loam.. ...........................

A two-horse team with plow and driver and an extra man to hold
will loosen about 250 CU. yd. of strong, heavy soil per 10-hour day
or about 400 of ordinary soil. With very hard material requiring
a pick-pointed plow with two teams and an extra man to ride the
beam, about 180 cu. yd. can be loosened. With wages a t $3.50
per day for team and driver and $1.50 for man, these data would
give the following costs for labor, except foreman, for loosening
with plows:
Stiff clay or hardpan ............. 5 . 6 cents per cu. yd.
Strong, heavy soils.. . . . . . . . . . . . . . . . 2 . 0 cents per cu. yd.
Loam.. . . . . . . . . . . . . . . . . . . . . . . . . . I . 25 cents per cu. yd.
Shoveling. Earth may be cast short distances with shovels,
the cost being about the same for limits of 5 to 10 ft. horizontally
or 4 to 7 ft. vertically. For somewhat greater distances it may be
recast, the unit cost being about 80 per cent if a platform or other
suitable bed is provided from which to shovel. This is frequently
done in taking material from a pit too deep for one cast. Platforms are also often used in mine tunneling and might well be used
in cuts when the material is broken down from a face. Shoveling
is also required in loading wheelbarrows, carts, wagons or cars
for transporting longer distances. The material should be thoroughly loosened and often a second plowing will be more than repaid
in the reduced cost of shoveling. The quantity loaded per man
will depend upon the material, the extent to which it is loosened,
the height to which it must be raised and upon so proportioning
the gang that the shovelers will not have to wait for either material
or vehicles. For loading into an ordinary wagon or cart, 24 cu.
yd. per 10-hour day is about the upper limit for light material
or material well loosened, 18 for material a t the face of a cut

ELECTRIC RAILWAY HANDBOOK

which has been broken down onto a g+ surface for shoveling,

and 15 for rather heavy plowed earth. At $1.50 per day these
would give the following costs for labor, except foremen, of
shoveling:
Light or well loosened material.. . . . . . 6 . 2 5 cents per cu. yd.
Face broken onto good surface.. . . . 8 . 2 5 cents per cu. yd.
Heavy plowed earth.. . . . . . . . . . . 1o.00 cents per cu. yd.

In Engineering-Contracling, 1909,it is stated that experiments

on a number of pieces of work involving the handling of thousands
of yards of excavation have shown the long-handled shovel much
superior to the short for handling earth. Men can do more work
with less effort and can stand more nearly erect. I t is stated that
the long handle is used in Euro and the West, and the short
handle in the East. The rounrpointed shovel enters the earth
easier than the y a r e and is generally used exce t in shoveling
from a platform w en a square-pointed scoop shoulBbe used unless
the material has to be cast a considerable distance.
Spreading and Dressing. A bankman will s read in 6-in. layers
about 15 cu. yd. of average earth which has L e n dumped from
carts or wagons, the cost therefore being about 2 cents a cubic yard
with wages a t $1.50 er day. For a railroad bank, it is usually only
necessary to keep e!t surface smooth enough to drive over until
grade is reached, in which case half a cent per cubic yard should be
sufficient if the work is so planned that the bankman will be kept
busy, i.e., a t least 300 cu. yd. must be delivered a t the bank each
day, If the earth is hauled in carts and dum ed over the edge of
the bank, about one-quarter of a cent per c 3 i c yard should still
be allowed for keeping the dumping place in order. Earth in large
quantities can be spread by a team and road machine or Shuart
grader for from one-half to three-quarters of a cent per cubic yard.
In dressing railroad earth slopes from 125 to 150 sq. yd.
per day is a fair average on construction work. This wouCk%:
the cost about I cent per square yard. The cost per cubic yard
would, of course, depend on the lightness of the work, the amount
of material handled not being a proper criterion on account of the
time spent in smoothing and skimmmg.
Superintendence, Timekeeping and Water Carrying. These
items vary with local conditions, but half a cent per cubic yard
will usually cover all three.
Wheelbarrows. The wheelbarrow is used for excavating small
quantities of material and in cases where it is impracticabl; to use
carts or scrapers on account of the cramped situation or deep mud
in which horses could not work. I t is also valuable for moving
stony soil over short distances. Run planks or "gangways"
should be provided and usually each man should load his own
barrow. The load is about %r cu. yd. for wheeling up slopes, as is
generally necessary, but it may reach K O for level ground. A
study of the available data would indicate that the time for loading a cubic yard may be taken a t 0.6 hour for the heavy soils and
0.5 for loam, while the time for wheeling, including dumping, adjusting the plank and other unavoidable delays, may be taken a t

HANDLING EARTHWORK

)f hour per cubic yard per ~ o o

ft. for wheeling up slopes and
hour for level ground. Adding the cost of picking gives the following costs for labor except foremen, for picking and loading, with
wages a t 15 cents per hour:

.......

Stiff clay or cemented gravel..

19.0 cents per cu. yd.
Strong. heavy soil ................. 16 5 cents per cu. yd.
Loam.............................. 1 2 . 5 cents per cu. yd.

to which should be added 3% to 5 cents for each xoo ft. of lead,

de ending upon the slope.
brag Scrapers. The ordinary No. 2 drag scraper is a steel scoop
weighing about IOO Ib., which will hold from % to cu. yd., place
measurement. I t is drawn by a two-horse team which will travel
s t the rate of 2.5 miles per hour, or cover roo ft. of haul per minute
including loading and dumping, the haul making allowance for
the extra distance required for turning at the pit and dump and
beiig greater than the lead or straight line &stance. An extra
man is required to hold in loading while the driver usually dumps
his own scraper. As a man can load for two or more teams, depending upon the lead, it is advisable to work the scrapers in gangs.
The material should be well loosened so that the scoop will fill
readily. J. W. Brown, Engineering Record, gives the following
data based on the average cost of scraper work in Iowa. In making low embankments from side ditches with 6-ft. berms he makes
the following assumptions: Distance, center of ditch to center of
bank, 33 ft. Seven to ten trips per cubic yard. S i t y cubic
yards per 10-hour day good average work. Plow team generally
required to loosen the material, one plow to six scrapers. Two
horses per plow required in light soil, four or more necessary in
compact soil, average three a t $6 per day with driver and man
to hold. Field expenses, except management, and maintenance
for loosening, loading and dumping, 360 cu. yd.:
Three-horse team with driver and plowman. loosening 16.00
Three men holding scrapers. loading.. .............. 4 . 5 0
One man dumping. ............................... I . 7 5
One foreman. ....................................
a . SO
Maintenance, plows and scrapers.. ................. 0.90

Total ..................................... $15.65

This gives 4.35 cents per cubic yard. The cost of hauling a t
$3.50 per day for team and driver is 5.83 cents, giving a total
cost of 10.18 cents per cubic yard for the lead of 33 ft. For double
the lead, or 66 ft., the yardage would be reduced to 40 per scraper
per day, requiring 9 teams or $31.50 for hauling, while for a lead
of loo ft. the number of teams must be increased to 1 2 or the cost
to $42, the other expenses remaining the same except that
maintenance would be increased to $1.08 and $1.44, respectively,
giving 8.75 cents and 11.66 cents for hauling and an increase of
0.05 cent and o.15 cent for maintenance. For the above prices and
material these data give the following rule for field expenses,
except management, and maintenance for the drag scraper: To
a fixed cost of 7.2 cents per cubic yard add 9 cents per loo ft. of
lead with 20 ft. as a minimum value to allow for turning. For
stiff clay 25 to 30 per cent should be added to the above.

10

ELECTRIC RAI1,WAY HANDROOK

Fresno Scrapers. The Fresno scraper is an improved forF of

drag scraper invented in California and used mainly in the v e s t .
'l'he four-horse scoop is 5 ft. long laterally and only about 15 in.
wide from cutting edge to rear, the former giving large capacity and
the latter easy filling. Two- and three-horse sizes are also made.
The horses are driven abreast by one driver. The four-horse
scraper weighs from 2 7 5 to 310 Ib. Many Fresno loads a t the dump
have been compacted into a box with a rammer and follnd to run
from 12 to 16 cu. ft. in average earth where the lead was n6t so
great that much material was lost in transit. The editofi-of
Engineering-Contracting give the following rules for estimatl?g
cost, based on a study of a large number of data: When the dally
wage of a driver is $ 2 and that of each of the four horses is Sf a
total of $6 per Fresno per lo-hour day, the average cost, not Including plowing, trimming or superintendence, will be as below:
To a fixed cost of 4 cents per cubic yard add 2 cents per loo it. of
lead. The fixed cost includes traveling the extra distance and the
slower speed in loading, the shifting to newly plowed ground, etc.
The hauling cost is based upon a traveling speed of 2 0 0 ft. per
minute when not delayed by loading, dumping, etc., and upon
an average load of
cu. yd., with 50 ft. as the minimum lead.
If the soil is not of a kind that heaps up and drifts well in front
of the scraper, the average load will probably not exceed H cu. yd.,
particularly on long hauls. This would change the rule tc: T o
a fixed cost of 4 cents per cubic yard add 3 cents for each I@ ft.
of lead. The editors believe that the horses can be crowded 50 as
to do about the same amount of work in an 8-hour day as in a rohour day, or that the cost per cubic yard would be but slightly
affected, but experience in the East with 8-hourdays would hardly
warrant this belief. They state that the cost of plowing 0rd1narily ranges from % cent to 1.5 cents per cubic yard, foreman's
wages, from % to I cent, and dressing roadbed and slopes about
cent per square yard of surface trimmed. If the cost of plowing, dumping, maintenance and superintendence be added on ?he
basis of $10.25 for 300 CU. yd. with % cent per cubic yard for rnalntenance, as previously given, the fixed cost would be increased by
3.1 cents giving: T o a fixed cost of 7.1 cents per cubic yard add?
cents per ~ o ft.
o of lead, or to a fixed cost of 7.1 cents per cubic
yard add 3 cents per loo ft. of lead, according as the material heaps
up and drifts in front of the scra
This allows the driver $2
per day and requires him to load K ' o w n scraper, thus increasing
the cost about % cent per cubic yard per roo ft. of lead, and reducing
i t 1% cents independent of lead, as compared with the drag scraper
figures previously given.
Wheel Scrapers. With the wheel scraper, the steel scoop is
hung between two wheels with broad tires and i t can be lowered
and filled, raised or dumped, while the team is in motion. The
capacity, place measurement, ranges from about 6% cu. ft. for the
No. I to 1254 for the No. 3, while the loads actually carried are
about 0 . 2 , 0.25,0.4 cu. yd., respectively, for the three sizes. These
loads can be increased for long leads by finishing with shovels
when the material does not fill readily. The dead load varies from

HANDLING EARTHWORK
350 to 800 Ib. according to size. A snatch team is generally used
in loading all but the No. I , even then shovelers are necessary if
the box is to be filled in tough clay. In scrapcr, as in other work,
the details must be carefully studied and given attention if economical results are to be secured. Thus the plow should be set
to cut l o to 1 2 in. deep or to such a depth that the scoop will be
heaping full after traveling but a few feet. The rear portion of
the a n will not fill well with shallow plowing. The furrows
shouh be close together, and if the soil is heavy it should be plowed
twice. The bottom of the cut should be kept level so that the scoop
will lie flat and not tilted. J. W. Brown, Engineering Record,
as the result of experience in Iowa, gives the same costs, not including scra r teams and drivers, for handling 360 cu. yd. of
earth with t c No. 1 wheeler as were given for the drag scraper
on page 9, except that the go cents for maintenance is increased
to $1.40. He assumes four loads per cubic yard and about 100
ft. of lead per minute while traveling, or 60, 40 and 30 cu. yd. per
scraper per day for leads of 100, 2 0 0 and 300 it., respectively, requiring 6, ? and 1 2 wheelers for transporting the 360 cu. yd. over
the respectwe distances. He also increases mairtenance slightly
for the longer leads. These values would give for the No. I wheeler
r 10-hour day, for average soil: T o a fixed cost of 7%

For leads over 300 ft. Mr. Brown uses No. 3 wheelers with two men
to hold a scraper (requinng one extra holder) and a two-horse
snatch team to aid in loading. To move the 360 cu. yd. per day
he used 8 wheelers for a lead of 400 ft., 1 0 for 500, 1 2 for 600,
14 for 700 and 16 for 800, the limit to which he considers it advisable to go with wheel scrapers. Adding $3.50 for the snatch team,
$1.50 for the extra man to hold and. Sago a s lead increases for
extra wear to the fixed cost, will give a total of Szr.15 per day, or
5.88 cents per cubic yard for the fixed charge. Add~ngto this
$3.50 per scraper for the different leads and dividing by 360, the
number of cubic yards moved, will give the cost for hauling. The
cost per cubic yard is given quite closely by the following: T o a
tixed cost of 5% cents add 2 cents for each 100ft. of lead within the
limits of 300 and 800 ft. For a No. 2 the cost would be given
approximately by the following: T o a fixed cost of 6)r cents add
2% cents for each of 100ft. of lead within the limits of 2 0 0 to 500
f t . Good roads are essential to economy, especially with the No.
4 wheeler. These are for average conditions. For light material
j4.50 per day could be saved for the No. I wheeler by having the
drivers hold their own scrapers, while for heavy clay, a three- or
four-horse snatch team with a n extra man to hook and unhook
would be needed for the No. 3, instead of the 2-horse team. Four
horses might also be needed for the plow team. Scrapers are used
to some extent for loading cars and wagons through a latform,
over which the teams are driven and the material dumpesthrough
a n opening.
Carts. The one-horse cart, although not used so much as
formerly, is economical for short leads when shovel loading is employed and is convenient in turning and in dumping over the end

12

ELECTRIC RATI.WAY HANDBOOK

of an embankment. Five is a suitable number of shovelers for

loading, two on each side and one in the rear. They can load a
cart with 55 cu. yd in 2 ) ; minutes, while about I minute is required
for turning ~ n dumping,
d
maktng a total of, say, 4 minutes per
trip, allowing for turning into place for loading. A driver can
attend two carts, by dumping one while the other is being loaded,
for leads up to 300 ft. For greater leads he can attend to two by
taking them both together to the dump. At $I per day for a
horse and $ I 50 for a driver this would give for I W ft. of lead per
minute and 3 trips per cubic yard:
N cent per 100 ft. of lead, driver to 2 carts.
1% cents per 100 ft. of lead, driver to I cart.
For the fixed cost,
Pour mlnutes per trtp
Shovelrng. average rnntenal
Plow~ng
Dump
Foreman and maintenance
Total per c u b ~ cyard

355 cents
9
cents
I

cents

cent

>i cent

16

cents

This gives for field expenses, except managemcnt and maintenance:

T o a fixed cost of 16 cents, add >$ cent for each ~ o ft.
o of lend.
With one driver per cart, the 4 minutes will cost 5 ccnts, giving:
o of
To a fixed cost of I 7 f 5 cents, add 1% cents for each ~ o ft.
lead.
This is on the su position that the number of carts is so proportioned to that of tRe shovelers that both can be kept busy, otherwise the cost may be much greater.
Wagons. These may have an advantage over carts for long leads
on account of the larger load, but they are a t a disadvantage
in turning The slat-bottom box used for grading with a n ordinary
wagon is 3 by 9 by I ft giving a capacity of I cu. yd. of loose
earth, or about o 8 cu yd place measurement. This is a full load
for tern rary roads over soft earth and up steep pitches, as in most
railroa$work. For long hauls over hard roads, as in road improvement and city work, add~tionalside boards are much used, increasing the load to 1% to rk cu. yd , place measurement. The slatbottom box has about 3 by 4-in slats for the bottom and requires
a man a t the bank to aid the driver in dumping, which takes about
1% minutes for the o 8-cu. yd or 3 minutes for the I$<-cu yd load.
Dump wagons which can be dumped and the bottom closed again
without stopping the team are rapidly coming into use. Enough
men should be put in the pit to load a cubic yard in about 5 minutes.
This would give a t 35 cents per hour for team and driver, 3 cents
for loading time, or about 4 cents total for loading and dumping
time with slat-bottom wagons. If inconvenient to use so many
shovelers an extra wagon may be loaded whlle the team is g o n g
to the dump so that by shifting the lost time can be kept about the
same. If the earth is lowed and shoveled, with foreman and
maintenance increased to 15i cents this would give for field expenses,
except management, and maintenance for a speed of 2.5 nliles per
hour, or 2 2 0 ft. per minute when traveling:

cent for each loo ft. of lead,

T o a fued cost of 1 7 cents, add
for loads of I cu yd , place measurement.
For different loads the amount to add for cach loo ft. of lead
would be as follo\vs
Load
Load
I ~ a
Load

of o 8 cu
cu.
of r
dof I 5 CII
cu
of z

ycl
yd
yd
yd

add
add
add
add

o
o
o
o

66 ccnt pcr roo ft.

loo ft.
I V O it.
26 ccnt per 100ft.

5 3 ccnt pcr
35 cent per

Frequentlj wagons can be loaded with scrapcrs dumping through

a platform chcapcr than with shovcls. The following data are
taken from Engrrtecrrrtg-CoWr~ltng, 1907, for the cost of extavating a street of a wcstern c ~ t yto a depth of about 2 I t , using a plow
and drag scrapers. A ro by 12-lt platform was built with a floor
of 2-in plank on 6 by 6-in stringers high enough to give a clearance
of about 7 5 ft An opcnlng z i t square was lcft in thc ccntcr
through whlch the mater~al was dumped automatically by the
front end of thc scraper catching on a clcat na~ledin front of the
hole Thls a ~ d e dbut did not do away ~ 4 t hthe dumpman 'rhcre
were two inclined runways The approach was steep and soon
banked with carth; the run off was on a 15 per cent gradient. The
strect gradlent was 6 per cent and the material was hauled down
gradc a n avcrage distance of 1 2 0 ft. in d ~ r e c tline. The platform
had to bc movcd from time to time. The wagon loads averaged
2 ru yd In place and a wagon was filled by scraper loads in less
than 6 minutcs, or a t the rate of more than 2 0 CU. yd per hour.
The labor cost, c ~ c c p foreman,
t
of loading per hour was as follows:
Onc plow tcam
One man holdtnr: plow
One man holdlng scrapcr
One man at dump
Ftve .craper teams
Total for

20

cu. yd

SO 40
0 20
0 20
0 20

....

a oo
-

$3 00

or I j cents per cubic yard. The S o I wheel scraper should give

better results than the drag scraper for this lead.
Power Shovels. The standard machine for loading large quantities of material is the power shovel I t is usually operated by
steam, but sometimes by electricity. I t will handle earth, loose
rock. and even cemented material without loosening Solid rock
must, of course, be hlasted and it will often prove economical to
loosen cemented material and loose rock. The ordinary shovel
is usually mounted on standard gage trucks and provided with
propelling chains The I m m swings through an angle somewhat
over 180 deg. Jacks outside the tracks are used a t the front
corners to give a broader base for stab~litywhile a t work. Some
of the smaller shovcls are balanced on a car and can saing through
a full circle. Some of thc makers mount these on traction wheels
for highway and street wvork, cellar excavation, etc. Shovels with
short booms arc also made for tunnel and mining work. The dipper
dumps through a door a t the bottom and for light material its
capac~tyis sometimes increased by providing an c\tension or lip
in front. For hard material dippers are fitted with steel teeth.
Heavy machines with smaller dippers than for earth are usual for

14

ELECTRIC RAILWAY HANDBOOK

hard digging and for handling large boulders. Steam shovels $re
usually operated by three men, the engine man or runner, the
cranesman and the fireman The engineman controls the raising
and lowering of the di per and the swinging of the boom, while t,he
canesman regulates t i e depth of cut or " blte, ' releases the dipper
from the bank when full and dumps the load. Pitmen, usually
four to six, prepare for and move the short sections of track forward, operate the jacks and chocks in moving, etc An excellent
analysis of the cost of steam shovel work and discussion of the
factors affecting the same are given in a Handbook of Steam Shovel
Work published by the Bucyrus Company, South Milwaukee,
Wis , it being a report by the Construction Service Company, based
on records and time stud~esgiven in full for 45 Bucyrus shovels
working under different conditions as to material, depth of cuttidg,
size of cars, number of cars in a train, management, etc A formt~la
is given for cost of loading cars, in cents per cubic yard, place me3surement, for shovel work only, including plant expenses, and labor
(except superintendence) and materials of field expenses, in which
d = time in minutes to load I cu. ft , place measurement
c = capacity of one car in cubic feet, place measurement
f = time shovel is interrupted to s t one car
e = time shovel is interrupted to cRnge trains
g = time required to move shovel
L = distance of one move of shovel in feet
d l = minutes per shift less loss for accidental delays
A = area of excavated section in square feet
R = cost per cubic yard on cars
n = number of cars per train
C = shovel expense in cents per shift
From these.
Using estimated values of C and A and the average values
given below (or estimated ones) for the other terms except M and d ,
a plate of cost curves may be plotted for any value of LA showing
the relation between R and d for vanous values of M. T o estimate C, a 514,000 shovel is assumed, with the following data.
Cost per year
$653 34
8ao on

Deprectatron. 4% per cent

Interest. 6 per cent
Repalrs. when worklng one s h ~ f t

Per year of 150 worklng days. or $23 29 per worklng

day
723 29
Shovel runner
5 00
Cranesman
3 60
Flreman
2 40
One-half watchman at $50 per month
I 00
SIXp~tmena t $1 50
9 00
One team haullng coal. water. etc half day, say
2 50
Two and a half tons coal at $3.50
8 75
011. waste, etc say
ost per day. C/IOO

STEAM SHOVELS
The depreciation is found by distributing the difference between first
cost, assumed a t $150 per ton, and scrap value, $10 per ton, over
the life of the shovel, assumed as 20 years. I t is stated that
"The cost of repaus should be apportioned to the work turned out
rather than considered as a function of the age of the shovel. I t
will be higher for rock than for earth-work and higher for badly
broken rock than for well-blasted material." I n assuming 150
working days per year, allowance has been made for bad weather,
lack of continuous work, transportation of plant, etc. The actual
number will be greatly affected by local conditions. The fuel
consumption assumed for the heavy shovel used checks fairly
well with some data given by Gillette, Handbook of Cost Data.
Hi values for coal and water per 10-hour day vary from Y, ton
and 1500 gal. for a 35-ton shovel with a 1%-cu. yd. dipper to 2%
tons and 4500 gal. for a go-ton shovel with a 3-cu yd di per
The average shovel move, L, was 6 ft. A varies with the L p t h
and width of cut. For example, values of 250, 500 and 1000 sq.
ft. are used in the Handbook in computing cost curves. The
larger the volume, LA, per shovel move the less the cost per cubic
yard. The width of cut and L are fixed by the reach of the shovel.
I n increasing the depth to increase A , the danger of landslides
should be considered as also the height of the loading track ~f
cars are handled in trains alongside the shovel. On the other
had, if the &pth r e a c h a crta.in miniw.zzsm, lrarying with the
material, such that the dipper will not fill in one raise, the cost will
also be increased by increasing d, the time required to load a cubic
foot. The time d, required to load I cu. ft depends upon the
material, the depth of cutting, the shovel and the capacity of its
dip r, as well as upon the skill and cooperation of the engineman
a n r c r a n e s m a n . These men must work together perfectly or
costs will be seriously affected. The results of the tests taken
show that the average time to load I cu. yd , place measurement, is
about 10% seconds for iron ore, 1 2 for sand, 1835 for clay and earth
and 31% for rock. These were with average dipper capacities,
place measurement of I cu. yd. for rock, 1% for earth and sand,
1% for clay and 2%
for iron ore. The average ratios of place
measurement to water measurement for the dippers were 0 9 4
for iron ore, 0.56 for sand, o 61 for clay, o 53 for earth and o 43
for rock. The time, f, for spotting cars is usually zero, as it is
done while the shovel is turning and digging. This is where the
frain is alongside and moved a car length a t a time without switchrng. The capacity, c, is taken as 4 cu. yd , water measurement,
or 2 5, place measurement, for ordinary contracting work where 10
car trains of side dumping cars are most common, or n = 10. The
time, e, the shovel was interrupted to change trains averaged 4
minutes. The average time, g, required to move the shovel
averaged 8 minutes. I t depends upon the skill and cooperation of the crew and pitmen, and should be done according t o a
definite schedule. The time lost by accidental delays, to be subtracted in finding M, averaged about 7% per cent for brick clay,
8% for sand, gravel and iron ore, 17% for earth, clay and loam
from railroad borrow pits and crushed stone from quarries, and 2 0

16

ELECTRIC RAILWAY HANDBOOK

for rock cuts on railroad work; with a m x i m u m of about 40 per

cent for borrow pits of earth and 56 for rock cuts. These delays
may be due to the condition of the material or to breakdowns or to
accidents. Thus, wet clay often clogs the dipper teeth and sticks
in the bucket. Delays are most frequent, however, in handling
rock, especially if it has not been properly broken. Large pieces
have to be "chained out" or broken by mud capping or blockholing and blasting, thus seriously delaying the shovel. Small
air hammer drills can be used to advantage in block-holing as the
hole can often be drilled on the side of the stone away from the
shovel and the stone broken by light charges. Every effort should
be made, however, to properly break the rock with the original
blasts. Holes a few inches too shallow or with the bottoms not
properly loaded often leave ridges which must be drilled and blasted
hefore the track for the shovel can be laid. Delays due to breakdowns should be minimized by keeping duplicate parts liable to
breakage on the job and training the men to make re airs quickly.
The shovel should be carefully inspected each nigRt and p a n s
liable to break the next day replaced. It should be noted that
the use of average values of the various factors will give only what
may be considered standard cost curves. But estimated values,
or actual ones from time studies, are easily used, thus making i t
possible to vary conditions and determine what plan of operations
gives the best results. The Bucyrus formula shows clearly that
delays, either e, f and g, or those which reduce the value of M,
increase R, the cost of loading, and they may also increase the cost
of transportation. The best efforts of the management should
therefore be directed toward reducing these, both by the use of
proper general design and plan of operation, and by careful supervision during the progress of the work.
Steam Shovels for Light Work. For light work or for loading
into wagons or small cars a small shovel is often preferable. I t is
estimated, for example, that the Thew 13-ton full swing shovel
with traction wheels and %-cu. yd. dipper can be operated for
$13.50 per 10-hour day, as follows:

.....................................

Engineman
$5.00
F~reman. ..................................... z . o o
Two pltmen.. ................................... 3 00
Fuel a t $4 per t o n . . .............................. 2 . 0 0
Supplies and repairs.. ............................
I . 50

Total.. ................................... f 13.50

I t is claimed that i t will excavate some 35 cu. yd. per hour in ordinary soil, while if the engineman does his own finng and but one
pitman is used the cost can be reduced to about $7 and the capacity will be about 25 cu. yd. per hour. Allowing I minute
delay per load in getting the I-cu. yd. wagons into place and
a little extra time for moving, this would give 2 0 cu. yd. per hour,
or z w per day for the full crew. Dividing $13.50 by 200 would
give 6% cents per cubic yard for loosening and loading. This
does not include delays, depreciation or interest.

ELECTRIC SHOVELS

Electric Shovels. When electric power is available, a considerable saving can be made by the use of the electric shovel, in
which electric motors displace the steam engine and boiler of the
steam shovel. Many electric roads are making good use of the
electric shovel, not only in gravel pits and grading, but in construction and reconstruction work in city streets, in the latter
case usually in connection with pavement plow, pneumatic tools or
the "skull cracker" for breaking up the pavement. During construction, the Brantford & Hamilton (Ont.) Electric Railway used
a n electric shovel of I to 136 cu. yd capacity in gravel pit work,
the depth of cut being about 14 ft., gravel loaded on flat cars
of 14 cu. yd. capacity each, frequently hauling away IOO loaded
cars daily.
The Rockford & Interurban Railway reports use of a 1% cu. yd.
electric shovel in 15-ft. gravel bank, with three work train motors
and thirteen 12-yd. center dump ballast cars; four per cent grade
out of pit, average daily mileage of work trains, 120; shovel operatcd
by three men, foreman's time divided between pit and surfacing
gang; work trains cleared passenger cars, shovel not operated to
capacity because of inability to keep cars s tted for loading;
total mileage per month, 6630. The same skoovel, operating in
12-St. bank of earth with hardpan a t bottom, one work train of
three 12-yd. side dump cars, average 50 miles daily with work
train, cleared passenger cars, shovel operated by two and three
men, 9240 cu. yds. moved per month.
The Milwaukee Electric Railway & Light Company, using a n
electric shovel to excavate track trench in city streets, reports
a season's output of 25,000 cu. yds. Part of this work was on
new construction, but on that part which was reconstruction excavation was done a t night, mainly during the time of owl car
operation. The excavated material was loaded into dump cars
on the adjacent track, the work trains clearing the passenger
cars, exce t a t the outer ends of lines stub service was operated
through tge early hours of the night, and passengers transferred
around the shovel operation.
The use of the electric shovel has become quite general, and
these instances are only given as examples. One more might
be cited, however, that of The Indianapolis Traction & Terminal
Company, which uses the electric shovel to excavate, grade and
subgrade in city streets. The ditch is 9 ft. wide by 24 in. dee
and an average day's work is 340 lin. ft., maximum 410 lin.
Work less than 300 ft. is due to obstructions such as gas and water
pipes, conduits, manholes, etc., which the shovel must work over
or around. The shovel crew consists of one foreman, one shovel
operator and eight laborers, moving a n average of 226 CU.yds.
daily. T o do an average day's work requires a move every SIX or
seven minutes, each move being four feet. With deeper digging and
less frequent moves a greater number of cubic yards can be handled.
I t makes no material difference as to the character of the excavation,
that is, clay, gravel, old paving concrete and old ties. The old rail
is pulled out with a crane car and then the shovel is put in after
removing any of the old paving that has value. No current charge

8:

ELECTRIC RAILWAY HANDBOOK

is made, but i t will average about 35 amperes on 550 volts. On

clay digging in 4-ft. bank the machine averaged 420 lin. ft. ditch
9 ft wlde, with same crew I n all cases material was loaded into
6-yd standard gage Western dump cars.
Hauling with Cars and Dinkey Locomotives. On new construction this is the standard method of hauling material excavated with
steam shovel. The usual gage is 3 it. The c a n are side dump of
3 or 4 cu yd. capacity the latter weighing about 6000 Ib. The
dlnkey weighs from 8 to about 30 tons; all on drivers. This light
weight allows the use of rails of from 16 to 40 Ib. per yard. With
5-ft ties about 6 by 6 in. in section, it makes a hght track which
can be easily shifted a t cut or dump as required. The following
rental rates were quoted by a n equipment company, July, 1911.
Four-cu yd 3-it. gage Western cars.
Frrst month
Second month
Thtrd to fifth month rnclusrve
S~xthto e~ghthmonth rbcluslve..
Each month thereafter

...........
.............
.............

................
.....................

$10 50
10 ao
9 60
9 oo
8 40

Twelve-ton dmkey.
Frrst month
Second month
Thrrd to fifth month rncluslve
Sixth to erghth month rncluslve
Each month thereafter

............... $141 oo
.........
131 oa
...
. I17 00
. . . . . . . . . . . I O I oo
. . . . . . . 93 00

The rolling friction on this light track is from 20 to 30 Ib. per ton
and probably more in starting on dirty track. The dinkey can
exert a pulling force of about one-fourth of its weight. The speed
is about 5 miles per hour when loaded and 8 to 9 when empty or
on down gradients with smooth track. One dinkey is often used
with a 1%-cu. yd. shovel for leads up to ~ o o oft. With six cars
and three or four dumpmen the train can be dumped in about 2
minutes so that good results can be obtained. For longer leads, a
second engine would be required for spotting cars, with cars enough
for two trains, one dumping while the other is loading. The length
of train and weight of engine should increase with the lead when the
work is heavy enough to warrant it
Hauling with Cars and Horses. Two-foot gage I-cu. yd. capacity
cars weigh about 1000 Ib each and 1%-cu. yd. cars about 1350,
so that one horse can draw three loaded cars if favored slightly by
the gradient for the heavier cars. Fifteen to 2 4 b rails are heavy
enough, with plank or round timber ties A side track is put in
a t the cut and two trains are used the same as for dinkey engines.
For hand loading both tracks should extend into the cut and be
used alternately to save work in switching. Allowing the driver
6 minutes for dumping and I minute a t the cut, the fixed cost,
a t $ I 50 for the driver and $ I oo for the horse, would be 1 cent
per cubic yard for the heavier cars, I cu. yd place measurement,
whlle the cost per roo ft of lead would be 0.14 cent, giving for the
cost of hauling per cubic y a r d To a fixed cost of I cent add o 14
cent for each 100 f t of lead. For level or slightly rising track,
requiring the 0.8-cu. yd., place measurement, cars, this would be-

19

HAULING EARTHWORK

come: To a fixed cost of 1% cents, add o 1 8 cent for each roo ft.
of lead. This assumes a dumping trestle and no trackwork or
depreciation. The trackwork would cost much less than for the
shovel and dinkey, especially a t the cut. One cent plus 0.2 cent
per l o o ft. of lead should cover ordinary conditions. With no
dumping trestle 1% cents per cubic yard should be added for spreading and about I cent for extra shifting of track. This would give
for the cost of transportation: To a fixed cost of 4 5 cents add o 34
cent for each roo ft. of lead, or to a fixed cost of 4% cents add o 4
cent for each l o o ft. of lead, according as 3 or 2 4 cu. yd. are
hauled per train. To this must be added the cost of loosening and
loading.
Power Shovel and Standard Equipment. Standard gage flat
cars and ballast cars are used on maintenance work in widening
cuts and fills, filling old trestles, reducing gradients, distributing
ballast, etc. The flat cars are unloaded by a plow drawn by cable.
The ballast or dump cars are dumped through doors operated by
air from the brake system. For the heavy cars, i n c h e d floors
rather than tilting bodies make them self cleaning. The ca acities
of the flat and ballast cars range from l o to about 30 cu. yd: The
cost of loading with power shovel and hauling on good track with
large cars is less than with the small cars and poor track used on new
work, provided the forces can be sc adjusted as to keep all busy.
Usually, however, the work must be done subject to interruption
from traffic so that a careful study of conditions must be made in
order to estimate costs. W. Beahan, First Asst Engr.,L. S. & M S.
Ry., in a letter dated October, 1 9 1 1 , places the cost of grading
for third and fourth track, using standard equipment and a haul
not exceeding 5 miles including loading, unloading and leveling
ready to lay the ties, about as follows per cubic yard:
Borrow pits or cuts w ~ t hIS-ft.face
So 1 1
Earth cuts. 3 to 1 0 ft. deep
.
0 15
Shale cuts, all blasted
o ar
Other rock cuts all blasted and requlnng breaking up
.
0 25
by blockhol~ng .

. .....

This includes labor and supplies as follows.

Foreman per month
Laborers per hour, 10-hour days
Steam shovel crew, 8 men per 10-hour day
Tram s e ~ c elabor
,
and supplies per day

$75

00

0 15
25 0 0
28 o c

Interest, depreciation, explosives and overhead charges are not

included The repairs for shovel probably are included as the labor
was performed by the shovel crew. Mr. Beahan states that in
moving a short distance where overhead obstructions will allow,
they sometimes let the di per rest on a flat car, remove the jack
arms and haul the shovefwlth the work train; but that usually
it is best to take the shovel down even for moving a short distance,
and they estimate that it will cost $loo to take a shovel out of one
cut and put it in another, although they can occa5ionally do it for
$50. A. J. Himes, Engr. of Grade Elim., N. Y. C. &St. L. R. R. Co.,

20

ELECTRIC RAILWAY HANDBOOK

gives the cost of moving a shovel about 3%miles through the City
of Cleveland in August, 191I, as follows.
LABOR
Taklng down
Monng
Settlng up

WORK TRAINSERVICE:
Placlng cars to load parts and help~ngto
take down
Movlng shovel and bunk cars

$95 64

$10

20

8 50

18 70

The cost of shipping the above shovel from Ashtabula to Cleveland and setting up is given as follows.
Frelght shovel and three cars at $19 50
Boarhng car and tool car at $22 50
Lost ttme shovel crew
Settlng up shovel
$163 31

On new work, the cost of moving from the railroad to the site
would be in addition to that of setting up, or taking down and settlng up, as above. This may be over highways or across country.
A track is required with force sufficient to take up and move forward
ahead of the shovel. The shovel can be moved w t h its own power
if the gradients are not too stee . If too steep for adhesion, one
end of a rope can be anchored agead, the other end wound around
the driving axle and the running gear started. No general estimate
of cost of moving can be given on account of the variation in conditions. The Bucyrus Handbook of Steam Shovel Work states that
a p t o n shovel was moved r6oo ft in 8 hours by the shovel crew,
16 men, foreman and one team at a total cost of $34
- . or 2 1 2 cents
per foot
Culvert Openings. The following run-off formulas for culvert
openings are taken from the report of Committee on Drainage of the
Illinois Soclety of Engineers and Surveyors, 1913 In these formulas, M = area of watershed in square miles; A = area of watershed
in acres; Q = maximum discharge entire watershed in cubic feet per
second; q = maximum discharge per square mile in cubic feet per
econd, c = coefficient; a = area of opening required in square feet:
TALBOTTa = c i / A " . c = o 6 for flat land; c = o 85 for moderate slope, c = 1.1 for steep slope
MYERS a = 642. c = I for flat land; c = 1.6 for hilly land;
c = 4 for mountains
A
PECK.a = -. c = 4 to 6 (Missouri Pacific Ry )
WENTNORTH.
a = A%.
& Western R. R.

Applicable to conditions on Norfolk

CULVERTS AND TRESTLES

K u 1 c a r . r ~ ~For occasional floods. q =

for rare floods. q =

+ 320 + 15

46 790
MURPHY.q = --

*
+ +

M
-

170

20;

+ 370 + 7 4
1

GRAY.q = 5 8 9 ~ 1 6
FANNINGQ = CM"

c = 130 to 2oa for New England and A

palachian watersheds; c = 60 to roo for eastern mid&
states watersheds, c = 1 2 to 50 for western tributaries of
Mississippi River north of Missouri River.

Wooden Trestles. The reasons for the common use of wooden

trestles on new work are summarized in a n editorial in the Engineering News as follows
I A well-built timber trestle, while it lasts, is a very solid and
safe structure, and it lasts normally in good condition for from 5
to 10 years while much hastily built masonry gives out in I or
2 years
2. There is more time to determine accurately the size of opening
needed and thus a v o ~ dneedless washouts, besides, well-built timber
structures are less likely to wash out suddenly
3. The t ~ m eof construction is shortened materially, often a n
important consideration
4 The masonry, when a t last built, is almost certain to be better
built and of better stone. Haul then is of less importance and there
will be more time to secure good materials. The roads are few on
which any large proportion of the original masonry is in good condition after r o years. This is especially true of the smaller structures, such as cattle guards and open culverts which are often so poor
a s to shake to pieces in a few months. The lesson that the smaller
the structure, the larger and better dressed must be the stones
composing it, if i t is to be durable, is one which engineers are slow
to learn
5 I t is easier to introduce long and high 6Us afterward to be
filled by train, or re laced by masonry or iron, and thus to secure a
better alinement ancfavoid rock cutting or other objectionable work.
6. A very large part of the total cost of the line in its permanent
form is postponed for 6 to 8 years past the trying years of early
operation, thus not only saving the ~ntereston the cost of the permanent work but going far to protect the company from the
danger of early insolvency, which has proved so deadly to many
overconfident companies

ELECTRIC RAILWAY HANDBOOK

7. The only necessary disadvantages are the liability to decay

and fire. To guard against the former is a mere question of inspection. The danger from fire is a real one and every year has its
record of accidents resulting therefrom, but if the danger is real
i t is small. There are few such accidents and those mostly from
gross carelessness. I n proportion to their number, accidents from
iron structures have been vastly more numerous and more fatal,
and the same is true in substance of small masonry structures where
the great liability to washouts is a serious matter.
The reasons given above will apply to-day in sections where
timber is cheap, the country and traffic undeveloped, and the
company with scarcely sufficient means to put the road in operation.
I n improvements in alinement and gradients, or in building extensions and branch lines in a fairly well-developed country by a
prosperous, well-established company, the conditions are different
and the tendency is toward masonry structures with solid floors
so as to give a continuous ballasted roadbed. In improvements
under heavy traffic, as in track elevation or depression, grade reduction, etc., timber trestles are usually necessary to carry the track
until the permanent roadbed is completed.
Quantities of Material and Cost of Wooden Trestles. The
editors of Engineering-Contracting from a carefully prepared and
tabulated bill of materials for the Northern Pacific Railway standard wooden trestle deduced the following formulas for preliminary estimates:
M = aao + 6H for H between zo and zg

M -

++
+

a40
8H for H between 25 and 50
240
9H for H between 50 and 75
240
IOH for H between 75 and 125
where
feet B.M. in trestle, including deck, per lineal foot.
H = average height from ground to a point 3% ft. below bas4
of mil.
=

The division into groups is due to the construction of high trestles

in stories, each story be~ngabout 25 ft. The bents are 15 ft. 9 in.
centers. Each has four 1 2 by 12-in. posts, the outside posts having
a batter of 3 in. per foot and the inside posts a batter of I in. per
foot. The deck consists of six 9 by 18-in. stringers, with 8 by 8-in.
cross ties 13%-in. centers, and 5 by 8-in. guard rails, a total of 164
ft. B.M. per lineal foot of trestle. For the deck there are 40 lb.
of wrought iron, 25 lb. of cast iron, and 25 lb. of galvanized iron,
a total of 90 lb. per rooo ft. B.M. of timber, or 15 lb. per lineal foot
of trestle. For the bents and braces there are about 35 Ib. of
wrought iron and a little less than 15 lb. of cast iron per 1000ft.
B.M., a total of 50 lb. per ~ o o oor 0.05 lb. per foot B.M. of timber.
This would give in pounds:
Iron per foot of trestle = 15 0.05 (M-16~).
If piles are used under the sills as for the Sante Fe, five would be
required for the heights up to 25 ft. and six above that height.
The average penetration will be from 12 to 18 ft., depending on the
soil, requiring about to-ft. piles.
For a pile trestle, four piles per bent, bents 16-ft. centers, with
2 0 ft. per pile allowed for penetration and cut off, lineal feet of

piles per foot of trestle = (20 H)/q, where H is the average

height in feet to a point 3% ft. below base of rail.
The sawed lumber per foot of trestle = 185 ft. B.M. for trestles
under 15 ft. high.
= 2 0 0 ft. B.M.for trestles
15 to 25 ft. high.
The iron weighs 16 11,. pcr foot of trcstlc; .to per cent is wrought,
30 cast and 30 galvanized.
With bridge carpenters a t $2.50 per day it is stated that the cost
for framing and erection, including the handling of the iron, should
rarely exceed $10 per loco St. B.M., while the cost of driving the
piles is placed a t 7 cents per lineal foot of pile (not per lineal foot
of penetration). Freight or freight and cartage must be added
to the cost of the material if delivery was not included in the
purchase price.
Steel Trestles. The general type of construction is with spans
alternately about 30 and 60 ft. without much regard to height of
trestle; longitudinal bracing is placed under the 30-St. spans joining
the vertical bents in airs forming towers capable of resisting the
longitudinal forces &e to starting and stopping trains on the
track. The open deck is carried by plate girders spaced to support
the ties on the upper flanges; the bents are in vertical planes and
the posts batter about z in. per foot. Fach post rests on a masonry
pier some 4 to 5 ft. square at the top and large enough a t the base
to reduce the unit pressure on the foundation to a safe value.
Anchor bolts are set to templet before the pier is built and masonry
thus adds to stability in the case of wind pressure strong enough to
produce tension in the
The following formurthnve been given for weight per foot of
steel in terms of height of trestle:
I. C. P. Howard, Eng. News, 1906, for Cooper's E 40 loading.
Weight per foot

520 lb. for height of

20

ft.,

= 1 2 0 0 Ib. for height of 60 ft.,

= 1530 Ib. for height of 90 it.
The above values with 20 per cent added for Cooper's E 50
loading.
3. Editors, Eng.-Cont., 1907, for two 116-ton engines followed
by 3000 lb. per foot, spans 30 and 60 ft. alternating.
Weight per foot = 600
12 times height.
I t is claimed that this has been used in estimating the weight
of many viaducts of different heights and has been found to give
very close results except for heights as low as 2 0 to 25 ft.
4. H. G. Tyrell, Eng. News, 1900, for two engines weighing roo
tons each followed by 4000 Ib. per foot. Unit stresses ro,ooo and
12,000 Ib. per square inch.
Weight per foot:
9 times span,
Deck plate girder = roo
= 550 for spans of 30 and 60 ft.
Bents and bracing = 9 times height.
2.

24

ELECTRIC RAILWAY HANDBOOK

5. Electric railway trestles for 25-ton cars, or

foot.
Weight per foot:

2000

Ib. per lineal

30 5 times span,
260 for spans of 30 and 60 ft.
Bents and bracing = 6 times height.

Deck plate girder

=
=

Combination Highway and Railway Bridges. The following

specification of the Massachusetts Public Service Commission js
widely used. I t is definitely stated as for purposes of design of new
bridges or strengthening of old ones, and that it may require modifications in considering old structures and the desirability of continuing them longer in service.
For the track load these specifications use a So-ton car with
wheel spacing of 5 ft., 15 ft., 5 ft. and a total length of 40 St. over
all. For roadway and sidewalks loads of I W lb. per square foot are
used for city bridges and 80 Ib. per square foot for country bridges
roo ft. or less in length. These uniform loads are assumed to cover
the full area of the roadway and sidewalks except a width of 9 ft. a t
each track. For longer spans these uniform loads are reduced r Ib.
uare foot for every 5 St. additional length up to 2 0 0 it., and
greater lengths 80 and 60 lb. per square footrespect~vel~
are used. For suburban bridges the floor is designed for the same
loads as the city bridges, while trusses and girders are designed as
for country loads. For highway bridges in city, town or country
the specifications require provision for an alternative roadway load
of a single 20-ton auto truck on two axles 1 2 ft. on centers and
wheels at 6-ft. gage; the weight assumed to be distributed 6 tons on
one and 14 tons on the other axle; the truck assumed to occupy a floor
space of 3 2 St. long and l o St. wide, the overhang being equal a t
front and back and a t the sides. With track and uniform roadway
and sidewalk loads impact of 25 per cent is added for floor beams and
stringers, while for girders and members of main trusses the impact
used varies from 25 per cent to l o per cent according to the loaded
length producing maximum live-load stresses, except that 40 per
cent is used for counters and floor-beam hangers. With the autotruck load 50 r cent impact is used for steel members which receive
their full loa8rom one panel point only and no impact is used for
wood floor or stringers. Tension stress allowed by these specifications is 16,000 Ib. per square inch of "structural steel." Other
allowed unit stresses in general correspond with those given in other
specifications using the same tensile stress, except that in direct
compression these specifications allow only 12,000 Ib. per square
inch of steel, reduced by the Gordon formula.
Ballast. R. C. Cram lists the materials used for ballast in
order of desirability as follows: (I) Broken stone; (2) coarse slag;
(3) screened and washed gravel; (4) chats, granulated slag and disintegrated granite; (5) burnt clay; (6) bank-run gravel; (7) cinders;
(8) chert and cementing gravel; (9)sand; (10) shells, and (11) earth.
Gravel is the material most commonly used both on steam and electric railways, but in city tracks crushed stone is used to a consider-

c: a

BALLAST

'

25

ably greater extent than gravel. I n selecting ballast the h s t

consideration should be to obtain a material as free from clay and
, order to afford an opportunity for water to drain
loam a s ~ o s i b l ein
off rapi ly. Crushed stone possesses most of the qualities of an
ideal ballast, while screened and washed gravel is a fairly close
second, and is quite extensively used in city track work. However,
bank-run gravel is in general use, due to its availability and comparatively low first cost.
Crushed stone should not be smaller in size than will pass through
a %-in. ring nor larger than will pass through a 2%-in. ring.
Larger sizes have too many voids, whlle the smaller sizes wear the
ties less, are less noisy, more easily tamped and give a better
surface with less labor.
Gravel varies greatly in quality in different localities, and
bank-run gravel will vary in clay, sand and gravel as follows: Dust
and clay, o to 2 0 per cent; sand, 5 to 60 per cent, and gravel, 35 to
go per cent. Good gravel ballast should not contain more than
ro per cent clay and 2 0 per cent sand, as greater proportions of
these materials seriously interfere with drainage. I t has been
stated that ties will last from two to three years longer in washed
gravel ballast as compared with bank-run; this is attributed to the
lesser amount of water retained.
Slag, being available only in the vicinity of blast furnaces,
has not been used as ballast by electric railways generally. Coarse
slag, however, is said to be almost as durable as crushed stone
and to equal i t in many ways. I n certain cases it is stated that
dry rot attacks the ties more rapidly in slag than in stone ballast.
Granulated slag is much inferior to the coarse variety, but is still
superior in drainage qualities to cementing gravel, cinders or sand.
Chats, disintegrated granite, burnt clay, sand and shells are
not much in use for ballast on electric railways, but cinders are
in quite general use, largely through their availability as a byproduct from power stations. There is a wide divergence of opinion
as to their value, due to the claim that sulphur in them tends to
decrease the life of the ties. Cinders stand eighth in the order
of desirability for ballast materials, but when of good quality
they serve a good purpose as sub-ballast and for ballast in yards
and sidings or upon main lines having comparatively light traffic.
The depth of ballast depends upon the nature of the soil in the
subgrade, the size and spacing of ties, the strength of the rail as a
beam, the weight of the wheel load and the number of loads. The
recommendations in the A.R.E.A. Manual for 1915 are as follows:
Minimum de ths of ballast for Class A traffic, 1 2 in.; for Class B
traffic, 9 in.;
r ! t Class C traffic, 6 in. The recommended minimum
depths given in the 1916 A.E.R.E.A. way committee report are
as follows:
Ballast material
Broken s t o n e . . .................
Washed gravel . . . . . . . . . . . . . . . . .
Bank N n gravel.. . . . . . . . . . . . . . .
Cinders. . . . . . . . . . . . . . . . . . . . . . .

Main tracks
8 in.
8 in.
1 2 !n.
12 ln.

Side tracks
and yards
6 in.
6 in.
8 in.

8 in.

26

ELECTRIC RAILWAY HANDBOOK

27

TIES

Ties. The following table, from information by the Government

Forest Service, shows the proportions of different species of woods
purchased for ties by electric railways in 1915:
Kmd of wood
Per cent
Whiteoak . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 6 . 7
Chestnut.. ....................................
11.8
Cedar.. . . .
. . . . . . . . . . . . . . . . . . . . . . . 11.6
Southern p ~ n e
. . . . . . . . . . . . . . . . . . . . . . . . . 10.3
Redoak
....................................... 10.3
- .Douglas fir.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7 .i
Redwood . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3
Western yellow p ~ n e .............................
.
a.5
Cypress.. .......................................
I .a
Eastern tamarack.. ..............................
I .O
All others .......................................
3.9
Total.. . . . . . . . . . . . . . . . . . . . . . . . . . . .

100.0

White-oak timber is by far the best and most largely used. I t

is but seldom treated, as it has been found that it is hardly economical to do so, owing to its very high resistance to decay. The
average life of white oak under heavy steam-road tra5c is about
nine years, while this is extended to from twelve to fifteen years
for moderate tra5c. Bur oak, rock oak and chestnut oak will last
from six to eight years. Other species of oak, such as black and
red oak, pin or swamp oak and water oak are inferior woods and
have a life of from four to five years if untreated. The several
species of pine are used on steam roads in quantities second only to
oak, but chestnut has taken second place on electric lines and the
pines take third place. While it is a soft wood as compared with
oak, pine is quite slow in decaying and long-leaf heart pine will average seven years and has been known to last twelve years. Some
pines, if high in pitch, will check badly, but long-leaf yellow pine
is much to be preferred for bridge timbers and bridge ties, slnce
it does not warp as much as oak. Chestnut is not used much
for bridge timber because of its tendencies to split and check, but
for ties it is nearly as durable as oak, having a life untreated averaging
seven years. Cedar is a durable species of soft wood and will
resist decay for from twelve to fifteen years, but it is apt to fail
from spike driving and nail cutting; it has an average llfe of tcn
years. Hemlock is a soft wood and is very short-lived when
untreated, averaging not over four years; its use continues to a
considerable degree because of its cheapness. Tamarack and spruce
have characteristics quite similar to hemlock and cost about the
same but have an average life of from five to six years. Red and
black cypress are soft woods, largely used in the south, and they
decay rather slowly. Cypress has an average life of nine years.
In California, redwood is used to a large extent. I t is classed as a
soft wood which resists decay quite well, lasting five years untreated
and without tie plates and twelve years when used with tie plates
and treated. The foregoing information on life of ties is based on
steam road conditions.
The manner in which the tie is cut out of the tree is generally
the basis for defining its kind (Fig. 3). A tie cut from a tree from

28

ELECTRIC RAILWAY HANDBOOK

which not more than one tie can be produced from a sectionis
called a "pole" tie and it is hewed or sawed on two parallel faces.
When made from a tree of a size that two or more ties can be made
from a section by splitting, the tie is called a "split" tie. An
inferior tie, named a "slab" tie, is sometimes made from the first
or outside cut of a log. A sawed tie has the two sides and two
faces sawed. The upper or lower plane surface is called the "face."
A "quartered" tie is one made from a tree of a size to yield four
ties per section. A "slabbed" tie is one sawed on only two faces.
If the two faces are of equal width, a slabbed tie is aka a ''pole"
tie, but should the lower face be wi$ than the upper, it is called
half-round" tie. A "hewed"
tie must be hewed on a t least
two surfaces other than the
ends. Tie specifications always limit the amount of s a p
Polo Tic QuarterPdTies Split Ties wood, and if the section shows
to)
(b)
CC)
more
than
n e c i f i e d-.
- - the
.
-. s
- r amount the tie is called a
"sap" tie. If the specified
amount of sapwood is exceeded
on onlv one or two corners. but
SplitTm Slabbed Ties WaneTie does not measure more than
one inch on either corner mea(e)
(C)
(dl
sured diagonally across the tie,
FIG. 3.-Types of tles.
it is classed a%a "heart " tie.
An "all-heart" or 6i~trict-heart"tie has no sapwood. A "wane"
tie is made from a tree too small to make a pole tie, by allowing
the original surface of the tree to show on one or more corners.
When a tie has been made from a tree from which th: resin or
turpentine has been extracted before felling, it is called a tapped"
tie. Ties which do not conform to the specificationsare "cull' ties.
The thickness of ties varies from 6 to 8 in.; the width from 6
to 12 in., and the length from 8 to l o ft., the greater length being
for bridge ties and tracks over marsh land. Electric railways commonly use a size of 6 in. X 8 in. X 8 ft., while 7 in. X 9 in. or 7
in. X 8 in. X 8 ft. to 8 ft. 6 in. long are sizes being used more and more
by steam roads. Spacingshould be considered as of more importance
than size. Two advantages are obtained by decreasing the s w i n g :
the unit pressure on all track material is decreased, and the carrying
capacity of the roadbed is increased correspondingly. The minimum spacing should not be less than the width of track shovels used.
The usual spacing varies from sixteen to eighteen for a l e f t . rail or
eighteen to twenty for a 33-ft. rail, indicating a variationof from 2640
to 3200 per mile. Bridge ties are usually 8 in X 8 in. X 10 ft.
spaced from 12 to 16 in. centers.
Treatment of ties for wood preservation is without doubt an
economical measure for increasing life and thus reducing maintenance charges, inasmuch as the labor cost is now so great a proportion of the total cost of tie renewal.
Fences. The American Railway Engineering Association
Manual gives specificatipns for three classes of smooth wire fenceg
- -

--

FENCES
Preference is given to smooth wire,
but if barbed wire is used, a heavy smooth wire, or a plank a t the top
of the fence is recommended. For the three classes of smooth wire
fence, galvanized No. 9 gage is used throughout except for the top
and bottom longitudinal wires of Class I which are No. 7 gage.
The longitudinal wires are all coiled; the spacing, commencing
a t the bottom, is Class 1:3, 4, 5, 6, 7, 8, 9 and 9 in.; Class 2:5,
6%, 7%, 9, 10 and 10 in.; Class 3: 14, 14 and 14 in. The bottom
wire is to be placed above the ground 3, 6 and 1 2 in., respectively,
for the three classes. The stay wires are spaced 12, 2 2 and 2 2
in., respectively. Intermediate posts are to be 8 ft. long and not less
than 4 in. in diameter a t the small end, and end posts 9 ft. long
and 8 in. in diameter; round posts are preferred. The posts are to
be set with the large end down, the end posts 4 ft. deep and the
intermediate ones 3 ft., with spacing from 1634 to 33 ft., depending
upon the nature of the ground and the service required. Gates
are necessary a t farm or private crossings.
In Bulletin No. 144 of the Railway Engineering Association,
it is stated that the tendency to use reinforced concrete posts is
increasing and that the figures prevailing for the most popular
form now on the market are from 18 to 2 2 cents. The prevailing
cost for wood posts of the most durable kinds of timber native to
the road is from 12 to 15 cents. Several forms of metal posts
are being made, and it is claimed by a large manufacturer that
they will have a life of a t least 30 years and can be delivered a t
reasonable distances for 2; cents f.0.b. line of road.
Camp, Track, estimates that under average conditions the labor
of building a mile of barbed wire fence four strands high, posts
.
with posts 12 ft.
16 ft. apart, is about 13 days work ( ~ c r h r day);
apart, 16 days; with top board and four wires, posts 12 ft. apart,
18 days. For a fence w t h a different number of wires allow about
8 hours labor for each wire. Experienced fence men working by
contract will build about 50 per cent more fence per day than the
same number of ordinary track laborers engaged on the work only
a short time each season. The average cost for labor in erecting
2 2 miles of Page woven wire fence, posts 17 ft. apart and set 3
to 3% ft. in the ground, was 17.2 cents per rod as shown by the
report of the fence gang of a certain ralroad. The surface was
generally rough and uneven and a great many anchor posts had
to be used. The cost stated covered the labor of loading and
unloading new material, removing the old fence and piling or burning it, and the time used in moving the fence gang from point to
point.
Snow Fences. Where much trouble is experienced from drifting snow, snow fences have been extensively used to protect cuts
and other places where snow accumulates. These snow fences
may be installed permanently or may be made of the portable
type. The standard portable snow fence of the New York State
Railways is shown in Fig. 4, from the Electric Railway Journal,
1910.
Street Railway Roadbed Construction. The construction of
roadbed in highways must necessarily differ materially from con4% ft. high with wooden posts.

30

ELECTRIC RAILWAY HANDBOOK

struction on private right of way, and as the conditions as to subsoil,

highway traffic, pavement, etc., vary so widely, it is not possible to
standardize such construction to so great an extent as has been done
in the case of the private right-of-way roadbed. A careful study
is necessary with respect to the bearing value of the soil in the

FIG.4.-N Y. State Railway's portable snow fence.

street, and upon this will depend not only the character and depth
of ballast, but whether or not some form of concrete foundation is
desirable. The subgrade and ballast should be rolled, especially
where the street is known to be on made ground or where there has
been much disturbance of subgrade due to foreign subsurface con-

TYPE B

m l ofDoubk Tmk

gLns

&mf&iLnce
wryhwyytlbys-HernyTmffKond
kcMiLst6n uph
btsondc S u r r e h i m

FIG.$.-Types of track foundations for city streets.

structions and cross trenches. When it is impractical to roll,

the subgrade should a t least be thoroughly rammed, soft spots
eliminated and ballast placed under tamping which should be continued until there is no movement observed under load. When a
concrete paving base is to be constructed, the ties should again be

TRACK FOUNDATION -RAIL

31

tamped just prior to the placing of the concrete. The installation

of surface drains and good pavements is necessary in order to prevent the infiltration of surface waters which eventually causes
disintegration of all forms of foundation.
Two of the types of foundation construction submitted by the
1914 Committee on Way Matters of the A.E.K.E.A. are shown in
Fig. 5. Type B has been adopted as a "Recommended Design."
Some city track is being laid on ties, either wood or steel, embedded
wholly in concrete.
Data relative to the standard track construction on some of the
principal street railways, as compiled by the San Francisco-Oakland
Terminal Railways in 1921, are shown in the accompanying tables,
pages 32 to 35, inc.

LongiMinol Section
Cmss Section
FIG.6.-Track drain. N. Y.State Rys.

Track Drains. The importance of caring Ior surface water makes

advisable the frequent use of track surface drains in connection with
street railway tracks. The type used by the New York State
Railways (Fig. 6) is inexpensive to handle and install, takes large
amounts of water from along the rail, does not clog easily either on
the surface or under ground, and may be cleaned with a shovel.
Selection of Rail. For open track, the selection of rail for
electric railways will follow the same general principles as for
steam railways, the rincipal factors being weight of car, axle
spacjng, tie spacing, eind of ballast, frequency and character of
servlce. However, the loading conditions resulting from the typical
steam locomotive driving wheel arrangement are not found with
electric locomotives or electric motor cars, the latter having less
average wheel loads and better distribution of weight, although
the center of gravity may be lower, thus producing greater lateral
thrust, especially on curves. The stresses in the rail and the ability
of the steel to resist them must be taken into account, as well as
the proportions of the rail which must be such as to properly distribute the load to the ties. The load due to movement must be
considered as well as the weight of the car a t rest; the former is
known as the dynamic augment, and is usually taken as 0.7 times
the static load for wheel loads less than 15,ooo Ib., or as 10,ooo Ib.
for wheel loads of 15,ooo Ib. or more. The Baldwin Locomotive
Works' rule for determining weight of rail is: "Each 10 Ib. per

DATACOMTUNG
STAND- TRACK
CONSTRUC~ON
ON S o n OF THE PRINCIPAL
RAILWAYS
THROUGHOUT
THE UNITEDSTATES
(Compiled by San Francisco-Oakland Tenninal Railways, April, 1921)

City

Type of rail

Albany. ....................... la2 1b.-7" gir. 95 lb.. 7" T.

Atlanta. .............. ......... 103 1b.-7"
gi;.. 80. 70 lb.
ASCE. 80. 70 Ib.. 7" T.
Baltimore. ............
Iaa. 1o~-lb.-7" gir.. I00 lb.
ARA.
Birmingham.
......... 10s 1b.-7",@.,, 80 lb. ASCE.
Boston. ..............
132 1b.--9
pr.. rza lb..
7" gir.. loo l . b . 4 " T.
Brooklyn. ............
~ a 1b.-7"
a
@I.............
Buffalo.
,...... 124 l b . - 9 " F ..............
Chicago.
I:? 1b.--9' pr.. raa lb..
11
91 lb.. 7" T. 85 1b.

.........

Type of joint

Tie rod

Continuous.. . . . . . . . . . . . . . . Flat.. ..........

Elec. weld. Cont. plates. H"XaU. ........
Thermit
Weld fish'and bose plates to None.. .........
ra~l.
Bolt plates. . . . . . . . . . . . . . . . Brace platea..
Lorain. Thermit, arc weld.. . 2>i"X'Pfr"

Tie plate

L3
None.
Flat.
6"X 1o"shodder.

..................
... Flat.
Brace plates.
....
.........
Cast weld.. ............... " X a". ........ None.
................
Continuous ................ R a t . . .......... Flat.
......................
Lincoln. Lorain elec. weld. ... z"XHe".. ...... h" shoulder,
As~E: '............
Cincinnati. .................... I40 1b.--9"
welded channels.. ...... Rail brace.. .....
flat.
Cleveland. ............ ......... 95 1b.-7"
........... Arc
?klr:.
Rivat and weld.. .......... None.. ........
Columbus. ............ ......... raa Ib.-1"
gir.. ........... Columbus joint, arc wdd. ... Only In macadam S " X 6
under
spec, X.'
Conn. Co.. ........... ......... 9" $.. 95 1b.-7''
T,80 1b.Continuous. arc weld.. ..... %"X
......
Lundre, t i 1 t e d
5' T. roo lb. ARA.
t~ e
Dallas R y . . .......... . . . . I b - 8 0 - 7
T A
I n d p l . . . . . . . . . . . . . N o n e . . ....... R. fi.'~.CO.
Denver. ....................... 80. 65 1b. A CE ............ Cont. arc weld in paved None ........... Shoulder.
Streets.
Detroit. ....................... 91 1b.-7"T ............... Cast weld ................. %"Xxf/" ....... None.
Indianapolis. .......... . . . . . . . . 95 1b.-81%21p gir.. 91 1b.Themit.. .................% " ~ af ........ None.
4
,' -.
Kansas City.. ................. 91 1b.-7"
T.. .............
Lomin. 96"X a". ........ None.
Lo$ lngeies.. ..............
13a Ib.-7"
gir., 1x6 lb..
None .......... Brace PIS...............
. I y1 gir.

vane.

i;.
w

5
cC

"

1%".

'

Z
U

....................

105 1b.-7''
gir.. . . . . . . . . . . . Bar With base plate welded
to rail.
................... IOO lb. ARA.. 95 lb., 7" T.... Cast weld. thermit, elec. %sl'X a". .......)Shoulder.
weld.
T.L.S. Co.. 93- Cast weld, themit, arc weld SB"XIW" ....... Plat.
Minneapolis & St. Paul.. ........ 91 1b.-7"
501.
gir., 101 1b.Lorain elec. weld.. ......... %" X a". ........ Brace plates.
N. 3. Public Service.. . . . . . . . . . . 116 1b.-7"
Memphis..

Milwaukee..

New Orleans..

.............

Oakland. ..........
Omaha. ...........
Philadelphia. . . . . . .
Pittsburgh. . . . . . . . .
Portland. O n . . . . . .
St. Louis.. .........
Washington (Cap. Trac.) ....

7"
Ios C ~ 7 1 gir..
)
loo, 80 1b.
ASCE.
141. 106 1b.--9" gir..
97 1b.-7"
I41 1b.--9' @I..
13 lb.
grr
804b.-7k. 7 2 l b . 4 " T.
13? 1h.--9" Dr., 103 l b . 7"
Dr.. too lb. ARA.
80 1b.-5" T,. ............

Channels

on gir.. angles %''round

on T.
....... Cont
..............
thermit arc weld.. ..
pr,. ........... Continuous.
Nich;ls ziac h'elec. weld. ...
............. T h e m l t . . ................

..............

.......I Brace plates.

...... Shoulder.
Vane
...........
Brace plates.
None. .......... None.
"Xa%".

........ PI. Rat sh.

.......... None.

X a".

#h
:e.
Continuom.
Nichols. ................. Round and flat..

Flat.

Bolted. .................. ) L f f X ~ H " .. .. . . . None..

r-

'p
C]

?q

m
0
z

DATACOVERING
STANDARD
T R W KCONSTRUCTION
ON SOMEOF THE PRINCIPAL
RAILWAYS
THROUGHOUT
THE U ~ I T ES
DT A T E S - - ( C O ~ Z U ~ C ~ )
(Cornplled by San Francisco-Oakland Terminal Railways, 4pril, ~ g a x )

I
Albany
Atlanta
Baltimore

1I

Cross tles
Yellow pine
6" X 8" u p p ~ n creos0t.1
c
6'l

x 8"-8'

untrrated wood

Ballast

Dram t ~ l e

stone & concrete

1 Nonetile
1 Cr
Crushed granite. 6'. deep
/ 6" cr stone under
1 6" B 8" ttle

Type of pav
( ~ r n n.
~block
te
Granite asphal*
~

611

tier

Blrmlngham

nu

wuod or

I Lr~ckon concrete

I'

Sheet asphalt gran blkllners

Gran blk oving
Concrete
i r i c k on concrete
w ~ t hasph filler

1 5''

Boston
Brooklyn
Buffalo

i
ione

4" tile

gran blk & sW"XJ '

wood blk laid on concrete
i C r a n ~ t eb l h cement joints
I Sandstone blk. on 6" wnsrete
I base

I Granite blk.
Oak

Cleveland
Columbus
Conn Co
Dallas Ry
Denver
Detroit

tle
Carneg~e Concrete 7" below t ~ e

,centers
M-2s
center8
c u t stee? c a m
Camroe 5
wh.
spec wk
6' Y 8"-8' chestnut
2'

12"

concrete under r a ~ l

oak Concrete and crushed stone

5z

4" t ~ l e

1 :: ::;:

6" gravel or cr stone under Some 6" trle

tte
Lonp leaf yellow plne
Gravel
4" t ~ l e
I n t twin steel. Ore Ion8 Gtav & hroken conc pvg Xonc
leaf
base I IS" deep
6" X r 0"-6' 8" Y 1 whlte oak Concrete slab 8" under tles
6" tile

G r a n ~ t eblk.
Granite
G r a n ~ t cblk.

3
Bnck.

H a s ~ m hlock, bnck
b~tumen
I Bltullthlc and bnrL
Std concrete slab

'

54

Cran nose Mk ad) toof rails bnck between.

and

?4

Indunapolis

conc slab 3'*dry mlx

ttle
conc ballast
l n t twln steel untreated wh Cr stone 6"-8' deep. Soltd
kansas C ~ t y
conc 6" deep
oak
6" x 8 " d redwood
Los Angeles
6" cr stone under tie
4" tlle
6"-8" cr stone
Creo me whtte oak
6" ttle
Memphis
Long feaf y plne. wh oak. Cr stone 8" deep
M~lwaukee
6" tlle
steel
6" X 8"-8' sawed
6" cr stone under ties
M~nneapolls& S t Paul
None
6" stone
Long leafy plne creo .oak
N J Puhltc Servlce
Some 4" & 6"
tlle
New Orleans
Long leaf Y ptne creo I n t Conc m t h gir rail Cr Small amt
stone with T-rall
twln steel
6" X 8"-8' redwood
9'' cr stone under tic
6" tlle
Oakland
6 " ~ 8 " - 7 ' red oak t r wrth 6'' cr
stone Conc slab Small a m t 6"
Omaha
where necesrnry
creo
tlle
5" x 9"-8' yellow plne
None ~n ctty work
Phlladelphla
None
6" x 8"-8' wh oak
8" cr stone under tie
P~ttsburgh
0" tlle
6" conc slab wlth I" of 1" 4" tlle
Portland Ore
stone
S t LOUIS
Oak. Steel ties ~n spec Solld conc
~n pnvemt 6" tlle.
const
Stone ~n unpvd sts.
Washington (Cap. Trac )
Stone
6" ttle

6 " ~ 8 " - 8 ' whlte oak

6"-8"

Gran block. hnck

! SErlck
ec dressed
gran

flsn blks.
flm)s k ~ n c
Acphalt on roncrctc
Asphalt hnck & WOO ' blk
Vartous clty r-sure nents

4
r

G r ~ n t t eblk pr-(erred
Gren hlk wlrl, cemc r grout

Cran & wood (creo ) blk

xt2" asphalt on Lonc

Vlt brk blk , gran hlk
conc base
Gran blk on concrete
Block stone.
Concrete.
bnck.
, Concrete,
under protest
I

U7ood blk

Asphalt and macadam

C1
9

54
w
313
Y
h

36

ELECTRIC RAILWAY HANDBOOK

yard of ordinary rail steel, properly supported by not less than

14 ties per 3-ft. rail, is capable of supporting a safe load per wheel
of 2240 Ib." R. C. Cram has suggested that in the application
of this ~ l toe an electric railway, the dynamic augment should be
added to the static load.

)----A
FIG. 7.-A.S.C.E.

ra11section.

FIG. 8.-Am.

Ry. Eng. Asection.

rail

For use in public streets, a 7-inch girder rail usually will be

necessary to care for paving requirements satisfactorily. With
modern joints and paving blocks not over four to five inches deep,
the 9-inch girder rarely will be required, as the 7-inch rail
with proper ties spacing will carry very heavy static and dynamic

9.-Am. El. R
91-lb., ,-in.

.f(:Fg'

FIG.10.-Am.

El. R Eng. Assn.

80-lb.. 7-in. $:rail.

loads. Where shallow pavement is used, the standard T-rail may

be used with economy. The recommendations of the Am. El. Ry.
Eng. Assn. are as fon?ws:
The selection of pIaln girder rails for use in paved streets requires
the most careful consideration of the type of pavement to beinstalled
and the vehicular traftic to be susta~ned. Particular care should

RAIL SELECTION

37

be taken with respect to the use of standard section rails in pavements, as it will often be found that plam girder rails of equal
weight are much better ada ted to a greater range in types of pavement which may be selecteifor use therewith. For use where the
type of pavement will permit, as with macadam or other shallow
pavements in wide streets having moderate vehicular traffic, three
standard rails are recommended. These are shown in the table on
page 39, and weigh 8 0 ~ 9 0and loo Ib. per yard, respectively.

38

ELECTRIC RAILWAY HANDBOOK

The first is the A.S.C.E. standard 80-lb. rail, while the others correspond to the Am. Ry. Assn. old standard Series A go and loo-lb.
rails. The three types are recommended in order to provide for
varying degrees of rail service as may be desired. For use in light
service with deep pavement, a 7-in. plain girder rail, weighing 80
Ib. per yard, as shown in drawing in Fig. 10, has been adopted.

PIG. 13.-Am. El. Ry. Eng. Assn.

prn order guard rail. Amencan
Standard E7-1923.

FIG.14.-Am. El. Ry. Eng. Assn.

7-111 order guard rarl. Amencan
Standard E6-1913.

FIG.rs -Jo~nt plates for Am. El.

Ry. Eng Assn. g-rn. prder rarls.
Amencan Standard EJ-1913.

FIG. 16 -Joint plates for Am. El.

Ry. Eng. Assn. 7-111 grrder ratls.
Amencan Standard Ez-192.3.

RAIL SELECTION

This section is identical.with L.S. Co. Section No 80-335 and B.S. Co.
Section No. 277. For use in heavy service, in connection with
dee block pavement, a plain girder rail 7 in. in height, weighing
91
per yard (Fig. 9) is recommended. This section IS identical
with L.S. Co Section No 91-375 and B.S. Co. Section No. 282.
For use in heavy service in connection with deep block pavements
in the congested sections of narrow city streets where the vehicular
traffic is largely confined to the pavement area to be maintained
by the railway, which conditions exist, as a rule, only in cities of
the largest class, the committee recommends the use of the Association Standard 7 and 9-in. girder grooved rails, as shown in
Figs I I and 12. The corresponding girder guard rails are as shown
in Figs. 13 and 14, and the mrder rail joint plates in Figs. 15 and 16.

Pb

Standard T Rails

Amencan Socrety of C r n l Enmneers (Prg. 7)

2%

8 % 4 ~ % 2%

42

4%

4%

2 x 6 '%A I I H Z Z ~ H Z

42

21
21
21

37
37
37

26 2
aa 9
19 6

2 4
a 4
2 2

1 % ~236
'%a
4 x 6 4 x 6 z1f6zW
'fbr1Wa z1M4 '%4
4 Y 4 Y 2%
1 9 J ~ ~ 1 H 4 2 1 l%% ~2
4 ! 2 ~ 4 f f ~ 2%

42
42
42

ar
21
21

37
37
37

16 9
11 5
11 9

2.2
2.1
a o

'Kg

42
42
42

21
21
21

37
37
37

9 8
8 0
6 6

r 8

80
75
70

26
~ A 4a
z
41Ks41Hea196t1M~~2h4zaW4

65
60
55
so
45
40

3%
3'fis;%sif4
355 356 1%

f r 1%
' k r 1 ~ ~
2 % 4 1 W ~ 1'964

'552
36

1.9

1 7

Amencan Railway Enmneenng Assocratron (Pig 8)

I

-IOO*(6
5%

*80t 5

Prg. 8.

Amerrcan Electnc Rarlway Enmneenng Assoc~atron

1% 1

5%

316 1 % ~396
2 x 6 '(6 1'%z3%a
6
1%

Prg. 7.

1 4.1 4
1%.
1

36 23 4 b 9 7 4 8 9 29'
36 224 039.8 38 7
0137
26 7 1.96

% I.4z

40

ELECTRIC RAILWAY HANDBOOK

As to the relative merits of the plain girder rail for use in paved
streets, either can be used satisfactorily as far as car operation goes,
and pavement can be installed with plain girder rails in such a manner as to be unobjec-

by the less paving maintenance c o s t s w i t h

groove girder rails. A
groove girder rail of
prqper design may
we~gh r z z Ib. per yard

girder will cost le& per

foot of track. Based
on wear of pavement
alone, the groove girder
FIG. 17.-Rail nomenclature.
rail is preferable for use
in narrow streets of
large cities under heavy steel-tired wagon traffic which is largely
confined to the railroad pavement area. When the plain girder
(high-T) rail can be used in small cities and towns where wide streets

permit the wagon traffic to keep away from the tracks, there is an
~ncreasing tendency toward the use of standard section (low-T)
rails weighing roo lb. per yard and of a depth of about 6 in. These

RAIL SELECTION

41

rails can be purchased a t prevailing prices for standard section rails

which are much less than the prices for plain girders (high-T).
Incidentally the 6 in.-IOOIb. low-T rail will permit the use of many
types of pavement which are suitable for moderate traffic.
There seldom should be occasion for the adoption of a section
other than one of the A.R.E.A. or A.E.R.E.A. standards, except
possibly in the need for a groove girder rail to carry M.C.B.
wheels, as the A.E.R.E.A. groove girder rail provides little head
wear when used with M.C.B. wheel flanges. Fig. 18 shows a
groove rail designed by the Pacific Elec. Ry. for use with M.C.B.
flanges.
Special Rail Head Section for C w e s . Special rails for curves
have been designed mainly to give additional metal available for
wear. The Manning rail, tried some years ago on the Baltimore
& Ohio R. R., had M 2 in. of additional metal on the inner side
of the head and H a in. less on the outer side. The special 110-lb.
rail of the Lehigh Valley R. R. has the head slightly wider and
considerably deeper than that of the standard rail. The special
feature of the so-called "frictionless" rail (Fig. 19) is a very narrow
head, and its purpose is to reduce the slip of the inside wheel,
which takes place in compensating for the greater length of travel
of the wheel on the outside rail. I t is claimed that there is a
diminution of friction and resultant wear to both the outer and
inner (frictionless) rails and the wheel flanges, while the reduced
friction gives a freer and smoother passage of the wheels, with
a reduction in power required to handle a given tonnage.
Rail Renewals. There is little or no standard practice regarding
the limits of wear for the various types of girder and high T-rail.
Allowable rail wear has been fixed by various engineers a t from 35
to 50 per cent reduction in area of the head. Many engineers
believe that with the grooved-girder and tram girder rails the limit
of wear has been reached when the wheel flanges ride on the floor
of the groove or tram, but others favor the opposite extreme in permitting wheel flanges to shear the tram com letely off. The
large investment in track compared with that in A e e l s might under
some conditions dictate a temporary change in the wheel flange
contour to prolong the life of track, especially where rolled steel
wheels are used, but safety of operation must bea controlling factor.
Limits of wear fixed for T-rails are usually based on steam-road practice. Conditions governing the life of T-rail on open track with
high-speed trains, however, are not analogous to those in paved
streets where speeds are comparatively slow, and the unmodified
application of steam-road practice to electric railway tracks in paved
streets is in utter disregard of the economics of the problem a t hand
and resultsin most extravagant track maintenance methods. Generally speaking, street railways cannot afford to use relay rails, except
in temporary work, on account of the expensive construction
employed and the difficulty experienced in making repairs, although
some relay rails have been taken from heavy trunk lines and laid
in extensions of light tra5c lines, but usually such practice is economical only if undertaken a t the time of pavement renewals. In
some cases the limiting factor of the life of the rail has been railway

42

ELECTRIC RAILWAY HANDBOOK

traffic, in others vehicular traffic, in still others, corrosion. Prc*

vision has been made in modern rail sections for these and othcr
factors which might influence serviceability. The general adoption
of the girder-grooved rail as a substitute for the tram rail with the
horizontal wagon-wheel tread is certain, except under extraordinary conditions, to eliminate vehicular traffic as a life-determining
factor. The shape of the groove also makes flange riding less
hazardous, and the depth of the groove insures a liberal wear
value in the head of the rail.
The wear of electric railway rails in streets is dependent upon
operating speed, wheel load of equipment, density of car and
vehicular traffic, use of brakes, frequency of stops, grades, g e n e r ~ l
alinement with respect to curvature, designof rails, design of wheele,
upkeep of wheels, use of sand, cleanliness of streets and manufactuie
and composition of rails. When pavement renewals become necessary, or the foundations, ties or joints have failed, the question often
arises whether it is economical to use the old rails in the new workBefore the economy may be determined, it is necessary to fix some
limit of wear, but with a limit fixed, the remaining life of the rail
may be estimated by obtaining the average head reduction for the
period the rail has been in service. If corrosion indicates that it
may limit rail lire in advance of wear, the rate may be determined
in a similar manner. The condition of the foundation, ties and
joints also affects the economy of renewal. When the problem can
be decided solely on the basis of economics, it resolves itself into
one of balancing interest, depreciation and maintenance of the new
rail against the old.
The report of the Am. Ry. Eng. Assn. committee for 1 0 1 0 states
that the Hverage rail wear o n a 6-deg. curve is about 106 per cent
more than that for a straight rail, and 25 per cent for a 3-deg.
curve, the variation being approximately as the square of the degree
of curve.
Tilted Rail. For the purpose of obtaining a full line contact
between the wheel and rail, some corn anies have tilted the axis
of the rail inward a t a slope of I : 25, wit( good results. The tilting
is accomplished by the use of a tapered tie plate, or, where steel
ties are used, by bending the channels a t the proper points dependin upon the gage and type of rails.
%ad Length. The standard length is 33 ft., but 60-ft. rails are
used to a considerable extent in aved streets, the increased cost
of the latter being oliset by the g s t ccst and maintenance of the
reduced number of joints. Specifications provide for the acceptance of about 10 per cent of rails of lengths shorter than the standard by whole feet down to about 25 ft. because the cropping of the
top of the ingot may prevent the remaining portion from cutting
into full rail lengths.
Composition of Rail Metal. The quality of the metal in the
finished rail will depend upon the chemical composition, the temperature of rolling A d the-work put upon the mital during rolling.
The chemical composition, to be determined from drillings taken
from the ladle test ingot, is to be as follows in the Am. Ry. Eng.
Assn. specification for carbon steel rails, 1 9 1 5 :

43

RAIL STEEL COMPOSITION

Elements.
Besserner process
Open-hearth process
per cent
70-84 Ib.
85-100 Ib.
70-84 lb.
85-100 Ib.
Carbon
. o 40 to o 50 o 45 to o 55 o 53 to o 66 o 62 to o 75
Phosphorus, not
t o exceed
o 10
o 10
0 04
0 04
Manganese
o 80 to I l o o 80 to:[ 10 0 60 to o 90 o 60 to o 90
&llcon, not less
than
.
o 10
o 10
o 10
o 10

The Am. El. Ry. Assn. specifications for carbon steel rails, 1922,
contain the following requirements, with the statement that i t is
desired that the percentage of carbon in an entire order of rails
shall average as high as the mean between the limits specified.
Elements.
per cent
Carbon
Manganese
Phosphorus.
Blicon. m a x I

mum

.
Carbon
Manganese
Phosphorus.
Slllcon

50-69 Ib.

7-84 lb
85-100 Ib.
101-120 11).
Besemer steel
o 37 t o o 47 o 40 t o o 50 o 15 to o 55 o 45 to o 55
o 80 to I 10 o 80 to I . 10 o 80 to I . 10 0.80 to I lo
0. 10
o 10
o 10
o 10
o

20

20

20

Open-hearth steel
o 50 to o 63 o 53 to o 66 o 62 to o 7s
0 6 o t o o 9 0 0 6 0 t o 0 9 0 0.60to090
0 04
o 04
0 04
o 20
o 20
o 20

20

n.62 to o 75
060to090
o 04

20

Carbon increases hardness and tensile strength and decreases

ductility. Manganese tends to prevent the coarse crystallization
due to phosphorus and sulphur, and raises the critical temperature
to which i t is safe to heat the steel. The effect of silicon is small
although in manufacture it acts like manganese as a tlux and tends
to prevent injury by oxidation. Phosphorus tends to produce
coarse crystallization and hence lowers the finishing temperature.
I t s effect when cold, up to about 0.12 per cent, is to increase strength
and hardness, but it renders the steel brittle under shock and should
be kept a t the lowest ~racticablelimit. I n the Bessemer process
a n acld lining is used in the converter and this prevents burning
out either phosphorus or sulphur. I n the open-hearth method
a basic lining is used and this permits the conversion of the phosphorus into a slag with lime, and the sulphur with lime and manganese ore. The basic open-hearth method thus allows the use of
cheaper ores and the reduction of phosphorus and sul hur to low
limits. I t also furnishes a more uniform product, as tKe melt can
be ympled and the pmportions corrected if found desirable before
pourlng.
Alloyed Steel Rails. The consensus of opinion seems to be that
the use of special alloy steels is not generally warranted for ordinary street service when cost differences are taken into account. In
certain special cases, as in subways, on elevated roads or m other
locations where the radius of the curves is very short and where
the renewal work is expensive, some alloy steel may be warranted.
I n general, special alloy steel can only be justified when its life
is three or more times that of ordinary open-hearth steel. Girder
and high T, as well as standard sections, can be easily rolled from

ELECTRIC RAXWAY HANDBOOK

44

the following special steels possessing the general characteristics

mentioned.
Steel containing

Kind
Titanium. . . . . . .

General properties anticipated

.I

0.1 oer cent metallic1 Less searenation. cleaner metal.

hence.loGer life.
titaxiium.
Nickel.. . . . . . . 3.5 per cent nickel.. . . Increased Ide.
Increased
life by being tough
Nickel chrome.. . Containing varying percentageruckel
and and hard.
chromium.
Manganese. .
About l a per cent mannanese.

..

.I

..
Electric.. .. . . . . .
High silicon. . . . .

1
I

Made in electric furnace.

About 0.35 per cent siliCOD

.
I..,.i..
O."...,.

steel. free from impuritiles. thus adding llfe.

med life. Much used in
md.
I

Ferro-titanium Steel. The addition of the alloy ferro-titanium

to either Bessemer or open-hearth steel is to-day the most common
method of seeking to prolong the life of steel rails without materially
increasing their cost. The alloy is, as the name indicates, composed
chiefly of iron and titanium, the latter being a chemical element
found in various ores and conspicuous for having great affinity
for oxygen. The manufacture of the alloy renders obtainable in
i t various proportions of titanium, so that a 15 per cent alloy
means that nominally there is that amount of titanlum present as
against 85 r cent of iron, but these figures are not fixed and allowance must
made for the presence of other ingredients, as carbon,
alumnum, etc. The theory on which the use of ferro-titanium is
advocated is very simple, hinging upon the affinity that titanium
has for oxygen, or largely u n the effect of chemical reactions that
occur from its addition to & ) m o l t e nsteel, resulting in a cleansing,
so that the name "scavenger" has often been applied to the alloy,
There are two brands of ferro-titanium all0 available for use.
The one known as the Rossi process is most Lequently used, and
differs principally from the other or Goldschmidt alloy in being
practically free from aluminum. The latter brand contains from
4 to 6 per cent of aluminum, which is a deoxidizing element often
used in casting steel to reduce piping and blow holes. The amount
of titanium alloy to be used is a matter of some argument, but in
short it may be said that enough should be added to thoroughly
saturate the molten metal with titanium. Theoretically, then, a
trace of titanium in the finished steel may be regarded as proof that
enough has been added to the molten steel to effect complete
deoxidation. Recent practice advises the addition of one-tenth
of I per cent metallic titanium to either Bessemer or open-hearth
steel, and while this figure may be taken as a safe minimum,
there is abundant reason to think that a larger amount would
be more satisfactory, especially under some conditions. The
use of the alloy requires close attention to detail in the steel works,
and should be attended with such supervision as will insure a
strict adherence to the recogrued principles and specifications

ALLOYED STEEI. RAILS

governing its use. Success due to the use of titanium has resulted
from a denser, more uniform and homogeneous metal, as might be
expected from the nature of its duty as a scavenger. Granted that
the metal is clarified by using titanium, the possibility of increasing
the carbon content without a material loss of ductility occurs, so
that of late many tons of rails have been made containing more
carbon than formerly whose wearing qualities are regarded as
greatly increased, and with no loss of shock-resisting qualities.
Recommendations as to the carbon content as well as to the other
usual elements had best be left to the individual case until more
definite information has been obtained. When used under proper
metallurgical conditions, titanium alloy increases the wearing qualities of rails, sufficient evidence having been produced as a result
of careful measurement to justify such conviction. Whether the
increased wearing quality is economically profitable to the ur
chaser, is a ditferent question, depending upon the first cost otthe
rails, and experience with grooved girder and high T-rails for street
and interurban uses, has not been sufficient to make accurate figures
obtainable.
Manganese steel is a high-carbon open-hearth steel containing
from 12 to 15 per cent of manganese. Steel having such a proportion of manganese present, when quenched in water from a red
heat becomes exceedingly hard and tough, but remains sufficiently
ductile to resist impact. The manufacture of rails of such composition has attained considerable headway in the last few years, and
it is now possible to produce any section if ordered in fair quantity.
The process is quite simple, involving principally the addition of
a large amount of ferro-manganese to the open-hearth metal.
The treatment of the ingots in the soaking pits must be attended
with great care, and finally after rolling and sawing to length, the
rails must be immediately Immersed in a tank of water. Naturally
the long lengths become very crooked when thus cooling, so that
a greater amount of cold straightening often becomes necessary.
A very tough and ductile steel results, the most objectionable
feature of which is the impossibility of sawing or drilling it, and
therefore, careful ordering to length and punching of all holes is
necessary. While the price of manganese steel rails quite precludes their common adoption for straight track, still for curves
and in special or hard service track the benefits derived are such
as to warrant their careful consideration.
Chrome-nickel steel does not flow under compressive stresses
as does manganese steel; the flange-bearing portions of the special
trackwork do not cut out as fast and after they are cut out can be
replaced by welding. This feature alone adds considerably to the
life of special work, for in manganese work after the flange bearing
is cut out, pounding begins which loosens the special pieces and
ultimately destroys the foundation. Joint plates do not wear
into chrome-nickel steel castings as they do into manganese steel,
and switch tongues stand up better a t the heel because this metal
does not Bow under the action of the wheels. Joints can be thermit
welded and thus do away with t h ~ stroublesome feature of special
trackwork and save the cost of machining to exact dimensions

46

ELECTRIC RAILWAY HANDBOOK

necessary for proper fit of joint plates. Repairs can be made

either by electric welding or thermit welding. Chrome-nickel
steel can be machined with the ordinary machine shop equipment,
and no special grinding machine, etc., are necessary in producing
special work of this material. As successfully used for special
work in Milwaukee, the following is the specified analysis (El. Ry.
JOUT.,1923):

Carbon . . . . . . . . . . . . . .
Chrominm.. ..........
Nickel . . . . . . . . . . . . . .
Silicon... . . . . . . . . . . . .

0.45 to 0.55 Sulphur............. 0.05 max.

0 . 6 0 to 0 . 8 0 Phosphorus..
0 . 0 5 mas.
a . 50 to 3 . oo Manganese.. ......... 0 . 6 0 to 0.80
0.25 to 0 . 3 5

........

High Silicon Steel. I t may be well here to mention the possibilities in using an increased amount of silicon In common open-hearth
steel for rails. Steel with as high as four-tenths of I per cent of
silicon, or double the high limit used in this country, is reported to
have given excellent wearing results on English roads.
Rail Joints. The common angle bar splice, Fig. 20, comes in
contact with the rail along the fishing surfaces under the rail head

1
FIG.20.-Angle

bar splice.

and on the rail base. By tightening the track bolts the bars are
wedged in so that shear and bending moment due to vrheel loads
will be transmitted across the
joint giving somewhat the effect
of a continuous rail. To increase FIG. ~ I . - - I O O
per cent type joint.
the strength and stiffness, the
lower flanges may be widened opposite the joint and extended
down below the rail base as in the "IOO per cent" type, Fig. 21,
the Bonzano, Duquesne, etc. In other forms a plate is placed
under the joint as an extension of the lower flange of one of the
angle bars, or as a separate plate locked to the lower flanges of
both bars. Figs. 22 to 26, inclusive, illustrate the special forms of
joints known as the Webber, Continuous, Wolhaupter, and Atlas.
A number of rail joints were tested a t the Watertown Arsenal
under the direction of the Committee on Rail of the Am. Ry. Eng.
Assn. and the results published in Bulletin 123, 1910. The span
was 30 in. Two joints of a kind were tested, one with a center

47

R A E JOINTS

load of 32,000 Ib. on the base, the other with an equal load on the
head. The rails were then inverted and the load increased t o failure
or to the capacity of the testing machine. Below are given the

FIG. 22.-Webber

FIG. 23.-Continuous mil joint.

rail joint.

data for the strongest and stiffest joint tested for the ~ a r l b rail
.
and for the strongest and stiffest angle bar joint for the loo-lb. and
for the 8 d b . rail.

FIG. 24.-Wolhaupter

rail joint.

PIG. 25.-Atlas

rail joint--section A-A.

Rail

Joint

100-lb.
I 80-lb.
-Area full section.. ........
Moment inertra.. .........
Section modulus, normal.. .

a l b l c
9.92

Elastic limit normal. .....

Elastic limit: inverted.. ...

Ultimate. inverted

87.500

............

The roo-lb. rail, a, was a n American Society section, the loo-lb., b,

and the 80-lb., c, Dudley New York Central sections. The section modulus normal is for the head and the inverted for the base
of the rail. The elastic limit and the maximum deflection are inserted from computation for comparison with the joints, with

48

ELECTRIC RAILWAY HANDBOOK

an assumed elastic limit of 60,aao lb. and a modulus of elasticity

of 30,000,ooo lb. I t may be noted that the joints compare more
favorably with the rail in section modulus than in stiffness.
The bolt holes for the splice bars and the notches for the spikes
to prevent rail creeping are usually punched, and this is one of the
reasons why soft steel, about 0.1 per cent carbon, is common,
but one of the pairs of bars tested contained 0.63 per cent carbon.
The length varies from about 2 ft. with 4 bolts to 3 ft. with 6
bolts, although a length of 3% ft. has been used. The short bar

PIG. 26.-Atlas

rail joint-suspended

type.

is used with a suspended joint midway between two ties, the bar
reaching from tie to tie. The long bar is used with a three-tie
joint, the ends resting on the outer ties with the joint on the center
one. The suspended joint is advocated as doing away with pounding action due to a solid support under the rail and it is required
for the splice bars which extend below the rail base a t the joint;
the three-tie joint is advocated as giving better support for the
joint (which is the first part of the rail to go down under tra5c)
than the two-tie support and it is used on a number of the heavy
traffic trunk lines. With long bars a little wear of the fishing

PIG.37.-Am.

El. Ry. Assn. &hole drilling for standard section mils.

surfaces or looseness of the bolts has less effect in allowing angular

motion, a consideration often overlooked. The track bolts are
usually
to % in. with round heads and elliptical section under
the heads to prevent turning in the elongated holes of the splice
ban. Various methods are used to prevent the nuts from rattling
loose, among the most effective being the spring washer and the
Harvey grip thread. The holes in the rails are made large enough
to allow for temperature changes, the bolts acting only in tension
to hold the fishing surfaces in contact with su5cient force to transmit bending moment.
Rail joints may be laid opposite or alternate. The former is
common in the West and in Europe. I t is advocated on the
ground that since the tendency of the track is toward low joints, if

RAIL JOINTS

49

they are put opposite no side lurch is given and the train is easier
on track and passengers. The motion, however, is unpleasant
and it is hard on draft rigging and on track, the blow if both sides
go down being heavier than for only one. On curves alternate
joints hold alinement much better as there isa solid rail opposite each
joint to prevent the track from kinking due to springing the track
somewhat to fit the curvature. Again, it is more expensive on
curves due to having to cut and redrill each inner rail, rather than
let the inner joints run ahead until a rail I ft. or more shorter than
the standard length can be used.
If a concealed type of rail bond is to be used, care must be taken
in the design of the rail joint that sufficient space is allowed for the
bond, even after wear has taken place on the joint.
The Am. El. Ry. Eng. Assn. recommends joints for the 7 and *in.
girder rails as shown by Figs. 15 and 16. The Committee on Way
Matters, 1910, states that failures due to poor design of plate may
be attributed largely to insufficient fishing surface or insufficiency
of metal in the web, either of which faults may cause a line contact,
or nearly so, a t the fshing surfaces, thereby materially decreasing
the life of the joint. The life of this style joint is materially
increased, a t small additional cost to the completed track structure,
when a plate is designed of such strength and stiffness that it
cannot be bent under any load ap lied when tightening the bolts,
particulhrly when supplied with cfiannel contact surfaces of such
width as the head of the rail will rrnit.
Welded Joints. By the use o E e l d e d joints a continuous rail,
which fulfills the three required elements of a perfect joint, is obtained, and if a permanent weld could be made, thus eliminating
the joint, the life of the rail a t the joint would.be equal to that of
the rail a t any other point. The early attempts to weld rails
were made by means of electricity. Later the cast weld was introduced and became very popular, and after another period an im
proved electric weld was placed on the market. A late development
in the cast weld line is the thermit weld, which is essentially a modification of the general principles of cast welding. This has been
followed, still more recently, by the arc welded joint.
Cast Weld. This is an early type of weld, sometimes called the
Falk joint, roduced by pouring molten cast iron into a mold around
the ends o t t h e abutting rails, the latter having been cleaned by a
sand blast or some other means. In many cases it is not a weld a t
all owing to the difficulty in preheating the cast iron sufficiently
to melt the steel of the rails. It was a good mechanical joint,
however, and frequently served admirably the functions of both
bor < and joint plates. The many failures that occurred in this and
ott'er types of welds have been attributed to numerous causes.
Whether the heating of the rail was sufficient to change its properties, and thereby result in excessive wear and breaks, has occasioned
a great deal of discussion and still is a mooted question. Expansion and contraction has undoubtedly been responsible for failures
in some cases, particularly where raiis were welded in hot weather or
laid in pavements having poor binding qualities. While the cast
weld jolnt has been superseded in most cities by various forms of

50

ELECTRIC RAILWAY HANDBOOK

improved electric or thermit welded joints, it is interesting to note

that in a few cities the results obtained were such as to justify their
continuous use down to this date, among these cities being Minneapolis, Detroit, Brooklyn and Milwaukee. In Milwaukee there are
many cast weld joints which are 20 years old, and which are in such
good condition that their location in the track can hardly be distinguished on inspection. A new type of cast weld joint has been
developed in Milwaukee which promises to give much better results
than the old rectangular form of weld formerly used and in which
many failures developed. In the new joint by keeping the weld
well below the head of the rail and taking precautions to keep the
head of the rail cool by means of a water-filled strong back while
portion of the rail retains its original
pouring the weld, the ~ n n i n g
~ r o ~ e r t i eand
s does not have so great a tendency to cup under
continuous traffic. The cast weld in this altered form is considered
as the standard in Milwaukee, although a considerable percentage
of repair joints are now being made by other methods.
Thermit Weld. This is essentially a cast weld, but a s the temperature attained by the reaction of aluminum and iron oxide is far
in excess of that produced in the ordinary cast weld, the ends of
the steel rails are melted down and an
obliteration of the joint actually results. In
making the improved form, known as the
thermit full section weld, the rails are spaced
three-fourths of an inch a rt, in the case
of new rail, and a gap of t E t width is provided by cutting or sawing In the case of
old rad. Into this space and between the
heads of the rails only is placed an insert
cut from a rolled section of similar analysis
to the rail itself. (See Fig. 28.) The web
and base are thoroughly melted and amalgamated with the thermit steel, but the entire
head of the rail is not, only the outside of the
head and the lip of the rail being melted by
the thermit steel. A weld of the entire rail
section is obtained, however, when the ther,
mit steel begins to cool and contract. The
proximity of this steel to the insert has
pIG.21.-~hemit full melted u the lower part of the insert and
sectlon weld.
heated &e upper part as well as the
abutting heads of the rails to such a high
temperature that when contraction sets in, thus drawing the rails
together with tremendous force, the effect is to butt weld the irsert
permanently into place between the mil heads. The advantige
claimed for this method is to be found in the fact that the metal in
the head of the rail coming into contact with the car wheels is not
melted and its physical properties not changed. As a result, it is
said, welds made in this way have shown no tendency to cup after
six years of hard service under heavy and frequent traffic, and the
breakage of welds made in this way has been very small. Where
breaks have occurred they have usually taken place either shortly

,.,

WELDED RAIL JOINTS

61

after welding, indicatihg something wrong with the procedure, or

else they have occurred during the h t winter.
Clark Joint. Thermit welds have also been used in connection
with mechanical joints. A shoe of thermit steel is poured around
the base of a bolted or riveted rail joint, thereby welding the
base of the rail, but in no wise changing the properties of the
head or running part. The joint needs no other bond, and is mechanically good. I t has met with marked success in Baltimore
and Cleveland, where it is known as the "Clark joint."
Electric Weld. The electrically welded joint was introduced
over 2 0 years ago and has found wider application than any of the
other modern t
I t consists of heavy bars or plates from two
to three feet i n x g t h spot welded to the web of the rail by the
use ?f electric current. The process requires a heavy and expensive ant and is usually canied out by contract on a compmtive$
large scale. For this reason it is not well suited lo
~nstallationson small systems. I t is well adapted to the reclaiming
of old track as well as for new work, and has been a plied on open
T-rail construction where e ansion joints are instalid a t intervals
to provide for expansion anTcontraction. The electrically welded
joint with spelter head support was superseded by a chock head
support bar joint which was introduced in 1916. A drop-forged
steel chock is welded over the bars a t the center to provide a head
support for the abutting rails. A weld is made between the back
of the rail head and the chock as well as above the bars. Electrically welded chock joints in which the use of the bars was dispensed
with were first used in Chicago in 1918. The newer butt-weld
process consists essentially in heating the ends, or abutting surfaces,
of the rails to be joined, by having them in contact with a liquid
flux which is heated above the welding point of steel by the passage
of a comparatively smaU electric current. When the rail ends have
reached a welding heat, they are forced together by means of a
vwerful clam while a hammering of the heated portion of the
rall across the gead gives a forging finish to the operation. Failures
of the electric weld have in general been confined t o fracturing of the
rail a t the end of the welded bar, and are no doubt the result of
strains introduced into the web of the rail from the localized heating.
Such breaks have been more prevalent in old than in new rails.
On new work with rails having no bolt holes the failures have been
reduced to a minimum which is not serious.
Arc Weld Joints, in which the plates are welded to the rail by
means of an electric arc, combine features of ease and.simplicity
of a plication, low cost, and high conductance. This method of
weling has also been applied to the welding of rails to steel ties
and also to the welding of bolted and riveted joint plates to the
base of the rail. There are three rincipal forms of arc welded
joints, two of which use the metalEc electrode process, and the
third, the carbon electrode process. One of the two using metallic
electrode process welds the joint plate to the web of the rail. The
other metallic electrode method and the carbon electrode process
weld the joint plate top and bottom to the head and base of the
rail respectively.

52

ELECTRIC RAILWAY HANDBOOK

Nichols Composite Joint consists of two plates which fit the

web of the rail and are riveted snugly to it, but which do not come
in contact with the iishing surfaces. The spaces thus left around
the head and foot of the rail are filled with molten zinc, which
expands upon solidifying and enters into all of the irregularities
of the rail surface. If the rail and plates are properly cleaned before
pouring the zinc, a good electrical contact is obtained, and a joint
of high conductance is the result. The process is rather expensive, and is warranted, therefore, only on lines bearing heavy
traffic. The fact that it has been a standard of construction in
the city of Philadelphia for a number of years is an indication of
its racticabiljty.
8omproxnjse Joints, or the joint between ends of two didmilar
rail sections are often points of trouble, due to the fact that it is
di5cult to make such joint plates with good contact for both rails,
except with special facilities. I t is too often the case that the
compromise joint is overlooked until actually needed and a rough

FIG. 29.-Am.

El. Ry. Eng. Assn. uniform method for designating

compromise joints.

blacksmith job is allowed to sutlice only to cause trouble later.

Some form of welded joint between the two rail ends is most
satisfactory, and where joint welding apparatus is not available
in the field, some companies weld together short pieces of the
two rail sections, and use standard joint plates to connect these
to the similar adjoining rails. A uniform method for designating
compromise joints was ap roved by the Am. El. Ry. Eng. Assn. in
19x2, and is shown by gig. 29
The observer stands between
rails a t 0,facing joints, All readings are made from left to right,
as in reading a book, and as indicated by arrows on the figure.
Joint X would read "one combination joint connecting Sec. A
worn % in., drilled R., with Sec. B worn % in., drilled S."
Allowance for Expansion in Laying RaiL The Am. R
Aun. (1912) recommends that in laying new rail, standardl; Engo
s o n shims shaU be used, that the temperature of the rail s h r g
taken by placing a thermometer on the rail, and that the openings
between 33-ft. rails shall be as follows:
Temperature
(degrees Fahr.)

Amount of.
openlng

de.

......................................... ) / i n .
2.5..........................................
In.
50- 7 5 ..........................................
I?.
75-100 .......................................... % a In.
Over loo....................................... dose
0-

25
50

The above applies, of course, to open track. The engineen on

street railway lines, through experience, have concluded that the

R A l L JOINTS

S3

spacing of rails with open joints is unnecessary in paved streets

and ascertained by practice that if the rails were butted tightly
together the damage to the rail end, due to the impact of the wheels,
could be minimized. This damage is in direct ratio with the width
of the opening between rail ends. Therefore, in order to obtain
a perfect fit of the rail ends, most rail specifications now require
that, in addition to the sawing, the rails be milled and hished, and,
if necessary, Ned so that a true and accurate rail end may be o h
tained. Further, the Am. El. Ry. Eng. Assn. has, in view of the
desirability of obtaining a joint with no intervening space between
the abutting rail heads, recommended, in its s cifications for the
manufacture of rails, in addition to the usual L h i n g of the rails
a t the ends, that the rails may be undercut Ht to Hzof an inch.
There is now in use in one of the western cities, on T-rail construction, a special form of rail end, the rail being cut to a vertical bevel
in the ratio of I to 6. This type of cutting enables the engineers
to obtain a perfect 6 t of the abutting rail heads and with the ordinary bolted joint has been found satisfactory. I t is questionable,
however, if joints of this type would be of any particular benefit
if laid on heavy tra5c lines.
Grinding and Compromising Joints. Many engineers believe
that all joints should be ground when first installed. There is no
question that a joint should not be left if there is any variation whatsoever between the running surfaces of abutting rail ends. Once
the wheels have an opportunity of pounding the rail the receiving
side will rapidly cu out and the track will fail. In a double track,
and where the tra&c is in one direction, the receiving rail, after a
period of years, becomes badly cupped. If the joints are ordinary
lates they may be offset or re laced with step joints so that the
pottom of the cup is on a levefwith the to of the delivery rail.
The plates are then driven tight and boltefup in place and the
adjoining rails ground level. I t is im rtant that the joints be
ground back in
to the deptrof the cup, this grinding
extending sometimes for a distance of 6 it. on each side. Rail
treated in this manner and having proper foundation and ties will
ride like new track. Rail could have its life considerably extended
if, as soon as the joints begin to pound, the plates are pulled up and
the rail so ground that the butting rail ends are a true, level surface. Such practice should be resorted to instead of neglecting
the joints to such an extent that nothing is to be done except to
lose the rail or apply some expensive method in reclaiming it, such
as inserting short pieces of new rail.
Rail Corrugation. With the increased use of solid forms of track
construction (concrete base, etc.), the advent of large cars, and the
necessity of rapid rates of acceleration and braking, this form of
maintenance trouble increased greatly. While the rail manufacturer
is ready to lay the cause to modem traffic conditions, equipment
and air brakes, many engineers place it with the rolling of the rail,
and a satisfactory reason has not yet been offered. Tbe curved
head rail section and tilted rail both tend to retard the appearance
of corrugation in new track. The corrugations must be removed
by some form of grinding, either by an ordinary tile or emery block

54

ELECTRIC RAILWAY HANDBOOK

set in a frame and operated by hand, or by one of the many forms of

rail grinding machines. The grinding should be done as promptly
as possible after the corrugations manifest themselves, as once
started the trouble rapidly grows worse
Miscellaneous Track Fastening Material. The Am. El Ry.
Eng Assn. (1922) adopted as recommended specifications the following, which are identical with Am. Soc. for Testing Mat. specifications as noted:
Low carbon steel track bolts
Quenched carbon steel track bolts
uenched alloy steel track bolts .
tee1 track splkes
Steel screw splkes
Steel tle plates.
..

A.S.T.M. A76-20
A.S.T M. AS+ZI

A S.T.M. ASI-21

A S.T M. A65-18
A6o--21
A67-20

A.S.T.M.
. A.S.T.M.

Perhaps the most important im rovement in the bolted rail joint

in recent y e a n and the one whicf will do more than any one other
thing to lengthen the life of the joint is the substitution of improved
bolts for the more inferior grades. The testimony in favor of bolts
having a high elastic limit, and a great ultimate strength as against
iron or even carbon-steel bolts is apparently conclusive, and the
sl~ghtadditional expense is completely justified by the decreased
maintenance and additional life of the bonds and joints, to say
nothing of the improved riding properties of the roadbed.
Track
blts

Square nuts
Bolts per 2 0 0 lb. keg

D~am..
*engt..ln.

)4%s
46 % % I 1% h k s % 9i H I 1%
~~nI~nI~n~~n.I~nI~nI~nIrnI~nI~n~

2N

770 568 441

730 540 420
694 515 402 290
662 493 384 277

3%

633 472 369 265

606 453 355 254
581 435 341 244
558 418 329 235

2
2%

4
4!4
4%
4%

Hexagon nuts
Bolts per 200 lb. keg

750 580 436

712 551 417 302
678 525 398 289

647 502 382

194
186
619 480 366
179
593 460 352
173 I30 95 570 442 330

227 166 125 92

219 160 121 90
87
114 84

. 2x1 155 117

. .

. 790 611 458

Tie rods are usually of (about

2"

276
264
254
243

..
. .
201
.
193
185
178 136 99

234 I71 132 96

226 165 127 9 3
218 160123 90
119 87
115 84

X M") flat steel or iron,

with (about M") round threaded ends, and nuts on the inside and

outside of the web of the rail. Some companies have used rods of a
circular section throughout, the advantage being that they can be
readily made in any machine shop. However, wh~lethe latter may
answer admirably in open track or where the paving is macadam,

65

TRACK FASTENTN GS

the rincipal objection is that their diameter, which cannot well be

ma& less than )( i n , is such that entirely too wide a joint is produced in block paving. The round rod also offers practically no
resistance to turning as does the flat rod, and this element might
tend to cause the nuts to loosen. Another advantage of the flat
rod is that it lends itself more easily to installation on account of
its greater flexibility. The function of the tie rod is to prevent
s
is more severe on curves,
spreading of the rails and, since t h ~ actlon
it is very desirable to install both tie rods and either rail braces or
brace tie plates on them. I t has been found that a spacing of tie
rods on 6-ft. centers has given very satisfactory results in most
cases, although some engineers recommend a spacing as low as
5 ft. Tie rods should be located as near the head of the rail as
Track splkes.
size. In.
5%
5

X %s

X %

6, :
3
2%

%6

5 !I

XM
X H

No. In keg of a w Ib.

375
400
450
530
600
680
720
900

1000
I 190
1240
1342

kegs per ml'e.

ties a ft. centers

29%
26
23%
20
17%

%
::

9
8%

774

possible in order to be of the greatest value. The vertical play of

the rails develops bending stresses in the tie rod in addition to the
direct tensile stresses caused by the spreading action. The greatest
bending occurs just inside the inside nut, and most of the failures
of tie rods occur a t this point or somewhere between i t and the
goint where the section changes from flat to round, although some
reaks have been known to occur between the two nuts. I t is
believed, however, that in the latter instances failure has been caused
by loosening of the nuts
su6ciently to permit
play between them I t
1s because of these
bending stresses that i t
is essential that a certain ductility in the
material be provided
for in the specifications.

Notes o n Track
Curves. I n American

PIC.30 -Designatron

of track curves.

practice a curve is des~ g n a t e dby the number of degrees of angular measure subtended
a t the center of the circle by a chord of loo ft. (Fig. 30). A r-

ELECTRIC RATLWAY HANDBOOK

56

deg. curve is a curve of such a radius that.a chord of loo St. subtends
a central angle of I deg ,while an ndeg. curve has a radius such that
a chord of loo it. subtends a central angle of n deg. The radius of
a ~ d e g curve
.
is 5730 ft. and the radius of an n-deg. curve is 0
2
5

ft. The maximum allowable curvature on new construction of

trunk line steam railroads now seldom exceeds 3 deg. In street
railway work, where curves of small radius are necessary, the
curves are often designated by the radius instead of the degree
of curve.
Super-elevation of Outer Rail on Curves. On curves the outer
rail is elevated sufficiently to neutralize the centrifugal force of
the train, which varies with the degree of curvature and the speed.

TABLE
OF SUPER-ELEVATION
FOR CURVES
Degree of
rune

-1 --- 1 Speed in1 miles per1 hour 1

10

20

30

40

50

60

--

80

Super-elevation of outer rail in inches

8%

I.. .... .I.. .....

The above table (Elec. Jour., 1908)gives the super-elevation for

different speeds and curvature. I t is not considered good practice
to elevate any curve more than 736 in.
Fig. 31 shows a chart of the super-elevation of outside rail on
curves as used by the Ohio Electric Railway Company. The
chart bears the notation that the speed of cars and elevation of
outside rails should be limited to figures of the chart shown inside
the heavy lines. Thus on a 4-deg. curve the track elevation
should not exceed 451 in., nor the speed of the car more than 40
miles per hour; on a 30-deg. curve the elevation should not exceed 5 in. nor the speed of the car more than 16 miles per hour.
Formula for Super-elevation. The amount of super-elevation
of outside rail on curves is given by the following formula. This
formula is an approximate one, but the error amounts to less than

SUPERELEVATION OF CURVES

one per cent in the case of a su relevation of 9 in. on standard

gage and is proportionately 1%
other super-elevrtioor

&

c = o.oooo1165GS2n
where e = super-elevation of outer rail, inches
G = gage of track, inches
S = speed of train, miIes per hour
n = degree of curve.
Middle Ordinatec for Chord ol3OFeet

Pm. jr.--Ohio Electric Ry. chart for super-elevation of outside ra~lon

curves.

Rail Bending. In shaping a rail for use in a given curve, it is

convenient to stretch a cord from end to end of the concave side
of the head and bend the rail until the length of the middle ordinate
(distance from center of cord to side of head) is equal to the value
given in the following table for the curvature and rail length in
hand. The middle ordinate for any other length of rail or degree
of curvature within the limits of the table may be found by
proportion.

ELECTRIC RAILWAY

58

Length of rail

Degree of
33 ft.

87
8

in.
9<s
f5

'3h

HANDBOOK

1%
1Tt.
x1$ls
1~44s
3

$4

30 ft.

I 28 ft. /

26 ft.

in.

in.

in.
$t
$5

3ia

?<a

36

91s

'3t.5
I5ts

'94s

19(.
17ts

156
176

1 24 ft. I 22 ft. I

T6

131

13t6
15i
13ts

13s
I

l3t8

15ts

1%

20 ft.

in.

34s
3f e

7/t8
9t s

36
34
H

Finding Degree of Curve. I t is sometimes necessary to find the

degree of a curve on a completed track; this can be done by the use
of a tape, by the following method: Some convenient point on the
outer rail as C (Fig. 32) is selected. From the gage side of the
outer rail a t C, sight along the gage side of the inner rail on the line

FIG. 32.-Finding

degree of curve.

CB. The point A , where the line C B produced cuts the gage side
of the outer rail again, is thus determined, and the distance AC
is measured. The length AC is now compared with the values
given in the table and the corresponding degrees of curve in the
table will be the value sought. The values given in the table for
AC are the long chords corresponding to a constant middle ordinate
(equal to the gage of the track which, in this case, is 4 ft. 8.5 in.).
The distance measured may vary several feet from that found in
the table due to the inaccuracy of the alinement of the track, but
since curves are nearly always made in even degrees, except in special cases where local conditions make i t necessary, the degree of
curve can almost alwarj be determined by the above method in a
very short time.

TRACK CURVES

6#

TABLEOF CHORDLENGTHSFOR OUTERRAILCO~~ESPONDING

TO
MIDDLEORDINATEOF 4 IT. 8~ IN.
Deg. of curve r0 z0 3 O
Length of AC 463 328 268
In It.

so

6"
4O
a32 208 190

Degree of Curve to Connect Two Tangents. (See Fig. 33.)

The radius of a circular curve required to connect two tangents
may be found by dividing the apex distavce (P.C.t in Fig. 33) by

.1

:
\

FIG. 34.-Laying

,-

out curve.

the tangent of one-half the

total deflection angle. Dividing 5730 by this radius
0
gives the degree of the curve.
PIG. ~3.-Curve to connect tangents. These rules may be expressed
as follows:
apex distance
radius of curve = tan. $5 total deflection angle
5730 X tan. $5 total deflection angle
degree of curve = apex distance
Laying Out a Circular Curve by Chord Deflection Distance.
(See Fig. 34, also Chord Deflection Distance, below.) Assuming
a chord length (roo ft. is generally used), lay off P.C.d on the tangent produced through P.C., such that
P.C.d = V
/
wlength)' - ( $ 6 chord deflection distance)'.
Locate A a t the extremity of the first chord so that d A perpendicular to P.C.d is equal to one-half the chord deflection. Produce
P.C.A to e so that Ae is equal to the chord length. Locate B
so that AB is equal to the chord length and eB is equal to the
chord deflection. Produce AB to f so that Bf is equal to the
chord length. Locate C so that BC is equal to the chord length
and fC is equal to the chord deflection.
NOTE:Results sufficiently accurate for many purposes may be
obtained by taking the distance P.C.d equal to the chord length
instead of taking it equal to
d(chord length)' - ( $ 6 chord deflection distance)Z

60

ELECTRIC RAJLWAY HANDBOOK

Chord Deflection Distance.

(length of chord)ZI
(radius of curve)
- -(degree of curve) X (length of chord)'

(chord deflection distance) =

573"

19
20

%::

:::::

/I

::

12::

66.76
68.40

The above table gives chord deflection distances for roc-it.

chord and even degrees of curve. Values of chord deflection for
other curvatures within the limits of the table may be obtained from
the table by proportion.
Track Spirals. Where the track passes from tangent to curve,
the direction and super-elevation change suddenly. To avoid the
shock and lurch of the train, due to an instant change in the relative position of cars, trucks, etc., an easement is made from
tangent to curve. This easement allows the train to take the path
of a spiral, LC., the degree of curve increases a t regular intervals
from tangent to curve until the maximum degree of curvature, or of
the simple curve, is reached. The Am. El. Ry. Eng. Assn. uniform
system of track spirals, adopted in 1923,is shown in Manual Section
W I ~ - 2 3 . The spirals of this series are of the Searles type, all
having the same angular functions and differing only in linear dimensions. The switch easements are derived from the plain spirals
in such a manner as to permit the conversion of a plain curve ~ n t o
a
branch-ofi with the minimum disturbance to existing work. At
least one switch easement is provided for each spiral which cuts in
without changing the alinement. A.E.R.E.A. recommended
dimensions for switches and mates are used throughout. The

61

EASEMENT CURVES

number which designates each spiral indicates the length of each

arc of that spiral on the center line of standard gage track. As
special trackwork is usually dimensioned on the inner rail of curves,
and staked out on the ground either on this line or on a n offset
line therefrom, the functions are tabulated for the inner rail of
standard gage track. For other gages, the same dimensions are
to be used for the inner rail, thus agreeing with the A.E.R.E.A.
definition of switch radii. I t is believed that a su5cient number of
spirals and swltch easements are included to cover all possible
requirements without further interpolation. If spirals are desired
for curves of still longer radii, those given in "Field Engineering"
by Searles & Ives, are recommended. The Searles type of spiral is a
multi-compound curve in which the radius of curve progresses in
inverse arithmetical ratio from arc to arc. I t IS constructed upon a
series of equal arcs and the angle subtended by the first arc will be
the common difference for the angles subtended by the succeeding
arcs. This form of curve has been proved by many years of experience to be the most suitable for the purpose of eliminating the shock
of transition from tangential to circular motion of a railway car.
I n the A.E.R.E.A. series the base spiral is one having successive
angles of I deg., 2 deg., 3 deg., 4 deg., 5 deg., 6 deg., 7 deg., 8 deg.
and equal arcs of 10 it. From this base a series is calculated by
proportion for the lengths of arcs given in the designations of the
~ndlvidualspirals. The dimensions of these spirals are calculated
and tabulated for the inner rail of standard (4 ft. 8% in.) gage
track shown. This alinement of the inner rail is to be followed
for aU gages t o simplify special trackwork manufacture and to agree
with the A.E.R.E.A. definition of swltch radii.
L. R. Brown, Engineer N. Y. State Kys., shows simple and
quite usable solutions of some compound curve problems in the
Elec. Ry. Jour., Vol. 55, p. 160,and Vol. 60, p. 704.
Laying Out Easement for Moderate Speeds. The following
method of laying out a n easement curve is from Trautwine's
"Field Practice of Laying Out Circular Curves for Railroads."

.I I.
~ I G .35.-Laying

out easement.

The process of laying out the curve is simple and the easement is
suficiently long for moderate speeds and curvatures. Let m,
Fig. 35, be part of the curve; and xs its tangent. Divide the chord
deflection distance (see Chord Deflection Distance) by 10. The
quotient will be cx. Set every stake of the entire curve inward this
distance, cx, so that the curve shall be removed to ca. The radius
of the curve is thus shortened by thelength r.r; but this is negligible.

62

ELECTRIC RAILWAY HANDBOOK

From x measure on the tangent l o o ft. to s; and from c measure 50 ft.

to the curve a t o. After stretching a cord from s to o, lay off the
eleven equidistant ordinates, if for rail laying; or only the middle one
(6), or lt and the two quarter-way ones (3 and g), if for gradirlg.
(Since xs is always loo it., and co 50 ft., the distance apart of t h e
eleven ordinates will always be nearly 1 2 . 5 ft.) The ordinates
themselves in feet are found for any curve by multiplying c z by the
following multipliers:
Ord.
2

Mult.
0.180
0.355

0.505

Ord.
4

Mult.
0.645
0.775
0890

Ord.
7
8
9

Mult.
0.975
0.990
0905

Ord.
10

I1

Mult.
0.715
0.430

Gage and Flangeway on C w e s . The following method of

determining the minimum flangeway required is from a report of
the Committee on Way Mattersof the Am. El. Ry. Eng. Assn., 1909.
Having a fixed rail groove, i t is desired to know if that groove
will properly serve under given conditions, and if not, how to alter
i t to meet them. Without discussing the question of wear on wheds
or rails, let us find the minimum groove in which a given wheel
flange will run, on the
assumption that the
e a u i ~ m e n t is eeomztri'calty perfect.u I n a
car traversing a curve,
the longitudinal axis of
the car truck is normal
to the radius of curvature of the track drawn
to the mid-point of the
truck, the wheels being
held rigidly parallel to
PIG.36.-Flangeway o n curves.
the a&. TKe
of the wheel
which determine the size and shape of the minimum groove are the
portions below the shaded sections shown in Fig. 36. The shaded
areas are sections of the wheels on a level with the wheel-tread. If
we project these portions of the flanges upon a radial plane, we obtain the contours of the minimum grooves, as shown a t A B . T o
obtain these projections (see ABC, Fig. 37), pass a series of planes
perpendicular to the axis of the truck. They will cut lines from
the surface of the flange. Project these lines by means of track
arcs upon the radial plane A B (shown rotated about line AB,
as a n axis, into the plane of the paper). The outline curve FCH,
tangent to all of these projections, is the contour of the required
rmnlmum groove. The gage-line of a grooved rail or of a groove is
customarily taken as the intersection of a horizontal plane through
the tread-line with the head side of the groove produced, the
rounded comer being ignored. On this basis, the position of such
a gage-line for the minimum groove as compared with a gage-line
as determined by the standard gage is determined as follows:
If in Fig. 36, with the mid-point 0 of the axle as center, we describe

Eez

FLANCEWAY IN CURVES

dl

the circle D-E, of diameter equal to standard gage and draw track
arcs tangent to D-E, one on each side and produce them to A-B
(Fig. 37), we determine the position of the standard gage-line.
Only one such determination-at mlr-1s shown in Fig. 37.
For wheels set to a gage % in. less than standard gage and m t h
the wheel gage-line taken a t a point on the fillet of the flange %

FIG.37.-Flangeway on curves.

in. below the level of the tread, it is found that the standard gageline and the minimum groove gage-line coincide for all flanges used
a t the present time.
The Am. El. Ry. Eng. Assn. recommends the following rules for
determining gage and flangeway on curves: Condition: W.heel gage
is assumed to be A.E.R.E.A. standard, namely, 4 ft 8% In., taken

FIG. 38.-Miminum

grooves for Am. El. Ry. Eng. Assn. flanges A and

between fillets of flanges, N in. below treads; track gage, q ft.

8% In.
Rule I. For A.E.R.E.A. standard wheel flanges and one wheel
base use standard track gage and determine the minimum flangeways required by graphical plotting of groove, cross-section of
wheel flanges, wheel base and radius of curve, as described above,
allowing >$ in. extra width In outer flangeway for irregulanbes
in wheel setting and special work.

64

ELECTRIC RAILWAY HANDBOOK

Rule 2. If various types of wheel flanges or various lengths of

wheel bases are operated, determine, by graphical plotting, the
inner flangeways required for both the maximum and minimum
conditions. Use inner flangeway to suit maximum conditions and
widen gage by the difference between the maximum and minimum
width of flangeways as above determined. Make outer flangeway
% in. wider than minimumouter flangeway for maximum conditions
plus the amount of the widening of the gage.
Rule 3. Rolled guard rails having too narrow flangeways should
be planed on gage side of inner rail and on guard side of outer rail
to obtain required width.
Rule 4. If inside flangeway be too wide, increase the gage the
difference between the width of the flangeway and the narrowest
groove required. Make the width of outside flangeway equal to
that required by the maximum conditions of outside radius, wheel
flanges and wheel bases, increased by the widening of gage and also
by 3 i in. to be allowed for irregularities in wheel settings and
special work.
Rule 5. I n expressing specification for gages and flangeways
measure gage on a plane % in. below the head of rail a t fillet, except
t h a t when steam railroad wheels are to be used give distance on
plane 96 in. below the head of rail, making special notation thereof.
Rule 6. Measure Aangeways horizontally from gage-line opposite
head of rail. Give vertical slope of guard if rails are to be modified.
Victor Angerer In a paper before the Keystone Railway
Club (1913) said that theoretically for track laid to true gage
every combination of radius of curve and wheel base of truck,
with a given wheel flange, calls for a specific width of groove
to make the inside of the flange of the inside wheel bear against the
guard and keep the flange of the outside wheel from grinding against
the gage-line and possibly mounting it. I t is manifestly impracticable to provide guard rails with such a variety of grooves or to
change the grooves of the rolled rail. The usual minimum of
1%
in. is wide enough to pass the A.E.R.E.A. standard flanges on a
6-ft. wheel base down to about 4;-ft. radius, and the maximum width
of 11% in. down to about 35-ft. radius. On curves of larger radius
the excess width should be compensated for by a corresponding
widening of the gage. If the groove in the rolled rail is too narrow
for given conditions, it must be widened by planing on the head
side of the inside rail, to preserve the full thickness of the guard, and
on the guard side of the outside rail to preserve the full head.
Unusual wheel bases such as 8 ft. or 9 It. may require widening of
the gage on some curves. This widening of gage is necessary only
to bring the guard intn play when the groove is too wide for some
one combination of wheel and flange. I n T-rail curves the guard is
formed of a rolled shaped guard, or a flat steel bar, bolted t o the
rail. I n special work and curves in high T-rail track a girder guard
rail is often used. This is desirable, as it gives the solid guard in
one piece with the running rail. The idea that a separate guard
can be renewed when it is w o n out does not work out in practice,
as it is usually the case that when the guard is worn the running rail
is also worn t o such an extent that i t will soon have t o come out

SPECIAL TRACK WORK

65

also. The only drawback to the girder guard rail in high T-rail
track is the compromise joint between the two. This, however,
may be properly cared for, as described elsewhere.
Hard Center and Solid Manganese Special Work. I n such
special work as frogs and crossings a t the present prevailing prices,
few lines can afford anything less durable for tracks in paved streets
than "hard center" work. I n this special work the pieces are
formed of rails held together by a casting into which they are partly
fused or welded, or they are made of an entire steel casting. In
either case a recess is left in the central part into which a hard metal
plate is set and fastened in various ways. The best metal for the
centers is manganese steel, although some work is still made with
chrome or tool steel centers. On account of the difficulty of machining manganese steel and because heating i t to some marked degree
destroys its toughness and wearing quality, certain limits are
placed on the methods of applying and fastening manganese steel
castings as centers in the main body of the special work. One
method is to use a pendent lug and key going through the lug and
body, pulling the center down into the recess. Another is to use a
combination of wedges applied to the edge of the center and between
it and the wall of the body portion. Another method is by use of a
nut and brass set screw between horizontally extending lugs on the
center and corresponding recesses in the body portion; and still
another is the method of simply bolting down the center by vertical
bolts running through both the body and center plate. I n all
except the last-mentioned method, zinc or spelter is poured in
between the lower side of the center and the body. I n the bolteddown center zinc is also used in the same way in some cases, while
in others the under surface of the center and the top surface of the
recess are finished by grinding and planing, and the two pieces are
bolted together without the zinc centers. Much stress has been
laid on the so-called renewability feature of special work, but any
mechanically jointed center that is easily detachable would, when
ut under repeated strains, be likely to detach itself. On the other
L n d , this renewability feature has been applied very little in practice, as where centers are renewed on account of wear, i t is very
difficult to make a good job out of the surfaces, even by a very large
amount of grinding of the adjoining parts on account of the necessarily unequal wear on the two sides of the rail. A good manganese
steel center should outlast the adjoining portion of the body o r rails,
particularly the curved portions which exist in nearly every piece
of special work. The wearing out of those portions of special pieces
which had been made of ordinary rail outside of the manganese steel
center led to the introduction of solid manganese steel special work
in which the entire frog, crossing or other piece was made in one
manganese steel casting with the arms cast and ground to correspond to the shape of the adjoining rail and joined to i t with the regular joint plate. The main drawback is the greater cost, but in many
rases the severity of traffic will warrant the greater initial outlay.
Special Work Speci5cations. The Am. El. Ry. Eng. Assn. has
adopted recommended specifications for special track work, the
various sections of the Manual being as follows:

ELECTRIC RAILWAY HANDBOOK

66

W x q - 1 6 Material for use in mrrnufacture c-f special work

W1o5-23 Solid nianganese steel spccial
for girder rail
W1o6-zj Cast steel hard center special wnrk
WIO? 2 3 Iron bound hi~rdcenter special work
WIOS-23 Plain bolted special work
WIOQ-21Rail bound insert type zpecial work
Flange-bearing Special Work. I n either the hard center or solid
type, the bottom of the grooves should be raised so that the wheel
when crossing the intersecting groove will have a bearing on the
flange on this raised floor, this flange hearing being a necessity with
narrow tread wheels. If the wheel treads are wide enough to span
the intersecting groove and still leave a good amount of bearing
of the wheel tread on the running surface and wide enough so that
when crossing a groove a t right angles the blow may be distributed
on a considerable area, this flange bearing is not necessary. Chicago
specifications for hard center work provide that the total length of
the level (flange-bearing) portion shall be twice the distance
between the points formed by the intersecting guard and gage
lines, and that inclines shall begin a t the ends of this level portion
and slope down uniformly to a depth of one inch a t the ends of the
manganese late, the length of the slope to be not less than six
inches. 0 t & r companies have adopted an approach slope as long
as IS inches to avoid the shock incident to abrupt rises in the
flangeway. I t is also common and good practice to raise the flangeway a t the approach to the first intersecting flangeway and continue
the shallow way clear through the crossing or series of crossings.
Switch Tongues and Mates. The important factor in the switch
is the tongue bearing or pivot which should Ix rotislrurted to hold the
R--+

"-4
---

Standard radius

Radius
for all
Rages
R

Dimensions

---

I3 and C
I

E and P

-I ---I-

roo ft. lateral

200 ft. lateral.. .
200 ft. equilateral
350 ft. equilateral .
50 ft. lateral*
75 ft. lateral*
.
loo ft. equilatera!.

hould be use on y where s cia conditions will not permit the use
of :Lndard dimen:ions1shown in R s t :our I~nes.
PIG.30.-A.E.R.E.A. standard switches and mates.

SPECIAI, TRACK WORK

67

heel of the tongue steady and to give the greatest possible resistance
against the knocking down of the tongue a t the heel end. I t must
also have means of taking up the wear so as not to allow the tongue
to get too Itwse, and still it must work loosely enough to permit
throwing devices, like electric track switches, to move the tongue
under all conditions. Special attention must be given to the tongue
switch to take up wear as it occurs. The tongue is now usually
reinforced by lateral flanges against the thrust of the weight of can.
Tongue locks, which are devices in which a spring holds the tongue
in one or the other position, should not be necessary in a properly
constructed tongue switch if the bearings are kept as snug and tight
as they should be, but are safeguards against a tongue which is too
loose or which has tendenciesof accidental throwing between trucks.

1-1 I I I
Toe
length

Angle

Heel Total
length length

--

4
5

6
7

FIG.40.-A.E.R.E.A.
standard 8
frogs for standard rall and M.C.B. ro
wheels.

1 4 ~ 1 5 'o"
I 1~z5'16"
9O31'38"
8~o'~6"
7 O 9'10''
~~43'29''

3'2"
3'4"
3'6"
4'5"
4'9"
6'0"

4'10"
5'8"
6'6"
7
8'3''
10'6''

8'0"
9'0''
10'0"
I z'o"
13'0''
16'6"

Dimensions of Special Work. The Am. El. Ky. Eng. Assn.

standard dimensions for switches and mates are shown in Fig. 39;
for frogs for standard section rails and M.C.B. wheels, in Fig. 40;
lor turnout and crossover frogs for girder grooved, girder guard,
plain girdcr and standard section rails, in Fig. 41. The Association
has also adopted, in its Manual W 1 0 - 2 1 , standard designs and engineering data for tongue switch construction in connection with
lateral or side turnouts, equilateral or diamond turnouts, and for
crossovers. The Association has also recommended distances

Grooved rail

FIG.41.-A.E.R.E.A.

Plain rail

standard frogs fdr turnouts and crossovers.

68

ELECTRIC RAILWAY HANDBOOK

between tangent track centers through special work layouts which

are believed to be sufficient in range to permit their use throughout
the country without requiring rad~calchanges in prevalhng center
distances. They have been prepared from the records of the special
trackwork manufacturers and are Intended prmarily for use In the
tangent track portions of special trackwork. I t is desirable, however, to adopt some one of these distances for tangent tracks generally. The d~stanceshave also been used in connection with the study
of the matter of standardization of branchoff frogs, as these cannot
be standardized unless uniform track centers are used. The
recommended track-center distances are as follows:
Dlstance between
track centers

Dummy gage or

davrl strho

Note that the above table indicates devil strip for standard gage;
i t is obvious that this will vary with other track gages.
Open Track Point Switch. As shown in Fig 42, the outer
rails of the track as seen in facing the turnout are continuous, one
along the main track and the other as the inner lead rail of the
turnout; each is kinked slightly to protect the points of the switch
rails which are placed between the continuous rails. The switch

FIG 42 -Open

track p a n t swltch.

rails are bent and planed so that the gage side of the head will be
straight and the other side conform to a given spread a t the heel
and thickness a t the point without cutting away the web. The
Committee on Track of the Railway Engineering Association, 1912,
recommends a spread of 6% in. between gage-!ines, a throw of 5 in.
at the fmt rod and a thickness of W in. a t the point which is afterward ground to 46 in. and the top comer rounded. A %-in reinforcing bar is riveted to the web on each side and carried back as far
as the heel connections will permit. The bottom of the switch rail is

OPEN TRACK SWITCH AND FROG

69

planed to fit on base of stock rail where bases overlap, Fig. 43a.
Sto blocks are used as shown, Fig. 42, to support the free portion
of t i e switch rail from the stock rail; which has a brace a t each tie.
The supporting plates for the two rails are planed with a step to
raise the top of the switch rail ! I in. to provide for hollow tires.
The top of the switch rail is planed down to be $6 in. lower than
the stock rail and this planing runs out or rises 9 i in. in the following
distances :
Length of
planlng
12 ft.

Lq"gth of
swttch rarl
33 ft.

The above lengths of switch rail are arranged for cutting from
33-ft. rails.

%,C

!s

1-

SstbD A-4

PIG. 43.-Open

track spring frog

(NO.11).

The committee claims that when the corres nding switch angle
exceeds one-fourth the frog angle, the s w i t c k i n t presents the
worst feature in the alinement and there is an economic loss both in
space occupied and in cost of turnout. On this basis the committee
recommends the following lengths of switch points:
16% it. for
ft. for
ft. for
11
ft. for

22

33

frogs over No. 6, and rncludlng No. 10

frogs over No. 10 and tncludlng No. 14
frogs over No. 14
frogs No. 6 and under when requred.

Frogs Nos. 8, 11 and 16 are recommended as meeting all general

requirements for yards, main track switches and junctions, with the
object of eliminating other numbers and reducing the stock pile.
The lengths shown in the drawings submitted are 13%. 1756 and
24 ft., respectively, for the three numbers. The rails are all bolted

70

ELECTRIC RAILWAY HANDBOOK

through the webs, using fillers or washers for spacers, while for the
r~gidfrogs the rail bases are riveted to a plate.
The standard rigid frog, with the various dimensions, is shown by
Flfi 44.

Actual

offronl

-1

;,6

de;:

9
Io

~n

;n

toe

I to I to I Heel

I heel 1 toe 1
1 n . 31 19 n . L
2 ~n
3
11%
an.

101Jfs
1055
8 0
5 0
3 0
5 3
1054s
3 356
9 0
6 o
5 856
10 0
3 10
6 a
6 6
9%
1 1 0
61055 4 155 7 3 3 9
936
12 o
8 o
o
9%
FIG 44 -D~rnens~onsof standard r~gidfrog for open track.
25
932
8 I0
7 10
621

11

I to heel I t o

7 0

;-

Spread

8}f

4 6

2 6
2 9
3 o
3 6

I TW

'

555

5%

556

'19k

Thes*h.g
is-%& for, main knewo~kwhere ihore ibbut xitJpr
tra5c on the side track on account of the better support of the wheel
treads a t the frog point. For the side track, the spring yields to
the ressure of the wheel flanges and the frog acts otherwise like a
rigilfrog. With hollow treads the wings receive heavy blows from
wheels coming in the direction to reach the heel or wings first unless
the latter are chamfered to inclined planes as shown in Fig. 43 TO
protect the frog point and prevent derailments, a guard rail is
necessary on each track with a flangeway of about 1% in to
the wheels past the frog. Where the turnout is on the outside of a
curve, a guard rail is often placed in advance of the switch point
to prevent the flanges from crowding and possibly getting b e h i d
the point when set for main track. Cutting the side track point
about 2 ft short allows the guard rail to reach more nearly opposlte
the main track point.
Derailing Smtches. Derailing switches may be used as additional safety devices a t points where i t is imperative that cars or
trains do not pass except under certain exact conditions. Under
favorable conditions cars may be expected to start unaided on a
grade of about soof r per cent, but wind w l i start cars down a n
easier grade, and heavy wind may start them on level trark The
setting of brakes should not be depended upon to hold cars that
are left alone; consequently, wherever there is likelihood that
cars on side track may start from gravity or be blown or e a s l l ~
ushed, the only safe policy is to provide for derailing them
Gefore they can get far enough to obstruct main track A derailing switch may conslst of a single moving rail connected with a
switch stand, although sometimes two moving rails will be used,
as in the stub switch. The rails should be set to guide the wheels

71

DERAILERS AND CATCH SIDINGS

away from the main track. Switch points for derails may be
made more blunt than those commonly used in turnouts, but for
deralls In side tracks old polnt rails too badly worn for maln track
service may bc used. The throw of the switchpoint should be such
that a deralled wheel may pass between the point and stock rails
without spreading them apart, and a t the heel of the point r a ~ the
l
nuts of the splice bolts should come on the gage srde. In side
tracks much used the derails should be connected with the main
line switch or switch stand, so that the operation of dosing the
swtch for main track opens the derall, and mce versa. This connection is usually made by means of throw rods and bell crank.
There are also various special forms of derailing devices other than
the common ones mentioned above. In order to hold the wheels to
the ties a long mard rail should be laid on the track about 8 in.
from the op site rail, extending from a point in advance of that
where the wEels are derailed.
Catch Sidings. For stopping runaway cars or trains on heavy
grades without derailing, resort is sometimes had to catch sidiags.
Cam gives an example of such provision as found on the Canadian
~ a c i l road
c
between Hector and Field, B. C., near the summit in the
Rocky Mountains, where there is a 9-mile grade of 4 4 per cent.
Along this grade there are spur tracks or "blind sidings ' r mile
apart, each tended by a switchman. Each spur track runs up into
Tlmbcr

.,-,.-

FIG.45.--Sand track for catch siding.

the mountain side several hundred feet on a very steep grade which
rises in the direction in which the grade of the main track falls.
Normally the switches are all set for the side-track and are not
closed for main track unless called for by whistle. Hence, if a train
or detached cars get beyond control and come down the grade they
are diverted to a heavy u grade a t the first switch, without giving
any signal. As the spee!at
which the runaway cars are likely to
enter such a siding is high the curvature of the turnout should be
easy and the angle of the switch points small. Wherever it is
feasible to do so, it would be well to have the switches for such
sidings turn from the outside of a curve In main track This
arrangement would permit of easy curvature in the turnout, or
perhaps enable the turnout to branch off a t a tangent. In lieu
of the upgrade arrangement the catch siding is sometimes buried
in sand to the depth of a few inches over the rails. Sand tracks
are more common in Europe than in this country. A cross-sectional
view of a sanded catch siding in use a t Dresden, Saxony, is shown
in Fig 45. The rails of the diverting track are laid gauntlet fashlon,
on the same ties with the main rails, and the stretch of diverting
track is provided at both ends with a switch for connecting with the
main line. Guard timbers or angle irons for retaining the sand
are placed a t both sidcs of each rall of the siding, which gradually

72

ELECTRIC RAILWAY HANDBOOK

dips deeper until it is covered by 2 or 3 in, of sand. The arrangement is considered very e5cient for the purpose. In this particular
instance the catch siding is 1640 ft. long and 1148 ft. of it is
covered with sand. In very dry weather the sand is kept damp.
The braking effect of sand sidings is discussed in Engineering
(London, England, 1897),and in the Bulletin of the International
Railway Congress (1899). The Street Railway Journal (1906)
gives the results of a series of tests on a catch siding as installed on
the New'Jersey & Hudson River Ry. & Feny Company's line,
where the catch siding is located on a 7 per cent grade and the
approach of the sand track for a distance of 1500 ft. is also a 7 per
cent grade. The catch siding is 180 ft. long. The car used in
making the test was one of the company's standard closed cars,
weighing 2 2 tons and equipped with M.C.B trucks and 33-in. cast
wheels. The results of the tests were as follows:

Speed
M'P'.H.
No. of
" , "test, " e
track

Depth of sand.
over head of tad
in.
7.35 in.

No.1
No. a

14.5

254

No. 3

15. o

80 ft. of
40 it. of

No. 4

23.0

19.5

$4 iq.
254 In.

80 ft. of I in:
roo It. of zf(l ~ n .

Dimce
car ran

80 ft.

180 ft.

r l o it.
180 ft.

Remarks

Free rolling.

Front truck left track

when leaving sanded
section, after sl. ht a p
,cation of air%rake:.
rolling.

&L

Brakes set m d wheels

slid on last 40 ft., of
test, leaving a coatlng
of sand $$ in. on rail.

On this road in climbing the sides of the Palisades, the tracks take
a zigzag course up the face of the cliff and ascend a t an average grade
of about 7 per cent. The catch sidings are installed on all the steepest grades and the switches are always left open so that it is necessary for a car to come to a dead stop a t this point, and for the conductor to get off to throw the switch for the main line. An automatic
tripping mechanism is used wherebythe car, after o ning the switch,
automatically resets the switch point for the catcrsiding.

.
hc. 46.-New

--

York Rapid Transit Subway.

water.

Typical 4-track section above

SUBWAY AND TUNNEL SECTIONS

R G . 47.-New

York Rapid Transit Subway.

Lexington Ave.

Typical z-track section,

FIG. 48.-Hudson & Manhattan concrete and iron tunnels.

FIG. 9.-New York Rapid Transit darlern River double tunnel.

FIG. 51.-Fourth

73

FIG. 50.-Pennsylvania R. R. North

River tunnel.

Ave. subway. Brooklyn.

74

ELECTRIC RAILWAY HANDBOOK

FIG. 52.-Market

St. subway, Philadelphia.

PIG.53.-Boston subway. Washington St. tunneL

FIG. 5s.-Berlin subway,

Schoneberg extension.

FIG. 5.1.-Metropqlitan railway of Pans.

PIG. 54.

- Cambridge

subway

Boston.

FIG. 56.-Buda-Pest
subway.

FIG. 58.-Hamburg subway in

districts with ground water.

ELECTRIC TRACK SWITCHES

PIG. 59.
PIG. 60.
PIG. 61.
FIG. 59.-City & South London Ry.
FIG. 60.-Great North & City Ry.. London.
FIG. 61.-Metropolitan & District Ry.. of London.

Electric Track Switches. In the operation of the Cheatham

switch, the car is run under a trolley pan (placed a little more than
a car length in front of the switch point) with current on if it is
desired that the switch be set in one direction, or with current off
Pole Box

PIG. 62.-Cheatham

electric track switch.

if it is to be set in the other direction. Referring to Fig. 62, assume

that the car is using current as it passes under the pan. As the
trolley wheel strikes contact E of the pan, :he current used passes
through solenoid A , the plunger of which throws the circuit changer
F into contact with plate G ; as the trolley wheel proceeds, it makes

ELECTRIC RAILWAY HANDBOOK

contact with C, and current for operatiqg the switch point passes
from the trolley wire through D, solenoid A , El trolley wheel C,
resistance J, F, G, and solenoid R, to ground, solenoid R moving
the switch point to the right (in this case) through switch rod S.
To set the switch int for the other direction, the car takes no
current (except ligKs and heater circuits) as it passes under the
pan; these auxiliary circuits require insufficient current to operate
solenoid A , the circuit changer F remains in contact with H , and
current flows from trolley through D, A , E, C, J, F, H, solenoid B
and solenoid L to ground, solenoid L moving the switch point to
the left (in this case). As current flows through solenoid B, its
plunger raises the weight W slowly against the damping effect of
mercury in the cup M, and if current flows for an abnormal length
of time, as in case of a car a t a standstill with trolley wheel in pan,
the tripper arm H is raised, allowing the circuit changer F to fly back
and open the circuit between F and H.
The operation of other electric track switches is similar to that
described above, the design of trolley pan, pole boxes and track
boxes varying in the various types. A small motor is sometimes
substituted for the solenoids in the track box. I t is most important that the track box be water tight, and in some types it is filled
with oil for better insulation of the contained electric parts. An
interlocking device may be added to prevent a following car from
operating the switch until the first car has passed over it.
Machinery of the Track Maintenance Department. On account
of the increasing necessity for economy in the use of manual labor
in recent years, there has been a general trend toward the adoption
of special tools, such as improved types of work cars, crane cars and
automatic dump cars; an increased use of power shovels both for
grading and for loading can; a rapid adoption of pneumatic or
electric tie tamping tools, electric arc welders and power drills in
regular maintenance, as well as construction work; a general rearrangement of storage yards, including installation of various forms
of labor saving devices therein.
The various forms of arc welders had their origin in the use of the
acetylene flame. With the arc welder, defective rail joints can be
repaired more quickly, with little or no disturbance to traffic, and
for about 10 per cent of the cost, as compared with the method of
cutting out the defective joint, installing an insert with two joints,
and the incidental restoration of vement and bonding.
Grinding equipment is a comprment to the welding equipment
and should be proportionate to it. Grinding apparatus may be
classified into four general type groups: ( I ) push files, handqerated,
reciprocating; (2) rotary grinder, flexible-shaft type; (3) rotary
grinder, machine type, mounted on small trucks; (4) reciprocating
grinder, machine type, mounted on small trucks or special carriage.
The larger machine-types and reciprocating grindercare so effectrve
and efficient that a t least one machine of each type may be used
advantageously on almost any property.
Pneumatic and electric tie tampers have shown remarkable savings. The pneumatic type is in more general use, but certain features commend the electric type, principally that of availability

TRACK MAINTENANCE MACHINERY

77

of power. Mechanical tie tamping can be done for half the cost
of manual tamping, the men used being of a higher grade but only
about 30 per cent in number. The tamping itself is much superior
t o that obtained with the best hand labor.

FIG.63.-Cleveland storage

yard.

FIG. 64.--Ground trolley contact. between tracks in Cleveland storage yard.

Power drills will cut the cost of drilling operations by 75 per cent
as compared with hand drills, the saving being in costs of labor, drill
bits and sharpening. Drill grinding machinery of the portable
type with special tool holders enable any type of drill bit to be

78

ELECTRIC RAILWAY HANDBOOK

resharpened by the average bborer o n the job much more quicldy

and better than ~fsent to the shop.
The electric shovel has conle into quite general use not only for
excavation and grad~ngon ex~ensionwork but for break~ngup concrete and many other forms of reconstruct~onwork Under favorable circumstances i t will reduce the cost of excavation by
two-thuds.
The pauement plow is an escient rridia the rewvalaf b k k paye
ments of all types I t has beep stated that i t will do as much workof
the most difficult kind in a n hour or less as five men can do in three
nine-hour days, and for less than 2 5 r cent of the man labor cost.
Both air drills and the so-called " S ~ cracker"
U
have been effective in removing concrete paving base in reconstruction work. Air
drills are used in connection with the com ressor apparatus for the
operation of tie tampers. The "skull cracier" is a casting of about
one ton in we~ghtattached co a derrick or power shovel arm and
rigged wlth a release trlgger 9 as to drop from a height of about
12 feet onto the concrete.
In concrete mixing apparatus three kinds have been developed:
(I) moderate sized batch mixers; ( 2 ) very large batch mixers with
self-loading hoppers mounted on trucks and self-propelling, generally
on the track, but sometimes on the roadway surface alongside; (3)
the assembly of complete mixlng plants on a train of two or more
cars. Because of its mobility, the small batch mixer operated by a
gasollne engine may be considered the most serviceable for all-round
use. The larger self-propelled mixers are particularly adaptable
on jobs where a large volume of concrete must be produced rapidly.
The special concrete mixing plant mounted on trains of cars has been
used In a few instances, its particular advantage being in the elimination of storage of concreting materials on the street.
Work cars have been most radically improved in design for various special operations, i t having been found that the use of haphazard old equipment and ordinary flat cars as work cars was most
uneconomical. The automatic side dumping car has very greatly
saved labor in unloading, as well as savlng time of equi ment on
the road and in lessening delay to passenger car traffic w h i i unloading between cars under regular serviceconditions. I t hasbeen authoritatwely stated that aut~nw.ticdump car equipment has w e d as
much as 30 cents per foot in the cost of track work. The locomotive crane and the p~llarcrane mounted upon c a n are of particular
value in handling special work installations, in load~ngand unloading r a ~ l sand in the general handling of materials.
The order of merit of various labor saving appliances, as indicated in a summary of replies to a questionna~reby the Elec. Ry.
Jour is as follows: ( I ) crane car; ( 2 ) air tamper; (3) arc welder; (4)
automatic dum car; (s)concrete mixer; (6) pavement plow; (7)
electric shovel; &) power drill; (9) acetylene torch; (10) rail grinder;
( I I ) stone crusher; (I 2) ballast spreader.
Hand Tools for Track Work. The following table prepared by
R. C Cram, Englneer Surface Roadway, Brooklyn Rapid
Transit, shows suggested lists of tools for track construction and
maintenance gangs of several sizes.

79

TRACK MAINTENANCE TOOLS

1 1 1 si 1
Construe- Repalre

Tools

10 men

..
Adzes
Axes
.
Bars--claw
.
Ban-Immg
Bars-tampmg
Benders-rall (furncsh as needed)
Boots-(furnrsh as needed)
Brooms-sw~tch
Brooms-push
Brushes-mre
Cars-hand
Can-push
Chalns
Chrsels-cold
Ch~sels-track
Chrsels-asphalt
Compressors-bond

1.2

20

8
4

12

4
3

I
2
I

4
a

20

4
4

4
4
4

20
3

I
I

5
5
5
5

4
a
a
a

4
4
I

4
4

I
I
I

4
a

6
3

I
I

ft

IOO

3
I

I
I

doz

..

2
20

Fries--hand-round

. .

a
3
I
I

z
2

15 f t

30 f t

6
a

IS
3

12

35

I
1

zo

50
2

..

Prcks
P~cks-tamp.
. . .
Plpewater
Plows
Punches-center
Puncheetrack
Punches-thread
Rammen-sand
Rammers-pav~ng
Rasps

Dnlls-56 rn ( m ~ nnumber)
Dnlls-Isis
tn ( m ~ n number)
.
~ n l l s - ~ ' m . ( m ~ ntrumber)
Drills-1% ~ n (m~n.
.
number)
D n f t pans

Dippers-water

Plles-hand-flat
F~les-hand-rall
P ~ l eholder--rat1
Flags (furn~sha s needed)
Framer--Hacksaw (at least
blades to each frame) .
Gagewtrack
Gagewwood (Insulated)
Gnnder-keystone tool
Hoes
Hook-ass
Ho-bber
Hydrant connectron
ack-track
ack-wreckrng
~ r ncrows
Lanterns
Lanternss~gnal
Level-boards
Level-pocket
Lrght clusters
Mattocks
Mauls-splke
O~lerr--steel ..
Old men
.
Palls

Sectron.
6 men

1 ; ;

I6

10
4

zso It

II i

z
2

5
4

..

80

ELECTRIC RAILWAY HANDBOOK

Tools

Ratchets
Reamers
R u l e s 4 ft foldrng
Saws--crosscut
Saws--hand
Scythes
pt
Shovels-track
ShovelsL H
Shovelpscoop
Shovels-wood

&

Taipeullns
TI^ rod-square
Wrencheehydrant
Wrenche-monkey
Wrenches-stillson
W r e n c h e ~ o mressor
wrenches-traca
W r e n c h e ~ t l erod
Wheelbarrows
Water barrels

6omen

1
I

Repalr
romen

(
I

spec
wk
1618

men

Sectron.
6 men

SECTION I1
CAR HOUSES AND SHOPS
Wbde it is not possible wlthln the llmts of thls volume to go into
any great detail on the subject of bulldmgs, ~t has been thought
desirable to Include some notes on the deslgn of car houses and shops
from typlcal structures and from tile reports of the Am El Ry
Eng Assn committees on Car House Deslgn (1908) and on
Buildings and Structures (1922), as well as some notes on the
layout of the repair shop whlch appeared orlgnally in the Electrzc
Razlway Journal, 1912
Layout of Tracks for Car House
Grade of Tracks. The grade of tracks ln the car house should be
somewhat above that of the tracks In the street to prevent the
entrance of water into the house
Convemence of Operating. In an op.ratlng car house the saving of tlme 1s of paramount ~mportance,and the speclal work should
be so planned that the incomng and outgolng cars wlll not ~nterfere
wlth each other, and the tracks should be constructed In the best
manner In a storage house inconvenience and loss of tlme may be
tolerated, and here, for the sake of safety on the main line, trahng
sw~tchesare preferred by many roads
Car-house Track Layouts. Figs I to 12, mcluslve, Illustrate
specla1 work deslgned to meet certaln conditions Where convenlent to have ~ t a, Y Incorporated In the speclal work will permt
the operation of cars from one end, as 1s necessary wlth cars of
some deslgns and IS convenient when operatlng sernce and other
cars in tandem Ths same object can be attalned by a loop, whch
is more convement but not always feasible, owng to a lack of room
Flg I shows a track layout where the car house 1s bulk on the street
hne and it is des~rahleto introduce as few frogs and sw~tchesin
the maln l ~ n eas poss~ble The three track bay arrangement is
objectionable on account of the roof span, which would requlre
trusslng Fig 2 shows a track layout whlch will permlt frequent
operat~ngon one of a palr of tracks Operating on the other track
requlres the use of a cross over In the street Flg 3 shows a layout
for an operatlng house whlch introduces a mnlmum amount of
speclal work In the street Flg 4 shows a gauntlet track layout
beslde the maln track to rmrumlze frogs and swltches In the main
track Flg 5 shows a simlar layout to Fig 2 except that the operatlng tracks are on the rlght hand slde of the house Ftg 6 shows
a lajout for a house on a through lrne operatlng In erther d~rect~on
The Y 1s formed by a cross over on the two center tracks Flg 7
shows a large operat~nghouse whlch calls for a long, narrow track

81

82

ELECTRIC RALLWAY HANDBOOK

approach. The Y is formed a t the entrance to the street. Fig. 8

shows a track layout for a storage house which is built well back
from the street. Fig. 9 shows a third track approach to a car house,
a hyout which requires a wide street. Fig. 10 shows a track layout
where the approach to the lot is restricted a t the street. Fig. X I

FIGS.I-5.--Car-house track layouts.

shows a common type of terminal layout; special work built with
4-in., 48-lb. T-rail. Inside of years the mates were all renewed
and within 6 years the speciai work was entirely worn out and
was replaced with 4%-in. rails with hardened centers according to
the layout shown in Fig. 1 2 . The car house has a capacity of 1 2 7

CAR HOUSE TRACK LAYOUTS

83

cars, and there are operated from it 336 c a n daily on five different
lines, or one car every go seconds during rush hours.
The track layout of a double-end car house for single-end cars is
shown in Fig. 13, which is the Luzerne car house of the Philadelphia
Rapid Transit Co.

Prcs. 6-10.-Car-house

hc. ra
track layouts.

The main feature of the track plan at the Forbes St. terminal of
the Pittsburgh Railways (Fig. I ~ is) a loop around the substation and
transportation buildings. One side of this looj, is a htltler track

84

ELECTRIC RAILWAY HANDBOOK

having all switch points facing the direction of car movement. Cars,
therefore, which approach the terminal from either direction may
easily be turned by as sing around the loop, and a t no time will

FIG. 11.
1'1~s. 11-12.-Car-house

FIG. 1 2 .

track layouts.

PIG. 13 -Luzerne car house. Philadelphia R a p ~ dTrans~tCo.

they encounter a facing switch point except when backing into the
car house for a trailer or for inspection a t theendof the day. This
arrangement also makes i t possible simultaneously to couple a

CAR HOUSE TRACK LAYOUTS

85

half dozen trailers to as many motor cars on as many tracks without

having one crew get in the way of another. I t is the intention to
store many trailers here between the morning and evening rush
hours, thus saving the dead mileage to their regular car houses a t
the ends of the several lines. Fig. 1 7 shows the layout of the
Hooker St. car house and yard of the Springfield (Mass) Street
Railway, typical of the outdoor storage plan which is gaining so
much favor, and an excellent example of combined operating and
light maintenance in a single plant
Housing Car-house Special Work. In general, car-housespecial
work requires no more protection from the weather than does special

,,,
,

FORBES

FIG.14.-Porbeq

STREET

St. Tenninal, Pittsburgh Rys.

work on any other part of the track. Whether or not car-house

special work, ladder track, for instance, should be covered will
depend upon the possibilities of keeping the car house sufficiently
warm during freezing weather. I n this connection, it should be
noted that ice should be melted from incoming cars in time for overhauling. An exposed ladder track will require a door for each
track or group of tracks entering the car house from the ladder track,
while a covered ladder track will rcquirc only onc door a t the

86

ELECTRIC RAILWAY HANDBOOK

entrance. For this reason, the necessary tem rature will be more
easily maintained with a covered ladder t r a c r A covered ladder
track requires the expense of extra wall and roof, but its trolley
wire construction may be simpler and more durable than that for
exposed special work.
Clearances. Frequent accidents have shown the necessity of
establishing proper clearances between cars and between cars and
posts, walls and other fixed portions of the building. The minimum
clearance should be 2 ft.; on curves i t is well to increase this on
account of the constantly in~reasingsize of can.
Spacing of Tracks. Clearances also determine the minimum
spacing of tracks. Eleven feet center to center is satisfactory for
a storage house, with 2 ft. additional if posts occur between tracks.
This spacing will also do for inspection section of pit room in a n
operatlng house, but where a great amount of work is to be done i t is
desirable to increase this distance.
Transfer-ways. I n closely populated districts where i t is not
feasible to have all tracks entering ones, i t is necessary to install a
transfer-table. This is equally necessary in a house from which
a large number of cars are operated, and where there is constant
shifting from closed to open cars and vice versa. For convenience,
i t should be located centrally. I t s position inside the building,
however, if i t passes through party walls (even though fire doors
be used in the openings) adds to the fire risk and will affect
the insurance rates. For this reason i t is often advocated that
the table be placed outside the building and a t the rear. When
so placed, however, i t requires the passing of cars through the pit
room -when being shifted and oftentimes requires the moving of a
large number of cars-more than if placed near the center of the
building. The objection that a transfer-table within the house
causes a loss of storage space can be answered by the recommendation of a flush transfer-table.
Cross-overs. While generally desirable to omit all special work
within the house, on certain types of layout cross-overs are necessary within the building for the convenience of operation. A righthand cross-over is preferred to a left-hand one as being the more
convenient.
Design of Car-house Building
Convenience of Operation and Workhg. For the convenience of
operation and worklng, the operatlng force should be placed a t the
front of the building. The starter, or whoever has charge, should be
placed whcre he can see all incoming and outgoing cars; he should be
in close touch with the lobby, in order to call the men assigned to
duty, and to preserve discipline among the men. The superintendent, or any other official interested in the operation of the cars,
having quarters a t the car house, should be placed in a n equally
advantageous position. That portion to be set apart for working
and repairs should be placed far enough back in the house to prevent interference from shifting cars.
Quarters Other than Operating. Besides the quarters for operating and repair forces, which are found connected with almost all car

CAR HOUSE DESIGN

houses, some roads
desire to p r o v i d e
quarters for such departments as machine
shop, b l a c k s m i t h
shop, paint shop and
road and line departments. If a power
station or substation
is to be located on the
premises i t is better
to have i t in a separate building from the
car house. Some substations, however, are
built as part of the car
house, but they are of
fireproof material. I t
is sometimes found
necessary to provide a
garage or stable under
the same roof as the
car house,but in such
a case it should beshut
off from the car house
by fire walls, and the
penetration of gas or
ammonia to the storage portion should be
prevented. W h e r e
operating houses are
located away from
populous centers i t is
often desirable for the
road to provide some
lace (a s e p a r a t e
guildingon account of
fire risk) for employes
to obtain meals, or
interest other parties
in roviding accornmogations of this
kind. I n similar 10cations quarters for
slee ing are also desiraLe for men who
have to be held for
work on snow or night
duty. These quarters
may be located in the
car house. Fig. 15
shows the layout of
the Middletown car
house and shop of the

67

88

ELECTRIC RATLWAY HANDBOOK

Connecticut Co. The offices and shops are arranged along one
side of the building. The first room to which entrance is obtained
is that for the conductors and motormen, where lockers, a drinking
fountain and other conveniences are installed. Following this are
the toilets, the master mechanic's office, storeroom, oil room, salt
and sand room, pump room, wheel room, blacksmith shop, machine
shop, coal storage and boiler room.
Fire Protection. The importance of providing fire protection
should not be underrated, and a t no time should i t be forgotten.
I n selecting the location for the house, proximity to buildings of
inflammable material or having contents that burn casily should
be avoided. Car houses should also be located where a good
hydrant service is obtainable, and as near to a fire station as
possible. If this is not done a private service with ample supply
of water must be installed. Provision must also be made for a
secondary supply of water, either in reservoirs under floor, or in
elevated tanks.
Fire Prevention. No road should expose more than a certain per cent of its rolling stock to the risk of destruction by any
one fire, as the loss of cars means the loss of revenue. Divide
the house by party and curtain walls where it is practicable to do
so without interfering with the operation of cars or increasing the
cost of the building to a prohibitive extent. The omission where
possible of combustible material is recommended, especially below
the grade of top of rails. I t s use should be avoided on outside walls
and cornice work. Sheathing partitions inside are undesirable,
and floors and roof, if they are to be of wood, should be mill construction, with heavy timbers and planking.
I n the construction of a car house, corners and recesses where
rubbish would tend to accumulate should be avoided. Special
care should be taken to avoid the accumulation of oily waste and
rags. Fireproof receptacles should be provided for such material
and they should be regularly and frequently emptied. Ample
light and cleanliness will aid greatly in reducing the possibility of
fire to a minimum.
Insurance Regulations. I t is recommended that the designer
familiarize himself with the regulation of the fire underwriters
and adopt their suggestions as far as practicable. Insurance
requirements are important factors in determining many details
of construction, and a consultation with the underwriters when
planning may save many expensive changes. The underwriters
stipulate that no section of the house shall contain more cars
than amount to the value of $200,000.
Automatic Sprinklers are to be preferred as the hest possible fire
protection in enclosed places where the sprinklers will be opened by
the heat of the fire. The very fire itself which is sought to be
extinguished sets in operation the influence which extinguishes it, a t
the particular spot where the fire is and a t no other, and entirely
independent of human action. This applies to all buildings whatsoever. There should always be two sources of water supply, and any
two of the following methods can be adopted: city water with adequatepressure; elevated tank; pressure tank; underwriter fire pumps.

FIRE PROTECTION

89

Auriliarg Protection. Sand pails, chemical extinguishers and

water pails should also be provided. Small hose lines are advisable
for reaching sparks and flames in places not reached by water from
the sprinklers.
Outside Protection. For open s aces, including yards, automatic
sprinklers are not usable, since t& heat will not operate to open
the sprinklers. For such places the available methods of protection
are: (a) Universal nozzles on standpipes; (b) standard fire hose and
nozzle; (c) open sprinklers set in operation by human agency.
Universal Nozzles on Standpipes are the best protection. There
is no possibility of delay during a serious fire through the bursting

of hose occasioned by kinks, or momentary excessive water pressure,

or by the accidental cutting of the hose by the running of the cars
over it in trying to remove them from car yards. The nozzle
should not be less than 1% in. or more than 1% in., if the water
supply and pressure can be had. Since the operator cannot move
such nozzles bodily to the place of the fire, the range of the nozzle
should not be more than loo ft. in order to secure the greatest degree

90

ELECTRIC RAILWAY HANDBOOK

F I R E PROTECTION
of &dency, which distance will necessitate as high a pressure a s
can be safely maintained. High water pressure is advisable for
the further reason that the operator may be a t a safe distance and
still place water upon the fire. Intense heat may prevent the operator from placing water upon the spot desired if the water pressure
is too low. This condition applles to nozzles on hose as well.
The universal nozzle should be located a t a height of from 10 to 12
ft. above the tops of cars. The pressure a t the nozzle should be I W
Ib. if possible. For this pressure the discharge for 1%;-in.and rjijn.,
1%-in. and 2-in. nozzles will be, respectively, 466,671, 904 and I 194
gal. per min. I t is believed that the effective way to extinguish
a fire, especially one which has gained much headway, is to concentrate one or more heavy streams of water upon one particular car in
the yard and maintain i t there until its effect is shown. This can
be readily accomplished with a standpipe and universal nozzle.
Pumps should be provided to supply a t least two nozzles a t one
time. Nozzles should be located so that their range circles will
overlap safely. Universal nozzles are the best protection for practical installations where cost is considered; they are certainly to be
preferred over standard hose and nozzle in yards where cars are
stored. Universal nozzles should be supplemented by small hose
and nozzle for getting a t sparks and flames not accessible to the
stream of the universal nozzle, also by water pails and chemical
extinguishers.
Fig. 16 gives the details of such a water tower as installed in the
yards of the Puget Sound Traction Light & Power Co. The towers
are connected to the city water main by a 6-in. supply line, and each
system is equipped with a steamer connection for the use of the city
fire department. I n order to cover the yards thoroughly, the
towers were spaced so that any car could be reached by a t least two
streams. Tests show that where the static pressure is only 65 Ib.,
the nozzle pressure is 47 Ib. and the discharge 255 gal. per minute.
With a static pressure of 125 Ib., the nozzle pressure is 88 Ib. and
the discharge 350 gal. per minute. Thus a car can be flooded with
water a t the rate of 400 gal. or 5 w gal. per minute, making it almost
impossible for fire to spread to neighboring cars.
I n a proposed layout for the Cleveland Railway the yard is
750 ft. X 454 ft., including car-inspection structure, offices and
utilities buildings. The 18 standpipes are located around the yard
a t intervals of go ft. The standpipe nozzle tops are 136 in. in
diameter and the supply mains are of 8-in. and 6-in. diameter.
Two hydrants are also installed to reinforce the interior sprinkler
protection of buildings. The water supply is forced through the
mains by a 750-gal. electrically driven underwriters' pump of centrifugal type, taking suction from a 75,ooo gal reservoir maintained
on the premises. Under maximum pressure the fire pump will
deliver rooo gal. of water per minute, which can be concentrated
on any one car in the yard.
Standard Fire Hose and nozzles are the next best available protection. Hose should be 2 ~ 4in. for ordinary cases. No~zles
should be not less than 1% in. Pressure should not be tw, great,
because men who are not professional firemen cannot handle the

92

ELECTRIC RAEWAY HANDBOOK

nozzle under great pressure. Pressure a t nozzle should probably

be 40 Ib. to 60 Ib.
Open Sprinklers operated by human agency are probably a very
e5ective means for checking an early fire, and perhaps for extinguishing a fire which has gained more or less headway, but the cost
appears to be too great for general use, although for small installat~ons,where the sections are not large and where the controlling
valves can be operated by hand the cost may warrant the use.

Details of Car-house Design

To design the most economical car-house building for a particular
service in a particular location the designer must be able to design
the several types of construction in detail and he must be able to
estimate the costs of the materials and labor necessary for the
finished structure. The general determination of the most econom-

-b
-

~ I G18.-Luzerne
.
car house. Philadelphia.
repairs.

Crowsection of bay for l~ght

ical design of car house consists in a determination of the most economical foundation, wall, form of roof and form of skylight. These
individual parts must finally be considered with regard to their
suitable relationshi to each other. Among the general decisions
which must be maxe a t the outset by the designer are: Shall the
building be short and wide or long and narrow? Shall two-, threeor four-track units be used? Shall the foundation be of stone,
brick or concrete? Shall the walls be of brick, terra-cotta, concrete
blocks, iron or concrete? Shall the roof framing be of wood,
unprotected steel, protected steel, reinforced concrete or a combination of these? These must finally be decided with regard to their

94

ELECTRIC RAILWAY HANDBOOK

95

CAR HOUSE DESIGN

suitable relationship to each other. At the outset the designer

should acquaint himself with the following: municipal regulations
and building ordinances, rules and requirements of fire underwriters, character of soil upon which the foundation is to rest and
the costs of materials and labor.
Figs. 18 and.19 show the construction of the Luzerne car house of

Conncte:
Quantity
Girder A. 26X I ~ X J ~ . I S
144
+ ................
82
Girder B. 5 X 8 X l q X lq.66a144 . . . . . . . . . . . . . . .
57
Kaees.3X4X14c1z . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
14
Slabs.14.66Xz8.gX3.~i1a-4.~5~4.66~~.sc12.
.. 104

Cost

Total concrete a t 20 cents... . . . . . . . . . . . . . . . . .

Slrrl:
Girder A. 6 - 1 % ~ I.. X 34. ........................
~ ......................
Girder B. 5 x 3 - % c . r . 17..
K n e w 2 x 2 - x c . r . X ~ . .........................

$51.40

steel

Fwmina:
2560 b. m. a t

257

.................

895.
344.
95.5
190.

............................

1524.5

Slabs.lX16X34Xarea!(sq.c.b

-"
lotal

2s cents per

Total cost . . . . . . .
Total cost per sq. f t .
PIG. 21.-Typical

1coo=65+5 for knees..

...

................

......................

45.74

70.00

$167.00
0.33

study sheet for car-house constnlction material.

ELECTRIC RAILWAY HANDROOK

the Philadelphia Rapid Transit Co. The girders
sts, roof slabs
and Boors are of reinforced concrete and the w % ? are of brick.
'Thc concrete portion was built by the tinit method, whereby the
concrete members are cast separately and lifted into place.
Fig. 20 shows the construction of the hliddletown car house of
the Connecticut Co. The floors and footings are of concrete, the
walls 3re of brick and the roof is of mill type construction framed
with yellow pine girders.
Foundations. If the soil is of good bearing rapacity little study
need be given to the matter of foundations. A car house i s not a
heavy structure, the walls being light, and the load of cars being
well distributed by trark work. 'The materials to be selected for
foundations are largely determined by the supply near a t hand.
Concrete is probably as common and as satisfactory as any. Where
soft yielding soil is encountered and piling found necessary, concrete
piles have been used. In such cases and under certain conditions
they undoubtedly prove more economiral than wooden ones.
Walls. Brick masonry or conrrete monolith seems to be the
most suitable construction. If a cheaper form of wall is wanted,
a heavy mill frame with a 2-in. cement plaster curtain makes a
good fire-resisting scheme and one which is inexpensive to maintain.
Corrugated iron is sometimes used, but is not recommended.
There are in the market several forms of asbestos boards, but their
cost does not warrant their selection in preference to masonry.
Roofs. A Rat type of roof is the only suitable one for a car house.
I t should be as close to the trolley wire as possible. Light can he
brought through the roof by means of skylights or monitors. The
Am. lil. Ky. Eng. .\ssn. committee strongly fa\.ored a mill type of
heavy construction for the roof and atlviscd against the use of stcel.
The use of concrete for roof construction may he advantageous
because of its fire-resisting quality. I t s cost, however, may bc prohibitive especially where long spans are necessary. Water should
be taken off by valleys rather than by guttersand taken down within
the building. For the covering of the roof there are a number of
materials, all of which have some merit, but none of them are any
more satisfactory than a tar and gravel roof~ngwell laid with firstrlass materials.
Fig. 2 1 by C. A. Neff and T. P. Thompson, Elcc/ric
Roilulay J o a r t ~ l ,1913, is a typical one of forty odcl study sheets
used by them in determining the proper rooting design for the
car house of the Virginia Railway & Power Co. These sheets
covered the various systems of steel supports with slal,s, tile,
platesas well as " monolithic" ant1 "unit built" structurcs.
Fig. 2 2 is a study sheet prepared by these engineers to sliow the
results of a n investigation of three different plnns of roofing three
units in one building. Scheme No. I shows the roof high on one
side and sloping all one way. Scheme No. 2 shows the roof high
in the center of the renter unit and sloping two ways, scheme
No. 3 shows the roof high over the center of each unit and sloping
two ways over each unit. In each schcme the slope per foot is
the same. As the necessary distance 1)etween track and trolley
was fixed a t 16 It., it was considered that all \\.all used above the
horizontal line 16 ft. above the tracks should be chargedagainst

97

CAR 1IOUSI:. DESIGN

each scheme in considering the relative cost. In scheme No.

3 there are only 2 ft. of brickwork above the trolley line a t the
outside wall, whereas in schemes Nos. I and 2 there are 3 ft. of
brickwork. This results from the fact that in scheme No. 3 a glrder
3 ft. deep in the center and only 2 ft. deep a t the wall line was
possible, whereas girders 3 It. deep throughout were required
in the other two schemes for the same strength. For purposes of
comparison, the walls were considered as r j-in. brick walls, the
figures being based on trick laid in the walls a t $ 1 5 per 1000 ant1
each wall being qoo It. long.

Scheme 1
Brirk Work aboveTrolley Liue Ludicatrd thu>..--.

(9+145+3)x BCOx I8 x ,015ctr.-UI84.W

..n

Scheme 2
Brick \Volt above Trolley L l n s Inldtr.rted tbur----_--(3+ 5+5+3)xBM x I8 1.015 cts.-W50.W

Scheme 3
Brick Work above Trolley Line l u d ~ c r t e dt h u . 8 . - - ~
~ 2 + 2 + ? + 2 l rBM 111) 1 .Dl5 CIS.-U728.W

SUMMARY
Scheme No. 3 saves over scheme No. I . . . . . . . . .
53.~56.00
S h e m e No. 3 saves ovcr scheme No. a , . .............. 1 . 7 2 8 . 0 0
Scheme No. 3 saves over form of roof on certain c3r houses
11uilt by others
. . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.580 oo
FIG.22.-Typical study sheet for car-house roofing schemes.

Posts. The 1908Am. El. Ry. Eng. Assn. committee favored the
adoption of posts for the roof supports in preference to trusses.
In the pit room they provide a means for supporting the car handling apparatus; they are convenient for holding the aisle sprinkler
pipes, and also the standpipes and hose. They can also be used to
support brackets for the fire pails. Where there are posts it is
easy to introduce curtain walls, either the entire height of building
or dropped down 6 i t . or more from the ceiling. Plaster concrete
partitions along lines of roof posts make excellent fire curtains.

Trolley Troughs. One of the most satisfactory ways of holding

u p the trolley \ \ i r ~ is
. I)y a suspendetl plank. T h e plank serves not
onl\ tlii.; I~ilrpn.~,.1)11t~ l \ o~ I I C I I the trollry polr Ir:~vest h r wire,
prrvrciI< 1Ii(. ~ r o u n t l i ~ iofg t.urrt.nt throufih contact of p ~ l rwith
sprinklrr pip(.\ r l c , in the I)~~ilding.
Bumpers. At the dead end of every track, :inti ,3 ft. iron1 the wall,
there should Iw n stop t o prrvrnt the carp doing tiamage. There
are st.vernl Li~iclsof stops, such a s the common (ast-iron shoe Imlted
to the rail o r fabtrnecl with clipping blocks, and bumping posts,
consistin:: of hard pine timber buried in the ground, or concrete.
'There i~ also the post srcurrcl with holt and str:ips to the rail.
Doors. Doors are necess:~ry for those p o r t i o ~ ~ofs the house
whit h have to be heated and those storage portions t h a t are seldom
used. 'I'hey are less necessary on other parts of the house and may
he omitte(1 if one is satisfied t h a t there is no fear of intrusion b y
thost, hrnt on thieving, sctting lire or other ~ii:~licious
mischief. It
is dcsir:tl)lc* to havc thc tloors swing out Iloth swinging and
roller typtss of doors :ire being used. 'The former has many points
in its favor, I)ut the latter is a great convenience where room
will not permit the use of sn~ingingdoors. Where swing doors a r e
used they may be partly glazed. Roller type doors are sometimes
motor driven for greater convenience and more rapid operation.
Floor Construction (Recommended by Am. 14. Ry. Eng. Assn.,
1911, Committee on nuildings and Structures).
( 1 r
/ / J I
Concrete Iloor with cement finish is hest for
perrn:inent construction. Otherwise crushed stone a n d screenings,
or ashes is ;tdvisrd.
(11) ('or Ships. C'onrrete floor with cement finish is rccommended, except in machine shop, if subjcctccl t o heavy service,
nherr creosoted wood block on concrete foundation should be
installed.
(0 Po;,bcr Ilorrrrs. For engine room, concrcte with cemcnt o r
tile finish is advised. For boiler room, concrete and cement finish
or a floor of hrick laid on edge in cement mortar is recommended.
( J ) O#rrs, E ~ ~ ~ p l o yRoom,
rs'
elc. Alaple or con~bed-grainyellow
pine wearing floor should he used. If wearing floor is supparted
by sleepers embcdded in concrete, it is always k t t o install a false
floor underneath the wearing floor.
( e ) Toilr! atrd lorkrr-room joors shoultl he of concrete with cement
finish, and connected with sink ant1 t r a p to drain, so t h a t the floor
may be conveniently washed and scrubbed down.
( j ) For ;,sailing rooms or shrlicrs and plutfortns, floors of concrete
and cement finish are recommended for permanent structures.
Rolled broken stone and screenings o r ashes may be satisfactorily
installed for cheap open shelters for temporary construction.
Lobbies. As already stated, the lobby should be under the eyes
of a supervising o5cial; and a n y plan by which the men are left
to themse!ves is undesiral>le. The lobby should be provided with
seats, tables, bulletin hoards, and the like; counters should be
arranged for the filinp of time sheets and other papers requiring the
signatures of conductors. All seats are Ixtter attached to the wall
for the sake of cleanliness. hletal lockers should be provided

CAR HOUSE bES1C.N

9D

dirertly in the lol~byor in an adjoining room. For those men who

are o f a cluict clislmsition a sel~aratt.room shoultl 11t. pro\ itlctl, whercreacling and the playing ~i c p i t t g.imcs can IJC intlulgc.tl in. 'I'his
room stloi~ldh:l\r ;i tal~li.fnr lrtter writing. Sc1):lrate rcmms are
also advi.;;~t~lcfor making out time shri,ts, reports nncl acri~lrnl
blanks. ,111 ~vartsof the lobl~i$,<
ant1 toilet3 shollltl I)r so tlcsigned
as to suffer as little as possil~lefro111 the destructive work ol idle
hands. Adjoining thr l o l ~ l ~shoulcl
y
be aml)le toilet facilities, with
rovisions to keep thcln in sanitary tondition. 'The floor shnuld
such as to admit flushing; partitions should not run to the floor;
as little woocl as possible should be used; sharp corntrs .zho~tlcl
be avoicled; and there sllould be g m 1 provisions lor light and air.
In the larger llouses there shotlld be equally wrll-planned toilets
adjoining the pit room, and possibly a private toilet for the officials.
If the house is small a central toilet with one or two water closets
under lock and key may answer for all ctnl)loyc.s.
Pits. The pit room, one of 1111. n ~ o \ timport:~nt p;irts ol t hv
house, shot~ldbe placed well Ijnck. ;~ntlthr :irr;r 41oulcl reprr\c.nt

&

PIG.23.-Section

showing drop floor I x s ~ ~ ptt.

le

a capacity of a t least one-third-or, better, one-half-of the car

storage. I t should be well lighted, to save artificial light, and
amply heated for the sake of the comfort of the men and for the
melting of the snow around the trucks. Pits should have an
average depth of 4 It. 6 in. from the top of the rail. Preferably,
the Boors of the pits should hc of concrete ancl slightly crowned
to shed water. The main Boor on one or more tracks should
be dropped next to the rail on each side of the pit to facilitate working about the truck and the running boarcls. Fig. 23 shows a drop
floor a t the side of the pit. This floor is 2 7 in. below the top of the
rail. I t can be less if preferrecl, and with some equipments 1 2 in.
proves satisfactory. The natural lighting for the pit room is
through the monitor of saw-tooth type with vcrtical glass. A cross,
or transverse pit, with the floor slightly Ix.low the floor of pit and
terminating in a room a t the side, is a convenience for moving
~
lor storage or, if
material from pit to pit. 'Thc room can I I usecl
well lightell, lor a work room. ,In olxning can be provided so that
material can be lifted to the floor above. By referring to Figs. 2 3

100

ELECTRIC RAILWAY HANDBOOK

to 29, several schemes of pit constructioq will be found illustrated.

(See also Figs. 18, 19 and 20.) A form of construction that
utilizes the rail for the support of the floor, as well as to carry
the cars, will be found simple and economical. I t also does not
pocket the heat under the floor and gives greater clearance for
working and for assing from pit to pit. In the form of pit construction showny! Fig. 29, a single pit suacient in size to cover
all tracks is constructed in the car-house floor. Tracks over
this pit are supported on columns made of a single 3 X 3 X
in.
angle with a gusset plate a t the top to serve as a point for attaching
a strut, two diagonal
- ' ~ r u ; ( _ . j
,;blank ~ ~ o o r tie rods and the rail
clips. Two a n g l e
lugs a t the bottom
serve to spread the
load on the concrete
pedestals and as a
means of anchoring
the s u p p o r t i n g
structure to the
pedestal f o u n d a tions. These track
columns are spaced
9 ft. 6 in. apart and
are built In bents
similar to bridge
FIG.
zh&ction
of ptt construction deepe; and buck braces. They
are paired to s u p
nore expenstve than usual supports.
port the rails of
parallel tracks, and are stiffened diagonally by round tie rods
with turnbuckles for adjustment. A 4 X 3 X xs in. angle
between columns serves as a strut to prevent overturning and as a
support for the walkway between the tracks. In order to provide
sufficient girder strength between the pit columns IOO Ib. A.S.C E.
rail has been used to sgan the 9-ft. 6-in. interval between them. The
method of fastening t e r a ~al t the columns is by means of clips bolted
to the horizontal leg of the structural angle strut.
An installation of push buttons by means of which errand boys
may be called to its will save much time.
Apparatus. &le the apparatus for the handling of the cars is
not a matter of car-house design, still there must be provision made
for its installation. Supports sufficient to take cranes and body
hoists must be provided, as well as revisions made for motor and
wheel hoists. An example of wheey drop pit and wheel hoist is
shown in Figs. 25 and 26 Motor hoists may be constructed similar
in design to Fig. 26, but care should be taken in location that the
wheel hoist does not tie up the motor hoist, as the operation of
changing motors or changing armatures in split-frame motors usually requires much less time than that of changing wheels.
Stock Room. The stock room should be placed convenient to the
pit room and should be of sufficient size to contain all the stock.
I t should not he so large as to make the care of stock difficult, but

CAR HOUSE DESIGN

101

102

ELECTRIC RAILWAY HANDBOOK

Bottom Plan of Can

on
Pier foundation may be omitteb
U steel rexnforcement i s US$&

PIGS.27, 28 -Types

PIC 28
of p ~ construction
t
wlth 8 - ~ nT rat1 on
supports.

pier

or column

CAR HOUSE DESIGN

Its

should be a well-lighted, neatly designed room which will inspire

the keeper with neatness (and neatness means economy where there
are a large number of small parts to be kept In stock). Broad
benches with lockers beneath them, drawers for screws and small
s
lettered and palnted will be a n object of pride
parts, and b ~ n neatly
as well as an incentwe to savlng
In the movable bin construction used in the stock room of the
Syracuse Rapid Transit Co the b ~ n are
s of galvanized iron and are
built up as follows: Carry~ngchannels are bolted vertically to the
wall a t 26-111 centers and are perforated for the insertion of the shelf
brackets a t any desirable intervals The shelving consists of honzontal sections 12 ft. long which have a series of holes 4 in apart
so that the vertical partitions can be riveted to the shelf a t intervals of 4 in or mulhples thereof.

FIG z g -Pit construction on structural steel.

Blacksmith Shop. If any repairs are contemplated, i t will generally be found necessary to prov~dea place for blacksmith work
Th~s.placeshould be a t the end of a track so that the trucks can be
run In The small type of blacksmth shop will require a work
bench and a n opportun~tyto install a few hand-drill~ng machines,
etc As the amount of work and the size of the house increases,
more machinery is required With small roads a fully equipped
machine shop is maintained for doing all the work of the road,
including a small but well equipped foundry for brass work.
Paint Shop. A small palnt shop for the renovation of cars is rovided as a part of the car house on 5ome roads The paint g o p
must be so located that it will be absolutely dustless If the building in which it is located is heated indirectly, the air should be blown
into the paint shop first Preferably, however, the paint shop
should be heated by steam, even though other parts of the house
use hot air.

104

ELECTRIC RAILWAY HANDBOOK

Wash Room. The use of the wash room and the washing of cars
with the hose is being abandoned by manyroads; instead, the entire
floor is drained, and with hose bibbs a t frequent intervals the cars
are washed a t any point by the use of pail and brush.
O
i
l Room. Wble some prefer a separate building for the oil
storage it can be made a part of the car house, if properly designed.
The floor should be dropped below the outside floors, dished to the
center and properly dralned. There should be barrel racks, shelves
for small vessels and waste receptacles, all of incombustible materials. Proper provisions should be made for heating the oil so that
it may be fit for use in the winter. Tank systems, having the
tanks buried outside in the ground and the oil piped into the
house, afford a good method of caring for oil and one not liable
to much waste. Am. El. Ry. Eng. Assn. Manual, Section Bzo621, describes a design of oil house which, it is stated, is economical
and convenient from an operating standpoint, and a t the same time
meets all municipal and insurance requirements.
Sign Room. I t is convenient to have racks for the storage of
signs used on cars and they should be convenient to the operating
section, so that the shifting of signs can readily be made. There
should be facilities for the pasting of signs for the dashers. This
work should be done in a room set apart for it.
Sand. I t is important to provide for the storage and drying of
sand. The apparatus for the drying should be laced close to the
operating end of the house and should allow circugtion of air as well
as of heat. The process of drying is slow a t the best, so that it is
well to have a large storage s ce adjacent to the heater. This
s ace should be filled with a i r x e d sand in the summer, and, with
t i e assistance of the heater, an ample supply can be kept on hand
during the winter. In the Luzerne car house of the Philadelphia
Rapid Transit Co., an elaborate sand-storage and distributing
equipment has been installed to make it easy for the crews to fill
their boxes with clean sand in the shortest possible time and without
placing bins throughout the building.
For this purpose the company built a sand house where the
sand brought from the drying plant in hop r bottomed c a n is
dumped and then raked by a motor-driven & L e t elevator to a
belt conveyor which distributes its load throughout the sandhandling panel extending across the rear of the car house. This
panel is 16 ft. wide and is lighted by means of steel-sash windows in
the rear wall. Outlets from the bottom of the sand storage are
provided a t the partition walls with a supply pipe for each side of
the partition. To obtain sand, it is necessary only to raise a
weighted valve and the sand receptacle will be filled.
Salt. Less space is required for the storage of salt, but full provision for its handling is just as necessary.
Heating Systems. The following systems are recommended
by the Am. El. Ry. Eng. Assn. (Manual Section B203-11):
(a) For car houses, a blower system, where the air is blown over
steam coils and through the building.
(b) Large car shops, the blower system, except in the paint shop,
where direct steam radiation is advised. For the small shop, either
direct steam or a hot water heating system.

CAR HOUSE DESIGN

105

(c) Isolated waiting rooms, only practicable to heat if attendedgenerally,standard coal stoves are advisable. If waiting room is large
and retentious, steam or hot water heating system may be used.
( d y Unisolated waiting rooms. Direct steam or hot water system
is general)y advisable.
( e ) Small isolated ticket booths and the like may be satisfactorily
heated with the ordinary electric car heater.
Lighting. The pit room should be well lighted, but a proper disposition of lights will make their number comparatively small.
One satisfactory and economical arrangement is to place lamps on
one side of pit about 18 it. on centers. Each lamp is portable and
enables the inspector to place i t close to any desired point, and a
hook attached to the lamp allows it to be hung up. Sufficient
light should be provided in those places where trolleys are usually
shifted. For the sake of the eyes, light should not be stinted in the
reading, report and other rooms attached to the lobbies. All
interior winng should be installed in conduits.
W i g . Rule 41 of the National Electric Code, reproduced in
Section B 2 ~ 3 - 2 3of the Manualof the Am. El. Ry. Eng. Assn., should
be consulted in connection with the installation of electric wiring
in car houses and repair shops.
Road and Lie Departments. When quarters for the road and
line departments are to be furnished, they should be placed so as to
have yard roomadjoining. The headquarters themselves should be
of ample size to take the more or less bulky materials that are used.
Proper provisions should also be made for tbe locking up of copper
and other valuable portions of the stock.
Express Accommodations. Companies doing or contemplating
doing a n express business may wish to establish a terminal a t the
car house. I n such cases the room should be so situated that the
express cars and teams receiving and delivering shall not interfere
with the operation of the regular cars, and will not be in the way of
future improvements of the house. I t should also be placed so that
the clerk in charge could, if he had spare time, do other duties connected with the general routine.

106

ELECTRIC RAILWAY HANDBOOK

REPAIR SHOP DESIGN

The Electric Railway Journal, in 1912, made a compilation of
figures from ten electric railway repair shops, showing ground and
building areas in relation to number of cars owned, and the relative
area of the various shop departments. Those figures and the
accompanying discussion were reproduced in the first edition (1915)
of the Electric Railway Handbook. The data was later taken up
by the Buildings and Structures Committee of the Am. El. Ry. Eng.
Assn , revised and enlarged, and, based upon it, the committee
proposed, in 1922, two typical shop layouts, one for 150 and the
other for 250 cars, which, the committee stated, represent the best
modern practice. Much of the following data are from the report
of that committee.
The following table shows the relation of total shop areas to
the number of cars for which each of 15 shops were designed, as well
as the two typical shop layouts proposed by the 1922 Committee.

Montreal Tramway-Youvllle
M~lwaukeeElectrnc CO.-~llwaukee.
New York State Rys.--Syracuse
Pacific Electnc Co.-Torrance
U n ~ o nTractron Co.-Anderson
Unlted Rys. & E. Co.-Balt~more
Blrmlngham Ry.. L & P. Co.-B~rm~ngham
Georg~aRy. & Pr Co -Atlanta
York Rallways Co -York

Knoxville Ry & Lt. Co -Kno~v~lle.

Lou~svllleRallway Co.-Lou~svllle
V~rglnlaRy & Pr. Co.-Norfolk
Dallas Ry. Co -Dallas
Newport News & Hampton-Hampton
Portland Ry. & Lt Co.-Portland
Average. 15 shops

A.E.R.E A. typlcal 150 car shop

A.E.R.E A. t y p ~ c a l250 car shop

'Thts tncludes a large number of

I
fretght cars.

Size of Buildings. The actual areas of the building or buildings

required for housing the various departments of the repair shop of
any electric railway are, of course, subject to a number of obscure
factors. Probably the easiest way to arrive a t a basis for comparison is to consider the areas in square feet per car served by the shop.

SIZE O F REPAIR SHOP

8dv

I t is granted that this is an exceedingly rough, approximate method,

but, in view of the fact that there is necessarily a certain relationship between size of shop and number of cars, the above table was
prepared to show the range through which this unit varies. I n
determining the areas given, only bulldings have been included.
No yard space of any kind is considered, and covered transfer
tables where they exist have been omitted.
'The preceding table indicates t h a t 2 0 0 square feet per car is
a fair average floor area, with 45-ft. cars. The same result may
be arrived a t in a somewhat different manner by assuming that
each car in the shop occupies a spaw 60 It. long and 16 ft. widc.
The length of 60 ft. is based on a 45-ft. car body with a 7-ft. 6-in
space a t each end for passageway, trucks and waste room. The
width is assumed arbitrarily from the track spacing existing in
many shops. This gives a total area of 960 sq ft required for
each car which stands in the shop. However, the whole of the shop
is not devoted to housing car bodies; the table on page r q shows
that only approximately half of the shop is used for housing cars
and the other half for machinery and floor work. If each car
requires 960 sq. ft. of space in which to stand, the total area of shop
for each car contained in it will be twice that, or approximately
1900 sq. ft. Then, if ro per cent of the equipment is held in the
shop a t any one time for repairs, accidents o r rebuilding, ro per
cent of 1900 sq ft. of shop area will have to be provided for each
car owned. This amounts to 190 sq. ft. of shop per car, which
checks fairly well with the figures indicated in the preceding table.
This is undoubtedly the most accurate method for arriving a t the
proper size of shop for any electric railway, as i t permits allowance
for the requirements of variations in length of car body as well as
for unusual conditions which might necessitate frequent shoppings
due to numerous accidents or excessive wear. The latter allowance
may be made by using judgment in selecting the expected percentage of cars held in the shop. I n crowded city streets, for instance,
where accidents are common, i t would not be unreasonable to
assume that 1 2 per cent of the cars owned might be held in the shop.
On the other hand, where equipment is first class and conditions
good, it is safe to reduce the figure to 8 per cent. This method can
be expreswd by the formula:
Shop area in square feet = 2 X L X S X C X P
where L = length of average car plus 15 ft.
S = track spacing (generally 16 ft )
C = number of cars owned by the railway
P = maximum percentage of cars held out of service in the
shop (generally ro per cent).
Ground Areas. I n the foregoing only the areas of buildings or
floor areas of the various departments have been considered. I t is,
of course, undesirable to build a shop upon a plot of ground of the
exact size of the building. There are many classes of material, such
as wheels, iron castings, lumber, scrap and the like, which can be
stored conveniently outside of the shop building, and this is undoubtedly better than to have them occupying valuable space under

'

EXECTRTC RAT1,WAY HANDBOOK

108

a roof, or to have them kept in a s t o r e y a ~ d

a t some distance from the
shop. I n addition to thls ~t is not unusual to make use of the shops
for storing and repairing track matenal so that a need exists lor a
plot of ground considerably larger than the shop buildings. Unfortunately no set relation appears to exist between building and
ground areas. However, the following table has been prepared to
give a conception of present practice in regard to the total area of
ground required. The figures have been worked out in three ways,
each column showing the same area expressed in different form

AREASOF SHOPGROUNDS
Area of

Mtlwaukee..
Balttmore
Seattle
Rochester

Anderson .
Mlnneapolls.
Syracuse
Chtcago

Sep-rate bulldlngs
Basement area not
included. N o
transfer table.
Separate bulldlngs.
Includes track maBasement area not
~n~luded.

I t w11l be noticed from this table that a marked difference in area

of ground required exists between those shops whlch are composed
of separate build~ngsand those in which all departments are grouped
under one roof or are located in two or three large build~ngsseparated only by narrow transfer tables. The former arrangement
offers a much greater protection against disastrous fires, permits
better natural lighting facilities and, especially for large shops, gives
an opportunity for greater flexibility in des~gn. However, it
appears that the spaces between the different buildings are not
needed aside from their value for fire protection, and that separating
departments in d~fferent buildings necessitates a considerably
greater ground area for the shops. The low figure of o 31 acre per
roo cars owned which is shown for Rochester is due to the fact
that this shop has a basement extending under the entire shop
building. This in effect makes a two-story shop and, as has been
mentioned before, gives the shop a n unusually large storeroom.
The area outside of the shop usually allowed for heavy storage has
in this case been put into a basement. The same arrangement
obtains in a limited degree a t Syracuse, w h e r e C y t of the shop
building is provided with a storage basement.
icago has a layout characterized by decidedly restricted building area in consideration of the total number of cars owned, so that it IS evident that the
minimum ground area required for a complete set of shops is somewhere between o 49 acre per IOO cars owned, as a t Chicago, and I 83
acres, as at Anderson. T h e latter shop, as shown in Fig. 39, is

109

RJiLA?'IVE SIIOP AREAS

gs$

N b

2-g
v,t-

r-m

-n

m * o
n v , -to

N "

G
0

vlm

v,ol

mm

*n

~n

* - a

mm
r n

a o
ns.
--

vl

C)

cl

-0

N -

o m

rr.

o r

Nv)

-0

0 -

0 0

.oo
q

-Cd

-- 53
-0

ro

O*;

ELECTRIC RAILWAY HANDBOOK

WASH

IRON
RACX

"---GO'---

ERECTING

( FUTURE
PIG.31.-Am.

EXTENSION

El. Ry. Assn. typical rhop for aso urn.

.-i

,I-

UMBER
I

112

ELECTRIC RAILWAY HANDBOOK

located on a triangular plot
of 'ground which is a t best
dficult to utilize. In round
numbers, it is probable that
r acre per loo cars owned is
sufficient, or a t least desirable, for a compactly designed shop. When separate
buildings are used for the
different departments, and
especially where considerable
quantities of track material
are stored, 2 acres per roo
cars owned appears to be the
least ground area which
should be provided.
Examinations of the shop
arrangements shown in Figs.
3 2 to 41 disclose the somewhat unexpected condition
that only a minority of the
! t e n contain distinct provision for future extension.
I t is, of course, possible to
, extend any shop provided
that sufficient ground area is
provided around it, but in
most of the layouts shown an
extension could only be made
by shifting departments or
adding to the sides instead of
' the length of the buildings.
Relative Area of Shop
Departments. The I 5 shops
studied by the Am. El. Ry.
Eng. Assn. 1922 Committee
on Buildings and Structures
were used in the compilation
of the table shown on page
log, where the relative floor
area occupied by the vanous
departments is shown, expressed as percentages of the
whole shop area. It will be
noted that with few exce
tions the relative areas of t E
various shops agree reasonably well with the average,
and with those of the typical
shops proposed by the Committee, the plans of which are
shown in Figs. 30 and 31.

REPAIR SHOP DESIGN

Miscellan e o u s
Considerations i n

minimum c l e a r ance should be 2

feet; on curves i t is
well to increase
this to provide for
increase in the size
of cars.
The distance between track cen-

and i t should not

be less than 15 feet
inanyevent. This,

For a moderate

large shop, a combination of the two

is often desirable,
as it gives additional flexibility
and speed in hand-

113

114

ELECTRIC RAILWAY HANDBOOK

ling cars through the shops.
For a small number of tracks,
the first cost of aladder track
will be lower,while for a large
number of tracks the transfer
table will cost less. I n designing a new shop layout, i t
will be well for the engineer
to make up. detailed estimates covenng both and
then make a study of the
advantages of each to ascertain which will be best
adapted to his particular
case. The advantages of
ladder tracks are: (a) Cars
can be more readily removed
from shop in case of fire;
( b ) do not have to depend on
one iece of apparatus to
h a n d i cars; (c) cost of operation is less; (d)
trucking over it;
maintenance for small number of tracks. The advantages of transfer tables are:
(a) Flexibility; (b) less space
required; ( c ) lower maintenance for large number of
tracks; (d) less first cost for
large number of tracks.
Two types of transfer
tables are now in service,
namely, pit type and flush
type. A quest~onnairesent
out to representative companies indicates that the pit
type is less expensive and
more desirable.
Some companies now using
transfer tables have had
trouble with them due to
using old discarded motors
to drive them and on account
of improper design or construction. If the transfer
table is to be used, i t should
be carefully designed and
built rugged enough to stand
hard service. Whether the
transfer tables should be in
the open or under cover

(ecz::

REIJAIK SHOP DESIGN

depends on climatic
conditions. I n c o l d
climates where there is
much snow and ice i t is
well to have it under
cover. Where pit type
tables are used, suitable drainage should be
provided.
I t is important to
make use of a number
of labor saving devices,
and in the design of the
buildings this should be
taken into consideration. Prominent among
them are wheel d r o ~it
and pit jacks, jib &ne
with electric hoist for
truck repairs, overhead
trolley and travelers,
hood for blowing out
motors, p i t w h e e l
grinder, truck transfer
tables, transveyor or
portable crane, oxyacetylene cutting and
welding o u t f i t , a r c
welder and pneumatic
tools.
Individual drive of
shop machinery has
more advantages over
belt drive, as far as
building construction
and natural lighting is
concerned. Especially
is building construction
affected when the machine tools are of heavy
type which mcansarigid
support for lineshafting.
I t is desirable to
locate the shops so as
t o reduce non-revenue
mileage and to obtain
a steam road connection. While the latter
is not absolutely essential, i t is a great convenience in unloading
car load lot materials.

115

+.

gc
?i

-5
91

9
V)

a'
C)

d
~ r ,

116

ELECTRIC RAILWAY HANDBOOK

FIG,38.-Minneapolis

FIG.39.-Andenon

shop arrangement.

shop arrangement.

REPAIR SHOP DESIGN

Basement Under Half of Burlding

FIG.40.-Syracuse shop arrangement.

FIG.41.-Montreal shop arrangement.

117

118

ELECTRIC R A E W A Y HANDBOOK

Of the methods of raising car bodies off trucks, electric hoists of

the screw type appear to be preferred as being the best adapted for
the purpose. The truck overhaul shop of the Am. El. Ry. Eng.
Assn. typical layout is accordingly so equipped. For large shops
electric traveling cranes can be used to advantage but for a moderate
size shop, the advantages gained are more than offset by the additional cost of the apparatus and building details.
In laying out the yard track arrangement, some provision should
be made so that cars may be readily turned.

SECTION 111

TRAZN MOVEMENT
Schedules. I n the preliminary determination of schedules for
a n electric railway the factors concerned bear directly on the
amount of travel between various points and the time intervals
during which such movements exist.
WhereS = schedule speed in miles per hour (including time for
stops)
H = headway in minutes
D = length of line in miles
N = number of cars if single cars be used, or train units
if more than one car be used
T = time in minutes occupied in running 1)etwcen terminals
during one single trip

Chart for Headway Calculations. If the schedule speed and

any one of the following factors-headway in seconds, headway
in feet, cars per mile-are given, the other factors may beobtained
quickly by reference to the chart developed by H. M. Wheeler

120

ELECTRIC RAILWAY HANDBOOK

(Fig. I ) . The curved line marked "cars per mile" is a hyperbola

and shows the relation between cars per mile, read on the horizontal
scale, and feet as shown on the vertical scale. The diagonal lines
converging a t the left and marked "distance between cars" show
for various headways in seconds (at the bottom of the chart),
the spacing of cars in feet, by the use of the diagonal line corresponding to the proper schedule speed.

PIG. 2.-Preliminary

interurban train schedule.

Graphical Train Schedules or Train Sheets. Many factors

entering into the proper construction and successful operation of railways are a t once apparent from the gra hical train schedule which
is plotted with time of day in hours anXminutes against distances
in feet or miles. I t is convenient if the coordinate drawn through
each hour division be made a heavy line on the coordinate paper
and if these sections be subdivided into sixths or twelfths, representing 10- or 5-minute intervals, respectively. It is customary
to designate the distance between stations in feet or miles, and
the location of any points of special engineering interest, such as
branch lines, railway crossings, city and township limits, etc. Slop
ing lines may then be drawn to represent the progress of a train from

TRAIN SHEETS

121

station to station, as shown in Fig. 2, reproduced from Harding's

"Electric Railway Engineering!'
The inclination between these
lines and the time axis depends upon and represents graphically the
schedule speed of the train. A chart made up of such straight lines
as shown In Fig. 2 is sufficiently accurate for a rough preliminary
study of traffic possibilities, power requirements and substation
locations, but before exact time tables can be adjusted and meeting
points determined, a very much more accurate and detailed

graphical train schedule must be drawn. Such a schedule involving

several different schedule speeds over various sections of road as
well as a representation of the time elapsed in making station stops
is shown in Fig. 3. Such a graphical schedule enables the engineer
to predetermine not only the number of cars necessary to maintain
a given schedule and the position of all such cars a t any moment,
but it also locates the meeting points a t the crossing of the schedule
lines, and when used in conjunction with the power curves of the
various cars, as described later, it aids in locating substations and
determining the average and maximum loads on substations and
power stations. Another illustration of the use of train sheets is

STOPS
shown in Fig. 4 drawn u p by W. Nelson Smith. In this case
some of the details of the electrical distribution system are also
shown. The particular schedules which have been illustrated
are relatively simple ones. With the addition of express and local
service and in some cases freight and mail trains, with the necessity
of meeting the schedules of trunk and branch lines, a graphical
schedule may become rather complicated, but by the use of a large
scale drawing such solutions are made with little di5culty.
Frequency of Stops. The following table shows the usual. frequency of stops as encountered in various classes of railway service:
Steam locomotive through service. . . .
Steam locomotive local service.. ......
Steam locomotive suburban service. . .
Electric interurban express.. .........

I
I
I
I

Electric interurban local.. ........... I

Electric suburban.................. ; I
City elevated or rapid transit.. ...... a
City surface lines.. ................. 5

stop in loo miles

stop in 20 miles
stop per mile
stop m 10 miles
stop in a miles
to a stop? per mile
to 3 stops per mile
to 10 stops per milc.

The location of stopping places in city streets is often determined

by the location of intersecting streets. Due to the important effect
of frequency of stops, both on schedule speed and power costs,
efforts should be made to spread stopping places as far as is consist e a t , i t L w . e ~ d r e ~ .~~.i~ebJiezvd~h~.thi~B~~-y~ir~P~~P~f.
.X;ll.bmet by a minimum s cing of 5 0 0 feet in congested business districts, 600 feet in c l o s e r built residence districts, and up to 1 2 0 0 feet
in suburban territory, depending upon the density of population. On
a street railway serving a number of cities and thickly settled suburban communitiesin eastern Massachusetts, a 27 per cent reduction in
stopping places resulted in a 1 2 per cent reduction in the actual
number of passenger stops made by cars.
Duration of stops varies approximately as follows for different
classes of railway service:
Through trains, steam.. ..................
Local trains. steam.. .....................
Interurban cars. electric.. ................
City rapid transit trains, electric.. .........
City surface cars. electric.. ...............

5 minutes

a minutes
KO to 30 seconds
10 to zo seconds
5 to 1 2 xconds.

The rate a t which passengers board and leave cars, together with
other minor conditions, practically determines the length of service
stops in street railway practice. The design of car and method of
fare collection is largely responsible for the rate of passenger movement, as is illustrated by Fig. 5. These curves show average stop
times as related to the number of passengers hoarding or alighting
(in whichever direction the maximum movement occurred) in different types of cars, and were made up by averaging a large number of
observations made in a number of cities.
Effect of Stops. The effects of frequency and duration of stops
are severe, both on schedule speed and energy consumption. The
general relation between maximum and schedule speed, as effected
by frequency of stops, is shown by Fig. 6 (Standard Hantlbook).
Figs. 7 to 10, inclusive, show more detailed information as appli-

124

ELECTRIC RAILWAY HANDBOOK

cable to a definite equi ment of a suburban type car with four

GE-203 motors, p i n . wkeels, 67/17 gears, weight with passengers
47,000 lb. The free running speed is about 31 m.p.h., line voltage
average 500, and 2 0 0 feet of coasting is allowed before application

FIG. 5.-Comparative passenger stop times.

Btcps per Mile

PIG. 6.-Relation

b e t w e maximurn and scheduled speed and stops Per

nule.

of brakes. Fig. 7 shows the effect of frequency and duration of

stops on schedule speed. Fig. 8 shows the additional time
required by one additional stop with various rates of acceleration
and braking, or conversely, the time which might be saved by the
elimination of one stop under such conditions. Fig. 9 shows the
effect of frequency of stops on total energy requirements, while
Fig. 10 shows the additional energy required by the addition of one
stop, or conversely, the energy which might be saved by the elimination of one stop.

1%

STOPS

Speed Limitations. Maximum speed may be limited for a given

equipment by the safe maximum armature speed (see tables, pp. 254
to 2 s 9 ) There also may be limitations applicable t o specific
locations, such as those due to grade crossings or other public high-

6 Becoud 8toym 8 Becond &tope 10 Becand atop# 12 8lCOnd &&OD# -

1.6 M.P. B.P. 8.

1.6 M.P. E . P . 8 .

BLopr

FIG.7.-Effect

Curve
Curie
Curie
Curve

B
B

Aaceleratlw,
Braking

per M i l e

of frequency and duration of stops on schedule speed.

Stop. ~ c MTle
r

PIC. 8.-Additional

time required for one additional stop. under various

condit~ons.

interference, those due t o track special work, and those due to

track curves. For track cunres with proper superelevation of the
outer rail, the safe maximum speed in miles per hour is usually
considered to be equal to the square root of the radius of the m e
in feet.

1%

ELECTRIC RrZJLWAY H A N D B N K

9.-Effect of frequency of stops and mtes of accelerat~onon energy per

car m ~ i e .

TRAIN RESTSTANCE

Train Resistance
Train resistance may be defined as the resultant of the forces,
exclusive of those which are evidenced by internal losses in motor
equipment, which oppose the motion of the train a t a constant
speed on a tangent level track in still air. For convenience train
resistance a t any speed is expressed as the number of pounds
tractive efiort a t the driving wheel treads necessary to keep.tBe
train moving a t that constant speed on tangent level track in still
air. From the results of tests by many investigators most of the
many train resistance formulas have been approximated, but
while many of these formulas have been shown to give values nearly
equal to those secured by test on certain equipment and track and
under certain conditions, i t is not safe to depend on any one formula
to give very close approximations in universal application. This is
because there are so many conditions affecting the result. The
variation of some factors, such as temperature change in the
bearings, may be comparatively slow, while that of others, such
as of air resistance, may be great from moment to moment during the operation of the train. One of the points of considerable
disagreement among the formulas for train resistance is the manner
in which train resistance varies with speed. At low speeds this is of
but little importance compared with other factors, but a t high speeds
i t becomes one of the most important factors in determining the
speed for economic operation.
The more recent train resistance formulas recognize three
principal ordinary components which are conveniently referred to
as ( I ) journal friction, (2) rolling friction, and (3) air resistance.
Journal Friction. Journal friction of the car wheel axles has
been found to depend upon the weight and speed of the train, and
as the weight of the train increases the journal friction becomes
of less importance per unit weight of train until i t reaches a lower
limiting value of about 3.5 lb. per ton weight of train. Lower
values of journal friction have been secured in tests. The 1go4
proceedings of the Am. Ry. Eng. and Maint. of Way Assn. states
that a coal car with qa to 50 tons of coal will not have a journal
friction of more than 2.5 Ib. per ton, while the same car empty will
have a journal friction of abont 5 Ib. per ton. As noted on page 143,
the condition of the lubricant has a considerable effect on journal
friction.
Rolling Friction. Rolling friction, including rail friction, increases with the speed of the train and depends upon the diameter
of the wheels, general design of the truck and the condition of
wheels and track. I t is greater as the wheel and rail surfaces are
more imperfect and increases with track irregularity and flexibility.
Track irregularity and flexibility are also important in setting up
oscillations and concussions, the damping of which consumes energy.
Jn this process flange friction is also increased. All of these effects
are so closely related that for ordinary work it is impracticable to
separate them.

128

ELECTRIC RAILWAY HANDROOK

Air Resistance. Tests to determine.the valuc of air resistance

have yielded the most widely differing results. I t s value depends
upon the speed of the train, the area of cross-section and lateral
area of the locomotive and each car, the shapes of the front and rear
ends of the train, and the nunlber of cars composing the train.
Effect of Individd Car Weight on Train Resistance. Train
resistance has been found to depend upon the weights of individual
cars making up the train, that is to say,
given two trains of the same total weight
but made up of a different number of cars,
the train resistance will be different for the
two. Tests indicate that the train of heavier cars will have the lesser train resistance,
and that this difference is not entirely due
to different air resistance. (See Journal
Friction, above, also Fig. 24 and Train Resistance of Freight Train, p. 136.)
Train Resistance Formulas. Two formulas for train resistance are in common
use as applicable to the operation of single
c a n and to trailer and multiple unit trains,
as well as to the usual locomotive and pasof senger car train, these being known as the
II,-Out,ine
projerted area conslderrd Armstrong and the Blood Formulas. AnIn alr resistance c a ~ c u ~ aother
formula, as develo d by the Unitlon.
venity of Illinois, is consigred more nearly
correct for a plication to freight trains. No
such definite formula has been deveLped for train resistance in tunnels or subways, although several tests have been made, the results
of which vary widely and are applicable only to conditions similar
to those of the tests.
Armstrong Formula. This formula is based on one of similar form
roposed by W. J. Davis after % series of tests on the Buffalo &
L i t p o r t Railway, and modified by A. H. Armstrong following
a number of later tests, including the Zossen tests in 1902-3,
Electric Railway Test Commission tests in IW-5, tests a t Schenectady in 1905-6, and by the University of Illinois in 1910, 1.914
and 1917. I t is suitable for application t o single c a n and to tra~ler,
multiple-unit or locomotive trains up to the usual limits of passenger
train operation. The formula for single cars is

where f = train resistance, pounas per ton weight of car

a = area of cross-section of car, square feet. This is the
area included within the outline of the car body and a
rectangle of a width equal to the average width of the
trucks and with its bottom line a t the heads of the rails.
(See Fig. I I)
b = constant component of journal friction
=

2 but limited to a minimum value of 3.5

*/w

TRAIN RESISTANCE

1%

c = coefbcient of variable component of journal friction

and rolling friction
= 0.03 for ordinary conditions and car weight; higher
values for cars weighing less than 30 tons, and for poor
track conditions; maximum, 0.07
d = air resistance coefficient
= 0.001j for cars with pointed ends of the extreme type
= 0.002 to 0.0025 for usual rounded end suburban car or
electric locomotive with rounded ends or sloping front
= 0.004 for cars with perfectly flat ends
S = speed, miles per hour
W = weight of car, tons (of 2 0 0 0 Ib.)
The formula for single cars under usual conditions is therefore

I t is assumed that air resistance is increased one-tenth by the

addition of each succeeding car in a train. For train operation,
therefore, add one-tenth to the air resistance term of the formula
(d$)

for each car after the first, provided all cars are of similar

cross-section. If cars vary considerably in cross-section, the total

air resistance term should be vaned accordingly.
For application of the Armstrong formula to cars of common
weight and trains of various numbers of cars, see Figs. 13, 1 5 , 17,
19 and 21. I n these charts the values assumed for the crosssectional areas correspond closely to actual practice, and are as
follows:
Weight of
car. tons
10

IS
20
2s

Area.

sq. It.

7s
83
88
93

30

100

40

113

50

110
I ao

60

Blood Fornula. The train resistance formula proposed by

John Balch Blood in his paper before the American Society of
Mechanical Engineers, in 1903, is as follows:

where f
E

train resistance, pounds per ton

coeaicient of slrdzng friction
= 7 for light electric cars
= 6 for medium weight electric cars
= 5 for heavy electric cars
= 4 for average passenger trains
= 3 for heavy freight trains
=
=

= cok,llit-icnt of rolling friction

= 0. I ,ifor l i ~ h ttr;tc!i con.truc.tion

= 0.12 lor 11e;rvy tr:lck c o n s t r u ~tion

= c.t)c.llicicnt of sitlc rcsisl;in,.c

= o.oo10 lor orclinnrily constructc~lcars

0.0014 for cars wit11 \~cstil,ulcs

coeflicient of hcad and stcrn rcsistanres
= o . r j for cars of smnll cross-section
= 0.30 for electric cars of medium cross-section
= 0.3 j for large electric or suburl);un tr:~ilis
= 0.40 for Inrgcst e\press trains
S = spcccl, miles per hour
I!' = weight of car, ton:. (of zobo \I).)
=

I n the use of thc Iilood formula, it is oftcn assurncd t11:lt tHe

corflit.icwts I.:, H , C and D vary directly with car weights; such
corfticients:~sused I)? many cngineersarc. shown I)y Fig. I z.

10

20

30
P
Ton8 \V\'slrbt of Car

SO

FIG. 12.-Cnrficients for Blood train reslstnnce formula.

For train operation, t h e formula as given above should be used for
the leatling car o r locomoti\~e, while for succeeding or trail c a p
the coeficient I), representing head and stcrn resistances, should
equal wro. Thc total train resistance, in pounds per total weigPt
of train, is determined by applying the formula t o each class (Or
weight) of cars in turn, taking such proportion of each result as tPe
weight of such car o r cars bears to that of the entire train, add
atllling these quantities. For instance, to find the train resistanie
ol a train made u p of t\vo j s t o n motor c a n with one 30-ton trailer
coupled between: total train weight = 130 tons; train resistance

TRAIN RI<SIST,\FU'CE

-Mile* per Boar

FIG. 13.-Train resistance, single car (Armstrong forrnuh).

M l ) t S per no01

FIG. 14.-Tnin resistance, single car (Bloorl formula).

132

E1,ECTRIC RAILWAY HANDBOOK

x ~ l per
o ~Eour

FIG.IS.-Train resistance, twc-car train (Armstrong formula).

133

T R A I N RESISTANCE

Mile. pcr Eour

PIG.18.-Train

'0

25
YUrsw-

50

resistance. three-car train (Blood formula).

75

25

100

Mile# pcr Eour

PIG.19.-Tram resistance, five-car train (Armstrong formula).

134

ELECTRIC RAILWAY ITANDBOOK

26

20

15

10

0
0

2.5

0.-Train

50
Milen per Honr

75

resistance, five-car train (Blood formula).

5" of
-

train resistance a t that speed for leading

130
car from p t o n car (Fig. 14)
30 of that for 3 e t o n trail car
130
50
(1;ig. 22)
- of that for so-ton trail car (Fig. 22).
I70
For applicaiion of the Blood formula to motor cars and trailers
of various weights, and to trains of various numbers of cars, see
Figs. 14, 16, 18, 20 and 22. The coefficients as shown by Fig.
12 have been used in the construction of these charts.
Train Resistance in Tunnels and Subways. The air friction
element of train resistance may be very materially increased when
the car or train is traveling in a tube, the extent of such increase
probably depending upon the relative cross-sectional areas of car
and tube, and the facilities for relief of differences in air pressure
due to the motion of the train. One authority has suggested that
a t any speed

135

TRAIN RESISTANCE

Pic. 21.-Trarn rcsrstancc. locomotive and train (Armstrong formula).

20

ls

st

r"

210

PI

PIG. 22.-Tram

25

50
uilen per Hour

resistance, trarl~ngcars

75

(Blood formula)

136

ELECTRIC RAILWAY HANDBOOK

open air train resistance values should be increased 2 5 per cent to

meet subway or tunnel conditions, but no such general rule can be
applied with assurance to the many differing designs of subways
and tunnels, some examples of which are shown in Figs. 46 to 61,

Milea por Hour

FIG. 23.-Train resistance in subways and tunnels.

I . Interborough Rapid Transit test-3-car train-I l o tons
Interborough Rap~dTransrt test-5-car train-180 tons
3. Interborough Rap~dTransit test-8-car train-300 tons
4. Interborough Raprd Transrt-~osar train-estimated from above
5. Boston-Cambridge subway test-3-car train-16s tons
6. Hudson-Manhattan tests In 4000 ft. t u b e 3 ton train-104 tons
7. Hudson-Manhattan tests in aooo ft. tube--?
ton tra,in-104 tons
8. Above tests modrfied to meet Brooklyn Raprd Translt subway and tunnel
conditions
2.

inclusive, Section I, pages 72 to 75, inclusive. I n the absence

of any general forniula such as those available for open air conditions, recourse must be had to the results of tests under conditions
as nearly as possible similar to those in hand, whenever such tests
have been or can be made. Fig. 2 3 shows the results of several
investigations on this subject.
Train Resistance of Freight Train. Due to the relatively small
cross-section of cars in proportion to length of train, the total air
resistance of freight trains is made up of proportionately more side
friction and less end effect than that of passenger trains. The
lower speeds further reduce the relative effect of air resistance, so
that bearing and rolling friction enter into freight train resistance
in a larger measure than into that of passenger trains. For such
reasons, the usual train resistance formulas are rarely a plied t o
long and slow speed freight trains. Fig. 24 (from ~ u l E t i nNo.
43 of the Univ. of 111. Eng. Exp. Sta.) shows the results of a series of
dynamometer car tests on freight trains, expressed as train resistance in pounds per ton. These are the most commonly used and
generally accepted a s most reliable data on train resistance of freight
trains, but i t must be remembered that as the tests were made by a
dynamometer car, the train resistance curves and formulas deduced
therefrom do not include the head on portion of air resistance. The

FREIGHT TRAIN RESISTANCE

137

14

g '2

,.E. lo

c-4

4
$ 8

a
.

dG

i; 4

, 1 41 1 8 1 11 2 1 1 81 1 a 1o ~n 1r z1 s1 s 1 n 1 a ~ a 1 1 ~ ~
Bpeee-M. P.B.

FIG. 24.-Train

resistance of freight trains.

following formulas are empirical, yielding values whose maximum

variation from those given by Fig. 24 is 0.5 per cent when used
within the speed limits of Fig. 24.
FREIGHTTRAINRESISTANCE
FORMULAS

++
++
W = 30 tons; f = 5.02 + 0.066 S + 0.00116 S*
W = 35 tons; f = 1 . 4 9 + 0.060 S + 0.00108 S*
W - 40 tons; f = 4.15 + 0.041 S $ 0.00134 S*
W = 4 5 t o n s j = 3.82 + 0.031 S + O . o o I ~ o ~ *
W = 50 tons: f = 3.56 + 0.024 S + 0.00140 S*
W = 55 tons; f = 3.38 + 0.016 S + 0.00142 S*
W = 60 tonsif = 3.19 + 0.016 S + o.00132 .Sf
W = 65 tons;f = 3.06 + 0.014 S + o.00130 S*
W = 70 tons; f = 2.92 + 0.021 S + 0.oo111 S*
W = 75 tons; j = 2.87 + 0.019 S + o.00113 St

When W = 15 tons. f = 7.15

0.085 S
0.0017s S*
W = 20 tons: f = 6.30
0.087 S
0.00126 S*
W = z5tons;f = 5 . 6 0 + 0 0 7 7 S + o . o 0 1 1 6 S *

The following formula gives a n approximation to the values of

train resistance given by Fig. 24 (the maximum difference being
9.5 per cent which occurs a t S = 2 1 and W = 55).

in which f = train resistance, pounds per ton

W = weight of car, tons
S = speed of train, miles per hour.
These curves (Fig. 24) may be applied to predict the probable
total train resistance of entire freight trains which are either homogeneous or mixed as regards individual car weights and which
have been in motion for some time, when the air temperature is
above 30 deg. F. and the velocity of the wind is not more than 20
miles per hour. Due to variation in make-up or external conditions,

EI.ECTRIC RAILWAY HANDBOOK

138

some trains may have a train resistance about 9 per cent in excess
of that given by Fig. 24, but this is of importance only in rating the
motive power for speeds under 15 miles per hour.
The above results were obtained from tests of 32 ordinary
freight trains in regular service of such make-up as naturally
resulted from the traffic conditions in the Champaign yards of the
Illinois Central Railroad. The chief characteristics of these trains
were as follows:
Total weight of train, tons.. .............
Average weight of cars composing the
tram, tons.. .......................
Number of cars in the train .............
Train length, feet. . . . . . . . . . . . . . . . . . . . .

Minimum

Maximum

747

2908

16.12

26

rzo

69.92
89
3480

The trains whose average car weights were less than 2 0 tons or
more than 60 tons were composed of c a n of nearly uniform weight;
while those whose average car weights were between 2 0 and 60
tons were either homogeneous or mixed as regards the weights of the
~ndividualcars. Presumably, the majority of the cars had journals
conforming to the specifications of the Master Car Builders' Association which for some years have required that the size of freight
car journals be.either 3% by 7 in., 4% by 8 in., 5 by 9 in.,or 5% by
10 In., depending upon the car capacity.
All the c a n had four
wheel trucks and it is safe to assume that all the car wheels were
33 in. in diameter.
The track on which these tests were made is on the Chicago
division of the main line of the Illinois Central Railroad. I t extends from Gilman to Mattoon, Ill., a distance of 91 miles. The
maximum grade against north-bound traffic was 29 ft. per mile,
and against south-bound traffic, 31.9 ft. per mile, and in the 91 miles
there was 7850 ft. of curved track. The track was well constructed and well maintained, and probably was such as one might
expect to find on main lines of first class railroads. About 94 per
cent of the track was of 85-lb. A.S.C.E. section rails laid in about
rgoo, and the remainder was of 75-lb. A.S.C.E. section rails laid
in 1894 and 1895. I t was laid on oak ties spaced 2 0 in. center to
center. About 83 of the 91 miles were ballasted with broken
limestone and the rest, which was in station grounds, was ballasted
with screenings or cinders. None of the data used in the construction of the curves was taken before the train had been in motion
a t least 10 miles. The speed during the tests ranged from 5 t o
35 miles per hour, the air temperature from 34 deg. F. to 82
deg. F. and the approximate average wind velocity during all
but one test was less than 2 0 miles per hour. The direction of
the wind relative to that of the track varied through 360 deg.
during the tests. Each train resistance was reduced to train resistance on level track by correcting for grade. The tests were
made by means of a dynamometer car and only take into consideration the train resistanceof the part of the train behind thelocomotive
tender.

Relation Between Vestibule Shapes and Train Resistance.

Fig.

25

gives approximate speed air resistance curves for a single

139

TRAIN RESISTANCE

PIG. 26.-Standard
vestibule.

PIG.27.-Parabolic

vestibule.

interurban car when tested with vestibules of various shapes at

speeds up to 60 miles per hour. These curves are from data given
in the report of the Electric Railway Test Commission, 1904, and
show the end air resistance which may be expected under ordinary
conditions of service on a level track, especially the variation in

140

+-

ELECTRIC RAILWAY HANDBOOK

resistance due to various shapes of car.ends. The tests on which

these curves are based were performed with a specially constructed
car which was operated on a tangent track of 80-lb. T-rails laid on
gravel-ballasted oak ties. The general method of making the tests
was to make determinations of speed, electromotive force applied
to the car and current taken by the car while running a t practically
constant s p e e d
From the average
---OW --4 k- 4 6 4 values of voltage
P"
and current and
the current e5ciency curve for the
motor equipment,
the power required
to balance train resistance a t that
speed was determined, and by
means of dynam*
meters t h k e n d
FIG 28.-Parabolic wedge vest~bule.
resistances w e r e
segregated.
The car body was of an interurban type 3 2 ft. long without
vestibules, the foundation was a pressed steel flat car of ~oo,ooo
Ib. capacity, the trucks were Baldwin locomotive, M.C.B. interurban type, the motors were four Westinghouse No. 85, the gear
ratio was 27 : 47. The total weight of the car was approximately
38 tons. The types ~f vestibules
--- -used consisted of a standard,"
"parabolic," "parabolic wedge"
and "flat." The dimensions of
these are shown by Figs. 26, 27,
2 8 and 29, respectively.
Wid Pressure. Wind pressure ,i
on a body varies with the velocity of the wind and the shape of
the body presented to the wind.
For convenience i t is given in
pounds per square foot of area projected on a plane perpendicular
to the direction of the wind. 4- -PIG. 19.-Cross-sect~on of car.
(Wind used in this connection indlcates relative motion of air and
a body in the air.) The following formula has been found to show
wind pressures agreeing with experimental results:
P = dV2
in which P = pressure on a plane surface normal to the direction
of the wind, pounds per square foot
V = actual velocity of the wind, miles per hour. (See
Wind Velocity, p. 141.)
d = wind coefficient. (See p. 141.)

-7

fnj

I.

141

WIND PRESSURE

Wind Velocity. When wind velocity is tested by hemispherical

cup anemometer
V = log-l[o.~og 0.9012 log v]
in which V = true wind velocity
v = actual velocity of cup centers.
Velocity given out by the Weather Bureau is obtained on the
assumption that
= 30.
This gives a velocity higher tban the actual velocity. By the
following table the true wind velocity may be obtained from the
velocity given out by the Weather Bureau:

Reported velocity
10

10

30
40
SO
60
70
80
90
100

True velooty
9 6
17.8
1S.7
33.3
40.8
48.0
55.1
61.1

69.a
76.1

Wind Coefficientfor Plane Surface Normal to the Direction of

the Wmd. Following are values of wind coefficient determined
by various methods:
Authonty
Wind coefficient (d
Weather Bureau.. ............................. o 004
W. J. Davls ................................... o 004
Mart~n........................................
o 004
Longley ..................................... o 0036
Srneaton . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.005

The value of 0.004 is robably the closest approximation.

Wind Pressure o n 0 & e r than Plane Surfaces. (From a paper
by Professor Kernot before the Australasian Association for the
Advancement of Science, 1893.) The following determinations
were made by placing small models in an approximately steady
jet of 10 in. by 12 in. cross-section.
M o d d w . At any given wind velocity, the ratio of the wind
pressure on a given body to the pressure on a plane surface normal
to the direction of the wind and having an area of cross-section equal
to the area of projection of the body on a plane normal to the direction of the wind is, for convenience, called the modztlrs.
Cube. The pressure on a cube was as nearly as could be measured
the same whether the direction of the wind was parallel to a sideora
diagonal and was 0.9 the pressure on a square card equal in size to a
face of the cube.
Rectangular Blocks. r = length measured in the direction of
the wind, y and z = dimensions in other directions.
Where x
Where x
Where y
Where y

= ay = zr

= 3y = 32
= zx = zs
= 3x = j e

. . . . . . .
..............
................
....................

modulus
modulus
modulus
modulus

= 0.8
=

0.7

= 0.9
= 0.9

142

ELECTRIC RAILWAY HANDBOOK

A block re resenting a tower and having a height equal to three

times its wi&h of base gave a modulus of 0.9 when the direction
of the wind was normal to one face. When the direction of t h e
wind was the same as that of a diagonal the effect was, as nearly a s
could be measured, the same.
Pyramid. A pyramid of square base, having a height equal t o
about three times its base gave a modulus of 0.8 when a side was
presented to the wind. When one angle was presented to the wind
the total pressure was increased by 25 per cent.
Cylinder. Cylinders having the elements of their curved surfaces normal to the direction of the wind gave a modulus of 0.52.
Octagon91 Prism. The rtqsure on a n octagonal prism was 10
per cent greater than on tKe circumscribing cylinder.
Cone. For a cone having a height equal to three times the diameter of the base, the modulus was o so.
Sphere. For a sphere, the mod~lluswas 0.36.
Hemispherical Cup. For a hemispherical cup (such as is used
on Robinson's anemometer): when the convexity was to the wind
the modulus was 0.36; when the concavity was to the wind the
modulus was I I S .
Retaining Surjaces. When a surface parallel to the direction of
the wind was brought nearly into contact with a cylinder or sphere,
the pressure on the latter bodies was increased by about 2 0 per cent,
owing to the checking of the lateral escape of air.
Sheltering Surfaces. When a 9-in. disk was used as a sheltering
surface and a 6-in. disk was placed 2 in. in front of it, the latter
received only two-thirds the pressure it endured if the larger disk
was removed. This reduction in pressure was perceptible, though
to a less extent, a t all distances up to 9 in.
Temperature Effects on Train Resistance. Train resistance
increases with an increase of journal lubricant viscosity; thus when

FIG. 30 -Decrease

Mile8 Dfatmce from start

of fre~ghttrain resistance as train progresses (Univ. 111.
tests)

the vls~osityof the journal lubricant is increased by the lowering

of atmospheric temperature an increased train resistance results.
Tests have indicated that a t the atmospheric temperatures of ordinary railway operation, train resistance decreases as the journal
lubricant becomes less viscous with rising journal temperature,

T R A I N RESISTANCE

143

other conditions being constant. When a constant journal temperature is reached on a moving train, the train resistance is at a
minimum for the then prevailing track conditions, speed of train
and temperature of the atmosphere. Figs. 30 and 31,from Bulletin
No. 59 of the University of Illinois Experiment Station, were
plotted from data secured from 32 tests. The tests were made
in 1910 with trains on the Illinois Ccntral Railroad as they
came in regular service Fig 30 indicates that the mean resistance
a t speeds of from 1 2 to 35 miles per
hour, and atmospheric temperatures
from 30 to 42 deg. F., became a
minimum when the train had been
in motion for about 35 miles. I t B
was found that in warm weather %
the minimum train resistance for a i14
similar train was reached when the' rs'
train had been in motion from 8
to ro miles. Fig. 31 gives a com- i;
parison of minimum train resistance ;12
values for the two atmospheric tem- 3
peratures and shows that the min- 5
imum train resistance in cold ?
weather in approximately 25 per Fl0
cent greater than in warm weather.
er TC*Q.
Fig. 32, as the result of a large
35 Mile. rom S t a r t
number of observations by A. W. 2
Baumnarten. Electrical Eneineer of 4
the ~Xicagd and Joliet -~lectric
Railway (El. Ry Jour ,1922),shows
6
a typical example of the variation
l6
in energy consumption which folMilea porEr.
lows variation in temperature. T h e
of temperature
tests previously quoted were made on train3r,-EBcct
rcslstance (Unlv. of I!].
on freight trains, while those on tests).
which Fig. 32 is based were made
in single passenger car operation, and show an energy requirement a t
freezing temperature about I z per cent greater than a t summer temperature. M. B. Rosevear, Superintendent of Distribution, Public
Service Railway (El. Ry. Jour., 1918),shows double this increase
between the same temperatures, but, as he states, his cars were
electrically heatcd, the heaters requiring more than 10per cent of
the total enerEy
-- used by the cars; this explains the apparent
-discrepancy.
Starting Resistance. All of the train resistance formulas which
have been given herein indicate a minimum value for train resistance a t the minimum or zero speed. I t is a fact, however, that a t
the instant of starting the tract~veeffort requ~redto start the car
or train from rest is very mnsiderably greater than that needed to
keep it in motion a t low speed. Such " friction of rest" or " starting
resistance" is of importance only in determining the peak power
requirement of starting, and in the design of starting resistors for
control. Tests made on a single car a t Purdue University (El.

144

ELECTRIC RAILWAY HANDBOOK

Ry. Jour., 1 ~ 1 5 )showed a train resistance a t the instant of starting

equal to four times that a t 5 miles per hour; this was reduced one
half as soon as the car began to move. I t is not unusual that the
starting resistance amounts to 4 0 pounds per ton, which with
fair to good conditions may be taken as the maximum; the minimum
probably is about 10 pounds per ton. These limits of 10 and 40
pounds per ton are recommended for use in the 1921 Manual of the
A m Ry. Eng. Assn., depending on loading, temperature, character
and condition of the track and train.

Prc. 32.-Variation of energy consumption with temperature.

Curve Resistance. (See pages 55 and 58 for designation and

measurement of track curves.) Curve resistance may be defined
as the resultant force, due to track curvature, which opposes the
motion of a train a t a constant speed on a curved track. I t is
equal to the tractive effort a t the driving wheel treads necessary to
keep a train moving a t a constant speed on a circular curve of level
track in still air, in excess of that necessary to keep the train moving
a t the same speed on tangent level track of the same character in
still air. Curve resistance is commonly expressed in pounds per
ton weight of train per degree of track curvature. For convenience
i t is also often expressed in terms of grade, that is, in terms of that
per cent grade whose grade resistance would be equal to the curve
resistance under consideration. Many values of curve resistance have been determined by experiment, hut the variables
concerned are such that no rational method for the general calculation of it has yet been determined. I n ordinary cases, the value of
curve resistance is small compared with train resistance and grade
resistance; also, but a small portion of the total track of any usual

CURVE RESISTANCE

145

railway is curved; therefore, in ordinary cases an accurate deternonation of curve resistance is of minor importance. Curve resistance
is due to the increased slippage between wheel (tread and flange)
and rail and increased friction in the moving parts of the train on the
curve. I t s value depends primarily on the coefficient of friction
between wheel and rail, length of truck wheel base, gage, flexibility
and condition of the track and the condition of the rail surface.
The value of curve resistance on moderate curves is nearly proportional to the degree of curvature. On curves of short radius it is
less per degree of curvature than on curves of greater radius.
On curves of very short radius, such as are usua!ly found in city
streets, conditions are such that no definite statement of the value
of curve resistance can be given.
Grade Compensation. I n order that the combined up grade
effect and curve resistance on a curve may be equal to the
grade effect on tangent track having a grade equal to that originally on the curve, the grade is often reduced by the amount
whose grade effect would be equal to the curve resistance. This
reduction of grade on curves is called "grade compensation."
Determination of C w e Resistance (and Grade Compensation).
The Committee on Economics of Railway Location ( 1 ~ 1 0 )of the
Am. Ry. Eng. and Maint. of Way Assn. reported tests a t North
Mountain Cut-off and Mt. Airy Grade on the Baltimore and Ohio
R. R. to determine the effect of curve compensation. The trains
used in these tests were made up of locomotive, dynamometer
car, 30 and 36 steel hopper cars (empty and loaded) and
caboose. On portions of the grade compensated a t the rate of 0.03
per cent per degree of curvature the combined up grade effect and
curve resistance was greater than on tangent track, while on
portions compensated a t the rate of 0.04 per cent per degree of track
curvature the combined up grade effect and curve resistance was
less than on tangent track. Assuming that the mean of these rates,
0.035 per cent per degree of track curvature, was the correct rate of
compensation, the curve resistance was 0.7 Ib. per ton weight of
train per degree of track curvature, and this is the coefficient most
generally used in curve resistance calculations,althoughitisincreased
by some engineers to as much as 1.0 pound per ton per degree.
The 1921 Manual of the Am. Ry. Eng. Assn. recommends that
grades be compensated 0.03 to 0.05 per cent per degree of track
curvature, depending upon relative lengths of curve and train,
upon location of curve with res cct to beginning of grade, upon
superelevation of outer rail, an$' upon train speeds. For electric
railway conditions, the following formula may be used:
where

c = 0.7 D
c = curve resistance, pounds per ton (of
D = track curvature, degrees

2000

Ib.)

This is on the assumption that each degree of track curvature is the

equivalent of 0.035 per cent of up grade. I t should be noted that
this formula does not apply to short radius curves with no superelevation of the outer rail such as those found in city streets and usually
.rated as "special work."

146

ELECTRIC RAZLWAY HANDBOOK

Grade Effect. The component of force along a line parallel

to the center line of the track, due to the action of gravity on a train,
is called grade effect. Grade effect is commonly expressed in
pounds per ton weight of train. If the direction of grade effect
is opposite to the direction of motion of the train (that is, if the
train is on a n up grade) the tractive effort will be decreased by the
value of the grade effect. If the direction of grade effect is
the same as the direction of motion of the train (that is, if the
train is on a down grade) the tractive effort will be increased by the
value of the grade effect.
Approximate Value of Grade Effect. The following formula
will give results slinhtly in excess of the true values, but close
enough for all ordinarv-traction work. The error is greater with
heavier grades, but reaches only 0.5 per cent for a 10 pe;cent grade.
G=zon
in which G = grade effect, pounds per ton (of 2 0 0 0 lb.)
n = number of per cent of track grade.
This formula may be stated in the form of a rule: Grade effect
is equal to 2 0 Ib.pfr ton weight of train per per cent of grade.
Actual Value o Grade Effect. The precise value of grade
effect may be expressed by the formula

2000

~'lo,-

+ n2

in which the symbols are the same as in the preceding paragraph.

I t may be noted that the above rule for obtaining the
approximate value of grade effect yields the precise value of grade
effect if the per cent grade used in applying the rule is the approrimate per cent grade obtained according to "Approximate Value of
Grade," below.
Actual Grade. The actual grade of a track is its rate of uniform
rise or fall. I t is equal to the difference in elevation between two
adjacent points on the grade divided by the horizontal distance
between them, i.e., the tangent of the angle of inclination of the
track to the horizontal. Grade is generally expressed in per cent.
Approximate Value of Grade. As it is easier to measure the distance along the track than along the horizontal, this substitution is
often made. This introduces no serious error in the case of the
small angles of inclination encountered in the usual railroad work
as the sine thus substituted is practically equal to the tangent.
Per Cent of Grade Determined by the Use of Tape Line. Fig. 33
(by G. hl. Eaton, Elec. Jotirnal, 1911) shows a method of
finding the per cent of grade by use of a tape line. The tape is
reefed through the end ring or eye, and a pin is stuck through the
tape a t a point beyond the 12-ft. mark such that when the pin
just clears the ring the I 2-ft. mark will lie in the center plane of the
ring. The tape is creasedat the 4-ft. and 9-ft. marks and is fastened
to the reel in such a way that the latter may be used as a plumb
bob. If possible. the ballast is scooped out to receive the reel
hangin:! ireely below the rail, as a t D. The tape is tied to the rail
a t ,I,the 4-ft. mark, and is suspended by a string or wire a t B
with the portion A B drawn straight. The portion of the tape

TRACK GRADES

147

from the ring to A is drawn straight and the suspension of B and

C are adjusted so that suspended tape and pin clear the ring. The
number of per cent grade is then approximately equal to z x in which
z is measured in inches. T h a t is, every half inch between the 1 2-ft.
mark and the surface of the rail a t G indicates approximately I per
cent of track grade.
Average Grade. The average grade of a section of track connecting two given points is the ratio of the difference in elevation
of these two points to their distance apart, measured in the same
unit. I t is usually expressed in per cent.

PIG.33.-Method

of finding approximate value of grade by use of tape line.

Ruling Grade. The ruling grade of a road or section of road is

that grade which limits the weight and length of train to be propelled over that road or section by a given motive power. I t s
value is fixed by the traffic and economic conditions encountered
on the particular section under consideration. The maximum
weight of train to be propelled by a given motive power and
limited by the ruling grade is that weight of train which the given
motive power would move up a continuous grade equal in per cent
to the ruling grade, a t some predetermined constant speed. The
grade of the greatest per cent on a section need not always be the
ruling grade for that section. The length of the grade, the rates
of those grades preceding it, and the speed of the train on entering
the grade are of as much importance as the rate of the grade, because
a grade which may be approached a t considerable speed by a given
train may be easily passed over while a longer grade of much lower
rate may be sufficient to stall the same train.
Momentum Grade. A grade whose operation is made possible
by the momentum of a train which approaches the grade a t a
considerable velocity is called a momentum grade (or velocity
grade). Grades of much greater rate than that of the ruling grade

ELECTRIC RAILWAY HANDBOOK

may be operated in this way in order to-save the expense of their
reduction. The argument against the very extensive use of the
momentum grade is the possibility of some unforeseen condition
such as the presence of a new crossing or a new stopping place
which would so reduce the speed of the train that the momentum
of the train would be insufficient to help it over the grade. The
possibility that such a condition will arise to limit train movement
IS more remote on an old road where stopping places are fairly
well established than on a new road in a rapidly growing community. This point is of importance in fixing the ruling grade for
a new road or in reducing the rate on a n old one. I n some cases the
necessity for reducing or removing a grade may be foreseen, but
the time of its occurrence may be so remote that i t will pay to
delay this reduction or removal till the change is made necessary
by developing conditions.
Virtual Profile. The electric railway engineer usually finds i t
convenient to study the relation of grades,-trains and equipment
together directly on the speed-time and distance-time curves (see
Run Curves) for the train and equipment operating in the required
direction on a typical run between whose limits lies the particular
section under investigation. Unless the train is o erated a t constant speed (as by three-phase induction motorsr care must be
exercised that the speed of the train shall become neither dangerously high a t the foot of a grade nor so low a t the top of a grade
that there is a probability that the train may be stalled by a strong
head wind or adverse conditions of track, or both. As a n aid in
adjusting speeds, locating points where power should be cut off,
brakes applied, or regeneration resorted to, or in studying the
shortening or reduction of grades for the satisfactory operation of
a given equipment and train weight over a given section, a virtual
profile is sometimes plotted. The virtual profile depends upon
the following six items: ( I ) actual profile of the section considered,
(2) track curves in the section, ( 3 ) equipment (motors and gears),
(4) train weight, (5) direction of motion on the given section,
(6) known speed a t the beginning of the section. I t is only possible
to construct a virtual profile for a section a t whose beginning the
speed of the train is zero or some other definitely known value.
Construction of the Virtual Profile. T o construct the virtual
profile, the actual profile for the track section is first plotted with
elevations as ordinates; then the virtual profile is the locus of all
points, the ordinate of each of which is equal to the sum of the
nding ordinate of the actual profile and the kinetic energy
rZde5Ce p. 1 ~ 9 (to
) the same scale) of the given equipment and
train, when a t that location and traveling in the direct~onconsidered. A virtual profile thus constructed will generally be curved.
For most purposes, however, it is sufficient to plot only the points of
the virtual profile corresponding to the ends of each grade (and
curve) and connect the successive points thus plotted by straight
lines (see Fig. 34). The virtual profile for a given train and section
of track touches the actual profile where the train is a t rest, it
diverges from the actual profile where the train is accelerating
(velocity increasing), i t is parallel to the actual profile where the

KINETIC ENERGY HEAD

149

velocity of the train is constant, and it converges toward the actual

profile where the train is retarding.
Kinetic Energy Head. The kinetic energy head (also known as
the velocity head) for a train a t any speed is the height to which the
kinetic energy of the train a t that speed would 11ft the train against
the force of gravity alone. I t is equal to the height through which
the train acted on by the force of gravity alone would have to fall
from rest to acquire a kinetic energy equal to that of the train a t the
particular velocity considered. With the falling mass there is no
effect of rotating parts, therefore in order to acquire an amount of
energy equal to that of the train a t any instant a mass equal to that
of the train and acted upon by the force of gravity alone must fall

FIG.34.-Sample vlrtual profile.

through a greater height than would be necessary in acquiring a
speed equal to that of the train a t that instant. The following
derivation of the formula for kinetic energy head is further explanatory of the term.
mu2
mo12
At any instant E = - = 2k

in which

E = kinetic energy of train a t speed S

m = mass of train
v = speed of train, feet per second
k = ratio of linear inertia to total inertia of
train (see age I 50)
ol = velocity orthe falling mass m, feet per second

therefore
ot2 =

f
which
k

a body falling from rest) gives

vz =
=-

( ? ? *.)

2gk

in which

h = kinetic energy head, feet

= weed of train miles Der hour
k = &me as above.

31=

when substituted in h = - (for

2E

150

ELECTRIC RAILWAY II.4NDl3OOK

Acceleration
Linear Acceleration. When acceleration is expressed in miles per
hour per second, and the force producing it in pounds per ton, the
familiar formula for acceleration becomes

or
A = o.orog81
where A = rate of linear acceleration, miles per hour per second
I = force required for such acceleration, pounds per ton
(of 2 0 0 0 Ib.).
Ratio of Force Required for Linear Acceleration to Force
Required for Total Acceleration. As the total e5ective inertia of
rotating parts is greater than their inertia when moved as a mass in a
straight line, allowance must be made for this fact in the consideration of train acceleration. This is particularly true in the case of
electric motor cars the geared armatures of which rotate a t considerably higher speed than the car wheels. The ratio of the "linear
inertia-to the "total inertia," including that of rotating parts, may
be expressed by K, in the expression:

in wbicb

force required for linear acceleration, pounds

per ton weight of train
F = total force required to accelerate the train,
pounds per ton weight of train
K = ratio of linear inertia to total inertia of train.
I = - - - -W
K =F
r,,R.e
=

in which

+ nWww($)' +

W = weight of car complete, car or locomotive

W , = weight of car wheel
W. = weight of motor armature
n, = number of car wheels
no = number of motor armatures
R, = radius of car wheel a t tread
R. = radius of motor a r m t u r e
r, = radius of gyration, car wheel
r. = radius of gyration, motor armature
e = gear ratio
F = total effective inertia ( = I i)
I = linear inertia
i = rotational inertia
= approximately 0.6 for the avenge car wheel

-+

(:)I

(2)'

approrimateIy 0.5 for the average railway

armature.

Substituting these latter two values the formula becomes:

The value of K for a train made upof units (motor cars, motor cars
and trail cars, or locomotives and trail cars) is equal to the sum of
the values obtained by multiplying the value of K for each individual
motor car, locomotive, or trail car by the ratio of the weight of
that particular car or locomotive to the weight of the whole train.
Ratio of Linear Inertia to Total Inertia in Practice.
For electric locomotive and heavy freight train,
K = 0 95
For electric locomotive and high-speed passenger train, K = 0.935
For high-speed electric motor car,
K = 0.935 to 0.91
K = 0.91 to 0.85.
For low-speed electric motor car,
Formula for Train Acceleration. The formula for the linear
acceleration of a car or train including rotating parts then becomes
A = 0.01og8 K F (symbols as before)
Inasmuch as an average value of K for electric motor cars is
about 91.1 (which makes o . o r q 8 K = o.or), it is common practice
to assume such a value, and for approximate u-ork to use the
acceleration formula as
A = 0.01 F
The force producing acceleration is made up of the tractive effort
of the motors plus the effect of down grade and reduced by the cornbined effect of train resistance, curve resistance, up grade effect
and braking, or such of those elements as may exist a t the moment.
Inasmuch as the constant in the acceleration formula has heen
deduced with the accelerating force F expressed in pounds per ton
weight of train, each of its elementsshould he used in the same terms.
As a formula, the above becomes
F=P-f-c+G-B
and the complete formula for train acceleration is
A =o.o1q8K(P-f-cf
G-3)
or for approximate work
A =o.o~(P-f-cf G-B)
where A = rate of acceleration, miles per hour per second
K = ratio of linear inertia to total inertia
F = net force for acceleration, pounds per ton (of 2 0 0 0 Ib.)
P = tractive effort of motors, pounds per ton
Tti
-W
T = tractive effort each motor, pounds
n = number of motors each developing T
W = weight of train, tons (of 2Ib.)
f = train resistance, pounds per ton (see pp. 1 2 7 to 144)
G = curve resistance, pounds per ton (see p. 144)
G = grade effect, pounds per ton (see p. 146)
= positive on down grades, negative on up grades
B = braking effort, pounds per ton

152

ELECTRIC R A L W A Y HANDBOOK

Motor tractive effort (P) will be present as a definite value in the

formula while power is being used; i t equals zero when power is off,
as during coasting; it may be obtained through T from the motor
characteristic curve, and will vary in amount depending upon the
current or speed. Train resistance (f) always will be present in the
formula; it will vary in amount for a given equipment depending
upon the speed. Curve resistance (c) equals zero on straight
track; and grade effect (G) equals zero on level track. Grade effect
(G) is entirely independent of speed, while curve resistance (c)
and braking effort (B) arc practically so. Braking effort (B)
equals zero except when brakes are applied. I n ordinary operation
both P and B rarely appear in the formula a t the same time, as
brakes are seldom applied while power is on. Constant speed is
the equivalent of zero acceleration, or when the conditions are such
that P - j - c +_ G - B = 0. For tangent level track, constant
speed is had when P = f (no braking), or in other words, when
the speed is such that the train resistance is just equal to the motor
tractive effort. When the formula gives a negative value for
acceleration, a decreasing speed is indicated; this usually will occur
when coasting (P = 0)or braking, and always will result when the
combination of motor tractive effort and down grade effect (if either
or both exist) are insufficient to balance the resistance factors f,
c, -G and B, or such of them as exist.
Graphical Representation of Acceleration. The slope of any
increment of a speed time curve plotted between time in seconds as
abscissas and speed in miles per hour as ordinates is numerically
equal to the rate of acceleration in miles per hour per second.
Relations between Acceleration, Speed, T i e and Distance.
Where A = rate of acceleration, miles per hour per second
S = speed, miles per hour
f = tlme, seconds
d = distance, miles
S2
then
t = - and d = 72ooA

Speed a t End of
Line Acceleration. T o obtain the
approximate speed a t the end of straight line acceleration, use may
be made of the following formula, which, as shown on page 244,
gives the approximate relation between the speed and tractive
effort of a direct current railway motor:

where S A = speed a t end of straight line acceleration

SM = free running speed
T A = tractive effort a t end of straight line acceleration
T M = tractive effort a t free running- s ~ e e d .
For example, assume a s e t o n car geared for a maximum speed of
60 miles per hour, and a n acceleration of 0.8 miles per hour per sec-

ACCELERATION

153

ond. The train resistance a t 60 miles per hour (see Fig. 13) is
26 Ib. per ton, which a t maximum speed 1s the same as the tractive
effort. During acceleration the tractive effort must be approximately (0.8 X loo =) 80 Ib. per ton for acceleration, plus an average of say 16 Ib. per ton for train resistance, or 96 Ib. per ton total.
The speed a t the end of straight line acceleration will then be
26 35
S = 60 X
= 38.8 miles per hour.

($)

Usual Rates of Acceleration in Practice.

Steam locomotives. freight service. .........

Steam locomotives, passenger service..
Electric locomotives, passenger service..
Electric motor cars, interurban service..
Electric motor cars. city service.. ..........
Electric motor can. rapid transit service..
Light weight safety cars.. ................
Highest practicable rates..

o. I to 0 . 1 mi. per hr. per sec.

..... 0.z to 0.5 mi. per hr. per sec.

.... 0 . 3 to 0 . 6 mi. per hr. per sec.

.... 0.8 to I .3 mi. per hr. per sec.

I .3 to I .8 mi. per hr. per sec.
.. I.
s t o z .o mi. per hr. per sec.
1.5 to 2 . 5 mi. per hr. per sec.

................ 1.5 to 3.0 mi. per hr. per sec.

The maximum possible rate of acceleration is limited by the available tractive effort which is dependent not only upon the motor
equipment and weight of train but also upon the coefficient of adhesion between wheel and rail.
Comfortable Rates of Acceleration. I t should be remembered
that the effect of acceleration on the comfort of passengers is not
so much a measure of the rate of acceleration as of the rate of change
of accelerating rate. Referring to Fig. 35, the rates of acceleration
and of braking as shown by the full lines may be uncomfortable to
passengers, due to the abrupt start, application of brakes, and stop. However, if the
starting resistance on the first controller
position be so proportioned as to give the start
shown by the dotted line a t A , and the brakes
be gradually applied and released so as to rePIG. 35.
place angles by curves as shown by the dotted
lines a t B and C, very high rates of acceleration and braking may
be employed without causing discomfort to passengers. The inherently high values of train resistance a t the instant of starting (see p.
143) tend to round out the initial start a t A .
Effect on Draft Gear and Other Car Equipment. The above
remarks, relative to comfort of passengers, apply equally to shocks
to draft gear in train operation and to other parts of the car equip
ment. To reduce these shocks it is necessary to keep down the
rate of change of accelerating rate.
Economical Rates of Acceleration. The most economical rate
of acceleration will depend upon (I) the frequency of service, ( 2 )
train weight and (3) the capacity of the equipment. If, with the
same schedule speed, the acceleration be increased, there will result
a slight saving in energy due to the lessened train resistance a t
the lower maximum speed attained. There will also be a very
important saving in energy due to (I) the possibility of a longer
coast, allowing braking to begin a t a lower speed (the losses in
braking being proportional to the square of the speed a t which
braking begins), and (2) with direct current equipments a saving
in rheostatic losses, which are inversely proportional to the rate

J"1,

154

ELECTRIC RAILW.4Y H.4NDBOOK

FIG. j7.-Varlatlon

of energy consumption wrth brakrng rates.

166

ACCELERATION

of acceleration. Such savings are effected each time a stop is

made, and therefore are much more important in frequent sto
service, not only as regards energy consumption, but also wit
respect tospeed. As illustrative of the relation between rates of
acceleration and braking on energy consumption, Figs. 36 and 37
have been reproduced from a paper by C. W. Squier (met.
Ry. Jmrr., 1918). The curves are based on a study of a 23% ton
car a t a schedule speed of 8% miles per hour, with a 620 ft. run
and 7 second stop Fig. 36 shows the effect of changing the rate
of acceleration, the braking rate remaining a t 1% miles per
hour per second. Fig. 37 shows the efiect of changing the rate
of braking, while the accelerating rate remains constant a t 134
miles per hour per second. I t will be noted that as the rate of
acceleration is increased, the relative energy saving becomes less,
and this will be found generally true. I n this case, had advantage
been taken of the possibility of increase in schedule speed (8 per
cent between 1.0 and 2 . 0 miles per hour per second), the energy
saving would have been less (only 3 per cent), but the combined
savings of platform wages (due to the reduction in car hours) and
power cost would have amounted to more than 6 per cent. Formulas for energy losses in rheostatic control are given on page 201,
and for losses in braking on page 195.
Relation Between Acceleration and Maximum Speed. As
noted in the section on Run Curves, the dotted line in Fig. 44
(p. 164) is the locus of the maximum speeds attained during a
run of I mile, assuming coasting, braking retardation and train
resistance constant. This figure, as well as Fig. 46 (p. 165)~illustrates the fact that for a given run the maximum speed usually is
reduced by increasing the rate of acceleration.

JOO

10

300

30

200

100 1 0
L.

:e ?u

e 2
'0

FIG.38.-Reduction

B2

40

6Ogecond,BO

100

120

140

of rnaxlrnurn currcnt by use of lower rate of acceleration

In multtple than In wrlcs.

Special Accelerating Method. Where the service is infrequent,

as in the case of some interurban railways, the acceleration is often
limited by the low tension distribution and substation capacity.
This limitation may, however, be greatly reduced with the retention

158

ELECTRIC RAILWAY HANDBOOK

of a fair average rate of acceleration and its attending benefits

by using a high rate of acceleration during series acceleration and a
considerably less rate during multiple acceleration. The effects
of such a procedure are shown by the speed-time curves and corresponding current curves in Fig. 38. In case A , constant current
per car is used throughout both series and parallel connections of
the motors; this is done by allowing the controller to rest in full
series position until the value of current is one-half of what it was
during the series acceleration. In case B, with a constant rate of
straight line acceleration, the value of the current per car when the
motors are in parallel is double that when they are in series; in other
words, the current per motor is kept constant.
I n case A , for this particular run the rates of
straight line acceleration are approximately
1.5 and o 6 mile per hour per second. In case
B, the straight line acceleration is approximately 0.9 mile per hour per second throughout. Practically the same energy is used in
both cases but the maximum value of current
in case A is only about 73 per cent of that
in case B. Another method of using two rates
of acceleration in the series and parallel positions of the controller is an intermediate one,
where the controller is so operated that the
current per car with motors in parallel is
greater than when in series, but less than
double, as in case B. The maximum current is then somewhere between the values
in cases A and B. An article by F. E. Wynne
in the Elecfric Journal states that for hilly
roads this third method is found more satisfactory than the method of case A . He also
states that on runs of % mile or less. the acceleration with constant current per motor
(case B ) is less economical of energy than
either of the other methods which are about
on a par with each other, but on longer runs
the energy consumption is practically the
Wel8ht
same with all three methods.
FIG. jg.--Irnprov~sed
Improvised Accelerometer. A readily imaccelerometer.
~rovisedaccelerometer is shown in Fig. 39.
I t consists of an ordinary 2-ft. four-fold
pocket rule, a piece of thread and a weight. Eighteen inches of the
rule is opened out straight and the thread is tied to this part a t a
point 196 in from the end; the weight is tied to the other end of the
thread, and the remaining 6 in. of the rule is opened and leveled
(extending in the direction of motion) till the plumb line cuts the
3-in point of the base thus formed. About 16.1 in. of thread then
hangs between the point of support and the base; consequently, each
increment of r it. per second, per second acceleration or retardation
will be indicated by a correspondingdeflection increment of approximately % in. a t the point a t wbich the thread and the base line

157

COEFFICIENT OF ADHESION

intersect. If the thread be attached a t a point 6% in. from the

upper end, and the upper part of the rule bent further over until
the plumb thread again cuts the 3-in point, the deflection of each 34
in. due to acceleration or retardation will indicate I mile per hour
per second.
Coe5cient of Adhesion. As used in railway work the coefficient
of adhesion is the coefficient of friction between driving-wheel
tread and rail. I t is equal to the ratio of the maximum
tractive effort (neglecting torque limit of motor equipmentf:?E
driving-wheel tread, to the normal pressure (called effective weight
on driver) between rail and driving-wheel tread. I t is usually
expressed in per cent of the effective weight on the driver. The
results of tests to determine the coefficient of adhesion vary considerably. This variation is probably due to the difficulty in
ascertaining, describing and duplicating the exact condition of
rail and wheel surfaces
I t is common steam locomotive ractice to assume 0.22 to 0 . 2 5
as the coellicient of adhesion with c i a n very dry rails. Because of
torque uniformity throughout a revolution of a driving wheel
higher maximum values are possible in electric railway practice,
but in order to take full advantage of this fact all sudden fluctuations in torque such as might be brought about by improper
arrangement of resistance taps or improper controller manipulation
must be avoided.
Coefficient of adhesion is a quite variable quantity depending
upon the condition of the rail and wheel. Neglecting the use of
sand, a clean dry or very wet rail gives the highest coefficient of
adhesion. Instances have been noted where the value of the
coefficient of adhesion with a thoroughly wet rail has been nearly
that for a perfectly dry rail, while with the presence of moisture
in amounts scarcely perceptible to the unaided eye it has been
found to be very low. Sand, while detrimental in its abrasive
effect, im roves poor adhesion and by its use very high values
may be ogtained with dry rail. I t is, however, unsafe to depend
U
I sand to improve adhesion with dry rail because the sand is
1' ely to be blown from in front of the wheel. When the wheels are
dipping the coefficient of kinetic friction between driving wheel
tread and rail is less than 0.10.
Approximate Values of Coefficient of Adhesion
o 25 to o 30; anth sand o 35 to o 40
Clean dry rail
Clean thoroughly wet rall
Greasy and motst rail
Sleet on rarl
Llght snow on rarl

o 18 to o 20; m t h sand o
o 15 to o IS,m t h sand o
0 IS
wlth sand o
m t h sand o
o 10

22 to o 2s
22 t o o as
20
15

I t is evident that the coefficient of adhesion is the factor which,

with a given weight on the driving wheels, very definitely limits the
tractive effort of a car or locomotive, and therefore also limits the
rate of acceleration or the grade which may be surmounted with a
given weight of train, or the weight of train which may be handled
over a given grade or accelerated a t a given rate. Conversely, with
a given weight of train and rate of acceleration or per cent of grade,
the coefficient of adhesion determines the required weight on the

158

FLECTRIC RAILWAY HANDBOOK

driving wheels. The coefficient of adhesion further determines the

maximum possible braking rate
In the consideration of any such problems where the coefficient of
adhesion is the Iimiting factor relative to rates of acceleration or
braking, due allowance must be made for transfer of weight between
axles in the direction opposite to that of positwe acceleration.
Weight Transfer Due to Acceleration. Due to the facts that
efiective tractioh is ap lied a t the points of contact between dnving
wheels and rails, and tRat both the center of gravity and the drawbar of the motor car or locomotive are located considerably above
tbe rail, there will be a transfer of the weight of the car or locomotive from the forward to the rear wheeIs during acceleration,
and a t any other time when sitive drawbar pull is being exerted
on trailing cars. Such weiggtransfer is simlar to, but opposite
in direction to that which occurs in braking, as discussed by
R. A Parke, A 1E.E., 1902, and abstracted at page 444. Formulas for such weight transfer due to acceleration and drawbar pull
of trailers are as follows:
Case of One Car or Loco,D,rerf~bno f M o f ~
motiue Alone with no trailing cars and hence no
drawbar pull.
Exterttal Forces Acting
Car Body in Accelnatzng. (Fig 4 0 . )

on

II =

PIG qo -We1

ht transfer rn accelerat~on
Ear body.

WlU

*g

PI = W
- -, - -W
- ,aj

2
gJ
in which H = horizontal retarding force on each truck center pin,
pounds (the same for each truck because the same
number and size of motors per truck, similarly controlled and providing the same tractive effort per
truck)
Pt = pressure between body and truck rear center plates,
pounds
PI = pressure between body and truck forward center
plates, pounds
(Center of gravity
W1 = weight of car body, pounds
being in a vertical axis midway between truck
cen ters)
j = height of center of gravity of body above center
plate surface, inches
1 = distance between center pins, inches
g = acceleration due to gravity, feet per second per
second ;= 32 2
a = rate of acceleration, feet per second per second
= (miles per hour per second) X r.467.

WEIGHT TRANSFER I N ACCELERATION

E*nd

159

Fwzes Acting on Fonoard Trarck i n Accderaling. (Fig.

41

PIG.41.-Wezghi

transfer rn accelerattnn.

Trucks.

in which R2 = total pressure betwcen rail and rear pair of wheels

of forward truck, pounds

RI = total pressure between rail and front pair of wheels

of forward truck, pounds

T* = total maximum tractive effort available between
rail and rear pair of wheels, pounds
TI = total maximum tractive effort available between
rail and front pair of wheels of front truck, pounds
WI = weight of each truck, pounds (Center of gravity
of truck being in vertical axis midway between
wheels)
W = total weight of car, pounds = WI 2W2
h = height of truck center plate surface above rail
b = wheel base of truck
d height of center of gravity of truck above rail
q = coe5cient of adhesion between wheel and rail
(o 25 may be used for this).

For signihcance of other symbols see page I 58.

The above formulas do not take into consideration weigbt transfer
due to the inertia of rotating parts (wheels, axles and armatures).
This egect may vary, depending upon the relative weights and relative arrangements of such rotating parts, the effect 1s small, however,
and in all practical cases, the formulas as shown will give results
within the limits of error in the assumptions necessary w ~ t hrespect
tr location of centers of gravity and coefficient of adhesion.

160

ELECTRIC RAILWAY HANDBOOK

Case of Car w Locomotire with Trailer

Trailing Cars. (Figs.

4 0 and 41 )

Drawbar pull D will be applied a t the coupler which is a distance

m above the llne of application of the forces H

W t P I H h Wzad
= ~ + ~ + ~ +

TI = ~ R I

gb

Tz = qR.

in which D = drawbar pull in pounds.

m = height of drawbar pull or coupler above center plate
surface, inches.
The significance of other symbols, the methods of deriving equations and the considerations neglected are the same as in the preceding case.

RUN CURVES

Run Curves
Speed is the desired end in the solution of most railway problems.
Revenue usually is in some proportion to the miles operated, while
many important expenses are more nearly proportional to time.
Within certain Iimts, therefore, efforts should be made to increase
miles operated in a given time, or to reduce the time for operation
over a given distance, either process resulting in increased speed.
Under some conditions, however, the attainment of increased speed
is costly, and under any conditions a critical point is reached where
the cost of a further increase in speed is not justified. I t is some
phase of this problem that is involved in nearly every engineering
study of train movement. Factors must be considered such as
size and headway of cars or trains, frequency and duration of stops,
track grades and curves, train resistance, acceleration and braking
rates, coefficient of adhesion, signalling, motor capacities, gear
ratios, control and starting resistance design, power and energy
requirements, feeder layout and power supply.
Speed Determination by Counting Poles or Rail Joints. The
approximate speed of a car may he determined by counting the
number of evenly spaced poles or rail joints passed in a given time.
The speed in miles per hour is equal to the number of poles or rail
joints passed in o 682d seconds, where d is the spacing of the poles
or rail joints in feet. Thus, approximate miles per hour speed =
number of
number of
number of
number of
number of
number of

poles passed ln 68 seconds

poles passed ln 75 seconds
poles passed In 85 seconds
Iolnts passed In 20 seconds
jolnts passed In 22% seconds
Iolnts passed In 41 seconds

(100 it. spacing)

(110 ft. spactng)
(125 ft. spacmg)
(30 ft. rads)
(33 ft. rank)
(60 ft. ra~ls)

The Speed-Time C w e is the convenient and useful medium for

the study of the interrelation of the many elements suggested above.
Plotted in ordinates of speed against time, its slope a t any point
is a measure of acceleration, the latter being the resultant of the
potential, kinetic or frictional forces which tend to aid or oppose
the motion of the car or train. A knowledge of such forces as are
applicable to the given problem enables the engineer, through the
acceleration formula, to plot the speed-time curve, and through
the latter to make further determinations relative to distance,
power, energy, losses, and the like. The detail of such a study, and
the exactness of the resulting determinations may de nd upon the
degree of accuracy desired, or upon the limitations o E h e available
data. Preliminary estimates from rough data need not nor cannot
be made with as much detail as to minor factors as is possible and
often desirable in specifying final equipment in costly installations.
I n ordinary street railway service where succeeding tri s are made
with stops varying in number and location, a n average fzngth, time
and grade may be taken and "typical" run curves usually are
employed for the solution of the problems usually encountered. In
rapid transit or heavy traction service where stops are a t definite
fixed stations, the predetermination of motor or signal requirements

ELECTRIC R.ULW.IY HANDBOOK

1G2

may involve a considerat~onof the actu:rl slops, profile and alinemeut, with the rcsulhnt run curves plotted in considerable detail
from one end OF the route to thc other, In e a ~ hdirection and for
rach class of service. 'The rnc:tho(ls cmplo!ed in Lhl. construction
of the spertl-tinw curve vary (onsicler;lbly with the requirements of
the I-ase in hand. W hcre but few typic 11 rulis arc t o be corlsidered
with the same equipment, or a study is LO be made of the effect
of a ~ h a n p ein motors, gear ratio, or accelerating rate, for instance,
the "step-by-step" mcthod IS most often employed, with succeeding
calculations of thc acceleration formula applied directly to the speedtime curve. Where a definite equipment is to be considered as
operating over runs of varying length with definitely fixed stops,
grades ant1 curves, it may be advisable to adopt some such method
as that proposed by C. 0.3lailloux, including the preparation of
working charts or trmplets as ultimate time-savers. Or, where
data is most meager, preliminary estimates may be made by the
use of "straight line" curves, or charts or tables of more or less
general application.
Data Required for Speed-Time Curves. The data listed below is
required for the construction of speed time curves, the exactness
depending upon the degree of accuracy desired. Letters to the left
refer to one of three commonly used methods as tisted below, a
capital letter indicating imperative need, lower case indicating data
desirable for exactness but not essential for close approximations.
A

B
C

ABC
ABC
ABC
ABC
abC

ABC
bc
bC

BC
A
BC
bC
be
bc
bc

BC

bc
bc

BC
bC
ABC
ABC
ABC
bC
aBC

Straight 11ne appros~mation-no definite motor under contemplat~on-typical run. See pp. rhj-16;.
Step-by-step method wlth motor characterist~c-typical run.
See pp. 167-176.
Debiled run wlth defintte stops. grades and curves. and w ~ t hgreatest
accuracy. See pp. 1;&184.
h'umber of r n o t ~ xcar5 In train
Kumber of trail cars In tratn
Welght of motor car or locomotive. e.iclus~veof load
Wclght of trail car, cxclusioe of load
Welght of locomot~veon drivers
Welght of averaKc load, motor J I I ~trail cars
Welght of masrrnum load. motor and trail cars
Cross-sectional area of lncomot~veand cars
iiumber of motors per motor car or ~ocomotive
Rate of accelerat~on
Character~;ticcurve of motor
Line voltaae a t train. averaKc. maxlmurn a n d lninimum
Electrical resistance of motors
Welght of motor armature
Dlameter of motor armature
Dlameter drlvine wheels
.
Welght d n r l n v wheels
Diameter and wcight of tratl~nvwheels
Gear ratlo of motors
Profile and altnement of tmck
L e n ~ t hof run
Tlmr of run
Number of stops
Locat~onof stops
D u r a t ~ o nof 5:ops
Layover a t ends of 11ne
Fpeed Iimrtations
Braking rate

BC
Determination of Typical Run. In selecting a typical run it is
advisable to divide the line into general sections, for instance,

city, suburban and interurban, and then use the approximate

average length of run in a section for the lrngth of the typical rnn
for that section. Jf there arc steep grntles on the line there should
also be a tfivision into grade srrtions ;~ntla typical run sc.lcctecl for
each of thcsr. The avcragc (quivalcnt grade, in per cent, may be
obtained 1)) he following formula, which assumes only one-half
the gr.~vitationalenergj on down grades to he saved, the remainder
being consumed in braking. Other assumptions may I)e found
necessary in certain cases, for instance, in city work, more than
half of this energy may be wasted in br:tking; in high spcetl limitetl
interurban service on private right of way, less.

wherc C = average equivalent grade, per cent

I = distance, feet
FII = summation of rises, feet
I I I = summation of falls, feet.
In determining the average length or time of a typical run, the
total distance or time should be divided by the total number of
equivalent stops. The term "equivalent stops" includes slowdowns
in addition to actual stops, as it is necessary to allow for the former
as well as for the latter. In the absence of data permitting a more
accurate determination, it is usual to make each slowdown where
),rakes are applied equivalent to one-half a stop in city service, or
one-third stop in suburban or interurban servire. The number of
equivalent stops is then the number of actual stops plus one-half
(or one-third) the number of slowdowns.
Wherever the frequency of stops, maximum speeds, or grades are
materially different on various sections of a line, the distanc.es, stops,
running times and grades shoulcl he segregated, ant1 separate typical
runs made u for each such section.
"Sbaight fine" Approximation of Speed-Time Curves. Several
authors have proposed "general" solutions of speed-time
problems, b a s e d u p o n
either an assumed general
motor characteristic curve
Y)O(I;
or an assum tion of uni- 7 4 0
tam2
a owl :
form rates o r acceleration :,
and retardation. As uni- 220
lmoj
form rates of acceleration $
are represented by straight A
lines in the s~eed-time
0
~9
40
00
YO
~ U J 1.u
,
Be, oud.
curve, the general solution
bsed on the latter assump- F I G . 4z.-Typlcal strawht I I I I ~ spccd nncl
d~stancecurves ( n r ~roasttny).
tion has been called the
"straight line method." It is most convenient for cruick and
rough~kpproximation, and involvcs no reference to motbr charncteristic curves. The following description, with Figs. 42 to 46,
inclusive, is from the Standard Handbook, 1 9 2 2 , section 16, paragraphs 39 to 47, inclusive.
The speed-time curve is shown in Fig. 42 in its simplest form,
acceleration being carried on a t constant rate up to the point of

"

1mo2

-~~

~-

164

FLECTRIC RAILWAY HANDBOOK

applying brakes, which are so applied as-to give a constant rate

of braking. The area enclosed within the triangle ABC is proportional to the distance traveled, the distance covered up to
any instant being shown by the distance-time curve. Thus, with
the constants chosen in Fig. 42, a maximum speed of 60 miles per
hour is required to obtain an average speed of 30 miles per hour,
that is, covering a distance of 5280 ft. in 1 2 0 seconds.
In practical o ration it is not possible to choose the rate of acceleration and braEng with such nicety as shown in Fig. 42, a greater
or less wriod of coasting
being rGuired. Introduc
ing coasting gives rise to
the form of speed-time
curve shown in Eig. 43,
showing three f r i c t i o n
rates: f = o, 15 and 30
lb. per ton respectively.
With no friction the speedFIG.43.-Typical straiqht line speed-time time curve ABCD is constructed, the speed being
curves
..(various coastlna
- resistances).
constant a t 46 miles per
r ton friction,
hour during the coasting period. With 15 lb.
the speed-time curve AEGD is formed, and wit&o lb. per ton
friction the speed-time curve AFHD. The introduction of friction occasions a falling off of speed during the coasting period
proportional to the friction value taken, which for the sake of
simplicity is here assumed to be constant a t all speeds.

FIG.&-Typical

straight line speed-time curves (various rates of accelerat~on).

The speed-time curves shown in Figs. 42 and 43 both indicate

the completion of the run of 5280 ft. in ~zoseconds,although in one
case the rate of acceleration was that produced by 65.7 Ib. per ton,
and in the other case by loo Ib. per ton. These curves are of equal
area, as the distance in each case is 5280 ft., and it thus becomes
possible to produce any number of speed-time curves for a given
distance and elapsed time by varying the rate of acceleration with
consequent variation in time of coasting.

STRAIGHT LINE SPEED-TIME

CURVES

165

A more extended set of curves is given in Fig. qq, for the same
distance of I mile covered in 1 2 0 seconds, the rate of acceleration
varying from 0.713 mile per hour per second as a minimum t o an
infinite number of miles per hour per second as a maximum. A
train resistance value of 15 lb. per ton is assumed constant a t
all speeds and the dotted curve A B is the locus of the maximum
speeds reached with the
different rates of acceleration. The highest maximum speed required is
obtained with no coasting, and the minimum
speed is obtained with an
infinite rate of accelera-tion. The tractive efforts
Second#
corresponding to the dif- FIG. 45.-Similar
straight line speed-time
curves
(var~ous
distances).
ferent accelerating rates
are given as including 15 lb. per ton train resistance, hence the net
tractive effort values corresponding to the rates of acceleration
indicated are 15 Ib. per ton less than the figures given.
Instead of plotting similar curves for distancesother than 5280ft.,
advantage may be taken of the fact that the area enclosed by the
speed-time curve is proportional to the distance travelled and the
coordinates are proportional to the square root of the enclosed area.
I t is convenient therefore to plot a full series of curves for one distance, preferably one stop per mile, that is, a distance of 5280 ft.
run, and apply the results so obtained to any other distance by using
a factor expressing the relation of the square roots of the distance
traveled. This is shown in Fig. 45, where ABCD represents
an area of I mile, or one stop per mile; AFZL two stops per mile
with a factor of
= 0.707; A E H K four stops per mile with a
= 0.5, and AGJM one stop in 1% miles with a
factor of G5
= 1.225.
factor of.&
Refernng to Fig. 44, it is obvious that a similar sheet could he
* prepared for any elapsed time other than 1 2 0 seconds, using the
same train resistance and braking values of 15 and 150 lb per ton
respectively. Fig. 46 is so constructed to show the time limits
imposed by 15 lb. per ton train resistance, and 1501b. per ton braking
effort for any length of run and any rate of acceleration. The
dotted curves indicate the locus of the several maximum speeds
reached with different accelerating rates for a run made in a given
elapsed time; thus the dotted curve terminating a t 80.7 lb. per ton
is a reproduction of the similar dotted curve, AB, given in Fig. ,++,
and shows the maximum speed reached with any rate of acceleration
for a run of 5280 ft. in 1 2 0 seconds with 15 lb. per ton train resistance and 150 Ib. braking effort. Similarly, the dotted curve terminating a t 100.4 lb. per ton shows the limiting maximum speeds
reached with any rate of acceleration when a run of 5280 ft. is
accomplished in I l o seconds with the same values of train resistance,
braking, etc.
The line CD shows the slope of a coasting line a t the rate of 15
lb. per ton train friction. Thus in a run completed in 120 seconds,

-.

166

ELECTRIC RAILWAY HANDBOOK

the minimum accelerating rate corresponds to 80.) Ib. per ton

(gross) with no coasting, braking commencmg as soon as accelekatlon ceases. If a higher rate of acceleration than 80.7 Ib. per ton
be used for a cycle completed in 1 2 0 seconds, for example 132 ib.
per ton (gross), coasting must be introduced between the acceleratlng and braking lines. This coasting line may be plotted, with 132
lb. per ton acceleration, by drawing a line parallel to the line CD
(Fig. 46) starting a t intersection of accelerating line, 132 Ib. per
ton, and dotted line to 80.7, and terminating at intersection with
braking line ending a t I 2 0 seconds.
100

; >1

$ ,5

.I.

4
i 60
.0
t

a
.-.. d:

z
11

'2

rl)

L2
180
D
240
Tame = Sec. +
PIG. 46.-General straight line speed-time curves (train resistance, 15 lb.
per ton; braking reststance. 150 lb. per ton).
0

09

By the use of Fig. 46 i t becomes possible to determine the tithe

required to make a run over any distance with any rate of acceleration, provided the train resistance is 15 Ib. per ton and braking corresponds to 150 lb. per ton retarding effort.
Example: Given a distance of 8000 ft., train resistance IS Ib. per
ton, braking effort 150 Ib. per ton, gross tractive effort 67.4 Ib. per
ton (including 15 Ib. per ton train resistance), what is the mifiimum time required to perform the run and what maximum speed is
reached?
Solution: From Fig. 46, minimum elapsed time with 67.4 Ib.
tractive effort is 130 seconds with no coasting.
Ratio of distances

$g

= 1.23.

FIence for 8000 ft. time of run = 130 X 1.23 = 160 seconds.
Maximum speed for 5280 St. = 55.6 miles per hour.
Hence for 8000 ft. speed = 55.6 X 1.23 = 68.5 miles per hour,,

STEP-BY-STEP SPEED-TIME CURVES

167

Jn actual practice, a certain amount of coasting is necessary;

hence, the run of 8000 ft. would be made in somewhat more than the
minimum possible l i m t of 160 seconds, or else the tractive effort
should be Increased to allow for some higher rate of acceleration
which would permit of some coasting.
The values chosen for train resistance and braking effort in Fig. 46
(15 lb. and 150 lb., respectively) are conservative and practical
operating values. The maximum speed during a service run is
influenced but little by the type of motive power and its characteristics, and will diEer but slightly from that indicated by Fig. 46.
This chart, therefore, constitutes a set of fundamental data by means
of which it becomes possible to approximate the several data in
any acceleration problem.
Step-by-step Method, Speed-Time Curves
In this method the formula for acceleration is applied successively
to the various conditions of speed, tractive effort, train resistance,
curve resistance, grade effects and braking, as such conditions
change and present themselves throughout the run. During successive increments of speed or time, the acceleration rates are assumed
to be constant, and are plotted as slopes. I t is only necessary that
such increments be made sufficiently small to obtain great accuracy
in plotting. I t should be remembered, however, that such formulas
as those for train resistance, which enter into the acceleration .formula, are based upon certain assumptions, and the measure of
probability of error in such formulas is the measure of maximum
possible accuracy in the resulting speed-time curves.
Definite motor characteristics are usually used in the construction
of speed-time curves. However, as the characteristics of railway
motors are very much alike, the characteristic curve of any motor
of the proper type and capacity may be used in preliminary work
without fear of serious error. In some cases it even may be advisable
to disregard any definite motor, and use a "general" motor characteristic such as described on page 242, and following.
I n the following, the plottlng of a s eed time curve is described
and illustrated by the solution of a de%nitiploblem. A method is
shown where the curve is plotted in steps a ter a calculation of the
acceleration rate for each step, and this is followed by a convenient
gra hical solution. I n each case a typical run has been considered.
&ep-by-step method of plotting the speed-time curve for a given
car or train starting from rest and running a given distance to a stop,
a t a given schedule speed. I t is desirable to divide the speed-time
curve into the following principal parts which are given in the
order in which they usually are constructed:
( I ) Acceleration with motor current controlled, called for convenience, Straight Line Acceleration. (oa, Fig. 47.)
( 2 ) Acceleration beyond Straight Line Acceleration, or "acceleration on the motor curve." (ab, Fig. 47.)
(3) Stand-still during stop, allowed for in the schedule speed.
(de, Fib. 47.)
(4) Braklng. (cd, Fig. 47.)
(5) Coasting. (bc, Fig. 47.)

168

ELECTRIC RAILWAY HANDBOOK

Scales. The speed-time curve shows.the speed of the train a t

any instant during the run and usually is plotted between time in
seconds as abscissas and miles per hour as ordinates.
( I ) Acceleration with Motw Current Controlled (Straight Line
Acceleration) oa.
miles per hour
Slope of 00 = A = ----- drawn to scale.
seconds
A = assumed acceleration, miles per hour per
in which
second, or the acceleration produced by the
assumed maximum current per motor, found
as below under (2).
Speed ai a. The approximate speed of the train a t a, Fig. 47, is
determined from the motor characteristic CUNe, and is either the
speed corresponding to the assumed value of current per motor
during straight line acceleration, or to the tractive effort T determined in accordance with the following:

in which T = tractive effort a t driving wheel tread for one motor,

pounds
W = weight of train, tons
n = number of motors
A = rate of straight line acceleration, miles per hour per
second
K = ratio of linear inertia to total inertia of train (See
page 150)
f = average train resistance of the train between the
initial speed and an assumed probable approximate
speed a t a (Fig. 47), pounds per ton weight of train
C = curve resistance, pounds per ton weight of train
G = grade effect, pounds per ton weight of train. (Use
(+) before this value for up grade and (-) for
down grade.)
( 2 ) Accelerafion Beyond Straight Line Acceleration (ab). The
curve ab is approximated by a succession of straight lines (ag, gh,
etc., Fig. 4 7 ) .
Slope to scale of each of these lines =
A = o.orogBK(P - f
C G)

- +

in which, for any part such as ag,

Tn

- = tractive effort, pounds per ton weight of train.

W

P, T, and f are taken as the mean of those values between the speeds
a and g. For this and other parts of the speed-time curve beyond
the straight line acceleration portion, T is obtained from the
motor tractive effort characteristic curve. The sign of G is (-) for

STEP-BY-STEP SPEED-TIME CURVES

169

up grade and (+) for down grade. The point b will be determined
by the final location of the coasting line bc.
Speed Increment. The amount of the speed increment between a
and g will depend on the allowable error in approximation. For the
same allowable error, this increment may be greater in the steeper
than in the flatter portions of the curve.
( 3 ) Slop Period, de. The time of run, in seconds, is
miles from start to stop X 3600
miles per hour schedule speed
od = oe - (stop period in seconds)
oe =

and

(4) Braking Line, cd. The slope to scale of the braking line
cd is equal to the assumed rate of braking retardation. T h e
point c is determined by the final location of the coasting line bc.
(5) Coasting Line, bc. The average slope to scale of the coasting
line bc is
A = o.o1098K(- f - C G)
in which f = approximate average train resistance between the
probable approximate assumed speeds a t b and c ,
pounds per ton weight of train.
K, C and G are the same as before.
The line bc is placed so that area (oabcd) is equivalent to the
distance to be covered by the run (see page 184).
Theoretically, the coasting line is curved, as the retarding forces
decrease as the speed decreases; practically no great error is involved
by assuming the retarding forces constant; these are often taken as
equal to the train resistance a t the speed a t which power is cut off,
assuming that the decreasing train resistance is offset by the neglect
of motor friction and windage.
See pages 1 7 1 and 175 for more definite location of coasting line.
As an example to illustrate the application of the formulas and
method described above, assume the following conditions:

Weight of car. 1 8 tons

Diameter of wheels. 30 inches; weight. 400 Ib: number. 8
Motor equipment, two GE-a47 motors; qenr Atio
Weinht of armature so.7 Ib.: d~ameter10%
. inches
Average llne voltage; 660 .
Average startlng current per motor. 70 amp.
Braking rate. 1.5 mi. per hr. per see.
Average distance between stops. 0.46 mile
Schedule speed. 17. mi per hr.
Average duration of stop, 5 seconds.

Required to construct the speed-time curve for a complete run,

neglecting track curves and grades.
Rnlio of Linear to Total Inertia. The value of K is 0.924, calculated from the above data by the formula given on page 150. In
the absence of such data relative to wheel and armature weights
and dimensions, a value for K may be assumed by inspection of
the table on page I j ~or, for less accurate work, the constant 0.01
may be substituted for o.01og8K in the calculations of acceleration
rates.

170

ELECTRIC RAILWAY HANDBOOK

Becondl

FIG. 47.-Typlcal

or,l.tI

Sectlon

otoa
a
a tog
g&h
h
h to]
7

jtok
k
ktol
1

l to

m
m to b

b
b to 9
9
9 to a
Q

tor
r
rtoc
c
c to d
q

For data, see p. 169; calculations below

run curves

Speed.
p

I
14 a
avg 1.5 6
I70
avg 1 8 0
19 o
avg 19 7
20 5
avg a1 a
22 0
avgzzs
23 0
avg 23 5
24 o
avg 24 5
as o
avg 24 5
240
avg 23 5
230
avg 22 5
220
avg 21 2

20s

1 1 d.

70 0
470
38 0
33 0
29 5
27 5
26 o
a4 5
0 0
0 0
0 0

Distance.

:1
1230

136 7 Ia 5

910

101 0 14 5

'

260

o 878

600

667 153

o ~ z a

460

51.1

03"

380

41 a 16 4

o 262

330

367 170

oaoo

290

32

0 149

250

27 8 18 o

o 18 o

-0

183

o 17 5

-0

178

0 I7 0

-0

I73

0 16 4

-0

166

-I

500

0 0

158

17 5

o 022
0 036
0 055
0078
0 111
0 I44

o 099

OI

,I,

0 a54
0 290
0 329
0 366
o 421
o 460

(I) Strazght Line Acceleraiion, oa. The s ed of the car a t a,

Fig 47, is 14 I ml per hr a t 70 amp ,as read
the characteristic
curve of the motor, wh~chis shown In Section I V as Fig 23, page 240.

Ern

The slope of oa = A = o o 1 q 8 K
=

o 01098 X o
I

924(

26 mi per hr per sec.

- 12.s)

STEP-BY-STEP SPEED-TIME CURVES

171

where T = 1230 lb. a t 70 amp., as read from the characteristic

curve of the motor, and where f = 12.5 Ib per ton average train
resistance between o and 1 4 2 mi per h r , as Interpolated wlth
sufficient accuracy for t h ~ sexample from Fig 13, page 131. On
a sheet of cross-sectlon paper, lay off approprtate units of time as
abscissas and units of miles per hour as ordinates, as in Fig 47,
and draw the line oa w ~ t ha slope of I .26 mi per hr per sec., terminat14 2
ing a t 14 2 miles per hour and - = 1 1 3 seconds.
I 26
( 2 ) Acceleration beyond Stralghl Lzne Accelerdion, ab.
Assume
an increment of speed of 2 8 mlles per hour Then between 14 2
and 17 o miles per hour, A

o o 1 q 8 X o 924

= o 878 mi per hr. per sec., where T and f are determined as above

a t the average speed of

14"

I7

= 15.6 miles per hour.

Draw

ag having a slope of o 878 mi. per hr. per sec ,and termtnatingat I 7.0
m~lesper hour Calculate the accelerating rates for succeeding
increments, and draw in a similar manner, continuing the construction of the acceleration line to a point somewhat beyond the point
a t which it is probable that coasting will begin.
(3') S'RJ? 4w*J,6%. T k
d~Ptit~,:I?L~JVLW&~,
:fa
mlles
from
start
to
stop
X-3600
o
460
X 3600 =
oe=
=
95
m~lesper hour schedule speed
17 4
seconds
and od = oe - de = 95 - 5 = 90 seconds
(4) Braking Line, cd. Draw cd of indefinite length, starting a t
90 seconds, zero speed, and havlng a slope of -1.50 mi per hr.
per sec , the glven braking rate.
(5) Coasting Line, bc. On a separate sheet of similar cross-section
paper, lay off axes w ~ t hunits of time and miles per hour exactly
ltke the ones first constructed and, beginning a t a speed somewhat
greater than the probable maximum, calculate and draw the coasting line for proper decrements of s eed down to a speed somewhat
lower than that a t which it IS probabe that braking will begin. The
decelerating rate A for each decrement is calculated exactly as for
accelerattng increments, a to b. Place the second sheet carrylng
the coasting line beneath the f i t sheet carrying the accelerating
and braking lines, and with base lines coinciding, slide it
horizontally to a posit~onwhere the coasting ltne, bc, is so placed
that the area (oabcd) == length of run = o 46 mAe Trace the coasting ltne on the first sheet, intersecting the accelerating llne at b
and the braking line a t c. Several tnals for the position of bc may
be required. The area may be determined by the use of a planimeter or by counting the squares. The number of squares, of the
size shown in Fig 47, needed under the complete curve = a =
3600 -- 46 3600 = 33.12, as derived from the formula
Is
I0 X 5
for d on page 184.

172

ELECTRIC RAILWAY HANDBOOK

I n the actual construction of the speed-time curve, a much larger

diagram than Fig. 47 should be drawn, the size depending upon the
degree of accuracy required.
Chart for Speed-time Curve Calculations. Fig. 48 is illustrative'of the form of chart used by S. B. Cooper to locate ints on the
motor acceleration and coasting portions of a speedl(lime curve.
To the right and left of the point 0,Fig. 48, are laid off positive and
negative values of tractive effort. These may be in pounds per
motor as shown in Fig. 48, or in pounds per ton as in the preceding
discussion. To a suitable vertical scale of miles per hour, curves
are plotted showing the relation between pounds tractive effort

and speed, for "net tractive effort" and "train resistance." To a

second vertical scale of miles per hour per second, straight lines are
plotted corresponding to level track and various up and down grades,
showing the relation between net tractive effort and acceleration for
the various grade conditions, by the acceleration formula. In
the upper right quarter of the chart, and to the same scale of miles
per hour per second, various reciprocal curves are plotted, showing
the relation between acceleration and time, for various increments
of speed (see "Chart of Reciprocals," page 179). In Fig. 48,
only one such curve is shown, that for an increment of two miles
per hour. I n the lower right quarter of the chart, and to the
same scale of time, a group of radial lines are plotted, showing the
relation between time and distance for various speeds. I t is obvious
that the "net tractive effort" and "train resistance" curves a p ly
only to one definite motor equipment and car weight; the otEer

GRAPHICAL SPEED-TIME CURVES

173

portions of the chart apply to any equipment. The use of the chart
1s illustrated by the dotted line which shows for this equipment
the time and distance required to accelerate from 28 to 30 miles
hour. Starting, as shown, at the average speed of 29 miles per
our, proceed horizontally to the net tractive effort curve a t D,
then vertically to the proper grade line a t E, then horizontally to the
acceleration scale at F and the reciprocal curve which indicates
the proper speed increment (in this case, 2 m.p.h.) a t G , then down
to the time scale a t K and the ro er average speed line (in this
case, 29 rn.p.h.) a t H, then to tge !eft to the distance scale a t I.
The points F, K and I indicate, respectively, a rate of acceleration
of 0.9 m.p.h.p.s., an increment of time of 2.2 seconds, and an increment of distance of 0.018 mile. The chart may be'used for the coasting portion of the speed-time curve by using the " train resistance"
curve instead of the "net tractive effort" curve, the remainder of
the procedure being the same as above. When a negative value for
acceleration is indicated by a location of the points E and F below
the base line, this ordinate value may be laid off above the base,
and time and distance determined as before described.
Graphical Method of Plotting Speed-time Curves. The following graphical method of plotting speed-time curves is adapted
from Bulletin No. go of the University of Illinois, by A. M. Buck.

FIG. 49.-Graphical

method of plotting speed-time curves.

Lay off units of current as abscissas and units of pounds tractive

effort and miles per hour speed as ordinates, the latter to the scale
desired for the final curve (as shown in Fig. 49). Draw the currentspeed and current-tractive effort curves of the motor under cons~deration. Calculate and draw the net current-tractive effort
curve. The net tractive effort for any current may be calculated
as follows:
T , = T - - fw
9%

where TI = net tractive effort per motor, pounds

T = gross tractive effort per motor, pounds

174

ELECTRIC RAILWAY HANDBOOK

f = train resistance a t speed corresponding to assumed
current, pounds per ton
W = weight of train, tons
n = number of motors in train.

At the right of the curves just drawn, erect the perpendicular

OF,and lay off to the right the horizontal scale of seconds for the

speed-time curve. Locate the point X to the left of 0 a t a distance

corresponding to t seconds on the time scale, determined as follows:

or, approximately :
f=-

where

1dW.M

F
S = any speed, miles per hour
F = the tractive effort, pounds, the ordinate value of
which is the same as that of S
W M = weight of train, tons er motor
K = ratio of linear to tot3inertia (see page 150).

T o construct the straight line portion of the speed-time curve,

OVI, erect a perpendicular a t I , , the average starting current,

cutting the speed curve a t S I and the tractive effort curve a t TI.
Draw T I M parallel to the horizontal axis, draw MX, and then Ova
parallel to M X and terminating on a horizontal line from S I a t V I .
Should the data for the problem include the initial accelerating
rate, draw XM first, a t the proper slope to represent such rate.
Then draw the horizontal M T I and the perpendicular T I Z I ,the
latter indicating the starting current as 1, and the speed a t the end
of straight line acceleration as S I . OV1 is then drawn as described
a hnvp
---.
-.

To construct the next increment of the speed-time curve, assume

a small increment of swed. giving the new meed S2. Determine the
average net tractive e'ffort'F~between S , a'nd Sz by erecting a perpendicular a t SA = '-.

Draw

TAN and NX. Draw a

line through V I parallel to N X , cutting a horizontal line from SZ

a t Vt. Then V I V Zis the new increment of the speed-time curve.
By assuming further increments of speed and continuing the process
outlined above, a complete accelerating speed-time curve may be
constructed, the accuracy of which may be made as great as desired
by roper choice of the speed increments.
&r the coasting portion of the curve, the ordinates representing
train resistance, which are plotted down from the gross tractive
effort curve, may be stepped off with dividers and transferred to the
line OY (produced below the base) to determine the corresponding
retardation. Speed increments may be taken as before, and the
coasting curve plotted. For the braking curve a n ordinate corresponding to the braking force must be obtained and added to the

SPEED-TIME CURVES

175

train resistance. I n this manner the entire speed-time curve may be

determined.
As Fig. 49 is drawn, it applies to level track conditions. I t is
evident, however, that other "net tractive effort" curves may be
drawn corresponding to the average equivalent grade or to various
per cents of up and down grades, so that by choosing the proper
net curve for reference, the slope of a section of the speed-time
curve may be determined for any grade. Fig. 49 is nothing more
or less than a graphical solution of the formula for acceleration.
Location of Coasting L i e when Brakes are to be Applied at a
Given Speed. Construct the acceleration portion of the speedtime curve and its corresponding distance-time curve, and carry

9.0

2.6
2.0

1.6 3
0

1.0

0.6

20

40

60

80

100

120

140 e

Tirno.8cconds

FIG. SO.-Method of joining retardation and acceleration curves.

both to a time somewhat beyond that a t which i t is assumed coasting will begin. These CUNeS should be plotted on tracing paper or
cloth, and will be similar to those shown by the solid lines in Fig.
50. (For Fig. 50 grades, see page 182.) Next, on a second sheet
of paper, and to the same scales, begin a t the end of the run and
plot the braking and coasting speed-time curves up to a speed somewhat in excess of that where it is assumed coasting will begin. The
distance-time curve for braking and coasting should also be plotted
on the second sheet, beginning a t the proper distance ordinate for
the end of the run (1.7 miles in Fig. 50). The curves on the second
sheet will appear as shown by the dotted lines in Fig. 50. Then
place the second sheet under the first, and keeping the base lines
coinciding, slide one under the other until the two speed-time
curves cut one another on the same vertical ordinate as that of the
point of tangency of the two distance-time curves (cc', Fig.. 50).
The speed-time and distance-time curve to the right of this ordinate

176

ELECTRIC RAILWAY HANDBOOK

on the second sheet (dotted lines, Fig. 50) may now be traced on the
first sheet and the curves are complete for the run, and show not
only the exact time a t which power should be cut off in order to
coast down to the given speed for applying brakes, but also the
total time required for the run under the assumed conditions.
Speed-time Curve for Train over Given Section of Track with
Varying Grades and Alinement. When i t is required to plot a
sped-time curve for a train over a given track, taking into considerat ~ o nthe effect of definite track grades and curvature, i t is convenient to divide the total distance into consecutively numbered
sections, each of which is uniform throughout its length with respect
to track grades and curvature. A tabulation should then be made,
with the following headings (compass points are relative only).
I. Consecutive number of section (beginning a t west end).
11. Length of section (feet or miles).
111. Percentage of track grade ("+" for down, " - " for up
grade in east-bound direction).
IV. Degree of track curvature.
V. Track curvature expressed as "equivalent grade" (as
explained below: always with " - " sign).
VI. Net equivalent grade, east-bound.
VII. Net equivalent grade, west-bound.
As the resistance due to track curves and the effect of grades are
functions only of the degree of curvature and the percentage of
grade, respectively, the degree of track curvature (Col. IV) may
be expressed in terms of "equivalent grade" (Col. V) as follows:
nc o.7n
= 0.035 n
g' =
=

zo

in which n = number of degrees of track currature

c = track curve resistance, pounds per ton per degree
of curvature. (See,p. 145.)
g' = "equivalent grade, per cent, which would cause
a resistance equal to the curve n.
If the values in Cols. I11 and V have been given " " or ' I - "
signs as above, their algebraic sums will give the proper values for
G I . VI; reversing the signs in Col. I11 and adding the values in
Col. V will give the proper values for Col. VII.
Using Cols. VI and VII values (multiplied by 20) for C', t h e
formula for acceleration becomes
A = o.o1og8K(P - f f G')
with but three variables instead of four, as before.
The speed-time curve may then be plotted as before described,
but the distance-time curve must be plotted a t the same time.
The distance-time curve, in connection with the distance values
as shown in Col. I1 of the tabulation, will show the time ordinate
a t which the acceleration formula must change by the substitution
of a new value for the equivalent grade, G'. I t is advisable to use
the method described on p. 175 to locate the coasting and braklng curve. The completed curve will be similar to those shown in
Figs. 50, 55 or 58.

CHART O F ACCELERATION COEFFICIENTS

177

Chart of Acceleration Coefficients. Where a considerable number of speed-time curves are to be plotted for the same equipment,
the "chart of acceleration coefficients" as proposed by C. 0.
Mailloux (A.I.E.E., 1902) becomes convenient. This chart,
a sample of which is shown by Fig. 51, is plotted with rates of acceleration as ordinates and miles per hour speed as abscissas. Curve
Tn
A, is the solution of A , = o . o r q 8 K P with P = - for the partic-

lclar equipment in hand, values of tractive effort being taken for

various speeds from the motor characteristic curve. Curve A ,
may be plotted between the limits of maximum possible motor current and maximum safe motor armature speed. The dotted line
A , corresponds to the current per motor (constant or not, as
assumed) during the period in which it is under control by rheostat
or otherwise. Curve A / is the solution of Ar = o.oro98K (-f) with
f the value of train resistance for the particular equipment a t various
speeds. Curve Art is the same as Curve A , reproduced above the
X axis. Curve AF is the solution of AF = o.01q8K ( P - f ) =
0.01098KP - o.o1og8Kf, and therefore, its ordinates are the algebraic sum of the ordinates of Curves A , and A,, or the ordinate
distance (which may be measured with a pair of dividers) between
curves A , and Art. Lines are drawn above and below and parallel
to the X axis, a t ordinate distances therefrom equivalent to solutions of Ac = o.o1og8K(+_G), and marked, as a t the right of Fig.
51, with the per cent grade to which they correspond.
If track curvature be expressed as "equivalent grade," as
explained on p. 176, the chart may be used for any solution of the
acceleration formula A = o.o1og8K(P - f - c G) for the particular equipment in hand. For the rate of acceleration, with
motor current applied, a t any speed and on any (equivalent) grade,
take the ordinate value (on that speed ordinate) between curve AF
and the line for that particular grade. For the rate of acceleration
during coasting, take the ordinate value between Ar and the grade
line. Ordinate values measured below AF or A / are positive, above
are negative. The speed abscissa on which curve AF cuts any grade
line indicates the speed a t which A = 0,or the maximum uniform
speed which will be maintained on that grade, with current on.
Likewise, the speed abscissa a t which Ar cuts any grade line indicates the maximum uniform speed a t which coasting will be maintained on that grade. The following solutions have been indicated
on Fig. 5 I :
Rate of acceleration, power on, a t 33 miles per hour, on 2 per
cent down grade = 1.32 miles per hour per second (length of
ordinate ab).
Rate of retardation, coasting, a t 37 miles per hour on 1 % per
cent up grade = 0.46 miles per hour per second (length of ordinate

cd) .

Xlaximum constant speed, current on, on 0.5 per cent up grade

49.5 miles per hour (e).
Maximum coasting speed on I per cent down grade = 44.0 miles
per hour ( j ) .
=

178

ELECTRIC R W W A Y HANDBOOK

CHART OF RECIPROCALS

179

I n the practical use of the chart, the ordinate value between curve

AF or AI and a grade hne is most conveniently measured by a pair

of dividers, transfernng the measurement to the A scale a t the left

of the chart. If some care be exercised in the original construction

of the chart, and kt be made to a sufficiently large scale, rates of
acceleration may be very rapidly determined to a n accuracy of
H o 0 mile per hour per second
Chart of Reciprocals. For use in connection with the above
described chart of acceleration coefficients, M r Mailloux proposes
a "chart of reciprocals," a sample of which is shown in Fig. 52.
Rates of acceleration are plotted as ordinates (to the same scale
3.o
2.8
2.6
2.4

g 2.2
"
2 2.0

3 1.8

g 1.6
1.4

.,.1i.2

-g 10..08

'

0.8

0.4
0.2
0'

0.2 0.1 0.6 0.8 LO W L4 L6 1.8 2.0 2.2 2.1 2.8 2.8 3.0 3.2 3.4 5.6 5 d 4.0
Time Seconds

FIG. 5 1 -Typtcal

chart of rectprocals.

as that used in the Chart of Acceleration Coefficients (Fig 51))

against time values as abscissas Varlous reciprocal curves are
then drawn, the true reciprocal curve belng marked " I m l e
per hour," the half reciprocal curve "0 5 mile per hour," the twice
reciprocal curve "2 miles per hour," e t c , between the limits of
meed increments desired. The chart is so drawn to take advanspeed increment
tage of the relation time increment =
and
acceleration
shows the time required to make a given increment of speed a t a
pven rate of acceleration. I n its practical use, the length of ordinate
representing the rate of acceleration is taken off the chart of acceleratlon coefficients with a pair of dividers, one point of which is then
moved along the X axis of the chart of reciprocals untll the other
point cuts the required speed increment curve; the abscissa on which
this occurs indicates the time increment drrectly. Thus the coordinates of the tcrminat~ngpoint of a speed-time curve increment are
definitely located, obviating the possible errors likely to result from

180

ELECTRIC RAILWAY HANDBOOK

WORKING CHARTS FOR ST. CURVES

ELECTRIC RAILWAY HANDBOOK

drawing this increment a t a slope corresponding to a rate of acceleration. Obviously, the chart of reciprocals is a general one and one
such chart may be used in connection with any number of charts
of acceleration coefficients, provided only that all are drawn to the
same scale of rates of acceleration.
Tracing Method of Construction of Speed-time Curves. Where
speed-time curves are to be drawn for a considerable number of
runs with the same equipment, especially with varying track grades
and lengths of run, a most convenient method is one described by
C. 0. Mailloux (Trans. A.I.E.E., 1902). A given motor starting current having been assumed, a number of accelerating speedtime curves are drawn, for level track and various percentages of
up and down grades; these are drawn to the same scales as have been
chosen for the final curves; corresponding distance-time curves are
also drawn on the same axes (Fig. 53). On a second sheet, to the
same scales, coasting speed-time curves and distance-time curves
are drawn, for level track and various grades (Fig. 54). Some care
should be exercised in constructing these two charts accurately.
Curves on both charts should be drawn between the limits of the
maximum grades to be encountered, and a su5cient number of
both accelerating and coasting curves should be drawn so that the
CUNeS for intermediate grades may be readily interpolated with
sufficient accuracy when tracings are being made.
Having tabulated the uniform sections of track, as suggested
on p. 176, the final speed-time curve is drawn on tracing paper,
over the proper one of the two charts described. T o trace the speedtime curve for any track section, move the chart under the tracing
paper (with time axes coinciding) until the speed-time curve on the
chart for the (equivalent) grade of the track section falls on the point
of speed and time known a t the beginning of that section. Trace
this curve to the point where its corresponding distance-time curve
indicates that the length of the track section has been covered, and
repeat for the various track sections tabulated. I t is advisable
that the retardation portions of each run be plotted backward from
the stop, as described on p. 175, and here also the same method
of tracing may be used.
The final distance-time curve may also be traced from these
charts, if, after the speed-time curve section has been drawn for a
given track section, the chart be moved under the tracing along a vertical ordinate, keeping time axes parallel, until the proper distancetime curve coincides with the terminal point on the end of the curve
as drawn for the preceding section; then trace for the proper distance.
The nln curve shown in Fig. 50 was constructed by this tracing
method from the charts shown in Figs. 53 and 54. In the d a t a
assumed, and given below, the breaks in grades have hecn made
much sharper than are practicable, simply in order to bring out in a
small figure the effects of changes in grade on the speed-time curve.
The data assumed for Pig. 50 are as follows:
Sect~on

Length

o-a

a 4

0.25

&d

0.30

20

d-.
0.95
Braking to begin at 40 miles per hour.

level

Grade

cent up
)$per cent down
I per

level

CONSTRUCTION OF DETAIL S.T. CURVES

183

ELECTRIC RAILWAY HANDBOOK

184

The speed-time curve shown section.alized in Fig. 55 illustrates

the use of charts similar to Figs. 53 and 54 as described above. T h e
sections oa, bc, etc., are of speed-time curves from a Chart of Accelerations similar to Fig. 53, and the sections oat, b'c', etc., are of distancetime curves corresponding thereto. I n the final tracing, b will be
joined a t a, d a t c , etc., in the speed-time curve, and b' a t a', d' a t
I,etc., in the distance-time curve.
Distance-time Curve. The distance traveled by a train during
any period of time is roportional to the area bounded by the
speed-time curve for tKe train during that period of time, the
ordinates through the extremities of that period, and the axis of
time. This distance is shown by the distance-time curve (Figs. 47
and 58) plotted between elapsed time in seconds and feet or miles,
on the same axis of time as that used for the speed-time curve.
Detenniaation of Distance from Speed-time Curve. A method
of obtaining the distance traveled during any elapsed time is
as follows: The portion of the speed-time CUNe for this periotl of
elapsed time is divided into increments of length depending upon
the degree of accuracy to which the speed-time curve is drawn
and the degree of accuracy desired. The distance increment
corresponding to each of these speed-time increments is calculated as below, and all the distance increments for any elapsed
time are then added together.
3600
in which d = distance increment, corresponding to any increment
of the speed-time curve, miles
S , = mean speed of train during that increment, miles
per hour
(t2 - t , ) = time elapsed while that increment accrued,scconds.
The same result may be attained by the use of a planimeter,
integrating the area above described, when

ats
--

3600
distance increment, as above, miles
number of seconds per unit of ordinate
s = number of miles per hour per unit of abscissa
a = area, in square units.
Units may be inches, centimeters, etc., as desired, but t, s, and a
must all refer to the same unit.
Graphical Construction of Distance-time Curve. The following
graphical method for the construction of the distance-time curve
is adapted from a description by J. G. Pertsch in the Sibley
Journal of Engineering in 1910. The process will be described for
one short section of the distance-time curve, j k (Fig. 56), assuming
that the speed-time curve is complete and that the portion O j of the
distance-time curve has been drawn. Through various points
a, b, c, etc., on the speed-time curve (including especially the points
of maximum and minimum speed) draw lines parallel to OX and
interceptingoy a t d, e, f,etc. Draw linesg j, hk, etc., perpendicular

where

=
t =

DISTANCE-TIME CURVES

185

to OX,and a t such positions that the small triangles formed on each

side of such lines by the speed-time curve and the horizontal lines
previously drawn, will be equal in area. (Note that triangle m
should equal triangle n, and v = w, but that triangles n and v are
not necessarily equal.) Locate point P, to the left of 0, and on OX
rolonged, as follows:
t d = any distance, miles
s = the speed, miles per hour, whose ordinate value on the
speed scale is the same as that of d on the distance scale

L?

OP

50

in seconds
value of t, on time scale.

40 30 20 10 0 10 20 30 40 50 80 70 80 90 100 110 1 1
Time. Seconds

PIG. 56.-Graphical

construction of distance-time curve.

Then draw the various lines Pd, Pe,Pj,etc. The average slope
of the distance-time curve between the perpendiculars g j and hk
will then be the same as Pe,and the portion j k may be drawn as a
straight line parallel to Pe and joining the portion Oj previously
drawn. The final distance-time curve will be a smooth curve
drawn tangent to the short straight line sections such as j k .
Current-time Curves. A current-time curve is drawn to indicate
the current per motor, current per car, or current per train a t any
instant. It is plotted between time values as abscissas and current in amperes as ordinates, and for convenience is constructed
on the same time axis with the speed-time and distance-time curves.
(For typical examples see Figs. 38 and 47.)
Current Curve for One Motor.
( a ) During the Sfraighl Line Accelerafion Period. During straight
line acceleration the average value of current will remain practically constant a t the value, shown by the motor characteristic curve,
necessary to give the required tractive effort for that rate of
acceleration, and it usually is so plotted (as constant). Actually,
the current as well as the accelerating rate will fluctuate above and

186

ELECTRIC RAILWAY HANDBOOK

below the average, inasmuch as the controlling resistance is cut

out in finite steps, but i t rarely serves any purpose to introduce
into the curves the refinement of following these fluctuations.
(6) During Acceleration Beyond the Straight Line Accekratwn
Period. From the time full voltage is applied to the motor until
the current is cut off, the value of the current a t any instant is given
by the motor characteristic curve a t the speed for that instant.
Current P e r Car. (Two- and four-motor equipments.) T o plot
the current per car the value of current a t any instant (for similar
motor equipments) is found from the value of current per motor
a t that instant, as follows:
( a ) TWO-motorEquipment (Series parallel control)
Series (current per car) = (current per motor)
Parallel (current per car) = 2 X (current per motor)
(b) Four-motor Equipment
Series (current per car) = (current per motor)
Series-parallel (current er car) = 2 X (current per motor)
Parallel (current per car7 = 4 X (current per motor)
E
A
T i e of Change from Series to
Parallel. I n series-parallel control,
the swing from series to parallel
should be made a t the time when
the counter electromotive force plus
the drop in one motor equals onc
half the line voltage. When the
current is being kept constant through
c
straight line acceleration, the voltage
across each motor rises in a straight
line from the instant of starting when
it is eaual to the I R drop, to the end
I
1
of straight line accelera'tjbn when it
0
S
is equal to the line voltage. The
Tims
relative times in series and in parFIG. time for series-par- allel may be determined as in Fig.

allel swlng.

57,
.. .

where OE and P A = line volts

OV = 34 line volts
OZ = ZR drop in motor
OP and EA = time for straight line acceleration
OS and VX = time for series operation of motors
X = point where line I A cuts VX.
Current Per Train. I n plotting the current per train having
more than one motor car the value of the current per train a t any
instant is equal to the sum of the current for all the motor cars a t
that instant.
Potential Curve. The line potential a t the car, or potential
across motor terminals, is generally plotted between time in seconds
as abscissas and volts as ordinates on the same time axis as is used
for the current curve. The value of the potential a t the car a t
any instant depends upon the potential of the generating station
and substation, the regulation of the generator, the efficiency and

ELECTRIC RAILWAY WANDBOOK

188

regulation of transformat~on,transmission, conversion and distribution, and load on the system I t s value may be obtained practically only by test. For preliminary work a n average voltage
a t the car usually is assumed to be constant; the corresponding
curve is a straight line arallel to the axis of time.
Power and Power A c t o r Curves. The poHer curve may be
plotted on the same time axis w t h other curves as above, and its
instantaneous ordinate (usually to a scale of kilowatts) is a t any
point proportional to the product of potential and current per motor,
car or train a t that instant. I n the case of alternating-current
equipments, the power factor must, of course, be used in the
determination of the power ordinate values, and in such cases,
the power factor curve usually is also plotted on the same time axis,
the instantaneous values being determined in connection with the
motor characteristic curve, where the power factor of the motor is
shown plotted against the current values.
Speed-distance Curves. Most run curves are plotted with
time as abscissas, as previously described, but in some cases, such
as when signal locations are being determined, the speed-distance
curve, with distance as abscissas, is of great value. The construction of speed-distance curves is preceded by that of speed-time and
distance-time curves, the combination of the data from which enables
the speed-distance curves to be plotted, point by point. Fig.
58 shows speed-time, distance-time and speed-distance curves
for a given run. I t will be noted that constant rates of acceleration
and braking, shown by straight lines in the speed-time curve, are
represented by parabolic curves in the speed-distance curve.
Traction Power Requirements
Value of Power Input Required by Train While Moving at Constant Speed. The approximate value of power input required
by a train moving a t a constant speed may be obtained by the formula:
- FWS
p = - -2FWS
1oooU
soou
in which P = approximate value of power i n p d required by
train, kw.
S = constant speed of train, miles per hour
F = total constant train, track curve and grade resistance (f C G ) for the train while running a t
speed S, pounds per ton weight of train
W = total weight of train, tons
U = efficiency of motor equi ment a t the proper speed,
decimally expressed. ~ E i may
s
be taken f r ~ mthe
characteristic curve for the motor equipment, or
from table, page 190

+ +

NOTE.Since

the value of F is always a n approximation, the use

of 2 instead of the more nearly exact value I 99 for the
constant is justified in this and similar formulas herein.

TRACTION POWER RBOIJIREMRNTS

189

Value of Power Input Required by Train While Moving at Constant Speed on Tangent Level Track. When the train is moving a t a
constant speed on R tangent level track C and G are zero and the
formula is

in which

= constant train resistance for the train while running at speed S, pounds per ton weight of train
P,W,S and U = same as given In paragraph immediately preceding.

The following tables show the power input required for motor-car
trains and for locomotive passenger train operation, a t constant
speeds on tangent level track. The tables are based on the above
formula for power input and the Armstrong formula for tram resistance. The efficiency of geared motors a t full speed is taken as 75
per cent for the motor-car trains; that of direct-current gearless
motors as go per cent for the locomotives. Cross-section of cars is
taken as listed on page 129.

TRAIN INPUTFOR CONSTANT-SPEED

RUNNINGON
LEVELTRACK,MOTOR-CAR
SERVICE

TANGENT

(Input values expressed tn k~lowatts)

Specd (mrles per hr )

Tram wetght

20 ton
3 0 ton
4 0 ton
5 0 ton
60 ton

10

120

I I 1
30

40

1
5 0 ! 6 O l 7 O I 8 O / 90

I100

--

car
car
car
car
car

2-20
2--30
2-40
2-50

ton
ton
ton
ton

cars
cars
cars
cars

3-20
3-30
3-49
3-50
3-60

ton
ton
ton
ton
ton

can
cars
cars
cars
cars

5-20
5-30
5-40
5-50
5-60

ton
ton
ton
ton
ton

can
can
cars
cars
cars

..
..

190

ELECTRIC RAILWAY HANDBOOK

TRAIN INPUT
FOR

CONSTANT-SPEED
RUNNINGON TANGENT
LEVELTRACK,LOCOMOTIVE
PASSENGER
SERVICE
(Input values expressed In kdowatts)

1 1 1

Gross

train
weight

zoo tons
3ootons
400 tons
500 tons
600 tons
700 tons
800 tons
900 tons
1,000 tons

17 5
26.1
32.0
41.0
50.0
59.0
69.0
77.0
85.0

Speed (miles per hr.)

3

4
,

1 I 1 I 1
60

5
.

277
41 2 75.8 I24 I %
60.5108.0172258
369
79 o 139.0 219 330
462
9 9 . 0 170.0 268 395
563
118.0 203 o 315 464
645
135.0 235.0 363 530
735
828
154 o 267 . o 410 5%
173.0 300 o 459 663
920
193.0 330.0 507 730 1,011

398
511
635
755
875
996
1.117
1.238
1.358

80

9~

530
689
840
995
1,150
1.305
1.460
1.615
1.771

710
905
1.100
1.295
1.189
1.682
1.878
2.070
2.261

: I

920

r,1co
1.405
1.645
1 . 8 ~ ~
2.132
2.37~
~ , 6 2 ~
~ . 8 6 ~

Value of power input required by train during straight line

acceleration for direct-current equipments with rheostatic c o n t r ~ l ,
L

5oou

ih which P

= approxiinate average of power input required dur-

Fa

= gross tractive effort of motors during straight liqe

ing straight line acceleration, kw.

acceleration, pounds per ton weight of train (see
page 15:)
W = total we~ghtof train. tons
S = speed at-instant when rheostat is entirely cut out,
miles per hour (see page 168)
U = efficiency of motor equipment, a t a load corresponding to the required gross tractive effort per motqr
decimally expressed. See motor characteristic t,;
table below.
Note that for the three cases of ( I ) parallel, (2) motors two in
series and (3) motors four in series, the speed S should corresporid
to the instant when rheostat is entirely cut out with such motbr
combination. As a rough approximation, the power input in
cases (2) and (3) are about Y and x , respectively, that in case ( I ) .

Efficiency a t full speed, per

i

* Gearless.

The maximum efhciency of railway motors of the geared tybe

occurs during maximum output, hence the values so quoted should
be used for calculation of the power required to accelerate a car or
train. An exception to this rule may be taken in locomotive work,

TRACTION POWER REQUIREMENTS

191

where it is customary to force the motors to nearly their maximum

rated output even after the train has reached its normal maximum
speed.
The efficiency when the car is a t full speed is generally lower with
motors of the geared type, owing largely to the losses in gears and
also in the magnetic circuit of the motors themselves. This lower
efficiency a t higher speeds does not hold true of motors of the gearless type.

FIG.60.-Power

required per ton weight of traln.

Graphical Solution of Power Equations.

general gr~phiralsolution of the equation P

Figs. 59 and 60 give n

FS
for the amount

5@'u

192

ELECTRIC RAILWAY HANDDOOR

of power required per ton weight of train. Fig. 59 is for values

of F up to 50 Ib. per ton. For speeds greater than 7 0 miles per hour
(given specd)
use the curvc for spced =
and multiply the resulting
(a
- ronsl:tnt)
.....- ......,
kw. pcr ton weight of train by t11at ons st ant. Fig. 60 is to be used
in determining the power necess.try while the train is accelerating
or in othcr cases where the valuc of F is greater than 50 Ib per ton
weight of train.
Power Required at Substation. The average or instantaneous
value of po\ver required a t the substation bus bars is equal to the
average or instantaneous value, respectively. of power required
a t the trains taking power from that substation divided by the
average efficiency (decimally expressed) of the distribution ant1
working conductor systems connecting the trains to the substation. In approsimating the power required at a given time on a
given section over which the trains have regular running times and
all definite stopping places, such as turnouts and stations, the
train sheet or graphical time table (see p. 120) may be used to
advantage as it will give the number and location of trains on the
section a t any time. This information together with the run curves
for the trains and runs included in this section cvill make possible
the closest approximation to the value of power required a t the
substation bus bars a t any instant. Where trains are not operated on regular running times with definite stopping places, so close
a n approximation to the value of power required a t any instant is
impossible, but the average value of power required during any
period may be approvimated to a degree of accuracy depending
upon the regularity, frequency and uniformity of operation over
this section. The average value of power per train required may
be determined from the run curve for an average train and typical
run on this section. Multiplying this average value by the number of trains on the section during the given period will give the
approximate average value of power required a t the substation
throughout that period. Any change in the headway, speed or
weight of trains will tend to change the power requirements of the
given section. For sections on which such changes are brought
about periodically, as in putting on extra cars and adding trailen
during rush hours, the power requirements may be calculated for
each of the different conditions of operation.
A rough general rule is to install one kilowatt in rated substation
capacity for every two horsepower in car equipments drawing
more than one-half their total current from that substation, and
one kilowatt per four horsepower for those cars which are taking less
than one-half current from that substation.
Power Required at Generating Station. In the operation of a
system without substations the value of power required a t the
generating station bus bars may be determined in the manner just
outlined for the determination of power required a t the substation
bus bars. The average or instantaneous value of power a t the
generating station bus bars required by a substation is equal to
the rorresponding average or instantaneous value, respectively, of
powcr required at the substation bus bars divided by the combined

ENERGY CONSUMPTION

1-

efficiency (decimally expressed) of substation, transmission from

generating station to substation and generating station transformation. The average or instantaneous value of power required
a t the generating station bus bars t)y srvcral substations is equal
to the sum oi the correspontling avcragc or instantaneolts power
values. resl)ectivcly, requircd a t thc generating station bus bars
by all the substations.
A rough general rule is to allow one kilowatt of generator capacity
for every two horsepower of rated motor capacity in car equipments.
This rule is based on the fact that the continuous current capacity
of a railway motor is about one-half of its rated one-hour capacity
and that the average efficiency of a complete well-designed system
is about 70 per cent or 75 per cent from power house to car, so that
if a motor is operated with a fair margin, one horsepower input
to the car equipment is equivalent to one kilowatt output from
the gencrator.
Load Curves. The load curve for a substation or generating
station may be approximated by plotting the instantaneous or
average values of power required a t the bus bars of the station
against elapsed time and connecting adjacent points so plotted. If
the loatl curves for the substations arc first constructed they may be
convenient in applying the method outlined in the second preceding
paragraph for dctermining the p w e r required a t the generating
station. A load curve for an existing load may be automatically drawn by a graphic recording wattmeter. A curve so
drawn gives instantaneous values of power, and from it average
values of power may be determined in the manner outlined for
determining the average value of power required by a train.
Energy Consumption
Energy Consumption for Traction. The net energy available
for traction remaining after the control, wiring, motor and mechanical transmission losses in the train have been supplied is consumed in the two following processes: lint, balancing train, track
grade and curve resistance; second, heating brake shoes and car
wheels while braking. The first continues while the train moves,
the second only while the train is in motion and brakes are a plied.
Net Traction Energy Consumed in Moving Train When grakes
Are Not Applied.
5280 X 745.6 F1vL
60 X 33,000
= 1.99 F1IrI. = approximately zFWL

"=

in which
r r = energy necessary to 1)alance train, track curve and

grade resistances while the train travels the distance considered, watt-hours
F = total average train, track curve and grade resistance
(f (;
C') for train while traveling the given
scction, pounds per ton weight of train

+ +

194

ELECTRIC RAILWAY HANDBOOK

L = distance traveled, miles

IF' = weight of train. tons.
NOTE: Since the valuc of F is always an approximation, the use
of z instead of the more nearly exact value 1.99for the
constant is justified in this and similar formulas herein.
Net Traction Energy Consumed Wile Train Is Coasting or While
Moving with Brakes Applied. The value of the total net traction
energy consumed by a train in passing with power off over a section
between two given p i n t s is e q ~ ~ to
a l the difference between Lhe
total values of kinetic energy of the train at these two p i n t s , that is
En = K t - R1
in rvhirh
1':=
~ net traction energy const~metl while brakes are
applied or train is roasting. w;~tt-hours
El = total kinetic energy of train a t brginning of section
considered. (See IS. 1)eIow)
E2 = total kinetic energy of train at en(l of section considered. (See En below)
En = ~oooll'lr
in which
E. = total kinetic energy of train a t any instant ? I , footpounds
11. = total weight of train, tons
11 = kinetic energy head for train a t the instant 18, feet.
Expressing kinetic energy head in tcrms of speed of train (see
p. 149)and reducing Emto watt-hours,

in which B. = total kinetic energy of train a t any instant n,

watt-hours
S, = speed of tmin a t that instant rz, milts per hour
K = ratio of linear inertia to total inertia of train. (See
p. 150)
Sul~stitutingE, in the formula for EB,
11-St? Il.S2
fia = -39.75K 39.75K

in which 1V = total weight of train, tons

SI = speed of train a t beginning of section considered,
miles per hour
S, = speed of train a t end of section consiclererl, miles
per hour.

ENERGY LOSS I N BRAKING

195

Net Traction Energy Consumed in Heating Brake-shoes and

Car Wheels in the Process of Bfaking.
Eb = 1 ' ; ~- c?

= --- (SS2- Sa2) - zFN'IA

39.75K
(S32 -- 0.0252W
K

or
in which

- 2Fff'h

1V = total weight of train, tons

Ii = ratioof linear inertia to total inertia. (See p. 150)
St = sl,cctl of train a t Iscginning of brak~ng sectlon
consitlcred, miln per hour

S , = speed of train a t end of braking section ronsiderctl, miles per hour. When train is brought to
a standstill by braking, S4 becomes zero
I4 = length of braking section, miles
F = total average train, track curve and grade resistance
(/ +_ C;
C') for train while traveling the braking
section, pounds per ton weight of train.
Approximate Formula for Energy Consumed in Heating Brakeshoes and Car Wheels
-,
in tbe Process of Braking. I t is often assumed
that while brakes are bcing applied the kinetic
energy of rotation of the
rotating parts just balanccs the train rcsis60
tancc, track curvc. and
gradc rcsistanccs (if the a
latter bc small). This 210
assumption gives rise to
the following approximatc formula:

&

1v

-- (S,' - S," or

40

(S3( S,?)
When brakes are applied until train comes
to a standstill, the formula becomes I<b =
0.0251v

-iVS12
or 0.025
40

!
2

rn30
20

W S ~ in
?
0

which the significance

Mi. ~ c Br.
r
8 p p d r l wbich Brakes are ADDllcd
of the symbols is the
FIG. 61.-Losses in brak~ng.
same as in the prcceding paragraph. Fig. 61 gives a graphical solution of this formula.
Total Net Traction Energy that must b e Available for Traction
whiie Train is Running with Power on Between Two Points. The

ELECTRIC RAILWAY HANDROOK

196

value of this energy is equal to the sum obtained by adding the difference between the values of the kinetic energy of the train a t these
two points to the net traction energy consumed by the train in
passing between the two points. I t is expressed by the following
formula:
e,

= -( S G~ Ss2)

+ zFWL
+ zFWL

39.75K
0.0252W (S2 - S*')
=-

in which

total weight of train, tons

(See
P. 1 5 0 )
Sa = velocity of the train a t the beginning of the section
over which the train is considered to run with power
on, miles per hour. The value of Ss is zero if
train starts from rest a t this point
Sc = the velocity of the train a t the end of the section
over which the train is considered to run with power
on, miles per hour
F = total average train, track curve and grade resistance
(f G C) for train while traveling the section
between the two given points, pounds per ton
weight ~f train
L = length of section between the two given points,
miles. This is the length considered over which
the train runs with power on.
Total Net Traction Energy Consumed during any Run.
Care I. Train coasts to standstill, brakes not applied.
Energy consumed = ep (See p. 193)
Case 2. Brakes applied during any part or number of parts of
run.
BB
Energy consumed, kilowatt hours = ep
in which
ep = energy consumed while brakes are not applied,
kilowatt hours. (See p. 1g3.)
Where several sections over which the train runs without bmking occur during the run, n is cal~ulatedfor each of these sections
and these values of ep are added together to give the value of ep
for use in this formula.
EB = energy consumed while the brakes are applied,
kilowatt hours. (See p. 194.)
Where several sections over which the train runs with brakes
applied occur during the run, EB is calculated for each of these sections and all those values of En are added together to give the value
of EB.
Approximation of Energy Consumption. For preliminary
determinations, approximate energy consumption may be obtained
by use of the general energy output charts describeti bt!ow, or by

= ratio of linear inertia to total inertia of train.

+ +

ENERGY CONSUMPTION

197

ELECTRIC RAILWAY HANDBOOK

198

use of the general operating characteristics shown in Section EV,

Figs. 1-4, inclusive, and described on page 224 and following
General Energy Output Chart. The general energy consudPtion charts, Fsgs. 62 and 63, are from the Standard Handbook and
are based on speed-time curves in Fig 46, and the above formllla
for total net traction energy. A constant braking rate of 150 lbper ton weight of train is assumed in both, and a constant train
resistance of l o and 15 Ib. per ton weight of train is used in the
curves of Figs 62 and 63 respectively. These curves give approxlmate values of energy &fit
watt-hours per ton mile, for a z-mile
run a t various gross tractive efforts for acceleration and times required to make the run; these values must be divided by the e8iciency of acceleration (see below) to get energy input. I n using the
general energy output charts the approximate value of watt-hot'"
per ton mile is given by the ordinate of the point on the p r o p =
gross tractive effort curve, whose abscissa is equal to T I in the
following formula. The curve selected should correspond to the
gross tractive effort during straight line acceleration.

TI =

T
7
V L

in which T = time to make run under consideration, seconds

L = length of run under consideration, miles
Efficiency of Acceleration. The efficiency of acceleration
is the ratio of the net energy available for traction (output) to
the gross energy delivered to the train in furnishing that net ener6Y.
It is-usuallyexpressed as a per cent. I t depends upon the eficienc?
of control. efficiency of car wirinn. motor efficiency. efficiencv of
mechanicil transmission, and the i;kr cent of straight'line accefetation(see p 199) obtaining during the parttczilar run under conriderafiOnEfficiency of Acceleration with Series Motors and Three-phase
Induction Motors. The following table from data given in the standard Handbook gives the per cent efficiency of acceleration a t various
per cents straight line acceleration for direct current series motors,
also for three-phase induction motors (connected in parallel)

,-

per cent straight

line acceleration

--

Rheostat~~
E
;roJ.i

Series
pardlel.
per cent.

Four series,
two qerles,

t,$!zt.

T~~~~~~~
moton.
parallel.
per cent,

43
46
49

56 o
59 o
62 o

60 5
62 5
65.0

40
43
46

70

52
56
58

61

67 o
69 o

67 o
69 o
71 o

49
sz
55

40
30
ao

62
65
68

71 0
72 5
73 5

72 5
73 5
74 0

58
6r

64

10
0

72
75

74 5
75 0

74 5
75 0

i2
75

100

90

80

60
50

ENERGY CONSUhIPTION

ESciency of Acceleration with Single-phaseCornmutating Motor

Euuivments. The efficiency of acceleration for trains e o u i o ~ e d

wkh*single-phase commutaiing motors may be taken at'

to
75 per cent. This value is a close approximation for the class of
service to which such cquipments are most often used, z e , a n
express or semi-express service on private right of way, with
necessarily small per cent of straight line acceleration.
Per Cent Straight L i e Acceleration. The per cent straight
line acceleration of a train on a given run is the ratio of the time
during which the motors take constant current to the total time
current is supplied to the motors, expressed in per cent. The
per cent straight line acceleration for runs of greater than a mile
in length may be less than 10; on shorter runs it is greater and may
even exceed 75 for very short runs. I t may be taken directly from
the speed-time curves for any particular case.

Btops per Yllo

FIG. 64.-Ernplrlcal

constants for energy appro~~rnatlon

(Del MarWoodbury).

Approximate Determination of Energy Consumption. The

following method and constants for determining the approximate
amount of energy consumed in train movement are by JV. A.
Del Mar and D. C. Woodbury.
Significance of symbols.
S = average running speed, excluding stops, miles per
hour
S,,, = maximum speed, miles per hour
L = distance between stops, miles
L, = distance traveled with power on, miles

number of stops per mile including one terminal

f = average train resistance, in pounds per ton (say

that corresponding to a speed of ro to 20 per cent
greater than the average speed)
C = average equivalent gradc, per cent (see page 163)
D = average curvature, degrecs

200

ELECTRIC RAILWAY HANDBOOK

= average motor and controlkr efficiencies as a decimal,

s,
= - (see Fig.

which is usually about 0.7

64)

L
M = - (see Fig. 64)
L,
OUTPUT
AT WHEELRIM A N D INPUTTO CARIN W,ITT HOURS
TON MILE

Due to kinetic energy

Due to grades..
Due to curves

S,'
36.zL

..

-I

.........

??J
2s

.SD~LP
L--

z.~d
M

57e

~_9~%&

- iT

........

Total.. . . . . . . . . . . . . . . .

PER

Actual energy output appro xi mat^ electrical

at wheel nms of c a n energy inpdt to Cars

......

Due to train resistance.. .

?.sS

L
Sum of above

Sum of $hove

In th'ls tablethe k t column gives the mechanical energy require&

a t the wheel rims. In the second column a n allowance of 70
per cent for e5ciency of motors, gears and control has been made.
(For values of J and M see Fig. 64.) I n cases where S+ and 4,
are known, more accurate results can be obtained by using them In
the formula of the first column.
Example: A multiple-unit train makes a n average speed, while
running, of 30 miles per hour, making a stop every 2 miles. The
average train resistance is 7 Ib. per ton, the average equivalent
grade 0.05 per cent, and the average curvature 0.75 deg.
Then S = 30
f = 7
G = 0.05
D = 0.75

N=

0.F

1 ] from Ti.

64.

Then, in watt-hours per ton mile,

Kinetic energy

1.36' X 0.5 X 30'

--

ZF

2.85 X 7---I .67

s7 x ' 0 ~ 5
1.67

Train resistance energy

Energy for grades

Energy for curves

= 2'85

o"S

I .67

= 33.3
=

12.0

1.7

1.3

Total energy input a t train = 48.3

Energy Consumed in Rheostat. The energy consumed in the

series motor control rheostat and appearing as heat therein depends
upon ( I ) the voltage drop across the rheostat, ( 2 ) the current
flowing through the rheostat and ( 3 ) the time during which this
current continues to flow.
Below are given approximate formulas for calculating the
energy consumed in the series motor control rheostat during start.
Cases I to 3 , inclusive, are for equipments arranged to give a
uniform rate of straight line acceleration. Case 4 is for equipments
arranged to give two rates of acceleration, viz., one rate of acceleration while half the motors are in series with the other half, and
another rate of acceleration while all the motors are in parallel.
(See P. 1 ~ 5 . 1
Since the resistance is arranged in steps, each of a finite value,
the value of the current will fluctuate above and below the average
as the controller is advanced and the motor speeds up. When the
acceleration approximates a constant rate, the mean of these fluctuating values of current will approximate the value of current
necessary to accelerate the train a t the constant rate.
For convenience in using these formulas, they are given for
rheostat energy consumption per motor. The second formula under
Case 2 is an exception. I t gives an approximati011 to the loss per
ton weight of train.
Signig.amtz nf Symlnl.~.
eR = energy consumed in rheostat during one straight line
acceleration period, watt-houw per motw
T = duration of straight line acceleration, seconds
I = constant or average current per motw during period of
straight line acceleration, amperes
E = working conductor (trolley or third rail) potential a t
the motor car or locomotive, volts
R = resistance of the motor, ohms. (See pp. 254-259.)
Case r . Otze or more motors, rheostatic control. All motors in
parallel throughout straight line acceleration period.
I T ( E- IR)
eR = 2

X 3600

Case 2. T z o - and fozrr-motor equipments, series-parallel. Twomotor equiptnents having motors in series throughout first part of
straight line acceleration period and in parallel throughout the
remainder of straight line acceleration period. Four-motor equipments having two groups of two motors in parallel in series throughout the first part of the straight line acceleration period and all
motors in parallel throughout the remainder of the straight line
acceleration period.
2ZzR2)
eR = -I T-( E 2 - z E I R

4 X 3600 ( E - ZR)

AVT

or (approximate formula), L = 3 ~ 0 0 ~

202

ELECTRIC RAILWAY HANDBOOR

in which

L = approximate loss in watt-hours per ton

A = current per car with series position of controller
V = M line volts - drop in motor
T = time of straight line acceleration in seconds
= weight of car in tons.
Case 3. Four nzotors, series, series-parallel, pardlel. All motors
in series throughout first part of straight line acceleration period,
two groups of two motors in parallel in series throughout the
second part, and all motors in parallel throughout the remainder
of the straight line acceleration period.
IT(3E2 - 4EIR
812R')
e~ = 16 X 3600(E - ZR)
Case 4 . Two rales of slraight line acceleralion (see p. 1551,
i.e., two values of constant current per motor during straight line
acceleration. Arrangement of motors same as Case 2.
Signijcan, e of Symbols, Case 4.
e~ = energy consumed in rheostat during one straight line
acceleration period, watt-hours per motor
TI = duration of first rate of straight line acceleration,
seconds. (Series connection)
T, = duration of second rate of straight line acceleration,
seconds. (Parallel connection)
I, = constant current per motor during period (TI) of
first rate of straight line acceleration, amperes (series
connection)
1 2 = constant current per motor during period ( T , ) o f
second rate of straight line acceleration, amperes
(parallel connection)
E = working conductor (trolley or third rail) potential
a t the motor car or locomoiive, volts
R = resistance of the motor, ohms. (See pp. 254-259.)
IITI(E- ~ I R ) ZzTLE
ER =
4 X 3600
Energy Consumption Reduced by Use of Grade a t Approach to
Station. A considerable saving of energy may be brought about by
causing a train to lift itself, a t a station, to an elevation greater
than its elevation between stations. By this process much of
the energy that would be dissipated as heat in brake-shoes and car
wheels in bringing the train t o a stop while maintaining a practical
schedule speed on a 'eve1 track is converted into potential energy
available to aid in accelerating the train while it departs from the
station. This method of conserving energy is especially applicable
to operation on elevated railways and subways, where the track
grades are to a greater evtent under the control of the engineer than
is generally the case with the surface railway. Fig. 6 5 shows the
speed-time curve and energy-consumption curve for a 154-ton train
on a typical average run of the Manhattan Elevated system, N. Y.,
and the corresponding data for the same train, with lighter motors

ENERGY SAVED BY GRADES AT STA'MONS

203

as explained further on, running the same distance on a track having

a 3 per cent up grade a t the approach to the station, and a 3 per cent
down grade a t the departure from the station. I t is to be noted
that the train operates a t the same schedule speed in both cases,
but in traversing the run having the grades it consumes much less
encrgy than i t docs in traversing the all level run. The rate of
straight line acceleration is the same in both cases. I n this example,
the 3 per cent grade a t departure from the station furnishes a force
of approximately 9240 Ib. of the 18,620Ib. (if we neglect rotational
inertia) necessary to accelerate the train a t the required rate of 1.33
miles per hour per second. The motors are then called upon to
furnish an accelerating force of approximately 18,620 - 9240 =
9380 Ib. or about half t h a t necessary on the all level run, SO they
need have only about 50 per cent the capacity required for the all

3 s Dowu Grade
Level
3 %Up Grado
RO Peer
1024 Feet
no Peer
FIG.65.-Example of energy saving by use of grades at stat~onstops.

level run. I n constructing the power curve, Fig. 65, for the train
operating on the run having the grades, i t was assumed that the
capacity of the motors and consequently their torque a t any speed
are one-half those for the all level run. A comparison of the
time areas indicates that in this case operation on the run
the grades demands only a b u t half the energy necessary for operation on the all level run.
Regenerative Braking. I t has been shown that the energy loss
in heating brake shoes and wheels forms a considerable part of the
total required for traction. Inasmuch as the electric motor is
reversible, that is, can give out electric power when mechanically
driven as a generator, i t is possible, with the proper control, to brake
a motor car or electric locomotive electrically, or nuke use of "regenerative braking," with reduced expense of brake shoe and wheel
maintenance, besides returning to the line a considerable portion
of the energy stored in the train during acceleration. Opposed to
the advantages of regenerative braking, especially as applied to
single car and rapid transit operation, are several facts. Some
arrangement of shunt winding or separate excitation must be applied

K%i

ELECTRIC RAILWAY

HANDBOOK

to the common direct current railway motor, which is of the series

type, to enable it to act as a generator. Further, the thermal
capacity of the motor must be increased due to the extra motor
losses entailed in electric braking and the less proportionate amount
of time with no load which may be utilized for cooling. The
additional weight of the excitation and control equipment often
will further increase the required capacity of the motors, and will
add considerably to the first cost. Even in frequent stop service,
not much more than r 5 to 2 0 per cent of the gross input can be saved
by regenerative braking, which may be more than balanced by the
first cost and cost of maintenance of the necessary increased motor
capacity and auxiliary apparatus. As any system of regenerative
braking depends upon the counter e.m.f. of the revolving armature,
i t is evident that when approaching zero speed there will be no
torque developed by the motor, and hence, regenerative braking
must be used in connnection with air brakes, with resultant decreased economy. For street car and rapid transit service, including interurban service, it can be said that there is small attraction
in regenerative braking, not because i t is not possible, but because
the added complications and expense bring no adequate return in
the cost of the energy saved.
Regenerative braking used in connection with locomotives for
heavy passenger or freight service on mountain-grade sections
offers many advantages far outweighing in importance the possibility of a small amount of energy saved. One of the disadvantages
attending the operation of long heavy trains on mountain grades
is the danger incurred in braking trains on down grades. More
accidents occur when trains are running down grade than when
operating up grade, due to the possibility of overheated brake shoes
and car wheels and also breaking apart of long trains when brakes
a5e released momentarily for the purpose of recharging the train
plpe. I n such service, a regenerative system of control offers
safer means of holding trains on down grades than exists with
air brake control, and its claims in this direction make it worthy
of very careful consideration for this class of service. The advantages have been listed as follows: saving of approximately 15 per cent
of total power required; elimination of brake shoe and wheel wear
and brake rigging troubles with material reduction in maintenance
charges; removal of difficulties encountered in operation of long
heavy freight and passenger trains on long grades due to inherent
operating characteristics of air brakes; reduction in wear of track
on grades and severe curves; increased safety to passengers and train
crews due to du licate braking systems; increased comfort to
passengers and reBuced wear on equipment due to constant speed
on grades and uniform braking when slowing down for curves and
stops.
Direct Current Regeneration. The system of regeneration used
with direct current railway motors by the Westinghouse Electric &
Manufacturing Company is diagrammatically shown as in Fig. 66.
This figure shows the armature of the main motor, A , the main
motor field, F , the balancina or stabilizing resistor, R, and the exciter, E. Inherently, a series direct current generator has no sta-

REGENERATIVE BRAKING

205

biiity of voltage, and some external means is necessary to control i t

when regenerating. I n the figure, the exciter, E, furnishes current
which flows in the path indicated by the dotted arrows. This
current excites the main motor field and causes the armature, A , to
generate a voltage opposed to the line voltage. The voltage of
E can be regulated either by regulating its field strength or by
varying resistance in series with its armature. When the voltage
of A overcomes the line, a current starts to flow from the ground
through R and A to the iine. Thus R carries both the main
regenerated current and the exciting current.
The stabilizing effect comes from the resistance R, since any
variation in the regenerated current will vary the drop across the
resistance and thus, assuming the voltage of the exciter, E, to be
fairly constant, will have a n immediate effect on the field of the
main motor to counteract such a variation.
For instance, if a train is regenerating down
grade with the exciter voltage set so as to give
the proper excitation to the main field to furnish
the necessary braking effort, and the line voltage
should suddenly drop, the regenerated current
would immediately tend to increase. However,
the increased current which tends to flow increases the voltage drop across the resistance R,
and directly reduces the exciting current, thus
maintaining a balanced condition as before. I t
is important in a direct-current regenerating system to have this regulating influence as nearly
direct as possible, since changes in line voltage
may a t times be so great and so sudden that unless the regulation is simultaneous with the varipic. M.-Direct
ation, a condition of unbalance between the current regeneraarmature and field of the motor may be sufficient- tive breaking.
ly great to cause flashing a t the commutator.
The excitation for the main motors is supplied by the separate
generator, E, driven either by a motor or from an idle locomotive
axle through gearing. The regulation of the braking is then obtained by varying the exciter field strength or by varying the
amount of resistance in the stabilizer circuit.
Fig. 67 shows the relation between speed, retarding effort and
am eres regenerated, plotted on the basis of constant line voltage.
Mifcs per hour and amperes regenerated are both plotted as ordinates against retarding effort as abscissas. The curves shown are
for a 3000 volt direct-current locomotive which has six 1500 volt
d r i v i ~ gmotors that regenerate in three different speed combinations.
The performance for the intermediate speed combination (of two
parallel circuits of three motors each) is shown in Fig. 67. The
numbers on the curves refer to the different control notches, the control in this case being accomplished by varying the stabilizing
resistance in steps. As long as the controller is on a certain numbered notch, the correspondingly numbered speed and current
curves indicate the regenerative erformance. The kilowatts
returned to the trolley may be foungfor any point by multiplying

206

ELECIRIC RAILWAY HANDBOOK

the amperes regenerated by the trolley voltage. I t will be noted

that the general shape of the speed curve is upward as the retarding
effort increases. This is just the reverse of the series motoring
speed characteristic of the same motor, and is necessary for a
stable condition of regeneration. As an example, suppose that the
locomotive is regenerating on a down grade a t 2 0 miles per hour
with the controller on notch No. 9. The speed curve shows that a
retarding effort of 18,000 pounds is being exerted; the current
curve shows that 2 0 0 amperes is being returned to the line, and this
current, multiplied by the trolley voltage of 3000, indicates that the
reg nerated power is 600 kw. Should the grade become steeper
an% the speed increase, the retarding effort will also increase

Frc. 67.-Relation

between speed retarding effort and regenerated amperes

on varlous control notches In d~rcctcurrent regenerative braking.

until a balanced speed is again reached. The shape of the curve

is also an indication of the regeneration performance. A flat curve,
or one which shows a gradual increase of speed with increase of
retarding effort, indicates that ordinary changes in retarding
effort will not necessitate a change of the controller handle from
one notch t o another, unless a change in speed is desired. A steep
curve, however, or one which shows a rapid increase of speed with
increase of retarding effort, indicates that any increase in retarding
effort required will necessitate movement of the controller handle,
if the same speed is to be maintained. The curves corresponding
to the high-speed notches will be found to be steeper than those
for the low-speed notches. This is because a t the high-speed notch
the value of the stabilizing resistance is the greatest, and with a

REGENERATIVE BRAKING

a07

large value of stabilizing resistance, small changes in. retarding

effort represent large changes in speed. This is due in turn to
the fact that, with a large value of stabilizing resistance, a small
change in regenerated or armature amperes produces a large change
in field amperes.
I n any direct current system of regeneration, it is not safe to allow
the armature current to become too great in proportion to the field
current, or the armature reaction may distort the field form to the
extent of causing a flash-over. I n Fig. 67 the upper end of the curves
are discontinued a t the point where the ratio of armature amperes to
field amperes becomes 2.5 to I . This ratio is determined from
sxperience, and depends on the ratio of the armature turns to the
held turns and on the individual flashing characteristics of the
motor used.
If the pcr ccnt efficiency is desired for any point on any curve,
i t may casily I)c found as follows:
output I'er cent Efficiency .=
input
1,inc volts >( amperes regenerated
X 503
-----Miles per hour X pounds retarding effort
The curves of Fig. 67, as mentioned, refer to only one of three
regeneration combinations. The higher speed combination makes
use of tlirce parallel circuits of two armatures each from trolley to
ground. The lowcr s ~ x c dcombination uses a single circuit of six
armatures in series.
Alternating Current Regeneration. Very successful installations
of regenerative electric braking have been made in connection with
three-phase or split-phase clectrifications using induction motors
on the locomotives, such as the Great Northern Railroad's threephase installation a t the Cascade Tunnel, the Italian State Railways in Italy, and the split-phase electrification of the Norfolk &
Western Railroad. The three-phase induction motor characteristics are such that i t lends itself very readily to this method of
control, as i t is only necessary to operate the motor slightly above
the synchronous speed to cause it to work as a generator; but i t
has the serious inherent disadvantage that electric braking cannot
be obtained a t any other speed. The regenerative CUNeS of a 115
ton 6600 volt three-phase freight locomotive are shown in Fig. 68.
Effect of Varjation in Line Voltage on D.C. Motor Operation.
The speed of a series motor a t any given current is directly proportional to the counter e.m.f., that is, to the impressed voltage
less the drop in the motor itself. The line voltage impressed on a
railway equipment has, therefore, an important bearing on the
performance of that equipment. Voltage affects schedules, energy
consumption and motor heating. The effects of low voltage are for
the most part undesirable, and hence demand serious consideration.
The effect on schedule speed is the most readily observed feature
of low voltage. I t is well known that in a specific service a lower
voltage involves a lower schedule speed, or a reduction in margin for
making up lost time, or both. If service conditions are quch that an
equipment is able a t a givcn voltage to make a certain length of

208

ELECTRIC R N L W A Y HANDBOOK

run without coasting, i t is found that a n increase in voltage produces

a comparatively small speed margin on very short runs, while on
long runs for the same voltage increase the speed margin may
be relatively great. This indicates that for the reasonable maintenance of schedules, i t is necessary on suburban and interurban
lines to maintain better line voltage conditions or apply equipments
with greater leeway in speed than would be necessary on a city line.
Since the maintenance of good voltage is relatively more difficult
and expensive on interurban than on city lines, the alternative of
higher speed equipment is usually adopted.

Kilowatt* El~ctricaIOUtpUt

FIG.68.-Performace

I l l o r a t t s XechaoLcal 0utp.t

of three-phase induction motors in regenerative

bralung.

On every run, the time during which current is drawn by the

moton is made up of two distinct parts, the period of straight line
acceleration and the period of running with the controller on full.
The speed of the car before the controller is on the full parallel
position depends on the rate of straight line acceleration, and is to
a certain extent independent of the line voltage. The more marked
effect of the change in voltage on the speed margin on longer runs
is due to the fact that the motors are running a t l i e voltage a
larger part of the time.
The effect of voltage on the operation of car equipments will be
examined from two standpoints: first, where the schedule can be

EFFECTS OF VOLTAGE VARIATION

!an3

adjusted to suit the voltage, and second, where the schedule must
be maintained irrespective of voltage. In the first case, the
schedule, energy consumption a t the car and the motor heating
will vary in the same direction as the line voltage. For example,
as shown by Fig. 69, under certain assumed conditions and with
a typical railway equipment, it has been found that on the bass
of constant speed margin, a reduction in line voltage from 550 to
450 results in a reduction in schedule speed of 5 per cent for an
average run of IOOO ft., and a reduction in schedule speed of 10

FIG 69.-Relatron

between voltage, energy consumption and schedule

speed In a glven run.

per cent for an average run of I mile. These percentages are approxlmately constant so long as the comparison is made on the basis of
the same percentage of speed margin for both voltages. Under
these same conditions, the heating current is reduced about 9 per
cent, both on a looo-ft. run and on a I-mile run, and this percentage
varies but little for different speed margins so long as the comparison
is made on the basis of a definite speed margin. Energy consump
tion is reduced 17 per cent a t the car on a 1000-it. run, and 13 per
cent a t the car on a I-mile run, the percentage being slightly greater

ELECTRIC RAILWAY HANDBOOK

when the comparison is based on runs without coasting than when
based on runs with suhcient speed margin. This reduction in
energy consumption a t the car is due almost entirely to the reduction in schedule speed. If the lower voltage were due to drop in
trolley and track circuits, the energy consumption measured a t ttie
substation would increase by 13 per cent on the I-ft.
run and
5.7 per cent on the I-mile run.
As an example of the conditions prevailing where the schedule
must be maintained, irrespective of voltage, it has been found th;\t
an increase in generator voltage from 4 5 0 to 550 results in a 28 p r
cent increase in percentage speed margin on a 1000-ft. run and a
32 per cent increase on a %-mile run. The same increase in voltage results in a 1 2 per cent decrease in energy consumption a t the
car on a 1000-ft. run and a 2 0 per cent decrease a t the car on a
>$-mile run. If the lower voltage is due to drop in the trolley and
track circuit, then with the higher voltage a t the car, the energy
consumption measured a t the substation decreases 28 per cent on
the 1000-ft. run and 35 per cent on the %-mile run. The energy
consumption decreases as the voltage increases because with higher
voltage the car accelerates more quickly, more coast is o b t a i n 4 ,
and the brakes are applied a t a lower speed. The lower energy
consumption is the result of less loss in the heating of wheels and
brake shoe?. Under any given set of conditions there is one voltage
giu+wa. .rn\=v.twsv . h i .iw a ~ r - r e d ,wKib
~wrs&seif ,f&yakage
is e~therraised or lowered.
Fig. 70 shows the variation in heating current, energy consump
tion and speed margin for several lengths of run from 0.1 mile to
2.5 miles, with voltages from 3 0 0 volts to 600 volts. For the four
short runs, the schedule is such as can be made a t 3 0 0 volts without
coasting. The short runs are taken on that basis since the cornpadson can be made over the entire range of voltage chosen. On the
two long runs, however, a t higher voltages the car makes the d i g
tance and coasts to a standstill in less time than required for the
run without coast a t 3 0 0 volts. I n order to obtain a wider range
of comparison, the schedule is taken as that which can be made a#
600 volts when coasting to a standstill without any braking.
While the effect on speed margin and energy consumption 3:
marked, the effect on heating current is comparatively slight.
The ability of a car to maintain schedules under varying track
conditions as measured by the speed margin and the energy
consun~ption are of primary importance to the operator. The
maintenance of an equipment is ;rffected to a certain extent by the
temperature rise obtained oh the motors in their average service.
It is doubtful, however, if the degree to which motor heating is
increased by reasonably increased voltage will ever be reflected
in any maintenance reports, provided running time is kept the same.
With higher voltage, the coasting time is longer, and brakes as a
result are applied a t a lower speed. This reduces the wheel and
brake shoe wear and tends to counterbalance the theoretical
increase in motor maintenance due to higher temperature.
Low voltage will often cause the overheating of a motor eqnifiment on long grades. On account of the lower speed resulting froin

E F F E C T S O F VOLTAGE VARIATION

211

decreased voltage, the time on the grade may become sufficiently

long to exceed the motor's thermal capacity, overheat the brushes,

or otherwise produce poor commutating conditions. This is
particularly true of locomotives, since they generally will be more
heavily overloaded on grades and operate a t lower speeds than
motor cars.
With extremely high resistance in trolley and track circuit, it is
sometimes possible to obtain a higher speed with the motors in

FIG. 70.-Relation

between voltagc and energy consumption for various

lengths of run and fixed schedule speed.

series than with the motors in parallel. Assume as an example the

following somewhat abnormal conditions: A single track section
of road laid with 60-lb. rails and fed from one end by a single No.
oooo copper trolley wire. The total resistance per mile would be
0.305 ohm. If the substation voltage is 625 volts, and a car with
quadruple motor equipment a t the end of the line, 5 miles from
the substation, is drawing 70 amp. per motor, the voltage a t the car
with motors in parallel is 198 volts. The voltage a t the car with
motors in series-parallel is 412 volts, and the voltage per motor

212

ELECTRIC RAILWAY HANDBOOK

206. The voltage per motor is, therefore, in the example chosen,
8 volts more with the motors in series-parallel than with the motors
in full parallel, and the speed of the car would be 5 per cent higher.
With any higher current per motor the voltage drop would be
greater, and the difference in the voltage per motor with motors
in series-parallel and with the motors in full parallel would increase
as the current per motor increases.
Effect of Coasting. With a given equipment on a run of a given
distance in a given time from start to stop, an increase in the rate of
acceleration, the rate of braking remaining the same, will (a) reduce
the time of running on resistance, which reduces the resistance
energy loss, and (6)reduce the speed a t which braking begins, which
reduces the energy lost in braking; the average speed remains the
same, and the maximum speed is changed little if any, therefore the
energy expended in overcoming train resistance is changed little
if any. Similarly, an increase in the rate of braking, the rate of
acceleration remaining the same, will reduce the speed a t which
braking begins and thus reduce the energy lost in braking. Either
or both of the changes, in rate of acceleration or in rate of braking,
will increase the coasting time, so that in general it may be said
that within the physical limitations of the equipment, the energy
consumption becomes less as the coasting time is increased.
Devices for Checking Motormen. I t is upon the principle outlined above that the operation of the Coasting Time Recorder is
based. The instrument records the time during which the car is
coasting, and is one of the devices in common use for checking and
comparing operation of motormen. Another such device is the
Current Time Recorder, which records the time during which current
is applied to the motors, the most economical run being assumed to
be that in which the current-on time is the shortest, as such result
presumably is obtained only by increasing the rates of acceleration
and braking to the maximum. The comparative operation of
motormen, as well as that of the car equipment, is more directly
indicated by the readings of a watthour meter connected in the
motor circuit. This directly measures the energy (which i t is desired to save), and is further used by many operators as an indicator
of inspection time, the various parts of the equipment being inspected
on a kilowatt hour rather than on a mileage basis. The Economy
watthour meter is of the mercury flotation type, reducing to a
minimum the troubles incident to carrying such meters on moving
cars. When used as an inspection indicator, it is equipped with
auxiliary inspection dials, one of which may be used as a n indicator
for brakes and controllers, one for oiling, and one for general inspection. The auxiliary dials are arranged for resetting to zero after
equipment inspections, and are equipped n-ith markers which may
be set to indicate proper inspection mtervals.
The indications of any of the devices named, and particularly the
watthour meter, serve as a check as to the condition of equipment,
giving warning of such troubles as short-circuited motor fields, too
c l o s ~brake shoe adjustment, etc. The saving in energy resulting
from the use of one or another of these devices has amounted to
from 10 to 25 per cent of the total, on numerous roads.

213

T R A I N OPERATION

Train Operation. The full realization of the advantages of

multiple-unit operation can be obtained only on roads having ample
substation capacity and a liberal feeder system, although the
requirements of these items can be kept within very reasonable
limits by the proper training of the motorman. Where multipleunit trains are employed, it is usually not necessary to have either
cars or equipment as large as if they were operated singly a t all

FIG. 71.-Motor

temperature, single car.

times. The predecessor and principal competitor of the multipleunit train is the combination of the motor car and trailer. If the
motors are of sufficient capacity, and especially if two-car trains
are to be operated for only short periods of time, trailer operation
may be more economical than multiple-unit operation. Care
should be taken that in such operation the motors are not overloaded since the excessive heating may do them damage. The
m
89

10

6 GO
cofao

Hw~,o
,

4,

20
10
0
L?b

S:%

10.-

1 l

8 4

h ~ n
. -

1
2
.
a 2m
P.U

6 0 1 8

FIG. 72.-Motor

6:%
8:s 10?8 1?:51
4;s 4%
0:PI
P.M.
P.Y.
10 11 12 101 14 15 10 11 18 19 20 11 t) 11 24
W el h7'1 E u

4.%
0

~~

Bovn b W

temperature. w ~ t hrush hour trailer.

effect on motors of pulling a trailer may be seen from the curves

in Figs. 71 and 72 which are reproduced from a paper by
Clarence Renshaw, Trans. A.I.E.E., 1903, Fig. 71 showing rise
i n temperature of motors with one car operating singly and Fig.
7 2 showing the rise in temperature with the car pulling a trailer
during the rush hours. I t will be noted that the presence of the
trailer for one trip in the morning when the motors were fairly cool
only caused the temperature to rise much more rapidly without

ELECTRIC RAILWAY HANDROOK

causing it to reach any higher value, but in the afternoon when the
motors had reached the temperature where they would otherwise
have remained, the addition of a trailer for about 3% hours raised
the temperature in this case approximately 2 5 deg. from 75
deg, C. The use of trailers for rush-hour service is also likely to
cause delay, due to the reduction in speed on account of the added
weight, and the inability of the trailer t o move by itself causes
further delay in backing i t up, dropping it, and possibly shifting
i t a t stub-end terminals. Otherwise, in certain classes of s e ~ i c e
trailer operation may be very successful without causing injury to
the motor equipment, especially by taking advantage of the thermal
capacity of the motors in attaching trailers to motor cars which
have been out of service for 5 or 6 hours and leaving them attached
for only a limited number of trips. It may be passible, also, to
compensate for the entla load by a reduction in the number of stops
which would ordinarily be made by the motor car without the trailer.
With multi le unit operation it is of course possible to apply
motors to tRe cars more efficiently, both as to size and gear ratio,
than if trjller operation had t o be provided for, b u t on the other
hand, the motor equipment on all cars, including those used as the
second car, represents considerable dead investment when train
operation is resorted to only infrequently. I n rapid transit sewice where train operation is the rule instead ot the exception, there
is of course no cjuestion as to the advantage of multiple-unit operation, thus distributing the motors among the axles a11 along the
train, reducing the capacity of the single-motor equipments, and
making available for traction a much greater proportion or all of
the weight of the entire train.
The Public Service Railway Company of New Jersey conducted a
series of experiments with the view of determining tbe meritsof
multiple-anit train operation as applied to city setvice; whether the
operation of two-car trains in city service was economical and
desirable; and the combination and number of motors best suited
for such operation. Comparative tests were made with six trains,
the following description being taken from Aera, 1913: The
trains tested comprised two single cars, one four-motor and one
two-motor, one train of two four-motor cars, one train of one fourmotor and One two-motor car, one train of two two-motor cars
and one train of a four-motor car and trailer. Motors were Westinghouse No. 307, with HL control. Somc of the operating results
are as follows:
Class of train

. .. '
.
. . .. .

Four-motor srngle car.. . .

Two-motor s~nglecar. . .. . . .
Two four-motor cars.. . . . . .
one twomotor car.
Two two-motor car;. .
.
Pour-motor car m d traller.

Per cent of
standard

runnrng time

per cent

lor . 4
r 06.5
loo.8

100.0
68.4

rorr.9

104.5
117 . 8

C.0

81.5

per cent

100.0

100.0

92.6

TRAIN 0PF;RATION

The table shows that the four-motor cars, either single or in

trains, kept fairly close to the standard running time. A train
made up of a four-motor car and a two-motor car lost a little time,
but not as much as the two-motor cars, either single or connected.
The four-motor car with trailer fell much further behind. The
energy consumption figures are taken from typical rush-hour trips
in the direction of travel, and it will be noted that the two-motor
cars show a saving of about 30 per cent in energy consumption
per car mile over the four-motor cars. This, because of the lighter
weight of the two-motor cars, corresponds to about zo per cent per
ton mile. The six-motor train shows a saving of about 14per cent
per car mile, or about 8 per cent per ton mile, as compared with the
eight-motor train. The motor car with trailer shows a slight
saving on a car-mile basis, but over z o per cent more energy consumption per ton mile than a train of two similar motor cars.
The six-motor train seems to be worthy of special consideration
for this class of service. I t s running time was very nearly as good
as that of the cight-motor train or of the single four-motor car, and
at the same time i t showed a saving of 7.4 per cent in energy consumption per ton mile and about twice as much per car mile. It
has none of the disadvantages of traiier operation and it would seem
to he entirely feasible to operate the two-motor cars singly during
the middle of the day for a larger part of the year and thus secure
the still greater energy ecooomy as shown by the two-motor car
tests. If, however, weather or rail conditions make i t advisable, the
four-motor cars could be m n throughout the day, while the combination of a two-motor and a four-motor car in the rush hours
would meet the conditions of heavy travel much more satisfactorily
than a train of two two-motor cars, and practically a s well as an
eight-motor train. The rcport on the Public Service tests concludes
that train operation in city service is desirable under certain conditions and during certain periods of the day. This is i ~ d i c a t e d
(I) by the facilities which train operation provides for handling
rush-hour loads and traffic of extraordinary proportions; (2) by
the relief offered through congested sections by operation in trains
instead of as single units; (3) by the reduction in platform labor
expense which train operation makes possible in comparison with
single cars.

SECTION I V

RAILWAY MOTORS
A.I.E.E. Standards. The following numbered sections (5101 to
5502, ~nc.)relat~veto the rating, ca acity and selection of railway
motors, are from the Standards of tKe Amencan Institute of Electrical Engineers, 1922 edition.
5101. Temperature Limits of Railway Motors in Continuous
Service. The following maximum observable temperatures are
permissible in the windings of railway motors, when m continuous
service.
Under extreme amblent temperatures ~t I S permlsslble to operate, for short
Infrequent penods. a t 15 deg C hlgher temperature than specrfied In this
rule
Owlng to space llmltat~onsand the cost of carrylng dead welght on vehicles.
~t IS considered good ractlce to operate propulsion mach~nery~t hlgher
temperatures than wourd be advlsahle In statlonary machtnes
Temperature
Class of matenal
By thermometer
By reststance
Cotton, zllk, paper and slmllar matenals
when so treated or rmpregnated a s to
Increase the thermal Irmrt, also
enamelled m r e
85 deg C.
1x0deg. C.
Mrca, asbestos and other matenals
capable of reslstlng hlgh temperatures. rn whlch any Class A (above)
material or b ~ n d e rIS used for structural purposes only. and may be
destroyed mthout Impalrlng the
lnsulat~on or mechanical qualltres
of the rnsulatlon
loo deg C.
130 deg. C.

5202. Nominal Rating of Railway Motors. The nominal rating

of a railway motor shall be the mechanical output a t the car or loccmotive axle, measured in kilowatts, which causes a rise of temperature above the surrounding air, by thermometer, not exceeding go
deg. C a t the commutator, and 75 deg C a t any other normally
accessible part after one hour's continuous run a t its rated voltage
(and frequency in the case of a n alternating current motor) on a
stand with the motor covers arranged to secure maximum ventilation wthout external blower The rise in temperature, as measured
by res~stance,shall not exceed roo deg C The statement of the
nominal rating shall include the corresponding voltage and armature speed.
In the absence of any specification as to the kind of rating, the
nominal rating shall be understood
5203. Continuous Ratings of Railway Motors. The continuous
ratings of a railway motor shall be the inputs in amperes a t which

217

218

ELECTRIC RAILWAY HANDBOOK

it may be operated continuously a t 36, % and full voltage resp"tively, without exceeding the observable temperature rises speciffed
M o w , when operated on stand test with motor covers and cooljng
system, if any. arranged as in service. Inasmuch as the same motor
may be operated under different conditions as regards ventilati0n.
i t will be necessary in each case to define the system of ventilation
which is used. In case motors are cooled by external blowers, the
flow of air on which the rating is based shall be given.
The tern rature rise In service may be very different from that on stafdtest. See c c . ssoz for the relation between stand-test and service tempemtuxes as affected by ventilation.
Temperature rises of windings
Ckss of material
B y thermometer By resistanci
Cotton, silk. paper and similar materials
when so treated or Impregnated as to
increase the thermal limit; also
65 deg. C.
85 deg. C.
enamelled wire.. ..................
Mica. asbestos and other materials
capable of reslstlng h ~ g htemperatures. in which any Class A (above)
matenal or binder is used for structural purposes on!y and may be
destroyed without rApairing the insulation or mechanical qualities of
%he in%u;\ati~n.
...................
%<a.
C.
xns ap8.C-

5204. Ratings of Field-control Railway Motors. The nominal

and continuous ratings of field-control motors shall relate to the"
performance with the operating field which gives the maximam
motor rating. Each section of the field windings shall be adequate
to perform the service required of it, without exceeding the specified
temperature rises.
5210. Ratings of Electric Locomotives. Locomotives shall be
rated in terms of weight on drivers, nominal one-hour tractive
effort, continuous tractive effort and corresponding speeds.
5211. The weight on drivers, expressed in pounds, shall be
the sum of the weights carried by the drivers and of t h e driven
themselves.
5212. The nominal tractive effort, expressed in pounds, shall be
that exerted a t the rims of the drivers when the motors are opefiting a t their nominal (one-hour) rating.
5213. The continuous tractive effort, expressed in pounds, s ~ l l
be that exerted a t the rims of the drivers when the motors are operating a t their full-voltage continuous rating, as indicated in Sec. 5203.
In the case of lo~omotivesoperating on intermittent service, the
continuous tractive effort may be given for H or Jvoltage,
i
but
in such cases, the voltage shall be clearly specified.
5214. The rated speed, expressed in miles per hour, shall be tpat
a t which the continuous tractive effort is exerted.
5337. Losses in Gearing and Axle Bearings of Railway Mot@.
The losses in gearing and axle bearings for single-reduction sin&geared motors varies with the type, mechanical finish, age and lubrlcation. The following values, based upon accumulated tests, shall
be used in the comparison of single-reduction single-geared motors,
Sec. 5339.

A.I.E.E. RAILWAY MOTOR STANDARDS

Per cent of input a t nominal
rating
zoo
150

>%

3
25

219

Losses in axle bearings and

single-reduction gearing
a s per cent of lnput
3.5
3.0

2.7
1.s

24
&*
8.5

Further investigation may indicate the desirability of giving separate

values of the losses for full and tapped fields, or low and high speed motom.

5338. Brush Friction, Armature-bearing Friction and Windage.

The brush friction, armature-bearing friction and windage, shall
be determined as a total under the following conditions: In making
the test, the motor shall be run without gears. The kind of brushes
and the brush pressure shall be the same as in commercial service.
Drive the machine idle as a series motor on low voltage. The
product of armature counter-electromotive-force and amperes a t
any speed shall be the sum of the above losses a t that speed. (See

"=. 5339.
5339:)
No-load

Core Loss, Brush Friction, Armature-bearing

Friction and Windage. The no-load core loss, brush friction, armature-bearing friction and windage shall be determined as a total
under the following conditions: In making the test, the motor shall
be run without gears. The kind of brushes and the brush ressure
shall be the same as in commercial service. With the fierd separately excited, such a voltage shall be applied to the armature
terminals as will give the same speed for any given field current as
is obtained with that field current when operating a t normal voltage
under load. The sum of the losses above mentioned is equal to
the product of the counter-electromotive force and the armature
current. The no-load core loss is obtained by deducting from the
total losses thus obtained the power required to drive the motor
a t corresponding s eeds as determined under Sec. 5338.
The core loss ofdirect current railway motors under load shall
be assumed to havz the values given below:
Per cent of input a t
nominal rating
aoo
150

rw

Core loss a s per cent of

no-load core loss
165

14s
130

75
11s
50
113
25 and under
raz
With motors designed for field control the core losses shall be assumed a s
the same for both full and permanent field. I t shall be the mean between the
no-load losses a t full and oermanent field. increased bv. the wrcentanes
- uiven
i n the above table.
Approximate Losses in D.C. Railway Motors. In comparing projected
railway motors, and in case it is not possible or desirable to make tests to

220

ELECTRIC RAILWAY HANDBOOK

determrne mechanical losses. the follomng values of these losses. detetmned

from the averages of many tests over a wrde range of slzes of s~ngle-reduct~on
slngle geared motors. mil be found useful. a s a proxlmatlons. They Include
axle-bear~ng. gear, armature-beanng, brush-fnctlon, wlndage, and strayload losses
Input rn per cent of that a t
nominal ratlng
100 or over

Losses as per cent

of ~ n p u t
5 0

7s

60
50

40
30
2s

&:.t
I7

The core loss of railway moton may also be determined a s spec~fiedfor

other machmes.

Characteristic Curves of Railway Motors

5401. General. The characteristic curves of railway motors
shall be plotted with the current as abscissas and the tractive effort,
speed and e5ciency as ordinates. In the case of alternating
current, the power factor shall also be plotted as ordinates.
5402. Voltages. Characteristic curves of direct current motors
shall be based upon full voltage, which shall be taken as 600 volts,
or a multiple thereof
5403. Field-Control Motors. I n the case of field-control motors,
characteristic curves shall be given for all operating field connections.
Selection of Railway Motor for Specified Service
5501. Data Required in Selecting Motor. The following
information relative to the senrice to be performed is required, in
order that an appropriate motor may be selected.
(a) W e ~ g h of
t total numher of cars In tram (rn tons of zooo Ib ) exclusrve

of electncal eaumment and load

( b ) Average we~ghtbf ioad and durat~onsot same. and rnaxlmum we~ght
of load and d u r a t ~ o nof same.
(c) Number of motor cars or locomot~vesIn traln. and number of trader
cars In tram.
(d) D~ameterof dnvrng wheels
(c) Welght on dnvlng wheels, exclusive of elcctncal equ~pment
W Number of motors per motor car
(g) Voltage a t tram m t h power on the motors-average.
maxlmum and
mlnrrnum.
......-~
-- - - ~
( h ) Rate of acceleratron rn mlles per hour per second.
(a) Rate of brakrng In mlles per hour per second
( I ) Speed Irm~tat~ons.
~fany (~ncludlngslowdowns).
( k ) Distances between stopplng polnts
( I ) Average d u r a t ~ o nof stops
(m) Schedule speed, lncludrng stops. In m ~ l e sper hour
r ton of zooo pounds a t stated speeds
(n) Tram revstance In pounds
( 0 ) Moment of Inertla of revorrng parts, exclus~veof electr~calequrpment
( 0 ) Profile and allnement of track
(cl) D~stancecoasted a s a percentage of the d~stancebetween stopplng
pomts.
(r) Durat~onof layover at end of run, if any.

A.I.E.E. RAILWAY MOTOR STANDARDS

221

5502. Method of Comparing Motor Capacitp with Service

Requirements. When l t IS not convenient to test motors under
actual specific service conditions, recourse may be had to the
followng method of determlnlng temperature nse from the standtests. The essential motor losses affecting temperatures in servlce
are those in the motor wlndlngs, core and commutator. The mean
service cond~tionsmay be expressed, as a close approximation, in
terms of that continuous current and core loss which will produce
the same losses and distribution of losses as the average in service.
A stand test with the current and voltage which will give losses
equal to those in service, will determine whether the motor has
sufficient capacity to meet the service requirements. I n service,
the temperature rise of an enclosed motor (Sec. 4044), well exposed
t o the draught of air incident to a moving car or locomotive, will
be from 75 to 90 per cent (depend~ngupon the character of the
service) of the temperature rise obtained on a stand test with the
motor completely enclosed and w t h the same losses. With a
ventilated motor (Sec. 4045 and 4046), the temperature rise in
service will be go to loo per cent of the temperature rise obtained
on a stand test with the same losses.
I n making a stand test to determine the temperature rise in a
specific service, it is essential in the case of a self-ventilated motor
to run the armature a t a speed which corresponds to the schedule
speed in service. I n order to obtain this speed, it may he necessary,
while maintaining the same total armature losses, to change somewhat the ratio between the Z2Rand core-loss components.
Calculataon for cornparang mdor capacaty with snurce rcquarrmenlr. The
heatlng of a motor should be detennrned, wherever osstble by testlng tt tn
semce, or m t h a n equivalent duty-cycle. When tKe s e m i , or eqcuvalent
duty-cycle tests are not practrcable, the rattngs of the motor may be uttlrzed
a s follows, t o determine ~ t temperature
s
nse
The motor losses whtch affect the heattng of the mndtngs are as stated
above, those tn the wtndtngs and tn the core. The former are proporttonal
to the square of the current. The latter vary m t h the voltage and current.
accordmg to curves whtch can be supplred by the manufacturers. The
procedure IS therefore as follows:
(a) Plot a tlme-current curve, a trme-voltage curve, and a trme-core loss
curve for the duty-cycle whtch the motor 1s to perform. and calculate from
these the root-mean-souare current and the average core loss.
(b) If the calculated r m s. semce current exceeds the contrnuous rattn
when run m t h average servtce core loss and speed. the motor 1s not su&
crently powerful for the duty-cycle contemplated.
(c) If the calculated r m s servlce current does not exceed the contrnuous
rat&, when run m t h average semce core loss and speed. the motor rs
ord~nanlysutable for the servlce.
In some cases. however. rt may not have sufficrent thermal ca aclty to
avold excesstve temperature nses dunng the penods of heavy g a d . In
such cases a further calculatron 1s requrred, the first step of whrch IS to
compute the equtvalent voltage whrch, anth the r m s. current, wtll produce
the average core loss H a v ~ n gobtalned t h ~ s .determme, a s follows. the
temperature nse due to the r m s servlce current and equrvalent voltage

Let

-1 $gg:zt?;

"se
core loss, kw
$ggE,t?;.nse
Po = core loss, kw.
p,

Then

m t h r m s servrce current and eoutvalent

wlth conttnuous load current corresponding
t o the eaurvalent servtce voltage

ELECTRIC RAILWAY HANDBOOK

(d) The th-l
CA city of a motor 1s approximately measured by the
ratlo of the electnpl Ess m kw a t ~ t normnal
s
(one-hour) capaclty. to the
correspondrng maxlmum observable temperature rise dunng a one hour test
starting a t ambrent temperature.
(c) Consider any
nod of peak load and determine the electncal losses
In hlowatt hours E n n g that penod from the electncal efliciency curve.
Ptnd the excess of the above losses over the losses m t h r m s servlce current
and eqtuvalent voltage. The excess loss, d ~ n d e dby the coefficient of thermal capacity. m11 equal the extra temperature nse due to the peak load.
Thls temperature nse added to that due t o the r.m.5. semce current, abd
eqtuvalent voltage, gwes the total temperature nse. If the total temperature nse m any such m o d exceeds the safe l ~ m ~the
t , motor 1s not suffic*enUy
powerful for the servlce.
V) If the temperature reached, due to the peak loads. does not exceed
the safe Ilm~t.the motor may yet be unsu~tablefor the servlce, a s the peek
loads may cause excessive sparlong and dangerous rnechan~cal stresses.
I t IF. therefore. necescary t o compare the peak loads w ~ t hthe short-penod
overload capaclty. If the peaks are also mthln the capacity of the motbr.
it may be cons~deredsmtable for the mven duty-cycle.

Relation between Ratings and Voltage. The change in rating of

a railway motor due to change in the voltage a t which the rating is
made is difficult of exact determination except by repetition of the
stand test a t the various voltages under consideration. In general,
however, i t may be said that the nominal, or one hour rating,
is almost proportional to the voltage for such ordinary changes in
the latter as are encountered in practice. The one hour rating of a
railway motor is a measure of the heat storage capacity of the motor.
Most nf the beat gsne~ztedk ~ w in
d mising t-be tempemture of the
motor, although part of it is dissipated by radiation from the franle
and on ventilated motors by the air blown out by the fan. The
amount of heat dissipated in this manner will vary from 1 5 to qo
per cent of the total heat generated on a one hour run depending
upon how well the motor is ventilated. For a given currex~t
the core loss and friction and windage will increase as the voltage
increases, since the speed varies with the counter-electromotibe
force. In a ventilated motor the ventilation, and hence the
amount of heat dissipated by this means, will also increase with the
speed. As a result, the one hour current ratings a t say, 600 and
500 volts, will be practically the same, and it is generally true that
the nominal one hour rating is proportional to the voltage.
The continuous ratings are given for x,?/, and full voltage, and
the difference is seldom more than a few amperes for either a
ventilated or non-ventilated motor. I t is therefore usually safe t o
interpolate between these values for an intermediate voltage.
Motor Capacity for Specific Service. As suggested in the
Standards of the A.1.E.E , the most satisfactory method of comparing motor capacity with service requirements is to make tests
under the actual service conditions or equivalent duty-cycle and
with the specific motor under consideration. Such tests are often
costly and frequently impossible or not practicable; recourse muat
then be had to some such method as outlined in the notes under
A I E E. Rule 5502 (page 221). This involves the plotting of run
curves (the data for which are required by Rule 5 5 0 1 ) ~the calculation of r.m s. current and average core loss, the consideration ~f
peak loads, thermal capacity, cornmutating ability, armature speed
and ventilation, all as outlined under Rule 5502.

CAPACITY O F VENTILATED MOTORS

223

Capacity of High Speed Self-ventilated Motors. I n the older

types of heavy, slow speed, totally enclosed motors, the temperature depends almost wholly upon the total loss in the motor,
since all losses are dissipated through the motor frame. The
self-ventilated motors of more recent design are arranged for
circulation of air through both armature and field windings by
mean5 of fans mounted on the shafts of small high speed armatures, and as a result there has been a considerable reduction in
weight of motor per nominal horsepower. I n the application of
these light weight, high-speed, self-ventilated motors, it must be
remembered that they have much less thermal capacity than the
older type enclosed motors, so that short time heavy loads are
of relatively much greater importance, and the average running
temperature must be considerably lower in order that the same
maximum temperature be not exceeded. With the self-ventilated
motor, the .temperature depends not only on the losses, but
on the armature speed, and the core loss plays a much smaller
part than in the enclosed motor. The temperature rise with
a given current is usually less a t a high than a t a low voltage, as
the increased ventilation due to the high speed more than offsets
the higher core losses. I t is therefore des~rableto consider the
average armature speed in service as well as the equivalent voltage.
If the average armature speed is below that required to develop
the average core loss, extra margin must be given. Conversely,
if the average s p e d is higher than that given a t the equivalent
voltage, the equipment will have a good margin. I t is also necessary to consider the possibility of operating a t slows eed under such
conditions as in bucking snow, when the losses will ge high and the
ventilation poor. T o sum up, i t must be recognized that the
high speed, self-ventilated motor should be a p lied with greater
care than the old enclosed types. The high speeffan will give good
vent~lationand normally gives a low tern erature rise. But when
heavy loads, low thermal capacity and row speeds are combined,
disastrous heating is likely to occur.
The practical result in the use of the ventilated type motor is that
in a given service a lighter motor may be used provided the lower
thermal capacity of the smaller motor is not a limit. In general,
in both city and suburban service, these lighter motors do meet
the service requirements where formerly the heavier enclosed motor
was used, and of course give a very considerable saving both in
first cost and in reduced weight of car equipment. For instance,
on a car where formerly a Westinghouse ror-B-2 motor would
have been used, i t is now found to be feasible to use the Westinghouse 532-B or even, possibly, the 5x4-C. By total weight of
motors and gears, the comparison is shown in the following table:
Type
101-B-2
3obCV-4
307
532-B
514-C

Hour ratlng
40 hp.. 500 volts. 520 1P m.
65 hp 600 volts. 700 r p m.
5 0 h p . 600 volts. 615 r p rn.
50 hp 600 volts. 665 r p.m.
40 hp 600 volts. 760 r p rn.

.
.

We~ght,pounds
2780
2660
2700

Ssp0
1770

If the user of IOI-B-2 motors desires to duplicate that motor

a s to thermal characteristics in a modern commutating pole

2 -

ELECTRIC RAILWAY H.4NDBOOK

motor, it will be necessary to select a motor of approximately the

same weight as the old motor. The total weight of the motor is a
rough indication of thermal capacity, but a better check is the size of
copper in mtbtors of the same speed. For overloads lasting only a
short titnc, say tcn minutes or less, the temperature rise is determined almost wholly by the thermal capacity of the copper windings.
For longer overloads the determining factor is the thermal drop
through the insulation. By areas of copper size and armature
weight the comparison is as follows:
Type motor
101-8-2
30&C\'-q
307
532-B
514-C

Approximate copper
area, circular rnils
13.000
1b.600

Approximate armature
weight. pounds
58s
61s

13.000

615

10.400
7.420

550
390

I t is evident from this table that from the standpoint of thermal

capacity only it would be necessary to select a motor of about
the size of the 306CV-4 or the 307 to match up with the 101-B-2.
Such a heavy motor is undoubtedly better able to stand heavy
short o\.erloads without injurious heating than are the lighter motors,
but the user pays for this capacity both in first cost and in operating
cost.
Cornmutating Ability. Allowable Accelerating Current. Few
non-interpole motors will commutate successfully more than IOO
per cent of their nominal rating, and the practical allowable current
used in accelerating these motors should never be considered as
greater than the one-hour rating. I n the early days the rating of
railway motors generally determined their commutating ability
and for this reason these motors were rated a t approximately their
nlaximum commutating ability.
For interpole motors the practical allowable current for acceleration is determined hy the heating effect of the current and the size
of the commutator. The design of the motor also has a great effect
on this value of current. Of two motors of nearly identical design,
the one with the larger cornmutator or better ventilation would
rate a greater accelerating current than one with a small commutator or poorer ventilation. There is no set percentage or comparison
between the one-hour or continuous rating and the allowable
current used for the acceleration of a modern interpole motor, but
it is safe to allow 50 to 7 j per cent more current with interpoles
than without, where the question is simply that of commutation.
Usually, however, commutation will not limit the accelerating
current of interpole motors, since practical!^ any of them will commutate double the one-hour rating current without serious sparking.
The limitation is usually a matter of heating or wheel slippage.
Preliminary Determination of Motor Capacity. S. B. Cooper,
General Engineer with the Westinghouse Electric and Manufacturing Co., has developed the general operating characteristics and approximate motor application curves shown by Figs.
I to 6, inclus~ve. These apply to direct current railway motors, and
are based on the great similarity between the characteristic curves
of such motors, as pointed out on page 242, and also making use of

FIG. I.-General
operating characteristic. 1ooCbft. run' 20 Ib. per ton
train resistance. acceleration and hraking, 1.5 m.p.h.s. !%lid lines-full
Jmd acce~eratio$;dash lim.9-arcekr.?tiun a t 135 per rent ful! JruQ. Speprl
margin based on time excluding stop, true speed margin therefore somewhat
less than l o per cent.

0 -1

10

20

30

do

Speed at End of $+mightLiw Accebmtion-d

FIG. 2.-General
operating characteristic. One mile run; 20 Ib. per ton
train resistance; acceleration and braking. 1.25 m.p.h.p.s.; coastingexpressed
a s per cent of total moving time.

226

ELECTRIC RAILWAY HANDBOOK

the fact that the ratio of the linear diniensions (speeds or t i m ~ )

of two similar speed time curves is equal to the square root of the
ratio of the areas (ciistances), as shown on page 165.
In using Fig. r or Fiq. 2 for the selection of an equipmcnt, an
approximation of the proper equipment must first be made, based
on the wright of car, length of run and schedule speed desird.
The accelerating tractive effort, calcul~tedas in ordinary speedtime curve work (page I S I ) , is I jo pounds per ton for Fig. I and 145
pouncls p r ton for Fig. 2. These valr~esmultiplied by tons per motor
give the ncc.rlerating tractive effort per motor. From the characteristic CLITVC of the motor to be used there can then be read the

Schrdule Speed-M~lesper Hour

FIG. 3.-Approxlrnate motor application curves for runs of 0.1 to 0.5
rnrle. showing holanctng speed. speed at end of stralght line acceleration.
r.1n.s. amperes per ton, accelerating ampercs per ton, watthoun per ton mile.
Based on full load accelerat~on;acceleration and braking. 1.5 rn.p.h.p.b.;
20 111. per ton traln resistancr; coastlnK to glve 1 0 per cent speed margln; t o
sec. stops. 500 volts. For other voltage. change r.m.s. and accelerating
current in'inverse proportion to voltage.

accelerating current per motor, I, and the speed, S, a t the end of

straight-line acceleration. Note also the approximate percentage
of full load or nominal rating. Fig. I shows curves for loo per cent
and 133 per cent of full load acceleration, and other values may be
interpolated; Fig. 2 is drawn for 133 per cent.
Inasmuch as Figs. I and 2 are drawn for 1000 It. and one mile
runs, respectively, a "linear correction factor" must be applied
to the speed or time values shown thereon, to make them applicable to runs of other lengths. These correction factors are:
brFig.1,

-.
----run in feet
I=+---,ength loo0

for Fig.

f = d ~ e n g t hrun in miles

2,

MOTOR APPLICATION CHARTS

Using

227

= d, the speed a t end of straight-line acceleration for the

clquivalent I-ft.
or one-mile run of Figs. I or 2, the curvesshow
the following:
watthours per ton mile (applicable to run of any length if shape of
sped-time curve is similar)
time, 1, for run of rooo ft. (Fig. I ) or one mile (Fig. 2)--multiply
by correction factor, f,for r i n of otber length

Frc. 4.-Approximate motor application curves for runs of 0.5 to 5 miles.

showlng balancing speed. speed a t end of stralght llne acceleratlon. r.m.s.
amperes per ton, accelerat~ngamperes per ton, watthours per ton mrle.
klased on acceleratlon a t 133 per cent full load; acceleratlon and brak!ng,
1.25 m.p.h.p.s.; zo lb.
r ton train resistance coasting 40 per cent of tlme
excluding stop. stops o E o sec. for one mile or iess. and 10 sec. pcr mile f o r
one to five miies. 525 volts. For other voltage. change r.m.s. and accelerating amperes insinverse proportion t o voltage.

per cent r.m.s. current, p. Multiply by accelerating current, I,

for r.m.s. current averaged over time of run excluding stop.
For r.m.s. averaged over total time, take:
time exc. stop
time inc. stop
Based on the general operating characteristics of Figs. I and 2,
the approximate application curves of Figs. 3 and 4 have been constructed, applying to runs of under and over one-half mile, respectively, and using data relative to accelerating current, acceleration
and braking rates, train resistance, and stop time, as noted in the
respective captions. From these curves may be read directly the
-

ELECTRIC RAILWAY HANDBOOK

228

maximum free running or balancing speed; the speed a t end of

straight l i e acceleration; the r.m.s. amperes per ton; the accelerating amperes per ton; and the watthours per ton mile.
If field control motors are used, Figs. I , 2, 3 and 4, using the permanent (or short) field curves for the motor, will give approximately
correct results except for r.m.s. amperes, rated horsz power, and

eed at Strikinq Full Field Curve

z e datskiking NonnaI Field Curve
FIG. 5.-Constant,

K. for corrections to field control motors.

energy consumption. The corrections for field control motors

may be made as follows:
H N = (amperes)l X seconds, with normal field (found
let
as above)
H s = reduction in ampere^)^ X seconds, due to field
then
control2
K to (IN 12)
X = as shown by Flg. 5, speeds read from motor characwhere
teristic a t accelerating currents
to = time of straight line acceleration (speed a t end of
straight line acceleration divided by rate of acceleration)
I N = accelerating current on normal field
IP = accelerating current on full field
then
HP = (arnperes)2 X seconds, with field control
= HN
Hs

and

4:

= r.m.s. current with field control

T = total time, including stop, in seconds.

where
For energy saving due to field control:
Let P N = kilowatts per car a t end of straight line acceleration
on normal field
Q = constant from Fig. 6
to = time of straight line acceleration (as above)

MAXIMUM MOTOR CAPACITY

then Es

and

229

energy saved by field control, watt-hours per run

=PN~oQ
E = energy saved, watt-hours w r ton mile
=

where N = number of stops per mile

and W = weight of car in tons.

EXAMPLE,'
&fio of fullfo normal

iidd acceimfing

currenf = 0.882

FIG. 6.-Constant.

Q. for energy saving by

field control.

Maximum Practicable Motor Capacity. The highest powered

cars in service have equipments rating a t average voltage 1 2 or I
horsepower per ton of loaded car, while the (seldom approached5
minimum is approximately 3 horsepower per ton. The maximum
horsepower per ton is most frequently approached with high speed
interurban equipments. The present average light weight double
truck cars in city service run about 7.8 horsepower per ton; in
interurban service, 8.5 horsepower per ton. The rated output of
commercial railway motors a t average voltage averages about 48
horsepower per ton of complete motor, up to 65 horsepower--above

230

ELECTRIC RAILWAY HANDBOOK

that rating, about 66 horsepower per ton. The mnimum weight df

complete car with average load which could be bulk to accommck
date such motors and safely operate a t high speeds would be
approximately three times that of the motors alone Hence It
IS practically imposs~bleto equip a car with more than about z D
horsepower per ton. Such cars and equipments would be very
expensive to build and operate, and it is doubtful whether it will
ever be practicable to exceed the present maximum of 1 2 or 13
horsepower per ton in equipping cars. The maximum possible
schedule speeds, a t instant of cutting off power, and energy consumpt~onfor various runs, with 1 2 horsepower per ton, are shown
by Fig 7. These curves are based on maximum accelerating and

FIG 7 -Max~mum practicable motor performances.

braking rates of 2 and 2 5 m p h p s , respectively, and a motor of

the ventilated field control type. They are intended only as il
gulde to the posslb~lltiesof equipments, as such performance usually
IS too costly to be justified
Approximation of Required Motor Capacity. I t has been noted
that a definite rallway motor equipment on a car of fixed weight
will attain approximately the same temperature rise irrespective
of the number of stops per mile demanded by service condition5
(although the schedule speed will be affected thereby). Some
engineers, therefore, for p u y of approximation, rate a railway
motor a t the tons train weig t per motor which can be moved a t B
given maximum speed (or gear ratio) and with a n allowable
temperature rise The following table shows the approximate
nominal one-hour horsepower rating of self-ventilated direct
current motors required for various car or train weights and maximum speeds, with a n allowable temperature rise of 65 deg. C.

SINGLE PHASE hIOTOR THEORY

Max
speed.
mph

231

Wclght of tram lncludlng passengers, tons

Theory of Single Phase Commutator Motor. Figure 8 shows a

four-pole stator with torque field c o ~ l sassembled in place. This
torque field is connected in series with the armature, as shown in Fig.

Armature

FIG 8 -Four-pole
slngle pha= cornmutatlng alternatrng current motor wlth
maln or torque field wrndlng

FIG 9 -Connection
of
armature and torque field.
a c motor

9. If an alternating current voltage is impressed across this circuit,

current wll flow and two flux fields will be set up, one by the
armature and one by the torque field winding. These two fluxes
will produce an inductance voltage a t right angles to the current.
Due to the current carrylng armature conductors find~ngthemselves
i n the flux of the torque field, they attempt to move out, and the
armature rotates. On account of this rotation, a rotational or
counter e m f , together with the resistance and iron loss drops
and inductance volts, buck the impressed voltage and are equivalent
to this voltage. An exammation of Fig. l o w~llindicate how high
field and armature inductance will cause low power factor, as shown
by the angle between the impressed voltage and motor current.
It is desirable, therefore, to reduce the inductance to a minimum.

ELECTRIC RAILWAY HANDBOOK

232

The useful flux is t h a t due to the torque field and must not be
disturbed. The flux set up by the armature must be eliminated,
if possible. By refinement in design, the leakage flux due to the
toll ends and through the slots is reduced to a minimum. The
major portion of the armature flux is counteracted or compensated
for by means of a winding placed in thefaces of the pole-this

Res~shmceL ass,p/us

Iron L ms duehupph

FIG. 10.-Relation between impressed and induced voltages. a. c. motor.

FIG. I I.-Single

phase motor with torque field and compensating windings.

FIG.I ?.-Connection of armature.

torque field and conductive compensating winding, a. c. motor.

FIG.13.-Connect~on of armature.
torque field and inductive compensating winding, a. c. motor.

SINGLE PHASE MOTOR THEORY

233

winding having just sufficient turns to neutralize the armature

reaction. The arrangement of this winding in the pole faces is as
shown in Fig. 1 1 . If the motor is to run on both alternating and
direct current, the winding should be connected as shown in
Fig. 1 2 , but if alternating current operation alone is required,
it may be connected as shown in Fig. 13. The armature
should then have zero impedance, for the same reasons that the

FIG. 14.-Resistanceleads
to commutator. a. c. motor.

FIG. 15.-Arrangement of resistance leads to

commutator ln a. c. motor armature.

PIG. 16.-Transformer actlon of torque field on armature coil short circuited

b y brush.

primary of a transformer has no counter e.m.f. when the secondary

is short-circuited.
Bad commutation on any motor is due to the brush short-circuiting two commutator bars of different potential, this resulting
in a flow of current which, when interrupted as the bar leaves the
brush, causes an arc or spark. I n the case of the resistance lead
type of motor, the commutation is improved by introducing a
resistance in the path of the short-circuit current (see Fig. 14,

234

ELECTRIC RAILWAY HANDBOOK

which shows that the resistance is only in circuit when the coil is
being commutated). The mechanical arrangement of the resistance
leads in the armature is shown in Fig. 15. The sparking voltage is
made of two components. One is a voltage due to rotation of the
commutated coil in a leakage flux field set up partially by the
main field and partially by the coil itself. This component exists
in all commutator apparatus whether alternating or direct current,

Wage due fo fram6rmer acfibn

PIC. 17.-Components

of sparking v o l t a ~ r n.
, c. motor.

and is, of course, in phase with the armature current because the
main field flux is in phase with the armature as well as the armature
reaction field. The other component is a voltage due to the
transformer action of the torque field flux on the coil being commutated (see Fig. 16). This last component, being due to transformer action, is a t right angles to the
motor or armature current. These two
1
components with their resultant are
shown in Fig. 17, in which i t should be
noted that the first vector depends entirely on rotation, while the second exists
even a t zero speed. I n the doubly-fed
motor, i t was attempted to remove the
sparking volts resultant vector and thereby allow the removal of the resistance
field
leads. This was done by creating a n
opposing voltage equal to the resultant
voltage shown in Fig. 1 7 . Thisopposing
voltage was produced by impressing o
Fm. 18.-Doubly-fed a. c. voltage across the compcnsating windmotor with impressed vo'- ing (Fig. 11). The diagram of connectage across compensatin~
tions was then as shown in Fig. 18.
winding.
The flux in the compensating field, or
cross field, must be a t right angles to the impressed voltage. A
voltage generated by rotation through this flux must be in phase
with the flux, consequently the commutated coil, on rotating or
moving through this flux, has a voltage generated in it which o
poses the resultant voltage shown in Fig. 17. I n Fig. 19 i t can
seen that the two voltages practically neutralize each other, and
consequently, resistance leads are not required.
Due to the fact that the cross field winding is distributed across
the faces of all the poles, considerable exciting current is necessary

SINGLE PHASE MOTOR THEORY

.Vdfoge acrossbars due fo mfafion

ofcoil through cross field flux

PIG. 19.-Sparking voltage ncutrnlizcd by cross field

PIG. 20.-Interpole

type a. r . motor.

Vplfage
--.------.Uolfage ------.r

*. .

--).

FIG. 21.-Connections

of interpole type a. c. motor.

235

ELECTRIC RAKWAY HANDBOOK

to maintain this flux field. The flux is .only useful in the zone of
commutation, therefore if all but the useful flux could be eliminated,
the power factor of the motor could be raised and the rating of the
motor increased, by elimination of heat due to e ~ c i t i n gcurrent
and iron loss due to cross field This was accomplished in the interpole type of motor. I n this motor the stator has a compensating
winding similar to that shown in Fig. I I , and connected as shown in
Fig. 1 2 , but in addition, there is a small winding around a tooth
in the commutating zone, which is called the interpole winding
(see Fig. 2 0 ) . The connection diagram is as shown in Fig. 21.
Study of Figs. 18,19, 2 0 and 21 will make clear how the commutation effect of the double-fed motor is obtained in the interpole
motor, but with a gain in rating. I t is evident from the foregoing
that a t zero speed no counteraction is obtained. I n the doubly-fed
motor the voltage impressed on the cross field is transformed over
into the armature, making this voltage equivalent to having been
applied on the armature plus torque field circuit. This complicates
the control to some extent, which is not true in the case of the
interpole motor.
General Strudural Differences between Single Phase Commutator and Direct Current Motors. The single-phase series commutator motor is a low voltage machine. For traction purposes
the armature composes a simple multiple wave winding with only
one turn between adjacent commutator segments. This is t o
limit the voltage induced by transformer action in the coil shortcircuited by the brush when undergoing commutation. The practical limit in the value of this transformer voltage likewise limits the
value of the torque field flux which induces it, below that field
strength ordinarily employed on direct current motors. For a
given torque the armature current therefore is relatively greater.
For these reason, (a) the armature diameter is larger as compared
with that of a direct current armature, (b) the commutator is
larger and wider and has a greater number of segments than that of
a corresponding direct current motor, (c) the field poles are short
and stubby as compared with the direct current motor, and ( d )
there are a comparatively large number of poles on the alternating
current motor. Other structural differences are (e) the laminated
core of the alternating current motor, (f) the use of a distributed
auxiliary winding on the alternating current motor in addition t o
the main field winding, and (g) the smaller air gap on the alternating
current motor. The field is laminated in order to reduce the
losses and associated heating due to the alternating flux in the iron.
Types of Single Phase Cornmutating Motors. The purpose of
the compensating winding, as shown above, is to neutralize crossmagnetization, thereby improving commutation and power factor.
Compensation is classified as conductive or inductive, depending
upon the process by which the current flowing in the compensating
winding is obtained. If the compensating winding is connected
to receive current from a n external source, as by connecting i t in
series with the main motor circuit, the compensation is said to be
conductive. If the compensating winding is closed on itself so that
a current due to a directly induced electromotive force flows within

SINGLE PHASE MOTORS

A rmoture

field Cod

Armature

held Cod

SmlesMofor

Sertes Motor

( c ) - I n t ~ r ~ l e T of S9nesMdor
n
t
h~huntpX5lstorE I C I ~ ~ W

(0)-Doubly Fed Type of Serm (f)-Dwbly FedTypoofhes

Motor- Runnlng Connectlons

Motor-Stort~ng
Co~nectms

(d )-Ide

oleTypoof SeriesMofor
w ~ f h ? m f ~ r mErc~tot~on
e~

Armutum

Armutum

( h)-Thornson Repulson

Armfure

Le/dCo~l

Motor

( ~ ) - S ~ r t eCompe.Jotd
s
Repuls~onMotd

(9)-Induction SerlesMotor

(j)-Wlnler-E~chber~

(k)-LabvrRepulslon M
o
b

Repulson Mutor

PIG. zz.-Vdnous

types of s~nglephase commutatrng a. c. motors.

238

ELECTRIC RAILWAY HANDBOOK

itself, the compensation is said to be inductive. On the New Yoik,

Westchester and Boston Railway, inductive compensation is us&On the New York, New Haven and Hartford Railroad, motors
intended only for single-phase operation are inductively COW
pensated, but those intended for both single-phase and direFt
current operation are conductively compensated. Insulation straln
to ground is made negligible in the inductive compensationWhen motors operate only on alternating current, the difference
in results of either inductive or conductive compensation 1s
inappreciable, but where the motor is to be operated with direct
current, conductive compensation should be used. There are three
general types of single-phase commutating motors, namely, the
series, repulsion and series-repulsion. Of these the former two are
in most general use in America (see Fig. 22).
Single Phase Motor Capacity. The single phase motor is essentially a high-speed motor, having a high copper loss and low core
loss, and therefore is adapted more particularly to constant-sped
running, and suffers in comparison with the direct current seri:s
wound motor when used for high rates of acceleration. There 1s
no lack of starting torque with the single phase motor, but a large
tractive effort is obtained only a t the expense of a large copper lo*,
so that single phase motors are unsuitable for rapid transit service
demanding high rates of acceleration, not because such motors capnot furnish the tractive effort required, but because the copper
loss incident to such high tractive efforts would heat the motors
unduly, or necessitate a motor much heavier than would be required
with direct current. I n interurban or express service with infrequent stops, the smaller core loss of the single phase motor brings
i t more nearly on a par with the direct current series wound motor
as regards output per pound weight of motor. I t is possible therefore to use the figures of the table on page 231, as applying to single
phase motor capacity, where stops are not more frequent than oPe
in 2 miles. For a higher frequency of stops,
special
study should
.
.
be made, as before outlined.
Three Phase Motor Capacity. Owing to the few applicatiofls
of three phase induction mitorsto motor cars, no generafitateme~t
can be made regarding the capacity of such motors for service
performed. The induction type of alternating current motor IS
not a t all ada ted for high rates of acceleration, owing to the poor
efficiency of t l e equipment during fractional speed running. Tbe
field of the three phase induction motor lies in the direction of a
service calling for constant effort a t constant speed, and the problems wherein this type of motive power can be considered are
special and infrequent as to place it entirely beyond consideration for
general motor car application, and to make its use in locomotives
the subject of special study.
Locomotive Motor Capacity. 1.ocomotive operation demands a
special treatment of the motor capacity subject, in most cases
requiring such special knowledge of motor construction as to prevent any general conclusions being drawn. With a locomotive
working a t 75 per cent or more of the slipping point of the driver%
there is a possibility of but small overload and hence the motor

THREE PHASE INDUCTION MOTOR

239

may be designed with lower density, in fact more like a stationary

motor. I n the absence of extreme overloads, the question of commutation becomes secondary, and the selection of a motor becomes a
matter of its ability to radiate the heat generated in the specified
servlce. The whole question of motor capacity for locomotives
is so intimately associated with mechanical problems of locomotive
design that it must be placed in the list of special problems requiring
the careful cooperation of the manufacturers. These remarks do
not apply to motors for switching locomotives, where the service
is intermittent and the consideration more like that in motor car
service, with weight on drivers and coefficient of adhesion of
increased importance in determining maximum possible tractive
effort or drawbar pull.
Three Phase Induction Motor. The three phase motor is called
a constant speed motor (see p. 356 for methods of changing speed).
I t has the characteristics of a direct current constant speed motor.
Neglecting control, the speed of the three phase induction motor
depends upon the frequency of supply, and the spced falls off but
little as the load on the motor is increased. The speed of a train
driven by three phase induction motors ordinarily remains nearly
constant up grade and down grade. 'The three phase induction
motor regenerates energy, forcing it back into the line when the
motor is driven above its synchronous speed. This process takes
place as t'ne 'train hescenhs a grabc, anh since fnis regenerailon'~i1ngs
a load on the motor acting as a generator, the train is not allowed
to accelerate beyond the point a t which the force due to gravity
is balanced by the forces it is overcoming in driving the motor
acting as a generator. Since the three phase motor is a constant
spced motor, in ascending a grade it will make a demand on the
power station proportional to the torque required and the losses to
the motor. This demand may, however, be reduced by the energy
of regeneration of trains descending. I n order to operate on long
up grades a three phase induction motor must have h ~ g hcontinuous
capacity.
The starting efliciency of three phase motors is low, so the motor
is not well suited to service having frequent stops. The tractive
effort may be held constant during the starting period, but this is
done a t the expense of energy lost in the motor starting resistance.
Since there is no clectrical connection betwecn the primary winding and the secondary winding of an induction motor, and there is
no commutator, this type of motor may be built to operate on voltages which are high compared with those on which nlotors of other
types are operated.
Speed of Induction Motor. The speed of an induction motor is
equal to its synchronous speed minis the slip. 'The value of the
slip increases a t practically a constant rate from nearly zero a t no
load to about 2 to 5 or 6 cent of synchronous speed a t full load.

-.

in which S = synchronous speed, revolutions per minute

N = number of pairs of ~w)lcsin the primary winding of
the motor

240

ELECTRIC RAILWAY HANDBOOK

= frequency

second

of applied electromotive force, cycles per

in which Sl = actual speed of motor, revolutions per minute

s = slip, per cent.
TABLE
OF SYNCARONOUS
SPEEDS

Number of pairs of poles

jnmotorpnnary
Pnqnency, cycles per
second

Synchronous speed, revolut~onsper

mlnute

15

25

60

FIG 23 -Sample dlrect current rallway motor characterrstlc curves

General Electric No. 247 motor, 3s h.p., too volts, gear 63. plnton IS, ratlo
4 20. wheels 30 In.

MO'I'OR CIIARACTERISTIC CURVES

241

Railway Motor Characteristic Curves show:

tractive effort (pounds)
speed
(miles per hour)
eaciency
(per cent)
and, in the case of alternating current motors:
power factor (per cent)
each a s a separate curve, plotted against
current
(amperes)
Such curves as furnished by the manufacturer a ply to a definite
motor only when that motor i s operatingat the d e g i t e voltage and
with the definite gear ratio and driving wheel diameter for which
the curves are drawn; Zor application to any other voltage, gear
ratio or wheel diameter, the curves must be changed as described
on page 246.

Pmpres

FIG.24.-Sample

d~rect current r a ~ l w a y motor character~st~c

curves.
West~nghouseNo. 3o6CVDq motor. 60 h.p 600 volts. gear 7 2 . pinron 15.
ratlo 4.80. wheels 33 In.

Typical characteristic curves are shown for direct current motors

as Figs 23, 2 4 and 2 9 ; for a single phase series alternating current
motor as Fig. 2 5 ; and for a three phase induction motor as Fig. 26.
There is a marked similarity in the characteristic curves of all
railway motors of the same general type. As illustrative of this,
Fie;. 27, has been prepared, in which the characteristic curves

242

ELECTRIC RAILWAY HANDBOOK

of a number of self-ventilated, interpole, direct current motors,

have been combined. The nominal ratings of the motors used In

preparing Fig. 27 range from 25 to rqo horsepower, and the current,
tractive effort and speed ordinates of all of them were reduced to
terms of per cent of those values a t the nominal, or one-hour rating.
Fig. 27 shows the maximum, minimum and average efficiency,
tractive effort and speed characteristics of all the motors considered,
and although it was derived from a study of self-ventilated, interp i e motors, it well illustrates the general shape of the characterlstics of all direct current motors. The shape of the typical
characteristics of the single phase series motor differ from that of the

Amperes

Frc. 25.-Sample

single phase alternating current railway motor characterlstlc curves. .Westinghouse No. 41zB motor. 1 2 5 h.p.. joo volts.
zs cycles, gear 60, prnlon 1 9 , ratlo 3.16. wheels 435, ln.

direct current in that its

wer varies much less with the speed,
while the induction motor g s a speed characteristic that is almost
Bat.
This marked similarity in railway motor characteristic CUNeS
has led to the use, by some engineers, of average curves, particularly for direct current motors, such as those shown in Figs. 27
and 28, for the purposes of preliminary calculations. In such use,
any desired values may be given to the nominal or one-hour rating
of two of the three elements, speed, tractive effort and input, these
being connected together and with efficiency by the formula:

MOTOR CHARACTERISTIC CURVES

I

243

- 1.99 ST
EU

where I = motor current, amperes

S = speed, miles per hour
T = tractive effort, pounds
U = efficiency,decimally expressed
E = voltage
Fig. 28, in which ordinates are shown as per cents of the nominal
rating, thus may be used for an approximahon of the characteristic
curves of any direct current motor, by assigning any values of

FIG.$6.-Sample three phase induction railway motor characteristic curve.

current, speed and input to the nominal rating, so long as the
above formula is satisfied. Generally, speed and tractive effort are
assumed and the current is derived from the above formula. If
these assumptions are made for the nominal rating of the motor,
the per cent ordinates of Fig. 2 8 may be expressed decimally and
used as multipliers, thus enabling the curves to be read in amperes,
miles per hour, and pounds tractive effort. Having determined
the desired values of nominal rating speed, tractive effort and current, a commercial motor nearest to the proper rating may be
selected, and knowing the clriving whed diameter, its gear ratio is
determined. For further calculations, the characteristics of this

244

ELECTRIC RAILWAY HANDBOOK

commercial motor may be substituted for the general characteristics

of Fig 28.
I t is interesting to note that the characteristics of the direct
current railway motor bear these inter-relations: the speed is approzid e l y inversely proportional to the cuhe root of the fractibe effort
and to the sqrrare root ofthe current. This rule is based on an qssump
tion of constant efficiency, SO that it should be used only for rough
preliminary calculations; the dotted CUNeS of Fig. 2 8 correspond to
the relations as above expressed, and show the extent of the error
likely to result from its application.

PIG. 97.-Radway motor characterrstrc curves. Current. tractlre effort

and speed values rn per cent of nomrnal rated values Maxtmum, mlmmum
and average for group of self-ventrlated. mterpole. dlrect current motors,
from as to 140h.p.

The approximate rule as given above lends itself to exprwion on

the so-called polyphase slide rule, as has been pointed but by
E E. Kimball. The polyphase slide rule shows four sets of logarithmic scales as follows, the latter named two being in addition to the
first two which are shown on the ordinary slide rule:
D scale: the bottom scale, running from I to 10 in the lesgth of
the rule
A scale: the top scale, running from I to 10 to roo in the lehgth of
the rule, and indicating squares of the values on the D
scale
E scale on the lower edge of the rule, running from I to 10 to roo
to r c m , and indicating cubes of the values on the b sale
CI scale: the middle scale-same as D scale but inverted or
reversed in direction.

MOTOR CHARACTERISTIC CURVES

245

If the various scales are set with the ends coinciding, the runner
may be set a t the point on the A scale correspondng with any
per cent of the nominal rated amperes current input; the corresponding point on the middle or CI scale will indicate the corresponding per cent of the nominal rated speed, and the corresponding
point on the cube or E scale will indicate the corresponding per cent
of the nominal rated tractive effort; all in accordance with the
approximate relation beween these values as indicated by the
above quoted rule. The polyphase slide rule in this way affords a
quick approximation solution of problems involving direct current
railway motor characteristics.

Per Cent of Rated Current

FIG.28.-Ra~lway

motor characteristic curves. Current. tractive

effort and speed values in per cent of nomlnal rated values. Sol~dline.
average of direct current motors. Dotted lines from formula and as read
from polyphase slide rule.

Gear Ratio is the ratio of the number of teeth on the driven gear
to the number of teeth on the driving motor pinion. The ratio of
the number of revolutions of the motor shaft per unit time to the
number of revolutions of the car axle driven by that motor is equal
to the gear ratio. The speed a t which a car having a given driving
wheel diameter will be moved by given motors a t a given current
and electromotive force depends (neglecting the variation of train
resistance) upon the gear ratio. The approximate formulas connecting armature speed, car speed, gear ratio and wheel diameter
are :

246

ELECTRIC RAILWAY HANDBOOK

where R = armature speed, revolutions per minute

S = car speed, miles per hour
G = gear ratio as defined above
d = driving wheel diameter, inches.
The sum of the diameters of gear and pinion is twice the gear
center distance and is therefore constant for a given motor. Then
so long as the diametral pitch and gear center distance are constant,
the total number of teeth in gear and pinion is constant, irrespective
of gear ratio. This fact should be remembered in finding the
number of teeth in gear and pinion for changing gear ratio. For
instance, a gear ratio of 091. (4 60) is to be changed to a ratio of
3.20. The total number of teeth is 84, and with the new ratio,
I ) or 4.20 times the number of pinion teeth.
this will be (3.20
The pinion will therefore have (84 + 4.20 =) 2 0 teeth, and the
gear will have (3.20 X 2 0 = ) 64 teeth.
Maximum Gear Ratio. The maximum gear ratio of any motor
is limited mechanically by one or both of two things. The minimum pinion must be strong enough for the capacity of the motor,
and the maximum gear must permit sufficient clearance below the
gear case for safe operation. I t is not practicable to use a pinion
so small that with the best available material the teeth are weak
or there is insufficient stock between the root of the tooth and the
bore or keyway. The permissible clearance under the gear case is
somewhat variable on account of the different characters of roadway in use. For example, a clearance which would be safe for low
speed on an unpaved track with T-rails might be too small for use on
streets with grooved rails and high crowned paving between rails.
The approximate clearance under the gear case in any instance is
readily determined when the number of teeth in the gear, the diametral pitch and the wheel diameter are known. The diameter of
gear over the ends of the teeth is equal to the number of gear teeth
plus two divided by the pitch. The outside diameter of the gear
case is approximately one inch greater, on account of the clearance
between teeth and case and thickness of case. Then the clearance
from the bottom of gear case to top of rail is one-half the difference
between wheel diameter and outside gear case diameter. A 69tooth, 3-pitch gear has an outside diameter of (69
2) t 3 =
23.67 in. For this gear the.outside gear case diameter is 23.67
I = 24.67 in. On a 33-inch wheel the clearance will be (33
24.67) + 2 = 4.165 in.
Changes in Characteristic Curves, Due to Changes in Gear
Ratio, Wheel Diameter or Voltage. When it is desired to apply
a series railway motor to some other gear ratio, driving wheel dlameter, or voltage than those to which the original characteristic
curves apply, the ratios shown by the following formulas may be
used. Note that voltage changes should be applied with care;
when small changes are made, it may be safe to use the approximate

+
-

CHARACTERISTIC CURVE CHANGES

247

formula in which i t is assumed that the speed ordinates vary directly

with the voltage; the exact formula, shown later, indicates that i t is
the counter electromotive force, and not the impressed voltage,
that controls.
The ordinates for the lraclive effurl curve for a series motor with
any gear ratio or driving wheel diameter may be derived from the
ordinates of the original tractive eforl curve for that motor as
follows:

in which T 1 = tractive effort ordinate for any current value on the

derived curve
T = tractive effort ordinate for the same current value
on the original curve
GI = gear ratio for derived tractive effort curve
G = gear ratio for original tractive effort curve
Dl = driving wheel diameter for derived tractive effort
curve
D = driving wheel diameter for original tractive effort
curve.
The approximate values of the ordinates of the speed curve for a
series motor with any gear ratio or driving wheel diameter operating
a t any other voltage differing slightly from that a t which the original
speed curve was obtained, may be derived from the original speed
curve as follows:

in which S1 = speed ordinate for any current value on the derived

CUNe
S = speed ordinate for the same current value on the
original curve
E1 = voltage for derived speed curve
E = voltage for original speed curve.
(NOTE: When the voltage is not changed, that is, when El = E ,
the value given for S x is exact. See p. 248 for formula for exact
change due to voltage change.)
Example of the Derivation of Tractive E J r l and S p e d Curves.
Fig. 2 9 shows tractive effort and speed curves (dotted) derived
from the original curves by the above process as follows: The
original gear ratio, wheel diameter and voltage are 2.78,33, and 500,
respectively, and i t is desired t o derive the tractive effort and
approximate speed curves for a gear ratio, wheel diameter and
voltage of 3.83, 36, and 550, respectively. I n this case, G1 =
3.83, G = 2.78, Dl = 36, D = 33, El = 550 and E = 500.
Therefore

and

248

ELECTRIC RAKWAY HANDBOOK

Thus in Fig. 29 the dotted tractive effort curve is the locus of

all points the ordinate of each of which is equal to 1.26 times the
corresponding ordinate on the original tractive effort curve, and the
dotted speed curve is the locus of all points the ordinate of each
of which is-equal to 0.87 times the corresponding ordinate on the
original speed curve.

38
36
34

1800
17lX

32

l5W

30

l a

>:

'

,:a. ;

m w o
28 pa'

5 24
loo

:lax,

XI 4 m < W

,80
i70

4 60
2 50

& 40

230
a0
10

;IS<

900

165 700
14

MC

12

so0

10

WO

3Kl
0

100

0 10 20 30 40 50 60 70 80 90 lOOllOlZOl301NJl50XgD
Amperem

FIG.29.-Illustrating

change of tractive effort and speed curves for a

change in gear ratio, driving wheel diameter and voltage. (GE-87 motor,
60 h.p.)

The actual change in the ordinates of the speed characteristic

of a series railway motor due to a change in voltage is proportional
to the counter electromotive force, as follows:

where I = current
r = resistance of motor
Sl = speed ordinate (at I current) on the derived curve
S = speed ordinate (at Z current) on the original curve
El = voltage for derived speed curve
E = voltage for original speed curve.
As the derivation of a new speed characteristic by the above
formula involves a se arate calculation for each value of current,
it is convenient to ma e use of a graphical method for its application, such as the one proposed by A. M. Buck, and illustrated by

CHARACTERISTIC CURVE CHANGES

249

Fig. 30. Here the original speed-current characteristic is for 6volts, and it is desired to plot a new curve for 500 volts. Lay off a
second scale of amperes on the same base line and with the same
ordinate values of amperes, but offset somewhat to the right, as
shown. On the ordinate representing zero current on the second
scale, and to any convenient scale of volts, the point E is a t boo
(the original voltage) and El is a t 5 0 0 (the new voltage). The points
C and D are located below 600 and 5 0 0 volts, respectively, by an
amount representing the It drop for the current ordinate on which
they are plotted (in this case 2 5 0 amp.). The lines EC and EID
are drawn. To locate any point on the new speed curve corresponding to S on the original CuNe, proceed as follows: Take the
point A on the line EC and a t the current on the second scale of

Amperes
PIG. 30.-Graphical

method of constructing speed-current motor characteristic for voltage differingfrom original curve.

amperes corresponding to the current a t the speed S (in Fig. 30,

150 amp. is used for illustration). Then the line AS produced
will cut the base line a t some point K. The point B is on the line E I D
and a t the same current ordinate as A. Then the point S1,
where the line KB cuts the current ordinate of S , is the point on the
new speed curve corresponding to S on the original curve. Other
locations of A and B corresponding to other currents and other
points S o n the original speed CUNe will result in other locations for
K and give other points on the new speed curve, in the same
manner.
A slight modification of Fig. 30 may be used to determine the
performance of a motor when in series with an external resistance.
In such case, locate F a t a distance below the line volts corresponding to Z(R t), Z being the current on the ordinate of which F is
located, R being the external resistance, and r the resistance of the

250

ELECTRIC R.IILW.4Y HAN1)ROUK

motor. The procedure will then be as before, except that the lwints
B will be taken on the line L'P instead of on E'D.

'The use of the method shown In Pig. 30, and described almve,
generally involves either a replot ting of the original speed-current
curveor a piecing out of the sheet on which it is plotted, toaccommodate the construction shown a t the right of Fig. 30. Should either
he inconvenient, the whole process may be performed within the
limits of the motor characteristic curve sheet by a modification of
the method as shown in Fig. 31. Here the base of the volt-ampere
diagram is taken the same as that for the speed-current curve. The
lines EC and E'D are located as in Fig. 30. The point N is located
so that E'N = OE, and the line N M is drawn a t an angle of 45
degrees. Then for any point S o n the original speed curve, take B

Amperes
FIG. 3 1.-(;8.1~11a~.#I 1111.1110d (11 C ~ ~ T I S I ~ U C I I F
I IP
RCP~I-CU~
noto
~ PorI II ~hdrartCr~ztlrtor \.#>Itaged ~ f f e r ~ front
~ i g orig~nalcurve.

(at the %%mecurrent) on the line E'D, project B to an intersection

with S d f at Y , ant1 draw PO. Then with a pair of dividers lay off
the speed value of S on the base line as OT, project T up to the
diagonal OP and thence across to the proper current ordinate a t S1,
which is the point on the new speed curve corresponding to S on
the original curve.
Effect of Variation in Driving Wheel Diameters. When all of
the driving wheels on a motor-car or train are not of the same
diameter, the characteristic curves of the motors will vary accordingly. Since the speed of all wheels must be the same in miles per
hour. the revolutions per minute of the armatures will not be the
same, the motors driving the larger wheels will take more than
their proper share of the load and consequently will show more
heating than those driving the smaller \vheels. The difference in
current a t various speeds may be determined from speed-current
characteristics of the motors, corrected for the different wheel

diameters, and the difference in heating m y then be found I)y

calculating thc diEerenccs in core loss and 12K loss clue to the
differences in currents and armature s1,eeds. Although the shape
of the spccd-current characteristic and the distribution O( losses
between core and copper varies somewhat with diKercnt designs
of motors, A. L. Rroornall has worked out the curves shown as
I'igs. 32 and 33 based on an average speed-current curve such as in
Fig. 28, and using the average distribution of losses as found in a
number of direct current railway motors. Figure 3 r shows thc
difference in current, and Fig. 33 the difference in heating which

PIG. 32.-Diffprencez in current due

to drfierenres In wheel diameter.

FIG. 33 -DiRerences
i n heat in^
due to d~ffrrencesIn wheel ~ l ~ a n ~ r t v r .

results from various differences in wheel diameter and a t diflerent

speeds. Mr. Broom11 p i n t s out that the permissible difference
in driving wheel diameter depends upon several considerations,
among which he mentions the margin in temperature rise when
the load is divided equally betwcen the moton, the relative number
of motors with the large and small wheels, and the ordinary operating speed of the car. If the car has three motors on large wheels
and one on small wheels, the condition is not so serious as with
three sets of small wheels and one large. If the service is such
that thc motor operates much of the time a t heavy loads, a given
per cent difference in wheel diameter is more serious than in service
where most of the running is a t high speed.
Resistance of Direct Current Motors. 'l'he electrical resistance
of a large number of direct current railway motors, lrom data

252

ELECTRIC RAILWAY HANDBOOK

furnished by the manufacturers, is shown in the tables on pages

254 to 259, inclusive.
E. E Kimball (Trans. Am. Inst. Elec. Engrs., Vol., 33, p. 1715)
proposed a method of determining the approximate reststance of a
&r&t current railwav motor from itsad5ciencv curve and the
following formula:
= (8K2 - S K I - I S ) E
1100 I
(approximately)
in which R = approximate resistance of motor, ohms
K1 = total losses a t I current, per cent
.Kz= total losses a t 2 1 current, per cent
I = any current not less than % nominal rating, but not
more than H largest current for which efficiency is
known, amperes
E = voltage a t which efficiencies are shown.
K Iand K? may be taken from the efficiency characteristic of the
motor, belng l o o minus the corresponding efficiency. The method
depends upon the assumptions that a t % nominal rating or more,
the gear and friction losses amount to about five per cent (see
A.I.E.E. rule No. 5339, . 219) and also that the core losses may
be taken as proportionar to the cube root of the current. The
accuracy of the method depends upon these assumptions and u
how accurately the efficiencies can be read. As errors in r e a g
efficiencies may be cumulative, the resulting error in a single
determination may be great, so that it is advisable to make check
calculations using different values of current. This method was
tested on nearIy a hundred direct current railway motors of various
types and capacities, the resistances of which were known; in 90
per cent of the trials the indications were within 10per cent of the
true values, in about half the trials within 5 r cent.
Resistance Losses due to Bmsh Contact.
calculations involving the IT drop in direct current railway motors, it is usual to
assume a drop of two or three volts due to brush contact resistance
and to use this drop irrespective of the current.
Resistance of Car Wiring. In calculations involving the total
resistance of the motor circuit, it is usual to assume the resistance
of the car wiring as 2 0 per cent of
w**-=Ithe motor resistance, or the total
resistance per motor including wira'z
ing as 1 2 0 per cent of the motor
resistance.
Measurements of Equivalent
Heating Current. The equivalent
FIG.34 -Watt-hour
meter con- heating current or root mean square
netted t o masure I'R loss.
current may be obtained by the
following method : A specially
calibrated watt-hour meter is connected to record the energy lost
in the motor field winding, as shown by Fig. 34, then the equivalent

heating current will be equal to

RAILWAY MOTOR DATA

s@

in which c = energy lost on motor field, watt hours

R = operating resistance of motor field, ohms
T = total time during which the equivalent current is
desired (sum of accelerating, coasting, braking and
stop periods), hours.
Another and very satisfactory method is by use of a "vacuum" or
heat-insulated jar containing a measured quantity of water in which
is immersed a known resistance, the leads to which are brought in
through the cork, in company with a thermometer and a stirring rod.
Since the heat generated is proportional to the square of the current,
the temperature of the water will be raised in the same proportion
(exce t for heat losses, which through the vacuum jar are very
small7. The instrument is calibrated by noting temperature
rises resulting from the passage of various constant currents for
given periods of time. In measuring large currents, a shunt
or transformer is used, so that only a given pro rtion of the current
is passed through the apparatus. By this met/% the r.m.s current
values may be measured on either direct or alternating current, and,
unlike many alternating current instruments, the accuracy is not
affected by frequency or wave form.
Railway Motor Data. Tables on pages 254 to 259, inclusive,
show trade numbers and ratings of Westinghouse and General
Electric railway motors, and include not only the modem types
but also a number of the older types which remain in use to some
extent. Electrical resistances are shown, whether or not commutating pole, type of ventilation, type of frame, number and size
of brushes, maximum safe armature speeds, maximum gear ratios,
maximum axle diameters, minimum wheel diameters, rail clearances,
and weights.

254

ELECTRIC RAILWAY HANDBOOK

----- .-.*.-.-.----.-.-.-...,+-++'

am-3

JO "dL1

+ ~ ' + '

'.5"

OOOOO

002$2

-uo!?q!wan

0
UI

4
"7

b a r n

adA.r.

."ti
::

;ij

00000

UOOOU

*"
$~~~~

' u

38444 833s3 $2;;;

55snz
N5n%zZ*$\Y ?xXXXXX XXXXX XXXXX XXXXX
XXXXX

xxxxx

XXXXX

Xxxxx

WES'MNGHOUSE MOTOR DATA

255

2
.

g g z g z g g g g g g g g ~ 5g g g g g g g s g g g g g g g

ammmm mmmmm mmmmm mmmmm mmmmm mmmmm

n n n n n

n-ronn

n n n n n

n n n n n

n--1-

0 0 0 0 0
0 0 0 0
dmmm3
n----

0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0
ommmca m - t ~ v d
n--n n n-n n n n n n-

O . * . I

r n n n n

- 2211.2 2%%22 22-??2 41211 X 4 f ?,Z ",%40,4

\n.h
A
n
7
X X X X
--

$6

lo lo w

Y1C CQ 10

10 10 Y) d +

-?pz.C
1 0 2 ' 1 dd% d p T - d

gXgg% g?gg%
8%888
4%%2%8828g Pi?%%%
Q F C C ~ Q C C I O Q a a m m o a m m m m b m Q ~ z~_~-ezc,
- o o o o
n100om m o o r n o -n- n-nn
--- - gzggg
Pg%gz ",--$mg
wW00
0 0 0 0 0 n 0 0 0 0 O*OOonmom nwwoc oa~m10
ooooo

n n n n
-n-

o-w-n
001000

- o n - r

dofdd

r o o o m

wet-mm
o

..-

o n o o d

.b

10
wN8%G

FZZZZ EZfZd a"dF23

Z m n n o oSf;-bo
p????
????N
????? 1>1?? ???1
7*???
$2

:;

-23;:s

CI

c,CICIH

: z . g .:gg:
.: .: :. .: :. :. .: .: .:% .: :. .,.
sgzoz ea ;as :zz;z
a:%::

O O O ~ O 0 0 .mm
mmooo
m w b w l ~d z - v r ~ . o w . n o

O O O O -

nw==w

hot-en

o n n n m

CI

::z:;oofi

:::a%
nwoww
ooooo

CI

gg223 Z22:% i?2888 $2824 8:%%2 38523

t - t - o v , ~ ) a w n w e CLOQQO P - ~ Q C ~a-m o o o
nor-r-e

s
y y> >
y? y4 qg
a.9-t-c

o o o o o
omorno

cmmoc
0 0 0 - 1
o o o o o

4
AmAAA
r--nn
00ooo

ooAAAk
noooo
ommoo

-0.-

xG%

-1

yp$q$

o&d.do;l
m o n n e
do000
o r - - o w ~ n nw w n m w

ELECTRIC RAILWAY HANDBOOK

256
a"",
("orJq

uo!lellluJA jo
-.

*****

* * K * l * *

3 0 Jd.4.L

$-2;;-

;gg;

>>

>

>>3

--

add^. mmmmmwm ~ s m w m mw m a L w m m m - w - a

---

--

- -

;
;
;
;
;
;
;;
;
;
;
;
;;g;
u l m m u l v s n

H U Y * * W

$g;g

>+>>

+>>>3>>

355)

* K * * K K

;;gig

U!lel

larmurur

.-

-.

* v ' A m n n

cnmmvrmm

;;;g
;;

-------

X X ' X ~::xr-xx-n'
~

L
X
L
:
~
"
X
~
"

y~-xa-z
n n ~ n - -

*nnm-r

r w . . w - - -r - - - r

xxxxxx

XXXXXXX XXXXX XXXXXX

XXXXXXX XXXXX
n

"a"

N"

--

N N"

mxzxx X X X X X

" " n * " " " om ooon o m

""

8 8 8 8 8 8 8 % % % % % %%%%P88 8 8 5 8 8
o, .g o, :-zOo_z"o,."o,----.ocO
:z:2:2
-$2
-. .-.
P ~ D W - W O e e e e ~e - m W e e

- .

0 - 0 0 0

FJFsT$F sFs3s zsssss FZ*>J

xS v o m 2* * xxxxx
FRXXXX
,K.~.YXYX
ooooo *ooooo o o o m o o

O!IEJ2-a -5%

T)"

0 0

aumJ JolOH

5s
x XXXX'f: X$:kXX$:
m n 0 0 0 o Q mC)r)C)O d * * ' P P P

-bY)*mV)

,u!..me!p l ~ a q a.U!N

V.

.uz - W E !alxe
~ TEN
i:

-.

- ---

r.
i:

.z * E

>

--

-,dzr

g,o~,<o,O,

..
r

.$

U. Y.

\"

00n000.0
r o n o o o c
onm ma a o
- n n n o n n
"0moo"o
00-Cr-a
vroonnrr n n n n--n n
o lS
0 n
mmopmn

asEslEaRpoe
_I

,.-

omo-o
a b~ -00
o n m o o n o n o o m m noonoo ononoo

0 0 n-

Crn

w o n n n 0 0 0 0 0 0 0
0 0 0 0 0
n o - n p . m w o o o o mmvlmmv,
I- OOOCLOCLQ Q O 0 0 0
l o o o o U P * * * *-?*?Lo_
mwoov) 0 0 0 0 0 0 0 0 0 0 0 0
nnont- noemem o o o o o o
I-c-wct- e a * b d b n n n e n n
nnnnn do0*oo - n w o - 2
n
c
nvrnm
n c a w w n ~ 1 * 3 0 0 denQwO -CaL O
mo

E
2

I
o,:z;a. . $.z."o.ozo; . . * ",,,<2
lh
.OOY)C.

. . .

--

-a

0 0

$3

OWW,:~:~
'5,8$>:

Zz

230g5b oo~~%~~~
aooap

~
-00r-8

r c

.d'H

- 9

~
, : ,2 ~
mz$**F:
m

-6

. "-"--,.,
-- d
o o o ~ o ~ o o o - o o o o o ~ o 0 o o 6 o ~ 6 ~ b

O -Om
Om
PD
0
mO0O0O 0t -8o8o 0m. o0

00
U
P0
O0
OnOoC -mog:%s
0
"NN

-Ion s g g g s g s s g g g : s : s z g g g g g s % s UI
6

m-vo-

$$Y4q$$$$$$

u<!X3q<$P,m - t - ~ w
gA&-ewm
OW***
L.r)momoo
mmmnmmm n m n - m

q--,-m.a

$$n$$$
2
~ r r n n n

CCCmCC
."VIQaOQ
m ~ ) m m m mn n n n m u

GENERALDATAON WESTINGHOUSE
SINGLEPHASE.\LTERNATING
I

J.ffrs.
K.3.

Nomlnal
I-hr. raring

11

Brushes

Max.

afe

Max.

----

CURRENT
?~~OTORS

So.

In.

1:

11;-.%
1.32-F

z:o
235
335
2 3 5

4
148-~I
151-8
156 1
403-A

250
loo
100
1.575

'
235

120

225
300

),

140
125
I72
150
310

215
6
600
680
640
360

i:!

630
440

16.420
,300
5,300
6 . 4,500
025

1
1
,

14,485
5.926
9,580
O.ao8
~4.375

. ..... .. ..
6%

$1
6

61

st$
5
5

I.
s1ko d;
..

36
33

4%.

,#

62
37%
50

6%
5)'

$i

23

... ..

~cn~lcss
496
'360

44ie

'917

'9(s
'968

7 93 i 7s

Bfis

810
1.500

3
3

HX,H

g,,

, Split

F
F

, Box

:::"

96x296

Egip
96xzd

1.000
1.350
1.000
r.400
830

3
4
3
4
4

StiXzffi
aiX2y
56x292
96Xzy
3BX13c.

1.600

3%

F
F

Box

BOX

Box
BOX

258

ELECTRIC RAILWAY HANDBOOK

x x x x x x x x x x xxx:

2:xxx

--nnn

nnnmn

xxnxx-zxzA.r: -8gxrg xxxxx

n-nnn

nnnmn

- -5 8 8 z z 0 0 0 0 0 0 0 0 0 0 00000
om-om
1r S
S 0 f g E S S 2 ZPB%B
r - n r - r r r
r r r - -~
*.,1*0
lo.,.,.
400CI
.,.,OCD

0 I

.ul ..rurrp

a p e .rely
-

5552

%s"$"~ 5 ~ ; s

w\c.

"nca
A

228Eg :%DO,;;%
z::z;
$2232
YO.?*:
9??~3.
:?:?
n . .. .. .. .. . .. .. .. .. : : : :* '19:::
lloAIl"d
: : : : : : : : : : ,. ., .. ..",
* . . .
. . .- . - -. .. .. . . . . . n
...
.. .. .. .. .. .. .. .. .. .. .. . . . . .. h.- .- ,,a.....
)10"f(
.. .. .. .. .. . . . .. .. .. .. .. .. " " p .. . . ..
. . . . . . . . . . . . . . . m * P . .. . .
.. .... .. .. .. .. .. .. .. .. .. .. .. .. h. .0 .0 .. .. ..
.
% l o i f f . .. .. .. .. .. .. .. .. . . . . . - m n . . . . . . . . . . . .. . .. - * * : : :
n n o m n wmnoo
o o o o m 00-01'n'd..rr :0"20"Li RGRZZ GR"9%,6 E'Dh%&!C:
smqo

' ~ ~ u e l s ! s aji e ~ ! ~ p a l a

za...

<

*
2W

. C *

q
zz
UgE
0 . 4
A
-

S<:

$52

PZ;:~

2 ' -

e
I

;-%=.-

o o o o o
OQP"*

n o o n 0
e-zzt0

8m 8m 8H 8m 8m 8n n8 m8 8m 8n 8n -8m8w8a8 o888:o
o0~3
m
0

GENERAL ELECTRIC MOTOR DATA

j p i p p

;*g;;

%***b

b****

gggy4
~

pi;;;

0 0 0 0 0

;=-la.

*****

$,b**>*

moo00

0 0 0 0 0 0

~ r n - rc o n - - - - ---

- & D = U - C ~ C O -

C C - o c

* ) 0 - 9 9 U

0 0

--

zzzzz
-

3bWOQ
q n m o n

z
XX
O ~ ' 0 0 ' 0-"dQb.-

00om0
omm-m
mnnQ-7
- 0 0 0 0 0 b00Fv)
*I-nmn
mnnmo
oooov)

0We0m
mmomo

-.

x
-we.
."l*-fy3?.0.

XX A%*
0.0

S Q O O ~ Wm

- cmett-m

mtP?,OQ
ooommn

wnmmo
t-t-~no
r?-v)no
0 0 0 0 0
h - m r n
-100o*1
m-m-v)
mob

o
naB1"2
eomae
e m wm
n o e o n

i
3

'3

E
X Z ~ " ~ Y ZuC
o o m t m d

r - n n n
me*omn

'2

CZBKSi 8I
mnm:
m
n
8%%86
sa2E ~ a o Og Z Q Z X %
or---. . . . . 0. 0.- m. -. . . . . . . .w.n .r -.c c. m.
n o o o
-. . -o?-m:oa
..
. n . z m m m z--:. 2
:fZi;

- W O O -

a'

.: .: :zm m4 4m4z 4e :: dn :Y;4,

dY;d&&&
r+bnmb
"LL
.- .. - o r - c ------.. .. . . o. n. r .o ..--: 409 0 0 0 o y o L O
.. -. .. Qo m* mmC;l*
-mar .
bF-oc t a m
$29
: z .:mz:
oo*)mm9
...
.. .. .. -.a. ~ m o .a n . b o o-m-*-m- ooo
OE
.. .. . n o A 4 o d : d : & A d ,.,n'&d;&
...
.
.It.,

,4-

'W"

9-

v)

.......a

-33sgs dO z ~g zQg
000-0

nm
o vo)oa e
Q
ca~
nma

m~
oo
ov
n )ova) n
czzom.-

;gnQz

o w
nn
oo
n
d
an

nn
v )mv v) m
C
l vw
l vo)

0 0 0 0 0
0 0 0 0 0
aQaaO

0 0 0 0 0 0
0 0 0 0 0 0
aQQQQ0

n
moon
am:--

A
CdAA
ba+*n

n n o n n

>X

ad

5:

6.2

(3

0,

.. 2

m44om 4 4 4 4 4
A&&&&
---nn
nnnnn

I-

0000::
,0000::
0 0 0 0 , ~ 0 0 0 ~
OQOQ-QQO-

w r

44044

<44++4

&CdA6 A&Jao J
v ) m m n ~aoaacc
nnnnn
n n n n n n

260

ELECTRIC R A E W A Y HANDBOOK

Two or Four Motors per Car. I n the choice between two and
four motor equipments for double truck cars, the determining factor
usually is the coefficient of adhesion, i.e., whether or not the available traction of two sets of driving wheels will give the required rate
of acceleration, although another consideration is that of reliability
or the proportion of total motive power lost by the disabling of one
motor. Where there are grades of more than five per cent, or where
rail conditions are poor, as in climates with bad winter conditions,
the adhesion limitation usually will dictate the four motor e q u i p
ment for sinqle car operation. With two motor equipments the
weight available for traction may be increased greatly by the use

Prc 35 -Comparat~ve we~ghtsof two and four motor equipme ~ t .

of maximum traction trucks. Where conditions are favorable

for the use of the two motor equipment, its less weight, energy consumption and maintenance cost frequently dictates its selection.
The relative weights of two and four motor equipments of various
capacities are shown by Fig. 35. The energy consumption of the
two motor equipment will be less than the corresponding four
motor equipment due to less weight and to the better efficiency of
the larger motors. The number of men required for inspection and
overhauling of motors is more nearly proportional to the number
of motors than to their capacities, while the relative cost of repair
material and replacement parts is not a t all proportional to motor
capacity; the resultant maintenance costs with two motor e q u i p
ments is estimated as 25 to 35 per cent less than with corresponding
four motor equipments.

SELECTION OF GEAR RATIO

Selection of Gear Ratio. If the gear ratio be increased, the
speed of the car a t a given current will be reduced, the tractive effort
a t the driving wheel treads will be increased, and the possible rate of
acceleration will consequently be increased. If the gear ratio be
reduced, the speed of the car a t a given current will be increased,
the tractive effort a t the driving wheel treads will be reduced,
and the possible rate of acceleration will consequently be decreased.
Thus, w t h a low gear ratio, greater current will be required to produce a given tractive effort and acceleration than would be requlred
with a higher gear ratio. (For effect of gear ratio and electromotive force on speed and tractive effort characteristics see p 246.)
The proper gear ratio to be used with a given car depends upon the
service the car is to render and the character of the track over
which i t is to be operated. Whether or not a gear ratio once
proper will be proper a t a future time will depend upon whether
or not the senrice requirements and character of the track remain
constant, for example, a route may a t one time require a few stops
widely separated and on it high maintained maximum speeds may
be permitted or necessitated; later this same route may require
many stops close together, thus permitting only low maximum
speeds. A gear ratio satisfactory in the first case would be
unsatisfactory in the second. Frequently a car is required to give
both city and suburban or interurban service or a combination
of the three. I t is in such cases that unsuitable gear ratios are
probably most often used-the common error being to use a gear
ratio lower than is necessary. This is brought about by not giving
the requirements of city service due consideration while seeking
high speeds. Unless field control be resorted to, a compromise
based upon the different classes of service required must be made.
Where i t is possible to use field control the low speed service will
largely determine the proper gear ratio. The proper gear ratio is
the one with which, other things remaining constant, the required
service is afforded a t a minimum cost for energy and maintenance.
As above noted, this will be determined by the conditions to be met,
but i n general the proper gear ratio i s the highest with which the
necessary schedule can be maintained (with allowance for low voltage and making up lost time) without establishing a dangerously
high armature speed on descending grades. Various authorities
have estimated that by the installation of proper gearing in existing
equipments from 5 to 1 5 per cent of the energy now used for traction may be saved. This saving is accompanied by a decrease in
maintenance costs and the schedules are not impaired by the change.
The following table from tests on the Brooklyn Rapid Transit
System shows that by increasing the gear ratio a saving of 10 per
cent in energy consumption was secured and the original schedule
was nearly maintained even though the number of stops per mile
was increased.

ELECTRIC RAILWAY HANDBOOK

I
Average number of stops r mile.. . . . . . . . . . . .
Average number of slow-gwns per mile. . . . . .
Schedule speed. miles per hour.. . . . . . . . . . . .
Weight of car, tons.. . . . . . . . . . . . . . . . . . . . . . . . .
Average passenger load. pounds.. . . . . . . . . . . . . .
Weight of car and average load, tons.. .........
Seattng capacity . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Watt hours per car mile.. . . . . . . . . . . . . . . . . . . . . .
Watt hours per ton mile.. . . . . . . . . . . . . . . . . . . . .

Tests with
3.58 gears

Tests with
4.12 gears

7 z
7.9
8.a
16.21

3040
17.77
38
2672
150.4

Where stops are infrequent, as in suburban and interurban service, the possibility of maintaining comparatively high maximum
speeds for a considerablelength of time, together with the possibility
of long coasting periods, make rapid rates of acceleration and braking of minor importance and a comparatively low gear ratio may
be used to advantage. Where stops are frequent, as in city service,
there is little or no operation a t high maximum speeds, high rates
of acceleration and braking are necessary and there may be little
coasting. If a low gear ratio is to be used in such city service, a
larger motor must be employed than would be necessary with a
higher gear ratio and there will be a waste of energy. The use of a
too low gear ratio with motors which would be large enough if a
higher gear ratio were used causes the motors to operate in series or
on resistance during a too great part of the time. This is equivalent
to operating the motors on a voltage lower than normal. The
motors are overloaded, and, in some cases, burned out, rheostats and
controllers are overworked and, in some cases, generating stations
and substations are overloaded and the load factor is abnormally
low. All this is accompanied by a waste of energy.
Figs. 36 to 41 by N. W. Storer show some of the effects of
different gear ratios when applied to a certain motor equipment.
The example is a specific one in which it is desired to determine
the best gear ratio to use with equipment already in hand. I n the
preparation of these curves the following conditions were assumed:
Weight of car loaded.. . . . . . . . . . . . .
Number of motors.. ...............
Size of motors.. ...................
Line voltage.. ....................
Size of wheels.. . . . . . . . . . . . . . . . . .
Length of run.. ...................
Schedule speed.. ..................
Length of stop.. .................
Rate of braking . . . . . . . . . . . . . . .

40 tons
4
75 h.p.
500 volts
33 ~ n .
0.6 mile
ao miles per hour
l o seconds
I . 5 miles per hour per second.

F i g . 36, 37 and 38 show the results obtained with a constant

accelerating current of 138 amperes for all gear ratios. Figs. 39, 40
and 41 show the results obta~nedwhen accelerating a t a constant
rate of 1.5 miles per hour per second for all gear ratios. Fig. 36
shows the time required to run with power on, time for coasting and
the time for the maximum accelerating current; or, in other words,
the time for running on resistance. The time for the maximum
accelerating current varies from 37.5 seconds, with the 2 : I gear

SELECTION OF GEAR RATIO

263

ratio, t o 6.8 seconds with 4.5 :I gears. At the same time, the rate
of acceleration varies from 0.8 mile per hour per second to 2.1 miles
per hour per second. Fig. 37 shows that the average voltage of the
motor while power is on is only 310 volts with 2 : I gears, while it
rises to 480 volts with 4.5 : I gears. Motors are overloaded when

G e uh t l o

FIG.36.-Typical

performance curves of car with different gear ratiosand

constant accelerating current.

the voltage drops badly, and this is just as true when the average
voltage applied to the motor is lowered by changing the gear ratio.
This fact will be more readily appreciated when it is seen that the
average rate of taking current from the line is 1.30 amperes per
motor with the 2 : I gear and only 52 amperes with the 4.5 : I . The
root mean square current per motor is 87 amperes for the 2 : I gear
and 49 amperes for the 4.5 : I gear. The curves show average watts

264

ELECTRIC RAEWAY HANDBOOK

loss in the motor in copper and iron and the total electn'cd h s
They show that the motm losses will be a minimum with the maximum gear reduction. (The maximum gear reduction possible
for this schedule would not be commercial as there is absolutely no
margin for making the schedule, but it is shown simply to point

Gear Ratio

FIG.39.-Typical

performance curves of car with different gear ratios and

constant rate of acceleration.

out the fact that the greater the gear reduction for a given schedule
the lower will be the losses in the motor.) The curves in Fig. 38,
however, show that the energy consumption is a minimum a t
about 4 : I gear reduction. The increase in watt-hours per ton
mile between 4 : I and 4.5 :I is coincident with and due to the
increase in speed a t the time the brakes are applied, as shown i n

SELECTION OF GEAR RATIO

265

Fig. 36. The rise in watt-hours per ton mile with the z :I gear ratio
is due to the increased loss in resistance and motors as well as in the
brakes. The motor loss, it will be seen, is controlled almost

FIG.40.-Typical

performance curves of car with different g a r ratios and

constant rate of acceleration.

m
70

.2

60
50

r 40

f a

a
10
1.

1.5

'22

FIG.41.-Typical

2.5$

3:

Gear Rnrlo

$35

_.

45

performance curves of car wlth different gear ratios and

constant rate of acceleration.

entirely by the root mean square current, as the iron loss is small at
all times and remains close to 300 and 350 watts for all gear ratios.
Fig. 39 corresponds to Fig. 36, Fig. 40 to Fig. 37 and Fig. 41 to

266

ELECTRIC RAILWAY HANDBOOK

Fig. 38. A comparison of these curves indicates that while a

constant rate of acceleration produces a nearly constant energy
consumption per ton mile, the motors will be very much more
overloaded with the lower gear ratios a t the conslant rate of
acceleralim than when accelerating a t the cmsfant current.
Commutating Poles. I n the railway motor the commutating
pole (also called the interpole) is a small pole piece placed between
the main poles with its wndings in series with the armature. The
total weight of copper in a commutating pole motor is nearly the
same as that in the non-commutating pole motor because the introduction of commutating poles makes possible a reduction in the
number of field turns. The function of the interpole is to improve
commutation a t all loads and voltages on the motor. I t accomplishes its purpose by providing a field which when swept through
by the armature coil
&.I
short-circuited b y
the brush will pmduce a complete current reversal in that
armature coil so that
there w i l l b e n o
spark when contact
is broken between
the commutator bar
I and the brush which
it is leaving. Incidentally the commutating r1.e .prevents or iminishes
the possibility of
flash overs by restricting rushes of
FIG.42.-Circuits of typical commutating pole current. The
motor.
mutating pole is permanently connected directly in series with the armature, and whenever, as for the purpose of changing the direction of rotation, the
direction of the current to the armature is changed the direction of the
current in the commutating pole winding must be changed. The
current in the commutating pole winding varies with that in the
armature, consequently, if the motor operates successfully a t one
load i t will operate successfully a t any other load up to the point
of saturation of the commutating pole. The commutating pole
operates most satisfactorily a t below saturation. At loads considerably above that whose current causes saturation of the
commutating pole the influence of the commutating pole is
annulled by the flux due to the current in the armature and with a
still further increase of load commutation would be worse than
with no commutating-pole.
Commutating poles make possible a reduction in the air gap,
with a consequent reduction in the weight of the motor, and raise
the limiting voltage between segments to a considerably higher
value, thus making possible the design of motors for operation on

FIELD CONTROL MOTORS

higher voltages. Further, the inductance of the commutating pole
windings tends to reduce the number of flashovers by limiting the
flow of surge currents. The practical elimination of sparking results
in many operating advantages. The short-time ratlngs of motors
have been materially raised, thus permitting the use of higher rates of
acceleration. More rapid acceleration in turn results in decreased
energy consumption and in many cases in decreased mptor heating.
Commutator maintenance costs are greatly decreased, as reflected
by the fact that the sale of replacement commutators by manufacturers has practically ceased since the general introduction of
interpole motors. Brush life has been increased about 50 per
cent on the average as the result of the decreased flashing and
sparking, and the accompanying decrease in commutator losses has
reduced motor heating to some extent. Briefly summarized, the
introduction of the interpole has made available a lighter, cheaper,
more reliable type of motor. At the same time, the artial removal
of important limitations in design has made possibk the develo
ment of successful types of high-speed, high-voltage and fie6
control motors.
The circuits in a typical commutating pole motor are shown by
Fig. 42.
Field Control. The circuits of a typical field control motor are
illustrated by Fig. 43. The idea of securing additional efficient
mnning speed by varying the effective field turns on a railway motor
is an old one, but on account of commutation difficulties it did not
prove successful until the commutating pole was developed. The
field control motor has the field arranged in two parts which are
connected in series with a lead brought out from the point where
they are joined together, FM, Fig. 43. In starting, the motor
current passes through both parts of the field in series, called the
full field connection, setting up a very strong field and developing a
large torque with a relatively small current. When desired, one
portion of the field is cut out, leaving only the permanent or short
field in the circuit, so that a weaker field results and higher s ~ e d s
are secured. This arrangement tends to economy and flex] lllty
in that less starting current is required and the motor has two
efficient running connections instead of one. Field control is
very generally used on heavy multiple unit equipments, locomotives
and for the largest car motors. I t is not uncommon to cut out as
much as one-half the field coils for the maximum speed.
Any well-designed commutating pole railway motor may be
adapted for field control by a proper arrangement of its field windIngs. To get the full benefits, however, the gears must be properly
selected. For interurban work the benefits of field control may
be secured by the use of standard high-speed armatures with a
larger gear reduction than usual. In most cases. also, sufficient
space is available to permit the extra field winding to be used.
Special armatures for use with field control are necessary only for
cases where the slowest s eeds and the maximum gear ratio are
required. I t usually will
found that where a motor of a given
size is used in city service with the maximum gear reduction and the
usual series-parallel control a slower speed armature may be used

&

268

ELECTRIC RAnWAY HANDBOOK

with the same motor frame and will make the same schedule with
a lower energy consumption when field control is employed. The
motor will have a lower horse-power rating, but the current used
will be correspondingly less, and, consequently, the motor will
have no more loss in it than with the motor of higher speed with
a larger rating. I n other words, the use of field control permits
Main Pole Field Coil Connections
Pinion End

Cornmutating Pole Field Coil Connection

Pinion End

AF+

Dotted lines show connections at pinion end

PIG. 4J.-Ci~itsof typical field control motor.

RAILWAY MOTOF VENTILATION

269

the use of a motor of a smaller r~tingfor a given service. Where

the maximum gear ratio is used in both cases, the same size of
frame must be used. However, where the gear reduction can be
increased for the field control motor it will frequently be found
that a smaller size of motor can be used a t a lower first cost and
with less weight to be carried around. A double saving will thus
be effected.
Ventilation of Railway Motors. The losses in a motor, which
appear in the form of heat, result in a motor temperature that is a
function of the magnitude of those losses, conditions of ventilation
and (in some cases) the weight of the motor. On loads essentially
constant over a time period of several hours, the weight of the
motor has little or no effect upon the final motor temperature.
Under these conditions, the temperature rise is almost entirely

FIG.44.-Internal

ventilationamature without fan-radial

No air inlet in frame.

vent ducts.

dependent upon the conditions of ventilation. On short time loads,

the temperatures a t any given time will depend upon both the
ventilation and weight of motor per watt loss. The greater the overload the greater will be the influence of the weight of the
motor in limiting the temperatures; since under these conditions
a great majority of the losses must be stored up in the copper and
iron of the motor. Also under these overload condit~ons, the
maximum temperature reached will depend upon the initial
temperature. A ventilated motor will naturally have a greater
continuous capacity than an enclosed one of the same size; or
conversely, of two equally rated motors, the better ventilated one
will weigh the less. Hence, for railway work, where the minimum
motor weight is desirable, the ventilated motor is a logical develop
ment. With regard to ventilation, there are three general classes
of railway motors, namely: totally enclosed, self-ventilated, separately-ventilated.
A totally enclosed motor is one so enclosed as to prevent circulation of air between the inside and outside of the case or frame,
but not sufficiently to be termed air-tight. Such a motor is sometimes called a non-ventilated motor, since all of the heat losses must
finally be dissipated from the external frame surfaces by radiation

270

ELECTRIC RALLWAY HANDBOOK

and convection air currents. A totally enclosed motor may,

however, have a system of internal ventilation such as shown in
Figs. 44 or 45. This is highly desirable since such a system of
ventilation will greatly aid in the transfer of the heat losses from
the armature to the frame and housing surfaces.

FIG. 4s.-Internal

ventilation-armature
with fan-longitudinal
d u c t s a i r inlet and outlets closed.

vent

A self-ventilated motor is one in which the ventilating air is

circulated through the machine by a fan, blower or centrifugal
device integral with the machine. Two common systems of selfventilated motors are shown in Figs. 46 and 47. I n both schemes,
the fan mounted on the inion end of the armature acts as a n exhaust
fan and pulls the air tirough the intake and ventilating passages

FIG. 46.-Series ventilation-armature with fan-longitudinal vent ductsair inlet at pinion end, top side of frame.

and discharges i t through the outlet openings in the frame. The

parallel or multiple system of ventilation as shown in Fig. 47 is
the better system since the air paths are shorter, and due to greater
cross-sectional area for the air flow more air will be pulled through
than is possible with the series system of ventilation as shown in
Fig. 46. Since the losses in the armature range from 60 to 70

RAILWAY MOTOR VENTILATION

271

r cent of the total, it is desirable to have the armature well ventig e d with the air entering i t as cool as ssible, and this condition is
best met by the parallel ventilation as s c w n in Fig. 47. On motors
of this type the entrance of foreign material, such as dirt, water or
snow, should be prevented as far as feasible. This is usually done
by some type of a baffle intake, one form of which is shown in Fig.
47. The entering air is forced to change its direction very abruptly
and on so doing the heavier foreign particles are dropped. The
parallel system of ventihtion as shown in Fig. 47 will Increase the
continuous rating of a motor 50 per cent or more over its totally
enclosed rating. A self-ventilated motor may be made totally enclosed by changing the intake and outlet covers to solid coven,
as shown in Fig. 45. I n this case the fan is beneficial in providing
internal ventilation. This practice is adopted by some of the
electric railway systems in the northern part of the country for

FIG. 47.-Parallel

ducts-air

ventilation-armature
with fan-longitudinnl
inlet at commutator end under side of frame.

vent

winter operation, where snow and ice troubles are especially severe,
but it should be done only when the air temperatures are low, or
the motor may be overheated.
A separately-ventilated motor is one in which its ventilating
air is supplied by a n independent fan or blower external to the
machine. With a self-ventilated motor, the amount of air which
will be pulled through the motor depends upon the size 05 the tan
and the speed of the motor; on slow speed the air circulated will
be very limited, and when the motor is stationary there will be no
air circulation outside of small natural convection currents. On
account of the above limitations, a separate blower is used where a
maximum rating is desired. Thus, separately-ventilated motors
are being used on the majority of locomotive motors. The air
supply to the motors is independent of the motor speed, and since
the air is usually taken from the locomotive cab, it is considerably
cleaner than air picked up near the road bed as with self-ventilated
motors. Where the motors are axle-hung the air connection between
the motor and the air conduit must be a flexible one, a common type
of connection being the canvas "accordion" connection. The
air intake to the motor ran he provided by one or more holes a t

272

ELECTRIC RAILWAY HANDBOOK

either end, although the more usual sition is at the commutator

end, since more room is available t e r e . The air paths through
the motor ducts are usually similar to those shown in Fig. 47;
there being a parallel air flow through both the armature and the
fields.
A recent development in connection with the ventilation of nonventilated type railway motors has been made in which an outside
fan is mounted on an extension of the armature shaft a t the commutator end of the motor. Air is drawn in a t the pinion end of the
motor, assed through and over the armature and between the
field coiE, over the commutator, then through openings in the specially designed commutator end housing, and finallv is exhausted
by the external fan mounted on the shaft extension: Some of the
results in connection with this method of revamping a non-ventilated type of motor a s worked out on two 6c-h.p. motors were as
follows: continuous rating of motors increased 30 to 40 per cent;
one hour rating of motors increased 7 per cent; from 38 to 40 per cent
reduction in tem ratures observed during a service test. Cases
are cited where
the addition of this ventilating device motor
cars have been able to handle trailers during msh hours, which
was not possible otherwise on account of excessive heating.

FIG.48.-Section of an assembled commutator.

Commutator Material and Construction. Hard-drawn copper

is recommended for commutators as possessing greater uniformity
in size and hardness over the forged bars and corres nding superiority in life and service. Attention is also called to tEimportance
of material and workmanship in the construction of commutators
being such as to insure a solid structure which will not shrink,
become loose, or get out of true. Built-up mica is preferable for
commutator segments, but it is very important that the building-up
should be even and compact. Assembled commutators should be
baked a t 230 deg. C. and compressed while hot to insure solidity,
damps being tightened before pressure is released.
Care of Commutators. The baud over the front V-ring of a
commutator should be wiped off every month. After cleaning,

COMMUTATORS
painting with an air drying varnish makes cleaning easier the next
time. Painting every six months is advisable. Flat spots, high or
low ban, ridges, burned spots, etc., should be smoothed up. Where
these are not bad, the motor need not be removed from the car.
A tool can be made by mounting a block of wood on a stick, one
face of the block being cut to the radius of the commutator and
lined with sand paper or stone. As the car is run by other motors,
this tool, held against the commutator, will smooth the rough spots.
The armature should be removed from the car and placed in a lathe
for turning the commutator, if its face is very rough. Holes left
by defective mica or pits in the side of bars can be tilled with commutator cement supplied by dealers.
Undercutting. A clean undercut commutator prevents high
mica, reduces the burning of the commutator bars, increases the
life of the brushes, and practically eliminates flashing. This results
in greatly reduced maintenance cost. The object of undercutting
commutators is to clean out the mica between the copper segments
from the face of the commutator to a depth of 964 inch. The completed armature, after being rewound, soldered and banded, has the
commutator trued up, and all excess solder removed from the
neck and face, and is then undercut. If undercutting were done
before soldering, the excesq solder might till up the grooves and
short-circuit the commutator bars.
The most efficient machines for undercutting are special lathes
with motor-driven or beltdriven circular saws, clamped on an arbor
which is mounted on a head that moves on slide rails. By means of
a hand-operated lever or a foot pedal controiled by the operator,
the revolving saw is carried over the face of the commutator. The
head carrying the arbor is fitted with an adjusting screw to adapt
the height of the saw to commutators of different diameter. An
air hose and hood is sometimes provided to carry off the mica
dust. A less expensive uipment consists of a n air or power-driven
(through a flexlble s h a z circular saw, which is guided over the
commutator face by hand. The armature is mounted so that i t
can be rotated readily between the centers of the special lathe, or
on two horses if the hand-guided saws are used. The circular saws
are revolved a t approximately 2000 r.p.m., and are drawn toward
the operator so that he can guide the cut. The cutting edge of the
saw should revolve in a direction toward the operator. After
slotting is completed, the face of the commutator should be thoroughly polished and cleaned of all particles of copper by means of
emery cloth. The special lathe can be so equipped that this polishing is done without removing the armature from thelathecenters.
I t is essential that all particles of mica be removed from between
the segments; thus i t 1s advisable to use a saw about 0.005 inch
larger than the thickness of the mica to be undercut. A small
diameter saw must be used in order to cut the slot to the proper
depth and a t the same time not cut into the neck of the commutator.
After the slots are sawed, it is sometimes necessary to go over them
with a small hand cutter to remove aU remaining particles of mica.
An undercut commutator should operate from one heavy inspection period to another. However, if an armature is removed from

ELECTRIC RALLWAY HANDBOOK

the motor frame for any other repairs, the commutator should be
carefully inspected, and if the mica is getting flush with the copper
it should be re-undercut. The commutator should be re-undercut
before it is worn flush, since the groove left will guide the saw and
make the work much easier. Where commutation trouble is
frequent, it is good practice to use a V-shaped hand tool to round the
edges of the under-cut grooves between bars to about
inch
radius. This can be done with the motors in the car. All particles
of mica, copper, or dirt should be removed from the grooves after
undercutting.
Commutator Repairs. A thin drift, driven in the top of the slot
in the neck toward the shaft between the side of the commutator
neck and the top filling piece, will loosen the filling piece so that i t
may be forced out by a gouge. Similarly, windings may be taken
out of the first bar. With one set of leads removed, there is enough
space left to bend one side of the next neck, thus permitting the
removal of the leads by means of the gouge. This can be done
without heating the soldered joints. Where a small number of
bars are to be replaced, it is not necessary to remove all the leads
from the armature neck or take the commutator from the shaft.
Stand the armature on end, commutator up. Mark each separate
piece so that it may be put back in its old position. After removing
the ring nut or bolts, take out the metal V-ring and mica V-ring.
If the bars are tight, tapping with a wooden or rawhide mallet will
loosen them. The new bar must be filed to the exact shape or
thickness of the one which it replaces to prevent the new bar or the
old bars next to it from becoming loose. Clean out the space where
the new bars fit and tap them in with a soft mallet
Detached parts of the commutator must be kept clean and dry.
When the parts are ready for rebuilding, sand paper the mica V-ring.
Clean the metal V-ring and the V in the commutator bars, and paint
the V in the bars with shellac. The shellac and brush must be
absolutely free from all dirt or moisture. After putting the mica
and the metal V-rings back in their original positions, the ring
nut or bolts should be drawn u p fairly tight. Painting the threads
of the ring nut or bolts with very thin white lead will make it easier
to remove these parts next time. The commutator should now
be heated to I I O deg. C. in a n oven where the air is dry, and the
ring nut or bolts drawn up tight while the commutator is hot. The
commutator should then be turned in a lathe. Test after these
repairs should be a t I I O volts alternatiag current to ground.
When i t is necessary to replace the rear mica V-ring, a number of
segments, or a complete set of segments, it is advisahle to remove
the commutator from the shaft. The method of taking down and
rebuilding is the same as before described, care being taken to heat
the commutator thoroughly to soften up the new mica parts, so
that the ring nut or bolts can be drawn up tight. T o this end,
when the commutator is assembled, i t should be put in a n oven and
heated to a temperature of 1 2 5 to 140 deg. C. While at this temperature the commutator should be placed in a press, using a pressure of 2 0 to 21; tons for a 50 hp. motor, and the ring nut should be
drawn up tight while under pressure. Complete sets of segments
are shipped, temporarily handed together, with the mica and copper

BRUSHHOLDERS

275

segments in their proper position. The complete set should be

assembled in the commutator as a unit, and the temporary band
should be removed just before the commutator is finally tightened.
Test after these repairs should be a t 300 volts alternating current
between commutator bars and 2 0 0 0 volts alternating- current to
ground.
Brushholders. The brushholder castings should have a t least
j/, inch air clearance from the motor frame. Railway motor frames
have machined pads for the clamping blocks to hold the brushholders
exactly in the neutral position. On replacing brushholders the
clamping blocks must be seated properly. The carbon box should
be parallel with the commutator bars, and the distance from the
center line of one box to the center line of the other box on a fourpole machine should be equal to one-fourth of the number of commutator bars. The coil short-circuited by the brush should be
midway between the main poles; ordinarily this means that the
brush will be o posite the mlddle of the main poles. This spacing
is adjusted a t t i e factory and will be correct if the clamping blocks
are seated properly and the brushholders are those belonging to
the motor and are not bent. The underside of the carbon box
should be kept within % to %. inch ( referably 36 inch) of the
commutator surface, so as to prevent Pbreakage of carbons. A
piece of tapered fiber % to Jig inch thick makes a convenient gage.
The nut whlch holds the clamping block should be carefully tightened
so that the brushholder will not come loose.
Spring pressure should be about 4% Ib. per sq. in. of contact
surface for large motors of say 150 hp., up to 7% Ib. per sq. in. for
small motors of 25 to 30 hp. Factors that may modify this
pressure are the amount of sparking, the play in the armature
bearing and the track conditions. T o prevent breakage and side
wear of the carbon, it should fit in the box as snugly as possible
without binding. This clearance should not exceed %. inch.
The shunt formerly used on the carbon has been replaced by a
shunt on the brushholder which connects the contact tip on the finger
and the main casting. This shunt must be large enough to carry
the current and make a good electrical contact where fastened.
If a good electrical path is furnished for the current through the
shunt, much of the side wear of the carbons and burning of the
carbon box is prevented. I t is therefore very important to keep
the shunts in good condition. The object of the contact tip is not
only to transmit pressure to the top of the carbon, but to conduct
current between the shunt and the carbon. I t s shape should be
such that i t will do the least damage to the top of the carbon and
provide a large contact area to carry the current. Cable leads
should be supplied with sleeves or terminals. These connections
should be kept tight to secure a good electrical as well as a mechanical contact. Set screws with jam nuts and lock washers should be
used. When carbon dust and dirt get into the working parts of
a brushholder, i t is likely to work stiffly. These parts should,
therefore, be kept clean and occasionally lubricated. If neglected,
they may become tight ant1 not allow enough pressure, or I~ecome
so badly worn that the parts get out of line and the tips press on the
carbon box instead of on the carbon. The keeping of railway

ELECTRIC RAILWAY HANDBOOK

brushholders in good repair and proper alinement is a large factor
in cutting down maintenance costs, as i t lengthens the Iife of the
carbons, commutators and motor windings by improving commutation, thus tending to prevent flashing between brushholders and
ground. For this reason i t is very important that brushholders
be regularly inspected and overhauled.
Carbon Brushes. When making a selection, the following few
fundamentals, in connection with operating conditions, should be
considered: commutation; life of commutators; life of brushes;
frequency of inspection; current density; cost per car mile. The
above factors will vary more or less with the design of the motor,
design of brushholder, correct spacing of brushes, condition of roadbed, condition of equipment, condition of armature bearing, condition of commutator surface, brush tension, weight of car, number of
motors per car, schedule speed, service conditions, etc. Since all
of these factors must be taken into consideration, the best and most
reliable results are obtained by making tests of recommended
grades in service under actual operating conditions. Carbon manufacturers generally are willing to assist the operator inmaking
service tests. The following general information regarding carbons
will help the operating man more intelligently to select carbons for
service tests.
Carbon is a non-metallic element found in both crystalline and
amorphous form. Natural graphite is carbon in a crystalline form,
and IS mined in many localities. Amorphous or non-crystalline
carbon may be obtained in the form of coke or lamp-black. Artificial graphite is obtained by heating amorphous or non-crystalline
carbon, such as coke, in a n electric furnace to change its structure
to a crystalline state. By the use of these materials, the following
general classes of brushes are made: carbon brushes, made of
crushed coke and binder; graphitized brushes, made of carbon and
then electrographitized; graphite brushes, natural graphite and
binder; metal graphite brushes, natural graphite with metal powder
and binder. With modem commutating-pole motors having commutators undercut, and considering costs per car mile, brush end
wear side wear, breakage, life, commutation, life and maintenance
of commutator, etc., these classes of brushes are best suited for
railway motor service as follows: graphitized brushes, best allround results; carbon brushes, next best; graphite brushes, very
special and limited uses; metal graphite brushes, not used. I n the
process of manufacture the most important operations, depending
upon the class of brush being manufactured, are as follows. crushing,
carbonizing (if it is done) and cooling; milling, mixing, cooling and
re-milling; moulding and packing in the furnace; gas baking;
electric baking or graphitizing. The two most common general
methods of manufacture are: (1) extruded or squirted-where the
material in the form of pulp is forced through a metal die under
pressure and then cut off to the desired length and baked a t a high
temperature to carbonize the bond and permanently set the
material; this method is used in making the cheaper grades of
carbons which do not have the strength to resist breaking and chipping in service; ( 2 ) moulded and machined-where the material is
moulded under heavy pressure into blocks, and then baked; the

CARBON BRUSHES
carbons are cut from these blocks and machined to exact size;
this method is used in making high grade carbons and gives a
brush of uniform texture and strength that is best suited for railway
work. I n the manufacture of various gradesof carbon brushes best
suited to meet the requirements of operating conditions, the following characteristics of the brush must be considered: contact drop;
coe5cient of friction; heat conduction; hardness; apparent density;
conductivity; resistance; abrasiveness; toughness.
For use in modern brushholders, the carbons should be not over
two inches long. That is, when new, they should not extend over
the top of the carbon box, for the following reasons: if longer, they
are subjected to a greater side pressure, due to the action of the
contact tip, which increases the side wear, tends to bind the carbon
in the box and reduce the direct pressure on the surface of the
commutator; if longer, they are discarded due to excessive side
wear before the added length can be used up in end wear; approximately the same mileage can be secured from the shorter carbon
and since the carbons are bought on a cubic inch basis, the cost is
less. The width is not so important. The brush can have as
inch clearance in the box without causing any trouble
much as
in service. Thickness is very important, the initial clearance
betwen carbon and carbon box should be approximately from
o 006 t 3 0.008 inch. If i t is much less, the carbon will tend to
stick in the box and bind, and if greater, i t will soon rattle in the
box, wearing away the side; i t will also tend to chip and break,
thus reducing its life.
When shunts or pig tails were used on brushes, i t was considered
necessary to copper-plate the carbon to provide a good electrical
contact for the shunt connection. With the present design of
brushholder having a heavy braided copper shunt from contact
tip to carbon box, shunted carbons have been discontinued, so that
copper plating is unnecessary; in fact, i t is objectionable on the
higher grades of carbons as i t tends to peel off in service and bind
the carbon in the box. For railway work, shunts or pig tails on
carbons have been practically discontinued, for the following
reasons: first cost, they were used with the cheaper grades of carhons
that had a comparatively short life, which required renewal of
carbons a t frequent intervals; inspection, the pit-men could not be
depended upon to maintain shunts during inspection; not reliable,
the design was such that shunts became loose and disconnected.
With these conditions greatly improved by the use of a higher
grade of carbon with longer life and requiring less frequent inspection, and by improved methods of fastening pig tails to carbons,
some advantages can be obtained by the use of pig tails; and in
certain specific cases, especially for very heavy current densities,
they have been adopted with a saving in maintenance. Foreign
practice tends to the more extensive use of shunts than is customary
in this country.
Flashing of Motors. Flashing of railway motors, also commonly
known as bucking or blowing, is primarily caused by poor commutation which results in a sudden breakdown of the insulation
over the face of the commutator from the brushholder to the
motor frame or ground. As a result, there is a sudden rush of

278

E1,ECTRIC RAILWAY HANDBOOK

heavy current which either opens the circuit breaker or hangs as

an arc, and badly burns the shortclrcluted parts. The results of
flashing are vaned. Somet~mesthe motor is w badly damaged
that i t is inoperative. I n thls case, the motor must be overhauled,
armature mndlngs repaired, commutator cleaned up, brushholders
and wiring around frame put in good condition. When the short
circuit is immediately cleared, the motor generally can be continued
In service to finish its run, but should be reported as defective, and
given a careful inspection for any damaged parts that need attention.
A detailed lavout of the various conditions that affect commutation
of a railway motor and, in turn, may cause flashing, are as below:

CVNDITIVNS
THATTLNDTO PRODUCEFLASHING
( Yon-commutmg pole type

Dcstgn

Sensrtlve neutral
H:gh voltage between commutator bars
H ~ g hbars
Rough face
Low bars
Loose bars
Flat spots
Poor undercutt~ngof mrca
Commutators
Sharp edges on commutator
bars
Poor surface condrtlon
a
Sharp corners on commutator
bars
D~rtvcommutators
B;Xe";pnng
Weak spnng pressure
Lack of spnng tenslon
Pressure finger sttcklng
Worn mechan~sm
Too far from commutator
I
I surface
Incorrect spacing between
Brushholders I Incorrect settlng
brushholders
Out of alrnement wtth pole
Worn carbon box
Loose In clamplng block
Too small rlearances to ground
Infenor grades
Too long
Length
Too short
Carhons
Loose In carbon box
Clearance
T ~ g h rn
t carbon box

I
/

Wrnd~ngs

Loose brushholder connection

Wrong connectron

( Short-clrcu~tedcot1

{ F ~ e l dcontrol-wrong

Commutat~ng-palefield

I
1

connectlon
nesrgn of corl bobbln or case
Short-clrculted corl
Reversed coll
( Faulty control
Rapld acceleration

Sudden voltage changes

Operatton

Reversed cod

Maln field

High speed runnrng

Hlgh trolley voltage
O>erheatrng of motors
Loose or worn beanngs.

Rough track
Flat wheels

Heavy operation
Breaks In thlrd rarl

;zz $::

~:nd,2;:

MOTOR MAINTENANCE

279

Removal of Armature from Box Frame Motor. Several methods

of removlng armatures from box frame motors are in common use.
A scheme used by a large operating company requires one short
and one long piece of pipe, bent for convenience. Two rings that
slip over the ends of the armature shafts are hinged a t one end of
each piece of i e, about six inches apart. The motor is set on the
floor and b o t i X)ousings removed. The rings on the long piece of
pi,y are slipped over the commutator end of the armature shaft,
e rings on the short piece of pipe are slipped over the pinion
w ~ l the
end of the shaft. The men a t the end of each piece of pipe lift the
armature bodily out of the frame through the pinion end opening,
and let it down on a grooved block made to receive it. The number
of men required to handle
a n armature by this method
depends upon its size and
weight.
A machine designed for removing armatures consists of
two heads fitted with centers-one
long, which is
stationary, and the other
short. which is adiustablemounted on the *end of a
base plate. The long center
must have a diameter slightl y smaller than journal size a t
the commutator end of the
armature, and should be long
enough to extend through the
.--- -.---length of the motor frame.
A carriage to receive the
motor travels on the base
plate between the two heads.
The motor is set on the car- FIG 49 --Yoke for armature removal.
riage of the machine, with
the commutator end toward the machine head with the long center
pln. The pinion end housing bolts are removed and the center
points are entered in the centers of the armature shaft and locked.
The pinion end housing is then removed and the carriare supporting the frame is pushed toward the machine head carrying the long
center pin. The armature being held rigid, the long center pin
passes through the commutator end bearing opening. In this
manner the frame is shifted far enough to clear the armature,
which can be lifted to the floor after being released from the center
plns. One man and a crane operator are required to handle an
armature by this method.
A very quick and simple way of doing this work, but one which
requires a careful and skilled crane operator, is as follows: The oil
from the bearlng housing oil wells must first be removed, after
which the motor is upended, commutator end down. The pinion
end housing bolts are removed, and a special threaded cap fitted
with an eye bolt is screwed on the pinion nut fit of the armature

iEz?

280

ELECTRIC RAILWAY HANDBOOK

~ , the armashaft. The crane hook is attached to'the eye b o ~and

ture is lifted out of the motor frame. When the armature is
being let down on the floor, it is necessary to rest the commutator
end of the shaft on a block to prevent damaging the commutator
and end windings as it turns to a horizontal position. With this
method one man and a crane operator can handle the largest motors
used on city or interurban cars. In Fig. 49 is shown a yoke that is
attached between the inion and housing, replacing the special
threaded cap. This scieme works well; the yoke is easily made
and can be used on any size motor.
The facilities and number of men available will, to a very great
extent, be the deciding factor as to which method will be best
suited to individual cases.
Armature Winding. Great care should be used to guard against
failure by ground, short circuit, open circuit, or loose bands. Sha
corners and roughness in the slots that might damage the c o x
should be tiled down and all chips and tilings removed before applying the insulating material to the core. In applying the insulating
material on the coil supports, the material should be evenly placed,
and no thin spots allowed. The coils should fit tightly in the slot,
and so that the top of the coil is about Hx inch above the band
groove. Fillers should be used between the coils if necessary to
meet these conditions, which will permit the bands to pull the coils
down tightly and a t the same time finally rest on the iron, which
allows the minimum movement of the coils in the slots. If the coils
are wound too high, the bands will not rest on the iron, and when
the insulation dries out in service, loose bands will result. The
coils should not be twisted, bent or abused any more than b
absolutely necessary to get them in place. Care should be taken
not to get the wires or leads crossed in such a manner that, when
pressure is applied in banding, short circuits will occur. The coils
on the end windings should be down, so as to make a solid foundation for the bands, but pounding should not be carried to the extent
that the coils or leads are damaged. Insulating protecting pieces
should be placed at all points where the coils cross, and where there
is danger of short circuit. I t is good practice to weave braid
between the Ieads directly back of the commutator, both on the
top and bottom layers. The leads should be tinned back to such
a distance that there is no untinned copper in the commutator neck.
The cotton sleeving on the leads should not be allowed to get into
the commutator slot, as it may hinder soldering to such an extent
that a poor connection will result. The tool used in driving the
leads into place should be free from sha comers that might nick
the Ieads, as this may result in a brokenTead.
Layout of Armatures and Commutators. Most direct current
windings used in railway work are of the wave or two-circuit type.
In Laying out such a winding it is necessary to know the throw of
the coils and the throw of the leads. The throw of a coil is the
distance spanned on the core designated in terms of the slots, that
is, if the coil throw is stated as r and 8, the bottom of the coil lies
in slot No. I, while the top of the coil is in slot No. 8. In a similar
manner, the throw of the leads is the &stance spanned on the com-

281

ARMATURE WINDINGS

mutator in terms of the commutator. When there is a n odd number of leads per coil, or if, when a dead coil is taken care of, an odd
number of leads remain, locate the center between the slots indicated by the coil throw. If the lead throw is a n odd number of
bars, this center will line up on a commutator bar, and if it is an

"

i
FIG. so.-Layout

"

of armature winding w ~ t han odd coil throw and an odd

throw of leads.

even number of bars, i t will line up on the mica between bars. This
bar or mica is the starting point for laying off the commutator.
If there is a n odd number of bars in the throw, take one less than
the number of bars and count off half of this number each direction

FIG. 51.-Layout

of armature winding with an even coil throw and an even

throw of leads.

from the starting bar, and this will give the h t and last bar of the
commutator throw. If there is an even number of bars in the
thfow, count off half the number in each direction from the starting
mlca. As the first coil put down will have an odd number of leads,

282

ELECTRIC RAILWAY HANDBOOK

the center one of the top and bottom leads should be placed in the
first and last bar of the throw as determined. Sample layouts are
shown in Figs. 50 and 51. A slightly different method is necessary
when there is an odd number of coils per slot and an even number of
slots, which case, however, very seldom occurs. If the lead throw
is an odd number of bars, the center, as indicated by the coil throw,
will line up on the mica and, if an even number of bars, it will line
up on a bar. If there is an odd number of bars in the throw, take
one less than the number of bars and count off half this number to
the left and one more than half t o the right, and this wiU give first
and last bar of the commutator throw. If there is an even number
of bars in the throw, count off half the throw to the right and one
less than half to the left. If there are two leads in the first coil,
No. I lead should lie in No. I bar, and if there are four leads in the
first coil No. 2 lead should lie in No. I bar.
The alinement of commutator bar and mica with armature tooth
and slot center lines is as follows: coil throw even, lead throw even
-mica with tooth; coil throw odd, lead throw odd-bar with slot;
coil throw even, lead throw odd-bar with tooth; coil throw odd,
lead throw even-mica with slot.
These rules are of value: the wave or two-circuit winding always
requires an odd number of commutator bars; when a n even number
of coils per slot are used, there will always be a n idle or dead lead;
when a n odd number of coiIs per slot are used and there are a n
even number of slots on the armature, there will always be a n idle
or a dead lead; when an odd number of coils per slot are used and
there are a n odd number of slots on the armature, there will never
be an idle or a dead lead.
Soldering Leads to CommuCator. For armatures operating
under normal service conditions, and not subjected to high temperature and unusual mechanical strains due to high speed, which
tend to throw solder from the commutator necks and armature
bands, half-and-half solder can be used with good results. When
motor equipments are subjected to high temperatures and high
speed, pure tin should be used to solder the armature leads to the
commutator necks, and to solder the armature bands. When
tin is used for soldering, it is necessary to have the clearance between
parts to be soldered as small as possible. I t is important that
the flux does not contain any acid, as the acid may get to the insulation of the coils and cause short circuits and grounds. A gpod,
cheap and safe flux is made by m i ~ i n gone pound of rosin in a
quart of either wood or denatured alcohol.
Banding Armatures. Since the coil insulation shrinks when
heated, it is necessary to shrink it as much as possible before the
final handing wire is applied. This is done by heating the whole
armature to about 75 deg. C., when the insulation becomes pliable
and can be prcssctl down into permanent shape. The hot armature
is thcn mounted in a lathe, protecting strips of cloth are placed over
the ctid windings, a temporary banding wire is wound over the
coils with enough tension to draw them down into place, and the
ends are fastened by soldering tin clips over the wire. The armature is then allowed to cool. After the temporary wires are

ARMATURE BANDING

283

removed, the armature is ready for the permanent banding. When

a banding machine is used, the tension in the wire is regulated by
passing i t over a train of friction pulleys mounted on the carriage.
The friction of the pulleys can be adjusted to any desired value by
the regulating screws. I n the absence of such a device, fair results
can be obtained by passing the wire two or three times around a
round wooden banding stick approximately two inches in diameter
and adjusting the tension by hand. When core bands are used, the
grooves are fitted with thin stripsof tin, which protect the coils from
the cutting action of the bands. I n starting the permanent banding,
a few turns are first made a t one end to secure the necessary tension.
All the banding groups are wound on continuously, to eliminate the
necessity of fastening the ends of each group as they are wound.
The bands are held together and the ends are fastened by means of
narrow tin strips placed under the wires, bent back over the top
and held by ure tin solder. These strips are inserted while the
wire is being
on and are located about every three inches around
the armature, with closer spacing a t the beginning and end of each
band. For the core bands, these strips are placed in the slots and,
being wider than the groove, they prevent any tendency for the
bands to slide around the armature. The ends of the band wire
groups are then cut and secured by being bent back outside one
clip and inside the next one. Pure tin solder is applied to the whole
surface of the bands, forming a solid web. The end windings are
secured by groups of wire which are wound on insulating hoods to
protect the coils. On the commutator end, strips of thin mica with
overlapping ends are usually placed on the commutator neck and
held in place with a few turns of twine. The drilling hood or head
is wrapped around the neck, extending about a n inch from the edge
and placed inside out. After this end is secured with twine, the
free end of the hood is pulled back over the windings, bringing the
outside of the hood a t the surface and making a neat folded-under
edge. This hood is also held temporarily with twine unti\ the wire
is applied. The other end of the armature is similarly covered
with a hood and banded.
The proper tension varies with the size of wire and the construction of the end windings. When the end coils have no rigid support
and extend out considerable distance from the core, the tension is
gradually reduced, as shown below:
POUNDSTENSIONFOR BANDINGWIRES

&

Drameter of wire.
rnches

Core banrls

End bandr

--

A t core

The best banding material is a high-grade steel piano wire,

having a final breaking strength of 200,000 pounds per square inch.
For temporary bands a cheaper grade can be used. Pure tin
solder should be used, as this gives a band that will hold together

284

ELECTRIC RAILWAY HANDBOOK

T i n clips and s t r i p s
f o r a longer t i m e t h a n half-and-half solder
should be of commercial s h e e t t i n a b o u t o 0x2 i n c h thick.
Broken Armature Leads. Some c o m p a n i e s h a v e experienced
a g r e a t d e a l of trouble f r o m t h i s source, i n s o m e cases c o n t ~ n u a la, n d
in o t h e r s sporadic T h e table below l ~ s t st h e p r o b a b l e c a u s e s
of s u c h trouble.

PROBABLE
CAUSESOF BROKENLEADS

f Poor workmanshrn

Loose armature wrndlngs

Loose armature core

Loose commutator.

' Internal

Poor bandlng
Loose cod support

Excessive current
Out of balance
Hrgh peripheral speed
Armature beanngs
Armature beanng housmgs

Worn key
Poor fit on shaft or s p ~ d e r
Poorly clammd
' Worn~key Poor fit on shaft or s p ~ d e r
Poorly clamped
Not thoroughly baked
' Not well soldered
Poor grade wlre
Too few bands
Worn key
Poor fit on shaft or s p ~ d e r
Overload

Rapid acceleration

Faulty handllng of car

Core unbalanced
Commutator unbalanced
Wlndlngs unbalanced

( Ez

'$2,"

Axle beanngs

External

Loose

Wofn

Gear case

Broken
Loose bolts

Axle caps

Broken
' Loose
bolts
' Badly worn

Gears and plnrons

'

Out of mesh
Eccentnc
Bottoming

Overloads

Leads damaged

' Speclal work

Track
Poor c o n d ~ t ~ o n s
Non-resll~entroadbed
R~gtdmountlng of motor
Faulty handllng of cars
Flat wheel
' Heavy currents.
Magnetic stresses
Shrinkage of ~ n s u l a t ~ o n
Loosening of band wlre
Overheating
.
Expansion and contractlon of wlndrng
Dnvlng into slot
Sharp comers on comLeads. n~cked
mutator slot
Actlon of heat In solderMetal changed
mg
Act~onof the solder
Metal crystallized by poundlng
Mechanically weak-no elastlcaty

.. .

Matenal (copper)

Locallzed stralns

Sharp bends In wlndlngs

Uneven planng of leads
Impro r bandlng
Poor E t n b u t r o n of band?

285

RAILWAY MOTOR FIELDS

Annature Record Tags. The use of record tags similar to that

shown by Fig. 52, of a standard size such as 5% X 2% inches,
preferably made of cloth, provides an accurate record of armature
troubles, armature repairs, location of armatures in the equipment,
and a means of analyzing armature and motor failures, it gives a
record of life of motor parts, and helps to weed out defective frames
and troubles due to incorrect winding and motor connections, requires little clerical work, and assists in figuring cost of repairs on
armatures. The suggested wording may be modified to suit
individual conditions.

H r p lhl. 1.; "Nh -tun

u t u rry.lndand put In e u
Tknrca,~ducd1.o.ol
tor r-d
uld Dl.

ARMARIBO TAQ
lWsri&DhWartb,
Eaur Pi.?m.n
and t y .rucfl.d

;1 EhETzzd

w o r t Dons

0 Q-

0 Eudsbtr.

grn-s~at.d

Bbdt

[II CE.

Dlppd d M d

D P.E.

k l y m

lbnrut.

b d u l o . ul P-r

Tested .rd. y p . o r r b - - - h t .
signed
b hr=m.a
ArmdCfr. Put I.

no.
N

Pkld-oC
WM.iq "land
-P

L
%

M nolPa
B

BUlbn
ht.

M&.

FIG. 52.-Suggested

armature tag.

Field Coils. I n general, the four coils have the same number of
turns on each. On a few machines, however, the to and bottom
coils have more turns than the two side coils, in orfer to shorten
up on the gear centers of the motor. On commutating pole
type railway motors, four commutating coils are most commonly used, in which each coil has the same number of turns.
With small size motors, owing to the mechanical limitations in
design, and to make use of some electrical advantages, the necessary commutating magnetic flux is obtained by the use of only two
or three commutating coils. There is no standard arrangement of
leads that can be applied to all field windings. I n the more modern
motors i t has been necessary to place the leads a t the end of the
coils on account of lack of clearance for the commutating coils. By
the use of open and crossed coils the wiring around the frame 1s
simplified. This requires carrying in stock two types of eoils, with
the added chance of getting coils mixed in replacing damaged coils.
T o provide against this, these coils are marked with an "X" for
crossed coils, and an "0" for open coils. I t is considered best

286

ELECTRIC RAILWAY HANDBOOK

practice in connection with modern motors and is recommended

for old motors, to use cables coming out of the body of the coil
instead of heavy brass terminals for the wiring around the frame
connections. The objection to the use of terminals is that they are
more likely to break off due to vibration, or to develop a loose connection and burn off the lead. Another disadvantage is the
difficulty of properly insulating the terminals after the connection
is made.
Fig. 53 shows a common arrangement of coils for a field control
motor (short or permanent coil B and added coil A), assembled
together and having four leads. Generally, the permanent coil
has about 60 per cent of the total number of turns, while the added
coil has the remainder. In some cases the proportion is as low as
50-50
As the added coils are worked only a part of the time, the
section of the copper conductor in this coil is made smaller, approximately 75 per cent of the section used on the permanent coil.
The following precautions are worth following while overhauling
or repairing field windings: coils and frame should be cleaned and
painted; coils should be placed on poles properly; coils should be

FIG.53.-Section

through maln field coil.

spring supported, and where necessary, backed up by washers of

either metal or fibrous material, well painted; in replacing coils,
temporarily tape the spring and washers to the coil to keep them
from working out of place and getting between the pole and pble
seat when assembled; clean the surface of the pole and pole seats,
to insure a good close fit when the bolts are drawn up tightly; do not
pound the pole in place with a s l e d g e u s e a block of wood or a piace
of soft metal, such as copper; be sure that the pole bolts are drawn
up tightly and that lock washersare under each nut; add some white
lead over the nuts after tightening, to prevent entrance of water;
sound the assembled pole, using a small hammer, to insure that She
poles are drawn up tightly; connections must be made as shown on
the winding diagram; check the polarity of each coil; the connections
between coils and with cable leads should be made by butting the
ends together, covering the joint by a copper sleeve and soldering
well; when the coils have terminals, the ends of the cable connected
to the terminal should have a metal sleeve soldered to the wire,
which is held in the terminal by screws securely tightened and locked;
the ends of the cables connected to the brushholders should be ptovided with metal sleeves or terminals that are to be securely clamped
and locked to the brushholder casting; all wiring around frame
should be securely anchored to the coils and tied down, to prevent
vibration and to keep the insulation from being rubbed or cut by
parts of the rotating armature; dip and bake the individual coils;

RAILWAY MOTOR FIELDS

287

to secure best results, dip and bake the entire frame after the coils
are in place and all connections are made; the motor leads should
be well protected by insulating bushings where they come out of
the frame.
Aluminum Field Coils. The Lind method of winding and insulating aluminum field coils for direct current motors was introduced
into this country about three years ago, after fifteen years' use in
European countries. The aluminum coils are wound with bare
square aluminum wire insulated with aluminum oxide. A layer of
paper is placed between coil layers to facilitate even winding and
to serve as a wick for distributing the oxidizing solution and a finish
coating of insulating varnish. Lack of other insulating materials
between turns and layers permits a considerably increased crosssectional area of conductor. The characteristic relations between
the cross-sectional areas and corresponding resistances of aluminum
and copper, and the per cent saving in weight, are such that for the
aluminum coil resistance to equal the copper resistance, the crosssectional area of the aluminum conductor must be increased 59
per cent, and when increased by that amount the saving in weight
IS 52 per cent. For railway motor field coils wound with round
copper wire, the aluminum cross-section can be made large enough
to give coils of equal or lower resistance than the standard copper
coils. For most strap-wound coils, the same is true, although for
some the resistance is increased somewhat, but for coils wountl
with rectangular or square copper, the copper resistance cannot t e
met within the usual tolerances.
Testing Polarity of Field Coils. A reversed main field coil wil:
cause the armature to run hot, due to the unbalanced magnetic
circuit; a reversed commutating coil will cause poor commutation
and flashing. The apparatus required for testing polarity is a
polarity detector, such as a small compass, a switch resistance (such
as several car heaters, grid resistors, or headlight resistor), a fuse or
circuit breaker, and, until the resistance has been determined
properly, an ammeter. The resistance should be such that the
current is sufficient to give a readable deflection on the polarity
indicator, but not above the rating of the motor. The motor may
have the armature in or out of the frame and, if a split frame, may
be open or closed. With thecoilsall connected in series, connect
the two field leads to the test circuit. When the switch is closed, the
polarity indicator, when held close to the ends of the coil, or to the
pole stud bolts on the outside of the frame, will reverse a t alternate
main poles; i. e., if No. I pole attracts the positive endof the polarity
indicator, No. 2 should attract the negative end, No. 3 the positive
end, and No. 4 the negative end. This test can be made by rubbing
a screw driver on the pole bolts and then holding it to the compass
needle or polarity indicator. If these conditions are not obtained,
the field winding connections should be changed. I n the case of a
commutatirg pole machine, it is important to have the pro
relation of polarity between the main and commutating field
To check this, connect the negative armature lead of the motor
to the positive field lead, the positive armature lead to the trolley
side of the test circuit, and the negative field lead to the ground side

par:

288

ELECTRIC RAILWAY HANDBOOK

of the test circuit, and close the switch. If the armature is in the
frame and the brushes are making contact on the commutator,
current will flow through all the windings; if the armature is not in
the frame, it will be necessary to short-circuit the brushholders.
With these conditions, the polarity of a main pole should be the
same as the polarity of the commutating pole next to it in a clockwise direct~on when facing the commutator end of the motor.
Tests of Rewound Armatures. The following list gives the tests
in the order of their application to completely rewound armatures
I,y the Ch~cagoKys. Co : ( I ) Breakdown test between commutator and sleeve as assembled off the machine; (2) breakdown
test between the commutator and sleeve, with commutator on
the shaft; (3) bar to bar test on commutator; (4) after coils are
in slots, with lower leads connected to commutator; breakdown
test, commutator to ground; (5) after coil.; are in slots, with lower
lrads ronnected to commutator; breakdown test between adjacent
leads, (6) after coils are entirely
ronnected in and before soldering;
breakdown test, commutator to
ground; (7) after toils are entirely
connected in and before soldering;
short circuit tests on armature
co1ls;(8)afterbanding,shortcircuit
and open circuit tests on armature
coils; (9) after banding, breakdown test to ground.

Locating Short Circuits in

Annatures by Direct Application
of Alternating Current. 'The
following method described by
FIG. SJ -Test of armature for short
CI~CUI~.
Leonard Work, Electric Journal,
1910,is stated by him to be rapid
and thorough (Fig. 54). Alternating current is applied to the
commutator through contacts separated by an amount equal to
the correct brush separation for ordinary operation of the motor.
While this continues, connection is made and broken between any
pair of adjacent b a n as a t A on the side of the commutator opposite
the points where thealternatingcurrent isapplied. If on such making and breaking there is a small spark, there is no short circuit
between these adjacent ban. If there is no spark, a short circuit
exists between these two bars The points a t which the alternating
current is applied to the armature are advanced around the armature and the test is continued between each pair of adjacent commutator bats. In making this test, approximately full load current
should be applied. The test is made with the armature removed
from the motor frame.
Armature Test by Full Load Current. I n theshops of the
Brooklyn Rapid Transit Company armatures removed from
the motor frames are tested for poor soldering, defective leads and
short circuits in commutator or armature winding by passing full
load current through the armature by way of a pair of brushes
set in the same relative posttions as for actual operation. The

289

ARMATURE TESTS

current flowing through the armature and the voltage drop between
adjacent commutator b a n are measured. Poorly soldered connections and abraded leads are burned off by the current The
testing outfit consists of a rheostatic controller and a set of car
resistances arranged to give arly desired current from 8 to jm
amperes, a n armature circuit breaker and switch.
Insulation Test of Armature Coils. Between windings and
ground
For roads using trolley:
New armatures

Old armatures

2500 xolts. n c
loo0 volt., a c

. . ... .

.. 55 srconrl;
seconds

For roads using third rail where voltage fluctuations havc to be

taken into consideration:
New armatureq
Old or partly repalred armatures

3000 volts. n.c 5 seconds

I 500 volts. a.c.. 5 seconds

Between armature coils before windings are connected:

New armatures
Old armatures

zzo volt$. d c 5 seconds

I l o volts, d.c.. s seconds

Equipment for Armature b h t i o n Testing. The Chicago City

Railway Company uses a motor generator set consisting of a single
phase, 12-kw., 230-volt, 125-cycle generator excited by 3 1.5-kw ,
125-volt belted exciter and driven I>y a 550-volt, 1875-rev. per
min. direct current motor. The testing electromotive force
is obtained by the use
of a 5-kilovolt-ampere
IW Im. T d Pw.
6000-vol t transformer
arranged to give a voltage of from 2 0 0 to 6000
(uoltn.cm I.".
con)

Yoke Transformer Test

(Fig 55). A "U"shaped two-pole laminated
iron core having its pole
Aruulur.
pieces cut to embrace
the armature as is done
F I G 5 s -YuLe transformer test for armatures.
by the motor field
pieces, and carrying a n
exciting coil about its yoke, is placed against the armature t o be
tested. When the coil on the yoke is excited by analternating
current, the resutting flux passes through the armature core and
its windings and induces a n electromotive force in the latter.
The armature coils which are in a position to link with the greatest
number of lines of magnetic force have induced in them the
highest electromotive force and if one of these coils be short-cirt this will be evidenced
cuited a heavy current will flow t h r o u g h ~and
by heating of the coil, or by the noisy vibration of a soft iron test
piece laid over the armature. Manifestations remain the same
when the two adjacent commutator bars to which the terminals of
a short-circuited coil are connected are short-circuited by having a

200

LLECTRIC RAILWAY HANDBOOK

p~eccof metal insrrted between them and the testlng current is

n~omrntar~l\
appl~ccl The test IS cont~nuedaround the armature
I)\ rt\ol\ in:: it so that one c o ~ al t a t ~ m ew~llcome under the Influ
( nct of thc \trongist held
The co~l\not found to be short-circu~ted
.ire nelt Ir3trcl for open c~rcuit r h ~ sIS done bv short clrcultlng
adjacent c o ~ ~ w ~ u t .bar,,
~ t o rpair a t a tlrnc and applv~llgthe test for
short c~rcultto the correspond~ngcoil An ~ n d ~ c a t i oofn a short
clrcu~t\\hen t h ~ sis done probes the absence of an opc? circuit
In the coil In testlng for opcn c ~ r c u the
~ t vibrat~ontest should be
uscd and the current
C ~ l r a u o s ~ e t cLOW
r
reudtn#
shollld not be kept on
Voltmetor or Telephone
long enough to damXecelvcr
age the cod under test
The alternat~ngelectromot~ve force applied to the tcst~ngcoil
may be ol>t.unecl by
ionnectlng it t o a convenlent lamp socket.
For ease In handling,
the test~ngyoke may
be bolted to the top
cross plece of a twowhecletl
warehouse
truck arranged w the
truck frame w~llstand
vertlcall\ when bppetl
up
Thrs arrangement permits the testing ) oke to be quickly
moved to the armat u r e s w h ~ c ha r e
mounted on wooden
horses of such he~ght
that thearmature core
center shall be even
w ~ t hthe center of the
test~ngyoke when the
truck IS t~ppedup
FIG 56-Bar to bar arrnaturc test
Bar to Bar Armature Test. Method of locat~ngopen circuits, short clrcults and
grounds In a closed cod armature (Fig 56) I n the following
d~scucs~on,
"meter" refers to the Instrument w ~ t hwhich the readlngs ar, taken, t h ~ smay be a galvanometer, a loareadrng voltmeter, or a telephone recelver If a telephone receiver 1s used,
"deflect~on" refers to the cllck of the telephone recelver when
the circuit IS made or broken through ~t Two or three cells of
batterv \\111 be suffic~entfor ordinarv work The battery IS first
connected across commutator bars correspond~ng to operating
brush posit~ons If the deflect~onsare the same when the meter 1s
connected to each palr of adjacent commutator bars, then there
are no open clrcults, short clrcu~tsnor leads ~nterchanged.

FIE,LD COIL TESTS

Optn Czrcrrrl The battery IS connected to two polnts on the
commutatol and the mettr 1s tonnectecl across each palr of atljaccnt
commutator bars In turn Zero tleflectlon ~ncllcatcsan open clrcutt
When thtre 1s but one open ilrcult on one slclc between the battery
contact point, there all1 be 3 great deflection when the mctrr IS
contlcct~d.rtross the adjacent commutator bars (as 1 7 and 18)
which Ire on etther slde of the open ctrcult When an open clrcult 1s located it should be closed temporar~lyby 3 drop of solder
between the two adjacent commutator bars between whlrh the
break occurs If other open clrcults, eulst they may be located
by shlftlng the battery contacts to other polnts and repeating the
above process
Shorf C ~ r r t r r f When the meter IS connected Eetwecn two
adjaccnt bars such as 1-2 and 3-4, between whlch there IS a short
clrcutt, the re will be a very sllght deflection
Inad5 Inlrrrlzcrng~d If as I~ctween l o anel r r the drflrrtlon 1s
rr\ersetl, the ~ r m a t t l r eleacls connectecl to hars lo ancl 1 1 are
Interch~ngeelfrom thelr correct positlon
Grolr~td The bqttery 1s connecte(l to two potnts of thr commutator, appro\lln~ttly opposlte each othcr, onc meter tcrminal IS connecttd to the ~ r m a t u r eshaft, and the other IS drawn
complctcly around the commutator If there IS only one grouncl
there w~llbe found two balanclng polnts, or polnts givlng the least
deRectlon Ne\t, the two battery contacts are to be shlfted a fcw
bars one way or the other When thls IS tlonc one balanclng point
alll be follnd a t thc same bar as bcforc, thus showlng the appro\lm,tte lotatlon of the ground The other halanclng p a n t wtll move
wlth each cl~angeof battery connectton ~f but one ground eulsts.
Insulation Test for Field Colls. I or roads uslng trolley
Yew fields

Old or p ~ r t ~ a l lrye p ~ r c d

.I

2500 volt\
c
looo volt, a c

5 seconrls
5 se~onds

Tor roads uslng thlrd rall where voltage fluctuations habe to be

taken Into cons~deratton
New fields
Old or partially repalred

3000 volts a c
1500 volts a c

5 seconds
5 seconds

F ~ e l dCot1 Test by Full Load Current. 1 1111load current 1s a p

plled to the field cot1 Temperature and voltage drop across
the c o ~ lare observed and comparcd with those for a field coil
known to be good
Fleld Coll Test by Transformer. During thls test the field cot1
becomes the secondary wlndlng of a transformer (I lg 57) The
field cod 1s placed on the laminated Iron core and the yoke placed
to complete the magnettc clrcurt Alternatlng current IS thcn applied to the prlmary The prlmary voltage applled and the ratlo
of the number of turns in the co11 under test to the number of
turns In the prlmary coil must not be such th.tt a voltage tlartpcrous
to the operator or insulation w~llbe Induced If there 1s no short
clrcutt In the field cot1 the current taken by the prlrnary at any
~ p p i ~ evoltage
d
w~llbe the same as wlthout the field roll In place but
w ~ t hthe yoke In place A short clrcult III the field cod IS e\~tlcncul
by a considerable current In the prtrnary, ancl ~f the appllccl \olt,~ge
be of sufficient \ d u e the fielcl coil will Iwctlrne appreciably heated by

292

ELECTRIC RAILWAY HANDBOOK

The terminals of a field

coil found to contain no
short circuit are then
luk.
connected together and
Pirl.1 Coil
the test for short circuit
U&. T d
is repeated on this field
coil. If there is no evidence of short circuit
when this is done then
4
L . m L b d lrcn Cur.
there is an open circuit
V"l1ul.v
within the coil. If, however, the test indicates
FIG. 57.-Transformer test for field coils.
a short circuit, this field
coil contains no open circuit. Fig. 58 from the Electric Railthe induced current flowing within itself.

%j*

PIG.58.-.4rrangement for testing field coil-Brooklyn.

FIELD COIL TESTS

293

way Journal shows the construction of a field-testing transformer

used in the shops of the Brooklyn Rapid Transit Company.

The framework A is made of wood. B is a slate panel 20 X
11% X I in., on which the circuit-breaker, ammeter and switch
are mounted. C, H and D, which comprise the magnetic circuit,
are made of several thicknesses
of HI-in. wrought iron dipped
Yetr
I
I
in shellac. After the shellac
was dried, the laminations were
placed between the two %-in.
side pieces and riveted together.
The primary coil is made up of
81 turns of No. 5 B. & S. double
cotton-covered wire with 2 7 turns
between T-I and T-2 and 8 1
turns between T-I and T-3. E
is a cast-iron counter-weight 698
in. in diameter and 6 in. wide p,,. S9.-lransformer
for testing
which assists in raising D and also
held cods.
in keeping it in the upper position while a field coil is being placed in position for testingor while
one is being removed. The arms. F. made of % X 2 in. flat iron,
bolted to 3,have the counterweight attached-at the other end
and are free to turn on a pin. D and counterweight E thus
revolve on the same pin, which is supported by two brackets,
one on each of the upright members of the
wooden frame.
Fig. 59 shows anarrangement for
I Adlsitablc Inductance d. Adlaatable lndudance B I other
making the transformer test of field
coils. C C are two
similar coils, so connected that the voltage induced in one
I
I
I shall oppose that induced i n the other.
The field coil to be
tested is laced on
one of the Pegs A (the
yoke not being used
in this case) and an
alternating rurrent is
Coel onder Test
Rlandard Coil
PIG. So.-Inductance balance for testing field appliedtothe rimary
coil. If the leld coil
coils.
under test contains no
short circuit, the sound in the telephone receiver will be very faint.
If, however, there i s a short circuit in the field coil, this will unbalance
the fluxes through C C and the sound in the telephone receiver will
he loud. The terminals of a field coil found to contain no short
circuit are then connected together and the test for short circuit is

",",zt

.;.;. ,

lDr-td'

294

ELECTRIC RAlLWAY HANDBOOK

repeated. A loud sound in the telephone receiver now indicates

that there is no open circuit in the field coil. A faint sound indicates that there is an open circuit in the field coil. If i t is desircd,
this apparatus, together wih the yoke but without the telephone
receiver, may be used to test field coils in the manner outlined in
the preceding aragraph.
Field Coil f e s t by Inductance Balance (Fig. 60). Instead of
the battery and buzzer indicated, a low-voltage alternating current
may be used. A standard coil known to be faultless and similar
to what the coil under test should be if perfect, also the coil to be
tested are connected as shown. The adjustable inductances A
and B are then adjusted to give a balance which is indicated by the
faintest sound in the telephone receiver. If A and B are then practically alike, the coil under test is without short circuit or open
circuit. If. however, B is less than A there is a short circuit in the
coil under test.
Insulating Materials for Railway Shop Use
(X.E.R.E.A. Miscellaneous Methods and Practices)
Definitions. Air Drying Varnishes. Linseed oil or other drying oils, with such driers and treatment as to dry quickly without
necessity for baking.
Baking Ywnisbs* Linseed oil or other drying oils, with y m
and driers. Requires baking.
Spirit Varnishes. Animal or vegetable gums, shellac, copal, etc.,
dissolved in alcohol, amyl acetate, benzine, etc.
Insulating Paints. Solutions of asphalt, and mineral gums in
carbon bisulphide, naphtha, etc.
Impregnaiing Compounds. Petroleum asphalt, melting point
105 to 1 x 5 deg. C., used for impregnating motor fields, etc. Parxffine and linseed oil are used as impregnating material for special
purposes as described below.
Air drying varnish should be composed of linseed oil or other
drying oils sufficiently treated and with proper amount of dryers so
that varnish will dry in 2 or 3 hours, leaving a smooth, flexible surface. Baking not required. Should be both oil and moisture proof.
Should not crack, soften or become tacky a t loo deg. C.
Baking Varnish. Composed of high-grade linseed oil, with sorne
gums and dryers. Baking varnish should not be used except baked
a t r o o to 1 x 0 deg. C. for 10 or 1 2 hours. Characterist~cs:high
melting point, flexibility, smooth, glossy surface, good filler, high
insulating qualities, oil and moisture proof.
Spirit varnishes are composed of animal and veactablc gums, such
as shellac, copal, etc., dissolved in spirits. Characteristics: oil
proof but not moisture proof; good stickers and binders, hut not
necessarily good insulators. I n gcneral they soften a t loo (leg. C.
NOTE:On string hand or wehbing covering mica a t the outer
end of the conlmutator it is recommended that no varnish be used
that will soften a t a temperature of loo dcg. C.
Insulating Paints. These are solutions of asphalt or mineral
gum in naphtha or other solvents. Characteristics are moisture

INSULATING MATERIALS
proof b u t not oil proof, lacking in toughness, good as a sticker and
paint, but not high in insulating qualities.
Impregnating Compounds. Asphalf Compounds. Composed of
asphalt. Characteristics: Different asphalts have widely different
melting points. The asphalt to be used is one in which the melting
point is above the operating temperature, preferably I 10 deg. C.
At ordinary temperature should be somewhat pliable, but should
not flow. At melting point should flow freely. I s moisture proof,
but not oil proof. Good as a filler, and has fair insulating qualities.
Asphalt should be practically free from volatile matter. In order
to get a good penetrating asphalt the committee recommends that
the asphalt shall not contain more than 5 per cent of constituents
insoluble in carbon bisulphide (CSz).
Para$ne is a mineral gum of a low melting point, suitable for
impregnating wood for outside or low temperature work. I s not
suitable for use in railway apparatus where working temperature
is over 50 to 60 deg. C.
Linseed oil is excellent for filling wood parts of railway apparatus.
The Darts should be boiled in raw linseed oil until fillkd andl then
bakeh to oxidize same, the wood to be finished with either a n air
drying or baking varnish. Wood treated in this manner is excellent
for outdoor work, exposed to climatic conditions. Linseed oil
hardens under temperature, does not melt, is moisture and oil proof
and is flexible and strong.
Tests for Insulating Materials. The following simple tests are
recommended for the various substances: Without a n elaborate
testing laboratory, i t will be impossible to find the exact values of
tests on the various insulating compounds, but these tests can be
made with little or no apparatus and serve as very good general
tests of the material.
Asphalt. The chief consideration in connection with asphalt
is the melting point. Asphalt can be obtained from a low to a very
high melting point. Some asphalts contain as high as 30 per cent
of inorganic matter. I n order to determine how much inorganic
matter, a certain definite amount of the asphalt can be melted and
burned away in an ordinary crucible or gas flame. The presence
of sulphur and the degree may be determined by melting the
asphalt and allowing a bright sheet of copper to remain in the
melted asphalt for 2 or 3 days. The tarnishing of the bright
copper will indicate the presence and quantity of st~lphur.
Linseed Oil. Linseed oil may be heated up to the boiling point,
and if there be any foaming and spluttering as it begins to boil, it
indicates presence of water in the oil. I inseed oil may have a
sediment usually known as "foots." The relative amount may be
approximately determined by allowing a sample of the oil to stand
in a test tube for about 24 hours and noting the amount of sediment
deposited. Regular boiled oil has a specific gravity of 0.945 or
higher, whereas the ordinary boiled oil which has been produced by
pouring into the bunghole a certain amount of drier, has a specific
gravity but little higher than raw linseed oil, this sperific gravity
being Q.931. Acids are used during the process of refinement. I n
~ r d e rto test the amount of acids, it would be well to immerse a
bright copper strip in the oil for several clays.

296

ELECTRIC RAILWAY HANDBOOK

Japans and Varnishes. These should be tested a t a temperature

which is ordinarily used in service. After the coating has thoroughly dried in the air or by baking, according to which treatment the material is designed, the flexibility of the coat may be
determined. I t would be well to test with a bright copper strip
for acids.
Shellac. Shellac contains more or less rosin. The determination of this requires elaborate analysis. The quality of the shellac,
however, may be determined by painting on bright tin, and by
making comparisons with other shellacs, the flexibility may be
determined. A very flexible shellac has little or no rosin. The
presence of rosin makes it brittle. I t would be well KO test with a
bright copper strip for acids.
Para$ne. The chicf di5culty in para5ne is water and sulphuric acid from the refining processes. The amount of water
can be determined by heating; foaming and sputtering on reaching
the boiling point indicates the presence of water. Sulphuric acid
can be detected by the immersion of a bright copper strip, or by
mixing some of the para5ne with boiling water and then muring
off the water and testing with litmus paper.
Impregnation of Motor Insulation. Dipping in varnish or insulating compound and then baking thoroughly fills all cracks and pores
in the insulation. This greatly reduces the possibility of breakdowns, which might be caused by moisture or other conducting
materials filling these cracks. Further, such treatment acts as an
effective bond to r e n t vibration of motor parts. Dipping
and baking, even o new motors, is an insurance against maintenance charges for rewinding. I t improves the insulation, fills
up the pores, keeps a smooth surface on the coils, and prevents
v~brationof motor parts. The equipment consists of a tank to
contain the di ping solution, an oven in which to bake, and means
of handling t i e apparatus. Heat the apparatus in a n oven a t
95 to 105 deg. C., for 1 2 to 24 hours, depending on the size of the
armature, to insure that all moisture is driven off and to warm the
armature for dipping. Dip in an oil proof and moisture proof
baking and insulating varnish a t approximately 0.840 specific
gravity. This depends, however, upon the varnish and upon the
thickness of coat desired. If the varnish is too heavy, thin it with
benzine. Dip armatures in the varnish in a vertical position, pinion
end down, and to such a depth that the varnish will come to the
commutator neck. Allow to soak for approximately five minutes.
If a tank is not available, results (though not so good) may be
obtained by turning the armature in a shallow pan with the varnish
deep enough so that the bottom of the slots will be completely
immersed. If this is done, the insulated creepage surface a t the
end of the commutator should be treated by repeated paintings of
the varnish. Turn until all the coils have been thoroughly soaked.
Field coils for direct current machines may be removed and dipped
and baked, but if desired, the frame may be dipped with coils in
place. Drain a t room temperature until all dripping ceases. The
apparatus should be placed in such a position that pocketing will
not occur. It may be necessary to turn during draining. Place

HANDLING COPPER
apparatus in oven in a position to avoid pocketing of varnish,
and bake a t 95 to 105 deg. C., for the following time: armatures
below 1 2 in. diameter, 48 hours; 1 2 to 30 in. diameter, 60
hours; over 30 in. diameter, 72 hours. If an armature is baked
in a horizontal position, it should be turned a t intervals during the
first half of the baking period, otherwise the varnish will drain
toward the lower side and throw the armature out of balance.
Handling of Copper. Copper cannot be given the same rough
treatment that other metals will stand, but requires some important
precautions in its handling and application. I t fails quickly under
localized stresses. Sharp bends, rough or nicked edges in copper
straps or wires, limited movement to take care of expansion and
contraction, all are points especially to be guarded against. This
is particularly true when the copper is subject to quick sharp
blows or vibration, such as are common with railway motors. All
bends in copper should be made with as large a curvature as is possible in the space available. The effects of vibration, expansion,
contraction, and centrifugal force show up first in sharp bends and
Gllets. Sharp bends should not be made a t the ends of copper
strap field coils or when making clamped or soldered joints, as
this may later be the cause of a motor failure. I t often happens
that armature coil failure is at a point where wires have been carelessly bent or crossed. The nickingof copper should also be avoided.
I t is very easy to nick copper with the sharp edge of a metal drift
or other tools as are used in connection with the winding of armatures.
I t is preferable to use a hard fiber drift and drive leads down into
commutator neck slot by using a copper filling piece placed on
the lead to receive the blow from the hammer. Another source of
trouble due to nicking of copper is found in field coil cable leads
breaking a t the point where the insulation has been cut off with a
knife, the break having been started by the knife nicking the strands
of the cable. Extreme care should be used in removing insulation
on all cables and wires of small cross-sections. For the same
temperature change, copper will expand more than iron or steel.
Further, it is usually found that, in a motor, the copper becomes
hotter than the other materials of which the motor is made. Therefore, it is necessary to provide means for the copper to expand and
contract to take care of the relative motion between the different
materials for the changes in temperature. A common error in
this respect is to anchor the wiring around frame connections to the
frame proper, when it should have been securely bound to the windings so that it would be free to move with the windings. I t is
obvious that this is more important with solidstrap conductors than
with flexible cables. Properly supported copper stands up well
against vibration. In fact, one finds it extens~velyused in certain
applications where vibration occurs, because of its good behavior
in this respect. But improperly supported, copper fails easily
under vibration. This point is frequently overlooked in connecting
car wiring cables to the motor leads when cleats are not applied.
Care should be observed to properly locate the cleats in supporting
the cables. I t often happens that the weight of a solid connector,
even though it may seem rather small, is sufficient under vibration

ELECTRIC RAILWAY HANDBOOK

to cause the copper strap or stranded cable to break a t a point just
behind the connector where stresses are localized. The bad effects
of tinning stranded copper beyond the point where the joint is made
must be emphasized. The solder should stop just inside the connector, so that the stresses will not localize where the strands are
not supported against vibration.
Copper is subject to a form of "sickness" which so far as has been
experienced- is peculiar to copper alone. This subject has been
thoroughly discussed by N. B. Pilling ( B e d r i c Journal, 1920).
All commercial copper contains a small amount of oxygen in the
form of copper oxide, without which i t has poor mechanical characteristics. When it is heated in a flame which is rich in free hydrogen, the hydrogen unites with oxygen forming free copper and steam.
I t is a peculiar characteristic that hydrogen will readily enter hot
copper, but the steam cannot get out. The copper is thus not only
weakened by the elimination of the copper oxide, but the highpressure steam expands, producing a spongy effect which still
further weakens the copper. This effect is, of course, greatest near
the surface. This peculiar form of sickness should be guarded
against by maintenance men. An experience due to this change
in structure will serve to bring out t h ~ lesson.
s
An armature was
being wound with coils having german silver resistance leads,
with copper tips brazed onto their ends. A number of coils had
been put in place when i t was found that the first one had to be
removed. In doing this, the copper tips were bent back to get them
out of the way. With only a s~nglebend, one of the tips broke off
in the workman's hands. On examining all the tips of the coils,
twelve more defective ones were found. Tests showed that the
copper from which the tips were made was of good quality. A
study of the process of handling the copper revealed that the
defective tips had been heated in a flame containing unburned
hydrogen. Since one cannot see, without breaking the strap or
wire, whether the copper has been affected by this sickness, i t
follows that to be safe, co per should not be heated in a flame
containing an excess of hyjrogen. This means that with a blow
torch the copper should be kept outside of the inner cone of blue
flame. When heating copper in a gas and air furnace, an excess
amount of air should always be used, as too little air will produce
an excess of free hydrogen. Smoke from such a furnace always
indicates too little air, and the mixer should be adjusted to give a
little more air than is necessary to prevent any trace of smoke.
Wherever possible, the copper should be heated without coming
into direct contact with the flame. I t has been common practice
to burn the insulation from old coils. This should not he done
where the coils are to be reinsulated and used again. The question
then arises, how to remove the old insulation. One operator places
the coils in an oven and passes steam through for 12 to 14 hours.
He finds that the insulation peels off easily while hot. Another
operator dips the coils in a weak solution of muriatic acid for u
time (approximately 2 4 hours) so that the acid weakens the insulation, but not long enough to give the acid a chance to eat into the
copper. The necessary time required can be established by check-

MOTOR LEADS

299

ing carefully and removing the coil when the brightening of copper
commences. After the acid treatment, the coils should be thoroughly washed in clear water.
C o ~ e c t i o n sbetween Car Wiring and Motors. A scheme of
connections and cable suspension which has been used with considerable success by a number of operating companies is shown in
Fig. 61, which illustrates the method employed for inside hung
motors; for outside hung motors, the general arrangement is the
same. The knuckle joint connectors are held firmly in place
between two cleats fastened to the car underframing. The cleats
holding the connectors should be placed as near as possible to the
bolster and the truck pin on the car underframing, in order to keep
the cable loop, as i t swings between the car underframing and the
motor, as tlose to its original length as possible in any position of
the truck. About three inches should be allowed on each side of the
connector between the cleats
to take care of creepage.
The cleats are split so that
they can be put in place after
the connections have been
made and insulating hose or
tube has been slip d over
cables
the connectors.
are clamped to the housing
with a pair of stationary
cleats as they come out of
the motor frame. T o prevent chafing of the cables
as they drag over the motor housing, a pair of loose

TE

allow free mbvement but also such that the cables will not touch
the motor housing a t any point. To give further protection to
the knuckle joint connectors against the weather, a box may house
the cables a t the point where the leads from the motors are connected to the leads of the car body.
T o simplify the changing of motors, the cables a t the cleats on
the car underframing should be arranged in the same sequence for
all motors, and the cables as they leave the cleats on the motors
should always be in the same sequence and should line up so that
when the motor is placed in the truck, the proper lead a t the cleat
on a car underframing will be directly over the lead to which it is
connected on the motor.
Broken Motor Leads. Assume that No. 3 motor has an open
circuit, due to either field or armature leads being broken or burned
off, and that the car is continued in service. With the control in
any of the series positions, No. I motor receives double the current
taken by the No. 2 or No. 4 motors, which means that No. I motor
has an increased torque and tends to take too much of the load.
With the control in any of the parallel positions, the load is equally
divided between the three motors, I, 2 and 4. Depending upon

ELECTRIC RAILWAY HANDBOOK

service conditions, such as headway, schedule, location of turnouts
on single tracks, congestion of traffic, etc., the amount of series
running will vary, and this will determine how much excess current
No. r motor wjjj receive. I n many cases cars are cont~nuedin
service with the motors operating in this unbalanced condition,
which is discovered only when they are given their regular light
inspections, which are usually from 7 to 15 days apart. Under
these conditions, No. I motor has had a number of short time
overloads, possibly resulting in overheating of the winding. Such
overheating may not cause the motor to fail a t the time the damage
is done, but the insulation may have been charred and its life
shortened. Some time later the windings on this motor may either
ground or short-circuit, due to the poor condition of the insulation, and when examined the motor will be reported as having
a roasted armature. I t is quite difficult to locate the direct cause
of the roasting, as the conditions responsible for this trouble may
have happened weeks before, and may have been remedied without
thought of the possible damage done to the windings of the running
mate of the motor with a broken or burned off cable. I n a service
test duplicating the above conditions, the brushes were taken out
of No. 3 motor. Run thus, the current in No. I motor was 80 per
cent greater, and that in Nos. 2 and 4 motors 10per cent less than
when all four motors were operating normally in the full series
position of thecontroller. I n full parallel, each of the three working
motors took 2 0 per cent more current than when four motors were
operating normally. In this test, the (experienced) motorman was
unable to tell by the operation of the car whether or not one motor
was out. The results of this test indicate that one motor could be
open-circuited, due to a broken or burned off cable lead, and the
car could be operated with the motors in this unbalanced condition
without the knowledge of the motorman; this would result in overheating the windings of the running mate of the motor with the
o r circuit. With the older and heavier type of motor on the car,
t ere was a sufficient mass of metal available to absorb the heat
resulting from short time overloading of the motor, due to unbalanced conditions resulting from a broken or burned off motor lead,
and consequently armatures were not so apt to be roasted. On
the other hand, the light weight, ventilated type of motor, weighing
much less than the old, heavy type motor, and operating under the
same service conditions, is unable to absorb so much additional
heat due to short time overload, and its temperature rises rapidly,
more often resulting in roasted armature windings. The always
important precautions against broken motor leads, such as proper
cleating, reinforced insulation, etc , therefore become of greatly
increased importance with the light weight ventilated type of motor.
I t may even be advisable to rovide for a daily inspection of motor
leads, although a hand t o u c l system of motor temperature observation as cars come in from runs should indicate a motor which
has not been working.
Railway Motor Gears and Pinions
Pitch. The diametral pitch (number of teeth pel inch of diameter of pitch circle) commonly used is 2.5; 3 is also used, but for

301

GEARS AND PINIONS

motors of not greater than 75 rated h p Larger teeth are used

on locomotives. Spur gearing has been of the standard 14% deg.
involute form.
Taper of Pinion Bore. A taper for bore of pinion in the proDortion of 1.25 in in d~ameterto I i t in length was adopted as st'andard
by the Am El Ry Eng Assn., 191I
Split Gears. The halves of eight-bolt gears are fastened together
by four bolts on each side of the hub, two of each four being placed
beside each other near the hub, and the other two as near the rim
as possible. There are two methods of bolting the halves of fourbolt gears together. One method is to place one bolt on each side
of the hub and the other two out near the rim Another method is
to place two bolts on each side of the hub, side by side between hub
and rim. A complaint in connection with split gears is that the
bolts are often stretched beyond their elastic limit.
Solid Gears. Generally speaking, solid gears can be made
stronger and lighter than split gears, moreover, they require no
key seating in the axle
Diameter Allowance and Pressure Required to Fit Solid Gears.
An allowance of o OOI in. for every inch of axle diameter should
require from 40 to 60 tons to force the gear in place, and such a fit is
ample to prevent the gear from slipping
Tooth Pressure. Experience indicates a pressure of ~ o o olb. per
inch width of gear face as a perfectly practicable value for continuous
rating of large gears, and that with special steel pinions and high
grade gears it is probably safe to exceed this figure. In the St.
Clair Tunnel locomotives the pressure is carried on a single gear
having a 6-in. width of face on the gears. The 11feof the pinions
is 40,003 to 50,000 miles. Wlth the normal loads a t which the locomotive operates on up grade the pressure is from 1500 to 2 0 0 0 Ib.
per inch width of face on the gears.
Factors Essential to Durability of Gears and Pinions. The most
important factors conduc~veto long life of gears a r e ( I ) Proper
steel properly treated; ( 2 ) proper alinement; (3) proper lubrication; (4) tight gear cases.
Specifications for Gears and Pinions. The Am. El. Ry Eng.
Assn has adopted standard specifications for case-hardened and
for quenched and tempered forged steel gears and pinions, which
contain the following provisions Blanks to be made from openhearth steel. Chemical composition of case-hardened gears and
plnions-

..... o -202 8perpercent-not

cent

Carbon
Manganese..
Phosphorus
Sulphur

..

0 . 5 0 per cent-not
0 60 per cent

less than o

1 2 per

~ e n nor
t more than

less than o 40 per cent nor more than

not over o 05 per cent
not over o 05 per cent

(Requirements for heat-treated gears and pinions include the last

two, relat~veto phosphorus and sulphur, only) The minimum
thickness of the rim of gears, measured $6 in from edge of rim,
shall be 96 in for 3 pitch, Tis in for 2 % pitch, and 34 in for 2 pitch.
Approximate Life of Gears and Pinions. The following information is from a paper by T. W. Williams a t a meeting of the

302

ELECTRIC RAILWAY HANDBOOK

Street Railway Association of the State of New York. Data from

gears and pinions ( 2 . 5 pitch) on 100-h.p. motors on two cars identical
as to weight, motor capacity, running schedule and route. T h e life
was estimated by subtracting the amount of wear on the pitch circle
of the teeth after running 50,coo miles from the original thickness
and then estimating how many more car miles the teeth could run
before 0.1 in. had been worn from each face of the teeth, on t h e
basis that the wear would be constant throughout their life. T h e
pitch of all the gears was 2 . 5 .
Description

L ~ f eof pin- Life of gear.

miles
ion, miles

Untreated high carbon pinion mnning with cast

70.000
steel gear.
Heat-treated high carbon pinion running w ~ t h ~ o o . o o o
cast steel gear.
1Ieat-treated hiph carbon pinion running with
250.000
heat-treated h ~ g hcarbon gear.
Case-hardened pinion running with case-hardened zoo.oo0
gear.

2zo.000
z4o.000
75o.000
650.000

Some of the modern types of gears and pinions have been showing
a life very considerably in excess of that indicated in the above table.
Material of Meshing Gear and Pinion Should B e t h e Same.
If soft steel gears be run against hard steel pinions, or vice versa,
the softer surface becomes charged with the gritty material which
finds its way into the gear case, and when so charged this surface
grinds away the harder surface, but when hardened gears and
pinions are run together this trouble does not seem to be so great.
Gear Teeth of Special Design. Long and Short Addendum Type.
The "addendum" of a gear tooth is its length above the pitch line;
the "dedendum" is the depth below the pitch line to which the
top of the meshing tooth enters. I n the standard type of involute
gearing, the addendum and dedendum are equal. I n the "long
and short addendum" type of gearing, the armature pinion has a
long addendum and short dedendum, while the axle gear teeth are
of the corresponding opposite design-short addendum and long
dedendum. This accomplishes a shorter pushing action as the
teeth go into mesh, and a longer pulling action out of mesh, with a
consequent less tendency to vibration. I t also results in a wider
base and greater strength of the pinion teeth, as well as more metal
between the roots of the teeth and the bore of the pinion. Where
only a portion of the gear and pinion equipment is of this type,
great care should be taken to avoid mixing with the standard type,
as, obviously, the two types will not mesh together properly.
Helical Gears. The helical gear overcomes inherent gear vibration by eliminating the so-called "stepping over" action. The
operation of this type of gear can best be realized by thinking of i t
as made up of a number of thin spur gears twisted on the shaft with
respect to each other. Helical gears transmit practically average
motor effort, with properly maintained bearings. The gears .tend
to wear evenly over the full tooth length, thus preserving the
original tooth form. The contact from the tip of the tooth to

GEARS AND PINIONS

303

the root and across the face is the pro rty inherent in helical gearing which produces the smoothness op",ear action. The action of
engagement from one side of the tooth to the other is practically
one of pure rolling. The action from the tip of the tooth to the
root is one of sliding and rolling. The percentage of rolling action
can be redetermined, within reasonable limits, and the teeth
can be &signed to give a maximum percentage of rolling. As a
result of this rolling contact across the face, there will be a t any
time (after the tooth has come into full mesh) tip, pitch tine and
base contact. In the case of the spur gear, it is in contact across
the face first a t the tip and then progressively from the pitch line
to the base. From the design of the involute form of tooth, there
is pure rolling a t the pitch line, while a t the tip and the base there
are sliding and rolling. The tendency is thus to greater wear a t
tip and base. A s the gear
wears it is continually
destroying the original
tooth form. The characteristics of helical gears
for railway motors, as
produced by the Westlnghouse company, are:
4% to z diametral pitch;
3% to 5 in. face; 7!6 deg.
helix angle; 2 0 to 2 2 3 4
deg. involute, with long
addendum on pinion and
s h o r t addendum gear
teeth. The advantages
claimed are: reduction of
gear vibration ; increase
jn life of gears; reduction
In gear noise; together
with the advantages of
the long and short adFIG. 62.-Gear tooth vernier.
dendum as before listed.
The end thrust due to angularity of teeth has, according to some
users, been productive of increased wear on bearing flanges and
thrust collars; others, however, find the end thrust only sufficient
to reduce the lateral movement of the armature, provide a cushioning effect to such lateral movement, and maintain sufficient end
bearing to insure an oil film.
Measurement of Gear Tooth Wear. A gear tooth vernier, which
is most convenient for the measurement of gear tooth wear, is
shown by Fig. 62, which illustrates its method of use. One vernier
enables the measurement to be made at the pro er distance from the
top of the tooth (usually on the pitch circle), wKile the other is used
for the measurement. Both verniers are scaled to read to o.oor in.
Fitting Railway Motor Pinions. Experience has shown that in
order to obtain satisfactory operation, pinions should drive their
gears through the "press fit" or "shrink fit" on the shaft and not
through the key. The key acts merely as a safety device should the

304

ELECTRIC RAILW.\Y HANDBOOK

pinion accidentally loosen. The desired fit for the pinion can be
had by heating or by pressing. The following points should be
observed when putting pinions on railway motor shafts with
taper fit: The shaft should be clean and free from burrs or swellings;
the pinion bore should be clean and free from burrs; the fit of
the pinion bore should be in contact with a t least three-quarters
of the surface of the taper fit on the shaft. This can be checked
by rubbing I'russian blue, thin red lead and oil, or thin lamp black
and oil on the pinion bore and fitting i t on the shaft. The pinion
then should be put on the shaft cold to make sure that the keyway
in the pinion is the proper size for the key mounted on the shaft,
and that the pinion does not ride or bind on the top or sides of the
key and will not ride the key when pressed further on. The keyway
on the inion can be 0 . 0 0 2 in. wider, but not less than the key.
be a t least
in. top key clearance in the pinion.
There s!ould
The corners of the key should be rounded, so as not to cut into the
fillet of the keyway. Pinions may be pressed cold o ~ l t othe shaft
with a wheel press. The pressure required will be 1 2 to 25 tons
for pinions on motors up to 1 2 5 horsepower and 40 to 80 tons for
pinions on motors of 1 2 5 horsepower or over. Pinions with bores
up to three inches that are pressed on cold should advance on the
shaft approximately W. in.; those with three- to four-inch bore,
96. in.; and those with four- to five and one-half-inch bore,
in.
This distance is measured from the point where the pinion is seated
firmly on the shaft before pressing.
To fit by heating, pinions for motors up to I 25 horsepower and up
to three-inch bore should be heated in boiling water for thirty
minutes, and those with three-inch or larger bore for sixty minutes.
When the pinion has attained the temperature of the boiling water,
i t should be taken out of the water and the bore quickly wiped clean.
Without allowing the pinion time to cool, i t should be tapped on the
shaft with a six or eight pound sledge hammer, using a heavy
piece of wood or copper between the pinion and the hammer. This
sledging is not to get n driving fit, but to make sure that the pinion
is well seated. Three or four taps evenly distributed around the
pinion end should be enough. The pinion nut with lock washer can
then be screwed home tight with a wrench having a lever arm of
three to four feet.
For motors of over I 25 horsepower, the pinions should be heated
with a gas flame, applied in the bore of the pinion in such a manner
a s not to touch the teeth of the pinion, as this might affect the temper. The flame should be so regulated as to take 45 to 75 minutes
to bring the pinion to a temperature of 125 to 1 5 0 deg. C. The
temperature can be measured by placing the bulb of a thermometer
against the pinion between the teeth. The surface of the pinion
where the bulb touches it must be made perfectly clean by rubhing
with emery cloth. I t is also important to protect the exposed part
of the thermometer by covering it with asbestos cloth so the flame
cannot touch the thermometer. When the pinion has reached
the correct temperature, the bore should be wiped clean and the
pinion put on the shaft in the same mani~eras suggested for motors up
to 1 2 5 horsepower.

>(.

INSPECTION ANT) 1.URRTCATION OE' GEARS

305

Any furnace in which the pinion is so located that the flamc

cannot touch the teeth can be used for heating pinions. The flame
can be regulated and the pinions kept at a temperature of r o o d e ~C.
.
for pinions on motors up to 1 2 5 ho-rsepower, or 1 2 5 deg. C. for 125
horsepower motors or larger, until the mechanic is ready to apply
them.
Inspection of Gears. The inspection of gears and pinions on
some predetermined basis is essential to their length of life, whether
they are inspected on a time or mileage schedule, as the service or
conditions they are operating undcr would have to be considered
when deciding the rules to be followed. Cars operated with small
motors in city service are generally inspectcd on a thousand mile
schedule, or on a time basis of from seven to thirty days. This
inspection, however, is only a casual one, as the gear casc is seldom
removed, and the small opening in the top half of the Gear case is
used to make this inspection as well as to apply the lul,ricatinl:
material. The alinement of gear and pinion is of lirst importmcc,
as well as proper meshing, for if they are permitted to run out of
true centers for any length of time the possibility of their being
properly interchangeable is destroyed, on account of the uneven
wear of the teeth. Two important points which affect the alinement
~f the Rear and pinion seriously arc the axle bearin~sand axle collar.
Jt the axle bear in^ is worn on collar end, or thc aslc cnllar has not
t~cenproperly adjusted, the motor wiH h a w a laiectl motion which
will carry the pinion out of alinement with the gear, and carry the
gear case over against the gear, which in a short time will destroy
the gear case. Again, if the axle bearings are worn in their inside
diameter, they permit the gear to drop away from the pinion,
thereby destroying the pitch line of the gear and pinion. The gear
case, which has been planned for the protection of the gear and
pinion, as well as to hold the lubricating grease, is designed to fit the
allotted space in the best possible manner, and while it is not all
that could be desired, it is essential that it be maintained in a
manner that will protect the gear and pinion, and especially their
lubricating material, from being contaminated with dirt and water.
Lubrication of Gears is regulated according to the service in
which the equipment is being operated. Small quantities of grease
e
at
applied frequently will give better results than l a ~ quantities
longer periods, both as to the cost of grease and life of gear and
pinion. The opinion of the master mechanics on different electric
railways as to the kind of grease best adapted for the purpose is
diversified, some preferring a sticky grease because of its adhering
to the gear and p~nionteeth, while others, realizing the amount of
energy lost in overcoming the resistance of a sticky grease, prefer a
grease of better lubricating qualities. The idea of a sticky grease
being a preserver of gears and pinions probably is a false one.
The oily grease not only shows a longer wearing quality, but the
temperature increase of the gear and pinion is nil as compared with
tempera,tures obtained with sticky grease undcr the same conditions. rhe eucessive wear on gears and pinions occurs during
the winter rnonths, ancl an investigation as to why this should hc
shows that sticty.grcases become so hard in zero weather that

306

ELECTRIC RAILWAY HANDBOOK

(4
FIG. 63.-lfechanical

assembly of armature.

ARMATURE SHAFT RENEWAL

307

the lubricant flakes off and leaves the gear teeth bare until such time
as the temperature of the gear and pinion increases to a point where
the grease becomes more pliable. If there is not enough lubricant
left on the gear to cover the teeth, they continue in service without
lubrication until inspection time, when another a p Iication of
lubricant is made, and if the weather continues t o l l , the same
condition prevails as before and the gears and pinions suffer on
account of it. The conclusions are that an oil grease should be used,
that i t should be applied in small quantities a t short intervals, that
maintenance of the gear case is of first importance, and that proper
attention must be given to motor alinement.
Renewal of Annature Shafts. There are four general methods of
mounting the commutator and core on a railway motor armature
shaft. On the more modem motors, especially of the smaller sizes,
the core is built directly on the shaft and held in place by a nut
a t the commutator end; this type is used mostly on the ventilated
type motors having longitudinal ventilating ducts through the
core and a fan a t the rcar end. Where space is available, the spider

--

FIG. 64.-Clamp

for armature core for shaft renewal.

type of construction is also used. The other two types shown in

Pig. 63 have been superseded, as i t is necessary to disconnect the
armature leads from the commutator when a shaft is to be replaced.
Some of the older type shafts have wiper rings screwed on the shaft,
but this practice has been changed to a shrunk-on type of wiper
on the more modem motors.
At (a)in Fig. 63 is shown the construction in which the armature
core and commutator bushing are mounted directly on the shaft and
depend upon the press fit, the key and the lock nut a t the commutator end to keep them in place. The fan and commutator end wiper
ring which have been shrunk on, are removed by heating with
blow torches (keeping the shaft cool by wrapping i t with wet
asbestos), and are driven off with a hammer and chisel. After
the fan and wiper are removed, clamp the core and commutator
together by placing over the shaft two pieces of iron pipe, one on
each end, large enough to clear the armature nut and long enough
so that plates placed over the shaft and against the pipes will clear
the end of the commutator and the entl of the coils a t the pinion
end, as shown in Fig. 64. Ilolt the plates together, using four or
more t)olts just clearing the outsitle of the armature. Take special
care to cut the pipes oB square ant1 to pull the bolts up evenly to
prevent warping the core when the shaft is removed. Allow clear-

ELECTRIC RAILWAY HANDBOOK

ance enough a t the pinion end between the plate and the shaft so
that the key will clear. Another form of clamp has rings machined
to fit the commutator and rear coil su port, with bolts through the
longitudinal vent ducts. When the cLmp is in place, back off the
lock nut a t the commutator end, press out the shaft by applying
pressure a t the commutator end, and replace with a new one.
With the new shaft in the core, apply and tighten the annature nut,
remove the clamp and shrink on the fan and wiper ring.
At ( b ) in Fig. 63 is shown the construction where the armature
core and commutator bushing are mounted on a spider, which is in
turn mounted on the shaft. The commutator end wiper and the
fan are shrunk on. To remove the shaft, take off the pinion end
wiper ring (or fan, if a fan is used) by heating as described above.
W ~ t hthis type it is not necessary to remove the commutator end
wiper, as it comes off with the core. Pressure is applied a t the
commutator end of the shaft in removing. The new shaft is pressed
in and after it is in place, the commutator end and pinion end
wiper or fan are shrunk on. This method applies both to spiders
with separate rear end bells and to spiders with rear end bells
cast integral.
At ( c ) in Fig. 63 the commutator spider and core are shown
mounted directly on the shaft with a nng nut between the commutator and the core. With this type of construction, it is necessary to lift the armature leads out of the commutator neck, remove
the wiper rings and pull the commutator first. The commutator
is provided with tapped holes for bolts to aid in this operation.
After the commutator is removed, clamp the core together, using
a modification of the clamping device shown in Fig. 64, back off
the ring nut, and press out the shaft, applying pressure a t the
commutator end. After the new shaft is in place, apply the ring
nut to secure the core, press on the commutator, shrink on the wiper
rings and reconnect the windings.
At'(d) in Fig. 63 is shown a type in which the nut is a t the pinion
end and there is a shoulder on the shaft between the commutator
and the core. In this case also it is necessary to disconnect the
windings. Remove the wi r rings and pull the commutator as
described above. Clamp t c core together, back off the ring nut
a t the pinion end and press out the shaft, applying pressure a t
the pin~onend. Press In the new shaft and apply the ring nut,
which should be drawn up tight before the clamp is removed.
Replace the commutator and reconnect, then shrink on the wiper
rings.
In connection with the replacing of shafts of railway motor armatures, the following points should be kept in mind: new shafts
should have fillets a t all changes in diameter; new shafts should
be made about o 004 in. larger than the original shaft a t the ress fit,
to insure the proper tonnage in pressing; check clearance getween
top of key and key seat in core to prevent binding a t this point;
chamfer the start of the commutator bore to allow the shaft to
enter straight; shafts should be pressed in a t approximately 2 0 to
25 tons on motors from 25 to 50 h p. and about 40 to 50 tons for
sizes above 50 h.p.; whencvcr possible, the armature nut should be

RAILWAY MOTOR BEARINGS

removed and replaced while the clamps are in place; shafts will
press out a t approximately 1% to z times the tonnage used when
they are pressed in, because of slight rusting and flowing of metal;
the press fit will vary, depending upon the material of the coresteel or malleable iron can safely stand a higher tonnage than
cast iron; a little white lead on the shaft a t the fit before pressure
on acts as a lubricant and prevents rust.

Motor Bearings

Ball Bearings. Among the advantages claimed for the use of

ball bearings, as stated by 0. R. Wikander, probably the one
which has received the greatest amount of attention, is that of
the decrease in starting current, due to the smaller starting friction
of ball bearings over plain types. With properly constructed ball
bearings, the wear a t the bearing should be considerably reduced,
and the danger of the armature wearing down in its bearings so as
to cause rubbing on the field pole faces is reduced. A reduction in
bearing wear might also make it practical to reduce the air ga of
the motors to a considerable extent, which would produce a ligfter
motor for the same power or a more powerful motor of the same
weight. Another advantage claimed for the use of ball bearings
is that, due to the narrow construction a t the bearings, it may be
possible to build the whole motor narrower, so that a motor of
greater capacity could be designed for a certain width between the
wheels. The use of anti-friction bearings also should produce
a considerable saving in the cost of lubrication, since this t
bearing should require lubrication only about every six m o n g i f
stead of weekly, as is the requirement for plain hearings. While some
installations have met with considerable success, others have failed
and have caused great inconvenience. Bearings with insufficient
load-carrying capacity have been applied, together with improper
design of the bearing housing. As a result of this last consideration,
bearings have been damaged while being mounted or when removed
for the purpose of inspection. Other failures have resulted from
stray current passing through the bearings, as ball bearings are
very sensitive to electric current, and even a very slight amount
passing continuously through a ball bearing will cause pitting and
premature destruction. Another difficulty which has been encountered in the use of anti-friction bearings is that due to heavy shocks,
particularly in the case of old and worn tracks. Ball bearings are
quite sensitive to shock loads, due to the fact that their capacity
to carry even temporary overloads is not much larger than their
continuous load-carrying capacity. While proper selection and
mounting may to a considerable extent overcome such failures, and
while several arrangements have been proposed for overcoming
them, still in spite of the great efforts made to design reliable ball
bearings for rallway motors, the margin of safety obtainable is not
quite satisfactory.
Roller Bearings. Properly designed roller bearings possess the
same advantages as ball bearings and the additional ones that
they do not appear to be subject to pitting by electric current to

310

ELECTRIC RAKWAY HANDBOOK

anywhere near the same degree as ball bearings, and that they
are capabie of sustaining comparatively heavy shock loads. I t
can be said conservatively that a roller bearing wrll carry about 60
per cent more load than a ball bearing of the same outside dimen'
sions and stand more than double the shock load. For the abo\le
reasons the results obtained with precision roller bearings in electflc
railway service have been on the whole far more satisfactory than
those obtained with ball bearings. Roller bearings have selddm
been used for electric railway motors in America. The reason for
this is that "precision" roller bearings, as they are termed in G'"
many, have not as yet been marketed to any large extent in tXe
United States. The term precision roller bearings is applied
those which are manufactured of as high grade material as the bcSt
ball bearings, are of equal precision in the workmanship and have
about the same coe5cient of friction. The commercial roller bedrings most widely marketed in America do not measure up to theSe
standards. They have a coe5cient of friction which is from tbV0
to five times larger than that of high class ball bearings and do not
meet the requirements for railway motor service. A descripti?"
of various types of roller and ball bearings, and the results obtain^^
from their use, is given by Mr. Wikander in Elec. Ry. Jous., v?].
60 (19221, P. 935.
Sleeve Bearings. This is the type of bearing most c o m m o n l ~
used. A sleeve or plain bearing can be either a split or solid cylinder
of a hard metal lined with babbitt. or a solit or solid cvlinder
babbitt or bronze material not lined: I t shbuld have ;n&gh ;letrance over the journal to allow a thin film of oil to form between the
journal and the bearing to float the journal. With this type
bearing, the running friction is low if the film of oil is constantly
maintained and is free from dirt, although the starting friction IS
comparatively high. The bearing will last for a number of ye@
if i t is not damaged mechanically.
To allow bearings to be removed without taking off the pinion?
most of the older motors are provided with split armature bearingS
a t the pinion end. Due to improved lubrication, with a resultPg
increased life of bearings, all modem motors use solid bearings at
both ends. This gives a much better mechanical design, which
does not require frequent renewals. The pinion end bearings sre
always larger than those on the commutator end. Axle bearingS
are always made in two halves, so they can be removed con\reniently. The two bearings are of the same size and in modem
t
motors are interchangeable.
T h e most common types of armature bearings are made of either
a bronze or a malleable iron shell lined with babbitt. The babbitt
lining of a bronze shell is less than the single air-gap in thickness, 9
as to save the armature core from rubbing on the pole pieces shodd
there be excessive wear or the babbitt melt out of the shell. l)epending primarily upon size of axle and size of axle bearing seat in tbe
motor frame, axle bearings are made of either bronze tinned Or
malleable iron lined with a soft alloy. Other factors, such as fiFt
cost, cost of repair, and the experience of the operating man, enter
into this selection.

RATLWAY MOTOR BEARINGS

311

Babbitt Metal used to line railway motor bearings consists either

of a tin base or of a lead base alloy; in other words, the bearing metal
should be an alloy composed of a t least 80 to 90 per cent tin or lead.
Both classes give good service if properly handled during the melting and pouring process. Where large numbers of bearings of
the same s i ~ eare used, sufficient to support the expense of special
tools, the surface obtained by broaching is found most satisfactory.
With many operating companies it is the common practice to
machine armature bearings after rebabbitting, although some
consider this unnecessary, and babbitt bearings to exact size.
Whichever method is used, it is customary to scrape the babbitt
to get a uniform bearing surface. Axle bearings, when not lined
with babbitt, are given a machined finish and this surface often
is tinned to fill up the small irregularities, thus helping the bearing
to seat itself while new. When axle bearings are lined with babbitt,
they should be machined to get best results in senice, although i t
is the practice of some operators to babbitt to exact size and fit by
scraping.
Method of Holding. In addition to a press fit of from three to
five tons, keys are commonly used to prevent armature bearings
from turning in the housing. Most of the older motors without
housings have their bearings held from turning by means of a dowel
in addition to the clamping action of the frame. Depending
largely upon the size and location of the lubricating openings, axle
bearings are held from turning by a dowel or a key, as well as by the
clamping fit of the axle caps. Dowels in the flange or shoulder
of the bearing shell are being used successfully in modern motors.
Special forms of lugs cast on the bearings, or plates inserted between
the two halves of the bearings, have given satisfactory results in
service, but are difficult to manufacture.
Clearanaes which are considered good practice for railway motors
using grease or oil waste lubrication, are:
Prom
Prom
From
From
Prom

2 ~ nu
. p t o and not rnclud~ng3 ~n
3 In. u p t o and not lncludlng 4 In
4 ~ n up
. to and not ~ n c l u d ~ n5g~n

5 ~ n u. p t o and not ~ n c l u d ~ n6g~n

6 ~ n up
. t o and not ~nclitdlng7 ln

Mlnimum
o 006 ~ n .
o 008 ~ n .
o O I O In.
o 0 1 4 In
o 016 ~n

Maumum
o 008 ln.
o oro ~ n .
o 014 In.
o 016 ~ n .
o 018 ~ n .

Oil grooves are machined, cut or moulded in armature bearings

so as to help the oil to enter and more evenly distribute Itself
throughout the bearings. I n general they are not required in the
case of axle bearings, on account of the slow s ed and low unit
pressure. The sharp edges a t the split of axle c a r i n g s should be
beveled to prevent wiping away of the oil film by these edges, but
this bevel should not extend to the ends of bearings, as it would
drain off and waste too much oil. I t is good p r a c t i ~ eto round off
or bevel the edges of the windows in both armature and axle bearings to encourage the flow of oil into the bearing surface.
Bearing wear, especially that of armature bearings, should be
carefully checked by periodic inspection to avoid armature rubbing
on pole faces. Such inspection, sometimes made visually, is more
satisfactorily and safely done by means of sweep gages inserted

312

ELECTRIC RAILWAY HANDBOOK

between armature and pole face. Where the direction of rotation,

especially with single-end cars, is such that the bearing pressure
is up, means must be provided to lift the armature so as to measure
the minimum air gap a t the top, as well as under the armature.
This may be done mechanically or by the application of power on
the first controller notch with the brakes set.
Lubrication of Motor Bearings.
Grease. Iieavy grease, stored in a box or cup over the bearing,
melts and runs on the journal as heat is developed by the friction
of the bearings.
Grease and Oil. Same as above, with the addition of an oil well
located under the bearing, from which oil is fed up to the bearing
by means of a felt wick.
Oii and Rinas. Oil well located directly under the bearing, in
which a small brass ring runs suspended on-the journal, and c&es
the oil to the bearing.
Oil and Waste. Oil well below and to one side of bearing, packed
with saturated waste which presses against the journal through a
window.
Vaseline Packed. The entire bearing packed in vaseline. Used
mostly in the case of anti-friction (ball and roller) bearings.
Circrrlating Oil. Where the oil is forced through the bearing by
means of a small pump.
Special Adaplions. Small oil cups placed in the grease box,
from which wicks are suspended down to the top of the journal.
Joggle Type. Where a grease cup is used as an oil well and a
metal ball or pin is placed a t the opening of the journal.
The older types of railway motors used the grease method, later
types grease and oil, and modem types, oil and waste.
Oil and Waste Lubrication. Use a good grade of mineral oil,
light oil in winter and heavy oil for summer use. For best results,
use a long fiber wool waste. Before using, it should be saturated
in oil for a t least 24 hours, and left on a screen or grating to drain
for several hours. The oil wells should be of ample capacity to
hold sufficient oil to last between inspection periods, should have
an accessible opening for inspection and refilling, and should be prot
cover, held in place by a strong spring or
vided with a t ~ g h fitting
bolts to keep out water, dirt or grit. Provision should be made for
proper drainage of the spent oil, and means provided to gage the
depth of oil a t regular inspection periods. Before the bearings are
packed, all water, dirt and small particles of metal should he removed
from the oil well. Saturated waste should be loosely packed
in the oil chamber and forced into place by a pronged rod of brass
or some other soft metal, so that i t will not injure the journal. I n
this manner, pack the waste up over the bearing window, forcing
i t in place, so that its springy action tends to hold the waste against
the journal. Well-designed bearings of this type, if in good condition, pro erly acked with a long fiber wool waste and using a
good grafe of oi[ should run from one to three weeks between oilings.
The time is determined largely by the system of inspection of the
other equipment on the cars, which makes it advisable for each
operator to work out in actual service the most suitable oiling sched-

RAILWAY MOTOR LUBRICATION

315

ule to fit the operating conditions of his equipment. One to two

gills of oil per bearing are required a t each o ~ l ~ nperiod.
g
This
varies with the size of motor and the location of the bearings and
upon the length of time between oiling periods and the service
conditions. I t is considered good practice to repack the bearings
every three months, a t this time removing all the waste, discarding
that which is glazed and charred, refilling the bearings with good,
clean, old waste, to which has been added sufficient new waste.
About once a month it is advisable to "tease up" the waste in the
bearings to make it more effective.
Grease is used in connection with the lubrication of some of the
old type railway motor bearings. Grease consists of a fatty soap
impregnated with a mineral oil. The solid part acts as a carrier
for holding the oil in position, and has little value as a lubricant.
Greases are usually graded according to their stiffness, from the
softest to the hardest, with numbers indicating their relative melting
point. A limesoap grease with a neutral reaction (which does not
show any traces of an acid or an alkali) gives best results in service.
To insure its being retained in the bearing, the grease should have a
melting point of 10 to 15 deg. C. above the normal operating
temperature.

FIG.65.-Rico

oiler for railway motors.

Adaptation of Oil Lubrication to Old Type Motom Orig&lly

Designed for Grease. A number of devices have been designed
for this purpose, among which is the " Rico" oiler, as shown by Fig.
65. The reservoir, 3, in. diameter and 10 in. long, with a capacity
of approximately 8 g~lls,is provided with a removable filler plug a t
the top. A bushing extends entirely through the reservoir for the
anchor bolt. In the interior of the reservoir are two tubes, one
inside of the other. The larger, or control tube, extends from a
point above the bottom of the reservoir up through the top, and is

314

ELECTRIC RAILWAY HANDBOOK

closed a t the upper end by a cap. The smaller, or feeder tube,

extends from a point near the center of the reservoir down through
the bottom. A vent tube connects between the control tube and
~
the anchor-bolt bushing. The feeder wick is inserted in t h control
tube, the short leg extending to the bottom of the reservoir, with
the long leg inside of the feeder tube and projecting out the lower
end. After filling the reservoir, inserting the filler plug seals the
interior against the atmosphere. The oil feed is regulated by the
number of strands in the feeder wick, different styles being used
according to the rate of feed desired.

Fro. 66.-Nose

suspension

of motor

FRONT VIEW OF CRADLB

FIG. 67 -Cradle suspension of motor.
Motor Suspension on Cars. Nose .Suspension (Fig. 66) is the

most common method of motor suspension on motor cars. Springs

in the motor nose suspension lessen shocks during starting or sudden
changes of torque. Approximately 60 per cent of the weight of
the motor is carried d~rectlyon the axles without spring support.

RAILWAY MOTOR SUSPENSION

315

Cradle Suspension (Fig. 67). The total weight of the motor

is hung by lugs on either side from a longitudinal horizontal bar
which a t the back end is spring supported from lugs on the arm
which carries the axle bearing and a t the front end by a cross beam
and the truck frame. This type of suspension is seldom used.
Spring Supports. The motor manufacturers do not ordinarily
furnish suspension parts other than the bracket or nose casting on
the niotor frame, and the suspension problem usually has been
dependent on solution by the truck manufacturers. A resilient
suspension for motors is very desirable, to relieve shocks and hammer blows; too short a spring gives too ngid a suspension. The
truck suspension should be so designed that the motors are reasonably free to move transversely with the axle; otherwise axle thrust
collars will wear rapidly and there will not be freedom of movement
between the motor and the truck frame. The suspension also
should be designed with as few wearing parts and of as simple design
as ssible, so that the motor can be removed from the truck
witrminimum labor. I t is generally conceded that three-point
suspension presents certain advantages over the other forms, but
unfortunately truck space frequently prevents its use.

Compress~onSprings
S e c t ~ o nA-A

FIG. 68.-Improved type? of motor s rlng suspension proposed by

A.E.R.E.A. Eqrupment Eornrnlttce. 1923.

316

ELECTRIC RAILWAY HANDBOOK

The 1923 Committee on Equipment, Am. El. Ry. Eng. Assn.,

states that the springs should be so designed that they will compress
1 in with the following loads and should compress 5 in. more
before going solid, thus allowing for a 50 per cent overload; this
allows a normal movement of I in and a total movement of 1%
in.
Motor. h.p.
25

35
40
50
65

Load In pounds

Downwdrd
2000
3000
3500
4000
5000

Upward
1200
1.500
1800
2000
2500

It has been found in practice that many suspensions, originally

designed to have from I in. to 1% in. motion, have been screwed
down by the repair men until they have practically no motion.
This condition could be avoided if shoulder bolts were used. Some
operators have removed the spring suspensions entirely, due to
excessive spring breakage, but very probably a t the expense of
commutation and excessive maintenance.
The Committee presented three types of motor suspension spring
arrangement, as shown in Fig 68, in each of which proper allowance has been made for lateral motion without spring distortion.
It is believed that while these designs or modifications of them may
cost slightly more than those then in common use, the increased
cost would be offset by reduced maintenance cost within a short
time.

SECTION V
CONTROLLING APPARATUS
Controllers for Direct Current Series Motors
Rheostatic Control. Rheostatic control is carried out entirely by
making changes in the res~stanceof the motor circuit to maintain
the desired current through the motor during acceleration. I n this
method there is no arrangement for series-paralleling. Rheostatic
control may be used with equipments of one or more motors, but
because of the greater energy economy secured by series-paralleling
in two or four motor equipments, the use of rheostatic control is
practically limited to single-motor equipments or two-motor
equipments on double voltage. The hand controller commonly
used to make the necessary connections for rheostatic control is
known by its trade name, the type " R" controller.
Series-parallel Control. The series-parallel method of control
is the method most commonly used with two- and four-motor e q u i p
ments. I n starting a two-motor equipment by this method the two
motors are first connected in series and in series with the control
resistance, then the control resistance is reduced by steps until
the motors are running in series on the working conductor potential;
control resistance is then added as the motors are connected
in parallel, after which the control resistance is reduced by steps
until the motors are running in parallel on the working conductor
potential. Four-motor equipments are usually arranged in two
groups of two motors ~ermanently connected in parallel with
each other and these two groups are controlled as the two motors of
a two-motor equipment. There are two general methods of making
the transition from series to parallel~connectionsin series-parallel
control, one opening
- the power circuit, while the other leaves i t
closed.
Type "K" Controller. The type " K" controller does not open
the power circuit during transition from series to parallel. This
type of controller is in most common use. The type " K " controller
is arranged to cut the current off half the motors during transition.
This is done by first shunting this half, then disconnecting i t
before placing the halves in parallel, and is known as the "shunt"
type of transition. Some forms of the type "K" controller were
designed for the use of the "bridging" type of transition from
series to parallel, but these have been superseded by those using
the "shunt" transition
Type '2"Controller. The type "L" controller opens the power
circult during transition from series to parallel connections of the
motors. This type of controller is little used.

317

318

ELECTRIC RAILWAY HANDBOOK

TYPES OF CONTROL

319

Type "BB"
Controller. The type "B' controller has the usual
power connections and in addition it is arranged t o make connections for electrical braking by operating the motors as series
generators to excite the electrically operated brakes. This method
of electric braking imposes an
additional load upon the motors,
thus necessitating larger motors.
S h ~Transition.
t
There are
G
three transition steps between
series and parallel, as shown in
Fig. 2. On the first transition
step, part of the resistance is
inserted in the circuit with the
motors still in series. On the
second transition step, one of
the motors is short-circuited or A Tr;;;hnhnnnto----.-~
shunted, which gives the name
G
to this method. On the third z B
transition step, the shunted mo- %
G
tor is disconnected from the GCT
,+,.-G
other motor and theshunt. On
G
the second and third transition
cCOIJ*CG
points, one motor is developmg no torque, but the other is
G
~
TJ
still working. Therefore, this
method has an operatingadvanJ
GLDE*CG
tage as compared to the open
circuit method in that only one
G
G
blT
of lhe torque is
F~.'2-nnkns
during transition instead of all type K controller, shunted motor
torque being lost. Series-par- transltlon.
alleling by this method is used
with most platform controllers of the K type and with unit switch
control except where automatic operation is required or the current
handled is comparatively large.
Bridging Transition. Figs. 3 and 4 illustrate diagrammatically
the sequence of connections in the bridging
method of control. 'This method has the advantage that both motors (or pairs of motors)
are developing torque a t all times through the
transition period. I t has been used in some
type X platform controllers, but on account of
heavy arcing a t controller fingers under certain
conditions in transition, its use has been
discontinued in platform controllers, and is
FIG. 3 -Dlagrarn
now confined to unit switch control where
ofbrldmngmethod of automatic operation is desired or heavy curcontrol.
rent is handled.
Three-speed Control. Fig. 5 illustrates the various combinrrtions of the three-speed motor control which was designed by P. N.
Jones of the Pittsburgh Railways Co. primarily for use with low
floor cars and small motors. The amount of rcsistance required is

Tv

J
~

320

ELECTRIC RAILWAY HANDBOOK

Sequence of ConnertlonB,

FIG. 4.-Cirniits

of bridging typr of contnbl.

y
"
1

9"
FIG.5.-Diagrams

of circuits. three-speed four-motor control,

reduced to a minimum by the use of four motors in series

first two controller points, and the special connections shownon the
in Fig.
5 for the tilth, sisth and seventh points.
Resistance Co~ections. There are three general methods
necting control resistance in the motor circuit, the series, the
and the combination method. I n the series method (Fig. 6!ara11e1*
the
resistance is connected in series a t the beginning and i t is
of the circuit by short-circuiting sections of it, progressivelJut Out
this method much radiating surface is made available r'. By
starting. In the parallel method (Fig. 7) the resistance in during
is reduced by addlng resistor sections in parallel, progressivcircuit
the sections already in circuit, thus reducing the total resirtie$;

REVERSING SERIES MOTORS

821

the circuit and finally short-circlriting all the resistance. This

requires smaller switches than the serirs method. 17ig. 8 is an
illustration of a combination of these two methods.

L l ~ p u e n c e01 colruectionli

Cluae l.b.c.d,r,

couseculllcly

P I G .6.-Series method of conlicrtlng control resistance.

PIG. 8.-Series-parallel

Sepueuee of conuectlonll
Close a;

Close b,c.d.e.l. cou~ecuQvnly

PIC;.7.-P.~rnllel
mrthod o l
ca~nnrctinl:ctbntrol resistnncc.

method of connecting control rr$.stance.

Controller Rating. The rating of a controller is based upon the

combined nominal I-hour rating of the motors it is to control.
This rating is generally given for a certain voltage and for a voltage
variation from this value, but within the insulation and arc-breaking
capacity of the controller, the horse-power capacity of the controller
will vary approximately as the voltage.
Reversing Series Motors. The direction of rotation of a series
motor may be reversed by inverting the relation of the direction
of the current in the armature to the direction of the current in the
field. This may be done by reversing the connections to the armature or by reversing the connections to the field. Older types of
controllers were arranged to reverse the conncctions to the brushes,
but controllers for use with commutating pole motors are arranged
to rc\.erse the connections to the field and leave the interpole
winding permanently connected on the ground side of the armature,
thus securing minimum insulation strains on the interpole winding.
Controllers built to reverse the connections to the brushes can be
arranged to reverse the connections to the lield.
In order to simplify the control with some types of field-control
equipments, the main field is permanently connected to ground and
the armature and commutatlng field are reversed as a unit, thus
throwing the commutating-pole winding next to the trolley.
With modem insulation and construction this is permissible,
a l t h o u ~ hless desirable than to retain the armature always next to
the trolley, in its relation to the commutating field.
High Voltage Direct Current Control. The control for I 200-volt,
1500-volt, 2400-volt and 3 m v o l t car equipments is essentially the

322

ELECTRIC RAILWAY I1ANDBOOK

DRUM TYPE CONTROLLERS

323

same as that dcsrribed above for 600-volt service e\cep,t that two
ly
pcrmancntly in series. Ser~es-parallcl
motors .we u s ~ ~ . ~c lunnccted
conncctlon oi thc two groups of motors provides for hall specd ancl
full speed. but when full srzcd is destred on 60evolt supply, a
cornmutating switch is p r o ~ i d e dwith the 12-volt
cquipmcnts
which operates to change the permanent connection of the motors
from two in series to two in parallel. Usually, however, two motors
remain permancntly in serles and provide half speed when 1200volt car5 run over terminal tracks supplied with energy a t 600 volts.
A dynamometer or motor-generator set run from the trolley supplies
current at a low voltage for actuating the switches. The air
compressor and heaters operate direct from the trolley circuit.

Drum Type Controllers. There are in use a very large number of

forms of drum type contmllers, varying in minor details of circuits.
The general types, with capacities, number of control points, and
weights, are as shown in the following tables, Minor variations
in type are designated by a sub-letter following the type Ictter and
number as "K-39-C." Unless otherwise noted, the ratings given
are for 600 volts, and the maximum peaks should be limited to 7 5 0
volts. Type U controllers are for electric braking scrvlce; type R
are rheostatic contn)l.

Type

1 1
N~ of

Max allowable capacltres, each

motor (nerthtr to be exceeded)
Houri rat~ng. Cont~nuous
ratlng amp

Number of pornts
Series

Appror.

Remarks

K-28

40

hletsll~creturn ctrcult.

240

Metallrc return c ~ r c u t .
K-35
4
4

70
65
75

66
60
65

4
5
6

4
3
4

230
a80
437

K-5 I

70

66

250

hfetallrc return crrcult

Metallrc return crrcurt
Two motors permanently ~n w l e s
for 1200 volts
For-tapped field motors

K-63
K-64
K-67

a
4
8

40
110
40

38
I05
37

4
6
5

3
4
3

I35
450
270

For tram o p r a t ~ o nfum~shed

.
nth
lrne breaker and separate motor

K-68

70

66

K-39
K-40
K-47

""I

K-52
B-50

'-"
B-54

"-"I

R-zoo

I.

I
4

11

$
60

so*

I1
1

:f
53

;2:

* Ratrng a t 500 volts. maxrmum peaks of 600 volts.

pub of 17s voltr.

t Ratlng a t aso voltrf -mum

II

4
4
5

II

11

For storage battery servlce.

For storage battery semce.

492

1:
245
I"

' :8"k:E:%:.
9 brakrng pornts

TWO

motors permanently rn senes.

$ Ratlng at 750 volts, maxrmum peaks of 1650 volts

All others rated s t 600 volts, maxmum peaks of 750 volts.

325

DRUM TYPE CONTROLLERS

mmm

mmm

mmm

mmo

mom

xx xmmm

nnn

xA*x
nnn

&nnn
*A%* A\*x
m n n A*A*x
nnn

tnn

00

O F 0

0 0 0

O d d

dad

-0c4

X l f

ST:^

5s
Qmn

- r -

rrr
r - w

+ma

&x& $A&

b b t

:;hO

51 ;
:
r 55

:om

YZX
: *A
cmo om-

:am

r - r

-#em

o o n

Cro

c - n

? p p g':?

X
xxx

o_oo

A
*v
n%

-3

,,I

-\-a,

2:

m r o

S
x x 1 ~ xrx
n m r n n m nn m r m

nr -

n r r

0??

.E

'""

Ghc 2
r r c

03*

mmo

r n r

o+r

???
x~ x x

n n

or

nn

dd

r - r

- -

e r o

?.;.?
rndg

326

ELECTRIC RAILWAY HANDBOOK

A*
Line Switch Equipment. Fig. 15 shows the circuits of
an auxiliary Ilne switch (or contactor) equipment. I t consists of a
ratchet attachment in the controller, combined switch and fuse for
protecting the line switch operating circuit, and a box containing
a line switch (pneumatically or magnetically operated) with the
over\oad relay and necessary resistance tubes. The energy for
operating the magnet valve of a pneumatic line switch, or the electromagnet line switch, is obtained from a n auxiliary circuit carried
through the contactsof the ratchet attachment in the drum controller and the control switch. This auxiliary circuit is called the control circuit. The pneumatic line switch is operated by compressed
air taken from the main reservoir of the air brake system on the

FIG. IS.-Circuits

of auxiliary line switch equipment.

car through a cut-out cock and strainer. The admission and release
of air from the line switch air cylinder is accomplished by a magnetically operated valve. The ratchet attachment which is usually
mounted near the lower bearing of the main drum operates to close
the control circuit contacts when the controller handle is placed on
the first notch. The auxiliary contacts then remain closed throughout all succeeding positions of the controller handle. If, however,
the controller handle is moved backward a n appreciable amount,
the ratchet opens the control circuit to the line switch operating
coil which causes the line switch to interrupt the main motor circuit.
I n order to close the line switch again, i t is necessary to move the
controller handle to the "off" position and then advance to the first
notch. I n case of an overload, or a surge due to a short circuit,
the overload trip will operate to open the control circuit, thus opening the line switch and cutting off power. I n order to re-establish
the power circuit, it is necessary to return the controller handle t o
the offposition and then move to the fust notch. The control

STARTING RESISTANCE DESIGN

327

circuit then is automatically restored and the switch comes into

normal operation. The regular operation of the line switch during
each service application of power insures that it is always in good
working order, and ready to operate in case of overload or short
circuit. This auxiliary equipment accomplishes the purpose of
making and breaking the main circuit in a properly designed switch
outside of the controller and underneath the car.
Circuit-Breakers. T o furnish protection against extreme overloads or grounds and short circuits, a n adjustable circuit-breaker
is connected between the trolley and controller. Circuit-breakers
are rated according to the current which they will carry continuously, and for any specific equipment are selected so that the
current rating of the breaker is approximately two-thirds of the
nominal one-hour current rating of one motor multiplied by
the number of motors in parallel in the equipment. The breaker
usually is set to trip a t a current 50 per cent greater than the
nominal current rating of one motor multiplied by the number of
motors in parallel per equipment. For example: a 600-volt car
equipped with four motors each rating 100amperes for one hour,
should have a circuit-breaker with a continuous rating of % X roo
X 4 = 267 amp. Ordinarily this breaker should be set to trip
a t 1.5 X loo X 4 = 600 amp. On locomotives it is customary
to set thc breaker to trip a t double full load on the motors. I n
the above case this would be 2 X roo X 4 = 800 amp.
Lightning Arresters. For protection against lightning and surges
which occur in the contact line (trolley or third rail), a lightning
arrester is connected next to the collector and a choke roil is connected in the power circuit between the liqhtning arrester tap and
the circuit-hreaker. If lightning or a line surge tends to produce a n
abnormal voltage a t thc car, which in turn tends to send a heavy
rush of current through the equipment, the choke coil limits the
current rush and piles up the voltage next to the arrester until the
Current finally jumps to ground through the arrester, the tension is
relieved, and n o m a l conditions are re-established.
Design of Starting Resistance for Direct Current Motors
The design of direct current railway accelerating resistance
involves three distinct problems: ( I ) determination of the total
amount of resistance required in the circuit when the controller
is on the first notch, ( 2 ) determination of the number of steps (and
their ohmic values) into which the resistance must be div~dedto
give uniform acceleration, and (3) determination of the current
carrying capacity of the various steps, necessary to prevent the
overheating of the resistance elements.
The total resistance required in the circztif at start, and the amount
of resistance that is cut out a t any given instant, depend upon the
permissible rate of acceleration. Assuming this it is possible to
determine the total tractive effort which must be exerted by the
motors. Dividing this total tractive effort by the number of motors
gives the tractive effort which must be exerted by each motor.
Knowing the gear ratio and wheel diameter, the value of the current
which will produce the above tractive effort can be obtaincd from

328

ELECTRIC RAILWAY HANDBOOK

the characteristic curve of the motor. Then, by means of the

line voltage, the total resistance can be determined. The value of
the resistance obtained by this method must be reduced by a factor
which varies from 0.7 to 0.8, i.e., the resistance must be multiplied
by a p roximately 0.75, to give the value of resistance which is actually &aced in the circuit. The necessity for using this factor is
due In general to two causes: first, the inductance of the motors,
and second, the regulation of the transmission system. By following the method outlined and by using a constant of 0.75, the amount
of resistance which is required in the circuit a t start can be determined, and then by subtracting the motor resistance (see page
252) from the total the external resistance which is required is
obtained.
Current Fluctuations During Straight Line Acceleration Period.
When resistance is cut out of the motor circuit in passing from one
controller point to the next,
there is a sudden increase in
current, resulting in a n increase of tractive effort and
a corresponding sudden in17'
crease in rate of acceleration.
(See Fig. 16.) This is evidenced by a jerk the violence
of which depends upon the
magnitude of current fluctuation, train resistance and inv
I 'A
ertia and the tractive effort
1 3 4 5
characteristic of the motor.
Po in l s
Excessive sudden increase in
PIG. 16.-Fluctuation of motor current rate of acceleration may cause
on silccesslve controller pants.
discomfort to passengers, i t
may produce too great mechanical strains in trucks, draw bars or
gears, or it may cause driving wheels to slip. The increase in rate
of acceleration between controller points should be kept below
about 0.6 mile per hour per second in passenger service, and still
lower in freight locomotive operation. For ordinary conditions of
city service, the corresponding maximum and minimum accelerating currents will be 10 to 15 per cent above and below the average,
with total current fluctuations of 2 2 to 33 per cent. The degree
of perfection to which the acceleration of a train is accomplished
depends upon the number of steps into which the starting res~stance
is divided, the proportioning of these steps and the speed and degree
of uniformity with which the steps are cut out. The proportioning
of the resistance steps demands a study of equipment and conditions,
as discussed below. The speed and degree of uniformity with
which the steps are cut out are cared for automatically in the
case of automatic control, but in the case of hand-operated
control, except where an auxiliary regulating dcvice is attached
to the controller, they depend upon the skill and efforts of the
motorman.
Analytical Method of Calculation of Motor StartingResistance
Steps. The following simple analytical method of determining

s'

STARTING RESISTANCE DESIGN

329

motor starting resistances requires as preliminary data only the

tractive effort-ampere curve and the resistance of the motor.
No assumed constants are used, and the formulas may be applied
to any type of series motor.
Let I I = current per motor a t instant controller makes contact
a t any point, amperes (maximum current)
I z = current per motor a t instant controller breaks contact
a t any point, amperes (minimum current)
T I = tractive effort per motor corresponding to I I , pounds
Tt = tractive effort per motor corresponding to I I , pounds

b = -TI11
T J,
E = line potential, volts
r = resistance of each motor, ohms (see page 2 5 2 )
RI = total resistance of circuit a t first series point of controller, ohms

R, = total resistance of circuit a t last series point of controller, ohms

R.

= total resistance of circuit a t first parallel point of

controller, ohms
total resistance of circuit a t last parallel point of
controller, ohms
r , = total external resistance in resistors a t first series
point of controller, ohms
ri = total external resistance in resistors a t last series point
of controllers, ohms ( = o)
ro = total external resistance in resistors a t first parallel
point of controIler, ohms
r, = total external resistance in resistors a t last parallel
point of controller, ohms ( = o)
s = number of series points on controller
p = number of parallel points on controller
Formulas for two motors in series-parallel connection:
Series connection

R,

RI =
R)

RI

E
E - b(E - I z R I )
I1
B - b(E - J2R2)
I1

Continue until
E - b(B

I?*

- 12K.-1)

I1

Parallel ronneclion

rl = Rl -

21

2r

R, -

2r

= R1
r5

= 21

ancl r, = R.

- sr

=o

330

ELECTRIC RAILWAY HANDBOOK

Continue until
E - b(E - 2ItRp-I)
-- I- and r, = R, - = o
R, =
211
2
As an example, it is desired to determine the starting resistances
to be used for an 18 ton car, with two 40 h.p. 600-volt railway
motors each having a resistance of 0.60 ohm, and series-parallel
controllers having five series and four parallel points. The maximum current per motor in any particular case depends upon the
weight of the car and load, the desired acceleration and the allowable motor current. The difference between the maximum and
minimum values of current is determined by the number of controller steps, which in turn depends on the allowable variation from
the mean acceleration. As a trial, take 60 and 79.5 amp. as the
minimum and maximum currents. The corresponding values of
tractive effort per motor, as read from the characteristic curve,
are 1000 and 1460 Ib., respectively, which will produce accelerations of 0.99 and 1.50 m p.h.p.s., respectively. (See formula,
Page 151 ) The change between steps, amounting to 0.51 m.p.h.p.s.,
IS found to be well within the limit of 0.6 m.p.h.p.s. mentioned
on page 328.
Series
TI12 = ------1460 X 60.0 =
Then b = -

T J I 1E
600

X 79.5

1.102

R I = E = G - - 7.547 ohms
rl = R I - 2r = 7.547 - 2 X 0.60 = 6.347 ohms
Ra = E - b ( E - I ~ R I-) 600 - 1.102(600 - 60 X 7.547)
I1
79.5
= 5.507 ohms
ra = Rz - 21 = 5.507 - 2 X 0.60 = 4.307 o h m
F - b ( E - ItRz) = 600 - 1 . 1 0 2 ( 6 ~ 60 X 5.507)
RI I1
79.5
= 3.810 ohms
r, = R I - 2 1 = 3.810 z X 0.60 = 2.610 ohms
E - b ( E IzRa) = 600 - 1.102(600 - 60 X 3.810)
R , = ------I1
79.5
. -= 2.399 0 h m ~
r, = R , - 2 1 = 2.399 - 2 X 0.60 = 1.199 ohms
E - G(E - I t R d - 600 - 1.105*(600 - 60 X ~ . ~ 9 9 )
Rr =
I1
no
= 1 . 2 0 0 ohms
r , = R , - 2r = 1.200 - 2 X 0.60 = o o ohm
Here X I
80 amp.. the maumum current to satisfy the condrtron rc
0.

'---

STARTING RESTSTANCE DESIGN

33 1

1.944 ohms
0.60
1.944 - - = 1.644 ohms
E - b ( E - 2ZZRo)= 600 - 1.102(600 - 2 X 60 X 1.944)

r. = R. -

Ip, =
2

211

79.5

= 1.232 ohms
= 1&

,- R
8 -

rd =

Here Is

'

0.60

- - = 0.932 ohm

= 1.232

-'= 0.640

0.60
- = 0.340 ohm

0.300 ohm
Rd -

= 0.300

0.60
-=

0.0

ohm

72.2 amp the maumum current to aatssfy the eon&tson rr

a.

If the equations for R. and R,, when using the assumed values
for I, and 11, do not equal

27 and

:I

respectively, or approach those

values within reasonable limits, it will be necessary to make new

assumptions as to limiting currents and recalculate. These limits
are not necessarily the same in series as in parallel, but the average
current must be kept the same if i t is desired to retain a uniform
rate of straight-line acceleration.
Following this calculation, it usually will be necessary to make
some compromise adjustments in the external resistance actually
to be used, first, to fit standard resistance grids, and second, when
required by the design of the controller to use the same leads for
both series and parallel points.
The foregoing applies, as stated, to a single pair of motors.
Should the equipment consist of four motors in palrs permanently
connected in parallel, or any other number or combination of motors,
the values for current, tractive effort and motor resistance must be
modiiled accordingly
Calculation of Starting Resistance Steps by Graphical Method.
The following graphical method of determining starting resistances
follows the same principles as the foregoing, and is shown in detail
bv Prof. A M. Buck in Bulletin No w of the University of Illinois.
where a proof of its correctness is demonstrated.

332

ELECTRIC RAILWAY HANDBOOK

AS in the foregoing example, it is desired to determine the

starting resistances to be used for an 18 ton car, with two 40 h p.
6-volt
railway motors each having a IesBtance of 0.60 ohm
series-parallel controllers having five senes and four parallel pints.
The maximum current per motor in any particular case depends
upon the weight of the car and load, the desired acceleration and
the allowable motor current. The difference between the maximum and minimum values of current is determined by the
number of controller steps, which jn turn depends on the allowable
variation from the mean acceleration. As a tnal, take 60 and 79.5

FIG. 17.-Graphical

method of calculatiog starting resictance steps.

amp. as the minimum and maximum currents. The corresponding

values of tractive effort per motor, as read from the chalgcteristic
curve, are ~oooand 1460 Ib., respectively, which will produce accelerations of 0.99 and 1 . 5 0 rn.p.h.p.s., respectively. (See formula,
page 151.) The change between steps, amounting too.51 m.p.h.p.s.,
is found to be well within the limit of 0.6 m.p.h.p.s. mentioned
on page 3 28.
Referring to Fig. 17, draw a diagram between motor amperes and
motor volts to a scale of not more than 50 volts per inch and in
about the proportions shown. If the line e m f . is 600 volts, then
when the two motors are in series, each will be taking 300 volts, less
the resistance drop. Draw the Lines S R and P T so that the ordi-

STARTING RESISTANCE DESIGN

333

nates RV and TU represent the IR drop in one motor a t any current

I. Here RV and TU = I R = 79.5 X 0.60 = 47.7 volts. Draw
the line S A to represent the I R drop per motor when the external
resistance is so chosen as to bring the motors (in series) to a standstill a t 79.5 amp. When the current has fallen to 60 amp., the I R
drop in one motor is represented by the ordinate NO, the drop in
its external resistance by LN, and its counter-e.m.f. by KL. If
the resistance is then reduced so as to increase the current to 79.5
amp., the counter-e.m.f. will be increased from KL to AB. To
determine the location of the point B and succeeding similar points,
plot to any convenient scale the points Y and W, which are the
values of tractive ejmt per ampere for the particular values of current
chosen. The tractive efforts are read from the characteristic curves
of the motor under consideration. In this case the tractive effort
per motor a t 60.0 amp. is found to be 1000 lb. and a t 79.5 amp.
1460 Ib., and the corresponding tractive efforts per ampere are
16.66 and 18.36, as plotted at Y and W, respectively, on the 60
amp. and 79.5 am . ordinates. The straight Iine WYX is drawn,
through W and
intersecting the current axis prolonged a t X,
131 amp. to the left of zero. The intersection has been omitted
in Fig. 17 to save space. The location of this point may be determined accurately by the use of the following formula:

f:

Where OX = distance to left of zero, in amperes

T , = tractive effort per ampere for minimum current 1,
Tt = tractive effort per ampere for maximum current It
(16.66 x $.I) - (18.36 X 60.0) =
In this case OX =
Pel18.36 16.66
If then the straight line XLB is drawn through X and L, intersecting AU at B, the ordinate A B will represent the counter-e.m.f.
developed when the current has been Increased from 60 amp.
t o 79.5 amp. without changing the speed. The ordinate YR gives
the ZR drop in one motor, and BR the drop external to this motor;
the latter, divided by the current, determines the new value of
external resistance per motor. The IR drop will then decrease
along the line BS as the current falls off, until, at point M, the current must be increased again. The same construction is repeated
until the two motors are in series without resistance. The current
which will be obtained when the last point of resistance is cut out
may be readily determined, since the ZR drop in the motor alone
is plotted as SR. When the last line radiating from X is drawn,
it will intersect this line a t some point as E. The abscissa determines the current, which, in this case, is 80.0 amp.
In changing to parallel, it is only necessary to move the axis of
reference for I R drop to P, the ordinate for 600 volts, andcontinue
the construction from that point in exactly the same manner as for
series. The diagram, having been so constructed, may be used
as follows to calculate the proper external resistances for the
equipment:

334

ELECTRIC RAILWAY HANDBOOK

. Prom Prg.

3
4
5

a
6

7
8
9

D
I3
F
G
H
J

o
81
149
205
252
291
404
498
557

I7

300
219
151
95
48
309
196
102
43

Total rcsrstance = I R drop

Resrstance per motor

795
79 5
79 5
795
800
79 5
79 5
79 5
720

3.77
2.76
1.90
1.20
0.60
3 89
2 4i
1 28
060

060
o 60
0 60
060
060
o 60
0 60
o 60
060

317
a 16
I 30
060
0 0
3 29
r 87
o 68
0 0

634
4 32
2 60
120
0 0
I 64
o 93
o 34
0 0

- current

The diagram may be used for any value of linc potential without
any further change than shifting the origins for the I R drop. For
different current limits i t is necessary to take ncw points of tractive
effort per ampere, thus getting a new location for X, but a small
variation in current limits may be made wlthout relocating this
point, and the error will not be great. As the resistors must be
used both for the series and the parallel connections, a certain
amount of adjustment usually must be made in the values of resistance found for definite current limits. The method of doing this is
to continue the ZR drop line corresponding to the actual resistance
until it intersects the corresponding dash line originating a t X.
The limiting current values wi 1 then, of course, not be those selected
a t the outset, but will be as indicated by such intersections.
Should the equipment consist of four motors in pairs permanently
connected in parallel, or any other number or combination of
motors, i t is obvious that the values for current, tractive effort
and resistance must be modified accordingly.
Starting Resistance with K Controllers. The following table
shows the division of resistance as commonly used with various K
type controllers:
DIVISIONOF TOTALEXTERNALRESISTANCE,
K CONTROLLERS
(Per cent on each step)

R7 R 8
R 8 -RY
Rg R I O
Rro R I I

335

GRID RESISTANCES

Current Capacities for Resistance Grids. Having obtained a

set of values for the resistance steps, the linal operation is to determine the current-carrying capacittes for these steps. Knowing the
time spent on each controller point, the current, and the time for
each complete cycle of the speed time curve, it is possible to determine the root mean square value of the current flowing through each
step of the resistance. The resistance element that is used should
!x able to carry this value of current continuously without overheatIng. Also it should have suffxient thermal capacity to withstand
the peak loads without dangerous heating. The capacities obtained
by this method are safe only on the assumption that the acceleration is automatic. Sf the equipments are accelerated by hand, the
capacities should be increased, the amount of increase depending
upon the type of service I t is obvious that the first grid to be cut
out can have a lower continuorts capacity than the last. Good
practice is to allow from I S per cent of the one-hour capacity of the
motors as the capacity of resistance elements for the first step, up to
40 or 50 per cent for the last step. For locomotive service the
capacities should be son~cwhathigher, ranging from zo to 50 per
cent. The following tables show the resistance and capacity of
various resistance grids.
-

8-111. u s t rron
Style
46,029
46 030
46:032

Ohms

Amp

I
i

1 z z: 1 $ 1
o 0 1 s roo o

Cm.alloy
--

Style

Ohmr

Amp.

I=

5-tn. allay

-style

Ohms

49.174

Installation of Grid Resistances. The following should be taken

into consideration with respect to the position occupied by the grid
resistance: the
should lie in the longitudinal plane of the car--this insures a maximum of ventilation, when ~t is necessary to
expose the grid resistance to wheel wash, a s lash guard similar to
the one shown in Fig. 18 should be used; w&re i t is necessary to
place the grid frames close together, due to lack of room, sufficient
space should be allowed between the top of grids and car underframing for making connections; clearance between bottom of
grids and top of rail should be not less than 10in., except that
for low floor cars this figure m n hr retlucc(l to 7 in.

336

ELECTRIC RAILWAY HANDBOOK

A method of insulating grid frames for 600 volts is shown in

Fig. 18. For ~ z h ovolts the same arrangement is used with the
addition of insulated bolts between the hangers and the straps to
which the grid frames are attached. Fig. 18 also shows a crosssection of the bolt arrangement. The heavy black lines indicate
the insulating washers and tube, while the dotted cross-sectional
pieces are porcelain insulators. On equipments requiring more
than five grid frames, it will be necessary to use heavier mounting
straps or more hangers. The use of three hangers is preferable to
heavy straps.
To satisfy the National Board of Fire Underwriters' requirements,
i t is necessary to observe certain regulations regarding insulation

Cross Section
of Insulokd Bolf

FIG. 18.-Installation of grid resistance.

as a fire protection. When the car underframing is composed

entirely or partly of wood over the place where the grid resistance
is hung, the regulations call for a fireproof non-metallic insulation
a t least N in. thick, covering a surface which extends 8 in. beyond
the resistance on all sides. The cable insulation must be removed
for a distance of a t least 6 in. from the terminals of the grid resistance and su ported in a manner to give a rigid construction and
to revent vipration.
8ables for Car Equipment. Wires in cables for car equipment
usually have 7 strands for sues smaller than No. I B. & S. gage, while
No. I B. 6r S. and larger have 19 strands. In the following table,
I-Motor, 2-Motor, etc., indicate that the wire carries the current
of a single motor in the former and of two motors in the latter case.
The numbers are American or Brown & Sharpe gage sues, except
where shown in circular mils.

337

POWER OPERATED CONTROL

Nominal rated
h.p. of motor

and
ground

Motor wires

Resistance wires

-I-

I /
I-

Motor Motor

4-

Motor

Power-operated Control
Following are some of the advantages of power-operated control.
These are secured with simplicity,
- reliability and low maintenance
cost.
(I) Multiple-unit operation is made practicable. Thus the
motive power may be distributed in small units without loss of
efficiency, and since all the wheels thus may be made driving wheels
a maximum tractive effort is made possible. A maximum closeness of speed adjustment is secured and strains on mechanical
transmiss~onand draft gear are reduced to a minimum.
(2) All circuit-breakers and other controlling devices which carry
the main motor current are removed from the platform and placed
beneath the car. This reduces the possibility of danger to passengers and stimulates more careful manipulation of the control.
Platform weight and space economy are effected.
(3) Switches are operated with heavy contact pressure and
reliable overload protection is secured.
The multiple-unit or power-operated type of control was designed
primarily for the control of motors in a service requiring that
cars be operated singly or with several cars coupled together in a
train and operated simultaneously. When several cars are coupled
together in a train, the train connections are so arranged that the
motors on all of the cars may be controlled from either end of any
car by a single operator, the cars being coupled in any desired
relation and with either end of any car connected to any other car
in the train. Although designed for train operation, power-operated
control is widely used for the control of single-car equipments where
the motors have a capacity of 50 h.p. or greater, owing to the
size and weight of hand type control used with motors of large
capacity. It is also coming lnto greater use for smaller equipments
because it removes the power circuits and circuit-breaking apparatus
from the car platform thereby removing a great source of annoyance and danger to passengers.
Power-operated control systems consist in general of two parts,
the first consisting of a series-parallel motor controller which may be

ELECTRIC RAILWAY HANDBOOK

composed of a number of switches or circuit-breakers, sometimes
called contactors, whose function is to effect the d z e r e n t electrical
combinations of the motors and regulate the starting resistance in
circuit with them. There is also a reverse switch for the motors,
called the reverser, which is actuated by the same means used to
operate the circuit-breakers or contactors. The second part
comprises master controllers which operate the motor controlling
contactors and reversers through the medium of train wires extending throughout the length of train so operated. A single operator
may simultaneously effect similar combinations upon the several
motor cars composing the train, thus giving to the train the same
advantages of large tractive effort and rapid acceleration obtained
in singlecar operation.
The two systems of power-operated or multiple-unit train control
in operation differ in the means employed to actuate the contactors
and reversers, in one the contactor being operated electrically by the
line current, while in the electro-pneumatic system the contactor
or unit switch is closed by compressed air and opened by a
spring, the necessary air valves being actuated electrically by the
master controller through a train cable using storage battery or line
current. Each system aims to produce the same result, that is, to
duplicate the performance of a car operated singly, but the systems
diBer in the means adopted to secure this end.
Power-operated or multiple-unit control is divided into two general
classes according to the method of acceleration, i.e., hand-operated
or automatic, the former being wholly under the control of the
operator as is the case with the hand controller, while the latter is
effected by the current limiting relays and interlocks on the contactors or unit switches.
General Electric Power-operated Control
General Electric Contactor Control employs magnetically operated contactors or circuit breakers which are opened and closed
in proper sequence by means of train wires which are energized
through a master controller.
The Master Control, Fig. 19, comprises those parts which switch
the control current operating the motor-control apparatus. The
master controller is operated by hand, and is located in the vestibule a t either end of the car. The motor control is local to each
car, and current for this circuit is taken directly from the trolley
or third-rail, through the contactors, starting resistance and reverser
to the motors, thence to the ground. Where i t is necessary to
operate with a gap in the third-rail system, i t is sometimes customary to install a train line so that any car may supply the motor
current for the other cars of the train. The master control includes
train wires made continuous throughout the train by means of coupl e e between the cars. On each car the operating coils of the motor
control are connected to the train line through a c u t m t switch, the
train wires being energized in proper sequence by the hand-operated
master controller on the platform. Current for the master control
is taken directly from the trolley or third-rail through the master
controller, which is being operated by the motorman, to the train-

POWER OPERATED CONTROL

ma

line, and thus to the operating coils of the contactors forming the
motor control on all cars of the train.
The Master Controller is similar in design to the original cylinder
controller in that it contains a movable cylinder and stationary
contact fingers through which current is supplied in pro r sequence
to the different wires of the train line, for energizing t e operating
coils of the motor control. The value of the current required is
very small, not exceeding 2.5 amp. for each car in the train. The
master controller is provided with two handles, one for operating
and one for reversing the train movement.
The Operaling Handle is provided with a button which must be
kept down except when the handle of the controller is in the off

PIG. 19.-4prague General Electric contactor control.

position, as releasing this button permits an auxiliary circuit to

open, cutting off the supply of current to the master controller,
thus de-energizing the train-line and opening the motorcontrol
apparatus. This button is intended to serve as a safety appliance
in case of physical failure of the motorman.
The Reverser Handk is connected to a separate cylinder which
establishes control connections for throwing the electrically operated reverser into either forward or reverse position when the mastercontroller handle is on the off point. The operating circuit for the
reverser is so interlocked that unless the reverser itself corresponds
to the direction of the movement indicated by the reverser handle
of the master controller, the line contacton cannot be energized.

ELECTRIC RAlLWAY HANDBOOK

The Conlaclor is a switch operated by solenoid coils, and each
contactor may be considered as the equivalent of a finger and its
corresponding cylinder segment in the hand-operated K-type
controller. I t consists of an iron magnet frame with an operating
coil and two main contacts, one fixed and the other connected to
the movable finger. These main contacts open and close in a
moulded-insulation arc chute provided with a magnetic blowout.
Interlocks are provided for making the necessary connections in
control circuits to insure proper sequence in operating the different
contactors. All of the contactors usually are mounted in a box
placed beneath the car. this box being provided with a sheet-steel
cover lined with insulating material.
The Reverser is a switch, the movable part of which is a rocker
arm operated by two electromagnets working in opposition. The
coils receive their energy from the master controller through the
train-line, and the connections are such that only one coil can be
operated a t a time. Leads from the motors are connected to the
main reverser fingers, and by means of copper bars on the rocker
arm, the proper relations of armature and field windings are established for obtaining forward or backward motion of the car.
Circuit Couplers between Cars are so designed as to give a corresponding connection of train wires, this being secured by means of
p!oper mechanical design of plug and sockets, it being rendered
ssible to insert the plug in the socket im roperly.
lrn8&gue General Electric Automatic h t i p l e - u n i t Control
provides for the acceleration of the train a t a predetermined value
of current in the motor, this feature being provided without preventing the manual operation of the master controller a t less than the
predetermined current if desired. The operation of the contactors
IS controlled from the master controller, but is governed by a notching or current-limit relay in the motor circuit, so that the accelerating current of the motors is substantially constant. This is
accomplished by having small auxiliary interlocking switches on
certain of the contactors, the movement of each connecting the
operating coil of the succeeding contactor to the control circuit.
The contactors are energized under all conditions in a definite
succession, starting with the motors in series and all resistance in
circuit; the resistance is subsequently cut out step by step; the
motors are then connected in parallel with all resistance in circuit,
and the resistance again cut out step by ste . The progression
can be arrested a t any point, however, by tge mastercontroller,
and is never carried beyond the point indicated by the position of
the master controller. The rate of the progression is governed by
the current-limit relay, so that the advance cannot be made a t a
rate so rapid that the current in the motors will exceed the prescribed limit. One of these relays is provided with each car equip
ment, so that while the contactors on each car of a train are
controlled from the master controller for the application and
removal of power, the rate of progression: through the successive
steps is limited by the relay on each car independently, according
to the adjustment and current requirements of that particular car.
Protection Againsl Failure of Pmuw. Another automatic protection for individual c a n is provided by a second or potential relay,

POWER OPERATED CONTROL

341

which has its coil connected to the lead from the collecting shoes
of the respective car. The contacts are so connected in the contactor circuits, that in case of failure of power to any car (such as
would be caused by passing over a dead section of rail), this relay
is de-energized and causes the control circuits on that car to be
thrown back to the series position with resistance in circuit.
When power is re-applied the control progresses step by step to
its former advanced position.
Train-wire Circuits. There are five circuits leading from the
master controller, and five corresponding train wires. The five
circuits comprise one for forward direction, one for reverse, one each
for series and parallel, and the fifth for controlling the acceleration.
When the master controller is moved to its first forward point a
direction wire is energized which throws the reverser to its forward
position, if it is not already so thrown. The reverser is electrically
~nterlockedso that it cannot be manipulated when the motors are
taking current. When the reverser has moved to the proper
position, connections are made by it from the direction wlre,
through the forward reverser-operating coil, to the coils of the
contactors which control the main or trolley leads to the moton.

FIG.20.-Type

PC maln enmne and magnet valves.

General Electric Cam Control. This is a later design of multipleunit control than the contactor type and is a more direct application of the rinciple of the drum controllers. In certain features
it resemblesgoth the drum controllers and also the contactor control.
That is, instead of the main controller shaft of the drum controller
with segments mounted thereon and turned by hand, there is substituted in the new control a cam shaft which is operated by means
of compressed air. (See Fig. 20.) This cam shaft operates the contactors by means of cams. I t has been possible with this design
to assemble the contactors, reverser, line breaker relays, etc., in
one box beneath the car body. The electric control circuits are
greatly simplified and positive serial action of the contactors has
been obtained without electrical interlocks. I t also weighs less
than the older type of contactor control.
Air-operafed Line Circuit Breaker and Reverser. The line breaker
and reverser are provided with individual air cyiiders and magnet
valves. These magnet valves control the air inlets and outlets
of cylinders which in turn close the line breaker and operate the
reverser. The operation of these valves is governed through
train wires which are energized through the master controller.

342

ELECTRIC RAILWAY HANDBOOK

Air-operated Cam S h f t . The contactors instead of having individual magnet valves and air cylinders similar to the line breakfr
and reverser are all operated by means of cams mounted on a
horizontal shaft which carries a pinion at one end. This pinien

engages with a rack as shown in Fig. 2 0 , which is moved back apd

forth by a double-acting air piston. Air is aqmitted or r e ~ e a ? ~
from either end of the air cylinder by means of magnet valves w b d
are energized through the master controller and train circuit.
Controller Air Cylinder. (See Fig. 20.) The ends of the controller
air cflider are fitted with magnet valves and are designated as the

POWER OPERATED CONTROL

"on" and "off" ends. The "on" end is the end of the cylinder
a t which air is admitted to turn the controller "on" and likewise
the "off" end is the end a t which air is admitted to turn the controller off. When the magnet valves are de-energized the normal
position of the "on" valve is closed to the reservoir and open to the
exhaust. The normal position of the "off" valve when the magnet
valve is de-energized is just the reverse, closed to the exhaust and
open to the reservoir. When the car is stopped the magnet valves
are both de-energized, t h a t is, air a t reservoir pressure is admitted
a t the "off" end, and the "on" end is open to the exhaust; thus
the piston is forced in the direction which turns the motor controller
"off," i.e.. toward the "on" end. When the master controller is
turned on and the reverser throws, the line breaker closes and both
the "on" and "off" magnetic valves are energized, that is, air is
admitted to the "on" side of the piston and exhausted from the
"off" side and the rack revolves the cam shaft making the first
motor connections. The piston will continue to move in the same
direction if both magnetic valves are kept energized, but if the current flowing in the main motor circuit should exceed a predetermined amount i t will operate a relay which de-energizes the "off"
magnet valve, thus clos~ngthe exhaust port and admitting air from
the main reservoir, which in turn will equalize the ressure on both
sides of the piston and stop its movement. When t i e current in the
-n%inl*rt6tf rd\s~~m~~M~~&~rnin&-v-d,~ut:~h~r:''
-15:' nsagner
valve is again energized and the motor controller moves on to the
next point.
Master Controller. There are three pointson the master controller;
the first point gives the first point series, the second all series positions and the third all parallel positions. There are five steps in
series on the motor controller, all of which except the first are
obtained automatically by moving the master controller to the
second notch. Likewise there are four ste s in parallel on the
motor controller, all of which are obtained gy moving the master
controller to the third point. The progression from one point to
the next will only take place as fast as the notching relay, which
energizes the "off" magnet valve, operates.
Advance Leuer. A third handle is located on the master controller
cap plate. This handle is known as the advance lever and is
used to advance the motor controller point by point if in emergencies the car does not accelerate a t the current for which the notching
relay is set.
Westinghouse Power-operated Control
Classijication According to Source of Current Supply for valve
Magnets. I n the Westinghouse power-operated or multiple-unit
control the two classes (automatic acceleration and h a n d q x r a t e d
acceleration) each may be again divided according to the source of
current supply for the valve magnets. That is, the source of this
current may be the line or a storage battery.
Hand-controlled Acceleration Current for Control Circuits Taken
from Line. (See Fig. 2 2 . ) The system having hand-contrdled
acceleration and taking current for the control circuits from

344

ELECTRIC RAILWAY HANDBOOK

POWER OPERATED CONTROL

345

the line is the sim lest type and is in most general use on interurban
and city surface
for both train and single car operation. I t
is known as type HL. I t is in effect the same as the hand-operated
drum type of control, except that instead of combining the various
circuit-breaking contacts upon a single drum, operated by hand, it
divides these contacts into separate contacton and actuates them
by means of compressed air taken from the air brake reservoirs
and controlled by magnet valves and train-line wires from a small
master controller. The system is made up of two general parts,
namely, the auxilialy control system and the motor control system.
Auxiliary Control Syslem. The auxiliary control system comprises a set of train wires, a control resistance, electrically operated
valves operating the pneumatic unit switches, a master controller
and auxiliary protective devices. Current for the control circuits
i s taken directly from the trolley or third rail through the master
controller to the control resistance, the latter being arranged with
low voltage taps for the valve magnet circuits.
Master Controller. The master controller is similar in design
to the drum controller in that it contains a movable cylinder
and stationary contact fingers are used to su ply current in proper
sequence to the difierent wires of the train i n e for energizing the
operating coils of the motor control. The value of the current
required is very small, not exceeding 2 amperes for each car in
the train. The master controller is provided with two handles,one
for operating and one for reversing the direction of train movement.
Reoerse Handlc. The reverse handle of the master controller
is connected to a se arate drum which establishes control connections for throwing t i e electro-pneumatically operated reverser to
either the forward or reverse sition when the main master controller handle is moved to the g s t accelerating notch. The control
circuits of the reverser are so interlocked that unless the reverser
itself corresponds to the direction of the movement indicated
by the posit~onof the reverse handle of the master controller, the
unit switches on that car cannot be energized.
Control Coupler. Train-line jumpers between cars are so designed
as to give a corresponding connection of train wires, this being
secured by proper mechanical design of receptacles and jumper
heads so that it is impossible to improperly
insert the jumper in
the receptacle.
Conlactors or " Unit Switches." (See Fia. z t . ) These are olaced
in a group, usually being assembledon a comr;ldn frame and $aced
in a box beneath the car. The unit switch constitutes a c~rcuitbreaker having a fixed and movable contact and provided with a
magnetic blowout, the movabIe finger being actuated by an air
cylinder energized from the brake reservoir and controlled by a
magnet valve connected electrically to the train wire system. The
switch finger is normally held open by a spring, which is compressed
on closing the switch. The high pressure used in closing the switch
is made use of to reduce the contact resistance, thus reducing heating and the size of the switch contacts.
Reverser. The reverse switch or reverser consists of an insulated
drum carrying two sets of contacts arranged to make contact with

tries

346

ELECTRIC RAILWAY HANDBOOK

stationary fingers when operated forward and backward by a pair

of ~neumaticallyo erated pistons actuated through electrically
o rated valves by tEe auxiliary or master controller. No magnetic
brwout is required as the reverser does not operate when carrying
current on account of a mechanical interlocking between the main
and reverse drums of the master controller. This reverser switch
usually is mounted a t one end of the switch group, but in some
installations it is enclosed in a separate sheet-iron box.

PIG. ~3.-Electro-pneumatic switch unlt.

Overbad Trip. An overload trip is mounted on one end of the

box containing the unit switches. I t is actuated by a coil in the
motor circuit and is arranged to o en the control circuits of several
switches ahead of the motors a n t m a i n circuit resistance when an
excessive overload on the motors occurs.
Conlrol and Reset Switch. The control and reset switch is a small
single-pole double-throw canopy switch used for disconnecting the
contro circuits from the line and for dosing the control circuit to
the reset coil of the overload trip, which thus can be reset from the
car platform.
Motw C1J-out Switches. The motor cut-out switches consist of
knife switches arranged in a manner similar to those on a hand or
dnfm controller and are mounted in a box on one end of the switch
PUPControl Resistance. The control resistance consists of a number
of flattened tubes wound with resistance ribbon and sup rted in an
iron-dad frame with all live parts protected from
weather.

POWER OPERATED CONTROL

347

The several tubes are connected in series for the full trolley voltage,
and low-voltage taps are taken off for the magnet valve circuits.
Fig. 24 shows the main and control circuits for a four-motor equip
Dent of 6c-h.p. motors with HL control, using a ten-switch group.

<<
.,

2 3;
Master Oontroller

To Juuctron Box
Control Bwllch

Cut our Switchea a.b,c.

are connected t o No.1 Eandle
Cut-out 8wltches d.c. 1. are
corseetcd to No.2 Handle

PIG. 24 --Scheme of clrcutts. West~nghouse hand-operated multtple-unrt

control. llne current operation, four motors.

Westinghouse Automatic Control. Westinghouse automatic

multiple-unit control d~ffersfrom the hand operated type in that the
acceleration of the motors is made on a predetermined current
which is governed by a secondary master controller (called a sequence
switch) and the setting of a current limit relay in the motor circuits,
instead of being entirely under the control of the operator. A
typical schematic diagram, shown by Fig. 25, is for heavy duty with
two 190-h.p. motors in subway service.

348

ELECTRIC RAILWAY HANDBOOK

Master Controlh. The master controller is essentially the same

HI, control except that i t has fewer notches and is arranged
with a safety device to set the brakes in case of ssible physical
failure of the motorman. The emergency or "dea$man9s" a plication can be secured pneumatically, or electrically if desiretwhen
eiectro-pneumatic air brakes are used. The main handle of the
master controller has an upward movement whenever the motorman's hand is removed; this releases the fingers on the upper positions of the reverse drum and cuts power off the control supply to
certain switches and further releases a pin valve, which causes the

as for

FIG. 25.-Main
and control circuits. Westinghouse battey operated
autornatlc electric-pneumatic control for two I 9 0 h.p. moton. tram operation.

air brake train pipe to be vented and brakes to be applied. The

master controller has three working notches and a n off position.
The scheme of operation consists of the following progression: the
reverse handle is first moved to correspond to the desired direction
of the train; the main handle is then depressed and moved from the
off to the succeeding notches to give the speed desired. With the
main handle depressed, the off position serves as a coasting
of the controller, as it then opens all motor switches but
apply the brakes. The first or switching position establishes connection with the train line so that the reverser is operated to the
desired position, and the unit switch group closes the circuit to
the motors with all resistance in. The second or full series position
of the master controller handle introduces the current-limiting

gz%

POWER OPERATED CONTROL

349

relay which r m i t s the regressive picking u p of the several unit

switches as t e current f$s below the predetermined value. With
the controller on this notch, n o further progression is made after
the current relay has permitted the closing of the proper unit
switches to effect full series position connection of the motors with
starting resistance entirely cut out. The third notch on the
master controller establishes parallel connections of the motors
by again bringing the current relay into activity, which permits the
progressive cutting out of starting resistance until full parallel
operation is attained, with full line voltage upon the motors.
I n former designs, the master controller was constructed to have
but one handle, which was notched to right or left of a central
"off" position for forward or reverse running, respectively. The
dead man's feature was then obtained by a spring which caused the
handle to return to the central off position if the operator's hand
was removed; this energized a n emergency train brake magnet
valve which set the brakes.
Unit Sutitches. The group of switches is similar to that used in
the hand-operated type of multi le unit control, with the addition
of a sub master controller (calfed a sequence switch), which is
automatically advanced under the control of the limit relay to
close the unit switches progressively so as to obtain a uniform
acceleration a t practically constant motor current.

FIG. 26.-Cross-sect~on of differential alr cylinders.

The Sequence Swikh is essentially a n electricaily controlled,

pneumatically operated drum that carries contacts arranged for
energizing the valve magnet coils of the switch group. I t consists
of two air cylinders with the pistons connected by a rack that
transmits the movement of the istons through bevel gears to the
drum carrying the contacts. &ee Fig. 26.) Compressed air is
admitted or released from the cylinders by means of a magnet valve
a t each cylinder. The magnet valves are similar in general construction to those in the switch group, but the operation of the "off ''
valve is different in that when this valve coil is energized air is
released from the cylinder, and when it is de-energized air is admitted.
The "on" valve admits air to its cylinder when the coil is energized
and releases air when it is de-energized. If both magnet valves
are energized simultaneously, movement of the drum results,
since air is admitted to one cylinder and releaser1 from the other.
With neither magnet energized, the air in the "on" valve magnet
cylinder is a t atmospheric pressure, while that in the "off" valve
magnet cylinder is a t reservoir pressure, thus holding the pistons

350

ECECTRIC RAILWAY HANDBOOK

. Energizing the "on" magnet

and drum in the "off"
balances the pressure in t ~ ? g p i s t o n sbut causes no movement.
Thus, to move the piston and drum in successive steps, i t is only
necessary to energize the "on" magnet from the control supply
circuit and the "off" magnet through the limit relay contacts. The
sequence switch may be mounted directly on the switch group or
apart from it.
The Limit Relay is of the solsnoid type, with a coil connected in
series with the main motors, in which operates a plunger that
carries contacts for closing the control circuit to the sequence switch.
The relay plunger lifts when the main motor current reaches a
certain predetermined value afid interrupts the control Current to
the "off" valve magnet thus halting the progression of the sequence
switch drum. As thk motor speed increases, the current falls
until a value is reached where the limit relay plunger again drops,
closing the operating circuit to the sequence switch. This performance is repeated until the control has reached the notch indicated on the master controller.
Line Relay. The line relay consists of a coil connected between
the line and the ground and rovided with a plunger, to which is
attached a contact disk. wKen line voltage is available, the coil
of the line relay is energized, lifting the plynger and causing the
contact disk to close a circuit across a paif of contacts. These
contacts are directly in series w t h the auxiliary control circuit of
certain contactors, and hence, if the line current fails for any
cause, this auxiliary contact is opened, thus opening the main
power circuits. Should the lioe switch be opened or should the
current be cut off from the line for any reason, the plunger of the
line relays drops, the auxiliary control circuit is opened and all
the contactors are opened. When the contactors are opened by the
temporary cutting off of the control current for any reason when
the master controller handle is pz?ially or wholly notched up, the
limit relay governs the progressive reclosing of these switches.
Line-operated Automabc Control. For the line-operated automatic control, type AL, a control resistance is used to obtain a low
voltage supply of control current. The control and reset switch
and control resistance are the same as those used for the handoperated control, type HL.
Battery Control. Battery control may be either the hand-operated or automatic t v ~ e . The former is known as type HB and
the latter as type AB: When B battery is used to supply the auxiliary control current for multiple unit control, the control resistance is replaced by a storage battery, supplying current a t from
14 to 32 volts. On account of the low voltage used, a plug switch
replaces the canopy type control and reset switch used on line current control equipments and the insulation of the entire auxiliary control circuit can be greatly reduced Battery current for the auxiliary
control circuits is particularly adaptable to high voltage systems,
either alternating current or direct current, on account of the elimination of the high voltage from the master controller. This type of
control is also employed on elevated and subway systyms, and in
trunk line service where close headway between trains is essential.

CABINET CONTROL

PIG.27.-Cabinet control arranged for installation in car vestib~lle.

352

ELECTRIC RAILW.\Y H:INDROOK

The use of a low voltage control which is independent of the line

voltage eliminates various causes for incorrect operation, making
this type of control particularly applicable for this class of service.
The storage battery used for the control can also be used for
emergency lighting, as well as for signal and brake systems. I n
large systems, where storage batteries can be economically maintained, the benefitsof a low voltage control system are most evident.
Unless it is desired to charge the storage battery from an outside
source a t the end of a run, it ran be charged automatically from the
line on a direct-current system or by a small motor-generator set
on an alternating-current system, the amount of charge being
regulated by means of an adjustable battery charging resistance
ant1 3 battery charging relay.
Cabinet Control. Some engineers are of the opinion that the
elements of 3 power-operated control system might be more
advantageously arranged than in the group mounting as usually
supplied by the manufacturers, both from the viewpoint of the space
occupied and the place of installation, thc latter especially considering ease of maintenance. There is an advantage in the
usual group mounting in that the manufacturer properly assembles
the various units and makes the various electrical interconnections,
thus relieving the user of that responsibility. However, the
proponents of the so-called "cabinet" mounting of control apparatus are willing to accept such responsibility in exchange for other
advantages. The various switching elements of the control apparatus are shipped separately, and installed in a "cabinet" or compartment especially designed and built into the car body for the purpose.
The cabinet may be built into the bulkhead back of the motorman's
position, as suggested by Fig. 27 (see also Fig. 18, p. 506), or i t
may consist of special compartments under the car, arranged for
the most economical use of the space there available, as well as for
the most convenient inspection and maintenance. It has been
suggested that the space beneath longitudinal seats might be used
for all control switching apparatus except the line switch which
breaks main circuits, which probably should be kept in its usual
location under the car.
Single-phase Series Motor Control
'The single-phase series motor (see pages 231 to 238) is controlled
by adjusting the electromotive force
1applied to the motor. The three methods
oi 'making this adjustment are: ( I ) By
resistance in series as in ordinarv directcurrent series motor control. ( z ) - auto~ ~
transformer. (3) By induction regulator.
The latter two are more efficient than the
first because by their use the heavy
resistance losses which would occur in the
resistors by the first-named method are
saved. The second is a widely used method
FIG.28.-Tap control for and is generally known as the tap potential
control (see Fig. 28) and the function of
a.c. motors.

SINGLE PHASE CONTROL

the control, whether it be hand-operated K type or power-operated,
is to successively connect the motor terminals to transformer taps
of increasing potential while starting.
Auto-transformer. The auto-transformer or single-coil s t e p
down transformer is wound for the trolley potential. Upon its
grounded side there are several taps brought out to facilitate
starting. I t is customary to connect two motors in series when
operating alternating current in order to approximate 500 volts
input and reduce the size of the contact surface required incontactors. Owing to the fact that the alternating-current motor
characteristic is more drooping than that of the direct-current series
motor, fewer starting p i n t s are required for alternating-current
control. As each point on the controller
with potential tap control constitutes a
running point a t full efliciency, it is not
necessary to use series-parallel connection
of motors as is done with direct-current
motors. If while passing from one compensator tap to the next without opening
the circuit the ga were bridged by a connection of very row impedance, a heavy
g
current would flow in the portion of the
g
corn nsttor winding thus short-circuited.
To
this current a t a Inw value an
hG.
zQ,- ~.r;con,rol
inductance or a non-inductive resistance for a.c. motors. with pre(termed, in this application, preventive ventive coils.
inductance and preventive resistance, respectively) is placed in the
circuit. Fig. 2 9 shows the application of the inductance (preventive coil).
Reversing. Reversal is effected by reversing the series field
windings as in direct-current control.
Hand-operated Control. The hand-operated control is identical
with the type K controller used with direct-current motors, except
that advantage is taken of the fact that alternating-current arcs
of considerable size can be broken in the air without the aid of
the magnetic blowout. Where cars are to be operated from both
alternating current and direct current trolley with the same
equipment, the hand+perated control is provided with magnetic
blowout for direct current running, using the same controller without the magnetic blowout for alternating current running, unless
the motor equipment be of large capacity.
Power-operated Control. The master controller is similar to the
d i ~ e c current
t
master controller. Its office is to energize the train
wlres in proper sequence. Current of proper potential is obtained
from transformer taps. The contactor a s~milarto the direct
current contactor, except that it is designed for operation on
alternating current and therefore has its magnetic parts laminated.
The contactors connect the motors with transformer taps or with
starting resistance grids, depending upon the character of the
supply current.
Combined Alternating and D i e c t Current Control. (Fig. 30).
Owing to the fact that alternating current series motors are wound

e,

354

TLECTRIC RAILWAY HANDBOOK

for potentials of aDprnximately 2 2 5 volts per motor, and that such

motors in order to meet the exactions of cornmerc~aloperation
must in some cases also be adapted to run with 600 volts direct
current, some additional features in control are required to perform
this double service of alternating current and direct current control.
When both alternatinp. current
and direct current rurkinp; are
to be rrccomplished with the
snmc control it is necesmry to
hwnce
dS-ltchn

A.C. O p c n U a .

PIG. 30.-Clrcuits

and sequence of sw~tches.New Haven twin-motor locomotive (eidht motors), alternating and direct currcnt operation.

effect the follo\ving changes when changing from alternating current to direct current and rrcc versa.
Change main line fuses or circuit breakers.
Change lightning arresters.
Change motor field winding connections.
Change transformer taps to resistance t a p s
Incidental changes in car wiring connections.
To elIect these changes with a minimum amount of delay, all the
necessary contacts are concentrated upon one cylinder in an
a~lviliarycontroller, so that a single movement will effect all changes
sirnultane~usly.

SINGLE PHASE CONTROL

355

Retaining Coil. I n changing from alternating current to direct

current trolley and vice versa, it is necessary to guard against the
possibility of wrong connections upon the car for the current
received, that is, to prevent disaster should connections be made for
b v o l t direct current operation and accidental contact be made
with high voltage alternating trolley. T o guard against this, the
main switch of the direct current, alternating current car equipment
is provided with a retaining coil so designed that it will open when
the motor current is interrupted. Where alternating current and
direct current trolley sections adjoin, a dead section may be left
between the two for a length not exceeding a car length, w, that a
car may pass from one section to the other a t full speed, in which
case the main car switch opens on the dead section through lack of
power to operate the retaining coil, ancl will reset automatically
for alternating current or direct current operation as the case may
be, after leaving the dead section.

Relay

FIG. 31.-Control

circuits for doubly-fed motors. Paoli

Philadelphia.

Electrification.

C o ~ e c t i o n to
s Air Compressor. The air compressor is geared
to a single-phase motor wound for operation with 500 volts altemnting current or direct current, and it is customary to wind the motor
fields in two parts connected in multiple for alternating current
and in series for direct current running, these changes being effected
through the medium of the commutating switch.
Control of Doubly-fed A.C. Series Motors. The doubly-fed
series alternating current motor is described on page 234. The
circuits of a typical control system for two such motors are shown
by Fig. 31.
Indnction Regulator. Due to great weight, low power factor and
lack of simplicity as c o m r r e d with the tap potential method, the
induction regulator met od is a t a disadvantage in controlling
alternating current traction motors. The induction regulator
ial type of transformer, so arranged that one winding may
r , " g s t e d relative to the other, mechanically. J t is commonly

356

l?LECTRIC,RAILWAY HANDBOOK

built like an induction motor having a coil-wound rotor and

a very short air gap which is permitted since there is little motion
and negligible bearing wear. The rotor winding is generally the
pnmary. The primary is connected across suitable taps of the
transformer, and the secondary, in series with the motor, is also
connected across suitable transformer taps (see Fig. 32). The
electromotive force induced in the secondary of the regulator due
to the current in the primary may be adjusted by turning the coils
relative to each other, and it may be made to buck or boost the
electromotive force applied to the motor directly from the transformer by any amount within the range of the regulator. Thus the
net electromotive force applied a t the motor terminals is adjusted.
On the single-phase locomotive of the Dessau-Bitterfeld line of the
Prussian-Hessian State Railway gradual change of voltage is obtained by means of an induction regulator which covers a range of one step
on the main transformer. The first
position of this regulator lowers the
transformer voltage by one-half step.
=%'I I ~ On turning the induction regulator the
voltage reduction gradually drops to
zero; that is to say, the resultant voltage is the mean value between the two
successive transformer taps. As the
1.37
turning is continued the regulator delivers a rising additional voltage which
FIG, jZ,-~n~uction
gradually reaches the value of one-half
contro~.
step and thus equals the next higher
step on the transformer. In thls way the voltage changes from step
to step are made without shock. The switching from one step of
a transformer to the next is automatically obtained by means of a
switch which is cou led to the induction regulator. The small
single-phase motor wkch is used for turning the induction regulator
is cut out automatically as soon as the voltage of the traction
motor has reached its highest or lowest value as the case may be.
This small motor and the various contactors in the motor circuit are
operated by low voltage current taken from the main transformer.
Three-phase Induction Motor Control
The three general methods of controlling the speed of a threephase induction motor are ( I ) rheostatic; (2) changing the number
of primary poles; (3) cascade operation. The first method is used
in connection with the other two, a t least during the straight line
acceleration period.
Reversing. A three-phase induction motor may be reversed by
interchanging two of the three leads which supply the energy to
the rimary windings of the motor.
d e o s t a t i c ContxoL In this method the roper tractive effort
during the straight line acceleration penoxis secured by adding resistance to the secondary winding of the induction motor.
For railway work the windings of the secondary are brought out to
slip rings and the added resistance in the form of grids or liquid is

INDUCTION MOTOR CONTROL

external to the motor circuit. The grid resistance is cut out in
steps as the motor speeds up and is short-circuited a t the last step.
Where liquid resistance is used, its value is reduced steadily, thus
securing a steady acceleration. The electrodes which dip into the
liquid are he!d stationary and in the process of reducing the resistance, the level of the liquid is raised, the influx being regulated
by a valve controlled by the engineman. Ordinarily the secondary
winding of the motor is three-phase; the external control resistance
is star connected and is customarily kept balanced throughout
adjustment. With a view, however, to simplifying apparatus this
external resistance is sometimes unbalanced. Such unbalancing
is used in the control of the four motors which drive each of the 115ton locomotives operating through the Cascade Mountain tunnel
on the Great Northern Railway. Each motor is three-phase, 25cycle, 500-volt, 475-h.p with a gear ratio 4.26 driving 60-in. wheels.
T h e following, relative to this particular installation, is from a
paper by Dr. Cary T . Hutchinson, A.I.E.E., 1909: "The control
system of each motor is separate; the circuits branch from the transformer and are independent through the resistance. There are
14 contactors in each motor circuit, 56 in all. . .
. Iron grid
resistances are provided for each motor; there are thirteenstep in
the control, but in order to reduce the number of contactors to the
lowest possible point a n unsymmetrical system is used. A change
is made in the resistance of one phase only, in passing from step
to step. This arrangement in effect treats the three-phase circuit
as a single-phase circuit; on each step of the control the torque
is the average of the three values of the torque of the separate circuits. The principal advantage of this is that 56 contactors do the
work that would otherwise require 128, thus effecting a great gain
in the simplicity of the control apparatus."
Because of the resistance loss in the added external resistance in
the secondary circuit the rheostatic method of induction motor control is inefficient. The net mechanical energy output of the motor
is equal to the energy input to the secondary of the motor, dirninished by this secondary copper loss a t that instant.
Changing the Number of Primary Poles. (See page 239 for
method of determining synchronous speed.) A motor may be
arranged so that the number of its primary poles may be changed,
thus giving more than one efficient running speed. The number of
poles may be changed by re-grouping the coils of the motor primary
winding or by using a n independent primary winding for each
number of poles desired. The re-grouping method is generally
preferable because in its application all the motor primary windings
are used.
Cascade Operation (Also referred to as Tandem Control or Concatenation). A great many variations are possible in concatenation. I n railway work two motors are used. Two motors may be
operated in direct concatenation to give half speed and then
operated in parallel to give full speed and the advantages thus
gained will be analogous to those gained in series-parallel control
of direct current motors. Concerning concatenation the following
useful comparison is drawn by B. G. Lamme: "This corre-

358

ELECTRIC RAILWAY HANDBOOK

sponds to two or more direct current shunt motors with their a r m tures connected in series. If both armatures are wound for equal
speeds, then the tandem or series connections give half speed, as
with the alternating current motor. If one direct current armature
is wound for a higher speed than the other, which corresponds to
two alternating current motors with different numbers of p o l e
then four speeds may be obtained corresponding to the windin6
connected cumulatively, differentially and to each motor used semrately. This corresponds t o the four speeds of the alternating current combination where two motors have different numbers of poles
and are connected in tandem to give speeds corresponding to the
sum or difference of the number of poles or are operated separately,))
For concatenation of two motors for railway service, the rotom
of the motors are mechanically connected, the primary of the f i ~ t
motor is connected to the supply, the secondary of the first motor
is connected to the primary of the second motor and secondary
of the second motor is connected to a n external resistance a t start.
As the concatenated set speeds up the resistance in the s e c o n d a ~ ~
of the second motor is reduced and finally this secondary is shortcircuited just as in the rheostatic method of control outlined above.
If the motors have the same number of poles and are operated in
direct concatenation they may, after havlng reached their normal
maximum speed (synchronous speed of the set minus the slip)
separated and each., having resistance inserted in its secondary,, may
have its primary connected to the supply. From this point t4e
resistance in the secondary circuits may be reduced as in the ordinary
rheostatic method of control outlined above, finally leaving the
secondaries short-circuited and the motors o erating in full parallel
on the line. Two running speeds using L t h motors are thsS
secured.
Cascade-single Control. This method is similar in operation
to the method just outlined except that the second motor (termed
the auxiliary in this use) is cut out entirely beyond the period of
concatenation.
Cascade-varallel-sin~le Control. This method is between
the two p;evious mechods (parallel and cascade-single). T K ~
motors have a different number of poles or the same number of polgs
with different gear ratios. This method is similar in operation to
the parallel method except that during acceleration beyondcoqcatenation the motor with the greater number of poles reachsynchronism and is cut out before the notor with the fewer
reaches its maximum running speed. If the motor having the
poles be cut out when the one having the greater number of poles
approaches synchronism the train will operate a t a running speed
between that for the set in concatenation and that for the motor
having the fewer poles when running free with its secondary shortcircuiied.
Combination of Concatenation and Changiq the Number af
Poles. I f several running speeds are desired thls may be accoqplished by changing the number of poles in one or both motors in
the manner previously outlined and then proceeding with one a f
the methods of concatenation.

e:i

CONTROL AI'PARATUS DIAINTIINANCE

359

Synchronous Speed of Two Concatenated Motors; Molors i n

Direct Concatenation. Two concatenated motors are said to be in
direct concatenation when they are so connected that they tend to
start in the same direction. The synchronous speed of a set so
connected is given by:
X 120
S = fin which S

PI + P2

synchronous speed, revolutions per minute

f = frequency of supply, cycles per second
pl and pz = the number of poles per motor, respectively.
Molms i n Differential concatenation. Two concatenated motors
are said to be in differential concatenation when they are so connected that they tend to start in opposite directions. The synchronous speed of a set so connected is given by:
I20
S = f-X
-PI - P 1
=

OVERHAULING AND INSPECTION O F CONTROL

APPARATUS
(A.E.R.E.A. Miscellaneous Methods and Practices)
Type K Controller. The equipment should be given a thorough overhauling for every
miles of service a s follows:
The controller should be taken apart, thoroughly cleaned, defective
parts replaced, wood scraped and shellaced and other parts of the
controller painted with insulating paint. The controller should
then be given a breakdown test of not less than 1500 volts alternating current for 5 seconds. With controllers in good condition,
periodical inspections on a basis of from 300 to 500 miles servicedepending upon the conditions of operation-will be sufficient and
the most economical. Careful attention should be given to the
fit of the controller and reverse handles on spindles.
Inspection of G E Type M Multiple-unit Control Apparatus.
Master Controller. Examine the master control fingers, trying
tension of fingers and polishing cylinder and fingers to secure good
contact. Scrape carbon dust from arc-deflector division plates.
Wipe dust off inside of cover. Examine controller throughout
for loose connections and see that each finger is adjusted to make
good contact. Lubricate fingers and cylinder with small amount
of vaseline. Press down button on controller handle to see that
the auxiliary fingers make proper contact, and polish up the same
when necessary. O
i
l sparingly the shaft bearings of cylinder
and reverser switch. Wipe off excess lubricant, especially around
the blow-out coil. When finished, try on each point to see if
contactors pick up properly, also throw reverser two or three times.
Blow out controller with air hose.
Contactors. Scrape arc chutes clean of carbon and copper dust
and examine the tips, filing up any that are blistered to make sure
of good contact on all. Press up under contactor fingers to see if
bottom tip wipes or travels on top tip a t least one-eighth of an inch.
If tip is worn thin or so a s t o give less than one-eighth of an inch

ELECTRIC RAILWAY HANDBOOK

wipe renew the tip. Try all screws which secure the tips to the
contactor arms, making sure that all screws are absolutely tight.
Examine for loose connections and try all screws and bolts to make
sure that they are all tight. A long screw-driver arranged to
turn with a wrench at bottom is desirable for this purpose.
Inspect interlock fingers and polish up tips to insure good contact.
If ~nterlocksare bent out of shape, bend back into such shape
as to make ro r contact. Examine all shunts to see that none
are partly LroKn or any in such condition that they will not
have full carrying capacity, thus preventing danger of burning
the hinge pins. See that contactors will lift sharply on each point
and drop freely when controller is shut off. Before shutting the
contactor box take air hose and blow out the contactors and box
thoroughly clean.
Reuwsers. Try all screws to finger and connections to see that
they are tight. Examine all fingers, filing fingers and contacts to
insure a good contact. Feel tension on each finger, and renew
if finger has been heated and softened. Make sure that each finger
will make square and firm contact with segments. Wipe off all
segment blocks, and if there is any indication of shorting across
between segments remove segment block and replace with one
that has been cleaned u and varnished. Throw rocker by hand
a few times to see that &e fingers will not catch. Throw reverser
electrically a few times to see that it throws with strength and
promptness without undue arcing and without rebound. When
reverse finger is not making good contact it sometimes can be
detected by tapping the finger with a screw-driver handle or by
placing a thin strip of paper between the finger and the contact
and noting the way the a r pulls.
Overhauling Type hf &?ultiPle-unit Control Apparatus. Overbasis as follows:
hauling should be on a 60,-mile
All coils removed from the contactor, reverser and circuitbreaker; boxes thoroughly cleaned and inted with an insulating
int. Interior of boxes cleaned an8apainted; contact strips
Etween coil frames inspected for loose contacts; all working parts
thoroughly inspected and worn parts replaced when necessary;
wires inside of contact box thoroughly painted and when reassembled given an insulation test of 1500 volts alternating current.
I t would then seem that periodical inspections on a basis of from
600 to 900 miles of service would be satisfactory.
The following points are suggested as requiring attention a t
such times. Examine for broken shunt straps and broken hinge
pms. See that interlocks are Eperly adjusted and that small
arcs do not form between the gers and disks, thereby burning
finger and disk, which would eventually cause a defective contact
a t this point and a dead car. Clean the disk and finger with fine
emery cloth. Keep the arc chutes and plates clear of all copper
caused by contactors breaking current. See that all connections
are tight. See that springs are not broken and are in good order
insuring good contact when closed. See that plungers do not bind
and that contactors break free when the current is thrown off.
Contact plates should not be worn so low that screws holdihg

CONTROL APPARATUS MAINTENANCE

361

them are burned. Blow out contactor box with compressed air.
Note condition of wiring in the box. Clean the master control
cylinder and use a small amount of vaseline on the fingers. See
that the handle is of proper fit and works perfectly free. The
adjustment of controller should be looked after very carefully, as
there are no adjustment screws on the contact fingers. Note condition of throttle. Clean throttle disks and fingers and see that
adjustment nuts are not loose. Do not clean throttle plunger
unless it shows signs of sluggishness. Great care must be taken
when cleaning plunger. Clean reverser and note adjustment and
condition of plates and fingers and that the reverse throws in r o p
erly. Use no oil or grease on contactor or reverser finger or prates.
A great deal depends on the close adjustment of interlocks. All
bearings on contactors and interlocks must he made loose. When
a contactor box becomes coated inside with a yellow coating
caused by the burning of co per, short circuits are very likely to
occur if this is not cleaned off.
Overhauling Electropneumatic Control Apparatus. Overhauling
should be on a 60,-mile
basis as follows:
Clean the drum and adjust fingers of master switch; inspect cab
switch terminals and see that they are held rigidly and no strands of
wire are broken. Repair, clean and carefully adjust line relay,
limit switch and battery relay. Limit switch should be adjusted
with ammeter. Take apart, clean, scrape and shellac drums of
motor cut-out switch and reverser; replace any parts that will not
make the mileage and adjust the finger tens~on. Strip switch
groups of all magnets, switch arms and moving parts, replace
w o n parts when necessary. Replace worn or broken arc shields;
adjust all magnet valves to operate a t proper voltage; replace
defective shunts; adjust and dean all interlocks and interlock
fingers; examine all insulation and make as good as new; examine
piston leathers and see that they are flexible and replace those badly
worn. Storage batteries should be cleaned of sediment and acid
strength adjusted. Grid diverters should be cleaned, the insulation renewed where necessary, and all connections tightened.
Control jumpers should be tested by passing 7 amperes of current
through them for 3 minutes, a t the same time giving jumper the
same motion that it has when in service. Clean and adjust circuitbreaker; thoroughly blow out all piping and air chambers connected
with the control.
On short period inspection the following is the practice on a road
having inspection periods based on a 600-mile service.
Master Switch. Clean and lubricate every tenth inspection.
Cab Switches. Inspect terminals every inspection day.
Close jaws of cab switch to fit tight each ins ction day.
Line Relay and Limit S.witches. Clean witl?crocus cloth each
tenth inspection. Inspect connections each inspection day.
Motor Cut-off Switch and Reuerser. Inspect finger tension and
oil drum contactor each second inspection and feel the terminals
to see if the wires are O.K.
Inspect interlocks each twelfth inspection.
Oil reverser switch toggle each tenth inspection.

362

ELECTRIC RAILWAY HANDBOOK

Circuit-breaker and Switch Group. Clean armature and valves

each tenth inspection. Inspect contacts each inspection day.
Clean arc chutes each inspection day. Blow out with compressed
air all switches and grid diverters each third inspection. Inspect
all grid diverter connections and oil all pistons each inspection
day, see that all terminals are tight and inspect wires. Wipe off
insulators. Inspect shunts and battery connections each inspection day. Add distilled water to take care of evaporation when
necessary. Test specific gravity each thirtieth inspection day.
Test battery relay and inspect terminals of battery switches each
ins ction day.
gaintenance of Controller Fingers and Contacts. The Westinghouse company's instructions 'nder this heading call attention
to the importance both of lubrication and contact pressure, a s
follows:
Lubrication. Large capacity fingers j% and I in. in width), such
as are used in the K type of controller, are made of copper and slide
over copper contact segments. Both finger and segment, being
of the same composition and comparatively soft, will wear excessively unless roperly lubricated. The quality of lubricant used
varies somewRat with the climate and temperatures, but vaseline
will be satisfactory for summer and for moderate winters; engine
oil is good for very cold winters. I t is erroneous to use large quantities of lubricant with the idea that the more used the longer i t
will remain on the contact segments, as the surplus soon wipes or
bums off and accumulates on arc barriers, fingers and drum castings,
collecting copper dust and dirt, with a resulting tendency toward
insulation failure. The contact segments also become sticky
and dirty. The best practice is to spread the lubricant as smoothly
over the segment as possible with a cloth, operate the fingers over
the segment several times, and then wipe around the finger and the
segments to remove auy surplus. Where non-arcing duty is performed, considerably less lubricant can be used, and i t should be of a
lighter grade.
Contact segments and fingers become roughened by arcing and
should be smoothed up carefully with emery cloth or a file before
lubrication is applied. A wire-drawn contact surface should be
carefully smoothed and wiped off. The safe mileage between
inspections depends on local conditions, and should be determined
by experiment for each service.
Small fingers
in. and % in. wide, such as used in the control
circuits of multi le unit apparatus, are usually made of different
material from tie-contact segments, thus causing less cutting.
The arcing is very slight, so that the wear is much less, and the
need for lubrication is not so urgent as on the copper fingers.
However, the same general methods should be followed.
Conlacl Pressure. The safe current load on a finger depends on
the width of the contact surface, the pressure a t the point of contact, and the mass and radiation of the finger and segment. The
capacity for a given width increases with the pressure, but too
heavy pressure causes excessive wear and stiff drums. Average
practicable finger pressure for general service (copper fingers on

CONTROLLER MAINTENANCE
copper segments) are listed below. For different contact materials
these values may be increased somewhat.
Size of finger
One inch . . . . . . . . . . . . . . . . . . . . . . . . . .
Three-fourths inch . . . . . . . . . . . . . . . . . .
One-half ~ n c h . . . . . . . . . . . . . . . . . . . . .
One-fourth inch .............................

Pounds
pressure
8

6
4

By means of a small spring balance and a wire stirrup the pressure

is easily checked, and inspectors soon become accustomed to the
feel of a finger with correct pressure. Pressure is varied by changing the bend in the flat finger spring. After bending, see that the
finger is making contact along its full width. Most fingers have an
adjustable stop, which limits the drop of the finger tip when it
leaves the contact, but this stop does not vary the finger ressure;
its sole purpose is to prevent stubbing. The drop should \e set a t
X s in. to % in. or enough to allow the finger to lift entirely free
from the sto when the finger is on the contact; this allows full
pressure at t f e contact surface. The lift and pressure should be
checked in all positions of the drum, as an eccentric drum or one
having worn bearings and shaft may have good finger pressure in
one sition and weak pressure in another position.
considerations mentioned above are equally important when
installing new fingers or contact segments. A new finger should
prrfenbly be ground in with emery cloth to give a contact area a t
east % in. width along the contact line, and the finger should make
contact over a t least three-fourths of its breadth.

TE

SECTION VI

CURRENT COLLECTING DEVICES

Trolley Wheels. The first essential of a good trolley wheel is
that it should be in perfect balance, and run perfectly true. The
necessity for this, as well as for good lubrication, is due to the location a t the end of a long and flexible pole, and to the high speeds;
a t 30 miles per hour a 4% in. wheel ( 2 % in. diameter a t contact)
runs a t about 4250 revolutions per minute; a t 60 miles per hour a
6 in. wheel (4% in. diameter a t contact) runs a t about the same
speed. As a simple test for balance on commercial wheels, it is
suggested that wheels to be tested be mounted in harps and the
wheel held against a rapidly moving belt. If the balance be only
slightly imperfect, i t will be found impossible to hold the wheel
against the belt.
The two designs or types of trolley wheels in most general use
are the alloy (Fig. I) and the "Ideal" (Fig. 2). The alloy wheel
is cast bronze, and should be of a composition to combine ample
conductivity with sufficient hardness to insure maximum mileage
and minimum wear on the trolley wire. Among the formulas
which have been used by various companies are the following.
Copper.. ..............
Tin ...................

Zinc...................
Copper.. ..............
Tm ...................
Zinc . . . . . . . . .

Lead ..................

Per cent
90.0
8.0
2.0

Per cent
91.38

6.5
2. o
o.xa

Copper ..............
Tin ...................
Zinc.. ................
Flux . . . . . . . . . . . . . . . . .
Copper.. ..............
Tin.. .................
Phosphor tin.. .........

Per cent
91.0
7.0

1.5
0.5
Per cent
96.0
3.0
1.0

The so-called "Ideal" trolley wheel is somewhat cheaper than the

alloy wheel, and has given good service. I t is built up with a
forged copper center with steel flanges pressed on either side, as
shown in Fig. 2.
For lubrication the bronze and graphite bushing is in almost
universal use, often supplemented by an oil well in the hub, as
shown in Fin. I. In some cases, this oil well has been very
- considerably enLrged, as in the special type shown in Fig. 3.
Dimensions of various sizes of trollev wheels. as standardized bv
the manufacturers, are shown in the table on page 367 In generil
the smaller wheels are used in city service where the maximum
speeds are comparatively low. Some years ago there appeared a
distinct tendency toward large trolley wheels for heavy and high-

365

366

ELECTRIC RAILWAY HANDBOOK

speed service, as high as 10 in. diameter having been used in some

instances. The great weight of such large wheels, the high base
spring tension required, and the resulting likelihood of increased
damage to overhead construction when the wheel left the wire,
caused a reaction to the present tendency to limit the diameter to
about 6 in.
The average mileage obtained from trolley wheels in city service,
as reported by 22 companies to the 1922 Committee on Equipment,
A.E R E A , was 8272 miles; in interurban service, as re rted by 2 0
companies, 5035 miles. Of 21 companies reporting g n i t e data

PIG. I --Standard alloy

trolley wheel

trolley

FIG 3 -Specla1 form

m t h large or1 wen.

on the hfe of trolley wheel bushings as compared to the life of the

wheel, experience varied from % to 3 times the life of the whqel,
the average l ~ f eof bushing being the same as that of the wheel;
eliminating four cases which seemed to be unusuaI (w,2, 3? and s),
the average of the bushing lives reported by the remalmng 1 7
companies was 75 per cent of the life ol the wheel.
The Standard Handbook for Electrical Engineers gives t.he
following for approximate life of trolley wheels:

. .....

Cttv service 2~ rntles wr hour maxlmurn

I 1.000 nules
Sub;rban&gce 35 m;les per hour maxtmum
. 6.000 rnlles
Interurbanservtce 50 m ~ l e sper hour maxtmum .
3.500 mtles
High-speed seMce 60 rntles per hour maxlrnum
z.ooo rntlea

am

TROLLEY WHEELS

DIMENSIONS
OF TROLLEY
WHEELS,FIGS.I AND

Dlmenslons. ln tnches

DIE,F

D ~ a m Wtdth

6
31
32

Alloy
Alloy
Alloy
Alloy

2
2
2%
2%

33
40-A
34
40-B

Alloy
Ideal
Alloy
Ideal

241
236

36
38
61

37

Alloy
Alloy
Alloy
Alloy

3
3

41
42

Alloy
Alloy

3
3

46

Alloy

3%

48

Alloy

Dlam.

:; t
z
;
:

bushIn,

;2
2%
2%

5
5

sf4

s $5
5%

5%

' d & ~ $3

ig
3%

6
6
6
6

Trolley Harp Details. Trolley harps are of malleable iron, the

weight being kept as low as ossible consistent with strength and
weanng ualities. Limiting gmensions depend upon those of the
trolley wteel to be carried. In most cases the harp is riveted to a
steel rod which is inserted into the
le and riveted fast. The
trolley axle pin is of case-hardened colEolled steel, and the contact
springs and washers are of s ring bronze. The current should be
conveyed from the wheel to &.harp through the springs and washers-not through the axle pin. I t is most important, therefore, to
see that contact washers and springs are functioning roperly, and
that the latter are securely riveted into good electricafcontact with
the harp. See Fig. 4 for details of one design
Current Carrying Capacity of Trolley Wheel. The following
values of current carrying capacities of trolley wheels when traveling a t various speeds are from the Standard Handbook for Electrical Engineers. The pressure between trolley wheel and trolley
is from 2 0 to 40 Ib.
Speed. m~lesper h r

Current capacity, amperes

1000

10

850

20
650

30
550

40
400

50
300

60
200

Trolley Wheel Defects. The following consideration of possible

defects in trolley wheels is from a paper by C. W. Squier,

368

ELECTRIC RAILWAY HANDBOOK

Electric Railway Journal: The greatest trouble with trolley

wheels arises from the difbculty of securing satisfactory lubrication
and of conducting the current from the wheel to the ha . Some
of the most frequent defects are as follows: sides bent,Troken or
chipped; double groove; holes in sides; flat spots; bushing hole too
large; burned wheels; sides loose and rim worn off.
Bent, Broken or Chipped Sides. Bent, broken or chipped sides
usually result from wheels coming off the wire and stnking some

part of the overhead construction. The number of such troubles

can be reduced by proper attention to the lubrication of the trolley
stands so as to insure their free swiveling, by making sure that the
wheels stand perpendicular to the car roof so as to make proper
contact with the trolley wire, by keeping the side bearings properly
adjusted so as to prevent wear by the swaying of the car when
operating a t high speeds, by renewing the trolley bushings and axles
before they become unevenly or excessively worn, and by keeping
the trolley tension properly adjusted.

TROLLEY WHEELS

369

Dorlble Grooves or Holes i n Sides. Double grooves or holes in

the sides arise from the condition that the wheel is not following
the wire properly, and this may be due to improper maintenance
of the overhead construction. These defects also occur frequently
on cars that are operated from one end on lines with frequent curves,
caused by the wire riding the flange while rounding a curve. If
the harps are not straight in the poles so that the wheel is maintained
at an angle to the wire, or if the side springs are weaker on one side
of the car than on the other so that the car body does not rest level
when loaded, the trolley wheels are liable to wear unevenly. When
a wheel is found wearing to one side it can be made to wear straight
in some cases by reversing it in the harp.
Flat Spots. Flat spots are caused by the sliding of the wheel on
the wire on account of imperfect rotation. Flanges frequently become slightly bent and rub against the harp, or the side springs
and washers may be too tight or exert too great pressure against
the side of the wheel. Flat spots usually start with a very slight
slippage between the wheel and the wire, but when a spot is once
started it increases rapidly.
Bushing Hole Too Large. Oversized bushing holes are caused
s
have become
bv lack of lubrication or bv shunts and s ~ r i n gwhich
worn and loose that the& pressure is to6 lig& to properly conduct
the current to the harp. A slight burning or pitting results from
the carriage of the excessive current through the bearing and axle.
This causes the bushing to bind on the axle, and if it is not a very
tight fit in the wheel, rotation will take place around the bushing
instead of the axle. Under such circumstances only a few trips
are required to wear out the bushing hole in the wheel.
Burned W h e l s and Sleet Troubles. Burned wheels are caused
when the wheel becomes se arated short distances from the wire
so that rapid arcing and Zstructive burning take place. Sleet
destroys the wheel very rapidly in this manner. At other times it
is found that if a wheel is out of balance its centrifugal force as it
rotates will break the contact with the wire. The trolley tension
on cars found with burned wheels should always be tested carefully. In sections where sleet storms are frequent, sleet cutters
are used to advantage. On large systems, however, the removal of
trolley wheels and the installation of sleet cutters assume enormous
proportions and come as an additional task for the shop forces a t
the very time when they are usually busy with snow equipment.
Large city roads usually prefer to depend on the frequency of the
service to keep the sleet off the wire, or they install sleet cutters on
but a few cars of each line, depending on these to scra the ice
clean and thus prevent excessive burning- on the wheels oKhe other
cars.
Sliding Contact Shoe. Several types of sliding contacts, designed
to be substituted for the trolley wheel, have been developed. One
of these, the Miller trolley shoe, is shown by Fig. 5. The forged
steel shoe has its contact surface polished smooth and to about the
same cross contour as the standard trolley wheels, but has a longitudinal curvature of much longer radius than the wheel, thus giving
a greater area of contact than the wheel. The contact shoe is

370

ELECTRIC RAILWAY HANDBOOK

hung in the special harp and electrically connected to it by a copper

shunt cable, as shown in Fig. 5. The sliding contact shoe eliminates
trolley wheel bearing troubles, including lubricat on; it makes
better contact with the trolley wire; it eliminates vibration due to
wheel unbalancing, and therefore should reduce arcing and noise;
it is claimed that its l i e is greater than the trolley wheel, and that
by its use there is less total wear on trolley wire, due to reduction
in wear due to arcing; one of the disadvantages noted is increased
diiculty in backing cars. H. Savage, of the Detroit United
Railway, reports life of the sliding contact shoe as about 2% times
that of the standard trolley wheel. He says that an examination
of trolley wire where contact shoes had been used continuously for
2 3 months showed that the wear on one-way wire was very
slight, the wire having a brownish
color, slightly glazed; while on wire
in two-way o eration the wear was
f
more noticeahe, but not excessive.
The brownish color and glaze do
not a pear on the two-way wire.
He arso found that on wire where
wheels and slides were both used
the wear was much more noticeable
hb'p
than if slides or wheels alone were
used. This was true also for a short
time after all trolley wheels were
replaced by slides. I t has been
pointed out, in connection with
trolley wire wear, that the sliding
contact is likelv to cause considerV
more wire wear at low than
Pic. 5.-Miller trolley shoe and
a t high speeds; this without doubt
harn.
. is partly- due to the fact that the
coefficient of friction is greater a t low than a t high speeds, as well
as to the generally larger currents taken through the contacts a t low
speeds.
Trolley Base. The functions of a trolley base are to provide
so flexible a support for the tro!ley pole that it will have free
movement both laterally and vert~callywhile exerting the pressure
necessary to keep the trolley wheel in contact with the wire. A
freely swiveling base is most essential in order that the trolley wheel
may follow the overhead line. To piovide this extreme sensitiveness, roller and ball-bearing bases have come into general use.
Both types operate satisfactorily when provided with properly
hardened parts. By using ball bearings a lower base can be
obtained than with roller bearings, but roller bearings give greater
wearing surfaces. Various pressure spring arrangements are in
use. Some bases have springs in tension and others in compression,
some bases have but a s~nglespring, while others have a battery of
spnngs. The following brief consideration of possible trolley base
defects is from a paper by C. W. Squier, Electric Railway Journal:
Brokcn or Weak Springs. A spring which is in tension breaks
most frequently at the end loops or a t the bend where the loop
r

TROLLEY BASE

joins the first turn of the spring. The number of such breakages
can be reduced by keeping the parts over which the springs hook
in good condition so that the loops have a maximum amount of
bearing surface. By the use of a magnifying glass one may often
discern small cracks in new springs a t the bend from the loop to
the first turn of the spring. These cracks are evidently due to the
method of manufacture, and while such springs are strong enough
to withstand heavy strains, they will ultimately break a t these
fissures. Springs are also weakened from the gradual loss of
their ability to resist elongation or to the compression which follows
the slow accumulation of a permanent set. In older types of bases
the full power of the springs may not be available for producing
tension at the trolley pole because of excessive friction in such
parts as the cross-pin that forms the up-and-down bearing for the
trolley pole, the cross-head in its guide, the side rod bearings which
carry the trolley spring
pressure and the trolley
springs themselves on their
----guides. I t is very di5cult
to keep these parts lubricated, as rain forces the oil
out on the roof of the car
and the windows and sills
often become bespattered
and soiled from oil. In
the later designs of stands
an eKort has been made to
provide bearings which do
away with the necessity
for constant oiling a t friction points, and the de- PIG. 6.-Bushings in trolley base yoke.
signen have also endeavored to provide for a uniform pressure of
the trolley wheel on the wire a t different elevations. Tension
springs are sometimes weakened by overstretching them during installation. One bad practice, for example, is to force a
screwdriver or other sharp tool between the spirals of the spring
and to use this as a lever to hook the eye over its post. This is
likely to force the spirals apart to such a distance that they will
not come back to their original position.
Worn Bearings and Pins. Worn bearings a t various parts of the
base are a constant source of trouble. In a great many cases the
desire to keep down the weight has left insu5cient material to
permit the boring and bushing of worn bearings. Some roads have
gone to the expense of making patterns of new castings which are
rovided with su5cient material to allow all wearing parts to be
pushed. Fig. 6 illustrates one method of bushing a yoke to take
care of excessive wear.
Nut, Bolt, Ball and Roller Troubles. All nuts should have lock
washers, and if di5culty is then experienced, cotter keys should be
added. Where ball bearings are used, the races are usually insulated from the socket casting to prevent them from carrying the
trolley current. Notwithstanding this precaution, the frequent

372

ELECTRIC RAILWAY HANDBOOK

discovery of burned ball races and balls shows that they do carry
current. This action may take place in several different ways.
For instance the contact shunts that carry the current from the
socket casting to the base plate occasionally get bent out of position
so that they bear on the race and socket casting a t the same time.
Current then passes across the contact face of the shunt to the
ball race and thence through the balls to the base plate, causing
the burning of the balls and race. Moisture and dirt collect on
the surface of the insulation, and the current then has an easy path
to the ball race. Again, contact shunts are sometimes torn off
entirely and current then passes over the sutface of the insulation
from the socket casting to the race. When new ball races are
installed care should be taken to see that the lower edge of the ball
race does not project beyond or come flush with the contact surface
of the socket casting. This precaution will prevent the contact
shunts from bearing on the race,
for then they cannot touch the
edge of the race while still remaining in contact with the faceof the
socket casting.
Burned or Broken Shunts and
Burned-of Leads or Tetminals.
Of necessity, trolley bases must be
made very low. As a result, the
distance from the socket casting to
the base plate is usually not more
than I to 1% in. This, then, is
all the space that is available for
contact shunts. I t is very difficult to get in this space a spring
PIG. 7.-Trolley. base with contact which is eficient enough to
ring.
give the necessary current-transmitting pressure of the contact shunt against the socket casting.
The springs soon take a permanent set and arcing then takes place
between the contact surface of the socket casting and the shunt.
As a result, the shunt is burned away or else the contact surface of
the socket casting becomes so rough from the arcing that the shunts
are torn off. The surface of the socket casting is thus destroyed,
and in most cases it is necessary to install a new casting. Fig. 7
shows a brass contact ring which was made to screw to the face of
the socket casting. As this ring is of brass, it forms a better conducting s d a c e than the steel casting, and if it becomes burned
it can easily be replaced a t small expense. This ring also permits
the re-use of socket castings that have become slightly burned, since
the lower face of the casting can be turned off for the reception of the
contact ring. The common form of terminal used on trolley bases
consists of a hole in a lug to receive the lead. The hole is provided
with set screws to clamp the lead in place. Leads become loose
from the working out of the set screws or because the screws are
not tightened when a new base is installed. If the screws stay
in place, they soon become so rusty that it is almost impossible to

TROLLEY MAINTENANCE

373

remove them when it is necessary to install a new base. If the

leads become loose, the arcing thereby initiated soon bums away
the lead and terminal. A better form of terminal which can be
quickly and easily removed consists of a standard soldered terminal
such as is used for the ground leads on motors. If the trolley lead
is soldered into this properly, there is no danger that the lead will
burn off a t this place. This terminal is then bolted to the base
plate, where a spot-finished surface is provided to give a good contact
surface.
Inspection and Lubrication of Trolley Wheel, Stand, Rope and
Retriever. (A.E.R.E.A. Miscellaneotrs Methods and Practices.)
Examine wheel and see that bushing and hub or spindle are not
unduly worn and that outer rims are not bent, nicked or worn out;
see that cotter keys holding spindle in harp are in good condition;
see that contact springs and washers are sufficiently tight between
harp and hub of wheel to form a good contact, but not so tight as to
allow the wheel to slide on the wire. See that the spindle is tight
enough in the harp to form a good contact, and that the spindle
holes in the harp are not too badly worn to prevent this contact.
Examine the harp and see that it is not loose on the pole and that
rivets holding same are tight. Examine pole for cracks, bends or
flaws, and see that it alines the wheel properly with the wire, and
if not, loosen the clamp bolts and turn with a pipe wrench until
wheel is in proper aljnement, leaving the wheel on the wire during
the operation. See that clamp bolt and nut holding same ahd base
bolts are all tight and in good condition; examine springs and see
that they have sufficient tension, allowing sufficient space between
the coils for compression when the pole is pulled down to the roof
of the car. In bases with more than one spring, see that they are
equalized on each side, both sides given the same tension. Give
s rings su5cient tension so that the wheel will have pressure of
a L u t 2 0 to 25 Ib. against the wire in city service and from 35 to
40 Ib. on high speed. Make this test where the wire is of standard
height. I t can be done by using a hook scale, or by hanging an
old brake-shoe or other weight to the trolley rope. See that trolley
board is securely fastened to the rocf of the car.
Trolley Lubrication. Lubricate trolley wheels a t each inspection
and wipe all surplus oil from wheel hub after lubricating. Lubricate bases when necessary. This can be determined by swinging
the pole from side to side below the wire. If the base operates
freely no lubrication is necessary. Great care must be taken not
to allow any surplus oil or grease to reach car roof while lubricating
the bases and wheels.
Trolley Rope. See that rope has a firm fastening with the harp;
that it is not chafed or showing signs of wear where it comes in
contact with the hood; that it has not been broken and that it has
no unnecessary knots in it.
Trolley Refrimer. Trip the retriever and see if it operates
properly; that the tension in the retriever spring is not so severe as
to break the rope or pull the trolley down so severely as to damage
the hood or roof; see that rope works freely when resetting, and that
it is of such length that it will not pull trolley from the wire where
the wire is high, such as at railroad crossings, etc.

374

ELECTRIC RAILWAY HANDBOOE

Exha Trolley Pole. See that all interurban cars are supplid
with an extra pole, fully equipped, and in good condition.
Trolley Pole and Pressure between Trolley Wire and Trolley
Wheel. Poles of such length that the longitudinal axis of the p ~ ~ l e
makes an angle of from 35 to 45 deg. with the axis of the trolley
on tangent track are in most common use and the most common
length is 12 ft. Height of car, height of trolley wire, aiinement of
track and other local conditions may require a longer or shorter
between the limits ro it. and 16 it. A light, flexible pole is prderable to a heavy one, as it adapts itself to irregularities in tbe
overhead system and does comparatively little damage to th?t
system when the trolley wheel jumps from the trolley wire. Tbe
pressure between trolley wire and trolley wheel should be such that
the trolley wheel will follow the trolley wire with as little wear ?S
possible. High-speed service requires a greater pressure then
low-speed service. When the roper pressure has been decid~d
upon, the adjustment should
made by measurement. This
adjustment has been commonly made by hanging a weight a t tbe
end of the trolley pole or by fastening a spring balance to the troll6Y
rope and applying the required tension through the spring balance,
then adjusting the trolley stand spring till the wheel is of the proper
height to just touch the trolley wire. Adjustment by use of the
weight has been found to be the most satisfactory.
In operating p ~ a t G t p~essure
~,
between t)ne troNey wheei a&
wire varies from 16 to 4 0 Ib., pressures from 16 to 25 Ib. beipg
usually considered sufficient for city operation, while 35 Ib. is generally used in high-speed interurban operation. As the alinemeflt
of trolley wire and condition of track approaches perfection, tbe
trolley wheel ressure a roaches the minimum, and no more
pressure shoulcfbe used tian is required to hold the trolley on the
wire a t the required speeds with pro rly lubricated trolley bases
Trolley Poles for Double Truck 8&s. The general practice is
to use two trolley poles on double truck cars when operated doubleend, the trolley base usually being mounted directly ovet the center
of the truck, although less offset in trolley wire is required when
the trolley base is mounted between the truck center and the efid
of the car. With two trolley les there is less likelihood of tbe
trolley wheel being pulled off
wire from rope friction over the
rear end of the car or a defective retriever; with two trolleys
the car is not entirely disabled by the failure of one; trolley wheel
replacements are made more easily when the base is located nearer
one end of the car; and the offset required in trolley wire on curves
is less as the trolley base is brought nearer the car end. With short
double truck cars in city service, where the distance between
truck centers is not great, and the trolley wheel may be easily observed from the rear platform, a single trolley pole may be used
to advantage.
Third Rail Collector. The third rail collector has the greatest
current collecting ca acity of any current collecting device used m
electric railway worE. In the section on electric traction of the
Standard Handbook for Electrical Engineers, it is stated that tests
have been made which indicate that current may be collected a t

tE

THIRD RAIL COLLECTOR

------------------- /g'---------------RG.8.-Third rail co11ector. New York submy.

376

ELECTRIC RAILWAY HANDBOOK

the rate of 2000 amp. from a single shoe a t a speed of 35 miles

per hour and 500 amp. a t a speed of 70 miles per hour. The two
general classes of collectors are that in which the contact pressure
is furnished by gravity and that in which the contact pressure is
furnished by a spring. The shoes are generally made of wrought
iron or cast iron and in some cases a steel wearing surface is used.
Fig. 8 shows the details of the third rail collector designed for
operation on overrunning protected third rail in the New York
subway. Contact pressure is furnished by a spring. The shoe is
arranged with a stop by which its downward movement is limited.
The maximum poss~bledrop is 156 in., but this may be reduced by
shims. Connection from the shoe to the supporting brackets is
made by a flexible cable wound around the shaft so that it will not
be affected by the motion of the shoe.
1 ~ ! 1 r o r n 4 w eLine

eetiq Bod

~ o n t a c d \ ~ e a r f i n gPI1 Roll
@

FIG. 9.-Third

rail collector. Philadelphia and Western.

Fig. 9 shows the automatic third rail collector for use on overrunning third rail of the Philadelphia and Western Ry. The pressure between the shoe and third rail is about 10Ib. and is furnished
by a spring which also serves to hold the shoe in the inoperative
psition. A tool-steel insert in the shoe is used for the wearing
plece. This piece is held in place by two rivets and babbitt. Durmg the winter the shoe is slanted slightly downward in order to
make contact with the far side of the third rail, as that part of the
rail is usually free from sleet. This slanting is done by means of
the eccentric stop, the four quarter turns of which will also take
care of a 1.5-in.reduction in diameter of the car wheels. There is
also a rack adjustment on the car trucks for raising and lowering
the collector beams. The automatic folding and unfolding of
the shoe is obtained in the following manner: Where the shoes
should either open or fold, as the case may be, two parallel metal

THIRD RAIL COLLECTOR

377

strips which are sloped in opposite directions are placed along the
tracks and set several inches apart. If the shoe is open, it will bear
against the rising strip until it assumes a nearly vertical position,

and the spring then acts to close it. The rising of the shoe, however, is coincident with the lowering of a tail-piece which is a t right
angles to the shoe. When the shoe is to be opened, this tail-piece

378

ELECTRIC RAILWAY HANDBOOK

will be within range of the other indined strip, so that it will be

forced upward whlle the shoe moves downward to make contact .
with the conductor rail. The eccentric stop and the tail-piece .
prevent the shoe from drop ing when unsupported by the third rail.
Fig. ro shows a third raifcollector and sleet brush of the Brook- *
lyn Rapid Transit Co. This collector was designed to eliminate
all links and castings, it is of wrought iron, and it allows a maximum
height variation of 2.5 in. The wrought iron shoe replaces a linksuspended casting which was too fragile. Copper shunts were used
on the casting, but the Bat springs
were found to be sufficient conductors
on the type illustrated. The sleet

Fig. I I shows a third rail collector

for use with either over-running or
under-running third rail. The tension
or pressure of the shoe on the rail is

Neutral Poritlon of Shoe

A-&I

A- 9%'

RG.11.-Third

For Lowut Horlzontil Position

For Highuc Horizonwi P d t i o n

rail collector for over- or under-running.

regulated by the pos~tionof the spring in relation to the pivot or

bearing points and may be set a t any reasonable amount.
Pantograph Collector and Bow Collector. The pantograph
collector differs from the bow collector essentially in the movement
and position of the contact shoe. The shoe of the pantograph
collector (Fig. 14) is controlled by a pantograph structure and
moves in a vertical plane through the center of the mechanism as it
rises and is depressed in action, while the shoe of the bow collector
(Fig.
13) is traded in a manner similar to that of the common trolley
wheel and moves in a vertical curve. When used in high voltage
operation, both types are insulated from the car by being mounted
on porcelain insulators. Both types are generally placed in action
and removed therefrom by means of a compressed air mechanism
and in high voltage service this is often arranged to lower the
collector automatically when the high voltage distributing box is

PANTOGRAPH COLLECTOR

379

opened. The scheme of the pneumatic operation of the low pantogra h on New York Central locomotives is shown in Fig. 12. Near
eacE master controller in the cab there is a valve by means of which
the pantogra h shoe may be raised or lowered. When air is
a p hed, the sloe is lifted so as to make contact with the overhead
raf. When air is released, the shoe drops; also if the shoe runs oB
the rail, it is tripped automatically and drops. Moving the
handle forward operates the pilot valve, by means of which a
slide valve is thrown to admit air from the reservoir to the cylinder
of the contact shoe device. Pulling the handle back operates
another ilot valve and the slide valve is thrown over to connect
the air clamber of the contact device to the exhaust. The handle
will spring back to the middle position from either direction.
There are two of these overhead contact shoes which are controlled
in common by either valve in the cab.

--z?

opuln' -I'.

PIG. 12.-Low

11.-.

pantograph collector. Ncw York Central.

Either pantograph or bow may be used when high speed, voltage

or amperage or overhead construction will not permit satisfactory
wheel trolley operation. In many places a pantogra h is used
in suburban service, but in city service the overheadl construction usually permits only a wheel trolley. This is provided for by
equipping each car with a pantograph in the middle and a wheel
trolley le a t each end. Both the pantograph collector and the
bow colrctor are built with many variations in the details of construction. The pantograph collector is built either with the simple
antograph and rigid contactor or with the pantograph base to
Follow great changes in trolley mire height and having the contactor
spring-supported to follow minor unevenness. The bow collector
may have a framework or pole or a pantograph base to follow variations in trolley wire height and from this the contactor is generally
carried on a lighter secondary trailing structure made flexible to

380

ELECTRIC RAILWAY HANDBOOK

follow minor unevenness. The following, relative to successful

current collector operation, is by Otis Allen Kenyon, Elec. Ry.
Journal:
The successful o ration of any system depends upon the ability
of the contactor t
o
k in conpct with the ,re.
Failure to do this
produces arcing and Lmmenng. The arcing destroys both the
collector shoe and the wire, while the hammering accelerates the
wear, kinks the wire and breaks the fastenings. The difficulty of
keeping the collector in contact with the wire increases very rapidly
with the speed, and the minimizing of this difficulty is one of the
important problems in high speed operation. The variations
in the position of the wire over the track and the swaying of the
car require the collecting device to be so constructed as to adapt

PIG. 13.-Bow

collector.

FIG.14.-Pantograph

collector.

itself to all such variations, thus forming a flexible connection

between the wire and the car. At low speeds it is comparatively
easy to design such a device, since all variations in position take
place so slowly that the inertia of the coUector does not prevent it
from following the wire. However, a t high speeds the effect of
every little variation is exaggerated and tends to produce vibration.
Large variations in position, such as are caused by passing under
bridges, etc , when made gradually, have little or no effect on vibration. The three principal causes of vibration are:
( I ) Unevenness of the contact wire, due either to imperfect
suspension or non-uniform wear.
(2) Vibration and swaying of the car or locomotive.
(3) Inertia of the contact device.
Contact shoes have been made of copper, steel, aluminum and
various alloys in attempts to get a considerable life of shoe together
with small wear on the trolley wire. A shoeof "U"-shaped section
containing lubricant in the form of grease has been found to reduce
wear and singing. Aluminum shoes are light, consequently they
have comparatively little inertia which makes possible a collector
structure of high natural period of vibration.

381

PANTOGRAPH COLLECTOR

Wear of the wire and shoe can be traced to two causes, mechanical
and electrical. The mechanical causes are due to pressure and
cannot be entirely eliminated. Non-uniformity of pressure is the
most serious trouble. The springs can be so designed as to make
the pressure constant for any position of the collector. However,
the speed a t which the collector changes its position affects the
pressure to an extent which can only be controlled by reducing the
~nertiaof the mass to be moved. Throughout Europe the average
pressure of the contact shoe against the wire is about 11 lb. The
electrical causes are secondary, being due to arcing, which is caused
by insu5cient contact surface. Arcing is especially destructive
in that once started the surface is left more predisposed to arcing
than before and conditions rapidly go from bad to worse. The

nuul .,a
Cootnet Roller
Bh.lt C. Balled Steel
z-3

Busbins Brass

lo----l
Wd.BL0Ck

FIG. IS.-Roller trolley. Key Route. Cal~fornia.

cause is due to imperfect contact and vibration and the resulting

wear is a function of the current rather than of the power. For
a given power, the higher the voltage the less d e s t ~ c t i v is
e the arc.
Soot deposited on the contact wire by steam locomotives increases
the wear to an astounding extent. The Swedish commission found
that the wear of the shoe on a soot-covered wire is about ten times
as rapid as on a clean one. The aluminum shoes on the various types
of collectors ran about 1500 miles on soot-covered wires. I t is estimated in the report that on clean wires they would run a t least
12,500 miles.
The practice of the Swedish tramways indicates an
average life of such shoes, when used in tramway service, of 12,500
miles before renewal, but if left until worn out they will run as
high as 62,000 miles.
Collecting Capacity of Pantograph. Tests were conducted by the
General Electric Company a t Erie, Pa., in July, 1923, in which
currents exceeding 5000 amperes were successfully collected by

382

ELECTRIC RAILWAY HANDBOOK

PIG. 16.--Slot Plow. Washington.

pressure of the pantograph against the trolley was 30 to 35 Ib. The

working conductor was two 4/0 grooved wires suspended from the
catenary messenger by clips spaced IS ft. apart on each wire
and alternating. Such currents are greatly in excess of those re-

SLOT PLOW
quired even in heavy locomotive practice, but the tests demonstrated the possibilities which could be met with the pantograph
slide and with the working conductor in excellent condition.
Roller Trolley. The roller trolley is good in current collecting
capacity, but it has more inertia than the pantograph shoe. For
these reasons it is suited to service at heavier currents and a t lower
speeds than those to which the pantograph shoe is particularly
adapted. Fig. 15 shows a roller trolley used successfuUy on the
Key Route, California. The roller is mounted on a antograph
wearing
frame and weighs, complete with spindle, 28 Ib.
surface is a tube of non-arcing brass, supported on a wooden roller.
The height of the trolley wire above the head of the rail varies from
14 ft. 6 in. to 22 ft., but by the pantograph the pressure of the roller
against the wire is kept nearly constant at about 34 Ib. The average mileage of the rollers is 55,000.
Top Contact Collector (Oerlikon Collector). Current is collected from either a trolley wire suspended from an inverted catenary or a trolley wire stretched tightly over insulators a t the side of the
track. The collector consists essentially of a curved hinged arm
which swee s over nearly a semcircle in a plane transverse to the
longitudinaraxis of the track. This collector is supported on insulators on the side of the car top. Normally the arm rests on top
of the trolley wire with a pressure of about I .5 Ib. On cross-overs
and in tunnels, where the trolley wire is carried over the track, the
arm swings toward the center of the car, and is depressed, making
contact progressively from the top around to the side and then
underneath the trolley wire. In addition, the saddle which cames
the arm is movable laterally, increasing the radius of action.
Slot Plow. The slot plow is arranged to collect current from conductor bars which are located about 6 in. apart and a foot below
the track rails. I t consists of a flat frame made of insulating
material suspended from the car and carrying the necessary conductors from the conductor bar to the car. This frame extends
through the $$-in. or %-in. slot between the slot rails and carries a
contact shoe a t either side of its lower end. I t insulates the contact
shoes from each other. The contact shoe is generally made of
chilled cast iron and is held against the conductor bar by a flat steel
spring by which it is fastened to the frame. The frame cames thin
wear plates of hard steel for protection against excessive wear in
rubbing over the slot rails. Electrical connection between the
contact shoes and the upper end of the plow is by fuse connectors
and rubber-insulated conductors. These fuse connectors are connected to the contact shoes and serve to protect the plow from damage b y short circuit from shoe to shoe. Fig. 16 shows a slot plow
used by the Capital Traction Co., Washington, D. C. The suspension of the plow from the truck must admit of free lateral movement in order that the plow may not be damaged by truck play and
unevenness of track and slot rails. Where the car moves between a
conduit section and a trolley section, the plow is removed or replaced by a man stationed in a pit a t the junction of the two sections.

TE~

SECTION V I I
TRUCKS

General Classification. Trucks are of two gcncr.tl cl.lsses,

namely, the single truck ant1 the double or swivel trutk, thc: I:~ttcr
sometimes called the bogie truck.
Single Truck Cars are intended for city street scrvicc ant1 whrrc
the m a ~ i m u ms1)ced does not esceed 2 j milcs per hour. 'The two
axles of the single truck arc more or less rigitlly allncd by side
frames, therefore the wheel base must be short in order to ncjiotiate
sharp curves. This linlits thc length of car to a maximum of almut
jo feet overall.
Double Truck C a n are equipped with two distinct trucks joinrtl
together by the car lmtly framing. Each swivel truck consists of
two or more axles centered in common side frames which are joinccl
together by cross pieces. The central cross members usually form
;I transom surrounding a spring supportc:l Imlstcr.
The Imlstcr
carries n center plate on which the truck s\vi\.els and upon whicli the
car body rests, and also side bearing plates which maintain the body
levcl. The usual type in electric service carries two axles; for very
high s eetl service or for very heavy cars, three-axle trucks m:ly IJC
t~csirlE~e.
Design of Trucks. I t has come to be generally recogni~edthat
definite engineering principles should be applied in the design of
trucks for electric railway service, ant1 that weight of trucks is not a
prime essential to easy riding qu:llitics. '1 numl>er of variatio~lsin
the type iind arrangement of springs and in the general design ancl
arrangement of truck partsatediscussed in thc following paragraphs.
I n addition to those matters of general design, the following lmints
have been mentioned by Norman 1.itchiicld as tl~ose which
should be consideretl in connection with truck clesign: weight of car
and of load, standing; shifting of weight front side to sitle, due to
centrifugal force; shifting of a portion of the load from the rear to
the front truck in braking, ant1 vice versa in acceleration; similar
shifting of a portion of the load from one pair of wheels to theother
pair in the same truck under similar conditions; the forces set up by
the motor during acceleration; the flywheel effect of rotating parts;
the forces set up by friction of brake shoes on the wheels during
braking; and the distortional effect in rounding curvc3.
Developments in Truck Design. Reductions in weight and in
maintenance, together with improvement in riding qualities, have
been the outstanding objectives in the development of truck design,
principally due to the dcvelopmttnts in light weigl~tcar c o n b t r ~ ~ t i o n
The ret~trnto the single truck drsign with the Dirney typc of safety
car, and the light weight construction ,demanded, produced n
marked development in the single truck. rbe principal differences
385

ELECTRIC RATLWAY HANDBOOK

in present types of construction found in trucks built by various
manufactum are in methods of constructing the side f n m e
and in the arrangement of spring to groduce easy riding. I n
addition, the layout of brake rigging an design of brake hangers
for the purpose of rliminating chattering and reducing maintenance
represent important variations in construction. Combined with
the arrangement of springs to rovide vertical flexibility, various
methods and devices are utiIii.2 for the purpose of allowing lateral
movement of the bolster and at the same time limiting the amplitude of its oscillations. Further refinements in design by various
manufacturers a n rimarily for the purpose of increasing rigidity
and durability anzdecreasing maintenance costs. Rather wide
departures from conventional designs have been made by several
operators who have felt the need for more fundamental develop
ments than were available in the products of the equipment maaufacturers, and this work is a t present in process of development;
Fig. 14 is an example.
Single Th&.
In the construction of single trucks a number of
designs are in use. The demand for reductions in weight and the
requirements of minimum maintenance and easy riding are reflected
in the variations in design. There is a tendency to avoid built-up
side frames and to secure the rigidity and freedom from maintenance offered by a one- iece side structure, as well as to avoid heavy
castings wherever possi&e and to substitute lighter weight forgings.
For the pu
of obtaining minimum weight, one type of construction u t x two deep channels of high grade steel assembled
so as to form a closed box structure filled with wood and welded
together along the upper comers. Of the one-piece side frames, one
type is machined from a solid flat billet, another is forged to shape,
and in another the side frames a n ressed out of a com ratively
thin plate, providing reinforcing Lnges and ribs a t t c critical
points.
The application of brake rigging to the single truck is by'several
arrangements designed to eliminate chattering and insure a true
and square contact between the shoe and the wheel. The method
of employing two ralkl link hangers, a lied either directly to or
adjacent to the siPfrune so as effcctiv$ to guide the rhoc in its
movement back and forth, as well as e design with a center
hanger back of the brake beam and a single brake hanger for each
shoe head a t the side, are both widely used.
The arrangement of springs is impor-t
in connection with the
results obtained in the opmtlan of singk trucks. With axles s u p
ported squarely in the truck fnnx a t the journals, the requirement
or negabating n h r p curve8 existing in m ~ scities
t
limita the allowable wheelbase to between 8 and 9 ft. as a maximum. Since the
car body of a single tmck car extends onniduably beyond these
lengths, the overhang each side of the wbeei line is comparatively
large and results in 8 tendency toward teetering and nosing in
operation. Various methods a n employed to overcome thin
condition, practically all of which center around the general i d a
of increasing the effective spring base by the extension of the truck
side frame a considerable distance beyond the whed line a t arch end,

&

SINGLE TRUCKS

387

to act as a support for spring. By the use of overhanging halfelliptic springs secured a t one end to the truck side frame, or

y r t n elliptic springs rigidly secured to the side frame close to

t e pedestal, the same effect is obtained. Fig. I shows a simple

spring arrangement with a half-elliptic spring fastened rigidly to

the j o u d box, utd short helical springs a t either end extending

to the car body support. I n this construction the effective spring
t m a is equal to the wheelbme of the truck. Fig z illustrates the
next step in providing a more flexible spring construction. In this
arrangement the truck ride frame is supported on helical springs

388

ELECTRIC RAILWAY HANDBOOK

mounted on extended ears on the journal box, and the half-elliptic

is, as in Fig. I, again sup rted on the journal box, but its inside
end is fastened by a long
to the truck side frame instead of to
the body. The outside end of the half-elliptic is connected to the
body through a long helical suspension spring, which allows considerable lateral and longitudinal motion. The long link employed for
fastening the inside end of the half-elliptic to the truck side frame
also contributes to a comparatively free movement. Fig. 3 illustrates the next step in the development of this type of truck, in
which the loose and flexible connection to the body a t the outside
end of the half-elliptic is eliminated, and the link connecting the
inside end of the latter to the truck frame is materially shortened.
The final development in this design to produce easy riding and a t
the same time limit excessive oscillation is shown in Fig. 4, where the
construction is similar to Fig. 3, except that a short compression
spring is interposed in the link connection t o the truck frame for

FIG. ,.-Driving

connections between single truck and car body.

the purpose of reducing the strains imposed on the parts a t this

point and giving a little added vertical flexibility. I n Fig. 5 the
side frame is carried on short helical springs resting upon ears
extending from the sides of the journal box. The main body su
porting spring consists of a stiff quarter-elliptic rigidly fastened to tl%
side frame adjacent to the pedestal. The connection to the body is
usually made through a short helical spring, flexibly connected t o
the end of the quarter-elliptic. A flexible device for absorbing lateral
strains between the body and truck and for maintaining the body
and truck in alinement is also shown in Fig. 5, as well as the details
of the quarter-elliptic spring fastening to the side frame, the brake
hanger arrangement, and the journal box construction. Still
another method of supporting the side frame on short helical
springs set on extension ears on the side of the journal box, is shown
in Fig. 6. The quarter-elliptic spring is fastened with a strap and
projects under a lip on the pedestal frame for firm anchorage.
This type of truck employs a centrally supported brake-beam hanger
with the brake hangers supported from gusset plates bolted to
the side frame. The brake hanger is of the half-ball type, in which
a semi-ball and socket arrangement with a compression spring on
the connecting bolt is utilized to compensate for wear.

DOUBLE TRUCKS

3ay

Two arrangements for transmitting the driving force between the

single truck and its car body are shown by Fig. 7, in addition to the
one shown in Fig. 5.
The Radiax truck is a n example of a sinde truck in which the
two axles are not carried rigidly parallel, 'but are allowed some
motion, so as to adjust themselves more nearly radially in passing
around sharp curves. Such construction permits a longer wheel
base to be used than would be possible with a rigid truck on the
same radius track curves. The principal feature of the Radiax
truck is the swing link suspension of the truck from the journal
boxes which is illustrated by Fig. 8, the double bolt arrangement
tending to return the axles to their original alinement when on
straight track.
Swivel or Double Trucks. I n this type of truck, modem construction shows side frames forged in one piece, made of a steel
casting, built up in the conventional
diamond arch bar construction, or built
up of structural shapes. The forged
side frame has the advantages of a onepiece rigid structure, in which by proper
design the metal can be distr~buted
accurately in proportion to the stresses
imposed. This type gives a very rigid
truck and freedom from maintenance
trouble on the side frame structure.
I n the application of cast steel to
modern trucks, the M.C.B. equalizer
bar type of truck, as shown in Fig. 9,
lends itself to the elimination of the
8.-Swmg
link of the
heavy pedestal castings, which were
"Radlax" truck.
used with earlier trucks of the cast
steel type. As here shown, the truck side frame and transom have
been cast in one unit, giving a n extremely tigid structure from the
standpoint of maintenance of alinement, and also allowing the
brake rigging and spring suspension parts to be conveniently applied. I
The arch bar type of truck gives both low weight and first cost.
I n this type of truck high grade heat treated steels of either high
carbon or alloy composition are being used to increase the structural strength of the built-up side frame. In some cases the arch
bar is of the conventional flat section, while in other trucks a channel-shaped member is used for the bar in order to obtain the
maximum sectional rigidity with minimum weight (Figs. roand 11).
I n the spring suspension and brake rigging arrangement, there is a
variation over a considerable range of design. The customary
full-elliptic spring under the bolster a p ears in a number of types,
but various features have been added g r the purpose of obtaining
improved riding performance. The heavy spring plank former1
used under such springs has in some cases been entirely eliminated:
the full-elliptic spring is carried in a swinging hanger or swing link
provided wlth an appropriate socket a t the lower end to form a seat
for the lower spring band, and the two hangers or stirrups are tied
together across the truck with a comparatively light structural

90

ELECTRIC RAILWAY HANDBOOK

PIG. 9.-M.C.B.

type swing bolster swivel truck.

FIG.10.-Arch bar type swivel truck with half-elliptic bolster springs supported from long lmk.
Y

L-1

?dc0

FIG.I I.-Modified arch bar

type swivel truck with helicnl springs at jou-I

under bolder.

boxes and combination half-elliptic and helical springs

W

392

ELECTRIC RAILWAY HANDBOOK

DOUBLE TRUCKS

393

member. Light helical springs mounted on top of the full-elliptic

and set into upper and lower spring caps which are provided
with lugs which come in contact after the capac~tyof the helical is
exceeded are utilized for the purpose of absorbing road shocks and
vibrations while the car is operating under light load. This light
action spring goes out of operation when the spring cap lugs come
in contact, a s d from that point on the total load is carried by the

Flc. 14.-Mlnneapolls

11ght-welght truck w ~ t hband brakes and hollow

axles

full-elliptic. Bolster guide links fastened between the truck transom and the bolster are utilized in some trucks of this general t y e
for the purpose of preventing the bolster from rubbing against t i e
transom sides, I n addition to the simple arch bar type of construction with the bar resting directly on the journal box, some recent
trucks utilize a type of construction in which helical springs are
provided betyeen the arch bar and projecting ears on the sides of
the journal box. The side frame, instead of being clamped down
around the box by the long bolts between the top and bottom bars,

394

ELECTRIC RAILWAY HANDBOOK

is clamped to tubular struts which ride through o nings cored in

OK

the journal box structure, so that the equivalent

pedestal construction is provided.
The use of half-elliptic springs camed longitudinally in links or
stirrups hung over the top of the side frame has been customary for
a number of years, as shown in Fig. 12. The light-load helical
spring described above is utilized with this construction and is
imposed between the top of the half-elliptic and the truck bolsterThe design of link in whlch the half-elliptics are carried is such that
the amplitude of the lateral swing is rapidly checked by the angularity which these stirrups assume. I n modem trucks the entire
theory of securing easy riding from the standpoint of absorbiig
lateral oscillation seems to be the provision of relatively shhrt
suspension links in the spring system, which are com arativqy
free a t the beginning of an oscillation, but which q u i d y assuke
an angularity that checks the amplitude of the swing.
Another type of spring arrangement which has been used to a
considerable extent, particularly in trucks of the modified arch bar
construction, is a type in which a half-elliptic spring mounted
crosswise is camed independently a t either s ~ d eof the truck in a
long stirrup supported from the transom. At the ends of the halfelli tic, comparatively long helical springs are set up vertically
anIdirectly support the truck bolster. This same general type of
construction is represented in Fig. I 3 by two half-elliptics mounted
crosswise a t either side of the truck, one fastened to the truck bolster
above and the lower fastened in short swinging links carried on a
comparatively large structural member extending across the truck.
This lower structural member is used to tie the two sides of tile
truck together, and the usual transoms above are largely made bf
wood. Between the ends of the half-elliptic springs describq,
vertical helical springs are set so as to hold the two half-elliptics a
considerable distance apart.

Electric Railway Axles

Standard Specifications for all types of electric railway axlies

are contained in the A.E.R.E.A. Manual. The following information is a n outline of some of the more important points covered by
the complete specifications.
Process of Manufacture. The steel shall be made by either tile
open hearth or electric furnace process, except that the open hearth
hammered steel axles may be made by only the open hearth proceg,.
Heat Treafmenl. I . Quenched and Tempered Carbon Steel
Axles: For quenching and tempering, the axles, immediately after
forging, shall be allowed to cool to a temperature below the critical
range under suitable conditions to prevent injury by too rapid coqllng. They shall then be uniformly reheated to the proper temperature to refine the grain (a group thus reheated being known as a
"quenching chargev) and quenched in some medium, under substantially uniform conditions for each quenching charge. After
quenching they shall be uniformly reheated to the proper t e m p s ture for tempering or "drawing back" (a group thus reheated beiqg
known as a "tempering charge"), and allowed to cool uniformly.

AXLE SPECIFICATIONS

395

2. Annealed Carbon Steel Axles. The procedure to be followed

in annealing shall consist of allowing the axles, immediately after
forging, to cool to a temperature below the critical range under
suitable conditions to prevent injury by too rapid cooling. They
shall then be uniformly reheated to proper temperature to refine
the grain (a group thus reheated being known as an "annealing
charge") and allowed to cool uniformly.
Turning. Quenched and Tem ered Carbon Steel Axles and Annealed Carbon Steel Axles: The krgings shall confonn to sizes and
shapes specified by the purchaser. Unless otherwise specified

Pro. 15.-A.E.R.E.A.
standard design of motor axles.
NOTE.-Capacities based on steel per A.E.R.E.A. Standard Sycification
for "Annealed Carbon Steel Axles. Shafts and Slm~larFormnas. Al! sizes
shown are finished sizes. Journals and motor-bearin fits to be burn~shed.
Length of motor-bearing fits to suit type of motor use&

by the purchaser, axles, shafts and similar round forgings shall

have a collar approximately 2 in. in width left rough forged on each
forging, and the remainder of forging shall be rough turned, with
an allowance of % in. on the surface for finishing. Unless otherwise
specified, all axles, shafts and similar forgings shall have ends
faced, with an allowance of % in. on each end for finishing, and shall
be centered for 60 deg. lathe centers, with clearance for points.
Chemical Composition. Quenched and Tempered Carbon Steel
Axles and Annealed Carbon Steel Axles: The steel shall conform
to the following limits in chemical composition:
Carbon. ..........................
Manganese.. ......................
Phosihys.. ......................
Sulp ur

.......................

Not
0.40
Not
Not

over 0.60 per cent

to 0.70 per cent
over 0.05 per cent
over 0.05 per cent

396

ELECTRIC RAILWAY HANDBOOK

Rc. 16.-A.E.R.E.A. stendard design of motor axles.

N~TE.-Capacities based on steel per A.E.R.E.A. Standard Specification
for Annealed Carbon Steel Axles, Shalts and Similar Forgms%"
All
sizes s h o a are h h e d am. Journals and motor-beanng fits to be bumished. Length of motor-beanng fits to s u t type of motor used.
DiagramShewing K y s
(Rrtomtpcndedf o r use
wifh+lrt Grars only)

PIG. 17.-Table

oi axle and gear data for standard A.E.R.E.A. axles-

397

AXLE SPECIFICATIONS

Tensile Tests. After final treatment, the forgings shall conform

to the following minimum tensile properties. When ap lied to
axles, the diameter indicated shall be taken as the finished 8ameter
of the motor bearing seat.

-1/
i

I
1
1
O
z
:
1
t$$zzS

Quenched and tempered

carbon steel axles

UP to 4 in.

&!

?, &

Annealed
carbon Hammered

:X:

in diameter
or thickness

Tensik strength. Ib. per

85.000
Sq. Ill.. . . . . . . . . . . . . . .
90.000
Elastic limit. lb. per sq. (
~ n .. .. . . . . . . . . . . . . . . .
55,000
5o.ooo
Elongation, in 2 in.. per
cent.. . . . . . . . . . . . . . . . Not under 20.5 Not under 20.5
Reduction of area, per
cent. . . . . . . . . . . . . . . . .
39
39

80.000 80,000
40.000

40,000

20

20

da

2s

A he&-treded axle is one which is allowed to cool after forging,

is then reheated to proper temperature, quenched in some medium,
and then reheated to the proper temperature to refine the grain.
The exact treatment of a particular steel depends upon its chemical
composition. If the chemical composition of an existing axle be
known, the axle may be properly heat-treated and then turned down
to a smaller size for service. The purpose of heat treatment of

FIG. 18.-A.E.R.E.A.

standard tensile test specimen for axle steel.

axles is to relieve the internal stresses set up by improper cooling,

also to so rearrange the structure of the steel that it wll resist the
action of repeated blows which, though each wpuld have an effect
within the endurance of the original untreated steel, such continued
repetition would cause the untreated steel to fail.
Axle Breakage. The principal reasons for removal of axles are
excessive wear and the occurrence of cracks or breaks. A few electric
railways that are operating high speed service are removing axles

ELECTRIC RAILWAY HANDBOOK

after they have reached a certain mileage limit regardless of appearance on inspection, experience having shown that axle breakage is
most frequent after this point has been reached. This mileage
limit vanes with various companies from 200,000 to ~oo,ooo.
Many electric railways specify a wear allowance, and axles are either
scrapped or turned down to smaller diameters when this wear
is reached. These wear allowances vary from Hein. to
in. of
diameter of bearing fit. One company stamps on the end of each
axle the date that it is placed in service. After four years of service,
all interurban axles are removed and annealed, regardless of the
mileage which they have made. Careful inspection is made of
wear, and tests are made for flaws. Most of the broken axles on
this system have occurred close to the wheel. The fillet a t this
point is tested for flaws or cracks by gouging up a groove in the fillet
with a small diamond pointed chisel; if it transverses a flaw, a
square chip will be thrown off. Another company found that
many axle breaks start a t the comer of the keyway a t the base of
the hole drilled to start the keyway cut; this has led to a change in
the practice of cutting keyways, so that they are now milled a t the
end instead of being drilled.
In testing axles for cracks, the general practice is first to place the
axle in a lathe and rotate it so that its trueness can be determined.
After this a hammer test of some form is usually a plied. One test
consists of painting the axle with a coat of black read and oil over
the entire surface. The axle is then wiped clean and again painted
with a coat of white lead and oil, after which the axle is struck with
an 8-lb. sledge on each end. If there are cracks, flaws, breaks or
other defects, the oil and black lead will show through the white
Many railways use this test except that the first coat of
lead and oil is omitted. Others report that if the axle ir
carefully cleaned and then struck several sharp blows on the ends
with a sledge, the oil which remains in any crack will be forced out
and thus indicate the position of the crack. Another method of
testing axles without removing the wheels consists of immersing the
axle with wheels mounted in a tank of hot oil. After removal, the
oil is carefully wiped off, and the axle is covered with whiting. The
axle is then supported on a block a t the center with the wheels
hanging free, and is struck several blows with a heavy sledge.
Wherever there is a crack, a tine thread of oil works up from the
crack under the vibration, and can be seen against the white.
Several roads report that they clean axles very carefully and then
go over them with a magnifying glass, while on one road the test
consists of taking a very fine cut with a sharp lathe tool over the
section where the crack is most likely to develop; if the axle has a
crack, it is then apparent to the operator.

: :?k

CAST IKON WHEELS

Weight
of car

City service
a)i-in. tread

Light interurban service.

3-in. tread

--

Weight
of car

/Smh/..C.S.ISpale)~.C.s./)
32.000
6 0 o
40.000
44.000

48.000
52.000
56.000
60.000
64.000
68.000

399

440
460
480
500
520
540
560
580
600
620

.......

. . . .

.......
480
500
520
540
560
580
600

490
I

530
550
570
590
610

......
.......
.......

.......
.......
5x0
530
550
570
590
610
630
650

urban
henrice.
&-in.
tread

P.CS.

-p

F6.z:
Fie.

Trd.

*.C.S.

3:

ss.ooo
60.000
65.000
680
700
70.000 \
75.000
720
80.000
740
85.000 .........
90.000 .........
95.000 .........
100,000 . . . . . . . . .

670
690
710
7 30

750
770
790
810
830
850

- -

Nore: weights of wheels in the above table are for cars operated under

normal condlt~ons.

Effwt of Dimensions on Weight of Cast Iron WheeL Considering cast-iron wheels of the same general design, a 36-in. wheel will
weigh approximately loo Ib. more than a 33-in. wheel, and a 30in. wheel will weigh approximately 75 Ib. less than a 33-in. wheel.
Increasing the width of tread affects the weight of wheel approximately as follows:
Diameter of wheel, in.
36
33
30
24
20

Increase,in weight of wheel

per lnch increase in width of
tread. lb.
60
50
40

sa
24

Wheel Grinding. Chilled cast-iron wheels which would have

been removed from service because of worn flat spots, slight c h i p
ping, worn tread, or sharp flange have been made to give from 25
to 40 per cent more service by grinding. By close attention to the
grinding out of flat spots track and truck maintenance costs have
been reduced in addition to securing the increased mileage.
Whether the wheel is worn out or another grinding will be economical may be decided from a comparison of the worn diameter

400

ELECTRIC RAILWAY HANDBOOK

WHEEL GRINDING

401

or circumference with the original de th of the chill. A high

abrasive such as emery or alundam ine!lt form of a block or wheel
has been used for the grinding. The best practice is to grind both
mating car wheels, thus maintaining the same diameter for both.
A method of grinding out flat spots employed on many roads is to

FIG. 19.-Griffin Wheal CO. P.C.S..cast-iron wheel. See p. 400 for dimenslon s.

use a brake-shoe of abrasive material until the fht spot is ground

out. This method c d s for dose watching in order that the grinding
shoe may be removed as soon as the flat spot is eliminated, and so
additional flat spots will not develop. A method which has been
used with success is to jack up the end of the car which carries the
flat wheel and s in that wheel against a block of the abrasive while
the other pair oPwheels is blocked. An adaptation of this method
is to place a block of emery, suitably retained with approach
inclines a t each end of the emery block, on the rail. A similar
device, with the emery replaced by a system of rolls, is placed on
the opposite rail. The car is run up on these blocks with the wheel
to be ground resting on the emery block and the mating wheel on
the rollers. All motors but the one driving the axle are cut out,
and with the car blocked in position, the motor is run, grinding the
wheel against the emery.
The use of the abrasive in the form of a grinding wheel in a grinding machine is most satisfactory. A separate grinding wheel is used
for each of the mating car wheels. The car wheels and grinding
wheels are revolved in opposite directions and the dust is conducted away by a blower. The car wheels are revolved a t 10 to
IS r.p.m.
The grinding wheels are from 14 to 18 in. in diameter by
from 2 to 2% in. in thickness and have a peripheral speed of about
5000 ft. per minute. In some machines the feed is automatic. The
power required depends upon the ra idity of grinding, but for
ordinary work it amounts to about ro R.p There are two general
types of grinding machine; the Aoor grinder and the pit grinder.
In using the floor grinder the axle carrying the pair of wheels to be
ground is removed from the truck and mounted in bearings in the
grinder, a sprocket wheel is fastened to the axle and it is revolved
by a chain driven by the motor which drives the wheels. In the

ELECTRIC RAILWAY HANDBOOK

pit grinper the end of the car carrying the flat wheel is raised, a
short piece of rall IS removed from beneath each of the wheels to
be ground, the grinding wheels are adjusted to the wheel treads in
place of the rails removed and the car motor driving the pair of
wheels being ground is run with the controller on the first notch
and a water rheostat is used in place of the regular grid resistor.
If trailer wheels are to be ground a sprocket wheel is attached to
their axle and is driven by the grinder motor. The removal of
ordinary flat spots requires from 10 to 40 minutes actual grinding.
I t has been found advantageous to grind chilled cast iron wheels
after mounting them and before putting them into service. Such
grinding will correct for possible defects and thus aid in securing
the most satisfactory wheel service. Wheels may thus be brought
to the same diameter, wheels out of round or warped may be trued,
and in case the axle has not been turned perfectly, the wheels bored
not in the center, or the wheel improperly pressed onto the axle,
grinding the wheel will make the wheel tread, flange and journal
coaxial.
Opinion varies as to what conditions warrant the purchase of a
wheel grinder. A majority of master mechanics, in a canvass made
by the Eleclric Railway Journal in 1923, indicated that on roads
having loo pairs or more of cast iron wheels, the economies obtained
would pay interest on the investment in a grinder. Individual
opinions varied from 50 pairs to 500 pairs of wheels. Wheel
grinders vary in price, depending on the degree of efficiency, and a
purchase should be made commensurate with the actual requirements of the railway.
Wheel truing brake shoes have the advantage that the wheels are
ground without removing cars from service. Frequently i t has
been found that wheel truing brake shoes do not leave the wheels
concentric with the journal. The method is slow compared to pit
or floor type grinders. I n this method some companies pay
particular attention to the increased coefficient of friction of the
brake shoe and proportion the brake leverage accordingly.
The main advantage of the pit grinder as compared with the
floor type grinder lies in the saving of expense in jacking up cars,
removing and replacing wheels, and handling wheels in and out of
the grinder. There is also a reduction in time that the car is withheld from service. Some of the disadvantages given are that the pit
grinder does not produce as satisfactory a job as the floor type
grinder, that the wheel treads are not ground concentric with
journals, and that the driving machinery is in the way of pit work.
Some master mechanics feel that a sufficiently rigid construction
could not be obtained from a pit grinder, and for that reason the
floor type grinder is best. Many roads grind their wheels before
placing them in service, in which case the floor type grinder is of
advantage. A provision for removing or caring for the dust
removed from the wheels is important, and several railways are using
an exhaust ventilating system. This improves working conditions
considerably, so that the job does not become so annoying or
unfavorable for the workmen.

WHEEL SPECIFICATIONS

403

Solid Wrought Carbon-steel Wheels for Electric Railway Service

(A.E.R.E.A. Slandard Specificafion)
Process. I. Steel for wheels shall be made by the open hearth
process.
Discard. 2. A su5cient discard shall be made from each ingot
to secure freedom from injurious pipes and undue segregation.
Chemical Composilion. 3. The steel shall conform to the following limits in chemical composition:
Acid steel
Carbon ............... o 60 to o 80 per cent
Manganese ............ o 55 to 0 . 8 0 per cent
Silicon. .............. 0 . 1 5 to 0.35 per cent
Phosphorus, not over.. . .
0 . 0 5 per cent
Sulphur, not over.. .....
o 05 per cent

Basic steel
0.65 to o . 85 per cent
o 55 to o 80 per cent
o . 10 to 0 . 30 per cent
0 05 per cent
0 . 0 5 per cent

Sample for Chemical Analysis (Ladle Analysis). 4. An analysis

of each melt of steel shall be made by the manufacturer to determine the percentages of the elements specified in Par. 3. This
analysis shall be made from a test ingot taken during the pouring
of the melt. The chemical composition thus determined, together
with such identifying records as may be desired, shall be reported to
the purchaser or his representative, and shall conform to the
requirements specified in Par. 3.
Check Analysis. 5. A check analysis may be made by the purchaser from a wheel representing each melt. The chemical composition thus determined shall conform to the requirements specified in Par. 3. A sample may be taken from any one point in the
plate; or two samples may be taken, in which case, they shall be
on radii a t right angles to each other. Samples shall not be taken
in such a way as to impair the usefulness of the wheel. Drillings
for analysis shall be taken by boring entirely through the sample
parallel to the axis of the wheel; they shall be clean and free from
scale, oil and other foreign substances. All drillings from any one
wheel shall be thoroughly mixed together.
Finish. 6 . ( a ) The wheels shall be free from injurious defects
and shall have a workmanlike finish.
( b ) Wheels shall not be offered for inspection if covered with
paint, rust or any other substance to such an extent as to hide
defects.
Workmanship. 7. The wheels shall conform to the dimensions
specified within the following permissible variations:
Neigh.! of Flange. 7. ( a ) The height of the flange shall not vary
from the dimensions specified more than $i. inch.
Thickness of Flange. 7 . (b) The thickness of flange shall not
vary from the dimensions s ecified more than
inch.
Radius of Throat. 7. ( c r The radius of throat shall not vary
from the dimensions specified more than 5i6 inch.
Thickness of Rim. 7. (d) The thickness of rim shall not vary
more than ;/, inch over nor more than $6 inch under that specified.
The thickness of rim shall be measured from the inner edge of the
rim to a base line drawn from the point of tangency of the throat
and tread, parallel to the axis of the wheel.

404

ELECTRIC RAILWAY HANDBOOK

Width of Rim. 7 . (e) The width of rim shall not vary from the
dimension specified more than $6 inch.
Thickness of Plale. 7. ( f ) The late may vary in thickness, but
the variation less than that speciled shall not exceed %, inch for
each % inch in the thickness of the plate.
Limil Groove. 7. (g) When a limit-of-wear groove is specified,
its location shall not vary more than )fs inch from that specified.
Diameter of Bwe. 7. (h) The diameter of rough bore shall not
vary from that specified more than 3<6 inch over or fB inch under.
When not specified, the diameter of rough bore shall be % inch
less than the diameter of finished bore, subject to the above
limitations.
Diameter of Hub. 7. (i) The diameter of hub shall not be less,
but may be % inch more than that specified. The thickness of
wall of the finished bored hub shall not be less than I inch a t any
point for bores 6 inches or under in diameter, nor less than 1%
inches for bores over 6 inches in diameter, unless otherwise specified.
The thickness of wall of the hub of any wheel shall not vary more
than M inch a t any two points equidistant from the face of the hub.
h g t h of Hub. 7. ( j ) The length of hub shall not vary from the
dimension specified more than )B inch.
Projection of Bock Face of Hub. 7. (k) The projection of back
face of hub shall not vary from the dimension specified more than
inch.
Black Spots on Rub. 7. (1) Black spqts deeper than hainch will
not be permitted in rough bore mthin 2 inchesof either face of hub.
Eccentricity of Bwe. 7. (m)The eccentricity between the axis
of the tread and the axis of the rough bore shall not exceed %,4 inch.
Block Marks on Tread. 7 . ( n ) Block marks shall not exceed
inch in height.
Rotundity. 7. (o) Wheels shall be gaged with a ring gage and
the owninn between the gage
- - and tread, a t any
- point,
. shall not
exceed
inch.
Plane. 7. (p) Wheels shall be gaged with a ring gage concentric
with and pe ndicular to the axis of the wheel.
(b) For a r p i n t s on the back of the rim equidistant from the
center, the variation from the plane of the gage when so placed
shall not exceed
inch.
Tape Sizes. 7. ( q ) Wheels with treads under .z inches in width
shall hot vary more 'than 6 tapes over nor more thBn 4 tapes under
the size specified. Wheels with treads 3 inches or over in width
shall not vary more than 9 tapes over nor more than 5 tapes under
the size specified.
Mating. 7. ( r ) The wheels shall be mated as to tape sizes and
shipped in pairs. The tape size shall be legibly marked on each
wheel.
Gages and Tapes. 8. The manufacturer shall provide suitable
gages and tapes which shall conform to the contour and dimensions
specified.
Bronding. 9. Wheels shall be stam ed with the manufacturer's
mark, date, heat number and seriar number, in such manner
that each wheel may a t any time be readily identified.

WHEEL T A P E GAGE

405

Insficttwn. 10. (a) The inspector representing the purchaser

shall have free entry, a t all times while work on the contract of the
purchaser is being performed, to all parts of the manufacturer's
works which concern the manufacture of the wheels ordered. The
manufacturer shall afford the inspector, free of cost, all reasonable
facilities to satisfy himself that the wheels are being furnished
in accordance with this specification. Tests and inspection shall
be made prior to shipment.
10.(b) The purchaser may make the tests to govern the acceptance or rejection of material in his own laboratory or elsewhere.
Such tests, however, shall be made a t the expense of the urchaser.
10. (c) All tests and inspections shall be so conductetas not to
interfere unnecessarily with the operation of the works.
10. (d) Wheels which show injurious defects while being finished
by purchaser, will be rejected and the manufacturer shall be notified.
10. (e) Unless otherwise arranged, any rejection based on tests
made in accordance with Par. 10 ( b ) shall be reported within ten
working days from the receipt of samples.
10. (f) Samples tested in accordance with Par. ro (b) which
represent rejected wheels shall be preserved for one month from the
date of the test report. I n case of dissatisfaction with the results
of the tests, the manufacturer may make claim for a rehearing
within that time.
Standard Wheel Circumference Measure (A .R.A.). The measure for tape sizes, referred to in above wheel specifications, is
shown in Fig. 20. To use this standard measure, place tape about
circumference of wheel, having the brackets in contact with flange.
The normal circumference for wheels of each of the different
diameters is indicated by the space marked "3." According to the
A.R.A. specifications, steel and steel-tired wheels should be
rejected if the sci-iFd lines on the head piece fall to the left of the
space marked "C or to the right of the space marked "xu;
cast iron wheels should be rejected if the scf;ibed line on the head
piece falls to the left of the space marked " I or to the right of the
space marked "5." The continuous markings on the upper side
of the tape, as shown in the figure, may be used for mating worn
wheels. The linear dimensions shown in figure represent measurements of actual circumference of the wheel and not straight length
of the tape. Graduations are spaced H in. apart with the tape laid
flat, and hence a variation of one tape size means approximately
a variation of j b in. The tape size of a wheel is determined by the
space in which the circumferential measure falls; for exam le, the
space between lines 157 and 158 on the upper side of tge tape
coincides with the space on the lower side of the tape, representing
Tape Size No. 3 for the 33 in. diameter wheel.
Rolled Steel Wheels. The general replies from fifty operating
companies to questions regarding solid steel wheels sent out by
the Committee on Equipment, A.E.R.E.A., were as follows:
The diameters when new range from 30 to 37% in. New rims are
from 2 to 3?/r in. thick. These are worn down by some companies
to as low as Sh in. Others wear down to 1% in., while the average
considers % in. t o be safe. The average number of turnings is

406

ELECTRIC RAILWAY HANDBOOK

three. Solid rolled steel wheels are considered safer than steel-tired
wheels for interurban service. They are also held to be more
economical, scrap considered.

Wheel Turning. The necessity for turning is most often determined by flange wear. A general statement of the flange problem
is as follows: When a train is moving on tangeflt track, the flanges
of the wheels, due to lateral oscillation of the train, strike against
the rails alternately on one side or the other; when the train moves

WHEEL TURNING
on curved track, the wheel flanges are forced against either the inner
or outer rail depending on the speed and the amount of superelevation of the outer rail. In both cases the car wheel flanges in contact
with the rail are factors in transmitting the various forces set up
within the car to the rail. Since the flange pressures produced in
traversing curves are the more severe, the pressures arising on tangent track need not be considered. First, the pressure or force
between rail and flange sets up internal strains and stresses within
the body of the wheel; in the case of a wheel subjected to severe
o rating conditions, the internal stresses may result in the failure
o r t h e wheel. Second, the ressure between the rail and flange
coupled with the rotation ancfsliding of the wheel on the rail results
in a grinding or wearing action and in consequent reduction in area
and strength of the flange. Third, the rail pressure, when of sufficient magnitude and if acting on a flange materially worn, may cause
the failure of the flange. The necessity for turning may also be
determined by flat spots on the tread or by the allowable difference
between the circumferences of two wheels rigidly fastened to the
same axle or geared or linked together. This allowable difference
is often taken to be % inch in ordinary electric railway work, and
a consideration of it is of greater importance on driving wheels than
on trailers.
New York subway wheels are turned in two Pond centerdrive
lathes. Of these one is driven by a 2eh.p. motor and the other.
which is one of a newer and heavier type of construction, by a 40-h.p
motor. The latter is equipped with air operated tail stocks and
both have the cutting tools clam
in the tool post b means of
compressed air. The output of t e former lathe is tweLe pairs of
wheels per day, while the latter machine has a capacity of twenty
pairs per day. The great stiffness and pulling power of the latter
mach~nepermits the finished cut on the wheels to be taken with a
forming tool which gives proper contour with one cut.
A. B. Creelman of the Youngstown Municipal Railway reports
the average life of steel wheels obtained without turning to be 150,000
miles. This high mileage is obtained by careful attention to various
factors affecting wheel wear beginning with the initial mounting
of the wheels on the axle. Wheels of exactly the same tread diameter are aired, axles are trued and straightened, and both the
wheel antgear seats are turned, if necessary, to secure full bearing.
The trucks are squared and brake rigging placed in first class sha
with the brake shoes hung to eliminate any side thrust on the trucE
when the brakes are applied. In operation the brake rigging is
kept in line, and all lost motion is taken up as it develops. Journal
bras? are renewed before the side motion or end play becomes
excesswe. The axles are kept square in the truck frame by
rearranging the shims in the journal box pedestals. Wheels that
begin to show flange wear on single end cars are removed to the
other end of the car, thus balancing wheel wear such as is obtained
on double end cars.
Other companies report various mileages between turnings ranging
from 20,000 to 65,000 miles, with in. to I in. as average material
turned off, and three to five average turnings in life of wheel.

R"

ELECTRIC RAILWAY HANDBOOK

408

Welding Worn Flanges. Satisfactory results have been obtained

by building up worn flanges by means of a welding process, .as
follows: The wheel with the worn flange is turned, removing j/, m.
of metal from the tread, and in so doing, y in. from the new flange
a t the base. The top of the flange is 6nished to proper contour,
which leaves a right angle groove in the flange about 36 in deep,
and parallel to the tread, and % in. deep at right angles to the tread
for a I in. flange. The flange is then fiUed out, using a !im.
welding rod, and turned to finish. Most all steel wheels are taken
out of service on account of sharp flanges, and usually there will be
one sharp flange on each axle, while the other will have become
larger than the original. The latter wheel is turned to the same
diameter and a new flange formed, unless both wheels have a sharp
flange, in which case both flanges are welded. Experience shows
that a filled flange will last as long as a turned flange.
Mounting Wheels. Wheels with flanges worn to or below the
limit of wear, as shown by the standard gages, should not be
remounted. The flange thickness of wheels fitted on the same axle
should be equal and never vary more than H e in. In mounting
wheels, new or second-hand, the standard wheel mounting and check
gage (Fig. 2 1 ) should be used, pressing on one wheel so that it is the
proper distance from the center 11neof the axle by having the pointer
of the gage a t the center line of the axle and the end ieces of the
gage resting only on the tread and contour of the wheefat the gage
line, allowing enough clearance to take care of the tolerance variation of the wheel. Then the other wheel is pressed on in the same
manner, and by proper mounting and use of the gage should thus
Pace the wheels at the pro r distance from each other and also
ocate them centrally on t e axle. The wheel seats on all axles
must be turned to a uniform diameter throughout the e n t i e
length of each wheel seat, and must be smooth and freefrom ridges,
so as to provide an even bearing for the wheel fit throughout. The
mounting of wheels on ales which have the wheel fit tapered is not
permissible. Mounting resses should be rovided with recording
pressure gages, and all w A e e ~not mounte8within the limits given,
or wheels that are forced against the shoulder, are to be withdrawn.
Mounting Pressure. The A.R.A. recommended practice is
shown in the following table.

I
A.R.A.
axle

MOUNTINGPRESSURES,
IN TONS

1
I

Cast won wheels

1
1

Steel wheels

d~arneter Mrnrrnurn Maxrrnurn Mlnrrnurn Maxtrnum

WHEEL MOUNTING

409

Removing Wheel from Axle. The pressure ordinarily required

to remove a wheel from its axle will vary from two to five times
that a t which the wheel was mounted. The wheel press is often
used to apply this pressure which amount generally lies between
150 tons and 300 tons. Heating the wheel and hammering while
the pressure is applied are often resorted to in starting the wheel.
A drop hammer has also been used alone to start the wheel which
was afterward removed by the wheel press.
Shrink Fits. The process of shrinking wheels and gears to axles
and tires to wheels IS outlined in the following citation of the
practice on the New York, Westchester and Boston Railway by
R. R. Potter: The parts are bored accurately to a certain
diameter in relation to the axle, micrometers being used in measuring. The part is then heated and placed on the axle without pressure. By this method there is no abrasion on the finished surfaces
of the axle, wheel or gear due to the process of application. Aleso,
as heat is applied to the part when it is removed, very little abraslon
occurs, due to the process of removal. In addition, the parts may
be replaced in then original positions a number of times, without
loss of holding power. The axles, a t the location of wheel seats, are
carefully finished and filed as smooth as possible. Wheels and
gears as well as tires are bored o oor in. smaller than the axle or
wheel center per inch of diameter of bore, and then they are heated
and slipped into place. After the wheels have been applied and
have cooled, they are tested in the wheel press a t a pressure of ro
tons per inch diameter of bore. The gears are heated a t center
with a torch using compressed air and city gas, but the rims are not
allowed to go above 300 deg. F., as measured with a thermometer set
with the bulb in a special cup which fits between the gear teeth.
I t is found that this temperature invariably gives sufficient expansion to permit gears to be slipped on without any pressure. Wheels
are heated with the same torch until a drop of water at the middle
of the spoke will boil freely when dropped onto it. Wheels sometimes require a slight pressure in addltion to the heating to move
them into place, but this is negligible in comparison with that used
in ordinary press fits. When wheels or gears are thus shrunk onto
the axles, the microscopic p i n t s of metal are not smoothed down
as they are when a wheel IS pushed on cold, and the result is that
these points interlock and make a very close and secure fit.
Holding Power of Steel Tire. In the New York subway where
wheels are heated excessively by frequent high acceleration and
heavy braking, the highest temperature of the tire of a wheel
just taken from service was found to be 145.4 deg. F., and the
corresponding temperature of the hub was 87.8 deg. F. After
this was investigated, the holding power of steel tires shrunk on
with the standard shrinkage allowance of 0.001in. per r in. diameter was calculated. The results are shown by the following table:

410

Shrinkage
allowance

ELECTRIC RAILWAY HANDBOOK

1 3 ; ~1 I
Temp.
of hub

Tire

2%

in. thick

6070 lb. per sq. in.

3970 lb. per sq. in.

Tire

rx in. thick

1b. per sq. in.

1700 1b. per sq. in.

2600

Percentage reduction in holding power caused by wear, 57 per cent.

Percentage reduction in holding power caused by heat, 35 per cent.
Percentage reduction in holding power caused by wear and heat,
92 per cent.
Life of Wheels. The actual mileage obtained from wheels varies
through quite a large range, depending on service conditions. The
wear of wheels resulting from r o l h g on the rails is very small.
Brake shoes wear away the metal more rapidly, and the loss from
chipped and broken flanges, the grinding out of flat spots on chilled
iron wheels, and the turning down of steel wheels for sharp flanges,
account for much of the wheel consumption. Thus the number
of stops made has an important bearing on wheel wear, and likewise the character of the line as regards grades andcurves. The
amount of special trackwork is also an important factor in the wheel
life. With the same depth of wear, the larger the wheel the greater
the amount of metal which can be worn 06, and hence increased
mileage for the life of wheel. In interurban service, with few stops,
a greater mileage can be expected from the same typeof wheel than
in city service.
Mileage

.......... ~S,OOO-~O,OOO
................
70.00~240.000
........... 2oo.000-3oo.000

City s e r v i c ~ h i i l e diron wheels..

City service--steel wheels..
Interurban servicesteel wheels..

Average
mileage
49.000
137.000

ago.ooo

The above table was compiled from a survey of sixty electric

railways by the Electric Railway lournol. By very careful attention to inspection and maintenance, the maximum mileages can
be increased.
The use of steel wheels is increasing. Many railways are not
equipped for proper maintenance of steel wheels and continue the
use of chilled iron wheels rather than go to the expense of purchasing
new shop equipment. Some railways l a v e found that, due to track
conditions, and where motors hang low, they cannot obtain the full
wear from steel wheels, and hence the chilled iron wheel may be the
more economical. The great advantage emphasized by most users
of steel wheels is the freedom from chipped flanges, and from the
trouble of removing f i t spots from cast iron wheels, as slight flat
spots in steel wheels will roll out in a few days running without
requiring attention a t the sho . Steel wheels usually are shopped on
account of worn flanges, a n 8 as this condition comes gradually to
steel wheels, they can be shopped a t convenience and the work
bunched to reduce maintenance cost. On the other hand, cast
wheels fail one a t a time and must be shopped a t once. Steel wheels

WHEEL DEFECTS

411

weigh less compared to cast iron wheels, and a slight saving in

energy results.
Wheel Defects. The following relative to common wheel defects
is from a talk by F. A. Beebe of the Grif6n Wheel Co. at the Hartford Shops of the Connecticut Co.:
Thin Flange. A thin flange may be due to one or more of the
following causes: ( a ) Wheels improperly mated as to diameters.
( b ) Improperly mounted as to gage. (c) Improperly put on axles;
that is, the distance from end of journal to gage line on one end of
axle being greater than the other. (d) I m p r o r l y shaped flange.
(e) Trucks out of true. (j)Car riding on side earings, preventing
free movement of the truck, bringing the work of turmng on the

FIG. 21.-A.E.R.E.A.

standard wheel mounting and check gage.

flange. ( g ) Track out of gage on curves. (Ir) Im roper elevation

of outer rail. ( i ) Running car in one direction; tgat is, when car
a t the end of the line runs around a loop instead of being turned.
( j ) Sanding one rail, causing one wheel to wear down faster than
its mate.
Warped Flange. A cast iron wheel flange warpage of about he
in. is allowable.
Eliminaiion o j Flat Spots. (a) See that the brake shoes are
properly adjusted. ( b ) Braking apparatus adjusted so that in
coming to a stop the wheel will not lock. (c) See that the percentage of brake pressure is in proper relation to the load on the wheels.
(d) An important point is the manner of hanging the brake. Most
brake hangers are placed below a line passing through the center
of the axle, parallel with the trucks, which allows the brake to fall
away from the wheel by gravity. This angle may be made too
large, so that with badly worn and loose parts the application

412

ELECTRIC RAILWAY HANDBOOK

of the brakes causes the shoe to crowd up against the wheel. This
makes a toggle joint, producing excessive pressure on one pair
of wheels in the truck causing them to stop rolling while the other
pair is still revolving. This is a prolific cause of slid flats and also
results in excessive wheel wear.
Wheel Gages. Fig. 2 1 shows the A.E.R.E.A. Standard Wheel
Mounting and Check Gage. This gage is made with a pointer for

the center linc of the axle so that

the first wheel may be pressed on
the proper distance from the center line of the axle and the wheels
will not only be located the proper distance from each other, but
will be located centrally on the
axle. The end ieces making
contact with the wKeels are machined so that they rest only on the
tread and contour of the wheel
of
a t the gage line, allowing enough
clearance to take care of the
tolerance variation of the wheel.
Frc. z2.-A.E.R.E.A. standard The pointer placed a t the center
gage. Measurements A has a thumb screw with which to
and B vary depending on dia- make adjustments for differences
meter of
to be gaged.
in the diameter of the wheels.
I n mounting and gaging wheels i t is understood that the gage line
is a t the point on the fillet of the flange j/, in. below the wheel
tread, and wheels should be gaged 5 i in. narrower than the gage of
the tracks; the track gage being measured between points ~ / 1in.
below the tops of the rails.
The A.E.R.E.A. Standard Plane Gage for Solid Wheels is shown
in Fig. 22. This is used to check the warped condition of wheels
by placing it on the back of the rim. The tolerances allowed will
be governed by the wheel specifications.

WHEEL GAGES

413

The A.E.R.E.A. Standard Rotundity Gage for Solid Steel Wheels

is shown in Fig. 23. As all standard wheels have a taper of I in.
in 2 0 in., in checking the wheel, the gage is used that will become
tight on the wheel a t about the center of the tread. The tolerance
governing this is included in the wheel specifications
Limit of Wear Gage. (A.E.R.E.A.Standard.) Designs for limit
of wear gages for the wheel contours which were adopted as standard
by the Am. El. Ry. Eng. Assn. in 1923 are a t this writing in preparation by the Equipment Committee of that Association; these
probably will be adopted as standard a t the 1924 convention, and
thereafter be shown in the A.E.R.E.A. Manual. T o use the limit
of wear gage, i t is fitted down over the flange contour, and if the
gage touches the tread of wheel and the guard side of the flange
contour, and does not encounter the flange contour a t any other

FIG.23.-A.E.R.E.A. standard rotundity gage. Measurement A varies depending on

diameter of wheel t o be gaged.

FIG. 24.-A.R.A.

standard limit of wear

gage.

point of the gage, the flange has been worn to the limit. I n this
case the flange is too thin or the angle of the flange is too s h a v .
If the gage touches a t the top of the flange and not the tread of the
wheel, the flange height is too great.
Fig. 24 shows limit of wear gage for A.R.A. Standard Tread
Cast Iron Wheels for cars under 80,ooo Ib. capacity. Similar gages
are used for cast iron wheels on cars of over 80,000 lb. capacity and
for steel wheels.
A.E.R.E.A. Standard Wheel Contours. (Fig. 25.) Investigation of the conditions of track, rail sections, etc., brought out that
the widths and depths of wheel flanges, measured on the level of
the tread, which appear to best cover all ordinary requirements,
are 1%. in. width in combination with 76 in. depth, 1% in. and I in.
widths in combination with % in. depth, and I in. width in combination with the 36 in. depth. A wide tread is desirable for
minimum wear on track, particularly in special trackwork, and as
practically all A.E.R.E.A. standard sections of rail permit the use
of a wheel tread 3 in. wide, i t is recommended that this should

414

ELECTRIC RAILWAY HANDBOOK

be given the preference to the gradual elimination of the narrower

wheel treads. Existing conditions in many localities, however, do
not now permit a wheel tread wider than 2% in., and the latter
therefore is also recommended as an alternate standard in connecin. and % in. depth of flanges. For interurban
tion with the ?i
lines using the % in. by IN, in. flange the 3% in width of wheel
tread is recommended. The taper of wheel tread in common use
formerly differed between steel wheels and chilled iron wheels, but for
the future the same taper of I in 2 0 is adopted for both. The
rounding off of the outer edge of the wheel tread in present practice
varies greatly. A certain amount of rounding is desirable to
prevent chipping or flowing out of the metal, but it should be
confined to the necessary minimum so as to not unduly reduce the

TAPERS:
Wheel Peed- /k?O
f l a n p T d 3 3 - 27O
flonp-GwrdSide-23

Fm. 25.-A.E.R.E.A.

Full Lines

- Sfeel wheels

Dof)edLincs-Chi//edlron
Whcds
h ~ e n s / o n Xnot
- less fhan,dN

standard tread and flange contours for steel and

chilled iron wheels.

width of the effective wheel tread. I t has been made H. in. uniformly in all standard A.E.R.E.A. contours.
The curved contours of wheel flanges now in use vary greatly
for the same width and depth of b n g e without apparent regard
to the track and wear conditions. For given or assumed fixed
conditions that affect the wear on wheel flanges, i.e., dimensions
of trucks or running gears, rail sections used, and proportionate
amount of straight and curved track of various degrees of curvature, the essential portions of wheel flange contours best suited to
these conditions can be theoretically and graphically determined.
For straight track of standard girder guard rails and the usual
play between wheel gage and track gage, the flanges would have
straight sides with a minimum angle from the vertical. On curves
this angle increases with the curvature and wheel base until maxi-

STANDARD WHEEL CONTOURS

415

mum practical conditions are reached. As the same h u e must

answer in average as well as extreme conditions, the contour
must follow a line that takes this into account. The hyperbolic
curves produced by the horizontal section through straight side
Banges must be modiied a t different heights according to the

CHILLED IRON WHEEL

Norma/ Flange

STEEL AND STEELTIRED WHEELS

FIG.26.-A.R.A.

standard wheel contourn

curves to be run through, the aim being to give as good wearing

surface as possible under the varying conditions, to reduce the wear,
and to give points of contact on both the gage side and guard side
as high up on the flange as practicable to reduce the leverage of
frictional resistance. Average extreme conditions must be the
principal guide. The graphical development under various condi-

ELECTRIC

RAILWAY HANDBOOK

tlons of wheel base and radius of curve, in combination with

maximum d~ameterof wheel of 33 m , produces practically stra~ght

lines through the reglon of contact between wheel flange and rall.
These llnes for the standard glrder guard rail have a practically
unlform lncllnatlon from the vertlcal of approxlmately 27 deg of
angle on the gage s ~ d eand 2 3 deg of angle on the guard side.
Theoret~cally,no roundlng off of the nose of the flange below the
polnts of contact is requ~red,and the flange could be carried down
on a contlnuat~onof the above angles to a sharp comer a t ~ t s
extreme depth Practical expenence in the operation of the cars
on track, with a certaln amount of unavoidable lrregulantles
and r b l e obstruct~onsin the flangeway, and the requirements
of w eel manufacturers, make a maxlmum rounding of the nose
des~rable Any flattening of the nose would increase the tendency
of the flanges to c h ~ pand to s t r ~ k efrog polnts or other irregularities
In the track The roundlng adopted for the four standard flanges
IS, therefore, the maximum permitted by the contour lines, but
still leaves sufhcient Ilne or area of contact against the rail on
curves. For the radlus of the fillet jolnlng the contour of h n g e
to the tread, the average practice has been followed, and the radius
made H In for all flanges Should a compound fillet, giving a
partlaliy curved tread, be decided upon or considered desirable in
the future, the necessary modifications can be made easily, theeffect
on the ho~rzontalpodtlon of the gage point being negllg~ble
On rolled steel wheels it is impractical in manufacture to produce
a curved contour for the back of the flange above the tread line.
The contours of the back of steel wheels above the tread line follow
a straight vertical Irne, the exact contour above tread line not
being essential I n chilled iron wheels it is desired from a manufacturing standpoint that the amount of material above the flange
proper be Increased and the standard wheel contours have such an
Increase by a continuation of the curved lines of the back of the
flange above the tread line. Such continuation on a circle of the
same r a d ~ u sas the back of the flange below the tread line gives
approxlmately H. of an inch extra width a t a height of
of an
inch above the tread line. The continuation of the back, however,
may be made on any larger radius or on a straight line tangent
to the contour a t the point of Intersection with the tread level, and
the width above the tread llne be increased further, and can be left
optlonal with the manufacturer. This difference between the steel
wheels and the chilled iron wheels does not in any way affect the
action of wheels on the track and the wheel flange proper below
the tread line is the same for both.
The A R A standard flanges for steel wheels and cast iron chided
wheels follow the same principles. (big. 26.)

STAWARDDIALENSIONS
FOR STEEL EELS (A.E.R.E.A.)
Flange contour
Tread
Flange he~ght

I A
3"

Flange thickness

1x6"

Rlm nldth

4Ha"

Hub length'

434!' or

1 a:: g

D~ameters.11 m to 36 m , lncluuve

i A I B I B 1 c
I

356"

3<"
19{s"

1%"

156"

396"

4%"

a:: g;;{

456" or

456'' or

g::

2:: &]

g;+]

Hub p r o e c t ~ o n * ~ t " ~ -3 ' -3 1 3" or

4' 854 " gage
l;;~~i'or
a" (33+)
sU(33+)! 1% s"(33+)
Hub

,;;go;;

p~o)ection*/3~%s"(31-)j1%s"(3r-)
a l % s " or
sf 2 5 2 " gage
or
s (53+J 1'%e"(~3+)

4 3
3"or
2" (33-t)
I

256"

96 "

I"

I"

I"

I"

4"

351"

4"

456" or 5"

455" or 5"

'

2Z

3"

3%''

9XI

456" or 5"

452" or 5"

I
356"

4" (3t -)
3.01
a" ( 3 3 f )

4%.

cn

9s"

3"(31-)
3" or
a" (33+)

F~guresIn parentheses indicate wheel dlsmcter, in inches.

Two figu~es.mchae rntervenrng ~ 1 s t ~ .

31"

4lfia"

454" or

1I

3"

2%"

1
I

3%''

3
z
356"

3%"

V)

.4iit1

Murur slip 1nhcat.a "01 under."

4%tf'

4101
Plua

am lndlcates

"or over."

2
-4

418

ELECTRIC RAILWAY HANDBOOK

Advantages of Large wheel. Following are some of the

advantages of the large wheel: It affords greater mileage between
turnings, gives lower journal surface speed with consequently
less wear and axle steel fatiguing stresses per car mile, causes
smaller stresses to be set up in wheel and rail surfaces, strikes
lighter blows on the track and is less likely to give trouble a t a
defective frog, witch point or rail. A large wheel makes possible
the use of a large gear and maximum gear reduction with a corresponding high motor speed which permits the use of a comparatively
small light weight motor and necessitates the least energy consumption. Special installations may be such that the above
advantages would be inferior to the advantages accompanying the
use of small wheels. Thus, the satisfactory design of many low
floor cars has demanded wheels of small diameter.
Truck Wheel Base. The use of 4 ft. wheel base trucks for urely
city service is desirable only by r e a n of the ease with whicR they
may be operated over short radius curves. The inside hung motor
truck of an approximate 6 it. wheel base is superior to outside hung
motor trucks of short wheel base for the following reasons: the lesser
displacement of tractive effort between the two pairs of wheeh
during acceleration and braking; greater factor of strength on
account of the location of motor suspension in close proximity
to the center of truck; less wheel flange pressures against rail heads
required to swivel trucks on curves, and also occurring during
"nosing" periods; and easier riding qualities. There is a less
liability to derail, less flange wear, less tendency to cause rail corrugations, less wear on track, and less maintenance, with the long
wheel base truck. Interurban service uses the long wheel base
truck, which is best for high speed operation. In a paper before the
Central Electric Railway Association, A. C. Vauclain of-the Baldwin Locomotive Works stated that the final determinauon of the
wheel base should, if curves permit, be based entirely upon the
transom width being suf6cient to allow a pro r width of bolster
and bolster springs, and the fixed length of t g motors measured
from their axle journals to their supporting noses.
Minimum Curve for Given Truck. Fig. 2 7 by Graham Bright,
El~crric Journal, 1911, shows the relation between wheel size,
wheel base and minimum curve. These curves are plotted according to the expression

R=-

(published by the Baldwin Locomotive Works)

sin a
in which R -- radius of sharpest curve that can be passed, feet.
W = wheel base, feet.
a = angle which flanged wheels make with d
sin a = 0.117
With diameter of wheel 2 0 in. to 24 in
25
30
= 0.107
31
40
= 0.090
41
5"
= 0.080
51
60
" 0.075
2

TRUCK WHEEL RASE

419

Ball Bearing Center PLates. The rincipal advantages of ball

bearing center pIates over smooth pkte bearings,r~rthat thrk
use is accompanied by a minimum of flange wear and energy consumption on curves.. They bring about less liability to derailment
and require less lubncation.

W i u a of Curve. Pwt
,tion between minimum curve, wheel diamreter and
wheel base.

ruck

ELECTRIC RAILWAY HANDBOOK

~ P Y o) h

onn

a,- aT
.,
dm-

ono

- - O

hOnmm

n m d
mom

-atmmm

"a 5
xxx xxr
h-m

LdY)rYX

00-

w h o
mm9

-nm

ddd

xxF
xP - hx- b
CWrn

STANDARD JOURNAL BOX DIMENSIONS

Mod~fiedType for Arch Bar

FIG. 28.-A.E.R.E.A.

standard journal boxes.

421

422

ELECTRIC RAILWAY HANDBOOK

'~crlf
~ e r h o n ~ - ~
PIG. 29.-A.E.R.E.A.

standard journal bearings.

Section 0-8

Section A-A
PIG.30.-A.E.R.E.A. standard journal thrust plates.

423

STANDARD JOURNAL PARTS

DIMENSIONS
OF A.E.R.E.A.

S T A N ~ A B D JOURNAL
(See Fig. aq)

In.

In.

BEARINGS

In.

5
6

7
8
9
I0
I1
I2

I3
14

"

IS

16
17

18
19
20
11

9g

DIMENSIONS
OF A.E.R.E.A. STANDARD
JOUBPJAL TERUST
PLATES
(See Fig. 3 0 )

Journal

Dimension

3%"X 7"
4NVX8"

ln.

I
2

5%

:aR
1"

4
5

7
8
9
I0

1 x 0
1%

1 %'

9i

In.

5" X 9''

sfll'X

In.

:g
1%

;%R

234

1%

10"

7%
1%

35'R
2
I

HR
Hs
Z%l

ELECTRIC RAILWAY HANDBOOK

424

PIG. 3 I.-.I.E.R.E.A.

standard journal wedges.

DIMENSIONSOF A.E.R.E.A. STANDARDJOURNALWSDGES

(See Figs. 31. 32. 33)

size

Dimension

I
/

3G,~;i,,

3%"X 7"

4%"X8"

5"Xg"

(Fig. 31)

(Pig. 31)

(Pig. 32)

(Fig. 33)

In.

In.

5X"X

10''

(Fig. 33)

In.

In.

in.

STANDARD JOURNAL WEDGES

p r ~ 32.-A.E.R.E.A.
.

standard journal wedges.

FIG. 33.--4.E.R.E.A.

standard journal wedges.

426

426

ELECTRIC RAILWAY HANDBOOK

A.E.R.E.A. Standard Journal Boxes, Bearings, Thrust Plates

and Wedges. Figs. 2 8 t o 33, inclusive, s h the A.E.R.E.A.
Standard Designs. They represent a development over a long
period of time to overcome various difficulties. One of the chief
difficulties was end wear on journals, which in the former types
was greatest on the journal brass itself and next on the stop on the
end of the axle. End wear was experienced chiefly where cars
operated a t relatively high schedule speeds and where there were
a comparatively large number of short radius curves. In the present
standards, end play has been overcome by casting guides integral
with the journal box, so that a thrust plate can be dropped into the

FIG.34.-"

Hoodcd " wedge.

guides and in close contact with the end of the axle, thereby imposing the lateral thrust of the axIe upon the thrust plate rather than
on the journal brass. The chief advantages of this method of
overcoming end play are that the stress set up by the lateral thrust
of the axle is more nearly in line with the axle itself, and that the
thrust plate is easily renewed. It also serves to hold the packing in
place in the lower portions of the journal and permits better
inspection of the journal box. The journal bearing now has a more
nearly full semi-circular cross-section, which gives ample bearing
on each side of the journal to take the horizontal thurst from the
brake shoes.
Hooded Wedge. Where the older types of journal box, bearing,
etc., are still in use, the end play may be overcome by use of a

STANDARD JOURNAL PARTS

427

hooded wedge, as shown in Fig. 34. The hooded wedge has a

renewable wearing plate which is located so as to provide an end stop
and wearing surface for the end thrust of the axle.

FIG.

35.-A.R.A.

standard journal b o x . wedge and bearing, assembled.

ILRA. Standard Journal Box, Bearing and Wedge.

Figure.3~
shows a n assembly of the A.R.A. Standard Journal Box, Beanng
and Wedge.
A.E.R.E.A. Standard Dust Guards. The A.E.R.E.A. standard
dust guards, as shown in Fig. 36, fit the standard A.E.R.E.A. journal
boxes. The guards are designed so that proper allowance is made
for clearance and so that they cannot be reversed when being installed
in the boxes.
Lubrication of Journal Bearings. (A.E.R.E.A. Miscellaneous
Methods and Practices.) Once every 30 days, providing cars do

Prc. 36.pA.E.R.B.A. standard dust guards.

ELECTRIC RAILWAY HANDROOK

not make more than 1 5 0 miles per day, or if on a mileage basis,
every 4500 miles. Journal beanngs in high speed service should
be lubricated weekly when cars are making from 300 to 500 miles
per day, or if on a mileage basis, every 2500 miles. I n both cases
the journal box covers, springs and bolts should be examined to see
that all parts are in good condition and that the cover is as nearly
dust-proof as possible, and that the waste is stirred up in the boxes,
and where it has worked forward, pressed back in place, and only
a certain amount of oil should be applied. This can only be done
by allowing a definite amount of oil for each car or by thoroughly
educating certain men to do this art of the work.
A.R.A. Standard Method of
Journal Boxes. Preparation
o j N m Packing. The waste should be loosened, placed in a saturating vat and kept completely submerged in car oil, a t a temperature
of not less than 70 deg. F., for a period of a t least 48 hours to insure
thorough saturation. I t should then be drained for the purpose of
removing the excess oil, until the packing is in a resilient or elastic
condition. Prepared packing should be turned over a t least once
each 24 hours, or the oil which has accumulated in the bottom of the
container should be drawn off and poured over the top of the
prepared packing.
Prebarnlion of Renovaled Packinn. All packing, when removed
from *journal bbxes for the purpo-se of periodical repacking or
renovating, should be piled into a container, avoiding contact with
the ground or any other place where it may pick up dirt, and taken
to the waste reclaiming lant. This packing must not be reused
until renovated. In recLiming packing it first should be picked
over carefully, and dirt, metal, etc., shaken out, the knotted strands
of waste pulled apart, and then placed in hot oil in renovating tank
for a short time, working it with a fork for the purpose of thoroughly
washing and loosening it. I t then should be rinsed in clean oil,
then drained for the purpose of removing excess oil.
Cleaning Boxes. Before packing a journal box the oil cellar
should be thoroughly cleaned of all dirt, sand, scale and grit, and if
water is present it must be removed. When new journal boxes
are applied, or when reapplying journal boxes, the interior of the
box, including the dust-guard well, should be so treated, and closefitting dust guards and lids should be applied.
Cleaning and Applying Bearings. Before applying journal bearings, they should be thoroughly clean, have a smooth bearing
surface, free from irregularities, and should have a proper bearing.
Under no circumstances is it permissible to use sand paper, emery
paper or emery cloth for the purpose of removing irregularities from
the bearing surface; a half-round file or scraper should be used.
Care must be taken that the wedge has a good contact on the crown
of journal bearing. The surface of the journal should be smooth
and thoroughly clean before the bearing is applied. When a plying
a journal bearing, a coat of lubricating oil should be applielto the
bearing surface of same. Never wipe the bearing surface of the
journal bearing with waste.
Applicafion of Packing. ( a ) Inner. In acking a journal box,
twist somewhat tightly a rope of packing anBplace it in the extreme

P~CL

JOURNAL BOX PACKING

429

back part of the box, as shown a t A in Fig. 35. Make sure that
i t is well up against the journal so as to roperly lubricate the fillet
on the journal and keep out the dust.
Main. Apply su5cient
r k i n g (preferably in one piece) to fill the space shown a t B in
Ig. 35. Take care to have this packing bear evenly along full
length of the lower half of the journal. The packing should not be
too tight, but should be tight enough to overcome any tendency to
settle away from the journal. The packing should extend to approximately the center line of the journal but not above a t any point,
and should be pressed down evenly a t sides that no loose ends may
work u under the journal bearings. (c) Outer. Apply a third
piece oPfirmly twisted packing as shown a t C in Fig 35, and ack
tightly in order to prevent displacement of the main pacl!ing.
There should be no loose ends hanging out of the box. as they would
tend to draw out the oil.

6)

Packing Knife

.I --___-1

&
L
d
,

Packlng Hook

PIG. 37.-Journal

box packing tools.

The committee which recommended the above practice also

emphasized the importance of observing several other factors as
follows, with the view of reducing hot boxes to the minimum:
Journals after being turned should be cylindrical, free from taper,
tool marks, ridges, corrugations and other defects, in other words
the turned journal should reflect first class workmanship The
necessity of fully protecting journals against rust and corrosion during storage and guarding against damage to journals as a result
of improper handling should be recognized. Care should be exercised in handling journal bearings to prevent tossing of journal
bearings against each other, thus nicking andneedlessly damaging
the smooth bearing surface of the babbitt metal lining.
Figure 37 shows a representative set of journal box packing tools.
Journal Temperatures. The following comparing journal temperatures reached a t different air temperatures are from tests by
the use of the University of Illinois dynamometer car which is
equipped with a recording thermometer, the bulb of which is inserted
in a hole drilled in the body of one of the journal brasses. The
car weighs 58,000 Ib. and is equipped with 4% by 8 in. journals:

430

ELECTRIC RAILWAY HANDBOOK

Average air temperature. deg. P . . .....................

A proximate average test speed, miles per hoar.. .......
dxirnurn temperature attamed by test car journal.
deg.P

.........................................

:36

15

116

98

Truck Overhauling. Complete overhauling means that the

trucks must be entirely dismantled, the various parts repaired or
excessively w o n parts replaced, and the trucks again assembled.
A thorough overhauling should place the trucks in the best of o erat
ing condition and as nearly as possible in their original con8itioi
The basis for truck overhauling varies among different railway
companies and conditions, some overhauling on the time basis in
order to schedule their work in advance, while large systems employ
the mileage basis, as it is really the mileage made by the equipment
that determines the need for repairs.
The work of overhauling may be divided into the following: (I)
measurements to determine necessary adjustments; (2) disconnecting brake rigging and motor leads; (3)lifting car body from trucks;
(4) placing trucks in
sition for overhauling; ( 5 ) inspection to
determine repairs nee&; (6) dismantling equipment and parts;
(7) repairing various parts; (8) reassembling; (9) running trucks
under car body; (10)lowering car body; (11) connecting brake
rigging and motor leads; (I 2) final measurements and adjustments.
Some roads clean the trucks by some method or other before dismantling. The measurements taken for c a n in city service may be
step height, wheel guard height, side bearing height, or measurements a t the corners of the car body to detect weak springs or
defective parts. Cars in interurban or rapid transit service require
measurements of platform and coupler height.
The following points were taken from the Westinghouse Railway Operating Data: (I) all parts of the truck and brake rigging
should be carefully gone over and the badly worn hangers, pins,
bolts, chafing plate, truck brake levers, and brake heads, should be
repaired or replaced; ( 2 ) truck side and center bearings should be
rigidly attached to thebolster; (3)height of bolster should be checked
by a standard gage supplied for the purpose; (4) new brake shoes
and turnbuckles should be supplied if these parts are badly worn;
( 5 ) adjust brake release spring; (6) wheels should be checked for
diameter and flange wear, and either t u n e d down or replaced by
new ones, if necessary; (7) pairs of wheels should be kept to the
same diameter to prevent crowding one rail or the other, witha
consequent destruction of flanges and an increased train resistance;
(8) diameters of pairs of wheels on the same car should not vary
more than onequarter inch, otherwise the motor on the larger
diameter wheels will be overloaded; (9) gears should be checked,
and renewed if they will not last until next overhauling period;
(lo) journal boxes, covers, bearings, check plates, etc., should be
inspected, and removed if badly w o n or broken; (11) axles should
be checked in a lathe, and if bent they should be straightened
or replaced by new ones; (12) if axles are badly w o n , or have
run their predetermined maximum mileage, replace by new
ones; (13)axle collars should be tightened and checked for proper
clearances.

SECTION VIII

BRAKING
Factors Controlling Length of Stop. The general facton which
determine the distance traveled by a train while i t is being brought
to rest on a tangent level track are: ( a ) weight and speed of train
a t the beginning of the braking period; (b) coefficient of adhesion
between wheel and rail (rail friction); (c) coefficient of brake shoe
friction; (d) maximum brake cylinder force; (c) time required to
attain this force; (j)efficiency of brake rigging in multiplying and
transmitting this force to the brake shoe.
Available Adhesion between Rail and Wheel. This factor is
subject to a wide variation on account of the varying condition of the
rails and wheels with changing weather. For example, the adhesion between a dry rail and wheel may be twice that of a wet rail.
The addition of sand to a slippery rail will increase the adhesion
from 15 per cent to about 25 per cent of the weight on the rails.
The maximum adhesion occurs when the wheels are rolling on the
rail, and this adhesion ra idly decreases as soon as slipping occurs
the force of brake shoe friction which
between wheel and rail.
opposes the rotation of the wheels must never exceed the force of
the adhesion between the wheels and rails which is keeping the
wheels rotating. (See Coefficient of Adhesion, page 157.)
Brake Shoe Pressure. The maximum retardation which may
be utilized in stopping a train by means of brakes is that which is
realized by so applying brake shoes to all the wheels that the resulting brake shoe friction shall be uniform and just insufficient to
overcome the static rail friction or adhesion. Thus the utilization
of the entire retarding force available as rail friction or adhesion
involves the application of a brake shoe pressure which shall ( a )
diminish as declining speed causes the coefficient of friction to
increase; which shall (b) increase as increased distance of frictional
contact causes the coefficient of friction to decline, and which shall
(c), when diminishing or increasing for such purposes, further
diminish or increase as reduction or increase of pressure itself
causes the coefficient of friction correspondingly to increase or
decline. I t is impracticable to so regulate the brake shoe pressure
that a t any instant it will be equal to a maximum as the resultant
of the above three, but since this resultant continues to diminish
during the process of bringing the train to rest, the maximum
possible retarding effect may be approximated by so diminishing
the brake shoe pressure toward the completion of the stop that the
wheels will not be caused to slide, nor the passengers and equipment
be subjected to undue shock.
43 1

%bus

432

ELECTRIC RAILWAY HANDBOOK

Emergency Braking. The above discr~ssionshows that an emergency stop for high speed is less e6c1ent than for a low speed, since
the maximum pressure which will not slip the wheels near the end
of the stop is applied at the beginning of the emergency application.
A shorter stop would result if the pressure during the first part
of the braking period was greater than that which would slip the
wheels a t low speed. This would require, however, some means
to decrease the pressure near the end of the stop in order thpt the
limits of rail friction are not exceeded and the efficiency of the stop
thereby decreased. The additional apparatus is not desirable
unless the conditions warrant the further complications. (See
"Practical Application of Principles," Page 436.)
Importance of High Rate of Retardabon. In long runs, the braking rate is of little moment, but in short runs it becomes an important factor and, as in subway work, may be of maximum importance.
In general, a high braking rate is advantageous because it allows
more coasting in any run, which means an earlier point of cut-off,
resulting in lower energy consumption and less heating of the
motors than would accompany a low rate of braking. Fast brak .
ing thus tends to minimize the size of motor for a given service.
The minimum length of a block and the consequent minimum headway is equal to the minimum distance in which a train can '&
brought to rest. Thus, increasing the rate of braking decreases
the a\lowab\e headway and inc~easesthe capacity of the track.
High Speeds and Wheel Failures. In the use of high speeds
care should be taken that the conditions of braking are such that
they will not be dangerous to car wheels. The temperature
attained a t the wheel tread during braking depends upon the pressure a t which the brake shoe is applied, the speed a t which the wheel
is turning, and the duration of the brake application. When
the wheel is turning a t high speed and the brake shoes are a p lied
with great pressure, the metal a t and near the surface o f t h e
tread suddenly becomes much hotter than that a short distance in.
Great thermal stresses are thus set up in the wheel and these may
cause fracture.

'

433

COEFFICIENT OF BRAKE SHOE FRICTION

Coefficient of Friction a t Various Speeds

Number of experlments from

which the mean
is taken

Rennie.

Speed
Miles per
hour

Feet per
second

Coefficient of friction
Extremes observed

I Mean

Maximum1 Minimum]

Static friction under

pressure of 180 Ib. per square inch.. . . . . . . . . . 0.300
pressure of 336 Ib. per square inch.. . . . . . . 0.347

From the above values R. A. Parke has developed the following formulas to represent the law of variation of the coe5cient
of friction with speed:
From the mean values
from the maximum values

0.326

= I + o,03532S

0.382

f = 1

+ 0.02933s

where f = coe5cient of friction

S = speed in miles per hour.
The latter formula gives values corresponding more nearly to
recent e erirnents.
The f3owing table furnishes a comparison of the values of the
coe5cient of friction as calculated by the above formulas, with
those secured by test (see above table):

434

ELECTRIC RAILWAY HANDBOOK

S d.
mlEper
hour

Coefficient of friction

Mean
Calculated

Observed

Maximum
Calculated

Observed

.............

10

0.326
0.277
0.241

0.330
o. a73
0.242

15
ao
2s

0.213
0.191
0.173

0.223
0 . I92
0.166

0.265
0.241
o. azo

0. 280
0.140
0.205

30
40

0.1~8
0.146
0.135

0.164
0.142
0.140

o.ao3
0.188
0.176

0.196
0 . I97
0.194

45
50
55

0 . 126
0.118
0.111

0.127
0.116
0.111

0.165
0.155
0.146

0.179
0.153
0.136

60

.............

0.074

0.138
0.131
0.12s

.............

70

0 . 105
0,099
0.094

80
90
100

0.085
0.078
0.072

0.114
0.106
0.097

.............
.............

35

65

.............
.............
.............
.............

0.382
0.333
0.295

0.340
0.~81

0.123

.............
.............

Relation of Coefficient of Friction between Wheel and Rail to the

Coefficient of Friction between Brake Shoe and Wheel and the
Effect of Sliding. In the Galton-Westinghouse tests it was found
that when the brake shoe friction overcame the rail friction and
caused the wheels to slide upon the rails, the coefficient of rail friction immedistely began to decline and then varied inversely as the
speed, in much the same way as brake-shoe friction. I t was found
to have a value of less than one-third the coefficient of brake shoe
friction when the brake shoe pressure was such as to allow the wheel
to continue revolving. From this, the retarding effect of a wheel
sliding upon a rail would be less than one-third the maximum available when the wheels are allowed to revolve without sliding.
COEFF~CIENT
OF FRICTION
BETWEEN STEEL TIRES AND STEEL
RAIL-WKEEL SLIDING. (GALTON-WESTINGKOCTSE
TESTS)
Coefficient of friction
0.242
6.8
0.088
13.6
0.072
31-3
0.070
34.1
0.065
0.057
40.9
47.7
0.040
54.5
0.038
0.027'
60.0
' This is from the mean of three experiments only.
Speed. miles per hr.
Just coming to rest

The rail friction is very materially affected by the condition of

the rail, being greatest when the rail is perfectly dry or very wet

COEFFICIENT O F BRAKE SHOE FRICTION

435

(as when washed by a hard rain) and least when the rail is quite
moist; but by the use of sand upon the rails, the effect of moisture
is ~racticallyeliminated. (See Coefficient of Adhesion, y g e 1~7.)
Coefficient of Brake Shoe Friction a s AfFected by D~stanceof
Application. Capt. Galton stated that a decrease in the coefficient
of brake shoe friction results from the time during which the brakes
have been kept applied, irrespective of any change in speed. He
gave the following table and explained that the va!yes given in the
column headed "Commencement of experiment are somewhat
different from those that have been g i v y in the table "Coefficient of Friction a t Various Speeds . . ., page 433, because they
resulted from the average of fewer experiments, but that the
effect of time reducing the coefficient of friction may be accepted
as correct.

Speed.

----

miles per Commence'Our

e$x",",Lt

Coefficient of friction

zeconds

20

27
37
47

60

:::;:

0.182

0.132
0.073

o.15a
0
0
0.096
0.080
0.063

After 10
seconds
0 . I33
0.070
0.058

After 1s
seconds
0.116
:;%

After 79
seconds
0 . m
0.072

...........
...........
...........1:::: .......
.......
-

The following is from a paper by R. A. Parke, A.I.E.E., 1902:

" I t was assumed by Capt. Galton that the decline of the friction is
as the increase of the time of contact, but a careful analysis of the
results shows that i t is a function of the product of the speed and
the time, or of the distance, through which the shoe rubs upon the
wheel." The following formula was developed by Mr Parke in an
endeavor to show the relation of the coefficient of brake shoe friction a t any point during the application of the brake shoe to the
initial coefficient of brake shoe friction a t the speed and pressure
a t which the brake shoe was first applied to the wheel and the distance traveled by the wheel during which the brake shoe has been
applied to the wheel:

f-+
0.0004721 f
0.002390l
in which h = coefficient of brake shoe friction a t instant considered
I = distance wheel has traveled in frictional contact with
the brake shoe, feet
f = coefficient of brake shoe friction a t s eed and pressure a t which brake shoe was applieg a t the beginning of the distance 1.
S. W. Dudley, in a paper before the A.S.M.E., 1914, on the
Pennsylvania-Westinghouse Brake Tests, 1913, stated that factors
=

436

ELECTRIC RAILWAY HANDBOOK

such as s
. pressure and time of action, which are ordinarily
considere to cause variations in cast iron brake shoe performance,
are effective chiefly as they affect the temperature of the working
metal of the brake shoe and wheel.
Practical Application of Above Principles. I n emergency application of modern electric railway air brakes, a high cylinder pressure
generally is held without reduction until the release is made. This
a t first appears contrary to the princi les established by the
Galton-Westinghouse brake trials to the eekct that, as the speed oi
a train is diminished by continued brake action, the effectiveness
of a given brake shoe pressure gradually increases, due to the
coefficient of brake shoe friction increasing with decreasing speed.
However, as pointed out by S. W. Dudley (Electric Journal, 1920)
this principle remains as firmly established today as when first
demonstrated. But i t is found that with emergency braking of
1 2j to 150 per cent maximum in modem practice (as compared with
the 2 0 0 to 300 per cent used in the Galton-Westinghouse trials)
and the amount of work done per unit of brake shoe bearing area,
the abrasion and heating of the brake shoe surface is such that the
coefficient of friction remains substantially constant for the major
portion of the stop. The increase in coefficient of friction and
consequently in retarding force for a given brake shoe and brake
cylinder pressure, which invariably occurs when the speed decreases
below 25 miles per hour, although considerable, either does not
become high enough to cause the wheels to slip a t all or not until so
near the stopping point that what slipping does occur is not detrimental. A significant fact in this connection is often not fully
appreciated. I t is better to take advantage of the maximum
possible retarding force throughout the major portion of an emergency stop and risk slipping the wheels for the last ten or twenty
feet, than to proportion the braking force so that the wheels will
not slip a t all, even on a bad rail, which may mean a collision when
a safe stop could otherwise have been made. For these reasons,
it has been found both desirable and feasible to hold the maximum
braking force obtainable in emergency applications without
reduction, until the train is stopped.
Distance Traveled by Train during Application of a Constant
Retarding Force. If a constant retarding force is applied to a train

in which 1 = distance traveled by train in coming to rest, feet

S = speed of train a t beginning of period, miles per hour.
a = rate of retardation of train, miles per hour per second
k = ratio of linear inertia to total inertia of train (see
page 151)
W = weight of train, tons

BRAKING DISTANCE

437

F = total retarding force applied to train, pounds

= Ph

or
= 2000bW
P = total braking pressure applied normally to wheel

treads by brake shoes, pounds

h = coefficient of brake shoe friction
6 = ratio of retarding force produced by brake shoes to
total weight of train
FI = combined train, track grade and curve resistnnce
= jw

j = train resistance, pounds per ton weight of train.

Neglecting the energy of rotation of wheels and armatures, the
formual is

sa
306
Braking Distance, Single Car Tests. The following results were
obtained from Bulletin No. 13, Engineering Experiment Station,
Purdue University. This bulletin publishes the results of emergency stop tests on four different cars:
=

I.
2.

Interurban car double truck. four motors.

City car, doubie truck, four motors.

3. C ~ t y
car. double truck. two motors.

4. Birney safety car.

As the result of a careful analysis of the test data, the following

conclusions were reached: (I) There is no one best way to stop any
car in an emergency. I n general, the best way will depend somewhat on the type of brake and motor equipment. (2) As a broad
rule, subject to occasional exceptions, i t may be said that the best
way to stop a car equipped with air brakes is to apply emergency
air and sand, simultaneously throwing the controller to the off
position. (3) The electrical methods of braking, reversed motors
and bucking motors, subject the equipment to very severe strains.
Because of the abruptness with which the braking force is applied,
the value of these methods in emergency braking is more apparent
than real, except in those cases in which the air brakes from some
cause are inoperative. (4) Of the two electrical methods, reversed
motors and bucking motors, the reversed motors method is the
most effective. (5) If the motors are reversed in conjunction with
emergency air and sand, skidding invariably results, and even
under good rail conditions, the results are not superior to those
obtained using air alone. (6) The electric braking methods are
more likely to produce erratic results than the emergency air, and
for this reason should not be resorted to as a general practice. (7)
Because the rail conditions under which these tests were made were
better than those ordinarily obtaining on city streets, the results
should be regarded as minimums rather than averages or maximums.
(8) The best way to stop a Birney safety car in an emergency is for
the motorman to raise his hand from the controller handle. This is
true only when the car is running with brakes released. With

ELECTRIC RAILWAY HANDBOOK

438

brakes applied the brake valve should be thrown into the emergency
posltion

DESCRIPTIONOF CARSUSEDI N

Car No

---

61

C~ty
42,758 lb.

38,507 1b.

No

Wheel d~ameter
Per cent braking power a t
50 Ib brake cyltnder
pressure (car empty)
Shoe area m contact, per
cent
Sander

I
(

244

(Interurban

1026

TESTS

82.139 1b.

qWelnht
g h t loaded
llnht load
NO. ?ruck;
motors
Type brakes

PuRDUE

City
1

Blrney
20.446 Ib.
16 OAS Ib.
1

4
Strarght &
Automatlc
35'952 ln
roa

55

Safety
2446 In.
58 o

BRAKINGDISTANCES,
PURDUETESTS
(Distances In feet)
See table above for descr~ptlonof cars and equipment

---

1 Is

Speed
mp'h

A A'

5
10

15
20
25
30
35
40
45

140
205
a90
380
500
650

k",?

Heavy load

Z P L P H P L P HL HL

HL H I ,

I2
15
36 55 32 37 27 30 30 35 24 a s
95 78 105 62 7 2 55 60 60 66 53 58
160 130 167 105 I17 92 98 98 I05 87 roo
245 ZOO 245 160 172 I35 I42 I45 I55 I25 140
azs 248
340 285
440
550
660

NOTE:

S & A Strarght and Automatic Air

A Automat~cAlr
E Emergency Alr
S Stralght Alr.
L P Low Pressure
H P Hlgh Pressure
H L Hand Llfted from Safety Controller

Braking Distance Xncluding Application Time. Neglecting the

energy of rotation of wheels and armatures, neglect~ngtrain resistance, and considenng the time required for the brakes to become
effective, when the wheels do not slip
I

= r 467.9

S
+ 3oPels
-

in which 1 = length of stop, feet

S = speed of train a t beginning of stop, miles per hour
1 = time a t b e ~ n n i n gof the stop during wh~chthe brakes
are to be considered as having no effect, seconds.

439

BRAKING DISTANCE

T h e value of 1 depends upon the equipment. I n the

Pennsylvania-Westinghouse 1913 tests on a twelvecar train i t was found to range from 2 o to 2 5 seconds
for the quick a c t ~ o nautomat~c(PM) brake and from
o 70 to o 85 seconds for the electro-pneumatic (UC)
brake
P = nominal per cent braking power (decimally expressed) corresponding to the average cyllnder
pressure existing for t h a t portion of the stop after
the brake is considered fully applied (The nomlnal
per cent braking power is the ratlo of the total shoe
pressure, calculated from the full service brake cylinder pressure, to the welght of the car, expressed in
per cent)
e = efficiency of brake rigging (decimally expressed)
h = coefficient of brake-shoe frict~on
I n the Central Electric Railway Association-Purdue University
tests of emergency braking on electric cars the following results
were obtalned

TIME
L ~ BETWEEN
c
STOPINDICATION
AND BEGINNING
OF RETARDATION
--

(For descrlptlon of cars and

ment, see page 4 3 8 )

equip

Car N o . .

Time from the stop lndlcatlon to the

beglnnlng of rctardatlon. seconds

---

-- (
61

244

---

--loaf,
---

(5s

Automatic alr

Stralght and automatrc alr

Stralght alr
Reversed motors
Buckrng motors
Reversed motors plus emergency alr
Hand llfted irom safety controller

NOTE I n the above tests the lnltlal time lndlcatton was made automat]cally In the trme clrcult of a chronokcaph In the alr brake tests whlle In the
electnc braking tests the lndlcatlon was made b the man1pulat;on of a sw~tch
by the chronograph operator upon the report o r a gun whlch was the slgnal
a t w h ~ c hthe motorman applied the brakes.

Brake Shoe Pressure and Braking Power. Nominal Brake Shoe

Presswe is that obtained by mult~plyingthe brake cylinder (or hand
brake lever) ressure by the lever ratlo of the brake n g o n g On
account of tRe friction in the brake rigging Levers this nom~nal
brake shoe pressure is never attained.
Actual Brake Shoe Pressure is the actual pressure exerted by the
brake shoes against the wheels, and is equal to the nominal pressure
divided by the efficiency of the brake ngging (the system of rods and
levers between the cylinder and the brake shoes).
Brakzng Power, usually expressed as a per cent, is the ratio
between the brake shoe pressure and the we~ghtof the car on the
wheels to which such ressure is applied. T o obtaln nomrnal
braking pourer, nominal grake shoe pressure should be used, or for

ELECTRIC RAILWAY IIANDROOK

440

actual b r a ~ i n ~ ' ~ o zactual

v e r brake shoe pressure as above described
should be used. It should be noted that the term " braking power"
is recognized to be a misnomer and should properly be replaced by
"braking force," hut as the former is the generally accepted term,
i t is thought best to conform to common usage herein.
The following values for braking power were given by E. H.
Dewson in the Electric Journal, 1905,to show the approximate
relation which the pressure applied to the brake shoes should bear
to the total weight of the braked wheels to produce a brake friction
equivalent to the adhesion of the wheels to the rails. Consideration is given to the fact that the coefficient of brake shoe friction
is less a t high speeds than a t low.
--

Approximate ratio of total pressure on brake shoes

to total weight on braked wheels,
Speed.
miles per hour

Coefficient of adhesion
0.30

Jo

60

2.48

4.14

0.25

2.07
3.47

0.20

0.15

1.65

1.24

2.77

2.08

Brake Rigging Efficiency. A part of the brake cylinder force is

lost in being transmitted to the brake shoe and as the leverage of
the brake rigging is calculated from the piston force i t is necessary
to make proper allowance for this frictional loss. The efficiency
of the rigging will depend upon the equipment, condition and pressure used. Tests to determine its value have been unsatisfactory.
Such tests have yielded various values, b u t they indicate t h a t in
most cases the value for ordinary equipment in good condition
probably lies between 80 and 85 per cent. Commenting on
the Pennsylvania-Westinghouse 1913 tests, S. W. Dudley,
A.S.M.E., 1914,states that they indicate that a t least 85 per cent
transmission efficiency could be obtained with either single shoe or
clasp brake rigging. These tests indicate that the following
features are important in securing the maximum overall brake
rigging efficiency: ( a ) protection against accidents that may result
from parts of rigging dropping on the track; (b) maximum efficiency
of brake rigging a t all times to insure the desired stopping with a
minimum per cent of braking power; (c) uniform distribution of
brake force, in relation to weight braked, on all wheels; (d) with a
given nominal per cent braking power, the actual braking power to
remain constant throughout the life of the brake-shoes and wheels;
( e ) piston travel to be as near constant as practicable under all
conditions of cylinder pressure; (f) minimum expense of maintenance and running repairs of brake rigging between the stopping
of cars.
The total retarding force acting on a car may be determined by
multiplying the weight of the car in tons by the retardation in

441

BRAKE EFFICACY

miles per hour per second and this product by 91.1. Jf from this
force, the train resistance be subtracted, the remainder will be
the retarding force produced by the brakes. The brake rigging
efficiency then will be:
Braking Force
Efficiency =
Braking Power X Coefficient of Shoe Friction

Brake E5cacy. As it is difficult to determine the coe5cient of

shoe friction with accuracy, the product of this coe5cient and the
brake rigging eBiciency are often used in comparing the etficacy of
braking systems. S. W. Dudley, A.S.M.E. 1914, used the following
values of eh, brake efficacy, as determined in the PennsylvaniaWestinghouse 1913tests.
K i d of brake rigging

30

80

Single s h a g

1 Flanged 1

Plain

1 Flanged

150
180

0.~41
0.129
0.118

0.169
0.154
0.141

0.108
0.099
0.090

o.rre
0.103
0.0~4

12s
150
180

0.103

0.122
0.112
0.102

0.074
o 068
o:o62

0.090
0.082
0.075

0.092
0.084
0.077

0.109
0.100
0.092

0.070
0.064
0.059

0.074
0.068
0.062

125

60

Clasp brake
Plain

Type of brake shoe

12s

I50
I 80

Value of data uncertain, due to non-uniform brake shoe conditions.

I n this test a single car (locomotive not attached) was stopped

from an initial speed of 60 miles per hour in 725 ft. This was the
shortest 60 miles per hour emergency stop made. The equipment
was electropneumatic, clasp brake, flanged brake shoes, using 180
per cent braking power. Under the same conditions the shortest
stop from 80 miles per hour was 1422 feet.
I n the Central Electric Railway Association-Purdue University
tests, the brake efficacies were calculated for a few cases and are
given in the following table. (For description of cars and equipment, see page 438.)
car

61
1026
1026

Speed
m.p.h.

;25.5

I/

~ r a ~ e
efficacy

1 i(ii )
0.192

Car

;:

speed
m.p.h.

1
22

/~;2~
:::"
0.131
0 . 104

442

ELECTRIC RAILWAY HANDBOOK

Retardation Produced by a Given Retarding Force. The rate of

retardation may be approximated by the following formula:
a = 0.01098

k(F
FI)
W

in which

a = rate of retardation of train, miles per hour per second

= ratio of linear inertia to total inertia of train (see
page 151)
I.V = weight of train, tons
F = total retarding force applied to train, pounds
= Ph

= 2ooobW

P = total braking pressure applied normally to wheel

treads by brake shoes, pounds

h = coefficient of brake shoe friction

b = ratio of retarding force produced by brake shoes to
total weight of train
F I = combined train, track grade and curve resistance
= jw
j = train resistance, pounds per ton weight of train

For a more rapid, very rough approximation in ordinary service,

neglecting the energy due to the rotation of the wheels and armatures and the energy required to overcome train resistance, the
formula is:

in which the symbols have the same significance as above.

Braking Distance Chart. The chart, Fig. I , by Gaylord
Thompson, A.E.R.E.A. 1914Committee on Block Signals, gives
a graphical approximation of the distance and time in which a train
will be brought to rest when retarded a t a constant rate. Example:
Assume that such a braking force has been applied to a car as
to produce a retardation of 3 miles per hour per second. It is
des~redto know how long a time will be required for the car to
come to a stop and in what distance i t will do so from a speed of,
say, 50 miles per hour. The solid line for 3 miles per hourper second shows that a speed of 50 miles per hour corresponds to a braking distance of 600 ft. This is indicated by the point A. Directly
below A , a t the point where a vertical line dropped from A intersects
the distance-time curve for 3 miles per hour per second and reading
on the right-hand scale, we find that the time required for the stop
is 16.5 seconds. That is, with this rate of retardation it will take

RATES OF RETARDATION

443

16.5 seconds to stop a car running a t 50 miles per hour, and in s t o p

ping it, a distance of 600 ft. will be covered. A similar procedure
shows that from 60 miles per hour the same braking rate will stop
the car in 2 0 seconds and 880 ft.

PIG. I.-1)istance

and time for braking.

Rates ofRetardation. The ejective rates of retardation obtained

in the Central Electric Railway-Purdue University tests are as
tabu'ated below, in miles per hour per second (for description of
cars and equipment, see page 438).
Car
No. 61

car
No. 244

Car
No.1026

Car No. 55

25
30
35
40
45

2.25 I 8 9 2 . 3 1 1.89 z 89 z 6 8 3 4 2 3 . 2 5 3 1 9 2 . 9 8 3 70 3 30
293266.
.
. .
2.~81942.32
a 362.08 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
a 35 2.13 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . ! ....
a 31 2 28

Note.-A

Automatic Air.
S & A Straight and Automatic Air.
t
E
Emergency Air.
S S t r a ~ g h Am.
H P High Pressure.
L P
Low Pressure.
H L Hand Llfted from Safety Controller.

The "effective" rates of retardation as shown above were calculated by using the distance from sto signal indication to full
stop and the speed a t the stop
indication. Thus thew

signs!

444

ELECTRIC RAILWAY HANDBOOK

values assume a uniform rate of retardation and include the lag

due t o personal element of motorman and due to the time i t takes
for the pressure to build up in the brake cylinder. I n the tests, the
actual braking rate varied throughout the braking period, in general
becoming the maximum a t the end of the braking period. For
example, the following maximum values were noted in the report
on the above cited tests:
Car 6 1 . at 1 o . s
Car 244,at 7 . 5
Car 1026,at 8 . 0
Car 5s. a t 8 . 2

m.p.h.. 4 . 9 m.p.h.p.s.
rn.p.h..4.97 m.p.h.p.s.
m.p.h., 5 . 1 m.p.h.p.s.
m.p.h.. 7.28 m.p.h.p.s.

Weight Transfer Due to Braking. The following is based on

the analysis by R. A. Parke, A.I.E.E., 1902:
Retardation caused by the application of brakes to the wheels
involves a redistribution of the weight which, where the retardation
corresponds to the maximum brake application, results in a very
serious loss of braking efficiency, unless means be provided for varying the brake shoe pressure to correspond with the changed wheel
pressures against the rails. Whatever the source of the retarding
force, its operative effect in retarding the motion of the car is the
same as that of an infinite number of small retarding forces each
engaged in retarding the motion of an elementary portion of the
mass of the car, and therefore, in order that a single retarding
force, or a combination of retarding forces, shall so operate,
without either changing the direction of the car's motion, or
producing rotation of the structure as a whole, or calling into
operation other forces to prevent such deviation or rotation,
the force or the resultant of the conibination of forces must be so
applied that it shall pass through the center of gravity of the mass,
in a direction opposite to that of the motion of the car. I n utilizing
the rail friction for the retarding force, while it has the proper
direction, it is applied a t the lowest points of the mass of the car and
must therefore either cause rotation of the entire structure or the
interposition of other forces which combine with the retarding
force to preserve the simple motion of translation. The car may be
considered as a single mass or as being composed of three separate
masses, according as it is provided with one rigid or two sw~veling
trucks. Rotation of the car by the eccentric retarding force does
not occur in either case; but, in the first case, this is because the reacting pressure of the rails upon the forward pair of wheels exceeds
that upon the rear pair of wheels in such measure that the contrary
rotative moment, thereby introduced, just balances that of the
eccentric retarding force, also applied by the rails to the wheels.
I n the case of the car with two trucks, the retarding force is applied
a t the lowcr extremity of the two trucks, being equa!ly divided
between them (if constructed alike); that portion of the retarding
force necessary to retard the mass of the trucks is absorbed in so
doing, and the remainder is applied by the trucks to the car body a t
substantially its lowest extremity. I n consequence, the center of
gravity of the car body being above the points of application of the
retarding force, rotation through the eccentrically applied retarding
force is prevented only by the resisting rotative moment of a greater
supporting pressure from the forward than from the rear truck.

WEIGHT TRANSFER IN BRAKING

446

Each truck is subject to the combined rotative moment of the

eccentric retarding force a t its lower extremity, and the eccentric
reacting force from the car body a t its upper extremity, and rotation
is prevented only by the contrary rotative moment of a greater
supporting pressure by the rails upon the forward than upon the
rear pair of wheels. Thus the very act of applying the brakes to
the wheels produces a new and very different system of wheel
pressures upon the rails, and i t is the wheel pressures under these
conditions which determine the available retarding force. AS
the total pressure of all the wheels upon the rails cannot vary. the
existence of a greater rail pressure for the forward than for the
rear pair of wheels of the truck implies the virtual transfer of a portion of the normal pressure from one pair of wheels to the other.
The brake shoe pressure upon the rear pair of wheels must be insuflicient to cause the wheels to slide upon the rails and must therefore
be cut down in proportion to the transfer of weight from the rear
to the forward pair of wheels. But as the forward pair of wheels
will become the rear pair when the car moves in the opposite
direction, the brake shoe pressure upon that pair of wheels must
also be limited in the same way. Thus, the braking pressure upon
each pair of wheels must be restricted to correspond with the
minimum pressure a t the wheels upon the rails, which occurs when
the wheels are the rear pair.
External Forces Acting on Car Body i n Braking. (Fig. 2.) The
following equations were obtained by taking the summation of the
horizontal forces, summation
Dtnollr of Y d l p
of the vertical forces, and moI
mentsof forces about the point
of application of PI
1
-

PI(.. 2.-Weight transfer in braking.

Car body.

in which H = horizontal retarding force on each truck center pin,

. pounds
(H is the same for each truck because the brake
pressure is to be limited for each pair of wheels to
correspond with the condition of minimum pressure
of wheels on rail, and hence each truck during braking will exert an equal retarding force for equal
brake pressure on all wheels of car)
Pp = pressure between body and truck rear center plates,
pounds
PI = pressure between body and truck forward center
plates, pounds
WI = weight of car body, y n d s . (Center of gravity
being in a vertical ax= midway between truck
centers)

ELECTRIC RAILWAY HANDBOOK

446
j

I
g

a
k

height of center of gravity of body above center

plate surface, inches
= distance between center pins, inches
= acceleration due to gravity, feet per second per
second = 3 2 . 2
= rate of retardation, feet per second per second
= (miles per hour per second) X 1.467
= ratio of linear inertia to total inertia of cars. (See
p. '51.1
=

External Forces Acting on Rear T r w k i n Braking.

hW2ad
R2 = W
- + P - fI2
2
b g b
R I -- w ~ + P , + ~ l rwoad
+2
2
b n b

PIG. 3.-Weight

transfer in braking.

(Fig. 3.)

Truck.

in which R2 = total pressure between rail and rear pair of wheels,

pounds
R 1 = total pressure between rail and front pair of wheels
of rear truck, pounds
T I = total maximum retarding rail friction available
between rail and rear pair of wheels, pounds
T 1 = total maximum retarding rail friction available
between rail and front pair of wheels of rear truck,
pounds
W , = weight of each tmck, pounds. (Center of gravity
of truck being in a vertical axis midway between
wheels)
W = total weight of car, pounds = W I 2 W 3
h = height of truck center plate surface above rail
b = wheel base of truck
d = height of center of gravity of truck above rail
q = coefficient of adhesion between wheel and rail
For significance of other symbols see above.
The above formulas do not take into consideration the weight
transfer due to the inertia of rotating parts (wheels and motor

BRAKE SHOE SUSPENSION

447

armatures). This effect may vary, depending upon the relative

weights and arrangements of such rotating parts; the effect is small,
however, and in all practical cases, the formulas as shown will give
results within the limits of error in the assumptions necessary with
res ct to locations of centers of gravity and coefficient of adhesion.
g a k e Shoe Suspension. The application of the brake shoes at
the outer face of the wheels results in an upward thrust of the brake
hangers, proportional to the brake shoe friction, upon the rear end
of the truck frame, and a corresponding downward drag upon the
forward end. However, by hanging the brake beams between the
wheels, instead of outside, and inclining the hanger links a t a proper
angle, the increased pressure and consequently the increased frictlon
of the brake shoes upon the forward pair of wheels and the diminished pressure and friction of the brake shoes upon the rear wheels,
due to the effect of the friction itself in causing the shoes to press
more or less forcibly upon the wheels through the angularity of the
hanger links, are made to correspond with and compensate for
the transferred weight from the rear to the forward wheels. In the
same manner that running in the opposite direction causes a reversal of the conditions for the transfer of weight, so, too, the rotation of the wheels in the opposite direction causes a reversal of the
effect of the inclined hanger links, and the increased brake shoe
pressure is always applied to the wheels carrying the increased
weight. The application of this method of inclined hanger links
is not without some difficu1ty. The chief trouble is that no constant
angle of the links can be maintained, as the wearing away of the
brake shoes, together with wearing and turning down or grinding
of the wheel treads, causes constant and considerable variation.
Thus, if the angle of inclination and the braking pressure be calculated for the conditions existing when the brake shoes and wheels
are new, the increased angle when the shoes bccome much worn
and the treads have been well turned off, would probably cause the
forward wheels to slide upon the rails. On the other hand, if the
calculations be made for turned or ground wheels and worn shoes,
the rear wheels would probably slide when the wheels and brake
shoes are new. I t is therefore necessary to compromise between
the extremes, in reference to the angle of inclination of the hanger
links, by inclining the hanger links a t the angle determined when
the brake shoes and wheels are each half worn, and the brake beam
force must then be so
established that neither >I- *
orxotlm
pair of whee!s shall be
caused to slide in the
extreme positions of the
hanger llnks.
Angle of Suspension
of Brake Hanger. (Fig.
4.) The following formula gives the angle between the brake hanger
FIG. 4.-Brake-hanger suspens~onangles.
link and the tangent
to the wheel a t the center of pressure of brake shoe face necessary to secure the maximum

448

ELECTRIC RAILWAY HANDROOK

retarding force with inside hung brake shoes and with motors driving all axles:

in which 4 = angle between brake-shoe hanger-link and the

tangent to the wheel a t center of brake-shoe surface
r = coefficient of brake-shoe friction (0.33 may be used
for this).
For significance of other symbols, see pages 445 and 446.
E f i c t of Roldional Znerlia. I n addition to destroying the energy
of translation existing in the moving car, the brake shoe friction
must also absorb the rotational energy of the wheels, axles and
motors. This inertia is entirely independent of that due to translation, and in destroying it the coe5cient of adhesion between
wheels and track does not enter. Practically, a greater braking
force must be used to produce retardation when the wheels and
other rotating parts are taken into consideration. This is taken
care of in the above formula by the value k.
Brakc-beam Pressure.(Mofors Driving A I1 Axles). (Fig. 4.)
q U r t ( l - r z tan2
4) cos 4
P =
4rk(W
~ y i v cos (a-I-4)

in which P = horizontal braking force applied to brake beam in

the direction of motion of the train, or, the sum
of the horizontal braking forces applied to one pair
of wheels in the direction of motion of the train,
pounds
a = angle between radius of wheel to center of bmkeshoe surface and the horizontal.
For significance of other symbols see preceding- -paragraph, also
pages 445 and 446.
Wear on wheel and brake shoe will brinn about a change in the
values of the angles a and ,j, and the dimension j, consequ&tly the
value of the force P will vary during the life of wheel and brake shoe.
The value of the force P should be determined for new wheels and
brake shoes, and again for worn wheels and brake shoes. The
lesser value thus obtained will be the maximum value of P that
may be used safely.
Brake Rigging Calculation. Fig. 5 gives formulas for calculating truck pull rod force, brake shoe pressure and the dimensions
of the lever carrying the brake shoe for the three general methods
of suspension. Figs. 6 and 7 give formulas which show the relations
between the brake shoe pressures on adjacent wheels and the pull
rod force. By an adjustment of the lever dimensions these brake
shoe pressures may be made to bear a desired relation to each other.
The following example of brake rigging calculation (see Fig. 9 )
is by E. H. Dewson, Elerlric Journal, 1905:
A car weighing 46,700 lb. without load has a weight of 32,200 Ib.
on the motor truck and 14,500 Ib. on the trailer truck. The motor

BRAKE RIGGING CA1,CULATION

ELECTRIC RAILWAY HANDBOOK

450

truck is equipped with inside hung brake shoes connected directly

to a double set of levers, no brake beams being used. The trailer
truck has outside hung shoes mounted on brake beams, conatquently a single set
of levers is used.
One hundred p e r
cent brake power on
the motor wheels requires a force of 8050
Ih. a t each brakeshoe. Ninety per
cent brake power
on the trailer truck
requires a force of
6515 Ib. a t e a c h
brake beam. T h e
total brake power
will be 45,250 ' l b ,
ru Xb
b
P X a
and as a 10-in. pisa= w
F ;
ton under a pressure
1:"
F.(a + a',
W' FI x c
of 6 0 Ib. per square
(
r
+
d
)
(C
d) b
inch exerts a force
F ( a +b)
= Ilr, ,,,hen O
-4
FP --of 4700 lb., the tatal
b.
leverage ratio will be
PIG. 6.-Braking pressures on adjacent palm of about 0.6 to I , which
wheels. I.
is satisfactory. The
levers supplied with the motor truck are 27 in. long and the shoes
are hung 7% in. from the lower end. Considering the dead lever
first with its fulcrum
a t the upper end we
know the delivered
force W and itsdistance b from the fulcrum, also the distance a from the
fulcmm to the a p
plied force F; substituting these values
in the equation F =

JV

P
(a
-

+ b)

,
TYXA

a =

(a + b )

c"--@

I:

F'

-q

x-a
b

JV, -. '---(' + )
d
IT*

(i 2:

mb

we have F =
8050 X 19.5 = 5814
27
lb.
the stressrod.
in
the asadjusting

the lower end of live lever

as the fulcrum we
wheels. 11.
have W = 8050 Ib.,
b = 7.5 in. and a = 27 in., consequently the pull a t the upper end
- 2236. The proof is that 5814 2236 = %SO.
F = 8 0 5 0 X 7.5 b = F X a
IV - 1.'

1 Y ~ I l " w bh e nc ~ = ~Considering

FIG.7.-Braking forces on adjacent pain of

27

BRAKE RIGGING CALCULATION

451

As a pull of 2236 Ib. must be exerted on each side of the trucks,

the stress in the truck pull rod will be 4472 Ib. For the drad lever
of the trailer truck we have W = 6525 Ib., b = 22.5 in. and a = 15
in., consequently the stressin the adjust6525 X 22.5 ing rod is F =
15
Ib. With the intermediate point of the
wl--,Llive lever taken for the fulcrum, W =
6525 Ib., b = 7.5 in. and a = 15 in.
and the stress in the pull rod F =
6525
7.5 = 3262.5 lb. The proof is

1s

WI==WCO~O

that 6525 3262.5 = 9787.5.

in which W Iis force normal
I t is now necessary to so roportion to the wheel treads at centhe push rod Lever that wit! a piston terofbrake-shoesurface.
force of 4700 Ib i t will deliver 3262 lb. PIG. 8.-Relation between
horizontal and normal brak(neglecting the half) to the trailer truck ing;orcer
pull rod and the cylinder lever that it
will deliver 4472 Ih. to the motor truck pull rod. For efficient
operation of the s!ack adjuster these levers should not be less than
30 in. long; with short levers the action of the adjuster would
cause excessive angular distortion. If the applied force a t the cylinder end of the push rod lever is 4700 Ib., and the delivered force
a t the pull rod is 3262 lb., that a t the push rod will be 4700

Car weighs 46.700 Ib.

ji.200

I+SW

~ bon
. motor truck
lb. on trader truck

100%

brake power on motor truck

90% brake power on trailer truck

FIG.9.-Calculations for a specific brake equipment.

Assuming that the pull rod pin is the fulcrum we
F Xa
h a v e F = 4700, a = 30 in., W = 7962 and b = ---- - 4700 X 30
W
7962
= 171x6 in. For the cylinder lever we have F = 7962 Ib., W =
3262 = 7962 Ib.

4472 Ib., and b

X_J_O = 16% in.

7962
..
With the ordinary brake handle, staff and chain supplied with
electric cars, 1200 Ib. is about as much force as the average man
=

30 in., therefore a = 2
4

452

ELECTRIC RAILWAY HANDBOOK

can exert on the hand brake pull rod. This must be multiplied to
4700 to give the same brake power by hand as by air pressure.
Using a multiplying lever 18 in. long and connecting to the push
rod pin by a chain fastened 10 in. from the fulcrum, the force
required a t the end of the lever will be

lo

:i700

= 2611 lb.

Assuming that the hand brake lever is 6 ft. long and pivoted a t the
center, we have F = 1 2 0 0 Ib , a = 36 in. W = 2611 lb aod b =
1 2 0 0 X 36
= 16.54in., or practically 16%in.
261I
Safety Stops for Brake Mechanism. The double truck cars on
the Holyoke Street Railway are equipped with a special safety stop

FIG. r o -Safety

stops for brake rigging. Holyoke.

for the brake rigging, originated by the engineers of that company.

The mechanism, as shown in Fig. 10, prevents the brake rigging on
one end of a car from beconing noperative through the breakage
of a pin in the mechan sm on the opposite end. There are two
provisions to hold the lever operated by the air cylinder in case of
breakage of connecting pins. One of these is to protect against
the breakage of the in connecting this lever with the rod operating
the brake equalizer gar on each truck. This protection consists of
a short chain one end of which is attached to the lever and the
other is bolted to a car sill. The other safety device is to protect
against the breakage of the pins which connect the tie rod w ~ t hthe
two levers operated by the alr cylinder. This protection, as shown,
consists of a duplicate tie rod close to the o r r a t i n g one and with
a slotted hole a t one end in which the brake ever pin can move, so
that the duplicate tie rod does not interfere with the movements
of the levers.
Variable Load Compensating Device. With light weight on
the wheels a low brake cylinder pressure is essential in order that
slipping of the wheels may not occur with no load. If the same

VARIABLE LOAD COMPENSATOR

453

brake cylinder pressure is used with a loaded car, stopping distances

will be excessively long, with increased danger on severe grades.
The use of the variable load compensating device limits the
p r e s s u r e ~ i n gfrom the operator's brake valve or from the emergency v ve to the brake cylinder to a maximum amount which
will not cause wheel sliding, whatever the live load on the car may be.
I t is thus possible to obtain a given braking ratio with the car
empty or fully loaded or with some intermediate weight. When
less than a full service application is desired, the motorman can
make a partial application with his brake valve and then graduate
the cylinder pressure to that desired. The variable load apparatus,
however, prevents getting into the cylinder any more pressure than

cyhllxkr

Trout fmnsan

FIG. :I.-lTariable load compensattng dev~cefor alr brake equipment

the load of the car will justify. Referring to Fig. I r , as the car
comes to a stop, the opening of the doors energizes two magnet
valves through a door contactor. The upper one of these magnet
valves exhausts air so as to unlock the pressure limiting valve
mechanism, while the lower magnet valve admits air to the strut
cylinder. A rocker arm is mounted on the body bolster and the
admitting of air to the strut cylinder extends the push rod of this
cylinder so as to bring the foot plate of this rocker arm in contact
with the truck transom. At the same time the other end of the
push rod is moved out corresponding to the load on the car and
adjusts the limit valve to its proper position. This limiting valve
is in effect an adjustable feed valve and regulates the maximum
pressure to the brake cylinder. With a certain light weight car, 85
per cent braking effort is obtained with approximately 34-lb. brake

454

ELECTRIC RAILWAY HANDBOOK

cylinder pressure, and the same car with a load of 21,oooIb., which
corresponds to 150 passengers, requires ~ ~ - l bbrake
.
cylinder
pressure to give 85 per cent braking effort. The closing of the doors
de-energizes the magnet, locks the limiting valve and exhausts air
from the strut cylinder. The strut cylinder spring then p ~ l l sthe
cylinder so as to lift the foot plate from the truck transom, and thus
the vibration of the car when it is moving is not transmitted to
the variable load apparatus. All air to the brake cylinder passes
through the limiting valve. As this is adjusted a t each stop to the
load on the car, the maximum pressure going to the brake cylinder
is proportioned to the load on the car. I n other words, it is possible
to secure a cylinder pressure that wiIl give a proper braking
ratio with a loaded car and yet hold the, cylinder pressure for a n
empty car down to a value that will not cause wheel sliding.
A full appreciation of the great value of this improvement is
obtained when one considers the effect of a 40 per cent increase in
the total weight, by means of the live load, upon the stopping
distance of multiple unit trains as heretofore braked. The weight
of the empty train determines the maximum braking power that
can then be applied to it, and under the old conditions an increase of
40 per cent in weight meant lengthening all the stopping distances
obtainable with the empty train by about 40 per cent when the
same train was loaded. With the variable load compensating
attachment, all trains can be stopped in the same minimum distances, irrespective of loading. As noted below, the acceleration
m any given
also may be maintained a t a uniform m a x i m ~ ~ for
grade condition, irrespective of loading. These important factors,
together with the shorter spacing of block signals involved, permit a
remarkable increase in the train capacity on a rapid transit railway.
As used on the New York Municipal subway cars, the variable
load cam nsating device is tied in with the selective acceleration
feature o E h e multiple unit control, by means of an extra winding
on the limit switch. The modification of current input to the
motors is controlled from a switch operated in connection with the
variable load corn ensating mechanism.
Importance of L o p e r Relatipn between Air Pressure, Piston
Area and Leverage. The following 1s from a paper by Fred Heckler,
C.E.R.A., 1907:
Pressfrre. If more than 2 er cent braking power per pound of
cylinder pressure is attempted: a very high braking power lor light
cylinder pressures is obtained and, therefore, the cars cannot be
handled without shocks a t low speeds and either the range betwecn
maximum and minimum braking power obtainable must be very
narrow or else wheel sliding will result when the maximum power
is used.
Brake Piston Area. If the ratio of cylinder piston area to cylinder
pressure is excessive, i t means either a low leverage, with great
shoe movement or high leverage with low pressure, which gives
a very narrow range between maximum and minimum braking
power.
Leverage. If the leverage is too low, it means excessive air consumption and too much shoe movement; if too high (that is, brake

466

BRAKE RIGGING STANDARDS

cylinder too small for weight of car, and i t is here that the principles
governing brake design are violated most frequently), smooth and
accurate handling of the car or train becomes impossible and the
shoes are constantly grinding on the wheels, consuming energy,
wearing out the shoe and causing loss of time; or else piston travel
must be lengthened out, thus greatly increasing the air consumption,
lengthening the time of application and release, and reducing both
service and emergency braking power. Besides, the high leverage
makes necessary a frequent adjustment of piston travel or a constant and very rapid decrease of braking power will result. Furthermore, high leverage, if made a t truck levers, necessitates low hung
brake shoes, which, when suspended from a spring supported part
of the truck, results in great increase of piston travel and resultant
decrease of braking power with the loading of the car; this always
occurs a t a time when readjustment of piston travel is impracticable
and when instead of a decrease of braking power a n increase is
greatly to be desired. I n addition, the danger of the levers fouling
is greatly increased, particularly where the truck leverage ratio is
high, and very frequent and careful inspection is required or the
total loss of the brake may result.
DIAMETLROF BRAKECYLINDER,TOTAI.LEVERAGE
RATIOAND
WEICIITOF CAR
Dlametcr
of cyllnder
in lnchcl

I 1
Force of

piston at

50 Ib.

Total
leverage
rat0

U'elght of car wlth brake power

equal to
90
per cent

100

per cent

110

per cent

Position of Brake Shoe on Wheel. I f the brake shoe is hung

too far below the center of the wheel and the arts are badly worn,
the application of the brakes may cause the {rake shoe to form a
toggle joint with the wheel, stopping the rotation of the wheel.
This makes braking inefficient and brings about unnecessary brake
shoe and wheel wear. I n the above-mentioned paper by R. A. Parke
i t is stated that the center of the brake shoe should be about 3% in.
below the center of the wheel. For the standard location adopted
by the Central Electric Railway Association see page 457 (" Centcr
on brake shoe ").
Brake Cylinder, Lever and Rigging S*dards
of the Central
Electric Railway Association. The following recommendations of
the Standardization Committee were adopted by the Association,
1911:

(A) Revision of Standard Air Brake Cylinders,Levers and Brake

Riggzng. Recommended: T h a t the standard air brake practice
should be according to revised print No. C-2 (see 1;ig. 1 2 ) as follows:
I. All braking power to be based on 50 lb. cylinder pressure.

This in order to avoid confusion when stating percentages of brak-

456

ELECTRIC RAILWAY HANDBOOK

ing power that may be figured on different brake cylinder pressures;

e.g., loo per cent braking power on 60 lb. cylinder pressure may be
taken to mean greater than I W per cent on 50 lb. cylinder pressure,
and it this is alwavs referred to on a common base, confusion will
be avoided.
2. All interurban cars to be braked a t roo per cent of light weight
on rails and motor axles, and 90 per cent on non-motor and trailer
axles

CYLINDER LEVERS

ALL PINS TO BE LESS IN DIAMETER TEAN DRILLED EOLES.

FIG. 12.-Central Elec. Ry. Assn. standard brake cylinders and levers.

BRAKE PISTON TRAVEL

467

3. All city cars to be braked a t 8s per cent on motor axles ana

75 per cent on non-motor and trailer axles. Seventy-five per cent
on 50 lb. is practically the same as go per cent on 60 lb.
4. Brake pipe pressure to be 70 1b. per square inch with automatic
equipments.
5. Governor aajustment-85
and IOO 1b. for automatic e q u i p
ment. Governor adjustment-50
and 65 lb. for straight air
equipment.
6. The standing piston travel adjustment-4 in.
7. Total truck leverage to be 6 to I for long wheel base trucks
tor inside hung motors, and 9 t o r for short wheel base trucks with
outside hung motors.
8. A 1 2 to I maximum total leverage is permissible when brake
shoes are hung not more than 2 in. below the center of the wheel.
If brake shoes are hung lower than this i t will be necessary to reduce
the maximum total leverage accordingly, and if brake shoes should be
hung 5 or 6 in. below the center of the
wheel a total leverage of ro to r should
be the limit.
9. The standard M.C.B. recommendations for maximum stress in levers,
rods and pins to be adopted as follows:
(a) Maximumstress in levers-23,Ib. per square inch. (b) Maximum
stress in rods, except jaws-rs,ooo lb.
per square inch, no rod to be less than
to ~F1%
:3---Standard brake shoe
36 in, (6) Maximum stress in jaws (d)
~ q l t Cent.
l ~ nElec.
~ RY.Assn.
be Io,ooo Ib. per square inch.
Maximum shear on pins-ro,ooo Ib. per square inch, single shear.
(e) Diameter of pins to provide a bearing value not to exceed 23,000
lb. per square inch, projected area.
10. Safety valve adjustments-10
lb. above maximunl governor
setting.
( B ) Center on Brake Shoe. Recommended: That center of
brake-shoe be set 2 in. below center of wheel. (See Fig. 13.)
Piston Travel and Shoe Clearance. "Total leverage" represents
the total number of pounds of nominal brake shoe pressure that
is obtained for each pound of force exerted on the brake cylinder
push rod. To obtain any desired degree of braking force or percentage of braking power with a specified air pressure in the brake
cylinder, the greater the value of total leverage employed, the
smaller is the diameter of the brake cylinder required, but with a
corresponding increase in piston travel. A very high total leverage
is objectionable because a slight shoe wear produces a great variation in piston travel. The increase in piston travel due to the wear
of brake shoes is equal to the average amount of wear of the shoes
multiplied by the total leverage employed. Proper operation of
the automatic brake depends largely on the relative volumes of
the auxiliary reservoir and the brake cylinder and thus it is important to use a reasonably low total leverage and so maintain a more
nearly constant piston travel with ordinary shoe wear. I n steam

ELECTRIC RAILWAY HANDBOOK

railroad service, where cars receive attention a t only long intervals
and where the piston travel is not adjusted regularly, a lower total
leverage is necessary than in street railway service, where systematic inspection and maintenance generally can be relied upon.
Moreover, in street railway service the amount of air consumed
would be excessive if the piston travel were long and the leverage
low, as brake applications usually are very frequent in such service.
I t is customary to adhere as closely as possible to the piston travel
standards which have been fixed by the best practice, and which
must be carefully maintained in service in order to secure satisfactory results. Standard piston travel should always be the
average of the maximum and minimum values actually obtained.
A further reason for adhering to established standards of piston
travel and braking power is that it is most essential to provide for
clearance between brake shoes and wheels when the brake is
released. There is always a certain amount of lost motion in a
foundation brake rigging, and the use of either a very high total
leverage or an excessively short piston travel might result in dragging the brake shoes. If there were no lost motion between the
push rod and the shoes, the shoe clearance would be exactly equal
to the piston travel divided by the total leverage. However,
owing to the clearances in pin holes and to the spring and "give"
of the various parts of the rigging, the shoe clearance actually is
considerably less than this value. The piston travel obtained
when a car is standing is usuaIfy an inch or two Iess than thrat
obtained when a car is running, because vibration and motion of
the various parts rovide for their better adjustment. Consequently allowance sEould always be made for this inequality, as it
is the correct "running" iston travel that must be maintained. if
the brakes are applied wgile the car is in motion and held on until
and after the stop is made, the piston travel thus measured will
be the "running" value.
A very important consideration in the installation and adjustment of brake apparatus is the motion or travel of the various parts
concerned. The location and dimensions of all moving parts mu:t
be so chosen that each lever and rod will always be unobstructed in
its movements, even when the brake shoes are worn down completely and the brake cylinder piston has traveled to its maximum
limit, or otherwise the brake easily might be rendered totally
ineffective. An additional requirement of good practice is that
all levers, except the truck levers, should stand a t right angles to
their connecting rods when the brake is fully applied and the
piston travel has the standard running value. This applies articularly to the two cylinder levers The connecting rods wiE t h ~ l s
maintain their required a proximately parallel directions much
more nearly than would otEerwise be the case, for obviously when
a lever is moved through half of a specified angle on each side of a
perpendicular position with respect to the connecting rods, the
resulting deflection of these rods will be much less than when the
lever is moved through the same angle but not equally on both sides
of the perpendicular. In order to determine the travel or the position of any moving part of a brake rigging, i t is merely necessary

CLASP BRAKE
to remember that when a lever (or standard bellcrank) turns about
one of its three points as a center, the two other points travel in
circular paths, and the length of the path passed over by each of
these polnts is in direct proportion to its distance from the center
about which the lever turns. I n cases where a lever has no positively fixed point, as for example a cylinder lever of live truck lever,
the travel of a specified int can be determined by considering as
being fixed, in turn, e a c rof the two points the travel of which is
known.
Clasp Brake. The following is based upon comments by
S. W. Dudley, A.S.M.E, 1914, after the Pennsylvania-Westing,
house 1913 tests: The use of the clasp type of brake rigging eliminates unbalanced braking forces on the wheels and so avoids the
undesirable and troublesome journal and truck reactions that come
from the use of heavy braking pressures on but one side of the wheel.
This not only has an im rtant effect on freedom from journal
troubles, but it also ena& the wheel to follow freely vertical
inequalities of the track. Heavy braking pressures on but one
side of the wheel causes uneven wear on the journal bearings.
Although the clasp brake rigging will produce better stops than a
single shoe brake rigging equally well designed (other conditions
being equal), its advantage in this direction is of less importance
than in the improved truck, journal and shoe conditions. The
use of two shoes per wheel permits a design of rigging which will
allow flanged shoes to be used without danger of pinching flanges
and causing excessive flange wear or non-uniform brake forces which
result when flanged shoes are used with rigid beam connections.
The use of two shoes instead of one per wheel will result in a higher
coefficient of friction and less wear per unit of work done. A comparison of the values of mean coefficient of friction for standard
and for clasp brake conditions indicates a decided advantage for
the clasp brake throughout the entire range of braking powers.
The gain in favor of the clasp brake with slotted shoes amounts to
about 40 f f i c e n t a t a braking power of 180 per cent, and IOO per
cent a t a raking power of 40 per cent, a n average gain for the
whole range of braking powers of about 70 per cent
From a brake shoe standpoint the advantage of using two shoes
instead of one shoe per wheel may be summed up as follows: First,
the clasp brake is associated with but one-half the wheel load and
consequently has but one-half as much energy to absorb, second,
the clas brake shoe is working a t only one-half the shoe pressure
a t whicg the standard shoe must work under the same braking
power; third, the available work area for the same amount of energy
to be absorbed is double.
A possible source of disadvantage when using two shoes
wheel is that a warped or p r l y bearing shoe is subjected to
pressure tending to force ~t into a good contact with the wheel.
For this reason, though the available shoe area is doubled when
using clasp brakes, the actual amount of working metal throughout
the stop may be less than with a single shoe, which is less capable
of resisting the tendency of the heavier pressure to cause a better
fit of shoe to wheel.

460

ELECTRIC RAILWAY HANDROOK

With lain solid shoes the durability will be increased go per

cent u n g r clasp brake conditions as compared with that under
single shoe conditions. With plain slotted shoes the durability
will be increased 33 per cent under clasp brake conditions w compared with that under single shoe conditions.
Automatic Slack Adjuster. The function of the automatic slack
adjuster is to maintain the brake shoe travel automatically a t a practical minimum while keeping the piston travel a t a minimum and
nearly uniform. By doing this, a maximum efficiency of brakes
is approached, the necessity of inspection for the adjustment of
brakes is eliminated, the brakes on all cars operate as nearly alike
Adjuster

Feed Pipe

bl~nder
Ratchet Nut
Check Pawl

L P a e k i n e Leather
L E x p a n d i n g Ring

PIG. 14.-American automatic slack adjuster.

as may be and their operation is nearly the same throughout the

life of the brake shoes, the energy consumption and brake shoe
wear due to shoes dragging are eliminated and a minimum of air
and the accompanying minimum amount of energy and compressor
duty are demanded. Automatic slack adjusters are usually
located either a t the brake cylinder or on the truck. Figs. 14 and 15
show typical adjusters located at the brake cylinder and Fig. 16
shows a typical adjuster located on the truck where it takes the
place of the ordinary turnbuckle brake rod connecting the bottoms
of the live and dead levers.
American Automatic Slack Adjuster. (Fig. 14.) The adjuster
cylinder is connected by the adjuster feed pipe to a port in the

AUTOMATIC SLACK ADJUSTERS

461

brake cylinder wall. When the brake piston uncovers this port,
compressed air flows to the adjuster cylinder, thus operating its
piston which in turn actuates a pawl engaging the ratchet wheel
and nut on the screw. Release of brakes allows the spring in the

FIG. 15.-Creco automatic slack adjuster.

FIG. 16.-Anderson

automatic slack adjuster.

adjuster cylinder to return the piston in the latter to its normal

position, resulting in the rotation of the ratchet nut on the adjusting
screw. The inner end of the screw carries a jaw which acts as
a fulcrum for the fixed lever of the brake cylinder.

462

ELECTRIC RAILWAY HANDBOOK

Creco Automatic Slack Adjuster. (Fig. 15.) Pin A traveling in

slot B of bracket F passes beyond the straight line which is made
the length of the deslred piston travel and goes into that part of the
braket slot which is a t an angle of 45 deg. from the straight line
or the line of normal piston travel. The action of the pin in traveling through the angled portion of the slot is to carry the ratchet
pawl over the teeth of the ratchet D, and upon the return of the
pin back to the angle with the release of the brakes to turn the
barrel C, which is rigidly attached to the ratchet, upon the screw
portion of the push rod proper, thereby lengthening the distance
E from the lever pin hole in the jaw to the end of cylinder G.
Hand Brakes vs. Air Brakes. Hand brakes are satisfactory for
ordinary operation on light cars and should be supplied for emergency on all cars. On one man cars pneumatic devices are often
required in the interests of safety; in such cases air brakes are used
even though the cars are light in weight. However, as the weight
of a car increases the difficulty of control increases till it reaches
such a value that power is necessary for ordinary braking. Just
what weight should divide the hand brake class from the power
brake class has been a matter of much investigation, many testa
and disagreements. The difficulties of such tests are briefly
summed up by J. N. Dodd of the Public Service Commission,
First District, New York, as follows:
"In the course of this search it became apparent that tests do not
always furnish a reliable criterion of the relative value of different
brakes. They merely state the stopping distance of the car under
the particular conditions that obtalned a t the time of the test.
Most of the conditions that affect this distance vary widely.
Among these varying conditions may be mentioned the brake-shoe
adjustment, the weight of the car, the condition of the rail, and the
human element. Most of these factors vary widely also during the
course of the day, and each of them may change independently of
any of the others. Thus, on account of the wear of the brake shoes,
the brake adjustment may chanqe materially even in the course of a
single trip. The weight of the car is continually changing, owing
to the varying number of passengers. The condition of the rail
may alter entireiy in the course of a few seconds. This changeis
often such that a visual inspection of the rail fails to reveal its
quality. Thus, a wet rail may provide an ideal surface for stopping
a car quickly, or it may offer the reverse. I n the same way, though
not to the same extent, the distance in which a car may be stopped
on a dry rail varies according to whether the rail is clean or covered
with dust or dirt. For this reason it is impossible to be sure that
the rail conditions are the same in tests on two different brakes
or that the rail conditions a t the time of the test correctly represent
average rail conditions under which the car must operate throughout the year. The human element also is extremely variable. I n
actual service there are many strong motormen and many who are
physically weak, many who are intelligent and mentally alert and
many the reverse, many in fresh physical and mental condition and
others may be tired from a day's work. During most tests the
motorman usually knows that he is soon to receive the stopping

463

HAND BRAKES

signal and, knowing what he is expected to do, is intent upon doing

that thing in the most efficient manner. During such tests, also,
the streets are usually bare of t r a 5 c and the motorman's attention
is not distracted by other duties such as making up lost time, keeping a lookout t o pick up passengers and obeying the conductor's
signals. Usually a picked motorman is chosen, selected for his
general intelligence and interest in his work. The motorman is
generally in fresh physical condition and therefore in good mental
condition and to that extent capable of responding to any demands
made upon him. The results obtained in service a t the close of a
day's work might be entirely different from the results obtained in
m y series of tests. I t is impossible to devise any series of tests
under conditions which even approximate the average conditions
existing in actual service because i t is impossible t o tell what the
average of these varying factors may be."
After much consideration the Public Service Commission, First
District, New York, required cars weighing more than 25,100 Ib.
to be equipped with air brakes.

&.eulsnr

r;;l

B-Spar

Gemred Stall

0-Spur
Qsucd BtaRlorr

D-Worm

(tumd 8 t d

FIG.17.-Typical hand brake schemer

Typical Hand Brake Sehemes. The more common hand brakes

may be divided into four types, namely: gearless staff, s ur geared
sta8, spur geared staffless and worm geared stafi. ~ [ escheme
of each of these is shown by Fig. 17. Although the drum on which
the chain is wound may be of uniform diameter throughout its
length, many of the drums now used are conical or eccentric so
that the slack chain may be rapidly taken up over the portion of
large radius and the h a 1 tenslon applied to the chain a t small
radius, thereby giving a maximum tension when it is needed. As
the crank or handwheel is turned the tension is transmitted through
this chain, thence through a rod, thence through a chain to the
brake lever system (see Figs. 5, 9 and 10). I n calculating the
relation of tension in the hand brake rod to the pull applied to
the crank, the distance b, Fig. 17, is the effective radius of the
drum when the brakes are fully applied. This distance varies
according to the shape of the chain, its method of winding and
the travel of the brake shoes. Tests made by G. L. Fowler

464

ELECTRIC RAILWAY HANDBOOK

on the cars of the Brooklyn Heights Company, using dynamometers in the truck pull rods, showed the braking pressure a t
the wheels to vary as much as 40 per cent with a given pressure
applied to the brake handle on different applications on the same
car. This was found to be due entirelv to the manner in which
the chsinrolled on the brake staff.
Gearless Staff Hand Brake (A, Fig. 17). The gearless staff is
the most common type of hand brake used on light cars.
X (pull on handle).
b
Spur Geared Staff Hand Brake (B, Fig. 17). An increase in
mechanical advantage over that of the gearless stag type is secured
by the spur gear.
a
T
(tension in rod) = - X - X (pull 01; handle).
b
t
Spur Geared Staffless (C,Fig. I ? ) . The force applied at the
handwheel is transmitted through a spur gear to the winding drum.
The ratchet is normally held a t release by gravity and is set by a
foot lever. Most satisfactory results have been secured with gear
ratios of 14 :23 and r 2 :36.
a
T
(tension in rod) = - X - X (pull on handle).

(tension in rod) =

Worm Geared Staff (D, Fig. 17). h worm on the handwheel

shaft drives a gear on the staff. Ordinarily a compression spring

holds the worm in mesh with the gear. Release is obtained by
means of a grip eccentric lever which disengages the worm from
the gear. The slack chain is taken up by a torsion spring.
Hand Brake Maintenance. Brake staff defects are due principally to the staff binding and not releasing freely, and are oft-n
caused by the drawbar rest being displaced. Brake chain t r ~ ~ b l e s
are largely due to the hand brake binding and jamming between the
brake staff and sill. One of the most i m p r t a n t points in the transmission of braking power from the bra e haodie to the wheels is
the winding of the brake chain on the staff. A close link chain
should be used and care taken to have sufficient lead to the chain
to allow it to roll on the staff without one turn binding or running
upon another, There also should be sufficient release spring pressure to pull slack chain promptly from the staff, so that the chain
will wind on the staff directly below the eye bolt. Great care should
also be taken to see that the lead of the chain is such as to prevent
its winding above the eye bolt and jamming against the platform
and rendering the brakes inoperative. Another paint ta be
guarded against is that of the chain a t the rear end of the car
catching on the snow scrapers and thus preventing an application of
brakes. Inspection of the brake chain should guard against badly
worn links or eye bolts, or the possibility of nuts working off the
eye bolts.
Types of Air Brakes. Straight air brake system, recommended
for single car operation only.

STRAIGHT AIK BRAKE

465

Emergency straight air system, suitable for two car operation,

particularly when one is operated single most of the time and with
a trailer added during rush hours.
Automatic air brake system, suitable for trains of three or more
cars.
Combined straight and automatic air brake system, for locomotive service and for operation of single cars or trains of several cars.
Electropneumatic air brake system, for service similar to that of
the automatic system.
Straight Air Brake System. (Fig. 18.) The straight air brake
system consists essentially of a source of compressed air (either a
tank filled a t intervals from an air compressor, motor or axle driven,
located upon the car, or rarely from a compressor a t charging
stations); a reservoir which receives the air from the compressor
or charging tanks and in which the pressure is maintained prac-

PIG.18.-Straight air brake

system.

tically constant by means of a governor which automatically controls the operation of the compressor; a brake cylinder, the piston
of which is connected to a system of brake levers in such a manner
that when the piston is forced outward by air pressure the brakes
are applied; an operating valve mounted in each vestibule by
means of which the compressed air is either admitted to or released
from the brake cylinders; a pipe system connecting the above
parts, including cut-out valves, hose, and angle fittings between
cars. In order to prevent any possibility of accumulating an
excessive ressure, a safety valve designed to open a t loo Ib. per
square
is placed in the air su ply system. d set of pressure
gages is usually supplied with eacK complete equipment m order
that the motorman may observe the pressure in the reservoir and
remedy any defects in the governing ap
To operate the motormanJsvalve the E:;is
inserted when the
valve is in lap position where the slot in the body of the valve is
enlarged for this purpose (and to prevent its removal in any other
position). In t h ~ sposition the valve is set so that air can neither

i$

466

ELECTRIC RAILWAY HANDBOOK

pass into nor out of the brake cylinder. Moving the handle to
the left places the valve in full release, that is, connects the brake
cylinder to the atmosphere and allows the air which holds the brakes
applied to escape, when the spring which is opposed to the air
pressure restores the piston and releases the brakes. To partially
release the brakes, which is necessary in braking in order to prevent
shocks as the car stops, the handle is moved to the left and returned
to lap position. This reduces the pressure on the brake shoes, but
does not entirely release them. To apply the brakes for a service
stop the handle is moved to the service application position which
is to the right of the lap position. This connects the reservoir
with the brake cylinder thr0ugh.a small port in the valve. The
handle should be left in this posltion until a sufficient amount of
pressure is built up in the brake cylinder to give the retarding e5ect
desired, when the handle should be moved back to the lap position.
As the speed of the car is reduced, brake cylinder pressure may be
reduced In a series of steps by moving the handle from the lap to the
release position and then back to lap, repeating this movement
until the stop is reached. At the point of stoppinq, there should
be only sufficient air in the brake cylinder to prevent the car from
rolling. Better braking results will be obtained by making one
application as described than by admitting only a small amount of
alr to the brake cylinder a t the beginning of the stop and increasing
the pressure as the speed of the car is decreased. The latter method
usually results in rough stops and a waste of air. The practice of
applying and releasing the brakes several times during the stop
should be avoided. Moving the handle further to the right connects the reservoir to the brake cylinder through a large openine,
thus causing the cylinder to fill rapidly and quickly apply the
brakes with maximum
Sand usually is applied to the
tracks as soon as the PhT%si:
turned to emrrgency to avoid
skidding the wheels. In descending grades a light application of
the brakes may be made and the handle returned to lap. A sufficient length of time should be allowed for car to feel the effect of
the brakes before applying more pressure. If speed is higher than
desired a second light application should be made and o eration
repeated as often as necessary until the desired speed is ogtained,
or until the car has left the grade.
The straight air system of air brakes, although only recommended
for single car operation, may be used when operating with a trailer.
The equipment for trail cars consists of a brake cylinder and system
of levers similar to the ones on the motor car, a length of pipe running the entire length of the car and provided with hose couplings
and cut-out cocks for connections to the forward and rear cars. I n
connecting up trail cars, all the hose couplings must be thoroughly
united to insure that air will apply throughout the entire train. All
the cut-out cocks must be opened except those on the rear of the
last car, and the front of the first car, which must be closed.
So far as single car operation is concerned, the straight air brake
system is very satisfactory, as the desired flexibility in the matter of
graduations of applications and release of the brakes with due regard
to the passengers standing can readily be secured, and this apparatus

STRAIGHT AIR BRAKE

467

is usually so simple in construction that the motorman may become

familiar with its operation to srich an extent that accurate stops may
be secured with a minimum amount of instruction. Installation
and maintenance costs are low, on account of the small number of
parts. I n trains of considerable lengtb, however, the response of
the brakes on the rear cars is too slow, since all the air must pass
from the main reservoir on the front car through the opening in
the motorman's valve to the brake cylinden of each car. As the
addition of each car adds to the volume of the brake system, the
main reservoir on the first car must be considerably increased in
order that the pressure will not be reduced to such a n extent that
the brake application will not be sufficient and result in overrunning
the desired stopping place. These latter objections would not be
sufficient to prevent the use of this type of air brakes on short
trains of two or three cars were it not for the fact that a broken
hose connection or leaky train pipe renders the brakes on the whole
train inoperative.

PIG. 19.-Emergency

straight air brake system.

Emergency Straight Air Brake System. (Fig. 19.) The

emergenfy straight air brake differs from the straight air brake in
the detalls of the motorman's valve, the addition of an emergency
valve, and the use of three pipe lines--emergency, reservoir and the
straight air application and rclease pipes in place of the single pipe
line. I n the ordinary operation of single cars or short trains,
the emergency valve is seldom brought into play. I t is necessary,
however, to provide a short direct passage from the reservoir to the
brake cylinder in order to insure the quickest pssible action in time
of emergency and to provide some means of automatically braking
the rear cars should a break occur in the train line. At other times
when it is desired to make a service application or release, the air
is admitted or exhausted through the motorman's valve the same as

468

ELECTRIC RAILWAY HANDBOOK

in the straight air brake. I n Fig. 19the motorman's valve is in the

service position and air from the main reservoir passing through the
main reservoir cut off valve (which is always open except in case
the reservoir pipe should break, when it closes and protects against
loss of the brake in emergency) to the reservoir pipe is admitted to
the straight air application pipe a t the motorman's valve and reaches
the brake cylinder through the emergency valve. The emergency
valve normally connects the straight air application pipe to the
brake cylinder, and service applications and release are made by
increasing or decreasing the pressure in the train pipe, as with the
straight air system.
The emergency position of the motorman's valve should be
used only when i t ii necessary to stop the car within the shortest
possible distance to save life or avoid accident. I n this
the straight air application and release pipe connection is
i n the brake valve, while the air in the emergency pipe is exhausted
to the atmosphere. This reduces the pressure on the outer face
of the emergency valve piston so that main reservoir pressure
acting on the other face forces i t to the extreme outer position,
carrying with it the slide valve. This movement uncovers a port
in the slide valve bush so that the reservoir has direct communication with the brake cylinder, in consequence of which air flows
q i c k l y from the reservoir to the brake cylinder until both pressures
become equalized. I n the same way, should a hose burst or
uncouple, or pipe break, the resulting drop in emergency pipe
pressure will Insure an emergency application of the brakes as
described. When releasing after an emergency application, by
placing the brake valve handle in release position, the pressure
is restored in the emergency pipe. The equalized pressure on
either side of the emergency valve piston permits the spring to
return the piston to its normal position, releasing the brakes.
The release is designed to take place slowly after an emergency
application to secure additional protection and to discourage the
unnecessary use of the emergency position of the brake valve'
handle.
The emergency straight zir brake system for a trailer includes
the brake cylinder, emergency valve, auxiliary reservoir, conductor's
valve and the necessary piping. An auxiliary reservoir is used on
a non-motor trailer to furnish a n independent supply of air for
applying brakes on that car when a n emergency application is
made. I t is connected to the emergency valve in the same manner
as is the main reservoir supply pipe on a motor car. The auxiliary
reservoir is charged from the emergency pipe by way of the emergency valve. The emergency pipe is connected to the reservoir
pipe on the motor car by means of the motorman's valve for all
positions of this valve except the emergency position.
The conductor's valve, an additional safety device, is connected
to the emergency pipe and may be located a t any convenient point
in the car, if desired, with a cord attached to its handle and running
the length of the car. When this valve is opened, the air in the
emergency pipe flows directly through it to the atmosphere, setting
the brakes in emergency.

:::",E

AUTOMATIC AIR BRAKE

469

Automatic Air Brake System. (Fig. 20.) The automatic air

brake system is constructed so that the brake will be applied
automatically in case of any accident which permits air to escape
from the system. I t differs from the straight air brake in that i t
requires a decrease in the train line pressure to apply the brakes,
and a n increase in pressure to release them, whereas in the straight
air system air is admitted to the train line to apply the brakes and
exhausted to release them. T o accomplish this, there is added to
each car, in addition to the straight air system, an auxiliary reservoir, in which is stored a supply of compressed air sufficient to
operate the brakes on that car; a triple valve to which the train line,
auxiliary reservoir and brake cylinder are all connected and which
serves to control the flow of air ( I ) from the train line to the auxiliary reservoir when charging, (2) from the auxiliary reservoir to
Duplex

FIG. 20.-Automatic air brake system.

the brake cylinder when applying, and (3) from the brake cylinder
to the atmosphere when releasing the brakes; and duplex air gages
which indicate simultaneously the pressures in the main reservoir
in the train line.
This system is capable of a great many refinements which may
be added or omitted as requirements of a particular service may
prescribe. The main points of difference between particular automatic air brake equipments will generally be found in the details
of the triple valves, and the addition of pressure maintaining and
reducing valves, which are essential in certain classes of grade work
in order to prevent brakes leaking off. For the sake of simplicity
these particulars have been omitted from this consideration.
Plain Triple Valve. Fig. 2 1 is a diagrammatic sketch of the
plain triple valve which is used only on com aratively short trains
or five cars or less. This figure represents t i e ttiple valve in the
release position. The brake cylinder is in communication with
the atmosphere by way of pipe B through ports 6 and 7, connected
together by valve 3, and the brakes are released. Air from the
train line enters the ttiple valve a t L, and the pressure holds the

470

ELECTRIC RAILWAY HANDROOK

ptlo.

iston I in the
shown; the by-pass at the top allows nu to
Peak into cham r R and enter the auxiliary reservoir, charging it
to train line pressure. This position of the triple valve prevents
air from flowing to the brake cylinder from the auxiliary reservoir,
as valve 3 closes port 8. When the pressure in the brake pipe is
reduced below that in the auxiliary reservoir, either intentionally by
manipulation of the motorman's valve or accidentally as in the
case of a broken hose or pipe, the greater pressure from theauxiliary reservoir moves piston 2 of the triple valve to the left, which
closes the by-pass just above the piston, moves valve 3 to the left,
closes port 6 so that communication from the brake cylinder to the
atmosphere is cut off, opens$rf 8 between the auxiliary reservoir
and the brake cylinder, an applies the brakes. T o release the
brakes, pressure is restored in the train line from the main reservoir,
From

Train
Lne

PIG. 21.-Diagrammatic sketch of triple valvc and conncctions.

thus increasing the pressure a t L above that remaining in the auxiliary reservoir and chamber R of the triple valve. Piston 2 of the
triple valve then moves back to the right as shown in the figure,
releasing the brakes and recharging the auxiliary reservoir a s
previously described.
A graduated release of the brakes may be obtained with this type
of valve by piping the exhaust from the triple valve to the motorman's valve where a movement of the valve handle will release the
air the same as in the straight air brake.
I n practice, the service application requires only a slight reduction
in train line pressure, and reference to Fig. 2 2 will show how the
graduufing valve functions to secure this end. I n a similar manner
to the previous simple explanation, a reduction in the train line
Fessure lowers the pressure in chamber h of the triple valve, and the
igher auxiliary reservoir pressure moves piston j to the left, thereby
opening communication between chamber h and the auxiliary
reservoir through feed groove i. Attached to the piston stem is a
pin valve 7, called the graduating & h e , which when seated, closes
communication between port w leading from chamber m to the

PLAIN TRIPLE VALVE

471

graduating valve seat in the slide valve and the service Ijort zleading
from the graduating valve seat to the face of the slide valve.
The first movement of the triple valve piston to the left unseats the
graduating valve 7 so that air in chamber m, entering port w , flows
to the service port z. There is a small amount of clearance between
the slide valve 6 and the collar on the end of the triple valve piston
stem, so that the first movement of the piston, which closes the feed
groove i and opens the graduating valve 7, does not move the slide
valve, but brings the collar on the stem against the end of the
valve. Further movement of the piston then causes the slide valve
t o move until it has closed communication between brake cylinder
port r and exhaust port p, and opened port r to the auxiliary reser,voir through ports z and w. The piston then comes into contact
with the ~ r a d u a t i n gstem 8. and the resistance of the graduating
s ring combined w&h
tEe reduction in auxiliarv reservoir nressurethen taking ilace
prevents further
m o v e m e n t of the
parts. The valve is
then in service position and air from the
auxiliary r e s e r v o i r
flows through the service w r t to the brake
cylinher, applying the
brakes.
While t h e
-'4..
pressure in the brake
FIG. 22.-Plain triple valve.
cylinder rises, that in
the auxiliary reservoir fallsand (the brake pipe reduction being 101b.)
tends to become lower than that in the trainline. Assoon, however,
as the pressure on the a u ~ i l i a r yreservoir side of the tri le valve
piston falls slightly below that on the train line side, tge higher
pressure causes the piston to move back toward its former (release)
position, until the graduating valve is seated, closing communication betwen ports np and z. This prevents further flow of air from
the auxiliary reservoir, the pressure in which is then practically
equal to that in the train line, and a t thesame timeprevents further
movement of the triple valve piston toward release position, because
the slightly higher pressure on the brake pipe side of the piston,
which was able to move the piston and graduating valve alone, is
not sufficient to move the slide valve also. Assuming that there
is no leakage, the brake ipe and auxiliary reservoir pressures will
remain balanced and the L a k e cylinder pressure held constant until
the brake pipe pressure is further reduced, in order to apply the
brakes harder; or increased,inorder to release the brakes. A further
reduction in pressure on the train line side of the triple valve piston
helow that on the auxiliary reservoir side causcs the piston and its
attached graduating valve to move as described for the first service
application of the brakes. The slide valve, however, is already
in service position, consequently as soon as the graduating valve is
~~

ELECTRIC RAILWAY HANDBOOK

opened, air from the auxiliary reservoir flows to the brake cylinder
and increases the pressure therein, thus increasing the pressure of
the brake shoes against the wheels. If the brake pipe reduction is
continued indefinitely, the auxiliary reservoir pressure will continue
to fall and the brake cylinder pressure rise until they "equalize."
This occurs a t about 50 Ib. cylinder pressure, when carrying 70
lb. train line pressure. After the pressures in the auxiliary reservoir
and brake cylinder have "equalized" in this manner, air ceases to
flow out of the reservoir and into the cylinder, because there is no
longer any difference of pressure to cause a flow. Consequently,
when the train line pressure is reduced below the "point of equalization," the brake cylinder pressure cannot rise above the "equalizing
int " even though the train line may be reduced far below 50 lb.
or this reason, therefore, nothing is gained by reducing the train
line pressure below the "equalizing point" as explained above.
Moreover, it is a needless waste of air and interferes with the proper
release of the brakes
An emergency application by means of the motorman's valve or a
broken hose or pipe causes a sudden and rapid drop in the train line
pressure. The pressure in chamber h, on the train line side of the
triple valve piston, is reduced at a rapid rate, and the resulting
difference of pressure is sufficient to move the piston and slide valve
immediately to the extreme left and uncover the brake cylinder port
r so that air from the auxiliary reservoir flows past the end of the
slide valve directly through port r into the brake cylinder until
the brake cylinder and auxiliary reservoir pressures are equalized.
The ressure obtained in the brake cylinder is no higher than when
a fulr service application is made, but the maximum pressure is
obtained more iiiickly.
hick-action Trifile Valve. The quick-action triple valve shown
in Fig. 23 is the same as the plain triple valve withadditional parts
and minor structural changes to permit the inclusion of these
additional parts and retain the features of the plain triple valve.
The preceding description of the plain triple valve, except for the
emergency application, applies equally to the quick-action triple
valve. When the piston and slide valve of the quick-action triple
valve move to emergency position, as described for the plain
triple valve, the ports in the slide valve registers with port r in the
seat, allowing air to flow from the auxiliary reservoir to the brake
cylinder. Port s is smaII, however, and in this position the slide
valve also opens port t in its seat, allowing air to flow from chamber
m through port I to the chamber above the emergency piston 8.
The other side of the emergency piston 8 is connected to the brake
cylinder, in which there is no air pressure, consequently the emergency valve is forced downward, pushing the emergency valve 10
from its seat and allowing the air in chamber Y above the check
valve 1 5 to flow past the emergency valve 10 to chamber X and the
brake cylinder. Train line air in a, below the check valve 15,
then raises the check valve against the resistance of its spring 1 2
and also flows to the brake cylinder through the passages mentioned.
During an emergency application, therefore, the quick-action triple
valve supplies air to the brake cylinder from the train line as well

QUICK-ACTION T R I P L E VALVE

473

as from the auxiliary reservoir. Port s is small, so as to restrict

the flow of air from the auxiliary reservoir to the brake cylinder and
thus allow as much air as possible to enter the brake cylinder from
the train line. Approximately 60 lb brake cylinder pressure is,
therefore, obtained on emergency applications, the air from the train
line increasing the cylinder pressure about 2 0 per cent above the
maximum obtainable with a full service application. Not only does
the air vented from the train line give a higher brake cylinder
pressure, but i t causes a local drop in train line pressure a t the triple
valve which causes the next triple valve to apply in "quick-action,"
and i t the next, thus transmitting the quick action from triple
valve to tri le valve serially throughout the train in a very short
time, with t i e result that all the brakes in the train are applied in a

PIG. 23.-Quick

aclion triple

valve.

fraction of the time which would be required if all the valves were
plain triple valves and all the train line reduction had to be made a t
the brake valve. The release after a n emergency application is
made the same as after a service application, except that i t requires
a longer time, the train line having to be recharged from zero to
slightly above the pressure in the auxiliary reservoirs before the
triple valve pistons can move to release position. The quickaction triple valve is designed to be used on freight trains of considerable length, its function being to apply the brakes on the rear cars
in emergency so quickly that the taking up of slack is avoided.
Combined Straight and Automatic Air Brake. This system
includes two sets of motorman's valves, operated by the same
handle, for the control of each system. The straight air brake,
operating with pressure between 55 and 70 Ib. per sq. in., applies
and releases the brakes on the front car independently of the brakes
on the other cars. The automatic brakes operate with air pressure
from roo to rrolb. per sq. in., and a p ly the brakes on the remainder
of the train independently of the grakes on the front car. The

ELECTRIC RAILWAY HANDBOOK

chief advantage of such a n arrangement is the ssibility of holding the brakes on the locomotive applied while tff: train brakes are
released for the purpose of recharging the auxiliary reservoirs.
Electropneumatic Brake System. I n the electropneumatic
brake, the ordinary features of the automatic air brake are retained,
but the application and release of the air pressure is governed by
electromagnetic control in a manner somewhat similar to that of the
electropneumatic control of the motor circuits in multiple unit
control operation. A special form of triple valve is used in which
the admission of air to the brake cylinders is governed by electromagnetic valves. By proper combinations of electric circuits, the
brake cylinder pressure may be built up to the maximum, may be
held in the cylinders, or may be wholly or artially exhausted. The
motorman's valve is composed of two &tinct parts, the electric
placed above the pneumatic part. The electric parts consist of a
rotary drum with contacts attached to the same shaft as the
rotary valve and with corresponding fingers on the body of the valve.
The position of the handle is identical for electric and pneumatic
release, service and emergency. With the electric features operative, the service magnet valves admit air from the auxiliary reservoir to the brake cylinder on each individual car a t the desired
rate, and a t the same time the train line pressure is reduced by the
motorman's valve so that if the electnc control is inoperative,
the pneumatic control will operate automatically to apply the
brake. If any of the service application magnets in the train are
inoperative for any reason, therefore, the brake will apply pneumatically on that particular car.
I n comparative tests of service stops from a n initial speed of 40
miles r hour, an 8-car rapid transit train with pneumatically
controEd brake was stopped in 40 seconds in a distance of 1290 ft.
I n this operation a graduated action was obtained by increasing
the train line pressure to release the brakes and then partially
reapplying them. The electrically controlled brakes brought a
10-car train to rest after 2 0 seconds in a distance of 700 ft. I n the
operation of the electrically controlled brake, a much more effective
graduation of the release was obtained. Due to this feature the
maximum braking effort can be increased to a value which would
be dangerous for the ordinary automatic brake. The final pressure
a t the end of the stop was less with electropneumatic control than
with the other type of control. The emergency stops in the same
tests showed the advantage of simultaneous action of the brakes
on all the cars to a greater degree than the service application tests.
The time was 2 2 seconds and distance about 625 ft. for automatic
air braking of the 8-car train, and I I seconds and 350 ft. for electro~neumaticbraking of the 10-car train.
The electropneumatic brake system adds to the pneumaticallyoperated brake certain advantageous features otherwise impossible
of attainment, namely: simultaneous and uniform response of all
the brakes in the train, which means the ability to obtain the desired
results with the least skill and experience, regardless of the length
of train; double protection against delays to traffic due to brake
failure, since the pneumatic brake is always in reserve ready for

ELECTROPNEUMATIC BRAKE

475

use, if required; maximum efficiency and safety due to simu~raneous

operation of all the brakes in the train, in both service and emergency application, and the ideal flexibility of manipulation;
economy in air consumption, and maintenance of brake cylinder
pressure a t will. The shorter service stop thereby made possible
means that power may be shut off sooner and the train allowed to
coast for a considerably longer time before applying the brakes and
making the stop a t the same point. Thus, for the same power
consumption, the electropneumatic brake makes i t possible to
maintain higher average speeds, shorter schedules and a n increased
traffic capacity with the same number of cars; or the same traffic
ca acity with fewer cars; or enables the same average speeds,
scKedu~esand capacity of road to be maintained with the ex
ture of less power. I t should be borne in mind that the e E : j ~
pneumatic control of the brakes is a control only and that i t must
depend for its fundamental safety and protective features upon the
pneumatic brake with which i t is associated.
For single cars or short trains, electric control of brakes, although
ideal and entirely practicable, has not appeared to be necessary or
justified by the requirements of such service. I t adds nothing to
the effectiveness of the individual brake, its only function being to
save time in applying the brakes by eliminating the serial time
element inseparable from a n impulse transmitted pneumatically
from car to car. Where trains are relatively short, therefore, pneumatic transmission is so nearly instantaneous that the sto IS
not materially lengthened thereby, nor is there sufficient slac& or
time element to cause objectionable shocks from this cause. When
the carrying capacity of a given trackage is taxed to its limits,
every means available for increasing the assenger miles per car
becomes of prime importance, and the vai)ue of electrical control
of brakes becomes apparent. This does not add directly to the
effectiveness of the brake on any one car, but eliminates the serial
application effect on a train of several cars. I n doing this, however,
i t accomplishes two results which are of consequence in long train
service, namely, the elimination of slack action between cars and the
saving of the time and distance through which the train would
otherwise run while the pneumatic brake action was being serially
transmitted from the head to the rear of the train. I t is quite clear
that these features become of greatest value where trains are long
and run on close headway with frequent stops and the traffic
capacity of the road severely taxed. For these reasons the electric
control has been found especially desirable for such service as in
the subway systems of Philadelphia and New York and the
elevated of ~ o s t o n .
Safety Devices of the Safety Car. Features that may be included
on the safety car to make operation safe, proper and easy, are:
(I) The car cannot be started with the doors open. ( 2 ) Opening
the doors causes a brake application sufficient to stop car. (3)
Unintentional removal of the operator's hand from controller handle
causes emergency brake application. (4) Brakes cannot "leak off"
while the car is standing with the doors open. ( 5 ) Automatic
emergency application results if either of the main pipes is ruptured.

4 76

ELECTRIC HAILW4Y HANDBOOK

BRAKE INSPECTION

477

(6) Power is cut off, sand is applied, and doors become hand operated whenever an emergency application is caused by releasing the
controller handle. (7) If service application "leaks off" while
changing ends, an emergency application results. (8) Rupture of
platform piping causes an automatic emergency application, and
the damaged iping is cut off so as to prevent exhaustion of the main
reservar.
To release brakes after a stop, the controller handle
must be ressed down. ( 1 0 ) Ends cannot be changed without
making I!rake application or an automatic emergency application
results. (XI) Sand is automaticauy applied in an emergency.
( 1 3 ) No increase in the number of manipulative handles.
(13)
Doors are air operated. (14) Sanding is accomplished by air
without the use of a special operating handle. (15) Intentional
and temporary release of the control handle may be made without
emergency application resulting. Fig. 2 4 shows a typical layout
of the various air devices and their connections to produce the results
enumerated above. Many cars are in operation with safety
devices which do not include all those here shown, or which include
some such devices in difIerent form. Reference may be had to
Eltdtic Railzvay Jmmal, Voi. 55, p. 788 et seq., for detailed
descri tion of the various devices ancf their operation.
B&
Inspeetion (A.E.R.E.A. Approved Practice). Start the
air ump and allow it to pump to its maximum capacity; see that
braEe valve handle is in release position, and where automatic air
is used see that gage hands show a difference between train line and
auxiliary of 20 Ib. If they do not, the governors need to be reset.
Ap ly brake to show reduction of 40 Ib.; place brake valve handle
to
position; see that air gage operates properly and that no leaks
are in or around the brake valves or ipes leading thereto; examine
all ipes, reservoirs, triple valves, cy!nders,
etc ,while brake is set,
anBsee tbat Bone are leaking and that brake does not release while
the brake valve handle is in lap position. If the cylinder piston
has a travel of more than 5 in. an adjustment of brakes is necessary.
t all shoes and see that they are in alinement with the wheel
MDe are broken, and renew those that will not give s u a cient wear until the next inspection. In renewing brake shoes
put shoes of the same thickness on opposite wheels, be they either
old or new. See that all brake shoe keys are in place and that none
are lost or broken; examine shoe heads and see that none are lost or
broken, and that all pins, bolts, etc., that hold heads to the beam
or tmck levers are not unduly worn; that all bolts, cotter pins, nuts,
e t c , are in good condition. Examine brake beams and see that
none are cracked, broken or bent, and that all bolts, pins and
holes are not unduly worn, and that aU cotter pins and nuts are in
place. Examine all hangers and pins connectmg brake shoe head
and beam to truck and see that all are in good condition, and tbat
none of the pins and hangers are unduly worn so as to cause brakes
to grab or chatter; spa[
attention must be paid to all cotter
pins in all parts of t e brake rigging. Examine all turn-buckles
and see that none of the threads are stripped and that all adjusting
and jam nuts are tight and in their proper places. Adjust brakes
so that cylinder piston will not travel more than 4 or 5 in. When

(8

z=t

478

ELECTRIC RAILWAY HANDBOOK

brake is in release, see that none of the shoes bind the wheels, t b t
release springs operate properly so that brakes will be free when
released. Examine all pull rods for cracks or flaws, lubricate all
pins in pull rods, levers and slides; set hand brake and see that it is
in good condition; see that brake staff and chain are not unduly
worn; that rod, pins, etc., are in good condition. Where slack
adjuster is used, see that i t is placed to its minimum of travel
before any adjustment of brakes; see that i t is operating properly
and that it has not traveled to a maximum position, leaving the
correct piston travel. Drain all reservoirs daily.
Maintenance of Motoxman's Valve. The seats of the rotary or
motorman's brake valve are subject to the collection of dirt largely
because of the passage across them of air which is more or less ladrn
with dust. When this collection causes the valve to work hard or
the valve seat becomes badly scored, an attempt to grind the surfaces with emery should not be made. Grinding with emery
generally causes leaks between the ports and a n attempt to grind ovt
cuts or scores will usually make the valve worse. When the condition of the valve becomes so bad that the valve cannot be kept in
service i t should be machined and scraped by hand to a perfectly
flat surface, using a face plate to locate the high spots.
Brake valves should be lubricated a t regular car inspection periods.
To oil a brake valve, it is necessary first to exhaust the air from the
valve; the oil should then be applied through the oil plugs, the valve
stem pushed down a few times, and the valve operated to work
the oil onto the various surfaces. Lost motion or play between
the handle and stem prevents the proper registration of parts and
should be eliminated.
Emergency valves, feed valves and triple valves should be completely disassembled and thoroughly cleaned a t regular overhaulidg
periods. The only part of the feed valve requiring lubricant is the
slide valve which should be lubricated with dry graphite. I n the
emergency and the triple valve, lubricate the slide valve with dry
graphite and the piston bushing with a drop or two of oil.
Inspection of Air Compressors (A.E.R.E.A. Approved Practice).
The frequency of inspection necessary for an air compressor depends
upon the service required of the compressor. The duties of air
compressors on city and interurban cars vary greatly. For instance,
a car may be equipped with air doors, electropneumatic control
and air brakes. The average number of stops of a car of this character in city service may be every 1500 ft. A compressor under
these conditions would necessarily have to be overhauled more
often than one which makes stops two or three miles apart, has
no air doors or electropneumatic control. The inspection period
for the former class of service should be 600 miles and that for
the latter class should be 1 2 0 0 miles. Air compressors should be
inspected as follows: Oil plug should be removed and oil added to
replace what has been lost in service each inspection day. Carbon
brushes removed and inspected each inspection day. Brush holder
tension inspected each inspection day. Brush holder wiped off
each inspection day. End of commutator wiped off each inspection
day. Hair should be taken out of hair strainer and strainer washed

AIR COMPRESSOR MAINTENANCE

479

in gasoline every thirtieth inspection day. Valves should be taken

out and cleaned in gasoline each thirtieth inspection day. Exterior
of pump should be wiped off each inspection day. Compressors
should be thoroughly blown out with compressed air each tenth
ins ction day.
Compressor Tests (A.E.R.E.A. Approved Practice). With
all air reservoirs empty and brake valve on release position, start
air compressor and note length of time taken in pumping up, and
pressure a t cutting-out point. A test for leakage should then be
made, leaving apparatus in same condition as above, and noting
number of pounds drop in one minute.
The Connecticut Company overhauls its rolling stock on a ~oo,ooo
mile basis. The overhauling of car equipment includes air compressors, which are tested on the cars before and after the overhauling.
For convenience in testing, a tank of exactly 5 cu. ft. capacity has
been mounted on a truck. This is provided with a gage and a
valve for bleeding off the air pressure. Compressors are tested for
the time necessary to pump up to a given pressure and also for the
number of revolutions. After connecting the tank to the compressor, the test consists of pumping up the pressure in the stand tank
to go lb. This is then reduced to 40 Ib. and again pumped up.
When the pressure reaches 60 Ib., a revolution counter is applied
to the armature shaft and the number of revolutions required to
pump from 60 Ib. to go Ib. is counted. A table has been prepared
from the manufacturer's rating of the compressor, to show the
number of revolutions under proper conditions. A glance then
will show whether there are abnormal conditions that require attention. After overhauling, compressors must conform to the requirements within certain definite limitations in order to be passed for
service.
Lubrication of Air Compressor (A.E.R.E.A. Approved Practice).
Air compressors should be lubricated weekly. Refore removing
plugs, wipe all dirt and surplus oil from around the openings, then
remove plugs and do not fill to an extent to overflow, thereby
allowing the oil to collect with the dust and dirt around the pump
frame and case. See that the plug threads are in good condition
and oil-tight when closed.
Overhauling Air Compressor (A. E.R .E.A. Approved Practice).
With a compressor designed for sufficient capacity for the service
i t is to perform, an overhauling for the former class of service outlined above (Inspection of Air Compressors) should be made every
60,000 miles, and for the latter class of service every 120,ooo miles.
I n overhauling, the compressor should be taken from the car to a
bench where the armature should be removed, oil drained from
crank case and from all bearings outside of crank case, and bearings
thoroughly washed out with gasoline. Crank shaft and connecting
rods should have slack taken up on them or bearings re-babbitted or
relined with bronze bearings, as the case may be, if wear is excessive.
Head should be taken from the compressor, ports scraped out, hair
removed from hair strainer and washed in gasoline, and valve
and valve seat should be re-ground. Armature should be blown
out, cleaned up and breakdown test applied. Also commutator

480

ELECTRIC R A Z W A Y HANDBOOK

trued up. Mica retaining rings should be painted with insulating

paint. Armature painted with a coat of oil-proof insulating
. Brush holders should be cleaned in gasoline and overauled, replacing worn t ~ p sand shunts that have broken strands.
F~eldcolls should be taken out and insulation carefully looked
over, and replaced if necessary. Inside of the motor shell
should be carefully cleaned with gasoline to get oil and dirt out
of shell The impregnation of pump field roils is recommended
as tending to eliminate field troubles and greatly prolong the
11fe of the fields The piston should be removed from pump, and
rlngs removed from groove and carefully cleaned; also, groove
should be carefully scraped out Springs from piston rings should
he tested to make sure that they have not lost their tension.
If, when pump is re-assembled and started to work under pressure
for 5 minutes, it be then disconnected and allowed to run
free and oil is discharged from the outlet, it is a n indication that
the rings are not tight enough in the cylinder, and springs of
greater tension should be put in the rings The insulators placed
between compressors and alr piping on car should be taken from the
car and cleaned and given a breakdown test of 1000volts. The oil
whirh was removed from the compressor when it was removed for
inspection should be run through a filter and returned to compressor,
enough oil being added to take the place of the dirt, etc., which i t
contained when removed In this connection care should be taken
t o use an oil which does not contain asphalt or which carbonizes
when the pump is given hard usage
Storage Air Brake System. I n the storage air brake system the
service reservoir is supplied with air from storage reservoirs (generally two) carried on the car. These storage reservoirs are charged
with air a t high pressure from compressor stations along the line.
The air is passed through a pressure-reducing valve between the car
storage reservoir and the service reservoir, and thus its pressure
is reduced to a value proper for service in the braking system.
Whether or not the storage system should be used rather than
the individual car compressor for a given service must be decided
from local conditions. The storage system may in some cases be
cheaper than the individual car compressor system, but the many
advantages in having the car self contained, as by the use of the
individual car compressor, make the use of the latter systemalmost
universal. Moreover, air operated auxiliaries demand air in addition to that for the brakes and such demands will, in most cases,
make the employment of the storage system impracticable.
Size of Air Compressor Required for a Given Service (A.E.R.
E.A Approved Practice). An air compressor should be of a size
such that it will not be required to operate more than one-third the
time. I n connection with the above, due consideration should be
given the air requirements of air operated doors, sanders and other
auxiliary equipment.
Air Used and Electrical Energy Required to Drive Compressor.
The data in the following table were taken from the Report of the
Electric Railway Test Commission, 1906, and were based on tests

rint

4Rl

AIR COMPRESSORS

made by the Commission on double truck city cars in operatton in

St. Louis

[
Average number of stops per mrle
Schedule speed of car, mlles per hour
Maxlmum speed of car (approxrmate). mlles
per hour
Average volume of free arr used, cubrc feet per
car per stop
Average volume of free alr used. cublc feet per
ton per stop
Electrical energy for compressrngair. watt-hour
pet car per stop
Electrrcal energy for compressing alr, watthour per ton per stop

Dry track

Wet track

4 1
9 5
16 o
I

68

o 076
6 74
0 306

The amount of air required per stop ior interurban car operation,
while greater, is not in proportion to the speed from which the stop
is made. As a matter o fact, it should not be greater merely
on account of the higher speed, as after the brakes are applied,
only enough air is needed to make up for leakage. As the duration
of the braking period may be greater for the interurban car, the
possibility of leakage is correspondingly increased The air
actually required by the brake in making the stop often is a small
percentage of the actual amount used, due to improper brake
valve manipulation, leakage of piping and brake cylinders, air
operated devices such as bell ringers, door engines, whistles, etc.
Capacity of Air Reservoirs. The air reservoirs should have a
capacity sufficient to supply air for three or four applications without reducing the pressure more than 1 2 or 15 Ib. Otherwise every
ordinary application of the brake will throw the compressor into
action, thus keeping the latter in a constant state of starting and
stopping, and causing unnecessary wear to both compressor and
governor The Westinghouse Traction Brake Company recommends the following sizes of reservoirs:
for 8 in. brake cylinders, 16 in. X 48 in. reservoirs;
for lo in. brake cylinders, 16 in. X 60 in. reservoirs;
for 1 2 in. brake cylinders, 16 in. X 7 2 in. reservoirs.
The lengths given above are overall.
General Characteristics of a Good Air Compressor. An air
compressor should be reliable, of light weight and compact construction, should be protected from dirt and water, should be so
constructed as to be easily inspected, lubricated, overhauled and
re ired, should run with the least possible amount of noise and
vication, should have high efficiency of operation and should have
a low cost of maintenance.
General Types of Air Compressors. The independent motordriven air compressor most commonly used for electric railway service may be considered to be of two general parts, namely, the motor
and the compressor. If the motor and compressor are geared
together, the machine is known as the geared type compressor. If

ELECTRIC RAILWAY H.\NDROOK

the motor shaft is direct connected to the compressor connecting
rod, the machine is known as the gearless type compressor. Because
of the absence of reduction gearing a gearless type air compressor
differs mainly from the gear type of the same capacity in that
the plston speed (revolutions per unit time) of the former is greater
than that of the latter, consequently the piston displacement of
the former is the less. Compared with a gear type air compressor
of the same capacity, a gearless type compressor weighs about onehalf, occupies about one-half the space and has a very low overall
height. The stresses in the gearless type are much lower, the
maximum transverse thrust on the crank being only about onefifth as much as that in the geared type.
Installation of Air Piping to Prevent Freezing. Great precaution
must be taken in the installation of air piping between the cornpressor and the main reservoirs to prevent freezing of thecondensed
moisture. This is especially true with pneumatically operated
apparatus which employs valves with small openings. Freezing
a t very low temperatures is not so troublesome as a t a little below
3 2 deg. F., because a t lower temperatures most of the moisture
has been frozen out of the atmosphere. The installation of the air
piping should be such that the maximum amount of moisture is
retained in the main reservoir. No pockets should exist where
moisture is likely to be retained, as freezing will occur whenever a
small body of water collects in the piping. The pipe between the
compressor and the main reservoir, as well as the pipe between
the two main reservoirs, should be a t least 25 ft long, and when the
length of car does not permit a straight run, the pipe should be made
up in a series of return bends, or with several pipes as multiplep~ths
connected into common headers. For a given length of piplng of
the same diameter in the two systems, the single path will require
a greater velocity and hence a greater loss in pressure for the
delivery of a given quantity of air. This means that, for a given
main reservoir, the compressor must pump against a greater
pressure with the single path system, because of the greater friction.
This higher pressure is accompanied by a higher temperature and
consequently, even though the radiating surface of both arrarlgements of piping are the same, the air from the single path of pipe,
in view of its higher initial temperature, must necessarily be warmer
when reaching the main reservoir than with the manifold arrangement. The extent to which this difference exists depends upon the
size of pipe and upon the quantity of air delivered in a given unit of
time.
The Cleveland Railway is using air coolers made from a 36 in.
section of standard 5 in. pipe, asshown by Fig. 25. I t isinserted in
the air line between the compressor and the main reservoir to cool
the air sufficiently in passage so that all excess moisture will be
dropped. The warm air, on being received from the compreswr,
enters the cooler and is deflected by a bafle so that it follows
along in the space between the wall of the 5 in. pipe and another
internal 1% in. pipe concentric with it. I n this way the air is
kept in contact w ~ t hthe larger pipe where the cooling effect is
obtained. The air leaves through twelve % in. holes in the further

AIR COOLING 1)EVlCES

483

end of the inner i p . Accumulated moisture is drawn out through

a drain cock in tRe bottom of the cooler. The only special castings
required are the internal baffle and the cap on the reservoir end
of the cooler. Both of t h e parts are threaded to receive the
inner r % in. pipe.
In connecting the feed pipe to the reservoir, particular attention
should be given to make sure that the connections do not give a
reduction of pipe area a t any point, as a change in area increases
the possibility of freezing. All piping from the reservoir to the
various pieces of apparatus should be arranged to drain back into
the reservoir as far as poss~ble,and when this is impossible i t should
drain away from the apparatus.
The best place to mount the intake is on the roof of the car, as i t
is then possible to obtain cool, clear air. With the intake mounted
inside of the car, a greater percentage of moisture is obtained, which
is always to be avoided.

'

Air Brake Hose. The bursting test of air brake hose is made by
subjecting it to a hydraulic pressure of zoo lb. per sq in ,and under
this pressure the hose should not show any signs of leakage nor
develop any defects, and the maximum expansion on the circumference should not exceed the following:
Nom~nalsize of hose
Mlnrmum expanston

. .

... .

2 m.~ n .

1%

x 6

1% I .

tn.

1%

~n.

% tn
The hose must then withstand the following hydraulic pressures
for a period of ten minutcu:
Nom~nals ~ z eof hose
Pressure. Ib per Sq. In

% ~n.

600

~n.
So0

1%

The above specifications were taken from the Manual of the Am.
El. Ry Eng. Assn. Other specifications such as refer to manufacture,. porosity test, friction test, stretching test, tension test, are
also lncluded
Magnetic Brake. A form of magnetic brake in which are
embodied both a track and wheel brake is shown by Fig. 26. A
track brake-shoe is laced between the two pairs of wheels and is
drawn to the rails
an electromagnet which is suspended from
the car, thereby not merely adding its friction to the friction of
the wheel brake, but also actually increasing the rail ressure of the
wheels to thc extent that the supporting springs for t i e track shoes
and magnets are in tension through the descent of the track shoes
to the rails. The electromagnet a, dividing the track brake-shoe b
into two parts, is secured by pins to the two push rods c, and suspended a t a proper distance above the rails by the adjustable springs
h. The push rods are secured by pins to the lower ends of the brake

By

484

FLECTRIC RAILWAY HANDBOOK

levers d, which are connected a t their upper ends by the adjustable

rod g and are pivoted a t an intermediate point to the brake shoe
heads e, carrying the wheel brake shoes, and the hanger-links f,
suspended from the truck frame. The push rods c are telescopic,
so that a movement of the track shoe toward the right, relative to
the truck frame, causes the wheel brake shoe a t the right to be
applied to the wheel and the connection g to be moved to the left,
thereby applying the wheel brake shoe a t the left, the stop i preventing the lower end of the brake lever a t the left from following
the track brake shoe. A relative movement of the track brakc shoe
to the left is accompanied by application of the wheel brake shoes
through corresponding movement of the parts in the reverse order.
The brake controlling device may be incorporated in the running
controller or may be a separate device, placed by its side and operatively interlocked with it, so that neither can be caused to inter-

FIG.26.-Magnetic

brake

fere with the operation of the other. I n the operation of the

apparatus, the current is supplied by the motors running in multiple
as generators with the fields reversed (the trolley current being cut
off) and is divided between the electromagnets and controlled by a
rheostat so as to cause the track brake shoes to be drawn upon the
rails with a force proportionate to the braking requirements. I n
order to limit the amount of current through the track brake coils
to prevent slipping of the wheels, it is common practice to provide a
regulating resistance which, in connection with a regulating relay,
shunts a portion of the field current when the braking current
reaches a n excessive value. Resistance is also inserted in the connection between the two motors in order to decrease any tendency
toward unbalancing. I n some cases it has also been found necessary to use a demagnetizing shunt to release the track brakes when
power is again applied in the forward direction. The frictional
resistance of the rails to the motion of the track shoes causes
the wheel brakes t o be applied with corresponding force. Thus, to
the ordinary retardation of the wheel brakes is added that of the
track brake and also the back torque of the motors, which latter,
however, is practically limited .to compensation for the rotative

MAGNETIC BRAKES
energy of the motor and car wheels. The force of application
depends primarily upon the current and upon the electromagnets
operating the brake shoes. The attractive force of the rails upon
the magnets is under the control of the motorman up to a limit of
about 150 lb. per square inch of brake shoe surface in contact
with the rails. The strength of the magnet is limited by the sectional area of the rail acting as armature, and where the weight of
the car makes a magnet of greater strength desirable, the track shoe
is divided into three parts, instead of two, and wound to form a
three-pole magnet, or two combined two-pole electromagnets with
one common pole. The friction of the track brake shoe may also
be adjusted to some extent through the angular inclination of the
push rods c, by which some of the weight of the car may be thrown
upon the track shoes, the levers d being correspondingly adjusted
to reduce the wheel brake shoe pressure in proportion as the weight
is transferred to the track shoe. The current declines with the
speed during a stop, thereby offsetting the increased coefficient of
friction a t the lower speeds. I n bad weather, when the condition of
the rails is likely to be accompanied by wheel sliding, the braking
force operating the wheel brake is correspondingly reduced, so that
the force of application of the wheel brake is automatically proportioned to the rail friction which rotates the wheels. But, in addition to this valuable feature, if by chance the wheels should slide
upon the rails, the interruption of wheel rotation is accompanied by
the cessation of the track magnet current, through which the
pressure of the brake shoes upon the wheel is instantly relaxed and
rotation of the wheels is resumed, without injury or serious loss of
time.
A large amount of special track work is a handicap to the operation of magnetic brakes, increasing maintenance on both the
magnetic brake shoe and the track. T o obtain an effective magnetic
brake i t is often necessary to design the truck especially for this
feature, which involves additional cost and often increases the
length of wheel base. The application of magnetic brakes on cars
has resulted in the development of many detail designs caused by the
differences in schedule, in car, truck and rail construction, in size
and characteristics of motors, and in desires of the local operator.
The magnetic brake is most useful on cars used in heavy grade
operation.
Braking by Regeneration. The process of saving energy of a
moving train that would otherwise be consumed'in heating the
brake-shoes and car wheels is discussed under Regeneration"
(see page 203). I n furnishing the energy of regeneration the speed
of the train is reduced without wear and tear and excessive heating
of brake rigging, brake shoes and car wheels. The possibility of
accident from these sources is thus reduced, and on long mountain
grades the safety of operation due to the braking feature of
regeneration may be of more importance than the saving of energy.
The application of air brakes during the rocess of regeneration
may reduce or destroy the regeneration anBits braking action and
might bring an excessive load on the brakes by causing the moton
to take current from the line, but this may be avoided by employ-

ELECTRIC RAILWAY HANDBOOK

ing an automatic interlock between the two systems so that the
"service" air brake cannot be applied while the regenerative
brake is in use.
Braging by Reversing Motors. A retarding force may be applied
to the train by reversing the motors and applying current to them
through a portion of the starting resistance. This method should
be used only in an emergency, as lt strams the car equipment, and
if the direction of rotation of the wheels is reversed the braking
action will be of low efficiency because of the low value of coefficient
of friction between the slipping wheels and rails.
Braking by Bucking Motors. If two series motors whose armatures are revolving but to which no current is supplied from a n
external source have their connections reversed and then placed
in arallel, they will tend to operate as series generators; the one
of Eigher potential will continue to act as a generator and drive
current through the other which will consequently act as a motor.
This action will retard the motion of the armatures. The steps
necessary to make use of this action in retarding the motion of a
car are: Open circuit hreaker, throw reverse lever to the position
corresponding to motion opposite to that of the car and move
controller handle to a parallel notch. On a four motor car, where
pairs of motors are permanently connected in parallel, the last step
noted is unnecessary. I t should be noted that if the current supply
to a car ceases as the car is ascending a grade and it is desired to use
this method of braking to keep the car from backing down the grade,
the reverse lever should be left in the position corresponding to the
forward motion of the car up the grade. This method of braking
should be used only in emergency.
Brake Shoe Wear. Concerning brake shoe wear, the 1910 Committee on Rrake Shoe Tests, M.C.B.A., recommended that on the
cast iron wheel the shoe wear be determined by making loo
applications of the shoe to the wheel, under a pressure of 2808 Ib.,
and a t a constant wheel speed of 2 0 miles per hour, a t each application the shoe to be in contact with the wheel during 190 revolutions
and out of contact during the succeeding 610 revolutions. That,
under these conditions, the shoe shall lose in weight not more than
0.8 Ib. for each ~oo,ooo,ocmft-lb. of work done. That, on the steel
tired wheel, the shoe wear be determined by making 10 stops from
an initial speed of 65 miles per hour and under a shoe pressure of
12,ooo Ib. Ten minutes shall intervene between successsive applications of the shoe. That, under these conditions, the shoe shall
lose in weight not more than 4 Ib. for each roo,ooo,ooo ft.-lb. of
work done.
That committee recommended the adoption of the following suggestions in place of the then existing specifications for brake shoes:
a. Shoes shall be tested for coefficient of friction and for wear
upon the Master Car Builders' Association testing machine, or upon
a machine with equivalent characteristics.
6 . Shoes shall develop upon the cast iron wheel, in effecting
stops from a n initial speed of 40 m.p.h., a mean coefficient of
friction of not less than 2 2 per cent when the brake shoe pressure
is 1808 Ib., 16 per cent when the brake shoe pressure is 6840 Ib.

BRAKE SHOE PERFORMANCE

487

c. Shoes shall develop upon the steel tired wheel, in effecting

stops from an initial speed of 65 m.p.h., a mean coefficient of
friction of not less than 1 2 % per cent when the brake shoe
pressure is 6840 Ib., 11 per cent when the brake shoe pressure
IS xz,ooo Ib.
d . No limitation is placed upon the rise in coefficient of friction
a t the end of the stop.
General Performance of Brake Shoes. The following deductions
are based on laboratory tests made in connection with the Pennsylvania-Westinghouse tests, 1913:
(a) The generation of the retarding forces and consequent
absorption of the energy of the moving train is dependent upon
but a very small quantity of brake shoe metal. ( b ) The actual
bearing area rather than the total face area of the shoe is the
important factor in brake shoe performance. ( 6 ) The magnitude
of the bearing area changes throughout the stop and is greatest
near the end of the stop. ( d ) The bearing area shifts continuously
from one portion of the surface to another during the stop. (c) The
principal factor in producing high friction for any given braking
condition is the frequent shifting of the bearing area from the heated
to the cooler spots over the face of the shoe. ( j ) Slotted shoes or
shoes that are cracked are more flexible than solid shoes and the
bearing area shifts more readily than in the case of solid shoes. (g)
With shoes of the same type and approximately the same hardness,
the wear per unit of work done is less with the slotted shoe than with
the solid shoe. The stops with slotted shoes were always shorter
and the mean coefficient of friction higher than with solid shoes.
(h) The shifting of the bearing area will tend to be more rapid if the
slze provides more available area for shoe bearing. (i) The greater
the pressure per square inch of bearing area, the lower will be the
mean coefficient of friction. ( j ) Flanged shoes provide more
available area for bearing than unflanged shoes.
"Squealing" During Braking. Opinions based upon tests and
given in testimonies a t hearings on street car noises agree that
"squealing" during braking is due to rubbing of wheel flange
against rail head as the freedom of the wheel to seek the path of
least resistance is restricted by the brake shoe pressure. The
greatest amount of noise is given off by steel wheels. Reduction of
the noise has been sought in lubricating the flange by the use of a
brake shoe of special composition or having inserts of a lubricating
material. Among these materials are lead and a composition of
graphite and asphalt. The ideal lubricant is the one which may be
used a t the least cost and which will not reduce the friction between
brake shoe and wheel.
Wear, Relative Efficiency m d Cost of Various Types of Brake
Shoes. The followinn conclusions were drawn from the results of
stand tests and a test'by a year's service with 800 brake-shoes on
the Brooklyn Ra id Transit System by George L. Fowler. ]:our
general types of sEoes were used in these tests; they are designated
as "A," "R," "C," and "D," respectively. "A" was a plain,
hard cast iron shoe containing 1.67 per cent combined carbon
and 1.36 per cent graphitic carbon. The microscopic structure of

488

ELECTRIC RAILWAY HANDBOOK

"A" shoe seemed to be between malleable iron and gray

cast iron in that the graphite appeared both as nodules and very
small Bakes. "R" was of cast iron, with ends chilled, containing
1 . 2 0 per cent combined carbon and 1.85 per cent graphitic carbon.
The microscopic structure of the " B " shoe would probably be
classified as mottled iron. I t differed, however, from the typical
mottled iron in the more even balance of the pearlite and graphitic
carbon by which the structure more nearly resemb'ed that of hard
iron. " C" was of cast iron with an expanded metal tilling. "D"
was of cast iron with chilled iron inserts in the part bearing on the
tread. The coefficient of friction and wear for each of the general
types of shoes given in the following table were secured from tests
in a brake shoe testing machine by applying the brake shoe under
a pressure of 6840 Ib. and a speed of 20 miles per hour:
COEFFICIENT
OF FRICTION
AND WEAR
the

Type of shoe

Cqefficient of
frict~on.per cent
39.8
31.2
41.9
29.1

Inches. wear per

roo.ooo.ooo ft.-lb.
1.79
0.61
1.83
0.47

B
c
D
Conclusions: (a) I n order to avoid excessive costs for the application of brake shoes the weight should be limited so that no individual shoe should weigh more than 24 Ib. T o avoid an excessive
loss of weight in the percentage of scrap the minimum weight of
individual brake shoes should be 2 0 Ib. (6) There appears to be a
close relationship between the microscopical structure of cast iron
and its wearing qualities. Hence the foundry practice should be
such as to secure a structure that is closely granular, of uniform
texture and with the combined and gra hitic carbon pretty evenly
balanced. The graphitic carbon s h o u l l b e in the form of nodules
rather than Bakes. (c) There is a general relationship between
the coefficient of friction as determined on the testing machine and
the stopping qualities of a brake shoe in service. (d) As far as the
safety of a car is concerned, owing to the ability to sto from speeds
common in surface car service, there is no apprecia!le
difference
between any of the shoes tested. The extreme variation a t 15
miles per hour averages but 5 ft., or about 7.7 per cent. At 2 0 miles
per hour it is but 17% ft., or 16.26per cent. At lower speeds the
difference is correspondingly less. (e) Brake shoe wear averages
from 3.75 to 6.5 Ib. per 1000 wheel miles. The scrap weights
of brake shoes on the surface lines should not be allowed to drop
below 6% Ib. per shoe. When the scrap weight has reached 5%
Ib. there is danger of cutting the head. The ideal scrap weight is
30 per cent of the original weight. (f) Shoes containing chilled
cast iron either integrally or in the form of inserts wear steel wheels
more rapidly than those of gray cast iron only. The order of wear
of the wheels with the shoes submitted in the total averages of
mileage per Hsin. of wear is: "A," 100;" C," 88.23;" D," 83.23;
"B," 8 0 . 3 2 . (g) The wear of shoes is largely dependent on the
character of the foundation brake rigging and the way in which it .
is maintained. (h) There is a wide range of mileage obtained with

STANDARD R R A K E PARTS

-'&

Section A-A

Section

8-B

27.-A.E.R.E.A. standard brake head, shoe and key for 3 in. and
in. tread, wheels 33 in. to 36 in. d~arneter.

ELECTRIC RAILWAY IIANDBOOK

FIG. 28.-A.E.R.E.A.

standard brake head, shoe and key for 23.; in treads

wheels ax In. to 26 in. diameter.

Broke Shoe Gage

Side o f Gaqe M o r k e d A - M u s t f i t 8ackofJhoc

8-I;h&

Mdfhfor

L,he C-Is MI" H e ~ g h ofSlof~nCenteroflu

f
Measurd fmmBuckofJhoe
Line D-1s Mar ~ e ~o ~
f ~ /h
o+
lnf
hfeosuredfm~ucho~.h

entero of&

Brake Head Gage

A-HeudMvs fAdmr f S,dPofGagc foFr.//DepfhandMustF;fRndius
8-L Mor D,j fancc Befween Lugs o f H e a d
C-/5 Mar Thickne~sofMPfalBetween
Face ofCenierfuysondKry6Jd
0-1s Mln Th,ckn~ssofhfefalBrfween h c e o f Cenferfuys und&Slof

PIG. 2%-Limit

gages for brake heacls and brake shoes for 3 in. tread
2!2 in. tread wheels 28 In. diam-

wheels 33 in. diameter and over, and for

eter and over (A.E.R.E.A. standard).

Brake Shoe Gage

S ~ d eo f Gage Marked A - Musf A f Back of5hoe
A- /l Max Ma'fh for Cenfer L ug
6 - / s Mn Wd+hfor CenferLug
Line C - k; Mln He9hf of5lof /nfenfer off ugMea~uredhrn6ackof~hoe
Line 0 - / s Mar Me&+ of5lof1nfenferofLu~
Measuredfrom Bockof~hoe

Brake Heod Gage

d - Head Musf Aclmlt Side o f Gage fo Full De fh and Musf hf Rod;,,<

B -1s MOID,.sfonce Between ~ e n f e r ~ ov f~Rs~ a d
C-/s Max Thickness o f Me fa/ Be f ween ace o f Center Lugs andKg5/of

D-1s: M l n

Th~cknessof M e fa/ Bdween Face ofCenferfugsandXeyS/of

FIG j o -Lrmlt gages for brake heads and brake shoes for 236 ~n and 3 m
tread wheels 26 rn drameter and under (A E R E A standard).

"

STANDARD BRAKE I'ARTS

I-"

A A AS max~mumund8"s m,n,mum wdth 16rcrn~rlug

L/m C i s m,mmum he~~tofslof~ncer*eroflug,mcosurpd

from buckofshoe
Line

D ~rno~~rnumhe~~tofslof
s
incenfcr o f l u ~ ~ m ~ . m u d
frombock ofshoe
muzimum fhjdness ofbndl~gson bockofshoe

Frr, j r -A

R A standard brake shoe ke) and gage

494

ELECTRIC RAILWAY HANDBOOK

individual shoes of the same type. ( i ) A steel wheel will N n about

8000 miles per )i6 in. wear in surface work. ( j ) The cost of brake
shoes only per rooo car miles is least for the cast Iron " B " brake
shoes with chilled ends. (k) The combinatipn of brake shoe ahd
wheel costs per 1000car miles places the A" and " B " brake
shoes about on a par. (1) If the price is based on the cost of brake
shoes per ~ o o ocar miles, a higher price can be paid for the plain
cast iron "A" brake shoe than for the one with ch~lledends, because
of the decreased cost of wheel wear. (m) Where steel wheels am
used, both thc wear of the wheel and the wear of the shoe should be
considered, for though a shoe may be very economical in itself, it
may have such a severe effect u n the wheel that when both
wheel and shoe are taken together t E result may not be economical.
A.E.R.E.A. Standard Brake Shoes, Brake Shoe Heads and Keys.
Figures 2 7 and 2 8 show the principal dimensions of two of the
A.E.R.E.A. standard brake shoes, brake shoe heads and keys,
as revised in 1921. Standard designs for wheels of other dimensions are shown in the A.E.R.E.A. Manual. The thickness
of the shoe is 1% in. If a thicker shoe is desired, the increased
thickness should be specified, but the other essential dimensions
should not be changed.
A.E.R.E.A. Standard Limit Gages. Figs. 29 and 30 show the
A.E.R.E.A. standard limit gages for brake shoe heads and brake
shoes, as approved in 192I.
A.R.A. Standard Brake Shoe, Key, and Limit Gage. Fig. 31
shows the principal dimensions of the A.R.A. standard brake shae,
key and limit gage.

SECTION I X

CARS

Resent Tendencies in Car Design. Of recent years there has

been a n increasing tendency to apply sound engineering principles
to the structural design of electr~crailway- cars and to give consideration to the great influence of car weight upon operating costs.
With a proper consideration of the stresses involved, and an increased
use of steel shapes and plates, the weights of car bodies have been
reduced very materially, with no sacrifice of strength or life, and
accompanied by considerably reduced maintenance costs.
The Birney "safety" car, a small single truck car of special
design for one-man operation, was the first long step toward light
weight. The first cars of this type, appearing in 1915, weighed
approximately 10,ooo Ib., completely equipped with special trucks
and motors. I n response to a general feeling that the original
construction of these cars was too light to afford proper life with
reasonable maintenance costs, there was a reaction to the present
generally used one-man safety car which weighs approximately
16,000 Ib. Corresponding reductions have been made in the weight
of double truck cars, both for city and interurban service. Both
trucks and motors have also felt this tendency to lighter weight,
as has been noted elsewhere. Figure 14, Section VII (page 393),
:or instance, illustrates a n application of hollow axles and band
brakes to trucks for light weight cars, while in Section IV may be
noted the reduction in motor weights by the proper application of
lentilation and other refinements of design.
There has been a further design tendency toward an elimination
of the truck as such, by combining the spring, wheel and motor
kssembiy with the car underframing, somewhat along the lines of
a I automobile chassis. In some cases the automobile design is
followed still further in the adoption of the two-part axle, differential,
ant1 longitudinal shaft drive, and in general making full use
of ball or roller bearings. These advanced designs are a t present
still in the design, or a t best the experimental stage, and appear to
b t more attractive in Europe than in this country. (Sce Electric
Railway Journal, Vol. 60, pages 825 and 828, and General Electric
Raiew, Vol. 19, page 881.)
The great importance of time lost in stops, especially in city
service, has brought attention to the matter of accelerating the
movement of passengers in boarding and alighting, and has had
considerable influence in the arrangement of entrances and exits
in city cars.
One-man operation, where traffic conditions are such that it is
feasible, affords a n opportunity for great savings in labor costs per

495

ELECTRIC RAILWAY HANDBOOK

car m~le. This also has occasioned numerous changes in details
of car design.
Car Weights a s Affecting Operating Costs. No general formulas
can be glven for the relat~onsbetween car weights and the various
items of operating expense Some expenses will vary almost directly
with car weight, wh~leothers have no relat~onto it. Some expenses
will be affected l ~ t t l eunless the change in weight applies to all cars,
while others will be affected proportionately by a partial change.
Only a careful engineering study can solve the problem for any
individual case; In general such a study will be along the following
lines :
Maintenance of W a y and Structures. Wear of rails and special
work w~llvary almost-directly m t h car weight, as will that of ties
and ballast, but depreciation will not be affected. Track labor in
proportion to the above Paving and other accounts, including
dlstnbution system and buildings, not usually affected
Maintenance of Equipment. There is no saving in these accounts
purely on account of reduction in weight of cars, except possibly
In wear on shop equipment, and in some cases minor savings due to
reduced welght and cost of repair parts. As a change to llghter
weight equipment most frequently includes a change to more
economically maintained equipment, the savings due to the latter
properly should be credited here.
Power If schedules remain the same, the amount of power in
kilowatt hours will vary almost directly as the weight of cars
(see Section 111, pages 193 to 2 0 0 ) . The variation in cost of power
with respect to k~lowatthours, however, will depend upon conditions
of power supply I f power is manufactured, no saving will be
effected beyond that In fuel and the cost of handling it, which, within
the general limits of the change in output due to changing car
weights, generally will be in proportion to the output. If power is
purchased, the change in total cost due to change in kilowatt hours
purchased may be estimated readily.
Conductzng Transportatron. There is no saving in these accounts
urely on account of a reduction in car weight, except the poss.h i t y of minor savings in car service expenses, such as lubrication.
If the change in car weight is accompanied by a change from twoman to one-man operation, there may be a large saving in wagts
of conductors and motormen, depending upon the daerential in
wages between two- and one-man operators.
General and Mzscellaneous Expense. The cost of accidents is
the only one of thls group of expenses that may be affected in a
change in car weights. The increased efficiency of modern saf 'y
devices which usually are included in new car equipment should
reduce the number and cost of accidents. Such a reduction, if i t
is effected, is not due purely to the change in car weights, and
estimates of such savings
- should be made with great care. if included
a t all
Fzxed Charges With a complete change to light weight cars,
longer tie spacing, lighter rail and lighter general track construction
may be used, with correspondingly less first cost. Whether or not
credit may be taken for a consequent reduction in fixed charges will

497

CITY CAR DESIGN

depend upon the circumstances of the individual case. I n a comparison of costs accompanying two plans for original construction,
one w ~ t hlight, the other with heavy cars, full credlt would be
given this item. Similar consideration applies to distnbution
system copper and capacities of power and substations, as all these
ar- dependent upon power peaks, which are directly affected by
car weights.
Decreased Weight of Single-end Cars. The A E.R.E.A.
1910 Committee on Equipment called attention to the large saving in weight which can be made by the use of single-end cars,
as follows, as applying to double-truck car of the semi-convertible
type.
Weight eliminated by making car singleend:
~

Step risers, hangers and brackets

Reversing mechan~smfor seats

Brake ngglng
Fender and han ers
Doors. door pocfets and operating mechanism
Sand box and mechanrsm
Brake valve
AII-brake plplng
Trolley equipment
Electnc equipment, wlnng. etc .
.
.
Headl~ght
Smtch cablnet
.....
Llfe guards
......
Snow scrapers
. .. . .
Total sanng 1n welght

Lb.
306
I 08
332
125
548
40
ax
100

140
292
35
75
130
225
2477

While the equipment advantages of single-end cars are admitted,

many operators hesitate to adopt a design which cannot readily
be turned back in case of blockades, which calls for either Y's
or loops a t the end of routes, and which is harder to handle quickly
in case of car house fires.
Light Weight Single Truck Cars. This design is commonly known
as the Birney "safety" car, as it was first developed by C. 0.
Birney of the Stone & Webster organization While minor changes
in design have been made by some roads, the car is of the general
design as shown by Fig. I. I t seats 3 2 passengers when arranged
for double-end operation, this capacity being increased to 35 by
utilizing the controller and brake space when equ pped only for
single-end operation. The car is for one-man operation and is
equipped with safety features which generally provide ( I ) that the
brakes cannot be released or the motor current applied until doors
are closed, ( 2 ) the doors cannot be opened until the controller is
a t the off position and the brakes applied, and (3) the controller is
arranged to require a conscious effort on the part of the operator
to operate it. Doors are operated and rails are sanded pneumatically, the air valve having operating positions so that one
handle operates brakes, sand and doors. When the air valve handle
is thrown to the emergency stop position, power is cut off, sand is
applied, and doors are opened, automatically. (See p 476.)
The small capacity of the car avoids the necess~tyfor a large
operating platform, and leaving passengers usually alight before

ELECTRIC RAILWAY HANDBOOK

*Cpsn"Y

FIG I -Typ~cal s~ngletruck l ~ g h twe~ghtcar

-k'$'ay/ 7
Eilt

PIG. 2 -Double

truck light weight car. m t h automat~cexlt gate for odeman operation. Chicago Surface Lmes.

PIG 3 -Double

truck l~ghtwe~ghtcar Eastern Massachusetts St. Ry.

c ~ t yand suburban servlce

PIG.4 -Ft

Wayne c ~ t ycar.

CITY CAR DESIGN

others are allowed to board; this when boarding and alighting are
through one door as shown by Fig. I , the usual design. The managers of some roads, however, have felt that the single door caused
undue delays a t passenger stops, and have provided a separate exit
door alongs~dethe one shown in Fig I , in such cases, the two doors
usually are operated by separate controls so that the operatorcan
handle them singly In Chicago, the exit door has been arranged
to operate automatically, being opened by the leaving passenger's
weight on a floor plate just ins~dethe door, provided a full brake
application has been made and the operating handle is in the dooropening position, the door closes automatically when the passenger
has cleared the ste preventing boarding passengers entering via
2).
the exit door (see
The single-truck double-end safety car as built in Chicago weighs
as follows:
Lb
Car body
Steel frame
Wood framrng, floors, roof. lrnrng and finlsh
Doors and operating mechanrsm
Car body apparatus and fitt~ngs

1.930
z.3ao
1,117

seats

1.662

700

~ L & a r y electrical apparatus. m n n g and condurt

Rarlrngs, srgns, curtarns ventilators. regrsters, drawbars, etc
Motor e a u r ~ m e n t
Motors, $rons and gear cases
Motor m u g and condurt
Controllers and resrstance
Trolley, arrester. fuse, choke corls, etc
Brake eqmpment
Compressor, reservorrs. governor, etc
Prpmg. valves and fittrngs
Brake cylrnder and foundahon nggrng
Hand brake and nggrng
Sand boxes and traps
Trucks
Pramrng spnngs, etc
Axles wheels and gears
Mrscellaneous
Parnt and varnlsh
Mrscdlaneous bolts. screws. etc

Complete car

Lb.
5.367

..

&6
707
2.809

665
574
191
94
59
9.780

1.583

4.693
a68

I10
158

16 375

The single-truck safety car may be used wherever a small car

unit is desirable, acd as compared with the former generally used
city car, makes posslble large reductions in costs per car mile, due
to one-man operation, as well as on account of small weight, as
outlined on page 496. The field of application comprises the following general classes of service- (I) lightly traveled lines on which the
substitution is made on a car for car basis, ( 2 ) lightly traveled lines
where materially increased service is furnished, (3) heavy service
lines radiat~ngfrom rapid transit or main truck surface lines; (4)
lines paralleling rapid transit lines; and (5) main surface lines.
I t usually is undesirable to substitute safety cars on a car for
car bass especially where large double-tmck cars were previously
operated. There are, however, many lines now operating with twoman cars where 3 substitution of safety cars for an equal number of

500

ELECTRIC RAILWAY HANDBOOK

FIG.5.-East St. Louis city car.

PIG.6 c O m a h a city car.

FIG.?.-Pittsburgh Railways Co. c ~ t ycar.

FIG.8.-Cleveland city car; front-entrance, center-exit type.

CITY CAR DESIGN

501

the old cars is the only proper course. Many cases can be cited
where railways began safety car operation by giving shorter headways on one or more of the busiest lines, and a t the same time or
shortly thereafter successfully substituted safety cars for old singletruck cars on a car for car basis on less important lines.
One of the most gratifying features of safety car operation is the
increased revenue which often has resulted from short headways
on lines previously regarded as unprofitable. The possibility of
increase in receipts resulting from more frequent service always has
been recognized. but the desirability of handling the greatest nurn-

PIG.9.-Brooklyn

city car. double-end front-entrance center-exit type.

ber of passengers per platform man has heretofore required the

use of large units running a t infrequent intervals, because when
shorter headways are operated with large -ars, the increased revenue
is counterbalanced by the increased operating expense. When the
safety car was first used to replace large double truck cars, it was
necessary in almost every instance to operate a t least 50 per cent
more mileage in order to handle the traffic properly. While some
increase in car riding generally was expected, the decrease in opera-

---.----.---.-.--------

FIG.~~.-Charlestoncity car for multiple unit operation.

ting expense was the chief reason for the adoption of safety cars.
Actual results of operation in many cases showed that the increase
in revenue was greater than the saving in operating expense.
The application of safety cars to lines rad~atingfrom rapid transit
lines or main trunk surface lines has been made in several of the
larger cities. Here the morning loads are picked up gradually and
discharged a t one point, and in the evening the majority of passengers board the car a t one point and leave in small groups. The
collection of fares can be simplified under these conditions by having
prepayment and post-payment areas a t the mpid transit terminals;

ELECTRIC RAILWAY HANDBOOK

all passengers boarding the car a t other points can pay as they
leave on inbound trips and as they enter on outbound trips.
In the application of safety cars to the main surface lines of large
cities the methods vary in different sections of the country, and there
is an even greater variation in opinions regarding their practicability
in this service. The most common use of safety cars on main surface lines is where the car mileage is increased from 50 to loo per cent.
The rapid acceleration and retardation of the safety car, the smaller
number of passengers per car and hence fewer stops, the low floor
design, the elimination of signals between conductor and motorman,
and the better view tra5c o5cers have of the door, combine to
speed up the movement of individual cars; on the other hand the
safety provisions of the door and step control and the collection of
fares by the car operator may tend to increase the duration of stops.
As previously mentioned, the latter tendency is counteracted in
some cases by the use of street collectors and by prepayment areas.
The congestion arising from the greater number of units also tends
to slow up traffic. This is not proportionate to the number of
units because of the features which tend to accelerate the movement
of the safety car, and because of their smaller size. I t is certain
that a number of safety cars can pass over a congested section of
city street more rapidly than an equal number of large doubletruck cars, and still more readily than an equal number of two-car
trains. Observations made of congested city service when safety
cars and double-truck cars operate on the same track show that the
safety cars are frequently waiting for a double-truck car to get out
of the way, and that in the neighborhood of 50 per cent more safety
cars than double-truck cars can be operated over a given section of
track in a given time. In general, a headway of 30 to 40 seconds
for distances up to one mile on surface lines is about the limiting
frequency even for safety car operation. Bearing in mind that no
material increase in riding can be anticipated by decreasing headways under three or four minutes, the use of safety cars for the
entire day generally does not seem advisable on the heaviest lines
even where their operation is physically possible. On many
heavy surface lines, safety cars are used in combination with doubletruck cars, short-routing the former during rush hours.
Light Weight Double-truck Cars. The successful operation of
the Birney single-truck one-man car directed attention to thelarger
double-truck car with respect to reductions in weight and possibilities of one-man operation. The table on page 508 shows several
instances where double-truck cars seating about 48 passengers have
been brought down to a tothl weight of approximately 30,000 Ib.,
or about 625 lb. per seat. These cars often are equipped with all
of the control and safety devices for one-man operation and are so
operated. When built with proper door arrangement, they frequently are operated one-man in the normal hours and two-man
in the rush hours.
Figs. 2 and 3 show the arrangement of light weight doubletruck one-man cars of the Chicago and Eastern Massachusetts
companies, respectively. Weights of the Chicago car are as follows:

SEAT SPACE PER PASSENGER

L b.

...........................................

Steel
Car
bod?rame.
.......
....................................
Wood framing, flqors. roof, lining and finish. .......

4.600
3,309
1 . 4 ~ ~

Doors and operat~ngmechanism.. .................

Car body apparatus and fittings.. ............................
Seats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
900
Auxiliary electrical apparatus. wiring and conduit.. .
346
Railings. signs, curtains, ventilators. registers, drawbars.etc .....................................
956
Motorequipment ..........................................
Motors, pinions and gear cases.. ..................
3.864
Motor wiring and conduit.. ....................... 1 7 5
Controllers and resistance.. ......................
8ga
Trolley, anester. fuse. choke coils. etc.. ............
a64
Brakeequipment ...........................................
Compressor, reservoirs, governor, etc..
953
Piping. valves and fittings.. ......................
Brake cylinder and foundation rigging.. ...........
Hand brake and rigging.. ........................
139
Sand boxes and traps.. ..........................
59
Tmcks ....................................................
Framing, bolsters. transoms, springs. etc.. ......... 4.658
Axles. wheels and gears.. ........................ 4.148
150
Tmck guards ...................................
Miscellaneous..............................................
Paint and varnish.. .............................
a25
Miscellaneous bolts, screws, etc..
a80

............

Lb.
9,361

a,zoa

5,195

1,831

:gad

.................

Complete car..

............................................

8,956

505

28.050

Arrangement of Seats. The transverse or cross seat is more

comfortable than the longitudinal seat, especially with rapid rates
of acceleration and braking. Approximately the same seating
capacity may be had with either type, but the longitudinal seat
leaves more room for standing passengers. Where large standing
loads are to be handled in rush hour service, the entire seating capacity often is in longitudinal seats, as in Fig. 12. In interurban cars
where the rides are long, all seats usually are transverse, as shown
in Figs. 15, 16, 17 and 18. The usual arrangement of the singletruck safety car also is with all cross seats, as in Fig. I. The general
practice in city and suburban service is to provide part cross and
part longitudinal seats, thus partly catering to the passenger's comfort, and partly taking advantage of the larger car capacity afforded
by the longitudinal seat. When both types are used, the cross
seats generally are grouped a t the ends of the car, as in Figs. 2,4, 6,
13 and 14, thus providing space for movement of boarding and
alighting passengers when the car is full. In some cases an arrangement is used wherein the cross seats are grouped on one side of the
car, with longitudinal seats opposite, as in Figs. 3, 5 and 7.
Seat Space per Passenger. In determining the seatlng capacity
of cars with longitudinal seats, it is customary to allow 17 to 18 in.
per passenger. The Equipment Committee of the Am. El. Ry. Eng.
Assn. (1911) recommended that I 7 in. be used in determining the
capacity of longitudinal seats; 17 in. is used by the Massachusetts
Dept. of Public Utilities and by the District of Columbia Public
Utilities Commission. The standard of the Public Service Commission, First District, New York, is 17.78 in., this being the average
of a number of observations by Commission inspectors. Kidder's

504

ELECTRIC RAILWAY HANDBOOK

FIG. XI.-Milwaukee city car for multiple unit operation.

PIG. 12.-New

Jersey city and suburban car.

FIG. 13.-Richmond city and suburban car.

PIG. 14.-Connecticut city and suburban car.

PASSENGER CAPACITY
Architect's and Builder's Pocket Book gives 18 in. as the length of
church pew considered as a "sitting."
Passenger Capacity. Seating capacity is determined by the
number of transverse seats, plus the capacity of longitudinal seats
figured as above.
Standing capacity is customarily determined by dividing the
floor area, measured in square feet, by 2, 3 or 4. In a longitudinal
seat car, 6 to 8 in. is deducted in front of the seats to allow for knee
room, and the floor area devoted to fare boxes, control apparatus,
etc., is also deducted. Standards adopted by the U. S. army are
based on the experience that an area of 3.3 j square feet per person
will permit free maneuvering. A report of Ford, Bacon and Davis
to the Pennsylvania State Commission in the matter of service
on the lines of the Philadelphia Rapid Transit Company, states that
4 square feet per standing passenger provides sufficient room for
comfort and for free movement through the car, and that the total
seated and standing capacity of the interior of the car will equal
about 3.5 square feet per total seated and standing passenger. In a
report of B. J. Arnold on transportation facilities in the city of
San Francisco, it was recommended that reasonable standards to
be applied to all types of cars are as follows: (I) comfortable standing, jo per cent in excess of cross seats, and roo per cent in excess
of longitudinal seats, plus platforms; (2) normal maximum capacity
3 square feet per standing passenger; (3) emergency maximum capacity, 2 square feet per standing passenger.
In connection with the determination of car capacity, it should be
noted that there are a certain number of passengers who will stand
by preference, irrespective of whether or not seats are available. The
number of such preferential standers varies with the conditions
existing in various localities. For instance, investigation has shown
that 2 1 per cent of any load will stand by preference in Madison,
Wisconsin; in La Crosse, 15.5 per cent; in Milwaukee, 19 per cent;
in Lincoln, Nebraska, 14 per cent; in Cincinnati, 15.5 per cent;
and in Springfield, Mass., about 10per cent.
Drop Platform Type. The greater number of city cars in use
today are of this type. Platform sills are carried by cantilever
suspension from the main sills and extend out to carry the end
platforms and vestibules. The floor of the end platform is thus
several inches below the floor of the car. In practically all cases
the platforms are enclosed by a vestibule with folding doors over
the entrance and exit steps. In later types of both city and suhurban cars the practice is to omit the bulkhead at the end of the car
body proper to facilitate passenger interchange. In two-man
operation, such cars may be operated with the entrance a t the rear
and exit a t the front, with prepayment or pay-as-you-enter operation, or front-entrance, rear-ex~t,with pay-leave operation. Typical cars of this type are shown in Figs. 4, 5,6,12,13 and 14.
Center Entrance and Exit Type. In this type, as illustrated by
Figs. 10 and 15, the combined entrance and exit doors are located
in the middle of the car. This type was advocated for city use
principally because by means of ramped floors or a depressed well
a t the middle of the car, the number and height of steps to the street

506

a E C T K I C KAlLWAY HANDBOOK

FIG. 15.-Kansas City interburban car.

PIG.16.-Lake Shore El. Ry. Co. (Ohio) interurban car.

.I%
.

..

FIG.17.-Union Traction Co. of Ind. interurban car.

FIG.18.-Chicago.

North Shore & Milwaukee interurban car,

C I l T CAR DESIGN
surface could be reduced very materially as compared with the
drop platform type. Experience with this type is congested city
operation, however, demonstrated that there was likely to be
conflict between boarding and alighting passengers, with a resulting
loss of passenger interchange time. Its use a t present, consequently, is practically confined to a few cases in interurban service
where passenger interchange time is not so important, to motor
cars which are frequently used in multiple unit service, and to
trail cars. In the two latter named cases, one conductor per car
can handle passengers most rapidly with the entrance and exit
combined or located side by side in the middle of the car. In the
Toronto trailers, there are three doors a t the middle of the car.
with control and railings so arranged that two may be used for
entrance and one for exit when the heavier movement of passengers
is boarding, and the reverse when the heavier movement of
passengers is alighting.
Front Entrance, Center Exit Type. This t , q e of car is also
eter Witt" car,
known as the "pay-as-you-pass" type or the
as it was originated in Cleveland during the time that Mr. Witt was
City Commissioner of Street Railways. For single-end operation
the car is illustrated by Fig. 8. Entrance is at the front end where
the doors are controlled by the motorman, and exit is in the middle
of the car through doors operated by the conductor who with his
fare box is stationed a t that point. Passengers board the car
with no delay incident to fare collection up to the point where the
entire front half of the car is filled. Fares are paid as passengers
ass the conductor's position. This may be done immediately on
Larding the car, if it is then convenient, or a t any time during
the trip while the car is in motion, in either of which cases the
passenger takes his seat somewhere in the rear half of the car.
Passengers who remain in the front half of the car pay fares as
they pass the conductor on their way to the exit door. When
cars are operated single-end exclusively, the arrangement of seats
is usually as shown in Fig. 8, with longitudinal seats in the front
half of the car and transverse seats in the rear half. This accomishes the purpose of providing a large standing area in the front
alf of the car, so that fare collection need not lengthen the stop
time where a large number of passengers are boarding, as might
occur with pay-enter farecollection. As the transverse seats in the
rear of the car are the more desirable, this arrangement also tends
to induce passengers to pay fares and occupy positions in the rear
of the car so long as space is there available, and this reduces the
stop time for alighting passengers as compared with pay-leave
operation.
When necessary to operate cars double-end, it is not possible to
concentrate the longitudinal and transverse seats as above noted,
but the arrangement is usually somewhat as shown by Fig. 9.
The relative arrangement of entrance, exit, andconductor's position,
used with this type of car, very materially reduces the time required
a t passenger stops. This is shown very definitely by Fig. 5 , Section
111, page 124, the data for which was obtained through a large
number of observations in several cities.

E'

508

ELECTRIC RAILWAY HANDBOOK

L(

.,'!

N "'0
o n n n o

'0"

.s(aaq4 .me!a

78as1ad . q

n n n n n

"Ud'0"
--nnon
N

~ m a m mu n m d m o m n m c m o b n o
-0-am
~ m d ' 0 ' 0 ate-n 0n0'0~ . o n d - t n '0'0'0'0~0~ n t - t - t'0-'0b'0
n

M t -

U O ~ J ~ ~ Y P&HL o o r :
lotoml&.d.H
s1oJom.o~
'91 'si030m pue
tqa!aA

SqJR1l

.q1-.ss~.15xa
198!a=Iwo.L
A1p3dw Bu!aw

d
91P!* II=aAo

'0'000"
n o m o o
m

""""d

-6

fk

WCnOQ

;<do&

ONmmC
)dad;);;)

dDoDdhad

COQWd

n n n n o

-nmnmo
"?"""

ooowo
"U*",P

NdOOd

Udddd

VddUn

n d d d d

*? ? "? . "

0
m0o 0
oo
0 n0

Odooo
ommdn

0 0 0 0 0
d00'00

om000
n - n n n

mmnon
n - o o n

nn
"n

. .n
. ., "
'

0 0 0 :o

""""d

. t 1'9116 ?*?'9?

2 : ~:

m
nn
-m
- -m
-t-

00000

00000

?????

?99"?

ooooo

OOQOO
m o ~ a oo o o d o

???t4

E z z z z mY4.m;: : f f f g

g;sg%

'$3%%0,%
9'0d-r~

m n n v n

aur,
X XX

t-mmmm

mmmmc

:eon-

ooomo

bz

,z,

--

?F"?Q?

00000

nomoc
Q????

~'00y-%
m t n d m

SEmz5

mmmmm

ooammm

z m o o o

a z o n m

93%nallI=aAO -

d
uO!te~aO
%!*ap m aauvqua

pus qqnop

10

al8u!s

a3pu.s

t%T%$ E3$;",%:%7'

--,.,.ee

nnnnn

'$=f;=f"o
o n n n n

YYUUW WYW$
ClrnQmQ

n n C n n

e
;;

W
I
www
U

;
;

W
-W
U

IWU

QQQQQ QmmQQ rnQQQL3

P.$,z.%E
.$.$,$+,$ 4,$,$,$.$.$.$.$.$.$
UUUUU UUUmU UUOUU OOUUO
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..
. . . . . . . . . . . . . . . . . . .
-. . . . .. .. . . .+ .. .. : . ..iI & ; i : : : :
s . 1 : .

dudmoa

10 61!3

.3.:>

f i + sW '5S -

2
*a$$

b,

mt-

:.u o;

U-.-

CX-,.$
.:em
6;ggg
gS&;

'on

-ill&

., * . p '2 : f i :i
26
: di 3 2 g
ae2$ . 2 C

EEaz$ ,
232::
gz$&
g$gh;
e u w
:

SURFACE CAR DATA

ao*',,,,

'ON

.ql 'aOlom puo

q r ~ n r.Ia3xa
~
?qB!afi

PP

P W P O P

PPPP

00000
0 "03 W "

00000
0 0Q 0 0

0000
000 0

"PO

? t l C ?

vr oo
n n n m n
N N

?'"??

nm
nnndo
~l N

??1?
N

Y) Y)

gddd

510

ELECTRIC RAILWAY HANDBOOK

Articulated Cars.

The

Milwaukee Electric Railway & Light Company pas

designed and put illto
operation a three-truck,
two-man train unit, in
which two separate car
bodies are carried by three
trucks, two of which $re
located in the usual p,ysitions a t the outer ends of
the cars, while the third is
located between the two
cars and carries the weight
of one end of each car.
The arrangement of the
ends of the two cars where
joined together, with the
vestibuling arrangement so
as to provide for an interchange of passeneers between the cars, and the
location of the middle
truck,are s h m in Fig. sg.
As constructed in Milwaukee, old cars being used, i t
was necessary to support
the old car bodies in substantially the same manner as they had been pieviously carried upon individual trucks, namely, a t
the car body bolsters. This
was accomplished by placing under each car an auxiliary steel underframe
beginning a t the outer
truck holster and continuing through to the point of
joining the two car bodies
together. By suitable design, this steel underframe
carries the weight of the
car a t the old car body balster, provides adequate
support for a greatly elllarged platform, and transmits the load to the middle
truck, where a kind of ball
and socket joint arrangement on the ends of the
two main auxiliary girdem

CITY CAR DESIGN

611

permits both to rest and pivot on the center-truck kingpin and

bolster. Enlarged platforms were desired in order to provide
space for the larger number of passengers to be handled and
to accommodate the heating equipment. The platforms were
connected by overlapping steel plates attached to the platforms, and arranged to permit the necessary joint movement on
curves. The vestibules were connected by diaphragms. The
motors of the cars were grouped on the outer trucks. There were
recovered from each pair of old cars two trolleys, two sets of air
and electric control, one air compressor and two standard trucks,
besides various minor equipment, such as dcstination signs and
headlights. Notwithstanding the addition of the third truck, new
steel underframes and enlarged platforms, the total weight of the
finished train is only slightly more than the combined weight of the
individual can. On account of the exposure of the middle truck
on curves, it was found desirable to reduce the lateral dimensions
of the truck and necessary clearance to a minimum, which was
accomplished by building a special truck with inside journal
boxes. A similar plan proposed for use in Detroit contemplates
the use of three car body units mounted on four sets of trucks. In
this case each of the car body units is made center-entrance and
exit, so that the middle trucks may be of standard design, and all
four trucks may carry motors. Experience in Milwaukee has
&urn hthree-truck,twa-ra llnits t~ b ~ v edvantages at (,I)
utilization of old equipment so as to make it entirely acceptable
for rush hour and special service, as well as for regular service on
heavy traffic lines; (2) economy in platform labor, as a motorman
and one conductor can handle the load of two full-sized double
truck cars; and (3) greatly reduced overhang of coupled ends of
cars on track curves. The disadvantages include ( I ) slowness in
passenger interchange as compared with single car units, and (2)
some difficulty in locating the trolley stand for proper operation
and still accessible to the conductor.
Double Deck Cars have been quite popular for city service in
Great Britain, but in this country the use of this type has been
confined to a few trial installations. Under congested traffic
conditions the double deck car occupies only about half the street
space per passenger, and it is capable of demonstration that its overaU cost of operation per seat mile is less than that of any combination of motors or motors and trailers of simlar capacity. Nevertheless its operation in American cities has been without the
marked success which such figures might indicate, due to the
apparent unwillingness of the street car rider to climb to the upper
deck. Such results are di5cult to understand in view of the
popularity of the upper deck seats in motor bus service in New
York and Chicago.
Rapid Transit Cars. I n the design of a car for elevated and
subway service the conditions to be met are considerably d~fferent
from those encountered by a car on a street railway, both as to the
limitations of size and the difficultiesof traffic to be handled. The
design also wiU differ from that of the rolling stock of a steam or
electrified suburban line, in which service a higher ratio of seat-

512

ELECTRIC RAILWAY HANDBOOK

TAEILE OF COXPARAT~VE
DATA FOR
Number
of cars in

train

Total
length

VARIOUS ELEVATEDAND
'rota\ capacrty
Seated

I
I
Island Railroad
drv~swn.

to

: : :a%
o

xoa 8

---3

247 0

0
o
o

64 5%
128 1056
193 334
257 9
312 154
386 7%
4510%
51s 6

Island Railroad 3

~xway-suburban 4
~ e n n s ~ l v a n i aeta- 5
tion.
6
7
8

a
3
4

0
o

6
7
8

52
56
104
160
a12
268
7I

142
213
284
355
416

497
568

SUBWAYC~~~.--(Conlinfled)
n
o
+

IS
13%
308
444
59.9
734

10.05

7.91

7.7

10.05
8.04
8.62
7.541

7.91
6.12
4-64
5.70

7.7
5.0
I 66

12.50

----

4.48

9.0

7.22

6.62

800
Boo

?.SO

I .80
2.01

I ~ W

1.03

--.-

I 86

371
558
744
930
11x6
1302
1488

400

. . . . . ....... ................... . . . . . . . . . . . .

2.52

IZOO

'

470
940
1410
I 680
3350
aBio

3190

i 376O

NOT I: Projection of couplers be,yond buffen i n cMe of Brookly? Rapid Transit. Boston, and Hudson& Manhattan assulrred
ai
a In., end for New York Munxclpal Raxway 1% in. NOTE a . The horsbpower per car of New York ~ u n i c i ~Railway
to
Corporation proposed rubway cars, 280 h.p. NOZB3: The wejght.of ench passenger assumed t o be I40 lb. NOTE.4: standing
capocity q u r e d a t I H q.ft. per passeneex. allow~ngp in. a t 10w1tudlMl ~ o t for
a knee roam. N o r r s: No deduct~onmode for
motornun I cab in u1culatin.g prrrcngcr capwlty.

DATAPOP
TABLE
OF COMPARATIVE

VARIOUS

ELEVATED
AND SWWAY

CARS.--(Continued)

Weight in pounds

Per cent

Total
Road and class of service

Interborough Rapid T r d t Subway-8tsel c a m

New York Municipal

Corporation-Subway.

Railway

r 55.740

77,870
51.410

207,150

101.530
75,070
103.060
178,130

r70,ooo
as5.000
340.000
415,000
sro.ooo
~95.000
68o.000
71.660
34.500
143.3a0
177.890
249,480
183.980
355.640
85.900
~71,800
~57.700
343,600
419,500
315,400
601,300

145.600
368.400
9r.aoo
%14,000
736.800
859.600
982.400
9j.110
55.1lo
186. 40
a41,%60
334.880
~9o.100
483.310
123.140
246.180
369.410
492.560
615.700
7380840
861.980

---Brooklyn

Rapid
TransitElevated, sepes motor car and
standard traller.

Boston-Combriae

Subway..

...

......

1268

1089

.......

I455
719
1455
1213
1274
1164
I213

1433
iI9
1433
1200
I 360
1153
1200

315

50.1

:.
49.9

5tj

*w
-.............
0

465
233
665
390
609
375
390

60.4

57.9

60.4
48.6
51.8
45.6
48.6

57.9
44.7
48.4
41.6
44.7

?4

1335

1193

313

58.0

55.6

'

zCn

TABLE
OF C O ~ ~ A P A T N
DATA
E POP VARIOUSELEVATED
AND SUBWAYC A R S . - ( C O ~ C ~ U ~ ~ ~ )
Weight in pounds
Total
Road and class of

Hudson & Manhattan-original

type. Tunnel service.

Hudson
&
ManhattanPennsylvania Newark service.

Light

(seated and
standing)

i4.100
148.200
222,300
296.400
370.500
444.600
~18.700
592.800
73.000

96.220
xga.40
288.660
384.880
481.100
577.310
673.540
769.760
95.110
190,240
185,360
380.480
475.600
570,720
665.840
760.960
ror.1~4
59.040
~02.248
261,288
362,412
421.452
130.440
260,880
391.320
521,760
6s~.200
782,640
913.080
1.043.520

/=
zr9.000
292.000
365.000
438.000
51 1,000
584.000

17940.000
.564
Long Island Railroad SubwaySuburban Flatbush Avenue dlvision.

159.128
199.128
278,692
718.692
--104.400
208,800
313.100
Long Island Railroad Subway417.600
Suburban Pennsylvania station.
sza,ooo
626,400
i30,800
hC.200
See notes on page 514.

Per foot

With max.

service

Per cent
welght on
drivers

Per max.

. capacity)
'

IS33

1684

469

56 4

'.

54 9

1510

1658

462

54 1

53

I
I

1530
715
1530
I245
I315
1190

516

61 o

860
ISSO
I335
I390
1290

448
466
434

61';
48 7
52 2
45 6

i8:i'.
45.4
49 a
42 2

1620

14iO

562

58 7

56 4

ISSO

::%

(flerofed Cors

/rwSv,rs)

.-..sz.',t. -.-...----------...-....-..

HUDSON AND H A N U A n A N

518

ELECTRIC RAILWAY HANDROOK

the heavy and light hours, respectively. Figures 20 to 24, inclusive,

show typical floor plans with door and seating layouts of various

FIG.2s.-Outline

plans for interurban passenger cars.

types of city rapid transit cars, and the tables on pp. 512 to 516,
inclusive, give various data relative to the weights, capacity, motor
equipment, etc., of these cars.
Interurban Cars. I n the design of interurban cars i t usually is desirable for the comfort
of the pasenger on the longer
trips to provide a full equipment of cross seats. I t is genb
l
z
z
=
J
erally deemed desirable also to
provide a separate smoking
compartment and a compartment for baggage and light
express matter, the necessity
or desirability of these comF---+-7
partments becoming greater as
the h a d w a y becomes longer
a n d the distance between terminals greater. I t is also desirable on long runs, and in
some states, compulsory, that
toilet facilities be provided.
The layouts of several cars for
heavy interurban service are
FIG.26.-Outline plans for freight and shown by Figs. r g , 16, 17 and
express cars.
18; the general dimensions and
weights of these cars are shown in the table on page 5 0 9
A Joint Committee on Transportation Engineering, A.E.R.E.A.,
found that when considered with regard to entrances there are eight
different general styles of interurban cars designed for train opera-

TYPES OF CAR FRAMING

519

tion. Those shown by Fig. 25, a to e, inclusive, are the most cornmon types; a, b, and c are double-end and the others are single-end
cars. m e r e there is any possibility of train operation, end doom
should be provided. Swing doors are preferable to sliding doom
for all but the baggage compartment because they allow more
rigid and strong bulkhead construction, cause less trouble and have
a lower maintenance cost.
For light interurban service several companies have adopted
cars as light as 15 to 18 tons, fully equipped, with layoutsand
dimensions similar to those shown for suburban service by Figs.
3, 1 2 , 13 and 14;the general dimensions and weights of which are
shown in the table on pages 508 and 5 0 9
Freight and Express Cars. Where shipments are made in less
than carload lots there should be two large doors on each side of
the car. Fig. 26 shows typical outline plans for freight and express
cars. C a n a, b, and c are motor cars and the others are shorter
designs for trailers. Car a, having one large and one small door on
each side, is satisfactory when the car is not so long that time is
lost in carrying freight between door and end of car. In b the small
end doors facilitate the handling of long pieces of freight. Car G
affords the most rapid handling of freight. In d, eand f the corner
doors permit assage between cars. Car / is rovided with an
end door for tge handling of automobiles and o J e r bulky freight.
Types of Framing. The various types of framing a t present in
use may be described briefly as follows:
First. Wooden framing with sills reinforced with steel plates;
this type is now practically obsolete, having been superseded by the
more extensive use of steel for the principal members, a scientific
application of pressed and commeraal shapes oi steel resulting in a
stronger and more durable frame.
Second. Composite framing using steel plates for the lower
panels, wood or thin steel for the upper side anels, and wood for
the posts and superstructure. A heavy steer plate lor the lower
outs~depanel of a car is desirable from a maintenance standpoint
on account of its durability and the resistance it offers against collisions and abrasions. With this type of construction a light wooden
superstructure is the most suitable, the upper side panel being of
poplar, agasote or thin steel, applied in sections so as to facilitate
renewals.
Third. A composite framing using built-up steel girder side
construction and steel floor framing with wood or thin steel as the
outside sheathing.
Fourtk. Semi-steel construction using thin steel sheets on the
outside of the body up to the window sill, steel floor members with
wood for side posts and roof construction.
Fifth. All steel body, including side sts with thin sheet steel for
outside sheathing and with a steel b o r late riveted to the top of
the pasts to form a girder of the side of tge car.
Sixth. The curved side body in which the side girder sheet is
curved as shown in Fig. 30. The widest dimension is a t the belt
rail, and the side posts are shaped in sharply. This gives pood
vehicular clearance with maximum width a t seat line and clearance

520

ELECTRIC RAILWAY HANDBOOK

a t the top of car to allow for side sway. The side posts between
windows extend only down to the belt rail and are not used to
support the roof structure. Wide steel end panels form corner
piers which support the weight and take the strain of the roof.
By the use of properly designed members in the steel construction
of the types last listed above, it is possible to build a light body, and
probably no other method produces the same strength a t a corresponding weight. An objection to these types of construction is
that if the thickness of steel is to be governed by the required
strength only, the material will be so thin as to be dented easily by
slight collisions. Riveting also becomes difficult, and unless corsitively eliminated, the life of the structure will
rosion can be
be com arativecshort. If steel of proper thickness to overcome
these ogjections is used, the weight of the car will be very greatly
increased. This em hasizes the extreme desirability of a design
r ! t frequent painting all steel parts so as to
to make accessible
eliminate corrosion. Some experimental construction has been

FIG.27.--Steel underframe of Boston 11ghtwelght double truck car.

made using monel metal and various forms of non-corrosive steel

for various parts. There is as yet no general agreement as t o the
advantages of such special material and in some cases inferiority
for structural purposes is considered.
Construction Details. Figure 27 shows the details of the steel
underframe of the Boston light weight doubIe truck car inchding a
section of the body bolster and showing the stiffening included to
care for couplers. Figs 2 8 to 31, inclusive, show details of upper
framing in four different cars and show various methods of combining
the steel frame and wooden parts. The Minneapolis car shown in
Fig. 31 uses a continuous steel channel combining side post stiffening and roof members.
Aluminum in Car Body Construction. Aluminum in its various
forms and alloys is rapidly becoming a n important material in car
construction. For use as stanchions and grab handles i t may be
obtained in the form of a hard drawn alloy which is quite rigid.
For electrical conduit i t may be obtained in a n annealed form. For
long unsupported stanchions i t is found that the I-in. size is not
quite stifl enough for the service, and usually i t is better practice to

ALUMINUM I N CAR CONSTRUCTION

521

use 1%-in pipe size for this purpose. Although aluminum retains
a fairly good finish after continuous use and eliminates the necessity
for paint~ng stanchions and
railings, it may tend, while
new, to blacken the hand of a
passenger. By careful clean?
ing after installation, this difficulty can, to a certain degree
a t least, be overcome. At
least one entire installation of
5~
aluminum air pipe has been
made, on a sample car built in
Chicago. Some difficulty was
\
experienced in cutting threads,
and it was also found di5cult
t o pull the pipe up tightly
enough to hold the air without breaking i t off a t the
threads. Some trouble was
anticipated due to the effects
of vibration, but after nearly
two years of service, no serious difficulty has developed.
Precautions were taken to prevent corrosion due to local
I
galvanic action where the pipe
enters steel or iron fittings, by
ainting the threads w i t h
'
Eeavy insulating varnish instead of the customary pipe
fitting compounds.
I n the form of the alloy
known as "duralumin," rolled
sheet aluminum is available
in heat-treated form, and develops a tensile strength of
50,000 to 60,000 Ib. per sq. in.,
which is ample for many car
structural requirements. The
alloy contains approximately
3 to 4% per cent copper, 0.4
to I per cent magnesium, o to
o 7 per cent manganese, and
$'
the remainder aluminum, with
some slight traces of iron and
silicon in the form of impurities. The ultimate production
of this class of material in the
form of heat-treated rolled sections may have a considerable
for car strutFIG. 18,-Construrt1on details, Baltltural p ~ q ? ~ s e s .
more llght we~ght slngle truck car.

-.

98
7

ECECTRIC RAILWAY HANDBOOK

--

PIG. 29.-Construction

details. Chicago llght weight double truck car.

CAR CONSTRUCTION DETATLS

523

Aluminum in a considerable range of alloys is available for

castings, and these are being applied quite generally and to a
constantly increasing extent for car body construction. The
castings may be heat treated, and in that form develop physical

PIG 30 -Construct~on deta~ls.curved s ~ d ecar.

properties which enable this material to be substituted for brass

and in some cases for malleable iron, a t considerable saving in
weight. On a recent lot of cars in Chicago, the controller frames
and tops were made of aluminum a t a saving of 150 lb. per car
and with no increase in cost.

524

ELECTRIC RAILWAY HANDBOOK

Painting. The Ekcfric Railway Journal sent out a questionnaire

to a large number of electric railways in 1923, requesting infor-

d H

cent, five to six years; and 4 per cent,

over SIX years. I t appears to be the
general practice to burn off paint from
wooden cars and to use a paint remover on steel cars. In reply to a
question on this subject, sixteen companies reported that they followed
this practice. Ten others reported
that they burned paint off, and seven
replied that they used a paint remover. Other replies stated that the outside paint is burned off and a aint
remover is used on the inside; t!at a
sand blast. together with a paint remover is used; that the wood is burned
off and the metal parts are sand
blasted; and that the panels are
burned off and a paint remover is
used on other parts.
Answers to a question regarding
the work done between general paintingsindicate that most railways touch
up and revarnish their cars between
what they consider general paintings,
without removing the old paint. Of
the forty answers received, thirty

Three replies stated that their prac-

Minneapolis car.

CAR PAINTING

they used practically the same process for both steel and wooden
cars.
An attempt was made to obtain information as to the extent that
spray painting was being used and the various parts that were being
painted by the spray system. Replies indicated that two-thirds
of the railways were not employing the spray painting system.
Of the remaining one-third that were using it, seven companies
replied that they were spraying the trucks and underbodies of can.
Two replied that they sprayed trucks only, two that they sprayed
platforms, hoods and trucks; two more that they sprayed box and
flat cars, and an additional two that they were spraying trucks
and work' cars. Of the remaining replies indicating the extent to
which the spray process was used, one company reported using the
spray method for the entire outside of the car. Another stated that
of the seven coats which were applied on the car, five of them were
applied by the spray method, and in addition, the roof, truck, pilot,
steps and bumpers were painted by the spray method. An investigation a t the shops of some of the companies which are employing
the spray method quite extensively indicates that the economies
result~ngfrom the use of this system are quite pronounced, and that
manufacturers are co-operating toward supplying paints and
varnishes which can be used and applied readily by the spray
method. S ray painting is particularly adapted for work in
restricted
and is economical in the quantity of material
needed.
Aside from the special systems which are used by some railways,
the various painting s stems can be classified in general into three
classes. These are ( r r the flat color and finishing varnish system,
(2) the color varnish and finishing varnish system, and (3) the
enamel system. Of the forty replies received in answer to the
questionnaire, fourteen railways re rted as using the enamel
system, thirteen used the color varnispoand finishingvamkhsystem,
seven used the &
ti color and finishing varnish system, two reported
using both the flat color and finishing varnish aud the enamel systems
and four reported the use of special systems.
After removing the paint or, where this is not entirely removed,
after the loose paint and that which shows signs of cracking has
been removed, those using the JWcolor and finishing varnish system
apply a priming coat. This is followed by a glazing or knifing
coat when holes and nailheads are puttied and spotted. In general
this is followed by two rough coats or surface coats. The number,
however, varies with different properties; some apply but one coat,
and others three. Next follow two or three color coats, and after
the lettering and striping usually two coats of finishing varnish are
applied. Some roads, however, use only one coat of finishing
varnish. Variations frequently will be found, as when cars are in
fairly good condition some of thevarious coats may not be necessary.
Where the color varnish and finishing varnish system is used, the
color varnish coats replace the flat color coats. Some roads use but
one coat of color varnish; others two, and some use a combination,
applying first a flat color coat followed by a color varnish coat.
After the striping and lettering, the cars are finished by the application of either one or two finishing varnish coats.

prates

526

ELECTRIC RAILWAY HANDBOOK

Where the enamel system is used, the enamel coats replace the flat
color or colored varnish coat. From the information received it
appears that the most general practice is to apply two enamel coats.
Some roads add a coat of finishing varnish over the enamel, and two
roads reported the use of a clear enamel coat over the colored enape]
coat. A large number of railways do not use finishing varnish with
the enamel system. There are various combinations of these
systems, as some roads use a flat body coat before applying the
enamel coat, and other combinations are possible depending u p ~ n
conditions.
As cars constitute the point of contact between the traveling
public and the railway company, it is essential that the gene&
a pearance of the interior and exterior be kept in a condition
to the eye. The use of paint to improve the appearance of
rolling stock is then a most important consideration. One of the
best mediums for selling transportation is to adopt bright, pleasibg
colors. Another reason for using brightly colored cars is that of
safety, and many companies adopt a combination of colors to give
the cars the greatest r i b l e visibility. In order to operate cqr.
a t high speed and a t t e same time to promote the safety of the
public, it is desirable that an approaching car should be distinguished
as far as the eye can see. .Other reasons for keeping rolling stock
properly painted are for bnitary effect and to improve llghtihg
r a d t k c ~Ja mdpr tna.wp~iainh ~ w f a ~ p a i nare
t s used by deshjc
railways for these two latter-named purposes, companies were
asked specifically if they used paint to improve lighting conditions
and for sanitary reasons. Eighty-three companies replied to ttlis
question, fifty-one of which (61.per cent) said that these were
im rtant considerations.
Lr Operation. In city and suburban service the operation of
the outside doors of both sliding and folding types generally is
some mechanical means operated from the motorman's or conductor's position. Such doors are frequently operated mecllaqically through an arrangement of shafts, rods and bell cranks or
bevel gears, and often are operated pneumatically by means of
door engines, consisting of cylinders with pistons properly connected to the operating mechanism, the air control being by
suitable valve a t the operator's position.
Car Couplm. The selection of coupler equipment is limited tIy
three conditions: (I) the short radius of track curves, ( 2 ) the
abruptness of changes in track grade, and (3) the constructic,n
and dimensions of cars. The center line radius of track Curves
should be as great as conditions permit, 40 ft. or greater if possible,
and never less than 35 it. If short radius curves must be used,
spiraled approaches offer effective corrective possibilities. The
greatest coupler swing on a given car and curve is when the car is
fully within the curve and the coupler anchorage of the adjacent
car is directly over the point of tangency of the track curve. The
amount of change in grade is of little importance if sufficient dlstance is available for making the change. I t is the change in gra9e
which must be made in a short length of track that affects coupler
clearance, Such conditions usually obtain a t railroad crossings

pressing

CAR COUPLERS
where street and railroad grades are permanently established a t
different levels, a t the approaches to some bridges, and a t points
where streets dip below railroads or other overhead structures.
Other special conditions must be considered such as temporary
crossovers, emergency hose bridges used a t fires, or transfer tables
with short approaches. As one car platform rises or falls with
respect to the adjacent platform, its coupler anchorage lifts or
depresses, and this movement affects the clearance between coupler
head and car buffer, as the straight line between anchorages must be
maintained a t all positions the two cars may assume. The projection of the coupler beyond car buffer when in center line
should always be su6cient to keep the hood rims of c o u p f $ ~ ~ ~
separated a t the bottom of the most abrupt track grade and should
n e v r be less than one and one-half times the coupler draft gear
travel. Coupler selection is dependent further upon car dimensions
and design. The coupler equipment demands a certain space under
the car platform, and the nature and importance of its service
requires that this clearance space may not be invaded even by such
important items as car steps, brake apparatus, fenders or door
operating mechanisms. When truck center spacing, as well as the
length of car, is fixed, the only two variables left are coupler length
and anchorage position, and this leaves small opportunity to secure
a favorable maximum swing of couplers or entirely acceptable
clearance space for car steps. 12re contour of car ends should
be such that coupled cars be separated as little as possible. The
buffer radius should approximate the coupler length and preferably
be slightly less than swinging length of coupler. Such platforms
r m i t close spacing of cars and im rove comer clearance on curves.
he stresses on car underframes wgen cou led in trains differ from
those of single car o eration, and sills g r anchorage mounting
should be capable o f withstanding and transmitting the loads
i m r d in normal operation and the shocks incident to coupling.
utomatic Couplers. The automatic coupler should have the
following qualities: (I) automatic cou ling without necessity of
reliminary adjustments; (2) simple anisafe means of uncoupling;
ample vertical and horizontal alining ability of cou ler head
in coupling; (1) no loose parts subject to loss or misp~cement;
( 5 ) ability to operate under all weather conditions and after long
periods of idleness; (6) freedom from relative movement a t coupler
faces; (7) strength to withstand abnormal operating conditions and
abuse; (8) provision for free vertical movement as well as horizontal
swing of coupler; ( 9 ) provision for limited torsional movement to
permit train to travel curves having elevated outer rail; (10)
simplicity of construction; (11) ease of repair. Unless interchange
of traffic demands a coupler of the M.C.B. type, the following
points should be added: (12) automatic coupling of air brake pipe
lines, and (13)
-provision for attachment of automatic electric
couplers.
The Tomlinson automatic coupler is one of the widely used types.
One of several modem forms is shown in Fig. 32. This coupler
consists of a cast steel head or drawbar fitted a t its face with a hardened steel forging in the form of a hook which constitutes the chief

r3)

528

ELECTRIC RAILWAY HANDBOOK

guide member for dining purposes during coupling, and by

snapping into engagement with a similar hook when coupler faces
are brought together acts as the tension member in haulage.
Projecting dowel pins at the coupler face serve to aline the air line
rubber gaskets which are above and below the hook on the center
l i n e of the coupler
face. A heavy draft
I
-----J--->
spring is housed withy..,
- -----I----,.=I?
in the enlarged rear
end of the coupler
body and a b s o r b s
both tension and compression shocks. The
coupler proper te-minates a t the real end
in a two-piece socket
clamp which encloses
the machined ball of
the coupler anchorage.
The two halves of this
coupler socket interlock to form a retaini
ing cup for lubricating
I
oil contained in a well
/
/
within the anchorage
ball. This ball and
/
socket joint provides
in one bearing for all
'------.>I
the necessary move--------I---ments of coupler head.
Adjusting bolts and
shims ermit the socket to f e adjusted for
any wear which may
occur. The uncoupling chain hangs from
the end of the projecting cam handle.
The carrier supports
the projecting coupler
head, when uncoupled,
a t a proper height to
FIG.32.-Typrcal coupler rnountrng on car.
facilitate c o u p 1i ng,
and is supported by
s rings which permit the coupler head to move downward, with
crearance above to permit lifting of the coupler from its spring
support. An overhead supporting bar, bent to curved outline, permits the carrier to swing to either side on track curves.
An efficient device for establishing electric circuit connections
between cars is an essential feature of modem automatic coupler
equipment. The most simple condition of motor car and trailer
operation requires the extension of buzzer and single stroke bell
I--

--

-.

LH/'

CAR COUPLERS
circuits from the motor car for signaling gurposp,
well as a
lighting and a heater circuit. Usually door s~gnalcucults are added
to increase speed and safety of operation. The operation of multiple unit motor cars in trains has always required some form of
multiple conductor jumper cables, the functions of which may be
combined with electric couplers. The electric couplers, as shown
in Fig. 33, operate on the push-button principle. The electrical
contact parts are insulated in molded composition blocks which are
mounted in metal cases and bolted directly to the machined side
faces of the coupler head. Their proper operation is assured by
perfect preliminary alinement and the straightforward coupling
movement of the mechanical couplers. This method of mounting
has proved to be most practical, as it combines easy accessibility
for installation and inspection with maximum track clearance. I t
is general practice to retain cut+ut cocks in the air brake pipe lines
back of the hose jumpers leading to coupler, and to insert disconnecting switches between car wiring and electric couplers. The
fact that both air line cocks and electric coupler switches must be
operated whenever cars are coupled or uncoupled leads to an
obvious combination and interloclung of these parts which simplifies
mounting arrangements and enforces their proper operation. The
disconnecting switch has a rotating drum, operated by pull rods
sttached to the ends of an external operating handle, which makes
and breaks contacts with metal fingers arranged on each side of the
drum. At each end this drum terminates in a socket which serves
to throw the air cock which is mounted between pipe lugs on the
switch end. The mounting is such that both air cocks are open
when the switch is closed and closed when the switch is open. By
this arrangement exposed electric coupler contact points on uncoupled couplers must always be dead, as the car could not operate
with the switch closed because the air cocks would then be open and
the air brake would act automatically to lock the wheels. The
disconnecting switch must be mounted in some accessible space
convenient for air piping connections and the attachment of
switch handle operating rods, and preferably it should be located
where wheel wash will be avoided and periodic inspection can be
made conveniently.
M.C.B. Type Couplers. There has been adopted as standard
by the A.E.R.E.A. for use on interurban cars a coupler of the vertical plane type which has the same contour lines of knuckle and
guard arm as, and which will automatically couple with, standard
steam railroad couplers of the M.C.B. type. The draft rigging
and drawbar supports for these couplers should be such that,
with sudden changes in grade, the vertical displacement of the
couplers with reference to each other will not be sufficient to
cause the knuckles to become disengaged. The length of the
radial coupler, measured from the center of the pocket-pin to
the pulling face, should be 54 in. This length will apply to both
interurban cars and to cars in city service. On city cars, where the
bumper arrangements will permit, a pocket casting should be placed
on the top of the bumper, the center of the pocket to be 35 in above
the top of the rail, and the casting to be of ample strength and

530

ELECTRIC RAILWAY HANDBOOK

properly braced, so that by means of a suitable bar city cars may be

coupled on a level with the automatic couplers of interurban cars.
For this purpose it is advisable a t resent to maintain a link slot
and coupling pin hole in the k n u k e of the automatic couplers.
The following are the A.E.R.E.A. standard specifications for
couplers for interurban cars where the interchange of equipment is
involved:
I. All couplers must be made to couple automatically by impact, and to uncouple without the necessity of going between the

PIG. 33.-Electrical
train line
connection (14 point) for automatic
car coupler.

. .
PIG. 34.-Contour
line,
standard coupler.

M.C.B.

cars, with M.C.B. and all other types cf M.C.B. contour couplers,
whether used by steam or electric roads.
2. A device should be adopted for holding couplers on- center
within the required limits when inter-coupl~ngw t h steam railroads.
3. An open knuckle for shackle bar connection should be used.
4. The draft gear, where possible, should meet M.C.B. requirements, and the drawbar anchorages should be equivalent in
strength to M.C.B. equipment and requirements.
5. Couplers must not uncouple when cars are being pushed
around a curve of 35-ft. center radlus.
6. There should be an arrangement to release and open the
knuckle without requiring the operator to pass between the cars.
7. The face of the knuckle vertically should be 16 in. maximum.

CAR COUPLERS

53 1

8. The height of the drawbar center should be 31% in, minimum and 34% in. maximum above the head of the rail.
9. The coupling center of couplers must have a minimum projection of 6 in. beyond buffer faces a t any point between the
working limits of couplers.

recommended location of end connections for

interurban cars.

ro. Couplers should be placed on both ends of the can.

The contour line for automatic couplers adopted as standard by
the Master Car Builders' Association is shown by Fig. 34.

532

ELECTRIC RAILWAY HANDBOOK

Standard Heights of Couplers, Platforms and Bumpers. Following are A.E.R.E.A. standards. Height of city car couplers, 2 0 in.
from top of rail to center of coupler. Height of interurban car
platforms, SI in. above top of rail. Height of bumpers for interurban cars, 43 in. from top of rail to bottom of bumper; for city cars,
31 in. from top of rail to top of bumper; width of city car bumpers,
6 in. Bumper on city cars to include pocket to permit coupling
with interurban cars. Where possib e, provision to prevent
telescoping in case of collision between city and interurban cars.
End Connections for Interurban Cars. The locations of the various end connections for interurban cars, adopted as recommended
practice of the A.E.R.E.A., are shown in Fig. 35. I n this figure
the locations shown for control train and bus line electropneumatic
and trailer Iight receptacles are such as to clear end doors where these

MOTOR

CAR

TRAILER

Sbnub,with m c s w r y r r s ~ s b ~foebe
, locuted soas rb k audtbk on tofh
s,&s of 6o/khrud5

o Signul line nnnnfors.as per prin f Abb-II-A.

E

Tmilrr li hf connectqrs asper print NaC-5-A, u d - l b r Iighf he&r rrnd

other ou%liory c1m1lS'
Signulsws oprmlW by bell cord.od connrcfing tmiler Nghf w;fh s w
/me bur

PIG.36.-C.E.R.A.

standard location of signal whistles and wiring.

are employed in the event conditions permit; if it is otherwise required, their position may be altered, but the re-location should be
on the vertical center, as shown. The receptacles indicated by
dotted lines are optional when not required; the trailer light
receptacles are optional when bus lines are used. The headlight
receptacle and bracket and trolley retriever bracket are optional
when not required.
The Standardization Committee of the Central Electric Railway
Association has proposed a standard location of electropneumatic
signal whistles and wiring, as shown by Fig. 36, and also a standard
trailer light connector, as shown by Fig. 37.
Miscellaneous Equipment for Interurban Cars. A Committee
on Maintenance and Inspection, A.E.R.E.A., recommended that
interurban cars should never be put in service without the following
miscellaneous equi ment :
Three sets of fags (red, white and green); telephone, where
standard to the road; classification and marker lam s where oil
h m p s are used; two trolley pickups; one coupler or pufimg bar; one

TRACK SANDERS

533

ull rope; coupling link and pin; extra supply of air pump-and
&ht fuses (also car and control fuses where used); fire extlngulsher
in working order; one extra trolley pole, fully equipped, on top of
the car; one extra trolley rope, or, better still, one extra retriever
equipped with rope; one trolley retriever in its place on the rear
dash; one headlight in its place on the front dash, for signal use.
Fuses and torpedoes should be on each car. The crew should have
both red and white Ianterns in good condition, and sufficient tools
to change the trolley pole or make other light repairs. Interurban
cars should also carry a wrecking outfit with axe, saw, jack and
crowbar.

FIG.3'1.-C.E.R.A. standard trailer light connector.

Track Sanders. Satisfactory results from the use of sand will

depend very largely upon the efficiencyof the sanding device itself
and the character of material with which it is supplied. Owing to
the restricted conditions on an electric car, it is usually very difficult
to install the older types of mechanical gravity sanders with anything like a satisfactory arrangement. This difficulty is a h augmented on account of the very sharp curves and other conditions
which usually prevail or a t least restrict the operation to some few
particularly bad locations. This has led to the general use of pneumatic sanders as being superior to the gravity type in that they are
more readily applied and have the advantage of distributing the
sand evenly and expeditiously a t the proper point on the rail
directly ahead of where the front pair of wheels makes contact with
the rails. The sand should be applied only in sufficient quantity to
give maximum traction and braking power, and it is especially
important that the application should be just previous to or a t the
time of the application of the brakes and before the braking power is
high enough to skid the wheels. This applies particularly to emergency brake applications and is one of the spec~aladvantages of the
pneumatic type over any of the gravity sanders. The pneumatic
type, however, requires very careful installation and arrangement of
the piping connections in order to insure reliable and positive results

534

ELECTRIC RAILWAY HANDBOOK

when called upon under the most trying conditions of weather and
roadway. I t a h is necessary to use a device requiring the
minimum amount of air and sand as well because of the tendency to
overload the compressor with minor pneumatic devices which were
not taken into consideration when the capacity of the air compreswr
was determined. The flexible connection between the sand box,
usually carried inside the car, and the discharge pipes attached to
the trucks is usually a source of considerable trouble and requires
careful attention. The style of the sand valve should be such as
to avoid useless waste of air. I t should also be located conveniently
near the brake valve so that the two operations of applying sand
and setting the brakes can be done in emergency
a t practically the
same time:
Character of Sand. The best quality of good sharp quarfi
sand, thoroughly dried and screened and free from dirt or soil,
should be used. Dirty sand is more susceptible to moisture and
consequently its tendency to clog up the pipes is greater, aside from
the harm it may do after reaching the rail. Lake sand has been
found quite satisfactory and is extensively used in pneumatic
sanders on account of its fine, even grain and freedom from foreign
matter. I t is also easily dried and screened and is generally economical. The character of the sand is worthy of more important
consideration than apparently is usually given to it. In some cases
it is entirely lacking in the essential qualities and has a tendency
not only to defeat the object for which it is applied, but actually to
assist in creating a more serious condition.
Car Heating. The usual methods of heating cars are ( I ) coal
stoves, direct method; (2) coal stoves, indirect method, similar to
hot air furnace used for heating of houses; (3) hot air heaters, air
blast, motor driven; (4) hot water heaters; (5) electric heaters.
The characteristics of the various methods should be considered with
regard to the following points: (I) ability to heat car to uniform temperature; (2) first cost, completely installed on car; (3) maintenance;
this will include repairs, renewals, replacements, etc.; (4) cost to
operate; in the case of hot water heaters, this will include fuel and
labor only; in the case of electric systems, power only, and in the
case of the hot air blast heater using coal, fuel, labor and power;
( 5 ) weight of system complete as installed on the car ready to
operate; (6) fire risk; (7) reliability; (8) regulation; this refers to
abllity to regulate the heat to outside temperature; ( 9 ) space occupied; (10) appearanfe; (11) attention required from car crew; (12)
cleanliness, whlch wlll include freedom from dust, ashes and obnoxious gases; (13) adaptability and relation to ventilation systems.
Aside from the ordinary coal stove, the three principal heating
systems are the hot air heater, hot water heater and electric heater.
Hot Air Heater. In this system the air is heated by a coal
fire and forced through suitable ducts along the side of the car by
motor-driven fans. By its use it is possible to secure quite unifom
heating of the car. The heat being applied along the floor line
results in dry floors, which is a strong point in its favor. The
first cost is about the same as that of an equivalent electric system
wired in conduit and equipped with thermostatic control. The cost

CAR HEATING
to operate, including only the items of coal, labor of attendance and
cost of electricity for the motor is comparatively low. The weight
depends upon type and size of car, but is practically the same as
electric, and less than hot water heaters. The fire risk is practically
the same as in the case of hot water heaters. The regulation is not
so good as it should be, but will undoubtedly improve as the apparatus is further develo ed. When this system is used in conjunction
with exhaust ventfation, the regulation is better. The space
occupied is a little greater than with hot water heaters. The
appearance compares favorably with other types of coal heaters.
Considerable attention is required from the car crew from time to
time in order to keep fire in proper condition, but where the heater
can be placed near the conductor or motorman, this is readily accomflished. I t is not so clean as the electric system, but compares
avorably with hot water. As this system is designed to provide
for ventilation, it is readily adapted to that end.
Hot Water Heater. This system possesses many valuable
characteristics, among which are independence of the electric
power supply, which is quite a consideration in interurban work
where long runs are made and the power supply is subject to interruption. By the use of this type of heater it is possible to heat
the car very uniformly. The efficiency of hot water heaters will fall
off materially if the pipes and coil are not kept reasonably free
from scale and other deposits. The first cost is greater than that of
the hot air system. The maintenance is higher than that of the
electric system. The cost to operate, including only the items of
fuel and labor, is approximately the same as for the hot air heater.
The weight of the hot water apparatus is high and has long been one
of its chief drawbacks, but the latest types show improvements in
this respect. The fire risk is substantially the same as for the hot
air heater. In case of accident there is a hazard from the fire in the
coal stove. The reliability is very good. The regulation is comparatively poor. I t takes some time for water to take up heat and,
conversely, it takes some time for water to lose its heat. The space
occupied is considerable and, except on single-end cars, this space
is valuable as seating or standing room. The hot water heater
with its expansion drum, water glass gage, etc., does not add to the
appearance of a car, except where it is practicable to partly enclose
the apparatus. Attention is required from time to time, but the
work is small and where the heater can be located close to one of
the crew, it does not take him from other duties. The hot water
heater as usually installed produces dust, and very frequently
obnoxious gases. The heating elements being pipes located one
above the other near the floor line, it is easy to adapt this system
of heating to any practical scheme of ventilation.
Electric Heater. I t is perfectly possible to secure uniform
temperature throughout the car. The first cost is lower than that of
any of the other modern systems. When the electric heater is
carefully installed with wiring in conduit, the maintenance is very
low, being considerably less than that of any of the other modern
systems. The cost to operate, which includes energy only, is vwiable, depending on the size of car, the range of temperature, the

ELECTRIC RAILWAY HANDBOOK

coat of energy, and whether the peak load on power stations comes
in the heating seasons or not. In general, the cost of operation, as
defined above, is high. There are many cases, however, where this
method of heating will show the greatest net economy. The weight
is the least of any of the modem heating systems, except in some
cases, where it is practically the same as the hot air heater. Where
the wiring is properly installed in conduit the fire risk is practically
nil. This system is very reliable. The regulation is best of all,
it being possible to follow closely and without trouble rapid changes
in outside temperature. The space occupied is verysmaU and is not
useful for standing or seating capacity. The appearance of such
parts as are exposed is very good. The electric heating system
requires the minimum amount of attention from car crew. This
type of heater is clean and free from dirt or obnoxious gases. The
heating units being subdivided and located under the car seats or
along the truss plank are readily adaptable to any practical system
of ventilation.
In the installation of electric heaters, it is preferable to have a
comparatively large number of heaters rather than a few, even
though t h e r w e r consumption is on the same basis, on account of
the better istribution of the heat. For localities where the temperature reaches zero or lower, it is well to have about 4.5 watts per
cubic foot of car body, otherwise it may be difficult to maintain a
comfortable atmosphere in the cars when low temperatures prevail.
Comparative Costs of Car Heating. As a comparison in costs
of heating a car by the three modern systems, the following estimates are given. The figures are based, in general, on results
obtained in practice.
BASICDATA(ALL SYSTEMS)
Length of car body to be heated (inside approx.
40 ft
Width of car body to be heated ([usidel approx.
8 ft. 3 j
a
7 ft. 6 m.
H a g h t of car body to be heated (mslde) approx.
Cubical contents of car body to he heated
(inside). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2400 CU. ft.
Weight of car without passenger load.. .......
34.000 Ib.
150 ml.
Average daily car mileage.. .................
Average season car mleage.. ................
22.500 ml.
ISO
days
Average length of heating season.. ...........
Average temperature during Oct.. Nov. and
Apnl.. ..................................
38 deg. P.
Average temperature during Dec.. Jan. and
Feb.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
24 deg. P.
Average lift of temperature above outside air.. ..
35 deg. P.
6 -s dm.
A v e r a ~ etemnerature t o be mainta~ned... . . . . .
- P.
Heat &its r k u i r e d per hour per cu. ft. of air to
be heated*. ..............................
I3
(above equals a heater capacity of about).. ...
10 kw.
Kw-hr. consumption per car mrle (a-motors
city service). ............................
2.15
Watt hours per ton mile.. ...................
132
460 Ib.
Weight hot air equipment, including air duct..
Weight hot water equipment, including piping
and water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1600 lb.
Weight electric heating equipment, including
anring and conduits.. . . . . . . . . . . . . . . . . . . . . .
Cost of coal per net ton a t car house.. ........
$0.01
Cost of current per kw-hr. a t c a r . . ..........
$0.015
Cost of haulage per Ib. per season of 150 days.. ..
Includes allowances for all rad+tion and air change losses; also for
ine5iciency of heaters due to re-heatlug through operation of thermostat.

b?.:;6

CAR HEATING COSTS

COST OF HOTAIR HEATINGSYSTEMPER SEASONOF 150 DAYS
Interest c h a r g e d per cent on $zoo per complete
equipment, installed . . . . . . . . . . . . . . . . . . . . . . . . . . .
Depreciation charge-7 per cent on $zoo cost per
equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Maintenance per equipment @ 0.007 per heating

hour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Attendance. including kindling of fires, removal of
ashes and handling of coal from bin t o car @ 12c
per day . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Current cost-to operate blower motor = 275 watts
per hour X 18 hours per day X ISO days X $0.01
perkw-hr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Fuel cost-full
heat for 90 days = 80 lb. hard coal
per day X 90 days = 3.6 tons X $13.00 per ton.. . .
Fuel cost-%
heat for 60 days = 60 Ib. hard coal
per day X 60 days = 1.8 tons @ $13.00.. .........
Fuel cost-banking fires 6 hours pcr day 15 Ib. hard
coal per day X 150 days = I ton hard coal @
$13.00.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
K i n d l i n g a t $0.03 per day for 150 days.. .........
H a u l a g e 4 6 0 lb. X $0.015 per lb. for 150 days.. ...
Labor--removing heater a t termination of heating
season and replacrng ~n fall.. ...................
Average cost per day for season of 150 days. ...
Increased insurance mte on car barns and
rolling stock is zc per $100.

$173.23
$ I . 16

COSTOF HOTWATERHEATINGSYSTEM
PER SEASON
Interest c h a r g e d per cent on $300 per complete
equipment installed.. . . . . . . . . . . . . . . . . . . . . . . . . . .
Depreciation c h a r g e 7 per cent on $300 per comp!ete equipment installed.. .....................
Ma~ntenance---per equipment @ $o.oog per heating
hour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Attendanceincluding kindling of fires. removal of
ashes and handling coal from bin to car @ b . r a
per day X 150 days.. .........................
Fuel cost-full
heat go days = go Ib. per day X
90 days = tons @ S13.oo.. ...................
Fuel cost-34
heat 60 days = 65 Ib. per day X
60 days = 2 tons @J $rg.oo.. ...................
Fuel cost-banking f i r e d hours per day X 15 Ib.
hard coal X 150 days = I ton @ $13.00.. .......
Kindling-at So.ox per day for 150 days.. .........
Haulage-1600
lb. (approx.) X $o.ors per Ib. for
150 days . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Labor--removing heater a t end of heating season
and replac~ngIn fall.. . . . . . . . . . . . . . . . . . . . . . . . . .
SIPS.50
Average cost per day for season.. . . . . . . . . . . . . . . . .
81.so
Increased insurance rate on car barns and roll~ng
stock is zc per $roo.

538

ELECTRIC RAILWAY HANDBOOK

COST OF ELECTRICHEATING
SYSTEMPER SEASONFOR 150 DxYS
Interest c h a r g e d per cent on $175 cost of 10 kw.
electric h e a t ~ n g complete, installed. including
wiring, condu~tsand thermostatic control.. . . . . . . s1o.50
NOTE: 10 kw. equipment is approximately equivalent
in heating capacity to the hot air and hot water
equipment covered by above tables.
Depreciation c h a r g e 7 per cent on $175 per
12.25
equipment ....................................
Maintenance--$o.oos per heating hour on 1775 hours
operating per season.. . . . . . . . . . . . . . . . . . . . . . . . . .
8.88
14.80
Haulage cost- 00 lb. @ $0.037 per Ib. for 365 days. .
Current cost-jull
heat = 10 kw. X 18 hours per
day X 90 days X $0.01 per kw-hr. i 80 per cent
162.01
time heat is on full.. . . . . . . . . . . . . . . . . . . . . . . . . . .
Current cost-76 heat = 10 kw. X 18 hours per day
X 60 days X $0.01 per kw-hr. + 60 per cent time
heat i s o n . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64.80
Current cost-To heat u p car before taking out of car
house in morning = 56 hour per day X 10 kw. X
1 2 0 days X So.01.. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.00
NOTE:This item must be included to compare with
hot air and hot water heaters in which fire is either
maintained all night or built u p early in the
morning before car leaves barn.
Average cost per day per season of ISO days..

....

$279.24
I .86

From the above i t is seen that, under the conditions assumed, the
ze\a<>vtt o b l wonomy 05 tht thtt pI;incipa\heafn~
sysktns is >%
follows: Hot air system, first; hot water, second; and electf'c~
third. If the conditions are different from those taken as typical,
different results will be obtained; for example, if, under the above
assumptions, the peak load on power stations came in summer tipep
then the electric heating system might show the greatest total
economy, due to the fact that in the latter case the proper chafge
for energy would be only the fuel cost, while if the peak comes In
the heating season, the charge for energy should include all filred
charges on power plant and distribution system. I t is, therefore,
readily apparent why i t is impossible to say off-hand what the cost
of car heating is unless all the conditions are known. As stated
before, in the choice of a system for heating cars some consideratton
other than that of total economy may govern, such as appearance*
space factor, ability to regulate, 6re hazard, or the requirements of
the State or municipality.
I n figuring the power consumption of electric heaters, the f?llowing method will probably give the most accurate results: ObtBln
from the weather bureau temperature readings for each winter
for several years. Plot a curve showing variation of temperato"
for each day of the heating season. Find what circuit in heaterf.?s
used for various temperatures and plot a power curve which ~ 1 1 1
indicate the average kilowatts per day.
I n the use of hot water or hot air heaters there is a tendency on the
part of the car crew to use less coal than is necessary to keep the
cars a t a uniform temperature during the time they are in service,
while with electric heaters the tendency is to put on three poiPts
where two points would suffice. This may give rise to false
ideas of the relative costs of the various heating systems.

CAR I-IEATING THERMOSTATS

539

Thermostatic control of Electric Heaters. A number of devices

have been designed for the automatic regulation of electric heaters.
the successful operation of which very greatly reduce the cost of
electric heating, in some cases the tests showing a saving of some 50
per cent. One type of thermostatic control is illustrated in Fig. 38,
by reference to whlch the operation of the control circuit is explained
as follows: Normally current flows through resistance A , the magnet coil of switch E and resistance B to ground, and switch E is held
closed. Current then flows direct from the trolley through the
blowout coil and the contacts of switch E and then through the
heaters to the ground. As the car warms up, the column of mercury
in the thermometer rises. When i t touches the upper platinum
contact, a current of very low voltage is shunted around resistance
B, through fuse C, the coil of relay D, and the thermometer, to

Platlnnm Contact

Yuw

II

mermometer

Rerlstaneo

I I
I I

I
Y,

FIG.38.-Electric

to Orolmd.

thermostat control for car heaters.

ground. Relay D is thus energized and its contacts dose. This

short-circuits the magnet coil of switch E, and that switch opens by
gravity, cutting current off the heaters. When the temperature
falls so that the mercury of the thermometer leaves the platinum
contact, the thermometer circuit is broken, and relay D is no longer
energized. Therefore, its contacts open, breaking the shunt around
the coil of switch E. The latter is then energized and closes, turning current on the heaters. A rise and fall of less than I deg. is
sufficient to turn the heaters off or on as required.
Before the adoption of thermostatic control for electric car heaters
in Chicago, tests were made to determine the probable energy
saving. Of the three types of regulators tested, two utilized a
mechanical expansion element operating a relay circuit through
which the main heater switch was controlled; the third a mercury
contact thermometer with relay circuit. One cf the types tested
possessed the advantage of double adjustmect so that can might be
partially heated when standing in the yard just before being put

540

ELECTRIC RAlLWAY HANDBOOK

into service. Another was so wired that current remained on the

heating circuits in case of any failure of the thermostat and thus
was under ordinary control by trainmen. Three of the testcars
had monitor roof with natural ventilation and three arch roof and
motor-driven ventilators to eliminate as much as possible the
effects of variance in car body design and ventilation. The current
used in heating these cars was compared with that used in standard
cars of the respective types not equipped with the thermostat.
With the latter, an effort was made to conform to the ordinance
requirement of 50 deg. F. as closely as possible, although it was
found that more heat was ordinarily used than authorized by
the operating department due to the impossibility of close adjustment by hand. The amount of energy used by the heaters represents in approximate figures from %O to f5 the average energy
requirement of the car for propulsion alone. During the coldest
winter weather this heating load averages per car as follows:
1st oint, 3.43 kw.; 2nd point, 6.87 kw.; 3rd point, 10.30 kw.;
resuiing in an increase in total peak load a t the power station from
15 to 2 0 per cent above non-heating requirements. The tests
showed that the average saving over non-automatic control, when
expressed in per cent, varied from nothing a t 15 deg. F. where full
heat was required, to a maximum where no heat was required,
viz., 50 deg. F.:
Natural omtilation, temperature zo dcg.. average
temperature 30 deg.. average
temperature 40 deg.. average
temperature 50 deg.. average
Forced venfilalion, temperature 20 deg.. average
temperature 30 deg., average
temperature 40 deg., average
temperature 50 deg.. average

saving t r
savlng 40
saving 56
saving 69
savlng 8
sanng a3
saving 44
saving 65

per cent
per cent
per cent
per cent
per cent
per cent
per cent
per cent

As some trainmen kee heat on above 50 deg. the results actually

showed some saving agove the 50 deg. limit. Expressed in actual
energy, the results were fairly constant throughout the range of
the test:
Natural ventilation. saving a . a kw. per car average
Forced ventilation, sanng I . a kw. per car average

Ventilation of Cars. Authorities d 8 e r as to the amount of air to

be supplied per person per hour, in order to provide a reasonable
standard of air purity. An ordinance of the city of Chicago on this
subject calls for the supplying of 350 cu. ft. per hour per passenger
(based on maximum standing and seated load); provided, however,
that the air in the car shall a t no time show more than 1 2 parts of
COI in I O ,parts.
~
I t was found possible to meet this requirement
by any one of several ventilating systems.
Monitor Deck Windows. The usual method of obtaining ventilation in electric railway cars has been to provide a monitor deck or
clear story in which there are a number of small windows that can
be o ened. The ventilation afforded by these small windows is
largeyy by dilution; that is, such air as may enter serves to freshen
the air in the car. The chief objection to the system is that, under
some conditions of operation, strong draughts are created which
are objectionable to passengers.

CAR VENTILATION
Types of Ventilators. Ventilation systems other than monitor
windows have been worked out principally along two lines: those
operated by the movement of the car through the air and those
operated by motors. The first are usually called the automatic
systems and the second, mechanical systems. Automatic ventilators are usually "exhauster" and should be so designed as to exclude
rain and snow and to prevent gusts of air coming into the car.
The action of automatic systems is, of course, variable, depending
on the velocity of the car and the direction and force of the wind.
The mechanical systems, of which there are two principal kinds,
the "exhauster" type and the "blower" type, are positive and
practically inde endent of motion of the car or velocity of the wind.
Mechanical ?&stem of Ventilation. A typical mechanical system consists of a motor-driven exhaust fan located on the vestibule roof, or other practical point, a n exhaust chamber formed in
the upper ceiling of the car, openings located a t various points in
the ceiling and intakes located a t several points in the floor and
connected to the electric heaters. The cold air thus is made to pass
over the heating surface before coming into the car body which is
generally appreciated as a very desirable feature. Tests show that
consumption of energy for heat is not increased by this method.
The fan-motor set consists of a s-h.p. motor, direct-connected to a
specially designed 9-in. cone fan. This fan will handle about
33,000 ft. of air per hour under normal conditions of line voltage.
The motor is connected direct to the goo-volt trolley circuit through
a standard combination snap switch and fuse, and is started and
stopped by means of this swltch. The motor and fan are mounted
in a suitable metal housing which is connected to the exhaust chamber. The fan discharges through rotected openings in each side
of the housing. The exhaust chamter in the upper part of the car
is formed by dropping down the ceiling about 4 in. from the roof
framing and is continuous from end to end of the car body. Communication between the car interior and the exhaust chamber is
provided by a number of openings, each containing a circular
adjustable register. By proper adjustment of the registers a
uniform velocity of air through all of them is obtained. There are
several intakes, half being located on each side of the car under the
seats in such a manner as to be readily connected to the electric
heaters. The connection between the screened opening through
the car floor and the electric heater is made with a pressed metal
duct. The size and number of the intakes is such as to permit of
a maximum velocity of the air of about 400 ft. per minute, which
is hardly perce tible to passengers.
Automatic &stem of Ventilation. The automatic systems
installed are of several diierent kinds, but all depend upon aspirator
action for their operation. One of these automatic systems which
has shown fair results comprises a number of exhausters located
along each side of the monitor roof and attached to panels placed in
the monitor or deck window openings. An opening in the panel
communicates with the exhauster. These exhausters are also
designed for use with plain arch roof cars. Intakes similar to those
described in connection with the mechanical system are located in

542

ELECTRIC RAILWAY HANDBOOK

The exhausters $re

rectangular sheet-metal boxes projecting outwardly from the ~ a n e l s
to which they are secured, having openings top and bottom, and
provided on the middle of each side face with V-shaped projections.
The V projections are placed horizontally on the faces of the
exhausters and split the air into two streams, one following upward
and the other downward. The air streams flow past the openirlgs
of the exhauster and by induction draw the air out from the car.
Location of Air Intake. There is a tendency to abandon intakes
near the floor line owing to street dust. On a number of European
lines, notably the Budapest suburban system, louver type ventilators
are installed over the side sash instead of using a monitor roof.
Car Lighting. Electric car lighting, formerly exclusively by
means of the carbon filament lamp, has recently been made much
more economical by the use of the tungsten lamp, which gives an
efficiency of 1.25 watts
3d per candle-power and has
a useful life equal to or
120'
greater than that of the
c a r b o n lamp. On ac'lo* count of its resistance
~ m . characteristics, the tungsten filament is much less
susceptible to changes in
so'
candle-power with varying voltage than the carbon, so that it will stand
a higher overvoltage than
the latter, and will also
M'
give s fair illumination
a t low v o l t a g e under
which the carbon lamp
'O*
gives practically no light.
Tests made by the Bay
State Street R a i l w a y
(Elec. Ry. Jour., 1912r9r3) indicate the econ-4
slur. -m.t)diWPdon frombu.
omy of using a small
6b"lt Y..d. l.mpb
number of large tungsten
& U d - l h o u . r . n m m u ulwuca (.ol.s~...tt
units rather than a large
ePwrW.-knumber of small units, as
PIG.39.-Light distribution-Tungsten lamp
was the almost universal
with and without reflector.
p r a c t i c e with carbon
lamps.
Fig. 39 shows the comparative light distribution from bare lamps
and lam s with reflectors.
on Equipment, A.E.R.E.A., 1914, stated that
The
the minimum acceptable intensity of car illumination is 1.5 footcandles, and in order to provide for a voltage drop to approximately 80 per cent normal, the car lighting system should be so
designed as to give an average intensity of illumination of at least
3.75 foot-candles at normal voltage in a plane 36 in. above the
floor. (See table, page 543.)
the floor a d provide a supply of fresh air.

,,-

--

committee

f5.B

CAR LIGHTING

36 IN. FROM CARFLOOR.

AT RATEDVOLTAGE.
FOR VARIOUS
LENGTHS
OF CARBODIES. (A.E.R.E.A. COMMITTEE
ON EQUIPMENT.
1914)

fr

Lamps

Length of car body

No. of

Kind

25 ft.

30 ft.

35 ft.

40 ft.

45 ft.

of car
--64-watt carbon. .................................
......
5

.I
...........................

23-watt Mazda.
Not sufficie nt watt age
Tno wid r space
j b w a t t Mazda. ..........................................
5&watt Mazda, prism
Prism
H. opal H.opal
94-watt Mazda. M.opa1 M.opal Prism P r ~ s m
L. opal L. opal H. opal H. opal
M. opal M.opal
64-watt carbon.
Not sufficlent watt age..
23-watt ~ a z d a
36-watt Mazda. H. opal H. opal H. opal
M.opal M.opal
L. opal L. opal M.opal
+watt Mazda. . . . . . . .
L.opal H. opal H. opal 11. opal
M.opal M.opal
L. opal

..............

zo

..........

.........

.......

.--Plvm

L. opal M.opal
Bare
Bare
Bare
L. opal
64-watt carbon. ..........................................
23-watt Mazda. Prism Prisnl Prism
Not su fficient wattage
H. opal H. opal H. opal
M. opal M.opal
L. opal
36-watt M d a . .......

94-watt Mazda. Bare

15

S b w a t t Mazda. Bare
10

*,+-watt Mazda.
64-watt carbon.

............................
.......... Not sufficie nt watt

23-watt Mazda. Prism

H. opal
M.opal
L. opal
36-mtt Mazda. Bare

Prism

Prism

Prism

Bare
Bare
a ~ e .. . . . . . . . .
Prism

' "'

H. opal H. opal H. opal H. opal

v
M.opal M.opal M.opal
L. opal
Bare
L. opal L. opal M,opal Prism
L. opal H. opal
M.opal
L. opal
Bare
Bare
+watt Mazda. ....... Bare
Bare
Bare
44-watt carbon. .......... Not sufficie nt watt age..
23-watt Mazda. Prism Prism Prism P r s ~ m P r ~ s m Prism

.........

&watt

Mazda.

544

ELECTRIC RAILWAY HANDBOOK

Storage Battery Lighting. Especially on interurban lines with

long headways, economy somctlmes d~ctatesthe use of less feeder
copper than will maintain a voltage which is a t all times satisfactory for car lighting. I n order to obviate the annoying effect of a
change in intensity of illumination caused by such fluctuations, the
Lehigh Valley Transit Company equipped a number of interurban
cars with ten cells of A-6 Edison storage batteries arranged in two
trays of five each, each tray weighing slightly under I W Ib. This
battery gives on discharge 240 ampere hours. The cars are
equipped with twenty 2mwatt, 16-c.p., 12-volt tungsten lamps with
holophane reflectors. The lamps are wired in multiple and connected to the battery through an ampere-hour meter which is cornpensated to run 2 0 per cent slower on charge than on discharge and
with a pointer showing the limiting capacity of the battery. The
battery and ampere-hour meter are arranged to be removable for
charging, which is done by a small motor generator set giving 50
amg,res a t 125 volts, which is set up in one corner of the cat house
an wh~chcharges six sets of batteries a t one time. The batteries
are pulled out of their compartment under the car onto a small truck
when the c a n turn in a t the car house on the last trip a t night, and
the sets are all placed in series for charging and cut out obe a t a
time as they come fully up to the charge, which is determined by the
individual ampere-hour meter. The twenty z e w a t t lamps on each
car require appro.xjmateJy 33 amperes and the batteries
deliver
that current for 7% to 8 hours continuously, which meets the
requirements for one day's run. The initial voltage of the ten
Edison cells is about 13 volts a t the beginning of discharge which
gradually drops to 10 volts a t the end of the discharge, and it is
stated that there is not a sufficient variation in the candle-power of
the lamps to be appreciated without the aid of a photometer, I t is
stated that the batteries give no more trouble in handling and charging than the ordinary arc headlight and that the system produces an
absolute direct lighting system in the cars unaffected by the variation
in or even entire failure of the power supply. Although the charging
is somewhat inefficient, the total energy consump
tion including all losses is still considerably less
than with the ordinary carbon lamp equi ment.
Electric Lighting of Train Marker $mPs. I
As it is absolutely essential that train marker
lamps remain lighted even during a period of 2
temporary failure of power supply, there has been 5
some hesitancy on the part of electric railways
in departing from the od-burning lamp for this
purpose. When electric lamps are used, some
method should be provided for maintaining the ,",!&
light during such periods of failure of power sup- Ing. Indianapolrs.
ply, and this is generally done by some arrangement of storage battery. A scheme as used by the Terre Haute,
Indianapolis & Eastern Traction Company, is shown diagrammatically by Fig. 40. A 7-volt storage battery is used in series with
four 1x0-volt, 16-c.p. lamps and a relay in the light circuit opens

c&;T$z

INDICATING TAIL LIGHTS

a 5

a n auxiliary circuit which consists of four ?-volt, 4-c.p. lamps connected in parallel and operating in series with the storage battery
when the trolley current is off. One 110-volt lamp and one 7volt lamp are installed in each marker.
Indicating Tail Lights. The Nichols-Lintern indicating tail
lights employ two lanterns, one provided with a red lens, the other
with a green lens. The lanterns are mounted on the ends of the
car, one on each side of the trolley catcher. The circuits controlling
the lamps in these tail lights are interlocked with the control
appqratus so that a red lamp is lighted when the car is stationary
or coasting, with power off; both red and green lights are burning
when the controller is in series or half-speed position; wit11 the controller in full mriltiple or full speed position the green light alone

FIG.41.-Wiring

diagram for indicating tail lights.

bums. Fig. 41 shows the connections' necessary to accomplish

these results. As shown in Fig. 41, the lights are arranged with an
auxiliary battery circuit which supplies current to a lamp in the
red lens in case the main power supply should fail, as when the
trolley leaves the wire; this feature may be included or not as
desired. The safety and economical features claimed lor this device are: in congested sections cars can operate closer to the car
ahead without danger of accidents; by having an indication as to
the operation of the car ahead, the motorman is able to coast more
and apply his brakes less than otherwise; automobile and truck
drivers behind cars equipped with these indicating tail lights
are able to regulate their speed to that of the car and thus a v i d
collisions.

SECTION X
TRANSMISSION AND DISTRIBUTION

A.I.E.E. Standardization Rules. Tlie L\.I.II.I.,. standards (1922

revision) which apply partictllarly to electric railnay tr.~nsmission
and distribution are as follows:
Contact Conductors. A contact conductor is that part of the
distribution system other than the tralfic rails, which is in immediate electrical contact with the circuits of the cars or locomotives.
Contarf Rail. General: A contact rail is a rigid contact
conductor.
Ovrrhead Contact Rail. An overhead contact rail is a contact rail
which is above the elevation of the maximum equipment line.
(The maximum equipment line 1s the contour w h ~ c hembraces crosssectlons of all rolllng stock under all normal operating c o n d ~ t ~ o n) s

Third Rail. A third rail is a contact conductor placed a t either

side of the track, the contact surface of which is a few inches above
the level of the top of the track rails
Center Contact Rail. A center contact rail is a contact conductor
placed between the track rails, having its contact surface above the
ground level.
Underground Contact Rail. An underground contact rail is a
contact conductor placed beneath the ground level.
Gage of Third Rail. Tlie gage of a third r.lil is the distance
measured parallel to the plane of running mils, Ijetween the gage
line of the nearer track rail and the inside g.lge line of the contact
surface of the third rail.
Elevation of Third Rail. The elevation of a third rail is the elevation of the contact-surface of the third rail, with respect to the plane
of the tops of running rails.
Third Rail Protection. A third rail protection is a guard for the
purpose of preventing accidental contact with the third rail.
Trolley Wire. A trolley wire is a flexible contact conductor,
customarily supported above the cars.
Messenger W i r e w Cable. A messenger wire or cable is a wire or
cable running along with and supporting other wires, cables or
contact conductors. A primary messenger is directly attached
to the supporting system. A secondary messenger is intermediate
between a primary messenger and the wires, cables or contact
conductors.
Classes of Construction. General: Overhead trolley constructions are classed as Direct Suspension and Messenger or C a t e ~ a r y
Suspension.
647

548

ELECTRIC RAILWAY HANDBOOK

Direct Suspension. A direct suspension is the form of overhead

trolley construction in which the trolley wires are attached,
by insulating devices, directly to the main supporting system.
Messenger or Catenary Suspension. A messenger or catenary
suspension is the form of overhead trolley construction in which the
trolley wires are attached by suitable devices, to one or more
messenger cables, which in turn may be carried either in Simple
Catenary, i . e . , by primary messengers, or in Compound Catenary,
i . e . , by secondary messengers.
Supporting Systems. General. Supporting systems for trolley
wires shall be classed as follows:
Simple Cross-span Systems. Simple cross-span systems are those
having a t each support a single flexible span across the track or
tracks.
Messenger Cross-span Systems. Messenger cross-span systems
are those having a t each support two or more flexible spans across
the track or tracks, the upper span carrying part or all of the vertical load of the lower span.
Bracket Systems. Bracket systems are those having a t each
support an arm or similar rigid member, supported a t only one
side of the track or tracks.
Bridge Systems. Bridge systems are those having a t each support
a rigid member, supported a t both sides of the track or tracks.
Transmission System.* When the current generated for an electric railway is changed in kind or voltage, between the generator
and the cars or locomotives, that portion of the conductor system
carrying current of a kind or voltage substantially different from
that received by the cars or locomotives,constitutes the transmission
system.
Distribution System.* That portion of the conductor system of
an electric railway which carries current of the kind and voltage
received by the cars or locomotives, constitutes the distribution
system.
Substation. A substation is a group of apparatus or machinery
which receives current from a transmission system, changes its
kind or voltage, and delivers it to a distribution system.
Standard Height of Trolley Wke on Street and Internban
Railways. (Same as A.E.R.A. Standard.) I t is recommended that
supporting structures shall be of such height that the lowest point
of the trolley wire shall be a t a height of 18 feet above the top of
rail under conditions of maximum sag, unless local conditions prevent. On trackage operating electric and steam road equipment
and a t crossings over steam roads, it is recommended that the
trolley wire shall be not less than a 1 ft. above the top of rail, under
conditions of maximum sag.
Standard Gage of Third Rails. The gage of third rails shall be
not less than 26 in. and not more than 27 in.
Standard Elevation of Third Rails. The elevation of third rails
shall not be less than 2% in. and not more than 3% in.
These definitions are identical in sense. although not in words. with those
of the Interstate Commerce Commlss~on,as glven In thclr Classification
of Accounts for Electric Railways.

WORKING CONDUCTOR CLEARANCES

549

550

ELECTRIC RAILWAY IfANDBOOK

Overhead Trolley Construction

The following notes on overhead trolley construction are principally taken from the Recommended Specifications of the American
Electric Railway Engineering Association.
Classes. Overhead trolley construction may be classed as ( a )
direct suspension, (6) catenary suspension. Direct suspension
comprises construction in which the trolley wires are attached, by
suitable devices, directly to the main supporting system. Catenary
suspension comprises construction in which the trolley wires are
attached, by suitable devices, to one or more messenger cables whch
in turn are carried (a) in simple catenary, by main supporting
system; or (b) in compound catenary, by secondary messengers
which in turn are camed by the main supporting system.
Supporting systems for direct or catenary suspension may be
classed as:
(a) Simple span, comprising those systems ha-g

at each point of sup-

port a slngle flexlble member attached at both srdes of the track or tracks.

( b ) Compound span, comprising those systems having at each polnt of

s u v ~ o r ttwo or mdre flexible members attached at both sides of the track
or-cracks, the upper member carrying part or all of the vertical load of the
lower member.
( c ) Bracket, comprising those systems having at each point of support an
arm or similar rigid member attached at one side only of the track or tracks.
( p ) Bridge. comprising those systems having a t each point of support a
rig~dmember attached at both sides of the track or tracks.

Supporting structures are the devices which sustain the supporting system, and may be poles, whether of wood, steel, or concrete,
towers, buildings, trees, or any other form of support, together with
their anchors, guys, braces and similar reinforcing attachments.
The type of supporting structure will be governed largely by local
conditions. I n general, natural wood (or concrete) poles are used
for all interurban construction and wherever else practicable; steel
poles may be used in streets if so desired; sawed poles and tree
attachments should not be employed, and building attachments
should be used only when local conditions make desirable, in which
case special precautions and construction will be necessary.
(See
.
page i52.1
Pole Framing. Before setting, wood poles should be roofed,
butts squared, entire pole rough shaved, knots smoothed, gains and
faces made, and roof, gains and faces given a coat of preservative or
paint. The size, number and location of holes, faces and gains
vary. I n general: holes, unless specifically noted otherwise,
should be of same size as bolt or rod for which intended. Faces
should be of su5cient area and of proper shape to receive their fittings, and should have about % in. margin outside; they should be
slightly hollowed to prevent rocking of fitting. Gains should be
square with axis of pole, % in. minimum depth, of width to secure
snug fit of arm, and slightly hollowed to prevent rocking. The
roof should have a pitch angle of 45 deg. and should be either a
wedge with edge parallel to line when pole is set, or a cone.
Pole Clearances. On private right of way and elsewhere when
practicable, side supports should be set with a minimum clear distance of 7 it. from center line of track to face of support a t level of

I'OLE SETTING
top of rail, and center Supports should have a minimum clearance
of 7 ft. from center of track, this clearance to be increased if necessary on curves to allow for rail elevation and car overhang. Where
curb lines are established, poles should be set just behind the curb
itself unless local ordinances or conditions prescribe other location.
Pole Spacing. Poles on tangents should be normally spaced not
less than go it. and not more than I I O ft. apart. Poles on curves
should be set as nearly as practicable in accordance with the
table on page 560.
Pole Setting-Depth
of Holes. Pole holes in level ground
should have depths as follows:
Length of pole (feet)

65

70

Depth of hole
In earth
In rock or concrete
6 It. o in.
5 ft. o in.
6 ft. o in.
5 f t . 6 in.
4 ft. 6 in.
6 ft. 6 in.
6 f t . din.
6 ft. o in.
7 ft. o in.
6 ft. 6 in.
7 f t . 6 in.
6 ft. 6 jn.
8 ft. o in.
7 ft. 0 In.
7 ft. o in.
8 ft. 6 in.
7 ft. o in.
9 ft. o in.

I n very compact soil pole holes may have depths intermediate

between those for same lengths in rock or concrete, and in earth.
The depth of a hole on sloping ground should be measured from the
lower side of the hole; and in very steep slopes and in loose or otherwise doubtful material the depth should be increased over the
standard depth by a n amount to be determined for each case on the
ground.
Barrel Holes. I n material which caves freely one or more barrels or the like may be used as casing, and driven down as the
material is dug from inside. Such barrels or casings may be of
wood or steel and should be of sufficient size to give clear tamping
room of a t least 6 in. around the pole without cutting the latter.
Rake. Wood poles with brackets should in general have a rake
from the track of 6 in. in 2 4 It.; steel poles with brackets, of 3 in. in
2 4 ft. Wood poles with span should have a rake from the track of
1 2 in. in 2 4 ft.; steel poles with span should have a rake from track
of 6 in. in 2 4 ft. When the strain is from the track, as with poles
on the inside of a curve, brace poles or head guys should be used,
and standard rake maintained. Double bracket poles should be set
without rake; other poles between tracks, and poles under outside
jurisdiction may be so set if necessary or required.
Keys. All span poles as well as bracket poles on curves of
radius less than 2400 ft., and such othcr poles as may be designated because of unusual load conditions, should be provided with
suitable keys of wood, stone, or concrete, a t least 4 in. thick with a
cross-section not less than 32 sq. in. One key 2 ft. long should
be placed on edge behind the pole a t the bottom of the hole; the
other key 4 ft. long should be placed on edge in front of the pole
just below the surface of the ground.
Refllling Holes. I n setting poles the excavated material, if
suitable, should be replaced in thin, even layers, and firmly aqd
thoroughly tamped, not more than one shoveller scrving' three

552

ELECTRIC RAILWAY HANDBOOK

tampers, until the hole is completely filled, after which the earth
should be well packed around the pole in a small mound, and if on
a slope there should be made on the lower side, a berm a t least 6 ft.
wide from pole to edge. In rock holes the broken rock should be
used to thoroughly wedge the pole in place. Black loam and similar
poor material should be replaced by
suitable material.
Concrete Settings. Poles subject
to heavy lateral strains which cannot well be met by guying or bracing may be set in concrete mixed
wet and consisting of one a r t Portland cement, three parts cffan sharp
sand, and five parts good hard gravel
or broken stone of size to pass
screen with holes 2 in. diameter,
and to be retained by screen with
holes % in. diameter. Concrete settings should have a diameter a t
least 1 2 in. greater than that of the
pole, and should completely fill pole
hole to a level 6 in. below surface of
the ground. I n parking strips the
authorities may require the concrete to finish a t this level, but
such latter practice is undesirable.
Wherever practicable the concrete
FIG.2.-Concrete pole setting.
should be finished as indjcated in
Fig. 2.
Span Wires Attached to Buildings. Building owners in crowded
districts sometimes prefer to have span wires attached to their
buildings rather than to have poles in front of the buildings. Eyebolts for the purpose are sometimes placed in the building during
construction. As a means of suspension such a n arrangement is
satisfactory but i t must be borne in mind that it subjects the build-

FIG.3.-Thimble

end attachment.

ing to unusual strains and the wires are more dangerous in case of
breakage as they are over the sidewalk and will sweep i t with more
force than when attached to poles.
Span and Guy Attachments. Span and guy attachments may
he made up with such of the following forms as may be desired:
Thimble End. (Fig. 3.) Thimble end may be made by bending
strand around thimble of proper size. The strand end should

SPAN WIRES

553

extend 15 in. beyond thimble, and should be secured by a three bolt

clamp close to thimble, and by a serving of about 10 turns of No.
1 2 galvanized wire I in. from end of strand.
Two-turn Wrap. Two-turn wrap may be made b y taking strand
around pole twice. If a t end of span or guy, the end of the strand
should extend 30 in. beyond the face of pole and should be secured
to main part by a three bolt
clamp ~q in. from face of
pole-and by a serving of
about 10 turns of No. 12 galvanized iron wire I in. from
end of strand. If hitch is a t
a n intermediate point in span
or m y i t should be secured
by '8-three bolt clamp on
the outside parts of the
strand. +he
bolt of
PIC.4.-Three-bolt clamp hitch.
---- center
- - ---d a m p being replaced by a lag screw into the pole as shown by
r lg. 4.
Close Tic. (Fig. 5.) Close tie may be made by bending strand
tightly around attachment, leaving a free end about 15 in. long.
One wire of this end should be unlayed back to the attachment and
sewed on main art and remainder of end; the other wires should be
in turn unlayefback to the last wrap and served over main part
and remainder of end until latter is a11 served on.
1

FIG.5.-Close

tie.

Temporary Tie. (Fig. 6.) Temporary tie may be made like

permanent tie of same kind, but end should be left long enough to
allow for adjustment and permanent make u p and should be
secured to main part by one or more servings of wire. I n making
temporary ties bend should be of as large radius as possible until
permanently secured.

PIG.6.-Temporary tie.

Anchors. (Fig. 7.) Anchors in earth should consist of a wooden

deadrnan 4 it. long, a t least 6 in. thick, and having a cross-section
not less than 48 sq. in., buried a t least 4 ft. below the surface with
not less than 2 ft. of rock if reasonably obtainable well packed
into hole, and the earth filling above thoroughly tamped. The
guy rod should pass through center of deadman, and must lie in line

554

E1,ECTRIC RAILWAY HANDBOOK

of pull of guy to prevent bending. Patent anchors of holding

capacity equal to the breaking strain of the strand to be used writh
them, and having rugged parts sufficiently large to allow a reasonable amount of corrosion without reduction in holding c a p a c j t ~ ,
may be used in place of rod and deadman where conditions are
favorable. Anchors in rock should consist of eye-bolt securely

CROSS-ARMS

555

leaded or sulphured for entire length of shank in hole )6 in. larger

in diameter than bolt, and inclined a t right angles to pull of guy.
In rock of sufficient strength to safely withstand the action, mechancal wedge type eye-bolts may be used, and the lead or sulphur
omitted. Where practicable guys may be anchored to adjacent
pole a t point not less than 7 ft. above ground.
Guys. (Fig. 7.) Guys should be used where practicable on
wood poles on curves of radius less than 900 ft., on poles to which
are attached strain plate guys, trolley guys and head guys; and
wherever any side strain exists. They should be of galvanized
seven-wire steel strand with thimble end a t anchor attachments,
and two-turn wraps a t proper height for attachment. The lead or
horizontal distance from pole a t ground line to guy a t same level,
whenever practicable, should be equal to the distance from ground
line a t pole to guy wrap. Guys located where there is a liability of
ersons or animals running into them should be made conspicuous
a piece of pipe 2 or more inches in diameter and 6 ft. long,
sli ped over guy, resting on anchor rod eye and painted white.
d e r e guy is already installed, a wooden casing, 4 In. diameter or
square and 6 ft. long, may be used in place of the plpe. The halves
should be well white leaded, and should clamp the guy tightly when
screwed together, the bottom resting on anchor-rod eye. Guy
hooks attached one on each side of pole a t level of guy wrap by a
Vn-rurrgP~-~d~
a% irg'tt~aq'm tu 'inn i3i pd113tmdd1tn -us& - & 1 m
local conditions compel the use of a lead less than one-fourth the
distance from ground line a t pole to guy wrap. Where guys cannot be run directly to secure attachments they may be carried high
on adjoining poles to a point where anchorage can be obtained.
Cross-arms. (Fig. 8.) Cross-arms should be given one coat of
preservative or two coats of paint before pins are installed. Pin
shanks should be dipped in the preservative or paint, and while wet
firmly seated in hole and secured by a wire nail 2 in long driven
through side of arm into pin. All cross-arms should be held to
pole by a through-bolt driven from back of pole toward and through
arm, having a washer a t each end, with hole a t back of pole p r o p
erly counterbored to secure good seat for washer. Cross-arms for
light duty service should be steadied by strap braces secured to
~ l bye a lag screw, and to side of arm away from pole by carriage
Its on center line of arm with nuts next to the braces and a washer
under bolt head. Arms 36 in. in length should have braces 2 0 in.
long, fastened to arm I z in. from center; arms over 36 in. long should
have braces 30 in. long, fastened to arms 19 in. from center. Crossarms for heavy duty service should be steadied by an angle brace,
fastened to bottom of arm by carriage bolt and to pole by a throughbolt. The lowest feeder, telephone or signal cross-arm should have
its center not less than 21 ft. above top of rail; other feeder, telephone or signal cross-arms should be spaced a t least 24 in. from
center to center. If the pole also carries a transmission line there
should be a clear distance of a t least 6 ft. between the top feeder,
telephone or signal arm and the lowest transmission arm. Double
arms for feeder should be used a t ends of curves and on intermedia~e
poles of curves of radius less than 500 ft. Double arms must be

gy

556

ELECTRIC RAILWAY HANDBOOK

HEIGHT OF TROLLEY WIRE

parallel and a t the same height. Both arms should be fastened
to the pole by the same bolt, and should be firmly tied together
by spacing bolts with nut and washer each side of each arm,
located on the center line of arm, 8 in. from end.
Choice of Supporting System. Bracket support on side poles
should be used for all single track where local conditions do not prevent, except for curves of radius less than 300 ft. Bracket support
on center poles may be used on double track where practicable.
Span support should be used on single track for curves of radius less
than 300 ft. and where local conditions do not permit use of brackets; on double track, including turnouts, not employing center
bracket poles; and for more than two tracks. Compound spans
should be used when necessary to support the overhead of a series
of tracks too closely spaced to permit poles between. Bridge
support should be used only in special cases.
Height of Trolley. Supporting structures should be of such
height that the lowest point of trolley wire in streets and on interurban lines will be a t a height of not less than 18 ft. above the top
of rail under conditions of maximum sag, unless local conditions
prevent; on trackage operating electric and steam road equipment
and a t crossings over steam roads the trolley wire should be not
less than 21 ft. above the top of rail under conditions of maximum
sag.
Bn&eb, B~aclretssboud be o f s ~ 5 c j ci ~~ t~ gt ot aiJ~w
h
3 jn.
between hanger or strain insulator and the end casting. When
poles are on outside of curve the length of bracket should be su5cient to allow for effect of pole rake and rail elevation. The eyebolt should be installed a t level of trolley wire; the pole casting with
center a distance above center of eye-bolt equal to distance center
to center between span eye and arm socket of end casting; and the
oversupport rod a distance in inches above pole casting of three
times the length of arm in feet. On wood poles eye-bolts and oversupport rod should pass through pole and pole casting should be
attached by two lag screws. On steel poles, eye for strand, socket
for arm and pole attachment for oversupport rod should be carried
by special fittings, clamping to pole. Eye-bolt should be installed
pulled out to full length; the nut, on wood pole, seated against a
washer; the arm should be given an upward rake from the horizontal of I in. in 4 ft. of length; the intermediate casting should be
clamped on arm so that trolley wire comes midway between it and
end casting; and the steel strand should be close tied into eye-bolt
and into end castings with 2 in. of slack to permit hanger installation. On steel poles strain insulators should be cut into strand on
either side of hanger, and between end and intermediate castings,
to give double insulation. Brackets on curves should be installed
as on tangents except that pull-over with attached strains should be
close tied in strand in approximately correct position, but ties a t
eye-bolt and end casting should be temporary until final dressing of
overhead.
Spans. Spans should consist of seven-wire strand, and in case of
steel poles, should have strain insulator cut in not less than 5 ft.
from pole; where pole carries high tension circuits, a strain insu-

558

ELECTRIC RAILWAY HANDBOOK

lator should be used of suitable strength and creeping surface.

Where a foreign line crosses close to the span two strain insulators
should be used, one a t either side of foreign line, to ensure that if
latter falls it shall be on a dead section. Spans on tangents should
be close tied into eye-bolts a t height above trolley wire not more
than one-tenth the distance from track center to pole but in no case
should the factor of safety for the span wire be less than two under
the conditions to be expected. On wood poles eye-bolts should be
a t least 12 in. below top of pole, and should be installed a t full
length, seated against washers. Spans on curves should be installed
as on tangents exce t that pull-over with attached strains should be
dose tied in stranfin approximately correct position and temporarily tied in eye-bolts until the final dressing of overhead. In
case of two or more tracks, strand between pull-overs should have
temporary tie a t one end until final dressing.
Trolley Wire. Trolley wire may be run out by mounting the reel
on an arbor on which it can turn freely, and leading the wire to an
anchor or to trolley already installed. Tension may be maintained
by a brake on rim or side of reel, but under no circumstances should
braking be done against the copper. The wire should be pulled to
proper sag and temporarily tied to brackets or spans by rope or
other soft insulating ties. Particular care should be taken to prevent twisting, kinking or bruising the wire. Parallel faced clamps
should be used; chains, cam come-alongs or other short grip devices
should not be employed. The sags should be as follows, and in
pulling up great care should be taken that the corresponding
tensions are not exceeded:
SAGAND TENSION
OF TROLLEY
WIRE (100PT. SPANS)
SIZE OF WIRE
41
Temperature.
deg. P.

--

000

0000

1450
1050
-

The table values are for spans of roo ft: for any other span the sag for
the tension given in the table is proportion61 t o the squares of the lengths.
For example for s n of 50 ft.. the sag for a given temperature is equal to
so squared &vide&
loo squared. or one-quarter the corresponding table
value for that temperature.

After the trolley wire nas been temporarily tied up with the
proper sags, and the line has been anchored, the line ears and
hangers may be located accurately and attached, clinch ears being
thoroughly closed down to give secure grip and smooth running
surface, and the mechanical ears well seated in grooved wire, the
clamp screws then being slightly upset to prevent backing out.
Trolley Wire Splices. Splices should be of a type and so installed

TROLLEY WIRE INSTALLATION

559

obstruction to the passage of trolley wheel. Grooved wire should

be kept in ~ e l l e c talinement, and if the splice is of the soldering
type, it should be thoroughly sweated on without annealing the
wire. The free ends of the wire may be bent back sharply at the
outlet, and cut off forming a hook with end $6 in. long.
Trolley Wire Guys. In bracket construction trolley wire guys
should be installed a t the ends of curves and on long curves and
tangents at equal intervals as nearly as ssible, but not to exceed
1500 ft. apart. Trolley wire guys s h o u l E e seven-wire steel strand
attached to strain plate supported a t a bracket by double pull-over
with proper insulation and led both ways to next adjacent poles.
Each guy should have a strain insulator cut in it 5 ft. from the
strain plate, and should be secured to proper ~ o l eby a two turn
wrap a t height of bracket arm. Where practicable, the strain of
these guys should be taken by anchor guys in the l i e of the pull;
if this is impracticable, high guys should be used. Great care
should be taken to ensure equal pulls on the guys, and especially
that the strain plate is not twisted out of line; the ties should not
be made up permanently until the final dressing of the overhead.

PIG.9.-Location

and arrangement of pull-offs in trolley wire curves.

Curves. (Fig. 9.) Curves should be made up with straight-line

clinch ears for round wire, or double clip mechanical ears for
grooved wire, attached to suitably insulated pull-over bodies.
Support should be by span except where rest of line is in bracket
and radius is greater than 300 ft. Between supports curves should
be held to line by seven-wire steel strand, with single body pulloffs for single track or for outside one of several tracks, and double
body pull-offs elsewhere, spaced as follows:

560

ELECTRIC RAILWAY HANDBOOK

Rad~usof curve
(feet)

Spacing of pull- Number of pulls

between supports
ofis (feet)

40
50
60
70

7
8
9
I0

80
90
100
135

11

I50
200-500
750
1000
1500 2000
Above 2000

I2

I3
I4
I5
20
25
33>5
50
100

4
4

4
4
4
4

4
4
4

D~stanceapart
of pole? (feet)
35
40
45
50
55
60
65
70

75
80
100
100

100

3
3

loo

The pull-overs in each span should have bodies and strains held
radially to curve by a lacing of seven-wire steel strand a t least 6 in.
away from trolley wire, which lacing, however, may be omitted
from any pull making an angle of 60 deg. or more with the ear
to which it is attached. With an odd number of pull-over bodies
the middle one should have pull-off strand to each pole. Intersections and complicated speual work, rticularly in dty streets,
will usually requlre special and indiviEal study and treatment.
Figures 10 to 16. inclusive, give typical overhead layouts for such
special work.

TROLLEY CURVE CONSTRUCTION

561

562

ELECTRIC RAILWAY HANDBOOK

TROLLEY CURVE CONSTRUCTION

563

564

ELECTRIC R W W A Y HANDBOOK

Offset of Trolley on C w e s . Curves should be dressed with

uniform deflection a t pull-overs and should offset to the inside of
curve an amount given by the expression

EH
C

S= - + R - . \ / R ~ + P ~ - Q ~ - L ~

in which

S = radial offset of trolley wire toward center of curve

,

E
H
G
R
P
Q
L

= super-elevation of outer rail

= height of trolley wire above rail

= track gage
= radius of curve
= distance from center of car to pivot of trolley base

= distance from center of car to center of truck

= horizontal distance from pivot of trolley base to

point of contact between trolley wheel and trolley

wlre.
NOTE:
All values in terms of feet.
The total offset should be uniformly tapered off from full value a t
inside easement point of track to no offset a t outside easement point.
I f track is not eased, start with full offset a t distance inside end of
curve as given below, and run to no offset a t point a t equal distance outside end of curve.
Radius of curve
Up to 100 ft.
roo to 500 ft.
So0 to 1000 ft.

Above 1000 ft.

Start off set easement

ft. from end of curve

20

40 ft. from end of curve

60 ft. from end of curve
100 ft. from end of curve

Frogs. Frogs should be installed with both main line and branch
trolley wires led straight through the latter to end 6 ft. beyond
frog in eye of strain insulator, to other end of which is dose tied a
seven-wire steel strand, secured to pole a t a level as nearly that of
frog as will allow safe clearancz over other wires. The frog itself
should be held on each side by a guy of seven-wire steel strand with
strain insulator attached to frog and secured to proper poles a t the
point of span attachment. Frogs should be temporarily located on
center line of main track and one-third distance from track switch
points to track frog point back from track switch points, and if
need be should be shifted to suit local conditions and equipment.
Until final location is made, trolley wire should be clamped just
firmly enough to prevent slipping without bmising or kinking.
Care must be taken that frog is not located too far back of track
switch points, as such location, while often giving satisfactory
running, will result in excessive wear of trolley wire.
Crossings. Crossings wherever practicable should be installed
without cutting either of the line wires, which latter should be
clamped just firmly enough to prevent slipping without bruising or
kinking until the crossing has been saisfactorily located. Where
wire must be cut, at least 3 ft. of free end should be left outside
clamp until the final adjustment, after which the end should be cut
off close to the clamp.

TROLLEY FEEDER TAPS

565

Feed Taps. (Figs. 17 and 18.) Feed taps should be a t

support and should consist of proper size triple braid. w e a t ~ S 2 :
stranded connectionfrom feeder, fecd yoke well soldered on a t
proper point, and straight-line ear soldered to the trolley wire and

Flu Joht-usm Pd*-

FIG.17.-Feed

tap in bracket construction.

bolted to the yoke. I n general, feed taps should be located every

1000 ft.
With bracket support the feeder connection should run
from the feeder to which it should be well soldered, to strain attached to span eye-bolt in pole, thence, replacing the usual steel

EJDBolt

ieed.1.

cu

FIG.18.-Peed tap in span construction.

strand through insulated intermediate casting to strain attached to

end casting of bracket. Bracket tube should be of sufficient length
to allow at least 8 in. of connection cable between feed yoke and end
strain, and intermediate casting should be so located t h a t feed yoke

ELECTRIC RAILWAY HANDBOOK

566

is midway between i t and end casting. With span support the

feeder connection should run from feeder to strain attached to span
eye-bolt in pole, thence, replacing the usual steel strand, to close
tie in strain 5 it. beyond trolley farthest from feeder. Span shodd
be completed by seven-wire steel strand close tied into strain and
into pole eye-bolt, and should sag not greater than one-tenth the
distance from trolley wire to pole.
Feed wire may be run out by mounting the reel on a n arbor on
which it can turn freely, and if practicable, run along the line on a (:ar
or wagon. Where local conditions necessitate pulling feeder on end
over the cross-arms, great care must be taken to prevent injuff,
especially to insulated feeder, rollers or snatch blocks of ample size
being used a t each arm. Feeder should be strung, for spans of
~ c ft.,
a with sags not less than those of the following table:
ALLOWABLE
SAGSI N FEEDERSPOR DIFFERENT TEMPERATURES
Temperature. Fahrenhe~t
(degrees)

Copper

(in.)

Alumrnum

(in.)
5

Po

10
13%
17
20%

120

21%

23%

26

30

60

10
16

The table values are for elther bare or weatherproof feeder of any size
from oooo to ~.ooo.ooocrrcular m ~ l s For spans other than I O O ft in Ienkh
the sag should be In the same ratio to tho table values as 1s the square of the
span to loo squared.

Feeder should be installed in the top groove of the insulator

on tangents and in the outer side groove on curves and a t angles, and
should be tied in with No. 6 soft drawn wire of the same metal as
feeder. Top tie (Fig. 19) may be made with two tie wires each 15
in. long. The first wire should be looped around insulator in side

FIG 19.-Top tie for feeder.

FIG.20.-S~de tle for feeder.

groove, the ends crossed and twisted once under cable, then brought
up, one each side and wrapped around cable, crossing above and
below until all is on. The other wire should be used to make a
similar tie on opposite side of insulator. Side tie (Fig. 20) may be
made with one tie wire 15 in. long which should be looped around
inside groove a t back and ends, carried under cable, then working
from center with one end each side, over top, down back, under,
up front, and so on for four complete turns, then around back of
insulator where ends should be twisted together securely. The
cable side of the insulator is front.
On tangents feeder may be carried on single arms having insulators on wood pins; on angles less than 10 deg. by single arms,

TROLLEY FEEDER INSTALLATION

567

and on angles greater than 10deg. by double arms, in either of the

latter cases having porcelain insulators on metal pins.
Splices in solid feeder should be made with an approved connector;
in stranded feeder should either be made with an approved connector,
or of the wrapped cable type, tapered uniformly to size of original
feeder a t ends. In insulated feedereither type should be smoothly
taped to the equivalent covering. The wrapped cable type splice
(Fig. 21) is made by stripping the ends to be spliced for 24 in.,
unlaying and brightening 18 in. of these bare ends, cutting out the
core strand of each, and passing wires of one end between those of
other end and laying parallel with main cable. The wires of one
strand either side of mddle of splice are then close served over main
portion and rest of wires of that end; a second pair of strands is
similarly served on, and so on, until the last pair of strands serve
simply on the main portion. The splice should then be sweated
full of solder and smoothly taped to equivalent of original insulation

FIG.21.-Cable

type feeder spl~ce.

if cable is insulated. Lead covered or specially insulated or covered

cables should be spliced under special instruction, and then only
by men experienced in that particular class of work.
Choice of Feeder. Bare feeder should be used on private rightof-way and wherever else practicable; weatherproof feeder where
required and where numerous trees or other obstacles would cause
grounds if bare cable was used.
Feeder Anchors. Where long spans, long heavy grades, or other
conditions cause unusually heavy strains, feeders should be
anchored by mechanical clamps secured to eye-bolt in cross-arm
through a suitable strain. Insulated feeder should be bared so
that clamp takes direct hold on the metal. If heavy pull is anticipated, take guy from cross-arm to next pole.
Section Insulators should be installed a t a span suspended from
hanger.
Section Switch Box should be bolted to back of pole, using bolts of
proper length, each with washer under head and nut, latter being
inside box, which should rest on cross-arm carrying feeder sectionalized. The feeder should be dead-ended in strains attached to
eye-bolt a t proper point in feeder cross-arm and then enter box.
If feeder drops from dead end point it should be carried lower than
bushings in box and then up to prevent drip from entering.

568

ELECTRIC RAILWAY HANDBOOK

Automatic Sectionalizing Switch. Fig. 2 2 shows the circuits in

the automatic sectionalizing switch of the General Electric Company. The switch is connected across the section insulator by
taps G and H. Circuit breaker B upon closing energizes section B
and current passes through tapG, contactor operating coil X to contact stud on relay which is then open circuited. On closing breaker
C, section C is energized and current asses through tap H and relay
operating coil W to ground, closing tge relay disk Y . This in turn
completes the circuit through the contactor operating coil X, causing the contactor to close. This completes the circuit across the
insulator, thus placing the feeders in multiple. The switch will not
operate until both breakers feeding the sections to which it is connected are closed. With a heavy load on one section (while the

FIG.22.-Circuits

of automatic sectionalizing switch.

automatic switch is in operation), current from the adjacent sections will be fed across the section insulators, thus increasing
materially the efficiency of the entire distribution. Where an automatic or hand-o erated sectionalizing switch is used, tests should &
made periodicaEy to determine any changes necessary in feeder
sizes in order to divide the load properly between the feeders.
Lightning Arresters where required by local conditions should &
installed a t feed taps, just below the feeder cross-arm, and should
be connected to the feed tap close to its attachment to the feeder by
solid insulated No. 4 copper wire. The ground connection should
be of solid No. 4 copper wire, securely fastened to the back of pole
and either (a) extended as a ground coil, (b) well soldered into a
pipe ground or ( 6 ) well soldered to the track rails. In any case
a t least 8 feet of the lower exposed portion should have a nonmetallic protection. Any changes in the direction of ground wire
should be made by easy curves.

LIGHTNING ARRESTER GROUNDS

Line lightning arrester ground wires should be connected to a
good earth ground and to the track rail, except:
( I ) Where the current flow (see note) on the connection from
track to earth would exceed an average of one-fourth ampere during
any twenty-four hour period, and
NOTE:In checking up this current flow the algebraic sum of currents

flowing first in one direction and then in &e other shall be used in deterrmning the current flow. I t is assumed that a resultant of more than onefourth ampere average m e~therdirection should be avolded.
(2) Where alternating current track circuit block signals of the
double rail type are used, the connection to the track rail should be
omitted.
Particular care must be taken to ensure that the ground is
effective; unless a good ground is secured the arrester cannot give
protection. Earth grounds should be secured as follows: Where
permanently moist earth is assured a t reasonable depth the ground
may consist of a one-half inch pipe driven a t least 3 feet into the
moist earth. Where there is doubt as to the condition of the soil,
excavate. If permanently moist earth is reached, install pipe
ground; if otherwise, install a flat coil containing 40 lineal feet of
solid No. 4 bare copper wire bedded in not less than 7 cubic feet of
charcoal.
Soldering. Where soldering is necessary, it should be done with
non-corrosive paste or with stearin; the use of acid or corrosive salts
should be strictly forbidden; and great care should be taken to prevent overheating and annealing.
Linemen. Only such men should be employed on overhead construction as have had experience and are so skilled in the work that
those details which make up good ractice will be attended to
without the necessity for specific antdetailed instruction. Such
details include: grading poles to bring tops to approximately the
same line; setting cross-arms square to tangent kne, and bisecting
the angle a t breaks; setting brackets square to line and all with
same rake; cutting in strains a t the same relative points; installing spans so that eye-bolts and strands line u setting chamfered
nuts with flat side to bearing; seating insuetor pins firmly to
shoulder; seating washers square with bolt hole; screwing in lag
screws a t least the last half; finishing off all splices, fastenings and
ends.

570

ELECTRIC RAILWAY HANDBOOK

Sag and Horizontal Pull in Span

Single Track Llne -The following table shows the sag resultIng m spans of vanous lengths when subjected to the tensions
gven, a standard span welght of 50 lb belng assumed This IS a
convenient standard welght for a 4.Es In steel strand span whch
supports a length of IOO f t of oo trolley with ear and insulator.

SACIN SPAN WIRE-SINGLE-TRACKLINE

Maximum Sag In Inches ( 6 ) for Weight (W) of 50 lb.

Length of span ( S ) In feet

Pull on span (P)

m pounds

500
600
700

1 1 1 1 1 1 1 1

30

35

40

45

50

55

60

65

70

75

80

9 0 1 0 5 1 2 0 1 3 5 1 5 0 1 6 5 1 8 0 1 9 5 2 1 022 5 2 4 0
7 5 8 8 1 0 0 1 1 312 513 8 1 5 0 1 6 3 1 7 518 8 2 0 0
6 4 7 5 8 6 9610711812913915016117 I

800
900
1000

5 6 6 6 7 5 8 4 ~ ~ I O J I I ~ I Z ~ I ~ I I
5 0 5 8 6 7 7 5 8 3 g21001081171zsr3f
4 5 5 3 6 0 6 8 7 5 8 3 9 0 981osr13rzo

I roo
I ZOO

4 1 4 8 5 5 6 1 6 8 7 5 8 2 8 9 95102109
3 8 4 4 5 0 5 7 6 3 6 9 7 5 8 2 8 8 94100
3 5 4 0 4 6 5 2 5 8 6 3 6 9 7 5 8 1 8 7 9 2

1300
I400
I 500
1600

3 2 3 8 4 3 4 8 5 4 5 9 6 4 7 0 7 5 8 1 8 6
3 0 3 5 4 0 4 5 5 0 5 5 6 0 6 5 7 0 7 5 8 0
2 8 3 3 3 8 4 2 4 7 5 2 5 7 6 1 6 6 7 0 7 s

1700
1800
I900

3 6 3 1 3 5 4 0 4 4 4 9 5 3 5 7 6 2 6 6 7 1
2 5 2 9 3 3 3 7 4 2 4 6 5 0 5 4 5 8 6 2 6 7
2 4 2 8 3 2 3 6 4 0 4 3 4 7 5 1 5 5 5 9 6 3

zwo

2 3 9 6 3 0 3 4 3 8 4 1 4 5 4 9 5 3 5 6 6 0
2 1 1 s 2 9 3 2 3 6 3 9 4 3 4 7 5 0 5 4 5 7
2 2 1 4 2 7 3 1 3 4 3 8 4 1 4 4 4 8 5 1 5 s

2300
2400
2500

2 0

2100
2200

2 3 2 6 2 9 3 3 3 6 3 9 4 2 4 6 4 9 5 2
1 9 2 2 2 5 2 8 3 1 3 4 3 8 4 1 4 4 4 7 5 0
1 8 2 1 2 4 2 7 3 0 3 3 3 6 3 9 4 2 4 5 4 8

Sags are computed

slngle track llnes

from the formula cl

In whlch W = so for

671

SAG I N SPAN WIRES

Double Track Lme. The followrng table shows the sag resulbng
In spans of varlous lengths when subjected to the tensons gwen, a
standard span welght of IOO lb be~ngassumed Thls 1s a convenlent standard we~ghtfor a %-In steel strand span whlch supports
a length of ~ o fot of two oo trolley w e s , insulators and ears.

SAGI N SPANWIRES-DOUBLETRACK
LINE
Maxlmum Sag In Inches (c) for Welght ( W ) of IOO Ib

a2; zq

Length of span (S) feet

'=-po

aa

gags
50
loo
150
zoo

301 5 1 401 451

120
60
40
30

orso
o 75
o 50
o 37

0180
o 90
o 60
5 45

0210
o rog
o 70
o 51

so(

0240
o rzo
o 80
5 60

5 s ) 6 0 1 6s

0270
o 135
o 90
o 67

7 0 1 751 80

0300 0330 0360 0390 0410 o

o 150 0 165 0 180 0195 0210 0
oroo o r r o o r 2 0 0130 or40 o
5 75 0 82 5 90 0 97 5 -05 0

950
300
400

24 o 30 o 36 o qa o 48 o 54 o 60 o 66 o 72 o 78 o 84 o
20 o 25 o 30 o 35 o 40 o 45 o 50 o 55 0 60 0 65 0 70 o
15 o 18 8 22 5 26 3 30 o 33 8 37 5 41 3 45 0 48 8 51 5

500
600
700

12
10

o 15 o 18 o 21 o 24 o 27 o 30 o 33 0 36 0 39 o 41 o
o 1 2 5 15 o 17 5 20 0 2 2 5 25 0 27 5 30 0 31 5 35 o
8 6 10 7 12 g 15 o 17 I 19 3 11 4 23 6 a5 7 17 9 30 o

800
900
1000

7 5
6 7
6 o

9 4 11 3 13 I 15 o 16 9 18 8 20 6 22 5 24 4 26 3

1100
I ~ O O

1300

5 5
5 0
4 6

6 8
6 3
5 8

8 2
7 5
6 9

9 6 1 0 9 1 2 3 1 3 6 1 5 0 1 6 3 1 7 7 1 9 ~
8 8 I O O 113 125 138 150 163 175
8 1 9 3 1 0 4 1 1 5 127 1 3 8 1 5 0 162

1400
1500
1600

4 3
4 0
3 8

5 4
5 0
4 7

6 4
6 0
5 6

7 4
7 0
6 6

8 6
8 0
7 5

9 6 107 118 129 1 3 9 1 5 0

g o 100 1 1 0 I 2 0 1 3 0 1 4 0
8 4 9 4 1 0 3 r r 3 1 2 2 131

1700
1800
1900

3 5
3 3
3 2

4 4
4 2
4

5 3
5 0
4 7

6 2
5 8
5 5

7 1
6 7
6 3

7 9
7 5
7 1

8 8
8 3
7 9

9 7 106115 Ia4
9 2 100 108 117
8 7 951031x1

2000
2100
aaoo

3 0
1 9
a 7

3 8
3 6
3 4

4 5
4 3
4 1

5 3
5 0
4 8

6 0
5 7
5 s

6 8
6 4
6 1

7 5
7 1
6 8

8 3
7 9
7 5

9 0
8 6
8 a

9 8 1 0 5
9 3 roo
8 9 9 5

2300
2400
2.500

2 6
2 5
2 4

3 3
3 1
3 0

3 9
3 8
3 6

4 6
4
4:

5 2
5 0
4 8

5 9
5 6
5 4

6 5

7 1

7 8
7 5
7 2

8 s
8 1
7 8

8 3 10 o 11 7 13 3 15 o 16 7 18 3 20 0 21 7 23 3
7 5 9 o t o 5 12 o 13 5 15 o 16 5 18 o 19 5 11 o

Sags computed from ) the

~ formula
- - S (c W J

and a

10 for double track.

:i : g
P

In which W

9 1
8 8
8 4
IOO

The values in the above table are plotted in Figs 2 3 and 24

For other sues of wlre strand or trolley the weight of the span will be
d~flerent The table on page 573 shows the "loadlng factor" or
ratlo of we~ghtof vanous spans to the welght of the assumed standard span of loo Ib The sag even In the above table, when multlpl~edby the proper "load~ng factor," wffl give the sag in any
des~redspan.

572

ELECTRIC RAILWAY HANDBOOK

FIG. 23.-Sag

and horizontal pull of double track span wires.

(See also

Fig. 24.)

35

30

-.
V

25

9.

3 16

10

6
0

loa,

WO
PIG. q.-Sag

wo

~ o * o n r ~ p p l l l o P o u n(P)
~

and horizontal pull. of double track span wires.

Fig. 23.)

(See also

SAG

IN SPAN WlRES

573

ELECTRIC RAILWAY HANDBOOK

Catenarp Construction
Endeavor to reduce the sag in overhead contact conductor by
supplying several points of support in a given span resulted in the
development of the catenary type of suspension. I n its elementliry
form it consists in suspending the contact conductor by means of
clips or hangers which are placed a few feet apart and which We
hung from one or more messenger wires which hang nearly ifl a
catenary curve between the main supports of the overhead construction or between intermediate supports suspended therefrom.
There are a great many varieties of this general form of suspensic?n,
varying from the simple catenary from which the trolley wlre
is suspended directly, to very elaborate suspension systems. The
purpose of the catenary construction is to support the conttct
conductor a t points close together while maintaining a great dlstance between points of support of the whole system, thus reducing
changes in height of the contact conductor to a practical minimdm
without introducing excessive tension in the suspension system.
The particular need for such a working conductor system is found
in high speed work. At the present stage of development the
advantages which might be derived from the use of the catenary
type of suspension in city work, except in special cases such as over
railroad crossings, are in general considered to be outweighed by
the complications in its application and operation. There $re
three general types of primary suspension, namely: (I) bracket,
(2) cross span and (3) bridge.
Figs. 2 5 and 26 show typical examples of working conductor suspension in heavy electric railway work.
Length of Span. Messenger spans for catenary construction $re
usually 150 ft. with wood pole su port, and 300 ft. on steel bridges;
hanger spacing is generally 15
for pantograph operation, and
30 ft. for wheel operation.
General Details of Suspension. The earliest catenary consisted
of a messenger from whkh the contact wire was suspended by a
series of equidistant hangers. This worked well for short spans, but
on long spans the great difference in length, and so in weight, of
the mid and end hangers led to the development of compound
catenary. The Siemens-Schukert type, one of the earliest forfls,
employs a main messenger, a secondary messenger suspended from
the main messenger by hangers approximately 2 0 ft. apart, and a
contact wire suspended from the secondary messenger by hangers
approximately 10 ft. apart. The secondary messenger supports
come a t the quarter points between the main hangers or "droppers,"
which are of wire, about No. 8 gage equivalent, and are in two
parts, linked eyes in the adjacent ends forming a flexible joint; these
hangers attach rigidly to both main and secondary messenger.
The secondary hangers attach rigidly to the contact wire, but
loop over the secondary messenger so they are free to rise 2 of 3
in. This has proved very satisfactory in a number of continental
installations, and on the Midland Railway of England.
Multiple Catenary. Two main messengers have vertical sag
and horizontal deflection toward each other. The hangers, spaced
10 ft. apart, are equilateral triangles of pipe, the two upper corners
attaching to the messengers while the bottom one holds the con-

8.

CATENARY CONSTRUCTION
tact wire; the entire system is a t line potential, the supports a t
the bridges being insulators as well. As first installed on the New
Haven road, the contact wire, No. oooo grooved copper, was
carried directly by the triangles. These points, however, were
exceedingly rigid, while the midspan was comparatively soft, and
when the speed of the train was such that the time interval between
hanger points was the same as the vibration period of the collector,
the latter chattered with increasing violence until changed conditions threw the pantograph "out of step."
Secondary Contact Wue. The trouble with the multiple catenary cited above was successfully obviated by the following improvement by E. H. McHenry: A steel contact wire of the same
section as the copper is suspended from the latter by clips midway between the hangers, which keep the wires 1% in. apart,
center to center. The clips are rigidly attached to both wires;
it was feared that with the heavy shoe pressure a looped clip such
as is used in the Siemens-Schukert form might permit the lower
wire to turn sideways, in which case the shoe would foul the clips
with disastrous results. The London-Brighton & South Coast
Railway, to secure the same end, uses hangers of comparatively
light wire, the small ones holding the wire by a loop, while the
larger ones have jointed sides to give the desired flexibility. On
this system the stresses are all low, the messenger having a sag of
6 ft. in vertical projection for a span of not quite 2 0 0 ft.
Twin Trolley. Twin trolley construction consisting of two trolley
wires hung side by side and alternately suspended from the same
messenger was first used on the Chicago, Milwaukee, and St. Paul
electrification. On account of the fact that there are two trolley
wires the current carrying capacity of the combination is greater
than with a single trolley wire and on account of the uniform
flexibility of the arrangement there are no hard spots causing the
collector to leave the wire and cause arcing. This type of construction is suitable for both high speed operation and heavy
current collection with pantograph collector.
McHenry-Murray Suspension. The extensive experience with
the pioneer and McHenry forms under heavy service, and careful
study of the behavior of other types, resulted in the McHenryMurray catenary employed on the extension of the New Haven
electrification. I n this the main messengers, one for each track,
have at the quarter points structural steel cross-bents which in
turn carry insulators from which the secondary system is hung.
This latter consists of the secondary messenger, carrying, through
hangers 10ft. apart, a air of wires, upper of copper and lower of
steel, held 1% in. center to center by clips wh~chare midway
between the secondary hangers. The main messengen have only
to carry the mechanical load of the secondary system. The value
of this grounded shield as a lightning protection has been strikingly
shown in several electrical storms which caused trouble on the
older forms, while the new construction in the same territory was
unaffected.
Catenary Hangers. A great variety of hangers has been tried,
but very few of the forms have persisted, the majority either tak-

PIG.as -Examples of typical catenary construct~onin Amenca.

CATENARY CONSTRUCTION

q
.

ELECTRIC RAILWAY HANDBOOK

ing too much time to install o r shaking loose it, service, not 3 few
doing both. That which seems to meet conditions best is a mechanical three-screw clip rigidly attached to a steel strap, the top
of which is formed into a loop 3 in. or 4 in. long, permitting the
contact mire to rise to that extent. The other types with cam,
screw, toggle, wedge and other locking devices, many of them
exceedingly ingenious, are for the most part of interest only as a
matter of history. Early forms of strap hangers were hinged a t the
lower end, but this permitted the loop to tip over, ant1 with each
lift to crawl further out on the messenger until all the lift was taken
up and the flexibility lost. T o facilitate installation several ingenious ways of making the loop have been devised. In one the end
is twisted back of the main stem, leaving a space of such shape that
the hanger screws onto the messenger; another snaps into
the spring of the shaft normally keeping the l o p closed; wR$(;
third, also of the snap type, has the tail of the strap so bent up inside
the loop that, once snapped on, the messenger cannot get into a
position to wedge its way out.
Catenary Brackets. Brackets are of two classes: those made up
of paired angles or channels, separated a t one end to go either side
of the pole, to which they fasten by a through-bolt, or to clasp a
casting lagged or bolted to the face of the pole; and those consisting
of a single I or T seating in a suitable pole casting. Paired member
brackets offer facilities for the attachment of supports, insulators
and the like; single member forms offer much less opportunity
for corrosion. Brackets which clasp the pole and are throughbolted to it restrict the length of line affected by a break to a
few spans either side, but the brackets a t the break are usually
badly crippled if not completely wrecked; brackets less rigidly
secured swing clear, and while dropping a much greater length of
line are usually injured little if any. Abroad, brackets on special
iron poles are used for spans up to 300 it., and frequently for
double track work, either with center pole construction or carrying
both trolley wires on the one long arm. In American practice twotrack brackets are not so frequently used, spans or bridges being
employed instead; the latter are also extensively employed on longspan foreign lines.
Cross Span Catenary. Cross span support has the marked
advantage of causing minimum interference with the view of the
motorman, but the sag must be great or the stresses will be very
high. I t has been used in some of the early light cantenary construction and the New York, New Haven & Hartford Railroad has
made extensive use of this construction for yard work, and to a
small extent for the main line, employing heavy lattice poles to sustain it. The span wire attaches to the main catenary messenger,
while a steady strand attached a t or near the contact wire prevents
overturning under pantograph pressure. I n some instances the
span has contained a section of structural steel carrying insulators
to which the messengers attach. The New Haven type has long
wooden strains in the steady strand, very materially stiffening it.
Bridge Catenary. For heavy work most engineers have preferred
a stiff cross member to the high or heavy poles required for cross

CATENARY CONSTRUCTION
span support. Bridges have a wide range of design, from the light
paired angle irons bolted to wooden poles, one on each side of the
track, as on the Midland Railway, England, and certain Continental
lines, and the light but rigid entirely structural material frames of
the Archbold-Brady type, to the very substantial structures
of the early New Haven work. The importance of the traffic
involved in the latter case and the lack of data as to behavior in
such service justified the use of a conservative design in the pioneer
installation.
Alinement of Contact Wire. Abroad the line is zigzagged from
side to side, usually about I ft. each way, to distribute the wear
over the top of the shoe. I n America the line is almost always
centered, probably for esthetic reasons, although it is found that
the natural side sway of the pantograph is ample to give all necessary side travel. Even with center location i t is essential that
track and overhead departments maintain close cooperation lest a
change in super-elevation without a corresponding line shift result
in the shoe being thrown entirely clear of the contact wire because
of unavoidable side play.
Turnout Diverters. At turnouts the diverging wire is lifted
abov6 the level of that of the main line, but for pantograph o era
tion early American catenaries used a gridiron of wires to &1 in
the space between wires to a point where the end of the shoe could
by no chance go over the branch wire. Later designs follow the
foreign practice of merely raising the branch wire, which, however,
is rigidly held to that elevation throughout the fouling space by
special double hangers.
Catenary Curve Dressing. Curve dressing is complicated, particularly on flexible simple catenary. T o prevent tipping there must
be two attachments: if made to the hanger rod the friction on the
messenger absorbs much of the flexibility; if one attachment IS made
to the hanger rod or contact wire and the other to the messenger
the pull-together effect of paired strands is equally undesirable.
The best solution is the use of a spreader which keeps the two
parts parallel for a t least 6 It. or 8 ft. from the catenary, beyond
which they come together in a common pull-off. On rigid hanger
lines this is unnecessary and the strands are attached to the messenger and to the bottom of the hanger rod or to an attachment to
the contact wire which, for pantograph work, is sufficiently offset
to prevent fouling. Sharp curves with several pulls per span
employ either a bridle between supports to which the pulls are
run radially, or else the pulls are run to the supports in fans; light
curves use either the bridle or fan method, or on heavy construction use a special pull-off pole. On light work flat curves offer so
little side pull that the weight of the strands and insulator causes
heavy sag; in one instance this was partly obviated by using a
light single strand, which a short distance from the catenary was
divided into two groups of three wires each, the seventh being
servcd on to prevent further unlaying. One group of three wires
was then taken to the messenger, the other to the trolley, while the
insulator was cut in a t the bridle. Foreign lines and a few American employ the rigid steady brace attached to the main support,

580

ELECTRIC RAILWAY HANDBOOK

and rarely to a pole specially set for it. The elimination of ypgial
pull-off poles is secured, and the disadvantage of the line loading
by the steady brace is obviated by extending the bracket to form
a hook to which a pull off strand IS attached. In case the pole IS
on the outslde of the curve the attachment is made directly to it.
The use of cross-bents on the Harlem Branch electrification of the
New Haven permitted the use of a bridle tied to these bents,
pulling them partly to the desired offset from the chord of .the
messenger, a further shift of the insulators on the cross-bents pnnging the contact wire over the track center. The most admirable
device, however, is the Murray curve hanger, the head of which
firmly grips the messenger, holding the shank a t an angle of
approximately 45 deg. from the vertical, while the lower end Is a
duplex clip, which is vertical. The lengths are such that the C ~ P S
are true to line, making the chords but 10 it., while the inclined
shank permits the contact wire to yield to the pressure of th2 shoe
against the torsion of the messenger and the pull of the contact
wire In connection with shortened support s acing to keep the
hanger lengths within reasonable limits, this %!erne,
employed pn
the recent New Haven work and now used on the PennsyI\~anla,
gives an almost erfectly true alinement.
~ e m p e r s t u r8hanges.
s
Changes in temperature fend to cause
corresponding length changes in the various members of catenary c o n s t ~ c t i ~but
s , tbe actual result is laxgely aflected by
character of the construction. The messenger, contractiflg or
expanding, tends to lower or rise, the movement being a maximum
a t mid-span, and zero a t supports, the contact wire tends to do
likewise, but a t a dlflerent rate because of its different sag. The
hangers tie the two together and compel equal movement, giving
to the contact wire in cold weather an inverted sag, so to speak,
and in hot weather slack wire between the hangers. On the Syracuse, Lake Shore & Northern Railway, the contact wire, which
was erected in winter, was given a very high initial tension and
an inverted sag of I ft. to the standard goo-ft. span, so that .the
heavy summer traffic should have a fairly tight straight hne.
With this one exception, which, it is said, did not entirely meet
expectations in this respect, American engineers have considered
compensation devices not worth the complication, a belief well
justified by the behavior of the lines which are in service without
such devices.
Messenger Wire Design. The Arner. Elec. Ry. Eng. Assn.
recommends a factor of safety of 4 between working and ultimate
tension of messenger wire, and that the messenger be pulled up to
such a tension when unloaded (measured by dynamometer) that
when the load of trolley and hangers is added the trolley wire will
hang in a straight line. Charts prepared by the Committee on
Power Distribution show the relation between sag and tension in
loaded and unloaded messenger, together with temperature correction. Figs 27, 28 and 29 reproduce such charts for steel messenger
with 180 and 300 ft. spans and for copper messenger with 300 ftspans, respectively. The Manual of the Am. El. Ry. Eng. Assn.
shows similar charts for steel messenger with 1 5 0 , 2 1 0 , 2 4 0 and 270
ft. spans.

CATENARY DESIGN

FIG 27 -Steel messenger sag t5nslon curves for 180 ft. spans.

FIG.z 8 . S t e e l messenger sag tenslon curves for 300 ft. spans.

Coef&~enfoflmear

PIG.29 -Copper messenger sag tension curves for 300 ft. spans.

581

ELECTRIC RAILWAY HANDBOOK

On these charts the true length of a parabolic curve of variob,
sags is expressed by the horizontal ordinate from the line of origi,
to the curved line. For instance, on the 300-ft. span diagram (Fig28) the true length of a parabola of 60 in. sag is represented by the
distance a 300 ft. The stretch of a steel messenger under its
own weight is represented by a similar distance measured to the
straight line marked I on the diagram. The stretch under multible
loads is marked by the lines 2, 3, etc., and intermediate ratios may
be obtained by interpolation. For instance, the stretch of a stgel
messenger loaded with 2.8 times itsown weight would be indicated
by the distance b measured from the origin to a line interpolated
on the diagram. Subtracting, a - b = c, then c
300 it. equals
the true unstressed length of the strand necessary to obtain the
given sag. I n this instance for a final sag of 60 in. c will have a
minus sign. Marking the quantity c on the edge of a card abd
keeping the card edge horizontal, move i t along the length cu%e
until i t fits between this curve and the line of unit stretch, and read
the required sag a t the left, in this case 38% in. This means that
a steel messenger which will be ultimately loaded with 2.8 times its
own weight should be strung to a sag of 38% in. in order to have a
final sag of 60 in. under full load. This would be the case w i q a
Tie in. steel messenger weighing 0.415 lb. per foot to support a 4i0
trolley weighing 0.641 Ib. and hangers weighing 0.1 lb. per foot
span. The total of these weights is 1.156 lb. per foot and the rabo
01'1 .I gdio 0.41 5 equal4 2.d:
I??
seess per square ~ncd-lirmessenqr
may be read in the u per line of figures. I n the case shown, tile
stress would be 2 2 , m f b . per square inch when loaded. This would
represent a pull of 2640 Ib. on the 0.1204 sq.in. area of a 746 in.
steel cable.
Temperature effects may be obtained from the triangular scale
at the left of the chart, and the quantity shown should be added to
or subtracted from the quantity c. Let it be assumed that the me+
senger is to be sagged when the temperature is 50 deg. belo,
normal. Measure the distance d on the temperature diagram
a t the 50 deg. point. As we are considering a lower temperatur
the wire is shortened, in other words d is a minus quantity.
is also minus, c and d may be added graphically and the marker
moved along the length curve until it fits between this curve ar d
the line representing stretch, and the required sag is read a t tte
left, in this case approximately 29 in. If the stringing temperature
were 50 deg. above the standard, t t e quantity d would have a
plus sign and c, being minus, would be subtracted from it. T4is
may be done graphically, the result being e which has a plus sign
Where the sign is plus the scale must be set above the intersectic,,
of the curve and the line expressing stretch. I n a similar manner
the effects of temperature on other spans may be studied.
.
(Fig. zb)
On the diagram for the copper messenger on 3 c ~ f tspan
there is indicated a similar calculation assuming a ~ 0 0 , o o oc . I ~ .
messenger, a 4/0 trolley and hangers weighing 0.1 Ib. per foot of
span, using the same letters to indicate similar quantities.
this case the ratio of total weight to weight of messenger would
1.5. If the normal sag in this case were to be 60 in. when loaded,

AS^^

. CATENARY DESIGN

583

for a stringing temperature 50 degrees below normal the messenger

should be pulled to a sag of about 38 in. and for a temperature 50
deg. above normal a sag of 63 in.
In a similar manner the effects of ice load may be studied, expressing the total load as a ratio to the weight of the messenger cable and
interpolating a line on the diagram to represent this ratio.
Catenary Hanger Spacing. The Amer. Elec. Ry. Assn. recommends that hangers be spaced 15 ft. apart where pantograph
trolleys are used; 30 ft. spacing may be used where pole trolleys
are used.
Catenary Hanger Length. Hanger lengths will depend on the
length of span and the minimum distance established between
lowest point in messenger and contact wire, and may be determined
from the formula:

Where h = length of hanger, inches

S = length of span, feet
d = maximum sag in center of span, inches
z = distance from center of span to hanger, feet
a = fixed distance between trolley wire and the lowest
point of sag of messenger, inches.

-50'-

CURVE CONSTRUCTION

PIG.30.-Types

of span using standard line material for trackless

used with swivel harp. wheel collector.

trolley.

Overhead for Trackless Trolley. The construction of the overhead working conductors for the trackless trolley bus usually is
uite similar to the standard 600 volt overhead trolley construction
page 550 et seq.), except that there are two parallel trolley wires,
one positive and the other negative. The two trolley wires generally are carried 2 4 to 30 in. apart, with a strain insulator cut into the
span wire between the hangers, although in one installation (Staten
Island) the spacing is as low as 14 in. on straight line and 15 in. on
curves, the two trolleys being carried by a single hanger, with

ELECTRIC RAILWAY HANDBOOK

porcelain insulators around the stud bolts. I n an installation a t
Baltimore there are two pairs of 2 / 0 round copper trolley wires,
the positive and negative wires of each pair being spaced 24 in.
apart. The trolley wire is hung from a span by the standard
straight-line hanger and clincher ear with porcelain strain insulators
cut in the s an between each positive and negative wire and also
between eacE outside wire and the pole. Two q/o feeders, one positive and the other negative, are run on cross-arms about two-thirds
of the total distance and each pair of trolley wires is tapped to the
corresponding feeder at intervals of approximately 400 ft., that is,
each wire is tapped to the feeder of the same polarity every 800 ft.
For the urpose of turning the busses, loops have been built a t
each e n t o f the line, using wood strain insulators and standard
pull-overs. The wood strain insulators are of proper length to
give the necessary clearance between the conductors. Fig. 30
shows both a straight-line and a curve-suspension span.
On some of the installations of trackless trolley line it has been
impossible to turn the busses on a loop and it has been necessary to
build a wye. The standard frogs and cross-overs are not designed
so that a troUey wheel mounted in a swivel harp will follow in the
desired direction and it has been necessary to put guides on the
frogs and cross-overs so that the trolley wheel will operate properly;
this difficulty also has been overcome by the use of pins or springs.

Heavy Electric Traction

There are three systemb of power distribution which may be
considered for heavy electric traction or the electrification of main
line railroads: (I) direct current, (2) singlephase alternating current,
and (3) three phase alternating current. These are described in
the Standard Handbook for.Electrica1 Engineers as follows:
Direct current eleclrijicatzon conslsts of direct current motor
equipment fed from overhead trolley or third rail which receives
power from substations containing rotary converters or motorgenerator sets. I t is the system in almost universal use in city
and interurban electric railways, but varies as to the voltage of the
direct current supply. As a result of perfecting the commutating
or interpole motor, it has been possible to wind motors for higher
voltages, and 1500-volt direct current motors are now in commercial operation. These motors are connected two in series when
trolley. Still higher voltage equipments are
fed from a 3-volt
ssible, as indicated by factory experimental apparatus
~ : f ' t K e s t e d , and every requirement of the heaviest main h e
service can be provided for a t moderate cost of locomotives and
distribution systems. The chief disadvantage of the direct current
system has been the relatively high cost of feeder copper and
substations and their low efficiency, but this has been largely overcome by the adoption of higher voltages and the automatic substation. The advantages of this system include simplicity, low cost
and high efficiency of the locomotives, the option of trolley or third
rail distribution, the ssible use of bi-polar gearless motor construction, and the use of &anced three phase power supply of any ire-

HEAVY FLECTRIC TRACTION

585

quency and with no serious telephone or telegraph interference.

Large direct current locomotives are in service upon the following
main line electrifications:
600 volts
P. R. R. New York Terminal.. .................
B. & 0.Tunnel Railway ......................
600 volts
New York Central Railway.. .................. 600 volts
Detroit Tunnel Railway.. ...................... 600 volts
Butte. Anaconda & Panfic Railway.. . . . . . . . . . . . z.400 volts
Canadian Northern Railway. . . . . . . . . . . . . . . . . . . 2.400 volts
Chicago. Milwaukee & St. Paul Raiiway.. ....... 3.000 volts

Single Phase. The use of a single phase alternating current permits the use of very high voltages on the trolley-as high in fact
as there is any reason for using. In this country the voltages vary
from 3300 to 6600 which are used on interurban lines and some of
the earlier main line electrifications, to I 1,000which is now acce ted
as the standard for main lines. In Europe voltages as hig[ as
15,ooo are in use, but it is felt in this country that this is higher
than necessary. With the exception of one small installation,
25-cycle current is used for all single phase railways in this country.
In Europe, however, the preference is given to 15 or 16% cycles. A
frequency of 25 cycles has the advantage of being a standard for
power distribution in many localities, and in lower cost of power
station and substation apparatus. I t has the disadvantage of
greater size and weight of commutator type motors, and thus
Iimits the output from a given space on a Iocomotive. On short
single phase lines the energy may be fed directly from the generators to the trolley, but for long Iines, it is necessary to use a high
voltage for transmission and feed the trolley from step down
transformers. These transformers are usually located in substations
but are sometimes of the outdoor type. I n either case regular
attendants are not required.
Two types of equipment are a t present operated from single phase
trolleys: (I) Cars and locomotives having commutator type motors
with series characteristics; and ( 2 ) split phase locomotives having
induction motors which derive the additional phases from a phase
converter in the locomotive cab. The phase converter consists of
a plyphase stator and a squirrel cage rotor, the combination
belng in effect an induction motor fed from single phase supply
and getting its polyphase field by means of its rotating armature.
Both types employ step down transformers on the locomotive so
that the motors have low voltage applied to them. The former
type is particularly suited for passenger service with both multiple
trains and locomotives. I t is also well suited for high speed freight
and switching service. In such service the variable speed characteristics are valuable. The split phase locomotive is particularly
desirable where heavy grades are encountered which necessitate
high tractive efforts and where the automatic regeneration of the
motors on down grades saves power and decreases the likelihood of
accidents.
The disadvantages attending the use of single phase current are
in the relatively high cost and low efficiency of motive power,
the cost of supplying single phase power and the inductive interference of the single phase trolley with neighboring telephone and

586

FLECTRIC RAILWAY HANDBOOK

telegraph circuits. The latter has been largely overcome by the

installation of series track transformers, the primary of which carries
the trolley current and the secondary the return current in the
track. The great advantages of the single phase system of distribution lie in the high efficiency of the substations and trolley and
the low cost of the distributing system. Large single phase locomotives with commutator type motors are in service on the foUowing main line electrScations: New York, New Haven & Hartford
Railway, Grand Trunk Ry., St. Clair Tunnel, Boston and Maine
Railway-Hoosac Tunnel, Spokane and Inland Empire Railway.
This system has also been adopted for the New York, Westchester
& Boston Ry., and the Pennsylvania Railway, both using multiple
unit trains, the former for New York and the latter for Philadelphia
suburban service. The Norfolk and Western Ry. is using split
phase locomotives with three phase motors for heavy mountain
grade service. Liquid rheostats are used in starting these locomotives. The Pennsylvania R. R. has also built one split phase
locomotive, with which it has been experimenting.
Auto-transformer Scheme of Single Phase Distribution. In the
auto-transformer scheme (also known as the semi-balanced system)
of single phase distribution developed on the New York, New Haven
and Hartford R. R., the contact conductor of the 1 1 , m v o l t
catenary system is connected to one side of a 22,-volt
single
phase line, auto-transformers distributed along the track are connected across this 22,00~~voIt
line, and the middle point of each of
these auto-transformers is grounded to the track rails. Thus this
system is similar to a three wire system except that the direct load
is on one side of the circuit, the other side of the circuit receiving
its share of the load through auto-transformers which act as balancers. By this system the o erating economies of a contact
conductor voltage of 11,a n 8 a transmission voltage of 22,000
are combined, and this has been done with the elimination of the
disturbance which the ordinary single phase system, having one
side grounded and the other side connected to the contact conductor, caused in adjacent telephone and telegraph circuits.
Three Phase. The three phase system of distribution is not used
to any extent in this country, chiefly on account of the difficulties
arising from the double overhead trolley which complicates the
yards and cross-overs. It is used extensively in Italy, and the
three phase locomotives in use there are the lightest electric
locomotives of their capacity in the world. Three phase locomotives are usually used with about 3300 volts on the trolleys and
this voltage is applied to the motors without the use of lowering
transformers. The Cascade Tunnel on the Great Northern Railway
has the only three phase locomotives in operation in this country.
High Conductivity Trolley Wire, Round and Grooved
(A.E.R.E.A. and A.S.T.M. Standard)
Material. The material of which high conductivity trolley
wire is made should be electrolytic or low resistance lake copper
wire bars of quality specified by standard specifications for electrolytic copper wire bars, cakes, slabs, billets, ingots, and ingot

TROLLEY WIRE SPECIFICATIONS

587

bars, or lake copper wire bars, cakes, slabs, billets, ingots, and ingot
bars, of the American Society for Testing Materials, 1921 Book
of Standards. Necessary brazes in trolley wire should be made
in accordance with the best commercial practice, and tests upon
a section of wire containing a braze should show a t least 95 per cent
of the tensile strength of the unhrazed wire. Elongation tests
should not be made on test sections including brazes. The wire
should be of uniform size, shape, and quality throughout, and should
be free from all scale, flaws, splits and scratches not consistent with
the best commercial practice.
Round Trolley Wire
(A.E.R.E.A. and A.S.T.M. Standard)
Dimensions. Dimensions of round trolley wire should be expressed as the diameter of the wire in decimal fractions of an inch,
using not more than three places of decimals; i.e., in even mils.
Wire should be accurate in diameter. Variations of I per cent
over or under nominal diameter are permissible.
Tensile Strength and Elongation. Round trolley wire should
be so drawn that its tensile strength will not he less than the minimum values given in the following table:
Are3, cir. mils

Tensile strength,
Ib.persq.~n.

Elongation in
~oin..percent

The elongation should be determined as the permanent increase

in length due to the breaking of the wire in tension measured
between bench marks placed upon the wire originally 10 in. apart.
The fracture should be between the bench marks and not closer
than I in. to either bench mark.
Inspection for Defects. For the purpose of determining and
developing defects which may be prejud~cialto the life of trolley
wire, owing to its peculiar service, as compared to that of copper
wire for other purposes, the following twisting tests should be made.
Three twist tests should be made upon samples 10 in. long between
the holders of the machine. The twisting machine should be so
constructed that there is a linear motion of the tail stock with
respect to the head, and the twist should be applied not faster than
10turns per minute.
All three samples should be tested to destruction and should not reveal under test any seams, pits, slivers or
surface imperfections not consistent with the best commercial
practice. At the time of fracture the wire should he twisting with
reasonable uniformity. Wire should not be considered satisfactory which does not withstand a t least 9 turns before breaking.
Resistivity. Electric resistivity should be determined upon fair
samples by resistance measurements a t a temperature of 2 0 deg. C.

588

ELECTRIC RAILWAY HANDBOOK

(68 deg. F.). The wire should not exceed in resistivity p . 7 7 b.

per mile ohm.
Density. For the purpose of calculating weights, cross-sections,
etc., the specific gravity of copper should be taken as 8.89 at 2 0
deg. C.

"0
NomlnaJ clr. mlla 133.200
Actual slea QltLU eq. in.
Actual welght Z O l Ib. per ml.

Nomlnal clr. mils 2U.m

Actual ares 0.W sq. In.
Actud welght 3382 Ib. per ml.

FIG.31.-Standard

$0
Nomlnal clr, mlla -1
Actual area 0.l314 sq. in.
Actual weloht 2669 Ib. per mi.

300.000 c.m.
Nomlnal clr. mils
Actual arsa 0.2350 w. in.
Actual weight 4?Bb lb. per mi.
sections of grooved trolley wire.

Grooved Trolley Wire

(A.E.R.E.A. and A.S.T.M. Standard)
Standard sections are shown by Fig. 31. These are known as
the "American Standard Grooved Trolley Wire Sections."

TROLLEY WIRE SPECIFICATIONS

589

Dimensions. Dimensions of grooved trolley wire should be

expressed as the area of cross-section in circular mils, the standard
sizes being as follows:
300 ooo cir. mils weighing 4.786 Ib. per mile.
211:60o cir. mils weighing 3.382 Ib. per mile.
168.100 cjr. mils wejghjn~2.669 lb. per mile.
133,200 clr. mils weighing 2.201 lb. per mile.

Grooved trolley wire may vary 4 per cent over or under in

w e i ~ h tper unit length from standard as determined from the
no&nal-cross-section:
Physical Tests. PhysicaI tests on grooved trolley wire should
be made in the same manner as those upon the round wire (see
page 587). The tensile strength of grooved wire should be a t least
95 per cent of that required for round wire of the same sectional
area; the elongation should be the same as that required for round
wire of the same sectional area. The twist test should be omitted.
Resistivitg. The requirements for resistivity are the same as
those for round wire of the same sectional area (see page 587).
Low Conductivity, High Strength Trolley Wire. At this time
( 1 ~ 2 no
~ ) specification for special trolley wire of low conductivity
and high tensile strength has the approval of the Amer. Elec. Ry.
Assn. The committee on Power Distribution in 1923, for the
purpose of giving member companies an opportunity to study
definite data and so promote discussion, but not as a recommendation, tentatively suggested certain limiting values for high strength
wire of 40 per cent conductivity as follows:
Tensile strength
round wire

Lb. per

Steel Trolley Wire. Steel wire is sometimes used as a secondary

contact wire in catenary construction, and also has been used as the
main trolley wire. In the latter case it must be supplemented by
a larger cross section of feeder than is required by a copper trolley,
but it has been claimed that the first cost is less, the hfe is longer,
the reliability is greater, the temperature coefficient is less and the
maintenance cost considerably less than copper. The Pacific
Electric Railway has had more than 100 miles of steel trolley in
use on b t h 600 and I 2 0 0 volt lines, and in both direct and catenary
suspension. In places where rapid accelerating occurs, tbe steel
wire appears to wear almost as rapidly as copper, a t other places,
considerably less. Galvanized steel wire of the 4/0 grooved section
was used, and feed taps were installed much more frequently than

590

ELECTRIC RAILWAY HANDBOOK

with copper troIley. I t was found that the maintenance on steel

trolley was far below that of copper, since steel is much harder.
The clips hold much better in the grooves, and the wire does not
break if the trolley pole hits a span. There are no breaks due to
crystallization a t the ends of ears, splices or switches. I n most
cases, steel wire that has been accidentally broken and shorted on
the rail can be restrung, whereas copper wire would become annealed.
Steel is easier to pull up and can be handled for double the length of
copper without sagging. Steel wire does not cut out on the sides
when out of alinement nor does it pound flat a t clips as does copper.
Practically all the wear is due to burning and arcing a t acceleration
points.
As the correct welding of splices in steel trolley wire is so important, the method developed on the Pacific Electric system is worth
description in some detail. The
equipment consists of a Prest-0Lite outfit of gages, welding torch
and tips. The torch is style H,
Shoring TWO Ends
and the size of the tip No. 5, this
Prepared for Weldlng
size being most satisfactory for
No. 4/0 double-grooved wire.
The welding rods are of z -in. or
%@-in. diameter nickel steel.
Fig. 32 shows the different stages
of a weld. First the opposing ends
of the wires to be joined are cut
Weld Halt Completed
a t an anale of A< deg. and a mace
of 56 in. -is left'8etGeen the& to
proGde for expansion. The flame
of the torch, which is next applied,
disposes of the galvanized scale
which coats the wire. As soon as
both ends of the wire are at the
melting point, the welding rod is
Flniahed f eld
introduced, care being taken to
keep both ends a t about the same
PIG.32.-Weld joint in steel trolley
mre.
temperature. As the nickel steel
is melted from the rod. it is fluxed
with the steel trolley wire by means of the flame, staking a t the
bottom and building across and up. As soon as all necessary metal
has been added, the weld while still red hot is filed down to a smooth
underrun and shaped up as much as desired. Then cold water is
poured on the wire, beginning about one foot on each side of the
weld and running toward the weld. Finally the weld is reached
and wetted until erfectly cold. The average length of time
to weld the wire in t t e air varies from ten to fifteen minutes, according to accessibility.
Bending, heat and resistance tests were made on the galvanized
wire by the Pacific Electric Railway, as follows: The bending test
showed that the wire could stand six right-angle bends in a vise
before breaking on the seventh bend. To find how much heat could
be withstood and how much current could be carried, alternating

SELECTION OF POLES

591

current was passed through the wire until it was red hot. After
cooling, the wire was found to be more flexible but still tough
enough for use as trolley wire. I t was given ten right-angle bends
without breaking. Direct current, averaging 750 amp., was passed
through the wire for five minutes. At the end of this time it was
heated to a cherry red and the galvanizing had burned off, but the
wire was otherwise uninjured. This latter test proved that under
normal senrice conditions the wire would not be damaged by
high temperature. However, even if the galvanized coating was
burned off by a heavy ground, this would not be so serious as the
annealing of grounded copper. Volt-ampere readings showed the
resistance to be 0.000342 ohm per foot a t IOO deg. F., or 6.53 times
the resistance of No. 4/0 copper a t loo deg. F. As steel is 10.6
per cent lighter than copper, its resistance per unit of weight is
j.g3 times that of co
Overhead L i e daterial. Standard swcifications for various
items of material for overhead constructio& 750 volt direct current,
direct suspension, have been recommended by the Amer. Elec.
Ry. Eng. Assn., and are shown in the Manual of that association.
The specifications cover iron and steel fittings, wood cross-arms,
bronze castings, seven strand steel cable, switch boxes, tree and
cable guards, wood insulator pins and brackets, porcelain strain
and feeder insulators, and wood break strain insulators.

Selection of Poles
The selection of the pole for trolley line construction is dependent upon the type of line which is to be supported, and involves
three elements, namely, height of pole, width of street for span
construction, and sag desired in span. The allowable sag having
been decided upon, the next step IS to find the horizontal pull for
the span wire. This is found as shown on pages 570 to 573, and
the pole of the desired length must then be selected to withstand
the pull thus determined. If the pole selected has a deflection of
say 3 in. for the required pull, then the pole must be given a 3 in.
rake if it is desired to pull it up vertically when the span is completed.

Steel Poles

in which x = deflection in inches, 18 in. from top of pole

P = horizontal pull, 18 in. from top of pole
I = length of pole from point of application of P to ground
line (6 ft. from base end), in inches
E = modulus of elasticity of steel (2g,ooo,ooo)
I = moment of inertia of base section.
Deflections from various loadings together with descriptions of
poles are given in the following tables. The variations of lengths of

8&2%gg$

N
UUUUU
mmmmIloom

- a m 0 1
-

~ - a~w m m - a y

W
mm-r-roa

,,az%s

ssss..s

heavy

oaumpo

m w ~ e w w

.+nu
I
w o g z ~ g8 ~ l i ; u w b - r
:~UO,W

Length of pole. ft

.lo.-

o w e ,

w
*

u m
u m
u m
um
u%
m

m*

kF82g

Standard

ard
We~ghtofpole.lh

I"""
""I"""
""

mmmmmm

mmmmmm

Butt

oooooo

o m o o a a

Top

-%=st.
-UUW01

g Z l l i E

!t*08L:

5848k

tB38$8

YEPgE

- - a +2
w u - 9
UWP

8 $ 2 8 8 % Zg2.9pii

kbuo I

889%~'$~ 8 %

D
5.

2
'F!

%
t#

-1u4-au-r

""

--

; $K

VI

n r

r UWP

wouwu

VOW

-"I

CI"ICtU

SCg4

8 C

* " I I U U

XUtI$S 8

I n - u v

-a
n

8
"

GI

GI.

7:

8
U

ea

a
5

-8 2

"""UU

00,

8 ZgE8
-UUUW

.,

tBZS8

23%

h, b2

--

I U U

a
a
"
0
a
-U
b

828

&
o

k w w a%+

S* U -U

, a ~ ~ ~z

888388

8
-5

U
P
U

w
P

(P

--

0"

Z3
a
g

%,

%3

gg

e3

D.'D

:
5
oB

S*
a7

-n
I I

~$ ~ H & ~
E

Greatest
~ safe load.
~ lb

'

P O 1 1 1 1 % & % P I % Length of pole. f t

8 0 0 8 a N V H A V M W X 3IXL3313

Z6Q

So9'1

r~

SzEcl

L6L
g6L
SSS

&

g9a1

&

EpI1r

IE'C S9.266'1 S9'1 CE' L

Lg'P W'E 26'2 FV'z 6%' L

or
or

61 261'1
61 0x6

9
9

L
L

V0.E CS.2 IS' L

' " '
" . . . " . . . . . I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .go.9
. . . . . .10,s
. . . . So'*
L
. . . . . . . . . . . . I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .Oy'PEg'E
. 9S'P LL'
16' L
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . LP'S
.
""6t'pOE'1
. . . . . . . . . . . .' . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .". " E g ' r L L
. . . . . . . . . . . .1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . " " g V ' z L

ox
01
01
01
01
or

61
61
61
61
61
61

$10'1
6EL
tpL
c8S
SSS
6EP

9
9

6
6
6
6
6
6

61
61
61
61
61
61

SCS'r
PLZal
gtf'~
990'1

6
6
6

61 S66
61 OZL
6 6 L

0C.S f9.P L6.0

........................

z.0'2 CL'I PP'I

bg.2 oC'Z 26'1
Sg'z '3'2 PO'Z
...........
0o.P C t " Z gg'z
&'P
go'P
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .oS'E
. . . . . 26'2
.g

............
............
. . . . ..I..

S I ' l g g ' ZL' CI'

E S ' I S I ' I F ' 61'
C9'I 22'1 20'1 02'
a . 2 zL'1 0'162'
ET'zSL'I 9P'I d '
0 L SI

L
L
L
L
L
L

. . . . . . . . . . . . . . . . . . . . . . . . . . . . S ' S 9 . P 9S'E L9.2 22'2 9' L

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0P.S
. SO'V LE'Eg9' L
g

lo g

11

~ o o v a~zo o I *~008'1
z
~009a1
~oov-I~ o o z o~I o o o ~ I ~ 0 0 9 ~ 0 0 s ~ ~
'0

'u! 'aiod 30 pua a a q uroaj ,,a1 pa![dds spunod u! p s o ~

'13 'sql8ua1
uo!z=s

ig

r
g
g

L
L
91'1 9
9

$
0,

.
.?

:/

a3
ae
tl

p a l a l s s s papsol p u s play u a q a suorlsayap a[qeqoad 'paa a a ~ ~

~ 0 1 ,,gr
3 palnssaur uo!loayap p u s pa!ldds'peo[-pun018
ui ,,o ,9 qaa aIod

....
8

....

.'.,

CE

EE

fr
EE

.9. . . . . . . ET

.S" '

S
V

"
9"

CE
EE
EE

6
6
g
8
L
1

....
01
....
".. 6

d
ZE
zf

,0.,

z.C

L
....

01

....
6

,,, ,

EF

tC

." ' rc
L

....

2
s

cr

I ":

ife gs f i fe ;8;

--do& alPP!N

ma

adrd jo p u g p a s aZ!S

?
i

g
....

salod )o uo!?duwaa

'$*
.C

STEEL POLE TABLES

CCCC

,.C,.,.CC

,.,.,.,.CC

CCPCCC

CPCCCC

596

ELECTRIC RAILWAY HANDBOOK

$
$2

Y'O?
-0

-4
9;

OQb
o n o

0
p
gO..h
:;

"
2:
hO

on

t-m*
o

- -

dbb 0

o a o
3. P. P. W

$
?

. t oro * * ,m
0

. .P . . 13
?C11
gz l o n o

*-

Y)

o o o
m a n

_o

"A

dd

'

rC*. *.*W d
ano- a
a- - - o m

rtrtir

'am

- o m c

20

a h n o

."
n

00

1,

rt

&

--.. .. .. .. ..

:. g.2
. .z%,n,l
. . . :
. q

...
..
..
...
..
..

. . . . . . .
.. .. .. . . . .. .. .. . .
.. ... . .. . .. . ... ... ... ... ... ...
.. .. .. .. .. .. .. .. .. ..
... ... ... ... ... ... ... ... ... ...
.. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. ..

-2
o

.
...
...
..
..

.
...
...
..
..

. . .
... ... ...
... ... ...
.. .. ..
.. .. ..

"

STEEL POLE PIPE TABLES

597

598

FLECTRIC RAILWAY HANDBOOK

the sections of a pole within commercial limits have but little effect
upon its strength, stZness, or weight. The section lengths given
conform closely to those usually employed. In general a number of
poles will satisfy the requirements, and choice must be made from
those of the proper length. Usually the lightest weight pole would
be selected unless the deflection was considered too great, in which
case the choice would be determined finally by the deflection.
Steel Pipe Used in Making Poles. All steel pipe used in making
up poles should be good quality of mild steel of uniform thickness,
free from bad welds, cold shuts, outside slivers, cracks, flaws, open
seams, rivets or other imperfections which would affect the strength,
life or appearance of the pole. The tables on pages 596 and 597
give data on steel pipe
which was used in making up the tables on
.Pages 59: to 595.
Joints m Steel Poles. The joints should be 18 in. in length and
should be made hot without reducina the diameter of the smaller
pipe more than 34 in., and without anyreduction in the thickness of
either pipe. The completed joint should caliper the same outside
diameter for its entire length, should be free from cracks or flaws,
and should be water tight. Any pole when dro ped three times
from a height of 6 ft. upon a solid wooden bloc[ on a rigid base
should not show any telescoping a t the joints.
Ground Sleeve. A ground sleeve, if required, should be made of
standard weight pipe 2 ft. in length; it should be shrunk on the
butt section with the bottom of the sleeve 5 ft. 6 in. from the base of
the pole.

Wood Poles
Classillcation According to Purpose, A3.R.E.A. Standard.
To determine the character of poles to be used in trolley line construction, they may be divided into three classes, A, B, C.
Class A. For span coilstruction on streets or rights of way where
a 35 ft. span is required, or for heavy feeder lines carrying from
one to six cross-arms.
Class B. For span or bracket construction where spans are not
more than 35 ft., or bracket line construction carrying two transmission circuits, one feeder arm, and two telephone and signal
arms.
Class C. For constructing telephone, signal and other light auxiliary lines where no side strain is required.

599

WOOD POLE SPECIFICATIONS

Chestnut Poles

M I ~ W
DIMENSIONS
OF CHESTNUT
POLESRJ INCHES
(A.E.R.E.A. STANDARD)
Length
of poles
(feet)

Class A

Class B

Circumference

Circumference

Six feet

Class C
Circumference

Six feet

Six feet

In.

Quality of Chestnut Poles and T i e of Delivery. A pole should

not have been cut over 2 years, nor should it have been cut when
dead. I t should not be shaved, nor should it have a crack or large
season checks or short crook or a crook or sweep in two planes, or a
short reverse curve. Poles over 30 ft. in length should have less
sweep than the shorter poles and as follows (A.E.R.E.A. Standard):

40 ft.
45 it.
so ft.
is ft.

60 it.

65 ft.
7 0 ft.

Sweep not ovet

1 0 In.
10 in.

14 jn.

15 In.

The sweep is measured between the 6 ft. mark and the top of
the pole.
All poles should have sound tops. Poles with double tops should
be examined carefully for split tops or rot where the two parts
joln. All poles should have reasonably sound butts. Hollow
butts should be carefully examined, and poles having them
rejected, if the hole runs over 4 it. Poles with hollow butts should
be rejected if there is evidence of decay a t the further end of the
hole inside the pole. There should be no sap rot. There should be
no "cat faces" unless they are sound and small and the pole has
an increased diameter a t the "cat faces." There should be no
"cat faces" near the 6 ft. mark or within ro ft. of the top. Poles
should be examined carefully for black knots, hollow knots or

600

ELECTRIC RAIL WAY HANDBOOK

projections caused by overgrown knots. Overgrown knots should

be trimmed off and examined for internal rot. Thereshould beno
evidence of rot a t knot holes nor should there be large hollow
knots. There should be no top, butt or black knot which has been
plugged. The pole should be free from woodpeckers' boles or
ev~denceof ants, worms or grubs.

Eastern White Cedar Poles

(A.E.R.E.A.STANDABD)

(izfes)
Circumference

of

from
(inches)
butt

Circumference

(i,'ph",,)frombutt
(inches)

Circumference

p
'(,ih"),

frombutt
(inches)

30. . . . . . . . . . .
35 . . . . . . . . . . . . .
40. . . . . . . . . . . .
45
5 0 ..............
55 ..............

..............

60..............
--

24
24
24
24
24
24

24

40

43
47
50
53
56
59

23
22
22
22
22
aa

a2

36

38
43
47
50
53
56

1894
1834
18%
18%
1834
1834

33
36
40

:
i
49

................

See page 598.

Poles should be of the best quality of either seasoned or live

green cedar of dimensions given in the above table. Seasoned poles
should have preference over green poles, provided they have not
been held for seasoning long enough to have developed any
objectionable timber defects. They should be reasonably straight,
well proportioned from butt to top, both ends squared, the bark
r l e d and all knots and limbs closely trimmed. They should not
e shaved. A pole should not have dead streaks covering more
than onequarter of its surface. A dead pole or one having dead
streaks covering more than onequarter of its surface should
be rejected. A pole having dead streaks covering less than onequarter of it surface should have a circumference greater than otherwise required. This increase in the circumference should be
sufficient to afford a cross-sectional area of sound wood equivalent to that of a pole of the same class.
Fire Killed or River Poles. No dark red or copper colored
gole which, wben s c n ed, does not show good live umber should
e accepted. No p o t should have more than one complete
twist for every 20 ft. in length nor should it have a large season
check. There should be no "cat faces" unless they are small and
perfectly sound and the poles have an increased diameter at the
"cat faces," nor have "cat faces" near the 6 ft. mark or within 10
ft. of the top. A pole should contain no sap rot, evidence of
internal rot as disclosed by a careful examination. There should be

601

WOOD POLE SPECIFICATIONS

no black knots, hollow knots, woodpeckers' holes, or plugged holes,

and no pole should show evidence of having been eaten by ants,
worms or grubs, except that poles containing worm or grub marks
below the 6 ft. mark may be accepted. A pole should have a
sound top with no pencil holes. A pole havlng a double top or
double heart should be free from rot where the two parts or hearts
join. A pole should be free from ring rot (rot in the form of a complete or partial ring). A le should not have a short crook or bend,
a crook or bend in two pf&s or a reverse curve. The amount of
sweep, measured between the 6 ft. mark and the top of the pole,
that may be present in poles, is shown in the following table
(A.E.R.E.A. Standard) :
Le th
35%.
40 it.
45 ft.
SO ft.
55 ft.
60 it.

x x in.
12

14.

Poles may have hollow hearts under the conditions shown in the
following table which is Standard of the A.E.R.E.A.:
--

--

Average diameter of hollow heart

0 f

Add to butt requirements of

z-in .........................
3-9.........................
4-p~
........................

Nothing

Gin.........................

3-?n.
4-111.
Reject
Reject
Reject
Reject
Reject
Reject
Reject

5-m.........................

7-in........................
en..

.......................

gp........................

10-!n. ........................
IX-m.........................
ra-in.........................
13-in. . . . . . . . . . . . . . . . . . . . . . . .

a-!n.

35. 0

Yd

45 ft. poles
Nothing
Nothing
Nothing
I-!n.
z-!n.
4-~n.
6-in.
Reject
Reject
Relect
Reject
Reject

50. 55.60

and

65 it. poles
Nothing
Nothing
Nothiii
Nothing
I-!n.
2-111.
3-jn.
4-!n.
5-!n.
~!n.
9-111.
Reject

Scattered rot, unless it is near the outside of the pole, may be

estimated as being the same as heart rot of equal area. Poles with
cup shakes (checks in the form of rings) which also have heart or
star checks may be considered as equal to poles having hollow
hearts of the average diameter of the cup shakes.

602

ELECTRIC RAILWAY HANDBOOK

Western White Cedar Poles

-.
("Red Cedar," "Western Cedar," or "Idaho Cedar")
MINIMUMDIMENSIONS
OF WESTERNWHITE
CEDARPOLESIN
INCHES-(CIRCUMFERENCE).(A.E.R.E.A. STANDARD)
Class B '

Class A
Length of poles (feet)

Top28

I
. . . .. . . . . . . . . . . . . . . . . .
...........................

30..
. .
35 ...........................
40
45 ...........................
50 ...........................
.............................

60 ...........................
65 ...........................

Butt

;!

Topaa

Top25
Butt

45
47
49

40
43
44

54

48

52

Class C

Butt

ii
40
41
43

See page 598.

NOTE: '.TOP" measurement being the circumference at the top of the
pole, and the " butt" measurement, the circumference 6 it. from the butt.

Poles should be of the best quality of either seasoned or live

green cedar. Seasoned poles should have preference over green
poles, provided they have not been held for seasoning long enough
to have developed any objectionable timber defects. Poles should
be reasonably straight, well proportioned from butt to top, both
ends squared, sound tops, the bark peeled, and all knots and
limbs closely trimmed. Large knots, if sound and trimmed close,
should not be considered a defect. A pole should not contain
hollow or rotten knots, nor should i t contain sap rot, woodpeckers'
holes, or plugged holes, nor show evidence of having been eaten
by ants, worms or grubs. A pole should not be dead nor should
i t have dead streaks covering more than one-quarter of its surface. A ole having dead streaks covering less than one-quarter
of its surgce should have a circumference greater than otherwise
required. The increase in the circumference should be sufficient
to afford a cross-sectional area of sound wood equivalent to that
of a sound pole of the same class. A pole should not have more
than one complete twist for every 2 0 it. in length nor should i t
have a large season check or crack. No pole should have a short
crook or bend, a crook or bend in two planes, or a reverse crook
or bend. The amount of sweep measured between the 6 it. mark
and the top of the pole should not exceed I in. to every 6 ft. in
length. A pole should have no "cat faces," unless they are small
and perfectly sound and the pole has a n increased diameter
a t the "cat face," nor have "cat faces" near the 6 ft. mark, or
within lo it. of its top. A pole having cup shakes (checks In
the form of rings) should contain no heart or star shakes which
enclose more than 10 per cent of the area of the butt. A pole
should not have butt rot covering in excess of 10 per cent of
the total area of the butt. The butt rot, if present, must be
located close to the center.

CONCRETE POLES
Reinforced Concrete Poles
Figs. 33 and 34 and the following formulas are from the Manual
of the Amer. Elec. Ry. Assn. The square section is more efficient
and economical than either the hexagonal or the octagonal. The
two latter were developed for situations in which the square pole
would not present a sufficiently artistic appearance.

PIG.33.-Proposed sections for concrete poles.

SQUARE
30~18

Deflection x = Ed'

96' (Approx.)
Area = I0
9P1
Pa
lo

Area = b'

Safeload, PI

(APP~OX.)

(Appro..)

OCTAGONAL

Area =

Safe load. Pa

47PL'
E.64

(Approx.)

81)'

-- .

I0
14ood'b (Appr0x.j
.l

I n which
r = deflection a t a point 18 in. below top of pole,

inches.

(See note below)

P = horizontal pull a t a point 18 in. below top of

I

from point of application of P to

the ground line (6 ft. from base end), inches
L = length of le from point of application of P to
the erounfine ( 6 ft. from base end).
. , feet
= rnod\lus of elashcity of steel used
= side of square pole or distance between parallel
faces of hexagonal or octagonal pole a t ground
line, inches
d = diameter of steel rods, inches
Four per cent of steel used in all sections.
NOTE: The above reference states that coefficients 30,40 and 47
of the formulas for deflection, respectively, should be increased 2 5
per cent for practical application.
= Ci:tTT$Ie

604

ELECTRIC RAILWAY HANDBOOK

The values in the table (pp. 605 to 607, from the "Miscellaneous
P r a c t i c e " of the
American E l e c t r i c
Railway Engineering
Association are calcuPIOOC r
: x"*d.n~.t.bb
hted on the pmposed
Brecl Collar lor Span
design of a remforcedconcrete pole of square
cross-section, shown
in Fig. 34.
The general results
of tests and e
of the Amer.
Eng. Assn. Committee
with concrete poles
indicate certain facts
as follows: failure of a
pole is always due to
L-stretching of the reinforcing rods on the
tension side; a failure
is always preceded by
the a p p e a r a n c e of
hair-line cracks in the
concrete on the tension side, a t rather freI
quent and regular int e r v a l s f r o m the
ground line up; it is
advantageous to use a
high grade of reinforcing steel to secure the
m a x i m u m tensile
strength; plain round
reinforcing rods are
essentially as satisfactory a s twisted o r
otlier rough rods bccause in general the
rods will elongate before they slip in the
8tr.p. Appmximatel7 mncrete; a 1a r g e r
ererJ20Dinn1eter'
number of small rods
of Reinforcement
ws
is p r e f e r a b l e to a
smaller n u m b e r of
large rods as a better
distributed reinforcement may be secured
for a given amount of
PIG. 34.-Proposed
design for sauare concrete steel and a greater
pole.
bonding contact surlace is presented to the concrete; the reinforcement need not be uni-

%?c

CONCRETE POLE TABLES

nwrs9nur :z:::
'91
a!peo[aps
7salea,9
0

."

- 02

0-

z z z g: : :~%::

z"n:","opgpgg

rooom
*n-o-o*m
am

605

~ U Y ) Yn~n e
. n n m oonmom
r-Uompgg ~ m - n u 0 n o d o r nnooo

- - 0 o a z C;ggzG

nuooo
nr-on0
--*-a

.. .. .. .. . . .. .. ... ."o, .. .. .. .. .. .. .. .. .. .. n
* .. .. . .. ..
. . . . . . . .
.
,
,
:
?
j.i:::
, , , t e r r t ; ; ; : : : . - ... ... ... ... ...
.. .. .. . .. .. .. .. .. , .m. . . .. .. .. .. . . . . . . .",..
0,
.....
. . . . . .. .. .

......

o
. . :f : : : : : : :. .:: ? ! .. .. .. .. ..
.. .. .. .. .. .. .. .. .., .. .. .. .. . .. . . . .em . . .- .- .
s - g _.:.::,_:-.,
. :::a:
:::::
.ed. " :. :. :. :. :. :. :. :. :. L, :. :.- :.-. :.-:.: :::;;
:-.: : : :
.
.
m
a
.
.
.
.
.
.
.
.
o
u
n
.
. .. .. .. .. .. .. ..
.: .;. ..: ..; ..:
.. .. .. .. .. .. .. :. Cmm
y??
m
'r
*
: :-:-: :
:: :,-ro
..-..-..- ..-..
-.. .. .. .. .. .. .. . o n :.. :..-:...:.
. -.. . .n*n
g
0
. . . . . : : :f? . . . .. .. .. .. ,.?* -n%o .. .. .. .. ..
2 j 0 -I-:._:::
. . ." " : : : : : : : : r - 0 : : : : :
c.

.r

. r n

.r

. O h

::.rr

.*O

&,-

". .. . .. .. .. .. m
oo
0
,. .. .. .. .. : :<?m
. . . . . .. .. o- nmo0
.....
.,P
,

Iiqc
.

.. .. ...-~
.. ..-.. .. O* nO oO "-,". .. ..-..--.
......

: :-::::

......
::::::

. . :. :. :. :.
::::::
. . . . . :~~~~
:i;dd
: : :-.
:::
.-no
.noon

:-:
-

.....

.....

.m?q?
:;roo
.Ov)v)O

a:;:

:;Moo
m o n o

s$azf
<d-dd

.....
:::::
: :-I-:-:
~ . . . . .
:::::
: : :-:..:

.....

....

.....

.....

: L:<-:~.

g . 7 :: : : g gSe"98;: ::::gz JssYg$

...
. . . , a n o o b o n : : :'??
9U g g
:. :
. :- m:$n ;odododrd- n :. .: : :; dn do do do dn .. .. . o o
S$
i ~ ; : ; " o ~ " : : : ; l g g ssz:a : : : 4 ~
OC\O*O

.0%

. O O W

. . .un
U d -.. ..-*a
oon
: :?MY
-

d s
Y

.loo
v)nO
0 om
Cnmu?
< i d d o

ddddd
omononmog
nn-ddddd
0-Oom

ZZt?*

dtiooX

:: :<id

. . o m o m nd od do dn dn : :. o:om
ow
: i;8<;
~ ~ 2 s :. :~;2

:.r*oo

neon*

0-0000

.qIn! alod X:"PobZ $%%hdZ 3-ZmOe"S

27,,8!a~

nm-rnwn

.. Sdi),.gggrxz
0

0
C -

.:.

.qZEu

nk

;r
I

n-

"'Oom

noubn

x--."-.-

7y.rTT
w - r n -

~ * w e ~m
-h o r n
I I . 4

loSnonrn-nn

24.-

""^

66600
onnno

2SFF8
ddooo

:"s

unnoo

4$854
on-0%

3 c - o ~gZSS2
nm
~ nt-o
oddno
--n

22,Z'ZE

-- ---.
- ""
I 1

"^"""

I I I I I

ucln

w n r n r

ou.o.ocm

ao-no
,4 CI

n C(

:*,*
,
""*
*
*

r?u.oob-

%$8;ixxxxx Wx CxhxOxn xN xr)*wlOlrW

x x x x x xxxxx
xxxxx
Q O W N c )
Od-Oh

bVl'330UUYy)r-

"..*

"8 -24.; Zrzrs:=:"z:

",.or-ma

111"

",or-ma0

nnoun

nor-Wh

~ -XnXo X
* XXXXXX;
n o t - m a ~ xxxxx

xxxxx
nocma

11-*-

w - r - -

&33Jq7811q

nmmmm
nnnnn

mwmmm
nnnnn

ooomolm
nnonnn

aaaaa
nnnno

OOOO
0000o

606

ELECTRIC RAILWAY HANDBOOK

zzzzz :
:
:
:
:z;;::z

haaJqIma7
'91
U! PB01 aJeS
1salea13

-.-. .

0 0m
m
01
O0- cna
n,

nmm
N N o

m
OO
t -m
* aa O
n
n
r i mm

r-gct-mo
o a m ca
n, - 0 A. ,
CnNnn-t

zzzzgz zzgzz
..no a,noN-c,* m,~ "- ~obmn bnooo m*onNm0e 02b

.. .. .. .. .. .. .. .. ' oo wm
. .. . .:R. .
: :. :. :.:.: . .: .: :. Cn n?
"
.
.
.
. . . . o . . . . . . .. .. .. n** * .. .. .. .. .. .. .. .. < mn *a
.- 0
0
.. ,.. :. :.?a :. :. :. : .: . .::::...q
: I : : : : :::??
. n r
. . . . . . . . .n- .. . .. m. r- . .. .. .. .. .. .. ..om-*
.. .. . w o . . . . . . . .m o n ... ... ... ........ . .: .:'?goo
.??
8

:R

: I : . . .. . .Oemn
:. :.: .: .: . :.: :.: .? ?
I I

N -

.. . :"
.n. . . . . .: .: :. Tnr 4-o n?b
. . . n n .. .. .. .. .. ... ... ..",or
2 - .: .: : m ?
.
.. .. .. .. .. .: .: :-<.0An ;lmo r nn
m
o
.. ..,nn
mm
. . . . . . .or*i' z :- : .?nS- fn? :. :. :.:.: . .: .:.q??+!
nn-:
:*,z
.
. . . . . .","no
2 00.X
. .,.,
.. .. .. .. .. .. .: ?o ?n?* ?a
0
A
: :&;
::: :: : .nn-0
..

.*r

.n-

C,

* n

gz

.-s &+
u.-

0 ,

.*nmm
.,*to

8
g

:::::

i ; ; d d

:::::

n n r - o o

:: : ::

%,!$B%,"

:;-Omp4

*onlo
& ; i d 6

3;"o:g

:. .: .: .: .:

:.:::

:. .: .: : :
I2
.. .. .. mmun,

--- m; ;adndmdm

:::

:. :. :co?
.mn

.0

moot-

b",-*"

4 6 6 6 6
n o m n n
-0.0 n m

,OlfS%
* n n

$
3 --

:@
%:ffz

d d d o o

.q~n!api
10 1 9 a p ~
u

C -

d d d d d
0.0-mn

B&

;e"

n o b n n

.. .. .. OmOa
. . .o n
:- 3
::".2 :
: . n o 0
0010100

.?

* - N

ZXX

m a o m n

z;;:%$
CQ????
o - - 0 0 0
om*oolo
F;s:$a

nw-ooo
mOm000
a b h c n n

q- M
y ?O?O?Of ?O

.(X,*

:. :. :. :. :. :.

::::::

::::

f f f

.00

if;

....nn
... ...

::

"

:C, =3::::

. . . . .: : : : :*,
..-

Onm8.m
,.jr;Jid
*ma
,.-am*
lomnac
t;;;dd

.Om0
. o a m
.mmm
:d.i;
.om0
. O no

nlo0m.o
r;;;dd
0nbd-m

::Po""
i ~ d d d r-am00
mobatoa,*r,
i d d d d

.. ..

d d d d d d

::

2
: z g ; gn *
""""1

ggza0:
.: :::,a%
:Y)??q * O N "

d d d d d d
r O n b O d

nm-tnor-

2 .++.-.+-

D P ~ O *

:???+!

.n-nn
.. .. .. .. .. ..
.: .: .: .: .: .: :.-tr-n8
r;;;d
.. .. .. .. .. .. 2 : z f :

o m o w o n
-0 br-o n

%q:g%
e n 4 6 6 ~
o o o o o ~
o
ZZ$z::

I
I n

c.-

.. .. .. .. .. .. ..: :. o:n"a- . . . . . . . .locab

: : : : : : ::r;;;
.. .. .. ... .. . .. .. moar % : : : : : : : :???
." .. ". .. . . ..,.Q"O
.on. . . . . . .n**n

......
::::::
. :. :. :. :. "oEZ%%
:. :. :. :. :. :.
aabh~qnr-lwq

.-

:<;;d

.n"clr
.*m
n

O
a
o
Yt

.....
.. .. .. .. ..
.....

.0*00n
.",too0
.onlone
:rir;;;d
.*om",*
-,c",oo
-,m",o*
: & ; ; i d

:q.q?q

Z
__

$2
.--

.. .. .. .. ..

::::::

A*-.,.*,-\

I I I I I I
n n - n n n

m a E : : z

"

'???

: n n u

:. n r o o
OmnOP'O
w n - o~ d d

?:;%$;

- - n o

"
a
,
.

$2%
Z d5::d d

66.566
d r m n n

: 3dkZd$d%
d
o l

fG%$s
menot-

22 Z5!%ZZ

-*.,%

I I

CI*"

n b n o h m

I I O I

* I n n -

caonno
n n - -

8:g 2 ~ x 2 2 ; x x x x x xxxxxx x x x x x x xxxxx

k2.z

~ m a o n on - t m a t I

P-

- - o- -- n+m- - ?

au.o

m-3

ZXXXXX
- 0 - 0 0 6

,wajqWa~

loor-ma

XXXXX
met-mca

mcao-no
n * w n
onnm*m
n n n n n n

n+mocm

n o n n n

mar-who
n

- r r ( - -

XXXXXX

XXXXXX

0-nn*m

- - n b . r

,.,1-.6.,-

&%%&;=;:

;;;;;;;;;;;,

111 -

-nmblo-

- --

XXXXX

a~ ~
n m t *
M

ZI;::

REINFORCED
CONCRETE
POLES-DEFLECTIONS
AND LOADS-Concluded

608

ELECTRIC RAILWAY HANDBOOK

form throughout the length of the pole, but may be stepped ofi as the
top of the pole is approached; a pole with uniform reinforcement
will break a t the ground line, while one with ta red reinforcement
will break a t some point above the ground, Epending upon the

llomv*u .r sg.W a r . s r Al l y 1~y . I"T* &.# P b

nmveiu u .
'
y
s 6Was.
11'~ 1 c 1 l%''Tc.p b~ P l u

A-

p I ~ " X %"L"
O.h.

P l c 35 -Typ~cal

pole construction. trolley and one transassion line.

ta r of the reinforcement; a concrete pole has a n element of

s a g y in that a failure of the pole will not in general allow i t to fall

to the ground; it is difficult a t times to pull over a g l e after failure,

ever though i t is inclined a t a large angle from e vertical.

NATIONN, ELECTRIC SAFETY

CODE

609

T
y
w h i b a y Transmission Line Construction. F i i . 35 to 38,
induslve, show typ~calrr,-volt,
and zz,ooo-volt pole h e constructions from the practice of the Connecticut Company.
National Electrical Safety Code. The U S. Bureau of Standards
has prepared and publ~sheda National Electrical Safety Code,
consisting of a set of rules for the safe construction and mainte-

FIG. 36-Typical pole construct~on, trolley and two 11.0oo-vo1t

transmlsskoa Imes.

nance of electrical apparatus and lines. In preparation and revision

of this code the Bureau had the cooperation and assistance of many
State industrial and public service commissions, municipal electrical
inspectors, engineers of operating and manufacturing companies,
committees of engineering societies (including the Amer. Elec.
Ry. Eng. Assn.) and representatives of the fire and casualty insur-

610

ELECTRIC RAILWAY HANDBOOK

ance interests, and of the electrical workers. The rules of the code
are divided into four parts. Part I refers to the installation and
operation of machinery, srvitchbaards, etc, in central stations and
substations. Part 2 contains rules for the installation and main-

FIG. 37.-Typrcal

pole-top constructxon. two z3.0oo-valt transmtss1on ltcies.

tenance of electrical supply and signal lines, both overhead and

underground. Part 3 contains rules for the installation and
operation of electrical apparatus and wiring in factories, ofices,
residences, or other places where electric light and power are

FIG 38 -Typ~cal pole-top constmctlon. one z2.000-volt Ime.

utilized. Part 4 contains rules for safeguarding employees working

about electric machinery or lines. A supplementary section gives
rules for protective grounding of circuits and equipment.

WIRE AND CABLE TERMINOLOGY

61 1

This code, while prepared for the primary purpose of insuring

safety, includes detailed data and descriptions of the best current
practice in construction and maintenance. Part 2, referring to
overhead and underground lines, is of especial value in including
general rules for pole lines, clearances and separation of wires,
cilculations of stresses, loads and strengths in connection with
poles, towers and lines, rules for crossings of overhead lines of various classes and quite complete tables of sags, stresses, tensions and
loadings of copper, iron and steel wires, etc.
Joint Use of Wood Poles. A Committee with representatives
from the Amer. Elec. Ry. Assn, Nat. Elec. Lt. Assn., Amer. Inst.
Elec. E?grs., and conferring with re resentatives of the telephone
rompames, pre red in 1914, ~pecilcationsfor the Joint Use of
Wood Poles, wEch specifications may be found in the Manual of
the Amer. Elec. Ry. Eng. Assn. Relative locations of wires of
various classifications are specified, together with vertical and lateral
spacings, k n g t h of spans, location for transformers and fixtures,
guy arrangement, etc.

Electric Wire and Cable Tenninologg, A.E.R.E.A. Standard.

NOTE.Same as recommended by U.S. Bureau of Standards. c~rcularNo. 37.
I . Wire. A slender rod or filament of drawn metal.
The definit~onrestncts the term to what would ord~nanlybe understood
by the term "sol~dwire." I n the definlt~on,the word "slender" 1s used rn
the sense t h a t the length 1s great tn cornpanson w ~ t hthe diameter. If a wlre
IS covered w ~ t hr?~ulat!vn ~t is pro rly called an ~nsulatedwnre. whrle pnmanly the term wrre &ers to t g metal. nevertheless when (he context
shows that the wire 15 Insulated the term "mre" wll be understood to rnclude the msulat~on.
2. Condlctkn. A wire or combination of wires not insulated from
one another, suitable for carrying a single electric current.

The term "conductor" is not to Include a combination of conductors Insulated from one another, w h ~ c hwould be su~tablefor canyxng several drfferent electnc currents. Rolled conductors [such as bus bars) are. of course.
conductors. but are not conudered under the termloology here glven.

3. Stranded Conductor. A conductor composed of a group of

wires or any combination of groups of wires.
The mres In a stranded conductor are usuaIly t w s t e d or bra~dedtogether.
4. Cable. ( I ) A stranded conductor (single-conductor cable);
or (2) a combination of canductas insulated from one another
(multiple-conductor cable).
The component codducto~sof the second klnd of cablt m a y be e ~ t h e r
sol~dor stranded, and t h ~ sk ~ n dof cable may or may not have a common
tnsulat~ngcovering. The first k ~ n dof cable IS a s~ngleconductor, whlle the
second ktnd 1s a group of several conductors. The term "cable" IS apphed
by some manufacturers to a solid wrre heav~ly~ a s u k t e dand lead covered;
t h ~ susage arlses from the manner of the rnsulat~on,but such,? conductor
1s not sncluded under t h ~ sd e f i n ~ t ~ oofn "cable."
The term cable" 1s a
general one and In pnctlce ~t 1s usually appl~ed2nly to the larger sizes. A
small cable IS called a "stranded wlre" or a "cord. both of wh~chare defined
above. Cables may be bare or ~nsulated,and thelatter may be armored wlth
lend or wlth steel wrres or bands.

5. Strand. One of the wires or groups of wires of any stranded

conductor.
6 . Stranded Wire, A group of small wires, used as a single wire.
A wlrc has been defined as a slender rod or filament of drawn metal If
such a fil~rnrnt1s ~ u b d l v ~ d einto
d ..ever31 smaller filaments or strands and

612

E1,ECTRIC RAILWAY HANDBOOK

used a s a single wlre. rt 1s ca!!ed "stranded wtre."

There ts,no sharp
d l n & n g llne of slze between a stranded wlre" and a cable
If used
a s a wlre, for example. In wnding xnduztance colls or magnets xt IS called a
str?nded,,wlre and not a cable. If ~t 1s substantlally xnsulated. tt 1s called
a cord, defined below.

IS

7 . Cold. A small cable, very flexible and substantially insulated

to withstand wear.
There is no sharp dlndlng llne In respect to slze between a "cord" and
a "cable." and llkemse no sharp dlvldlng llne In r e s p c t to the character
Usually the msulaof tnsulat~orrbetween a "cord" and a "stranded wlre
tlon of a cord contams rubber.

8. Conc~nlricStrand. A strand composed of a central core surryunded by one or more layers of helically laid wires or groups of
w1rs
9. Concenlric Loy Cable. A single conductor cable com sed of a
central core surrounded by one or more layers of helically E d wires.
10. Rope Lay Cable. A single conductor cable composed of
a central core surrounded by one or more layers of helically laid
groups of wires.
Thls lond of cable differs from t h e preced~ngin t b t the maln strands
are themselves stranded.
11. N-Conductor Cable. A combination of N conductors insulated from one another.

I t is not intended t h a t the name a s here m v y be actually used. One

would rnstead speak of a "3-conductor cable." a 12-conductor cable." etc.
In referring to the senera1 case, one may speak of a "rnultrple-conductor
cable" (as In definrtton No. 4 above).
1 2 . N-Condrulor Concentric Cable. A cable composed of an
insulated central conducting core with (N- I) tubular stranded
conductors laid over it concentrically and separated by layers of
insulation.

Usually only a-conductor or 3-conductor. Such conductors are used 12

carrylng alternating currents. The remark on the expression " N-conduct~r
glvea for the precedmg definltlon applles here also.

13. Duplex Cable. Two insulated

twisted together.

single-conductor cables

They may or may not have a common lnsulatlng covenng.

14. Twin Cabk. Two insulated single-conductor cables laid

parallel, having a common covering.
IS. Triplex Cable. Three insulated single-conductor cables
twisted together.
They may o r may not have a common ~nsulatlngcovering.

16. Twisted Pair. Two small insulated conductors twisted together, without a common covering.
The two conductors of a "tw~sted paw" are usually substantlally lnsulated, so that t h e comblnat~onrs a specla1 case of a "cord."

17. Twin Wire. Two small insulated conductors laid parallel,

having a common covering.

613

RUBBER INSULATION SPECIFICATIONS

Rubber Insulated Wire and Cable for Power Distribution

(A.E.R E.A. Standard)
Grades. Two grades of rubber insulation for distribution wires
and cables are designated by the A.E.R.E A as "Grade A Rubber
Insulation " and "Grade B Rubber Insulation." Grade A is a high
grade insulation for use on circuits of working pressure not exceeding 15,ooo volts. Grade B is for use where a cheaper insulation
IS desired and should not be used on circuits of working pressure
in excess of 7 0 0 0 volts.
Conductor Size and Stranding. The combined area of the wires
when laid out straight and measured a t right angles to their axes
should not be less than the specified gage or circular mils. The
stranding should be concentric and in accordance with the following
table:
Size

a.000.000 C m
I.SOO,OOO c.m
I.OOO,OOO c.m
6oo.000
500.000
400.000
oooo

oo
a

c rn

cm
c.m.

A.W.G
A W.G
A.W.G

7 and smal>er
For

I I
aena' use
9

61

6I
37
37

Cables for
other than
aerial use
127
91
61

I9

61
37
37

19

I9

19

7
7

'

~ntermedmteslzes, use strandlug for next larger nze.

Condzcctor Malerial. The conductor should be of annealed copper,

conforming to the standard specifications of the American Society
for Testing Materials. I t should be rovided with a heavy uniform
coating of commercially pure tin, witgout projections. The tinning
may be tested as follows: Samples of the wire should be thoroughly
cleaned with alcohol and immersed in hydrochloric a c ~ dof specific
gravity 1.088 a t 15.5 deg. C. (60 deg F.) for r minute. They
should then be rinsed in pure water and immersed in an aqueous
solution of sodium sul hide of specific gravity 1.142 for 30 seconds
and again washed. ~ R i operation
s
should be repeated three times.
At the end of the fourth ~mmersionin sodium sulphide the wire
should show no sign of blackening. The sodium sulphide solution
should contain an excess of sulphur and should have sufficient
strength to thoroughly blacken a piece of untinned copper wire in
5 seconds.
Separator. The separator may consist of soft cotton yarn (which
may be braided). or of paper or muslin tape. The separator should
allow the insulation sufficient contact with the conductor to prevent
the conductor sliding in the insulation.

ELECTRIC RAILWAY HANDBOOK

614

Insulation Thickness. The thickness of insulation for potentials

up to 7000 volts should be the minimum allowed by the National
Board of Fire Underwriters, as summarized in the following table:
Working voltage

A.W.G.

(B.&S.)

0-600

1300

2500

3500

5000

7000

Thickness of insulat~on.6 4 t h of one inch

8
8
8

12
12
12

16
16
16

la

8
8

12
12

16
16
16

14

xa
10

3
3
3

4
4
4

6
6
6

4
5
5

6
6

6
4

z
1

4
4
4
5

12

16

1a
12

16
16

5
5

6
6

7
7
7

8
8
8

Ia
Ia
IZ

16
16
16

zas.ooo
j00.000
400.000

6
6
6

7
7
7

8
8
8

9
9

12
12

12

16
16
10

500.000
6oo.000
700.000

6
7
7

7
8

8
9

12

00
000
oo00

Cir. mils

8oo.000
900.000
~.ooo,ooa

1,2SO,Ooo
1.5oo.ooo
1.750.000
a.wo,ooo

8
8

10
10

12
I2

16
16
16

10

9
9

10

II
12

10

12

16
16
16

11
II

14

18

10
10

II

14
I4
14

18
I8

10

11

I8

(For higher voltages, the thickness of insulation will depend upon

the conditions of service.)
Insulolion Repairs and Joints. If exigencies of manufacture
require repairs or joints in the insulation, the work should be done
in such a way as to leave the repaired part or the joint and all parts
affected by ~t as strong and durable electrically as the remainder
of the insulation. The repairs or joints should be properly vulcanized in a mold of approximately the same diameter as the remainder
of the insulation.
Insulation rubber compound should be thoroughly and properly
vulcanized. The vulcanized insulation should be homogeneous in
character, tough and elastic and should be concentric about the
conductors and fit tightly thereto. All laps or seams in the insulation should be as strong mechanically and electricauy as the rest of
the insulation.

RUBBER INSULATION SPECIFICATIONS

615

" ~ r o d cA Rubber Insuldion." ( I ) A 30 per cent fine Para or

smoked k
t latex Hevea rubber compound with mineral base
should be furnished. I t should contain only the following ingredients: Rubber, sulphur, inorganic mineral matter, refined solid
paraffin or ceresine.
( 2 ) I t should not contain either red lead or carbon.
(3) The vulcanized compound should conform to the following
requirements, when tested by the procedure of the oint Rubber
Insulation Committee (see Proc. A.I.E.E., Jan, 1914 .
(a) Results to be expressed as percentages by weight of the whole
sample.

.......................

Rubber
Waxy hydrocarbons ...............
Free sulphur. .....................

Maximum Mlnlmum
33 o
4 o

30

o. 7

(b) Results to be taken between the limits given in proportion

to the percentage by weight of rubber found:
Limits allowed for 30 per cent rubber compound:

........

Saponifiable acetone extract..

I
Unsaponlfiable reslns
.......... o
Chloroform extract
............ o
Alcoholic potash extract.. ............ o
Spenfic granty
...................

35
45

0.55

90

55
1

75

60

Limits allowed for 33 per cent rubber compound:

Saponlfiable acetone extract.. . . . . . .
Unsaponlfiable resans
.
Chloroform extract
.........
~lcohollcpotash extract.. ...........
specific granty
.

............

I 50
050

I.w
0 60

. . .

I 67

(4) The acetone solution should not fluoresce.

(s) The acetone extract (60 c.c.) should not be darker than

a light straw color.

(6) Hydrocarbons should be solid, waxy and not darker than a
light brown.
(7) Chloroform extract (60 c.c.) should not be darker than a
straw color.
(8) Failure to meet any requirement of this specification should
be sufficient cause for rejection
(9) Contamination of the compound, such as by the use of
impregnated tapes, should not excuse a contractor from conforming
to this s cification.
"
G
r
X
Z
J
Rubber Insulation." The vulcanized rubber cornpound should show, when analyzed by the procedure hereinafter
specified, not less than 27 per cent of rubber gum.
Five chemical tests should be made of the vulcanized rubber
compound as follows. Acetone extract, alcoholic potash extract,
chloroform extract, ash, total sulphur
The sum total of the results of these five tests should not exceed
73 per cent by weight of the total compound.
The ash test should be supplemented by ,tests to determ!ne the
quantity of substances other than vulcan~zed rubber whlch are
combustible but not soluble in acetone, alcoholic potash, or chloroform, and any such substance, if any, should be counted as ash.

616

ELECTRIC RAILWAY HANDROOK

The tests should be made in accordance with the Underwriters

Laboratories' specification.
Contamination of the compound, such as by the use of impregnated tapes, should not excuse a contractor from conforming to
this specification.
Tensile Strength and Elungatim of Rubber Znsulufion. A sample
which may be of entire, segmental, or approximately rectangular
cross-section, should be cut from the insulated conductor by means
of a shar knife. The sample should be bent in every direction
to magni& and reveal any surface incision or imperfection which
may exist. A portion of the sample without such defects and
having a free length of not less than 2 in. should then be stretched
a t the rate of 1 2 in. per minute until it breaks; another sample
should be stretched to three times its length and immediately
released. Marks should be placed on the insulation before its
removal from the conductor. The compound should conform to
the following limits:
Thickness
Property

I '1 '

Tensile strength. lb. per

sq. in. (each sample).
E l o n a t i o n m *-in.
length atfreak (times original length) .......................
Elo a t ~ o n S.
seconds
after r2ease when stretched
at the rate of 11 in. per min20%
ute to three times itslength.

................

loo0

..........

..........x

..........

Electrical Tests. Each length of conductor insulated with the

compound should be tested for dielectric strength and insulation
resistance before the application of any outer covering other than
tape or braid, after a t least 1 2 consecutive hours' submersion in
water and while still submersed. The conditions and conduct of
these tests should conform to the Standardization Rules of the
Amer. Inst. of Elec. Engn.
Dielectric Strength. An alternating-current voltage of not less
than twenty-five nor more than one hundred cycles and approximately as closely as possible to a sine wave should be applied
between each conductor and the water for a period of 5 minutes.
For a jc-minute test, 80 per cent of the s cified voltage should
be used. Lead sheathed cables after being E s h e d should have the
test voltage applied between their conductors and also between
all of the conductors and the sheath. The voltage to be a plied
should be in accordance with the table on page 617 The contfuctor
must not show any weakening of its insulation or any other injury
under this test, which is to be made before the test for insulation
resistance.
Insdotion Resistance. The insulation resistance should be
measured after the high voltage test and following a I-minute

TESTVOLTAGES, KILOVOLTS
FOR

5-MINUTE TESTS
Use roo per cent of the following voltages for Grade A
Use 80 per cent of the following voltages for Grade B

. one inch
Thickness of insulation. 6 4 t h ~ of
Size of conductors
a

1 3

1 4

1 s

1 6

I I /
8

' 1

I . . . . . . .
a.ooo.ooocir.mils . . . . . . . . . . . . . . . . . . . . . . ..
1 75o.000 cir. mils.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
~ ~ S o o . o o o c i r . m. i.~.s. . . . . . . . . . . . . . . . . . . . . . .' . . . . . . . . . . .
~ : a ~ o , o ocir.
o mils. . . . . . . . . . . . . . . . . . . . . .
...........

~.ooo.ooocir. mils..
750.000 cir. mils..
5oo.ooocir. mps..
a~o.ooocir. mlls.

. . . . . .".".'./. . . . . .
...........
. . . . . . . . . . .1 ' " ' . . . . . . . . . . . . . . . .
. . . . . a . 5 5.0
........ ...I::::
4.0

6.5

5.5
6.5
7.5
9.0

. . . ..... 4.5
A.W.G..
. . . . . . . . . 5.0
OOOA.W.G
I""'"'
. . . . . . . . 5.0
ooA.W.G
0A.W.G. . . . . . . . . . . . . . . . . . . . . . . . 5.5

7.0
7.5
7.1
8.0

9.0
9.5
9.5
10.0

0000

.......
.......
........
.........

-1'.
,:::::I

........

1A.W.G ...................
2A.W.G ...................
A.W.G ...................
A.W.G ...................

10 I 1 a 1 1 4 1 1 6 l 1 8 z o I 2 2 I 2 4 1 ~ 6 / ~ 8

5.0
6.5
7.0
7.5

10.5
11.5
12.0
xa.5

19
19
rg
19

a2
a2
az
a2

Z44
24

26
26
26
26

8.0
9.0
10.0

13.0
14.0
14.5
15.5

19
19
19
19

a2
aa
az
a2

24
a4
a4
24

a6
a6
26
26

11.5
11.5
11.5

15.5
15.5
15.5
15.5

18
18
18
18

a1

a3
23
a3
23

25
25
25
a5

a7
27
a7

15.5
1s.o
14.5
14.0

18
17
17
16

21

a3
a2
a2
ao

25

27

11.0

12.0

4.0
4.0
4.5
5.0

6.0
6.0
6.5
6.5

8.0
8.0
8.5
8.5

10.0
10.0
10.0

11.5
11.5

..............
.............. 32.5. 0

5.0
5.0
5.0

7.0
6.5
6.0

8.0
8.0
7.5

9.5
8.5
8.5

11.0
10.0
9.5

11.5

14A.W.D............... 2.1
16A.W.G
1.0
2.5
x8A.W.G
1 . 0 2.5

5.0
4.5
4.5

6.0
5 5
5.5

7.0
6.5
6.5

8.0
7.5
7.5

9.0
8.5
8.5

n.0
105
10.5

8A.W.G..
1oA.W.G
1aA.W.G

............ 3 . 0

.........
.........

10.0

12.0
12.0

13.0
12.0

21

ax
21
19
19
18

24

a2

28
28

31
31
31
31

33
33
33
33

a8

30
30
30
30

31
31
31
31

33
33
33
33

34
34
34
34

27

28
a8
28
28

30
30
30
30

31
31
31
31

32
3a
32
32

28
26
26
25

30
a8
a8
a6

31
29
29
a8

3a
30
30
29

28

28
28
28
28

23

:: ::

34
34
34
34

....................................
. . . . . . . . . . . .I . . . . . . 4................

. . . . . . . . . . . . ' . . . . I .....................
. . . . . . / .1
...........
...................
. . . . . . . . . . . . . . . . . . . .1:::: :... ........

.....................

For intermediate sizes use the vdtages corresponding to the next larger slze, hnving the same thickness of insulat~on.)
For a 30-minute tast use 80 per cent of value for 5-minute teat.)

30
30
30
30

620

ELECTRIC RAILWAY HANDBOOK

ougbly saturated with a dense waterproof compound having the

properties hereinafter specified. The impregnated braids should be
uniformly covered with a continuous layer of suitable weatherproof
compound which should adhere firmly to the braids. The outer
surface should be thoroughly slicked down.
Tape. The tape should be of cotton thoroughly saturated with a
waterproof rubber compound and of proper weight and width.
wedherproof Compound. The weatherproof compound should
be such as to have no injurious action upon the insulation of the
braid and it should be insoluble in water; it should be sufficiently
elastic between the temperature of 4.5 deg. C. (40 deg. F.) and 32
deg. C. (go deg. F.) so that the completed wire can be bent to a
radius of ten times its outside diameter without showing cracks in
the finished surface of the wire or squeezing out of the weatherproof
compound. The weatherproof compound should be of such nature
that it will not crack when the finished wire is subjected to a temperature of - 23.3 deg. C. (- ro deg. F.), and that it will meet the
requirements when the finished wire is subjected to the following
melting test.
Melting Test for Wealherproof Compound. Short samples of the
finished conductor should be placed on a piece of clean white paper
in an oven and should be subjected to a temperature of 52 deg. C.
(125 deg. F.) for
hour. The com und should not become
sufficiently fluid to form a ridge upon t e paper r p t i b l e to the
fingers, or, in case the compound should be absor ed by the paper,
to show a greasy or oily spot upon the paper, nor should the compound show a tendency to flow toward the bottom of the conductor, thus exposing the cotton fiber of the braid a t the to .
A p p l i c d i a of Braid and Taje. If the conductor is to be &shed
witb a braid, all conductors smaller than No. 8 A.W.G. should be
double braided, No. 8 and larger to be double or single braided as
required. The outside braids should be not less than %, in. thick
and the inside braid not less than %id in. thick. Single braid
should be not less than H, in. thick. When multiple conductor
cables are specified, the rubber insulation on each conductor should
be covered with a layer of tape overla ping not less than one-quarter
of its width. In the case of two conluctor cables the separate conductors thus finished should be laid side by side and an outer covering of single braid not less than %. in. thick should be then
applied. In the case of three or more conductors the taped conductors should be twisted together with a suitable lay of a t least one
complete twist in 24 in., the interstices being filled with sufficient
jute to make the core of circular cross-section. Over this Gore
should then be applied a tape with an ovorlap of a t least % in. and
the outer covering of braid not less than
in. should then be
applied.
Lead Cable Sheath
Composilion and Thickness oj Lead Cable Sheath. If the wire
or cable is to be furnished with a lead sheath instead of an outer
covering of braid, as provided in the preceding paragraph, there
should be tightly formed about the core a lead sheath of uniform
thickness not less than that indicated in the following table:

r'

s,

CABLE SHEATH AND ARMOR

Thickness d
sheath
% 4 ill.
%r ~ n .
jn.
1 . 2 5 ~ 1 . 9 9.9. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A4 !n.
z.ooo--2.699. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
964 !n.
a . 700 and over.. . . . . . . . . . . . . . . . . . . . . . . . . . % 4 In.

Diameter of
core. ln.

................................
0.300-0.699.. ...............................
0.700-1.z49. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
0-0.299..

A sheath having an internal diameter of less than 2 in. should

consist of commercially pure lead; a sheath having an internal
diameter of 2 in. or more should consist of an alloy containing not
less than 98 per cent of commercially pure lead and not less than
I per cent of commercially pure tin.

Cable Annor
Proledion Under Armor. Rubber insulated cable covered with
tape, braid, or other suitable protection, or the lead sheath should
be run through a hot asphalt compound, served with a layer of jute
yarn, run through hot asphalt again, and then laid with galvanized
wire armor.
Size of Armor Wire. The proper size of armor wire will depend
upon the conditions of service; the latitude allowed in the following
table represents the difference arising from such difference in service
conditions. The armor wire should be the minimum size.
Diameter of cable under
jute bedding. inches

Armor wire, U. S.
(steel) W.G.

Jute bedding under armor measured IU finished cable

$40 in. minimum

340 in. minimum
He in. minimum
He in. minimum
Me in. minimum
540 in. minimlam

Lay of Armor. The armor should be applied closely without

appreciable space between adjacent wires. The lay should be from
e ~ g h to
t twelve times the pitch diameter. Successive layers of jute
or jute and armor should be laid in opposite directions. The
direction of lay is the lateral direction in which the wires run over
the top of the cable as they recede from an observer looking along
the axls of the cable.
Finish of Armored Cabk. If an outer covering is required, the
armored cable should be run through hot asphalt compound, served
with a layer of the best three-ply 14-lb. hard twisted jute yam in
a close short lay, run through hot asphalt compound, then served
with a second layer of three-ply 14-lb. jute yam, run through hot
asphalt compound and finally run through some material to prevent sticking.
Material of Armor Wire. All armor wire should consist of galvanized mild steel wire of uniform diameter, free from all cracks,
splits, or other flaws; i t should be of such softness that it will easily
take a permanent set and give great pliability to the cable, permitting it to be coiled evenly and smoothly and showing no tendency
to rise or spring in the flakes.

ELECTRIC RAILWAY HANDBOOK

I :I

.... .... .

0 0 0
n o 0

0
o
m
n

0 0 0
no0

0
o
h
n

0 0
0 0
nm
n n

0 0 0 0
o n o n
ant-o
m o o t

0 0 0

0
n
o
n

0-2

=n

0 0 0 0
n o o n
nhdtnomm

8n 8n 8h

woo
- o n

0 0 0 0
onmo
nn
n
nm
nm
-

0 0 0 0
0 0 0 0
onnm mnno
~n nc n m
t c
nm
o
n -n
nm

0 0 0
n n o

0 0 0 0

0 0 0

.-a:

C SZ,":

0 0 0 0

P8Z8
R80% ,0:0
nwnn
nomm m t P
o d ~ mn u a n
n n n n
nnnm

n o n o
nnt-o

ttt-

om00
n n n n
v)oem

0
o
0
n

nnom

n n n n

n o 0 0

0 0 0 0

$
:%

0 0 0 0

ebb

n n n n

2":

0 0 0
n n n
0-n
n n n

0%89 8889 m
PO8
ao

$282
:*20*
n n
n n n n

o m t

0 0 0 0
,,am

1R8

",omY)

onnu)

n o 0 0

0 0 0 0

0 0 0 0

0 0 0

omom

n o n n

too00

0 0 0 0

0 0 0 0

onum
O N O C
m m t t

no00
~ o n

,"RZfi S$C;S O,: :";,

P-

n n
t-n
m t

Y)

. .g9
Y)

mm

n n n n

nmm

0 0 0 0
mnon
-x-~on
nnmn

000
nu0
nmn
nnm

n o 0 0
r-non
t n 0 0

0
o
n
n

0 0 0
mon
t m o
wnn

9%:
0 0 0

0 0 0 0
n n o n
091-o O n o m
n n n n

Z88
n t m

--

22::
:2s:
2%:
n n n n
n n n n
mmm

0 0 0 0
no n m w
~ G O Cm m d n
n

0 0 0 0
onom
t t n n

---

0 0 0 0
o n n o
n

I - C ~ O

0 0 0 0
own-

0 0 0
0"GOC

nom

n n n

0 0 0 0
",",no

0 0 0
VInw
m ~ C ; 1 zW
MP
" O N
0 0 0
nmm
mnm
n n

0 0 0
wow
vlmO
- n n
0 0
0

zz

YI

d m m d

-d-d

rn
0
0

::

cii

dddd

-d&d& &&dd &dd& ddd

#ddd

g555

2444

4444

4444

<d4

0000
0 0 0 0

0 0 0 0
0 0 0 0

0 0 0 "

" 0 0 0
0 0 0 0
0 0 0

O W N *

amon
n n

2-22

0000
omolo

?typ:

0 0 0 0
o n o n
qfcnc)

0
n
n

C,C(.In

C(

2 5 - 5
" 0 0 0

9999

9999

9O

PZ%G Z31

0 0 0 0
onmn
wan
n n n n

n o o o

"
,

0 0 0 0
omon

no00
t-nno
n o t G
n n n n

n
.n

--

0 0 0 0
om00

0 0 0
lono
GO*
t n n

Z:1%

Y)

1 1

0 0 0 0
o n 0 0
o o a o
o
nr n o +

---

N+r.

-. . .

RUBBER INSULATION SPECIFICATIONS

619

electrification with a continuous e m.f. of not less than roo nor more
than 500 volts, and the results corrected to the standard temperature
a t 15 5 deg C. (60 deg. F). All tests for insulahonreslstance
should be made a t a temperature withln xo deg C. (18deg F ) of
t h ~ sstandard tem erature. The insulation resistance of each conductor of multipre conductor cables should be the insulation
resistance measured between each conductor and all the other conductors connected to the sheath or water. The insulation resistance should be not less than given in the table on page 618
The insulation resistance a t any given temperature should be
reduced to that a t 15 5 deg C. (60 deg F ) by mult~plyingby the
coefficientin the following tables corresponding to that temperature
for Grade A and Grade B insulation:
Temperature.
deg C.

coefficient

Temperature.
deg. C.

Coeffictent

7
8
9

0
0
0

67
70
74

16
I7
18

I
I
I

I0

I 18

02
07
12

77

II

o 81

I9
20

Ia

0 85

2I

13
14
15
IS 5

o 89
0 93
o 98
I 00

22
23
24
25

Temperature.

Coeficlent

Temperature.

46
47
48
49

0 69
0 71
0 73
0 75

60
6I
61
63

SO

I II

51
52
53

54
55
56
57

58
59
60

I
1

23
30

35
42
49
55

I
1
1

Coefficient

I 00
1

03
05
08

77
79
81
0 83

64

2
67

I I4
I I7
I 20

85
o 88
0 91
0 92

68
69
70
71

I
I

0 95
0 97

72
73
74
75

I
1

I 00

I
1

23
26
30
33

37
40
I 4 4
1 48

Weatherproof Braid
Weatherproof braid should consist of the required number
of fine, smooth, closely woven braids, all of which must be thor-

622

ELECTRIC RAILWAY HANDBOOK

Tenrile S1rengih of Armor Wire. Armor wire should have a

tensiIe strength of not less than 60,000 lb. per square inch of crosssection and an elongation of not less than ro per cent in 10 in.
Flexihlzty of A r m o r Wire. The armor wire should admit of bending around a spindle of ten times the diameter of the wire and back
agaln without developing cracks of the galvanizing wbich are visible
to the naked eye.
Steel T a p e Armor. If the cable is to be armored with steel tape
the core covered with tape braid or other suitable protection should
be Nn through a bath of hot asphalt compound, served with a
layer of 14-lb. jute yarn, spun on w t h a close short lay, run through
hot asphalt compound, armored with a steel tape; armored with
second steel ta e; run through hot asphalt compound, servedwith
a layer of I-&.
jute yam with a close short lay, run through hot
asphalt compound and finished by running through some material
to prevent sticking. The layers of jute should be applied in the
reverse directions. The space between adjacent turns of steel
tape should not exceed one-tenth the width of the steel tape.
Thickness o j Armor and Jzcie under A r m w . The tape and jute
under armor, after armoring, should conform to the following table:
Cable drameter
before armonng,
rnches

Maxrmum w ~ d t h Mlnlmum thrckness each tape.

steel tape,
inches
rnches

0 45
o 4 6 0 75
0 7 6 1 00
r 01-1 40
1 41-1 70
I 71-2 00
a or-over

0 so
0 75
I 00
I

2s

I 50
I 75
2 00

Mrnrmum thrckness under

armor. ruches

0 02

0
o
0
o
0
0
0

02

0 03
0 03
0 04
0 04
0.05

06
06
07
07
08
08
09

Single Conductor, Paper Insulated, Lead Covered Cable for

Volts. The cable is composed of a single conductor, covered
with paper tape to the required thickness. The paper, soaked in
a suitable insulating compound, is covered with a lead sheath of
the required thickness. The completed cable and the material
of which it is made should conform to the requirements of the
following.
Condzrctor. The conductor should consist of annealed copper
wires, free from splints, flaws, or other defects, concentrically
stranded together in reverse spiral layers, and having an aggregate
cross-sectional area when measured a t right angles to the axes of
the individual wires a t least equal to the area of the specified size,
and in accordance with the following table:
1200

1 No.

Size of conductor

of wrres

~p

o 162 1n.-0 204 In drameter

o 229 in.+. 2j8 ln d~ameter
0.289 r n . 4 460 1n d~ameter
300,000-500.000 cu. mlls
6oo.oo~1.OOO.o~clr. mrls
~ . ~ o o . o o ( r ~ , ~ wcn.
, o omils
o
I 6oo.000-and
larger cir. mils

7
19

37
6I
91

127

I zoo-VOLT CABLE

623

Intermediate sizes take the stranding of the next larger listed size.
Each of the indimdual wires should have a resistivity of not more
than 888.55 ohms (mile, pound), a t 20 deg C.
1-ulation.
The highest grade of pure manila rope paper ta e
should be applied helically and evenly to the conductor until tge
required thickness of paper is obtalsed. All moisture should be
expelled from the paper by baking in suitable ovens; it should then
be thoroughly saturated with the insulating compound of the
required kind and quality. The paper tape must be applied to the
conductor in such a manner, and the insulating compound must be
of such a nature, that the cable will be capable of being bent to a
radius of 12 in. when wound on reels and taken therefrom and
put into place, a t any temperature between o and loo deg. F.,
without damage to the insulation or sheath. The insulating compound should contain no substance which will subject the paper or
conductor to deterioration, and i t should be of such composition as
to maintain the insulation in a soft, plastic state a t all seasons of
the year.
Sheath. A sheath of uniform thickness, not less than % in. for
cable under 1% in. core diameter, and not less than 964 in. for cable
with core diameter 1% in. and larger, should be tightly formed about
the core. A sheath having an internal diameter less than 1% in.
should consist of commercially pure lead. A sheath having an
internal diameter of 1% in. or more should consist of an alloy containing not less than 98 per cent of commercially pure lead and not
less than I per cent of commercially pure tin.
Electrical T a t . Each length of cable should be given the following dielectric strength and insulation resistance tests when mechanical operations of its construction are completed.
Lligh Voltage Test. An alternating current of 5000 volts should
be applied between the conductor and the lead sheath for a
period of 5 minutes. The cable must not show any weakening of
~ t insulation
s
or any other injury under this test, which is to be made
before the test for insulation resistance. The frequency of the test
voltage should not exceed loo cycles per second and should
approximate as closely as possible to a slne wave. The initially
applied voltage should not be greater than the working voltage, and
the rate of increase should not be over ~ o per
o cent in 10seconds.
Znsuldion Resistance. Immed~atelyafter the dielectric test, a
test should be made ior insulation resistance. The measurement
should be taken after I minute electrification, using an e m.f. of
not less than roo volts, and should be not less than 50 megohms per
mile a t 60 deg. F
High Voltage, Three Conductor, Paper Insulated, Lead Covered
Cable. The cable should be composed of three comer conductors.
each covered with paper tape to 'the required thidiness, and then
twisted together with a suitable lay. The interstices should be
rounded out with jute and the whole wrap ed with a paper belt to
the required thickness. The paper, soakelin a suitable insulating
compound, should be covered with a lead sheath of the required
thickness. The completed cable, and the materials of which it is
made, should conform to the following.

624

a E C T R I C RAILWAY HANDBOOK

Conductors. Each conductor should consist of annealed copper

w e , free from sphnts, flaws, or other defects, concentncally
stranded together in reverse spiral layers, and having an aggregate
cross sectional area when measured a t right angles to the axes of the
individual wires a t least equal to the area of the specified size and
in accordance with the following table
Stze of conductor

o 1 6 2 ~ n - o204 rn drameter
o z a g In -0 258 rn drameter.
o 2 8 9 r n . 4 460 nn dtameter

I .i

Each of the individual wlres should have a resistivity of not more

than 888 5 s ohms (m~le,pound), a t 2 0 deg. C.
Insulaltmr. The hghest grade of ure manila rope paper tape
should be applied hehcally and eve& to each of the conductors
until the required thickness of paper is obtained. The conductors
should then be tmsted together, m t h a suitable lay of a t least one
complete twist in each 24 in of length of cable. Theinterstices
of the core so formed should be filled in with jute laterals, of the
proper hmenslons to form a true cylinder. The paper tape should
then be applied helically and evenly over the core until the required
~ c k n e s is
s obtalned
All molsture should be expelled from the paper and jute by baking
in sultable ovens, they should then be thoroughly saturated with
the insulating compound. The paper tape should be applied to
the conductors and the core in such a manner, and the insulating
compound should be of such a nature, that the cable wll be capable
of being bent to a rahus of 18 in when wound on reels and taken
therefrom and put into place, a t any temperature between o and
xm deg. F., without damage to the insulation or sheath. The
insulating compound should contain no substance whch wdl subject the paper, jute or conductors to detenoration, and it should
be of such composition as to maintain the insulation in a soft,
plastic state a t all seasons of the year.
Shereath. A sheath of uniform thickness, not less than that indicated in the following table, should be tightly formed about the core:
Diameter of wrc
Th~cknessd
m mrls
Under a o o o

.........................
...
. .. .. .. . .

awa699.
1700 and over

sheath
$5 1%

1%
% r ~ n

964

A sheath having an internal diameter less than 2 in should consist of commercially pure lead. A sheath having an internal diameter of 2 in. or more should cornst of an alloy contaming not less
than 98 per cent of commercially pure lead, and not less than I
per cent of cornmerclally pure tin.
Electrical Test. Each length of cable should be given the following dielectric strength and insulation resistance tests when mechanical operations of its construction are completed The conditions
and conduct of the tests should conform to the recommendations
of the Amer Inst of Elec Engrs

626

DUCT CONDUIT SYSTEM

Didectric Strength. An alternating current voltage should be

applied between each conductor and all the others connected to the
lead sheath, for a penod of a t least 5 minutes. The voltage to
be applied should be that corresponding to the combined thickness
of the conductor and belt insulahon in accordance with the follow~ng
table
Voltage
5.mo
6.250
7.500
10.000

Thickness of
lnsulatron
964 In.
451 1n.
$fa In.

The cable should not show any weakening of its insulation or

any other injury under this test, whlch is to be made before the
test for insulation resistance
Znsulatzon Reszstance Immediately after the dielectric test, a
test should be made for insulation resistance, each conductor being
measured against all the others connected to the lead sheath The
measurement should be taken after I minute electrification,
using an e m f. of not less than roo volts, and should be not less
than 50 megohms per mile a t 60 deg.

Duct Conduit System

Tile Duct. (Fig. 39.) Each piece of tile duct should be of
thoroughly vitrified and glazed t ~ l e ,whole, sound, straight from
end to end, w ~ t hsmooth Interior surface, free from bhsters, sharp
corners and obstructions The bore should be not less than the
nomnal diameter ordered, and should be symmetrical with, and
the ends should be cut off square with, the longitudinal axis of the
duct. The wall of the duct should be not less than $d in. thick a t
the thinnest place, but the duct should average 96 in. thick. The
inner edges of the ends of each piece of duct should be chamfered
pff so that sha edges will not be encountered when cable is drawn
In. The comp7)eted duct should have the necessary mechanical
strength and toughness to prevent chipp~nga t the ends, and breakage in ordinary handhng The outer surface of all duct should be
scored in such manner as to give cement a hold on the surface of the
duct.
Fiber Duct. Fiber duct is made of a finely divided wood pulp
or fiber, thoroughly impregnated m t h bituminous insulating compound to e v e walls of hard, compact, homogenous matenal The
duct should not be affected by acids, alkalies, or moisture and should
be free from all substances wh~chmight corrode or injure the lead
sheath or rubber compound of a cable. The duct material should

626

ELFCTRIC RAILWAY HANDBOOK

not puncture when an altcrnalng current of not less than 2 5 nor

more than roo cycles per second and approxlrnattng as closely as
poss~bleto a sine wave ~sappl~edbetween the Inner and outer surfaces for a penod of a t least 5 mm Prev~ousto the appl~cat~on
of t h ~ test,
s
the duct to be tested may be immersed In water for not
more than 48 nor less than 24 hr The voltage to be app11ed
To be Pomi#hed fn Accordanca
with Cornpan 'a Btsudard

Bwcifidatioos

'/i

mln.Tblckue#a
Are
,
%'

PIG 39 -T~le duct. square and round

shall be that correspond~ngto the nomnal th~cknessof tbe duct
wall m accordance m t h the follomng table
Nomlnal thtckness
of duct wall

Voltage

627

CONDUIT DUCT SPECIETCATIONS

The circumference, includmg the jomts, should not vary a t any

po nt more than
In from a true clrcle a t a temperature of 65
deg C (150 deg F ) or less The th~cknessof the duct should not
be more than M p In less or f i eIn greater, a t any po~nt,than the
nomnal th~cknessas follows
~ O M I N A L THICKNESS
OF DUCT
WALLFOR EACHTYPEINCHES
Nomnal size of duct
1% ~n

~n

Socket jotnt
X tn

Dnve jo~nt
M an

Screw jomt
4<,, ~n

fi

Each plece of duct should be capable of passlng a mandrel 36 In

long and of cross-sect~on$6 m less than the nonunal d~ameterof the
duct A stra~ghtedge l a ~ dlengthw~seon the concave srde of a
5 foot secbon of duct should not show an offset of more than %
In Jolnts should be constructed as heremafter spec~fied
Sockeljotnls should have a mort~seor female jolnt on one end and
tenon or male jolnt on the other end of each plece of duct The mortlse and tenon should be mach~necut to produce a snug fit 36 m
long, shghtly tapered and free from projecting surfaces whlch would
prevent the jomt from be~ngproperly assembled The thickness
of the duct left after the mort~seand tenon have been turned shouid
not be less than %, m less than one-half the nomnal thckness of
the Pipe
D~mejorntsshould have smooth machme cut tapers on each end
of each plece of duct The taper should be four degrees on the
d~ameter For each jomt there should be furmshed a sleeve, of
the same matenal as s ufied for the duct, machne cut to an Internal taper a t each e n r t h e taper belng the same as that s ec~tied
for the duct The mlmmum th~cknessof the sleeve shoultnot be
less than one half of the nomnal th~cknessof the duct The
tapers on the duct and the sleeve should be so cut that when the
jo~ntIS made u the ends of the duct WIII ne~thertouch nor be sepaated by more 8 a n H Ln
S c r m jotnfs should have machne cut threads on each end of
each plpe For each jotnt there should be furn~sheda sleeve of
the same matenal as specified for the duct, havlng machne cut
thread to g v e an easy fit on the threads of the duct The mlnlmum
th~cknessof the sleeve should be not less than threequarters of
nomnal th~cknessof the duct The threads should be so cut and
the ends of the duct should be so faced that the ends of the duct
Pnll butt m t h a firm water tlght jotnt when the jolnt 1s screwed
up firmly by hand, uslng a jolnt compound furnished by the manufacturer The threads should be four to the Inch Bends should
have left hand threads and sleeves for bends should have one end
w t h left hand thread All other threads should be nght hand
C o n s t ~ c t ~ oofnCondmt. A bed of concrete to the depth of not
less than 3 and preferably 4 In should be put In the bottom of the
trench and the surface brought to grade The duct should be i a ~ d
l&e bnckwork ~n cement mortar In such a manner as to break

628

ELECTRIC RAILWAY HANDBOOK

joints both horizontally and vertically. Care should be taken in

selecting cement for the mortar used in this work, on account of
the possibility of seepage through the joints. A quick-setting
cement is recommended, and the mortar should be as dry as practicable to obtain adhesion. After the top layer of ducts is in place,
all exposed joints should be thoroughly covered by a coating of
cement mortar, applied with a trowel. A 30 in. mandril should be
drawn through each duct as i t is laid; the mandril to be of a diameter
W in. less than the interior diameter of the duct. The mandril
should be left in each duct until the next succeeding duct is laid.
The space between ducts and sides of ditch, which space should in
no case be less than 3 in., and for a depth of 3 or 4 in. above the duct
should be filled with cement concrete. Each layer of concrete
should be thoroughly rammed as i t is put in. The top of the finished conduit should not be less than 2 ft. below the surface of
the street.
Where fiber conduit is used, the ducts should be laid one tier a t a
time, using spacing combs about every 3 feet, and in such a manner
as to break joints both horizontally and vertically. All joints
should be so assembled and so sealed with approved compound
that they will be water tight. No duct should be laid until the concrete base or concrete over the previous tier has set firmly. After
each layer of duct is in place all space between ducts and all space
between ducts and side boards, which latter in no case must be
less than 3 inches, and the space for a distance of 1% inches above
the top of all but the top tier of ducts is filled with concrete and
thoroughly rammed so as to secure a complete concrete encasement
for each separate duct of a minimum thickness a t any point of I inch.
After the concrete has partly set the combs are removed and the
holes filled with concrete. I n no case should a comb or spacer be
left in the concrete.
Manholes. Manholes should be laced a t street intersections
or turns and wherever necessary t o m a k e cable joints. The distance between manholes should not exceed 500 ft. A manhole
should be kept well ventilated, dry and clean. Fig. 40 shows a
typical brick two-way manhole, from the Miscellaneous Methods of
the American Electric Railway Engineering Association. The
top of the manhole casting should be set so as to conform to the
grade line of the street and the space between the bottom of the
casting and the covering slab should be filled with brickwork laid
in cement mortar. All bricks used should be of the best quality,
whole, sound, perfect, and hard burned throughout. Every brick
should be thoroughly wet by immersion in water previous to laying.
In laying, each brick should have a full, close joint of cement
mortar made a t one operation on its bed, ends and sides. All joints
should be thoroughly trowel struck. If the manhole is to be constructed of concrete, its walls, floor and ceiling should be of
thickness not less than shown in Fig. 40.
Specifications for Electrical Conduit Construction. The Manual
of the Amer. Elec. Ry. Eng. Assn. contains a complete recommended form of specifications for electrical conduit construction.
Fig. 41 shows a typical conduit record sheet.

MANHOLE CONSTRUCTION

FIG.40.-Typical

design for 4 ft. X 7 ft. two-way manhole.

Limiting Clearance Lines for Third Rail Structures, A.E.R.E.A.,

A-R-E-A. and A.R.A. Standard. Fig 42 gives the standard locations of the limiting clearance lines for third rail and permanent
way structures and rolling equipment. The space within the lines

AT, B T , C T , DT, ET, FT, and AT, J T , KT, LT, M T should be

reserved for third rail structures on tangent track or curves of radius
greater than 800 ft. (For curves of radius less than 800 ft see notes
in Fig 42 ) Permanent way structures on tangent track or curves
of radius greater than 800 ft. should not be nearer the third rail

T H I R D RATT, LOCATTON

631

structure than lines AS, JS, KS, LS, MS. (For curves of radius
less than 800 ft. see note in Fig. 42.) Rolling equipment under
conditions of maximum wear and deflection on tangent track or
curves of radius greater than 800 ft should not be nearer the thirdrail structure than lines A E , BE, CE, DE, EE, FE, GE. (For
curves of radius less than 800 ft see note in Fig 42.)

Third Rail Gage and Elevation. The definitions of these terms,

as shown on page 547, have also been adopted as standard by the
Am El Ry. Eng. Assn. Standard dimensions, as adopted by the
Amer. Inst of Elec. Engrs , are given on page 548
Composition and Resistance of Third Rail. The following tables
and conclusions regard~ngthe composition of steel and iron conductor rail are from a paper by J A Capp before the American
Institute of Mining Engneers, 1903. The table on page 632 gives

632

ELECTRIC RAILWAY HANDBOOK

RESISTANCE
AND COMPOSITION
OP STEEL AND IRON*
Percentage composition

Mn

v1

22.72 19.0 7.58 13.20 0.33 1.27 0.09 0.05 0.05 1.79 0.19
ao.90 20.0 8.27 12.12 0.17 1.09 0.09 0.05 o . o o 4 1 . 4 0 4 0 . 1 ~
2 1 . ~ 925.0 8.27 12.09 1.40 0.2220.01 o.oz00.082 1.7340.112
19.87 19.0 8.65 11.55 0.20 0.95 0.10 0.08 0.05 1.38 0.23
19.80 19.0 8.68 11.51 0 . 4 3 0.77 0.10 0.04 0.0661.406o.ao6
6 19.80 19.0 8 . 6 9 11.51 0.36 0.80 0.10 0.04 0.0 7 I 3470.187
19.81
20.0 8.69 11.51 0.22 1.08 0.10 0.05 0.0% I : ~ I O O . ~ X O
7
8 19.69 19.0 8.73 11.
0.74 0.58 0 . 0 4 3 o . o ~ 6 o . ~ o1.5990.279
9 18.95 25.0 9.29 1 0 . 3 1.61 o. I47 o.oI5 0.018 0.092 1 . 8 8 ~
0. 125
10 18.17 19.0 9.46 10.56 0.41 0.72 o.0390.041o.11
1.32 0.190
11 17.27 19.0 9.96 10.04 0.36 0.87 0.08 0.09 0.04 1.44 0.21
I2 17.10 19.01o.o6
9.94 0.37 0.73 0.09 0.04 0.06 1.29 0.19
I3 17.10 19.5 10.06 9.94 0.23 0.80 0.046 0.033 0.016 I .o95 0.065
14 16.96 19.0 10.14 9.86 0.30 0.95 0.0630.01 0.01 1.3330.083
15 16.95 19.5 10.14 9.86 0.29 0.99 o.084o.01 0.01 1.3840.104
16 16.32 19. o 10.55 9.48 0.23 0.89 0.053 0.01 0.005 I . 193 0.073
I7 16.25 19.5 10.59 9.44 0.26 0.83 0.0530.01 0.004 1.1570.067
18 16.21 20.0 10.62 9.42 0.28 0.65 0.08jo.06 0.05 1.1230.193
I9 16.0919.010.60
9.36 0.22 0.68 0.0770.07 0.05 1.og70.197
20 16.09 19.0 10.69 9.36 0.16 0.66 O . O ~ ~ O . O ~ O ~ . ~ I ~
21 15.32 19.0 11.24 8.90 0.33 0.49 0.0680.05 0.02 0 . 9 ~ 8 0 . 1 3 8
22 14.57 19.5 11.82 8.46 0.31 0.45 0.10 0.04 o.oa6o.ga60.166
8.42 0.25 0.41 0.10 0.04 0.03 0.83 0.17
23 14.49 20.011.88
tr. 0.774 0.17
24 14.73 23.5 11.88 8.42 0.1440.46 0.09 0.08
8.36 0.1880.48 0.09 0.08
a s 14.6a 23.5 11.96
tr. 0.83 0.17
26 14.15 19.0 1a.17 8.22 0.22 0 4 6 0.024 0.34
tr. 0.838 0.058
27 14.03 19.0 12.26 8.16 0.192 0.57 0.024 0.34
tr. 0.82 0.058
28 13.86 19.012.41 8.06 0.16 0.48 0.091 0.04 0.01 0.781 0.144
a9 13.83 19.5 -2.44 8.04 0.10 0.55 0.08 0.05 o . o ~ 4 0 . 8 0 4 0 . . 1 ~ 4
30 13.80 19.0 12.57 8.02 0.14 0.41 0.11 0.05 0.0090.7190.169
31 13.67 19.0 12.58 7.95 0.23 0.48 0.024 0.01 0.023 0.767 0.057
32 13.64 19.012.61
7.93 0.24 0.57 o.oz90.01 o . o o 3 0 . 8 5 0 0 . o ~ a
33 13.9024.012.63
7.92 0.10 0.35 0.04 0.02 0.05 0.46 0.11
34 13.31 19.012.92
7.74 0.25 0.37 0.04 0.03 o.0180.7080.088
35 13.30 19.5 12.94 7.73 0.23 0.49 o . 0 ~ 4 tr. o.oo40.7480.oa8
36 13.27 19.0 12.97 7.71 0.19 0.37 0.09 0.05 0.01 0.71 0.15
37 1 3 . a ~19.0 12.99 7.70 0.27 0.41 o.oa40.01 0.001 0.7150.035
38 13.18 19.0 13.05 7.66 0.28 0.28 0.027 0.0340.04 0.661 0.111
39 13.1819.013.05
7.66 0.07 0.40 0.08 0.07 o.orjo.6330.163
7.60 0.28 0.42 0.022 0.04 0.008 0.770 0.070
40 13.07 19.0 13.16
41 1a.87 20.0 13.27
7.48 0.16 0.38 0.08 0.04 o . 0 0 9 0 . 6 6 9 0 . 1 ~ ~
42 1a.73 20.0 13.52 7.40 0.15 0.45 0.011 0.033 tr. 0.6440.044
43 12.69 19.013.55
7.38 0.19 0 . a ~0.0250.04 o.oj40.4990.099
44 12.53 19.0 13.74 7.28 0.215o.aa o.os10.113 ..... 0.5990.164
45 11.01 19.0 15.63 6.40 0.05 o. 19 0.054 0.059 0.03 0.383 0.143
7.82 0.15 0.0680.13 0.02 0.15 0 . ~ 1 8 0 . 3 0
46 13.8025.512.78
47 13.82 26.0 13.37
7.48 0.15 0.064 0.036 0.02 0.13 0.400 0.186
48 13.10 26.0 13.50 7.41 0.16 0.0740.12 0.027 0.10 0.481 0.247
nil 0.13 O . W ~ O . O ~ ~ O . ~ ~
49 11.54 25.5 14.07 7.11 0.08
50 11.92 25 5 I 80 6.76 0.17 0.027 o.0740.oza 0.077 0.3700.173
51 10.82 z4:o 1%:21 6.17 0.058 0.10 0.014 tr. 0.012 0.184o.oa6
sz 10.80 25.5 16.34 6.12 0.16 o.018o.049o.011 0.015 o . z ~ a o . 0 7 5
LSS 11.40 17.0 15.00
6.68 0.0500.1800 O I ~ O . O I I 0.02 0.2740.044
B
11.00 17.0 15.57 6.44 0.0300.03610:065 0.0160.14 0.287 0.221
s c ~10.35 17.0 16.55 6 . 0 6 o 028 tr. 10.004 0.005 0.07 o. 107 0.079
' Serial numbers I t o 45 inclusive are steel.
Serial numbers 45 to end of table are iron
I

a
3
4

THIRD R A E CONDUCTIVITY

633

temperature and specific resistance a t that temperature, conductivity and resistance, compared to that of copper, and the composition
for each of several samples of steel and iron tested. The samples
having serial numbers I, 2, 4, 7, 11 and 1 2 were standard T-rails.
Nos. 24 and 25 were cut from T-rail used for conductor on the
Aurora, Elgin and Chicago R. R. Nos. 46,47 and 48 were ordinary
reiined bar-iron; 49 and 50 were special refined bar-iron for staybolts and similar use, 51 and 52 were Swedish and Norway iron
respectively.
The values in the tables showing the variation of resistance with
manganese were selected from the table on page 632 in studying
the influence of manganese on resistance, and indicate that the
effect of manganese in increasing resistance gradually increases
with the percentage of manganese present, within the limits represented by these samples. Messrs. Barrett, Brown and Hadfield
(Trans. Royal Dublin Society, Val. VII, series 2, Part IV) found
the resistance to increase a t k t very rapidly, with constantly
increasing percentage of manganese, then more and more slowly,
until 7 per cent manganese, after which a further increase in the
percentage of manganese produces little or no increase in resistance.
WITH MANGANESE
RESISTANCE
OF STEEL.VABIA~TON

~~~~ I
a
4
7
13
16
19
25

a6
27

3I
35
36
43
44

(Carbon from 0.1 7 to 0.23 per cent)

%:='
Ioo
0.92

Resistance

112
- -. 1 2
~

11.55
11.51

0.89
0.68
0.48
0.56
0.57
0.48
0.49
0.37
0.21
0.22

9.48
9.36
8.36
8.22
8.16
7.95
7.73
7.71
7 38
7.28

C cent
a r
per
0.1,
0.20
0.22
0.23
0.13
0.22
0.188
0.22
0.191
0.23
0.23
0.19
0.19
0.215

1
1

P&SeitSi

--

o.ru
0.23
0.210
0.065
0.073
0.I97
0.17
0.058
0.058
0.057
0.028
0.15
0.099
0.164

RESISTANCE
OF STEEL. VARIATION
WITH MANGANESE
(Carbon from 0.27 to 0.33 per cent)
Sample

Manganese.
per

+
+ si
per cent

Resistance
13.20

40

0-49
0.45
0.41
0.28
0.42

8.90
8.46
7.70
7.66
7.60

0.33
0.31
0.27
0.28
0.28

0.083
0.I04
0.193
0.138
0. 166
0.035
0.111
0.070

634

ELECTRIC RAILWAY HANDBOOK

The values in the tables showing the variation of resistance Htith

carbon were selected from the table on page 632 in studying the
influence of carbon on resistance. W ~ t huniformly increasing
carbon the resistance of unhardened steel a t first rises very rapidly,
but the rate of increase gradually drops until it reaches a straight
line a t about 0.2 per cent C., which continues up to limits of the
carbon listed.
RESISTANCE
OF STEEL. VARIATION
WITH CARBON
(Manganese from 0.15 to 0.30 per cent)
Sample
numkr

1 sz(1

Resistance

M ~ : : ~

1 &'

:e$si

RESISTANCE
OF STEEL. VARIATION
WITH CARBON
(Manganese from 0.4 to 0.5 per cent)
Sample
number

Carbon.
per cent

Rnistanee

Manganese.
per cent

+ S + Si

per cent

The following is also from Mr. Capp's paper: "The elements

phosphorus, sulphur and silicon were not present in su5cient
quantity in any of the samples tested to permit us to draw any
curves showing their influence upon the resistance. . . . In
commercial steels the percentages of all three of these elements is
so small that their effect on resistance may generally be neglected."
A study of the several tables given in the paper shows that manganese preponderates in influencing the resistance of steels,
that for lowest resistivity, this element must be present in very
small quantity, much smaller than is usual in merchant or structural steels. While all the other elements must be present only in
very small percentages, so great is the preponderance of the influence of manganese, that they may be tolerated in quantities which
the steel makers would consider reasonable, without unduly
increasing the resistance. For a satisfactory third rail the lowest
possible resistance (from 6 times to 6.5 times that of copper?) is

635

THIRD RAIL CONDUCTIVITY

not necessary; and the'great cost of making such extremely pure

steel is not warranted. In fact, such extremely ure steels would
probably be so soft that the frictional wear of tRe collecting shoe
would be excessive and the life of the rail unduly short. Assuming,
then, that a rail from steel having a resistance not greater than eight
times that of copper (13.8 microhms a t 2 0 deg. C.) would be desirable for conductor rails, the figures tabulated would seem to indicate
that the following extreme composition would be permissible:
Per cent
Carbon up t o . ...................................
o.a
Manganeseupto ................................
0.4
Phosphorus up t o . . .............................. 0.06
Sulphurupto ...................................
0.06
Silicon u p t o .................................... 0.05

This composition, however, would be extreme, and any overstepping of bounds might result in too great resistance; therefore
for resistance up to eight times that of copper, the specified analysis
should be:
.

Carbon not to exceed.. ...........................

Manganesenottoexceed .........................
Phosphorus not to exceed.. .......................
Sulphur not to exceed.. ..........................
Silicon not to exceed. ............................

Per cent
o. 15
0.30
0.06
0.06
0.05

The following suggested third nil composition is from the

Standard Handbook for Electrical Engineers:

Carbon not to exceed.. ...........................

Manganese not to exceed.. .......................
Sulphurnot toexceed ............................
Phosphorus not to exceed.. .......................

Per cent
o. 12
0.40
0.05
o . 10

The resistance of a third rail of the composition given above' will

be approximately
- 7.75
. . - times that of an equivalent area of commercia1 copper.
Relative to conductor rails installed bv the Underaround Electric
Rys. Co., London., S. B. ~ o r t e n b a u i h(paper ~ I . E . E . , 1908),
says: "The resistance of these rails was about 6.4 times that of an
equivalent area of copper and the chemical composition substantially
as follows:
Carbon .........................................
Manganese......................................
Sulphur. ........................................
Phosphorus.....................................
Silicon.. ........................................

Per cent
0.05
0.19
0.06
0.05
0.03

I t is interesting to note that the cost of this special conductor

rail was no more than the standard track rail."
British Standard Method of Specifying Resistance of Steel
Conductor Rails. The British Engineering Standard Committee
has adopted ( 1 ~ 1 the
~ ) following: In specifying the resistance of a
steel conductor rail the value shall be given in microhms (millionths
of an ohm) and shall be expressed as the resistance in microhms a t
a temperature of 60 deg. F. (15.6 deg. C.) of a rail of the same
matenal as the conductor rail in question, having a length of I

636

ELECTRIC RAILWAY HANDBOOK

yd. and a weight of roo lb. I t follows that if the resistance of a

rail weighing loo Ib. to the yard be R, microhms, that of a rail
weighing W Ib. to the yard will be

microhms.

Conversely,

if the resistance of a rail I yd. in length weighing W Ib. to the yard

be R
. microhms, that of the corresponding roo-lb. rail will be
WRu microhms. I n either case we have the relation roo% = WR,.

I00

For the purpose of reducing observations made a t temperatures

other than 60 deg. F. to the standard temperature, a mean coefficient
of increase of resistance with temperature of 0.26 per cent per
deg. F. (0.47 per cent per deg. C.) shall be employed unless otherwise specified. Where greater accuracy is required for a wide
range of temperature, the coefficient of the actual piece of rail
should be used.
Temperature Coefficient of Conductor Rails. Commenting on
the above, the British Committee notes that the temperature
coefficient of the conductor rails tested a t theNational Physical
Laboratory for the purpose of the report varied from 0.28 per cent
per deg. F. (0.51 per cent per deg. C.) for a rail of 15.9 microhms
per IOC-lb.yard to 0.24 per cent per deg. F. (0.43 per cent per deg.
C.) for a rail of 20.5 microhms per IOC-lb. yard. Taking the
extreme case of a rail of 16 microhms per loo-lb. yard tested a t
40 deg. F. (4.5 deg. C.) or 80 deg. F. (27 deg. C.), the use of the
mean temperature coefficient may introduce an error of k0.6
per cent. For a smaller range of temperature the error would be
pro rtionately less.
Xtional Physical Laboratory Tests. The above-mentioned
rewrt of the British Engineering Standards Committee gives the
following table showing the relafion between the resistance, temperature coefficient and composition of some samples tested a t the
National Physical Laboratory for the purpose of the report.
(Sample I is electrolytic iron; 2, 3, 4 and 5 are conductor rails;
6 is a piece of ordinary track rail.)

Third Rail Sections. Both T-rails and rails of specially rolled

sections are used for third rall. The section should be such as to
offer ample surface for current collection, and it should have a n

THIRD RAIL SUPPORT

637

area of cross-section proportional to the conductivity desired. The

following table gives the weight per yard of the T-rail used for third
rail in several installations:
Rold

Weight of third
rail. lb. per yd.

......................
..........
:::::::
........................

Englewood Elevated Chicago..

Metropolitan West side Elevated. Chicago..
Northwestern Elevated Chic o
Gqand Ra ids. Grand ~ a v e n %~u;k&dn:::.::
~ l c h i g a nfJnited ~mlways,.
interborough Rapid Translt Co.. New York.. ..........
Interborough Ra id Transit Co. (Westchester branch)
~ackawanna& Gyoming valley..
Wilkes-Barre & Hazelton.. ..........................
Boston Elevated..
Aurora, Elgin & Chicago.. ..........................
Interborough Rapid Transit (elevated division). .......
Long1slandR.R..
P w e t Sound Elec. Ry.
Scioto Valley Tract~onCo.
Seattle-Tacoma Interurban..
Pennsylvallla Tunnel & Terminal R. R

.
....................
.................................

.................................
.............................
..........................
........................
................

The section shown in Fig. 43 was used in the New York Central
& Hudson River R. R. electrification. Fig. 44 shows an inverted
channel section devised
by S. G. Redman and
C. H. Merz. London.
This channei is of irregular form, as the
non-contacting flange is
used to secure more
cross-sectional conductivity and to keep the
contact-making flange
in place
Third Rail Support.
The r908 report of the
Committee on Power
Distribution, A.E.R.E.
A. states "The spacing
of third rail supports
varies without respect
to the type of rail. The
spacing most used is
that. of 10. it., with a
maxlmum in some Instances of 11 ft. and a
minimum of from 5 ft.
to 6 ft. These locaI
tions a r e apparently
PIG. 43.-N.
Y. C: type underiUnning
t h r d rail.
governed by the lengths
of third rail and the standard tie spacing in use. The weights
of brackets vary between 9 lb. with the over running rail and
13 lb. to 2 0 lb. for the underrunning rail. The committee can
st% no reason for such great diversity in spacing of supports,

638

ELECTRIC RAILWAY HANDBOOK

and would recommend that a spacing of 10 ft. be used on all

third rail construction where p f t . conductor rail is used, and
11 ft. where 33-ft. conductor rail is used.
In all cases but one
(reported), the support for the bracket is an extended tie, which is
also a part of the track structure. I n the one case referred to

[email protected] channel
th~rdra~l.

FIG. 4s.-Thlrd

rail and support.

Ph~la.& Western.

(subway of the Philadelphia Rapid Transit Co.), it is entirely independent of the track structure. I t might be well in this connection
to note that the road referred to reports an absolute lack of insulator
breakage, undoubtedly accounted for by the above condition of
supports independent of track structure." Fig. 45 shows the method
of supporting the overrunning
T-section third rail on the Philadelphia & Western Ry. Fig. 46
f shows
the method of supporting
the underrunning T-s e c t i o n
third rail of the 1200-volt interurban division of the Central
California Traction Co. Fig.
43 shows the method of supporting the underrunning third rail
on the New York Central &
Hudson River R. R. Fig. 44
shows the method of supporting the Redman-Merz underrunning third rail. The rectangular base of this channel is
supported on a flat insulator,
FIG.46.-Undemnnlng (T)third rail, an Intermediate bracket and a
Cahforn~a.
foundation insulator. The c a p
ping of the conductor channel may be of fiber, stoneware, or other
material keyed into the conductor as indicated.
Third Rail Insulation. The 1go8 repor! of the Committee on
Power Distribution, A.E.R.E.A. states Three kinds of insulation are used with the overrunning type of rail: wood, reconstructed granite and composition. Four kinds of insulation are

i
!

cL

THIRD RAIL INSULATION

639

used in the underrunning type: wood, porcelain, semi-porcelain

and composition."
A third rail insulator should have sufficient mechanical strength
to stand up under conditions of severe mechanical shocks and vibration. I t must also support the rail in such a manner that expansion
and contraction of the rail can take place without tipping or damaging the insulator. The insulator should be made up of a small
number of pieces and there should be no bolts used which can loosen
from vibration, shrinkage of insulation, etc. The insulator must
provide adequate electrical insulation for the voltage on which it
IS used. I t should be so designed that a liberal water drip is provided so that there will not be a continuous leakage path from the
conductor rail to the ground during times of rain and moisture.

FIG.47.-Third

rail insulator. Chicago

Elevated Railroads.

PIG. 48.-Third rall insulator. Lackawanna & Wyomlng

Valley.

The insulating material should be such that it will stand up under

vibration and shocks. The experience of the Chicago Elevated
Railroads indicates that maple wood blocks, impregnated in paraffin for a period of several hours, meets the electrical requirements
for 600 volt service in a very satisfactory manner. The top and
bottom castings of the insulator illustrated in Fig. 47 are one piece
affairs, easily assembled, are ample in thickness so they do not rust
out readily, and the means of holding the rail down in the chair is
simple; the clips can be removed by breaking off and the insulator
re-used by inserting a new clip. Two types of insulators are used
in the installation shown in Fig. 45. The one-piece insulator is
dry-process porcelain and the two-piece insulator is wet-process
porcelain. A cast-iron cup with lag screw holds the insulator
In place on the tie and a cast-iron cap on top of the insulator holds
the third rail in position. A canvas pad between this cap and the
insulator, together with provision for vertical movement, reduces
the vibration which is so destructive to the porcelain. In the

640

ELECTRIC RAILWAY HANDBOOK

installation shown in Fig. 46 a porcelain block is used a t each bracket.

In the installation shown in Fig. 43 the third rail is loosely held
in each bracket by non-charring, moisture-proof insulator blocks.
Fig. 48 shows a two-piece porcelain insulator used on the Lackawanna & Wyoming Valley R. R. The two pieces of porcelain are
cemented together as shown and then mounted on a wooden pin.
Third Rail Protection. Methods of protecting overrunning and
underrunning T- and U-section third rail are shown by Figs. 43 to

XNSULATVlL A N D WOODEN PllOTECIION BHXATLZINO

i
BEOTION B-B

ALTERNATE FIBER PROTECTION

PIG.49.-Protection

SHEATHING

for N. Y. C.-type third nit

45, respectively. The details of two types of protection usecl on the

New York Central type third rail (Fig. 43) are given in Fig. 49.
In this type a sheathing of wood or fiber reaching from bracket
to bracket embraces the rail head, reaches nearly to the lunning
face of the rail and extends outward from the web, thus fotmine
.a
petticoat.
Operation of Third Rail in Snow. In Feb., I@, tests of the
operation of third rail in snow were made on the New York Central
& Hudson River R. R. After medium heavy sngw had fallen,

641

T H I R D RAIL OPERATION

without drifting, to a n average depth of 1 7 in., a flanger was sent

over the track. 'This packed the snow against the third rail. The
test was made by observing the operation during the passage of a n
electric locomotive on several trips. During the first passage
over the overrunning unprotected third rail there was very little
flashing or trouble. Due to the formation of ice and the ironing out
of snow, operation grew worse on succeeding passages until it became alniost impossible. Results with the overrunning protected
third rail were about the same as with the overrunning unprotected
third rail. The first passage scooped the snow away for 25: or 3
in. under the contact surface of the underrunning protected third
rail. Each succeeding passage tended to clean the contact surface
and operation was practically free from trouble.
Operation of Third Rail in Sleet. I n March, I@, tests of
the operation of third rail after a sleet formation were made on the
New York Central & Hudson River R. R. The sleet formation
continued for a day in a wind having a velocity of about 2 miles
a n hour and temperature ranging from 2 8 to 33 deg. F. The
thickness of the sleet averaged about 3 in. The effect of the
sleet on the overrunning unprotected third rail was so bad that
running was impossible after 3 hours. I n places where the wind
caused the sleet to distribute entirely over the running surface of
the rail the contact shoes arced badly and the operation was unsatisfactory. There was no sleet formation on the running surface of
the underrunning rail and there was no sign of arcing.
n.nsb~.

FIG.50.-Third

B-

1-0 M

rail sleet shoe. Michigan United.

Mechanicnl Third Rail Sleet Remover. A type of third rail sleet

shoe designed by W. Sims, and which is giving satisfaction
on the Michigan United Railways, is shown in Fig. 50. It is of the
same design as the regular third rail shoe, but has four steel cutters
set diagonally in its face and cast integral with the body of the shoe.
When these cutting edges become worn or damaged they are chipped
off, and the shoe is kept in regular service until it is worn out. The
sleet-cutting shoe is mounted on a vertical iron shaft which passes
through guides and is attached to the piston of a n air cylinder which
has a 3-in. stroke. A spiral spring around thls shaft holds the shoe
off the rail except when air pressure is put on the cylinder to press
the cutters against the rail. The air supply is taken from the train
line through a %-in. three-way valve and pipes extending under the
car to a point convenient for connecting with the cylinder which

142

ELECTRIC R A n W A Y f1ANDB00K

operates the shoe. This connection between the pipes and

the cyllnder is made with ~ $ 4 ft lengths of 96-ln atr hose,
secured a t each end with hose
I
clamps. An ordinary straight
valve is placed in each supply
pipe to enable the motormm
to use one or both shoes as OCcasion demands.
The thlrd rail sleet-remaving device shown in Fig. 51,
(leveloped by the Metropolitan
West Slde Elevated R. R. CD.,
was found by that company
to give satisfactory service ~ n der all operating and weather
conditions. The device consists of a piece of oak 2%
in. X 544 in. X 4 ft. 1% in.,
a t one end of which crusher
rolls are placed and a t the
other end are scrapers. The
crushingrolls, 3% X 3% in. in
size, are right and left spiral
toothed, made of cast crucible steel, hardened, but not
machined. These rollers are
carried on I X 6-in. steel pips,
which are held in place in the
steel slide-rod casting by cotter
pins a t each end. At each end
of the oak board is a cast iron
cam and wooden handle, which
is used in raising o r lowering
the rollers or scrapers from or
to the working position. The
scraper blades are 3% X 3%
X 296 in., and are made of took
steel. There are four of these
b l a d e s bolted t o cast steel
pivots so that they may take
the scraping position for either
a backward or forward movement of the car. The scrapers
are supported on a slide rod
provided with the elevating
cam similar to the crushers.
Immediately inside the guide
plates which s u p p o r t the
crusher and scraper slide rods are two cast iron corrugated plates
which are used in adjusting th e slleet remover for different truck

SLOT CONDUIT SYSTEM

643

heights and wheel wear. Bolted to the top of the oak insulating
timber is a n adjusting leaf spring, the ends of which are fitted into
slots provided a t the top of each slide rod. At first this spring provided roo Ib. pressure a t both the crusher and scraper, later the
crusher end pressure was increased to 2 0 0 Ib. by adding another leaf
to the crusher end of the spring. This device may be either bolted
to the third rail shoe beam or held in position by means of a casting
which is fitted to the car spring seats. A )i-in. flashboard made of
ash and running the full length of the sleet remover is inserted
between adjusting plates and the oak timber t o provide additional
insulation.
Calcium Chloride Third Rail Sleet Removal. On the Aurora,
Elgin & Chicago Ry. a solution of calcium chloride made by dissolving caustic chloride of calcium in warm water in the proportion
of 5 Ib. of caustic chloride of calcium to I gal. of water has been
l
the prevention
used in the removal of sleet from the third r a ~ and
of the formation of a coat of sieet on the rail. The solution was
kept in a +gal. tank in the motorman's cab and was led through
a rubber tube to a %-in. pipe from which it was squirted upon the
third rail a few inches in front of a steel sleet brush. The flow of
liquid was controlled by a globe valve. The pipe was grounded to
the truck frame. There were four sleet brushes per car and each
was provided with a pipe. In operation only the forward pipe on
the third rail side was used. The flow of liquid was regulated
accoraing to the speed 01 the train and about r gal. per mile was
required. The sleet brush immediately behind the pipe spread the
liquid uniformly over the surface of the rail. A thin sleet formation
was dissolved and direct contact was made with the third rail. A
thicker sleet formation was not melted a t once, but was so loosened
that i t was scraped off the rail by the passage of one or two sleet
brushes. The formation of new ice was prevented as long as the
liquid remained on the rail, but after about 2 hours the liquid
was removed to such an extent that another application was necessary. This process removed the sleet and kept the trains in o
tion, l y t i t had a bad effect on the insulation of the third r a i r t d
car wlnng. This led to the development of a side contact shoe
having a cross-section similar to an inverted V. By riding on the
edges of the third rail this shoe either removes the sleet or makes
contact where the sleet is thinnest.
Conduit System. Figure 5 2 shows a section of cable conduit as
rebuilt for electrical operation, and Fig. 53 shows a standard section
of conduit, both as used in Washington, D. C. These are from a
paper by J. H. Hanna, Aera, 1913. The essential elementsof
this conduit system consist of cast iron yokes supporting both
wheel and slot rails and two steel "T"shaped conductor bars
supported from the bottom flange of the slot rail by insulators.
The wheel rails are fastened with four hook Imlts a t each yoke seat
with liners driven back of the bolts to allow accurate lining and
gaging of the rails after the yokes have been concreted. The
conduit proper is of concrete with concrete manholes a t each insulator, spaced 15 ft.apart. Insulators consist of malleable caps with
lugs by which the insulator is bolted to the slot rail porcelain insula-

644

ELECTRIC RAILWAY HANDBOOK

tors and forged studs cemented together, the latter supporting the
conductor rail by means of adjustable malleable clips or brackets.
The slot rail weight is 67 Ib., and conductor bar, 22.4 Ib. per yard.
Yokes are spaced 5 ft. apart and weigh about 350 Ib. each. The
conductor bars are connected with underground feeder cables a t
intervals varying from 800
ft. for lines having dense
traffic to much greater intervals in outlying districts.
The distribution system is
identical for positive a n d
negative sides and is cont r o l l e d a t substations by
double-pole, double-throw
switches, which make it possible to reverse the polarity
if necessary on account of
mounds on different sides of
section at Yoke
aifferent circuits.
The conduit system is a
FIG.32.-Cable conduit rebuilt for electric operation.
constant source of trouble to
both the railway company and the traveling public. I t is by far the
least reliable of all the systems. This is because of the difliculties of
ordinary operation and the excessive number of accidents which are
possible with this system and against which it is impossible to provide. Neglecting accidents, for the moment, this system is the
most difficult to keep in ordinary operation. The ordinary diffi-

PIG.53.-Slot

conduit. Washington.

culties reach a maximum with rain and snow storms. Snow tends
to fill the conduit and must be removed by pushing it to manholes
by scrapers. These manholes must be kept clean. The removal
of snow interferes with traffic. To return to the purely accidental,
various pieces of metal are washed or accidentally or maliciously
dropped through the slot and these foul the plow and cause short

TRACK BONDING

645

circuits. Among the most common articles which cause this trouble
may be mentioned automobile chains, hoops, rods, pipes and cables.
Defective switches or plow guides in the slot often make the car and
the plow start down different tracks. The results are bent plow
bars, broken yokes, bent and grounded plows. Locating a ground
in this system is often a slow process during which trafficis tied up.
Track Bonding
The following consideration of bonding deals with the application
of electrical conductor to maintain good conductivity from rail to
rail a t the joints of the track circuit. This conductor usually
serves no mechanical purpose a t the joint. I t should be noted that
in city work the practice of joining the rails together by welding is
increasing and that the use of additional electrical conductor is
unnecessary where this is the practice. (See pp. 4931.)
Classification of Bonds. Bonds may be classified according to
the materials of which they are made, whether of solid copper,
copper ribbon, or wire stranded copper; according to place of
application to the rail, whether to the head, web or base, and if
attached to web, whether "concealed" between joint plate and rail
or "exposed" by being run over the outside of the joint plate;
according to the method of attaching the terminal to the rail,
whether compressed or expanded, brazed, soldered or amalgamated.
Compressed and Expanded Terminals. The bond terminal is
forced into intimate contact with the walls of a hole drilled in the
rail. The two general types of this terminal are commonly known
as the "compressed (or stud) terminal" and the "pin terminal,"
respectively. The terminal of the compressed terminal bond is
forced into contact with the wall of a hole in the rail by a screw or
hydraulic press. The pin terminal is tubular and this tube is
expanded into contact with the wall of a hole in the rail by driving
a tapered steel pin into the bore of the tube. The following from
the results of tests by the Chicago Board of Supervising Engineers,
1911, gives briefly some of the important po~ntsrelative to the
installation of these types of bonds:
First. Averaging the thirty-two tests on each type, the hydraulic
compressed bond shows the least resistance, viz., about 96.65 per
cent of the pin terminal type. Stated in terms of conductiv~ty,
the in termnal bond is 96.65 per cent and the hand-compressed
bonc! 98.64 per cent of the hydraulic compressed bond considered
as the standard, loo per cent.
Second. The best form of terminal is that in which the flow of
copper is into and not out of the bore, that is, one in which the flow
is restricted. In this respect the compressed terminal is superior,
although not so easy to apply as the pin terminal, and it also insures
a much better mechanical attachment.
Third. A certain pressure between copper and steel is essential
to good electrical contact. Conductivity improves up to about
35,000 Ib. per square inch. With the pin terminal, however, it is
found that around 20,000 lb. the copper begins to flow out of the
bore and no higher contact pressure is possible on this account. This
maximum is somewhat dependent upon the texture of the copper

646

ELECTRIC RAILWAY HANDBOOK

and the lubricant used, and upon the manner in which the pin is
driven into place.
Fourth. General precautions to be taken in bonding as largely
developed from these tests are as follows: Use only accurately
ground drills, entered a t right angles to the web, and finish smooth.
The surface of the rail web should be cleaned for W in. around the
bore, and the latter should be thoroughly cleaned and dried before
bonding; thus the head of the bond will propcrly abut the web surface. In driving the pin a heavy lubricant should be used, first
with a drift pin of the proper size (MI in. for a %-in. terminal),
care being taken to avo!; striking the bond itself.
Brazed" bond terminals which are genBrazed Terminals.
erally flat are brazed to a spot on the rail which has been pre~ased
by grinding. The heat is supplied by a blow torch or heavy electric
current. The brass used to make the brazed joint is used in a
solder-like strip. A bond brazed electrically is sometimes called
an electrically welded bond. The current is supplied a t from I
to 6 or 8 volts from a transformer which, on a direct current road,
is supplied by a motor-generator set operated on trolley voltage.
Soldered Terminals. Flat terminals are soldered to the rail or
the end of a compressed or expanded terminal bond is soldered to
the rail after the terminal has been compressed or expanded into
place. The area over which the solder is to adhere to the rail is
prepared by grinding. Heat is supplied by a blow torch or, in the
case of the comp~essedterminal bond, the bond may be soldered by
the Thermo bonding process, in which the necessary heat for soldering is obtained from the heat of reaction of a special compound
which is ignited in a graphite cup held against the rail on the side
site that to which the solder is to be a plied.
OPE;kctric Weld Terminals. The demanB for something better
than a soldered contact and a substitute for the purely mechanical
contact has been responsible for the wide adoption of the electric
weld bond within the past few years. Although it has been used
for the most part as a head bond it is coming into general use for
concealed application to the web of the rail. There seems to be
practically no question regarding the permanency of the contact
that these bonds make with the rail, but some criticism has been
directed against some of their other features, particularly to the
breaking of the ribbons and to the inconvenience of using the bonding car on tracks over which traffic must be maintained. The
question also has been raised as to what effect if any the welding
heat has on the steel of the rail. Regarding injury to the rail by
heat there seems to be very little definite information. A number
of engineers have expressed a question or fear regarding this point,
but with possibly one or two exceptions no company has definitely
reported any broken rails from this cause. In the absence of more
definite complaint on this score it is safe to say that the injury to
rails resulting from the heat generated in welding rail bonds is so
small as to be negligible and can be practically disregarded in the
selection of a type of bond.
Oxy-acetylene Weld Terminals. This type of bonding is accomplished by welding bonds with forged or cast terminals to the rail,

TRACK BONDING
usually the head, and employing pure or Buxed co per to build
up the terminal of the bond along the rail head. So Far, it has not
found a very wide application, although it has been used in Minneapolis and St. Paul for a number of years with marked success.
I t possesses practically all of the advantages of the electric weld
bond and is free from the objection of requiring an ex ensive equipment and of interfering with traffic. Although it &es not make
quite as strong a contact with the rail as does the electric weld bond,
the contact is permanent and does not deteriorate with time. No
injuries to rails have been reported from the use of the flame which
apparently does not produce a higher temperature in the rail than
is generated by the electric weld.
Bonding Manganese Rail. As manganese steel is too hard to
drill, its use has led to some difficulties in track bonding. Such
material is used almost wholly in special work. Some companies
depend wholly on the cable connections around the special work,
as described on page 657, with no bonding of the separate pieces
except the mechanical joints. Others report that soldered, brazed
or welded bonds are satisfactorily attached to the manganese steel.
Other companies specify the insertion of soft steel plugs in each
manganese steel piece. into which holes may be drilled for com~ressed
ter&nal bonds;-in sdme cases only one ;oft steel plug is ihserted
in each piece, and a copper connection is run from this place to the
jum r cable.
I& Bond. The material, form and structure of the ideal bond
are briefly presented by C. W. Ricker, A.I.E.E., 1905, as
follows: To get the necessary conductivity, bonds are nearly
always of copper, about the only exception being those of tin
amalgam. To reduce cost and resistance, they must be as short
as practicable, and the manufacturing cost must be kept low,
to preserve the scrap value as near the first cost as possible. For
durability they must be flexible enough so as not to break or lose
contact by the allowable relative motion of the rails. They must
be formed so they may be applied to the types of rails in ordinary
use, in such position as to be protected from accidental damage and
from theft. They should be readily accessible for inspection and
repair. The cost of application must be kept low, and to this end
it is very important that the process shall be so simple and easy
that no highly skilled labor or extraordinary care is required to
install them with certain and uniform results.
Failure of Bonds in Service. Failures of bonds may be placed
in three classes, namely, (I) breakage of bonds, (2) disintegrationof
bonds, (3) impairment of contacts. The following discussionof
these cascs is from the above noted paper by C. W. Ricker:
Breakage of Bonds. Breakage may occur because of defects in
manufacture, as in copper bonds with welded terminals the strands
may be weakened by overheating where they enter the terminal;
and a slight but continuous motion of the joint will cause them to
break, one by one, a t this place. Long-continued jar and repeated
flexures will produce fatigue in the metal. Such breakage in the
case of either welded or solid bonds is of course most frequent where
the flexure of the bond due to rail movements is too great for its

648

ELECTRIC RAILWAY HANDBOOK

flexibility, which means ill-selected bonds or badly-kept track. A

less common manner of breakage occurs in laminated concealed
bonds which are too large for the space between the joint plate and
the rail, the bond shank is pinched and the working of the joint
under passing wheels tears off the outer strands by a kind of ratchet
effect, working them into the narrowing space a t top or bottom of
the rail web and sometimes squeezing them out of the joint in thin
ribbons. Bonds secured under the base of the rail may be frozen
in the ballast and torn off by the movement of the rail.
Disintegration. The surfaces a t the imperfect welds in composite
bonds corrode, increasing the resistance greatly and loosening and
weakening the bonds so that they may be pulled apart. Tin
amalgam used a t contacts or in masses, hardens and shrinks, losing
flexibility and contact with the bonded surfaces. I n the case of
amalgam plugs enclosed in cork boxes, the cork sometimes breaks,
allowing the soft amalgam to run out.
Zmpai7ment of Contact. By far the most important cause of
impaired contact is oxidation. This is greatly facilitated by the
presence of moisture, so that the slightest crevice into which moisture may penetrate and lodge is dangerous. Soft-soldered contacts
underground are not to be trusted, especially on track laid in streets,
which is sure to be wet with dirty water, though there is no apparent
reason why soldered contact entirely above ground should not be
durable, if all traces of corrosive flux are removed. Amalgamated
steel surfaces are not durable and soon rust in track exposed to
dampness. Expanded or compressed terminal bonds, whlch have
not been properly applied, may be loosened by the movement of
the rail, and well-soldered, brazed or welded bonds may be loosened
or torn off by the same means if they are too rigid. No mention has
been made of accidental breakage of bonds. Of course, bonds
which are improperly located may be knocked oE by rolling stock,
and various local external conditions may operate to destroy any
kind of bond.
Length of Bond. A short rail bond is more economical of copper
than a long one, but is more likely to damage from vibration of a
loose'joint. A short concealed bond should not be used where the
rail joint is not rigidly supported. The length of a bond to be
concealed between joint plate and rail is, under certain conditions,
somewhat dependent upon the spacing of joint bolts. The length
of bonds in use varies from 6 to 10 and more inches. The 10-in.
bond is in very common use. Electrified steam roads are using
bonds 16 to 2 4 in. long. Long bonds for spanning splice bars on
small rails should be about 5 in. longer (formed) than the splice
bar, and about 6 in. longer than the splice bar on large rails.
Resistance of Bonded Rail Joint. Neglecting the very unreliable
conductivity between rail and rail by way of mechanical joint
plates, the resistance of a newly bonded joint may be approximated
as follows:
Brazed, Soldwed or Welded Bond. Add I in. to the length of
the bond shank conductor between terminals. The resistance of
the bonded joint is approximately equal to the resistance of such a
length of the shank conductor.

649

STEEL RAIL RESISTANCE

m o a n
Ohc-Y)

4m.m.m
a",--

t--*a
*we-

=?moot
c)av)r
ma?,
w

$3
$4

o m e d
qmm?

UV)IOY)

o o m a
om--

....
r

850

ELECTRIC RAILWAY HANDBOOK

Compressed w Pin Terminal Bond. T o the resistance of a

length of shank conductor equal to the length from center to center
of terminals measured along the shank conductor add twice the
contact resistance of one terminal.

FIG.54 -Resistance annealed copper r a ~ bonds.

l

The resistance of the shank conductor may be determined from

l
and diameter of
Fig. 54. Knowing the th~cknessof the r a ~ web
termnal hole in web, the contact area (considering only area of hole
wall) may be determined from Fig. 55. For thickness of wpb

6.51

RAIL BOND RESISTANCE

see rail dimensions on pp. 3 6 3 9 . Common outside diameters of

bond plug of compressed or expanded terminal bonds are as follows:
S ~ z eof bond.
A.W G. (B & S.)
and clr mils

D~ameterof plug.
~nches

$6. %
+4. %. %.
%. %
%

0
00

y.

MH)

0000
300.000
500,000

Tests made by Messrs. Hall, Smith and Starbird at the Worcester

Polytechnic Institute indicate that the contact res~stancebetween
soft steel and pure copper a t pressures from 20,000 to 30,000 Ib.
per square inch has values from % to I microhm per square inch.
o m 2

U.OaXX)l ohm per sq

.-

.-

18
a

ro

inch

'.

"

..
.,

" " "

"

0000005

"

"

c.

'8

a.

"

"

'6

'a

'1

a'

m 1

0.0m
0

1
L
Cunts~tares of terminal ho1e.wuar-a incher

FIG.56 -Contact

resistance

These values are plotted in Fig. 56. The contact resistances of

bond termnals vary from each other, depending upon the many
factors in the rocess of installation. The value, o 8 microhm
(0 0000008 ohm7, may be used in the practical calculation of the
resistance of track return where bonds are carefully installed.
Resistance of Track Return. The resistance of a given section
of track using both rails as return conductors is equal to one-half the
resistance of the sum of all the joints plus one-half the resistance of
the rail between joints contained in one of the rails of that section.
T h ~ smay be determined by the following method: From one rail
length (ordinarily 30, 33, or 60 ft ) subtract the length of one joint
(distance between centers of the terminals of one bond installed),
and by the table, page 649, find the resistance of the part of rail
remaining (in the absence of more definite information the ratio
of resistance of steel to that of copper may be taken as 1 1 for
ordinary running rail) To this, add the resistance of one bonded
joint (see Resistance of Bonded Rail Joint, page 648). The resis-

652

ELECTRIC RAILWAY HANDBOOK

tance of the given section of track will be equal to one-half this

sum multiplied by the quotient obtained by dividing the length
of the given section by the length of one rail. This may be conveniently expressed as follows:

in which R

.=

resistance of section of track, ohms

L = length of section, feet

1 = one rail length, feet
1, 5 length of one joint, feet
r = resistance of rail, ohms per foot. See page 649.
r, = resistance of one bonded joint, ohms.

The allowable value of the resistance, R, depends upon the current

flowing and the allowable voltage drop. Its value is determined as
follows :
in which R = the allowable resistance of the given section of
track, ohms
e = the allowable voltage drop from one end to the
other of the given section, volts
I = current flowing in track return, amperes.
The allowable voltage drop may depend upon the satisfactory
operation of apparatus or it may depend upon local regulations or
upon the economical amount of power loss. The power loss may
be found as follows:
el
p =IOOO

in which p = power lost, kilowatts.

(Significance of remaining symbols as above.)
Size of Bond. A common allowance for the area of cross-section
of an ordinary joint bond is 500 cir. mils per ampere carried by the
bond. Short bonds may be operated a t a much greater current
density. The latter is of value in handling rush loads of short
duration.
The proper size and number of rail bonds per joint appears to be
a subject regarding which very little definite knowledge exists, and
the present practices of operating companies seem to be based upon
arbitrary rules rather than upon theoretical considerations. W.
A. Del Mar, in a letter pubhshed in the Electrical World, 1909,
commented upon the absence of information and standard practice
upon this point, and attempted to determine the proper ca acity of
bonds by the following three more logical considerat~ons: Make
the bond large enough to conform to proper mechanical conditions.
( 2 ) Determine the magnitude of the continuous and maximum
currents in the rail and make the bond large enough to prevent
undue heat. (3) Make the bond large enough to give a t least 90
per cent efficiency to the return circuit; that is, make the conduct-

8)

SIZE OF RAIL BOND

653

ance of the bonded rails a t least go per cent of the conductance of a

theoretically continuous rail. The following equation is given by
Mr. Del Mar for the effiziency of the return circuit:
Ll
EJ. =
L
L I + ~ ( / -K- )

where K = average e5ciency of bond, or the ratio of the conductance of the bond to the conductance of an equal length
of rail
L, = length of rail
L2 = length of bond between terminals.
That a short copper bond attached to heavy masses of cold steel
cannot attain a dangerously high temperature is obvious even when
carrying currents of the magnitude found on heavily loaded tracks.
The contact resistance of one terminal of a mechanically applied
bond in good condition is in the order of o.ooooo5 ohm. When
carrying a current of 500 amp., which is greatly in excess of
currents ordinarily found in rails, there would be a dissipation of
only 1% watts, and with 1000 amp. a dissipation of 5 watts per
terminal. This is, of course, in addition to the heat generated in
the copper of the bond, but as far as the contact is concerned there
seems to be no practical limit to its capacity to cany current.
Parshall in England, in an article on "Earth returns for electric
tramways," published in the Jmtrnal of the Insticution of Eleclrical
Engineers, 1898, stated that experience with pressure contacts in
central station work had demonstrated that I W amperes per square
inch was the safe limit, but suggests that 50 and even 2 s amp.
per square inch would be found more advisable for rail bonds.
As these figures are exceeded in practically all installations in this
country they cannot be regarded in any way as a practical limit for
current density. A number of companies, including some in large
cities, install only single bonds, while some employ double bonding
only when necessary to reduce the potential gradient on the tracks.
This would indicate that double bonding is not considered necessary
solely from the standpoint of capacity, but that its adoption is
demanded by other considerations. One of the usual reasons for
its use is to insure safety, and not place reliance on a single bond in a
permanent track where the repair of a bond would mean tearing up
the avement. Under extreme conditions double bonding is
justil%ble solely from the economic standpoint, and the factors
which determine thyse conditions may also be used to determine the
economic replacement of deteriorating bonds. (See below.)
These conditions can be determined when the constants of a given
system are known and the cost of power and the average current
i n the rails are obtainable.
Economic Replacement of Bonds. The length of life of a rail
bond in place may be anything up to the length of life of the rails
i t connects, depending upon the degree of perfection of the installation and the condition and care of track. The life of a bond may
terminate a t actual breakage or a t the point a t which its resistance

ELECTRIC RAILWAY HANDBOOK

654

becomes such that the voltage drop becomes too great or the saving
of energy brought about by its renewal will justify the total cost of
such renewal. The chart, Fig. 57, gives a method of determining
the resistance of a bond in terms of feet of adjacent rail at which
a bond should be renewed, considering financial economy only.

1 1 g I a0

m a $ualan!nw

1
*

-.0
X

%2 24
,$

-6 i?
.+2

"4

3:
E. 1;

<E" .ga
U

z3
22 g8
$2
-.e lta
.u
98
a

8 6
k! 8

tW
:I.
P r0

"Y

5i

5"
d

=,,:$i;
66 6 6

&A & A

I
dQs
rl

+ +_

EZ

s z ss

&&,a:

"Xz

Ob

fi

s P'

i 3'

; : ; ;

$??:
WON Qag

::
;

;
:

ECONOMIC REPLACEMENT OF BONDS

655

I t shows the increase in resistance over that of a well-bonded joint

a t which the energy saving amortizes the cost of replacement in a
given number of years. The use of the chart is explained by the
following example, solution of which is shown by the dotted lines
on Fig. 57: R.m.s. current in track, 450 amp.; weight of rail,
go lb. per yard; cost of energy saved, b . w z g per kw.-hr.; cost of
bond renewal, $2.00 per bond; estimated life of bond, 7 years;
interest 5 per cent. Chart shows that joints in this case should
be rebonded when resistance is (approximately) 27.6 ft. of rail more
than resistance of newly bonded joint.
The U. S. Bureau of Standards, in its Technological Paper No. 62,
gives a similar method of calculation as follows:
Let W = weight of rail per yard
I = root-mean-square current in rail over a 24-hour period
p = cost of power per kilowatt-hour in dollars
P = cost of ~nstallinga bond
n = number of years bond will last
L = reduction in joint resistance resulting from installation
of bond, expressed in feet of adjacent rail
r = rate of interest paid on invested capital;
then the resistance of the rail is very close to o.oo~/Wohm per foot
and the annual saving of energy in dollars due to installing a bond
would be

The annuity required to retire the investment, P, on a new bond,

a t the end of n years, its period of usefulness, is

P(=)R" -

where R =

+L
I00

When the annual power loss without the new bond exceeds this
annuity, it is obvious that the installation of a new bond would be a
matter of economy. The limiting condition would be obtained
when the energy charge is equal to the annuity. Equating these

(:--:))

two we get: 0.00876 z2Lp

-- = P -- from which any quantity
W
may be obtained provided thk others.are known. If we employ
the equation to deteimine a t what current double bonding becomes
economical, we have

In order to apply this equation with six independent variable

factors it is necessary to assume a set of values which would obtain
under normal conditions. Let it be required to determine whether
one or two 4/0, 10-inch compressed terminal bonds should be
used on IW lb. rails being newly installed under the following
conditions:

ELECTRIC RAILWAY HANDBOOK

656

= cost of energy
= So 01 per kw.-hr.
P = cost of lnstalhng bond =
60
n = life of bond
= 1 2 years
r = rate of interest
= 5 per cent
The reslstance of one Ielncn 410 bond, including contact resistance, is very close to o -55
ohm, and would probably average
more. Two such bonds in parallel would have approximately onehalf of this, or o -275
ohm, which is also the decrease In the
resistance of the joint resulting from the installation of the
second bond. As a loo Ib. rall has a resistance very close to
0-1
ohm er foot, L would be 2 75 ft. Substituting these
values in the agove equation we find that I 2 is equal to 15,660 or
Z = 125 amp.
The Bureau of Standards has examined numerous railway load
curves and has found that the ratio of the root-mean-square current to the all day average ranges from I 25 to I 4 If we use the
lower of these values, whlch is more applicable for the heavily loaded
lines with whlch we are concerned in this discussion, we get loo
amperes as the all-day average value of the current. With a load
factor of 40 per cent this would give 250 amp. per rail as the
value a t the peak period. W e the values here assumed are normal in every respect, they are undoubtedly on the side tending to
make the llmiting current small. A lower cost of energy, a shorter
l ~ f eof the bond, and a higher cost of installing a bond will tend to
give a larger current as the point double bonding becomes economical As the values assumed for these three variables are
obviously near the l i m t In the other direction, it is difficult to conceive how the economy point would be reached a t any allday
average current value much less than loo amp.
The value of the resistance whlch a deteriorating bond must reach
before economy ~ 1 justlfy
1
its replacement can be determ~nedfrom
the above equation in which the reduction in the j o ~ n treslstance
resulting from the installation of the new bond becomes the unknown quantity. Transforming the equation for this purpose we
find that

The following concrete example will illustrate the applicationof

this formula. Let us assume:
W, the weight of rail per yard = roo Ib.
P, the cost of replacing a bond = $0 80
p, the cost of energy
= o 005 per kw.-hr.
n, the life of the new bond
= 10 years
r , the rate of interest
= 5 per cent
Let us also assume that the current is of such a value as to give
a drop of o 8 volt per 1000 ft as an allday average on a perfectly
bonded rail This is not ordinarily exceeded, even in regions ahere
the problem of electrolysis does not exist The resistance of a
loo lb. rail is O O I ohm per 1000 ft which would limit the

667

SPECIAL WORK BONDING

average current to 80 amp. Taking I 3 as the ratio between

the root-mean-square and the all day average current we get for our
equation I = r o q amp The replacement cost of a bond is usually greater than the cost on new work and is therefore taken at 80
cents. One-half cent per kllowatt hour for power may seem low,
but as it is assumed to be the cost of energy which will be saved by
the application of the bond, it would not be logical to load it with
fixed charges and operating costs other than the fuel. Upon this
basisit is high rather than low. A replacement bond is often installed
on old track which is partially worn out and which may be entirely
replaced within a few years. Ten years, therefore, is considered
as a liberal life for the new bond. The scrap value of bonds is small
and is neglected in this discussion. Substituting thesevalues in the
above equation we find tbat L = 13 4 ft. As the new bond itself
will test equal to from 3 to 6 ft of rail this means that the old bond
will test equal to about 18 ft. of rail before a new one can be
installed m t h economy. Electrolysis and voltage conditions ordinanly, therefore, demand a better return circuit than economy
itself can dictate. I n fact, it is doubtful if economy alone in many
circumstances will justify any but the cheapest and simplest type
of bonding.
Cross Bonding. I n order to maintain the current in the two or
four rails of the track return circuit as nearly equal as possible,
and to care for a possible open bond, the rails should be bonded
together by cross bonds. These make contact a t the web of
the rail or at the outside of the rail head. The spacing of cross
bonds depends upon the density of traffic. I n city work it is customary to place them about three per ~ o o oft They are spaced
from this to 1500 ft , but commonly xooo ft in interurban work.
Where track circuit s~gnalsystems are used the matter of cross
bonding must be considered in its relation to the signal system.
Cross bonding may be entirely prohibited or, in the case of alternating current slgnal systems, it may be provided by inductive bonds
(see page 773).
Bondmg at Special Work. Special work must be removed comparatively frequently, its joints are subject to great vibration and
offer construction dificulties to
bonding, it may be of steel having low conductivity, and it contains many joints. Thus, the
bonding of joints through spec-

soldered Comectloa

PIG 5 8 -Bonding around swtch. PIG.59 -Bonding around branch-off

ial work in a manner similar to that on ordinary track is unreliable. It has been found satisfactory to bond around special

658

ELECTRIC RAILWAY HANDBOOK

work by means of bare stranded copper (500,000 to 1,000,ooo

cir. mil cable is commonly used for this work). Figs. 58 to 62,
inclusive, give schemes for bonding around several typical pieces of
special work. It is the practice of many roads to bond through
as well as around special work. This is necessary in city work and
is done by connecting each piece of rail in the special work to the
cable which bonds around thespecial work, or by bonding the joints
in the special work as in straightaway track construction. As in the
case of cross bonding, where track
signal systems are used, bonding
a t special work should be considered
FIG. 6o.-Bonding around cross- in its relation to the signal system.
over.
Much careshould be taken to insure
good contact a t the terminals of cables used in bonding around
special work. The following briefly outlines the tests and results
of the Chicago Board of Supervising Engineers, 1911, in the determination of the proper method to use in attaching such auxiliary
cables:
The tests comprised five different types of terminals.
First. Welded connection (Fig. 63) in which the weld is made
under current of 15,oooamp. to 25,000 amp. applied with flux and
spelter and a final compression of
about 15 tons.
Second. The same type of joint
with a considerable reduction in
the final pressure applied.
Third. Welded connection in

=
d o d
ilrctlon

FIG.61.-Bonding

around Y.

FIG.62.-Bonding

around crossing.

which the cable is enclosed by a mold containing a n overflow

chamber and clamped against the freshly cleaned web of the
rail. Especially prepared copper is poured into this mold.
Fourth. Bolted type terminal (Fig. 63) constructed especially
for this experimental work and so designed that the thickness of
copper and the length of iron bolt are inversely proportional to the
expansion coefficients of copper and iron, so that the changes in
temperature will not tend to loosen the connection. This type

BONDING PRECAUTIONS

659

was tested under varylng compressions with both increasing and

decreasing loads and plain as well as amalgamated surface.
Fifth. A multiple-pin type terminal, which was made up from
standard bond terminals.
An analysis of the general results of the tests indicates:
First. A very large improvement in the conductivity--over 50
per cent-of Type I, by reducing the final compression as noted
under the description of Type 2.
Second. Under a pressure of about 75.000 lb. or 50,000 Ib. per
square inch the improvement in conductivity practically ceases.
Third. Bolted type terminals show ten to fifteen times greater
unit conductivity than either of the welded type terminals and a t
about one-half the cost per terminal.
Fourth. The cable weld (Fig. 63) showed the greatest mechanical strength-13,200 lb. shear per square inch of contact.

FIG.63.-Welded

and bolted cable terminals.

Fifth. A bolted type terminal can be developed which will be

suitable for installation in limited numbers, while for a large number
of connections the cable weld is most desirable on account of its
higher mechanical strength and conductivity a t lower cost.
As a result of these tests it was decided to adopt the electrically
welded terminal as the standard. This standard is now in force
on the Chicago Surface Lines.
Code of Bonding Precautions. T o insure proper attention to
details essential to proper bonding the bonding department should
be provided with a code of bonding precautions. The following
is an example of such a code by Howard H. George, Electric
Railway Journal, 1914, for the installation of pin terminal bonds:
I. Every roadmaster and foreman should see t h a t one or more men in
each gang are taught the proper way of installing bonds, and should be
sure that any bondrng done thereafter is performed by these men.
2. When renewlng rarl or jornt plates on srngle track In operation. care
should be taken not to open or disconnect both rails a t the same time. a s
this would open the return circuit by which the current returns from the
cars to the power hoase. When it is absolutely necessary to open both ra~ls.
a long copper jumper should be installed to connect the open ends so t h a t
the path of the return circuit shall not be interrupted. This applies more
particularly to road ends and interurban 11nes.
3. Whenever any track is opened up and any ground wires for electric
lights. llghtnlng arresters, or other electrical apparatus which should be
connected to the rail, are found disconnected, they shou!d be reported at

ELECTRIC RAILWAY HANDBOOK

once to the bond Inspector or d~stributiondepartment. so t h a t they may be
repalred before the track 1s closed up. This 1s very Important and should
recelve careful attcntlon.
4. No bond holes should be dnlled untll just before the bonds are ready
t o be ut In. There are, of courze tlmes when ~t 1s d e ~ r o u sto have the
holes
before the rall 1s placeh on the tles. When thls occurs. ~t 1s
necessary t o place a tlght-fitt~ng lug In the hole a s soon a s ~t 1s dnlled to
avold any posslble ~ntroducttono r molsture. To dnll a hole a day or two
before and not rotect ~t from molsture means a film of rust In the hole.
whlch will great& Increase the resistance of the jolnt
5. Old bonds should never be used agaln because they become battered
u p In drlvlng them out. Then. when they are put tn agaln they wrll not
make good contact wlth the rall, whlch means a poor bond Where a bond
IS removed from the rarl ~t 1s not advisable t o use the same hole I U putting
In a new bond. unless some precaut~onarymethods are used. The proper
way 1s to dnll a new hole, but a s thts 1s not allowable In some types of rarls
ream out the old hole and use a bond m t h a s
lal large-stzed termmal.
6. Great care should be taken w ~ t hthe d n l E s e d In maklng bond holes
If a n improperly ground drlll IS used the hole wtll be Irregular and oval
shaped, thus glvlng a poor contact between t h e termlnal and the rad. All
dul and broken drtlls should be carefully boxed, labeled and sent to the shop
t o be reground, where the company has tnstalled a speclal machtne for the
Durwse t o do the work oerfectlv and a t much less exnense than could
b s i l b l y be done by hand.'
7. I n dnlllng bond holes never use 011 to lubncate the dnlls. I t 1s better
not to use anything, but where ~t 1s absolutely necessary t o use a lubncant.
nothln more than a soda solut~onshould be em loyed.
8. lfoles, after belng dnlled. should be carefully cleaned of any chl s.
and mped dry of any solutlon that may have been used to lubncate the dr1Es.
The holes must have a smooth and dry surface so t h a t the bond termlnal
mll make a good contact all around.
g. Wlth a proper-stzed hole. t h e bond terrmnal m l l make a very snug
fit, not small enough to have t o be dnven m t h a heavy maul nor large enough
t o be put In eastly m t h the hands. I t should requlre a couple of taps wlth
a hammer welghlng about 3 Ib. Wlth a heavy hammer or s p ~ k emaul. the
head of the bond ternunal 1s very llkely to be battered, and the taper punch
t
bend the termlnal.
struck on the slant. causlnn ~t to s ~ l l and
10. After the bond terminals are In posltlon. always d n v e the long steel
taper punch entlrely through the termtual. t a k ~ n gcare to strrke the punch
squarely on the head. The small end of thls punch should be d ~ p p e dtn some
h n d of heavy grease, such a s track grease. just before ~t la drlven through
each termlnal. The grease mll lubncate the sldes of the punch, thereby
expanding the termlnals and not dramng the copper m t h the punch.
1 1 . Dnve Into each of the expanded temlnals one of the short d n f t plus,
thus expanding the cop r a llttle more. Thlr pm should be drlven In untll
~t 1s lust flush m t h the g a d of the bond termknal.
d b y strarghtemng out the bond con12. The bond should then be sha
ductors. and formlng them so that t E y will not be cut by etther the track
bolts or t h e spltce bars. If ~t 1s a 36-m. bond. ~tshould be so shaped that
~t wlll In no way tnterfere wlth the removal of the spltce bar.
13. The bond, and particularly the bond termlnals on both sldes of the
rail are to be patnted w ~ t hsome good weatherproof palnt, care belng taken
t o bee t h a t the palnt fills the space back of the termlnal heads.

trilled

Bond Testing. An inspection to determine the condition of the

joint between bond terminal and shank may be made by sectioning
the termina!~in various planes, polishing the cut surfaces with a
smooth file, emery, and crocus to remove all burrs, and then
etching with a mixture of strong sulphuric and nitric acids, when
the defective welds will show as fine black lines, and the actual
welds and the form of the various component parts of the bond a t
that surface can be traced bv the different colors of the metal after
etching.
Bonded Joint Testing. Joints should be tested immediately
after installation of bonds. At this time compressed or expanded
terminal bonds should be tested for electrical resistance, and

BOND TESTING

661

soldered, brazed and welded bonds should be subjected to both

hammer and electrical resistance tests. Throughout the life of
the bonds, joints should be kept under inspection to a degree
depend~ngupon the cond~tionand character of the roadbed, the
intensity of traffic, the cost of energy, danger from electrolysis,
and the voltage drop allowable for good operation or permitted by
local ordinance. There are two general methods of test~ngbonds:
one, an aggregate test, is a test todetermine the resistance of a whole
section of track, thereby giving a general idea of its condition; the
other is an indimdual test of each ioint in the track.
Aggregate Test of Return ~ucuit. (See "Determination of
Potential Drop in R a i l s , " r g e 695.)
Individual Joint Bond esting. An individual test is necessary
to determine the exact condition of the individual bonds. Many
methods and various types of apparatus have been devised fir
such testing. The resistance of the bonded joint is generally compared with the resistance of adjacent rail by comparing the voltage
drops across this joint and in the rail, accompanying the flow of
current through joint and rail
General Classification of Bond Testing Apparatus :
Hand Instruments
One-man Instruments
S~ght
Double mllltvoltmeters
Roller dlrect readlng
Standard for returns wlth 40-50 amperes
Hlgh senslblltty type uslng battery ~mpressed1 0 - ~ aamperes
Sound
Conant
Crown
Two-man Instruments
Sight

Dxfferentlal mllhvoltmeter. balanced by measuring adjacent rall

m t h steel ta
Double rmlllvormeter. balanced In same manner
Test C a n
Two classes of ropulslon
Self- rope11eB
~ra&r
Two classes current through joints
Clrculatlon from low-voltage motor generator set
Resistance and propelling current
Four classes lndlcatlon
Autornatlc recordlnn and marhna
Autornatlc slgnallngand rnarklnf
Autornatlc slgnallng and hand rnarklng
lndlcatlng and hand marklng.

Typical Hand Bond Testing Instrumen?s. The instrument

consists essentially of two parts: h s t , a foldlng framework fitted
with three steel rail contacts 3 ft. apart as shown in Fig. 64; and
second, a standard d~fferentialvoltmeter controlled by a two-point
switch (in the circuit through the unjointed rail). As shown in
Fig. 64, the frame is so placed on the rail that the joint is included
beween two of the terminals with an equal length of solid stock
rail between the center and the other outer terminal. If the voltmeter reading indicates a joint conductivity of ~ o oper cent or
over, this is the only observation required. If less than roo per
cent, the voltmeter is connected up with a detachable prod by

662

ELECTRIC RAILWAY HANDBOOK

which contact is made a t successive points along the rail until a

balance appears on the differential voltmeter. The conductivity

dssembled Plea

FIG.64.-Contact frame for hand bond testing.

of the joint may be then computed from the relative rail distances
included between
An average of about 2 0 0 joints may be
tested in- a day of 8
hours on a fairly free
t r a c k , which corresponds to slightly over
a mile of single track a
day per man.
The scheme of conR
nections of a n o t h e r
t n i c a l hand bond test&'(the Roller Smith) is
shown in Fig. 65. I t
consists of a T-shaped
bar, C , similar to that
shown in Fig. 64, upon
which three contacts are
mounted, and connected with the electrical
mechanism as shown.
The contacts are made
of short pieces of hacks a w blades h e l d i n
clamps. B y rocking
the handle H of the bar
a t right angles to the
rails the blades s a w
through the dirt and
FIG.65.-Hand bond testing circuits.
scale on the rails and
insure good contacts. The principle of operation is that of
the Wheatstone bridge. By turning the handle K the resistance

663

ELECTROLYSIS

of the two legs of the circuit is varied. The indicator attached to

the handle K moves around a scale which is so graduated as to read
directly the resistance of the section containing the bond in terms
of the resistance of unbroken rail when the needle deflection is zero.
Bond Test Car. A test car makes possible a rapid and complete
report of the entire track by making a continuous record of track
condition and location as the car proceeds a t a speed of 8 or 10
miles per hour. As a motor-generator set may be carried, its
operation may be independent of the current in the track due to
ordinary traffic and i t may be operated where traffic conditions
are such as to make hand bond testing highly difficult or impossible.
Tests may be made by i t with great rapidity, thusaffording knowledge of the true condition and location of bonds of a whole system
in a very short time.
Electrolysis
The electrolytic corrosion of earthed metallic conductors r o ceeds a t a rate dependent upon the rate of flow of the current w ~ c h
leaves the metallic conductor for the earth.
Theoretical Rate. With a n oxidizing anode the weight of anode
corroded by I ampere in I second is equal to the electro-chemical equivalent of the metal of the anode. The amount of metal
(pounds) corroded by I ampere in I year is equal to the electrochemical equivalent of the metal (grams) multiplied by 60 X 60 X
24 X 365 X 0.0022046, the latter figure being the conversion
factor between grams and pounds. The product of these factors is
69,524, which may be termed the conversion factor between the
electro<hemical equivalent of a metal as usually expressed in
grams, and the theoretical amount of such metal, in pounds, which
would be corroded by the electrolytic action of I ampere in I year.
The values are as follows:

Metal

Iron (ferrous). .........................

Iron (femc)'. .........................
Lead.. ................................
Copper ...............................
Aluminum.. ...........................
I

Electrochemical
equivalent,
m

Corroded by
one ampere
per year.
pounds

o.000~894
0.0001929
o.ooro7r8
0.0003293
o .oooog~s

ao. I
13.4

74.5
22.9
6.5
I

Less commonly met with than ferrous.

Variations from Theoretical Rate. McCollum and Logan show

results of some very extended studies by the U. S. Bureau of Standards on the variations from the above theoretical rate. Some
of their conclusions are as follows, their term "efficiency of corrosion" being based on the above theoretical rate as IOO per cent.
They state that the current density has a marked effect on the
corrosion of iron in soils, the efficiency of corrosion being in general
greater as the current density is lower. I n saturated soil the cor-

ELECTRIC RAILWAY HANDBOOK

rosion may vary between 2 0 and 140 per cent for a range of current density varying from about 5 to 0.05 milli-amperes per square
centimeter. Moisture content also has a marked effect on efficiency of corrosion, it being in general greater with increased
moisture content up to saturation of soil. Temperature changes
within the limits commonly encountered in general practice have
no marked effect on corrosion efficiency, neither has the depth of
burial of pipes, other conditions remaining constant. The amount
of oxygen present has no appreciable effect on the efficiency of
corrosion in the case of iron immersed in liquid electrolyte, but it
has a marked effect on the end products of corrosion. If the
corrosion is rapid and supply of oxygen small, there will be a preponderance of magnetic oxide, while if the rate of corrosion is low
and the sup ly of oxygen abundant, the ferric oxide will predominate. 8wing to the fact that the supply of oxygen around
pipes buried in earth is always more or less limited, the character
of the oxides formed gives some indication as to the rate of corrosion, and thus indirectly as to the cause of the corrosion if local
conditions are properly considered. The efficiency of corrosion was
found not to be a function of the voltage, except in so far as the
current density may be affected. Voltages as low as 0.1 to 0.6
volt showed practically the same efficiency of corrosion as 5 to
10 volts or higher.
Corrosion tests on a large number of different
kinds of soil from widely different sources with average moisture
content and current density indicated that corrrosion efficiency
between 50 and I I O per cent may usually be expected under most
practical conditions.
Resistance of Soils. With a given potential difference between
two earthed metallic structures, the amount of current which will
flow between them is dependent not only on the contact resistance
but also upon the resistance of the soil. Relative to the latter,
McCollum and Logan state that the resistance of soils varies
throughout a very wide range with variations in moisture content,
the resistance of comparatively dry soil being of the order of
several hundred times the resistance of the same soil a t about
saturation. Above saturation increase in moisture content has
but little effect on the resistance of the soil. The resistance of the
soil varies greatly with temperature within the ordinary range
encountered in ractice. In the case of the soils tested the resistance a t 18 deg. gelow zero C. was over two hundred times as great
as a t 18 deg. above zero C. Even a t about freezing temperature
the resistance will be several times that a t summer tem eratures.
This has an important bearing on the magnitude of the etctrolysis
trouble that may occur a t different seasons and also indicates that
where practicable voltage surveys should not be made when
extremely low temperatures prevail. These authors give a table of
the specific resistance of soils as found in Philadelphia, Washington,
St. Louis and elsewhere, showing it to vary between the extremes
of 400 ohms per cubic centimeter for a St. Louis blue clay with
26 per cent moisture, and 2,340,000 ohms per cubic centimeter for
Washington air dry red clay with 4 per cent moisture. The majority of soils tested show resistivities of between 1000and sax,
ohms per cubic centimeter.

ELECTROLYSIS
Contact Resistance. Polarization and film resistances a t the
surface of the pipes may be an important factor in current flow.
As soon as an electromotive force is applied to a buried pipe the
current flow drops off rapidly with time, especially during the first
few minutes, due to the setting up of counter electromotive forces
and the formation of film resistances. McCoUum and Logan show
the effective resistance as a function of time after the application of about 6 volts between two short lengths of cast iron pipe
buried about 12 ft. apart. The initial resistance of about 18
ohms practically doubled within half an hour after the voltage
was applied and after that the resistance remained practically
constant. In this case the effect of polarization and film resistance
was practically as great as the total soil resistance between the pipes.
The Character of the Electric Railway Roadbed is an important
factor in determining the extent of leakage of stray current into
the earth. A well drained rock or concrete roadbed may in general
be expected to offer much higher resistance to the leakage of current
than one in which the construction is such that a large amount of
moisture is retained.
The U. S. Bureau of Standards presents the following conclusions
on this subject. Roadbeds constructed with solid concrete ballast
and vitrified brick or other nonporous pavements have a low leakage
resistance to earth, which is affected only moderately by seasonal
and weather changes. There is little difference between wood and
steel ties in their effects on the resistance of roadbeds of this kind.
Insulating layers of bituminous materials are not of practical value
in reducing leakage currents from such roadbeds. The resistance
of 1000feet of single roadbed of this type is from 0.2 to 0.5 ohm
under ordinary conditions, but may be double or treble these values
when the ballast is frozen to a depth of I foot or more. For double
roadbed of this type the resistance is approximately 70 er cent of
that for single roadbed, or the leakage from double tracg would be
about 40 to 50 per cent greater than from single track. Roadbeds
constructed with a foundation of clean, crushed stone under a
concrete paving base have about three times the resistance of those
with a solid concrete ballast. Roadbeds with a full crushed stone
ballast and a Tarvia finish have a very high leakage resistance, of
ft. of single track. The leakage
the order of 2 to 5 ohms for ~ooo
from a double roadbed of this and other high resistance types is
from 80 to IOO per cent greater than from single roadbeds. The
resistance of earth roadbeds in which the ties are embedded and
therefore kept in a moist condition, is much lower than that of
open construction roadbeds, being from I to 1.5 ohms for r o o o ft.
of single track under normal conditions and considerably more when
the ground is frozen. The resistance of roadbeds of open construction is subject to wide variations, depending on the condition
of the ties and ballast. In very dry weather with good ballast the
resistance will be 10 to 15 ohms and even more for ~ooofeetof single
track, but when wet will drop to from 3 to 5 ohms. Cinders, gravel
and particularly crushed stone, when used as ballast in open track
construction, will produce very high resistance roadbeds. Earth
has a tendency to keep the ties moist and therefore increases the

666

ELECTRIC M I L W A Y HANDBOOK

leakage. Open construction track is often considered as being

insulated from the earth, but this is not strictly true, even though
the leakage may not be in harmful amounts when compared with
other types of construction. Assuming a potential difference
between a track and the earth of 5 volts and a leakage resistance of
10 ohms per rooo ft. of single roadbed, the total leakage per mile
of track would be 2.64 amp. This small leakage current would
not ordinarily be harmful to underground structures in the vicinity
of the track. Zinc chloride and other chemical salts used as preservatives render ties highly conducting and greatly increase leakage
currents from tracks. Unless combined with some other material,
such as creosote, these salts gradually leach out, particularly in
damp climates, and eventually their influence on the resistance of
roadbeds disappears. Creosote has very little effect upon the
resistance of wood ties, but a treating material consisting of 75
per cent gas oil and 2 5 per cent creosote appears to increase their
resistance materially. No direct comparison, however, wasmade
between these two treatments. Open track construction on
which ties treated in this manner were employed had a leakage
resistance about twice as great as similar roadbeds with untreated
ties and about four times as great as roadbeds with chemically
treated ties.
The Bureau of Standards suggests that electric railway companies
can do much toward reducing leakage currents from their tracks by
observing the following regarding roadbed construction: Solid
concrete ballast should be abandoned, and clean crushed stone
should be used as a foundation under ties. This type of construction is approved by the Am. El. Ry. Eng. Assn., as ~t gives greater
resiliency to the track and is cheaper than the full concrete ballast.
Where crushed stone or gravel is used it should be kept clean by
proper coverings or pavements. If earth, sand, or street dirt is
permitted to filter into ballast of this character, its function as a n
insulating material is greatly impaired. Salt, which is often used
to prevent frogs and switches from freezing, will greatly reduce the
resistance of roadbeds and should be avoided if possible. I n open
construction rails should be kept out of contact with the earth.
The roadbed should be well drained to prevent fine material from
washing into the ballast and to keep the ties as dry as possible.
Vegetation should be kept down, as this tends to make the roadbed
moist and to fill the ballast with foreign material. Zinc chloride
and similar chemical preservatives should be avoided where the
escape of stray currents is objectionable or where block signals
are used. A treating mixture of creosote and gas oil improves the
insulating properties of wood ties.
Self-corrosion is very generally regarded as being due primarily to
the presence of local galvanic currents a t the surface of the corroded
metal, due either to physical differences between adjacent points on
the surface of the metal or to foreign conducting substances in the
soil. For example, if a piece of carbon in the form of coke be
embedded in the soil in contact with the pipe surface a t one point
there will be a differenceof potential between the coke and the pipe,
and a current will flow from the pipe through the moist soil to the

SELF-CORROSION

667

coke and return to the pipe. through the contact point between the
two. The action here is exactly analogous to the action in a primary battery in which a piece of zinc and a piece of carbon are
immersed in the electrolyte and connected together through an
external circuit. A current flows through the electrolyte from the
zinc to the carbon, thereby corroding the zinc. The electromotive
force given by iron or steel when embedded in ordinary soils in
contact with coke is usually about 0.6 volt, and this is sufficient to
give rise to very rapid corrosion and pitting of the iron surface.
Action of this kind is frequently encountered where pipes are
embedded in soils containing cinders in which particles of coke may
be found. Galvanic action of this kind may take place as a secondary action following electrolysis from stray currents. If stray
currents are discharged from a pipe, producing initial corrosion, the
corroded iron thus carried into solutlon in the soil comes in contact
with oxygen dissolved in soil water which results in the precipitation
of iron oxide on the surface of the pipe. This iron oxide is a conductor of electricity, and i t also exhibits a n electromotive force against
the iron, as in the case of the piece of coke. Hence, where iron
oxide is thus deposited on the surface of the iron a t any point,
corrosion may continue to some extent even though the stray currents which initiated the trouble have been removed. Selfcorrosion may also occur from galvanic action due to physical
differences between different points of the surface of the iron itself,
or due to differences in the electrolyte in the soil near to adjacent
portions of the iron surface. Unfortunately, self-corrosion generally manifests itself in a manner very similar to that caused by
stray currents. Pipe corroded under conditions such that no stray
currents could exist exhibits pitting very similar to that caused by
electrolytic corrosion. Because of this similarity in appearance it
is not possible in general to determine by inspection of a corroded
pipe whether or not the corrosion was caused by stray currents
or by local galvanic action. Owing to this fact it not infrequently
happens that cases of pipe corrosion are charged to the railway
companies when the, damage was actually due to local causes
arising from the nature of the soil or of the pipe, or both.
The only sure way of determining whether or not stray currents
are causing corrosion in a n y particular case is by making proper
electrical tests to determine whether or not the pipes are actually
discharging current into the earth. I n a case where serious corrosion has been caused by stray currents and the cause of these
stray currents later removed, the only certain way of determining
whether or not the previous corrosion was caused by stray currentsor
b y local influences is by making actual corrosion tests in the soil
under the same average conditions of moisture and using the same
kind of iron as was previously found corroded. I n the absence of a
test of this kind i t is not possible to fix with certainty the cause of
the damage.
Electrolysis in Concrete. Following an extended investigation
in this matter, the U. S. Bureau of Standards presents the following
conclusions: I. The observations of previous investigators that
the passage of current from a n iron anode into normal wet concrete

ELECTRIC RAILWAY HANDBOOK

caused the destruction of the test specimen by cracking the concrete
were only partly confirmed. This effect was found not to occur in
most of the specimens tested when the potential gradient was less
than about 15 volts through a distance of 3 in., or about 60
volts per foot. These figures must be considered as but roughly
approximate as they de end much on conditions. 2. Of the numerous theories that have geen advanced for the cracking of reinforced
concrete due to electric current that one which attributes it to
oxidation of the iron anode following electrolytic corrosion has been
fully established. The oxides formed occupy 2.2 times as great a
volume as the original iron, and the pressure resulting from this
increase of volume causes the block to crack open. 3. Metals which
do not form insoluble end products of corrosion and all noncorrodable anodes never cause cracking of the concrete as a result of the
passage of an electric current. 4. The mechanical pressure
developed a t the iron anode surface by corrosion of the Iron has
been measured in a number of cases and has been found to reach
values as high as 4700 lb. per square inch, a value more than
sufficient to account for the phenomena of cracking that have been
observed. 5. Suggestions of some engineers that co per-clad steel
or aluminum be used as reinforcing material have geen shown to
be impracticable, since the copper coating is readily destroyed
and the aluminum is attacked by the alkali in the concrete. 6.
Corrosion of iron anodes even in wet concrete is very slight a t
temperatures below about 45 deg. C. (113 deg. F.). 7. For any fixed
temperature the amount of corrosion for a given number of amperehours is independent of the current strength. 8. The lack of
corrosion of the iron a t temperatures below 45 deg. C. is due to the
inhibiting effect of the Ca(OH)* and possibly other alkalies in the
concrete. 9. The rapid destruction of anode specimens of moist
concrete a t high voltages (60 to 100volts or more) is made ossible
mainly by the heating effect of the current, which raises tRe temperature above the limit mentioned above. If the specimen be
artificially cooled no appreciable corrosion occurs, and no cracking
results. 10. The potential gradient necessary to produce a temperature rise to 45 deg. C. with consequent corrosion, in the specimens used, was about 60 volts per foot. For air-dried concrete it
is much higher. This shows that under actual conditions corrosion
from stray currents may be ex ected only under special or extreme
conditions as noted below. ~l?esefigures are but roughly approximate since they will vary greatly with the conditions, such as the
size, form, and composition of the specimen, but they serve to show
the order of magnitude of the voltage required to produce trouble.
1 1 . Since the passivity of iron in concrete is due chiefly to the
Ca(0H)r present it appears probable that old structures in which
the Ca(OH)* has been largely converted into carbonate will be more
susceptible to the effects of electric currents than comparatively
new concrete with which the foregoing experiments have been made.
The increase in the eficiency of corrosion would, however, be a t
least partly offset by the increase in the resistance of the concrete
which would accompany the change. 12. The addition of a small
amount of salt (a fraction of I per cent) to concrete (as is frequently

ELECTROLYSIS I N CONCRETE
done to prevent freezing while setting) has a twofold effect, viz., it
greatly increases the initial conductivity of the wet concrete, thus
allowing more current to flow, and it also destroys the passive
condition of the iron a t ordinary temperatures, thus multiplying
by many hundreds of times the rate of corrosion and consequent
tendency of the concrete to crack. Salt, therefore, should never
be used in structures that may be subjected to electrolytic action.
Further, reinforced concrete structures built in contact with sea
water, or in salt marshes, are moresusceptible to electrolysis troubles
than concrete not subjected to such influences. 13. Specimens of
normal wet concrete carrying currents increase their resistance a
hundredfold or more in the course of a few weeks. 14. The rise of
electrical resistance is probably due to a number of causes among
which are the precipitation of CaCOa within the pores of the concrete, thus plugging them up. A slight amount af salt tends to
prevent this precipitation and interferes with the rise of resistance,
thus still further emphasizing the detrimental effect of salt. 15.
Contrary to the observations of previous investigators there was a
distinct softening of the concrete near the cathode. This begins at
the cathode surface and slowly spreads outward, in some cases as
far as one-fourth inch or more. After exposure to the air this
softened layer becomes very hard again, but remains brittle and
friable. 16. The softening effect a t the cathode noted above
caused, under the conditions of the experiments, practically complete destruction of the bond between reinforcing material and the
concrete, reducing it to a few per cent of its normal value. 17.
Unlike the anode effect which becomes serious in normal concrete
only on comparatively high voltages, the cathode effect develops
a t all voltages used in the experiments, the rate being roughly
proportional to the voltage in a given specimen. 18. In general
the cathode effect occurs under conditions which may not infrequently occur in practice and is therefore probably a more serious
matter practically than the anode effect about which so much has
been written. This trouble is unlikely to be serious, however,
except where the concrete is wet and the potential differences
rather large. 19. The softening of the concrete a t the cathode is due
chiefly to the gradual concentration of Na and R near the cathode
by the passage of electric current. In time the alkali becomes so
strong as to attack the cement. 20. Softening a t the cathode is
increased by increasing the Na and K content of the cement, and
reduced by diminishing this content, a t least within the range
below 10 per cent of the total salts. 21. The softening of the
concrete has never been observed except very close to the cathode,
the main body of the concrete remaining perfectly sound. Numerous tests show conclusively that the crushing strength of the main
body of the concrete is not reduced even when the potential gradient
is maintained a t 175 volts r foot for over a year. 22. Because of
the cathode effect noted a E v e , the proposal to protect reinforced
concrete buildings by maintaining the reinforcing material cathode
as a by battery or booster would be much more dangerous than no
protection a t all. 23. Aside from slight heating, which is usually
negligible, the only effect which an electric current has on unrein-

ELECTRIC R W W A Y HANDBOOK
forced concrete is to cause a migration of the water soluble elements.
Consequently, in the absence of electrodes, the ultimate effect of
current flow on the physical properties of the concrete is not materially different from that of slow seepage, which also removes the
water soluble elements. Nonreinforced concrete buildings are
therefore immune from trouble due to stray earth currents. They
might, however, be injured by the grounding of power wires within
the structure since these or the inclosing conduits would then
act as electrodes. 24. Conditions arise in practice which give rise
to damage due to stray currents, but the danger from this source
has been greatly overestimated. While precautions are necessary
under certain conditions, there is no cause for serious alarm. 25.
If reinforced concrete could be thoroughly waterproofed, it would
greatly increase its resistance and diminish accordingly the
danger from either the anode or cathode effects. I t should be
emphasized, however, that waterproofing to prevent electrolysis is a
much more difficult matter than waterproofing to maintain a
moderate degree of dryness, because of the much higher degree of
waterproofing required in the former case. I t has been found that
practically all of the wate roofing agents now on the market
that are intended to be mixerwith the concrete areof little value a s
preventives of electrolysis. Waterproofing membranes, etc.,
applied to the surface can be made more effective and when properly
applied may have considerable effect in preventing the entry of
earth currents into the concrete. 26. Painting or otherwise coating
iron with an alkali resisting metal preservative before mbedding it
in concrete may serve to minimize the dangers of electrolysis,
but no such coating has been found that does not prevent the proper
formation of the bond between the concrete and iron when the
concrete sets. 27. In order to insure safety of reinforced concrete
from electrolysis the investigation shows that potential gradients
must be kept much lower in structures exposed to the action of salt
waters, pickling baths, and all solutions which tend to destroy the
passive state of iron. 28. All direct current electric power circuits
within the concrete building should be kept free from grounds.
If the power supply comes from a central station the local circuits
should be periodically disconnected and tested for grounds and
incipient defects in the insulation. In the case of isolated plants
ground detectors should be installed and the system kept free
from grounds at all times. 29. All pipe lines entering concrete
buildings should, if possible, be provided with insulating joints
outside the building. If a pipe line passes through a building
and continues beyond, one or more insulating joints should be
placed on each side of the building. If the potential drop around
the isolated section is large, say, 8 or 10 volts or more, the isolated
portion should be shunted by means of a cop
. 30. Lead
covered cables entering such buildings should r i % l t " e d from the
concrete. Wooden or other nonmetallic supports which prevent
actual contact between the cable and the concrete will give sufficient isolation for this purpose. Such isolation of the lead covered
cable is desirable for the protection of the cable as well as the building. 31. The interconnection of all metal work within a building is

FLECTROLYSIS

TESTS

an advantage where practicable, provided that all pi lines entering

the building are equipped with insulating joints anKead cables are
taken care of as indicated in the preceding paragraph, but the
grounding of such interconnected metal work or any part of it to
ground plates or to pipe lines outside of the insulating joints is to
be strictly avoided. 32. In making a diagnosis of the cause of
damage in any particular case, the fact that a fairly large voltage
reading may be obtained somewhere about the structure should
not be taken as sufficient evidence that the trouble is due to electrolysis. The distance between the points, and particularly the
character of the intervening medium, are of much greater importance than the mere magnitude of the voltage reading. As a precautionary measure, however, all potential readings about a
reinforced concrete structure should be kept as low as practicable.

Electrolysis Tests

Usual Polarities.

With the common arrangement of connecting

the positive terminal of the railway generator to the trolley and the
negative to the rails, the general path of stray currents is from
the rails through earth to pipes or cable sheaths a t places distant
from the power station, through the ipes or cable sheaths and
from these th~oughthe earth back to t i e rails or other return conductors in the vicinity of the power station. Where current tends
to flow from the rail to a buned pipe or cable sheath the rail has
a positive potential with reference to the buried conductor which
is negative. This is generally known as the negative district.
Where current tends to flow from a buried conductor to the rail
the buried conductor has a positive potential with reference to the
rail, and this is usually known as the positive district. The intermediate district where the potentials may shift from positive to
negative is sometimes called the neutral district.
Potential Survey. As the polarity of the earthed conductor is
indicative of the tendency of current to flow, and electrolytic corrosion only takes place where current leaves the current conductor
for earth, the first set of tests is generally a set of potential readings
and is called potential survey. I t is permissible and the usual
practice to use hydrants or service connections for a contact to
underground piping system. These readings usually are made
with a low-reading voltmeter, preferably with the zero indication
in the center of the scale, and readings are taken in a large number
of places throughout the system between the rails and the buried
conductors (pipes or cable sheaths). The voltmeter used should
preferably have a high resistance to minimize the effect of accidental
poor contact. A Weston high resistance zero center voltmeter
with ranges of 1.5, 15 and 150 volts is a very satisfactory instrument. Readings should be taken a t intervals of a few seconds for
several minutes a t each point and notation made of the location
and of the maximum, minimum and average readings. I t is
desirable to plot these potential readings graphically on a map similar
to Fig. 66, where the differences of potential between city water
mains and street railway tracks are shown graphically plotted with
the latter as a base line. I t is also desirable to show on this map

672

ELECTRIC RAILWAY HANDBOOK

the size and location of the various pipe systems. A more detailed
map of this character is shown in Fig. 67 which is an actual potential
survey in the Grand Avenue substation district of the Chicago Railways Company. This figure is reproduced from the Fourth Annual
Re rt of the Board of Supervising Engineers, Chicago Traction.
g t e n t i a l Itendings Mere1J Indicative. I t should be remembered
always that the potential difference between pipes and rails, even
if large, is not conclusive evidence of stray currents, but is only
an indication of the points a t which current may be flowing from
rails to pipes or from pipes to rails or between other conducton.
In fact, a high potential reading is generally an indication of high
earth resistance and consequently a small current flow rather than
of a large current flow.

."
FIG. 6 6 . 4 m p l e potential survey.

E. R. Shepard (Elec. Ry. Jour., 1923) points out that (I) lateral,
as against longitudinal, potential gradients in the earth are directly
responsible for stray current electrolysis. (2) The function of
underground structures in the longitudinal transmission of stray
currents is small and almost negligible in comparison with the part
played by the earth, unless such structures are electrically dralned
to the railway return circuit. (3) Piping systems in general assume
the average potential of the earth In which they are buried and
limited
rtions of such systems should be considered as equipotentiarbodies and not subject to the same potential gradients
that exist in the earth. (4) Potential gradients in the earth are in
general very much smaller in a direction rallel to street railway
tracks than a t right angles to them.
The familiar over-all
voltage curves showing the potentials of tracks, earth and pipes
are misleading, as such potential proses cannot be made to a ply
to the earth as a mass nor to the structures as a network witgout
rigid qualifications. Such profiles are applicable only to line
conductors such as a wire or an electric railway line.
Mr. Shepard says that in some cases water mains have been found
to be collecting current on one side and discharging on the opposite

POTENTIAL SURVEYS

673

674

ELECTRIC RAILWAY IIANDROOK

side, although potential difference measurements show them to be

strongly positive to the track. An example of this kind was obs~rved
in New Orleans in 1 9 2 2 , and is illustrated in Fig. 68. A 6-in, cast
iron main 3 ft. deep and
an electric railway track a t a
distance of about q ft. was 1.65 volts positive to the track. T h e
main was uncovered and the end of the trench squared up. A polar
diagram of current discharge was obtained by taking readings a t
1 5 deg. intervals with the non-polarizing electrodes and the earth
current meter. The main was found to be collecting current ofl one
side and discharging it toward the track on the other side, the
current density a t different points being shown in the diagram in
milliamperes per square foot of pipe surface. This condition 1%
typical of many other observations made on pipes paralleling
electric lines.
I<
. -- - - -------- - - 3'6"- -------- - ------

-1
II

*!
-9

F:
I

FIG.68.-Polar earth current diagram showing transverse electrolyis.

Direction and Relative Magnitude of Current Flow in Underground Conductors. Tests to determine the direction of current flow
in underground conductors may be made by measuring potential
differences between two points in the underground conductor.
A zero center Weston milli-voltmeter with scales of ro and roo
millivolts is a satisfactory instrument for this test and connections
may be made to the cable sheath in two adjacent manholes or on
the piping system between hydrants or service connections roo or
z m ft. apart. These readings which are clearly indicative of the
direction of current flow may be used in the calculation of the
amount of current flow in the case of cable sheaths where
the resistance per foot can be quite accurately known and is not

CURRENT

now

IN PIPES

675

seriously affected by joint resistances. In the case of pipes, however, such readings can only be used as an indication of the relative
magnitude of currents because, especially in the case of the ball
and spigot joint usually used in cast iron pipes, the resistances vary
so greatly that it is not possible to make any accurate assumption
as to the resistance of a considerable length of pipe, including the
joints.
Current Flow in Pipes. I n only one way can an accurate
determination be made of the current flow in underground piping
systems, this method with its modifications being a determination
of the potential drop in a continuous length of pipe between
joints, and the application of Ohm's law, knowing the resistance
of the pipe between the points of contact. Knowing the weight
of the pipe per foot, exclusive of hubs or joints, the resistance per
foot may be calculated by dividing the resistance in ohms per
pound-foot by the weight per foot. The following is a table of
resistances in ohms per pound-foot of various pipe materials:
Cast iron. ................. a.00144 ohm per pound-foot
.
o ooo18r ohm per pound-foot
Wrought ~ r o n ...............
Steel ...................... 0 .wo%K ohm per peund-foot
Lead........................ 0.00048 ohm per pound-foot

A table showing the resistance per foot of various sizes and

weights af Amaimn Water Works kiwiation cast-ima pi
American Gas Institute standard cast-iron pipe, standard wrougKl
iron pipe and standard pipe steel is shown (pp. 677 et seq.) in order
to simplify these calculations. In making these current measurements, it is necessary to have a perfect metallic contact between
the pipe and the milli-voltmeter leads, and it is therefore necessary to expose the pipe where current measurements are to be
made. The best contact is obtained by soldering the leads directly
to the pipe, especially where readings are to be taken over a considerable period of time. I t is sometimes convenient to carry
these rnilli-voltmeter leads as rubber-covered wires through a
conduit to a box a t the curb so that later readings may be taken
conveniently. In such cases it is necessary to have a calibration
of the milli-voltmeter including such leads.
Where measurements are to be made in a p i p , the weight or
resistance of which is not definitely known, and it is required that
the determination of curreht be made with accuracy, a determination of thp resistance and current may be made by one of the
methods described by Dr. Carl Hering (Trans. A.I.E.E., 1912)
ns follows:

I=A

FIG. 69.-Determination of current flow in pipes.

FLECTRIC RAILWAY HANDBOOK

The fundamental principle is as foUows: Let P, Fig. 69, be a
part of an underground pipe which has been uncovered and through
which an unknown current I is flowing as shown; a t first let it be
supposed that this current is steady, and of c o u m a direct current.
Let D be a sensitive galvanometer, milli-voltmeter or any other form
of detector of small differences of potential, connected as shown;
there should preferably be no variable resistance like an unbonded
pipe joint between the two contact points. Let A be an ammeter,
B a few cells of storage battery and R a n adjustable resistance; the
shunt circuit containing them is connected as shown anywhere
outside of the points of a plication of the voltage detector, the
farther away the better-tiey may even be on the other side of a
joint. To find the current flowing in the pipe adjust the resistance
R until D reads zero; then there will no longer be any current flowing in the shunted part of tbe pi e, hence the reading of the
ammeter wiU give the current I la t i e pipe.. The current may be
said to have been sucked out of the pipe by the battery, and made
to flow through the ammeter, where it can be measured; as far as
the current jn that short section is concerned, the pipe circuit has
in effect been electrically cut in two as though an insulating joint
had been introduced. If D is a galvanometer with proportionate
deflections, instead of a mere detector, then by taking a deflection
immediately after tbe shunt circuit has been opened a reading
proportio~ateto the drop of voltage for that curtent will be obtained.
The instrument D is thereby calibrated to read the pipe currents
directly and can be used for this purpose thereafter; the test witb
the battety current is therefore merely of the nature of a preliminary calibration and need be carried out only once for each station.
U in addition this voltage instrument is calibrated to read directly
in volts (usually in terms of miUi- or microvolts as for instance a
milti-voltmeter) then if the deflection reduced to millivolts is
divided hy the current it will give the resistance of the pipe in
milliohms between the two points of application of the voltmeter,
hence i t enables the true resistance of the pipe to be measured.
In this method shown in Fig. 69 there should be no current leaving or entering the pipe between the points of application of the
shunt, hence the ho e should be free from water and there should
be no moist earth in contact with that part of the pipe. I t is
assumed in this method that the short length of pipe between the
shunt contacts is too small a fraction of the wbole circuit of the pipe
line currehts to alter these currents when that section is practically
cut out of circuit, as it is when the current flows through the shunt.
This assum tion is probably absolutely safe in all cases in practice.
As this metgod is not in any way concerned with the nature of the
circuit beyond the single pipe length under test, nor with the rest
of the ath of the current, it can be applied t o the most complex
networE of pipes, even when the pipes are interconnected elsewhere or bonded to the track. I n measuring the pipe resistance
by this method, or in general when the voltage drop is measured,
it is of course assumed that the current is constant while the
two successive readings are taken, henceit isrecommended to repeat
the two readings a number of times.

CAST IRON PIPE RESISTANCE

TABLEFOR D E T E W N A ~ O N
OP COBBENTFLOW
FROM

6 77
ON

PIPMG

MILLNOLTDROPALONG CONTINUOUS
LENGTH
O F PIPE BETWEEN JOINTS
* Dlstance

L
E
K

between contacts m feet

reading m m~ll~volts
Constant from table
Cument Bow in amperes.

e 1-tmment
P

STANDARD
CASTRON PIPE
(Based on Resistance of 0.00144ohm per 1b.-ft.)

14 $N
G

C
E

. . . . . ......
.... .......

zoo

4
4

6
6

N
N

A
E

C r

300
400

lo3
109

400
3 96
416

1 7 2

1 2 0
12 5

86
5 00 44 :
......,
500 404
5
I30

ao o
20 o
213
21 3
22.8

13 9

604

272

18.9

roo

6 02
5 98

27 8

C
I

G
H

800

A ' roo
....
C

304
347

.... .....

qj

...
86'
. ...

99:

690
6 90
6 90

. . 43

700

4 0
3 9%

00

500

173

200

'

14.9
15.7
16.9

. .. . ..
300
.... .130
....

6
8
8
8

412
408
402

:I
...... .:::.
... .......

480

............ 44 88 00
........... 4 8 0
.....
xoo
4 80
A
43
..... ......
G
, 5 0 0
.,............ s o o

18 o
18.9

11.7

13.1

::%

14 8
15.8

16 9
18

29 I

19 3
ao 9

...a

5
8

6 10
6 08
6.03

32 4
31 9

22

34.8

24.9

7 38
7 38

6 08
6 00

42 8
45.2

29.7
31.4

90s
905
905

s a t

815
8.13

355
379
38.7

247
263
269

7 10
7 ro
710

a2

28.0
9 05 8 09
40 3
29 6
g 05 8 03
42 7
307
9 0 5 199
443
* W = Amencan Water Works Assoclat~onStandatd
N = New England Water Works Assoclat~oaStandard
G
Amencan Gas Institute Standard.
f As used by the Amencan Water Works Association and the New Emlaad
Plater Works Asncclat~on.
8
8

W
N

B
E

zoo

'

678

ELECTRIC RAILWAY HANDBOOK

STANDARD
CASTIRON P ~ ~ ~ - ( C o n t ~ n u e d )
(Based on Resistance of o 00144 ohm per Ib -ft )
Classlficat~on

8
8
8

W
N
W

8
8
8

W
W

8
8
I0
10
ro
10
10
1.9

C
G

.N
N

N
N

W
W
W
W
N
N

I2

W
N
N
W
N
W

ra
ra
xz

11
Ia

700
800

304
347

9 60
9 60

8 10

65 o
69 o

45 I
48 o

490
51 o
51 9

340
35 4
36 r

51 9
549

$7 9

36 I
381
4s- 2

11 10 9 96
1140 1020
11 40 10 16

58 9
636
65 5

40 9
441
45 5

11

40 1 0 1 4
1140 1006
XI401004

665
705
715

462
490
497

N
N

N
W

IOO

43

86
130

400

....
173

500
600

2x7
260

I1 40 10 00
11 60 10 12
X I 60 10 w

73 5
78 7
84 6

51 I
54 6
58 8

700
800

304
347

11 84 ro 12
11 84 ro w

92 4
98 5

64 I
68 4

r3 20 ra aa
x3 zo 11 14
x j zo 12 12

61 I
65 9
67 o

41 J
45 7
46 5

43

13 20 11 12
13 20 12 06
13 20 I I 98

67 o
70 6
75 3

46 5
49 o
51 3

IW

B
E
C

P
D
H
E
F

10 10
1004
p@

zoo

B
....
A
C

00

300

I I 10
X I I O
.'I 2 0

r.
E
F
G

1110 1016
X I 10 10 12
r I 10 ro 10

12

37 a
394
42 t

12

53 6
567
60 6

xa

8 04
810
8 w

N
G

XO

Ia
ra
ra

9 30
942
9 4a

I0
I0

12
12

217
260

B
E
C
P

1a

500
600

W
W

10
10

33 3
34 5
35 5

E
F

10

47 9
49 6
51 2

400

10
10

10

8 18
8 14
8 I0

I30

10

10

173

9 30
9 30
9 30

300

aoo

86

300

I30

13 ao 1 1 96
x j 50 12 20
13 50 I2 14

76 4
81 9
85 5

53 o
56 8
59 4

900

173

1350 1 2 1 2
x j 50 ra 04
13 50 Ia w

866
91 5
93 8

60a
63 6
65 I

500
600

217
260

1350 1196
13 78 ra 14
1378Iaoo

962
104 o

668
72 3
779

See footnotes, page 677.

1x20

CAST IRON PIPE RESISTANCE

679

STANDARD
CAST IRON P ~ ~ ~ - ( C o f t l i n u e d )
(Based on Resistance of o 00144ohm per 1b.-ft )
Actual
K-current
d~rnenstons We1 h t for one
per ft
rnrlllvo~t

Class~ficatlon

*Assoeter,
Inches

dard

I1
la

W
W

I4
I4
14

N
N
W

14
14
14

N
N

-G

E
C
P

14

53 4
57 I
57 1

aoo

86

15 30 14 08
1530 1398
15 30 13 98

87 9
948
94 8

61 0
658
65 8

300

I30

400 J

173

15 65 14 17
15 65 14 15
15 65 I4 07
15 65 14 or
'r.3 99

'rs 65

'

71 4
7s 0
76 a

103
I08
109
115
119

rzr

'

80 o
81 8
83 9

E
P

so0
600

a17
a60

IS 98 14 18
15 98 I4 00

133
145

93 4
I01

G
H

700
800

304
347

16 31 14 18
16 32 14 00

160
172

I11
130.

roo

43

16
16
16

G
N
N

16
16
16

W
N
N
W
N
W

G
D

18
18
18

75 8
82 3
82 3

16
16
16

16
16

1s 30 14 24
15 30 14 16
15 30 I4 16

W
N
N
W

16
16
16

43

86 7
91 4

125
133

I S 65 14 as

16
16
16

100

D
B

14
14

14 08 I 1 14
1408 1100

14
14

304
347

14
14
14

700
800

. 17
40
17 40

16 30
16 a0
17 40 16 ao

w 9
98 9
98 9

68 6

17 40 16 16
17 40 16 I0
17 40 16 oo

102.
107
115

70 7
74 1
79 6

B
E
F

100

86

17 40 16 00
17 80 1 6 3 0
17801620

115
115
133

79 6
87 x
926

300

130

400

173

17 80 16 20
17 80 16 10
17 80 16 oa

133
141
147

91 6
98 1
101 3

W
W
W

E
F

500
600

217
a60

17 80 16 00
18 16 16 zo
18 16 16 oo

149
165
181

I03 S
114 S
I15 I

G
H

700
800

304
347

18 54 16 18
18 54 16 oo

201
11.5

139 5
149 0

A
43

19 a5 18 r r
I9 15 17 99
19 50 18 az

104
115
118

70 5
79 8
82 a

86

19 50 18 11 127
19 50 18 00 138
19 50 18 00 138

88 s
95 8
95 8

A
W
roo
18
C
N
18
D
N
B
18
W
200
See footnotes, page 677.

ELECTRIC RAILWAY HANDBOOK

STANDARD
CASTIRONPIPE-(C&~?LUC~)
(Based on Resistance of 0.00144 ohm per 1b.-ft.)
Actual

Class~fication

dimenslorn

Out-

Weight
per ft.
e-xclu-

Inside slve of

difz;hub-Ib.

inches
18
18
18

N
N
W

18
18
18

w
W

18
18

W
W

ao
20
20

10
10
20

ao
20
ao

ao
10
20

20
20
20

l%'

24
14
24

N
N

w
W

N
G

14
14
24

14
14
a4

14
14

df%T

con.tinuour
PI-

amperes

300

130

18.10
19.70 17.98
19.91 18.18

148
159.
161.

103.
110.4
113.

D
E
F

400
500
600

173
117
260

19.92 18.00
20.34 18.20
20.34 18.00

178.
103.
220.

123.8
140.5
152.6

G
H

700
800

304
347

20.78
20.78

18.22
18.00

14%
264.

170.
183.3

122.

11.60

20.10
19.98
20.26

134.
137.

84.6
93
95.4

...... ...... ........ 21 60

...... ........ 21.60
C
D
...... ........ 21.60

20.24
20.16
10.02

140.
I47
161.

97.0
101.5

P
C

B
A

B
E

F
C
D
E

G
H
A

...... ........

. ..... . . .. . . . . 21.30
. . ... . . ...... . 21.30
loo

43

1x1.

zoo

86

21.60 2o.00
21.90 20.20
11.90 20.06

163.
175.
189.

1x3.
1.72.
131.

300
400
500

130
173
1x7

11.06 ao.22
20.00
22.06
22.5420.14

191.
241.

131.
148.
167.

600
700
8W

160
304
347

21.54
23.02
23.02

20.00
20.24
20.00

265.
295.
319.

184.
205.
221.

. . . . . . . . . . . .. 25 40 24.12
23.96
24.28

156.
174.
187.

108.
121.
130.

25.80

24.28
24.20
24.04

187.
196.
2x5.

130.
136.
149.

25.80

24.01
24.20
24.04

217.
234.
253.

ISX.
163.
176.

. .... . .. . . ... .
. .. . . . ... . . ...

.. . . . .

25.40
.... .. ...... .... . ........ .. 25.80

A
C
D

43
...... ........
25.80
...... ........ 15.80

B
F

... ... .. . . . . . . 26. lo

loo

zoo

86

112.

300
400

130
173

26.32 24.14
16.31 24.00

158.
286.

179.
198.

E
F

soo
600

217
a60

26.90 24.28
26.90 24.00

318.
361.

128.
251.

215.
145.
257.

149.
170.
179.

N
W

it
30
30
30

...... ........ 19 70

K-current
for one
millivolt

N
G

...... ........ 16.10

......
A
........ 31.60
...... ........ 31.60
B
...... ...... ........ 31.74

See footnotes, page 677.

30.18
29.98
30.04

CAST IRON PIPE RESISTANCE

(Based on Resistance of 0.00144 ohm per Ib.-ft.)

Actual

Classification

dimensions

Weight

K-current
for one

Nomi-

amperes

30
30
30
30
30
30

N
W

C
D

3 1 . 7 4 29 98

266.
277.
306.

185.
192.
213.

86

32.00 29.94
30.20
30.00

312.
337.
367.

217.
234.
255.

130

loo

43

...... ........ 3 2 . 0 0 30 18
..... ~ 1 . 0 0 29.98

. . . ... . ..
zoo

...... ........ 32 40
...... ........ 32.40

N
N

30
30

W
W

C
D

390
400

32.40 30.00
32.14 30.00

367.
422.

255.
292.

30
30

W
W

E
F

500
600

217
a60

33.10 30.00
33.4630.00

479.
537.

333.
373.

35
36
36

36.20
36.00
36.06

287.
326.
345.

IW.
026.
239.

43

37.9635.98
36.26
30.04

358
373.
412.

248.
259.
286.

86.

383036.00
36.20
38.70 35.98

418.
459.
497.

ago.
319.
346.

36
36
36

36
36
36

36
36
36
36
qa
42
42

N
N

W
W
W
N

A
. .. . .. ..... ... 37.80
B
. .. . .. ... . . ... 37.80
...... ...... ........ 37.96
A
C

zoo

300

I3

F
A

42
40
42

B
E
F

42
42

48
48
48
48
48
48

N
N

.....

........ 38.70
130

...... .......
qoo

173

38.70 35.96
39.16 36.00

502.
581.

349.
404.

500
600

217
a60

39.6036.00
40.0436.00

666.
753.

463.
523.

42.26
42.00
42.06

368.
422.
452.

256.
293.
314.

42.00

@.5041.96

465.
480.
538.

323.
333.
374.

86

44.50 4 1 . 9 4
. 4S.10 41.30
. 45. ro 42.04

542.
600.
654

376.
416.
454.

130
173

4 5 . 1 0 41.02
qs.584a.oz

657.
763.

456.
530.

48.30
48.00
48.30

459.
529.
608.

319.
367.
@a.

47.98
47.98
48.00

608.
608.
678.

@a.
422.
471.

. . . . . . .. 4 4 . 0 0
...... . . . . . . . . 4 4 . 00
......

...... ...... ........ @ . l o

42
41
42

N
N

loo

. ... . . .. ...... 38.30

.. . . . . . . . . . . . . 3 8 . 3 0

A
C

loo

43

44.20

..... . . . . . . . u
50 . 4 2 . 2 4

..............
200

'

. .. . . . . .. . ...
. . . . . . . . . . ...
300
400

...... ........ 50.20

...... ........ 50.20
. . . . . . . . . . . . . . 50.80
...... ...... ........ 50.50
A
roo
50.50
43
D . . . . . . . . . . . . . . 50.80
A

B
C

W
N
See footnotes, page 677.

ELECTRIC RAILWAY HANDBOOK

682

STANDARD
CASTIRONP I P E - ( C M ~ G I ? ~ ~ ~ )
(Based on Resistance of 0.00144 ohm per 1b.-ft.)
Actual
dimensions

Classification
Nomi *Assonal
diam- ciation
staneter.
inches dard

W
N

W
W

N
N
W
N
N

B
E
F
C
D
A

C
D
B
E
F

C
D

N
N

W
N
W

,"A&
yb
*

aoo

86

50.80 47.96
48.30
48.00

686.
757.
828.

477.
526.
575.

300
400

130
173

51.40 47.98
51.9848.06

832.
961.

578.
667.

550.
650.
731.

388.
452.
508.

750.
840.
845.

521.
583.
586.

........
......
...... ........ 51.40
jI.40
...... ........ 56.40
...... ........ 56. o
loo

43

54.34
54.00
56.8653.96

......
........ 57.10
...... ........
zoo

86

54.36
57.10 54.02
57.10 54.00

.......
..... ................ 57.80
57.80
300
400

130
I73

54.26 946.
54.00 1041.

657.
723.

57.80 54.00 1041.

58.40 53.94 1230.

723.
854.

...... ........ 62.60

62.60

60.40
60.00
60.02

664.
782.
836.

460.
543.
581.

C
B
D
E
C

...... ........
63.40
..zoo
... . ...86
.. ... 63.40
63.40

60.40 910.
60.06 10x0.
60.00 1028.

632.
701.
714.

64.20 60.40 1160.

...... ........ 64.2060.201220.

806.
848.

...... ........64.10
64.82

W
W

loo

300

43

130

889.

400

173

60.00 1280.
60.06 14.55.

roo
aoo
300

43
86
130

75.34 72.08 1178.

76.00 72.10 1415.
76.88 7a.10 1745.

819.
983.
xarl.

43
86

87..54 84.10 144.5.

88.54 84.10 1878.

1005.
1304.

W
.A 100
W
zoo
B
See tootnotes, page 677.

84

84
-

inches

... . .. ..... .. . 62.80

;:%ls

2 : 7::.
~ psq. m.~ 22.
~:
F;
;
-lb.~

N
W

Out-

wei ht K-current
for one
per 't.
mill~valt
drop
per
Inside e $ ~ foot of
con~inuou,
h-f;

10x0.

STEEL PIPE RESISTANCE

683

STANDARDSTEEL(OR WROUGHT
IRON)PIPE
(Based on resistance of steel--o.oooa~ohm per 1b.-It. Based on
resistance of wrought iron--o.ooo18r ohm per 1b.-ft.)

Nominal
diameterinches

Outside
diameterinches

..........

..........
H..........
$4 ..........
16..........
H..........
.....
.....
$4..........

# ..........

f..........
i
% ..........
I
1

...........
...........
...........

IW .........
134

.........

134 .........

.........

134

I M.........

. p.

laeight

loot
e d -

*Classification

0.405
0.405

K-current for
one millivolt
drop per foot of
continuous
pi-amp

1 I1 1 lwFough.
t

Actual dimensions

Inside
diameterinches

Iron

'

0.269
0.215

0.244
0.314

1.16
1.50

1.31
1.70

0.540
0.540

0.364
0.301

0.424
0.535

2.02
2.55

2.90

S
X
S
X
XX

0.675
0.675

0.493
0.423

0.567
0.738

3.51

3.07
4.00

0.840
0.840
0.840

o.62a
0.546
0.252

0.850
1.1.71

4.05
5.18
8.16

4.60
5.88
9.28

1.050
1.050
1.050

0.824
0.742
0.434

1.13
1.47
2.44

5.38
7.03
11.6

6.11
7.98
13.2

1.315
1.315
1.31s

1.049
0.957
0.599

1.68
2.17
3.66

7.99
10.3
17.4

9.09
11.8
19.8

1.660
1.660
1.660

1.380
1.278
0.896

2.27
3.00
5.21

10.8
14.3
24.8

12.3
16.2
28.2

1.900
1.900
1.900

1.610
1.500
1.100

2.72
3.63
6.41

12.9
17.3
30.5

14.7
19.6
34.1

2.375
2.375
2.375

1.067
1.939
1.503

3.65
5.oa
9.03

17.4
23.9
43.0

19.8
27.1
48.8

2.875
2.875
2.875

2.469
2.323
1.771

5.79
7.66
13.69

27.6
36.5
65.2

31.4
41.5
74.2

7.57
10.2
18.6

36.0
48.8
88.5

41.0
55.6
tor

XX
S

X
XX
S
X
XX

2.0

2.30

a
a ...........
a

X
XX
S
X
XX

2H .........
2% ........

X
XX

3 ...........
3 ...........
-3

X
XX

3.500
3.500
3.500

3.068
2.900
2.300

4.000

4.000
4.000

3.548
3.364
2.728

12.5
22.8

43.4
59.6
Iog

49.3
67.8
I24

4.500
4.500
4.500

4.026
3.826
3.152

10.8
15.0
27.5

51.4
71.3
131 .

58.4
81.1
149

154 .........

...........

...........
2% .........

3H .........
351.........
354 .........

..........

4
4 ..........
4 ..........

'S

X
XX

.
.
.

X
XX
S

X
XX

Standard pipe
Extra strong pipe .
Double extra strong pi pc

.
.

ELECTRIC RAILWAY HANDBOOK

684

(Based on resistance of steel-e.oooar ohm per Ib.-ft. Based OR

resistance of w r o u g h t i r o n - o . o o o r 8 1 ohm per )b.-ft.)

inches

4 % .........

5.000
5.000
5.000

4.~06
4.290
3.580

12.5
17.6
32.5

59.8
8
155 .

67.9
95.3
176

5.563
5.563
5.563

5.047
4.813
4. 063

14.6
20.8
38.5

69.7
98.9
183.

79.2
111
209

6.625
6.625
6.625
7.625
7.625
7.635

6.065
s.761
4.897
7.023
6.625 .
5.875

19.0
28.6
53.1
23.5
38.0
63.1

90.3
136.
253 .
I12 .
181
300

155 .
288 .
I27
206
341

8.625
8.625

8.071
7.981

14.7
28.5

8.695
8.625

7.625
6.875

43.4
72.4

9.625
9.625

8.941

8.625

33.9
48.7

S
S

IO.?SO
10.750

10.19~
10.136

31.2
34.2

..
a06 .
'235.
391.
345 .
161 .
184 .
264 .
232 .
149 .
169.
163.
185.

10.750
10.750

10.020
9.750

40.5
54.7

192.
261

11.750
11.750

zr.000
10.750

45.6
60.1

?rt.
a86 .

.
247 .
326.

19

..........
..........
............

S
s
X

12.750
r2.750
12. 750

12.090
12.ooo
11.750

49.6
65. 4

to8 .
236 .
311 .

354

............

S
X

14.000
14.ooo

13.250
13.000

54.6
72.1

256 .
39r

............
............

260 .
343 .

S
X

rg.ooo
1 s .000

14.250
14.000

58.6
71.4

279 .
369.

317
420

1 6 . ~ 0

rs.as0
15000

61.6
82.8

4>5

.........

X
XX

4H .........
5 ...........

...........

'

S ..........

xx

6 ...........
6 ...........
6 ...........

...........
7 ...........
7 ...........

XX

...........

8
8

...........

X
XX
S

........... x
........... X X
Q ...........I
S
9 ...........
X
8

..........

lo
I0 .......,..

..........
..........
11 ..........
rr ..........
I0
10

1'2

12..........

............
...........

.
.

- 16.~00

S = Standard pipe
X
Extra strong pipc.
Double extra sttong p i p

XX

43.8

.
.
,nR.
136.

..
I03 .

..

134
15s

2x9
297

;g:.

..
298 .
339 .
394 .
449 .

LEAD CABLE SHEATH RESISTANCE

685

C. W. Kinney, Elalric J o u r o l , 1909, describes an approximate method of determining the amount of current flow in a
pipe, the resistance of which was not known. The determination was made as follows: The water pipe was exposed for about
6 ft. of its length. Three places on the pipe, 2 % ft. apart, were
filed bright and the drop in voltage, as indicated by the voltmeter,
between the 6.sf and seco~d&ts
and then between the second
and third points, was noted. After these readings had been taken,
an ammeter reading was taken between the rail and the middle
point on the water pipe. While the ammeter was still connected,
voltmeter readings were taken as before. The voltage drop between
points I and 2 showed less than one-half of I per cent of the original
value. Thus, the voltage drop between p i n t s 2 and 3 could be
assumed to be due entireIy to the current Bowing through the
ammeter. The normal flow of current is the pipe was then calculated by a simple proportion; that is, the current flow during the
first test bore the same relation to the ammeter reading that the
voltmeter deflection between paints 2 and 3 during the first test
bore to the deflection of the voltmeter needle when the ammeter
was connected, since the voltmeter was of uniform scale type. I t
was, therefore, not necessary to determine the actual value of the
divisions on the voltmeter scale in order to find the amount of
current flowing in the ipe.
Rcsisk~ceof and 8 r m t in h a d and L u d Alloy Pipes urd
Sheaths. The resistance of "lead" pipe and "lead" cable sheath
will depend upon the alloy that b used in its construction. Practical results indicate that the resistance of pure lead is about 480
microhms (0.00048 ohm) per pound-foot, and the resistance of
"lead" cable sheathing containkg 3 per cent tin is about 530
microhms (0.00053 ohms) per pound-foot. The curve sheet,
Fig. 70, is arranged for the rapid approximation of resistance and
currentger millivolt drop per foot for lead and lead-tin alloy pipa.
The fo owing example wdl illustrate the method of using these
curves :
What is the resistance per foot of sheath and the current per
millivolt drop per foot along a cable sheath of a lead-tin alloy
containing 3 per cent of tin, the resistance of the sheath being
0.00053 ohms per wynd-foot? The outside diameter of the sheath
is 2.28 In. and the thckness of the wall is %, m.
Starting a t the " 2.28" point on the "Outside Diameter-Inches"
scale (Fig.
70), follow the vertical to intersection with the oblique
0.156 ( = decimal equivalent of ma) "Thickness-Inches"
line.
From this point foIlow the horizontal to the intersection with the
"Weight per foot-Lead-3
per cent Tin-pound"
a t the
5.07 Ib. point. Now starting with the 5.07 lb. p i n t on the Weight
per foot, pure lead, pounds," follow the horizontal to the left to
the intersection with the 0.0053 "ohm per pound-foot" oblique.
From this point drop a perpendicular. The intersections of this
perpendicular with the horizontal scale indicate that the resistance
of the sheath is about 104 microhms per foot and the current is
about 9.6 amperes per millivolt drop per foot along the sheath.
NOTE:
If the cable sheath or pipe is of pure lead, the horizontal

seal:

686

ELECTRIC RAEWAY HANDBOOK

NON POLARIIZNG ELECTRODES

687

sbould be canied straight through, instead of making the correction on the vertical scale. Such a case is shown by the arrows
starting at 0.5 in. outside diameter, Fig. 70.
Determination of Amount and Distribution of Current Leaving
Undereround Metallic Condictog. T h e generally acu, women AQW
c e p t e d method consists in
determining the amount of current Bow (by o n e of t h e
methods outlined above) in
two locations on the same
61ridmg
line of underground pipe. The
difference in the amount of ZIIX Rod
old5foppr
current on the pipe between
these two locations will then
be the amount of current leaving (or entering) the pipe between the two locations.
Normal Electrode. T h e
Haber normal electrode. also
called non-polarizing electrode, G1ass
consists of a rod of zinc which
is envelowd in a wet paste of
zinc suldhate contain& in a
dass tube which h a s h a d
cemented to it a t the bottom FIG. 71.-Haber n o r m a 1 electrode.
Cross sectlon.
a porous clay cell. The other
cnd of the tube is closed with a stopper from which the zinc rod is
supported; an insulated wire is led from the end of the zinc rod
through this stopper to the upper end of a wooden rod which also
enters the stopper and serves for the
e&rLmds purpose of handling the electrode.
A capillary tube is also run through
&n.
the sto per in order to have the interFmme
ior of tge tube a t normal atmospheric
pressure. The zinc sulphate paste is
e made by adding saturated zinc sulphate solution to fine zinc sulphate
crystals until the mixture has attained
a semi-fluid condition. A sketch
f
showing the construction of this
Earth
device is shown by Fig. 7r.
E a r t h Current Collector. T h e
.%rdRubber Haber earth current collector con%den
hame
Frame
sists of two thin copper sheets laid
PIG. 7 2 -Haber earth current one upon the other with a thin sheet
collector. cross
of mica or other non-absorbent
insulating material between them.
These two plates are gripped in a hard rubber rim which forms
part of a square wooden frame. A paste made by mixinapowdered copper sulphate crystals with a 2 0 per cent aqueous solution
of sulphuric acld is spread over the exterior surfaces of each

?,fibber

%%

688

ELECTRIC RAILWAY HANDBOOK

of the two sheets of copper, the paste being enclosed on each

exterior surface by a covering of parchment paper or of some similar
tough permeable membrane. Insulated wire leads of suitable length
are run from each plate through the frame to connect with the
measuring instrument. The opening in the frame may conveniently
be square. Four inches is a convenient dimension for the sides of
this square opening as this will yield a n area of one-ninth of B square
foot which IS approximately equivalent to a square decimeter.
The detailed construction is shown in Fig. 7 2 . When using the
instrument, the spaces between the parchment paper and the
outer edges of the wooden frame are first filled with closely packed
soil taken from the spot where i t is intended to make the measurement, and the frame is then placed in a position perpendicular to
the flow of current which i t is desired to measure and conlpletely
buried in earth removed in the course of making the excavation to
reach the structure whose condition is to be determined. A suitable low resistance miiliammeter can then be connected to the two
terminal wires and observations of the current flow made.

FIG. 73.-Princ1pIe and clrcults of earth ammeter.

Earth Ammeter. This instrument, designed for the measurement

of earth currents, was developed by the Bureau of Standards and
described in the Elec. Ry. Jmrr., 1921. I t s purpose is to afford a
means for the accurate determination of the polarity of pipes with
respect to earth and for the quantitative measurement of current
density a t any desired point in the earth. If a measurement be
made of the resistivity of the earth a t any particular point, and if
then a measurement be made of the voltage drop between two
points a known distance apart, within the same region in which

EARTH AMMETER

fm-

the resistivity has been measured, these two measurements wiII

permit a calculation of the current density in the earth in the region
immediately under investigation. The method described involves
something of the principle here stated, although in its actual carrying out neither the resistivity of the earth nor the true potential
drop between two points is determined. The principle of this
method of measuring earth currents can best be understood by
reference to Fig. 73, which is a diagrammatic illustration of the
elements of the apparatus. Let us assume that the pipe is discharging current m all directions as indicated by the arrows z.
Four electrodes 3, 4, 5 and 6, may be imbedded in the earth immediately adjoining the pipe, on whatever side the current intensity is
to be measured, or placed against the wall of an excavation made
near the pipe. An excavation is here assumed tentatively to simplify the explanation of the principle of the method. I t later will
be shown how the method can be applied without making excavations of any kind. For convenience theseseveral electrodes may be
mounted on a single insulating frame 7. Two of these electrodes,
for example 3 and 6, may be connected to a suitable voltage indicator
10, which need not read in any particular units.
Sup se, now,
a current Z be caused to flow between the terminals 4 a n c through
the earth from the battery 8, which current will be measured by
the ammeter 9. It will be evident that this current distributes
itself in all directions through the earth and produces a certain
voltage drop between the terminals 3 and 6 due to the resistance in
the earth immediately surrounding the group of electrodes. This
voltage drop between the terminals 3 and 6 mill be indicated by
the voltmeter 10 and will be proportional to the current Bowing
between the terminals 4 and 5 and to the resistivity of the surrounding earth. If EOis the voltage between the terminals 3 and 6 and
if 00 is the corresponding deflection of the voltage indicator 10
we have
80 = KEo
where K is the constant of the voltage indicator 10 which includes
the effect due to the resistance of the leads and the electrodes 3
and 6. Further it will be seen that Eo is proportional to the current
I sent between the electrodes 4 and 5 and to the resistivity r of the
surrounding earth, or
Eo = AZr
where A is a constant depending upon the geometrical arrangement
of the group of electrodes. Substituting this value of EOin the
first equation, we have
00 = K A I r
Here it is assumed that the voltage drop across the terminals 3 and 6
is due solely to the current sent through the terminals 4 and 5.
In order that this may be true, conditions must be such that no
other current Bowing through the earth a t the time the measurement is made will in any way affect the apparatus. For the present,
we will assume that this is actually the case; it will be explained
later how it is realized in practice. After the above measurement of
I and the corresponding 00, the circuit of the battery 8 is opened,

690

ELECTRIC RAILWAY HANDBOOK

after which the voltage drop E, between the voltage terminals 3

and 6 would be due solely to the current i which is flowing through
the earth, or
El = irL
where L is the distance between the terminals 3 and 6, i is the mean
current density in the region between the terminals 3 and 6 and r,
as above, is the resistivity of the earth. The corresponding deflection of the instrument ro is 01 and we will have

h = KE, = RirL
From these equations we have
eo - K A I r - AZo
01
KirL
iL
Solving for i, we have
. A10
$=1
LO0
As stated above, A is a constant depending upon the geometrical
form of the electrode group 3, 4, 5 and 6. This can be determined
once for all for a given electrode group by immersing the electrode
in a medium such as water through which a current density of
known value is sent. Under these circumstances, if we perform
the two measurements indicated above and substitute the values in
the above equation, i being in this case known, we can once for all calculate the value of A , and as soon as the distance L between the two
A
electrodes 3 and 6 is known, the proportional factor - becomes

Calling this factor R for brevity, we have

. RIB
*=1
00 ,
In this equation, i is the current per unit area, or the quantity which
is to be measured, and R is the known constant. T o obtain the
value of i, we have therefore to perform the two operations mentioned above, namely, to send a known current I through the two
electrodes 4 and 5 and a t the same time measure the corresponding
deflection 00 of the instrument 10, this being done in a manner
described below, such that the instrument ro will not be affected
by any earth current other than that which flows from the battery
8 through the terminals 4 and 5. We then disconnect the battery
8 and measure the deflection 01 of the instrument 10 due solely to
the earth current i. These three values, 00, I , and 01 are then substituted in the above equation and the value of i calculated.
As stated above, the indication of the voltage indicator ro is a
function of the resistance in series with its leads, and therefore of the
resistance of the electrodes 3 and 6 and of the earth immediately
surrounding them. I n practice i t is found that this resistance is
often very high and quite variable, so that the instrument 10 does
not in general give a true value of the voltage impressed in the
earth between the two electrodes 3 and 6, and often not even a n
approximation to the true value. It will be observed, however,
known.

EARTH AMMETER
that the resistivity r of the earth in the region in which the test
is being made and the constant K of the voltage indicator ro disappear from the equation from which the earth current i is calculated. I t will be seen, therefore, that in making this measurement, neither the resistivity of the earth, nor the true value of the
voltage drop between the electrodes 3 and 6 need be known.
This constitutes one of the important advantages of the method of
procedure above described.
As stated, in carrying out the first of the two operations above
described, it is essential that some arrangement be provided
whereby the deflection 00 will be due only to the current I which
flows through the terminals 4 and 5 and will not be influenced by
any earth current already flowing. This can be accomplished in a
very simple manner, by an arrangement shown in Fig. 73, which
shows also a complete wiring diagram of the test set. In this
arrangement, two commutators, 11 and 12, mounted on the same
shaft, are employed. These commutators are so mounted on the
shaft that rommutation takes place on both a t exactly the same
instant, and are provided with a crank whereby they may be
rotated by hand a t a suitable speed. The commutator 11 is interposed between the battery 8 and the test terminals 4 and 5 , while
the commutator 1 2 is interposed between the terminals 3 and 6 and
the voltage indicator 10. I t will be seen that an alternating current
flows through the earth from the terminals 4 and 5 and impressesan
alternating voltage on the terminals 3 and 6 which are being
commutated simultaneously with the current through the leads
4 and 5, which gives rise to a unidirectional voltage on the voltage
indicator 10. This instrument being of the direct current type will
therefore give a deflection 00 proportional to the current I sent
through the terminals 4 and 5. At the same time, any unidirectional
voltage impressed on the terminals 3 and 6 due to an earth current
will be commutated so frequently that it will exercise no appreciable effect on the voltage indicator, and hence the reading of
the latter will be just the same as if for the time being the earth
current to be measured did not exist. After the measurement of
the current 10and the deflection 00 is made under these conditions,
a double-throw switch 13 is reversed, which, as will be seen from
Fig. 73, disconnects the battery 8 from the terminals 4 and 5 and
a t the same time eliminates the commutator 1 2 from the circuit
between the eIectrodes 3 and 6 and the voltage indicator 10.
In the new position of the switch the voltage between the electrodes
3 and 6 due to the earth current i will produce a corres
deflection in the voltage indicator 10 which is then read as tE:%
01. These three values 130, I and el are then substituted in the equation, and the value of the earth current i is calculated in any
desired units, depending upon the value of the constant R used.
The electrode group 3, 4, 5 and 6, mounted on the insulating
support 7, may be permanently buried in the earth in the region in
which it is desired to measure the earth current a t any time, or
the four electrodes may he placed against the wall of an excavation,
so that all four terminals make contact with the earth, while a
measurement of current intensity in the earth adjoining the wall of

692

ELECTRIC RAILWAY HANDBOOK

the excavation is being made. The constant R of the instrument

will, however, be different in the two cases, but can be determined
once for all for the two types of measurements. I n most cases,
however, where ~t is desired to measure the current density discharged from a pipe a t any given point, it is unnecessary to make
an excavation. For measurements of this kind, a special type of
four-terminal electrode has been designed which can be placed down
in a hole extending from the surface of the earth to the pipe, as
shown in Fig. 74. This hole may be made by means of an auger,
or by simply driving a pipe or rod of suitable size into theearth,
and then removing the
rod from the hole. The
four electrodes 3, 4, 5
and6 are then put down
in this hole and the
measurement is made in
exactly the same manRubbrr
-... mn$uluh
1;!1
: r m m dak drdl rad
ner as described above.
The electrodes 4 and 5
through which the test
current is sent from the
battery in the first part
of the test can be made
from a n y o r d i n a r y
metal, such as iron and
copper. The electrodes
3 and 6, h o w e v e r ,
should be made on the
well-known principle of
the nonpolarizable electrode, that i s , t h e y
should comprise a cup
having an electrode a t
@.
~...
k c / , t ec h
py
the base of copper, the
cup being fill<d- with a
FIG. '14.-Electrodes for earth ammeter.
concentrated solution of
copper sulphate. This electrolyte is confined in the cup by a
stopper of wood or other porous material. Fig. 74 shows a typical
form.
It is well to have clearly in mind just what current is measured by
this apparatus. It is the mean current per unit area, as for instance
per square centimeter or per square inch, in the earth in a small
region immediately surrounding the center of the four-electrode
group. A good approximation will be had by stating that it gives
the mean current density throughout the volume of a s here having
a diameter approximately equal to the distance petween the
two potential electrodes of the four-electrode group. I t will thus
be seen that by using electrodes of very small sue the current
density in a very small volume of earth can be studied. By the
use of electrodes an inch or an inch and a half apart, the mean
current density in a sphere as small as three or four inches in
diameter can be definitely determined. For most ordinary pur-

TRACK LEAKAGE RESISTANCE

poses, however, it is found desirable to use a two- or three-inch
spacing of the electrodes, in which case we secure the mean
current density in a volume of perhaps half a cubic foot of earth
immediately surrounding the center of the electrode group.
The voltage indicator 10 used in this test set must be of very
special design to have an extremely high current sensitivity. The
instrument used gives a full scale deflection for one microampere,
and was designed and built especially for this apparatus by the
Rawson Electrical Instrument Company of Cambridge, Mass.,
which company is now manufactunng the complete sets. The
instrument has been in use for some time a t the Bureau of Standards, where it has been subjected to careful tests and experimental
work, and has been found to be a very convenient, economical and
accurate means of m e a s u ~
- athe current intensity discharged
from buried pipes.
Leakaee Resistance between Railwav Tracks and Earth. The
determiZion of the average resistance of the leakage path between
railway tracks and surrounding earth is often very desirable, particularly where it is necessary to determine what over-all potential
drops may safely be permitted in the track return. I t will be evident
that if the resistance of the leakage paths is very high it will be safe
to allow higher potential drops on the track than if the leakage
resistance be low, although the voltage drop which may be considered safe is not directly proportional to the average resistance
of the leakage path. The Bureau of Standards has made a
considerable number of tests on various types of roadbeds, during
the course of which two methods have been found satisfactory.
One of these methods consists of inserting insulating joints in tbe
track at, two points xooo ft. or more apart and bonding around
these with a heavy bond designed to be conveniently opened a t any
time for testing. The test is made a t night when no tra5c is on
the line, the shunt around the insulating joint be~ngopened; a
low-voltage battery then is applied between the isolated section of
track and a suitable earthed terminal giving a very low resistance to
ground. For this purpose the railway track on either side of the
isolated section can be used. These tracks have substantially the
potential of a point on the earth quite remote from the track
section under test. When this connection is made the current
flowing from the battery to the isolated section of track must
practically all pass off through the track roadbed in this section, the
eakage around the insulating joints being very small compared to the
total leakage through the roadbed of the section under test; and
thii current is measured simultaneously with the potential difference
between the track and a second earth terminal which should
preferably be remote both from the isolated track section under
test and from the earth terminal which is carrying the current of the
battery. The resistance of the leakage path between the isolated
track section and ground is then calculated from the ammeter and
voltmeter readings.
The second method of testing which has given satisfactory results
in some cases eliminates the necessity of inserting insulating joints
in the track, a procedure which is often quite d~fficult,especially

694

ELECTRIC RAILWAY HANDBOOK

where cross bonds or space bars are frequently used. In this

method two batteries are required and the arrangement is shown
in Fig. 75. The batteries are stationed from one to several thousand feet apart and cannected as in the preceding test, one terminal
being connected to track and the other to an earthed terminal some
distance away. I t is desirable to connect the positive terminal
of both batteries to the track and the negative terminal to earth,
since tbis represents thepolarity existing in practice where a current
is leaking from the trac into the earth. From an examination of
the figure it will be seen that a great deal of current flowing from
each battery will flow off in the directions away from the section

PIG.7s.-Diagram of connections between a double track railway ljne

and adjoining pipe system. for measuring leakage resistance of a sectlon
of roadbed.

under test, as indicated by the arrows, but a certain amount,

corresponding to the total leakage of the current on the section
between the batteries, will Bow into this section. If now we measure the millivolt drop on short lengths of the rails a t the points A
and B, just inside the points a t which the batteries are connected
to the tracks, we can calculate from this millivolt drop on a
measured length of rail of known weight the approximate current
which is flowing into the section under test from each end. The
sum of these two currents will then be the total leakage current
from the test section, At the same time a voltmeter is used to
measure the potential difference between the section under test
and a remote point in the ground; and from tbis voltmeter reading

and the total leakage current the resistance of the section can
be calculated.
Determination of Drop of Potential on Rails. As the difference
of potential between two points on the rails of an electric railway
system is in a way a measure of the tendency to cause stray earth
currents, and as many municipal and other public requirements are
based on this figure, this test is an important one, but not difficult
to make, the essential matter being that of getting a potential
wire between the two points on the rail and then measuring the
drop with an ordinary voltmeter. Where the company has telephone or signal circuits which may be disconnected temporarily
these are often used for this purpose. I n some cases a trolley feeder
is used, but a spare feeder is not often available during the time
when such tests are to be made. Often it is possible to arrange with
the local telephone company for its cooperation in furnishing spare
wires from a terminal box in one locat~onthrough proper connections a t the telephone exchange to a terminal box in another
location, and from these terminal boxes a continuation of the potential wire may be run to the voltmeter and rail. In using such circuits through a tele hone exchange, it is important that all connections be removeJon the terminal board which might serve to
impress telephone battery current on the particular circuit in use.
If the instrument used be of high resistance and if the telephone
or signal wires used as potential wires be of copper, it is rarely
necessary to allow for the resistance of the latter in correcting the
voltmeter readings.
Limiting Rail Drop Should Be Average Rather Than Maximum.
In many municipal or other public requirements a limit has been
laced on the potential drop in rails, and in many cases this limit
!as been expressed in terms of a maximum permissible drop. That
this is unfair and that the limit should be expressed as an average
drop is brought out by Messrs. McCollum and Logan in their
paper on electrolytic corrosion (Trans. A.I.E.E., 1913) as foIIows:
"It is evident that if the total amount of damage which results
is proportional to the average current, then the limitation of the
average voltage is more logical than the limitation of the peak load
voltage, since in the former case the cost of meeting the voltage
limitation in any given case is proportionate to the danger involved
irrespective of the station load factor; whereas if the voltage a t peak
load is the deter mini^^ factor, the cost of complying with the requirement depends not only on the danger involved, but on the load
factor of the system, and the poorer the load factor, the
greater its cost will be. The rate of damage does not increase as
fast as the voltage increases, because of the tendency toward lower
corrosion efficiencies a t higher current densities. This indicates
that, with a given average all-day current, the actual amount of
electrolysis that would occur would be less with a bad load factor
than with a good load factor, and hencepoints to the undesirability
of penalizing a high ~ e a of
k short durat~on. I t would appear very
much more logical, therefore, in so far as the damage itself is concerned, to make the average all-day voltage the basis of the limitation rather than the voltage a t time of peak load."

696

ELECTRIC RAILWAY HANDBOOK

Methods of Reducing Earth Potentials and Currents, thus

Mitigating Electrolytic Corrosion
The methods which have been most often used with success in
this country to reduce earth potentials and currents have been
three; the insulated negative return feeder system, the insulating
joint system, and the drainage system.
Insulated Negative Return Feeder System. I n this system the
tracks are drained of current a t radially disposed points about
the power station by insulated negative return feeder cables. The
negative bm bar is not connected to the tracks nor to ground in
the vicinity of the power station except through a resistance approximating that of the negative feeders. Pressures as nearly equal as
are required may be maintained a t the track end of the return
feeders by the use of rheostats on short feeders and boosters on long
ones, or by greatly increasing the amount of copper in long feeders
without the use of boosters or rheostats. The cost of the system
increases very rapidly as the permissible maximum variation
between the various portions of the track decreases. In a few
American cities insulated negative return feeders have been installed
in a portion of the city so as to secure an equipotential condition of
the track in an area where the amount of underground cables or
pipes subject to damage is the greatest. This method is feasible
and most often used where the railway company's power station
or substation is located very close to the center of distribution of a
system composed of several or many lines radially disposed about
the center of distribution.
The Insulating Joint System requires that cable sheaths or piping
systems should have insulating joints a t frequent intervals. If the
insulating joints be suficiently numerous, this system may give
good results, but there is some difficulty in applying them to an
extensive system of underground cables, especially where the
number of manholes is large and manholes crowded. Unless
the number of insulating joints is sufficiently large to break up the
potential around such joints to a very low value around each, there
is also danger of current leaving the pipe or cable sheath on one
side of the insulating joint and entering it on the other to such an
extent that more or less serious corrosion is caused on the psitive
side of the joint. If every joint in the piping system were an
insulating joint, this method would probably be an ideal one, as it
would so increase the total resistance of the pipe line that the
current flowing along it would be practically nil. This is illustrated in the case of some cast iron gas mains where cement joints
of high resistance are used and on which mains practically no
current can be found.
Pipe Ih+age System. In this system the pipe or cable sheaths
e
are connected to the negative side of the
in the p o s ~ t ~ vdistrict
railway return circuits through low resistance cables. This system,
as its name implies, drains the stray currents from the pipes or
cable sheaths through metallic conductors so that practically no current leaves the pipe by way of the earth path. The drainage cables
refera ably should be connected directly to the negative bus bar

ELECTROLYSIS MITIGATION
a t the power station or substation, and should be of very low
resistance. In many cases a meter is installed in the connection
to the negative bus, and in cases where the station shuts down for
a portion of the day there should be a switchin the bus connections
and the circuit should be opened when the station is not in operation. If the drainage system is properly installed, it will maintain
the piping system and cable sheaths throughout their length a t a
potential lower than that of the rails. An objection which has been
raised to the drainage system is that by reducing the total resistance
to the flow of stray currents it thereby increases their volume and
thus increases the damage which may be expected from "joint
electrolysis," or the electrolytic action which may take placearound
a high resistance pipe joint due to the stray currents leaving the
pipe on the positive side of the joint and returning on the negative
side. Some evidence is on record of such damage, but the cases
where such damage has occurred are extremely rare and where noted
it has been caused either by isolated exceptionally high resistance
joints or excessive current flow on the pip~ngmains. Many cases
are on record where a current density of 5 to 10 amp. per poundfoot or more has been carried on cast iron water mains for years
with no evidence of joint electrolysis. The drainage system has
been used by the Bell telephone companies throughout the country
for the protection of their cables for many years. In some cases
city authorities have compelled the installation of the drainage
system and in many cases they have allowed its use where the
water system is owned by the city. The drainage system is very
simple and much cheaper to install than the insulating joint method
on a system of cables or pipes which are already in place in the
ground, and is the system which has been in most common use in
American cities either alone or in connection with the insulated negative return feeder system.
Three Wire Distribution. This system, using parts of the troIley
wire as positive and parts as negative, with the rails as the neutral
member, greatly reduces the current in the rails, as well as the rail
potentials, and thus minimizes stray earth currents. According
to W. Nelson Smith, the first instance of this of which there is
any record was a t Portland, Oregon, about the year 1891, when
several small trolley systems in that city were unified under one
management and supplied from one source of power, on the three
wire system. This was done a t that time partly to keep two separate trolley lines entirely separate from each other as regards their
power supply, and partly to get some saving in copper feeders,
although just how this latter result was to be secured is not clear
from the information available; but it was noticed then that the
districts operating on the three wire system were immune from
damage by electrolysis of water pipes, while those traversed by
other lines operating on the two wire system were frequently
annoyed by this trouble. The next demonstration of this kind was
in Brisbane, Australia, about ten to twelve years ago, where the
manager of the system, who was a former Edison engineer from
the United States, changed it to sectional three wire operation for
the express purpose of preventing electrolysis damage, in which

698

ELECTRIC RAILWAY HANDBOOK

he was entirely successful from the start. I n 1915 the Pacific

Electric Railway stopped electrolysis damage to water pipes
in their interurban territory surrounding Los Angeles by changing
to three wire operation, subsequently enlarging their first installation to cover more of their territory. I n Omaha, Nebraska, the
trolley system was changed to three wire operation in 1916, and the
installation enlarged in 1917. I t is worthy of note that in Omaha
the gas mains were injured by stray current only on streets where
there were railway tracks. The soil there is of clay and of very
low resistance, but it does not carry any amount of alkaline salts.
In 1917 one of the substation districts in Milwaukee was changed
to three wire operation with very successful results in reducing
track potentials. A survey made in 1919, under the direction of the
Bureau of Standards, resulted in the recommendation that the
three wire system be ultimately extended to other districts. I n
1918 the three-wire system was adopted a t Wilmington, Delaware.
One of the special problems there was the protection of a very
important longdistance underground telephone cable system in the
outskirts of the territory, a negative booster with a rather long
feeder being applied for this purpose. I n 1919 the three wire
system was applied a t Winnipeg, first in the Fort Rouge district,
then in the St. Boniface, and successively thereafter in the Mill
Street, North End, Sherbrooke and St. James districts, all of which
.had more than one machine in each substation. The Logan Avenue
district, having onIy a single machine in the substation, was continued on the two wire system and the installation of insulated
track feeders in that district was extended and improved.
Surface Insulation of Underground Metallic Structures. The
U. S. Bureau of Standards, after an extended investigation, presents
the following conclusions: "Such pipe paints, dips, and wrappings
as have been brought to our attention are, with practically no
exception, of no value whatever for protecting pipes from electrolysis when applied in the positive areas near the power houses. If,
however, they are applied in negative areas they may be of considerable temporary value in reducing the current picked up by the
pipe, and in that way indirectly they may reduce damage in positive
areas. We wish to emphasize the fact that the results of these
tests are not to be considered as throwing light on the valueof these
coatings for protecting various metals from natural soil corrosion,
as the tests were designed solely for the purpose of testing their
value as a means for protecting against electrolysis from stray
currents, where the forces tending to corrode the pipes are of much
greater magnitude than those producing galvanic action, which is
largely responsible for slow corrosion of iron in soil. Whatever
use may be made of such coatings in the negative areas for reducing
the amount of current flow in the pipes it should always be looked
upon as a secondary means of mitigation,pnly and not depended
upon as a chief means of protecting ipes
Electrolysis from Alternating 8 w e n t . J. L. R. Hayden
(Trans. A.I.E.E., 1907) draws the following conclusions relative
to alternating-current electrolysis: I t is not a phenomenon like
direct-current electrolysis, on which definite quantitative general

TRANSMISSION LINES
laws can be formulated, but is of the character of a secondary
effect; that is, the action of the positive half wave is not quite
reversed by the action of the negative half wave, leaving a small
difference, rarely exceeding one-half of I per cent of the electrolytic action of an equal direct current. Alternating current electrolysis varies from practically nothing to somewhat less than I
per cent of direct current electrolysis, varying with the chemical
nature of the electrolyte and being practically independent of the
current density. He states that protection from alternating
current electrolysis may be absolutely obtained by the superimposition of a very small quantity of direct current upon the
alternating, the amount of direct current being only 1.5 per cent of
the alternating current.
Transmission Lines
Voltage Determination. The determination of the proper
voltage for the transmission of energy requires a cost study made
up of two parts, namely, the cost of the energy lost and the cost of
the line, the transmitting and receiving apparatus. The amount
and thus the cost of power lost in a given transmission line in
delivering a given amount of power a t the receiver end of that line
decreases if the voltage a t which that power is transmitted be
increased. Such an increase in voltage necessitates an increase in
the cost of insulation, apparatus and protective devices. Thus,
under a given set of market conditions, the voltage at which a
given amount of power will be delivered to a distant point a t the
least cost is limited to that value a t which the sum of the cost of
losses resulting and cost of line, transmitting and receiving ap ara
tus necessary is a minimum. Kapp's modification of Ke vln ,s
law of economy is: "The most economical area cf conductor is
that for which the annual cost of energy wasted is equal to the
annual interest on that portion of the capital outlay which can be
considered proportional to the weight of metal used." As the
distance a t which energy is transmitted increases, the cost of transmission line conductor and the losses in the line become of increasing
importance relati e to the cost of transmitting and receiving
apparatus and the continuance of a given economy demands an
increase in the voltage of operation. The amount of conductor
required for a transmission line to transmit a given amount of power
with a given loss on the line is reduced 75 per cent each time the
voltage a t the recciver is doubled.
The more common transmission voltages are 2200, 4400, 6600,
11,000, 13,200, 22,000, 33,000, 44,000, 66,000, 88,000, 110,000,
140,000.
Spacing of Transmission Line Conductors. Transmission line
conductors must be spaced far enough apart so that there will be
no danger of arcing nor sparking between them and no danger of
appreciable loss due to creepage or corona. The spacing must be
given particular attention with regard to the possibility of the conductors swinging dangerously near together in the wind, especially
where the direction of the wind is not a t right angles to that of the
transmission line.

/'.

700

ELECTRIC RAILWAY HANDBOOK

Power Loss and Regulation of Transmission Line. The values

of the power loss and regulation depend upon the amount of power
delivered, the voltage a t which it is delivered a t the receiver end,
the power factor of the load, the number of phases and the frequency
of the system, the length of the line and the material, temperature,
size, arrangement and distance apart of the conductors. For an
approximate calculation of the power loss and regulation accurate
enough for all ordinary electric railway work, it is sufficient to take
into consideration the following items: Amount of power delivered,
number of phases, voltage a t receiver, power factor a t receiver,
resistance of the line and inductive reactance of the line. (Note
that capacitance of the line is here omitted.)
Calculation of Power Loss and Regulation for Single Phase
Two-wire Line. For three-phase see page 705.

Z2R

Power Loss. (kilowatts line loss) = -I000

in which I

current in line, amperes

=-P X 1 o o o
E cos 0
P = power delivered a t receiver, kilowatts
E = electromotive force a t receiver, volts
cos 0 = power factor of load = power factor of receiver
R = resistance of line, ohms
= 2 X resistance of one conductor, ohms
Example: To find the power loss in line delivering 2 0 ~ 0kw.
at the receiver over a single-phase two-wire transmission line
60,000 ft. long composed of No. I A.W.G. (B. 8 S.) hard drawn
solid copper wire; spacing of conductors, 36 in.; frequency, 60
cycles per second; electromotive force a t receiver, 22,000 volts;
power factor a t receiver, 0.90 (current lagging).
Solution:
2000 X 1000
(Current in line) = 22,000 X 0.90
= IOI amperes
The :resistance per 1000 ft. of No. I A.W.G. hard-drawn solid
copper conductor is 0.1272 ohms; therefore:
(resistance of line) = R = 2 X 60 X 0.1272 = 15.26 ohms
101' X 15.26
(kilowatts loss in line) =
= 155.7
=

loo0

a-E
(per cent regulation) = loo X E
-(see Fig. 76).
in which &

clectromotive force a t generator end of line, volts

= ~ ( E c ~ s B + I K ) ' +(

current in line, amperes

=-P X 1 o o o
E cos 0

E ~ I (COSB)'+ZX)*

TRANSMISSION LINE CALCULATIONS

701

= power delivered a t receiver, kilowatts

E = electromotive force at receiver, volts
cos 0 = power factor of receiver (load)
R = resistance of line, ohms
= 2 X resistance of one conductor, ohms
X = inductive reactance of line, ohms
= 2 X inductive reactance of one wire, ohms
Following are three methods of calculating regulation: I. Solution
by means of above analytical method. 11. Solution by the use of
the "Mershon Diagram," (Fig.
77) introduced b y Ralph D.
Mershon, American Electrician,
1897. 111. Solution by a " d r o p
factor" method based on that
of Charles F. Scott and Clarence
P. Fowler, Electric Journal, 1907.
I . Direct Analytical Method of
Cdculating Regulation. Example:
To find the per cent regulation
of line delivering 2 0 0 0 kw. a t the
receiver over a single-phase, twoFIG.76.-E. M. P.. diagram for
wire transmission line 60,000 ft. ,ingle
phase line.
neglong composed of NO. I A.W.G. lected.
(B. & S.) hard-drawn solid copper
wire; spacing of conductors, 36 in.; frequency, 60 cycles per second;
electlomotive force a t receiver, 22,000 volts; power factors at
receiver, 0 . 9 (current lagging).
Solution:

'""

1ooo
(current in line) = I = P
-XE cos B
2 0 0 0 X1000
-22,000 x 0.90
= IOI amperes.
The resistance per 1000 ft. of No. I A.W.G. hard-drawn solid
copper conductor is 0.1272 ohms; therefore
(resistance of line) = R = 2 X 60 X 0.1272
= 15.26 ohms.
The inductive reactance per 1000 ft. of No. I A.W.G. solid
conductor, spaced (interaxial distance) 36 in., is 0.13218 ohms;
therefore
(inductive reactance of line) = X = 2 X 60 X 0.13218
=; 15.86 ohms

Eo

= d ( E cos 0

+(aa.ooo x 0.90
= 24098

+ IR), + (Ed=o(cos

101

+ +X)*.

x rs.z6)* + ( r z . o o o d r + o . g o l + l ~x~ 15.86)l

702

ELECTRIC RALLWAY HANDBOOK

(Per cent regulation) = loo X
= loo

EQ-E

- 22,000
x 2q,og8
----------22,000

= 9.5 per cent of receiver voltage.

ZI. Method of Cakulaling Reguldion by Use of Mershon Diagra*.

The method consists of first determining the resistance drQp
(called for convenience, "resistance volts") and the electro_
motive>orce required to balance the inductive reactance (called for
convenience, "reactance volts"); these are then expressed in per
cent of receiver voltage and applied to the Mershon diagram
(Fig. 77) from which the per cent regulation is then read directly.
Example: T o find the per cent regulation of a line deliverihg
2 0 0 0 kw. a t the receiver over a single-phase, two-wire trabsmission line 60,000 ft. long composed of No. I A.W G. (B. & S )
harddrawn solid copper wire; spacing of conductors, 3 6 . ih;
frequency, 60 cycles per second; electromotive force a t receiver:
22,000 volts; power factor at receiver, 0.90 (current lagging).
Solution:
PX1000
(current in line) = I = E cos 0
- 2 0 0 0 X I000
2 2 , 0 0 0 X 0.90
= IOI amperes
The resistance per rooo ft.of No. I A.W.G. hard-drawn solid
copper conductor is 0.1272 ohms; therefore
(resistance of line) = R = 2 X 60 X 0.1272
= 15.26 ohms
("resistance volts") = ZR = 101 X 15.26
= I541 volts
(per cent of receiver voltage) = 1541 x 100 = 7.0
22,000
The inductive reactance per 1000 ft. of No. I A.W.G. solid
conductor, spaced (interaxial distance) 36 in., is o 13218 ohms;
therefore
(inductive reactance of line) = X = 2 X 60 X 0.13218
= 15.86 ohms
("reactance volts") = ZX = lor X 15.86
= 1602 volts
1602 X 100 (per cent of receiver voltage) =
- 7.3
22,000
Now on the Mershon diagram (Fig. 77) from the point a t whirh
the vertical line a t load power factor 0 . y ~
cuts the arc o, lay off the
per cent resistance volts (7) horizontally to the right. At the
end of the resistance volts line thus laid off, lay off the per cent
reactance volts (7.3) vertically upward. The end of this line intgrsects the arc 9.5, which is the per cent regulation.

TRANSMISSION LINE CALCULATIONS

FIG. 77.-Mershon diagram.

703

704

ELECTRIC RAILWAY HANDBOOK

I I I . "Drop Factor" Method of Calculating Regulation. In this

method the regulation is found by multiplying the "resistance
volts" by the "drop factor." This "drop f a c t y " is the ratio
which the di5erence between "generator volts and "receiver
volts" bean to the "resistance volts." I t depends upon (I) the
ratio of the "reactance volts" to the "resistance volts" (the ratio
of the reactance of the line to the resistance of the line), (2) the
power factor of the load, and (3) the ratio of the "~esistancevolts"
to the "receiver volts." The "drop factor" IS gven on page 705
for "resistance volts" equal to 10 per cent of the "receiver volts."
'The ratio of the "resistance volts" to the "receiver volts" has a
comparatively small effect on the drop factor, consequently
this table may be used, with small error resulting, in cases where
the "resistance volts" do not exceed 15 or 2 0 per cent of the
"receiver volts."
Example: To find the per cent regulation of line delivering 2 0 0 0
kw. a t the receiver over a single-phase, two-wire transmission line
60,000 it. long composed of No. I A.W.G. (B. 8 S.) hard-drawn
solid copper wire; spacing of conductors, 36 in.; frequency, 60
cycles
second; electromotive force at receiver, 22,000 volts;
power actor a t receiver, o.go (current lagging).
Solution: The inductive reactance per 1000 ft. of No. r A.W.G.
solid conductor, spaced (interaxial distance) 36 in., is 0.13218
ohm.
The resistance per 1000 ft. of No. I A.W.G. hard-drawn solid
copper conductor is 0.1272 ohm.
Therefore

reactance
-=-

0.13218
0.1272
= 1.04
(resistance of line) = R = 2 X 60 X 0.1272
= 1.5.26 0h111s
resistance

P-x 1-

(current in line) = I = E cos 19

= 2 0 0 0 X 1000
22,000 X 0.90
= zoz amperes
("resistance volts") = I R = lor X 15.26
= 1541 volts

From the table opposite, the "drop factor" for a load power factor of go per cent and the ratio of reactance to resistance of 1.q
is found by interpolation to be 1.39, therefore
(regulation volts) = ("resistance volts ") X ("drop
= 1541 x 1.39
= 2142 volts
2142 x 100
(per cent regulation) = --22,000

factor")

705

TRANSMISSION LINE CALCULATIONS

Ratio of
reactanw to
resistante

Drop'factors for power factors of (cumnt lagging)

100%

( 95%

190%

185%

180%

70%

( 60%

140%

(20%

0.10
0.20
0.30

1.00
1.00

1.00
1.01
1.05

1.00
1.01
1.05

0.94
0.98
1.02

0.88
0.92
0.99

0.80
0.86
0.93

0.70
0.82
0.89

0.60
0.67
0.74

0.30
0.40
0.50

0.40
0.50
0.60
0.70

1.00
1.00
1.01
1.01

1.08
1.11
1.15
1.18

1.10

1.14
1.18
1.23

1.08
1.13
1.19
1.24

1.04
1.10
1.15
1.21

1.00
1.07
1.14
1.20

0.93
1.01
1.09
1.17

0.81
0.92
1.01
1.11

0.60
0.70
0.b
0.91

0.80
0.90
1.00
1.10

1.02
1.03
1.04
1.05

1.21
1.25
1.28
1.32

1.28
1.33
1.37
1.41

1.29
1.34
1.39
1.44

1.28
1.34
1.40
1.45

1.27
1.35
1.41
1.48

1.24
1.31
1.39
1.47

1.20
1.29
1.38
1.46

1.01
1.11
1.20
1.30

1.20
1.30
1.40
1,So

1.06
1.07
1.08
1.10

1.35
1.39
1.43
1.47

1.46
1.51
1.55
1.60

1.50
1.55
1.61
1.67

1.51
1.57
1.64
1.70

1.55
1.62
1.70
1.77

1.54
1.63
1.71
1.80

1.55
1.64
1.72
1.81

1.40
1.49
1.59
1.70

1.60
1.70
1.80
1.90

1.10
1.13
1.15
1.17

1.51
1.55
1.59
1.63

1.65
1.70
1.76
1.82

1.74
1.79

1.87
1.95
2.04
1.11

1.80
1.90

1.91

1.85
1.92
1.99
2.06

1.90

1.85

1.77
1.84
1.91
1.98

a.08
2.16

2.08

2.00
2.10
2.20
2.30

1.18
1.10
1.22
1.23

1.68
1.72
1.77
1.82

1.87
1.92
1.98
2.03

1.96
2.03
2.09
2.15

2.04
2.10
a.17
a.23

1.14
2.21
2.29
1.37

2.19
1.28
2.37
1.45

a.25
1.35
2.45
2.53

2.18
2.28
2.38
1.48

2.40
2.50
2.60
2.70

1.25
1.27
1.30
1.32

1.87
1.91
1.95
1.99

2.09
2.14
2.20
2.26

1.22
2.28
2.34
1.41

1.30
a.37
1.44
a.51

2.44
1.52
1.60
1.68

2.53
1.60
2.67
1.74

2.62
2.71
1.80
2.98

1.58
1.67
2.16
2.86

2.80
2.90
3.00
3.10

1.35
1.37
1.40
1.41

1.05
2.10
2.15
a.ao

1.31
1.39
a.45
2.51

2.47
2.54
2.60
2.66

2.57
2.64
2.71
1.80

2.76
2.83
1.90
2.97

2.81
1.91
3.00
3.10

3.07
3.15
3.23
3.31

2.95
3.05
3.15
3.15

3.20
3-30
3.40

1.45
1.48
1.51

a.26
2.31
2.36

2.57
2.63
2.69

1.73
1.80
2.87

1.87
2.93
3.00

3.05
3.11
3.20

3.20
3.30
3.39

3.39
3.47
3.56

3.35
3.45
3.54

3.SO
3.60
3.70

1.53 2.42
1.57 2.4'1
1-60 2.52

1.74
1.80
2.86

1.94
3.00
3.07

3.08
3.15
3.23

3.27
3.35
3.43

3.48
3.56
3.65

3.65
3.75
3.85

3.63
3.73
3.80

1.00

1.99

I.-

~alculationsof Power Loss and Regulation for Thne Phase

Three-wire Line. The power loss and regulation for a given threephase, three-wire line having its wires at the vertices of an q u i ateral triangle are equal to those, respectively, for two single-phase
two-wire lines of the same size of wire and the same spacing and
supplying a load equal to that supplied by the given three-phase
line.
Example: To find the power loss and the per cent regulation
in a three-phase, three-wire line 60,000 ft. long composed of No. I

706

ELECTRIC RAILWAY HANDBOOK

A.W.C. (R. & S.) hard-drawn solid copper wire tlclivering 4000
kw. a t the receiver; conductors a t the vertices of an equilateral
triangle and spaced 36 in.; frequency, 60 cycles per second; electromotive force a t receiver, 2 2 , 0 0 0 volts; power factor a t receiver,
0.90 (current lagging).
Solution: The power loss and per cent regulation in this line
will be equal to the power loss and per cent regulation, respectively,
in delivering 2 0 0 0 kw. a t the receiver over each of twosingle-phase,
two-wire transmission lines 60,000 ft. long composed of No. I
A.W.G. (B. & S.) harddrawn solid copper wire; spacing of conductors, 36 in.; frequency, 60 cycles per second; electromotive force
a t receiver, 22,000 volts; power factor a t receiver, 0.90 (current
lagging). Solving such a single-phase line as on pages 7 0 1 to 704,
the loss is 155.7 kw. and the regulation is 9.5 per cent. Thatis,
the loss and regulation in the above three-phase line are 3 1 1 4
kw. and 9.5 per cent, respectively.

Positive Feeder System

The design of distribution feeders and the location of substations
must be considered together in determining the best distribution
system for any particular track, equipment and condition of traffic.
There is no general method of locating substations and designing
distribution feeders. Each system must be dealt with according
to its particular requirements and limitations. A change in track,
equipment, condition of tra5c, any or all of them, may not be
sufficient to justify a relocation of substations unless the portable substation be used, but may be such as to necessitate a redesign of feeders. It should be noted that the use of the portable
substation demands a provision for ample transmission supply to the
int a t which the substation is to be connected to the transmission
R e . The distribution system must be so designed that the maximum drop in potential to cars a t all points on the line shall not exceed that which will leave a su5cient voltage for the satisfactory
operation of traction motors and auxiliary apparatus a t such points.
In addition to this, the conductors must be of such a size and length
and must be so related that each will carry its share of current
without undue heating. The original distribution should be designed to fill these requirements a t the minimum yearly cost possible for the particular installation being investigated and this
should h attempted in all future developments of that installation.
I n many cases the limitations met in such future development will
prevent the attainment of such a low yearly cost as was possible in
the case of the original distribution a t the beginning of its operation.
I n either case the quality of service required should first be considered, and the financial considerations made secondary.
The design of feeders and the location of substation must be
according to load and its distance rather than distance alone, thus
for the same drop in potential the distance from substation to a
given load would have to be less than that with a smaller load.
That the energy be delivered with no greater than the maximum
allowable drop in potential must be the first provision. That is,

POSITIVE FEEDER

SYSTEM

the energy must be delivered a t a potential a t least sufficient to

accelerate the trains, enable them to climb grades, operate auxiliary
apparatus such as air compressors and contactors and furnish
satisfactory lights, heat and ventilation.
A graphical train schedule (see page 120) is useful in the study
of an interurban distribution system. This shows the location
and speed of trains a t any and all parts of the day, it also shows
the location of stations, turnouts, grades and curves. Trolley,
feeder, transmission line layouts and substation locations may be
added to the graphical schedule for convenience in the future.
From the schedule the condition of regular tra5c which will cause
the greatest drop in potential may be approximated. The condition of greatest demand upon a given substation may also be
approximated from the schedule.
Where local regulations require that an insulated return such as
may be provided by a double trolley or conduit system shall be
used, the supply and return portions of the distribution are generally designed alike. Where the track rails are used for a
return the supply feeders must be so designed that the
drop in the supply feeders shall not be greater than the
between the total allowable drop from substation to car and the
drop in the track return.
Distribution systems divide into two general classes, namely,
interurban and city. These classes are characterized by the frequency of service. The interurban distribution must be designed
to limit the maximum drop to starting trains or trains operating
under very severe grade conditions. In city distribution where
the current required to start an individual train alters the average
current but slightly, the distribution is designed for the average
current demanded by the trains operating on the section under consideration. Such design will be further influenced by concentrated
rush loads of considerable duration.
The position and equipment of substations will often be determined by prominent load centers together with real estate and other
uliar to the problem in hand. Where, howbusiness conditions
ever, there is not s u c c g u i d e it may be necessary to decide upon the
number of substations necessary to feed the section under consideration. The number of stations with which there will result the least
total annual cost will in this case be the best. This may be decided
upon by calculating the total annual cost for each of several numbers of stations. For convenience in arriving a t this total cost the
value of each of the eight following parts may be determined and
these added together:
I. Annual charges on substation buildings and land.
2. Annual charges on substation equipment.
3. Annual charges on distribution conductor (feeders, trolley
wire, third rail, their supports, etc.).
4. Annual charges on transmission system.
5. Annual cost of substation attendance.
6. Annual cost of substation losses.
7. Annual cost of losses in distribution conductors.
8. Annual cost of transmission losses.

:::&
:

ELECTRIC RAILWAY HANDBOOK

Note that the recent practical development of the automatic
substation makes this an important factor in such determinations.
Whether or not the substations will be automatic or manually
operated may make a considerable difference in the values assigned
to numbers 2, 5 and 6 as listed above.
H. F. Parshall, British Institution of Civil Engineers, 1914,
draws the following conclusions from an elaborate study based
largely upon operating costs of the Central London Railway.
(Note that in this consideration, Parshall evidently did not have
the automatic substation in mind.) They illustrate briefly the
influence of the important factors which determine the cost of a
distribution system. With the given energy consumption per unit
of length of line that follows from a given train movement, the
capacity of the substations increases directly with the distance
between them. The energy loss in distribution conductors of a
given section varies with the cube of the distance between substations. The cost of attendance is within wide limits independent
of the size of the substation. The cost of the plant per kilowatt
falls off with the size of the units, but the maintenance and renewals
per kilowatt are more or less constant. With rotary converter
substations and a working voltage of 600,and for certain assumed
energy consumption, the most economical substation spacings were
8% miles, 5% miles and 3% miles for train service of six, twelve
and twenty-four trains per hour, respectively. For a working
voltage of 1200, the substation spacings were X I miles, 7% miles
and 5 miles, respectively, while when 2400 volts was adopted
the most economical substation spacings were 16 miles, 1 2 miles
and 8% miles for the three train services, respectively. With the
present arrangement of rotary converter substations, there was little
advantage in a higher voltage than 2400 for the track conductor.
The economy of higher voltages was shown to be approximately the
same whatever the train service. As between 602 volts and 1 2 0 0
volts there was a saving of 14 per cent in the total annual costs of
the distribution system; as between 1200 volts and 2400 volts there
was a further saving of 7 per cent or 21 per cent as between 600
volts and 2400 volts. If the working voltage be further increased to
3600,there was a decrease in total annual expenditure on substation
and overhead conductor equipment of only 3 per cent, which would
be less than the additional cost of the rolling stock. For single
phase distribution a t 5000 volts the most economical substation
spacings were 31 miles, 24 miles and 16 miles for train services of
two, three and six trains per hour, respectively. At 10,ooovolts
single phase, the most economical substation spacings were 45 miles,
34 miles and 26 miles for the same three train services, respectively.
With three phase distribution a t 5000 volts, the most economical
distances between substations were 38 miles, 31 miles and 18 miles
for the same respective train services. I n most of these last cases,
however, the economical distance between substations thus determined was greater than would be permissible in practice from considerations of both traffic operahon and voltage drop. Further,
in the case of the single phase operation, the lower pressure of

CITY DISTRIBUTION SYSTEM

709

5om volts was found to be the most economical for certain services
and the higher pressures of 10,000 volts, 12,ooo volts and 15,000
volts in vogue in Europe were explained by consideration of voltage
drop.
W. S. Murray, in commenting on the above quoted paper,
points out that Mr. Parshall apparently implied a power factor for
the single phase distribution of from 40 to 45 per cent. Mr. Murray
states that this is but little more than one-half of that which would
obtain in actual practice and points out that the single phase
substation spacing should be considerably greater than that given
in Mr. Parshall's paper if the proper power factor and the same
size train units are assumed.
City Distribution System. The design of a city distribution
system may be well outlined by a typical case, namely, the rehabilitation of the Chicago surface lines. The following is from the
Second Annual Report of the Board of Supervising Engineers,
Chicago Traction:
"It was the .sense of the Board that the direct current feeder
copper for the Chicago Railways Company and the Chicago City
Railway Company should be figured on a basis of 75 amp.
between the dlrect current bus bars of substations or power
and the point of delivery a t the car; it being understood that it is
the intention of both roads to carry the voltage somewhat lower
than 600 volts a t the station bus bars until such time as, through
the elimination of low voltage motors and otherwise, they are
able to raise the voltage to 600 volts. After discussion the following resolution was unanimously adopted:
"Resolved, That the system of secondary or direct current electrical conductors or feeders for the Chicago Railways Company and
the Chicago City Railway Company shall be calculated and plans
made therefor by the Chief Engineer
of the Work on the following
basis :
" I . That the direct current bus bar a t power houses or substations will be operated a t approximately 600 volts.
" 2 . That an allowance of 40 kw. in power house and substation capacity for each standard double truck car of the type
approved by the Board of Supervising Engineers, weighing approxi~ t e l y26 tons light, or its equivalent, will be provided a t each
dxect current bus bar.
"3. That in calculating the copper for current carrying capacity
an allowance of 75 amp. for each standard double truck car, as
described above, or its equivalent, shall be allowed.
" 4 . That an average drop of 50 volts will be allowed between
the direct current bus hars and the center of gravity of the trolley
section, due provision being made for suitable tie lines to take care
of emergency cases.

pziE

This refers to the average maximum for the 2-hour morning and evening
rush periods. In majority of cases the drop is less than 5 0 volts.

"5. That the carrying capacity of insulated underground cables

shall be calculated upon the following basis:

710

ELECTRIC RAILWAY HANDBOOK

?$,d

paw,
covered

IMN)
$00
.
600
500
425
375
....................................

I.OOO.MW) cjic. m/ls cable amp..

~ o o , o o oc~rc.m ~ l cable
s
amp.. .
350,000 circ. mils cable amp.. .
4/0 cable amperes

Triple braided

ysfierfie

1a50
625
325
325

"In arriving a t the kilowatts and amperes per car stated in the
foregoing resolution a series of tests was made jo~ntlyby represestatives of the Chicago City Railway Company, the Chicago Railways
Company and the Division of Electrical Transmission and
Distribution.
"Tests were first made upon a single caf by equipping i t ~ i t h
instruments and stationing observers u n ~t to record the results
in actual operation on digerent kinds oEeruice.
"Tests were also made upon groups of eighteen to seventy-six
cars operating on an isolated trolley section of a mile or more in
length by stationing observers to note the cars entering and leaving the section and also to take readings on station switchboards of
the current and voltage on the feeders to the sections a t IS-second
intervals.
Tests on Single Car. Car tested: Chicago City Railway Company's pay-as-you-enter car No. 5446. Scale weight: 55,800 Ib.
or 27.9 tons. Motor equipment: Four 40-h.p., No. G. E.--80
Gear ratio: 69 : 17 or 4.06 :I
motors. Motor control: No. K-28-E.
with 33-in. wheels. Air brakes: With 16-ft. compressor set for
range of 60 to 85 Ib. Heaters: Electric, consisting of 14 truss plank
heaters, 5 panel heaters, 2 platform heaters. Lighting: Eighteen
16-c.p. slde lights, three 32-C.P. center lights, one 16c.p. platform
light, one 32-c.p. headlight.

Passengers. crew and observers.. ...... 17.O

Weight in tons. car a n d live load. ..... 29.09
Schedule speed in m ~ l e sper hour.. . . . . .
7.69
Volts a t car.
450.0
Heater energy, kw.:
a .865
First point (5 tests).
Second point (5 tests).
5.006
Third point (6 tests). ..............
7.62
I ,603
Li hting energy. kw. (5 tests). ......
~ o t a ?energy, kw. a t car:
Motors and compressor.. ........... 28.21
Motors compressor and lights.. .....
29.82
Same with I point heat. ............ 31.69
Same with a points h e a t . . .......... 34.83
Same with 3 points heat.. .......... 37.56
Amperes a t 550 volts a t car motors
compressor. lights and a points heat.'
63.33

........................

...............
.............

Minimum Maximum

SUMMARY
0; SECTIONTESTS,
JUNE 11, I&
T w o HOURSMAXIMUMSERVICE
--

NU"~

on section..

.........
66.3

1oo.a
44.0

73.00
32.5
35.8

" I t was assumed that the average maximum condition for which
feeders should be provided would be when a car was operating with
motors, compressor, lights and two points on heaters. On this
basis the individual and section tests compare as follows per car:

I Minimum I Maximum I
Individual car
Kw. at car..
Amperes at car..
Section tests:
Kw. at station..
Amperes at station..

......................
..................
..................
...............

34.83
63 .33
36.9
66.3

5 3 . 12

96.58
55.0
100.2

Ave:age
43.81
79.66
42.4
73.0

"The calculations for feeder requirements for the Chicago City

Railway system are briefly outlined as follows:
" I. From the proposed operating schedules the total number of
cars which were required during the rush hours was distributed
and plotted,upon a skeleton map of the system, which is called a
'Spot Map. The afternoon maximum period is usually the heaviest service period, so that the car distribution for 2 hours of what
is styled the 'P. M. Rush' was used on this map.
" 2. The trolley sections were then drawn and the number of cars
on each multiplied by 75, which gives the total average maximum
load for each individual trolley section in amperes. This amount
was placed in a small circle a t the center of gravity of the trolley
sectibn.
"3. A study was then made of the proper location of stations.
The best probable locations were selected and a calculation of load
centers was made by finding the combined center of gravity of the
loads about a given station. If a given system was to be fed by a
single power house, the system load center was also determined,
which showed the most economical location so far as distribution
of copper was concerned for the generating station. If the locations chosen were not the most economical for distribution of copper,
studies were made of comparative costs for other locations where the
company might have property or where real estate for substation
purposes might be obtained to advantage.
"4. After the station locations were definitely settled and the
sections to be fed from each station were decided upon, a 'spider
diagram' was added to the drawing, which then became a drawing

712

ELECTRIC RAILWAY HANDBOOK

of record, and showed at a glance what sections were fed from any
given station and what average maximum Ioad was to be expected
upon that section (see Fig. 78).
" 5 . The most desirable routes for the cables were then determined, drawn upon a diagram and the distances measured from
station bus bar to the load center of each trolley section.

FEEDER CALCULATIONS

713

" 6 . The size of each feeder was then calculated in accordance

with the requirements of Resolution No. 562 (page 709). The
known elements are:
A. Load in amperes.
B. Distance in feet.
C. Drop in potential, 50 volts more or less.
D. Unit sizes of cables and carrying capacity of each.

714

ELECTRIC RAILWAY HANDBOOK

From these the size of cable was calculated or read directly from
the curves without calculation (see Fig. 79).
"7. A certain number of the more important trolley sections are
fed from two separate stations in such a way that in case of the shutdown of one station or of accident to one feeder, the cars on the section could still be operated from the other station, or by means of
the other feeder. These are designated as ' tie-sections' and in addition to the above advantages, are so proportioned and calculated
that on the whole system in case of the shutdown of one or two
stations, a certain proportion of the cars can be carried on the
remaining stations by interconnecting through these tie lines.
"8. Where the ordinances required feeders to be placed underground, i t was necessary to lay out underground conduit lines. A
diagram is used for this; the number of cables over a given section
being re resented arbitrarily by the numerator of a fraction, and
the numter of ducts by the denominator. Extra ducts are installed
in all conduit lines where practicable, to provide for future growth
without tearing up pavements. The percentage of extra ducts will
vary for different locations, depending upon the estimates of future
requirements."

Feeder Calculations

''

Significance of symbols in following Cases I to X, inclusive:

(C.M.), (C.M.),, (C.M.), = cross-sectional area of conductor a t
2 0 deg. C., 68 deg. F., expressed in circular mils. [NOTE:This may
be the cross-sectional area of one solid conductor or the combined
cross-sectional areas of two or more conductors of equal length
operating in parallel to give the equivalent cross-sectional area of
one large conductor. This includes the case of combined working
conductor (trolley wire or third rail) and feeder tied together a t
short intervals.]

= potential drop in feeder as indicated below, volts.

I = total current uniformly distributed along section, amperes.
I. and Zb = current taken from conductor a t points a and
b, respectively, amperes.
I*, ZB, 11,Iz,Z, = current as indicated below, amperes.
K = direct current resistivity of conductor, ohms
per circular mil foot.

c, e.,

K ohms at 20
deg. C . . 68 deg. P.
International Annealed Copper Standard. solid. ........
10.37
International Annealed Copper Standard, stranded.. ....
10.58
Hard-drawn copper, approximate average, solid . . . . . . . .
10.65
Hard-drawn copper. approximate average. s t r ~ n d e d ....
.
10.86
Hard-drawn aluminum, average, solid. ................
16.4
Hard-drawn aluminum, average, stranded.. ............
16.7
Siemens Martin steel, stranded.. . . . . . . . . . . . . . . . . . . . . .
119.7
Third rail or slot contact rail, steel;. ... : . . . . . . . . :. . . . . .
83.0
NOTE: This is for steel of a ressstlvlty e ~ g h ttlmes that of the Intemational Annealed Copper Standard. See page 6 3 5 For sizes, weights and resistances of bare solid and stranded copper and stranded aluminum see
tables on pages 728 to 733.

FEEDER CAI.CUI,ATTONS

715

I , 11, 12, 13, 1 4 , 11 = length as indicated below, feet.

r, 7'1, 1 2 , 7'3, T I , r 6 = resistance per foot of conductor, ohms.

= resistance per foot of track (two rails) return,

ohms.

S,SI,S1 = source of supply, this may bc generating station,

substation or another feeder.
Case I. (I'ig. 80.) Load concentrated a t a point on conductor
of uniform cross-section.
a

Z.Kl

(C.A.i.) = -eo

FIG. 80.

(Volts drop to point a) = e. = Z.rl.

Curves for Calculating Drop and Capacity of Feeders. Fig. 79
gives curves by which the solution of problems of Case I may be
rapidly approximated. These curves are plotted for copper having
a resistance of 10.56 ohms per circular-mil foot. By the application
of proportion they may he used for other resistivities. The following examples illustrate the use of the curves:
(a) Find the drop on a 500,000 cir..mil cable, 5000 ft. long
and carrying 450 amp. Follow the 5000-ft. ordinate up to the
~00,000cir..mil radial line, then horizontally until the 450-amp.
ordinate is crossed. This intersection a t 47.5 gives the volts drop.
( b ) Find the circular mils cross-section of a cable to carry 8 w
amp. 8000 f t . with 30 volts drop. Follow the 8oo-ampere ordinate
up to the 3wvolt line, then horizontally until the 8000-ft. ordinate
is crossed. The location of this intersection gives the size as
2,250,000 cir. mils.
(c) Find the distance a 1,500,ooo cir.-mil cable will carry 600
amp. with a 40-volt drop. Foilow the boo-amp. ordinate to
the 40-volt curve, then horizontally to the 1,500,ooo cir.-mil
line. The ordinate through this line gives the distance as 9350 ft.
Case 11. (Fig. 81.) Load concentrated a t a point, two conductors of tliffcrent cross-section in series. (NOTE: This solution is
convenient when the size of conductor as tletermined by Case I
lies 1)ctwcen two sizcs avnilal,le ant1 it is desirable to so construct the
line that the drop to point a will be equi~lto that for the size of
wire as tleterminetl by Case I.)
1 , = ..c.. - -J o r J
f,.rl - farz
C" - l o r , /
-

I,

also

I U r 2- la,rl

I, - I - I ,

716

ELECTRIC RAILWAY HANDBOOK

Case m. (Fig. 82.) Total load of I amperes uniformly distributed, conductor of uniform cross-section.
IKI
(C M . ) = 26

(Volts drop to end of section)

r =

I 11

Case 1 ' . (Fig. 83.) Load concentrated a t a point, conductors

as in Fig. 83.
(Volts drop to point a) =

(Amperes in conductor A ) =
IA

I.,

(rdl + r.1,)
[---(~212+ r ~ +t

r ~ 4 )

(Amperes in conductor B) =

Prc. 83.

Case V. Load concentrated a t a point, conductors as in Fig. 84.

(Volts drop to point a) =

(Amperes in conductor A ) =

--

FEEDER CALCULATIONS
(Amperes in conductor B ) =

also

IB = I o

- IA.

Case VI. (Fig. 85.) Load concentrated a t a point, conductor bf

uniform cross-section (C.M., of r ohms resistance per foot), source
of supply at each end of conductor, sources of supply a t the same
potential, track return assumed to be of uniform resistance per foot.
I.Kl,lt
(C.M.) = e, E
Iorll12
(Volts drop to point a ) = en = 1

(Amperes in section 11,supplied by St) = I I

1.12
-

f.[l

(Amperes in section 12, supplied by S2) = I 2 = also

I 2 = In

- II

Frc. 8s.

Case VII. (Fig. 85.) Load concentrated a t a point, conductor

not of uniform cross-section but of cross-section (C.M.)l having a
resistance of r, ohms per foot from source SI to point a and of
cross-section (C.M.), having a resistance of r2 ohms per foot from
source Sr to point a, source of supply a t each end of conductor,
sources of supply a t the same potential, track return assumed to
be of uniform resistance (R ohms per foot).
(Volts drop to point a) =

(Amperes in section I, supplied by S1) =

718

ELECTRIC KAILWAY HANDBOOK

(Amperes in section l2 supplied by Sy) =

also

I2 = I . - 11

Case VIII. (Fig. 86.) Load concentrated a t two points, conductor of uniform cross-section (having a resistance of r ohms per
foot), source of supply a t each.end of conductor, sources of supply
at the same potential, track return assumed to be of uniform
resistance per foot.
(Volts drop to point a) = e. = Ilrll
(Volts drop to point b) = er, = Z2rlz
(Amperes in section 11 supplied by S 1 ) =
13)
Id*
I l = Zdlz
-

+ +

(Amperes in section 12 supplied by S I ) =

12 = Za11 Ib(lt [S)
1
also
12 = lo I* - Il
(Amperes in section 13) = Z 3 = z1 - 1.
(NOTE:In case the valuc of 13 thus obtained is minus, this minus
slgn indicates that IS is supplied by S z . )
Case IX. (Fig. 86.) Load concentrated a t two points, conductor not of uniform cross-section (sections 11, l2 and 1s having
resistances of r l , r2 and r3 ohms per foot respectively), source of
supply at each end of conductor, sources of supply a t the same
potential, track return assumed to be of uniform resistance (R
ohms per foot).
(Volts drop to point a) = ea = Ilrlll
(Volts drop to point b) = a = 12fi12
(Amperes in section ll supplied by S 1 ) =

(Amperes in section 12 supplied by S Z ) =

I, = I .
IF,- T I
(Amperes in section !J) = 13 = I I - I .
(NOTE:
In case the value oi I S thus obtained is minus, this minus
sign indicates that I , is supplicd by S2.)

also

FEEDER CALCULATIONS

719

Case E (Fig. 87.) Total load of I amperes uniformly distribu ted, conductor of uniform cross-section, source of supply a t
each end of conductor, sources of supply a t the same potential.

Irl
8
(Cross-section of conductor for drop e to center of section) =
IKl
(C.M.) = -8c
I
(Amperes supplied by SI)= 11= ;
(Volts drop to center of section)

(Amperes supplied by St) = I2 =

Economical Voltage Drop in Conductors. Fig. 88 affords a means

of rapidly approximating the most economical voltage drop per mile
of conductor for a given cost of energy lost and given annual
charges against the installation. These curves were derived
according to Kapp's modification of Kelvin's law, which is: "The
most economical area of conductor is that for which the annual
cost of energy wastetl is equal to the annual interest on that por-

ELECTRIC RAILWAY HANDBOOK

tion of the capital outlay which can be considered proportional to
the weight of metal used." Example: When the annual charges
are $450 per mile per r,ooo,ooo circular mils of conductor and the
cost of energy lost is $0.007 per kilowatt hour, what is the most
economical voltage drop per mile of conductor? Beginning a t the
$450 point on the horizontal scale, follow the vertical to its intersection with the $0.007 curve. From this point follow the horizontal to the scale on the left. The intersechon with this scale shows
that ?he most economical voltage drop per mile in this particular
case IS 20.
Negative Return Systems
Uniformly Distributed Load, Track Extending in One Direction
(No Network). The curves and formulas, Figs. 89 to 95, inclu-

sive, from a paper by G. I. Rhodes, A.I.E.E., 1907, are theoretical, -giving convenient comparisons of voltage drop, earth
currents and amounts of auxiliary negative conductor necessary,
for various return systems in which the load is uniformly distributed
over the whole line which extends in one direction only from the
power station. I t is also assumed that the only resistance in the
path of the earth current is the contact resistance between rail and
ground. This is a safe assumption, as it gives the highest possible
values of earth current. These curves show that the way to obtain
a minimum of stray current from the grounded rails of a single
trolley electric road is to insulate the negative bus bar, and to employ
two or more insulated return feeders either so proportioned in
resistance or provided with negative boosters as to produce equal
potentials a t their connecting points to the rails. The following
seven general cases are considered: I. No copper in return circuit.
11. Copper of uniform section bonded to the rails a t short intervals.
111. Copper distributed to give uniform drop, bonded to the rails
a t short mtervals. IV. A single insulated negative feeder connected
to rails a t middle of the line and a t the power station. V. A
single insulated feeder connecting the rails to the negative bus bar
only a t the middle of the line. VI. Several insulated feeders.
VII. Several insulated feeders with e ual potentials a t all feed
points. In Figs. 89 to 91, inclusive, %e ordinates of the curves
are proportional to the voltage drop (loo per cent voltage drop
= -prL
- - drop to end of line, Case I) and the area beneath the
2Si

curve is a measure of the earth or leakage current.

The significance of symbols is as follows:
i = current a t any point on line
I = total current
v = potential a t any point referred to bus bar as zero
1 = distance from power station
L = total length of line
Lo = distance to point a t which, with bus bar insulated, rail
and earth are a t same potential
r = resistance per foot of return

NEGATIVE RETURN SYSTEMS

resistance per cir. mil-foot of copper = 10.37 ohms
weight per c i . mil-foot of copper = 0.00000302 lb.
equivalent conductivity of track rails in cir. mils of copper
Sc = cir. mils of return copper a t power station
W = weight of return copper
A = area representing leakage current with negative bus bar
grounded
a = area representing leakage current with negative bus bar
insulated.
I. No return coDwr:

11. Keturn coDper of uniform section:

111. Copper distribited for uniform drop:

us-

PI' a straight line up to the point a t which 1 = L1

Si
Se
Sc
,but for values of 1greater than L1,

<

si+sc

IV. Single insulated return feeder from middle of line.

nection to rails a t power house:

Con-

722

ELECTRIC RAILWAY HANDBOOK

V. A single insulated return feeder from middle of line. N o

other connection to rail:
L
From I = o to I = -

and from 1 =
6s.

A =

*[

+S

to I = L

3Si

2L

L
I.. = -and L.

=-

2&

(The weight of copper is the same as in Case IV.)

VI. With several insulated feeders the potential curve will take
a form similar to that of Case V in which the curve between any
two feeders is a portion of the parabola of Case I. If the feeders

FIG. 89.-Potential curves.

Cares I-V.

FIG.w.-Potential

of negative
referred to bus bar. Case VII.

are of uniform size, the general xorm of the potential curve will
approach the straight line of Case I11 as the number of feeders is
increased indefinitely. With a single feeder the area measuring
leakage current is larger than for the case with the copper distributed for uniform drop, and with increasing number of feeders
this area will approach it as a limit. With feeders graded in
size to give uniform potential a t the feed points, the curves will be
of the forms in Case VII.
VII. Several insulated feeders with equal potential a t feed
points. In this case the uniformly equal potenhals a t the ends of
feeders may be secured by properly adjusting the sizes or by means
of negative boosters or by resistances in series. By adjustment of
sizes, the potential of the track must necessarily be greater than
that of the bus bar, and if there is a solid ground a t the power
station there will be a leakage current due to this difference; but

723

NEGATIVE RETURN SYSTEMS

with the use of boosters or resistances the potentials a t the ends

of the feeders can be maintained uniform with the negative bus
bar. For the best results with a given number of feeders it is
evident that they must be connected to the rails a t such points
as will give equal maxima on the potential curves. This may be
accomplished in two ways:
I . No connection to rails a t power station, and distance between
feed points twice the distance from the power house to the first
point, and from the last point to the end of the line (see Fig. go).
2. Connection to rails a t power house with the distance between
feed points and from the power station to the first feed point equal
to twice the distance from the last point to the end of the line
(see Fig. 91).
I. With a s i n k feeder and no connection a t Dower house. the
feed point for equal mavima will he a t the cent& of the line'(see
Fig.
the potential curve being made up of the upper portions
of the parabola of Cast= I, also 5hown in Fig. go.
With the bus bar grounded the area beneath this curve is
4 1 3si J
With two feeders the feed points will be one-fourth and threefourths of the distance from the power house and the area will be

With n feeders the feed p i n t s will be a t

-f-2

A.

.-

2n

-I

zn
2n
of the distance from the power house to the end of the line, making
2n

With one feeder the point of connection to the rails will be a t

HL, making

2.

With two feeders the point of connection will be a t 1 = N L and

u,making

With n feeders the feed points will be a t 1 = A L, 2L

2 n + 1

2 n + 1

21
2n

+ I I,, making

Making a connection to the bus bar, in addition to using feeders,

bas the effect in the equation A

,.4I';[:-

of increasing the

724

ELECTRIC RAILWAY HANDBOOK

number by half a feeder. This will explain the points as the curve
in Fig. 95 of ?+,IS, 2% feeders, etc.
If the bus bar is insulated, the equation for area representing
leakage current is a =

PIG. 91.-Potential of n ative

referred to bus bar. ~ a s 8 1 1 .

PIG. 93.-Relative earth current.

negative bus grounded.

PIG.ga.-Weight

of copper In
negative feeders.

FIG. 94.-Relative earth current.

negative bus insulated.

Cross-section of Copper d Power Stdion. Fig. 92 gives the

cross-section of the copper a t the power station in the several cases,
to give equal weights of copper, the unit of weight being wSiL,
or a weight of copper of length L, which gives a conductivity equal
to that of the rails.
Leakage Currenf. Fig. 93 gives the relative leakage currents
for the several cases (in which the bus bar is grounded) plotted to
amount of copper used. The unit of leakage current is taken as that

NEGATIVE RETURN SYSTEMS

725

which occurs on a line without negative copper, of which the bus

bar is grounded. Fig. 94 gives the relative leakage currents
with the bus bar insulated. Fig. 95 gives the relative leakage currents for Case VII with bus bar insulated, plotted with the number
of feeders as abscissas, the points representing half feeders being as
explained above.

FIG. 9s.-Relative

earth current. negat~vebus insulated. Case VII.

Negative Return System for Network The proper negative

return system for a network, such as is made up of the return paths
in a city system, bay be approximated by the following method:
The locations of the points a t which the load currents which are
assumed to return to the power station are determined. The
current and voltage drop in each branch of the network are then

Frc. 96.-Connections between rails and negative feeders.

calculated and the probable amount of auxiliary conductor is

then estimated and added to the network. The values of current
and voltage drop in each branch of the new network thus found
is calculated. The process of adjustment and calculatjon is continued until the voltage drops are reasonably below the limit placed
for good operation, energy economy, to provide against electrolysis
or as demanded by local regulation.

726

ELECTRIC RAILWAY HANDBOOK

Connections from Track to Overhead Line. The conductor

used from the rail to manhole or foot of the pole is bare, weatherproof, or rubber covered stranded conductor, with or without lead
sheath. Where the rails are in fairly good electrical contact with the
earth, it is not necessary that the conductor be insulated. Where
the conductor is likely to be disturbed mechanically, it may be
encased in pipe. Fig. 96 shows a typical scheme of connections
between rails and negative feeders. The size of the longitudinal
cables will depend upon the load to be cared for, but for reliability
they should be divided into a t least two cables per track as shown.
The number and size of cross bonds also will depend upon the load.
They should be spread out along the track as shown.
Based on: Conductivity of aluminum 61 per
cent. of that of copper; welght of alum~num

OO.t of wpr, O..U pl lb.

lu ,b u of AIullnmm, p. d.

FIG. 97.-Comparison

bring b
, uu ~IAludnun,
p r out.

of costs. copper and aluminum conductor.

Aluminum Conductor Compared with Copper Conductor. Since

aluminum has a greater temperature coefficient of linear expansion,
its use demands a greater height or greater number of poles or towers
and greater spacing of conductors to prevent the conductors from
sagging dangerously low or swinging too near together. Since
aluminum weighs less it will not require such heavy supports and
will cost less to string, but because it is softer more care must be
taken in handling it, in order that i t may not be badly scratched.
Aluminum wire of the same conductivity will gather less sleet but
presents a greater surface to the wind. Due lo its greater radius,
eakage from it is less and the skin effect is slightly greater. Because
of its lower melting point, there is greater danger from arcing. T h e
initial saving (or loss) in the use of aluminum conductor of such a

COPPER WIRE TABLES

727

size as to have a conductivity equal to that of the copper wire

which would be required, may be rapidly a proximated by the use
of Fig. 97. The figure is manipulated as ~ollowsfor the example
shown: From the 15-cent point on the "Cost of Copper" scale
follow the vertical to its intersection with the heavy diagonal. Follow the horizontal through this point to its intersection with the
"Cost of Aluminum" scale. Follow the diagonal through this
point to its intersection with the horizontal through the 29-cent point
and from this intersection follow the vertical down to the "Saving by
the use of aluminum" scale. Here it is found that, jw the example
shown, the iirst cost of the aluminum is 3.7 per cent less than the
first cost of the necessary copper would have been.

Copper Wire Tables

A value for the standard resistivity of annealed copper was

adopted in 1910 by the U. S. Bureau of Standards and the American
Institute of Electrical Engineers, and it has also been agreed upon a t
an International Conference between representatives of United
States, Germany and France. I t is also of interest to note that the
resistivity and temperature coefficient and density of copper on
which the following wire tables are based were adopted in 1913 as
International Standards for all countries by the International Electro Technical Commission. The value is known as the Annealrd
Copprr Standard and is equivalent to 0.15328 ohm (meter, gram)
a t 2 0 deg. C.
This value in various units is equal to:
875.20 ohms (mile, pound) a t 2 0 deg. C.
(The term "ohms (mile. pound)" is exactly equivalent to "pounds per
mile-ohm." which is sometimes used)

1.7241 microhms-circular mil, a t 2 0 deg. C.

0.67879 microhm-inch at 20 deg. C.
10.371 ohms (mil, foot) at 2 0 deg. C.

ELECTRIC RAILWAY HANDBOOK

728

RESISTANCE
AND WEIGHTOF STANDARD
ANNEALEDCOPPER
WIRE
(American Wire Gage (B. & 5.))

oooo
ooo

460.
410.
365.

a~z.ooo. 641.
168.000. 508.
133,000. 403.

3382.
2682.
2127.

o.o?9o
.o618
.0779

0.259
.326
.411

32s.
289.
258.

106,ooJ. 319.
83,700. 253.
66,400. 201.

1687.
133i.
1061.

.I a4

,0983

.519
,654
.825

3
4
5

229.
204.
182.

52.600. 159.
41.700. 126.
33,100- 100.

841.
667.
529.

,197
.a48
,313

I .oq

6
7
8

162.
144.
128.

26.300.
20.800
16.500.

79.5
63.0
50.0

420.
333.
264.

.395

2.09
2.63
3.32

9
II

114.
Ioa.
91.

13.1oo.
10.400.
8.230.

39.6
31.4
24.9

209.
166.
132.

.79a
.999
1.26

4.18
5.28
6.65

ra
I3
14

81.
72.
64.

6.530.
5,180.
4,110.

19.8
15.7
12.4

104.
82.8
65.6

1.59
2.00
2.53

8.38
10.6
13.3

15
16
I7

57.
51.
45.

3.260
a.580.
2.050.

9.86
7.82
6.20

52.0
41.3
32.7

3.18
4.01
5.06

16.8
21.2
26.8

18
19
20

40.
36.
32.

1.620.
1.290.
1.020.

4.92
3.90
3.09

26.0
20.6
16.3

6 39
8.05
10.1

33.7
42.5
53.6

21
23

28.5
25.3
22.6

810.
642.
509.

2.45
1.94
1.54

12.9
10.3
8.14

12.8
16.1
20.4

67.6
85.2
107.

24
25
26

20.1
17.9
15.9

404.
320.
254.

1.22
,970
,769

6.46
5 I2
4.06

25.7
32.4
40.8

136.
171.
216.

00

o
I

10

22

.156

131
1.65

WIRE RESISTANCE AND WEIGHT

729

RESISTANCE
AND WEIGHT
OF STANDARD
ANNEALED
COPPERWIRE
(Concluded)

(American Wire Gage (B. & S.))

I

wdght
Gage
No.

Cross
Diameter section in
circular
in mils
mils

ponnd.

Pe'rd~w
Per mile

II

Resistance a t zoo C.
( = 68O F.)
ohms per
1000 feet

lohml

pex
mile

36
37
38

5.0
4.5
4.0

25.0
19.8
15.7

,0757
.<boo
.0476

,400 41s.
,317 523.
.a51 660.

2190.
2763.
3483.

39
40

3.5
3.1

12.5
9.9

.ON7

.IW

832.
,158 1050.

4392.
5539.

.OZW

NOTEI : The fundamental resistivity used in calculating the table is the

Annealed Copper Standard. viz.. 0.153 28 ohm (meter. p m ) a t 20 deg. C.
The temperature coefficient for this particular res~strvltyrs ate = 0.003 93.
or ao = 0.004 27. However. the temperature coeffictent is proportio.na1
to the conductivity, and hence the change of resistivity per degree C. 1s a
constant, 0.000 so7 ohm (meter, gram). The "constant mass" temperature
coeffic~entof any sample 1s
0.000 597
0.000 005
A' = resistivity in ohms (meter, gram) a t to C.
The density is 8.89 grams per cubic centimeter.
NOTE2: The values l v e n m the table are only for annealed copper of the
standard resisttvity. The user of the table must apply the proper correction
for copper of any other resistivity. Hard-drawn copper may be taken as
about 2.7 per y a t higher resistivity than annealed copper.
N 9 s 3: TblS table is the same as Table VIII page 48. Circular 31. ~d
Edition of the Bureau of Standards except th& Cross-sectron in square
inches (s omitted. resistpces are shown a t ao deg. C. instead of 25 deg. C.
and 65 deg. C., and R s r s t a n c s and Weights per mile have been added.

Weight ~n pounds

Per
1000
feet

700 ooo

0 0 0 0 0

s as0
0

1
(

II

( - 6 8 at
F )looC.'

Ohms
Per mrle per 1000
ret

2'1 720
do80

Class A
Bare, insulated or weather*roof cable for a e r ~ a use
l

1 1 1

006 23
6
6

Ohms

NUFDrameter
ber of
of mres.
w r r e ~ ~ mr~s

6L

136 7
3

Class B
Insulated cable, for other than
aerral use

1 1 1
Outaide
Numd~ameter, ber of
mrls
wrresa

::: /

Dtameter
of wlres,
mr~s

Outsrde
drameter.
mrls

I15 7

1.504
1*460

I 1

212,000
168 ooo
133,000
106 ooo
83.700
66.400

NOTEI: The fundamental resrstrvrty used In calculatrng the table

IS the Annealed Copper Standard VIZ o 153 28 ohm (meter

The temperature coefficrent 1s Ato
o 003 93, or Ao 0.004 27 The denslty 1~8.89$rams per cubrc centrmeter:
IS rn accord m t h the standards a%pted by the Standards Committee of the Amencan Instrtute of Electrical
Engrneers In June 1916. In respect to the "number of wrres
The value$ grven for "Ohms per 1000 feet" and "Pounds per roo0 feet"
are a per cent greater than for a solld rod of cross-sectron equal t o the total cross-sectron of the mres of the cable. Thrs Increment of
a per cent mean9 t h a t the values are correct for cables havlng a lay of I 111 15 7. For any other lay. equal to I rn n, resistance or mass
may be calculated by rncreaslng the above tabulated values by
2 ) per cent.
NOTE3 POTrntermedrate s l u s use strandrn for next large;ske.
Conductors of oooo A.W.G. and smaller are o?ten made solrd and t h s table of stranding should not be rnterpreted a s excluding thi.
practrce.
Class A cable. srzes oooo and ooo A.W.G. IS usually made of seven (7) strands when bare and nlneteen (19) strands when msulated or weather-proof.

gram) a t

20

de

NOTE2 'fhrs table

(% -

8
Y

E9

'OCI
.rsI
'I61

LIZ'
09C:
CLS

88'1
LC.t
66'c

I9 100'
CO COO'
9s COO'

'OSO'C
.oss0c
' 0 9 ~ ~ ~

.IPC

'POP
'P8C

116'0
SP'I
OF'C

8L'C
9L'P
00.9

CC COO'
LO too.
CI SOO'

~011'P
'osr6s
'ofS'9

'PgP
'019
'OLL

89"
8 s
92 6

LS'L
SS'6
O'CI

LP 900'
SI 800'

Fore.

'OCt'8
'OOP'01
'OOIICl

'16
'COX
'PI1

OCIO'
9910.
90~0'

'OoS'pr
'008'OC
'ooF'9c

'8CI
.w1
' C91

L'tI
P.Cc
C'LC
I 4

i
B

C'SI
1'61
I'tC

'oS6'1
'oSt'c
'060'C

c.6S
C'td
'OSI

P.oE
V.gC
P.8V

PIS'
got.
FCC'

' 006'C
'oc6.P
'oor'g

'8Cc
'6LE
'cog

0.19
6'9L
o'L6

9Sc'
roc '
191.

'oag6L
'098'6
.oOP'ZI

'LS6
'OCSI
'OCPC

'CCI

gcl'

'PSI
.S61

t080.0

101.

I
I

(('s q ' a) aSe3 al!M uw3uamy)

%
b

('3 .89 ao) '3 ,oz

LV

3 x 1 nnwmn?y
~
NMvaa-aavH

'SP
'IS
'LS

:;:
'18

ALUMINUM WIRE

Chms
C b
ooo

wmo
-nb

G.? a. .a. ~;xx

Ylnc

OW*

r r r

nno

&&

0-0

0-0

WLO

xs; ;s! x x s

gsg

~8

m ~ l bn m m
m-n
n - o

000
PPn

-no
o o o

mnn
0 0 0
o o o

0 0
0 0
o o

.t

n m c
tYlQ

4G-i &A& &Gd ;ado

do
2;;: m
o b mwm o n
n b w 0-2
2:

en

o & ;&A
6
6 6 on& d d i n
O O n 0-0
n a n r b o ON*

919

0wm
-00
o o o

m o w

dm0

n-"

-wv,

OW-

dd;
-00

d
& & "r 6
n r -

~ l h

l nr

...

?"'O

1?C" ? g ? 0 4 0 1?'O

o o n

900

mmn
nnn

o h n
ow*

b n -

-00

b~w 0
n n n

c m o

own

r n n

n n n

r-r

nnn

0-0

mh o r n

0-0

-8.4

& i + 44

OPY~ o h m
000 000

00
OP

734

ELECTRIC R A I L W A Y HANDBOOK

APPROXIMATEWEIGHT
OP WEATHERPROOF
COPPER
CONDUCTOR
(American Steel & Wire Co.)
Double braid

Triple braid

INSULATED WIRE

ALLOWABLE CUBBENT-CABBYING
CAPACITY
OF

COPPERWIRESAND CABLES
National Board of Fire Underwriters-1913

Size A.W.G.

(B.& S.)

clrcular mds.
I8
16

Rubber insulation.
amperes

Other insulations.
amperes

10

:t

IS
10

a0
25.

10

15
35
50
5s

30
50
70
&

6
5

70

&
Po

90

rw

100

12s
150

22s

100

0000

150
x75
21s

a?%
315

200.000
300.000
400.000
500.000

200
275
325
400

300
400
500
600

600,000
700.000
800.000
900.000

450
500
550
600

680
760
840
920

650

0
00

BBB

215

1.000.~
I.I00,000
I.~OO,OOO
1.300.000

690

730
770

1.000
1.0&,
1.150
1.220

1.4oo.MM
~.sw.ooo
1,600.000
I.700,000

810
850
890
930

1.290
1.360
1,430
1.490

970
1.010
1.050

1.550
1.610
1.670

1.800.000
1.900.000
2,000.000

736

ELECTRIC RAILWAY HANDBOOK

LOADING
OF BABESTRANDEDCOPPERCABLEALONE,WITH ICE
COAT,AND WITH ICE COATAND WIND

ICE AND WIND LOADS

Strands of
conductor

l.ooo.000
1.750.000

Load~ngper lmeal foot

91
91
91
gr

0.1482
0.1387
0.1284
0.1172

2.000

1.906

1.781
1.656

7.008
6.193
5.380
4.508

8.579
7 705
6.813
5.863

2.000
1.937
1.858
1.771

8.809
7 945
7.064
6.114

1.ooo.000
950,000
900.000
850,000

61

0.1280
0.1248
o.1~15
o.1191

1.531
1.468
1.437
1.406

3.674
3.503
3.332
3.162

4.950
4.745
4.549
4.360

1.687
1.645
1.625
1.604

5.130
4.922
4.831
4.645

800.000
750.000
700.000
650,000

61
61
61
61

0.1145

1.375
1.343
1.312
1.250

1.992
2.822
2.650
2.43

4.170
3.977
3.789
3.5~

1.583
1.562
1.541
1.500

4.461
4 271
4.WO
3.847

6oo.000

61
37
37

0.0992
o.Iarp
o.116a

I 23

sso.ooo
500.000

1:15%
1.108

2.235
2.064
1.894

3.32s
3.105
1.908

1.489
1.437
1.405

3.556
3.4ar
3.230

450.000
400.000
~SoPoo

37
37
37

0.1103
0.1040
0.0973

1.062
1.031
0.968

1.724
1.553
1.345

1.705
2.515
2.267

1.375
1.354
1.312

3.03s
2.856
2.620

300.000
250.000
a11.600

19
19
19

0.1256
0 1147
0.1055

0.921
0 875
0.812

1.174
985
0.800

2.067
I 849
1.624

1.281
I 250
1.208

2.431
2.232
1.024

1.s00.000

1,150,000

61
61
61

0 . 1 1 ~

0.1071
o.1032

738

ELECTRIC RAILWAY HANDBOOK

I.OADINGOF BARE STRANDEDA L ~ U CABLE

Y
ALONE,
WITH ICE COATAND WITH ICE COATAND WIND
Loadlng per lineal foot

r.soo.ow
1.575.500
1.431.000
1,351,500

61
61
61
61

1.438
1.406
1.359
r.ja8

1.461
1.393
1.317
1.143

1.679
a.591
1.485
2.391

1.615
1.604
1.573
1.552

3.133
3.047
a.941
a.851

1.272.000
1,19a.~oo
1.113.000
1.033.500

61
37
37
37

.z81
1.250
1.103
1.156

1.171
1 . 8
1.025
0.950

1.21.195
1.095
1.990

1.511
1.500
1.469
1.438

1.149
1.658
2 559
1.455

954.0~
874.500
795.000

37
37
37

1.109
1.063
1.016

0.877
0.805
0.732

1.888
1.787
1.684

1.406
1.375
1.344

3.324
a.a55
a.1~5

715.500
636.000
~s6.500

37
37
19

o.&
0.906
0.859

0.658
0.585
0.511

1.587
1.469
1.366

1.313
1.171
1.140

a.&
1.943
1.845

477.000
3~7.500
336.410

19
19
7

0.781
0.719
0.656

0.439
0.365
0.310

1.244
1.131
1.037

1.188
1.146
1.104

1.720
1.610
1.515

ICE AND WIND LOADS

739

LOADING
or TRIPLERRAIDWEATHERPROOF
STRANDED
ALUMINUM
CABLEALONE,WITH ICECOATAND WITH
ICE COAT A N D WIND

I _ _ I_ In._ I
.....

Loading per lineal foot

-1

/ -(

p
p

Lb.

Lb.

Lb.

Lh.

Minimum Sngs Mowable for Bare Conductor (A.E.R.E.A.

Recommended Specification). In the tables, ages 740 to 742, are
given the sags a t which bare conductors shourd be strung in order
that, when loaded with the specific requirement of 56 in. of ice and a
wind load of 8.0 Ib. per square foot of projected area a t o deg.
F., the tension in the conductor will not exceed the allowable value
of the ultimate strength of the conductor. The sags given for r 20
deg. F. are greater in every case than the vertical component of the
sags a t o deg. F. under the maximum wind and ice load. The
physical constants of the conductor used are as follows:
Modulm of elasticity
Copper. hard drawn. solid or stranded. . . . . . . . . . . 16,ooo.ooo
Copper. soft drawn. solid.. ..................... I z.ooo.ooo
Aluminum, hard drawn.. . . . . . . . . . . . . . . . . . . . . . 9.ooo.000
Temperatvre coefficient
Cmpper ......................................
Aluminum...................................

o.oooo6
o.oooo1i8

740

ELECTRIC RAILWAY HANDBOOK

Span, feet

.:f1 I I I I I ( 1
I25

As@k, Temp.
(B.&S.) Fahr'

-20
o
20

40
60
80
roo
120
ooo

-20

o
20

1:
80
IOO

120
00

-20
0
ao
40
60
80
100
120

In.

1 :

In.
5
5
6
6

8
9

13

20.

;j

:;27.

10

8
ro
12

13
I5
17
20

a
a
3

3
4

5
5

8
9
10
12
13
IS
18

4
4
5
a
2
3
3
3
4

6
6

5
6
7
8
3
4
4
4

d7
9

8
10
I2

7
7
9
ro
12

280
:

5
5
6

11

See page 739.

In.

5
6

5
5
6
7
8
9

120

300

In.

3
3
3
4
4
5

o
20

100

250

400

500

6
.
.

1 1 1 1 1 1 1 1

2
a
3

-20

200

sags

~ n .

0000

150

I3

11

21

::

ra
14
16
19
23
9
10
11

13
15
18
21
25

18
20
24
27
31

~n. ~ t . ~ t . ~ t .

3.5

6.
6.5
7.
8.
8.5
9.

10.
10
11

5
12.
13.
13.5
14.5
IS.

31.
35.
40.
46.

4.5
5.
5.5
6. 10.
7. 10.5

13
15
I7
19
22
2s
29
34

21.
23.
25.
29.
33.
38.
43.
49.

7.
4.
4.
7.5
4.5 8.5
5
9.
6.
9.5
6.5 10.5
7. 11.
7.5 12.

12.
I2.S
13.5
14.

I4
16
18
a1
24
28
32
37

23.
26.
29.
33.
37.
43.
48.
54.

4.5
5.
5.5
6.
6.5
7.
8.
8.5

11.5
12.
12.5
13.5

IS.
15.5
16.
17.
17.5
18.
18.5
19.5

16
18
21
24
a7
32
37
42

2.5
2.5
3.
3.5
4.
4.5
5.
5-

5.5
6.5
7.
7.5
8.
8.5
9.
9.5

11.5
12.
ra.5
13.
14.
14.5
IS.
15.5

18.5
19.
19.5
ao.
20.1
21.5
22.
22.5

9.
9.5
10.
11.

1s.

15.5
16.
17.

SAG TABLES

Span. feet

ELECTRIC R W W A Y HANDBOOK

MINIMUM
SAGSFOR STRANDED
BARE ALUMINUMWIRE

0000

wo

-20

ao
40
60
80
roo
120

a
2

-20
o
20
40
60
80
loo
Iao

oo

I
I

a
a
3

13
1

2
4
7
lo
14

-20

a
2
3
5
8
12
15

o
20

40
60
80
loo
120

14
18

2
10

0
20
40
60
80
roo
120

-20

o
zo
40
60
80
100
120
'See page 739.

10

10
r3

a
2

-20

2
a
3

a
2
4
7
ro
14
17

1:9
13
17

3
3
5
I:

16
zo
2s
3
4

s
7

11

16
20
25

2
2
3
4
7
ra
16
19

14
19
24
28

a
3

4
6

4
6

10

14
18
21
3
4
5
9
13
18
a1
a4

d9

13
18
23
27
31
7
11

16
21
25
29
32
36

5 11.
6 15.
8 21.
11 27.
17 34.
a2 41.
27 46.
32 52.

2.5
3.
3.5
4.5
5.
5.5
6.
6.5

11.
5.
5 . 5 12.
6.
12.5
13.
7.
7.5 13.5
8.
14.
8 . 5 14.5
9.
15.

12.
17.
24.
31.
38.
43.
49.
54.

3.
3.5
4.
5.5
5.5
6.
6.5
7.

5.5
6.5
77.5
8.
8.5
9.
9.5

5
6
8
12
18
23
29
33

13.
13.5
14.
14.5
15.
15.5
16.
16.5

29.
29.5
30.
31.
31.5
32.
33.
33.5

.aa.
22.5
23.
23.5
24.
24.5
2 5
25.5

33.5
34.
34.5
35
35.5
36.
36.5
37.

8 . 5 16.5 28.
41.
28.5 42.5
17.
9.
17.5 29.
9.
43.
29.5 43.
9.5 18.
10.
18.5 29.5 43.5
10.5 19.
30.
44.
11.
19.5 30.5 44.5
11.5 20.
31.
44.5

6
8
12
18
24
29
33
38

a.
2.5
3.
3.5
4.
4.5
5.
5.5

9
14
20
26
31
35
39
43

3 . 5 7 . 10.5ar.
21.5
7. 11.
4.
4.57.511.522.
8
.
12.
22.
5.
22.5
8 . 5 12.
5.
5.58.512.523.
6.
9 . 13.
23.
6.
9.5 13.5 23.5

20
2s
30
34
39
42
45
49

9.
5.
5 . 5 9.
5.5 9.5
6. 10.
6.5 10.
6.510.5
7. 11.
7. 11.

5.
5.5
6.
6.5
7.
7.
7.5
8.

19.
19.5
20.5
21.
21.5
22.
22.5
23.

.....
.....
.....
.....

36.5
36.5
37.
37.
37.5
38.
38.
38.5.

.....
.....
.....
....

.....
.....
.....
.....
.....

13526.543.5
27.
43.5
14.
1 4 . 5 27.
4.
14.5 27.5 44.
27.5 44.5
15.
15.5 28.
44.5
1 5 . 5 28.
45.
16.
28.5 45.

.....
.....
.....

bma
on- o n
0 0 0
0 0 0

f:

zg
a

743

REACTANCE TABLES

mnm
rotm o b
t-

-nn

000

000000

dddddd

da-mwm
n-or-0.0
tn-mbm
mbobn$
0 0 0 0 0 0

-0nmmm
wmnnbn
~ t - w h o bbbbnn
0 0 0 0 0 0

o m ~ m n b
b-omoh
m n ~ e o nnwnaa
0 0 0 0 0 0

t-0."

0 0 0

0 d o o d d

- 0 0
nmn
--n

ObhaOaeo--m
m h z - n ~ dm r m m n o
n-nmon
mnaemh
nommbb d b - s b t t

0 0 0

? U O ? O ?

0 0 0 0 0 0

..

O O ? O O ?
000000

0ddddd
hoOc-m
nont-a
n*wom%

a????!

0 0 0 0 0 0
m-0QOn
- a m ~ m w
- n a n c a
n n w n n n
0 0 0 0 0 0

dd0ddd

000

0 0 0 0 0 0

-an
m n b
-rer - n
3 0 0

r m m n m o m ~ a o o nc o m n a r
ommnab
nmabnm c o - n - a
-ro-JeCLnbnorc
O-mbQm
nmmmmt d b t b t t n - w n n q
O?q?oy
?????0
??.a000
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

000

om*
t-bm

0 0 0 0 0 0

a:. . . g a ~ s a a~a;gga
..

-*noow

0 0 0 0 0 0

0 0 o d d d d d 0 0 0 d
?not-on
om-nnb
o a ~ m no n d b m QOQh-m
not-moo

at--

n o m f t e

R28821
MRbLO
nnnnea
~ b t n n n

mmmbbb
0 0 0 0 0 0
0 0 0 0 0 0

nrto m

000

nt-name
ODY)OI-P~

ggzggg

ont-bt-a
It-no*-

eat-bmo
hQ -fn0

9 3 9 o1P1%%$
38Z
o o o o ? o
GS$ZZ%
q ? ? o.o . %92126
oooooo

0 0 0

d o o o d o

d o o o o o

dddddd

at-m
mmo
n n a

O-OmQm 0 b t - m m o
a ~ b n m on m h h m t ~ m n n b aO - n m b n

nmomt-a
mnomb-

aaaaaa ggaaaa
....

:.z .: . ;;???a

c h - n d ~

o o o o o d

dddddd

0 0 0 0 0 0

m a 2

m o n o w c

nmaeR-a

bcnnOm

qq0

?0??0?

?qqq.q

0.a0000
0 0 0 0 0 0

en-

"7-0

t - h b - b ~
t-*)norm

0 0 0

dddddd

d d d o d d

0 0 0 0 0 0

0-0
ammomn
n o h a n - h o e
0nn m n h - t m
ow- 0 0 0 0 0

m-bnntnnaanm
t-who-n
mmmbtb

nnmoda
nnt-o-m
tat-awn
bdbbnn
0 0 0 0 0 0
00ddd0

000

---

2694%2
ZSgZZP
2%- zo nZmZmG mz m
E m
b b b b b ~ a b n m m

0.00

0 0 0 0 0 0

b n n ~ ~m m
ad-nh
mOt-eQb a m m - n a
Q

999 ?33?00

000000

0 0 0 0 0 0

d00000

000

.ma
.me

:qq

:oo
.Om
.OOQ
.Om

nOmbbO
~O
-Q
IO
w b n n t n
t-bmonb at-moo-noomm
mmmmbb
0 0 0 0 0 0 o o q q q q
0 0 0 0 0 0 d d o o o o
*a-m-m
- O n m m ~
b o a
onbbm
\OmlOl%.-c) WQCCOhO

.oo

0 0 0 0 0 0

:zz

..

moo-nm ncmonm
;:
.:
. zaaaaa aaagaa gggaaa
...
09-

0 0 0 0 0 0

y;;

ncnhnb

......

;a?

-*oamn
b-amoc
mnaco-

btbdnn

o q q q q q
0 0 0 0 0 0
om-mn*
mont-C*
n$LO.OmIl(%

~ :oo"aa;:
T

0 0 0 0 0 0

d d d d d d

ELECTRIC RAILWAY HANDBOOK

m r -mo ena o~ ror-mr-o
r,
amnnao
:?2::o,
%o,EKKE
oooooo
0 0~0 0 0 0
0 0 0 0 0 0
0 0 0 0 ~
0 0 0 0 0 0
dddddd ddddod dddddd
-rCLnmbPI
b W - n O 1 0
O-nm0
m 0 n O 0 n ~
-mnnr-rn ec n
n w ~ n m
c-loo
mr-ntcr
g22ssz 2sE:z:
:22222
::KzzK
.OOOO? 939399 oooooo o o o o o o
0 0 0 0 0 0
000000
dddddd 000000
3 4 ~ ~ 3g,-ttggz
%
anovcm en~-mmY)OO.S-~

mOeeQb

menm-a

n0Onmm m

s2sJspg ERz:
r-

nr,vnst- m m o o o - BGQ%ZX %%X2%%

" ????%?
?sssss
ssssss ssssss
.....

oooooo

000000

dddddd dddddd

:54$f5

s z g s z g gN ams0 sOgmgo nataQr m

s cO sc
.ggo,z'gg 8.g.gggg 3r-r-cccr3 3 9 9 3 ?e r?-?r ?r r?- ?I dddddd dddddo oooooo oooooo
f
m
mnno
n a m a &
s
~
z
g
z
z
~ o o o * *
s::%o ,,,,n,
8 g z Z E :?:;z%
n000;88 O Q O
O ~ e c c h
cr-rr-cc
9 9 9 3 . . ??????
'g????? q q q 3 9 9
neo-om

- onmoon

000000

000000

000000

000000

mncnom

bneonm

??g&%

wgzsg

$!%$!!

0 0 0 0 0 0

000000

000000

mmOc2n m 0 a* E
,em,

m D n Q t 0

n O n 0 n 0

zgzfi24

a?!???

oneman

$f$O:E

sz$z;g

000000
000000
dddddd 000000
g $ f i z % l 8%8EZE z E 8 % 8 3 l:Z"*"
emon*n
. o w ~ ~ ' gOmQQb t n O r-emm00 0008:::
9 9 9 3 , ? ??'g?$?
'gg?gg? ? ?p?. qr eq qb ~ -

?aaa$?
oooooo

f;%f92

Gs::E2

000000

noaoma

S;SXg2?

ssssss ssss8s s q s s s
dddddd Gddddd ddoooo

gz;g!a

:$s!$;Q
0

hOQOOO

n n n m *

OQ

n.0Y)n.I)-

%'??'?!
oooooo

QOO
9 9 9 e? 9 3? 9 4 8 8 8 8 8 8 8 8 8 % 8 8
oooooo 000000 dddddd dddddd
an t n S 0 %emtor- h m m e o n O m t m n t
O
m
r - r~en amr - rn- enm0
Z ' O , ~ E $ $ O %- nOo~r m
,
m ~t m
t nn m

3339?3
oooooo

88?8?!
ddodoo

!?????

000000

:gs$$%
2SIZ:X
ZfiZ298 4K3F0 2
~ f 2 ' 0 , ; z 0 0 0 - n n ~ m m d b " nnOOOr-

?!????

!????? ??????
000000

000000

v0nm-t-

*"l-noY)

+.0rnma

mnnQho

::32$2

SS8gg:

?a???!

!?!???

??????

000000

000000

000000

0 0 0 0 0 0

0 0 0 0 0 0

dddddd

aaaa??
....

000000

E:gzg2,

;%$S%s

745

REACTANCE TABLES
nno

ro*ao*

e,*eaou
nno-zoo

nm~
n

e o m -roc
nb o o n a
m

0 0 0
000

. . . . . . dddddd
0 0 0 0 0 0

fme

noarmlo

0mo000

- - - n o n

000-"-4

1 - - 9 . 0 1 -

l d d d dddddd dddddd
00-

hmncm-

???

0rclr-n
0 0 0 0 ~ - n n r n r r

en-redh

%2S
3
:
2
:
1
*bn o r - m h o o

000

ddodod dddddd

000

000000

0" 2%

am 0 e m r
0 Chnm 0 "

.-

e*
n
oo
*

000
b * N

o Vlm

bo
n
" 0
b

939
.
~

k
E

E!

0
o

g0

g
g

000

mmo
nm"
ncom
nno

???
000

:zs

dddddd dddddd

on.,-0.0

dco

nnO

.::
.:
.:
.:
.:
.

n
d o w n ob
Ohm
W ccmm

oocnno

nmco nb

no-now
one0 n a

dddddd

000000

000000

noomnr-

n a o h n ~

zdz:: d d d d d

- ",

0 0 0 0 0 0
One

0 0 0 0 0 0

1-ggzg2
vooo-w

??????
0 0 0 0 0 0

:gzpgl

g.s.g.z.z .z . 2::222
....

000000

pa,ggr

dddood ddoooo

. w e

n d m o b h
b o w n o *

:oo

oo-moo

3E3H%
??;yyT

0 0 0 0 0 ~

.oo

:??

..

a $ 2mZ e S S g S Z E S

$!mgf;

:a:: 2;gg~ "z"o.8

:??

no m o m n

g. g.2.z.z .z .

n~e m m e,

??? 949999
--

E2Cz1:

c.as~-g
0 0 0 0 0 0

dddddd

nmooon
oomn

2gc-azZ
qqqqqm
000000

000000

m--cob

Zk$:ZZ
22-:::
dddddd

INDUCTWE
REACTANCE
Intersxlal

dFsth";p.

Ohms per IWO Ft. (for one wire) Solid Conductor 60 Cycles -Continued
Size of w i r ~ A W.
. G . (B.8 S.)

O I I I I I s

Ohms per

1000Ft.

Interaxial
distance,
inches 500.000 1450.000 1
o.oIr9
0.0185
0.0224

INDUCTIVE
REACTANCE
(for one conductor)--Stranded Conductor 2 5 Cydes

Circular mils--Size of condu~to*A.

7 0000

W.G.(B.& S.)

000

00

0.0124 0.0130
0.0190 0.0196
0.0229 0.0235

0.0151
0.0273
0.0290
0.0329
0.0357
0.0395

0.0256
0.0278
0.0295
0.0334
0.0362
0.0400

0.0262
0.0284
0.0301
0.0340
0.0368
0.0406

0.0423
0.0444
0.0461
0.0476
0.0489
0.0500

0.0428
0.0449
0.0466
0.0481
0.0494
0.0505

0.0434
0.0455
0.047a
0.0487
0.0500
0.0511

0.0440
0.0461
0.0478
0.0493
0.0506
0.0517

o 04 7
0:04$8
0.0485
o.0500
0.0513
o.05~4

0.0456
0.0477
0.0494
0.0509
o.ogaa
0.0533

0.04640.04~~0.0~86o.o~g7o.o~08o.o~aoo.o5~1o.oS~a

0.0510
0.G528
0.0543
0.0555
0.0566
0.0576

0.0515
0.0533
0.0548
0.0560
0.0571
0.0581

o.0~21
0.0539
0.0554
0.0566
0.0577
0.0587

0.05a7
0.0545
0.0560
0.057a
0.0583
0.0593

0.0534
0.05~1
0.0567
0.0579
0.0590
0.0600

o 0543
0'0~61
o:os76
0.0588
0.0599
0.0609

0.05510.0562 0.0~730.05840.0~950.06070.06180.06ag
0.0569 0.0580 0.0591 o 0602 0.061g0.06a5 0.0636 0.0647
0.0584 o.05g5 0.0606 o:o6r7 0.0628~0.06400.0651 0.0661
0.05960.0607 o.0618o.06ago.06~oo.o6~~o.o6630.0674
0.0607 0.0618 0.0629 0.0640 0.0651 0.0663 0.0674 0.0685
0.0617 o.o6a80.06~go.06~0o.o661o.o67~
0.06840.0695

0.0585
0.0594
0.0601
0.08
0.0615

0.0590
0.0599
0.0606
0.0613
0.06ao

0.0596
0.0605
o.0612
0.0619
0.0616

o.06oa
0.0611
0.0618
0.0625
0.c63a

o 0609
0:0618
0.0625
0.063.
0.0639

0618
0.062~
0:0634
0.0641
0.0648

0.06a60.0637 0.06480 06590.06700.0682 0.0693 0.0704

0.0635 0.0646 0.0657 0'0668 0.0679 0.0691 0.07oa 0.0713
0.0642 0.0653 0.066~0:0675 0.0686 0.0698 0 0709 0.0710
0.064. 0.0660 0.0671 0,06810.0693 0.0705~0:07~6
0.0717
C.06560.0667 0.0678 o.o6890.o7oo0.0711/0.0723 0.0734

0.0485 0.0496'0 0507 0.0518 0.05~90.0541 0.0552 0.0563

0.05oa o.o513lo:oga4 0.0535 0.0546 0.0558 0.0569 0.0580
0.0517 0.0528~0.0~39
0.0550 0.0561 0.0373 0.0584 0.0595
0.0530 0 . 0 5 4 1 ~ 0 . 00.0563
~ ~ ~ 0.0574 0.0586 0.0597 0.0608
0.0541 0.055a 0.05630.0~740.0~850.0597 0.06080.0619

EL ECTRIC RAILWAY HANDBOOK

748

n - d
c m n

nr-OO

qqq

0 - n e - d
a t - - d m
m h 6 O - n
q 0 q ; n n

0 0 0

odoodd

n f c

mlwmtco - n w - o
m
o o n n

b o a
mt-c

3.9

?s?;ry

.................
.................
o - m a a e
o n a o n m
m m m d t d

u m m m e n
r n m - t o
nnneeo

dddddd

dddddd ddddd

m d o n n a
- m o o -lo
n o m b d d

t o n n 6 n
m o o OW b
bnnnoo

--men
~ r n t
ececc

d t n m n
Q ~ n Oo
aor-r-c

0 0 0 0 0 0
dddddd dddddd ddddd
8:;
Z
Z
P
2
Z
Z
%%5E8"
%:Egg
p n m n n o ~ o o o
,, ? ? ? a? o. ao?o; o" "o o n "n "n "o" m" "~ 3'2,8;222
"""""
o
o
o
o
o
o
o
d
d
d
d
d
d
o
f
i
d
d
d
d
d
ddddd
-

0 0 0

not-

" " " H " "

- 0

0-3

lot-

qq?

0 0 0
WOO

dot-

g
0

0
n
m
c

0
m
o
m

0 0 0 0
dmmo
r o m o
m h o r

oonmmlo
n c n * c o
rrnmmmb

o n - c n m
Or-omam
edmlonlo

Oomnrole;
GQQIDO

0 0 0 0 0 0
dddddd
n ~ n n - o o m n o o m n
o $ m - n o
n w n r n n

ddddd

;";y""""""""

'""""
n w - m d

0 0 0 0 0 0

dddddd dddddd ddddd

,"$",Bfz
O,z,","zf
%'O,gzS
W O ~ N O O

r n n n n n

c)b-tdlon
n n n n - n

0 0 0 0 0 0
h o r n 0 0

dddddd

dddddd ddddd

a f n m m n

o n r r - n m

O O m n r

??? 99399y

yty11"

" " " M " "

*""""

0 0 0
-9
w a r6
n

???qyr

m o o o m
ggm,=)z~
993 ????yr
ZnZZ22 2ZPZZ

anO

6 0 r m o m
000aon
a041660n

0 0 0
- 0 0

2%

0 0 0

%&7$g
qqqqyy

:Z9,q:$Z

0 0 0 0 0 0

--woe
w r - n -

2$1X%:: 2%%13%%PZ8
0 0 0 0 0 0

dddddd ddddd

22% %2XZZ% :,"%%E"h1,"84%

g8nO;ZZ
!
?
?;
?
;
z
.
z
.
:
:
.
:
:
.
.
.
zJzJz:Zz
.
.
.
.
.
.
.
2.2 .2 2
. 2.
0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0

; ?a?

20::n:0 ~,022gf:
)t-r-mbo

0 0 0

0 0 0 0 0 0

dddddd

dddddd

*no

nmemme
0 - n d n 0

n m n n - o
t-no00lo
on-,,,
y y y ; ; ;
0 0 0 0 0 0

mlooomr
m n e 6 c
nmm,+d
y;#.#.y
0 0 0 0 0 0

? ? ? e? ? ? m? S !

0 0 0

0 0 0 0 0 0

M " " " " "

g $ 2 e~~m, nmomo ~n mm n* woi tb- mmmro b

mdlo

rn
O

24::go"
a:bB%: :
ommaw-$ -tnmnn

??q?,,y

d O a
O
n

8 n n n n n

q??

???
0 0 0

n o -

" " W " " "

"""""
ddddd
n n - m e

4w--..3PEZ
ddddd

m oo . no
el oo c- n~ nt -o omn m
* au m

z.z.z .z .: :. . .: Z. :. Z. ~. Z.

SZZ::

o o o o o o

ddodd

0 0 ~ v l 0 0 a o n m m ~ )o n t - b m m
nC1ommo
0 0 m a 0 0
nmo-mt-

oomvlr
e n o m 0

oooooo

oooooo

-*a

p833g

o o o

o o o o o o

dddddd dddddd ddddd

t l o 0 0 n m
Omocanb
q q q q o o

*lot-0-0
naodr-o
O o r r n n

en-

m b m
N d H

ooo h m a

ttxz:

: : z J x z zzz::

nr-nnom
mW000
~ Q O I Q O
~ N d O h
n n o w m m dddbd

. . n-nnnn --nnnn n-"-ddd 000000 dddddd dddddd ddddd

0 0 0

WOOD PRESERVATION

749

Preservative Treatment of Poles, Cross-arms and Ties. A joint

committee of the Amer. El. Ry. Eng. Assn., made up of members of
the Committees on Way Matters, Buildings and Structures, and
Power Distribution, reported as follows in 1 9 2 2 , and its recommendations were adopted. (The various specifications referred to
may be found in the A.E.R.E.A. Manual.)
After a careful study of service records and other available data,
the joint committee on wood preservation arrived a t the following
conclusions:
I. That some form of creosote oil is the best timber preserving
agent now in genera1 use for all purposes, by reason of the fact that
its preservative effect is not greatly affected by either rainfall or
temperature, and in addition it has a lubricating effect on the wood,
which diminishes the injury due to mechanical wear; this combination of qualities places it a t the head of all timber preservatives.
2. That where for economical reasons coal tar creosote oil or
water gas tar oil are not available, or other conditions of maintenance
will not justify the expense of creosote treatment, the adoption of
zinc chloride or some other preservative is, without question,
justified in the treatment of timber. Climatic conditions will go
further in determining the economy of these treatments than in any
other, as one can generally figure on doubling the life of untreated
timber by their use, and in dry climates this life probably will be
extended.
3. That in localities where the rainfall is excessive and with humid
atmosphere, where good zinc chloride treatment would be unfavorably influenced by leaching, and in any climate where the checking
of the timber is likely to be excessive, or the mechanical abuse of
the fiber is extreme, and where it is not considered possible to secure
a straight creosote treatment, the introduction of some lubricating
agent with the zinc chloride will have a beneficial effect in retarding
the destruction of the timber from the above cause. In such treatments, the use of a poorer grade of creosote oil is justified. However,
the committee calls attention to the undesirable features of zinc
chloride as a wood preservative when the timber is used in electric
railway construction (see pages 666 and 775).
4. That because of its extremely poisonous nature to humans, as
well as its corrosive action on steel, the use of bi-chloride of mercury
as a wood preservative is not recommended.
5. The full-cell or empty-cell pressure treatments of timber have
proven to be the most efficient in prolonging its life and should
therefore be used wherever possible.
6. That where the expense of some form of pressure treatment
is not warranted by the probable results obtained, the opentank treatment, using either hot bath only or hot and cold
baths, offers the next method of securing maximum penetration
and impregnation.
7. That where facilities for open-tank treatment or dipping are
not available, the results obta~nedthrough brush treatment or
spraying will more than justify the expense of such treatment.
8. That while the committee recognizes the fact that sodium
fluoride or compounds containing sodium fluoride or other fluorides

750

ELECTRIC RAILWAY HANDBOOK

possess many desirable qualities as wood preserving agents, experiments with its use in this country have not yet reached the point
where sufficient data is available on which to base definite
conclusions.
g. That in view of the extent of the study of service records as
well as laboratory experiments and research work which led up to
the adoption of specifications for the three grades of creosote oil,
creosote-coal-tar mixture, water-gas-tar mixture and water gas-tar
distillate by the American Railway Engineering Association and the
adoption of these same specifications by the Amer can Society for
Testing Materials and the American Wood Preservers' Association,
both of which organizations collaborated with the American Railway
Engineering Association in their preparation, and also considering
the widespread use of the specifications in the United States,
the committee recommends their adoption as Recommended
Specifications.
10. That the committee recommends the adoption of Recommended Specifications for carbolineum and similar oils for opentank and brush treatments, as these treatments, being superficial
in character, require an oil of this type in order to give the best
results.
1 1 . That, in view of the adoption of standard specifications for
coal tar oil, distillate oil and refined water-gas-tar for the preservation of wood block by the American Wood Preservers' Association
in cooperation with the American Society for Municipal Improvements, and the extent of the use of such specifications throughout
the United States, the committee recommends their adoption as
Recommended Specifications.
12. That the committee recommends the adoption as Recommended Specifications of the "Specifications for Preservative
Treatment of Wood," which have been adopted as standard by the
American Railway Engineering Association, as these specifications
cover the principal methods of impregnating timber by the various
pressure processes.
13. That the committee submits in addition to these specifications for pressure treatment, certain other specifications for the
treatment of timber by the open-tank and brush methods, for the
benefit of those companies having occasion to use such methods,
which specifications represent the result of a careful study of
existing data and experience with these methods of treatment,
and which we recommend for ad9ption as Recommended
Specifications.
14. The American Railway Engineering Association, the American Society for Testing Materials and the American Wood Pre&cognizing the necessity of standardizing
servers' ~~ssociation,
methods of ana!ysis of creosote oil in connection with standardization of specifications for the oils themselves, have, after a large
amount of study and investigation, adopted standard specifications
for "Standard Methods of Sampling and Analysis of Creosote Oil."
The committee appreciates the varying results which can be
obtained from the analysis of the same oil by different methods,
and while there may be some features of other methods offering

WOOD PRESERVATION.

751

advantages over those referred to above, in the absence of facilities

or time for research and experimental work of its own, and realizing
that the specifications adopted by the three national associations
above represent the consensus of opinion of some of the best
authorities on the subject in the United States, recommends their
adoption as Recommended Specifications.
15. There are also included specifications for the analysis of
carbolineum and similar oils, since i t was felt that the retort method
of analysis did not give sufficiently accurate determinations of the
fractions distilling off a t the various lower temperatures, the limits
of some of these fractions being held down to quite small percentages under the specifications. I n the flask analysis with the bulb
opposite the condensing tube, the temperature recorded is always
the temperature of the vapor actually distilling off into the condensing tube, whereas in the retort analysis a t the beginning of the
distillation the temperature of the vapor which is recorded is that
one-half inch above the surface of the oil, while, as the distillation
progresses, this distance constantly increases and the temperatures
recorded are those of different parts of the vapor. There is some
difference of opinion with respect to the question of precision of
fractionation between a retort and the usual type of distilling flask.
There is a difference between these two devices with respect to
the temperatures a t which different parts of the distillate pass over
to be condensed, that is, any given percentage of the sample will
pass over a t a lower temperature when the retort is used. The
committee believes, however, that where the limits of the fractions
are held to comparatively small limits the flask analysis offers the
most scientifically accurate method of making this determination,
and recommends the adoption of these specifications as Recommended Specifications.
16. The American Wood Preservers' Association, working in
conjunction with the American Society for Testing Materials and
the American Society for Municipal Improvements, has adopted
specifications for "Creosoted Wood Block Pavement" as their
standard, and the American Society for Municipal Improvements,
acting alone in the matter, has adopted specifications covering
the method of laying the block. The committee is convinced that a
large percentage of wood block aving failures has been due to
incorrect methods of laying the brock. We feel that the principle
of placing a sand cushion under wood block and of laying the block
with sand joints is wrong and has been the cause of a large percentage of the ultimate failures of such pavement, and also that the
use of dry sand-cement cushion is not the best practice. While we
feel that wood block is not a desirable type of pavement in street
railway tracks, we recognize that where its use is required, either
by choice or by reason of municipal requirement, i t should be laid
in the most approved method possible and in such a way as experience has proved will insure the best results. We feel that the
time has come to eliminate the use of a sand or mortar cushion
and sand joints, and that the best results can only be expected
where the blocks are laid on a smooth concrete base of adequate
depth and bedded in a thin coat of coal-tar pitch or other suitable

752

ELECTRIC RAILWAY HANDBOOK

water-proofing material and the joints poured with either a pitch

or asphalt filler. I n making this statement, we realize that it is a
radical departure from the long established method of laying wood
block, but the results being obtained where wood block has been
laid by thls method have more than justified the additional expense
which it involves, and the process of manufacture makes it possible
to keep the variation in depth of blocks within such limits as
will make the method practical. We therefore recommend the
adoption of these specifications as Recommended Specifications.

SECTION X I
SIGNALS AND COMMUNICATION

The requisites of signals on an electric railway vary with many

conditions; whether the system is double or single track; whether
it is a city, suburban or interurban system; whether the track is
straight or has many curves; whether the track is level or has
many grades; speed, headway and braking possibility of trains.
and whether direct current or alternating current is used to
the trains. The selection of proper signal apparatus and arrangement for a given system, therefore, calls for an individual
study of that particular system. Development due to such studies
has roduced a great many types and varieties of signal ap aratus.
System. The methods and rules by means of wgich the
movements of a train on a section of track are controlled relative
to the movements of another train or other trains on the same
track and section constitute a block system.
Block. A section of track the use of which by trains is thus controlled constitutes a block.
Block Signal. A block signal is a signal which controls the use
of a block.
Home Block Signal. A home block signal is a signal which controls trains entering and using a block.
Distant Block Signal. A distant block signal is a signal used in
connection with the home block signal to regulate the approach of a
train to that home block signal.
Signal Location and Arrangement. The most important features
influencing the location of signals are: (a) Maximum braking distance of trains as affected by their speed and weight and by the
grades encountered; (b) location of sidings, stations, curves and
like physical features of the right of way; (c) headway of trains
to maintain the schedule; (d) the distance a t which signals can be
easily read and interpreted.
Signals are usually, where possible, placed upon the right-hand
side of the track looking in the direction of the traffic which they
govern.

BL

---

-------

------.

.I

\&,

5% --------------------------

-C

FIG.I.-Curve

protection

Curve Protection. (Fig. I.) Signal I governs west-bound cars

and Signal 2 east-bound, the dangerous or obscured track lying
between the two signals. The dotted lines show the control limits
of the two signals, or, in other words, the extent of the track which
controls the signals. A car anywhere between Signal 2 and point
3 will cause Signal z to remain in the "Stop" indication, while

763

754

ELECTRIC RAILWAY HANDBOOK

Signal I is affected only when the car is between Signals I and 2.

Should two following cars approach a protected curve dase together,
the second car would be stopped by the signal until the first car
had passed beyond the limits of control of that signal.
Preliminary. Were Signal 2 controlled only up to Signal I ,
it would then be possible for two cars to simultaneously pass Signals
I and 2, each receiving a proceed indication.
This may be
avoided by the use of a setting section, or preliminary, extending
beyond one of the Signals, as 1-3. This preliminary section should
not ordinarily be less than loco ft. in length and a length of 1500
ft. is preferable. By the use of this preliminary a car running
west sets Signal 2 to stop after passing point 3, and in case an eastbound car has not already passed Signal 2, i t would be stop ed by
it. In case, however, the east-bound car had assed s i g n 3 2 before the one west-bound passed point 3, ~ i g n a f would
l
be set to
stop, being controlled to Signal 2, and the west-bound car would
be stopped at Signal I . The fact that Signals I or 2 are in the
proceed indication gives the governed car the right to proceed
only up to the opposing signal, or through the block, for if, as in the
last-mentioned case, an east-bound car passes Signal 2 before the
west-bound car passes point 3, the latter will stop a t Signal I and
the opposing car, hav~nghad a proceed signal a t 2, might meet
with an accident. If a preliminary were used on both ends, or the
control of each signal were extended out beyond the other, two opposing cars might get into the preliminary sections a t the same
time. In this event both signals would assume the stop indication
and neither car could proceed. I t is therefore desirable to locate
the signals, in a case of this kind, a t least within braking distance
from the point of curve so that if one car is standing a t a signal an
opposing car rounding the curve will be able to stop before reaching the signal.
*
--------- - - ----<- -s

-----

- - --

.a'-

\t7_

/3-----------A,
2 \-------a
------------ *
+

FIG.2.-Intermediate signals.

Intermediale Signals. (Fig. 2.) Where the use of a preliminary

(see preceding paragraph) is prevented, intermediate signals
are sometimes used. The purpose of these intermediate signals
(I and 4) is to afford protection if two trains running in opposite
directions should happen to pass points z and 3 a t the same time.

S i i Schemes for Suburban and Interurban Service

The selection of a signaling scheme for a given service calls for
a special study of each case. The following general schemes (Figs.
3 to 13, inclusive) are illustrative of common present day practice.
single-track Suburban Railway Headway 5 to 30 Minutes,
Specd Not Exceeding ao Miles per Hour. Employing either noncounting or car-counting signals with trolley contactor control.

'

SIGNALING SCHEMES

755

(Figs. 3 to 6, inclusive; in which S indicates Set, LC., counting

into block, R indicates Restore, i.e., counting out of block. Dotted
lines from a signal contactor indicate control by that contactor.
Arrow heads pointing toward signals indicate setting movements;
arrow heads pointing toward contactors indicate restoring movements in the direction as indicated by the arrow head.)

Scheme I. (Fig. 3.) East-bound cars take the siding and westbound cars remain on the main track. An east-bound car from
A to B, by I, sets B to stop, and A is changed to register car into
A-B. Upon reaching siding a t B the car is counted out of A-B a t
2. If C is in a permissive indication the car may proceed under 3,
thereby counting itself into C-D. If, however, a west-bound car
is in C-D, C will be at stop and the east-bound car will remain in
the siding until the west-bound car has passed C which will be
restored a t 6. For regular movements east-bound cars count
into A-B and C-D a t I and 3, respectively, and count out a t 2
and 4, respectively. West-bound cars count into DC and B-A
a t 5 and 7, respectively, and count out a t 6 and 8, respectively.
The contactors are shown connected for irregular movements as,
for instance, an east-bound car operating 2 counts out of A-8, but
if for any reason it should back over 2 it would count into B-A ; but
as it again proceeeds eastward it will count out again. West-bound
car movements are controlled as illustrated for east-bound, except
that the west-bound waits, if necessary, on the main track opposite the siding.
&

R
*-----.
* S
s <&' 97

R
*-----.
S
6

S~Gf$ig
'-43

R
FIG.4.

Scheme 11. (Fig. 4.) Cars in both directions are permitted to

pass sidings on the main line; car In both directions are permitted
to run through sidings. An east-bound car having operated I
sets B to sto and A to register it into A-B. If C is in a permissive
indication tRe car passes B and counts itself out a t 7 and proceeding forward counts itself into C-D. If, however, C is a t stop,
indicating a west-bound car in D-C, the east-bound car runs into
the siding and counts itself out of A-B a t z. When D-C is restored by west-bound car passing 6, the east-bound car then proceeds into C-D by operating 3. The reversal of these movements

756

ELECTRIC RAILWAY HANDBOOK

may take place, that is, a west-bound car reaching the siding
may pass it on the main track, or if B is a t stop the car wiU enter
the siding a t the east end and wait there until the east-bound car
has cleared B-A a t 7. East-bound cars count into A-B and C-D
a t I or 8 and 3 or 6, respectively, and count out of A-B and C-D
a t 2 or 7 and 4 or 5, respectively. West-bound cars count into
D-C and B-A a t 4 or 5 and 2 or 7, respectively, and count out of
D-C and B-A a t 3 or 6 and I or 8, respectively.

Scheme 111. (Fig. 5.) East-bound cars enter the siding. An

east-bound car approaching siding at B has signal B a t stop protecting it. Upon reaching the siding if C shows a permissive
indication, the car may proceed on the main line and count itself into
C-D a t 4 and count out of A-B a t 5. If, however, C is found
at stop, the car will pull into the siding and count itself out of A-B
a t 6. The west-bound car then passing C counts itself into A-B
a t 5 and counts out of D-C a t 4 and roceeds through B-A; the
east-bound car will then back out of tpe siding, count itself into
B-A a t 6 and run on to the main line. Proceeding eastward it
will count itself into C-D a t 4 and continuing forward will count
itself out of A-B a t 5. A west-bound car upon reaching the siding
and finding B at stop will wait on the main line until the east-bound
car takes the siding.

FIG.6.

..
I

Scheme IV. (Fig. 6.) Either east- or west-bound cars take the
siding, observing the rule that cars are to head in and back out.
An east-bound car in A-C has C protecting it and upon reaching
siding a t which C is located and finding D a t permissive, may
proceed on the main track and count into D-E a t 5; continuing
forward it will count itself out of A-C a t 7. If, however, D is a t
stop, the car will head into the siding and count out of A 4 a t 6 .
The west-bound car then approaching C and finding it clear counts
into C-A a t 7 and continuing forward restores E-D a t 5. The
east-bound car in the siding now backs out on to the main line and
in passing under 6 counts into C-A; then proceeding eastward on
the main line counts into D-E at 5 and counts out of C-A at 7.
As the arrangement is symmetrical from the middle of the siding,

767

SIGNALING SCHEMES

movements the reverse of the above are controlled in an obvious

manner.
Double-track Suburban or Interurban Railway, Headway I
to 10 Minutes. (Figs. 7 to ro inclusive.)
I
%eat

-D

----Indieate.
Eomc Control
-----Indicate1 Dirtnnt Control

FIG.7.

Scheme I. (Fig. 7.) Three-position semaphore or light signals

with track circuit control. An east-bound car passing I sets it to
stop, in which position it remains until this car has passed out of
A-B. Upon passing insulated joints B, 3 indicates stop, I
indicates caution. When the train proceeding eastward passes
insulated joints a t C, I will go to clear, 3 to caution and 5 to stop.
A similar explanation holds for a west-bound train.
This is the signalling scheme used (with 3-indication color light
signals) on the Washington, Baltimore & Annapolis Ry. between
Baltimore and Washington, where heavy cars are operated a t
speeds up to 50 mi. per hr. Polarized track circuits are employed,
which results in the track relay having three positions corresponding
to the three signal indications, and consequently no line wires
are required between signal locations except those for transmission
of power and the special circuits which may be used for switch
indcators. This typical circuit is regarded as the predominating
system of control and type of signal for double-track railway
installations.

Scheme 11. (Fig.

8.) Two-position signals with overlap track
circuit control. An east-bound car passing I sets it to stop when it
passes insulated joints a t A. The signal remains in this position until
the car passes insulated joints a t D. As the car passes 3, insulated
joint C, this signal will also go to stop, and while the car is in C-D
both I and 3 will be a t stop. The fact that I is controlled beyond
3 to D prevents a car obtaining a clear indication a t I and receiving
a stop indication a t 3, with the preceding car just beyond this
signal. A similar explanation holds for a west-bound car. There
are a number of such installations in service, but it is unlikely that
this scheme will be followed to any extent in connection with future
new work.

758

ELECTRIC RAILWAY HANDBOOK

Scheme 111. (Fig. 9.) Trolley contactors. Three indications

by lights alone, a pilot light being used in advance of the
main signal for an indication that the signal has operated. An
east-bound car passing A, in case there are no cars between A and

-indicates Rome Control

-----Iudlcatei Dimtnnt C o t l f d

P l G . 9.

C, changes I from neutral to proceed. The fact that the pilot

light, which is normally extinguished, lights up, shows the motorman that he has counted into the block and gives him the right to
proceed. I is a t stop while the car is in A-B. Upon passing under
B, I is changed from stop to caution and 3 is set to stop, these
indications remaining while the car is in B-C. Upon passing under
C, I goes to clear or neutral, 3 changing from stop to caution and 5
being set to stop. The fact that the contactors are located immediately a t the signals instead of in advance of them, as is customary
for single-track operation, makes the pilot light necessary, SO that
the motorman can be assured that he has counted into the block.
West-bound cars operate in a similar manner.

FIG.10.

Scheme IV. (Fig. 10.) Trolley contactors. Light signals

giving two indications. An east-bound car approaching A does
not pass under it unless I is at neutral. If this is the case, the car in
passing under A changes the signal from neutral to proceed, the
signal remaining a t this indication until the car passes under B ,
at which point the proceed indication is extinguished and the stop
displayed. The car in proceeding eastward and approaching C, ~f
it finds 3 a t neutral, proceeds under C and if the signal changes from
neutral to proceed, roceeds into C-E. The car in passing under D
restores I to neutraf changing 3 from proceed to stop. West-bound
movements are made in a similar manner.
Single-track Interurban Railway, Headway Not Less than I
Hour, Speed 40 to 60 Miles per Hour. (Fig. 11.) Absolute blocking, semaphore signals, only one track circuit from siding to siding,
continuous track circuits, following cars blocked from siding to
siding, no track circuit preliminaries being used, intermediate
signals replacing the preliminaries. Light signals may be used
if desired. Cars on slding within clearance limits do not affect

SIGNALING SCHEMES

759

signals. An east-bound car passing A sets r, 4 and 6 to stop and

protects itself against both following and opposing movements.
3 does not go to stop immediately as the train passes it, due to the
fact that the track circuit is not cut opposite the signal. I t does
go to stop, however, after the train reaches some point as C in
advance of this signal. Proceeding eastward, as the car passes D,
4 goes to clear. West-bound cars are, however, blocked a t 6.
In passing the insulated joint at F, r and 3 behind the car change
from stop to proceed, allowing a following car to enter the block
recently occupied by the first car. 6 will also change to clear, due

to the fact that until a following car enters the block a t A the block
is unoccupied. I n case there had been a meeting point a t siding Y,
the east-bound car would, on approaching 5 , have found it a t stop,
due to the west-bound car with which it had to meet, being between
Land 5 , because 5 is controlled to L. The purpose of 3 and 4, and
7 and 8 is to protect against the possibility of two opposing c a n
passing I and 6 simultaneously, both in this case be~nga t clear.
If this occurred, however, each car would receive the stop indication
a t 3 and 4 and in this way the advance signals 3 and 4 take the
place of the additional track circuit usuallv used as a preliminarv
section.
This is the system which was installed on the W. B. & A. Ry.
between Annapolis and Naval Academy Junction.
Single-track Interurban Railway, Headway 15 Minutes, Trains
in Several Sections, Speed 40 to 60 Miles per Hour. (Fig. 1 2 . )
Absolute blocking, semaphore signals. following cars blocked

-Indieaten

Block8 for Fallowing Car8

8Im.I Control lor Following Carl

------ Indicates
Indicates 81m.1 Control fur Oppollng Cum

practically one-half the distance between sidings, no track circuit

preliminary being used, intermediate signals re lacing the preliminaries. Light signals may be used if desired. $rs on siding inside
clearance limits do not affect the signals. This scheme is somewhat
similar to that shown above, except that the intermediate
signals have been moved farther into the block and arranged
so that following cars can operate more closely than is possible with

760

ELECTRIC RAILWAY IIANDBOOR

that arrangement. An east-bound car passing A sets I, 4 and 6 to

stop, thereby protecting itself against both following and opposing
movements. Proceeding eastward, 3 goes to stop as the car passes
C, I and 4 go to clear as the car passes D, thereby permitting a
following car to pass I in the clear position, the distance between
3 and 4 being braking distance a t maximum speed. If there be a
meeting point a t siding Y, the first east-bound car will find 5 a t
stop, due to the fact that the west-bound car that it is meeting, is
between L and 5, 6 being at stop against the west-bound car. The
east-bound car takes siding a t Y which clears 6 for the west-bound
car providing no following east-bound car is between A and 6.
If a second east-bound car, however, followed the hrst into A-F
after the first car passed D, 6 will remain a t stop until the second
east-bound car reaches siding Y and gets into clear, when 6 will
clear for the west-bound movement. The two cars on siding Y
now back out and the first proceeds eastward, 5 being clear. The
second east-bound car, however, waits a t 5 (whlch was put to stop
when the hrst east-bound car passed it) until the first car passes J,
when 5 will again clear. The second east-bound car may then proceed. 3 and 4, and 7 and 8 serve double purposes; they not only
divide the blocks for following movements but take the place of a
preliminary track section in the same manner as explained on
page 759; i.e., if two opposing cars passed I and 6 simultaneously,
both receiving the clear indication, they will be stopped by 3 and 4,
respectively, in which case one car would have to back up to the
nearest siding. This scheme uses but one track circuit from siding
to siding, continuous track circuits being used.
This system of signaling is commonly known as the T. D. B.
system and is a most popular system of single-track signaling where
high speeds are involved and continuous track circuits are employed.
Either semaphore or light type signals may be employed in this
system, the tendency being to favor P-indication color light signals.
For the last five years wherever this system is recommended
it is based on the use of two track circuits. With average roadbed
conditions where the ballast resistance would not be less than 4
ohms per xooo feet, an end fed track circuit can be operated successfully up to approximately ro,oooft. If one half of the distance
between sidings would exceed 10,ooo ft., it is recommended that a
center-fed track circuit arrangement be installed.

Heavy Electric Traction Single Track. The Chicago, Milwaukee

& St. Paul Ry., throughout its 660 miles of electrification, has installed what is known as the A. P. B. system, a typical signalcontrol
limit diagram being shown as Fig. 13. The operation is as follows:
Upon passing signal 35-0 an east-bound train sets signals 35-0,
32-5 and 29-9 to stop and 28-7 to caution. As the train passes
signal 33-6 that signal is a l set
~ a t stop. When the train passes

SIGNALING SCHEMES
signal 32-5 the aspect of signal 28-7is changed from caution to stop,
and when the rear end of the train has passed signal 32-5,that signal
can clear up. As the train enters the track section between signals
31-2and 29-9,signal 31-2is set to stop. When the rear end of the
train has passed signal 31-2,the control circuit for signal 35-0is established so that this signal assumes the caution aspect. The train
next passes signal 29-9 and upon shunting out the track relay for
the section between signals 29-9 and 28-7 opens the circuit for the
relay a t signal 31-2. When the rear end of the train has passed
signal 29-9 the circuit for signal 33-6 is established and this signal
assumes the caution aspect. In this connection note that signal
33-6 indicated caution with an east-bound train between signals
29-9 and 28-7. Signal 33-6 will indicate stop with a west-bound
train in the track section between signals 29-9 and 28-7 because a
stick relay will be open when a west-bound train receives a proceed
indication a t signal 28-7 and the contacts of this stick relay will remain open with the west-bound train in the track section between
signals 29-9 and 28-7. The control of signal 33-6as described above
allows a following east-boud train to pass signal 33-6as soon as the
rear end of the first east-bound train has passed signal 29-9. AS the
following east-bound train could leave the siding and pass signal
35-0 as soon as the rear end of the first train had passed signal
31-2, it is evident (on account of the short block between signals
35-0 and 33-6 and the long block between signals 31-2and 28-8)
that if both east-bound trains are moving a t approximately the same
speed, the second train might be delayed a t signal 33-6 exce t for
the control of signal 33-6by a stick relay. When the first
train has passed signal 29-9, this will indicate clear, provided the
following train is not east of signal 35-0, but if the following train
is approaching signal 33-6, then signal 29-9 will indicate stop.
When the h s t east-hound train has passed signal 28-7 that signal
will indicate caution if the following train has not entered the track
section between signals 32-5and 31-2,and if the following train has
entered this track section then signal 28-7 will indicate stop. The
control for the signals for west-bound movement can be understood
from the above. I t will be noted that the signals controlling movements into the passing siding are controlled by the track section
which covers the siding plus the first track section beyond the siding.
This is to prevent opposing trains attempting to use the main track
at the siding at the same time. The clear indication of each signal
is controlled from the indication of the next signal in advance in
all cases. It is possible for opposing trains to pass signals 35-0 and
28-7a t the same time, but these trains would be stopped by signals
33-6 and 29-9. The inferior train would then have to back out of
the block and allow the superior train to proceed. This condition
would not occur unless one of the trains attempts to pass thmugh
the block when that train has orders to meet the opposing train a t
the siding. The signal controls are arranged so that if shifting movements are being made a t one siding a train can enter the block from
the siding at the opposite end of the block. This allows shifting
movements to be made up to the last moment without delaying regular trains.

east-Lurid

762

ELECTRIC RAILWAY HANDBOOK

S
i Indications and Aspects.
(I) A.E.R.E.A. Standard use of the three fundamental indica-

tions. In all signaling the use of three fundamental indications:

(a) Stop; (b) proceed-with-caution; (c) proceed.

Green

4t 4

Green

Pmcood.next 81(lnaI a t atop

Proceed
wltb CmntloD

h e e d ondar Control
( Prepared to Btop Short of

m y Obstroetlon)

Frc. r4.-Aspects in three-psition signal~g. R.E.R.E.R. standard,

(2)
A.E.R.E.A. Standard use of semaphore signals. W>ere
sempahore signals are used they shall be so arranged as to Ibdicate (indicating in) three positions, in the upper left-hand quadrant.
Altsrnatirss

Btop and Btay

7 bed P

Btop and Proceed

Proceed

I
Proceed w ~ t h

Distant S l ~ a I m

Procscd. next Signal at 8Lop

C.otlon

Proceed under Colltml

(Prepared to Stop Short
of any O b ~ t r u c t l o n )

PIG.~g.-As~ectsin two-position signalling. A.E.R.E.A.. proposed practice.

(3) A.E.R.E.A. Standard. Aspects in three-position signaling

(see Fig. 14).
(4) A.E.R.E.A. Miscellaneous Methods and Practices. Aspects
in two-position signaling (see Fig. I S ) .

SIGNAL INDlCATIONS AND ASPECTS

763

764

ELECTFUC RAILWAY HANDBOOK

( 5 ) Aspects in position light signaling. Fig. 16 shows the

various indications with their respective interpretations in one of
the most recent developments in modern signal practice. The
position light signal is installed on the Philadelphia-Paoli and the
Chestnut Hill branch of the Penna. R. R. and IS growing in favor
throughout the country.
Spectacle. Fig. 17 shows the design of spectacle for left-hand
upper quadrant go deg. semaphore having 6 in. openings and
6% in. lenses for 3-position semaphore signals and for use with 3
ft. 6 in. pointed blade, adopted as standard by the A.E.R.E.A., 1914.

PIG.I 7.--Spectacle design for three-position semaphore. A.E.R.E.A.

Clearances for Block Signal. Figs. 18 and 19 show A.E.R.E.A.

standard clearance diagrams for semaphore signals, for use on
electric roads where steam road equipment is operated.
Discernibleness of Light Signal in Sunlight. There is a difference
of opinion as to what can be claimed for the range of the various
types of light signals; this is largely due to the fact that there
never has been adopted a standard constituting a basis of comparison. Manufacturers of daylight signals consider that there are two
general classifications, long range and medium range, and it is
generally considered that the long range signals provide a positive
and distinct indication under the most unfavorable sunlight conditions a t a distance of 3500 ft. While the medium range signals
provide a positive indication a t a distance ranging from 1500 to
2000 ft., the long range signals are employed where heavy trains

SIGNAL CLEARANCES

PIG.18.--Standard signal clearance diagram.

ELECTRIC RAILWAY HANDBOOK

i!

-$

767

MANUAL SIGNALS

are operated a t high speeds, such as over the entire electrification

of the Chicago, Milwaukee & St. Paul. The medium range signals
have been found to be adequate for electric railways and interurban
lines where comparatively light trains are involved, even though
operated a t high speeds.
Manually-operated Lamp Signal. (Fig. 20.) This is one of the
s of signals and is in very common use. The lamps
operate on trolley potential and but one additional
signal wire extending thmngh the Hatk is d.
A membm of
the crew of a train about to enter a block throws a switch, thereby $splaying a light a t the distant end of that block with a
caut~onsignal a t the entering end. When the train reaches the
end of the block, a member of the crew throws a switch a t that
leaving end, thereby removing the signals set by hlm a t both
ends of that block. If, however, two trams running in the same
direction occupy one block a t the same time, the signal should not

~ ~ $ ~ ~ ~ & x ~
Trolley

PIG.10.-Manually-operated lamp signal.

be removed until the last train lesves the block. I n some cases,
one lamp only is used in the signal box a t each end of the circuit
while the remaining lamps of the series are installed a t appropriate places along the section as an assurance to the motorman
that his protection remains as set.
The signal should be arranged to revent unauthorized manipua n . This is commody done ey p&odsing the o erating
handles. The following scheme developed by the Rhode 1sLnd Co.
for the protection of hand-operated signals also insures the removal
of the reverse handle from the car while the signal is being set:
Snap switches are used instead of knife switches. A round piece
of metal having a square hole in one end is substituted for the
ordinaryrubber buttonon the snap switch. The switch is mounted
so that this square hole is slightly above an opening in the bottom
of the signal box. The motorman's reverse handle, having one
end squared for the purpose, is used as a key to turn the switch.
Manual Block System. The manual block system is one in which
trains are controlled by signals operated manually, upon information received by telegraph, telephone or other means of communication. It consists of a series of block sections with a block

ELECTRIC RAILWAY HANDBOOK

control station a t each end of each section, a t which there are placed
signals to be used by the operator to contrd the trains entering
and using said block. I n the operation of this system. infonnation concerning the condition of a given block IS obtained by
communication between the operators a t the ends of said block,
and trains are instructed as to movement by means of the signals.
The safety of operation under the manual block system depends
upon the accuracy with which orden are received and transmitted, there being no check, either electrical or mechanical, to
prevent the dis lay of a wrong signal.
Controlled dmual Block System. The controlled manualblock
system is the same as the manual block system just described
except that the signals a t the ends of each block are electrically interconnected in such manner that cooperation of the signalqen a t
both ends of the block is required to display a clear signal. I t
is a great improvement over the manual block signal, but still
does not prevent two operators from making simultaneous error.
This is overcome by the continuous track circuit, so arranged
that a train having entered a block will automatically put the
proper signals to stop and hold them there until the train is entirely
out of the section, regardless of any attempt on the part of the
operators to clear them. Its use on many electric lines is proh~bitedon account of the large expense involved in maintaining
operators a t such frequent intervals.
Automatic Block System. The automatic block system is a
system in which the signals govern the entrance to each block
and are automatically controlled by the train as they proceed, in
such a manner that head-on and rear-end protection ~safforded.
A number of forms of the automatic block systems are being used,
their essential differences appearing in the manner in which the
control of the signals by the train is effected. These different
forms of automatic systems include the following methods of
control: (I) Trolley or auxiliary rail contacts, which are operated
by passage of trolley or third rail shoe and thus operate the signal
circuits when a train passes; (2) continuous track circuit, based
on a fundamental principle of a train controlling, a t all times, the
signals governing the specific portion of the track occupied, and
utilizing the track rails as a signal circuit; (3) short track circuit,
which utilizes short insulated sections of the track a t the ends of
the block for setting and restoring signals.
Trolley Operat* Signals. Trolley operated signals are generally
operated by electrical contact through the current collector of the
car or by electrical contact through a switch which is opened or
closed mechanically by the current collector in passing.
Contact Type Trolley Operated Signal. Fig. 2 1 shows the
relation of the essential parts of a typical trolley contact signal
(the Nachod). The mechanism of the signal itself may be divided
into three parts: (I) The indicating, consisting of the lamps, color
lenses and opaque color disks; (2) the intermediate, represented
by the relay, which converts the impulses of current caused by
the passage of the car into signal indications; (3) the actuating,
cons~stingof the overhead trolley contact switches, which by the

TROLLEY OPERATED SIGNALS

769

passage of the car determine the setting and the clearing of the
signals. The trolley switch consists of a light angle-iron frame
supporting a t its ends through insulating blocks, two flexible
inclined contact strips, the trolley wire being withdrawn during the
length of the contact strips, which are formed to receive the wheel
without shock. One strip is connected to trolley, the other to
ground through the relay. The wheel, beinga metallic conductor,
bridges them and sets the signal whether the car is taking power
or not a t the instant of passing the switch. One signal wire in
addition to the trolley wire connects the apparatus a t the ends of
the block. Fig. 21 shows the parts in position for no train in the
block, in which case each end of the signal wire is grounded through
a red lamp R. With no current in the coils. the armatures drop.
The first entering train sends proper current through magnet A
in the relay a t the entering end, operating a two-way revolving
switch, which transfers that end of the signal wire to trolley through
a white light W and a magnet D. This light and the red one a t the

other end now burn in series through the signal wire. Successive
following trains turn the revolving switch so that the contacts
overla further, but make no change in the electrical circuit.
Each k v i n g train energizes magnet C to break temporarily the
signal circuit a t P I , permitting magnet D in the first relay to drop
its armature and revolve the switch in the reverse direction. When
the same number of impulses has been made on magnet C as on A,
that IS, when all the trains that have entered the block have left it,
the signals are cleared and the connect~onsare as shown. The
color disks are brought to an indicating position by magnet D
and one m shunt with R, not shown. A no-voltage magnet, not
shown, is interlocked with magnet D to prevent a motion of the
armature of the latter should the power fail with trains on the
block. This is a signal of the absolute, permissive type. To illustrate the permissive feature, suppose a train has entered the block
and set the signals, and before it leaves, a following train arrives
a t the end of the block showing white The motorman will understand by this that there is a t least one train ahead going in the
same direction. As his train runs under the contact he notices
the flash of the white light that his train causes. This is an indication that his train is counted in or registered in the signal relay;
and so on for a number of trains following each other, each receiv~ng

770

ELECTRIC RAILWAY HANDBOOK

its signal that there are other trains ahead of it and that it has
counted in. When the first train arrives a t the end of the block,
in running under the contact switch, it causes a blinking of the

light, but the signals are set as before. They so remain until the
last train dears them, leaving the block ready for trains in either
direction. If a trrun should enter the block from one end and back

TROLLEY OPERATED SIGNALS

771

out by the other track a t the same end, using the single track merely
as a Y or cross-over, the signals will be cleared when the train leaves
the block. This makes a very flexible operation, as in the complicated case where a number of trains enter the block from one end,
and are continuously entering and leaving, some leaving by one
end and some by the other, but the signals protect all the trains
in the block and are cleared only when they all leave it. The
signal may also be operated as an absolute block system by leaving
the hand-clearing switch open, which is equivalent to opening the
line wire. For instance, the motorman of a work train, entering
the block closes the block a t both ends by leaving the hand-clearing
switch open. Then no permissive signals may be obtained from
either end. To leave the signals for normal operation again, he
must close the switch when leaving the block. In case of failure
of line voltage, the signals reappear on its resumption with the
same indications.
Fig. 2 2 shows .the circuits of another typical trolley contact
signal (the United States).
The right-hand front magnet is the setting magnet which rotates
the registering wheel one step for each energization. The lefthand front magnet is the restoring magnet which a t each energization rotates the registering wheel one step in the direction opposite
to the above. The front or main contact bar is in its right-hand
position when the block is clear, but is thrown to its left-hand
position when the first car is registered into the block on the registerlng wheel and remains in the left-hand position until the last car
is registered out. The middle contact bar or alternating switch
is operated by a star cam on the registering wheel so that it stands
in its right-hand position when each odd car is registered in, and
in its left-hand position when each even car is registered in. When
it is in its right-hand position, it causes the right-hand green and
white semaphore to be displayed, and when it is in its left-hand
position it causes the left-hand green and white semaphore to be
displayed so that the proceed indication will be given only when the
registering wheel operates. The rear contact bar, normally in its
right-hand position, is momentarily thrown to its left-hand position
when the rear right-hand magnet is energized, but as soon as the
latter is de-energized falls to its normal position. This magnet
is energized by the restoring trolley contactor.
The trolley wheel of a car entering an unoccupied block from the
left strikes the setting contactor, thus completing a circuit from
trolley wire, through fuse 4, through setting (right-hand lower)
magnet, resistance and fuse G to ground. This moves the registering wheel one notch and moves the main (front) contact bar to the
left-hand position. As soon as the trolley passes beyond the contactor the circuit previously made is opened and the armature of
the setting magnet falls back by gravity into its normal position.
A circuit is then completed from the trolley wire a t the right-hand
end of the block, through the wire indicated by the heavy line,
fuse r in the right-hand signal, resistance, red lamp, pick-up magnet,
front left-hand magnet, the two right-hand contacts and main
antact bar, the contacts and the rear contact bar, rear left-hand

ELECTRIC RAILWAY HANDBOOK

locking magnet, fuse 3, signal circuit line wire, fuse 3 in the lefthand signal, rear left-hand locking magnet in the left-hand signal,
rear contact bar, upper contacts and main contact bar, around
through contacts and middle contact bar, right-hand permissive
semaphore magnet, green light, pick-up magnet, thence to ground.
If a second car passes over the contactor, setting magnet will be
energized in the manner above outlined for the first car and the
registering wheel will be rotated one more notch, the middle contact bar or alternating switch will be thrown to the left-hand position. The circuit through the signal a t the right-hand end of the
block will then be completed as above outlined, but the circuit
through the left-hand signal will be completed as follows: From
fuse 3, through rear left-hand locking magnet, rear contact bar,
upper left-hand contacts, main contact bar, left-hand contacts
and middle contact bar, left-hand semaphore magnet, left-hand
green light, pick-up magnet, thence to ground. This will cause
the opposite green semaphore disk to show. When the trolley
wheel passes the restoring contactor a t the right-hand end of the
block, a circuit is completed through fuse 5, rear right-hand magnet, resistance, thence to ground. This moves the rear contact
bar to its left-hand position, opening the signal circuit previously
described and completing a c~rcuitacross the left-hand pair of
contacts. This action completes the restoring circuit as follows:
From the trolley wire a t the left-hand end of the block, through
fuse I, resistance, red lamp, pick-up magnet, restoring magnet,
lower left-hand contacts and main contact bar, fuse 2, restoring
circuit line wire, fuse 2 in the right-hand signal, left-hand pair of
contacts and rear contact bar, thence to ground. This energizes
the restoring magnet in the left-hand signal and rotates the registering wheel one step in the direction opposite to that in which it was
rotated by the setting magnet. As the trolley leaves the restoring
contactor the circuit just described will be opened, rear right-hand
magnet in the right-hand signal will be de-energized and the rear
contact bar will fall back to the right-hand pair of contacts. This
places the signal circuit in its normal condition. When the last
car is counted out of the block, the main contact bar in the lefthand signal is thrown to its right-hand position. Both signals
are then neutral. If a car enters the block from the right-hand
end, the circuits will be completed in the same manner except they
will be fed in the opposite direction, that is, from the left-hand
end of the block.
Single-rail Track Circuit Alternating Current Signal System.
Fig. 23 shows the relation of the essential parts of a track circuit
alternating-current signal system in which one rail (the "block
rail ") is insulated in sections and is used to carry only current of the
signal system. The other rail is used to conduct the train propulsion
direct current and the current of the signal system. Alternating
current whose potential is reduced from that of the signal mains by
transformers a t the block is used to operate the signals. A noninductive resistance is placed in series with the secondary of the
transformer which supplies the current for the track signal circuit.
A non-inductive resistance is likewise placed in series with the signal

ALTERNATING CURRENT TRACK CIRCUITS

773

relay and an inductance is shunted across the signal relay. The

ses of these resistances and the inductance are to reduce to a
::r@ble
value the direct current which, due to the drop in POtential of the train propulsion current in the block, flows in the
relay and the track transformer, also to limit the flow of alternating
current through the track transformer when the rails are bridged
by wheels and axles of a train in the block. The magnetic circuits
Trolley or Third Rail

PIG. zj.-Single-rail alternating current block signal.

of inductance and transformer each include an air gap to diminish

the magnetic effect of the small direct current which passes through
their windings. Wheels and axles of a train in the block shunt the
signal current, thus causing the relay to release its armature and
set the signals.
Double-rail Track Circuit Alternating Current Signal System.
Fig. 24 shows the relation of the essential parts of a track circuit

I ITranaformcr

-11-

A.C.8imal Mainsh

PIG. 24.-Double

&

rail alternating current block signal.

alternating current signal system in which both rails are used i n

common by the signal alternating current and the train propulsion direct current. This common use is made possible by the inductive bonds which, while offering little resistance to the train
propulsion current, maintain a considerable alternating current potential difference between the rails when there is no train in the

774

ELECTRIC RAILWAY HANDBOOK

block supplied by the track transformer. While this considerable

alternating current potential difference continues the armature of
the relay holds the signals a t the normal indication, but when,
as by the bridging action of the wheels and axles of a train in the
block, this alternating current potential is greatly reduced, the
armature of the relay is released and the signals are set. The inductive bond consists of a single winding of heavy conductor on a
laminated iron core and having a tap brought out a t its middle
point. (The magnetic circuit for use with direct current for
propulsion includes an air gap.) The bond is connected across the
rails at one end of an insulated section of track and the middle tap
is connected to the middle tap of a similar bond connected across
the rails a t the adjacent end of the adjoining insulated track
section. To limit the flow of current supplied by it, the track
transformer is provided with an adjustabie leakage path between
its primary and secondary windings.
This system is also used where alternating current is used for
train propulsion. For this service the frequency of the current in
the signal circuit is greater than that of the train propulsion current.
The impedance bonds have no air gaps. They offer little impedance to the train propulsion current, but they offer sufficient impedance to the signal circuit current. The relay is constructed
to o erate only with current of the higher frequency.
Track Circuit Signal System. In this system only short
insulated sections of track a t the ends of the block are used in the
signal operation. There have been a few signal installations emloying the short track circuit principle, or what is more commonly
inown as the "trap circuit" arrangement, but this has been resorted
to only in such cases as where the roadbed was such as not to permit
the installation of continuous track circuits. While it represents a
possibility, it has a small place in modern signaling. The following
outline is based on the Kinsman system. Special provision for
signal operation is made between adjacent rails of but one side of the
electric railway track, thus the other rail is left continuous as if it
were to serve only the needs of train propulsion. The track circuit
includes two adjacent insulated sections, each two rail-lengths long.
The sections are insulated by three insulated joints. The section
farther from the center of the block controls the setting of the
signals, while the second section controls the release of the signals.
The return circuit of the train propulsion current is carried around
these insulated sections by an auxiliary conductor, either rail or
cable. The relays are operated by current from a primary battery.
When the block is clear the normal position of the signal arm is
vertical. The train in entering the block first passes over the
setting section and operates the setting relay. This cuts out a
release relay and makes both insulated sections a setting section.
If a train is leaving a block it first passes over the release section,
cutting the setting relay out of circuit and making both sections a
release section. A stepby-step controller a t each signal registen
the movement of the trains into and out of the block and controls
the circuit operating the signal. The controller switch is not thrown
back to its normal position until the last train leaves the block.

SL

ZINC TREATED TIES

The position of the signals is controlled by the positian of the
controller switch. When the block is unoccupied this switch
is a t normal position and the signals a t both ends are in vertical
position. As a train enters a block the setting relay operates the
controller a t the entering end and the controller switches open.
This breaks the circuit, de-energizing the signal a t the distant end
of the block, allowing it to fall by gravity to the horizontal or
stop position. As a signal assumes the stop position it opens
the circuit. When an entering train leaves the insulated section,
the signal a t the entering end is de-energized and falls by gravity
to the 45-deg. or caution position and is held in this position
by a slot magnet which is operated by the switch on the controller.
As each succeeding train enters the block it passesonto the insulated
section and the signal moves from the 45deg. position to the vertical position; then after the train has passed over the insulated
section the signal falls back to the 45-deg. position, and the signal
a t opposite end of block goes to horizontal position. As each train
passes out of the block the release relay operates and in turn actuates the controller a t the entering end of the block one step back
toward its normal position. As the last train passes out the controller switch is operated which completes the circuit through the
signal a t both ends of the block, causing both to operate to
the vertical position, which indicates that all trains are clear of
the block.
Insulating Joints. Where, as a t the end of a block, it is necessary to insulate adjacent rail ends from each other, a piece of insulating fiber is placed between the rail ends, insulating bushings are
placed around the bolts and insulating fiber is clamped between the
rail and the joint plates, the latter having been planed down or
specially made for the pu
Effect of Zinc-treated les on Two-rail Track Circuits. The
following is the consensus of opinion obtained in letter replies. I t
was given by the Committee on Signals and Interlocking, 1913,
A.R.E.A., and accepted as information by that association: (I)
Track circuits a mile in length are rendered inoperative by the
extensive use of zinc-treated ties. (2) Track circuits 2 0 0 0 ft. in
length may be operated successfully, even with SO per cent or
more of ties so treated. (3) 10 per cent to 15 per cent renewals
per year will not materially affect such length circuits. (4) Where
renewals are made of 15 or 2 0 adjacent ties, the leakage is much
greater than where they are made singly a t uniform distances, i.e.,
with 15 per cent renewals (every sixth or seventh tie). ( 5 ) While
the surface salts are present, more leakage occurs during wet
weather than with untreated ties, as these wet salts form a
better conductor than ordinary wet wood. (6) In dry hot weather,
the salts are drawn to the surface and const~tutea more or less
perfect conductor. (7). After a period varying from 3 months
to a year, these salts d~sappearand subsequently no interference
is noticeable.
Interlocking. Interlocking plants are assemblages of switches,
switch-operating and switch-locking devices, and signals so interconnected that their movements must succeed one another in a

y.

ELECTRIC RAILWAY HANDBOOK

redetermined order. They are used a t crossings or junctions

of one line with another, a t drawbridges, a t cross-overs, or in
yards where the number of switches or the frequency of their movement is so great that hand operation would not be sufficiently rapid
nor safe. At railroad crossings or drawbridges, signals are frequently
interlocked with derailing devices, this protection being required by
the laws of several states where trains are permitted to pass
over without stop ing. Interlocking plants are either mechanical
or power-operated: In those of the former type, pipes or wires
connect the switches, locks, signals, or other operated units with the
levers that operate them. In the latter type, pneumatic cylinders
or electric motors are connected to suitable lock and switch movements or signal mechanisms. Electric circuits controlled by the
levers in the interlocking machine regulate the operation of these
switch and signal mechanisms.
In all interlocking machines the desired sequence of operations is
secured by means of the dogs, bars, or tappets which constitute the
mechanical locking. These parts are so interrelated that if any
lever in themachine is reversed (that is, moved from its normal
position), the act of the signalman in unlatching this lever will
cause parts of the locking so to operate that no other lever in the
frame can be moved which would allow a train movement conflicting with the train movement controlled by the first levef- The
mechanical locking between the levers of the machine also rovides
for the movement of the levers in proper sequence when t\e route
for a train is being set up, by assuring that the switches must first
be properly set and must then be locked in the proper position before
the signal governing the route can be cleared.
In power-operated interlocking machines where the only physical
connections between the operated units and the levers in the machine
consists of insulated conductors which carry small currents, there is
no direct assurance, when a switch lever is operated, that the switch
itself has moved. To insure that it is moved properly and will
be locked in a position corresponding to that of the lever, all powerinterlocking systems provide for what is called a return indication.
When the lever in the interlocking machine is moved to cause a
switch or signal to be thrown, it cannot a t first be moved through
its com lete stroke nor through su5cient length of stroke to
release t i e mechanical locking between itself and conflicting levers.
The lever can only be moved far enough to close the circuit which
will cause the desired motion of the switch that is to be operated.
The switch mechanism itself must complete the throwing of the
switch, must completely lock it in the proper position, and must then
close contacts that will cause current to flow back to the lever in
the tower and energize a magnet to release a dog, which has heretofore limited the motion of the lever. After the dog has been
released the operator is permitted to complete the stroke of the
lever, which will in turn release the mechanical locking between
the switch lever and others in the frame, allowing movement of
the latter to be made in roper sequence.
Long bars of steel, calPed detector bars, are held by clips along
the outside of the head of the rail a t switches to prevent operation

DISPATCHERS' SIGNALS

777

They are of a length greater

than the maximum distances between any of the adjacent wheels of
a train, and are so mounted that they must move upward above
the level of the top of the rail before the switch may be unlocked.
They generally are mechanically connected to, and operated simultaneously with, the lock plungers which pass through holes or notches
in the lock rods of switches to lock them in proper position. This
method of protection may prove unsatisfactory owing to the
increasing width of rail heads, and especially on electric roads,
the relatively narrow wheel treads of which may fail to engage
the top of the detector bar while the vehicle is passing over it.
Electric track circuits controlling electric locks mounted on the
levers of the interlocking machine itself, are becoming quite generally used as a substitute for detector bars in the prevention of
switch movement while trains are passing.
Dispatchers' Signal Systems. These are systems in which provision is made whereby a dispatcher may set signals a t desired points
along the track. Telephone communication between train crew
and dispatcher is also generally provided for. The main point of
difference between dispatchers' signal systems which set fixed
signals is the method whereby the desired fixed signal is selected.
Impulse Systems. In the Blake signal system several signals
are connected to the same circuit and each signal contains a pendulum of a length differing from the lengths of all other pendulums in
the signals on that circuit. In the dispatcher's o3ice there is a
pendulum of the same length as each of the pendulums in the signals.
When the dispatcher sets one of his pendulums vibrating it opens
and closes an electric circuit, thus sending impulses of current out
on the signal circuit. These impulses of current pass through
electromagnets in the signals and, in a few seconds, the pendulum
of the length of the one started by the dispatcher is made to vibrate
by the electromagnet to such an extent that it mechanically trips
the release which sets the signal. As the signal sets it closes a
sounder circuit to the dispatcher's office, thus giving indication
that the signal has been set. There are other signal systems in
which a spring or weight-driven selector in the dispatcher's office
sends out lmpulses which will operate a desired signal. The selector
signals are generally reset manually a t the signal or by cooperation
of the dispatcher and someone at the signal.
Crossing Protection. Automatic highway crossing protection
may be afforded by an audible or a visihle signal, or by a combination of these, to warn highway tra5c. Which should be used at a
given crossing depends upon the tra5c conditions on both the
railway and the highway, also upon the particular surroundings of
the situation. The audible signal is given by either agong havinga
frequency of about 2 0 0 strokes per rninutc or by a horn. The visible
signal may be given in the daytime by a semaphore or swinging arm
which is painted so that it is conspicuous against its background
and a t night by electric light or oil lamp-the latter being arranged
with screens operating to expose the proper danger signal. To
avoid unnecessary stops or slowdowns a t a crossing, a signal
operating with the highway protection device is sometimes placed
of the switch while a train is passing.

778

ELECTRIC RAILWAY HANDBOOK

to indicate to the motorman whether or not the crossing device has

operated. Such a signal is commonly arranged to indicate "danger" normally and "clear" when the crossing device is operating.
Crossing protection signals may be operated by any of the means
used for ordinary block signals with the exception that in any particular installation the means employed must be such as not to
interfere with the proper working of the block signal system.
The three-aspect "automatic flagman" of the Union Switch and
Signal Co. is shown by Fig. 25. This signal indicates the
approach of a train by swinging a red banner on which appears
the word "Slop," and displaying a red light attached to the
banner. When no train is approaching the banner is held to
one side as shown by the dotted lines, between two screens upon
which is painted "Look and Listen," and the lamp suspended
from the banner is not lighted. I f
the circuit through the holding magnets is broken, but the apparatus is
otherwise in good condition, the
banner will swing irrespective of
the approach of a train. If the
circuit is broken through the operating magnet but not through the
holding magnet the banner will be
retained in its extreme position
between the screens until a train
approaches, when i t wiU be released and ultimately assume a
vertical position with the banner
stationary but fully displayed. If
the current is totally cut off the
mechanism or if theoperatingparts
PIG.as.
become disconnected, the banner
will also assume the vertical position and be fully displayed. The operating mechanism is enclosed
in a waterproof case and consists essentially of operating magnets for driving the swinging banner and holding magnets for
retaining the banner a t one end of its arc of travel. A circuit controller provides for the selection between the pairs of operating
magnets. The signal is designed to operate normally a t ro volts
D.C. I t requires a n average of 0.4 amp. for swinging the banner,
and 0.4 amp. for the 5-watt 12-volt lamp, making an average of
0.8 amp. drawn from the battery while the banner is swinging.
The holding magnets are of 1000 ohms resistance and require
normally 10 mil-amp. when the flagman is latched in the clear
position. The control is so arranged that i t is unnecessary to break
the operating circuit through relay contacts; the only current passing through these contacts is that of 10 mil-amp. required for
holding the mechanism in its latched position. The signal can be
equipped with a bell and with either an oscillating or fixed lamp.
If desired, the fixed lam can be arranged to burn oil and to give
flashes of light as the ganner swings to and fro, as shown a t B,
Fig. 25.

ROAD CROSSING PROTECTION

779

The signal as described above is controlled by the continuous

track circuit principle, which may be considered as the most reliable,
where it is feasible to use it. Several schemes using other methods
of control are described below.
Trolley Switch Contact. A setting switch is placed in the trolley
wire a t the approach to the crossing and a restoring switch is placed
in the trolley wire at the crossing. The trolley wheel of a car passing the setting switch closes a momentary contact to energize the
setting magnet of a relay in the crossing mechanism, causes its
armature to be attracted and mechanically latched in position.
This operation closes a permanent feed from trolley across contact
points on the relay to the bell and through resistance to ground.
Upon reaching the crossing, the trolley wheel on the car closes a
momentary contact in the restoring switch, which, being connected
to the restoring magnet of the relay, causes its armature to be
attracted; which operation unlocks the mechanical latch above
mentioned, allowing the armature of the setting magnet to fall by
gravity and opening the trolley feed, thus restoring thesignal to the
inoperative condition. The circuits may be arranged to illuminate a danger sign at the crossing or for other lamps or light arrangement such as the use of red lights on the highway each side of the
crossing
as will best suit the particular
- or such other arrangement
case.
Third Rail Contact. (Sedwick.) A section of third rail is laid on
the side of the track opposite to that of the regular third rail which
supplies the train propulsion current. When a car passes this
point, the contact shoes of the car, being all four in parallel, make
contact between the regular third rail and the above-noted insulated section on the other side of the track. There is a circuit
from the insulated section through a relay and thence to ground.
The relay has two sets of coils. When a car is approaching the
crossing it passes over the section of rail which makes the contact
so as to energize one pair of coils in the relay. The armature of the
relay is then pulled over so as to make contact and establish a circuit from the regular third rail through a bank of five lamps to
ground. An alarm bell is connected in shunt with one of these
lamps. The lamps are lighted and the bell is rung until the car
passes a second section of third rail which is beyond the crossing.
In passing this section a circuit is established through the other
pair of coils on the relay, which results in opening the alarm bell
circuit and putting the signal out of operation until again energized by a car approaching the crossing.
Plunger Contactor. (Sauer and Johnson.) In this signal, the car
wheels, in passing, depress the plunger of a track instrument,
located close beside the rail. This plunger is held up by a stiff
spring and can be depressed only by the car wheels. The car
wheels, in depressing this plunger, operate a lever which pushes an
armature up in a solenoid. This armature, when at the upper
limit of its travel, makes contact between two fingers, thus completing the circuit from the trolley or third rail through the solenoid coil. The coil thus being energized, holds the plunger u of
itself. The same circuit passes through the alarm bell and ban! of

ELECTRIC RAILWAY HANDBOOK

lamps. The alarm bell circuit is broken by the passing of the car
wheels over a similar track instrument on the other side of the
crossing, which momentarily breaks the circuit and causes the
solenoid of the track instrument on the other side of the crossine
to let go.
Oscillator. (Protective Signal Manufacturing Co.) The oscillator is a contacting device enclosed in a waterproof iron casing, so
arranged that it may he fastened mechanically, in a rigid way,
to the base of the runnin rail. I t consists of a spring arm held at
one end and carrying at t i e other end a magnetic coil in series with
the line circuit connecting the oscillator and the signal mechanism
a t the crossing. Normally this contact is broken, but the movement or swing of the arm by vibration causes it to open the circuit
momentarily a t its outer end and thus start the operation of the crossing bell. The bell is kept ringing during the interval between the
time that the train passes over the oscillator and the time when it
reaches the crossing by means of a magnetic timer consisting of a
ratchet wheel revolved by means of a magnetically operated pawl
and retent, connected in series with the bell circuit. This is
usually set to operate for 30 seconds after the oscillator has closed
the line circuit.
Magneto Mechanical. (Hoeschen Manufacturing Co.) This
apparatus operates independently of an external supply of electric
current. One end of a lever pivoted near a rail extends under the
rail. The other end carries the armature of a pair of induction
coils having permanent magnets for cores. The passage of a train
over the outer end of this lever depresses it, thus lifting the armature from the magnets. This induces a momentary electromotive
force in the induction coils and the resulting current is conducted
through line wires to the release magnets which control the bell
motor a t the crossing. The bell motor consists of a gear movement of three wheels used in c~nnectionwith three motor springs.
These springs are secured to the main driving shaft of the motor
and are wound by a rod connected directly with a lever resting
against the underside of the rail a t a point opposite the motor.
When the rail at that point is depressed by the wheel of each car of
a passing train, a reciprocating motion of the winding rod (motor
lever) is obtained, thus winding the springs and restoring the releasing lever which controls the motor. The connecting rods, extending from the winding rod to the pawl plates on which are
carried the dogs for turning the ratchet wheel are arranged to wind
the ratchet wheel on both the upward and downward stroke of the
winding rod. The passing of one car will mind the ratchet more
than 4 in. The motor gearing is so arranged that less than this
amount of winding is needed to ring the bell for the passing of two
following can. The motor stores sufficient energy to ring the bell
for abcut loo cars before rewinding is necessary.

Automatic Train Control

The necessity for an automatic device to stop a train increases
as train speed is increased and as headway is shortened, hecoming
greatest where the possibility of failure to promptly see and obey a

AUTOMATIC TRAIN CONTROL

781

danger signal is greatest. Automatic train stops are generally

so arranged that the train may "key by" in an emergency. That
is, a member of the train crew may arrest the operation of the stop
in order to let the train pass into the block protected. Aside from
the location of regular stopping places, local conditions of traffic
and right of way and the type and method of signaling in use,
the climatic conditions under which the train stop is to operate
must be given serious consideration in deciding upon the proper
train stop for a required service.
The Joint Committee on Automatic Train Stops, American Railway Association, Nov., 1913, called attention to the necessity of
studying an automatic train control system with a view to avoiding the introduction of new elements of danger which might be
greater than those which it is the purpose of the installation to
overcome. This was because no automatic train control device, so
far as was known, could be universally applied without adding
elements of danger in train operation. The following requisites
of construction, installation snd operation of an automatic train
controlling system are from the report of that committee:
Failure of any essential part will cause the application of the
brakes.
Proper operation under all conditions of speed, weather, wear,
oscillation and shock.
A train traveling a t a speed conforming to restrictions may pass
a trip in the tri ping condition without the brakes being applied.
Prevention o f t h e release of bakes after application has been
made until the train has been stopped or its speed reduced to a
predetermined rate.
Stop or speed control accompLished within a predetermined
distance.
Operation without interference with the application of the
brakes by the motorman's valve and without a reduction in the
e5dency of the brake system.
Operation without interference with fixed signals.
In 1924automatic train control is being actually installed by the
steam railroads and there is no question but that in the very near
future large installations of automatic train control will be in service.
The fact that the Interstate Commerce Commission has ordered
49 of the principal railroads to install this system of control, serves
to show the importance which is being~.
placed upon this kind of
protection for tfa5c.
Union Switch and Signal Automatic Train Control from Continuous Track Circuit. This system provides three fixed swed limitshigh, medium and low, corkponding closely to the thr& indications
"proceed," "caution" and "stop" of the automatic block signaling
system. The cab equipment may include a speed limit indicator
showing the permissible speed limit established by the train control
system, and a speedometer to show the speed a t ~ h i c hthe train is
runmng. The control of the speed is by means of a centrifugal
speed governor, connected by a propellor shaft type of drive to an
axle of the locomotive. The control of the brakes by the governor
is purely pneumatic, by means of the same type of valves as are

782

ELECTRIC RAILWAY HANDBOOK

used in existing air brake equipment. The governor is controUed

by two magnets, which are, in turn, controlled by a three-position
a-C. relay on the locomotive, designated the "train control" relay.
This relay is of the same type and functions in the same manner as
the track relay of the block signaling system. The system involves
no clearance difficulties, as no apparatus is required on the trackway
and the engine-carried apparatus is within the equipment clearance
lines. Brakes are applied if the train passes a signal indicating caution. However, there is also a valve in the engine cab, which can be
operated by the engineman to prevent the train speed being unduly
restricted by a brake application, provided he operates it just b e f i e
passing a signal indicating caution; thus, a t each caution signal the
engineman is given the opportunity to show that he is awake and
alert, and, by so doing, retains control of the train so long as he
functions properly. If the engineman fails to acknowledge a caution signal by the operation of the valve, the train is automatically
brought to a stop. Inasmuch as the information is conveyed to
the engineman a t the caution signal, a service brake application
only is required and given. This system does not make use of
automatic emergency brake applications under any condition, but
it does not prevent the engineman himself making such a n application in the regular manner. The use of the closed circuit principle
gives a foundation for obtaining maximum safety. The section of
track which governs the train includes the track rails directly in
advance of the train, which are used to conduct the signaling current
which operates the train control apparatus located on the engine.
Plnnser Operatin((

Operation

-Train to Paso

FIG. 26.-Automatic

train stop. Pennsylvania tunneL

Union Switch and Signal Automatic Train Stop. Fig. 26 shows

the type of automatic train stop used on the Pennsylvania Tunnel &
Terminal installation in New York. This stop consists of a tripping
arrangement on the roadway and a valve under the car which can be
operated by the tripper to set the brakes. The tripper arm carrying
the tripper, held in position by two helical springs, is mounted
on a rocker arm and is operated in connection with the signal so
that when the latter is a t "clear" the tripper arm is depressed, by a

TELEPHONE DISPATCHING

783

solenoid or compressed air, against the ties and thus holds the
tripper out of the path of the train mechanism. When, however,
the signal is a t "danger" the tripper arm takes the vertical position and upon being .struck by the cushion spring on the train
mechanism the tripper revolves about its axis on the tripper arm
and operates the valve in the air brake system as indicated in
the figure. The tripper is of light and flexible construction in order
that it may not be broken by fast trains. Where weather or other
conditions prevent satisfactory operation under the train the stop
is placed overhead.
Suggested Clearances for Automatic Train Stop. Fig. 2 7 shows
the limiting clearances for automatic train stop as suggested by the
191I Committee on Heavy Electric Traction, A.E.R.E.A.

PIG.27.-Clearances for automatic train stops.

Telephone
General Methods of Dispatching by Telephone. Where train
dispatching is done by telephone, the telephone is either contained
in a booth beside the trackor it is carried on the train. In the latter
case, connection with the line is made to terminals situated beside
the track or by the connections provided for the cab signal. Where
the telephone is the only method of signaling, the motorman or conductor of a train calls the dispatcher from whom he then receives his
running orders. Where a dispatcher's signal system (see p. 777) in
addition to the telephone is used, the dispatcher sets a signal a t
which the train stops and the crew communicates with the dispatcher. The motorman or conductor may receive the order and
repeat it to the dispatcher directly or the motorman or conductor
may receive the order, write it, meanwhile making a carbon copy
which he gives to the other who reads it back to the dispatcher.

ELECTRIC RAILWAY HANDBOOK

The latter method insures the least chance for error in receiving
the order.
Telephone Dispatching in City Operation. The following is a
general outline of the system of telephone dispatching used in:
Rochester, N. Y., where, due to congestion, traffic handling is
unusually complex: The chief dispatcher and several assistants
are stationed a t a special dispatching telephone switchboar&
located in the railway company's office building. This switch^
board is connected by means of wires leased from the tele~hori;
company to telephone sets located at every line terminal an&
a t main intersections. Telephone sets are also located a t canac:
lift bridges, grade crossings and other points where delays art!
liable to occur. Each motorman is required to report immediate19
upon his arrival a t a terminal. The act of taking the receiver of'
the hook a t the terminal telephone immediately illuminates thc
line lam a t the dispatcher's switchboard. The dispatcher answenthe call pressing a key corresponding with the lamp, which place
him in direct connection with the motorman. After receiving thc
call the dispatcher gives the motorman the time for his departur,
from that terminal and any special instructions that may be needed
The system in this way gives the dispatcher the exact location o
all cars and advises him of any gaps in the service. I n case of
gaps he can order out trippers to fill in until service again becomes
normal. The switchboard used a t t h e dispatcher's oftice is
made up in the form of a flat-top desk hav~ngmounted on it three
turret apparatus cabinets. Two turrets include straight dispatching line equipments, while the third is a combination dispatcher's
and commercial board. During the rush hours three dispatchers
are required, but ordinarily one or two are sufficient. Twa of the
turrets contain forty dispatching lines each, while the third carries
twenty dispatching lines and also a multiple of sixty commercial
lines appearing in the railway company's private branch exchange.
This arrangement permits the night dispatcher, whose duties are
light, to care for the commercial business, thereby eliminating a
night operator for the commercial or regular private branch
exchange board. Space is provided on the top of the desk for schedules, train sheets, etc. A tier of drawers accommodates past
records, special schedules and a supply of dispatching stationery.
The train sheets, arranged by tram number and also 1n chronological order, are always before each dispatcher so that the whole
trafficsituation a t any moment can be read a t a glance. Not only
is the best of service under normal operating conditions assured by
this system, but a means for rapidly straightening out traffic tangles
caused by delays is also ~rovided. The system gives the railway
company a permanent record of every train movement, late cars and
times of accidents. There is also available in cases of court actions
much exact testimony which woxld otherwise be based upon the
inaccurate memory of witnesses. Furthermore, train crews
promptly report trouble with c a n and give many details which they
would not bother to make out in written form.
Local Battery Telephones. h c a l battery telephones may be
divided into two general classes, namely, series and bridging.

py

TELEPHONE INSTRUMENTS

785

Swies Telephone. (Fig. 28.) Series telephones give good service

where there is only one telephone per line. They have been used
where there are several telephones per line by connecting them in
~erieswith the line, but for such service bridging telephones are
more satisfactory. The hook-switch, H, is shown in its raised position, so as to connect the receiver, R, and the secondary, S, in
eries in a circuit between the binding posts, LL, of the instrument;
nd a t the same time the transmitter, T, the primary, P, of the
ducti ion coil and the battery, B, are connected in a local circuit by
aemselves. This is the condition for receiving or transmitting
peech. When the hook is depressed, as when the telephone is not
I use, the bell or ringer, R', is connected across the binding posts,
L. This circuit would also include the generator, G, but for the
rct that it is normally shunted out by the springs, g and g'. The
ontact between these springs is automatically opened, however,

P I G . 28.-Series

telephone.

F I G . 29.-Bridging telephone

when the generator is operated and then thcre is current from the
generator to the line through the bell, R'. The springs, gg', form
what is called an automatic shunt for the generator, their function
being to remove the resistance of the generator armature from the
circuit a t all times save when the generator is in use.
Bridging Telephone. (Fig. 2 9 ) Bridging telephones give good
service where it is necessary to have several telephones per line.
For this service these instruments are bridged across the h e , that
is, they are connected in parallel with each other. In the bridging
telephone the arrangement of the receiver, induction coil, transmitter and battery is identical with that of the series telephone
(Fig. 28). The ringer, however, is bridged permanently across the
line and the generator is placed in a circuit across the line which
is normally open, but which is closed automatically by the spring,
g', when the generator is operated.
Telephone Disturbance. Annoying inductive effects of adjacent
alternating currents may be sufliciently reduced by transposing the
telephone line wires relatively to each other so that the influences on
one side of the line shall oppose and nearly equal the influence on
the other.

INDEX
PAGE

Acceleration ................. 150

coefficients charts of ........ 177
economical rates of . . . . . . . . 153
e5ciency of . . . . . . . . . . 198
energy . . . . . . . . . . . . . . . . . . 153
fluctuations of current in
328
formula . . . . . . . . . . .
151
graphical representation of .. I52
m a u m u m speed . . . . . . . 155
ratio of linear to total ...... 1.50
relation t o speed time distance
... 151
speed a t end of straight line . 151
straight line. per cent of .... 199
usual rates of . . . . . . . . . . . . 153
we~ghttransfer in . . . . . . . . . 158
Accelerometer . . . . . . . . . . . . . . 156
Adhesion. coefficient of ....... 157
A.I.E.E. standards on railway
motors . . . . . . . . . . . 2I7
transmission and distribution 547
Air brake. automatic .
469
combined straight and aut&
matic . . . . . . . . . . . . 473
emergency straight . . . . . . . 467
hose . . . . . . . . . . . . . . . . . 483
storage system ............ 480
straight . . . . . . . . . . . . . . . . . 435
types of . . . . . . . . . . . . . . . . 454
Air compressor capacity ...... 480
inspection . . . . . . . . . . . . . . . . 478
lubrication ............... 479
overhauhng ................ 479
tests ...................... 479
types of ................... 481
Air coolers ................... 481
friction .................... 140
gap practical minimum ..... 31I
pjpe aluminum . . . . . . . . 521
plping t o prevent freezing
482
reservoirs. capacity.. . . . . . 481
resistance . . . . . . . . . . . . 128
used in braking ............ 480
Alloyed steel rails . . . . . . . . . . . 43
Alternating current see Three
phase Slngle phase etc.
Aluminum air pipe . . . . . . . . . . 521
compared t o copper conductor .................. 716
field coils .................. a87
in car construction . . . . . . . 520
wire sag tables
. . . . . . . 739
=.ire tables .
. . . . . . 731
Anchors in earth and rock . . . 553
Arc welded rail joint
. . . . 51

..

. .

Pacs
Arc welders .............. .76. 1x5
Area. ground shop
.
. 107
relative shop departments
109 Ira
shop. relation to number of
c a n . . . . . . . . . . . . 106
Armature banding
. . . . . . 181
bearings . . . . . . . . . . . . . . 309
clearance . . . . . . . . . . . . . . . 311
leads broken ............
184
record tags . . . . . . . . . . . . 285
removal from box frame
motor . . . . . . . . . . . 179
shaft assembly . . . . . . . 306
shaft renewal . . . . . . . . . . 307
speed. car s p d . wheel diam
gear ratlo . . . . . . . . 245
data on commercial motors 153
tests
188
windings and connections ... 180
Armstrong formula for train
resistance . . . . . . . . . . 128
Axle breakage . . . . . . . . . . . . 397
capacities ............... 396
design. standard . . . . . 395 396
diameter maxlmum commercial motors ......... 253
hollow . . . . . . . . . . . . . . . . . 393
specifications .............. 394
tests for cracks or flaws . . . . 398

..

..

......................

.............

Babbitt bearings
311
Ballast ...................... 14
depth of ................... 15
. sections................... 16
Ball bearings . . . . . . . . . . . . . 309
Beanng fnctlon. armature . . . . 219
losses in railway moton .
218
thrust plates. standard . . . . 416
wear. armature
......... 311
Bearings armature ......... 309
babbitt . . . . . . . . . . . . . . .
311
ball . . . . . . . . . . . . . . . . . . . . 309
journal standard . . . . . . . . 426
motor .................... 309
roller . . . . . . . . . . . . . . . . . . . 309
495
Birney safety car ..........
Blacksmith shop. car house 103
Block signals . . . . . . .
753
Blood formula for train ksistance . . . . . . . . .
129
B o g ~ etruck . . . . . . . . . 385 389
Bonds. bolted terminal . . . . . . 658
brazed . . . . . . . . . . . . .
646
classification ............... 645

787

.
.

INDEX
PAGE
Bonds, compressed termanal
645
cross
657
double
656
economrc re lacement pornt 653
electnc w e d
646
farlure of
647
rdeal
647
~nducttve
773
rnstalktron
659
length of
648
manganese rall
647
oxyacetylene weld
646
rail
645
resrstance
648
s ~ z eof
652
soldered
646
speclal work
657
Bond test car
663
testrn
660
BOW trotey
378
Brake. am. types of
464
160
automatrc arr
band
393
clasp
459
combmed strarght and automatlc alr
473
Y m E r t r o n for vanable
452
cyhnder. standards
455
efficacy
441
e~ectx+ntumat>c
47 4
emergency stra~ghtalr
467
hand, matntenance
t v ~ r c a schemes
l
vs arr
i65
hanger angle
447
head lrmrt gages
494
lnspectlon
477
leverage. plston area and
pressuie
lever standards
mametrc
.psibn area. alr pressure and
leverage
454
travel
457
n g ng calculatrons
448
g u b l e truck
389
effic~ency
srngle truck
standards
safety stop
shoe clearance
fnctron and applrcatron
dlstance
a t vanous speeds
l r m t gages
posrtlon on wheel
pressure
actual nomlnal
standard
suspensron
vanous types
wear
slack adjusten
stralght arr
vanable load compensation
Brakrng
affectrng wheel fa~lures
alr used

--

2%

PAGE
Bralong. apphcatron trme
438
buclung motors
486
&stance
436
chart
442
emergency
432
rn practrce
436
energy consumed
194
power, actual. nomrnal
439
rate and energy
I54
regeneratrve
203. 485
retardation by mven force
442
~n tests
443
reversed motors
486
squeahng m
487
werght transfer tn
444
Bndges. comblnatron rallwaY
and hlghway
24
Bruckner's curve
4
Brush contact resrstance
252
holder
275
spnng pressure
275
Brushes. carbon
276
Buckrng motors for brahng
486
Burldrngs. shop srze of
106
Bumper, car. standard height 532
track
98

Cabrnet c o n t ~
352
Cab\c. a\\owab>e current ca+rtres
735
armor speclficatron
620
control srzes of wlres
336
copper. werghts. reststance.
strandrng
730
Ice and wrnd loadrng tables
736739
paper rsulated. specrficatlon
622
sheath lead specrficatton
620
strandrng table
730
Car bodles electnc ho~stsfor
118
capacrty
505
constructron
519
alumrnum rn
520
couplers
526
desrgn
495
door. automatrc
499
operatron
equrpment for power shoves
framlng
heatlng thermostats
vanous methods
house blacksmrth shop
clearances
desrgn
86
doors
electnc ho~stsfor
heatlng systems
lrghtrng
lobby
or1 room
parnt shop
p1ts
roofing schemes
srgn room
special track work
track layouts

IND
I

Car house. wash room

mnng
l~ghtlng
palntlng
seat arrangement
ventllatlon
wexghts
and operating costs
mnng
motor leads
reslstance of
size wtres
Cars. appearance. Importance
of

articulated

center entrance and exlt

dlmens~ons.typical
&stance between
double deck
double truck l ~ g h welght
t
drop platform
end connectlons
express and frerght
front entrance. center exlt
laterurban
m~scellaneous equipment
for
l ~ g h welght
t
motor capacrty, typlcal
~srnbn.7
a d dun ana
one man
495.
w r m~le
P e t e r - ~ ~type
tt
r a p ~ dtransst
stngle end, decreased we~ght
single truck
llnht welaht
steeT
we~ghts
Carts
Cascade control
Cast weld rall lomts
Catch sldmgs
Catenary brackets
bndge
COnStmCtlOn
class1ficatron

hangers
lengths
spaclng
messenger. sag and tens~on
span. length
trolley temperature effect
~haractenstlc curves motor
220.
average or general
changes In gears wheels or
formZtaf:r
approxlmatlon
slrmlanty of
s l ~ d erule approxlmatlnn
Chord deflect~ond~stances
Chrome nrckel steel
Clrcut breakers
Clark rall lomt
Clasp brake

PAGE
Clearances armature
automatic tram stop
brake shoe
car house tracks
gear case
pantograph trolley.
pole
shoo track
s~gdal
Cleanng nght of way
Coastmn. effect of
energy consumed tn
speed tlme curve. construetlon of
txme recorder
Coefficrent of adheslon
Commutatlng abll~tyof motors
poles
Commutator connect~ons
leads. soldennn .
mate&
conitruct~on care
repatrs

311
783
457
86
246
549
550

I13
764
a
a12
I94
175
212
157
224
266
280
282

undercutting

C o m w s ~ t ~ oof
n rail metal 42.
Compromise rall jotnt
.
Concatenated motors. synchronous speed of
Concatenation of lnduct~on
mtncs

Concrete, electrolysis In
mlxlng apparatus
78
poles
603
alumlnum and
Conductor.
copper
726
contact. A I E E. defin~t~on
547
rail. see T h r d rall
spacmg. transrmss~on
699
Condult. duct
625
Contact conductor. A I E E
debtlon
547
rall. A 1 E E. defimt~on
547
underground
643
Contactor, auulhary
326
control
338
Control apparatus lnspectlon
and overhauling
359
automatic
340. 347
cab~net
352
cables, size of mre
336
cam
341
cascade
35i
comb~neda c and d c
353
concatenation
357
contactor
338
hlgh voltage d c
321
~nducttonregulator
355
pneumatlc
343
power operated
337
resistance connections
320
des~gn
327
graph~calmethod
332
gnds, current capacrtles
335
usual values
334
rheostat~c
3x7
serles parallel
317
slngle phase
352
three phase
356
three speed
319

'DEX

PAGE

Control transition shunt and

bridging ................ 319
AB .................. 350
. .................... 348
B ...................... 319
HB ..................... 350
HL ..................... 345
K ...................... 317
..
L ....................... 317
PC ..................... 341
Controller data .............. 323
dimensions ................ 32j
finger contact pressure ...... 362
lubrication . . . . . . . . . . . . . . . . 361
rating . . . . . . . . . . . . . . . . . . . . 321
types commercial . . . . . . . . . . 323
weights . . . . . . . . . . . . . . . . . 324
Copper handling of .......... 297
resistivity standard ........ 727
sickness................... 298
wire sag tables ............. 739
wire tables ................ 727
Core loss. motor ............. 2x9
Corrugation. rail ............. 53
Coupler automatic ........... 526
he1 ht standard ........... 531
type, standard ...... 529
Cranecar ................... 78
electric car house .......... I I 8
Creosote ireatment of wood ... 749
d , k t e d - m m m t ...... n;Cross arms . . . . . . . . . . . . . . . . . . 555
Crossing protection ........... 777
Culvert openings .. .......... 20
Current, allowable ~n wires and
cables ................... 735
heating, measurement of . . . . 152
jn lead cable sheaths and pipe 685
In plpes ................... 675
in track average and r.m.s. 656
-time curves . . . . . . . . . . . . . . . 18s
time recorder .............. 212
Curve protection. signals . . . . . . 753
resistance . . . . . . . . . . . . . . . . . 144
track. designation of ........ 55
flangeway. . . . . . . . . . . . . . . 62
gage . . . . . . . . . . . . . . . . . . . . 62
grade compensation for . . . 145
minimum radius for given
truck . . . . . . . . . . . . . . . . . 418
special rail sections for .... 41
superelevation of rail ..... 56
tan ents, to connect ...... 59
to &d degree of ......... 58

TyE

.
. .

.
M.E.B..

Definitions transmission and

distribution. standard ....
wire and cable standard . . . .
Derailing switches . . . . . . . . . . . .
Dinkey locomotives. . . . . . . . . . .
Distance determination from
speed time curve . . . . . . . . .
relation to speed. time acceleratlon ................
-time curve ................
graphical construction . . . .

547
61I
70
18
I 84

I52
184
184

Distribution system. feed,"

design . . . . . . . . . . . . . . . . . .
Doors. car. automatic ........
operation of ...............
Doors car house .............
Double truck . . . . . . . . . . .
Drag -pen
............ J.8.5a:
Dratnage system pipe ........
Drains track ................
Drifting. see coasting
Duct condmt ................
p m p can ..................
ura umin alloy .............

lob
4*

S26
98
8 9

31
"578
sax

Earth ammeter Bureau of

Standards ............... 688
Haber ....................
Earth resistance ............... 687
Earthwork freehaul .......... '564
loosenin and handling .... 4
overhauf ...................
I
shrinkage . . . . . . . . . . . . . . . . .
Ease.ment curves .............
Economics of railway location
Economy watthour meter .....
Efficiency of acceleration ...... 'I1
Electric hoists for car bodies ... t98
Safety Code. National ......
&a&
................ ,-,:,6
track switches .............
traction heavy ............
weld rail joint .............
Electrification. railroad .......
single phase ...............
three hase
....... :
Electroc emcal eqmvalents . . .
Electrode. normal. Haber .....
Electrolysis ..................
alternating current .........
concrete . . . . . . . . . . . . . . . . . .
corrosion rates of . . . . . . . . . .
current measurements . . 674
mitigation .................
potential merely indicative . .
survey ..................
self corrosion ..............
tests ......................
transvme .................
Elevated cars ................
Energy. acceleration ..........
braking ............... 154
coasting . . . . . . . . . . . . . . . . . . .
consumption
approximate
determination ..... 196 I
charts . . . . . . . . . . . . . . . . . .
effect of stops . . . . . . . . . . . . . . 1 9 ~
loss.in rheostat .............
savlng b y field control
moton . . . . . . . . . . . . . . . .
by grades a t stations . . . . . '28
train movement . . . . . . . . . . . . :pa
Equivalent grade . . . . . . . . 163
St0pS .....................
Expansion allowance a t rail 63
joints . . . . . . . . . . . . . . . . . . .
EX ess c a n . . . . . . . . . . . . . . . . '
.'5
g p r t m e n t quarters . . . . . . . . 1i95

$:

42

.........

.
.

4:

. $2

INDEX

............

Feed wire anchors

567
bare or insulated ........... 567
installation ................
sags ......................
Feeder calculations ...........
drop and capacity chart . . . .
economical voltage drop .....
negative
. . connection to rails
design ..................
sectionalizing switch automatic .................
system design ..............
taps to trolley .............
Fences ......................
snow ...................... 29
Ferro-titanium steel .......... 44
Fiber duct .................. 625
Field coils ................... 185
aluminum ................. 287
testing polarity ............ 187
tests ...................... 291
Field control motors ........... 267
286
arrangement of coils
connections ................ 32 I
Fire hose .....................
91
prevention and protection ... 88
Plange-bearing special work ... 66
Flangeway of track cum- .... 62
Ploor construction ........... 9 8
Foundation track ............ 29
Freehaul of earthwork
4
Freight cars ................. 5 19
department quarters........ 109
Fresno scrapers .............. I o
Friction. air ................. 140
armature bearing. motor .... a 19
brush motor .............. a19
coefficient of ............... I 57
journal .................... 117
starting ................... 143
tem rature effect
141
tracrcurve ......
144
Frogs. track ................. 67
trolley .................... 564

........

........

.':.::I: : ::

Gears and pinions ............ 300

life ....................... 30 1
material ................... a02
Gear case clearance ........... 346
helical .................... 301
inspection and lubrication ... 305
long and short addendum
302
losses ..................... a18
pitch ..................... 300
ratio ...................... 24s
maximum ..............: 246
data
on
commercial
......
motors .............. 253
selection of .............. 161
shrink fits ................. 409
specifications .............. 301
split and solid ............. 301
teeth . soecial desim ........ 302
tooth piessure . . .7 ......... So1
vemer .................. 303

...

Generating station capacity ....

Grade actual and approximate
average . . . . . . . . . . . . . . . . . . .
equivalent . . . . . . . . . . . . . . .
compensation for curves . . . . .
crossing protection . . . . . . . . .
effect on tractive effort . . . . .
equivalent . . . . . . . . . . . . 163
momentum . . . . . . . . . . . . . . . .
ruling . . . . . . . . . . . . . . . . . . . . .
station ener saving by . . . .
Grading classi%tion of ......
Graphs run curves . . . . . . . . . . .
Grinding equipment . . . . . . . . . .
rail jomts .................
Ground sleeve for steel poles
Guy attachments ............
Guys .......................

r9;
146
147
163
145
777
146
176
147
147
201
3
161
76
53
598
552
555

Haber earth ammeter .........

earth current collector . . . . . .
Haber normal electrode . . . . . . .
Hard center special work ......
Headway . . . . . . . . . . . . . . . . . . .
Heaters for cars ..............
Heating current. measurement

687
687
687
65
119
534

..

.
..

of . . . . . . . . . . . . . . . . . . . . . . a51
systrms ~ D Jcar houses ...... 104
bcket booths . . . . . . . . . . . . . . 105
waiting rooms . . . . . . . . . . . . . I05
Hoists electric for cars ....... 118
Hose. air brake .............. 483

Ice and wind loading wires and

. . . . . . . . . . 7.....
76710
cables
..............
Impregnating compounds . . . . . 294
Impregnation of motor rnsula- t i o n . . . . . . . . . . . . . . . . . . . . a96
Inertia ratio of linear to total 150
Induction motor . . . . . . . . . . . . . 139
control . . . . . . . . . . . . . . . . . . . 356
Inductive reactance tables 743-748
Insulated joints in underground
pipes and cables . . . . . . . . . 696
ne ttve return . . . . . . . . . . . . 696
l n s u g i n g joints. rail . . . . . . . . . 775
materials . . . . . . . . . . . . . . . . . . 294
tests . . . . . . . . . . . . . . . . . . . . 295
Insulation impregnation ...... a96
of
underground
surface
structures . . . . . . . . . . . . . 698
test armature coils . . . . . . . . . a89
field coils . . . . . . . . . . . . . . . . 191
Interlocking
switches
and
signals . . . . . . . . . . . . . . . . . . 775
Interpole motors . . . . . . . . . . . . . 266
Interurban cars . . . . . . . . . . . . . . 518
end connections ............ 532
miscellaneous equipment . . . . 532

f..

ack pit ................ IOI 1x5

o~ntlyused poles ............ 611

INDEX
PAGE
Joints. insulated in underground pipes and cables . . . 696
rail . . . . . . . . . . . . . . . . . . . . . . .
arc weld .................
cast weld . . . . . . . . . . . . . . . . AO.
Clark ................... 51
compromise . . . . . . . . . . . . . 52
drilltng standard . . . . . . . . 48
electric weld ............. 51
exnansion allowance ...... iz
f i d i n g . . . . . . . . . . . . . . . . 53
~nsulating............... 775
Nichols ................. 52
opposite or alternate ...... 48
standard . . . . . . . . . . . . . . . . 49
suspended or supported . . . 48
thermit weld ............ SO
welded . . . . . . . . . . . . . . . . . .
io
]ournalb&ring end play . . . . . . d i 6
hooded wedge . . . . . . . . . . . . . 426
lubrication.. . . . . . . . . . . . . . . 427
thrust p!ates ............... 426
Journal boxes. standard ....... 426
. wedges standard ........... 426
Tournal dust mards .......... a27
. friction . . . .............. ia7
packing tools .............. 429
temperatures .............. 420

. .

!:

Pacn
Lubrication motor ........... 312
oil for old type motors . . . . . 313
trolley .................... 373

r.

Kinetic energy head .......... 149

.
.

Lead cable sheath resistance

and current . . . . . . . . . . . . . . 685
specification . . . . . . . . . . . . . . . 620
Leakage resistance track to
earth ................... 693
Length. rail ................. 42
Lightmg. car ................ 542
carhouse ................. 105
shop ...................... 105
Li htning arresters ........... 327
Ene. ground for ............ 569
installation . . . . . . . . . . . . . . 568
Line departmept quarters . . . . . 10.5
I.inemen. qual~ficat~ons
. . . . . . . 569
fine switch auxiliary . . . . . . . . . 326
transmission. calculations . . . 700
Load curves for power or substation . . . . . . . . . . . . . . . . . 193
variable brake compensat.ion 452
Loading. ice and u m d mres
and cables . . . . . . . . . . 736-739
Lobby car house ............. 98
Location. repair shop . . . . . . . . . I I5
substation . . . . . . . . . . . . . . . . . 71 1
Locomotive crane . . . . . . . . . . . . 78
Locomotives
direct current 584
electric ratings of .......... 218
motor capacity. ........... 238
single e s e ............... 585
s lit
se ................ 585
tKree phase . . . . . . . . . . . . . . . . 586
Lubrication air compressor . . . . 479
controller . . . . . . . . . . . . . . . . . 35'2
journal ................... 427

. .
.

Machinery track maintenance 76

Mailloux' method speed time
curves . . . . . . . . . . . . . . . . . . 177
Manganese rail bonding . . . . . . 647
............... 6r
s w c ~ a work
l
steel . . . . . . . . . . . . . . . . . . . . . . 45
Manholes conduit . . . . . . . . . . . 628
Marker lamps. electric . . . . . . . . 544
Mass diagram . . . . . . . . . . . . . . .
1
Mershon diagram . . . . . . . . . . . . 702
Metal rail composition of . . . . 42
third rail comwsition and resistnnce ............... 631
Meter watthour, for car e q u i p
ment ................... 212
Mjller trolley shoe ............ 369
hfrnneapolis light weight car . 393
Motors accelerating current
allowable . . . . . . . . . . . . . . . . 224
A.I.E.E. standards ......... 217
application charts . . . . . . . . . . 227
armature bearing friction . . . . a 1 9
p x n n $ . . . . . . . . . . . . . . . . 309
rame
removal of
armature .............. 279
brush friction .............. 2x0
bucking . . . . . . . . . . . . . . . . . . . 271
for biaking . . . . . . . . . . . . 486
capacity. approximation .... 230
high speed self-ventilated . a13
induction ............... a38
locomotive .............. a38
maximum practicable ..... a29
preliminary determination aaq
service requirements ...... a21
single phase .............. 238
typtca cars
. . . . . . . . . . . 508
characteristic curve.... 220. 241
average or general . . . . . . . . 242
changes in gears. wheels or
voltage . . . . . . . . . . . . . . . a46
formula for approximation 244
s i m ~ l a n t yof . . . . . . . . . . . . . 241
slide rule approximation . 244
commutating ability . . . . . . . . 224
concatenated
synchronous
speed . . . . . . . . . . . . . . . . . 359
connections t o car wi+;'ng.... 299
core loss .................. 219
data on commercial ........ as3
field control . . . . . . . . . . . . . . . a67
connections .............. 321
energy savmg ............ 228
flashing . . . . . . . . . . . . . . . . . . . 277
four or two per car . . . . . . . . . 260
gearing and axle bearing
losses . . . . . . . . . . . . . . . . . 218
gears and ini ions ........... l o o
ihduction . ................ 239
capacity ................ 238
insulation impregnation ..... 296
leads . . . . . . . . . . . . . . . . . . . . . 299
broken. effect of .......... a99

.
.
.
.

.
. ..

INDEX

PAGE
Motors locomotive capacity .. 238
lubrication . . . . . . . . . . . . . . . . 312
oil lubrication of old types . 313
performance eflect of differences
~n
wheel
diameter .............. a50
in series with external res~stance............... 249
on low voltane ........... 207
ratings . . . . . . . . . . . . . . . . . . . 217
commercial .............. 253
resistance. . . . . . . . . . . . . . . . . 251
reversed for braking ........ 496
reversing ........ .T ........ 321
speed characteristic changes
in voltane ............. aa8
selection of . ..............
single phase capacity .......
structural features ........
theory of ................
types of .................
startlng resistance design . . . .
graphical method .........
usual values
.. . . . . . . . . . . .
support. springs . . . . . . . . . . . .
suspens~on.. . . . . . . . . . . . . .
temperature limits . . . . . . . . . .
with trailers .............
three phase ................
two or four per car .........
ventilation ................
voltage and rating relation . .
weight data commercial . . . .
weights, comparative. two
and four per car . . . . . . .
windage ...................
Motormen devices for checking
performance of ...........
Multlple unit control .........

. 609

National Electric Safety Code

Negative connection track t o
feeder . . . . . . . . . . . . . . . . . .
return insulated ...........
Nichols rail joint .............
Normal electrode Haber ......
Nozzles universal on standp ~ p e...................
s

.
.

..

726
696
52
687
89

...........
.
..............

Oil room car house

IOA
One man car ............ 495 499
safety devices
475
Overhaul of earthwork
4

........

Paint shop ..................

Painting car ................
Pantograph clearances ........
trolley ....................
Paper ~nsulated cable specification .................
Passenger capacity ...........
movement a t stops .........
seat space .................

103
524
549
378
622
505
123
503

PAGE
Pavement errooote wood block 751
plow
78
Pinlons ..................... 300
bore taper ................. 391
installation ................ 303
life ....................... 301
specifications .............. 301
Pipe. air aluminum .......... 521
cast iron tables . . . . . . . . . . . . 677
cukent flow in . . . . . . . . . . . . 675
drainags system . . . . . . . . . . . 696
resistance. cast Iron and steel 673
steel and wrought iron tables 683
steel ole tables . . . . . . . . . . . 596
Pitch o f ears and pinions ..... 300
Pits . car
............. 99
jack ................. I O I
shop ......................
wheel drop ............ IOI
grinder .................
Platform height standard
Poles cedar. eastern ..........
western ...................
Poles. chestnut ..............
clearances .................
concrete
framing ...................
joint use of ................
preservative treatment ......
rake of ...................
setting and depth of hole ....
spaclng ...................
speed
determination
by
counting ..............
steel ......................
ground sleeve ............
jomts ...................
for tables ..........
w
s ... ..................
Potentlal drop m rails. . . . . . . .
Power input a t constant speed
operated control ...........
requirements acceleration . . .
power station ............
substation ...............
train movement ..........
shovels .................... 13
c a r e uipment for ........ i 9
stationload curves ......... 193
capacity ................ 192
Preservative treatment poles
cross-arms and ties ..l .... 749
Profile. virtual ............... 148

.....................

..

souse..

.....

..................

Raid alloyed steel ............ 43

bending . . . . . . . . . . . . . . . . . . .
bonding ..................
connection
to
negative
feeder
contact see Third rail .
corrugation ................
grindln equipment . . . . . . . . .
groove% or plain girder ......
joints . . . . . . . . . . . . . . . . . . . . .
arc weld ................
cast weld ...............
Clark ...................

57
645

................ 726
53
76
40
46
51
49
51

INDEX

..

PACE
Rail joints compromise ....... 52
rrllmg standard ......... 48
electric weld ............. 51
expansion allowance ...... 52
grinding ................
53
insulating ............... 77 5
Nichok ................. 52
opposlte or alternate ..... 48
resistance.. . . . . . . . . . . . . . 648
speed determination by
counting .............. 161
standard ................ 49
suspended or supported ... 48
thermit weld ............ 50
welded . . . . . . . . . . . . . . . . . . qp
length .................... 42
manganese. bonping ........ 647
metal, composrtron...... 42 631
potential drop ............. 695
renewals . . . . . . . . . . . . . . . . . . 41
section special for curves 41
standard ................
selection of ................ .
tee. standard
39
third see Third rail
tilted ..................... 42
welding equipment ......... 76
Railless trolley. overhead construction . . . . . . . . . . . . . . . . 583
Railroad electrification ........ $84
Rapid transit cars ............ 51 I
Ratings of motors ............ 2x7
Reciprocals chart of
179
Reaction. inductive tables 743-748
Regeneration
alternating
current . . . . . . . . . . . . . . . . . 207
direct current . . . . . . . . . . . . . . 204
Regenerative braking . . . . 203. 485
Resistance brush contact . . . . . 252
car m n n g . . . . . . . . . . . . . . . . . 252
connections control ........ 320
cop r mre ................ 728
current capacities ..... 335
motors .................... 25 I
motor performance in series
With .................. 249
rail joint .................. 648
metal ................... 631
roadbed ................... 665
soils ...................... 664
starting. design ............ 327
energy loss in ............ 2or
graphics! method ........ 33 I
mstallatlon .............. 335
usual values
334
steel for third rail .......... 631
track ..................... 651
to earth ................. 693
Resistivity. copper standard ... 727
Return. insulated negative . . . . 696
system design . . . . . . . . . . . . . . 720
Reversed motor braking ....... 486
Reversing series motors . . . . . . . 321
Rheostat. see Resistance starting.
Rico motor oiler ............. 313
Right of way ................
a
Roadbed construction .........
contact resistance

Z
.............. ..

gr~g

..........
. .

.............

..........

PACE
Road crossing protection ...... 777
Roller bearings ............... 309
Rolling friction .............. 127
Rooting. carhouse ............. 97
Rubber insulated mrc and
cable s~ecifications....... 613
Run curves: ................. 161
typical .................... 162
S

Safety car .............. 495 499

Code. National Electric . . . . . 609
devices for one man car ..... 475
in bright colored cars ....... 526
Sag catenary messenger ...... 580
f~
.and aluminum wire . . 739
mre .................. 566
span wlres ................. 570
troUey wire ................ 558
Salt storage ................. 104
Sand drvinn
. .................. IOA
storage
track
track for catch siding .......
Sanders .....................
Schedule m
l
. a.~ h i c a...........
speed .....................
Scrapers drag ...............
Presno ....................
wheel .....................
Seat arrangement in cars ......
%=uce p e passenger
~
.........
corroston . . . . . . . . . . . . . . . .
Series-parallel change. time for
Shoo area .. relatlon to number of
can....................
area various departments I09
design .....................
ground areas ..............
heating systems ............
lighting ...................
location ...................
pits ....................... 99
plans. typical ......... 110 I11
relative area of departments ............ I09 I12
size of buildings ............ 106
track arrangement .......... 113
centers .................. I13
transfer table .............. I 13
typjcal plans .......... 1x0 I I
m n n g . . . . . . . . . . . . . . . . . 105
Shovels electric ........... 17 78
power .....................
steam .....................
Shrinkage of earthwork .......
Shrink fits wheels and gean ...
Sidings. catch . . . . . . . . . . . . . . . .
Sign room car house .........
Signals ......................
a.c. track circuit ...........
A.P.B. system .............
aspects standard ...........
automatic block ............
train control .............
clearances . . . . . . . . . . . . . . . . .
crossing protection . . . . . . . . .
curve protection ...........

....................
.....................

.
.

.
.

.
.

INDEX
P

.........
.......
...............

Signals. dispatchers'.
hand operated .............
indications. standard
interlocking
location ...................
manual block ..............
Nachodty pe ..............
position light aspects .......
schemes suburban and interurban . . . . . . . . . . . . . . . . .
semaphore spectacle. standard ...................
standard use .............
short track circuit ..........
sunli ht. discernibleness in
T.D.%, system .............
trolley operated ............
U . S. type .................
Silicon steel .................
Singie phase control ..........
electrification ..............
motors ....................
Single truck .................
Skull cracker ................
Slack adjusters ...............
Slide rule approximation of
motor characteristics . . . .
Slidin contact shoe . . . . . . . . . .
slot
current collector . . . .
Snow fences .................
Soil resistance ...............
Span catenary length ........
wlre ......................
attachments ..............
buildings ................
. sag and tension ...........
Special track work ...........
flange bearing
hard center ................
manganese ................
Speed. acceleration. time distance ...................
armature commercial motors
car and armature wheel
diameter and gear ratio
c h a r a c t e r i s t i c s , motor
changes for voltage .....
desirablllty of . . . . . . . . . . . . .
determination counting poles
or rail joints ...........
-distance curves ............
limi$tions ................
maumum, and acceleration
stops. effect of .............
straight line acceleration a t
end of . . . . . . . . . . . . . . . .
synchronous
concatenated
motors ................
induction motors ........
-time curves ...............
chart of acceleration coefficients ...............
for calculations ........
of accelerations . . . . . . . .
of reciprocals . . . . . . . . . .
of retardations .........
coasting line location .....
data required

..

p70w

.............

............

..

PAGE

speed-time
curves definite
stops grades alinement . 176
distance
determination
.

general . . . . . . . . . . . . . . . . . 165
od . . . . . . . . 173
alllous meth
method . . . . . . . . I77
a;aphical
step-by-step method . . . . . . I67
straight ilne approximation . . . . . . . . . . . . . . . . . . 163
tracing method .......... 182
Spirals track ................ 60
t o l a y o u t ................. 6 1
Spray paintin i3............... 515
Springs. douh e trucks ........ 389
sin le trucks ............... 386
sprinklers automatic . . . . . . . . . 88
Standpipes and universal nozzle 89
Starting see Acceleration.
friction resistance . . . . . . . . . . I43
resistance.
electrical. see
Resistapce. starting .
Station capacity power . . . . . . . I92
grades energy saved by ..... 201
load curves ................ 193
Steam shovels ...............
I3
car equipment for .......... I9
Steel poles
591
p u n d sleeve.............. 598
lolnts ..................... 598
{ails alloyed ............... 43
composition and resistance
.aa..631
.
trestles . . . . . . . . . . . . . . . . . . . 13
trolley wire ................ 5 8 9
Stock room .................. I 0 0
Stops distance betwqen . . . . . . . I13
d~stancetraveled In making 431
duration of . . . . . . . . . . . . . . . . 123
eSect on speed and energy 123
equivalent numher . . . . . . . . . 163
frequency of . . . . . . . . . . . . . . . I13
location of . . . . . . . . . . . . . . . . . I23
Straight line acceleration per
cent of . . . . . . . . . . . . . . . . . . I99
s p e e d a t m d o f . . . . . . . . . . . . 151
Straight line speed-time curves . 163
Stranding table. copper cable 730
Substation capacity . . . . . . . . . . 192
location . . . . . . . . . . . . . . . . . . . 706
Subway cars . . . . . . . . . . . . . . . . . 511
sections .................. 7-75
train resistance in, . . . . . . . . . I34
Super-elevation of rall on curves 56
Surface insulation. underground
structures . . . . . . . . . . . . . . . 698
Switches derailing . . . . . . . . . . . 70
interlocldng ............... 7 1 5
open track ................ 6 8
track ..................... 66
electric .................. 75
Synchronous speed, induction
motors .................. a39
concatenated . . . . . . . . . . . . . . 359

~~

..................

T
Tail lights. electric ........... 544
Tee rails. standard
39

...........

INDEX
PAGE

Telephone dispatching . . . . . . . . 783

disturbance . . . . . . . . . . . . . . . . 78s
local battery . . . . . . . . . . . . . .
Temperature effect on train
resistance . . . . . . . . . . . . . . .

journal bearing ............

Tension . catenary messenger
span mre .................
trolley wire ................
Terminology electric wire and
cable. standard, . . . . . . . . .
Thermrt weld rall joints.. . . . . .
Thermostat for electric heate n . . . . . . . . . . . . . . . . . . . . . 539
Third rail A.I.E.E. standards . 547
clearances standard
630
composition and resistance . . 63 I
gage and elevation 547 548 631
~nsulation. . . . . . . . . . . . . . . . . 638
operation in snow and sleet . . 640
protection ................. 640
k t i o n s ................... 636
shoes ..................... 374
support ................... 637
temperature coefficimt...... 636
Three phase control .......... 356
\ocomotlves ................ 586
motors .................... 239
Three wire distribution
697
Ticket booths. heating ........ 105
Tie rods for track............ 54
Ties. classification ............ 27
preservation ............... 749
iizc . . . . . . . . . . . . . . . . . . . . . . .
28
species of wood ............ 27
tampers. electric and pneumatic ................. 76
Tile duct .................... 62 5
Tilted rail ................... 42
Tire. steel. holding power ..... 409
Timber preservation .......... 749
Tools. journal packing ........ 429
track ..................... 78
Tower. water ................ 89
Track bolts .................. 54
bonding ................... 645
car house .................. 8 I
special work ............. 85
centers. shop .............. I I 3
connection to negative feeder 726
~onstructlon... . . . . . . . . . . . . 32
current average and 1.m.s ... 656
curves designation ......... 55
degree to connect tangents 59
to find ................ 58
flangeway ............... 62
gage .................... 62
laying out ............... 59
drains ..................... 31
drills ......................
77
electrical resistance ......... 651
to earth ................. 693
fastening material .......... 54
foundation ................
29
frogs ......................
67
spring .................. 69
layouts. carhouse ........... 81
maintenance machinery ..... 76
~ n d e r...................
s
533

..

........

. .

.......

.. .

PAGE

Tra* shop ................. 1x3

splkes .....................
55
spirals .................... 60
switches ................... 66
derailing ................ 70
electric .................. 75
interlocking .............. 77r
open track ............... bB
tools ...................... 78
Trackless trolley overhead construction . . . . . . . . . . . . . . . . 583
Traction. h e a w electric ....... 584
wwer requirements ......... 188
Trails. standard ............. 39
Trailer operation ............. 213
Train control ................ 337
automatic ................. 780
movement ................. 119
resistance ................. 127
car end sha effect. . . . . . 138
wei h t e E c t ........... 128
f ormuyas................ 128
freight trains ............ 137
starting ................ 143
tem rature effect ........ 142
tracrcurves ............. I44
tunnels .................. 134
sheets ..................... 120
stop. automatic ............ 780
Transfer table, car house ...... 86
shop ...................... 2 2 3
truck ..................... I15
Transmission lines ............ 699
calculations . . . . . . . . . . . . . . . . 700
construction, railway ....... 609
voltage ................... 699
Trestles. steel ................ 23
wood . . . . . . . . . . . . . . . . . . .
21
Triple valve plain ............ 469
quick action ............... 472
Trolley base ................. 370
bow ..................... 378
brackets ................... 557
construction classification . . . 550
contact secondary . . . . . . . . . 575
feed taps .................. 565
harp ...................... 367
inspection lubrication and
maintenance ........... 373
line material ............... 591
pantograph ................ 378
current capacity ......... 381
pole ...................... 374
pressure ................... 374
roller ..................... 383
Jectionalizing switch automatic . . . . . . . . . . . . . . . . . 568
section insulators ........... 567
switch box .............. 567
shoe sliding . . . . . . . . . . . . . . . 369
slot conduit ............... 643
trackless overhead construction .................. 583
trounhs ................... 98
und<rground ............... 643
wheels .................... 365
clearances, standard . . . . . . 5 9
current capacrry . . . . . . . . . 387
defects .................. 367

INDEX

PACE

Trolley wheels dimensions .... 367

mileage . . . . . . . . . . . . . . . . . 366
wire curves ................ 559
offset ................... 564
frogs........................ 5 64
grooved standard sections ... 588
r . y s ...................... 559
e ~ g h................
t
548 557
installation . . . . . . . . . . . . . . . . 558
sag and tension . . . . . . . . . . . . 5 8
specifications . . . . . . . . . . . . . . 5a6
splices . . . . . . . . . . . . . . . . . . . . 558
steel ................. 575 589
twin ...................... 575
Trnck arch bar .............. 389
bogie ................. 385 389
center plates ball bearing . . . 419
classification ............... 385
design .................... 385
double ............... 385. 389
l i ht weight ............... 393
~.c.B.
type ............... 390
minimum
curve for ......... AIR
~-~
overhauling ............... 430
s i n ~ l.....................
e
386
smvel ................ 385 389
transfer tahles ............. 115
wheel base ................. 41 8
Tunnel sections ............ 7-75
train resistance in .......... 134
Typical run ................. 162

.
.

U
Underground contact rail ...... 643
duct conduit . . . . . . . . . . . . . . . 625
slot plow . . . . . . . . . . . . . . . . . . 383
Universal nozzles on standpipes 89

Vacuum jar for measuring heating current ..............

Varnish. insulating ...........
Ventilation car ..............
motors. commercial .........
methods . . . . . . . . . . . . . . . . .
Vestibule shape and train
resistance . . . . . . . . . . . . . . .
Voltage changes in motor speed
characteristics ...........
transmission ...............
variation effect on operation

wnp u s.....................

..

............... 730

138
248
699
207

718
253
260
158

...

weatheroroof coooer wire ... 7aa

...
wheels .................... 399
Welded arc rail joint ........ 51
electric rail joint .......... 51
joints rail ................. 49
thermit rail joint
50
Welding eqrupment
76
Wheel base truck ............ 418
cast iron types and weights 399
check gage . . . . . . . . . . . . . . . . 411
contours standard ......... 413
defects . . . . . . . . . . . . . . . . . . . . 411
diameters effects of differencesin ............... 250
gear ratio armature and
car speed. formulas ..... 245
minimum. commercial motors
r r

.. .

..

..........
...........

..
.

a
rz
-""

drop pit .............. IOI I IS

failures due to braking ...... 432
flanges standard ........... 413
weldinn
................. 408
.
gages ................. 405 411
grinder. pit ................ 1x5
grinding .................. 399
large. advantages
418
life ....................... 410
limit of wear gage
413
mounting
408
gage .................... 411
plane gage ................. 412
removal ....................409
rolled steel ............ 405 410
scrapers ................... I 0
shrink fits ................. 409
steel ................. 405. 410
standard dimensions ...... 417
tapegage . . . . . . . . . . . . . . . . . 405
tire. holding power ......... 409
treads standard ........... 413
trolley .................... 365
turning ................... 406
wrought steel, specifications 403
8
Wheelbarrows ...............
W i d and ice loadings wires
and cables ........... 736-739
pressure .................. 140
velocity ................... 141
Wire allowable current capaFty ................... 735
alurmnum
nzes weights
resistances . . . . . . . . . . . . 732
car control cables . . . . . . . . . . 336
copper sizes weights resistances ................ 728
rubber insulated specifications . . . . . . . . . . . . . . . . . 613
sag table copper and alum~num
............... 739
sizes A.W.G. ............. 728
tables
727

..........
..........
.................

253
294
540
a53
269

Waiting rooms, heating .......

Washroom car house . . . . . . . . .
Water tower . . . . . . . . . . . . . . . . .
Watthour meter for cars . . . . . .
Weatherproof insulation specificat~on
.................
Weights. aluminum wire . . . . . .
cars . . . . . . . . . . . . . . . . . . . . . .
effect on operating costs . .
train resistance . . . . . . . I 28
p l ? S ............... 499 503
slngle end ............... 497

..........Pac.
324

Weights. controllers
COD^^ cable
mre ....................
motors ....................
two- and four-motor e q u i p
ments ................
transfer in acceleration ......

. ....................

798
Wire. terrmnology. standard

INDEX
PAGE

6 1I
trolley see Trolley wire
weatherproof specrficatron 619
werghts
734
mnd and Ice loadrng 736-739
W l m g car
336
car house
10s
Witt car
SO7

PAGS

Wood poles

598
7 49

preservation
Work cars

78

Z ~ n cchlonde treatment of wood 749

of tres, undesirable features
666 749. 775

THE FOLLOWING PAGES

ARE REPRINTED FROM THE
1915 EDITION

GENERAL ELECTRIC MOTORS

235

Standud General Electric Railway Motors

NON-COYYUTAT~N~
DIRECTCUPPENT500 VOLTS
POLETYPE,
Motor

No

Weight of motor
Wc~ght.
tncludln~gear and total e q u t p
gear case, pounds ment, pounda

H.p.

1 K-ro

K-11

1.900

as

4.760
8,865

as

a
4

K-10
K-la

1.700

4.360
8.06s

Boo

as

a
4

K-10
K-11

Y.950

4.860
9.06s

81

30

a
4

K-YO
K-~a

a.oao

5.000
9.345

78

35

a
4

K-10
K-a8

1.~80

6.080
~1.6~0

10001

3s

a
4

K-ro
K-a8

a.185

5.330
10.aoo

*18~

37

K-YO
K-a8

1.111

5.410
10~60

67'

40

a
4

K-YO
K-a8

1.450

5.860
11.r60

-80

40

a
4

K-ro
K-a8

2.850

6.r 1.860

88

40

a
4

K-10
K-28

3.070

7.080
13.630

701

40

a
4

K-10
K-a8

2.750

6.460
1a.460

SJ1

45

a
4

K-II
K-14

*,8so

6.1~0
13.66s

57'

50

K-11
K-14
Mult. u n ~ t

3.030

7.110
14.38s
14.74s

K-36
K-35
Mult. untt

%all

6.765
13.750
14.171

a
4
4

K-36
K-s

3.290

7.811
14.910
15.861

a
4
a
4

K-a6
K-34
Mult. untt
Mult. untt

3.3b

a
4
a
4

K-a8
K-34
Mult unlt
Mult. unlt

3.533

Name

601

4
4
Po1

50

4
4
'98

50

-87

60

74'

65

,f:,",

MU&.

unit

I:%
8.6~0
86.207
6.530
1s.h
9.360
16.ags

I These moton have been superseded by later types. but o w ~ n gt o t h l r

wtdespread use. data concerntng them are tncluded for cornpartan.
*Charactcr~strccurves. Figs. 6 t o 9.

ICLECTRIC RAILWAY 1IANDROOK

236

GE 500-VOLTNOX-COMMUTATING
POLE MOTORS.-Confind
Motor
Name

Weight of motor
Weight
including gear and total e q u i p
gear case, pounds ment, pounds

H.p.

I -

4.137

Mult. unit
Mult. unit
Mult. unit

-66

16

+t

1P:X

10.580
19.670

160

I :I
1 I

Mult. un/t
Mult. unlt
Mu:L unit
Mult. unit
~ u l t unit
.
Mult. unit

II
(

I1

5.420
5,170

6.250

14.M~
27.5ao
13,585
16.520
i~.tro
3l.aOO

GE 600-VOLTD.C. COYYUTATINC
POLE TYPERAILWAY
MOTORS
Moror
Name

H.p.

I I
1-40.

TYP

,Weight 01 motor
Weight.
~ncludinggear and t o W e q u l p
gqar c n u . pounds ment. pounds

2.180

Mult. unit

10.670
1.150

5.550
10.270
10.550

Mult. unit
2'M.

Mult. unit
2.875

Mult. unit

Mult. unit

13.200
13.450
I.-

13.390
1.731

Mult. unit
Mult. unit

6.450
12.100
12.310

6.710
13.610
13.100

'Characteristic curves Pig. lo. I r and 12.

t This motor has b e c ~ r u p c n e d e d ' b ylater types. b u t owing t o its r i d e spread use. data concerning ~t is included for comparison.

GI;;I\I.ERtU. ELECTRIC AIOTORS

237

GE VOLT D.C. CO~XWATING

POLETYPERAILWAY
Morons.Continued
hf otor

"'"1 1 i ::: :1

Name

Weight of motor
Weight.
mcludtng gear and total equipgear c w . pounds ment, pounds

H.p.
lo

70

r
J

1s

4
2

Mult. unit
Mult. unit

3.380

15.710
7.150
8.690
ts.uo

i
'

K-36
K-34

Mult. unit
Mult. unit
K-3s
K-34

Mult. unit
Mult. unit
Mult. unit
Mult. unit

i
I

225

$15

122

140

I Mult. unit I

107

165

5.160

211

135

6.000

109

175

Mutt. unit

--

10.~00
16.760
10.710
I9.900
13.050
25.950
14.800
29.780
50.930

10.400

GE 600/1~00-V0L~
D.C. COMML'TATIXG
TYPE
MOTORS
-- - --- - - POLE
-- -RAILWAY
.- - - - - --- -

le4:

~ u t t unit
.

hfult. unit

3.-

Mult. unit
Mult. u n ~ t

3.930

Mult. unit
Mult. unit

5.160

Mult. unit

--

I;

unit

*Characteristic curves. Pigs. I 3 to 17.

4.100
5.160

GE 1100-VOLTD.C. C O M M U T A TPOLE
IN~ T
"5

3.850

MOTORS

~ ~ R A I L W A Y

3.850

9.850
10.000

ELECTRIC RAILWAY HANDBOOK

238

Standud Westinghouse Railway Motors

D l n e c r - C ~ ~ ~ ~E N IO N - C O ~ ~ UPOLE
T A TYPE
T~~

I 1-

( D n ~ g n e dprior t o 1904)

Alotor

Name

H.p.

m
E
!
,

Weight d motor
Weight.
'including genr and total equipgear c u c , pounds ment. pounds

J:gl

'la-A

25

2
4

K-ro
K-12

I a-A

30

a
4

K-10
K-la

................

*bp

30

a
4

K-lo
K-12

1.950

4.900
9.140

*49

31

a
4

K-10
K-18

1.925

4.900
9.300

91-A

33

a
4

K-lo
K-a8

1.165

5,580
10.500

68-C

40

a
4

K-lo
K-28

1.280

5.600
10.~60

101

40

r
4

K-lo
I(-28

2.64s

'

2,100

5.400
10.140
5.400
10.140

6.340

I a.010

101-C

40

a
4

K-lo
K-a8

2.780

6.700
I J,700

38-B

50

a
4

K-ro
K-a8

3.380

5.810
10.960

'93-A

60

a
4

K-11
K-14

3.440

8.080
16.360

56

60

K-11
K-14

3.0-

7.100
14.400

Unit switch
Unit s w ~ t c h

3.44'3

8.771
15.820

4
1x1

7s

76

7S

a
4

Unit switch
Unit switch

3,840

9.57s
17.410

85

75

a
4

Unit switch
Unit s w ~ t f h

4.500

10.895
ao.060

111

a
4

Unit switch
Unit switch

4.300

10.530
ao.015

Unit switch
Unit switch

4.680

a
4

Unit switch
Unit switch

b.SS0

K-to-A
K-a8-B

1.780

K-36-

r.7&

'119

125

*I 13
*IOI-B-a

101-D-2

40

50

.a
4
a
4

:d::
IS0795

Special
6.635
11.6~0
7.225
13.065

Unit switch
Unit swltch

*Characteristic curves. Figs. 18 to 14.

6,705
12.695
7,245
13,100

WESTINGHOUSE MOTORS

239

WESTINGHOUSE
D .C. NON-COYYUTATWG
POLE TYPERAILWAY
Morons.-Codinucd

1 Ij 1 ; 1
I

Name

H.P.

1 I

*11a-8

75

[It.O

-114

3.440

Unrt switch

60

(Designed since I W ~ )
i
to
r
weight.
including g r .nd total equipgear c u e . pounds ,mcnt, pounda

Motor

i
4

a
4

Unit swltch

/K-3s-G
K-$4-J?

Unrt swrtch
1 Unit switch

3.4s
'

8.S8S
15.811

8 . ~ 0 ~
16.245
8.860
16;ooo

10,130
19.500
10.530
19.81s

Unit switch
Unit switch
Unit rwitch

13.080

Spceinl

WE~TINCHOUSE
D.C. COYMUTATING
POLERAILWAY
MO;ORS

.C

Motor

*eifht,of motor

mc U ~ I W

and sear case.

pounds

318

37

313-A

a
4
a
4
a
4
a
4

30

40

33

Weight.
total
equipment.
pound.

K-IS-M
K-28-U.

Unlt nw~tch
Unit r w i k h
K-11-M

K-aWU
Unit switch
Unit r w i k h

Unit switch

306

60

so

1316

60

so

l Unit switch l

13.390

4
a
4
a
4
z

( Unit rwitch

13.450

K-36K-3s-&

.
Un/t switch
Unrt switch

Characteristic curves. Pig% 15 and 2 6 .

t Characteristic curves. Fig,. a7 nnd a8.

I1,I;CTRIC RAILWAY IIANDBOOK

240

\VESTINGHOUSE
D.C.COYMUTATING
POLE RAILWAY~IOTORS.Continrrcd
Motor

305

3.550

8.115
16.370
9.000
16.360

310
8.780
15.920

Unit switch
304
Unit switch
Unit switch
-317

1.-

A
a

1 1.
4 0

.....

(;

. . . .I

:I

16.965
9.110
16.800

Unit switch
u n i t switch

-1

Unit switch (
u n i t sw/tch
Unit swltch

4.150

10.230
19,asO

4.685

11.430
a I ,630

Unit switch
Unit switch
Unit switch
Unit swltch

W E ~ T I N C H O U1 2~0E0 Motor

--

-I

Type
control

6 0 0 (~ ~ 0 0 ~ .

I :: I ::: I : I
I
II

* 3 0 8 ~ D - 260
~
260

215
215

IS.S~
Spcc~al

Wright of motor Weight.

total
including gear
and gear cast, equipment,
pounds
pounds

- ----

'JIJ

1500-VOLT D.C. HAILWAYMOTORS

AND

No.

6.400

U 'it switch
k i t switch
Unit switch
Unit rwltch

Charactarutic curves. Figs. a 9 t o 35.

3.90"

Special
Special

4.150

Special
Special

I
I '

4'6",

Special
Special

6.740

Spccjal
Specral

I\'. IS. & AI. A N D ALLIS CIIALAIERS AIOTOKS

241

WESTINGHOUSE
S ~ G L E - P R A S E A.C. RAILWAY
MOTORS
Por Motor Cars

Name

'

*rjS

H.p.

1.32

132-A

75
loo

IJI-F
*148-A
'1.3~

4
4

loo

11s

'156
409
409-D

175

-4x0
.409-C'
I51
.I

:::

/ 1

240

I75

137
jo
403

4
4
a
4

4 {in

'~03-A

'

Weight of motor
Weight:
and total equ,vincluding
~ case,
l gear
gear
pounds meot, pounds

specid
Special
SpcciJ

Uhit switch
Unit switch
Unit switch

4.500
. 5.100
S.Si5

Unit switch
Unit am itch
Unit switch

5.575
S.7*'3
6.100
6.000

Special
Special
Special
Special

Unit
Unit
Unit
Unit

7.iW
'1.846~
6.700
7.846'

Special
Special
Special
Special

switch
rwikh
switch
switch

Par Locomotive,
Unit switch
Unit switch
Unit switch
Unit switch
Unit switch

I
1

16,soo'
10.600

Special
Special
Specid

15.700
16.41 .

Special

Unitswitch\

21.2z.1

Unit switch

19.800'
19.800'
41.200

Special
Special
Specid

Weights here include drive details.

1 M o t o n arranged for quill drive.
1409-C motor is arranged t o o p r a t e in p u n . W e ~ g h tgiven is for one
pair of motors complete with drive details.

Motor
Name

I
*JOZ

H.p.

ss

i
I

I S-3
.S4

Chiracteristic curves. Pigs. 36 t o 46.

Weight 01 m o t w
Weight.
lincluding gear and/ total e q u i p
gear case. pounds (merit. pounds

3.060

7.640
14.030

Motors for Low Floor Cats. blotors commonly known a s "Low

Floor," "Dachshund" and "Baby," for use on low Boor cars built
with small wheels, are of smaller diameter than but d o not differ
in general electrical d a i g n from motors ordinarily used with 33411.
wheels.
P r e s s e d Steel Motor. Certain railway motor parts formerly
made of cast steel are now made of pressed steel, for some
of the motor capacities of most common application. The
10

242

ELECTRIC RAILWAY HANDBOOK

weight of a 40-h+. pressea steel motor is about 1 2 per cent. less

than a cast steel rnotor of the same capacity.
Railway Motor Characteristic Curves. Figs. 6 to 46, inclusive,
give the operating characteristics of some of the commercial motors
when operating a t particular voltages with particular gear ratios
and driving-wheel diameters. T h e ordinates for the tractire cjorl
curve for a series motor with any gear ratio or driving-wheel diameter
may be derived from the ordinates of the original l r o ~ l i weffort
curve lur that motor as Idlows:
in which T 1 = traiiive effort ordinate for any current vaIue on the
derived curve
T = tractive effort ordinate for the same current value
on the original curve
G' = Rear ratio for derived tractive effort curve
C = gear ratio for original tractive effort curve
Dl = driving-wheel diameter for derived tractive effort
curve
D a driving-wheel diameter for original tractive effort
curve.
T h e approximate values of the ordinates of the s p e d curve for a
series motor w ~ t hany gear ratio or driving-wheel diameter operating
a t any other electromotive force, difTering slightly from the electromotive force a t which the original speed curve was obtained, may
be derived from the oricinal sbeed curve as follows:
any current value on the derived
curve
S
speed ordinate for the same current value on the
original curve
E1 = c.m.f. for detived speed curve
E = e.m.f. for original speed curve.
( N o a : When the electromotive force is not changed, that is,
when E' = E, the value given for S1 is exact.)
Example of the Dersoaliow of Traclivc E j o r l and Speed Ciwaes.
Fig. 48 (p. zrf5)shows tractive eflort and speed curves (dotted) derived from the original curves by the above process as follows: The
original gear ratio, wheel diameter and voltage are 2.78, 33, and
500, respectively, and it is desired to derive the tractive effort and
approximate spted curves for a gear ratio, wheel diameter and
voltage of 3.83, 36, and 550, respectively. I n this case,GL P3.83,
G = 2.78, D1 = 36, D = 33,-El = 550 and E = 560.
Therefore

MOTOR CHARACTERISTICS

248

Thus in Fig. 48 the dotted tractive effort curve is the locus of

all points the ordinate of each of which is equal to 1.26 times the
corresponding ordinate on the original tractive effort curve, and the
dotted s p e d curve is the locus of all points tbe ordinate of each
of which is equal to 0.87 timu the corresponding ordinate on the
original speed curve.

[NOTE:Where, as for preliminary investigation, o d y one or very

few points on the derived characteristic may be desired, it may be
advantageous to obtain the desired derived values by the above
method of calculation directly without actually plotting the derived
curve.)

PIG. 6.-CeheraI

Electric No. 5.5 railray motor. 37 h p..

15. cur w. mtto 4-60. wheels 30 m

SW

volt* pi*

ELECTRIC RAKWAY HANDBOOK

244

ill 'uob
'24''TXa
loo 24"lm6

..,*. t Z ! , , l ~
a,pco
m

..

E y iI
l4C1m
0

. a

*40

us800

10 10

a0

\ o : a zm
0 4

10 20

CO 40

4 60

10

loo 110

&mporsl

RG.7.dGeneral

Electric No. 80 railway motor. 40,h.p.. SOO'*O~~*

17. g u r 6% ratio 4.06. wheels 33 in.

RG.8.-General

Electric No. 98 railway motor. 50 h.p.. ~ o volts.

o
pinion
16. s e u 71. ratio 4.43. wheels 33 in.

"

MOTOR' CHARACTERISTICS

wll
mo
aa,

'fsoo

tam
14

ua)

2t 62200
100;m".
,DO
18 ;I.W
ao:lcfrsa,
; m i 1 4 2lu.a
;60
IIL'rn

7 50

10
g 4 0 . s

:JO

loo0

WO

10

400

10

rm

0 lo P

l o

50 60

m eo 90

l m u o l m m l 4 o ~ l ~

A m ~ m s

Electric No. 87 railway motor. 60,h.p.. loo volt#. pinion

16. genr 71. ratw 4.44. wheels 33 an.

PIG.9.-General

ClOO

uul
# ) * m

1m

8 0 5 2 3 m

m arm

is0 224 5 2 -

5-w i 2I0 5 .
zoa,
f- 40 3- 16 . 1600

ijoan5m
m

10

400

0
Ampere.

PIG. 10.-General

Electric No. 73 railway motor. 7s h.9.. 500 volt.. pinion

17. s u r 73. ratio 4.30, whalr 33 in.,

ELECTRIC.RAILWAY HANDBOOK
mlm
eo

60

(OUJwm

lo

i w g soo0;60

5" a'

G50
El

-;40

moo m

10

10

0
Am-

xl.--Cener~l .Electric No. 66 f.ilray motor. Ias h.p., soo volts.

pmon 17. pear 61, ratlo 3 . ~ 9 wheels
.
33 in.

(000

*I1C

s8 aa,
2400

mo

mrnm

Ma

n le$nl-a
mtmprm

6 lo!

:
sY-

$
2
0
,;,

ZI~I

C1

SalO

lea,

;co

i,

QO

I 2.-General

M B O
AmVera

Electric No. a03 !ailway motor. s o h.p.. boovolt.. pinion

I S . pear 69. ratlo 4.60 wheels 33 la

MO'XQR CHARACTERISTICS

247

248

ELECTRIC RAILWAY HANDROOK

PIG 15-General

Electr~cNo. 205-8 r a ~ l r a ymotor. l o o h.p.. 600 volts,

plnlon I ) . gear 57. ratlo 3 35, wheels 33 In.

FIG 16.-General Elcctr~cNo nos-A ra~lwa motor 80 h p 600 volts

plnlon 17. gear 67, rFtlo 3 35. wheels 33 In 6'ound for operating two In
SerleS on r loo %oltr

MOTOR CHARACTERISTICS

Prc 17 -General Elcctrtc No. 207-A rallwa motor 140 h p .boo volb.
22. gear sp. rat16 2.68. wheels 36 In.
b o u n d for operatton two

ptnton

sen- on r loo vol t...

2MI

ELECTRIC RAILWAY HANDBOOK

loo m

am

90

(08010

1 m

70

1100

4 : hi

w rot60 dl5

1100

I@*

rn

(0

Imo
8m

10

SO
10 I0

YIO

Po

10

0
Ampoem

Rc.

1 8,.-Westinghouse

No. IY-A-15 r u l w a ~motor. ?S h.p.

panaon 14. gear 68, ratlo 4.86. whuls 33 rn.

hc. 19.-Wntinghouu No. 69 railway motor. 30 h . ~ 5.0 0 volt., c in ion

I & scar 68. ratlo 4.86, whuls 33 ~ n . Continuous u p . c ~ t y15 amp. at 3 0 0

-dl4 23 M P at 400

VO~U

MOTOR CHARACTERISTICS

'lo
65

I0

60

25

50

,za
a
1;;

so.

40

P"E"

a
~ t ~ ;
: e .
*sogzs

10 40

20

30 15
6 20

10

10

l o P m * n o m m m m m u o
Amwre8

PIG.20.-West~nghouse

No. 49 railway motor. 3s h.p.. soo volts. pinion

14. gear 68. ratlo 4 86. wheels 33 in. Continuous c a p u ~ t y 30
. amp. at joo
volts. 28 amp. at 400 volts.

No. 9 3 4 railway motor. 6 p h p.. soo vol(.,

plnlon 17. gem 70, ratlo 4.12. wheels 33 In.

Rc. 21.-Westinphouse

FIG. 11.-Westinghouse No. 1 1 9 railway motor. 1 1 s h.p,. sso volts

pinlon to, gear 35, ratio 2.75. wheels 33 in. Contindous capac~ty.9s amp.
at 300 volts. 85 amp. at 460 volts.

W
UI)

..Q! .

y)

m51'Q.xi

l00"10

30
100

.*

20.

10
0

0
Amwrea..

.Westinghouse No. r 1 3 railway motor. 195,h.p.. 550 volts. pinion

19,gear 64, ratio 3.37. whecls. 36 In.

MOTOR CHARACTERISTICS

263

Ampere6

-Westinghouse No. 1 0 1 - 8 railway motor. 40, h.p..

pinton 17. gear 67. ratio 3.94. wheels 33 tn.

loo

am

180

woo

160 240

KGQ1:
Y

so005

loci

Ci

8020

30

60

10

40

I0

20

BOD:

la0

10

0
0

Ampere6

FIG. 15.-Westinghouse
NO. 111-B r d l r a motor. 1 5 h.p!. sw volts.
pinion 19, gear 70. ratio 3-68, wheels 36 in. (3ontinuous capacfitr. 60 amp.
at 300 volts. 5s amp. at 400 volts.

ELECTRIC RAILWAY HANDBOOK

254

16.-Wut~nghour No. r 14 rai!way morpr. 160 h.p.. 550 d b

gear 57. ratlo 1.85, wheels 33 ra Contmuow c a p r l t y . x a a a m p

20.

votes, r r a unp. st gar v o k

-Wert~nghoux No. 3IwB. rulway motor. 50 h.p..

ptnlon 17, near 69. ratlo 4.06, wheels 33 la.

MOTOR CHARACTERISTICS

1:

mi m a
100 100 b

4
d
sm,

*m

m m a

?o

GO Q l

so so
a 1o 10
aJ
a206

ari3
1DO

D m a

8 lo
0 8o
lo

wm1oQm1(omm5o

Amg.n#

RG.~d.-Westiogboase No. 3x6 r u l w ~ ymotor. 75 b-p.. 600 rdU, p i d m

15, c u r 71. ratio 4.73. wheclr 30 r a

too

roo

190 m
1m

40

14OL.m

2 1s: a%

smof

3 1mii- so*4

ax03

m a m
(0

a
m

9 10

100

10
8
0

6 4 1 0 0 U O 1 0 0 ~
Amma

rrc. ap.-Watiqhouse No. 317 rulway motor. 9q h.p. 600 volts,

pidon 19.seu 70. rstm 3.68. whecla 33 In.

256

ELECTRIC RAILWAY HANDBOOK

Prc. 30.-Wert~nehouse No. 301-A railway motor. rqo h p.. boo

plmon ar. gear 56, rabo 1.67. w h c l s 36 tn.

Amp.ram

-Wcstmnhousc No JOI ratlway motor. 17s h p.. 600 volt..

18, gear 59, ratto 3 28, whulr 36 t n . .

hIOTOR CHARACTERISTICS

PIG. 3%-Wcst~nghourc No. 333-8 rri!way motor. 11s h p.. 606 volta.
plnton 10. g e u SO. ratm 3.0s. wbeels 36 tn. Wound for operatron two In
ten- on rroo v d t r

mL#)m

00

16a

en

140

lo

"I

40

-3

sot

@!

mi m ~
a
e

<

DDO5

80- 40=m
(0

.no gr .30
SO

co m z n

lano

0 u
Amvane

PIG. 33.-West~nghouse No. 311 r u l w a ~motor. I l o h p.. TSO mltr.

plnlon 16. gear 61. ratlo 3 81. wheels 36 m. Wound for oprrmtron tm In
rma on ISOO volts.
17

258

ELECTRIC RABWAY HANDBOOK

FIG. 34.-Wtatinghourc No. 3 2 1 railway motor. 140 h.p., 750 volt!.

pinlon 16. gear 61. ratlo 3.81. wheels 36 in. Wound for operatiqn two in
rtricr on ISW volt^.

AG. 3s.-Wertinrbourc No. 3od:q rulway motor. .as b.p. (250 h.p.
with forced vcnt~lation)boo volts. pinion 19.gmr 1 2 , ratio j.79. wheel. 36

in. Wound for operation t m in rner on l l w volts.

MOTOR CHARACTERISTICS

EG.
36.--Watinghouse
No. 13s-B railway motor. 380 amp.. 110 volt..
as-cycle, single-phase. alternating-cumnt. pin~on 17, g e u 60. ratio 3.53.
wheels 36 In. Cont~nuowCapacity I70 amp. a t 1.30 volts.

100aQK'

m m3Qe

m m a

m:

70

140 I

"
0

1M

.m,w

40

80 20

10

Gals

4Q

2250:

Ica;

~5

10

500

lorn)

'ls0

.
0

4 50e1004B
i P 3 a

lsoot

~oomzmsom.%aumwm~~
Amptrea

PIG. 37.-Westinghouse NO. t3z-C r d w a y motor. 430 unp.. 13s rdb.

2s-cycle, rlngle haw. alternrtingsurrrnt. plnlon ro. gear 63. raw 3.15.
wheels 36 an. &ntlnuour eapsctty roo amp. at 135 volt&

ELECTRIC RAILWAY HANDBOOK

PIC. 38.-Westinghouse No. 148-A, railway motor. 550 amp. (forced

vent~lation).225 volts. 25-cycle. alternating-cumnt. pinion 15, gear 66. ratlo
2.64, wheels 38 in. Continuous capacity 315 amp. at 150 volts.

Prc. 39.-Westinghouse No. 133 railway motor. 600 amp. (forced

vent~lat~on).
135 volts. as-cycle. alternat~ng-current.ptnion 1 7 , gear 71. ratio
Continuous c*pacity. 600 amp. at a j s volt..

4.24. wheels 38 m.

MOTOR CHARACTERISTICS

Plc. 40.-Westinghouse
No. 156 nilway motor. 600 amp. (forced
vent~latlon). 2 1 0 volts. 2 cycle. slternating-current. pinton 25. gesr 7 4 .
ratlo 2.96. wheels 4 2 in Ebatinuour capacity. 450 nmp. at rIo volts.

PIG.41.-Wut*nehol~e NO. 4-C

locomotive motor. 650 amp. (forced
ventllat~on). 275 volts. 25-cycle. alternatingsumnt. pinion la. g u t 92.
ratio 4.18. wheels 65 in. Contrnuous cnpulty. Soo amp. at 275 volts.

262

ELECTRIC RAILWAY HANDBOOK

RG.4a.-Westlnghouse No. rjo locomotive motor. rrjo amp. (tctr~.d

vent~lat~on),
a20 volts. a5sycle. dternat~ngsummt.geulcss, wheels 61 In.
Cont~nuouscapacity. 860 amp. a t aao volta.

FIG. 43.-Westinghouse No. 403-A locomotive motor. Iamp.

(forced ventilation). 300 volts. 25-cycle. alkrnat~ng-current.pinion 14, @err
Contmuoua capactty, 930 amp. at joo v o l t .

89. ratto 3.71. wheel* 63 in

263

MOTOR CHARACTERISTICS

Po0

BW

24 100

lea,

22

00

80

1Ga) 20

40,

2i!

<1100;18

~ l a m + l 6;so
ilmgl4

=":

tm

8
iea,llzs*

i a a 10 JO
loo

a m

10

'0

10

10 20 30

Y)

54 60 70 80 WlC0000111

Amwree
Rc. 44.-Alli+Chnlmen No. 301 railway motor. 40 h p.. 5 6 0 volts. pinion
15, gear 69. rmt~o4.60. wheels 33 in.

m
a

tcOO g
pm

30

2000

28

1sm a
3d 1 m . u

50.

so
m

e!
m
(0

Eiroo,jaero
zlaapqa,
imo;ie;j50
gaS1sY*

3Q

$
;

20

*a

14 30
(00 12 20
m 10 10
0 0 0

10

lorn

aasmm m

0
~ ~ I W U K O U

Ampere.

PIG. 4s.--Alllr-Chalmen No. sox rulway motor. s o h.p.. boo volts. pinion
I S . gear 71. ratlo 4.73. wheels 33 lo

ELECTRIC R A L W A Y HANDBOOK

204

PIG. 46.-Allis-Chdmers No. 301 ra~lwaymotor. 5s. h.p.. sw volts. pinion

16. gear 73. ratto 4.56, wheels 33 I~L.

LtO
LGO

:L50

;Ltll
2 L30

LZO
LIO

2 LW
=. a a

I am

3 a70
am

:: as0

aco

j aw

a20
a 10

'0'

25

50
15
100
12J
lk4
5015
200
Horse Power of Yaror(Nornlnml One E o u r R*tlm()

2S

PIG. 47.-~~proxirnate resistance of 500-volt direct-current moton at 75. C.

ALTERNATING-CURRENT MOTORS

36

1em
1700

31

lea,

SB

265

?a Ism

a Irm
28

1300

26
24 ~1100

lw

w
,eo:

:n

4";"

-: l o y

m$sm

!so

163 700
11 6W
12 m

:40

10

GO

4M

c30
20

8
6

10

100

0 10 2050 4 0 5 0 60 70 80 #OlWllOL~uOlrOutlMQ
Amperel

FIG. 41.-Illustrating

change of tractive effort and s p e d curves for

change In gear ratlo. d r ~ v ~ nwheel
g
diameter and voltage. (GE-17 motor.
60 h.p.)

Alternating-current Railway Motors

General Structural Difference between Single-phase Cornmutating Alternating- and Direct-current Motors. The structural
differences between single-phase alternating-current series moton
and direct-current motors are summarlzed by hlessrs. R. E.
Hellmund and E. W. P. Smith in the Electric Journal, 19x2, as
follows:
( I ) The laminated core of the alternating-current motor;
(I) Comparatively large number of poles of the alternating-current
motor;
'(3) The use of a distributed auxiliary winding on the alternatingcurrent motor in addition to the main field winding;
(4) The larger armature diameter as compared with that of a
direct-current armature;
( 5 ) Short and stubhy poles made possible by the small number of
field ampere-(urns on the alternating-current motor;
(6) The large ancl wide commutator of the alternating-current
motor and the larne number of segments as compared with
direct-current machines;
(7) Small air gap of the alternating-current motor.
To these should he added preventive leads between the armature
winding and commutator in the alternating-current motor.
The use of the laminated field structure, the auxiliary or com-

ELECTRIC RAILWAY HANDBOOK

pensating winding and the preventive leads in the single-phase
motor are the diflerences essential to the single-phase motor.
Laminated Field Structure. The field is laminated in order tc
reduce the losses and associated heating due to the alternating flux
in the iron.
Compensating Winding. The purpose of the compensating
winding is to neutralize cross-magnetization, thereby improvinq
commutation dnd power factor of the motor. Cornpensation is
classified as conductive or inductive, depending upon the process
b y which the current flowing in the compensating winding is obtamed. If the compensating winding is connected to receive current from an external source, as by connecting it in series with the
main motor circuit, the compensation is said to be conductive. If
the compensating winding is closed on itself so that a current due
to a direclly induced electromotive force flows within itself, the
compensation is said to be inductive. Gn the New York, Westchester and Boston Railway inductive compensation is used. On
the New York, New Haven and Hartford Railroad motors intended
only for single-phase operation are inductively compensated. On
the same road, motors intended for both single-phase and directcurrent operation are conductively compensated. Insulatiod strain
to gr und is made negligible in the inductive compensation.
~ n i u c t i v eand Conductive compensation. \Vhen motors operate
only on alternating current the diflerence in res~iltsof either inductive or conductive compensation is inappreciable, but where the
motor is to be operated with direct current, conductive compensation should be used.
Preventive Leads. Preventive leads in a single-phase commutating motor are added resistance between commutator windings
and commutator bars. Their purpose is to restrict the current
flowing in the armature coil when short-circuited by the brush.
They add their resistance effect to that of the carbon brush bridging
the commutator bars. I n so restricting the current they improve
commutation and they may improve the efficiency of the motor.
Whether or not eficiency is improved by the preventive leads
depends upon the balance of the following two effects: (a) I'R loss
due to the passage of the armature current through the preventive
leads on its way to the armature winding. This loss varies directly
with the resistance of the leads. ( b ) IIR loss due t o the passage of
current set up in the short-circuited armature coil as a result of the
electromotive force induced by the field. This loss decreases as the
resistance of the preventive leads increases. The preventive leads
are of greater value during starting and ac'celeration.
General Types of Single-phase Comdrutating Motors. There
are three general typesof single-phase cornmutating motors, namely,
the series, repulsion and the induction series. Of these the former
two are in most general use in America.
Connections of Single-phase Commutator Motors. The conventional diagrams, Figs. 49 to 5 5 , inclusive, show some of the mcthods of connecting armature, field and compensating coils of singlephase commutator motors. Since the windings on armature and
field of the repulsion motors are not electrically connected, opera-

ALTERNATING-CURRENT MOTORS

267

tion of this type is pocsible on comparatively high voltages if the

armature be so constructed that a low voltage is induced in it.

PIG. 49 -InP I C . 50.-Conductrvel corn- durtrvely

con+
pensated) serres pensated
series
motor.
motor

PIG. 51.-lnductron s m u
motor.

P I C . s 2 .-

Thomson
repulsion motor.

PIG. 53.-Series
compen<ated
repulsion motor.

PIG.54.-WlnterE~chbergrepulsion
motor.

Frc. 5s.-Latour
repulr~onmotor.

Three-phase Induction Motor. The three-phase motor is called

a constant speed motor (see p . 383 for methods of changing speed).
I t has the characteristics of a direct-current constant speed motor.
Neglecting control, the speed of the three-phase induction motor
depends upon tlie frequency of supply and the speed falls off but
little as the load on the motor is increased. The speed of a train
driven by three-phase induction motors ordinarily remains nearly
constant up grade and down grade. The three-phase induction
motor regenerates eneray, forcing it back into the line when the
motor is driven above its synchronotts speed (see p. 213). This
process takes place as the train descends a grade, and since this
regeneration brings a load on the motor acting as a generator the
train is not allowed to accelerate beyond the point a t which the
force due to gravity is balanced by the forces it is overcoming in
driving the motor acting as a generator. Since the three-phase
motor is a constant speed motor, in ascentling a grade i t will make
a demand on the power station proportional to the torque required
and the losses to the motor. As outlined on p. 2 1 4 this demand
may, however, be reduced by the energy of regeneration of trains
descending. In order to operate on long up grades a three-phase
induction motor must have high continuous capacity.
The starting efficiency of three-phase motors is low so the motor
is not well suited to service having frequent stops. The tractive
e6ort may be held constant during the starting period, but this is
done a t the expense of energy lost in the motor-starting resistance

?@

ELKCTRIC RAI1,WAY HANDBOOK

Since there is no electrical connection between the primary winding and the secondary winding of a n induction motor, and there is
no commutator, this type of motor may be built to operate on voltages which are high compared with those on which motors of other
types are operated.
Speed of Induction Motor. The speed of an induqtion motor is
equal to its synchronous speed minus the slip. The value of the
slip increases at practically a constant rate from nearly zero a t no
load to about 2 t o 5 or 6 per cent. of synchronous speed a t full load.

in which S = synchronous speed, revolutions per minute

N = number of pairs of poles in the primary winding of
the motor
j = frequency of applied electromotive force, cycles per
second

in which St = actual speed of motor, revolutions per minute

s = slip, per cent.
-

--

Number of pairs of polcs In

I

Prcquency. cycles per

second
IS
a$
60

1 1 1 1 1 1

Synchronous speed. revolutions per

mlnute
450
300 22s
150
1::;
750
500
37s 300 250
3600
1800 1200 900
6*

118

2f4
514