EiBgineermgf

library
.' ..*. v -'
 DESIGN OF

POLYPHASE GENERATORS AND MOTORS

McGraw-Hill BookCompatiy
Pujfi&s/iers offioo/br
Electrical World The Engineering and Mining Journal

Engineering Record Engineering News

Railway Age G azette American Machinist

Signal LnginGer American Engineer

Electric Railway Journal Coal Age

Metallurgical and Chemical Engineering
Power
 DESIGN OF
POLYPHASE GENERATORS
AND MOTORS

BY

HENRY M. HOB ART,

u
Consulting Engineer, General Electric Company; Member American Institute
Electrical Engineers; Institution Electrical Engineers; Institution Mechanical
Engineers; Member Society for the Promotion of Engineering Education

McGRAW-HILL BOOK COMPANY

239 WEST 39TH STREET, NEW YORK
6 BOUVERIE STREET, LONDON, E. C.

1913
 Engineering
Library

;'

COPYRIGHT, 1912, BY THE

MCGRAW-HILL BOOK COMPANY

THE. MAPLE PKKS8- YOUR. PA

.
 PREFACE

DURING several recent years the author has given courses

of lectures at three technical schools in London, on the subject
of the design of electric machinery. These three schools were:
the Northampton Institute of Technology; Faraday House;
and University College. Various methods of procedure were
employed and these ultimately developed into a general plan
which (so far as it related to the subjects of Polyphase Genera-
tors and Polyphase Motors), has been followed in the present
treatise. It was the author's experience that the students attend-

ing his lectures took an earnest interest in calculating designs

of their own, in parallel with the working out of the typical
design selected by the author for the purpose of his lectures.
At the outset of the course, each member of the class was assigned
the task of working out a design for a stipulated rated output,
speed and pressure. Collectively, the designs undertaken by the
class, constituted a series of machines, and co-operation was

encouraged with a view to obtaining, at the conclusion of the course

of lectures, a set of consistent designs. If a student encountered
doubt concerning some feature of his design, he was
difficulty or
encouraged to compare notes with the students engaged in design-
ing machines of the next larger and smaller ratings or the next
higher and lower speeds. Ultimately the results for the entire
group of designs were incorporated in a set of tables of which
each student obtained blue prints.
" "
At two of these three colleges, the sandwich system was
in- operation, that is to say, terms of attendance at the college

were " sandwiched " with terms during which the student was
employed in an electrical engineering works. The result of the
author's opportunities for making comparisons is to the effect
that students who were being trained in accordance with the
"
sandwich " system were particularly eager in working out their
v

257829
vi PREFACE

designs. Their ambition to obtain knowledge of a practical

character had been whetted by their early experiences of prac-
tical work; they knew what they wanted and they were deter-
mined to take full advantage of opportunities for obtaining what
they wanted. One could discuss technical subjects with them
quite as one would discuss them with brother engineers. There
was no need to disguise difficulties in sugar-coated pills.
It will be readily appreciated that under these circumstances
there was no necessity to devote time to preliminary dissertations
on fundamental principles. The author did not give these
lectures in the capacity of one who is primarily a teacher but he

gave them from the standpoint of an outside practitioner lectur-

ing on subjects with which he had had occasion to be especially
familiar and in which he took a deep interest. It is the function
of the professional teacher to supply the student with essential

preparatory information. In the author's opinion, however, a

considerable knowledge of fundamental principles can advan-
tageously be reviewed in ways which he has employed in the
present treatise, namely; as occasion arises in the course of

working out practical examples. Attention should be drawn to

the fact that various important fundamental principles are there
stated and expounded, though without the slightest regard for
conventional methods.
Of the very large number of college graduates who, from
time to time, have worked under the author's direction in connec-
tion with the design of dynamo-electric machinery, instances
have been rare where the graduate has possessed any consider-
able amount of useful knowledge of the subject. Such knowledge
as he has possessed on leaving college, has been of a theoretical
character. Furthermore, it has been exceedingly vague and poorly
assimilated, and it has not been of the slightest use in practical
designing work. One is forced to the conclusion that teachers
are lecturing completely over the heads of their students far more
often than is generally realized. The author is of the opinion
that the most effective way to teach the design of electric
machinery is by step, through the actual
to lead the student, step
calculations, simultaneously requiring him to work out designs
by himself and insisting that he shall go over each design again
and again, only arriving at the final design as the result of a study

of many alternatives. It is only by making shoes that one learns

 PREFACE vii

to be a shoemaker and it is only by designing that one learns

to be a designer.
While the author has not hesitated to incorporate in his treatise,
aspects of the subject which are of an advanced character and of
considerable difficulty, he wishes to disclaim explicitly any pro-
fession to comprehensively covering the entire ground. The
professional designer of polyphase generators gives exhaustive
consideration to many difficult matters other than those dis-
cussed in the present treatise. He must, for example, make
careful provision for the enormous mechanical stresses occurring
in the end connections on the occasions of short circuits on the

system supplied from the generators, and when a generator is

thrown on the circuit with insufficient attention to synchronizing.
He will often have to take into account, in laying out the design,
that the proportions shall be such as to ensure satisfactory opera-
tion in parallel with other generators already in service, and it
will, furthermore, be necessary to modify the design to suit it
to the characteristics of the prime mover from which it will be
driven. This will involve complicated questions relating to
permissible angular variations from uniform speed of rotation.
The designer of high-pressure polyphase generators will find it
necessary to acquire a thorough knowledge of the properties of
insulating materials; of the laws of the flow of heat through them;
of their gradual deterioration under the influence of prolonged

subjection to high temperatures and to corona influences. He

must also study the effects on insulating materials of minute
traces of acids and of the presence of moisture; and, in general,
the ageing of insulation from all manner
of causes. The designer
of polyphase generators can employ to excellent advantage a
thorough knowledge of the design of fans; in fact, a large part
of his attention will require to be given to calculations relating
to the flow of air through passages of various kinds and under
various conditions. The prevention of noise in the operation
of machines which are cooled by air is a matter requiring much
study. There occur in all generators losses of a more or less
" "
obscure nature, appropriately termed stray losses, and a
wide experience in design is essential to minimizing such losses
and thereby obtaining minimum temperature rise and maximum
efficiency. Furthermore, mention should be made of the import-
ance of providing a design for appropriate wave shape and insur-
viii PREFACE

ing the absence of objectionable harmonics. To deal compre-

hensively with these and other related matters a very extensive
treatise would be necessary.
While it will now be evident that the designer must extend
his studies beyond the limits of the present treatise in acquaint-

ing himself with the important subjects above mentioned, the

familiarity with the subject of the design of polyphase generators
and motors which" can be acquired by a study of the present
treatise, will nevertheless be of a decidedly advanced character.
It can be best amplified to the necessary extent when the problems
arise in the course of the designer's professional work, by con-

sulting papers and discussions published in the Proceedings of

Electrical Engineering Societies.
Just as the design of machinery for continuous electricity
crystallizesaround the design of the commutator as a nucleus,
so in the design of polyphase generators, a discussion of the pre-
determination of the field excitation under various conditions
of load as regards amount and phase, serves as a basis for acquiring
familiarity with the properties of machines of this class. The
author has taken the opportunity of presenting a method of
dealing with the subject of the predetermination of the required
excitation for specified loads, which in his opinion conforms more
closely with the actual occurrences than is the case with any
other method with which he acquainted. In dealing with the
is

design of polyphase induction motors, the calculations crystallize

out around the circle ratio, and this may be considered the nucleus
for the design. The insight as regards the actual occurrences
in an induction motor which may be acquired by accustoming
one's self to construct mentally its circle diagram, should, in the
author's opinion, justify a much wider use of the circle diagram
than at present the case in America.
is

The authordesires to acknowledge the courtesy of the Editor

of the General Electric Review for permission to employ in Chap-
ter VI, certain portions of articles which the author first published
in the columns of that journal, and to. Mr. P. R. Fortin for
assistance in the preparation of the illustrations.
HENRY M. HOBART, M.Inst.C.E.
November, 1912.
 CONTENTS

PAGE
PREFACE . . . . v

CHAPTER I

INTRODUCTION 1

CHAPTER II

CALCULATIONS FOR A 2500-KVA. THREE-PHASE SALIENT POLE GENERATOR 3

CHAPTER III

POLYPHASE GENERATORS WITH DISTRIBUTED FIELD WINDINGS 99

CHAPTER IV
THE DESIGN OF A POLYPHASE INDUCTION MOTOR WITH A SQUIRREL-
CAGE ROTOR 105

CHAPTER V
SLIP-RING INDUCTION MOTORS 195

CHAPTER VI
SYNCHRONOUS MOTORS VERSUS INDUCTION MOTORS 202

CHAPTER VII
THE INDUCTION GENERATOR 213

CHAPTER VIII
EXAMPLES FOR PRACTICE IN DESIGNING POLYPHASE GENERATORS AND
MOTORS 225

APPENDIX 1 247
APPENDIX II 252
APPENDIX III 256
APPENDIX IV 257

INDEX 259
ix
DESIGN OF POLYPHASE GENERATORS
AND MOTORS

CHAPTER I

INTRODUCTORY

POLYPHASE generators may be of either the synchronous or

the induction type. Whereas tens of thousands of synchronous
generators are in operation, only very few generators of the induc-
tion type have ever been built. In the case of polyphase motors,
however, while hundreds of thousands of the induction type are
in operation, the synchronous type is still comparatively seldom

employed. Thus it is in accord with the relative importance of

the respective types that the greater part of this treatise is devoted

to setting forth methods of designing synchronous generators

and induction motors. Brief chapters are, however, devoted to
the other two types, namely induction generators and synchronous
motors. The design of an induction generator involves consider-
ations closely similar to those relating to the design of an induc-
tion motor. In fact, a machine designed for operation as an induc-
tion motor will usually give an excellent performance when
employed as an induction generator. Similarly, while certain
modifications are required to obtain the best results, a machine
designed for operation as a synchronous generator will usually
be suitable for operation as a synchronous motor.
In dealing with the two chief types: the author has adopted
the plan of taking up immediately the calculations .for a simple
design with a given rating as regards output, speed, periodicity
and pressure. In the course of these calculations, a reasonable
amount of familiarity will be acquired with the leading principles
involved, and the reader will be prepared profitably to consult
2 POLYPHASE GENERATORS AND MOTORS
advanced treatises dealing with the many refinements essential
to success in designing. A
sound knowledge of these refinements
can only be acquired in the course of the practice of designing as
a profession. The subject of the design of dynamo electric machin-
ery being continually discussed from various viewpoints in the
is

papers contributed to the Proceedings of engineering societies.

Attempts have been made to correlate in treatises the entire
accumulation of present knowledge relating to the design of dyna-
mo-electric machinery, but such treatises necessarily extend
into several volumes and even then valuable fundamental outlines
of the subject are apt to be obscured by the mass of details.
No such attempt is made in the present instance; the discussion
is fundamental outlines.
restricted to the
There are given in Appendices 1 and 2, bibliographies of

papers contributed to the Journal of the Institution of Electrical

Engineers and to the Transactions of the American Institute of

Electrical Engineers. These papers deal with many important
matters of which advanced designers must have a thorough
knowledge. After acquiring proficiency in carrying through the
fundamental calculations with which the present treatise deals,
a study of these papers will be profitable. Indeed such extended
study is essential to those engineers who propose to adopt designing
as a profession.
 CHAPTER II

CALCULATIONS FOR A 2500-KVA. THREE-PHASE SALIENT POLE

GENERATOR

LET
us at once proceed with the calculation of a design for a
three-phase generator. Let its rated output be 2500 kilo olt
amperes at a power factor of 0.90. Let its speed be 375 revolutions
per minute and let it be required to provide 25-cycle electricity.
Let it be further required that the generator shall provide a ter-
minal pressure of 12 000 volts. We shall equip the machine
with a Y-connected stator winding. The phase pressure will be:

12000 12000

The Number of Poles. Since the machine is driven at a speed

of (- =)6.25 revolutions per second, and since the required

,60
periodicity is 25 cycles per second, it follows that we must arrange
for:

25
r- = \ 4.0
) cycles per revolution,
.^o /

Consequently we must provide four pairs of poles, or (2X4 = )8

poles.
Denoting by P the number of poles, by R the speed in revolu-
tions per minute, and by ~ the periodicity in cycles per second,
we have the formula:

In this treatise the power factor will be denoted by G. Since

for our machine the rated load is 2500 kva. for (7 = 0.90, we may
3
4 POLYPHASE GENERATORS AND MOTORS

also say that the design is for a rated output of (0.90X2500 = ) 2250
kw. at a power factor of 0.90.
So many alternators have been built and analyzed that the
design of a machine for any particular rating is no longer a matter
which should be undertaken without any reference to accum-
ulated experience. From experience with many machines, design-
ers have arrived at data from which they can obtain in advance
some rough idea of the proportions which will be most appro-

priate. It is not to be concluded that the designing of a machine

by reference to these datais a matter of mere routine copying.

On the contrary, even by making use of all the data available,

there is ample opportunity for the exercise of judgment and origin-
ality in arriving at the particular design required.
The Air-gap Diameter. Let us denote by D the internal
diameter of the stator. Usually we shall express D in centimeters
(cm.), but occasionally it will be more convenient to express it
in millimeters (mm.), and also occasionally in decimeters (dm.),
and in meters (m).
Since the internal diameter of the stator but slightly in excess
is

of the external diameter of the often convenient

rotor, it is
"
briefly to describe D as the air-gap diameter," but we must not
forget that strictly speaking, it is the internal diameter of the
stator and is consequently a little greater than the external
diameter of the rotor.
The Polar Pitch. Let us further denote the polar pitch,
(in cm.), by T. By the polar pitch is meant the distance, measured
at the inner circumference of the stator, from the center of one
pole to the center of the next adjacent pole. The polar pitch T
is the very first dimension for which we wish to derive a rough
preliminary value.
*

The values of T given in Table 1 are indicated

by experience to be good preliminary values for designs of the

numbers of poles, the periodicity and the output shown against
them in the Table.
The table indicates that the polar pitch T should have a value
of 70 cm.

Since the machine has 8 poles, the internal periphery of the

stator is :

8X70 = 560 cm.

CALCULATIONS FOR 2500-KVA. GENERATOR

TABLE 1. VALUES OF T AND .

6 POLYPHASE GENERATORS AND MOTORS

Consequently for D, the diameter at the air gap, we have :

In general, we have the relation

The Output Coefficient. be seen that in Table 1, each

It will
value of T is accompanied by a value designated as This .

" "
quantity is termed the output coefficient and its general
nature may be explained by reference to the following formula:

..
_ Output in volt amperes
(Dia. in decimeter)
2
X (Gross core length in dm.) XR'
In Table 1, the appropriate value for is given as 1.78.

$-1.78.

Transposing the output-coefficient formula and substituting

all theknown quantities, we have:
2 500 000
Gross core length (in dm.) =
17 32x375x1 73
= 11.8 dm.
= 118 cm.
The term \g is employed to denote the gross core length in cm.
Thus we have :

D
and \g are two of the most characteristic dimensions of the
design, and indicate respectively the air-gap diameter and the
gross core length of the stator core.
It is not to be concluded that we shall necessarily, in the com-

pleted design, adhere to the precise values originally assigned

to these dimensions. On the contrary, these preliminary values
simply constitute starting points from which to proceed until
the design is sufficiently advanced to ascertain whether modified
dimensions would be more suitable.
Discussion of the Significance of the Output Coefficient
Formula. Let us consider the constitution of the formula above
 CALCULATIONS FOR 2500-KVA. GENERATOR

given for ? the output coefficient. Two of the terms of the denom-
inator are D2 D
and \g. Their product, or 2 \g, is of the nature of

a volume. If we were to multiply it by we should have the vol-

ume of a cylinder with a diameter D and a length \g. But leav-

- does not alter the nature of the

ing out the constant expression

It is still of the nature of a volume. Thus:

t _ Output in volt amperes

(A volumetric quantity) XR'

Therefore, = output per unit of a volumetric quantity

2
(D \g) and per revolution per minute (R). In other words, ?
is the output per unit of D \g per revolution per minute.
2

Obviously the greater the value of the more are we obtaining

,

from every unit of volume. It is now easy to understand that

in the interests of obtaining as small and low-priced a machine
as is consistent with good quality, we must strive to employ the
highest practicable value of ?, the output coefficient.
While in most of our calculations we shall express D and ~kg
in centimeters, occasionally be more convenient, on account
it will

of the magnitude of the results involved, to express these quan-

tities in decimeters (as in the output coefficient formula), and,

in other instances, in meters. In Table 2, are given some rough

values connecting the total net weight with D2 \g.
TABLE 2. RELATION BETWEEN D*\g AND TOTAL NET WEIGHT.

D*\g (D and \g in
meters) .
8 POLYPHASE GENERATORS AND MOTORS
The Peripheral Loading. Having determined upon the per-
iphery (xXD or PXT), the next point to be decided relates to the
"
so-called peripheral loading." The peripheral loading may be
defined ampere-conductors per centimeter of air-gap
as the
periphery at the rated load of the machine. The range of appro-
priate values has been arrived at by experience with successive
designs. Such values are given in Table 3. They are only rough
indications, 'but they are of assistance in the preliminary stages
of the preparation of a design.

TABLE 3. THE PERIPHERAL LOADING.

Rated
Output
(in kva.).
 CALCULATIONS FOR 2500-KVA. GENERATOR
We can now ascertain the number of conductors per pole per
phase by dividing the ampere-conductors by the amperes, as
follows :

^ A
Conductors = 583 48.6.

This result must be rounded off to some value which will be

divisible by the number of slots per pole per phase.
The Number of Slots. The performance of a
machine is in
most respects more satisfactory the greater the number of slots
per pole per phase. But considerations of insulation impose
limitations. For 12 000-volt generators we may obtain from
Table 4, reasonable preliminary assumptions for the number of
slots per pole per phase, for three-phase generators, as a function
of T, the polar pitch:

TABLE 4. RELATION BETWEEN T AND THE NUMBER OF SLOTS PER POLE

PER PHASE.

T the Polar Pitch in cm.

10 POLYPHASE GENERATORS AND MOTORS

considerable as to be a serious impediment to the escape of heat

from the armature conductors. Consequently it is necessary
to proportion these conductors for a lower current density than
would be necessary for a machine for lower pressure. A current
density of some 3.0 amperes per sq.mm. will be suitable. The
current per conductor is 120 amperes. Consequently each con-
ductor must have a cross-section of some:

120
= 40 sq.mm.

Let us 12 mm. wide by 3.3 mm. high and

make each conductor
let us arrange above each other the ten conductors which occupy
each slot. Each conductor will be insulated by impregnated
braid to a depth of 0.3 mm. Consequently the dimensions over
the braid will be:

12.6X3.9.

The ten conductors will thus occupy a space 12.6 mm. wide by

10X3.9 = 39.0 mm. high.

The Slot Insulation. Outside of this group of conductors

comes the slot lining. In Table 5 are given suitable values for
the thickness of the slot lining for machines for various pressures :

TABLE 5. THICKNESS OF SLOT INSULATION.

Normal Pressure in Volts.

 CALCULATIONS FOR 2500-KVA. GENERATOR 11

For our alternator the slot insulation should have a thickness

of 5.6 mm. all round the group of conductors. The dimensions
to the outside of the insulation will thus be:

2X5.6+12.6 = 11.2+12.6 = 23.8 mm.,

by:

2X5.6+39.0 = 11.2+39.0 = 50.2 mm.

Thus the insulated dimensions are:

23.8X50.2.

The Dimensions of the Stator Slot. Allowing 7 mm. at the

top of the slot for
a retaining
wedge and allowing 0.3 mm. addi- -24.1mm

tional width for tolerence in

assembling the slotted punchings,

we arrive at the following dimen-
sions for the stator slot :

Depth = 50.2 +7 =57.2 mm.;

Width = 23.8+0.3 = 24.1 mm.

The above design for the slot

L.1
is shown in Fig. 1. This can
t
only be regarded as a prelimin- FIG. 1. Stator Slot for 12 000-volt,
ary design. At a later stage of 2500-kva., 375 r.p.m., 25-cycle,
the calculations, various consid- Three-phase Generator.
erations, such as tooth density,
may require that the design be modified.
The Slot Space Factor. The gross area of the slot is:

57.2X24.1 = 1380 sq.mm.

Thus the copper only occupies:

= 29.0 per cent.

of the total available space.

12 POLYPHASE GENERATORS AND MOTORS
" "
We express this by stating that the slot space factor is

equal to 0.29. The slot space factor is lower the higher the work-

ing pressure for which the alternator is designed.

THE MAGNETIC FLUX

Having now made provision for the armature windings, we

must proceed to provide for the magnetic flux with which these
windings are linked. Let us start out from fundamental con-
siderations.

Dynamic Induction. A conductor of 1 cm. length is moved

in a directionnormal to its length and normal to the direction
of a

magnetic field, of which the density is 1 line per square centimeter.

Let the velocity with which this conductor is moved, be 1 cm. per
second. Then the pressure set up between the two ends of this
conductor will be 1 X 1 X 1 X 10~ 8 volt. If the conductor, instead
of being of 1 cm. length, has a length of 10 cm. then the pressure
between the ends becomes

10 X 1 XI X10~ 8 volts.

If the density instead of being 1 line per sq.cm.,

field is 10 000
lines per sq.cm., then the pressure will be

10 X 10 000 X 1 X 10~ 8 volts.

If the speed of the conductor is 1000 cm. per second instead of

1 cm. per second, then the pressure will be

10 X 10 000 X 1000 X 10~ 8 volts.

Obviously with our 1 cm. conductor with a velocity of 1 cm. per

sec.,and moving in a field of a density of 1 line per sq. cm.,
the total cutting of lines was at the rate of:

1 X1X1=1 line per second.

But with our 10 cm. conductor moving in a field of a density of

10 000 lines per sq.cm., and with a velocity of 1000 cm. per
sec., the total cutting is at the rate of:

10 X 10 000 X 1000 = 100 000 000 lines per second.

 CALCULATIONS FOR 2500-KVA. GENERATOR 13

In other words, the pressure, in volts, at the terminals of the

conductor equal to the total number of lines cut per second,
is
~8
multiplied Let us now, instead of considering a straight
by 10 .

conductor, take a case where the conductor is bent to constitute

the sides of a turn. Let this turn occupy a plane normal to a
uniform flux. If in this position, the portion of the flux passing
through this turn is 1000000 lines (or 1 megaline), then if in
1 second the coil is turned so that its plane is in line with the

direction of the flux, there will in this latter position be no flux

linked with the turn. The flux linked with the turn will, in 1
second, have decreased from 1 megaline to 0. The conductor
constituting the sides of this turn will have cut 1000000 lines.

If, during this 1 second, the rate of cutting is uniform, then the
pressure at the terminals will be

1000000XlO- 8 = 0.0100 volt.

Thus 1 turn revolved in second through 90 degrees from a

1

position where it is linked with 1 megaline into a position in

which it is parallel to the flux, will have induced in it a pressure of
0.0100 volt. Obviously at this speed, 4 seconds are required
for this turn to make one complete
revolution, i.e., to complete
one cycle. The thus only 0.25 cycle per second.
periodicity is

If the speed is quadrupled, corresponding to one cycle per second,

then the pressure will be

4X0.01=0.0400 volt.

The Pressure Formula. If instead of having one turn, we have

a coil of T turns, and if a periodicity of
this coil revolves at ~
cycles per secondhand if, instead of being linked with 1 megaline,
the flux is M
megalines, then the pressure at the terminals of the
winding is :

This is the general formula for the average e.m.f. induced

in the windings of an alternator. But usually we are concerned
with the mean effective value and not with the average value.
14 POLYPHASE GENERATORS AND MOTORS
For a sine-wave curve of e.m.f. the mean effective value is 11
per cent greater than the average value. Consequently for the
mean effective value of the e.m.f., the formula becomes

F = 1.11X0.0400XTX~XM,

or

XM.

Modification of Pressure Formula. The formula only holds

when the turns of the winding are (when at the position of
all

maximum linkage of flux and turns), simultaneously linked with

the total flux M. This is substantially the case with the windings
of any one phase of a three-phase machine. A reference to Fig.
2 in which are drawn the windings of only one of the three phases,
will show that the windings of one phase occupy only one-third
of the polar pitch and so long as the pole arc of the field pole does
not exceed two-thirds of the polar pitch, there is complete linkage
of all the turns with the entire flux. But in a two-phase machine
(often termed a quarter-phase machine), the windings of each
phase occupy one-half of the polar pitch and in such a case as
that shown in Fig. 3, in which are drawn the windings of only
one of the two phases, it is seen that the innermost turns of one
phase are not linked with quite the entire flux.
The Spread of the Winding. We express these facts by stating
that in a three-phase winding, the spread of the winding is 33.3
per cent of the polar pitch and that in a two-phase winding the
spread of the winding is 50 per cent of the polar pitch. In view
of the allowances which must sometimes be made for incomplete
simultaneous linkage of flux and turns, it is preferable to replace
the formula:

by the formula:

V=KXTX~ XM.
CALCULATIONS FOR 2500-KVA. GENERATOR 15

\j

\J

3
16 POLYPHASE GENERATORS AND MOTORS
The coefficient K for various winding spreads and for various
ratios of pole arc to pitch may be obtained by reference to the
curves in Fig. 4.

10 20 30 40 50 60 70 80 90 100
Spread of CoiLin Percent of Pitch

FIG. 4. Curves for Obtaining Values of K in the Formula V = KXTX~XM.

Full and Fractional Pitch Windings. In the above statements,

it has been tacitly assumed that the windings are of the full-
pitch type. In a full-pitch winding, the center of the left-hand
side of agroup of armature coils is distant by T, the polar pitch
from the center of the right-hand side of the group of armature
coils. Thus in Fig. 5, is shown a full-pitch winding. But it is
sometimes desirable to employ windings of lesser pitch. These
are called fractional pitch windings. In Figs. 6 to 11, are shown
windings in which the winding pitches range from 91.5 per cent
down to 50 per cent of the polar pitch. For such windings it
is necessary to introduce another factor into the pressure formula

and we can employ the factor K' which we may term the wind-
ing-pitch factor. If the winding pitch is x per cent then K'

will be equal to

sn
 CALCULATIONS FOR 2500-KVA. GENERATOR 17

MEliJ^
FIG. 5. Diagrammatic Representation of a Full-pitch, Three-phase Winding.

FIG. 6. Diagrammatic Representation of a Three-phase Winding in which

the Winding Pitch is 91.5 per cent of the Polar Pitch.

FIG. 7. Diagrammatic itepresentation of a Three-phase Winding in which

the Winding Pitch is 83.5 per cent of the Polar Pitch.

M H M M H M \
B
\ \
B
\ M W \
c
\ \
c
\ M \
c
\ \
A

FIG. 8. Diagrammatic Representation of a Three-phase Winding in which

the Winding Pitch is 75 per cent of the Polar Pitch.

A A
H r r r r r

FIG. 9. Diagrammatic Representation of a Three-phase Winding in which

the Winding Pitch is 66.7 per cent of the Polar Pitch.

C\ \0\ \A\ A Lt \A\ \B\ IB] Lc \B\ \C\ \C

FIG. 10. Diagrammatic Representation of a Three-phase Winding in which

the Winding Pitch is 58.5 per cent of the Polar Pitch.

r r

FIG. 11. Diagrammatic Representation of a Three-phase Winding in which

the Winding Pitch is 50.0 per cent of the Polar Pitch.
18 POLYPHASE GENERATORS AND MOTORS
Thus when the winding pitch is 80 per cent, the winding -pitch
factor is equal to

sin X90 = sin 72 = 0.951.

Digression Regarding Types of Windings. An examination of

Fig. 5 will show that the conductors contained in any one slot
belong exclusively to some one of the three phases. On the contrary
in the case of fractional-pitch windings, there are a certain propor-
tion of the slots containing conductors of two different phases.
As a consequence, fractional-pitch windings should be of the two-
" "
layer type. This leads to the use of lap windings such as are
employed for continuous-electricity machines. But full-pitch wind-
" " " "
ings may be carried out either as lap windings or as spiral
windings, and consequently as single-layer or as two-layer windings.
The two varieties are shown diagrammatically in Figs. 12
and 13 on the following page.
While dealing with this subject of windings the occasion is

appropriate for explaining another terminological distinction.

Windings may be carried out in the manner termed " whole-
coiled," in which, for any one phase, there is an armature coil
opposite each field pole, as indicated in Fig. 14, or they may
be carried out as half-coiled windings, there being, for any one
phase, only one armature coil per pair of poles instead of one
armature per pole. A half-coiled winding is shown in Fig.
coil

15. usual to employ half-coiled windings for three-phase

It is

generators where the spiral type is adopted. But lap windings

are inherently whole-coiled windings as may be seen from an
inspection of Fig. 12.
It is mechanically desirable to employ fractional-pitch windings
in bipolar designs, otherwise the arrangement of the end
as
connections presents difficulties. In most instances of designs for
more than two poles, a full-pitch winding is quite as suitable as
any other. Let us employ a full-pitch winding in our 2500-kva.
machine.
Estimation of the Flux per Pole. We can now estimate the
flux required, when, at no load, the terminal pressure is 12 000
volts. This pressure corresponds to
12000 =
6950 volts per phase.
y=r-
CALCULATIONS FOR 2500-KVA. GENERATOR 19

FIG. 12. 6-Pole Lap Winding with 72 Conductors.

FIG. 13. 6-Pole Spiral Winding with 72 Conductors.

20 POLYPHASE GENERATORS AND MOTORS
Therefore

F = 6950.

We have determined upon employing 10 conductors per slot

and 5 Consequently for T, the number
slots per pole per phase.
of turns in series per phase, we have, (since the machine has 8

poles) :

FIG. 14. Whole-coiled Spiral FIG. 15. Half-coiled Spiral

Winding. Winding.

Thus we have:

6950 = 0.0444 X 200 X 25 XM.

Therefore

6950
M = 0.0444X200X25
31.3 megalines.

31.3 megalines must, at no load, enter the armature from each

field pole.
 CALCULATIONS FOR 2500-KVA. GENERATOR 21

The I R Drop at Full Load. At full load we must make an

allowance for the IR drop
in the armature winding. We have
previously ascertained that the full-load current is

I = 120 amperes.

Estimation of the Mean Length of Turn and of the Armature

Resistance. We must now estimate the armature resistance.
A convenient empirical formula for estimating the mean length
of one turn (mlt) is:

Appropriate values for K for machines for various pressures are

given in Table 6:

TABLE 6. VALUES OF K IN FORMULA FOR MLT.

Terminal Pressure for

which the Stator is
Wound.
22 POLYPHASE GENERATORS AND MOTORS

machinery), to correspond to a temperature of 60 Cent. At this

temperature, the specific resistance of commercial copper is :

0.00000200 ohm per crn. cube.

At 60 Cent, the resistance per phase, is, for our 2500-kva. machine :

Resis. = 0.00000200X656X200
-
~~o~3Qfi~

= 0.665 ohm.

Consequently, at full load, the IR drop per phase is :

120X0.665 = 80.0 volts.

Thus at full load the internal pressure is:

6950+80 = 7030 volts per phase.

At full load then, the flux entering the armature from each pole is :

7030
"0.0444X200X25
= 31.6 megalines.

In this particular case, the internal drop is so small that it is

hardly worth while to distinguish between the full-load and no-
load values of the flux per pole. But often (particularly in small,
slow-speed machines), there is a more appreciable difference
which must be taken into account.
The Design of the Magnetic Circuit. We are now ready to
undertake the design of the magnetic circuit for our machine.
In the stator, the magnetic circuit must transmit a flux of 31.6
megalines per pole. This is the flux which must become linked
with the turns of the armature winding. But a somewhat greater
flux must be set up in the magnet core, for on its way to the arma-
ture this flux experiences some loss through magnetic leakage.
The ratio of the flux originally set up in the magnet core to that
finally entering the armature is termed the leakage factor.
 CALCULATIONS FOR 2500-KVA. GENERATOR 23

Leakage Factor. In our machine, we shall assume the leakage

factor to be 1.15; is to say, the quantity of flux taking its
that
rise in the magnet core is 15 per cent greater than the portion
finally entering the armature and becoming linked with the turns
of the armature windings.

Consequently the cross-sections of the magnet cores and of

the yoke must be proportioned for transmitting this 15 per cent
greater flux.

1.15X31.6 = 36.3 megalines.

Material and Shape of Magnet Core. We shall, in this

instance, makethe magnet core of cast steel and we shall employ

a magnetic density of 17 500 lines per sq.cm. Consequently we
shall require to provide a cross-section of

36 300 000
sq.cm.

A circular magnet core of this cross-section, would have a

diameter of

1X208U = ri .

51.4 cm.

At the air-gap, the polar pitch, T, is equal to 70 cm., but it

becomes smaller as we approach the inner ends of the magnet
cores and there would not, in our machine, be room for cores of
a diameter of 51.4 cm., even aside from the space required on these
cores for the magnet windings. Let us take as a preliminary
assumption for the radial length of the magnet cores (including

pole shoes) 28 cm. This gives a diameter at the inner ends of

the magnet cores (neglecting the radial depth of the air-gap),
of

178- 2X28 = 122 cm.

*

The polar pitch at this diameter is :

24 POLYPHASE GENERATORS AND MOTORS

Obviously, then, we could not employ a magnet core of circular

cross-section, for we have seen that its diameter would require to
be 51. 4 cm. Out of the total available circumferential dimen-
sion of 48.0 cm., let us take the lateral dimension of the magnet
core as 26 cm., and let us constitute the section of the magnet
core, of a rectangle with a semi-circle at
each end, as shown in Fig. 16. The diam-
eter of the semi-circle is 26 cm. Conse-
quently the area provided by the two
-26 cm-
semi-circles amounts to:

= 530sq.cm.

This leaves

2080 -530 = 1550 sq.cm.

to be provided by the rectangle. Thus the

FIG. 16. Cross-section
length of the rectangle (parallel to the shaft)
must be:
of Magnet Core of

8-pole, 375 r.p.m.,

1550
2500-kva. Three- = 59.6 cm. (or 60 cm.).
phase Generator. 26

The overall length of the magnet core is, then, 26+60 = 86 cm.
The pole shoe may be made 114 cm. long, thus projecting:

114-86 =
14 cm.

at each end of the magnet core. Let us make the pole arc equal

to 42 cm. This is = 60 per cent of the pitch at the

|^X100 J

air gap. We may allow 8 cm. for the radial depth of the pole
shoe at the center. This leaves 28 8 = 20 cm. for the radial
length along the magnet core which is available for the magnet
winding. The depth available for the winding is at the lower
end of the magnet core:

48-26
- =11 cm.
 CALCULATIONS FOR 2500-KVA. GENERATOR 25

If we make our magnet spool winding of equal depth from top

to bottom, then, allowing a centimeter of free space at the lower
end of the spool, the two dimensions of the cross-section of the
spool winding will be :

20 cm.XlO cm.
It remains to be ascertained at a later stage whether this space
is sufficient accomodate the required ampere-turns.
to
The Cross-section of the Magnetic Circuit at the Stator Teeth.
The next part of the magnetic circuit which we should investigate,
is the section at the stator teeth. We must first determine upon
suitable proportions forthe ventilating ducts. A
suitable number
of ventilating ducts may be arrived at from the data in Table 7.

TABLE 7. DATA REGARDING VENTILATING DUCTS.

Peripheral
Speed in Meters
per Second.
26 POLYPHASE GENERATORS AND MOTORS
We shall provide 20 ducts and each duct shall have a width of
15 mm. Thus the aggregate width occupied by ventilating
ducts is 20X1.5 = 30 cm. But 10 per cent of the " apparent
"

thickness of the core laminations is occupied by layers of varnish

by means of which the sheets are insulated from one another in
order to prevent eddy currents.
The Net Core Length. The length of active magnetic material,
after deducting the space occupied by the ventilating ducts and
by the insulating varnish, may be termed the net core length and
may be designated by \n. For our 2500-kva machine we have :

<r = ;113-30)X0.90 = 79.2 cm.

t

The width of the stator slot has been calculated and has been
ascertained to be 24.1 mm. There are 15 teeth per pole (since there
are 5 slots per pole per phase). The tooth pitch at the air-gap is

. mm.
lo

Consequently the width of each tooth is

(46.6-24.1 =)22.5 mm.

The aggregate width of the 15 teeth is

15X2.25 = 33.8 cm.

Thus the gross cross-section per pole, at the narrowest part

of the stator teeth, is

79.2X33.8 = 2680 sq.cm.

But only a portion of this section will be employed at any one

time for transmitting the flux per pole, for the pole arc is only
60 per cent of T. The portion directly opposite the pole face is
0.60X2680 = 1610 sq.cm.

On the other hand, the lines will spread considerably in crossing the
gap, and this spreading will increase the cross-section of the stator
teeth utilized at any instant by the flux per pole. Let us in
our machine take the spreading factor equal to 1.15. Conse-
quently we have :

Cross-section of magnetic circuit at stator teeth =

1.15X1610 = 1850 sq.cm.
 CALCULATIONS FOR 2500-KVA. GENERATOR 27

The density at the narrowest part of the stator teeth is:

31 600 000
= 17 100 lines per sq.cm.
1850

This is a suitable value; indeed a slightly higher value, say any-

thing up to 19 000 lines per sq.cm., could have been employed.

Density and Section in Stator Core. For the density in the
stator core, back we may take 10 000 lines per sq.cm.
of the teeth,
The 31.6 megalines per pole.
total flux in the stator is This flux,
after passing along the teeth, divides into two equal parts and
flows off to right and left to the adjacent poles on either side,
as indicated diagrammatically in Fig. 17. Consequently the
magnetic cross-section required
in the stator core is:

31 600 OOP
= 1580 sq.cm.
2X10000
Since \n is equal to 79.2 cm.
the radial depth of the stator
core back of the slots must be

FIG. 17. Diagrammatic Representation

!lT 20cm -

of the Path of the Magnetic Flux in

the Magnet Core, Pole Shoe, Air-gap,
The depth is 5.72 cm.
slot
Stator Teeth, and Stator Core.
Consequently the total radial
depth of the stator core from the air-gap to the external periphery
is 20.0+5.7 = 25.7 cm. D, the internal diameter of the stator
punching, is 178 cm. Consequently the external diameter of
the stator punching is:

178+2X25.7 = 229.4 cm. (or 230 cm.)

In Fig. 18 is given a drawing of the stator punching, and in Fig.

19 is shown a section through the stator core with its 20 ventilating
ducts.
Even now we are not quite ready to consider the question of
assigning a suitable value to the depth of the air-gap. But
pending arriving at the right stage of the calculation, let us pro-
ceed on the basis that the depth of the air-gap is '2 cm. This
permits us to list a number of diametrical measurements, as
follows :
28 POLYPHASE GENERATORS AND MOTORS

Preliminary Tabulation of Leading Diameters.

External diameter of stator core 2300 mm.

Diameter at bottom of stator slots (1780+2x57 = ). . . 1894 mm.
Internal diameter of stator (D) 1780 mm.
External diameter of rotor (1780-2X20) 1740 mm.
Diameter to bottom of pole shoes (1740-2X80) 1580 mm.
Diameter to bottom of magnet cores (1580 2X200). . 1180 mm.

FIG. 18. Dimensions of Stator Lamination for 8-pole, 375-r.p.m., 2500-kva.,

Three-phase Generator.

-15 mm
FIG. 19. Section through Stator Core of 8-pole, 375-r.p.m., 2500-kva., Three
phase Generator, showing the 20 Ventilating Ducts.
 CALCULATIONS FOR 2500-KVA. GENERATOR 29

An end view of the design of the magnetic circuit so far as we

have yet proceeded, is shown in Fig. 20.
We are now ready to work out the cross-section of the so-
" "
called magnet yoke which completes the magnetic circuit
between the inner extremities of the magnet cores. We may
employ a density of 12 000 lines per sq.cm. Here also the flux
from any one magnet core divides into two equal halves which
flow respectively to the right and to the left on their wav to the

FIG. 20 End View of Design of Magnetic Circuit of 8-poie, 375-r.p.m., 250G

kva., Three-phase Alternator.

adjacent magnet cores on either side. Consequently for the

magnet yoke, we require a cross-section of

36 300 000

For the dimension parallel to the shaft, let usemploy 120 cm.
The radial depth of the magnet yoke should consequently be:

1510
12.6 cm.
120

Since the external diameter of the magnet yoke is1180 mm.,

the internal diameter is 1180-2X126 = 928 mm. The complete
30 POLYPHASE GENERATORS AND MOTORS
 CALCULATIONS FOR 2500-KVA. GENERATOR 31

list of leading diameters of the parts of the magnet circuit, is,

(pending arriving later at a final value for the air-gap depth),

as follows :

Extended Preliminary Tabulation of Leading Diameters.

External diameter of stator core 2300 mm.
Diameter at bottom of stator slots 1894 mm.
Internal diameter of stator (D) 1780 mm.
External diameter of rotor 1740 mm.
Diameter to bottom of pole shoes 1580 mm.
External diameter of magnet yoke 1180 mm.
Internal diameter of magnet yoke 928 mm.

These diameters and the leading dimensions parallel to the

shaft, are indicated in the views in Fig. 21.
Mean Length of Magnetic Circuit. The next step is to ascer-
tain the lengths of the various parts of the magnetic circuit, i.e.,
of the mean path followed by the magnetic lines. This mean
path is indicated in Fig. 22. The mmf. in the field winding on

FIG. 22. Diagram showing the Mean Path Followed by the Magnetic Lines
in an 8-pole Generator with an Internal Revolving Field.

any one of the magnet poles has the task of dealing with just
one-half of the complete magnetic circuit formed by two adjacent
poles. The lengths which we desire to ascertain are the lengths

corresponding to such a half circuit as indicated by the heavy

lines in Fig. 22.
32 POLYPHASE GENERATORS AND MOTORS
It is amply exact for our purpose to take the lengths in the
stator core and in the magnet yoke as equal to the mean circum-
ferences in those parts, divided by twice the number of poles.
The mean diameters in these parts are;

230.0+189.4
-
Mean diameter in stator core = =209.7 cm.

magnet yoke = 118.0+92.8

- = 105.4 cm.

The machine has 8 poles. Consequently:

Mean length magnetic circuit in stator core = 7 =41 cm.

length magnetic circuit in magnet yoke = = 21

' ''

Mean cm.

We also have:

Mean length of magnetic circuit in teeth =5.7 cm.

Mean length of magnetic circuit in magnet core = 20 cm.

The pole shoe is so unimportant a part that we may in the cal-

culation of the required magnetomotive forces (mmf.) neglect
it. As to the air-gap, we must, for reasons which will be under-
stood later, still defer taking up the calculations relating to it.

We may now make up the following Table :

TABULATION OF DATA FOR MMF. CALCULATIONS.

Designation of Parts
of Magnetic Circuit.
 CALCULATIONS FOR 2500-KVA. GENERATOR 33

circuit, we must consult saturation data of the materials of which

these parts are built. Appropriate values are given in Table 8:

TABLE 8. MAGNETOMOTIVE FORCE PER CM. FOR VARIOUS MATERIALS.

Density in Lines per

Square Centimeter.
34 POLYPHASE GENERATORS AND MOTORS
the tooth is 5.7 cm., consequently the total mmf. required for
the teeth is

5.7X57 = 325 ats.

Similar calculations may be made for the other parts. It is con-

venient to arrange these calculations in some such tabular form as
the following:

Designation of the Part of the

Magnetic Circuit.
 CALCULATIONS FOR 2500-KVA. GENERATOR 35

so complicate the matter as to considerably invalidate any theo-

retical deductions. But working backward from experimental
observations it has been ascertained that, independently of the
various relative dispositions of the windings and of other factors
of the design, results consistent with practice may be arrived at

by taking as the resultant mmf

of the three phases, 2.4 times the
.

mmf. of each phase. Consequently, for our design we have,

at full load, an armature mmf. of

2.4X3000 = 7200 ats.

It is only at zero power-factor that these armature ats. have the

same axis as the field ampere-turns. If, when the power-factor of

the external load is zero, the output is120 amperes, then, if the cur-
rent lagging, the resultant
is mmf. acting to send flux round the
magnetic circuit is obtained by subtracting 7200 ats. from the
excitation on each field pole. provided a mmf.
If each field coil

of 15 000 ats., then the resultant mmf. would, for this lagging
lead of 120 amperes, be

15 000 -7200 = 7800 ats.

If the 120 amperes were leading and if the power-factor were zero,
then the resultant mmf. would be

15 000+7200 = 22 200 ats.

For 120 amperes at other than zero power-factor, the armature

mmf. does not affect the resultant mmf. to so great an extent.
In a later section, we shall deal with a method of determining the
extent of the influence of the armature mmf. when the power-
factor is other than zero.
Even at this stage it is very evident from the phenomena that
have been considered, that the armature mmf. will at heavy loads,
and especially at overloads, exert a less disturbing influence the
greater the mmf. provided on the field spools, and that for a given
all-around quality of pressure regulation, the higher the armature
mmf., the higher should be the field mmf. A rule usually leading
to suitable pressure regulation, at usual power-factors, is to employ
in each field spool a mmf. equal to twice the armature strength.
A lower ratio is often employed; indeed it is often impracticable
to find room for field spools supplying so high a mmf. But let
36 POLYPHASE GENERATORS AND MOTORS
us endeavor to adhere to this ratio in the case of our example.
Thus our field mmf should be
.

2X7200 = 14400ats.
But we have seen that the iron parts of our magnetic circuit only
require a total mmf. of 2840 ats. How then can we employ a
total mmf. of 14 400 ats. and not obtain through the armature

winding, a greater flux than the 31.3 megalines which we have

found to correspond to the required pressure of 12 000 volts
(6950 volts per phase)?
We can so design the air-gap as regards density and length, as
to use up the remaining 14 400 2840 = 11 560 ats., in overcoming
itsmagnetic reluctance.
The Estimation of the Air-gap Density. First let us estimate
the air-gap density. We have
T= 70cm.
Pole arc = 0.60X70 = 42 cm.

Length of the pole shoe parallel to the shaft = 114 cm.

Area of pole face = 42X1 14 = 4800 sq.cm.

Density in pole face = 31300000 =

.
Konllines
6520
.

per sq.cm.

Many designers employ complicated methods for estimating the

density in the air-gap. These methods involve introducing
" "
spreading coefficients to allow for the flaring of the lines after

they have emerged from the pole face. They also involve cal-
culations of the area and density where it enters the
of the flux
armature surface. The author's own
custom, in the case of alter-
nators, is to usually take the air-gap density as equal to the pole-
face density, though it is quite practicable afterward to use
one's judgment in taking a somewhat higher or lower value
according as the other conditions indicate that the pole-face
density would be lower or higher than the air-gap density. In
this machine, we need make no corrections of this sort, but may
take the air-gap density as 6520 lines per sq.cm. This will
require a mmf. of

^X 6520 = 5200 ats.

per cm. of radial depth of the air-gap.

 CALCULATIONS FOR 2500-KVA. GENERATOR 37

The Radial Depth of the Air-gap. Evidently the air-gap

should have a radial depth of:

2.22 cm. =22.2 mm.

Since such calculations are necessarily only very rough, we shall

do well to modify this and make the air-gap only 18 mm. deep.
If, when the machine is tested, we find it desirable, we can
increase the air-gap by turning down the rotor to a slightly
smaller diameter.
The radial depth of the air-gap, in mm., may be denoted byA.
For our machine

A = 18.
Revised Tabulation of Leading Diameters. Having now
determined upon a value for the radial depth of the air-gap, let
us again tabulate the various diameters in our machine :

External diameter of stator core .................... 2300 mm.

Diameter at bottom .................
of stator slots 1894 mm.
Internal diameter of stator (D) .................... 1780 mm.
External diameter of rotor (D 2A) [Revised from
earlier table] ................................. 1744 mm.
Diameter to bottom of pole shoes ............ ...... 1580 mm.
External diameter of magnet yoke .................. 1180 mm.
Internal diameter of magnet yoke .................. 928 mm.

Saturation Curves at No Load. The next step is to work

out data from which we can construct a no-load saturation curve
for our machine, i.e., a curve in which ordinates will indicate the
pressure per phase, in volts, and abscissae will indicate magneto-
motive force per field spool in ats. We
have previously obtained
one point on this curve, namely the point for which we have
found ordinate, 6950 volts; abscissa, 14 400 ats. The process can
be considerably abbreviated in obtaining further points. Let us
work out the mmf. required for 6000 volts, and for 7500 volts.
With these three points we shall be able to construct the no-load
saturation curve.
38 POLYPHASE GENERATORS AND MOTORS
For the air-gap mmf we
. obtain the desired results by simple
proportion. Thus

For 6000 volts: 'Air-gap ats. = 560 = 10 000 ats.

For 7500 volts: Air-gap 560 = 12 500 ats.

But for the iron parts it is necessary first to obtain the flux density

by simple proportion and then from Table 8, on page 33, obtain

the corresponding values of the mmf. Thus for the teeth we have :

For 6950 volts: mmf. =57 ats. per cm. (from p. 34).

For 6000 volts :

Density = ^
fiOOO
X 17 100 = 14 800 lines per sq.cm.

Corresponding mmf. =16 ats. per cm.

7500
For 7500 volts :
Density X17 100= 18 400 lines per sq.cm.
6950

Corresponding mmf.
= 125 ats. per cm.
Since the length is 5.7 cm., we have:
Total mmf. for teeth:

For 6000 volts : 16 X 5.7= 90 ats.

For 6950 volts : 57 X 5.7= 330 ats.
For 7500 volts : 125 X 5.7 = 720 ats.

In the same way we estimate the corresponding values for

may
magnet core, magnet yoke, and stator core. It is needless to
record the steps in these calculations. The results are brought
together in the following table:
 CALCULATIONS FOR 2500-KVA. GENERATOR 39

The no-load saturation curve in Fig. 23 has been plotted from

these results.
8,000

7,000

6,000

75 5,000
>
.s

2
| 4,000

3,000
|

2,000

1,000
40 POLYPHASE GENERATORS AND MOTORS
in the calculations. Since, however, the occurrences assumed to
take place are qualitatively in accordance with the facts, it is
believed that the admitted defect is of minor importance.
Before proceeding to explain the method in applying it to our
2500-kva. design, let us bring together the leading data which we
have now worked out. This is done in the following specification:

SUMMARY OF THE NORMAL RATING OF THE DESIGN

Number of poles 8
Output at full load in kva 2500
Corresponding power-factor of external load .90
Corresponding output in kw 2250
Speed in r.p.m 375
Periodicity in cycles per second 25
Terminal pressure in volts 12 000
Number of phases 3
Connection of phases Y
Pressure per phase =- . . . . . 6950
( J

THE LEADING DATA OF THE DESIGN

External diameter of stator core 2300 mm.
Diameter at bottom of stator slots 1894 mm.
Internal diameter of stator (D) 1780 mm.
External diameter of rotor 1744 mm.
Diameter at bottom of pole shoes 1580 mm.
External' diameter of magnet yoke 1180 mm.
Internal diameter of magnet yoke 928 mm.
Polar pitch (T) 700 mm.
Gross core length (Xgr) 1180 mm.
Number of vertical ventilating ducts 20
Width each duct . 15 mm.
Per cent insulation between laminations 10 per cent
Net core length (1180-300) X0.90 (Xn) 792 mm.
Length pole shoe parallel to shaft 1140 mm.
Pole arc 420 mm.
Area of pole face (114X42) 4790 sq.cm.
Extreme length magnet core parallel to shaft. . . 860 mm.
Extreme width magnet core 260 mm.
Area of cross-section of magnet core 2080 sq.cm.
 CALCULATIONS FOR 2500-KVA. GENERATOR 41

Length of yoke parallel to shaft ............... 1200 mm.

Radial depth of yoke ........................ 126 mm.
Cross-section of yoke ................ ....... 1510 sq.cm. .

Number of stator slots ...................... 120

Number stator slots per pole per phase

Depth slot ................................. 57 mm.

Width slot ................................ 24 mm.
Width slot opening ......................... 12 mm.
Sketches of the design have already been given in Figs. 20
and 21. These preliminary sketches show wide-open slots. Let
us, however, employ slots with an opening of only 12 mm. in
accordance with the above tabulated specification.
The stator winding consists of 10 conductors per slot. The
bare dimensions of each conductor are 12 mm. X 3.3 mm., and
the 10 conductors are arranged one above the other as already
indicated in Fig. 1 on page 11. The mean length of one armature
turn is 656 cm. The winding is of the type which we have termed
" "
a half-coiled winding. That is to say, only half the field
poles have opposite to them, armature coils belonging to any
one phase. A
diagram of one phase of a typical half-coiled
winding for a 6-pole/ machine has been given in Fig. 15 on p. 20.
L
Q\
and. is seen to havef )
= three coils per phase. Since our 2500

kva. machine has 8 poles, there are four coils in each phase.
Each side of each coil comprises the contents of five adjacent slots.
Since each slot contains 10 conductors, there are (5 X 10 = ) 50 turns
=
per coil, and consequently (4X50 ) 200 turns in series per phase.
Denoting by T the number of turns in series per phase we have
^ = 200.
In Fig. 24 is given a winding diagram for one of the three
phases, and in Fig. 25, is given a winding diagram containing all

three phases.
/1 9
For a pressure of - 000\
- = 6950 volts per phase on open
(
\ v3 /
)

circuit, the armature flux per pole (denoted by M), is obtained

as follows:
42 POLYPHASE GENERATORS AND MOTORS
We may take the leakage factor as 1.15. Consequently the
flux in the magnet core and yoke is:

1.15X31.3 = 36.0 megalines.

FIG. 24. Winding Diagram for one of the Three Phases of the 2500-kva.
three-phase Alternator.

The estimation of the no-load saturation curve has previovsly

been carried out in earlier sections. A
summary of the component
and resultant magnetomotive forces (mmf.) for 6000, 6950, and
7500 volts, is given in the following table:
 CALCULATIONS FOR 2500-KVA. GENERATOR 43

A no-load saturation curve passing through these three points

has been given in Fig. 23, on page 39.
The Armature Interfering mmf. The current per phase at
rated load of 2500 kva. is:

1 ~ 2500000 " 12U '

3X6950

FIG. 25. Complete Winding Diagram for all Three Phases of the 2500-kva.
Generator.

We have seen that there are 200 turns in series in each phase.
Consequently there are:
200 or
-Q-
= 25 turns per pole per phase ,

o
and
25X120 = 3000 (rms.) ats. per pole per phase.

It has already been stated that many designers resort to

theoretical reasoning in ascertaining from the mmf. of one phase,
44 POLYPHASE GENERATORS AND MOTORS
the resultant mmf. exerted by the three phases. But, in practice,
the distribution of the stator and rotor windings, the ratio of the
pole arc to the pitch, and other details of the design, so complicate
the matter as considerably to invalidate any theoretical deduc-
tions. But working backward from a very large collection of
experimental observations, the conclusion is reached that inde-
pendently of the various relative dispositions of the winding and
of other features of the design, results consistent with practice
are obtained by taking the resultant mmf. of the three phases as

equal to:

2.4 times the mmf. of each phase.

Consequently for our design, we have, at full load, an armature

mmf. of:

2.4X3000 = 7200 ats.

It is only at zero power-factor that these armature ats. have

the same axis as the field ampere-turns. If, when the power-

factor of the external load is zero, the output is 120 amperes,

then, if the current is lagging, the resultant mmf. acting to send
flux round the magnetic circuit is obtained by subtracting 7200
ats. (the armature mmf.) from the excitation on each field pole.
If,with the power-factor again equal to zero, the current is lead-
ing, then the resultant mmf. acting to send flux round the magnetic
circuit is obtained by adding 7200 ats. (the armature mmf.) to the
excitation on each field pole. For this same current of 120
amperes, but at other than zero power-factor, the armature mmf.
does not affect the resultant mmf. to so great an extent.
Later we shall consider a method of determining the extent
of the influence of the armature mmf. when the power-factor
is other than zero. consequence of the preceding
It follows as a

explanations that the armature mmf. will exert a less disturbing

influence on the terminal pressure the greater the mmf. provided
on the field spools, and that for a given required closeness of

pressure regulation the higher the armature mmf., the higher must
also be the field mmf.
The modern conception of preferable conditions is not based
on such close inherent pressure regulation as was formerly con-
sidered desirable. The alteration in conceptions in this respect
 CALCULATIONS FOR 2500-KVA. GENERATOR 45

does not, however, decrease the importance of having at our dis-

posal means for accurately estimating the excitation required
under various conditions of load as regards pressure, power-factor
and amount.
The Field Excitation Required with Various Loads. The
required excitation is chiefly dependent upon three factors :

1. The no-load saturation curve of the machine;

2. The armature strength in ats. per pole;
3. The inductance of the armature winding.
In our 2500-kva three-phase machine, the armature strength
is equal to 7200 ats.
at rated load
The Position of the Axis of the Armature mmf If the arma-
.

ture winding had no inductance, then an external load of

for

unity power-factor, the axis of the armature magnetomotive

force would be situated just midway between two adjacent

poles; that is to say, there would be no direct demagnetization.

At the other extreme, namely for the same current output at zero
power-factor, the axis of armature demagnetization would cor-
respond with the field axis. The two cases are illustrated dia-
grammatically in Figs. 26 and 27. In our machine, when loaded
with full-load current of 120 amperes at zero power-factor, the
demagnetization would amount to 7200 ats. and this demagnet-
ization could only be offset by providing 7200 ats. on each field
pole. For power-factors between 1 and 0, the axis of armature
demagnetization would be intermediate, as indicated diagram-
matically in Fig. 28.
But we are not concerned with imaginary alternators with
zero-inductance armature windings, but with actual alternators.
In actual alternators, the armature windings have considerable
inductance. At this stage we wish to determine the inductance
of the armature windings of our 2500-kva alternator.
The Inductance of a i-turn Coil. Let us first consider a single
turn of the armature winding before it is put into place in the stator
slots. If we were to send one ampere of continuous electricity

through this turn, how many magnetic lines would be occasioned?

If the conductor were large enough to practically fill the entire

slot, then with the dimensions employed in modern alternators,

the general order of magnitude of the flux occasioned may be
ascertained on the basis that some 0.3 to 0.9 of a line would be
46 POLYPHASE GENERATORS AND MOTORS

Direction of dotation

FIG. 26 Diagrammatic Representation of Relative Positions of Axes of Field

mmf and Armature mmf for a Load of Unity Power-factor Neglecting
. .

Armature Inductance.

Direction of Rotation

FIG. 27. Diagrammatic Representation of Relative Positions of Axes of Field

mmf. and Armature mmf. for a Load of Zero Power-factor.

Direction of Rotation

FIG. 28. Diagrammatic Representation of Relative Positions of Axes of Field

mmf. and Armature mmf. for a Load of Intermediate Power-factor.
 CALCULATIONS FOR 2500-KVA. GENERATOR
linked with every centimeter of length of the turn. Taking the
mean value of 0.6 line per cm., then since in our design the
length of a turn is 656 cm., the flux occasioned by 1 amp. of
continuous electricity is

656X0.6 = 394 lines.

The inductance (expressed in henrys) of a 1-turn coil is equal

8
to 10~ times the number of lines linked with the turn when 1 amp.
of continuous electricity is flowing through the turn. Conse-
quently the inductance is, in this case, equal to

10~
8
X 394 = 0.00000394 henry.

The Inductance of a Coil with More than One Turn. The

inductance of any coil equal to the product (when 1 amp. of
is

continuous electricity flowing through the coil), of the flux

is

linked with the coil and the number of turns in the coil. This
definition is framed on the assumption that the entire flux is linked
with the entire number of turns. Where this is not the case,
appropriate factors must be employed in order to arrive at the
correct result.
In a two-turn coil, the mmf is, when a current of 1 amp. of
.

continuous electricity is flowing through the coil, twice as great

as in a one-turn coil of the same dimensions. Consequently for
a magnetic circuit of air, the flux will also be twice as great,
since in air the flux is directly proportional to the mmf. occa-

sioning it. But since this doubled flux is linked with double the
number and turns is four times
of turns, the total linkage of flux
as great. In other words, the inductance increases as the square
of the number of turns.
In Fig. 24, it has been shown that the winding of any one
phase of our eight-pole machine is composed of four coils in series.
Let us first consider one of those four coils. Each side com-
prises the contents of five slots. Since there are ten conductors
per slot, we see that we are dealing with a fifty-turn coil. On the
assumption that the incomplete linkage of flux and turns
sufficient
isprovided for by calculating from the basis of only 0.5 line per
centimeter of length, instead of from the value of 0.6 line per centi-
48 POLYPHASE GENERATORS AND MOTORS
meter of length which we employed when dealing with the one-
turn coil, we obtain for the inductance the value:

= 0.0082 henry.
502X^1x0.00000394

The Inductance and Reactance of One Phase. The winding

ofone phase comprises four such coils in series, and consequently
we have:

Inductance per phase = 4 X 0.0082 = 0.0328 henry

The reactance is obtained from the formula:

Reactance (in ohms)

where the periodicity in cycles per second is denoted by ^ and the

inductance in henry s by 1. We consequently have:

Reactance per phase = 6.28X25X0.0328 = 5. 15 ohms.

The Reactance Voltage per Phase. For our machine the

full-load current per phase is 120 amperes. Consequently when
carrying full-load current we have:

Reactance voltage per phase = 120X5. 15 = 618 volts.

The Inductance and Reactance of Slot-embedded Windings.

But up to this point we have considered that throughout their
length the windings are surrounded by air. In reality the wind-
ings are embedded in slots for a certain portion of their length.
For this embedded portion of their length, the flux, in lines per
centimeter of length, set up in a one-turn coil when one ampere
of continuous electricity flows through it, is considerably greater
than for those portions of the coil which are surrounded by air.

Suitable values may be obtained from Table- 9 :

TABLE 9. DATA FOR ESTIMATING THE INDUCTANCE OF THE EMBEDDED LENGTH.

No. of Lines
per cm.
Concentrated windings in wide-open, straight-sided slots .......... 3 to 6
Thoroughly distributed windings in wide-open, straight-sided slots 1.5 to 3
Concentrated windings in completely-closed slots .......... ....... 7 to 14
Thoroughly-distributed windings in completely-closed slots ........ 3 to 6
Partly distributed windings in semi-closed slots ................... about 5
 CALCULATIONS FOR 2500-KVA. GENERATOR 49

"
In order to illustrate the sense in which the terms con-
"
centrated," thoroughly-distrib-
"
uted" and partly-distributed"
windings are employed in the
above table, the three winding
diagrams in Figs. 29, 30 and 31
have been prepared. Evidently
for the windings of any one
phase of our machine, the
value of 5 lines per cm. of
embedded length is sufficiently
representative.
The embedded portion of the
length of a turn is equal to twice
the net core length. For our FIG. 29. Concentrated Winding.
machine we have :

Embedded length = 2X79 158 cm.;

=
Mean =
length of a turn 656 cm.
"
Free
"
length (i.e., the portion in air)
= 656 - 158 = 498 cm.

ury~ui^n^njij"yru^^

FIG. 30, FIG. 31.

Thoroughly Distributed Winding. Partly Distributed Winding.

We have calculated the inductance which our coil would have

if the entire 656 cm. of its length were surrounded by air (i.e.,
" "
were free length). We can now readily obtain the value of
50 POLYPHASE GENERATORS AND MOTORS
that part of the inductance which is associated with the actual
" "
free length. It amounts to:

iX 0.0328 = 0.0248 henry.

656

" "
The inductance of the embedded length is

j^XJ^X 0.0328 = 0.0790 henry.

The total inductance per phase is

0.0248+0.0790 = 0.104 henry.

It is interesting to note that

0.0248.
0.104
X 100 = 23.8 per cent

of the total inductance, is, in the case of this particular machine,

associated with the end connections.
Our estimate of the inductance has been so seriously interrupted
by explanatory text that it is desirable to set it forth again in a
more orderly form, and taking each step in logical order :

Mean length of turn .................. 656 cm.

" "
Freelength ...................... 498 cm.
"
Embedded " length .................. 158 cm.
" "
Flux per ampere-turn per 0.5 line for free length,
j
"
centimeter of length [
5.0 lines for "embedded length, J

" "
Flux per ampere-turn
(249
j
?go ^
lines for
for
M
free
embedded
length,
length<

Total flux per ampere-turn (

= 249+790=) 1040 lines.

Number of turns in one phase per pair of poles (i.e.,

per coil) ................................. 50
 CALCULATIONS FOR 2500-KVA. GENERATOR 51

Inductance of one coil

2
(1040X50 X10~
8 =) ...... 0.0260 henry
Number of coils (also pairs of poles) per phase ..... 4
Inductance of one phase (4X0.0260 = ) ........... 0.104 henry
Reactance of one phase at 25 cycles (6.28X25X0.104 = ) 16.3 ohms
Reactance voltage of one phase at 25 cycles and 120
amperes (120X16.3) = ..................... 1960 volts

Physical Conditions Corresponding to this Value of the

Reactance Voltage. This value of 1960 volts for the reactance
voltage, is of the order of the value which we should obtain

experimentally under the following conditions:

Twenty-five-cycle current is sent into the stator windings from
some external source, while the rotor (unexcited) is, by means of a
,

motor, driven at the slowest speed consistent with steady indica-

tions of the current flowing into the three branches of the stator

windings. Under these conditions, some 1960 volts per phase

would be found to be necessary in order to send 120 amperes into
each of the three windings.
Theta (0) and Its Significance. The value of the reactance
voltage thus determined, enables us to ascertain the angular
distance from mid-pole-face position at which the current in the
stator windings passes through its crest value.
Let this angle be denoted by 0. For a load of unity power-
factor, is the angle whose tangent is equal to the reactance
voltage divided by the phase voltage. Thus we have:
reactance voltage
6 = tan !
J
r .

phase voltage

The conception of may possibly be made clearer by stat-

ing that it represents the angular distance by which the center
of a group of conductors belonging to one phase has traversed

beyond mid-pole-face position when the current in these conduc-

tors reaches its crest value.
Theta at Unity Power-factor. For our example we have
(for unity power-factor) :

tan~ 1 0.282 = 15.9.*

*
In making calculations of the kind explained in this Chapter, the Table
of sines, cosines and tangents in Appendix III. will be found useful.
52 POLYPHASE GENERATORS AND MOTORS
The diagram is shown in Fig. 32. Strictly speaking, we ought
to take into account in the diagram, the IR drop in the armature.

FIG. 32. ^Diagram Relating to the Explanation of the Nature and Significance
of the Angle Theta (6).

The resistance per phase (at 60 Cent.), is 0.685 ohm. Conse-

quently for the full-load current of 120 amperes, we have:

IR drop = 120X0.685 = 82 volts.

The corrected diagram (taking into account the IR drop),

is shown in Fig. 33. In this diagram we have:

6-tBa-g^=tan->^taa-0^8*i5^

FIG. 33. More Exact Diagram for Obtaining 0.

Relation between Theta and the Armature Interference.

The armature demagnetization for any value of 6 is obtained by
multiplying the armature strength by sin 0.
For 120 amperes the armature strength is :

7200 ats.

We also have:

sin 6 =sin 15.8 = 0.270.

 CALCULATIONS FOR 2500-KVA. GENERATOR 53

Under these conditions (120 amperes output at unity power-

mmf. is equal to:
factor of the external load), the armature

0.270 X 7200 = 1940 ats.

The Hypothenuse of the 6-Triangle Has no Physical Exist-

ence. emphasis on the fact that
It is desirable to lay strong
the vector sum of 7032 volts and 1960 volts does not represent
an actually-existing internal pressure corresponding to an actual
flux of magnetic lines. The quantity which, earlier in this
"
chapter, has been termed the reactance voltage ", is made up
"
of two parts, associated respectively with the embedded "
" "
length and with the free length. While the portion associated
"
with the embedded " length manifests itself in distortion of
" "
the magnetic flux, the portion associated with the free length
acts in the same manner as would an equal inductance located
in an independent inductance coil connected in series with a non-
inductive alternator. (More strictly, it is only that portion of
" "
the free length which is associated with the end connections
which should be thus considered, and the portion associated
with the ventilating ducts should be placed in a different cate-
gory. But in practice the small margin provided by taking the
" "
entire free length, is desirable.)
"
We have seen that the inductance of the " free length of
the windings of our 2500-kva. machine is 23.8 per cent of the total
inductance ;

0.238X1960 = 466 volts.

The True Internal Pressure and Its Components. The three

components of the total internal pressure of our machine, when
the external load is 120 amperes at unity power-factor, are, per
phase :

Phase pressure . . . 6950 volts

82
"
IRdrop
' '
Reactance drop 466
54 POLYPHASE GENERATORS AND MOTORS
When these are correctly combined, as shown in Fig. 34, the
internal pressure is ascertained to be:

V?032 2 +4662 = 7050 volts.

The influence of the reactance voltage is thus (for these par-

ticular conditions of load), practically negligible, so far as concerns

occasioning an internal pressure appreciably exceeding the result-

ant of the terminal pressure and the IR drop.

FIG. 34. Pressure Diagram Corresponding to 6950 Terminal Volts and

120 Amperes at Unity Power-factor.

Total mmf Required at Full Load and Unity Power-factor.

.

From the no-load saturation curve in Fig. 23 on p. 39, we see

that we require:
15200 ats.

to overcome the reluctance of the magnetic circuit when the

internal pressure is:

7050 volts.

We require further:
1940 ats.

to offset the armature demagnetization for these conditions of

load (120 amperes at unity power-factor). Consequently we
require a total mmf. per field spool, of:

15 200+1940 = 17140 ats.

That is to say, for full-load conditions (6950 volts per phase

and 120 amperes at unity power-factor), we require an excitation
of:

17 140 ats.

The Inherent Regulation at Unity Power-factor. We can

now ascertain from the saturation curve the value to which the
 CALCULATIONS FOR 2500-KVA. GENERATOR 55

pressure will rise, when, maintaining constant this excitation of

17 140 ats., we decrease the load to zero. find the value of We
the pressure to be:
7350 volts.

Thus the pressure rise occurring when the load is decreased

to zero, is:

/^-6950 X100 \
xxtAft = KQ
per cent.
( 6950 J5.8

This is expressed by stating that at unity power-factor the inherent

regulation is:

5.8 per cent.

ESTIMATION OF SATURATION CURVE FOR UNITY POWER

FACTOR AND 120 AMPERES

Let us now proceed to calculate values from which we can

plot a load saturation curve extending from a pressure of volts
up to a pressure of 7500 volts for an external load of 120 amperes
at unity power-factor. We already have one point; namely:

17 140 ats. for 6950 volts.

For this unity power-factor, 120-ampere saturation curve, the

terminal pressure will be varied from up to a phase pressure
of say 7500 volts while the current is held constant at 120
amperes. The diagrams for obtaining fop 7500, 5000, 2500 and
volts are shown in Fig. 35. For these four cases we have:

-1
0.258 = 14.5 sin 14.5 = 0.250

2. = tan~ 1 = tan- 1
0.386 = 21.1 sin 21.1 = 0.360
oUo-^

3. 6 = tan- 1 ~? = tan- 1
0.758 = 37.1 sin 37.1 = 0.605

= tan- = 87.5 = 0.999

^ =tan~
1 1
4. 6 23.9 sin 87.5
56 POLYPHASE GENERATORS AND MOTORS

7500

5000

2500

,82

FIG. 35. Theta Diagrams for 120 Amperes at Unity Power-factor.

Since the armature current is, in all four cases, 120 amperes,
the armature strength remains 7200 ats. The armature demag-
netization amounts, in the four cases, to :

1. 0.250X7200 = 1800 ats.;

"
2. 0.360X7200 = 2590
3. 0.605X7200 = 4350
"
"
4. 0.999X7200 = 7200
 CALCULATIONS FOR 2500-KVA. GENERATOR 57

2500

FIG. 36. Pressure Diagrams for 120 Amperes at Unity Power-factor.

The Armature Reaction with Short-circuited Armature.

It isinteresting to note that in the last diagram in Fig. 35,
i.e., in the diagram relating to zero terminal pressure (short-
circuited armature) the angle is practically 90. Con-
sequently the armature reaction with short-circuited armature,
is practically identical with the armature strength expressed
in ampere-turns per pole.
The Required Field Excitation for Each Terminal Pressure.
The field excitation at each pressure, comprises two components.
The first of these components must be equal to the armature
demagnetization (in order to neutralize it), and the second com-
ponent must be of the right amount to drive the required flux
through the magnetic circuit in opposition to its magnetic reluct-
ance. This latter value may be obtained from the no-load
saturation curve in Fig. 23 (on p. 39), and must correspond to the
58 POLYPHASE GENERATORS AND MOTORS
four pressures obtained from the four diagrams in Fig. 36-
These four pressures are:

1. V7582 2 +4662 = 7590 volts.

2. V5082 2 +4662 =

3. \/2582 2 +466 2 =

4. V82 2 +4662 =

Thesaturation ats. for these four pressures are found from

Fig. 23 (on p. 39), to be as follows:

1.
 CALCULATIONS FOR 2500-KVA. GENERATOR 59

7,000

6,000

| 5,000
ri

4,000

-cfy

2,000

.1,000

N ^<ooQ033<o S o gf $ $ $ s
mmf per Field Spool in ats

FIG. 37. Saturation Curves for 7 = 120 and (r = 1.00 and for 7=0.

ESTIMATION OF SATURATION CURVE FOR UNITY POWER-

FACTOR AND 240 AMPERES
Let us now construct a saturation curve for unity power-
factorand 240 amperes output, i.e., for twice full-load current.
The diagrams of Figs. 35 and 36 are now replaced by those
of Figs. 38 and 39.
The reactance voltage for the 0-diagrams of Fig. 38 is now
twice as great as before, since the current is now 240 amperes
in place of 120 amperes.
The reactance voltage is now:
2X1960 = 3920 volts.

The 7.R drop is now :

2X82 = 164 volts.

60 POLYPHASE GENERATORS AND MOTORS
 CALCULATIONS FOR 2500-KVA. GENERATOR 61

The armature strength is now;

2x7200 = 14400ats.

Consequently in the four cases, we now have for the armature

demagnetizing ats. :

1. 0.456X14400 = 6580 ats.

"
2. 0.605X14400 = 8700
"
3. 0.826X14400 = 11900
4. 0.999X14400 = 14400
"

The internal inductance pressure is now:

2X466 = 932 volts.

The four internal pressures and the corresponding saturation

ats. are:

Internal Pressures. Sat. Ats.

1 . V7664 +932 = 7720

2 2
29 500

2. V51642 +9322 = 5250 9 350

3. V26642 +932 2 5150

4. V 1642+9322 = 945 1700

The total required ats. are shown in the last column of the
following tabulated calculation:
62 POLYPHASE GENERATORS AND MOTORS
The the last column are the basis for the unity
values in
power-factor, 240-ampere saturation curve shown in Fig. 40.
8,000

c3 J N c c eo co > o P
mmf per Field Spool, in ats ,

FIG. 40. Saturation Curves for Various Values of 7 and for (r = 1.00.

The unit}' power-factor, 120-ampere curve is also reproduced

from Fig. 37 and the 7=0 curve from Fig. 23.

SATURATION CURVES FOR POWER-FACTORS OF LESS THAN

UNITY
Let us now return to a load of 120 amperes, but let the power-
factor of the load on the generator be 0.90. Let us estimate the
required mmf. under these conditions, for terminal pressures of
7500, 5000, 2500, and volts, and then, from these four results,
let us plot a 0.90-power-f actor, 120-ampere, saturation curve.
The angle 0, i.e., the angle by which the conductors have passed
mid-pole-face position when carrying the crest current, is now
obtained as follows:
 CALCULATIONS FOR 2500-KVA. GENERATOR 63

When the power-factor of the external load is 0.90, the current

lags behind the terminal pressure by 26.0, since cos 26 = 0.90.
The 0-diagram for 120 amperes and a terminal pressure of
6950 volts, is now as shown in Fig. 41. The entire object of this
diagram is to obtain the angle 0, i.e., the angle by which the con-

FIG. 41. Theta Diagram for 7 = 120 FIG. 42. Pressure Diagram for
and G = 0.90. 7 = 120 and G=0.90.

ductors have passed mid-pole-face position when the current is

at its crest value.

AB = BCsm2Q
= 6950X0.438 = 3040
AC= 0.90X6950 = 6250

= AB+BE
~
AC+DE
^3040+1960
"6250+ 82
5000
6332
= 0.790.
Therefore
6 = 38.3
sin 38.3 = 0.620.
Therefore :

Armature demagnetizing ats. = 0.620X7200 = 4450 ats.

64 POLYPHASE GENERATORS AND MOTORS
The diagram for obtaining the internal pressure is shown in

Fig. 42. By scaling off from this diagram, we find that the internal
pressure is 7250 volts. From the no-load saturation curve we find :

Saturation mmf. for 7250 volts = 16 700 ats.

Thus the total required mmf. for a phase pressure of 6950

volts with a load of 120 amperes at a power-factor of 0.90, is:

4450+16 700 = 21 150 ats.

For loads of other than unity power-factor, the most expedi-

tious method of arriving at the results is usually that by graphical
constructions. In the chart of Fig. 43 which relates to the
graphical derivation of the saturation curve for 120 amperes at
0.90 power-factor, the diagrams in the right-hand column relate to
the determination of the internal pressure. The first, second,
third and fourth horizontal rows relate respectively to the diagrams
for phase pressures of 7200, 5000, 2500 and volts.
The left-hand vertical row of diagrams relates to the construc-
tions for the determination of 6 for these four terminal pressures.
From the internal-pressure diagrams in Figs. 42 and 43 and
from the no-load saturation curve in Fig. 23 we find:

Phase
Pressure.
 CALCULATIONS FOR 2500-KVA. GENERATOR 65

Theta Diagrams Pressure Diagrams

FIG. 43. Theta and Pressure Diagrams for 7 = 120 and = 0.90.

We are now in a position to obtain the total ats. for each value
of the phase pressure. The steps are shown in the following table :

Phase Pressure.
66 POLYPHASE GENERATORS AND MOTORS
These valuesfor 120 amperes at 0.90 power-factor and those

previously obtained for 120 amperes at unity power-factor, give

us the two load-saturation curves plotted in Fig. 44. We see
that for a phase pressure of 6950 volts, when the current is 120
amperes and at 0.90 power-factor, the required excitation is
21 150 ats. From the no-load saturation curve we find that an

8,000

7,000

I I -
of ? > oo- $ 3 3 g ? g g g o-
mmf per Field Spool, in ats

FIG. 44. Saturation Curves for 7 = 120 and for (7 = 1.00 and 0.90.

excitation of 21 150 ats. occasions, at no load, a phase pressure

of 7600 volts

7600-6950
6950
X 100 = 9.4.

The inherent regulation at 0.90 power-factor is, for this

machine, 9.4 per cent. In other words, if, for an output of 120
amperes at a power-factor of 0.90 we adjust the excitation to such
a value as to give a phase pressure of 6950 volts, and if, keeping
 CALCULATIONS FOR 2500-KVA. GENERATOR 67

the excitation constant at this value, the load is decreased to

zero, the pressure will rise 9.4 per cent.

ESTIMATION OF SATURATION CURVE FOR 120 AMPERES

AT A POWER FACTOR OF 0.80

Now let us carry through precisely similar calculations for

120 amperes at a still lower power-factor, namely, a power-factor
of 0.80. We shall first estimate the required mmf. (at 120 amperes
and 0.80 power-factor) for phase pressures of 7200, 5000, 2500 and
volts, and from these four results we can plot the required satu-
ration curve.
We have the relation;

cos" 1 0.80 = 37.0.

The reactance voltage and the internal IR drop remain the

same as in the diagrams of Fig. 43. From these data we readily
arrive at the diagrams of Fig. 45.
From the internal-pressure diagrams in Fig. 45 and from the
no-load saturation curve in Fig. 23, we arrive at the following
results :

Phase
Pressure.
68 POLYPHASE GENERATORS AND MOTORS

-82

,82

?=60.2

Theta Diagram Pressure Diagram

FIG. 45. Theta and Pressure Diagrams for 7 = 120 and G = 0.80.

From the data in the two preceding tables, we can obtain the
total ats. for each value of the phase pressure. This is worked
through in the following table:
 CALCULATIONS FOR 2500-KVA. GENERATOR 69

Phase Pressure.
70 POLYPHASE GENERATORS AND MOTORS

Volts

in

Pressure

Phase

,
 CALCULATIONS FOR 2500-KVA. GENERATOR 71

THE EXCITATION REGULATION

The inherent regulation is not the only kind of regulation
which it is necessary to take into consideration in connection
with the performance of a generator. There is also the
"
excitation regulation." This, for a given power-factor may
be defined as the percentage increase in excitation which is
required in order to maintain constant pressure when the out-
put is increased from no-load to any particular specified value of
the current.
For our design we have estimated that for a phase pressure
of 6950 volts at no load, the required excitation is :

14 400 ats.

For this same phase pressure but with an output of 120 amperes
per phase, the required excitations are :

17 140 ats. for G = 1.00

21 150 ats. for G = 0.90
22 600 ats. for G = 0.80
25000 ats. for G =

The corresponding values of the excitation regulation are:

= 19 -
per Cent f r G = IW
oi i KH _ 14400 \
X 100= 46 8 P er cent for
'
= 0.90
14 400 )

000-14
(25
Excitation Regulation Curves. Curves plotted for given
values of G, and of the phase pressure, with excitation as ordinates
and with current output per phase as abscissae, are termed
excitation regulation curves. We have values for such curves
so far as relates to 7 = and 7 = 120, but with respect to higher
values, we have but one point, namely :

7 = 240
Phase pressure = 6950
G = 1.00
Excitation = 23 000.
72 POLYPHASE GENERATORS AND MOTORS
Let us work out corresponding values for 7 = 240 and with
the other power-factors, namely,

G = 0.90, G = 0.80 and = 0.

OS68

SUIBJSBJQ

For these power-factors and also for G=1.00, the theta and
pressure diagrams are drawn in Fig. 48. With the values obtained
from these diagrams the estimates may be completed as follows :
CALCULATIONS FOR 2500-KVA. GENERATOR 73
74 POLYPHASE GENERATORS AND MOTORS
Saturation Curves for 240 Amperes. In the course of the
previous investigation we have had occasion to obtain the excita-
tion required for a phase pressure of 6950 volts and with an output
of 240 amperes. These values are:

G
 CALCULATIONS FOR 2500-KVA. GENERATOR 75

VOLT-AMPERE CURVES
From the data in Figs. 23, 46 and 50, relating respectively
to saturation curves for 7 = 0, 7 = 120 and 7 = 240, we can
" "
construct curves which may be designated volt-ampere

8000

7000

.2 5000

I
c

1
g 4000
\ \

3000

1000

20 40 60 80 100 120 140 160 180 200 220 240 260

Current^per Phase (in Amperes)

FIG. 51. Volt-ampere Curves for Various Power-factors and for a mmf. of
17 140 ats. per Field Spool.

curves, since they are plotted with the phase pressure in

volts as ordinates and with the current per phase, in amperes,
as abscissae.
For any particular volt-ampere curve the excitation and the
power-factor are constants. For the volt-ampere curves in Fig.
51, the excitation is maintained constant at 17 140 ats., the mmf.
required at 6950 volts, 120 amperes and unity power-factor.
76 POLYPHASE GENERATORS AND MOTORS

Comment is required on the matter of the value at which the

volt-ampere curves cut the axis of abscissae. This is seen to be
at 254 amperes. For this current, the mmf. required to over-
come armature demagnetization is obviously:

254
X 7200 = 15 300 ats.
120

Since the excitation is maintained constant at 17 140 ats., there is

a residue of :

17 140 -15 300 = 1840 ats.,

and this suffices to provide the flux corresponding chiefly to the

reactance of the end connections. We have seen on pp. 51 and 53
that the reactance of the end connections amounts to :

(16.3XP-238 = )3.88 ohms.

Consequently for 254 amperes the reactance voltage is:

254X3.88 = 990 volts.

The IR drop is:

254X0.685 = 174 volts.

The impedence voltage on short-circuit with 250 amperes is

consequently :

\/990 2 +174 2 = 1000 volts.

From the no-load saturation curve of Fig. 23, we see that

1800 ats. arerequired for a phase pressure of 1000 volts.

THE SHORT-CIRCUIT CURVE

When the stator windings are closed on themselves with no

external resistance, then the field excitation required to occasion
a given current in the armature windings must exceed the
armature mmf. by an amount sufficient to supply a flux corre-
sponding to the impedance drop.
The impedance is made up of two parts :
 CALCULATIONS FOR 2500-KVA. GENERATOR 77

THE REACTANCE OF THE END CONNECTIONS

AND
THE RESISTANCE OF THE WINDINGS

We have seen (on p. 51) that the reactance of one phase at

25 cycles, is 16.3 ohms. Furthermore we have seen (on p. 53)
that the reactance of the end connections is 23.8 per cent of this
value, or:

0.238X16.3 = 3.88 ohms.

Also we have seen (on p. 22) that the resistance per phase, at
60 C., is:

0.665 ohm.

Consequently, at 25 cycles, the impedance is :

V3.88 2 +0.665 2 = 3.94 ohms.

For any particular value of the current, the impedance drop

isobtained by multiplying the current by 3.94 ohms. Thus for
100 amperes we have an impedance drop of:

100X3.94 = 394 volts.

From the no-load saturation curve of Fig. 23, we find that for
a pressure of 394 volts per phase, a mmf of 700 ats. per field spool
.

isrequired.
There are 25 turns per pole per phase. Consequently for a
current of 100 amperes per phase, the armature mmf. amounts to:

2.4X25X100 = 6000 ats.

Thus to send 100 amperes per phase through the short-circuited

stator windings, there is required a mmf. of:

700+6000 = 6700 ats. per field spool.

78 POLYPHASE GENERATORS AND MOTORS

Making corresponding calculations for 200 and 300 amperes

we arrive at the following results, which are plotted in Fig. 52
:

Current in
Armature.
 CALCULATIONS FOR %500-KVA. GENERATOR 79

those given in the table on p. 42 to the extent indicated in

the following table:
80 POLYPHASE GENERATORS AND MOTORS
 CALCULATIONS FOR 2500-KVA. GENERATOR 81

eight values of the excitation given in the preceding table. We

have:

G.
82 POLYPHASE GENERATORS AND MOTORS
This is a consequence of the plan of limiting the pressure rise
by the saturation of the magnetic circuit. The inherent regula-
tion for both air-gaps, has, for full-load current of 120 amperes,
the values given in the following table:

7350 _ 6950
-
G = 1 .00 Inherent regulation = -^ X 100 = 5.8 per cent.

0,0

employing the smaller air-gap, we

If there are objections to
must look elsewhere for Let us examine for instance, into
them.
the question of the amount of current which, with normal excita-
tion of 9140 ats. (corresponding to 6950 volts and' 120 amperes
at unity power-factor) could flow on short-circuit. Without
going into the matter of the precise determination of the satura-
tion mmf., let us assign to this quantity the reasonable value of
700 ats. This leaves:

9140 -700 = 8440 ats.

for offsetting armature demagnetization. We have seen that the

resultant armature mmf. is 2.4 times the mmf. per pole per phase.
Thus we have:
8440
mmf. per pole per phase = =3500 ats.
-^T
Since we have 25 turns per pole per phase, the short-circuit
current is only:
3500 ,. A
-Q^-
= 140 amperes.
Zo

Thus with an excitation of 9140 ats. (the value corresponding

to full load at unity power-factor), a current overload of only:

16.7 per cent

suffices to pull the terminal pressure down to volts.

 CALCULATIONS FOR 2500-KVA. GENERATOR 83

Similarly with an excitation of 14 400 ats., (the value for 120

amperes at 6950 volts and 0.80 power-factor with a 6-mm. air-
gap), the short-circuit current is of the order of:

14400-1000 00 ,

=224 amperes.
-23X25-
Even with this greater excitation, the current on short-circuit
than twice full-load current.
is less

While the modern tendency is toward designing with limited

overload capacity, it is nevertheless impracticable to employ
80001
'Excitation Coutant at 14.400 ats,
9,140 ats.

7000

6000
\
>5000
.S

l\
4000

3000

2000

1000

20 40 60 80 100 120 140 160 180 200 220 240

Current per Phase, in Amperes.

FIG. 55. Volt-ampere Curves for 2500-kva. Alternator with 6-mm. Air-gap.

generators whose volt-ampere characteristics turn down nearly

so abruptly as do those in Fig. 55 which represent these two
values. In Fig. 55, the right-hand portions of the curves have
been drawn dotted, as it has not been deemed worth while to carry
through the calculations necessary for their precise predetermina-
tion. >

Let us now revert to our original design with the 18-mm.

air-gap which we have shown to possess the more appropriate
attributes.
84 POLYPHASE GENERATORS AND MOTORS

THE DESIGN OF THE FIELD SPOOLS

Our generator's normal rating is 2500 kva. at a power-factor

of 0.90 and a phase pressure of 6950 volts. The current per
phase is then 120 amperes. For these conditions the required
excitation is:

21 150 ats. per field spool.

The field spools must be so designed as to provide this mmf .

with an ultimate temperature rise of preferably not more than

45 Cent, above the temperature of the surrounding air.
The question of the preferable pressure to employ for exciting
the field, is one which can only be decided by a careful considera-
tion of the conditions in each case. The pressure employed in
the electricity supply station for lighting and other miscellaneous
purposes, is usually appropriate, although it is by no means out
of the question that it may be good policy in some cases to provide
special generators to serve exclusively as exciters. These exciters
should, however, be independently driven. In other words,
their speed should be independent of the speed of the generator
for which the excitation is provided. It is the worst conceivable

arrangement to have the exciter driven from the shaft of the alter-
nator, as any change in the speed will then be accompanied by
a more than proportional change in the excitation.
In general, the larger the generator or the more poles it has,
the higher is the appropriate exciting pressure. But it is difficult
to make any statement of this kind to which there will not be
many exceptions.
Let us plan to excite our 2500-kva. generator from a 500-volt
circuit and let us so arrange that when the machine is at its
ultimate temperature of (20+45 = ) 65 Cent., 450 volts at the
slip rings shall correspond to an excitation of 21 150 ats. The
remaining (500450 = ) 50 volts will be absorbed in the con-
trolling rheostat. It would not be prudent to plan to use up
the entire available pressure of 500 volts when obtaining the
mmf. of 21 150 ats., for this would leave no margin for discrepan-
cies between our estimates and the results which we should

actually obtain on the completed machine.

 CALCULATIONS FOR 2500-KVA. GENERATOR 85

Thus we have 450 volts for the eight spools in series, or:

450
r- = 56.3 volts per spool.
o

In Fig. 56 is shown a section through the magnet core and the

spool.

1149mm--

Section on A-B Looking in Direction of Arrows

FIG. 56. Sketches of Magnet Pole and Field Spool for 2500-kva. Alternator.

The inner periphery of the spool is

26X71+2X60 = 82+120 = 202 cm.

The outer periphery is 46X^+2X60 = 144+120 = 264 cm.

For the mean length of turn we have :

mlt. = 202+264 = 233 cm.

86 POLYPHASE GENERATORS AND MOTORS

Suppose we were to provide our normal excitation of 21 150 ats.

by means of a single turn carrying 21 150 amperes. For this

excitation the pressure per spool is 56.3 volts. Consequently the
pressure at the terminals of our hypothetical turn carrying
21 150 amperes is also the 56.3 volts allocated to each of the
eight spools. The resistance of the turn must consequently be:

56 3
'

0.00266 ohm.
21150

We have mlt. =233 cm.

Therefore since the specific resistance of copper per centimeter

cube, at 65 Cent., is 0.00000204, we have:

233X0.00000204
Cross-section of the conductor =
U.UlLJbo

= 0.179 sq.cm.

Now if we were to provide the entire excitation by a single

turn per spool as above suggested, the loss in field excitation
would be:

500X21150
kw
l
-

This over five times the output of our machine. Consequently

is

would be low say some 18 per cent. But also,

its efficiency

we should be running our conductor at a density of :

21 150
-f
= 118 000 amperes per sq.cm.
u. 17_^-
1 y

and would fuse long before this density could be reached. It

it

would, in fact, fuse with a current of the order of only some

1000 amperes. Also the loss of 10 600 kw. in the field spools
would suffice, even if the heat could be uniformly distributed

through the whole mass of the machine, to raise it to an exceed-

ingly high temperature. Even a very spacious engine room

would be unendurable with so great a dissipation of energy taking
place within it.
 CALC ULA TIONS FOR 2500-K VA . GENERA TOR 87

So let us look into the merits of employing 10 turns per spool

instead of only one turn per spool.
Since we require an excitation of 21 150 ats. per spool, the
21 150
current will, in this case, be only = 2115 amperes.
Since the mlt. is equal to 233 cm., the 10 turns will have a
length of
10X233 = 2330 cm.

The resistance of the 10 turns must now be:

-0.0266 ohm.
fff
Then we have :

2330X0.00000204
- = ni7n
Section =
Q ,.
0.179 sq.cm.

This is the same value as before. In fact, for, firstly, a given

pressure at the terminals of a spool; secondly, a given excitation
to be provided, and thirdly, a given mlt., the cross-section of the
conductor is independent of the number of turns employed to
provide that excitation, and it is convenient to determine upon
the cross-section by first assuming that a single turn will be
employed. Obviously the greater the number of turns per
spool, the less will be the current, and since the terminal pres-
sure is fixed, the less also will be the power. Thus with our
second assumption of a 10-turn coil, and only 2115 amperes in
the exciting circuit, the excitation loss is reduced to:

500X2115
- =1060kw -

and the efficiency of our 2500-kva. machine would rise to over

65 per cent.
With the endeavor to obtain a reasonably low excitation
loss, it is obviously desirable to employ as many turns as we can

arrange in the space at our disposal. We have already seen that

this space provides a cross-section of 10X20 = 200 sq.cm. Our
conductor has a cross-section of 17.9 sq. mm. and should thus
88 POLYPHASE GENERATORS AND MOTORS
have a bare diameter of 4.77 mm. The curves in Fig. 57 give the
thicknesses of the insulation on single, double and triple
cptton-
covered wires of various diameters.
If, in this case, we employ a double cotton covering, the insu-
lated diameter will be 4.77 +(2X0. 18) =5.13 mm. The term

0.3

w
I 0.2
^
4P
a

0.1

2.0 4.0 6.0 8.0 10.0 12.0

Diameter of Bare Conductor, in mm
FIG. 57. Thicknesses of Insulation on Cotton-covered Wires.

" "
space factor as applied to field spools, is employed to denote
the ratio of the total cross-section of copper in the spool, to
the gross area of cross-section of the winding space. Attainable
" "
values for the space factor of spools wound with wires of
various sizes and with various insulations, are given in the
curves in Fig. 58.
For the case we are considering, we ascertain from the curves
that the space-factor may be equal to 0.55. That is to say: 55
 CALCULATIONS FOR 2500-KVA. GENERATOR 89

per cent of the cross-section of the winding space will be copper

and the remaining 45 per cent will be made up of the insulation
and the waste space. Thus the aggregate cross-section of copper
will be 200X0.55 = 110 sq.cm. Consequently the number of
turns is:

110
= 615.
0.179

0.5

0.4

0.2

0.1

1234 5678 Bare Diameter of Wire in mm

FIG. 58. Curves showing "Space Factors" of Field Spools Wound with
Wires of Various Diameters.

For the normal magnetomotive force of 21 150 ats. per spool,

the exciting current is:

21 150
= 34.4 amperes.
615

The total excitation loss is:

500X34.4 = 17200 watts.

90 POLYPHASE GENERATORS AND MOTORS
But of this loss of 17 200 watts:

crj

^rX 17 200 = 1720 watts

ouU

are dissipated in the field regulating rheostat, and only :

X 17 200 = 15 500 watts

are dissipated in the field spools. The loss per spool is thus:

15500
1940 watts.
o

The next step is to ascertain whether this will consist with a

suitably-low temperature rise. The peripheral speed of our
rotating field is:

375
=
-^- 35.0 meters per second.

At this speed this very open general construction with

and with
salient poles, it be practicable to restrict the temperature
will
rise to some 1.3 rise per watt per sq.dm. of external cylindrical

radiating surface of the field spool. We have:

External periphery of the spool = 26.4 dm.
Length of spool = 2.0 dm.
External cylindrical radiating surface = 26.4 X 2.0 = 52.8 sq.dm.

Watts per sq.dm. = = 36.8.

OZ.o

Ultimate temperature rise = 36.8X1. 3 = 48 Cent.

This is a high value and would not be in accordance with the

terms of usual specifications. But the modern tendency is to
take advantage of the increasing knowledge of the properties
and to permit higher temperatures in
of insulating materials

low-pressure windings, provided offsetting advantages are thereby

 CALCULATIONS FOR 2500-KVA. GENERATOR 91

obtained. This is the case in moderate-speed and high-speed

polyphase generators. The entire design profits in great measure
by compressing the rotor into the smallest reasonable compass.
In designing on such lines, the space available for the field spools
is necessarily restricted. In the case of the design under con-
sideration, we could, by decreasing the air-gap, decrease the
required excitation and consequently also the temperature rise
of the field spools. But we have already seen that the char-
acteristics of the machine would be impaired by doing this.

For a power-factor of 1.00 the mmf. required per spool field

isonly 17 200 ats. The loss in the field spool decreases as the
square of the mmf. Consequently were the machine required
for an output of 2500 kva. at exclusively unity power-factor, the
temperature rise would be only;

1720Q\ 2
X 48 = 32 Cent.
,21 1507

Thus for an putput of 2500 kilowatts at unity power-factor,

the field spools are actually 33 per cent cooler than for an output
of only 2250 kilowatts at a power-factor of 0.90 (i.e., for an out-
put of 2500 kva. and a power-factor of 0.90).
If on the contrary, the machine were required for an output
of only 2000 kilowatts but at a power-factor of 0.80 (again 2500
kilovolt-amperes), the temperature rise of the field spools would be:

It is impressive to note that although this output (2000 kilo-

watts at 0.80 power-factor), is 20 per cent less than an output
of 2500 kilowatts at unity power-factor, the temperature rise
of the field spools is:

K4 _ 00
^f^X 100 = 69 per cent

greater. The losses and temperature rise in the other parts of

the machine will be the same for both these conditions, since the
kilovolt-ampere output of the machine is 2500, in both cases.
92 POLYPHASE GENERATORS AND MOTORS

methods of design, and

It is the object of this treatise to explain

consequently no particular advantage would be gained by carry-

ing through modified calculations for the purpose of providing
field spools which would permit of carrying the rated load at

0.90 power-factor with a temperature rise of only 45 Cent, instead

of a temperature rise of 48 Cent. An inspection of Figs. 21 and 56
shows that there is room for more spool copper should its use
appear desirable. It would also be practicable to decrease the
internal diameter of the magnet yoke from the present 928 mm.
down to say 878 mm. and increase the radial length of the
magnet core (and consequently also the length of the winding
space), by 25 mm,

THE CORE LOSS

In polyphase generators, the core losses may be roughly pre-
determined from the data given in Table 10.

TABLE 10. DATA FOR ESTIMATING THE CORE Loss IN POLYPHASE

GENERATORS.

Density in Stator
Core in Lines per
Square Centimeter.
 CALCULATIONS FOR 2500-KVA. GENERATOR 93

Area of the surface of an annular ring with the above external

and internal diameters is equal to:

2 - 1782 = 16
) 700 sq.cm.
|(230

Area of the 120 slots = 120X5.72X2.41 = 1650 sq.cm.

Net area of surface of stator core plate = 16 700 1650
= 15 050 sq.cm.
In = 79.2 cm.

Volume of sheet steel in stator core = 79 2X15

' 050
=1.19 cu.m.
1 UUU UUU

Weight of 1 cu.m. of sheet iron = 7.8 (metric) tons.

Weight of stator core = 7.8Xl. 19 = 9.3 (metric) tons.

On the basis of a loss of 4.0 kw. per ton, we have:

Core loss = 4.0X9.3 = 37.2 kw.

Friction Loss. No simple rules can be given for estimating

the windage and bearing friction loss. The ability to form some
rough idea of the former can, in a design of this type, only be
acquired by long experience. It must suffice to state that, for
the present design, a reasonable value is:

Windage and bearing friction loss = 20 kw.

Excitation Loss. The loss in the exciting circuit is made up

(seep. 90) of:

Loss in regulating rheostat = 1.7 kw.

Loss in field spools = 15. 5 kw.

Stator I2 R Loss. The resistance of the stator winding (see

p. 22) is:

0.665 ohm per phase.

94 POLYPHASE GENERATORS AND MOTORS
and consequently we have:

Stator PR loss at rated load = 3 X 12

-

^' 665 =
1
28.6 kw.
1UUU

Total Loss. The total loss at full load is the sum of these
various losses:

I. Stator 72 # = 28 600 watts

II. Field spool I 2
R = 15500 "
III. Field rheostat PR = 1700 "
IV. Core loss = 37200 "
V. Friction loss = 20 OOP
"
' '
Total loss at full load=" 103000
Output at full load =2250000 ' '

Input at full load = 2 353 000 ' '

Full -load efficiency = 2250

^^X 100 = 95.6 per cent.

CONSTANT AND VARIABLE LOSSES

Of the five component losses, the last four remain fairty con-
stant at all loads, whereas the
loss (the stator
first loss) PR
varies as the square of the load. It is true that the sum of the
second and third losses decreases slightly with decreasing load,
but in the present machine the total decrease is only in the ratio
of the mmf. at full load and no load. The twommf.are:

21 150 ats. at rated load.

14 400 ats. at no load.

The corresponding values of the total excitation loss are :

17 200 watts at full load (and 6950 volts per phase)

and

200 - 1 70 watts at no load < and 695 volts P er

phase)
 CALCULATIONS FOR 2500-KVA. GENERATOR 95

Thus the sum of the last four losses decreases from:

17 200+37 200+20 000 = 74 400 watts at full load

down to

11 700+37 200+20 000 = 68 900 watts at no load.

This decrease only amounts to

74400-68900
74400
X 100 = 7.4 per cent.

Thus, taken broadly, we may take the last four component

"
making up an aggregate which we
losses as may term the con-
stant loss," and in contradistinction we may term the first com-
"
ponent the variable loss." In our design we have:

Variable loss =28 600 watts.

Constant loss = 74 400 watts.

If weignore the 7 per cent decrease in the constant loss, we may

readily obtain the efficiencies at various loads. The method
will be clear from an inspection of the following estimates:

EFFICIENCY AT ONE-FOURTH OF FULL LOAD

Variable loss = 0.25 2 X 28 600 =1 800 watts "

Constant loss = 74400
"
Total loss at one-fourth of full load= 76200
= 564 "
Output = 0.25 X 2 250 000 000

Input =640200 watts

nnn
rj(at one-fourth load) =
Kfi4.
~ = 0.882.
96 POLYPHASE GENERATORS AND MOTORS

EFFICIENCY AT HALF LOAD

Variable loss = 0.502 X 28 600 = 71 50 watts

= "
Constant loss 74400

Total loss at half load = 81550 '

"
Output at half load = 1 125 000

Input =1 207 000 watts

ij(at half load)

EFFICIENCY AT 50 PEK CENT OVERLOAD

Variable loss = 1 .502 X 28 600 = 64 500 watts

"
Constant loss = 74400

"
Total loss at 50 per cent overload = 138 900
"
Output at 50 per cent overload =3 380 000

Input =3 519 000 watts

ooorj
rj(at 50% overload) = = 0.961 .

LOAD CORRESPONDING TO MAXIMUM EFFICIENCY

When the variable losses have increased until they equal the
constant losses, the efficiency will be at its maximum. The
corresponding load is :

400
X 2250 = 1.61X2250
= 3640kw.

The efficiency is then:

0.962.
3640+74.4+74.4 3789
 CALCULATIONS FOR 2500-KVA. GENERATOR 97

From this point upward, the efficiency will decrease.

This brief method of estimating the efficiencies at several
loads, gives slightly too low results at low loads and slightly too
high results at high loads. But the errors are too slight to be
of practical importance; in fact the inevitable errors in deter-

mining the component losses are of much greater magnitude.

In Fig. 59 is plotted an efficiency curve for the above-cal-
culated values which correspond to a power-factor of 0.90.

1000 2000 3000 4000 5000

Output in Kilowatts

FIG. 59. Efficiency Curve of 2500-kva. Alternator for G=0 90.

DEPENDENCE OF EFFICIENCY ON POWER-FACTOR OF

LOAD
Let us consider the load to be maintained at 2500 kva. but
with different power-factors. The field excitation will be:

For G= 1.00:
17200
X 17 200 = 14 000 watts.
21 150

For G = 0.90:
17 200 watts.
For G = 0.80:
22500
21 150
X 17 200 = 18 300 watts.
 POLYPHASE GENERATORS AND MOTORS
The total losses become:

G Total Losses
1.00 99 800 watts
"
0.90 103 000
"
0.80 104 100

The outputs, inputs and efficiencies become:

G.
 CHAPTER III

POLYPHASE GENERATORS WITH DISTRIBUTED FIELD

WINDINGS

THE type of polyphase generator with salient poles which

has been described in the last chapter has served excellently
as a basis for carrying through a set of typical calculations.

Salient-pole generators are chiefly employed for slow- and medium-

speed ratings. But for the high speeds associated with steam-
turbine-driven rotors with distributed field windings are
sets,
practically universally employed in modern designs. It is not

proposed to carry through the calculations for a design of this

type. While there are a good many differences in detail, to
which the professional designer gives careful attention, the
underlying considerations are quite ofthe same nature as
in the case of salient-pole designs. In Figs. 60 and 61 are
shown photographs of a salient-pole rotor and a rotor with
a distributed field winding. The former (Fig. 60), is for
a medium speed (514 r.p.m.), water-wheel generator with a
rated capacity for 1250 kva. The latter (Fig. 61), is for a 750-
r.p.m., steam-turbine-driven set, with a rated capacity for 15000
kva. The former has 14 poles and the latter 4 poles. The
former is for a periodicity of 60 cycles per second, and the latter
for a periodicy of 25 cycles per second.
An inherent characteristic of high-speed sets relates to the
great percentage which the sum of the core loss, windage and
excitation bears to the total loss. In our 2500-kva. salient-pole
design for 375 r.p.m., the core loss amounted to some 37 000
watts, the windage and bearing friction to 20 000 watts, and the
excitation to 17 000 watts, making an aggregate of 74 000 watts
" "
for the constant losses, out of a total loss at full-load of
103 000 watts. But in a design for 2500 kva. at 3600 r.p.m.,
(i.e., for very nearly 10 times as great a speed), the core loss,
99
100 POLYPHASE GENERATORS AND MOTORS

bearing-friction, windage, and excitation would together amount

to some 72 000 watts out of a total of some 80 000 watts. A

FIG. 60. Salient Pole Rotor for a 14-pole, 1250-kva., 60-cycle, 514 r.p.m.,
3-phase Alternator, built by the General Electric Co. of America.

representative distribution of the losses for a 2500-kva., 0.90-

power-factor, 12 000- volt polyphase generator would be:
Armature PR loss 8 000 watts
"
Excitation PR loss 9 000
Core loss 32000
Windage and bearing friction loss 31 000
"
Total loss at full load 80 000
Output at full load 2 250 000
"
Input at full load 2 330 000
Efficiency at full load 96.6 per cent
 WITH DISTRIBUTED FIELD WINDINGS 101

A result of the necessarily large percentage which the

"constant" losses bear to the total losses, is that the efficiency
falls off badly with decreasing load, In this instance we have:

Variable losses = 8 kw.

Constant losses = 72 kw.

The efficiencies at various loads work out as follows:

Load. Efficiency.
88.5 per cent
"
i 93.9
"
1.00 . 96.6

FIG. 61. Rotor with Distributed Field Winding for a 4-pole, 15 000-kva.,
25-cycle, 750 r.p.m., 3-phase Alternator, built by the General Electric
Co. of America.

On pp. 94 to 96 the efficiencies of the 375-r.p.m. machine for

this same rating were ascertained to be:

Load. Efficiency.

i 88.2 per cent

"
93.1
"
1.00. .
. 95.6
102 POLYPHASE GENERATORS AND MOTORS
Fora 100-r.p.m., 2250-kw., 0.90-power-f actor, 25-cycle
design, the efficiencies would have been of the following order:

Load. Efficiency.

J 88.0 per cent

i "
92.6
"
1.00 94.6

The values may be brought together for comparison as follows :

EFFICIENCIES

Load.
 WITH DISTRIBUTED FIELD WINDINGS 103

rise, and so to proportion the passages as to transmit the

air in the quantities thus ascertained to be necessary.
In our 3600-r.p.m., 2250-kw. generator, the losses at full
load amount to 80 kw. If the heat corresponding to this loss
is to be carried away as fast as it is produced, then we must

circulate sufficient air to abstract:

80 kw.-hr. per hr.

A convenient starting point for our calculation is from the

basis that:

1.16 w.-hr. raises 1 kg. of water 1 Cent.

The specific heat of air is 0.24; that is to say, it requires

only 0.24 times as much energy to raise 1 kg. of air by 1 degree
Cent, as is required to raise 1 kg. of water by 1 degree Cent.
Consequently, to raise by 1 degree Cent., the temperature
of 1 kg. of air, requires the absorption of :

1.16X0.24 = 0.278 w.-hr.

One kilogram of air at atmospheric pressure and at 30 degrees

Cent, occupies a volume of 0.85 cu.m. Therefore, to raise 1
cu.m. of air by 1 degree Cent, requires:

If, for the outgoing air, we assume a temperature 25 degrees

above that of the ingoing air, then every cu.m. of air circulated
through the machine will carry away:

0.327X25 = 8.2 w.-hr.

We must arrange for sufficient air to carry away:

80 000 w.-hr. per hour.

104 POLYPHASE GENERATORS AND MOTORS
\

Consequently we must supply:

80000 =
9800 cu. m. per hour;
8.2

or:

-- = 163.
9800 1ft0
cu. m. per minute.

In dealing with the circulation of

air it appears necessary to
make the concession of employing other than metric units. We
have:

1 cu.m. = 35.4 cu.ft.

Therefore in the case of our 2250-kw. generator, we must

circulate :

163X35.4 = 5800 cu.ft. per min.

 CHAPTER IV

THE DESIGN OF A POLYPHASE INDUCTION MOTOR WITH A

SQUIRREL-CAGE ROTOR

THE polyphase induction motor was brought to a commer-

cial stage of development about twenty years ago. Many tens
of thousands of such motors are now built every year. The design
of polyphase induction motors has been the subject of many
elaborate investigations and there has been placed at the disposal
of engineers a large number of practical rules and data.
The may proceed from any one of
design of such a motor
many and each designer has his preferred method.
starting points
The author proposes to indicate the method which he has found
to be the most useful for his purposes. It must not be inferred
that any set of rules can be framed which will lead with certainty
to the best design for any particular case. The most which
can be expected is that the rules shall lead to a rough preliminary
design which shall serve to fix ideas of the general orders of dimen-
sions. Before he decides upon the final design, the enterprising
designer will carry through a number of alternative calculations
in which he will deviate in various directions from the original
design. A consideration of the several alternative results at
which he will thus arrive, will gradually lead him to the most
suitable design for the case which he has in hand.
The method of design will be expounded in the course of
working through an illustrative example.

ILLUSTRATIVE EXAMPLE

Let it be required that a three-phase squirrel-cage induction

motor be designed. The normal rating is to be 200 hp. and the
motor is to be operated from a 1000-volt, 25-cycle circuit. It
is desired that its speed shall be in the close neighborhood of
500 r.p.m.
105
106 POLYPHASE GENERATORS AND MOTORS
Determination of the Number of Poles. Denoting the speed
in revolutions per minute by R, then the speed in revolutions
-p
per second is equal to ~~. If we denote the number of poles

by P, and the periodicity in cycles per second by ^, then we

have:

In our case we have :

~ = 25 72 = 500.

Therefore :

2X60X25
500

" "
If we
design the motor with 6 poles, the synchronous
speed will be 500 r.p.m. At no-load, the motor runs at practi-
" " " "
cally its synchronous speed; that is to say, its slip is
" "
practically zero. The term slip is employed to denote the

amount by which the actual speed of the motor is less than the
" "
synchronous speed. An
appropriate value for the slip of
our motor at its rated load, is some 2 per cent, or even less.

Taking it for the moment as 2 per cent, we find that, at rated

load, the speed will be 500-0.02X500 = 490

r.p.m.
The
variations in the speed between no load and full load
are so slight that at many steps in the calculations the speed
may be taken at the approximate value of 500 r.p.m., thus
avoiding superfluous refinements which would merely complicate
the calculations and serve no useful purpose.
Rated Output Expressed in Watts. Since one horse-power
isequal to 746 watts, the rated output of our*200-h.p. motor
may also be expressed as 200X746 = 149 200 watts.
Determination of T, the Polar Pitch. The distance (in cm.)
measured at the inner circumference of the stator, from the center
of one pole to the center of the next adjacent pole, is termed
" "
the polar pitch and is denoted by the letter T. Rough pre-
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 107

liminary values for T are given in Table 11 for designs for 25 and
50 cycles for a wide range of outputs and speeds. These values
have been found by experience to be appropriate.

TABLE 11. PRELIMINARY VALUES FOR T (THE POLAR PITCH) FOR USE IN
DESIGNING THREE-PHASE INDUCTION MOTORS

Rated Output.
108 POLYPHASE GENERATORS AND MOTORS

THE OUTPUT COEFFICIENT

The next step relates to the determination of a suitable value

"
for ?, the Output Coefficient," which is defined by the following
formula :

in which

W = Rated output in watts, (which is equal to 746 times

the rated output in h.p.),
D = Diameter at air-gap, in decimeters, i.e., the internal
diameter of the stator,
Xg = Gross core length, in decimeters,
R = rated speed, in revolutions per minute.

As in the earlier chapters of this treatise, the symbols and D

Xgf will sometimes be employed for denoting respectively the

air-gap diameter and the gross core length, expressed, as above,

in decimeters, but more usually they will denote these quantities
as expressed in centimeters. The student can soon accustom
himself to distinguishing, from the magnitudes of these quan-
tities, whether decimeters or centimeters are intended, and
thus will not experience any difficulty of consequence, in this
double use of the same symbols.
Values of 5 suitable for preliminary assumptions are given
in Table 12. In this table, is given as a function of P and T.
For our design we have :

p = 6, T = 32.5 cm.

The corresponding value of ? in Table 12 is about 2.0. But

letus for our design be satisfied with a less exacting value and
take:
= 1.80.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 109

Therefore :

1492
1.80.

149 200

TABLE 12. PRELIMINARY VALUES FOR (THE OUTPUT COEFFICIENT) FOR

USE IN DESIGNING THREE-PHASE SQUIRREL-CAGE INDUCTION MOTORS.

(The figures at the heads of the vertical columns give the numbers of poles.)

r, the
polar
pitch (in
cm.)!
110 POLYPHASE GENERATORS AND MOTORS
Therefore :.

(in dm.) = r-166 dm. or 43.0 cm.

D and \g and T are the three characteristic dimensions of the

design with which we are dealing.
D 2 \g(with D and \g expressed in decimeters), is also, in
itself,a useful value to obtain at an early stage of the calculation
of a design. We have, for our motor;

PRELIMINARY ESTIMATE OF THE TOTAL NET WEIGHT

A rough preliminary idea of the total net weight of an induc-

tion motor may be obtained from a knowledge of its D 2 \g.
The " Total Net Weight " may be taken as the weight exclusive
of slide rails and pulley. In Table 13, are given rough repre-
sentative values for the Total Net Weights of induction motors
with various values of D2 \g.

TABLE 13. VALUES OF THE TOTAL NET WEIGHT OF INDUCTION MOTORS.

Total Net Weight in

Metric Tons, (i.e., in
Tons of 2204 Lbs.).

10 0.27
20 0.40
40 0.73

60 0.98
80 1.20
100 1.40

150 1.90
200 2.30
250 2.70

300 3.00
350 3.25
400 3.45
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 111

PRELIMINARY ESTIMATE OF THE TOTAL WORKS COST

The Total Works Cost will necessarily also be a very indefinite

quantity, susceptible to. large variations with variations in the

proportions and arrangement of a design. It is even more

greatly dependent upon the equipment and management of
the Works at which the motor is manufactured. Neverthless,
some rough indication appropriate for squirrel-cage induction
motors is afforded by the data in Table 14.

TABLE 14. TOTAL WORKS COST OF SQUIRREL-CAGE INDUCTION MOTORS.

Total Net weight of Total Works Cost per

Motor in Metric Tons. Ton, in Dollars.

0.20 310
0.40 300
0.60 290

0.80 285
1.00 280
1.50 270

2.00 260
2.50 250
3.00 240

3.50 230
4.00 225

For our 200-h.p. motor, we have D 2 \g = 166.

From Table 13, we ascertain that the Total Net Weight
is some 2.00 tons. From Table 14, it is found that the Total
Works Cost is of the order of $260. per ton. Consequently we
have Total Works Cost -2.00X260 = $520.

ALTERNATIVE METHOD OF ESTIMATING THE TOTAL

WORKS COST
An alternative method of estimating the T.W.C. of an induction
motor is based on the following formula:

TWC (in dollars) =KxDXfrg+0.7i),

112 POLYPHASE GENERATORS AND MOTORS
where D, Xg, and T are given in centimeters. K is obtained
from Table 15.

TABLE 15 VALUES OF K IN FORMULA FOR T. W. C.

Air-gap Diameter, D.
in Centimeters.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 113

machine should be sold and it is not unusual to find the selling

price over double the Total Works Cost. This is one of the
penalties of the competitive system of supplying the needs of
mankind.
The Peripheral Speed. It is well, before proceeding further
with the design, to calculate the peripheral speed. Let us denote
by S, the peripheral speed, expressed in meters per second,

R
'

100
A 60'

For our design :

'

xX62X500 =
meters P er second.
100X60

In this instance the peripheral speed is very low, and does

not constitute a limiting consideration from the point of view
of mechanical strength. For other speeds, ratings, and peri-
odicities, the preliminary data as derived from the rules which
have been might lead to an undesirably-high peripheral
set forth,

speed. Consequently it is well to ascertain the peripheral speed

at an early stage of the calculations and arrange to reduce D
and T in cases where the electrical design ought to be sacrificed
in some measure in the interests of improving the mechanical

design.

PERIPHERAL LOADING

We shall next deal with the determination of the number of

conductors to be employed. The product of the number of
conductors and the current per conductor, (i.e., the ampere-
"
conductors), constitutes a quantity to which the term peripheral
"
loading may be applied. Designers find from experience that
it is desirable to definite ranges of values for the
employ certain
peripheral loading per centimeter of periphery, measured at the

air-gap. In Table 16, are given values which will serve as pre-
liminary assumptions for a trial design.
114 POLYPHASE GENERATORS AND MOTORS
TABLE 16. PRELIMINARY ASSUMPTIONS FOR THE PERIPHERAL LOADING OF
AN INDUCTION MOTOR.

5 (in Centimeters).
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 115

Consequently the total peripheral loading is:

62XxX372 = 72 500 ampere-conductors;

or

72 500
~- = 24 200 ampere-conductors per phase.

From this product we wish next to segregate the ampeers and

the conductors. We may do this by dividing the ampere-
conductors by the full-load per phase. We cannot
current
estimate the precise value of the full-load current per phase
until the design has been completed, as its precise value depends

upon the efficiency and power-factor at full load. But we cannot

complete the design without determining upon the suitable
number of conductors to employ. Hence it becomes necessary
to have recourse to tables of rough approximate values for the
full-load efficiency and the power-factor of designs of various
ratings. Such values are given in Tables 17 and 18.

EFFICIENCY AND POWER-FACTOR

For the case of squirrel-cage induction motors for moderate

pressures, and in the absence of any specially exacting require-
ments as regards capacity for carrying large instantaneous over-
loads, we may proceed from the basis of the rough indications
in Tables 17 and 18. From these tables we obtain:

Full-load efficiency = 91 per cent.

Full-load power-factor = 0.91.

The required estimation of the full-load current may be

carried out as follows:

Horse-power output at rated load = 200

Watts output at rated load = 200 X 746 = 149 200
Efficiency at rated load =0.91
1 4Q 200
Watts input at rated load - = 164 000
u.yi
116 POLYPHASE GENERATORS AND MOTORS
164 OOP
Watts input per phase at rated load = 54700
3
Power-factor at rated load 0.91
700
Volt-amp, input per phase at rated load = 54 =
-777^- 60 200
u.yi
Pressure between terminals (in volts) = 1000
Phase pressure

60200 =
Current per phase at rated load 104
577

TABLE 17.PRELIMINARY VALUES OF FULL LOAD EFFICIENCY, IN PER

CENT, FOR POLYPHASE SQUIRREL-CAGF INDUCTION MOTORS. THE
VALUES GIVEN CORRESPOND TO THOSE OF NORMAL MOTORS.

Rated
Output
in Horse-
power.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 111

We must round off this value to some suitable whole number,

say 39 or 40, taking whichever leads to the best arrangement
of the winding. An inspection indicates that we should consider
the following alternatives:

40 conductors arranged 8 per slot in 5 slots.

(~o~= )

'

" ^2_\
40 10
U /

39 13 P \o
(u r
40 20

TABLE 18. PRELIMINARY VALUES FOR FULL-LOAD POWER FACTOR OF POLY-

PHASE SQUIRREL-CAGE INDUCTION MOTORS OF NORMAL DESIGN.
118 POLYPHASE GENERATORS AND MOTORS
stated that the quality of the performance of the motor is higher,
the greater the number of slots per pole per phase. But the
overall dimensions and the weight and the Total Works Cost
increase with increasing subdivision of the winding amongst
many slots, and consequently the designer should endeavor to
arrive at a reasonable compromise between quality and cost.
The Slot Pitch. We may designate as the slot pitch the dis-
tance (measured at the air-gap) from the center line of one slot
to the center line of the next adjacent slot. Since this quantity
is usually small, it is generally convenient to express it in mm.
Good representative values for the stator slot pitch are given
in Table 19. The values in the table may be taken as applying
to designs for moderate pressures. The higher the pressure,
the more must one depart from the tabulated values in the direc-
tion of employing fewer slots,

TABLE 19. VALUES OF STATOR SLOT PITCH FOR INDUCTION MOTORS.

T, the Polar Pitch

(in cm.).
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 119

=\
)
4 slots per pole per phase (pppp). Thus we must
o
(12 /
revise the stator slot pitch to:

= 27.1 mm.
\&

The Total Number of Stator Slots. Since our machine has

we shall have:
6 poles,

12X6 = 72 stator slots.

The appropriate slot layout will be based on 10 conductors

per slot and :

4X10 = 40 conductors pppp.

and
6X40 = 10 _,
^ 120 turns in series per phase.
&

Let us denote the turns in series per phase by T. Then:

^=120.

THE PRESSURE FORMULA

In our discussion of the design of generators of alternating
we have become acquainted with the pressure formula
electricity :

V=KXTX~XM.
In this formula we have:

y = the phase pressure in volts;

il = a coefficient;
T = turns in series per phase;
~ = periodicity in cycles per second;
M = flux per pole in megalines.
For a motor, the phase pressure in the above formula must,
be taken smaller than the terminal pressure, to the
for full load,
extent of the IR drop in the stator windings. But at no load
120 POLYPHASE GENERATORS AND MOTORS
the phase pressure is equal to the terminal pressure divided by V3.
Therefore,

1000
Phase pressure = :r-, = 577 volts.

The K
depends upon the spread of the winding
coefficient
and the manner For the conditions
of distribution of the flux.

pertaining to a three-phase induction motor with a full-pitch

winding we have:
# = 0.042.
For other winding pitches, the appropriate value of may K
be derived by following the rule previously set forth on pp. 16
to 18 of Chapter II, where the voltage formula for generators
of alternating electricity is discussed.
Since our motor is for operation on a 25-cycle circuit, we have:

Thus at no load we have:

577 = 0.042X120X25XM;

M = 4.57 megalines.

THE MAGNETIC CIRCUIT OF THE INDUCTION MOTOR

In Fig. 62 are indicated the paths followed by the magnetic

lines in induction motors with 2, 4, and 8 poles. One object of
the three diagrams has been to draw attention to the dependence
of the length of the iron part of the path, on the number of

poles. Thus while in the 2-pole machine, some of the lines

extend over nearly a semi-circumference, in the stator core and
in the rotor core; their extent is very small in the 8-pole design.
As a consequence, the sum of the magnetic reluctances of the
air-gap and teeth constitutes a greater percentage of the total
magnetic reluctance, the greater the number of poles.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 121

Fig. 63 has been drawn to distinguish that portion of the total

flux which corresponds to one pole. It is drawn to correspond
to a 6-pole machine.

2 Pole 4 Pole 8 Pole

FIG. 62. Diagrammatic Sketches of 2-, 4-, and 8-pole Induction-motor Cores,
showing the Difference in the Mean Length of the Magnetic Path.

At this stage of our calculations, we wish to ascertain the cross-

section which must be allowed for the stator teeth. crest A
density of 15 500 lines per sq. cm. is appropriate for the stator

FIG. 63. Diagrammatic Representation of that Portion of the Total Path

which Corresponds to One Pole of an Induction Motor. [The heavy
dotted lines indicate the mean length of the magnetic path for one pole.]

teeth in such a design as that which we are considering. A skilled

designer will, on occasions, resort to tooth densities as high as
19 000, but it requires experience to distinguish appropriate cases
122 POLYPHASE GENERATORS AND MOTORS
for such high densities and the student will be well advised to
employ lower densities until by dint of practice in designing, he is
competent to exercise judgment in the matter. In general the
designer will employ a lower tooth density the greater the number
of poles. This is for two reasons: firstly, as already mentioned
in connection with Fig. 62, the magnetic reluctance of the air-

gap and teeth constitutes a greater percentage of the total reluc-

tance the greater the number of poles; and secondly, (for reasons
which will be better understood at a later stage), a high tooth
density acts to impair the power-factor of a machine with many
poles, to a greater extent than in the case of a machine with few
poles.
In Figs. 64 and 65 are shown two diagrams. These represent
the distribution of the flux around the periphery of our 6-pole
motor at two instants one-twelfth of a cycle apart. Since the
periodicity is 25 cycles per second, one-twelfth of a cycle occupies

a second. After another g^th of a second

(
^= Wfoth f

the flux again assumes the shape indicated in Fig. 64, but dis-
placed further along the circumference, as indicated in Fig. 66.
In other words, as the flux travels around the stator core, its
distribution is continually altering in shape from the typical

form shown in Fig. 64, to that shown in Fig. 65, and back to that
shown in Fig. 66 (which is identical with Fig. 64, except that it
has advanced further in its travel around the stator) Successive
.

positions of the flux, each ^th

second later than its predecessor,
are drawn in Figs. 67 to 70. Comparing Fig. 70 with Fig. 64,
we see that they are identical except that in Fig. 64, a south
flux occupies those portions of the stator, which, in Fig. 64,
were occupied by a north flux. In other words, a half cycle has
occurred in the course of the (dhr^iroth of a second which has
elapsed while the flux has traveled from the position shown in
Fig. 64, to that shown in Fig. 70. A whole cycle will have occurred
in u^th of a second (the periodicityis 25 cycles per second) and ;

the flux will then have been displaced to the extent of the space
occupied by one pair of poles. At the end of the time occupied
by 3 cycles (*\ths of a second) the flux will have completed one
/ (\ \
revolution around the stator core, since the machine has =
(
9 )

3 pairs of poles.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 123

Gap

Air

the

around

Density

Flux

the

to

Proportional

Figures

Arbitrary

are

nates
124 POLYPHASE GENERATORS AND MOTORS
In estimating the magnetomotive force (mmf.) which must
be provided for overcoming the reluctance of the magnetic circuit,
we must base our calculations on the crest flux density. This
corresponds to the flux distributions represented in Figs. 65,
67 and 69. It can be shown * that the crest density indicated
in these figures is 1.7 times the average density. In other

FIG. 71. Diagram Illustrating that the Crest Density in the Air-gap and
Teeth of an Induction Motor is 1.7 Times the Average Density.

words, the crest density with the flux distribution correspond-

ing to the peaked curve in Fig. 71, is 1.7 times the average
*
This is demonstrated, step by step, on pp. 380 to 390 of the 2d edition of
the author's "Electric Motors" (Whittaker & Co., London and New York,
1910;,
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 125

density indicated by the rectangle in the same figure. Each

tooth is, in turn, located at the center of the rotating flux and is
in turn subjected to this crest density.
Thus for our assumption of a crest density of 15 500 lines per
square centimeter we shall have :

15 500
Average density in stator teeth = ^
= 9100 lines per sq.cm.

Since we have a flux of 4.57 megalines, we must provide, per

pole, a tooth cross-section of:

4570000 = _ nn
500 sq.cm.

There are 12 stator teeth per pole. The cross-section of each

tooth must thus be:

- = 41.6 sq.cm.

Before we can attain our present object of determining upon the

width of the tooth, we shall have to digress and take up the
matter of the proportioning of the ventilating ducts.

VENTILATING DUCTS

The employment of a large number of ventilating ducts in

the cores of induction motors, renders permissible, from the
temperature standpoint, the adoption of much higher flux den-
sities and current densities than could otherwise be employed,

and thus leads to a

light and economical design.
In Table 20 are given rough values for the number of ducts,
each 15 mm. wide, which may be taken as suitable, under various
circumstances of peripheral speed and values of \g.
In our case, where the peripheral speed is 16.2 meters per second
and Xg is 43, the table indicates 1.8 ducts per dm., or a total of
126 POLYPHASE GENERATORS AND MOTORS

1.8X4.3 = 7.7 ducts, to be a suitable value. Eight ducts will

be employed and they will require .8X1.5 = 12.0 cm.

TABLE 20. VENTILATING DUCTS FOR INDUCTION MOTORS.

Peripheral Speed in
Meters per Second.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 127

Diameter to neck of tooth =62.0+1.0 = 63.0 cm.

Tooth pitch at neck = = ^7.5 mm.

~72~

Required cross-section of tooth at neck = 4 1.6 sq.cm.

Width of tooth at neck = -= = -7^77- = 14.9 mm.

An 279

Width of Slot. The slot will have parallel sides, and its width
will be:

27.5 -14.9 = 12.6 mm.

This will be its width when punched. Owing to inevitable slight

inaccuracies in building up the stator core from the individual
punchings, the assembled width of the slot will be some 0.3 mm.
less, or

12.6-0.3 = 12.3 mm.

"
This allowance of 0.3 mm.
termed the slot tolerance."
is

Having now determined the width of the stator slot, it would

appear in order to proceed at once to determine its depth. But

this depends upon the copper contents for which space must
be provided. Consequently we must now turn our attention
to the determination of:
The Dimensions of Conductor. The current
the Stator
density in the stator conductor determined upon as a compro-
is

mise amongst a number of considerations, one of the chief of

which is the permissible value of the watts per square decimeter
(sq.dm.) of peripheral radiating surface at the air-gap. This
value is itself influenced by such factors as the peripheral speed
and the ventilating facilities provided, hence the current density
will also be influenced by these considerations.
The value of the watts per square decimeter of peripheral

radiating surface at the air-gap, cannot, unfortunately, be ascer-

tained until a later stage when we shall have determined not
only the copper losses, but also the core loss. If, at that later
stage, the value obtained for the watts per square decimeter
of peripheral radiating surface at the air-gap, shall be found to
be unsuitable, it will be necessary to readjust the design.
128 POLYPHASE GENERATORS AND MOTORS
Table 21 has been compiled to give preliminary representa-
tive values for the stator current density for various outputs
and peripheral speeds, for designs of normal proportions.

TABLE 21. PRELIMINARY VALUES FOR THE STATOR CURRENT DENSITY.

Rated Output
in h.p.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 129

yet been determined, but the width of the available space is seen
to be 12.3 2X1.4 = 9.5 mm. Thus we see that a space 9.5 mm.
wide is available for the insulated conductors.
From Table 21 we ascertain that the stator conductor should
be proportioned for about 300 amperes
per square centimeter. Since we have, I 12.3 mm >

for the full-load current, 7= 104 amperes,

the cross-section should be :

-9.5 -1.4mm

104
= 0.347 sq.cm. or 34.7 sq.mm.
oUU

In the interests of securing a suit-

able amount of flexibility in the process

of winding, let us divide this aggregate
cross-section into two conductors which
shall be in parallel and each of which
shall have a cross-section of some:

- = 17.4 sq.mm.

The diameter of a wire with a cross-

section of 17.4 sq.mm. is:
FIG. 73. Stator Slot of 200
H.P. Induction Motor,
showing Insulating Tube
D= 4.70 mm. in Place.

The bare diameter of each wire would, on this basis, be

4.70 mm.
In Table 23 are given the thicknesses of insulation on suitable
grades of cotton-covered wires employed in work of this nature.
We see that our wire of 4.70 mm. diameter would, if double
cotton covered, have a thickness of insulation of about 0.18mm.
Consequently its insulated diameter would be:

4.70+2X0.18 = 4.70+0.36 = 5.06 mm.

But the width of the winding space is seen from Fig. 73 to be

only 9.5 mm. The natural arrangement in this case, would be
130 POLYPHASE GENERATORS AND MOTORS
to place the two components side by side. Consequently the
insulated diameter must not exceed: .

= 4.75 mm.
-^-

TASLE 23. VALUES OF THE THICKNESS OF COTTON COVERING.

Diameter of Bare
Conductor (in mm.).
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 131

and there is no reason, at this stage, to anticipate that we have

exceeded permissible values. However, this must be put to the
test at a later stage when we shall have sufficient data to estimate
the temperature rise on the basis of the watts lost per square
decimeter of peripheral radiating surface at the air-gap.
Since we must provide for 10 conductors (20 component
wires) per slot, the height
of the winding space must
be at least

10X4.75 = 47.5 mm.

But it will be impracti-

cable to thread the wires

into place with complete
avoidance of any lost space.
So let us add 5 per cent to
the height of the winding
L4mm
space, making it:
10 Ins. Conds. 47.5 mm
1.05X47.5 = 50 mm. 1.4mm

The slot with the wires 2.2mm

in place, is drawn in Fig. 2.5mm
74. It is seen from this 54.0mm
figure that the total depth
is 54 mm. The slot open-
ing is 6 mm. wide.
The Slot Space Factor.
The total area of cross-
section of copper in the FIG. 74. Stator Slot of 200 H.P. Motor
slot amounts to 10X0.302 = with Winding in Place.
3.02 sq.cm.
The product of depth and punched width of slot is equal to
1.26X5.40 = 6.80 sq.cm.

Space factor of stator slot = 3r 02

1
^ = 0.445.
b.oU

It is to be distinctly noted that this slot design is

merely a
preliminary layout. Should it at a later stage not be found to
fulfil the requirements as regards
sufficiently-low temperature-
132 POLYPHASE GENERATORS AND MOTORS
rise at rated load, it will be necessary to consider ways and means
of so modifying the design as to fulfil the requirements.
Preliminary Proportions for the Rotor Slot. For reasons
which will appear later, there will be a number of rotor slots
not differing greatly from the number of
stator slots and these rotor slots will be of
about the same order of depth as the stator
slots, but considerably narrower. The result
will be that the rotor tooth density will be
fully as low or. even lower than the stator
tooth density. Let us for the present, con-
sider that the rotor slots are 54 mm. deep
and that the crest density in the rotor teeth
is, at no load, 15 500 lines per square cen-

timeter.
Let us further assume for
present the
that the nearly wide open,
rotor slots are
the shape of a rotor slot being somewhat as
FIG. 75. Rotor Slot indicated in Fig. 75. In the final design,
of200 H.P.Squirrel- the width of the rotor slot opening may
cage Motor. readily be so adjusted as to constitute about
20 per cent of the rotor tooth pitch at the
surface of the rotor, the tooth surface thus constituting some
80 per cent of the tooth pitch.
Determination of Cross-section of Air-gap. The stator
tooth pitch at the air-gap is:

620 Xic
= 27.1 mm.
72

Stator slot opening = 6.0 mm.

The stator tooth surface thus constitutes:

27 l 6
X 100 = 78.0
'

271 per cent

of the stator slot pitch.

Considering the average value of this percentage on both

sides of the air-gap, we find it to be :

78.0+80.0 =
79.0 per cent.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 133

If there were no slot openings, the cross-sections of the surfaces

from which the lines of each pole emerge and into which they
enter, could be taken as Xn X T.
But the slot openings bring this cross-section down to

0.79XXnX-r.

In crossing the air-gap, however, the lines spread out and

may be considered to occupy a greater cross-section when half
way across. They then gradually converge as they approach
the surfaces of the teeth at the other side of the gap. To allow
for this spreading, we may increase the cross-section by 15 per

cent, bringing it up to 1.15X0.79XXnXT.

For our 200-h.p. induction motor we have:

\n = 27.9 cm., T = 32.5 cm.

Cross-section of air-gap = 1.15X0.79X27.9X32.5 = 825 sq.cm.

Average air-gap density (at no load) = 4 570 000

5^ = 5550 lines per
825
sq. cm.
Crest density = 1 .7 X 5550 = 9450.

Radial Depth of Air-gap. Let us denote the radial depth

of the air-gap in mm. by A.

Appropriate preliminary values for A are given in Table 24.

TABLE 24. APPROPRIATE VALUES FOR A THE RADIAL DEPTH OF THE AIR-
GAP FOR INDUCTION MOTORS.

D, the Air-gap
Diameter,
134 POLYPHASE GENERATORS AND MOTORS
i

For our 200-h.p. motor, the air-gap diameter is 62 cm. and

the peripheral speed is 16.2 mps. Consequently the radial depth
of the air-gap is A = 1.3 mm.
Preliminary Magnetic Data for Teeth and Air-gap. We
have now obtained (or assumed) the densities in the teeth and in
the air-gap and we have the lengths of these portions of the mag-
netic circuit. These data are:
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 135

taken as equal to the circumferences corresponding to the mean

core diameter of the stator and rotor respectively, divided by 2P,
where P
is the number of poles. But as a step toward obtaining
these values, it is necessary to obtain the external diameter of
the stator core discs and the internal diameter of the rotor core
discs. These data, in turn, are dependent upon the densities
which should be employed behind the slots in the stator and rotor
respectively.
Densities in Stator and Rotor Cores. The densities to be

employed and rotor cores are quantities which may

in the stator
be varied between wide limits. In general, however, the core
densities should be lower, the greater the value of T the polar pitch,
and ~ the periodicity. For preliminary assumptions, the values
given in Table 25 will be found suitable.

TABLE 25. DENSITIES IN STATOR AND ROTOR CORES.

Periodicity in Cycles
per Second.
136 POLYPHASE GENERATORS AND MOTORS

Consequently :

Radial depth of stator punchings (exclusive of slot depth)

229

Radial depth of rotor punchings (exclusive of slot depth)

176
= 6.3 cm.
27.9

External diameter of stator punchings = 62.0+2X5.4+2X8.2

= 62.0+10.8+16.4
= 89.2 cm.
Internal diameter of rotor punchings
= 62.0-2X0.13-2X5.4-2X6.3
= 62.0-0.26-10.8-12.6
= 38.3 cm.
Diameter at bottom of stator slots = 62.0+2X5.4
= 72.8 cm.
Diameter at bottom of rotor slots = 62.0 -2X0.13 -2X5.4
= 50.9 cm.
89.2+72.8
Mean diameter of stator core ~ =81.0 cm.

L
Mean diameter of rotor core
'^ =44.6 cm.

Length of sta. mag. circ. per pole = '

-=21.2 cm.
zXo

Length of rotor mag. circ. per pole=

'- = 11.7 cm.
^ X

Compilation of Diameters. It is of interest at this stage to

draw up an orderly list of the leading diameters:

External diameter of stator core 892 mm.

Diameter at bottom of stator slots 728 mm.
Internal diameter of stator (D) 620 mm.
External diameter of rotor (D 2A) 617.4 mm.
Diameter at bottom of rotor slots 509 mm.
Internal diameter of rotor core. . . 383 mm.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 137

Sketch of Magnetic Portions of Design. We are now in a

position to make a preliminary outline drawing of the magnetic
parts of our machine. This has been done in Fig. 76 which
shows the leading dimensions of the magnetic portions of the
design and indicates the locations of the end connections of the
stator windings.
Magnetic Reluctance of Sheet Steel. In the design of the
induction motor, our magnetic material is exclusively sheet steel.
For this material the mmf. data in the first two columns of the

table previously givenon page 33 will give conservative results.

Tabulated Data of Magnetic Circuit. We now have the
lengths of the magnetic paths, the densities, and also data for
ascertaining the mmf. required at all parts of the magnetic circuit.
Thus for example:
Density in stator core = 10 000 lines per square centimeter;
Corresponding mmf. from column 2 of table on page 33
= 4.6 ats. per centimeter
Length of magnetic circuit in stator core (per pole) =21.2 cm.
Mmf. required for stator core
= 4.6X21. 2 = 98 ats.

As a further illustration we may give the calculation of the

mmf. required for the air-gap:
Crest density in air-gap = 9450 lines per sq.cm.

Corresponding mmf. =X 9450 = 0.8X9450 = 7550 ats. per cm.

Length of magnetic circuit in air-gap =0.13 cr .

Mmf. =
required for air-gap 7550X0. 13 =980 ats.

These illustrations will suffice to render clear the arrangement

of the calculations in the following tabulated form:

TABLE 26. ARRANGEMENT OF MMF. CALCULATIONS.

Part.
138 POLYPHASE GENERATORS AND MOTORS
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 139

\
X 100 = 68.5 the air-gap mmf.
(980since
Tjoy j,

is 68.5 per cent of the total required mmf. The mmf. required
for air-gap and teeth is:

980+119+119^., 1218
X 100 =
1427
-

^^X100
inA = 85.5
per cent

of the total requiredmmf. Attention is drawn to these percent-

ages to bear out the correctness of the assertion on page 134
that considerable inaccuracy in the estimation of the mean length
of the magnetic circuit in the stator and rotor cores will not

seriously affect the accuracy of the result obtained for the total
mmf. per pole. Consequently the use of the rough but time-
saving rule to divide by twice the number of poles, the mean
periphery of these cores, is shown to be justified.
Resultant mmf. of the Three Phases Equals Twice the mmf.
of One Phase. It is a property of the three-phase windings of
induction motors that the resultant mmf. of the three phases is
twice that exerted by one phase alone. Consequently in our
1427
design, each phase must contribute a mmf. of 75
= 714 ats.
2i

In our design, T, the number of turns in series per phase, is

120
equal to 120. The design has 6 poles. Thus we have = 20
-^-
turns per pole per phase.

Magnetizing Current. The magnetizing current per phase

which will suffice to provide the required 714 ats. must obviously
amount to :

714
= 35.7 crest amperes

or

35 7
= = 25.2 effective amperes.

Since the full-load current is 104 amperes, the magnetizing cur-

25 2
rent is -rX 100 = 24. 2 per cent of the full-load current.
140 POLYPHASE GENERATORS AND MOTORS
No-load Current. The no-load current is made up of two
components, the magnetizing current and the current correspond-
ing to the friction and windage loss and the core loss, i.e., to the
energy current at no load. It may be stated in advance that the
energy current at no load is almost always very small in compar-
ison with the magnetizing current. Since, furthermore, the mag-
netizing current and the energy current differ from one another
in phase by 90 degrees, it follows that their resultant, the no-
load current, will not differ in magnitude appreciably from the
magnetizing current. The calculation of the energy current is
thus a matter of detail which can well be deferred to another
stage. But to emphasize the relations of the quantities involved,
let us assume that the friction, windage and core loss of this motor
will later be ascertained to be a matter of some 4500 watts. This
\
j=- J577
(1000
volts. Consequently the energy component of the current con-
1500
sumed by the motor at no load is -^== = 2.6 amperes.
The no-load current thus amounts to V25.2 2 +2.6 2 = 25.3 amperes-
In other words the no-load current and the magnetizing current
differ from one another in magnitude by less than one-half of
one per cent, in this instance. Although of but slight practical
importance, it may be interesting to show that they differ quite
appreciably in phase. For we have for the angle of phase differ-
ence between the no-load current and the magnetizing current:

9
tan- 1 =" fin
tan- 1 0.103 = 5.9.

Thus the current in this motor when it is running unloaded,

lags (90 5.9 =) 84.1 behind the pressure; in other words, its

) 0.104.
=
power-factor is equal to (cos 84.1
Ratio of the No-load Current to the Full-load Current.
y, the
It is convenient to designate by y, the ratio of the no-load current

to the full-load current. For our motor we have:

T-Sg-0.242.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 141

Practical use will be made of this ratio at a later stage in the design
of this motor.

H, the Average Number of Slots per Pole. It is also con-

venient to adopt a symbol for the average of the number of slots
per pole on the stator and rotor. We have definitely determined
upon the use of 12 slots per pole on the stator. We have reserved
to a later stage of the calculations the determination of the precise
number of rotor slots. However it may here be stated that it
ispreferable to employ a number of rotor slots not widely differ-
ing from the number of stator slots. Thus as a preliminary
assumption we may take 12 as the average of the numbers of slots
per pole on statoi and rotor. Designating this quantity by H
we have ;

THE CIRCLE RATIO

We have now all the necessary data for determining a

quantity for which we shall employ the symbol a and which,
for reasons which we shall come to understand as we proceed,
"
we shall term the circle ratio." This quantity is of great
utility to the practical designer. Although a cannot be pre-
determined withany approach to accuracy, it so greatly
assists one's mental conceptions from the qualitative stand-
point as to make ample amends for its quantitative uncertainty.
We have seen that at no load, the current consumed by an
induction motor lags nearly 90 degrees behind the pressure.
Let us picture to ourselves a motor with no friction or core loss
and with windings of no resistance. In such a motor the no-
load current would be exclusively magnetizing and would lag 90
degrees behind the pressure.
Let us assume a case where, at no load, the current is 10
amperes. The entire magnetic flux emanating from the stator wind-
ings will cross the zone occupied by the secondary conductors (i.e.,
the conductors on the rotor) and pass down into the rotor core.
If the circumstance of the presence of load on the motor were not

to disturb the course followed by the magnetic lines, then the

magnetizing component of the current flowing into the motor
142 POLYPHASE GENERATORS AND MOTORS
would remain the same with load as it is at no load. If the motor
were for 100 volts per phase, then, in this imaginary case where
the flux remains undisturbed as the load comes on, we could cal-
culate in a very simple way the current flowing into the motor for
any given load. To illustrate; let us assume that a load of 3000
watts is carried 'by this hypothetical motor. A load of 3000 watts
corresponds to an output of 1000 watts per phase. Assuming a
motor with no internal losses, the input will also amount to 1000
watts per phase. Since the pressure per phase is 100 volts,
the energy component of the current input per phase is

= \ 10 amperes. Since the magnetizing component is 10

-r^r
(1000 j

amperes the resultant current per phase is Vl0 +10 = 14.1

2 2

amperes. The vector diagram corresponding to these conditions

is given in Fig. 77. The resultant current lags behind the
terminal pressure by

tan" 1 1
1.0 = 45.
jptan-

The power-factor is :

cos 45 = 0.707.

Let us double the load. The energy component of the cur

rent increases to 2X10 = 20 amperes, as shown in Fig. 78.
The total current increases to

Vl02 4-202 = 22.4 amperes.

The angle of lag becomes :

tan- 1 1
0.5 = 26.6.
ijptan-

The power-factor increases to:

cos 26.6 = 0.894.

POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 143

Let us again double the load, thereby increasing the energy

component of the current to 40 amperes, and the angle of lag
to 14.0 as shown in Fig. 79. The total current is now 41.3

amperes and the power-factor is:

740
.0^ 0.97.
Ul.3

10

FIG. 77 FIG. 78 FIG. 79

FIGS. 77 to 79. Vector Diagrams Relating to a Hypothetical Polyphase

Induction Motor without Magnetic Leakage.

At this rate we should quickly approach unity power-factor.

The curve of increase of power-factor with load, would be that
drawn in Fig. 80. It is to be especially noted that in the diagrams
in Figs. 77, 78 and 79, the vertical ordinates indicate the energy
144 POLYPHASE GENERATORS AND MOTORS

components of the total current and the horizontal ordinates

indicate the wattless (or magnetizing) components of the current.

10 12 14 16 18 20 22 24 26
Output in Kilowatts

FIG. 80. Curve of Power-factor of Hypothetical Polpyhase Induction Motor

without Magnetic Leakage.

It would be very nice if we could obtain the conditions indicated

in the diagrams of Figs. 77, 78 and 79. That is to say, it would
be very nice if the magnetizing component of the current remained

FIG. 81. Diagrammatic Representation of the Distribution of the Magneto-

motive Forces in the Stator and Rotor Windings of a Three-phase
Induction Motor.

constant with increasing load. But this is not the case. As the
load increases, the current in the rotor conductors (which was
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 145

negligible at no load), increases. The combined effect of the cur-

rent in the stator and rotor conductors is to divert a portion of
the flux out of the path which it followed at no load. In Fig.
81 are indicated diagramatically a few slots of the stator and rotor
windings. Considering the 12 left-hand conductors, the current is
indicated as flowing (at the moment) down into the plane of the
paper in the stator conductors and up out of the plane of the paper
in the rotor conductors. This has the same effect (as regards the
resultant mmf. of the stator and rotor conductors), as would be
occasioned by the arrangement indicated in Fig. 82, in which
the stator and rotor conductors constitute a single spiral.
Obviously the mmf. of this spiral would drive the flux along the
air-gap between the stator and rotor surfaces. The reluctance
of the circuit traversed by the magnetic flux thus increases grad-
ually (with increasing current input), from the relatively low
reluctance of the main magnetic circuit traversed by the entire flux
at no load, to the far higher reluctance of the circuit traversed
up
by practically the entire magnetic flux with the rotor at stand-
the pressure at the terminals of the stator windings being
still,

maintained constant throughout this entire range of conditions.

Consequently the magnetizing component of the current consumed
by the motor increases as the load increases. Thus instead of
the diagrams in Figs. 77, 78 and 79, we should have the three
diagrams shown at the right hand in Fig. 83. The corresponding
diagrams at the left hand in Fig. 83 are simply those of Figs.
77, 78 and 79 introduced into Fig. 83 for comparison. In both
cases, the no-load current is 10 amperes. But in the practical case
with magnetic leakage, the loads calling respectively for energy
components of 10, 20 and 40 amperes (loads of 3000, 6000 and

FIG. 82. Diagram Indicating a Solenoidal Source of mmf. Occasioning a

Flux Along the Air-gap, Equivalent to the Leakage Flux in an Induction
Motor.

12000 watts) involve magnetizing components of 10.6, 12.1 and

18.2 amperes.
146 POLYPHASE GENERATORS AND MOTORS

10.0 10.6

10.0 12,1

10.0 18.2

No Magnetic Leakage Magnetic Leakage

FIG. 83. Vector Diagrams for Hypothetical Motor without Magnetic Leakage
(at Left), and for Actual Motor with Magnetic Leakage (at Right).

The total current inputs are increased as follows :

Load (in Watts).

POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 147

The power-factors for the two cases are:

Load in Watts.
148 POLYPHASE GENERATORS AND MOTORS
In Fig. 85, the hypothenuses of the right-hand diagrams
of Fig. 83 havebeen superposed, and their right-hand extremities
are seen to lieupon the circumference of a semi-circle with a
diameter of 200 amperes. In Fig. 85 the magnetizing current
of 10 amperes is denoted by A B. The diameter of the semi-
circle is BD. We have:

L _

10 20 30 40 50 CO 70 80 90 100 120 130 140 150 160 170 180 190 200 210
Wattless Components of the Current

FIG. 85. Circle Diagram for a Polyphase Induction Motor with a No-load
Current of 10 Amperes and a Circle-ratio of 0.050.

" "
The quantity which we termed the circle ratio and which
we designated by the symbol c, is the ratio of AB to BD. For
this case we have:

AJJ - 10
-0050-
050
-

B5-266- '

a = 0.050.

The semi-circle in Fig. 85 is the locus of the extremities of the

vectors representing the current flowing into the stator winding.
If, for any value
of the current input, we wish to ascertain its

phase relations, we draw an arc with A as a center and with the

value of the current as a radius. The intersection of this arc
with the semi-circle, constitutes one extremity of the vector
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 149

representing the current, and A constitutes its other extremity.

The horizontal projection of this vector is its wattless component
and the vertical projection is its energy component. Consequently
we also have the convenient relation that the vertical component
is, for constant pressure, a direct measure of the power absorbed
by the motor. We shall give further attention to these important
relations at a later stage.
The circle ratio is a function of \g, Knowing
T, A and H.
Xgr, a, A and H we can obtain a rough value any motor. for a for
In other words, having selected these four quantities for a motor
which we are designing, we can obtain a. Having calculated
A B, the magnetizing current, by the methods already set forth
on page 139, we may divide it by a and thus obtain BD. For I

we have:

We are now in a position to construct, for the 200 h.p. motor

which we are designing, a diagram of the kind represented in
Fig. 85.
We must first determine a. We have:

= 32.5 cm.;
T
=
A 1.3 mm.;
H = 12.
"
Knowing these four quantities, the circle factor," a, may be
obtained from Table 27. For our motor we find from the table,

a = 0.041.
The values of a in Table 27 apply to designs with intermediate
proportions as regards slot openings. Should both stator and
rotor slots be very nearly closed (say 1 mm. openings), the value
of a would be increased by say 20 per cent or more. On the
other hand, were both stator and rotor slots wide open, a would
be decreased by say some 20 per cent below the values set forth
in the table. It cannot be too strongly emphasized that we can-
not predetermine a at all closely. We can, however, take O.C41
as a probable value for a in the case of our design. If, on test,
the observed value were found to be within 10 per cent of 0.041,
150 POLYPHASE GENERATORS AND MOTORS

<N l> IQrH CO C^ <N 00 CO <N

CO CD
0000 o o o 00
o t^. Tfl IO

ooooo OOOOO ooooo ooooo ooooo

rH 00 (M rH t^ 00 >O O
00 1O
OOOOO *f iO CO CO l>

ooooo ooooo ooooo ooooo ooooo

<M O5 IO (M C5 C^T OS !>
co t^ oo oo cb rt< U3 O CO
oooo
fcO
ooooo ooooo ooooo ooooo ooooo
00 Oi O O o co r^- 1> oo
O O r- rH (

ooooo ooooo ooooo ooooo ooooo

88S23 O
O 00
O "* t^-
O <N O!> co oq 01 o
oooo T (
CO l^ t^ TP 10 co t^

ooooo ooooo ooooo ooooo ooooo

00 <N SCO'*
O rH rH <N <N O rH rH ec O rH rH <N <j O rH rH *l i O rH rH <N <N

OO
O rH
<N Tt< (M OOO <M 1>- CO rH (N rH ^ GO
OOO
00 OS (N
rH rH rH O O O rH
00 00
T 1
T i
t^oooiO
OO O rH T^l -HH

ooooo ooooo ooooo ooooo 00000

CO (M O O -*f <M 1C rH 00 CO -^ TH !>. CO
O O rH (M
i i rjn
iO t^ OOOOOiO
O5
O OO O rH
i i I-H
i ' T 1

OOOOO ooooo ooooo ooooo ooooo

rf (N T-I ^
O T^ O O CO ^ OO
oo O5 o
OO
(

(N CO
rH rH rH rH
i (

ills rH <M i>- 1^-

OOOOO ooooo ooooo ooooo ooooo

00 CO
C^ CO
>O CO "* 00 T^I
00 OS CO rH C^l
CO rH 00 00
IQ CO CO OO 00
O
OO rH rH rH O O O rH rH
ooooo ooooo OOOOO oddod ooooo
(XXfft SOO^ 00 (N OO^
O iH rH W W O rH rH >'l ffl O rH rH ei ffl O rH rH W S-1

a-
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 151

O ^ OS
tO CO Tfri
"
<M
2222 I"- i-< i-Hl> lO OS "*
CO ^t
OOO ^t lO
OO
1 1

SSS
CO 'M'
OO CM CM CO CO -*

ddddd 00000 ddddd ddd

CM 00 <M OS
*
CO "*
8SSS to to

doddd 00000 00000 00000 00000

^
oo ^O
os r* co CO ** i I

^^
CO tO CO CO TJH Tt< tO CO
8SSp8
ddddd ddddo d d odd 00000 00000 I

O CO CO GO
^ T^ iO CO CO
t i t^ co 1^ co
CO Tf iO tQ CO
i 1
CO 00 rf OS
i

CO CO Tt< lO lO
I

CO CO CO CO CO opopo ppppp
ddddd ddddd ddddd
00 ffl SO O rt< 00 <N SO O Tt< 00 (N SO O*
iO i i CO CO OS
O
TF Tt< *O 1C
888SS
ooooo ddddd ddddd ddddd
t> * l-H 1 I

9^
o
ddddd ddddd
co
i to CM o CO '^ '^ ^O CO

ooooo CO Tfl i-H OS CO

ddddd ddddd ooooo GO CO C^l OS t^*
iO O CO N- OfJ CO T^ tO O CO

ddddd 99999
ddddd ooooo dddoo ooooo
O O
CO CO GO rt<
CO
CM
f~-

ddddd d o' d d d do odd ooooo I

oo ffi so o^
O rH r-i (N CM O iH iH <N CM OrHrHcMcM Or-lr-lcMCM

CO CM tOO Ttn 5C CO O to OS T^
ii } ^t to O to CO
1

coco-^ ^ to
ooooo ddddd joddoo ooooo ddddd
O
GO CO CM CO O OS CM OS to CO
ooooo ppppp rt* Tf( to CO CO

ddddd ooooo O O O ooooo

i
-^ Os to O to CO CO 1
s
- l~-
t^-
TJH
CO Os CO ^^
to to CO t^-

ooooo ddddd ddddd ooooo OOOOO I

(N GO rf O cO CM OS CO <M OS
ooooo tO to CO 1>- Is-

ddddd ddddd ddddd ooooo ooooo

CO C^ i

CO CO CO T~H OS tO GO Os CO -^
1>-GO OS i

t--GO (

tO CO t^ GO GO CO Tf< lO CO !>
co co co >O O'
odd oo'od ooooo ddddd ooooo
00 CM SO 00 CM SO O^ QC CM CP O *(

00 rH T- CM CM d i-i TH <N' CM O T-i TH CM CM

152 POLYPHASE GENERATORS AND MOTORS
the result should be considered to be as close as could reasonably
be expected.
This indication of our inability to closely predetermine a
should not lead to a disparagement of its utility. Most practical
purposes are amply satisfied, in commercial designing, when we
can, in our preliminary work, construct the circle diagram with this
degree of accuracy. It will be seen in the course of this treatise
that the efficiency and power-factor can be closely predetermined
in spite of this degree of indeterminateness in the circle ratio.
At the author's request, Mr. F. H. Kierstead has recently
analyzed the test results of 130 polyphase induction motors,
and from these results he has derived a formula for estimating
the circle ratio. The range of dimensions of the 130 motors may
be seen from the following :

A varies from 0.64 mm. to 2.54 mm.

Xgr 10 cm. to 61 cm.
T 11 cm. to 84 cm.
H 7 to 32
D 20 cm. to 310 cm.

Fifty-eight of the motors were of American manufacture and

the remaining 72 were of British, German, Swedish, Swiss, French,
and Belgian manufacture. Our object was to obtain a formula
which would yield an approximately representative value, regard-
the detail peculiarities of design inherent to the independent
less of
views and methods of individual designers and manufacturers.
For squirrel-cage motors Kierstead's formula is as follows:
N
'0.20 0.48 3.0
c = i
. ,

C is a function of A, and may be obtained from Table 28.

TABLE 28. VALUES OF C IN KIERSTEAD'S FORMULA FOR THE CfRCLE RATIO.

A
(in mm.).
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 153

Of course it is realized that the width and depth of the slot,

width of the slot openings, the arrangement of the end con-
nections, the degree of saturation, and other factors which have
not been taken into consideration, necessarily materially modify
the value of a, the circle ratio. Indeed, all experienced designers
know that the introduction of some extreme proportion with
respect to dimensions additional to those taken into account in
Kierstead's formula, will be accompanied by quite a large increase
or decrease in the circle ratio. Consequently, the formula must
be taken as corresponding to average proportions. When any
extreme departure is made from these average proportions, it
will be expedient for the designer to employ his judgment as to
the extent by which he should modify the value of the circle
ratio as obtained from the formula.
The
investigation is at present being continued with a view
to arriving at a term to be introduced into the formula to take
into account the slot dimensions. In its present form, however,
Kierstead's formula constitutes a valuable aid to the designer of
induction motors. Table 27 was derived before Kierstead under-
took his investigation and is based upon data of 71 out of the
130 motors analyzed by Kierstead.
Below is given a bibliography of a number of useful articles
relating to the estimation of the circle-ratio.
"
I. Chapter IV (p. 29) of Behrend's Induction Motor."
II. "The Magnetic Dispersion in Induction Motors," by Dr. Hans Behn-
Eschenburg; Journal Inst. Elec. Engrs., Vol. 32, (1904), pp. 239 to 294.
"
III. Chapter XXI on p. 470 of the 2d Edition of Hobart's Electric Motors."
IV. "The Leakage Reactance of Induction Motors," by A. S. McAllister,
Elec. World for Jan. 26, 1907.
V. "The Design of Induction Motors," by Prof. Comfort A. Adams, Trans-
actions Amer. Inst. Elec. Engrs., Vol. 24 (1905), pp. 649 to 687.
VI. "The Leakage Factor of Induction Motors," by H. Baker and J. T.
Irwin, Journal Inst. Elec. Engrs., Vol. 38, (1907) pp. 190 to 208.

yand a the Two Most Characteristic Properties of a Design.

We have now determined the two most important characteristics
of our design. These are:

y, the ratio of the no-load current to the full-load current;

a, the circle ratio.
For our 200-h.p. design we have:
Y = 0.242;
a = 0,041.
154 POLYPHASE GENERATORS AND MOTORS
The Circle Diagram of the 6-pole 200-H.P. Motor. For the
full-load current we have:

7 = 104 amperes.

No-load current = yX/ = 0.242X104 = 25.2 amperes.

This value of 25.2 amperes is plotted as AB in Fig. 86.

Diameter of circle = AB =777:17

25/2
= 615 amperes.
a 0.041

In Fig. 86, the diameter of the circle is made equal to 615

amperes.
Significance also attaches to the length AD which, in Fig.
86 equals (25.2+615 = ) 640 amp. AD represents a quantity

Wattless Component of Current

FIG. 86. The Circle Diagram of the 200-H.P. Motor.

"
which may be termed the ideal short-circuit current." It is
that current which, if the stator and rotor windings were of zero
resistance and if there were no core loss, would be absorbed by
each phase of the stator windings if the normal pressure of 1000
volts (577 volts per phase) were maintained at the terminals
of the motor, the periodicity being maintained at 25 cycles per
second. In other words, the reactance, S, of the motor, under
these conditions, is:

= = 0.902 ohm per phase.

POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 155

Estimation of the Full-load Power-factor. It will be remem-

bered that up to this point we have employed for the full-load

power-factor, a preliminary value taken from Table 18. This

preliminary value was Gr = 0.91.
In Fig. 86 the line AE has been drawn with a length of 104
amperes, from A
to the point E
of intersection with the circle.

EF, the vertical projection of this line, is found to be 97 amperes.

Thus the diagramatically-obtained value of the power-factor is

Let us take it as 0.93 and let us revise our data in accordance

with this new value. We have:

Full-load efficiency (YJ)

= 0.91 (original assumption) ;

Full-load power-factor (G) = 0.93 (revised value) ;

Full-load current (I) = ^91

u.yj
X 104 = 102 amperes;

No-load current =25.2 amperes;

25 2
X = -r~ = 0.247 (revised value) ;
i(jz

a = 0.041 (unchanged).

The diagram, drawn to a larger scale, and with the slight

modification of employing this new value of 102 amperes for I
(the full-load current), is drawn in Fig. 87. It is seen that the
vertical projection of I is now 95. This gives us for the power-
factor :

confirming the readjusted value.

The Stator I 2 R Loss. We shall estimate the stator I2 R loss
by first estimating the mean length of one turn of the winding,
and from this obtaining the total length of conductor per phase.
From this length and the already-adopted cross-section of the
156 POLYPHASE GENERATORS AND MOTORS

conductor, the resistance per phase is obtained. We may denote

the resistance per phase by R. The stator PR loss at rated
load is equal to 3I 2 R.

^w
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 157

For our 1000-volt 200 h.p. design, we find the value:

# = 3.0.
Therefore :

mlt. = 2X43.0+3,OX32.5 = 183.5 cm.

For the number of stator turns in series per phase we have:

T=120

Consequently the length of conductor per phase equals:

120X183.5 = 22000 cm.

The cross-section of the conductor has already been fixed at:

0.302 sq.cm.

The PR loss at 60 Cent, is desired. At this temperature,

the specific resistance of commercial copper wire may be taken
at 0.00000200 ohm per cm. cube.
Thus at 60 Cent, the resistance of each phase of the stator
winding is:

22000X0.0000020

For the stator I2 R loss at full load we have:

3 X102 2 X 0.146 = 4540 watts.

Stator IR Drop at Full Load. Instead of a flux corresponding

/1000 \
to (
=- = 577 volts per phase, we shall, at full load, have a

lesser flux. It will, in fact, correspond to :

577-102X0.146 = 577-15 = 562 volts.

158 POLYPHASE GENERATORS AND MOTORS
This is an internal drop of:

~X 100 =
Oil
2.6 per cent.

Strictly speaking, we ought, therefore, to. take into account

the decreased magnetic densities with increasing load. Cases
arise where it would be of importance to do this, but in the present
instance such a refinement would be devoid of practical inter-
est and will not be undertaken.

THE DETERMINATION OF THE CORE LOSS.

For the statorcore, the best low-loss sheet-steel should be

employed notwithstanding that its cost is still rather high. The
advantage in improved performance will much outweigh the very
slight increase which its use occasions in the Total Works Cost.
In the rotor, the reversals of magnetism are, during normal
running, at so low a rate that the rotor core loss is of but slight
moment. It is consequently legitimate to employ a cheaper
grade of material in the construction of the rotor cores. But in
practice it is usually more economical to use the same grade as
for the stator cores notwithstanding the absence of need for the
better quality. By the time the outlay for waste and the outlay
for wages and for general expenses are added, there will be but

trifling difference in the cost of the two qualities.

The data given in Table 30 are well on the safe side. Individ-
ual designers will ascertain by experience in how far they can
rely upon obtaining better material.

TABLE 30. DATA FOR ESTIMATING THE CORE Loss IN INDUCTION MOTORS.

Density in Stator
Core in Lines per
Square Centimeter.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 159

For our 25-cycle motor, we have employed in the stator core,

a density of 10 000 lines per sq.cm. and we shall consequently
estimate the core loss on the basis of 4.0 watts per kg. of total
weight of stator core.

ESTIMATION OF WEIGHT OF STATOK CORE

External diameter of stator core = 89. 2 cm.

Internal diameter of stator core =62.0 cm.

The gross area of a stator core plate, i.e., the area before deduct-
ing the area of the slots, is equal to :

2
-62.0 2 )=3260 sq.cm.
|(89.2

Area of 72 stator slots = 72X5.4X1. 26 = 490 sq.cm.

Net area of stator core plate =(3260 -490 = ) 2770 sq.cm .

The volume of the stator core is obtained by multiplying this

area by \n, i.e., by 27.9.

Volume of sheet steel in stator core = 2770X27.9 = 77 400 cu.cm.

Weight of 1 cu.cm. of sheet steel = 7.8 grams.

Therefore :

Weight of sheet steel in stator core = - = 603

kg.
J.UUU

The accuracy with which core losses can be estimated is not

such as to justify dealing separately with the teeth and the main
body of the stator core. It suffices simply to multiply the net
weight in kg. by the loss in watts per kg. corresponding to the
density in the main body of the stator core. Therefore:

Stator core loss = 603X4.0 = 2410 watts.

Core Loss in Rotor. The periodicity of reversal of magneti-

zation in the rotor core is so low that there should not be much
core loss in the rotor. But to allow for minor phenomena and to
keep on the safe side, it is a good rule to assess the rotor core
160 POLYPHASE GENERATORS AND MOTORS
loss at 10 per cent of the stator core loss. For our machine we
have :

.
Rotor core loss = 2410 X 0.10 = 240 watts.

Input to Motor and to Rotor at Rated Load. We cannot

yet check our preliminary assumption of an efficiency of 91 per
cent at rated load. On the basis of this efficiency, the input to
the motor at its rated load is:

200X746
=164 000 watts.
L/.y -L

The losses in the stator amount to a total of :

4540+2410 = 6950 watts.

Deducting the stator losses at full load from the input to the
motor at full load we ascertain that:

164 000 - 6950 = 157 050 watts

are transmitted to the rotor.

A Motor is a Transformer of Energy. A motor receives
"
energy in the form known as electricity." An account can be
rendered of all the energy received. In the case we are consider-
ing, the full-load input is at the rate of 164 000 watts. Energy
flows into the motor at the rate of 164 kw. hr. per hour.* In cer-
tain instances more convenient to make some equivalent state-
it is

ment with other power and time. Thus we may say that
units of
energy flows into the motor at the rate of 164 000 watt seconds

per second. The amount of energy corresponding to the expendi-

ture of one watt for one second is termed by the physicist, one
joule.

1 joule
=1 watt second.

It requires 4190 joules to raise the temperature of I kg. of

water by 1 cent. Thus we have:

1 kg. calorie (kg.cal.) =4190 joules.

* amount of energy correspond-

The proposal to designate as 1 kelvin the

ing to 1 kw. hr., is gradually gaining favor amongst European engineers.

POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 161

The energy expended in lifting a weight of 1 kg. through a

height of 1 meter is equivalent to 9.81 joules. In other words

1 kg.m. = 9.81 joules.
The energy need not necessarily be expended in lifting a weight.
If a force of 1 kg. is exerted through a distance of 1 meter, an
amount of energy equal to 9.81 joules is expended in the process.
The amount of energy corresponding to the expenditure of
1 h.p. for 1 sec. is equal to 746 joules; or

1 h.p. sec. = 746 joules.

It is convenient to bring together these equivalents:

1 watt second = 1 joule;

1 h.p. second =746 joules;
1 kg.m. =9.81 joules;
1 kg.cal. =4190 joules.
1 kelvin =3 600 000 joules.

The rated output of our motor may be expressed as:

1. 200 h.p.;
la. 200 h.p. sec. per sec.;
2. 149 200 watts;
2a. 149 200 joules per second.

3. = 15 2 kg.m. per second.

/149 200
=\ .
kg.cal. per second.
\ 4190 /

Let us concentrate our attention on Designation 3 for the

rated output. In accordance with this designation, the motor's
output at its rated load is

15 200 kg.m. per sec.

If the shaft of our motpr is supplied with a gear wheel of 1

meter radius through which it transmits the energy to another
engaging gear wheel, and thence to the driven machinery, then
162 POLYPHASE GENERATORS AND MOTORS
for every revolution of the armature, the distance travelled by a
point on the periphery of the gear wheel is

2Xx = 6.28 meters.

At the motor's synchronous speed of 500 r.p.m., the peripheral

speed of the gear wheel is

500
6.28 = 52.3 meters per sec.
X-gQ-

(Although such a high speed would not, in practice, be

employed, it has been preferable, for the purpose of the present
discussion, to consider a gear wheel of 1 meter radius.)
Since at full load the motor's output is :

15 200 kg.m. per sec.

and since the peripheral speed of the gear wheel is 52.3 meters

per sec., it follows that the pressure at the point of contact between
the driving and the driven gear teeth is

This is the force exerted at a radius of 1 meter. We say

" "
that at full load the motor exerts a torque of

290X1 = 290 kg. at 1 meter leverage.

Hadthe radius of the gear been only 0.5 m. instead of 1.0

meter, then the force would have been 2X290 = 580 kg., but the
" "
torque would still have been equivalent to

0.5X580 = 290 kg. at 1 meter leverage.

In dealing with torque it is usually convenient to reduce it

to terms of the force in kg. at 1 meter leverage, irrespective of
the actual leverage of the point of application of the force.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 163

At rated load our motor will not run at the synchronous speed
" "
of 500 r.p.m., but at some slightly lower speed. If the slip
is 1 per cent, then the full-load speed will be

0.99X500 = 495 r.p.m.

" "
If the slip is 2 per cent, then the full-load speed will be

0.98X500 = 490 r.p.m.

Since the rated output will nevertheless be 200 h.p. in these

two cases, the torque will be; not 290 kg. at 1 m. leverage, but

/ 290
=
'

( rTnn )
293 k S- * r tne case * * P er
u.yy

and

= 1 296 kg. for the case of 2 per cent slip.

(290
i/O
\J

The torque (i.e., the force at 1 m. leverage) is only one com-

ponent of the energy delivered. The other component is the

distance traversed by a hypothetical point at 1 meter radius

revolving at the angular speed of the rotor. For 2 per cent slip,
the distance traversed in 1 sec. by such a hypothetical point, is

0.98X52.4 = 51. 3 meters.

The energy delivered from the motor in each second is thus

296X51.3 = 15200kg.m.

The power (or rate of deliverance of energy) is

15 200 kg.m. per sec.;

or,

9.81 X 15 200 = 149 200 watts;

or,

149 200
= 200 h.p.
746
164 POLYPHASE GENERATORS AND MOTORS
The torque exerted by the rotor conductors is greater than
that finally available at the gear teeth. The discrepancy cor-
responds to the amount of the PR
loss in the rotor conductors
and to the amount of the rotor core loss and the windage and
bearing friction. The rotor core loss and the friction come
in just the same category as an equal amount of external
load. Thus if 3 h.p. is required to supply the rotor core
loss plus friction, then the output from the rotor conductors
is 203 h.p. as against the ultimate output of 200 h.p. from
the motor.
But the PR loss in the rotor conductors comes in an altogether
different category. The loss can only come about as the result of a
cutting of the flux across the rotor conductors. In other words,
the rotor conductors must not travel quite as fast as the revolving
magnetic field. Consequently the rotor will run at a speed
" "
slightly less than synchronous; there will be a slip between
the revolving magnetic field and the revolving rotor. It is only
in virtue of such slip that the rotor conductors can be the seat
of any Thus the torque is inseparably associated with
force.
" " " "
the slip and the slip will be greater the greater the load.
" "
As the slip and torque increase, the rotor PR loss also increases
and the speed of the rotor decreases.
If the PR loss in the rotor conductors amounts to 1 per cent
" "
of the input to the rotor, then the slip will be 1 per cent.
" "
If the PR
loss is increased to 2 per cent, then the slip increases
to 2 per cent. If, finally, the PR
loss amounts to 100 per cent of
" "
the input to the rotor, then the slip will be 100 per cent,

i.e., the motor will be at rest, but it may nevertheless be exerting

torque. For such a condition it is desirable to regard matters
from the following standpoint. If the rotor is suitably secured
so that it cannot rotate; then if electricity is sent into the motor
a certain portion will be transmitted by induction to the rotor
circuit, just as if it constituted the secondary circuit of a trans-
former. The input to the rotor will under these conditions
PR loss in its conductors and the core loss. There
consist of the
isno other outlet for the energy sent into the rotor and it all
becomes transformed into energy in the form of heat in the rotor
conductors and in the rotor core. Since under such conditions
the rotor core loss is negligible in comparison with the rotor
PR loss, we may regard the input to the rotor as practically
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 165

identical with the rotor PR loss. The rotor PR loss is 100

per cent of the input to the rotor and the slip is 100 per cent.
The Locus of the Rotor Current in the Circle Diagram. Let
us consider the hypothetical case of a rotor with a number of
conductors equal to the number of stator conductors. We can ,

describe this as an arrangement with a 1 1 ratio of transformation. :

Any actual induction motor can be considered to have its equivalent

with a 1 : 1 ratio, and it is convenient and usual to investigate
certain properties of induction motors
by considering the equivalent rotor wind-
ing with a 1 1 ratio. :

The vector diagram of the stator and

rotor currents is shown in Fig. 88 for

our motor with a l:l-ratio

200-h.p.
rotor. The diagram is drawn for full
load and consequently the stator current,
AE, is equal to 102 amperes. The rotor
current, AF must have such direction
and magnitude that the resultant, AB,
shallbe equal to the no-load magnetizing
current, which we have already found to
be 25.2 amperes. AF is consequently
equal, as regards phase and magnitude,
to EBand is found graphically to amount
to 96 amperes. In practice, it is more
convenient to represent the rotor current
by lines B as the origin and
drawn from
connecting B
to the points where the FIG. 88. Vector Diagram
corresponding primary vectors intersect Indicating the Primary
the circumference of the semi-circle. In and Secondary Currents
in the 200-H.P. Induc-
Fig. 89 are drawn the stator and rotor
tion Motor at its Rated
vectors for two values of the stator cur- Load.
rent, namely 80 amperes and 300 amperes.
The corresponding values of the rotor current are 72 and 286
amperes.
The Rotor 2
R Loss of
the 200-H.P. Motor at Its Rated Load.
I

postpone to a later stage the design of the

It is still desirable to
rotor conductors. But let us assume that the full-load slip
will be 2.0 per cent. Then the rotor loss at full load will PR
be 2.0 per cent of the input to the rotor. The input to the rotor
166 POLYPHASE GENERATORS AND MOTORS

has been ascertained (on page 160) to be 157 050 watts. Thus
the rotor PR
loss at full load is 0.02X157 050 = 3140 watts.
We
have now determined (or assumed) all the full-load losses
except windage and bearing friction.

//

20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340
Wattless Component of Current
FIG. 89. Diagram Indicating Stator and Rotor Current Vectors.

Friction Losses. It is very hard to generalize as regards

friction losses. A reasonable estimate may, however, be made
from Table 31.

TABLE 31. DATA FOR ESTIMATING THE FRICTION Loss IN BEARINGS AND
WINDAGE IN INDUCTION MOTORS.

Z> 2X0 (D and \g in dm.)

POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 167

For our motor we have :

= 166.

From Table 31 we ascertain by interpolation that the friction

loss may be taken as 1300 watts.
Output from Rotor Conductors. The output from the rotor
conductors is made up of the 200-h.p. output from the motor,
the friction, and the rotor core loss. The two latter amount to

40\
=
(i^M=2.06h.p.
Thus we may take the output from the rotor conductors as

200+2.06 = 202 h.p.

Referring back to page 163 we find that for 2 per cent slip
and 200 h.p., we
require 296 kg. at 1 meter leverage.
Consequently the torque required to be exerted by the rotor
conductors is :

202
-X296 = 299 kg. at 1 meter leverage.

In other words, the full-load torque exerted by the rotor conductors

is299 kg.
The Torque At full load, the input to the rotor is
Factor.
157 050 watts. Thus the input to the rotor in watts per kg.
"
of torque developed, which we may term the torque factor," is:

157050__ .

~299~

This factor will be useful to us in studying the starting torque.

The " Equivalent " Resistance of the Rotor. We have on
p. 157 made an estimation of the stator resistance and have
ascertained it to be 0.146 ohm per phase at 60 cent.
The stator PR loss at full load is 3 X102 2 X 0.146 = 4540 watts.
The rotor I 2 R loss at full load is 3140 watts.
168 POLYPHASE GENERATORS AND MOTORS
We shall employ a squirrel-cage rotor (for which we shall
" "
soon design the conductors) and we may consider its equivalent
resistance to be:.

X 0.146 = 0.101 ohm.

Without introducing any serious inaccuracy we may (when

the motor is running in the neighborhood of synchronous speed),
ascertain the rotor PR loss for any value of the stator current
" "
(except for very small loads) by multiplying this equivalent
resistance by three times the square of the stator current.
Rotor at Rest. But when the rotor is at rest, the currents
circulating in its windings are of the line periodicity and the
conductors have an apparent resistance materially greater than
their true resistance. Attention was called to a related phe-
nomenon in a paper presented by A. B.
Field, in June, 1905, before
the American Institute of Electrical Engineers and entitled
"
Eddy Currents in Large Slot-Wound Conductors."* Recently
the application of the principle has been incorporated in the
design of squirrel-cage induction motors to endow them with
desired values of starting torque. The multiplier by which the
apparent resistance may be obtained from the true resistance may
be found approximately from the formula:

Multiplier = 0. 15 X (depth of rotor bar in cm.) X Vperiodicity.

In our case we have:

Multiplier = 0. 15 X 5.4 X V25 = 4.05.

This multiplier only relates to the embedded portions of the

conductors. The portions of the length where the conductors
cross the ventilating ducts are not affected, nor are the end rings

subject to this phenomenon. A rough allowance for this can be

made by reducing the multiplier to 0.8X4.05 = 3.24.
Thus at standstill the " apparent " resistance of our rotor will
be 3.24X0.101=0.327 ohm.

"Vol. 24, p. 761.

POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 169

The " " Resistance of the Motor. At

Equivalent standstill,
the motor, regarded as a whole, and without distinction between
primary and secondary may be considered as having a resistance
of (0.146+0.327 =) 0.473 ohm per phase.
On page 154 we have seen that the reactance of the motor is:

S= 0.902 ohm per phase.

For the impedance of the motor at standstill we have:

Vo.4732+0.902^1.03 ohm per phase.

THE STARTING TORQUE

In starting this motor we should not put it at once across
the full pressure of 1000 volts but should apply only one-half or

one-third of this pressure. Let us examine the conditions at

standstill when half pressure is applied.

We shall have ^ = 288 volts per phase.

The stator current will be:

OQQ
= 280 amperes per phase.

For the rotor PR loss we have 3X280 2 X0.327 = 77 000 watts.

The torque developed is obtained by dividing this value by
the torque factor, i.e., by 525.

77 000
Torque = _ = 146 kg. at 1 meter leverage.

The full-load torque is 296 kg.

Consequently with half the normal pressure at the motor, we

shall obtain

146
49.5 per cent of full-load torque.
170 POLYPHASE GENERATORS AND MOTORS
We obtain the half pressure at the motor by tapping off from
the middle point of a starting compensator. The connections
(for a quarter-phase motor), are as shown in Fig. 90. With this
arrangement, the current drawn from the line will be only half of
the current taken by the motor. Since the motor takes 280
amperes, the current from the line is only

/280
- \ ,
I
s-= ) 140 amperes.

)
=\ 1-37 times full-load current from the
TQO
(140
line, we can start the motor with 50 per cent of full -load torque.

FIG. 90. Connections for Starting Up an Induction Motor by Means of a

Compensator (sometimes called an auto-transformer).

This excellent result is achieved by employing deep rotor con-

ductors and does not involve the necessity of resorting to high
slip during normal running.
Circle Diagram for These Starting Conditions. For half
pressure, the magnetizing current will, strictly speaking, be a
than half its former value of 25.2 amperes, since the
little less

magnetic parts are worked at lower saturation. But for simplicity

~ 2 \
= } 12.6 amperes. The circle ratio, <j,
(25
varies a little with the saturation, but let us take it at its former
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 171

value of 0.041. We can now construct the circle diagram shown

in Fig. 91, in which:

AB 12.6 amperes;

= amperes;

AD = 12.6+308 = 321;
AE = 280 amperes.
16U
150
172 POLYPHASE GENERATORS AND MOTORS
This is in good agreement with the value of 77 000 watts which
we obtained for the rotor PR loss by applying the analytical
method.
EF is, for constant pressure, a measure of the input to the motor
and is to the scale of

113 000
= 865 watts per ampere.
lol

[It can also be seen that this would be the case from the cir-
cumstance that at a pressure of 288 volts per phase, the input
is equal to :

(3X288X7) = (865 7) watts].

The rotor PR loss of 78 700 watts may then be represented

by the height FG corresponding to (

= )
91.0 amperes.
\ oDO /

and the stator PR loss may be represented by the remainder,

GE, corresponding to

=. amperes.

It also necessarily follows that FG (and corresponding vertical

heights for other conditions similarly worked out), is a measure
of the torque. When
used in this way, the scale is

146
j-r-
r = 1.60 kg. (at 1 meter leverage) per amp.

Some General Observations Regarding the Circle Diagram.

It is this adaptability to the forming of mental pictures of the
occurrences, which renders the circle diagram of great importance
in the design of induction motors. All the various calculations
involved in induction-motor design may be carried through by
analytical methods but it is believed that these exclusively analyt-
ical methods are inferior in that they disclose no simple picture
of the occurrences. It is well known that in practice the locus
of the extremity of the stator-current vectors is rarely more than
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 173

a very crude approximation to the arc of a circle and that the

designer can only consider as rough approximations the results
he deduces on the basis of the circle assumption. But when
employed with judgment the plan is of great assistance and suffices
to yield reasonable results.

THE SQUIRREL CAGE

Let us now design our rotor's squirrel cage. As yet we have

merely prescribed that the full-load I R loss in the squirrel cage
2

shall be 3140 watts, that the slots shall be 54 mm. deep and that
we shall employ a number of slots not differing very materially
from 72, the number of stator slots.
The Number of Rotor Slots. Were we to employ 72 rotor
" "
slots, the motor would have a strong cogging tendency. That
is to say, if, with the rotor at rest, pressure were applied to

the stator terminals, there would be a strong tendency for the

"
rotor to lock," at a position in which there would be a rotor
slot directly opposite to each stator slot. This tendency would
very markedly interfere with the development of the starting
torque calculated in a preceding section. The choice of 71 or
73 slots would eliminate this defect, but might lead to an unbal-
anced pull, slightly decreasing the radial depth of the air-gap at
one point of the periphery and correspondingly increasing its
depth at the diametrically opposite point. This excentricity
once established, the gap at one side would offer less magnetic
reluctance than the gap diametrically opposite, thus increasing
the dead-point tendency. But by selecting 70 or 74 slots, this
defect is also eliminated. Let us determine upon 70 rotor slots
for our design.
It may in general be stated that the tendency to dead points
at starting will be less.
1. The smaller the greatest common divisor of the numbers
of stator and rotor slots.
2. The greater the average number of stator and rotor slots

per pole.
3. The less the width of the slot openings.
4. The greater the resistance of the squirrel-cage.
5. The deeper the rotor slots.
 174 POLYPHASE GENERATORS AND MOTORS
The two factors will be better understood
influence of the last
if it is
pointed out that they determine the rotor loss at start- PR
ing, and we have already seen that the starting torque is pro-
portional to the loss PR
in the rotor. The fluctuations in the
starting torque arising from variations in the relative positions
of the stator and rotor slots, will obviously be a smaller percentage
of the average starting torque, the greater the absolute value of
the average starting torque. Thus if the average starting torque
isvery low, a small fluctuation might periodically reduce it to
zero; i.e., the< motor would have dead If, on the other points.
hand, the average starting torque same fluctuations,
is high, these
superposed on this high average starting torque, would still leave
a high value for the minimum torque, and there would be no
dead points.
The Pitch of the Rotor Slot. The external diameter of the rotor
is (620-2X1.3 = ) 617.4 mm.
The diameter at the bottom of the slots is :

(617.4 -2X54
= ) 509.4 mm.

Consequently the rotor slot pitch at the bottom of the slot is :

509.4 Xx
= 22.8 mm.
70

The depth of the rotor conductor can be practically identical

with the depth of the slot. Therefore depth of rotor conductor
= 54 mm.
Ratio of Transformation. We have 72 stator slots and 10 con-
ductors per slot; hence a total of 720 stator conductors; as against
only 70 rotor conductors. The ratio of transformation is thus:

720 : 70 = 10.3 : 1.

We have already estimated that for a 1 : 1 ratio, the rotor

current would, at full load, amount to 96 amperes. We are now
able to state that with the actual ratio of 10.3 : 1, the current
in the rotor face conductors will be 10.3X96 = 990 amperes.
At each end of the rotor core, the rotor conductors will termin-
ate in end rings. It will be desirable, for structural reasons, to
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 175

space these end rings (in this particular motor), 2.0 cm. away from
the ends of the laminated core. Consequently the length of each
conductor between end rings will amount to

43+2X2 = 47 cm.

The PR loss in the 70 rotor conductors will be equal to that

in a single conductor of the same (as yet undetermined) cross-
section, but with a length of

70X47 = 3290 cm.

We have seen that we wish, at full load, to have a loss of 3140

watts in the squirrel cage. Were the loss in the end rings negli-
gible (the entire 3140 watts being dissipated in the 70 face con-
ductors, then their section would be so chosen that they should
have an aggregate resistance of:

= 0.00320 ohm.

Since at 60 cent., the specific resistance of commercial copper

is 0.00000200, the required cross-section would be:

3290X0.00000200 =
2.06sq.cm.
0.00320

Since the depth of a rotor conductor is 54 mm., its width would

thus require to be:

= 3.82 mm. -W,

But we cannot afford to provide so much material in the end
rings as to render them of practically negligible resistance. Let
us plan to allow a loss of 628 watts (20 per cent) in the end rings,
the remaining (3140 628 = ) 2512 watts occurring in the slot con-
ductors. This will require an increase in the width of the slot
conductors, to:

= 4.78 mm. or 4.8mm.

176 POLYPHASE GENERATORS AND MOTORS
Thus the slot conductors will have a cross-section of 54
mm.X4.8 mm. The rotor slot will also be 54 mm. deep and
4.8 mm. wide, and there will be no insulation on the rotor con-
ductors.
The End Rings. The object in minimizing the loss in the
end rings will have been divined. Since we want to have the
apparent resistance at starting, as great as possible for a given
" "
slip at normal load, we want to concentrate the largest prac-
ticable portion of the loss in the slot conductors since it is these
conductors which manifest the phenomenon of having, for high-
periodicity currents, a loss greatly in excess of that occurring when
they are traversed by currents of the low periodicity corresponding
"
to the slip."
The Current in the End Rings. It can be shown* that the
current in each end ring is equal to:

Number of rotor conductors

- ,
^ ; X current per slot conductor.
iuX number of poles

For our motor we have:

Full-load current in each end ring = -^X 990 = 3660 amperes.

Since we have allowed 628 watts for the loss in the end rings ;

we have a loss of 314 watts per end ring, and we have:

Resistance for one end ring = - = 0.0000234 ohm.

^2

Each end ring will have a mean diameter of:

61.7+50.9 =
56.3 cm.
^

and a mean circumference cm. of 56.3x = 177

Consequently the cross-section of each end ring is :

177X0.00000200 Q
0.0000234

*
The proof is given in Chapter XXIII (pp. 490 to 492) of the 2d Edition
of the author's
"
Electric Motors " (Whittaker & Co., 1910).
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 177

Let us make the cross-section up with a height of 54 mm.

and a width of 28 mm. A conductor of this cross-section will, at

Section on A.B Looking in Direction

of Arco.ws ,

FIG. 92. Outline of Squirrel-cage Rotor for 200-H.P. Polyphase Induction

Motor.

25 cycles, be subject to a slight increase in resistance, due to

ordinary skin effect, but this will not be of sufficient amount to
add much to the starting
torque. The increase in re- >j k-2.4mm.
sistance will in fact amount 22 mm
to about 10 per cent.
A
sketch of the rotor is

given in Fig. 92, and a section

of a slot and two teeth, in

Fig. 93. Since the slot pitch

<-4.8 mm
at the bottom of the slot is
22.8 mm. and since the slot
is 4.8 mm. wide, we have a
tooth width of 22.8-4.8 =
18.0 mm.
With 11.7 teeth
(?=)
per pole, the cross-section of FIG. 93. Rotor Slot and Teeth for 200-
the magnetic circuit at the H.P. Squirrel-cage Motor.
bottom of the rotor teeth is

11.7X1.8X27.9 = 588 sq.cm.

178 POLYPHASE GENERATORS AND MOTORS
The crest density at the bottom of the rotor teeth is:

4 570 000
588
X 1.7 = 13 200 lines per sq.cm.

THE EFFICIENCY
We have now estimated all the losses in our 200-h.p. motor.
Let us bring them together in an orderly table :

At full load we have :

Stator PR loss 4 540 watts

2410 "
Stator core loss
Rotor PR loss 3140 "
Rotor core loss 240 "
' '
Friction and windage loss 1 300

Total of all losses 11 630 watts

' '

Output at full load 149 200

Input at full load 160 830 watts

14.0 onrj
Full-load efficiency = - X 100 = 93.0 per cent,
lou ooU

Our assumption for the full-load efficiency was 91.0

original
per cent. Consequently, strictly speaking, we ought to revise
several quantities, such as current input, stator PR loss, rotor
PR loss, and ultimately obtain a still closer approximation to
the efficiency. But the object of working through this example
has been to convey information with respect to the methods of
carrying out the calculations involved in the design of an. induc-
tion motor ;
and there is no special reason to undertake the above
mentioned revision.

THE HALF-LOAD EFFICIENCY

At half load, the energy component of the full-load current
input, will, sufficiently exactly for our purpose, be halved. This
energy component is:

GXl = 0.93X102 = 95 amperes.

POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 179

At half load, the wattless component will be slightly in excess

of 25.2 amperes, the no-load current. Let us take it as 29
amperes. The total current input at half load will be:

55.5 amperes.

The stator PR loss at half load will be:

/55.5\ 2
X 4540 = 1340 watts.
U027

+* 34
Dia. 308-A
32
I
^30
* 28
4J
g 26
I 24

I 22
2
7
&
<3 18
W 16

14

12

10

2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44
Wattless Component of Current

FIG. 94. Diagram for Obtaining Rotor Current at Half-load.

The readiest way to obtain the corresponding value of the

secondary current is to make the circle-diagram construction

indicated in Fig. 94. From this construction we see that for a
1 : 1 ratio, the secondary current would be 47.5 amperes. We
180 POLYPHASE GENERATORS AND MOTORS
have seen on page 165 that at full load, when the 1 1 secondary :

current was 96 amperes, the rotor PR loss was 3140 watts.

Consequently at half load, the rotor PR loss will be:

X 3140 = 770 watts.

Thus at half load we have :

Stator PR loss '

1 340 watts
Stator core loss 2 410
"
Rotor PR loss 770
"
' '
Rotor core loss 240
' '
Friction and windage loss 1 300

Total of all losses 6 060 watts

half load. 74 600
"
Output at

Input at half load . 80 660 watts

74- fiOO
Half-load efficiency = ^
oU DOu^X 100 = 92.5 per cent.

Making similar calculations for other loads we obtain the follow

ing inclusive table of the losses at various loads :

TABULATION OF LOSSES AND EFFICIENCIES AT 60 CENT.

Percentage of rated out-

put
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 181

Estimation of the Temperature Rise. In making a rough

estimate of the heating of the motor the first step consists in cal-
"
culating the watts per square decimeter (sq.dm.) of radiating
surface at the air-gap."
The " radiating surface at the air-gap " is taken as the sur-
face over the ends of the stator windings. While this surface
will vary considerably according to the type of winding employed,
a representative basis may be obtained from the formula:

" "
Equiv. radia. surface at air-gap

100

1 90
2
P4 80
d

I 70
.2

G
I
50
20 40 60 80 100 120 140 160 180 200 220 240 260 280
Output in Horse-power

FIG. 95. Efficiency Curve for 200-H.P. Polyphase Induction Motor.

The k in this formula is a factor which is a function both of

the normal pressure for which the motor is built, and of the polar
pitch, T. Suitable values will be found in Table 32.

TABLE 32. DATA FOR ESTIMATING THE RADIATING SURFACE.

Rated Pressure of the Motor.

182 POLYPHASE GENERATORS AND MOTORS
For our design we have:

D = 62.0;
X0-43.0;
r = 32.5;

"
.'.
Equivalent radiating surface at air-gap

= xX62.0X (43.0+29.3)
= xX62.0X72.3
= 14000 sq.cm.
= 140 sq.dm.

The loss to be considered is, at full load:

11 630 watts.

Thus we have:

Watts per sq.dm. = T- =83.4.

The data in Table 33, gives a rough notion of the temperature

rise corresponding to various conditions.

TABLE 33. DATA FOR ESTIMATING THE TEMPERATURE RISE IN INDUCTION

MOTORS.

Peripheral
Speed in
Meters per
Second.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 183

ipheral speed has already (on page 113) been ascertained to be

16.2 meters per second.
From Table 33 we find that the value for the thermometric-

ally determined temperature rise is:

0.35 per watt per sq.dm.

This gives us a total (thermometrically determined) rise of

83.4X0.35 = 29 Cent.

Although in the predetermination of temperature rises, no

close accuracy is practicable,still, the margin indicated by the
above result is so great that the designer should rearrange the
final design in such a manner as to save a little material at the
cost of a slight increase in the losses.
Since a rise of fully 40 cent, is usually considered quite con-
servative (45 often being adopted), an estimated rise of 35
would leave a sufficient margin of safety.

THE WATTS PER TON

Another useful criterion to apply in judging whether a motor

israted at as high an output as could reasonably be expected, is
"
the watts per ton." The weight taken, is exclusive of slide rails
and pulley. In the case of a design which has not been built,
the method of procedure is as follows :

First estimate the weights of effective material.

Weight Stator Copper.

mlt. = 183.5 cm.

Total number of turns = 3 T = 3 X 120 = 360;

Cross-section stator cond =0.302 sq.cm.;
Volume of stator copper = 183.5X360X0.302
= 19900 cu.cm.;
Weight 1 cu.cm. of copper =8.9 grams;
w . , .

Weight stator copper = 19900X8.9

J-UUu
= 177 kg.
184 POLYPHASE GENERATORS AND MOTORS

Weight Rotor Copper.

Total length rotor face 1
_ O^yU
QOon
/ 1 -n-rr\ '
CHI.;
conds (see page 175) J

Cross-section =5.4X0.48 = 2.59 sq.cm.;

Volume of rotor face conds = 3290X2.59
= 8550 cu.cm.;

Weight rotor face conds = 8550X8.9

= 76 kg.;
Mean circum. end ring =177 cm.;
Cross-section =15.1 sq.cm.;

Weight two end rings

= 2X177X15.1X8.9
1UUU
= 47.5 kg.;
Total weight rotor copper =76+48 = 124 kg.;
Total weight copper I_ 177
, .
i x f -L I I
,
l9d
|~ JL^iTt
(stator plus rotor) j
= 301 kg.

Weight Stator Core.

This has already been estimated (on page 159) to be 603 kg.

Weight Rotor Core.

External diameter rotor core =61.7 cm.;
Internal diameter rotor core =38.3 cm.;

Gross area rotor core plate = 2

-38.3 2 )
|(61.7
= 1840 sq.cm.;
Area 70 slots = 70 X X
5.4 0.48 = 180 sq.cm. ;

Net area rotor core plates = 1660 sq.cm.;

A w = 27.9 cm.;
Volume of sheet steel in rotor core = 1660X27.9
= 46 300 cu.cm.;
Weight of 1 cu.cm. of sheet steel =7.8 grams;
46300X7.8
Weight of rotor core

= 360 kg.;
Total weight sheet steel (stator
,
603 +360 = 963 kg.
plus rotor)
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 185

Thus we have:

Weight copper = 301 kg.;

Weight sheet steel = 963kg.

Total net weight effec. material = 1264 kg.

From experience in the design of motors of this type it has been

determined that an adequate mechanical design is consistent with
" " "
obtaining for the ratios of Total Weight of Motor to Weight
of Effective Material," the values given in Table 34.

TABLE 34. DATA FOR ESTIMATING THE TOTAL WEIGHT OF AN INDUCTION

MOTOR.

D
Diameter
at Air-gap
(in cm.).
186 POLYPHASE GENERATORS AND MOTORS

Consequently we have :

Watts per ton =

=ry|^ 5900.
The designer must gain his own experience as to the attainable
"
values for the watts per ton." It may, however, here be stated
that 8000 watts per ton and even considerably higher values are

FIG 96. An 8-pole, 50-H.P., 900-r.p.m., Squirrel-cage Induction Motor of

the Open-protected Type. [Built by the General Electric Co. of America.!

quite consistent with moderate temperature rise for a motor of

this size and speed, if constructed in the manner generally
"
designated the open-protected type." In Fig. 96 is given a
photograph of an open-protected type of squirrel-cage induction
motor.
The Breakdown Factor. It has been stated that so far as
temperature-rise is concerned, the motor is rather liberally
designed and that in a revised design it should be possible to save
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 187

a little in material. We must, however, take into account the

limitation imposed by the heaviest load which the motor is
capable of carrying before it pulls up and comes to rest. The
ratio of this load to the rated load is termed the breakdown
factor (bdf.) and the present custom is to require that a motor
shall have a breakdown factor of about 2. In the case of our
200-h.p. motor, ought to be capable of carrying an instantaneous
it

load of 2X200 = 400

h.p. without pulling up and coming to rest.
It has previously been explained that the vertical projection of
the vector representing the stator current in the circle diagram,
is, for constant terminal pressure, proportional to the watts input

to the motor. Thus the current corresponding to the maximum

input is the current AM in the diagram in Fig. 97. The power
input corresponding to this current may be represented by MN,
the vertical radius of the circle. We have:

= VAN2 +MN2

= V333 2 +3082
= V206000
= 454 amperes.

Since the phase pressure is 577 volts, we have:

Power input = 3 X 577 X 308

= 534000 watts.

Consequently when used as a measure of the power input,

the vertical ordinates are to the scale of:

3X577 = 1730 watts per ampere.

But it is not the input which we wish, but the output. The
PR losses at the full-load input of 102 amperes, amount to:

4540+3140 = 7680 watts.

188 POLYPHASE GENERATORS AND MOTORS

Consequently when the input is 454 amperes, the I 2 R losses

are:

~ X 7680 = 152 000 watts.

The core losses and friction aggregate:

2410+240+1300 = 3950 watts.

240

220

.180
160
ft
140

100

W 80

60

40

20

A
20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 380 400
Wattless Component of Current

FIG. 97. Diagram for Preliminary Consideration of the Breakdown Factor.

The total internal losses amount to:

152 000+4000 = 156 000 watts.

Thus the output corresponds to that portion of the vertical

radius MN, which is equal to :

534 000 - 156 000 = 378 000 watts.

This is the vertical height, PN, laid off equal to :

-- =
378 000
218 amperes.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 189

This corresponds to an output of:

378 000
= 505h.p.
746

This vertical radius does not, however, correspond to the point

of maximum output. Let us draw, as in Fig. 98, the vectors
corresponding to:

430, 400 and 370 amperes.

.320

300

^240
2220

|200
"olSO

160
I
0,140
g
120

60

20

20 40 60 SO 100 120 140 160 ISO 200 220 240 260 280 300 -SO 340 360 380 400-420440
Wattless Component of Current

FIG. 98. Revised Diagram for Determining the Breakdown Factor.

The corresponding vertical ordinates are:

307, 300 and 290 amperes.

The inputs are:

1730X307 = 531 000 watts,

1730X300 = 520 000 watts,
1730X290 = 502 000 watts.
190 POLYPHASE GENERATORS AND MOTORS
The / 2 # losses are:

2
X7680 = 137 00 watts >

2
X 7680 = 118 000 watts,
Y~j
(4on\

= 101 000 watts.

Adding the remaining losses of 4000 watts, we obtain for the

total losses in the three cases:

141 000, 122 000, and 105 000 watts.

Deducting these losses from the respective inputs, we obtain as

the three values for the output:

531 000-141 000 = 390 000 watts,

520 000-122 000 = 398 000 watts,
502 000 - 105 000 = 397 000 watts.

Obviously the output is a maximum at an input of 400 amperes

and then amounts to:

398 000

Thus for our motor the bd.f. is:

S
A rough, empirical formula for obtaining the breakdown
factor is:

* -f.
For our motor we have :

y = 0.248;
0.041.

0.4X0.24?
.'.
bdf= 0.041
= 2.42.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 191

The use of this empirical formula is, in general, quite sufficient,

as it is rarely of consequence to be able to estimate the bdf. at
all closely.

THE POWER-FACTOR

It is a very simple matter to obtain a curve of power-factors

for various values of the current input. In Fig. 99 are drawn

320

240

220

200

180

160

140

120

100

80

60

40

20

Wattless Component of the Current

FIG. 99. Diagram for Determining the Power-factors.

vectors representing stator currents ranging from the no-load

current of 25.2 amperes, up to the break-down current of 400
amperes. In each case, the power-factor is the ratio of the
vertical projection of the stator current, to the stator current.
The values of the currents and of the vertical projections are
recorded in Fig. 99, and also in the first two columns of the
following table in which the estimation of the power-factor is
carried out, and also the estimation of the efficiencies.
192 POLYPHASE GENERATORS AND MOTORS

I.
POLYPHASE MOTOR WITH SQUIRREL-CAGE ROTOR 193

variation is too small to be of consequences at ordinary loads and

may for most purposes be neglected. For constant apparent
resistance, percentage slip the percentage which the rotor
is

PR loss constitutes of the input to the rotor.

The Determination of the Rotor I 2 R Loss at Various Loads.
By constructions similar to that illustrated in Fig. 94 on p. 179,
we may obtain the values of the rotor current for various values
of the stator current. A number of such values is tabulated
below:

Stator Current.
194 POLYPHASE GENERATORS AND MOTORS
to the current in the face conductors, the squirrel cage may be
" "
considered as having an equivalent resistance of:

3140
= 0.00324 ohm.
985 2
In the following table, the rotor I 2 R losses, the input to the
rotor, the slip and the speed have been worked out:

6
 CHAPTER V
SLIP-RING INDUCTION MOTORS

PROBABLY considerably over 90 per cent of the total number

of induction motors manufactured per annum are nowadays
of the squirrel-cage type. The strong preference for this type
is on account of the exceeding simplicity of its construction and

the absence of moving contacts. Nevertheless occasions arise

when it is necessary to supply definite windings on the rotor
and to connect these up to slip rings. Sometimes this is done
in order to control the starting torque and to obtain a higher

starting torque for a given stator current than could be obtained

with a squirrel-cage motor and sometimes the use of slip rings
isfor the purpose of providing speed control by regulating external
resistances connected in series with the rotor windings. This
isa very inefficient method of providing speed control, but cases
occasionally arise when it is the economically correct method to
employ.
Since the controlling resistance is located external to the motor,
there is no longer occasion to study to provide a deep rotor con-

ductor in order to obtain a desired amount of starting torque.

With freedom from this restriction, the density at the root of the
rotor tooth (for a design of the rating we have discussed
would
in Chapter IV) be taken higher than the 15 500 lines per sq. cm.
adopted for the squirrel-cage design. We shall do well to employ
a somewhat shallower and wider slot, for it must be observed that
it now becomes necessary to provide space for insulating the rotor
conductors.
Let us employ 70 slots, as before, but let the slots have a depth
of only 30 mm. and a (punched) width of 8 mm. The diameter
at the bottom of the slots will now be:

617.4-2X30 = 557.4 mm.

195
196 POLYPHASE GENERATORS AND MOTORS
The rotor slot pitch is now:

557.4 XTC OK n
=25.0 mm.
-njQ

Width tooth at root = 25.0 -8.0 = 17.0 mm.

For the squirrel-cage design, the slot pitch (see p. 888) was
22.8 mm. and the width of tooth at the root, was:

22.8 -4.8 = 18.0 mm.

Thus the tooth density be increased in the ratio of 17 to

will

18, but since the length of the tooth has been decreased in the
ratio of 54 to 30, the mmf required for the rotor teeth will be no
.

greater than for the squirrel-cage design.

The thickness of the slot insulation will be 0.8 mm.
Allowing 0.3 mm. tolerance in assembling the punchings, the
assembled width of the slot will be:

8.0-0.3 = 7.7 mm.

After deducting the portions of the width occupied by the slot

insulation on each side, we arrive at the value of :

7.7-2X0.8 = 6.1 mm.

for the width of the conductors, of which there will be 2 per slot,
arranged one above the other. Of the depth of 30 mm., the
insulation and the wedge at the top of the slot will require a total
allowance of 6 mm. Consequently for the depth of each conduc-
tor we have:

30-6 = \
)
,
12 mm.

Thus each conductor has a height of 12 mm., a width of 6.1

mm. and a cross-section of :

12X6.1=73 sq.mm.
 SLIP-RING INDUCTION MOTORS 197

A drawing of the given in Fig. 106.

slot is

A
two-layer, full-pitch, lap winding will be employed. There
are 70 slots and (2X70 ) = 140 conductors. The winding will
be of the 2-circuit type. It is impracti-
cable in this treatise to deal with the
extensive subject of armature windings.
The reader will find a discussion of the
laws of 2-circuit windings on page 3 of
"
the Author's Elementary Principles of
"
Continuous-Current Dynamo Design
(Whittaker <fc Co., London and New
York), and also in Chapter VIII of Hobart
" "
and Ellis' Armature Construction
(Whittaker & Co.). It must suffice here
to that such a winding is usually
state

arranged to comply with the formula:

FIG.
F = Py2. "We

In this formula, the total number of conductors

is denoted by
" "
F, the number by P, and the
of poles winding pitch by y.
The conductors are numbered from 1 to 140 as indicated in the
diagram in Fig. 107. The winding pitch, y is that quantity t

which is added to the number of any conductor in order to ascer-

tain the number of the conductor to which the first conductor
is connected.
For our case we have:
198 POLYPHASE GENERATORS AND MOTORS
cannot contain equal numbers of turns. The nearest to this
consists in subdividing the windingup two
circuits each
into
comprising 22 conductors and four circuits each comprising 24
conductors. We have

22+22+24+24+24+24 = 140.

FIG. 107. Winding Diagram for the Rotor of a Slip-ring Induction Motor.

Let us start in with conductor number 1 and designate this

beginning by A\, as shown in Fig. 107. After following through
the 22 conductors:

(1, 24, 47, 70, 93, 116, 139, 22, 45, 68, 91, 114, 137,
20, 43, 66, 89, 112, 135, 18, 41, 64),
 SLIP-RING INDUCTION MOTORS 199

we interrupt the winding and bring out a lead at the point Am.
We then start in again, indicating the point as B\ and proceed
next through conductor number (64+23 = ) 87. After passing
through 22 more conductors we come out again at a point which
we designate as Bm. 'the next circuit starts at C\ and ends at
Cm. The remaining three circuits are A^A n B^B n and
, C<iC n -

We now have six circuits. An inspection will show that these

can be grouped in three pairs which have, in Fig. 107 been
drawn in red, black and green. The red phase comprises A\A m
connected in series with AnA% forming A\A^. In the same way,
the black phase B\B^ is obtained, and the green phase CiCz. By
considering the instant when the current in Phase A is of the value
1.00 and is flowing from the line toward the common connection,
while the currents in Phases B and C are of the value 0.50 and
are flowing from the common connection to the lines, it is ascer-
taineol that theends A%, B\ and Cz must be brought together to
form the common connection, the ends A\,B% and Ci constituting
the terminals of the motor.
For the mean length of turn of this rotor winding we have the
formula :

mlt.~2XH-2.5Tj

T = 32.5;
2.5T = 82;
mlt. = 86+82 = 168 cm.

The entire winding comprises (70X2 = ) 140 conductors and

70 turns. Consequently the average length per phase is:

70X168
^
o
- =3920 cm.

Resis. per phase (at 60 Cent.)

x N
= 3920X0.00000200
- zr^ = 0.0107 ohm.
U. /o

Since there are 720 stator conductors, the ratio of transforma-

tion is now :

^-1 5. 15 to 1.00.
200 POLYPHASE GENERATORS AND MOTORS

Referring to p. 165, we see that for a 1 1 ratio, the rotor :

current corresponding to the full-load stator current of 102

amperes, is 96 amperes. Consequently the full-load rotor current
for our wound rotor, is:

5.15X96 = 495 amperes.

The full load PR loss in the rotor is:

3X4952 X0.0107 = 7800 watts.

/7800 \
This is
(5777:= )2.5 times the loss in our squirrel-cage design.
\O-L4:U J
The increase due partly to the waste of space for slot insulation
is

and partly to the long end connections inevitably associated with

" "
a wound rotor.
These long end connections also involve additional magnetic
"
leakage at load, and the circle ratio of a motor with a wound"
rotor may usually be estimated on the basis of a 25 per cent
increase in the values in Table 27, on pp. 150 and 151. Conse-
quently for the present design we have :

<j= 1.25X0.041 =0.051.

We still, have (as on p. 140), y

= 0.242.
Hence for the circle diameter we now have:

TV
Dia. oft
-
i
circle = 0.242X102
^r^\
= 485 amperes
u.uoi

" "
The ideal short-circuit current is now:

0.242X102+485 = 510 amperes.

as against the value of 645 amperes for the squirrel-cage design.

We now have :
 SLIP-RING INDUCTION MOTORS 201

From an examination of these various ways in which the

properties of the motor have sufferedby the substitution of a
wound rotor, it will be understood that the squirrel-cage type
ismuch to be preferred, and that every effort should be made to
employ it when the conditions permit. In addition to the
defects noted, it must not be overlooked that the efficiency and
power-factor of the slip-ring motor are lower and the heating
greater. Furthermore the squirrel-cage motor is more compact
and is and cheaper. It will also be subject to decidedly
lighter
less depreciation, and can be operated under conditions of
exposure which would be too severe for a slip-ring motor.
 CHAPTER VI

SYNCHRONOUS MOTORS VERSUS INDUCTION MOTORS

THE circumstance that a type of apparatus, possesses particu-

larly attractive features, is liable to lead occasionally to
disappointment, owing to its use under conditions for which it is
rendered inappropriate by its possession of other less-well-
known features, which, under the conditions in question, are
undesirable.
While the squirrel-cage induction motor is of almost ideal
simplicity, there are many instances in which its employment
would involve paying dearly for this attribute. It is, for instance,
a very poor and expensive motor for low speeds. When the
rated speed is low, and particularly when, at the same time, the
periodicity is relatively high, the squirrel-cage motor will
inevitably have a low power-factor. In Figs. 108 and 109 are
drawn curves which give a rough indication of the way in which
the power-factor of the polyphase induction motor varies with
the speed for which the motor is designed. While qualitatively
identical conclusions will be reached by an examination of the
data of the product of any large manufacturer, the quantitative
values may be materially different, since each manufacturer's
product is characterized by variations in the degree to which
good properties in various respects are sacrificed in the effort
to arrive at the best all-around result. It is for this reason that
instead of employing data of the designs of any particular manu-
facturer, the curves in Figs. 108 and 1C 9 have been deduced from
the results of an investigation published by Dr. Ing. Rudolf
Goldschmidt, of the Darmstadt Technische Hochschule, in an
excellent littlevolume which he has published under the title:
" "
Die normalen Eigenschaften elektrischer Maschinen (Julius
Springer, Berlin). Whereas from the 50-cycle curves of Fig. 109,
it will be seen that, at low speeds, motors of from 50 to 500 h.p.

202
 SYNCHRONOUS MOTORS vs. INDUCTION MOTORS 203

rated output, have power-factors of less than 0.80, the curves

in Fig. 108 show that the power-factors of equivalent 25-cycle
motors are distinctly higher.
It is desirable again to emphasize that the precise values
employed in the curves in Figs. 108 and 109, have, considered
individually, no binding significance. Thus, by sacrificing
desirable features in other directions and by increased outlay
in the construction of the motor, slightly better power-factors

may sometimes be obtained. On the other hand, the power-

factors on which the curves are based, are, in most instances,
already representative of fairly extreme proportions in this respect;
and few manufacturers find it commercially practicable to list
low-speed motors with such high power-factors as are indicated
by these curves. The purchaser would rarely be willing to pay
a price which would leave any margin of profit were these power-
factors provided; and consequently it is only relatively to one
another that these curves are of interest. They teach the lesson
that for periodicities of 50 or 60 cycles, it must be carefully kept
in mind that, if low speed induction motors are used, either the

price paid must be disproportionately high, or else the purchaser

must be content with motors of exceedingly low power-factors.
The power-factor, furthermore, decreases rapidly for a given low
speed, with decreasing rated output.
On the other hand, for a 25-cycle supply, these considerations
are of decidedly less importance. With a clear recognition of
this state of affairs, a power user desiring low-speed motors
for a 60-cycle circuit will do well to take into careful considera-
tion the alternative of employing synchronous motors. If he
resorts to this alternative he can maintain his power-factor at

unity, irrespective of the rated speed of his motors; but he must

put up with the slight additional complication of providing for a
rotor which, in addition to the squirrel-cage winding, is also
equipped with field windings excited through brushes and slip-
rings from a source of continuous electricity. The precise
circumstances of any particular case will require to be considered,
in order to decide whether or not this alternative is preferable.
While it has long been recognized that the inherent simplicity
and robustness of the squirrel-cage induction motor constitute
features of very great importance and justify the wide use of
such motors, nevertheless we must not overlook the fact that
204 POLYPHASE GENERATORS AND MOTORS
induction motors are less satisfactory the lower the rated speed.
The chief disadvantage of a low-speed induction motor is its low
power-factor as shown above. Let us take the case of a 100-h.p.,
60-cycle motor. The design for a rated speed of 1800 r.p.m.
will (see Fig. 109) have a power-factor of 93 per cent, whereas
the design for a 100-h.p., 60-cycle motor for one-tenth of this
speed, for a speed of only 180 r.p.m., will have a power-
i.e-.,

factor only some 76 per cent. The power-factors given

of
above correspond to rated load. For light loads, the inferiority
of the low-speed motor as regards low power-factor is much
greater, and this circumstance further accentuates the objection
to the use of such a motor.

LOO

0.90

0.80

0.70

0.60
100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400^1500
Hated Speed in_R.P.M.

FIG. 108. Curves of Power-factors of 25-cycle Induction Motors of 5 H.P.,

50 H.P., and 500 H.P. Rated Output, Plotted with Rated Speeds as
Abscissae.

For a long time there has existed a general impression that a

synchronous motor could not be so constructed as to provide
much starting torque. Otherwise it would probably have been
realized that, for low speeds, it would often be desirable to give
the preference to the synchronous type. For with the synchro-
nous type, the field excitation can be so adjusted that the power-
factor shall be unity; indeed there is no objection to running
with over-excitation and reducing the power-factor again below
unity, thus occasioning a consumption of leading
current by the
synchronous motor.
By the judicious admixture (on a single supply system) of
 SYNCHRONOUS MOTORS vs. INDUCTION MOTORS 205

high-speed induction motors consuming a slightly-lagging current

and low-speed synchronous motors consuming a slightly leading
current, it is readily feasible to operate the system at practically
unity power-factor; and thereby to obtain, in the generating
station and on the transmission line, the advantages usually
accruing to operation under this condition.
Now the point which is beginning to be realized, and which
it is important at this juncture to emphasize, is that the synchro-

nous motor, instead of being of inferior capacity as regards the

provision of good starting torque, has, on the contrary, certain
inherent attributes rendering it entirely feasible to equip it for

much more liberal starting torque than can be provided by

0.60
100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500

Bated Speed in R. P.M.

FIG. 109. Curves of Power-factors of 50-cycle Induction Motors of 5 H.P.,

50 H.P., and 500 H.P. Rated Output, Plotted with Rated Speeds as
Abscissae.

" "
efficient squirrel-cage induction motors. The word efficient
has been emphasized in the preceding statement and for the
following reason: by supplying an induction motor with a
squirrel-cage system composed of conductors of sufficiently
small cross-section, and consequently sufficiently-high resistance,
any amount of starting torque which is likely to be required,
can be provided. Unfortunately, however, the high -resistance
squirrel inherently associated with a large
cage is loss in PR
the rotor when the motor is carrying its load. This large PR
" "
loss not only occasions great slip but also occasions very low
efficiency. Furthermore, owing to this very low efficiency,
206 POLYPHASE GENERATORS AND MOTORS
the heating of the motor would be very great were it not that
such a motor is rated down to a capacity far below the capacity
at which it could be rated were it supplied with a low resistance
(i.e., low starting torque) squirrel-cage system.
But when we turn to the consideration of the synchronous
motor, we note the fundamental difference, that the squirrel-cage
winding with which we provide the rotor, is only active during
starting, and during running up toward synchronous speed.
When the motor has run up as far toward synchronous speed
as can be brought about by the torque supplied by its squirrel cage,
excitation is applied to the field windings and the rotor pulls in
to synchronous speed. So soon as synchronism has been brought
about, the squirrel-cage system is relieved of all further duty;
and it is consequently immaterial whether it is designed for high
or for low resistance.
Consequently with the synchronous motor we are completely
freefrom the limitations which embarrass us in designing high-
torque induction motors. In the case of the synchronous motor,
we can provide any reasonable amount of torque by making the
squirrel-cage system of sufficiently high resistance. One difficulty,
however, presents itself: while, at the instant of starting, we may
desire very high torque and may provide it by a high-resistance
squirrel cage, the higher the resistance the more will the speed,
up to which the rotor will be brought by the torque of the squirrel

cage, fall short of synchronous speed. It would in fact, be desirable

that as the rotor acquires speed the squirrel cage should gradually
be transformed from one of high resistance to one of low resistance.
If we could accomplish the result that the squirrel cage should,
at the moment the motor starts from rest, have a high resistance,
and if it could be arranged that this resistance should gradually
die away to an exceedingly low resistance as the motor speeds up,
then the motor would gradually run up to practically synchronous
speed and would furnish ample torque throughout the range from
zero to synchronous speed.
We can provide precisely this arrangement if, instead of making
the end rings of the squirrel cage, of copper or brass or other non-
magnetic material, we employ instead, end rings of magnetic mate-
rial, such as wrought iron, mild steel, cast iron or some magnetic

alloy. Let us consider the reason why this arrangement should

produce the result indicated. Just before the motor starts, the
 SYNCHRONOUS MOTORS vs. INDUCTION MOTORS 207

current? induced in the squirrel-cage system are of the full perio-

dicity of the supply. In the end rings, these currents will, with
usual proportions, be very large in amount; and since the currents
are alternating and since the material of the end rings is magnetic,
there will be a very strong tendency, in virtue of the well-known
"
phenomenon generally described as skin effect," to confine the
current to the immediate neighborhood of the surface of the
end rings. The current will be unable to make use of the full
cross-section of the end rings; and consequently, even though
the end rings may be proportioned with very liberal cross-section,
the net result will, at starting, be the same as if the end rings were
of high resistance. But as the motor speeds up, the periodicity
of the currents in the squirrel cage decreases, until, at synchro-
"
nism, the periodicity would be zero and there would be no skin
effect." In view of these explanations, it is obvious that the
impedance of the end rings will gradually decrease from a high
value at starting to a low value at synchronism.
Now we are more free to make use of this phenomenon in the
case of synchronous motors than in the case of induction motors,
when the synchronous motor is run at full
for as previously stated,

speed, the squirrel cage is utterly inactive (except in serving to

" "
minimize surging and to decrease " ripple " losses); whereas
the induction motor's squirrel cage is always carrying alternating
current (even though of low periodicity); and this alternating
current flowing through end rings of magnetic material, occasions
a lower power-factor than would be the case with the equivalent
squirrel-cage motor with end rings of non-magnetic material.
Even in the case of induction motors, excellent use can be made
of constructions with end rings of magnetic material, in improving
the starting torque, but there is unavoidably at least a little sacri-

fice in power-factor during normal running.

It should now be clear that there is a legitimate and wide

field for low-speed synchronous motors and that these motors will

be superior to low-speed induction motors, in that, while the former

can be operated at unity power-factor, or even with leading cur-
rent if desired, the latter will unavoidably have very low power-
factors. Furthermore these low-speed synchronous motors,
instead of being in any way inferior to induction motors with
respect to starting torque, have attributes permitting of providing
them with higher starting torque than can be provided with
208 POLYPHASE GENERATORS AND MOTORS
induction motors, without impairing other desirable character-
isticssuch as low heating and high efficiency.
Of course there always remains the disadvantage of requiring
a supply of continuous electricity for the excitation of the field
magnets. Cases will arise where this disadvantage is sufficient
to render it preferable, even for low-speed work, to employ

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Ratio of Apparent to True Resistance

FIG. 110. Robinson's Curves for Skin Effect in Machine-steel Bar with a
Cross-section of 25X25 mm.

induction motors, but in the majority of cases where polyphase

motors must operate at very low speeds, it would appear that
synchronous motors are preferable.
In the curves in Figs. 110 and 111 are plotted the results of
some interesting tests which have recently been made by Mr.
" "
L. T. Robinson, on skin effect in machine steel bars. By
 SYNCHRONOUS MOTORS vs. INDUCTION MOTORS 209

means of these and similar data, and by applying to the design

of the synchronous motor the ample experience which has been

acquired in the design of induction motors, the preparation of

a design for a synchronous motor for stipulated characteristics
as regards torque, presents no difficulties. The smooth-core
type of field with distributed excitation is to be preferred to the
salient-pole type.
We have considered the relative inappropriateness of the
induction motor for low-speed applications. Conversely it is
a particularly excellent machine for high speeds. Its power-

10 20 50 60 70 80 90 - 100

Periodicity in Cycles per Second

FIG. 111. Robinson's Curves for Skin Effect in Machine-steel Bar with a
Cross-section of 25X25 mm.

factor is higher the higher the rated speed; and when we come
to very high speeds we
obtain, in motors of large capacity, full-
load power-factors in excess of 95 per cent. In such instances
the simplicity of squirrel-cage induction motors should frequently
lead to their use in preference to synchronous motors.
Reasoning along similar lines, in the case of generators, the
induction type offers advantages over the synchronous type in
many instances. It cannot, however, replace the synchronous
generator even at high speeds, for as explained in Chapter VII, it
requires to be run in parallel with synchronous generators, the
210 POLYPHASE GENERATORS AND MOTORS

latter supplying the magnetization for the induction generators

and also supplying the lagging component of the external load when
the latter 's power-factor is less than unity. Notwithstanding
these limitations, there is a wide field for the induction generator;
and the above indications may be useful for guidance in showing
its appropriateness in any specific instance.

now be agreed that the properties of synchronous motors

It will
lend themselves admirably to the provision of high starting torque.
In addition to the means previously described whereby a synchro-
nous motor not only high starting torque but may also
may have
automatically run close up to synchronous speed, so as to fall

quietly into synchronism immediately upon the application of

the field excitation from the continuous-electricity source, there
is also available a phenomenon described by Mr. A. B. Field, in a
"
paper entitled Eddy Currents in Large Slot- Wound Con-
ductors/' presented in June, 1905, before the American Institute
of Electrical Engineers (Vol. 24, p. 761).
Mr. A. B. Field analyzed the manner in which the apparent
resistance of slot-embedded conductors varies with the periodicity.
Subsequent investigations show that the Field phenomenon,
while harmful in stator windings exposed constantly to the full
periodicity of the supply system, may be employed to consider-
able advantage in the proportioning of the conductors of the slot
portion of a squirrel-cage system. Both in synchronous and in
induction motors, this is an important step. Take, for instance,
the case of a 60-cycle induction motor. At starting, the periodicity
of the currents in the squirrel-cage system is 60 cycles, and the
Field effect may, with properly-proportioned conductors, be
sufficient to occasion an apparent resistance very much greater
than the true resistance. The motor thus starts with very much
more torque per ampere, than were the Field phenomenon absent.
As the motor speeds up, the periodicity decreases and the Field
effect dies out, the apparent resistance of the squirrel-cage grad-
ually dying down to the true resistance. Thus, whereas the full-
load running slip of an ordinary squirrel-cage motor must be high,
its heating high and its efficiency low, if it is to develop high start-
ing torque, we may, by using the Field effect, construct high-start-

ing torque motors with low slip and high efficiency.

In Fig. 112 are shown respectively the rough characteristic
shape of a curve in which torque is plotted as a function of
 SYNCHRONOUS MOTORS vs. INDUCTION MOTORS 211

the speed for :

A, a permanently low-resistance squirrel-cage
motor; B, a permanently high-resistance squirrel-cage motor;
and, C, for a squirrel-cage motor in which the apparent resist-
ance gradually changes from a high to a low value as the motor
runs up from rest to synchronism. As regards the application
of these curves to a synchronous motor, it will be seen that, in the
first case, (Curve A), the starting torque is rather low; but that

the torque increases, passes through a maximum, and falls slowly,

remaining quite high until the speed is close to synchronism.
This motor, while unsatisfactory at starting, has the property

10 20 30 40 50 60 70 80 90 100
Speed in Percent of Synchronous Speed

FIG. 112. Curves Contrasting Three Alternative Squirrel-cage Constructions

for a Synchronous Motor.

of pulling easily into synchronism on the application of the

excitation fr.m the continuous-electricity source. In the second
case, (Curve J5), while the synchronous motor starts with high

torque, the torque falls away much more rapidly; and when the
torque has fallen to the value necessary to overcome the friction
of the motor, the speed is several per cent below synchronous

speed, and the application of the continuous excitation, if it

suffices to pull the rotor into synchronism, will do so only at the
cost of an abrupt and considerable instantaneous drain of power
from the line. In the third case, (Curve C), there are present
212 POLYPHASE GENERATORS AND MOTORS
the good attributes of the first two cases, the bad attributes
being completely eliminated.
It is thought that with the explanations furnished in this

Chapter, the user will be assisted in determining, in any particular

case, whether it is more desirable to take advantage of the
extreme simplicity and toughness of the low-speed squirrel-cage
induction motor, notwithstanding its poor power-factor, or
whether he should employ the slightly more complicated syn-
chronous motor in order to have the advantage of high
power-factor.
Another way of dealing with the situation, which in certain
cases of a low-speed drive is preferable, is to employ a high-
speed induction motor (which will consequently have a high
down to the low-speed load. Thus,
power-factor), and to gear it
from the curves in Fig. 109, we see that, if we require to drive
a load at 200 r.p.m., a 50-h.p. motor will have a power-factor of
only about 0.70 whereas, if the motor were to drive the load
;

through 5 to 1 gearing, the motor's own speed would be 1000

r.p.m.,and its power-factor would be over 0.90. Not only would
the induction motor be characterized by a 20 per cent higher
power-factor at full load, but would be higher and it
its efficiency

would be much and cheaper.

smaller, lighter At half load the
advantage of the high-speed motor in respect to power-factor is
still more striking, the two values being of the following order:

Rated Speed.
 CHAPTER VII

THE INDUCTION GENERATOR

FOR
the very high speeds necessary in order to obtain the best
economy from steam turbines, the design of synchronous genera-
tors is attended with grave
difficulties. The two leading difficul-
obtaining a sound mechanical construction at
ties relate, first to
these high speeds, and, secondly, to the provision of a field winding
which shall operate at a permissibly low temperature.
It is very difficult to provide adequate ventilation for a rotating
field magnet of small diameter and great length. The result
has been that for so high a speed as 3600 r.p.m., the largest
synchronous generators which have yet given thoroughly satis-
factory results are only capable of a sustained output of some
5000 kva. Even at this output the design has very undesirable
proportions and the temperature attained by the field winding
is undesirably high.
But an Induction generator is exempt from the worst of these
difficulties. The conducting system carried by its rotor consists in
a simple squirrel cage, the most rugged construction conceivable.
Thus from the mechanical standpoint the induction generator
is an excellent machine. Furthermore, the squirrel cage can be
so proportioned that the full-load PR loss is exceedingly low.
Hence the rotor will run cool, and is, in this respect, in striking
contrast with the rotor of a high-speed synchronous generator.
Notwithstanding these satisfactory attributes, induction gen-
erators are rarely employed. The chief obstacle to their extensive
use relates to the limitation that they must be operated in parallel
with synchronous apparatus. Induction generators are dependent
for their excitation upon lagging current drawn from synchronous
generators, or leading current delivered to synchronous motors
connected to the network into which the induction generators
deliver their electricity.
213
214 POLYPHASE GENERATORS AND MOTORS
The practical aspects of the theory of the induction generator
have been considered in a paper (entitled " The Squirrel-cage
Induction Generator ") written in collaboration by Mr. Edgar
Knowlton and the present author. This paper was presented
on June 28, 1912, at the 29th Annual Convention of the American
Institute of Electrical Engineers at Boston. It would not
be appropriate to reproduce the descriptions and explanations
in that paper, since the purpose of the present treatise is to set
forth the fundamental methods of procedure employed in design-

ing machinery. Some brief extracts from the paper in question

are given in a later part of this chapter, and the reader will find
it profitable to consult the original paper in amplifying his
knowledge of the practical aspects of the theory of the induction
generator.
Although the induction generator is chiefly suitable for large
outputs, we can nevertheless, in explaining designing methods,
employ as an illustrative example the case of adapting to the
purposes of an induction generator, the 200-h.p. squirrel-cage
induction motor which we worked out in Chapter IV.
The full-load efficiency has (see p. 178) been ascertained to be
93.0 per cent. Consequently at full load the input is ;

200X746

Let us, in the first instance, assume that the machine is suitable
for a rated output of 160 kw. when operated as an induction
generator. Since, when employed for this purpose, there is no
longer any necessity for taking into consideration any questions
relating to starting torque, let us reduce the rotor loss by PR
widening the rotor face conductors. In the original design of
the induction motor the slip corresponding to 200 h.p. was 2.0
per cent. This slip corresponds to a rotor PR
loss of 3140 watts.
This loss is made up two components, which are 2512 watts
of
in the face conductors and 628 watts in the end rings.
The tooth density in the rotor is needlessly low, other con-
siderations, not entering into induction generator design having
determined the width of the slot. There is now nothing to pre-
vent doubling the width of the rotor conductors. This gives
 THE INDUCTION GENERATOR 215

for the dimensions of these conductors a depth of 54 mm. and

a width of 9.6 mm.
This alteration will reduce the rotor loss at 160 kw. output to

+628 = 1884 watts.

The slip will now be:

= 1.2 per cent.

1^X2.0

In the design of an induction generator, the slip should be

made as small as practicable, in order to have a minimum rotor
loss and consequently the highest practicable efficiency and low

heating. In large induction generators there is rarely any

difficulty in bringing the slip down to a small fraction of 1 per cent.
The slip of the 7500-kw. 750-r.p.m. induction generators in the
Interborough Rapid Transit Co.'s 59th Street Electricity Supply
Station in New York, is only about three-tenths of 1 per cent at
rated load.
Since we have reduced the squirrel-cage loss by :

(2.0-1.2 = )0.8

per cent, the efficiency of our 160-kw. induction generator will be

(93.0+0.8 = )93.8 per cent.

It is not this small increase in efficiency which is particularly

to be desired, but the decreased heating. The total losses are

decreased from :

(100.0-93.0 = )7.0 per cent of the input

to

(100.0-93.8 = )6.2 per cent of the input.

216 POLYPHASE GENERATORS AND MOTORS

If, for a rough consideration of the case, we take the heating

to be proportional to the total loss, this result would justify us
in giving careful consideration to the question of the feasibility
of rating up the machine in about the ratio of :

6.2 : 7.0.

This would bring the rated output up to

But before finally determining upon increasing the rating,

it would be necessary to examine into the question of the heating
of the individual parts, since in decreasing the rotor heating it
does not necessarily follow that it is expedient to increase the
stator heating to the extent of the amount of the decreased loss
in the rotor. On
the other hand, in the case of very high speed
generators, the heating of the rotor conductors will, practically,
always constitute the limit, owing to the difficulty of circulating
air through a rotor of small diameter and great length. Con-
sequently in the case of very high speed generators the increased
rating rendered practicable by the substitution of a low-loss
squirrel-cage rotor for a rotor excited with continuous electricity,
will often be much greater than in the inverse ratio of the respec-
tive total losses for the two cases.
The
author's present object is, however, to point out that the
induction generator has inherent characteristics usually permit-
ting of assigning to it a materially higher rating than that which
is appropriate when the same frame is employed in the con-
struction of an induction motor.
Since the slip is 1.2 per cent, the full-load speed is

(1.012X500 = ) 506r.p.m.

The Derivation of a Design for an Induction Generator from

a Design for a Synchronous Generator. Let us now evolve an
induction generator from the 2500-kva. synchronous generator for
which the calculations have been carried through in Chapter II.
 THE INDUCTION GENERATOR 217

This machine was designed for the supply of 25-cycle electricity.

It had 8 poles and operated at a

speed of 375 r.p.m. Let us first

plan not to alter the stator in any
respect except to employ a nearly-
closed slot of the dimensions indi-
cated in Fig. 113. This alteration
from the slot proportions employed in
the synchronous generator is necessary
in order to avoid parasitic losses
when the machine is loaded. The
necessity arises from the circumstance
that, unlike a synchronous genera-
tor, an induction generator must be
designed with a very small air-gap.
Otherwise it would have an unde-
sirably-low power-factor and would FIG. 113.
H
Stator Slot for
require the supply of too considerable a Induction Generator.
magnetizing current from the syn-
chronous apparatus with which it operates in parallel. In the case
with which we are dealing, we may employ an air-gap depth of
only 2 mm. Thus we have:
A -0.20.
The stator has 120 slots.
Let us supply the rotor with
106 slots. Each rotor slot may
be made 50 mm. deep and 25
mm. wide. The slot is closed,
as shown in Fig.114, by a solid
steel wedge with a depth of 10
mm. The conductor is unin-
-25-mm- sulated and is 40 mm. deep by
25 mm. wide, its cross-section
FIG. 114. Slot for Rotor of Induction
thus being:
Generator.
4X2.5 = 10sq.cm.
Since there are (120X10 = ) 1200 stator conductors and
(106X1 = )106 rotor conductors, the ratio of transformation is:

1200
= 11.3.
106
218 POLYPHASE GENERATORS AND MOTORS
The energy component of the full-load current in the stator

winding is Neglecting the magnetizing component

120 amperes.
of the stator current, we may obtain a rough, but sufficient,

approximation to the value of the full-load current per rotor

conductor. This is :

(11. 3X120
= ) 1360 amperes.

The current density in the rotor conductors is thus:

1360
= 136 amp. per sq.cm.

The gross core length is:

g
= cm.

Allowing 8 cm. for the projections at each end, the total length
of each rotor conductor is

(118+2X8 = )134cm.

The aggregate length of the 106 rotor conductors is

(106X134 = ) 14 200 cm.

The corresponding resistance, at 60 Cent, is:

The PR loss in the rotor face conductors at full load is

1360 2 X 0.00284 = 5250 watts.

At full load, the current in each end ring is (see p.

X 1360 = 5700 amp.

 THE INDUCTION GENERATOR 219

Let us give each end ring a cross-section of 40 sq.cm.

The mean diameter of an end ring must be a little less than

D, i.e., a little less than 178 cm. Let us take the mean diameter
of the end ring as 165 m. (

The resistance of a conductor equal to the aggregate of the

developed length of the two end rings is, at 60 Cent. :

The full-load loss in the two end rings is :

5700 2 X 0.000052 = 1680 watts.

The total loss in the squirrel cage, at full load is :

5250+1680 = 6930 watts.

If our induction generator were to be rated at 2500 kw., this

rotor loss would be :

6930X100 ftOQ
=0 28perCent
' -

2100^00

We have seen (p. 90) that the loss in the rotor of our 2500
kva. synchronous generator is (at a power-factor of 0.90 and
consequently an output of 2250 kw.) 15 500 watts.
Thus the efficiency is considerably greater in the case of the
induction generator rating of 2500 kw. Indeed, since the hottest
part of the synchronous generator is its rotor, we can easily rate
up the machine, when re-modelled as an induction generator,
to 3000 kw, the slip then being:

=00.33 per cent.

is now so small as to be of but little

Since, however, the air-gap
service as a channel through which to circulate cooling air, the
preferable design for the induction generator would consist in
a modification in which, instead of employing numerous vertical
220 POLYPHASE GENERATORS AND MOTORS

ventilating ducts, the airis circulated through 120 longitudinal

channels, one just below each stator slot. Two other methods
of air circulation which have been employed on the Continent

Fio. 115. A Method of Ventilation Suitable for an Induction Generator.

of Europe for synchronous generators and are especially appro-

priate for induction generators, are shown in Figs. 115 and 116.
In that indicated in Fig. 115, the air from the fans on the ends of
the rotor is passed to a chamber at the external surface of the
 THE INDUCTION GENERATOR 221

armature core. This chamber opens into air ducts in a plane

at right angles to the shaft. Suitably shaped space blocks lead

the air, in a tangential direction, to axial ducts just back of the
stator slots. The air then flows axially through one section, to
the next air duct, and then outwardly in a tangential direction
to a chamber at the outer surface of the core. This chamber is
adjacent to the one first mentioned and leads to the exit from
the stator frame. Looking along the axis of the shaft, the air
flows in a V-shaped path, the axial duct back of the stator slots

FIG. 116. An Alternative Method of Ventilation Suitable for an Induction

Generator.

being at the apex. Thus the two legs of the V are separated
axially by a single armature section.
The other method (Fig. 116) which is also independent of the
radial depth of the air-gap, consists in dividing the stator frame
into cylindrical chambers placed side by side. The air is forced
into a chamber from which it first passes radially toward the shaft,
then axially to adjacent air ducts, and finally outwardly to a
chamber alongside the one first mentioned. This last chamber
communicates with the outer air.
222 POLYPHASE GENERATORS AND MOTORS
Since the loss in the squirrel cage is 0.33 of 1 per cent, in the
case of this induction generator, the speed at rated load will be
1.0033X375 = 376.2 r.p.m.
For the 2-mm. air-gap which we are now employing, the
mmf. calculations for the phase pressure of 6950 volts, may
(without attempting to arrive at a needlessly exact result) be
estimated as follows:

Air-gap 1300 ats.

Teeth 700 "
Stator core 300 "
Rotor core. . 300 "

Total mmf. per pole ............. =2600 ats.

Thus each phase must supply 1300 ats. There are 25 turns
per pole per phase. Thus the no-load magnetizing current is:

1300
= 37 amperes.

Without going into the estimation of the circle factor it is

evident that the wattless component for full load of 3000. kw.
will be a matter of some 45 amperes.
The current output at full load of 3000 kw. is:

3 000 000
= 144 amperes.

The total current in each stator winding) at full load is:

\/45 2 +144 2 = 151 amperes.

The power-factor at full load is:

144

It must be remembered that this design is for the moderate

speed of 375 r.p.m. For high-speed designs (say 1500 r.p.m.
 THE INDUCTION GENERATOR 223

at 25 cycles) the power-factor, in large sizes, may readily be

brought up to 0.97.
In the paper to which reference has already been made
"
(Hobart and Knowlton's The Squirrel-cage Induction Genera-
tor "), mention is made of a comparative study which has been
carried out for two 60-cycle, 3600 r.p.m. designs, one for a 2500-kw.
synchronous generator and the other for a 2500-kw. induction
generator, both for supplying a system at unity power-factor.
The leading data of these two designs were as follows :
224 POLYPHASE GENERATORS AND MOTORS
it with a slip at rated load, of, say, two-tenths of 1 per cent. This
corresponds to a rotor periodicity of only (0.002X60 = )0.12
of a cycle per second, or 7.2 cycles per minute, and under such
conditions there is no necessity for employing a laminated core.
On account of the small air-gap of induction generators, the
value of the critical speed of vibration is especially important.
If possible, the critical speed should be at least 10 per cent
above normal. If other important reasons require employing a

critical speed below normal, it should be considerably below,

care being taken that the second critical speed is also removed
from the normal, preferably above it. With such a design the
rotor should have a very careful running balance before it is
placed in the machine. A damping bearing could be used to
prevent the rubbing of the rotor and stator if for any reason the
machine should be subjected to abnormal vibration.
It should be noted that in cases where the critical speed must
be below the normal speed, the air-gap cannot be so small as
would be preferred from the standpoint of minimizing the magnet-
izing current. A consideration tending to the use of a shaft with
a critical speed below the normal running speed relates to the
lower peripheral speed thereby obtained at the bearings.
 CHAPTER VIII

EXAMPLES FOR PRACTICE IN DESIGNING POLYPHASE

GENERATORS AND MOTORS

IN connection with courses of lectures on the subject-matter

of this treatise, the author hashad occasion to set a number
of examination papers. In the present chapter some of these
examination papers are reproduced and it is believed that they
will be of service in acquiring ability to apply the designing
principles discussed in the course of the preceding chapters.

PAPER NUMBER I

1. The leading data of a 12-pole, 250-r.p.m., 25-cycle, 11,000-

volt, 3000-kva., Y-connected, three-phase generator are as follows:

Output in kilovolt-amperes 3000

Number of poles 12
Terminal pressure 11 000 volts

Style of connection of stator windings . . Y

Current per terminal 157 amperes
Speed in r.p.m 250
Frequency in cycles per second 25
Gross core length of armature (\ g ) 100 cm.
Total number of slots 108
Conductors per slot 10
(All conductors per phase are in series) .

The no-load saturation curve is given in Fig. 117.

What is the armature
strength in ampere-turns per pole
when the output 157 amperes per phase? What, at 25 cycles,
is

is the reactance of the armature winding in ohms phase? Plot

225
226 POLYPHASE GENERATORS AND MOTORS
saturation curves when the generator is delivering full load
current (i.e., 157 amperes) for:

I. Power-factor = 1.00
II.
" =0.70
III.
" =0.20

For a constant excitation of 11 250 ampere-turns per field

spool, what will be the percentage drop in terminal pressure when

8000

2000 4000 GOOO 8000 10,000 12,000 14,000 16,000

Ampere Turns per Pole

FIG. 117. No-load Saturation Curve for the 3000-kva. Three-phase Generator
Described in Paper No. 1.

the output is increased from amperes to full-load amperes,

to 157 amperes:
i.e.,

I. At power-factor = 1 .00
II.
" =0.70
III.
" =0.20
 EXAMPLES FOR PRACTICE IN DESIGNING 227

For a constant terminal pressure of 11 000 volts, what will be

the required increase in excitation in going from amperes out-
put to full load (i.e., 157) amperes output

I. At power-factor = 1.00
II.
" =0.70
III.
" =0.20

In calculating theta (6) for this generator, work from the data
given on pp. 45 and 48.
2. A 100-h.p., 12-pole, 500-r.p.m., 50-cycle, 500-volt, Y-con-

nected, three-phase squirrel-cage induction motor has a no-load

current of 23 amperes and a circle ratio (a) of 0.058. The core
loss is 2200 watts, and the friction loss is 1400 watts. At rated
load the PR losses are

Stator 2850 watts

Rotor. 2850
"

[By means of a circle diagram and rough calculations based

on the above data, plot the efficiency, the power-factor, and the
amperes input, all as functions of the output, from no load up
to 100 per cent overload.

PAPER NUMBER II

1. In a certain three-phase, squirrel-cage induction motor, the

current per phase at rated load is 60 amperes. The no-load

current is 20 amperes. The circle ratio (c) is 0.040. Construct
the circle diagram of this motor. What is its power-factor at
its rated load? What is the current input per phase at the load
corresponding to the point of maximum power-factor, and what
" "
is the maximum power-factor? If the stand-still current
is 500 amperes (i.e., if the current when the full pressure is switched

on to the motor when it is at rest, is 500 amperes), ascertain

graphically by the aid of the circle diagram the power-factor at

the moment of starting. If the stator windings are Y-connected
and if the terminal pressure is 1000 volts (the pressure per phase

-
consequently being =577 volts), what would be the input
v3
228 POLYPHASE GENERATORS AND MOTORS
to the motor, in watts, at the moment of starting under these
conditions? Describe, without attempting to give quantitative
data, the means usually employed in practice to start such a
motor with much less than the above large amount of power and
" "
with much less current than the current given above.
stand-still
2. A
three-phase, Y-connected, squirrel-cage induction motor
has 48 stator slots and 12 conductors per slot. The terminal
pressure 250 volts (the pressure per phase consequently being
is

144 volts). The motor has 4 poles and its speed, at no load, is
1500 r.p.m. Estimate the magnetic flux per pole. The squirrel
cage comprises 37 face conductors, each having a cross-section
of 0.63 sq.cm. and each having a length between end rings of 20
cm. Each end ring has a cross-section of 2.45 sq.cm. and a mean
diameter of 20 cm. Calculate the PR loss in the squirrel cage
when the current in the stator winding is 17.3 amperes. If,
without making any further alteration in the motor, the cross-
section of the end rings is reduced to one-half, what will be the
PR loss in the squirrel cage for this same current? What general
" "
effect will this change have on the slip ? On the starting
torque? On the efficiency? On the heating?

PAPER NUMBER III

Determine appropriate leading dimensions and calculate as

much as practicable of the following design:
A 25-cycle, 250-volt, three-phase, squirrel-cage induction
motor for 750 h.p. and a synchronous speed of 250 r.p.m.
Calculate as much of this design as time permits. If, however,
you do not get on well with the entire design, then take some

particular part of the design, say the estimation of the magneto-

motive force and stator winding and carry it out in detail.
While there is still time, bring together the leading dimensions
and properties in a concise schedule. Make use of the printed
specifications (to be had on request) if desired.

PAPER NUMBER IV

A no-load saturation curve

is given in Fig. 118. This applies
to a three-phase alternator with a rated output of 850 kw. at
 EXAMPLES FOR PRACTICE IN DESIGNING 229

2880 volts per phase and 94 r.p.m. 32 pole, 25-cycle and unity
power-factor. At rated load the reactance voltage is 945 volts
per phase, and the resultant maximum armature strength is
3900 ampere-turns per pole.
Estimate the inherent regulation of this machine for rated full-
load current of 98.5 amperes at unity power-factor. Also for
this same current, but at a power-factor of 0.8.

Pressure

Phase
230 POLYPHASE GENERATORS AND MOTORS
But the chief consideration is that you should demonstrate your
ability to make a rough estimate of the most probably correct
design.)
1. Design a 4-pole, Y-connected, 30-cycle, three-phase, squir-

rel-cage induction motor for a primary terminal pressure of 1000

volts (577 volts per phase) and for a rated output of 100 h.p.

FIG. 119. Rough Indication of the Saturation Curves Called for in Paper
No. IV.

2. Design a 50-cycle, Y-connectea poiypiiase generator for

a rated output of 1500 kva. at a speed of 375 r.p.m. and for a ter-
minal pressure of 5000 volts (2880 volts per phase).

PAPER NUMBER VI
For the induction motor shown in Fig. 120:
1. Make
a rough estimate of a reasonable normal output to
assign to the motor.
2. Estimate the no-load current.

Estimate the circle ratio.

3.

Estimate the breakdown factor at the output you have

4.

assigned to the motor.

5. Estimate the losses and efficiency at the output you have

assigned to the motor.

6. Estimate the temperature rise at the output you have

assigned to the motor.

7. Estimate the power-factor at various loads.
 EXAMPLES FOR PRACTICE IN DESIGNING 231

(NOTE. If, rightly or wrongly, you consider that some essential data
have not been included in Fig. 120, do not lose time over the matter, but
make some rational assumption stating on your paper that you have done
so and then proceed with the work.)

Rotor Slot

Cond.
6.5 x 2.5mm

THREE-PHASE SQUIRREL-CAGE INDUCTION MOTOR.

Number of poles 12
Terminal pressure 1000 volts
1000
( .'. Pressure per phase . .
=
578 volts
V*
Synchronous speed in r.p.m. 500
Stator connections Y
Stator. Rotor.
Number of slots 180 216
Dimensions (depth Xwidth) 25 X 10 5 mm.
of slot . 21.5X8.0 mm.
Slot opening 3 mm. 3 mm.
Conductors per slot 6
Dimensions of bare conductor. 2.5 X6.5 mm.. . . 4.5X16.0 mm.
Number of end rings 2
Section of end ring 20 X20 mm.
Dimensions in centimeters and millimeters.

FIG. 120. Sketches and Data of Induction Motor of Paper No. VI.

PAPER NUMBER VII

Design the following three-phase, squirrel-cage induction

motor:
Rated output = 30 h.p.;
Synchronous speed = 1000 r.p.m.;
Periodicity
= 50 cycles per second;
Pressure between terminals = 500 volts;
Y-connected stator winding.

I .'. Pressure per phase = = 288 volts.

^= )
232 POLYPHASE GENERATORS AND MOTORS

Proportion the squirrel-cage rotor for 4 per cent slip at rated

load. Carry the design as far as time permits, but devote the
last half hour to preparing an orderly table of your results.

NOTE. If, rightly or wrongly, you conclude that some essential data have
not been included in the above, do not lose time over the matter, but make
some rational assumption stating on your paper that you have done so
and then proceed with the work.

PAPER NUMBER VIII

For the induction motor design of which data is given in Fig.
121, estimate the losses at rated load. Draw its circle diagram.
Stator Slot
2.16
|< |

3,Ducts
each 1.3 wide

Dimensions in cm
Rotor Slot
Dimensions in cm. Rotor Slot.

Output in H.P . 220

Number of poles 12
Connection of stator windings . Y
Periodicity in cycles per second 50
Volts between terminals .5000
Stator Winding.
Total number of stator conductors . . . .3564
Number of stator conductors per slot . .
33, each consisting of two
components
Dimensions of stator conductor (bare) . . . 2 in parallel, each 2.34 mm.
diameter.
Mean length of stator turn , 159 cm.
Rot-ir Windings.
Total number of rotor conductors .... . 288
Number of rotor conductors per slot. . 2
Number of phases in rotor winding. . . 3
Dimensions of rotor conductor (bare) . . 16X9 mm.
Mean length of rotor turn . 150 cm.

FIG. 121. Sketches and Data of the 220-H.P. Induction Motor of Paper
No. VIII.

Plot power-factor and output, using amperes input

its efficiency,

as abscissae, from no load up to the breakdown load. Estimate

 EXAMPLES FOR PRACTICE IN DESIGNING 233

the starting torque and the current input to the motor at starting,
when half the normal voltage is applied to the terminals of the
motor.
NOTE. If you are of opinion that sufficient data have not been given you
to enable all the questions to be answered, do not hesitate to make some
reasonable assumption for such missing data.

PAPER NUMBER IX

Design a 300-h.p., three-phase, 40-cycle, 240-r.p.m., squirrel-

cage induction motor for a terminal pressure of 2000 volts. Let
the stator be Y-connected. Obtain y, the ratio of the no-load to
the full-load current, a, the circle ratio, and bdf., the 'break-
down factor; and further data if time permits. Employ the
last half hour in criticising your own design, and in stating the

changes you would make with a view to improving it, if you had
.time.

PAPER NUMBER X
During the entire six hours at your disposal, design a three-

phase squirrel-cage induction motor, to the following specification:

Normal output in h.p 150

Periodicity in cycles per second 50
Synchronous speed (i.e. speed at no load), in r.p.m 250
Terminal pressure in volts 400
Connection of phases Y
Pressure per phase in volts . 231

The last two hours, i.e., from 3 to 5 P.M., must be devoted

to tabulating the results at which you have arrived, and to the
preparation of outline sketches with principal dimensions. As
to the electrical design, the following particulars will be expected
to be worked out or estimated:

1. Ratio of no-load to full-load current, (y).

2. Circle ratio, (a).
3. Circle diagram to scale.
4. Breakdown factor (bdf.).
5. Per cent slip at rated load.

6. Losses in stator winding, at rated load.

234 POLYPHASE GENERATORS AND MOTORS
7. Losses in rotor winding (i.e., squirrel-cage losses at rated
load).
8. Core losses.

9. Friction.
10. Efficiency at rated load.
11. Power-factor at rated load.
12. Estimate of thermometrically determined ultimate tem-
perature rise at rated load.

(NOTE. The students are permitted to bring in any books and notes
and drawing instruments they wish. They are also permitted to fill out
and to hand in as a portion of their papers, specification forms which they
have prepared in advance of coming to the examination.)

PAPER NUMBER XI

Any notes or note-books and other books may be used, but

students are put on their honor not to discuss any part of their
work in the lunch hour.
A three-phase, squirrel-cage induction motor complies with
the following general specifications:

Rated load in h.p = 90

Periodicity in cycles per second =40
Speed in r.p.m. at synchronism =800
Terminal pressure = 750
Connection of phases =Y

(1) Given the following data, estimate y, the ratio of no-

load to full-load current:

Average air-gap density =3610 lines per sq.cm.

Air-gap depth =0.91 mm.
Air-gap mmf -f- total mmf
.
= 80 .

Total number of conductor on stator .... = 576 .

(2) Given the following data, estimate a, the circle ratio:

Internal diameter of stator laminations =483 mm.

Slot pitch of stator at air-gap =21.1mm.
Slot pitch of rotor at air-gap =24.8mm.
Gross core length; \g =430 mm.
 EXAMPLES FOR PRACTICE IN DESIGNING 235

(3) Draw the circle diagram. Determine the primary

current :

(a) At point of maximum power-factor.

(6) breakdown load.

(4) Plot curves between:

(a) H.p. and YJ, the efficiency.

(6) H.p. and G, the power-factor.
(c) H.p. and slip.

(d) H.p. and amperes input.

Given the following data:

Section of stator conductor =0.167 sq.cm.

Dimensions of rotor conductor =1.27 cm.X0.76 cm.
(One rotor conductor per slot)
Length of each rotor conductor =49.5 cm.
Diameter of end rings (external) =45 em.
Number of end rings at each end = 2, each 2.54 cm. X 0.63 cm.
Total constant losses =2380 watts.

(5)Calculate the starting current and torque when com-

pensators supplying 33, 40 and 60 per cent of the terminal pres-
sure are used.
(6) Calculate new end
rings so that with a 2 1 compensator
:

the starting torque shall be equal to one-half torque at rated load.

(7) Designating the first motor as A and the second one

(i.e., the modification obtained from Question 6) as B, tabulate

the component losses at full load in two parallel (vertical columns) .

Estimate the watts total loss per ton weight of motor

(exclusive of side-rails and pul'ey).
(8) the squirrel-cage of the first motor A is replaced
If by
a three-phase winding having the same equivalent losses at
full load, the resistance per phase which would be
calculate
required, external to the slip rings, in order to limit the starting
current to 70 amperes. What would be the starting torque,
expressed as percentage of full-load torque, with this external
resistance inserted?
236 POLYPHASE GENERATORS AND MOTORS
PAPER NUMBER XII
For the three-phase induction motor of which data is given
below, make calculations enabling you to plot curves with
amperes input per phase as abscissae and power-factor, efficiency
and output in h.p. as ordinates.
Rated load =60 h.p.
Periodicity in cycles per second .
= 50
Speed at synchronism = 600 r.p.m.
Terminal pressure = 550 volts
Connection of phases =Y
Average air-gap density =3900 lines per sq.cm.

Air-gap depth =0. 9 mm.

Total mmf. -f- air-gap mmf = 1. 2
Total no. of conductors on stator = 720
Circle ratio =0.061
I2 R losses at rated load =2840
Constant losses = 2050
PAPER NUMBER XIII
The data given below are the leading dimensions of the stator
and rotor of a 24-pole three-phase induction motor with a Y-con-
nected winding suitable for a 25-cycle circuit. Ascertain approx-
imately by calculation the suitable terminal voltage for this
induction motor and give your opinion of the suitable rated out-
put. Proceed, as far as time permits, with the calculation of
the circle ratio and of the no-load current, and construct the
circlediagram. much as possible
Abbreviate the calculations as
by reasonable assumptions for the less important steps, thus
obtaining more time for the steps where assumptions can les
safely be made.
DIMENSIONS IN MM.
Stator.
External diameter 2800
Internal diameter 2440
Gross core length 385
Net core length 299
Winding 3-phase, Y-connected
Number of slots 216
Depth Xwidth of slot 51X21
Conductors per slot 12
Section of conductors, sq.cm 0.29
 EXAMPLES FOR PRACTICE IN DESIGNING 237

Rotor.
External diameter 2434.5
Internal diameter 2144
Winding 3-phase, Y-connected
Number of slots 504
Depth X width of slot 35X9.5
Conductors, per slot 2
Section of conductor, sq.cm 0.811

[This paper to be brought in for the afternoon examination (see Paper

No. XIV), as certain data in it will be required for the afternoon examina-
tion.]

PAPER NUMBER XIV

The 24-pole stator which you employed this morning (see
Paper No. XIII), for an induction motor, will, if supplied with
a suitable internal revolving field with 24 poles, make an excellent
three-phase, 25-cycle, Y-connected alternator. What would
be an appropriate value for the rated output of this alternator?
Without taking the time to calculate it, draw a reasonable no-
load saturation curve for this machine. From this curve and from
the data of the machine and your assumption as to the appro-
priate rating, calculate and plot a saturation curve for the rated
current when the power-factor of the external circuit is 0.80.

PAPER NUMBER XV
Of two 50-cycle, 100-h.p., 500-volt, three-phase induction
1.

motors, one has 4 poles and the other has 12 poles.

(a) Which will have the higher power-factor?
" " "
(6) efficiency?
" li " current at no load?
(c)
" " "
breakdown factor?
(d)
Of two 750 r.p.m., 100-h.p., 500-volt, three-phase induction
motors, one is designed for 50 cycles, and the other for 25 cycles,
(e) Which will have the higher power-factor?
" " " current at no load?
(/)
li " " breakdown factor?
(0)
2. Describe how to estimate the temperature rise of an induc-

tion motor.
3. Describe how to estimate the magnitude and phase of the
starting current of a squirrel-cage induction motor.
238 POLYPHASE GENERATORS AND MOTORS

PAPER NUMBER XVI

Answer one of the following two questions.
Question I. For the three-phase squirrel-cage, induction motor
of which data are given in Fig. 122:

Rotor Slot
h-i3,H
mm
Terminal pressure 750
Method of connection Y
Pressure per phase 432
Speed in r.p.m SOO
Full load primary current input per phase. . . '. 59
Number of primary conductors per slot 8
Periodicity in cycles per second 40
 EXAMPLES FOR PRACTICE IN DESIGNING 239

Question II. For a certain three-phase, slip ring, induction

motor the following data apply :

Periodicity in cycles per sec 50

Speed at no load in r.p.m 500
Y (ratio of no-load to full-load current) .36
a (circle ratio) 0.0742
Rated output in h.p 300
Terminal pressure 700
Pressure per phase 405
Connection of phases Y
Stator resistance per phase (ohm) 030 .

Rotor resistance per phase (ohm) 022 .

Total core loss (watt) 4080

Friction and windage loss (watt) 2000
Ratio number of stator to number of rotor conductors ... 1 28 .

(a) Draw the circle diagram.

(6) Plot curves of efficiency, power-factor, output in horse-
power and slip, all as a function of the current input per phase.

PAPER NUMBER XVII

1. For an armature having an air-gap diameter, Z) = 65 cms.
and a gross core length \g = 35
cms.
What would be a suitable rating for a 500-volt, 500-r.p.m.
machine of these data:
1st. As a 25-cycle induction motor.

2d. As a 25-cycle alternator.

Select one of these cases andwork out the general lines of the
design as far as time permits.
2. In Fig. 123 are given data of the design of an alternator

for the following rating:

2500 kva., 3-phase, 25-cycle, 75 r.p.m., 6500-volts, Y-con-
nected. The field excitations for normal voltage of 2200 volts and
for 1.2 times normal voltage (2640 volts), are given. From
these values the no-load saturation curve may be drawn.
Estimate (showing all the necessary steps in the calculations).
(a) The field ampere-turns required for full terminal voltage
at full-load kva. at power-factors 1.0 and 0.8; and the pressure
regulation for both these cases.
240 POLYPHASE GENERATORS AND MOTORS

(b) The current for normal speed and with

short circuit
no-load normal voltage.
field excitation for

(c) The losses and efficiency at rated full load and J load
at 0.8 power-factor. Also the armature heating.

Scale 1-20
25 65 V<>lt
Mh -

Y Connected A.C. Generator

Data
No of Conductors per Slot 9
True Cross Section of 1 Conductor 1.29 sq.cm.

Field Spool Winding

Turns per Spool 42.5

Mean Length of 1 Turn 1620mm
Cross Section of Conductor 1.78 sq. cms

Saturation- Ampere Turns at 6500 Volts 8000

, .. .. 7800 ... 13000
Air Gap Ampere Turns at 6500 Volts goOO

FIG. 123. Data of the Design of the 2500-kva. Alternator of Question 2

of Paper No. XVII.

3. In Fig. 121 are given data of the design of an induction

motor for the following rating :

40-h.p., 600-r.p.m., 50-cycles, 500-volt A-connected, three-

phase induction motor.

Scale 1-10

Stator Slot opening 4mm

Rotor Slot opening l.Smn

40 H.P. 600 R.P.M. 50f\J500 Volt, 3 Phase Induction Motor

Stator Winding Rotor Winding
Connection Winding A Squirrel Cage Type All Dimensions
No.of Conductors pet- Slofr 18 Total No.of
No. of Bars in Millimeters
39
Cross Section of 1 Conductor 0.0685 sq.cms Cross Section of Bar sq.cm
1.0
*
Friction and Windage Losses 300 Cross Section of End Ring 3.0 sq.cm

FIG. 124. Data of the Design of the 40-H.P. Motor of Question 3 of

Paper No. XVII.

Estimate (showing all the necessary steps in the calculations).

(a) Circle ratio ( a) ;

(b) No-load current in per cent of full-load current;

(c) Breakdown factor ;

(d) Maximum power-factor.

 EXAMPLES FOR PRACTICE IN DESIGNING 241

(e) Losses and efficiencies and heating at \ load and at rated

full load.

(/) Slip at full load.

PAPER NUMBER XVIII

(You may use notes, curves or any other aids.)

1. The leading particulars of a certain induction motor are
given in Fig. 125.

80. H.P., 600 r.p.m., 50 CYCLES, 500 VOLTS, A-CONNECTED SQUIRREL-CAGE INDUCTION
MOTOR
Data:
Number of stator slots 90
f Depth 36.0
Slot dimensions < Width 11.0
[ Opening 6.0
Number of rotor slots Ill
f Depth 21.5
Slot dimensions { Width 6.5
[ Opening 1.5
WINDINGS:
Stator
Conductors per slot 12
Cross-section of conductor 0. 138 sq.cm.
Rotor:
Bars per slot 1
Cross-section of bar 1.0 sq.cm.
Cross-section of end ring 3.6 sq.cm.
All dimensions in millimeters.

FIG. 125. Sketches and Data of the 80-H.P. Induction Motor of Question 1
of Paper No. XVIII.

(a) Estimate y ti\e ratio of the no-load current to the current

at normal rating.
(6) Estimate a the circle ratio, and also estimate the maximum

power-factor.
(c) Estimate bdf., the breakdown factor.
242 POLYPHASE GENERATORS AND MOTORS

(d) Estimate the copper losses at normal rating.

(e) Estimate the core losses at normal rating.
(/) Estimate the watts per square decimeter of equivalent gap
surface.
(g) Estimate the efficiency at normal rating.
(h) Estimate the power-factor at normal rating.
(i) Estimate the T.W.C. (the total works cost).
(a) Deduce the leading proportions for a three-phase
2.

alternator for a normal rating of 1500 kw., 1000 r.p.m., 50 cycles,

6 poles, 11 000 volts. Do not go into detail, but go far enough to
give an opinion as to the suitable values for D, \g, number of
slots, conductors per slot, and field excitation at no load. State
brieflyyour reasons for choosing each of these values.
What, in a general way, should be the changes in the
(6)

general order of magnitude of these quantities for a design for

the same rated output and voltage at

(1) 50 cycles, 250 r.p.m., 24 poles,

and the changes necessary in this second design in order to obtain

a design for

(2) 25 cycles, 250 r.p.m., 12 poles.

PAPER NUMBER XIX

Work up a roughoutline for a design for a 25-cycle, three-
phase generator to supply at a pressure of 10 000 volts (Y-con-
nected with 5770 volts per phase). The generator is to have a
rated capacity of 3000 kva. at a power-factor of 0.90, and is to
be run at a speed of 125 r.p.m. Design the machine to give good
regulation of the pressure, and work out the inherent regulation
at various power-factors. Work out any other data for which
you find time, and devote the last half hour to an orderly tabu-
lation of your results.

PAPER NUMBER XX
Work upa rough outline for a design fc/r a 25-cycle, 8-pole,
three-phase, squirrel-cage induction motor for a rated output
of 80 h.p.
The terminal pressure is 600 volte, i.e., 346 volts per phase,
 EXAMPLES FOR PRACTICE IN DESIGNING 243

and the s tat or windings should be Y-connected. The load will

not be thrown on the motor until it is up to speed.
During the last half hour prepare an orderly tabulation of
the leading results which you have found time to work out.

PAPER NUMBER XXI

Commence the design of a three-phase, 100-h.p., squirrel-
cage induction motor with a Y-connected stator for a synchro-
nous speed of 375 r.p.m. when operated from a 25-cycle circuit
with a line pressure of 500 volts. (The pressure per stator wind-
500
ing is consequently ^==288 volts.) Try and carry the design

as far as determining upon the gap diameter and the gross core
length, the number of stator slots, the number of stator conductors
per slot, the flux per pole and the external diameter of the stator
laminations and the internal diameter of the rotor laminations.
Then tabulate these data in an orderly manner before proceeding
further. Then, if time
permits, make
the magnetic circuit
calculations and estimate the magnetizing current,

PAPER NUMBER XXII

(Answer one of the following two questions.)

1. Which would be the least desirable, as regards interfering
with the pressure on a 50-cycle net work, low-speed or high-
speed induction motors? Why?
Of two 1000-volt, 100-h.p., 750-r.p.m., three-phase induction
motors, which would have the highest capacity for temporarily
carrying heavy overloads, a 25-cycle or a 50-cycle design? Wliich
would have the highest power-factor? Which the lowest current
when running unloaded?
Of two 1000-volt, 100-h.p., three-phase induction motors
for 25 cycles, one is for a synchronous speed of 750 r.p.in. and the
other is for a synchronous speed of 150 r.p.m. Assuming rational
design in both cases, which has the higher power-factor? Which
the lower current when running light? Which the higher
breakdown factor? Which the higher efficiency? Three 100-
244 POLYPHASE GENERATORS AND MOTORS

h.p., 1000- volt designs have been discussed above. Their speeds
and periodicitiesmay be tabulated as follows:
DpdirnatJrm Synchronous speed Periodicity in cycles
inr.p.m. per sec.
A 750 25
B 750 50
C 150 25

Assume that these are provided with low-resistance squirrel-

cage windings. Make rough estimations of the watts total loss
per ton of total weight of motor for each case.
2. A three-phase, Y-connected, squirrel-cage induction motor

has the following constants:

External diameter stator core 1150 mm.

Air-gap diameter (D) 752 mm.
Internal diameter rotor core 324 mm.
Diameter at bottom of stator slots ... 841 mm.
Diameter at bottom of rotor slots 710 mm.
No. of stator slots 72
No. of rotor slots 89
Conductors per stator slot 5
Conductors per rotor slot 1

Width of stator slot 19 mm.

Width of rotor slot 12 mm.
Space factor stator slot 0.51
Space factor rotor slot 0.80
Gross core length (Xgr) 190 mm.
Net core length (Xn) 148mm.
Polar pitch (T) 394 mm.
Depth of air-gap (A) 1.5 mm.

Peripheral speed 38.6 mps.

Output coefficient () 2.37

The pressure per phase is 1155 volts, the pressure between

terminals being 2000 volts.
What is the rated output of the motor in h.p.?
What is the speed in r.p.m.?
What is the periodicity in cycles per second?
What is the flux per pole in megalines?
What is the stator PR loss?
 EXAMPLES FOR PRACTICE IN DESIGNING 245

What is the stator core loss?

What cross-section must be given to the copper end rings
in order that the slip at rated load shall be 2 per cent?
Estimate the efficiency at J, J and full load.
Estimate a the circle ratio.
Estimate Y the ratio of the magnetizing current to the current
at rated load.
What is the power-factor at }-, J and full load?
What is the breakdown factor?
Estimate the probable temperature rise after continuous
running at rated load.
Remember that some of these questions are most readily
solved by constructing a circle diagram and combining the
dimensions scaled off from it, with slide-rule calculations.
 APPENDIX I

A BIBLIOGRAPHY OF I. E. E. AND A. I. E. E. PAPERS ON

THE SUBJECT OF POLYPHASE GENERATORS.

1891.

M. I. PUPIN. On Polyphase Generators. (Trans. Am. Inst. Elec.

Engrs., Vol. 8, p. 562).

1893.

GEORGE FORBES. The Electrical Transmission of Power from

Niagara Falls. (Jour. Inst. Elec. Engrs., Vol. 22, p. 484.)

1898.

A. F. McKissiCK. Some Tests with an Induction Generator. (Trans.

Am. Inst. Elec. Engrs., Vol. 15, p. 409.)

1899.

M. R. GARDNER and R. P. HOWGRAVE-GRAHAM. The Synchronizing

of Alternators. (Jour. Inst. Elec. Engrs., Vol. 28, p. 658.)

1900.

B. A. BEHREND. On the Mechanical Forces in Dj^namos Caused

by Magnetic Attraction. (Trans. Am. Inst. Elec. Engrs., Vol. 17, p. 617.)

1901.

W. L. R. EMMET. Parallel Operation of Engine-Driven Alternators.

(Trans. Am. Inst. Elec. Engrs., Vol. 18, p. 745.)
ERNST J. BERG. Parallel Running of Alternators. (Trans. Am.
Inst. Elec. Engrs., Vol. 18, p. 753.)

247
248 POLYPHASE GENERATORS AND MOTORS
P. 0. KEILHOLTZ. Angular Variation in Steam Engines. (Trans.
Am. Inst. Elec. Engrs., Vol. 18, p. 703.)
CHAS. P. STEINMETZ. Speed Regulation of Prime Movers and
Parallel Operation of Alternators. (Trans. Am. Inst. Elec. Engrs.
Vol. 18, p. 741.)
WALTER I. SLIGHTER. Angular Velocity in Steam Engines in Rela-
tion to Paralleling of Alternators. (Trans. Am. Inst. Elec. Engrs.,
Vol. 18, p. 759.)

1902.

C. 0. MAILLOUX. An Experiment with Single-Phase Alternators on

Polyphase Circuits. (Trans. Am. Inst. Elec. Engrs., Vol. 19, p. 851.)
Louis A. HERDT. The Determination of Alternator Characteristics.

(Trans. Am. Inst. Elec. Engrs., Vol. 19, p. 1093.)

C. E. SKINNER. Energy Loss in Commercial Insulating Materials
when Subjected to High-Potential Stress. (Trans. Am. Inst. Elec.
Engrs., Vol. 19, p. 1047.)

1903.

C. A. ADAMS. A Study of the Heyland Machine as Motor and Gen-

erator. (Trans. Am. Inst. Elec. Engrs., Vol. 21, p. 519.)
W. L. WATERS. Commercial Alternator Design. (Trans. Am. Inst.
Elec. Engrs., Vol. 22, p. 39.)
A. S. GARFIELD. The Compounding of Self-Excited Alternating-
Current Generators for Variation in Load and Power Factor. (Trans.
Am. Inst. Elec. Engrs., Vol. 21, p. 569.)
B. A. BEHREND. The Experimental Basis for the Theory of the

Regulation of Alternators. (Trans. Am. Inst. Elec. Engrs., Vol. 21,

p. 497.)

1904.

A. F. T. ATCHISON. Some Properties of Alternators Under Various

Conditions of Load. (Jour. Inst. Elec. Engrs., Vol. 33, p. 1062.)
H. W. TAYLOR. Armature Reaction in Alternators. (Jour. Inst.
Elec. Engrs., Vol. 33, p. 1144.)
MILES WALKER. Compensated Alternate-Current Generators.
(Jour. Inst. Elec. Engrs., Vol. 34, p. 402.)
J. B. HENDERSON and J. S. NICHOLSON. Armature Reaction in
Alternators. (Jour. Inst. Elec. Engrs., Vol. 34, p. 465.)
DAVID B. RUSHMORE. The Mechanical Construction of Revolving
Field Alternators. (Trans. Am. Inst. Elec. Engrs., Vol 23, p. 253.)
B. G. LAMME. Data and Tests on a 10 000 Cycle-per-Second Alter-
nator. (Trans. Am. Inst. Elec. Engrs., Vol. 23, p. 417.)
 APPENDIX 249

H. H. BARNES, Jr. Notes on Fly-Wheels. (Trans. Am. Inst. Elec.

Engrs., Vol. 23, p. 353.)

H. M. HOBART and FRANKLIN PUNGA. A Contribution to the
Theory of the Regulation of Alternators. (Trans. Am. Inst. Elec. Engrs.,
Vol. 23, p. 291.)

1905.

WILLIAM STANLEY and G. FACCIOLI. Alternate-Current Machinery,

with Especial Reference to Induction Alternators. (Trans. Am. Inst.
Elec. Engrs., Vol. 24, p.^51.)
A. B. FIELD. Eddy Currents in Large, Slot-wound Conductors.
(Trans. Am. Inst. Elec. Engrs., Vol. 24, p. 761.)
W. J. A. LONDON. Mechanical Construction of Steam-Turbines and
Turbo-Generators. (Jour. Inst. Elec. Engrs., Vol. 35, p. 163.)

1906.

J. EPSTEIN. Testing Electrical Machinery and Materials. (Jour.

Inst. Elec. Engrs., Vol. 38, p. 28.)
A. G. ELLIS. Steam Turbine Generators. (Jour. Inst. Elec. Engrs.,
Vol. 37, p. 305.)
SEBASTIAN SENSTIUS. Heat Tests on Alternators. (Trans. Am.
Inst. Elec. Engrs., Vol. 25, p. 311.)
MORGAN BROOKS and M. K. AKERS. The Self-Synchronizing of
Alternators. (Trans. Am. Inst. Elec. Engrs., Vol. 25, p. 453.)
E. F. ALEXANDERSON. A Self-Exciting Alternator. (Trans. Am.
Inst. Elec. Engrs., Vol. 25, p. 61.)

1907.

B. A. BEHREND. Introduction to Discussion on the Practicability of

Large Generators Wound for 22000 Volts. (Trans. Am. Inst. Elec.

Engrs., Vol. 26, p. 351.)

ROBERT POHL. Development of Turbo-Generators. (Jour. Inst.
Elec. Engrs., Vol. 40, p. 239.)
G. W. WORRALL. Magnetic Oscillations in Alternators. (Jour. Inst.
Elec. Engrs., Vol. 39, p. 208.)[This paper is supplemented by another
paper contributed by Mr. Worrall in 1908.]

1908.

M. KLOSS. Selection of Turbo- Alternators. (Jour. Inst. Elec.

Engrs., Vol. 42, p. 156.)

S. P. SMITH. Testing of Alternators. (Jour. Inst. Elec. Engrs.,
Vol. 42, p. 190.)
250 POLYPHASE GENERATORS AND MOTORS
G. STONE Y and A. H. LAW. High-Speed Electrical Machinery.
(Jour. Inst. Elec. Engrs., Vol. 41, p. 286.)
R. K. MORCOM and D. K. MORRIS. Testing Electrical Generators.
(Jour. Inst. Elec. Engrs., Vol. 41, p. 137.)
G. W. WORRALL. Magnetic Oscillations in Alternators. (Jour. Inst.
Elec. Engrs., Vol. 40, p. 413.) [This paper is a continuation of Mr.
Worrall's 1907 paper. |

JENS BACHE-WIIG. Application of Fractional Pitch Windings to

Alternating-Current Generators. (Trans. Am. Inst. Elec. Engrs.,
Vol. 27, p. 1077.)
CARL J. FECHHEIMER. The Relative Proportions of Copper and
Iron in Alternators. (Trans. Am. Inst. Elec. Engrs., Vol. 27, p. 1429.)

1909.

S. P. SMITH. The Testing of Alternators. (Jour. Inst. Elec. Engrs.,

Vol. 42, p. 190.)
J. D. COALES. Testing Alternators. (Jour. Inst. Elec. Engrs.,
Vol. 42, p. 412.)
E. ROSENBERG. Parallel Operation of Alternators. (Jour. Inst.
Elec. Engrs., Vol. 42, p. 524.)
E. F. W. ALEXANDERSON. Alternator for One Hundred Thousand
Cycles. (Trans. Am. Inst. Elec. Erigrs., Vol. 28, p. 399.)
CARL J. FECHHEIMER. Comparative Costs of 25-Cycle and 60-Cycle
Alternators. (Trans. Am. Inst. Elec. Engrs., Vol. 28, p. 975.)
C. A. ADAMS. Electromotive Force Wave-Shape in Alternators.
(Trans. Am. Inst. Elec. Engrs., Vol. 28, p. 1053.)

1910.

MILES WALKER. Short-Circuiting of Large Electric Generators.

(Jour. Inst. Elec. Engrs., Vol. 45, p. 295.)
MILES WALKER. Design of Turbo Field Magnets for Alternate-
Current Generators. (Jour. Inst. Elec. Engrs., Vol. 45, p. 319.)
GEO. I. RHODES. Parallel Operation of Three-Phase Generators
with their Neutrals Interconnected. (Trans. Am. Inst. Elec. Engrs.,
Vol. 29, p. 765.)
H. G. STOTT and R. J. S. PIGOTT Tests of a 15 000-kw. Steam-
Engine-Turbine Unit. (Trans. Am. Inst. Elec. Engrs., Vol. 29, p. 183.)
E. D. DICKINSON and L. T. ROBINSON. Testing Steam Turbines
and Steam Turbo-Generators. (Trans. Am. Inst. Elec. Engrs., Vol. 29,

p. 1679.)
 APPENDIX 251

1911.

J. R. BARR. Parallel Working of Alternators. (Jour. Inst. Elec.

Engrs., Vol. 47, p. 276.)
A. P. M. FLEMING and R. JOHNSON. Chemical Action in the Wind-
ings of High-Voltage Machines. (Jour. Inst. Elec. Engrs., Vol. 47,
p. 530.)
S.' P. SMITH. Non-Salient-Pole Turbo-Alternators. (Jour. Inst.
Elec. Engrs., Vol. 47, p. 562.)
W. W. FIRTH. Measurement of Relative Angular Displacement in
Synchronous Machines. (Jour. Inst. Elec. Engrs., Vol. 46, p. 728.)
R. F. SCHUCHARDT and E. 0. SCHWEITZER. The Use of Power-
Limiting Reactances with Large Turbo-Alternators. (Trans. Am. Inst.
Elec. Engrs., Vol. 30.)

1912.

H. D. SYMONS and MILES WALKER. The Heat Paths in Electrical

Machinery. (Jour. Inst. Elec. Engrs., Vol. 48, p. 674.)

W. A. DURGIN and R. H. WHITEHEAD. The Transient Reactions of
Alternators. (Trans. Am. Inst. Elec. Engrs., Vol. 31.)
A. B. FIELD. Operating Characteristics of Large Turbo-Generators.
(Trans. Am. Inst. Elec. Engrs., Vol. 31.)
H. M. HOBART and E. KNOWLTON. The Squirrel-Cage Induction
Generator. (Trans. Am. Inst. Elec. Engrs., Vol. 31.)
E. M. OLIN. Determination of Power Efficiency of Rotating Elec-
tric Machines. (Trans. Am. Inst. Elec. Engrs., Vol. 31.)
D. W. MEAD. The Runaway Speed of Water-Wheels and its Effect
on Connected Rotary Machinery. (Trans. Am. Inst. Elec. Engrs.,
Vol. 31.)
D. B. RUSHMORE. Excitation of Alternating-Current Generators.
(Trans. Am. Inst, Elec, Engrs., Vol. 31.)
 APPENDIX II

A BIBLIOGRAPHY OF I. E. E. AND A. I. E. E. PAPERS ON

THE SUBJECT OF POLYPHASE MOTORS

1888.

NIKOLA TESLA. A New System of Alternate-Current Motors and

Transformers. (Trans. Am. Inst. Elec. Engrs., Vol. 5, p. 308.)

1893.

ALBION T. SNELL. The Distribution of Power by Alternate-Current

Motors. (Jour. Inst. Elec. Engrs., Vol. 22, p. 280.)

1894.

Louis BELL. Practical Properties of Polyphase Apparatus. (Trans.

Am. Inst. Elec. Engrs., Vol. 11, p. 3.)
Louis BELL. Some Facts about Polyphase Motors. (Trans. Am.
Inst. Elec. Engrs., Vol. 11, p. 559.)
Louis DUNCAN, J. H. BROWN, W. P. ANDERSON, and S. Q. HAYES.
Experiments on Two-Phase Motors. (Trans. Am. Inst. Elec. Engrs.,
Vol. 11, p. 617.)
SAMUEL REBER. Theory of Two- and Three-Phase Motors. (Trans.
Am. Inst. Elec. Engrs., Vol 11, p. 731.) .

CHAS. P. STEINMETZ. Theory of the Synchronous Motor. (Trans.

Am. Inst. Elec. Engrs., Vol. 11, p. 763.)
LUDWIG GUTMANN. On the Production of Rotary Magnetic Fields
by a Single Alternating Current. (Trans. Am. Inst. Elec. Engrs., Vol. 11,

p. 832.)

1897.

CHAS. P. STEINMETZ. The Alternating-Current Induction Motor.

(Trans. Am. Inst. Elec. Engrs., Vol. 14, p. 185.)

252
 APPENDIX 253

1899.

ERNEST WILSON. The Induction Motor. (Jour. Inst. Elec. Engrs.,

Vol. 28, p. 321.)

1900.

A. C. EBORALL. Alternating Current Induction Motors. (Jour.

Inst. Elec. Engrs., Vol. 29, p. 799.)

1901.

CHAS. F. SCOTT. The Induction Motor and the Rotary Converter

and Their Relation to the Transmission System. (Trans. Am. Inst.
Elec. Engrs., Vol. 18, p. 371.)

1902.

ERNST DANIELSON. A Novel Combination of Polyphase Motors for

Traction Purposes. (Trans. Am. Inst. Elec. Engrs., Vol. 19, p. 527.)
CHAS. P. STEINMETZ. Notes on the Theory of the Synchronous
Motor. (Trans. Am. Inst. Elec. Engrs., Vol. 19, p. 781.)

1903.

C. A. ADAMS. A Study of the Heyland Machine as Motor and

Generator. (Trans. Am. Inst. Elec. Engrs., Vol. 21, p. 519.)
H. BEHN-ESCHENBURG. Magnetic Dispersion in Induction Motors.
(Jour. Inst. Elec. Engrs., Vol. 33, p. 239.)

1904.

B. G. LAMME. Synchronous Motors for Regulation of Power Factor

and Line Pressure. (Trans. Am. Inst. Elec. Engrs., Vol. 23, p. 481.)
H. M. HOBART. The Rated Speed of Electric Motors as Affecting
the Type to be Employed. (Jour. Inst. Elec. Engrs., Vol. 33, p. 472.)

1905.

R. GOLDSCHMIDT. Temperature Curves and the Rating of Electrical

Machinery. (Jour. Inst. Elec. Engrs., Vol. 34, p. 660.)
D. K. MORRIS and G. A. LISTER. Eddy-Current Brake for Testing
Motors. (Jour. Inst. Elec. Engrs., Vol. 35, p. 445.)
P. D. IONIDES. Alternating-Current Motors in Industrial Service.
(Jour. Inst. Elec. Engrs., Vol. 35, p. 475.)
254 POLYPHASE GENERATORS AND MOTORS
C. A. ADAMS. The Design of Induction Motors. (Trans. Am. Inst.
Elec. Engrs., Vol. 24, p. 649.)
CHAS. A. PERKINS. Notes on a Simple Device for Finding the Slip
of an Induction Motor. (Trans. Am. Inst. Elec. Engrs., Vol. 24, p. 879.)
A. S. LANGSDORF. Air-Gap Flux in Induction Motors. (Trans. Am.
Inst. Elec. Engrs., Vol. 24, p. 919.)

1906.
\

J. B. TAYLOR. Some Features Affecting the Parallel Operation of

Synchronous Motor-Generator Sets. (Trans. Am. Inst. Elec. Engrs. ,
Vol. 25, p. 113.)
BRADLEY McCoRMiCK. Comparison of Two- and Three-Phase
Motors. (Trans. Am. Inst. Elec. Engrs., Vol. 25, p. 295.)
A. BAKER and J. T. IRWIN. Magnetic Leakage in Induction Motors.

(Jour. Inst. Elec. Engrs., Vol. 38, p. 190.)

1907.

L. J. HUNT. A New Type of Induction Motor. (Jour. Inst. Elec.

Engrs., Vol. 39, p. 648.)
R. RANKIN. Induction Motors. (Jour. Inst. Elec. Engrs., Vol. 39,
p. 714.)
C. A. ADAMS, W. K. CABOT, and C. A. IRVING, Jr. Fractional-Pitch
Windings for Induction Motors. (Trans. Am. Inst. Elec. Engrs., Vol.
26, p. 1485.)
R. E. HELLMUND. Zigzag Leakage of Induction Motors. (Trans.
Am. Inst. Elec. Engrs., Vol. 26, p. 1505.)

1908.

R. GOLDSCHMIDT. Standard Performances of Electrical Machinery.

(Jour. Inst. Elec. Engrs., Vol. 40, p. 455.)
G. STEVENSON. Polyphase Induction Motors. (Jour. Inst. Elec.

Engrs., Vol. 41, p. 676.)

H. C. SPECHT. Induction Motors for Multi-Speed Service with
Particular Reference to Cascade Operation. (Trans. Am. Inst. Elec.

Engrs., Vol. 27, p. 1177.)

1909.

J. C. MACFARLANE and H. BURGE. Output and Economy Limits of

Dynamo-Electric Machinery. (Jour. Inst. Elec. Engrs., Vol. 42, p. 232.)

S. B. CHARTERS, Jr., and W. A. HILLEBRANDT. Reduction in
 APPENDIX 255

Capacity of Polyphase Motors Due to Unbalancing in Voltage. (Trans.

Am. Inst. Elec. Engrs., Vol. 28, p. 559.)
H. G. REIST and H. MAXWELL. Multi-Speed Induction Motors.
(Trans.Am. Inst. Elec. Engrs., Vol. 28, p. 601.)
A. MILLER GRAY. Heating of Induction Motors. (Trans. Am. Inst.
Elec. Engrs., Vol. 28, p. 527.)

1910.

R. E. HELLMUND. Graphical Treatment of the Zigzag and Slot

Leakage in Induction Motors. (Jour. Inst. Elec. Engrs., Vol. 45, p. 239.)
C. F. SMITH. Irregularities in the Rotating Field of the Polyphase
Induction Motor. (Jour. Inst Elec. Engrs., Vol. 46, p. 132.)
WALTER B. N YE. The Requirements for an Induction Motor from
the User's Point of View. (Trans. Am. Inst. Elec. Engrs., Vol. 29,
p. 147.)

1911.

T. F. WALL. The Development of the Circle Diagram for the

Three-Phase Induction Machine. (Jour. Inst. Elec. Engrs., Vol. 48,
p. 499.)
N. PENSABENE-PEREZ. An Automatic Starting Device for Asyn-
chronous Motors. (Jour. Inst. Elec. Engrs., Vol. 48, p. 484.)
C. F. SMITH and E. M. JOHNSON The Losses in Induction Motors
Arising from Eccentricity of the Rotor. (Jour. Inst. Elec. Engrs.,
Vol. 48, p. 546.)
H. J. S. HEATHER. Driving of Winding Engines by Induction
Motors. (Jour. Inst. Elec. Engrs., Vol. 47, p. 609.)
THEODORE HOOCK. Choice of Rotor Diameter and Performance of
Polyphase Induction Motors. (Trans. Am. Inst. Elec. Engrs., Vol. 30.)
Gus A. MAIER. Methods of Varying the Speed of Alternating
Current Motors. (Trans. Am. Inst. Elec. Engrs., Vol. 30.)

1912.

J. K. CATTERSON-SMITH. Induction Motor Design. (Jour. Inst.

Elec. Engrs., Vol. 49, p. 635.)
CARL J. FECHHEIMER. Self-Starting Synchronous Motors. (Trans.
Am. Inst. Elec. Engrs., Vol. 31.)
H. C. SPECHT. Electric Braking of Induction Motors. (Trans. Am.
Inst. Elec. Engrs., Vol. 31.)
P. M. LINCOLN. Motor Starting Currents as Affecting Large Trans-
mission Systems, (Trans. Am, Inst, Elec, Engrs., Vol. 31.)
 APPENDIX III

TABLE OF SINES, COSINES, AND TANGENTS FOR USE IN THE CALCULATIONS

IN CHAPTER II.

Angle.
 O001>tOTH
000
(M b- OS lO (M CO i

03 T i TP r~ CN Ososcoosoo TH !> '

IO T-H T-H
O
I

T-H T-H T-H (M (M CO to GO to CO

T-H
to
CO CO 1> GO
i T1 !> CO
i
OO
T^ OS !>

o,a 23i
O
^
oOoO CO O S
T-H
ooo 3:
gg 00000 00000
OS OS
oO to oo
00 T-H to to
co
C<J GO OS O O O O !> T-H
O O O
CO
OS OS <M
O
TH CO <N T-H
T to
i

GO CM
T-HCSji-HCOt^-
T i
CO Oq T-HO OOOOOO
TH T i

tOOOOO OOOOOO
O
OS
to OOO to to O T-H to TH TH CO T i
C-lto
OO >O
lO O OS C^COT-Ht^TH (MtOt^OSfM COT-HI>THT-(
TH co d (MT-HT-H GO TH CO (M CM T-H T I
O
O OS 00
-^H
to O CO O
iO to Os O to T-H T i
<N O I> O O
to oo oo oq T-H Oco <N COt^-
ddddd doT-i,-H"co'
GOT-H GO

CO CO s*
O^ 10 r^* f^ oco oo oo 10 OOO O to
t>
tocoo
1

OS<N t^CO <N IO IO OS GO

O l> CO
T-H
I-H T-H <N (M CO T^ OI^ O t^- CO (M tO

do ooo O OT-H T-H CO

CO <M CO rH iO

o t^O T^ O5
T)H 10 >OO
T-H O T-H OS T-H ...
(M rH TH 00 1>
t^i-H IO ii OS go .

>ot^o ooo
P T-H C^ (M COCO

ddddd rn' TH' <N'

IO T-H OS i-H
CO (M CO CO IO

fc
>O CO 1> C<J C^l OS TT
O to TH
CO OS CO OS l^
T-H T-H <N cq co
ooo^ coco
co OSThi co
TJH co os r^ GO CO Os
CO OS
T i 03 to CO
I>-

O (N
ddddd d d o' T-H ci
(M CO TH CO <N

10
TH t^ 1^ Tfl
00 O O CO O O O OtOOOO
-TH rHO
CO OOCOTtH <N t Tt TH i
T-H GO t^ <M >O (M
tO <M (M CO TH COGOCO
oo 'I-H <M
(N

T-H GC CO <N OS "* o co c^ O

TtH CO
T-H T-H
O >O T-H
(M C^ CO
T 1
CO TH
^ 1C GO (M <M
t>- CO OS
OO >O T-H T-H
00 I> TjH

00000 <M (M CO'*

O
to TH OS
o coos co co to T-H r^
GO OS (N t^ CO
o O
10
CO
oO t^-oo co
OQ CO O 0000to CO l^ CO
T-
OS(M b- COO
T i 1 1

CO <M O OOT-H T-H

T-HQOOQ OOOOQ OOOOO OOOOOO

OO
TH
TJH 1 i :

<N TH T-H T-H TH OO

tOtOT-H(MOO COOSOSOSCT-
C<J O OOO
COCOCOCOfN C-l T I
O OS 00 00 l> CO to
(M T-H T-H OO dooo
t>- CO to TH (M
(NrHOOO ddddd
(M T-H O OS 00 CO iO TH CO (N TH
odd ddodod
 INDEX

Air Circulation, see Ventilation Calculations for

Air-gap Slip Ring Induction Motor, 195-
Density, Method of estimating for 201
Synchronous Generators, 36 Squirrel-cage Induction Motor,
Diameter, Explanation of term, 4 105 et seq Design of Squirrel-
(of) Induction Generators, 217, 219, cage for, 173-178
222-224 Synchronous Generator, 3 et seq

(of) Induction Motors, 132-134 Specification for, 40, 41 Deriva-

Radiating Surface at the, Data for tion of Design for Induction
estimating, 181 Generator from, 216-222
(of) Synchronous Generator, 36, 37 Circle Diagram of Squirrel-cage
A.I.E.E., Bibliography of papers on Motor, 154, 170
Polyphase Generators, 246- General Observations regarding,
251 on Polyphase Motors, 252- 172
255 Locus of Rotor Current in, 165
Apparent Resistance of Squirrel-cage Circle Ratio,
Rotor, 168 Bibliography of, Papers on, 153
Armature (in) Induction Generator, 222
Inference, 34et seq; 43 et seq; 57 Kierstead's Formula, 152, 153
Relation between Theta and, 52 (in) Slip-ring Induction .Motor,
Magnetomotive Force, Axis of, 45, 200
46 (in)Squirrel-cage Induction Mo-
Reaction with Short-circuited Arm- tors,Estimation of, 141 et seq
ature, 57 Formula for estimating, 152,
Resistance, Estimation for Syn- 153
chronous Generator, 21 Values of, Table, 150, 151
Strength, see Armature Interfer- Circulation of Air, see Ventilation
ence Compensator
Auto-transformer, Connections for Connections for starting up an
starting up Induction Motor by Induction Motor by means of,
means of, 170 170
Axis of Armature Demagnetization, Step-up, for Induction Generator,
45,46 223
Concentrated Windings, Inductance
Bibliography of Papers on Calculations in, 48, 49
Circle Ratio, 153 Conductors
Polyphase Generators 246-251, Copper, Table of Properties of,
Polyphase Motors, 252-255 Appendix, 257
Breakdown Factor in Induction Rotor in Induction Generators,
Motors, Determination of, 186- 214, 215, 217-219 in Induction
191 Motors, 167
259
260 INDEX
Conductors Continued Diameters of
(in) Slip-ring Induction Motors, Squirrel-cage Induction Motors,
196-199 136
(in) Squirrel-cage Induction Synchronous Generators, Tabula-
Motors, 113-117, 127 tion of 28, 31, 37
,

(in) Synchronous Generators, 9, 10 Distributed Field Windings, Poly-

Constant Losses in Synchronous phase Generators with, 99 et seq
Generators, 94-96 at high Dynamic Induction, Discussion of,
speeds, 99-101 12, 13
Copper Conductors, Table of Pro-
perties of, Appendix, 257 Eddy Current Losses in Rotor Con-
Core ductors as affecting the Torque,
Densities in Induction Motors, 135 168, 169, 177, 192, 194
Length in Efficiency (of)
Induction Generators, 218 Curves, 97, 181, 192, 193
Induction Motors, 108 Dependence of on Power-factor of
Synchronous Generators, 6, 26 Load, 97, 101, 102
Loss in Induction Generator, 219
Induction Motors, Data for esti- Squirrel-cage Induction Motors,
mating, 158 et seq Methods of Calculating, 178 et
Synchronous Generators, 92, 99, seq.; 192, 193. Values for, 115,
100 116
Cosines, Table of, Appendix, 256 Synchronous Generators, Methods
Cost, see Total Works Cost of Induc- of Calculating, 95-98; 101, 102
tion Motors End Rings in
Crest Flux Density in Induction Induction Generators, 218, 219
Motors, 121 Squirrel-cage Rotor, 174-176 Use
Critical Speed of Vibration, 224 of Magnetic Material for, 206,
Current 207
Density suitable for
Induction Motors, 127, 128 Energy, Motor Transformer of, 160
Synchronous Generators, 9, 10 Equivalent Radiating Surface at Air-
(in) End Rings of Squirrel- cage, gap, Data for estimating, 181
176 Equivalent Resistanceof Squirrel-cage
Ideal Short-circuit, 154, 200 Induction Motor, 167, 168, 194
Magnetizing of Induction Motor, Examples for Practice in Designing
139, 140 Polyphase Generators and Mo-
(at) Rated Load tors, 225-245
Squirrel-cage Induction Motor, Excitation
Values for 115, 116 (of) Induction Generators supplied
Synchronous Generator, Estima- from Synchronous Apparatus on
tion, of, 8 System, 213
Curves Loss in Synchronous Generators,
Efficiency, 97, 181, 192, 193 93-96; 99, 100
Excitation Regulation, 71-73 Pressure for Synchronous Gener-
Power-factor, 192, 193 ators, 84 et seq.
Saturation, 74 No-load Satura- Regulation Curves, 71-73
tion, 37-39
(of) Slip, 194 Field, A. B. on Eddy Current Losses
Short-circuit for Synchronous Gen- in Copper Conductors, 168, 210
erators, 76-78 "Field Effect" for improving Start-
Speed, 194 ing Torque in Squirrel-cage In-
Volt-ampere, 75, 83 duction Motors, 168; 210-212
 INDEX 261

Field Excitation, 45 et seq. Calcula- Half-coiled Windings, 18-20; 41-43

tions for Synchronous Generator,
57,58 I.E.E., see Institution of Electrical
Field Spools, Design for Synchronous Engineers
Generator, 84-92 Ideal Short-circuit Current, 154; 200
Flux per Pole, Estimation of in Impedance, 76, 77
Synchronous Generator, 18-22 Inductance of Armature Windings of
Formula Synchronous Generators, 45-50;
Circle Ratio, Kierstead's Formula 53
for estimating, 152, 153 Induction Generators, Design of, 1;
(for) Current in End Rings of 213-224
Squirrel cage, 176 Derivation of Design from Design
(for) Equivalent Radiating Surface of Synchronous Generators, 216
at Air-gap, 181 222
Field Effect, 168 Speeds, Appropriateness for ex-
(for) Mean Length of Turn, 21, ceedingly high, 213
156, 199 Synchronous Generators versus,
Output Coefficient, 6, 7, 108 209, 210, 213
Peripheral Speed, 113 Ventilating, Methods of, 220. 221
Pressure Induction Motors, 1
Discussion leading up to Deriva- Slip Ring, 195-201
tion of, 13, 14 Squirrel-cage, Design of, 105 et seq.

(for) Squirrel-cage Induction Magnetic End Rings, Use of,

Motors, 119 206, 207
Winding Pitch Factor in, 16, 20 Open Protected Type, 186
Reactance, 48 Slip Ring, Discussion of the rela-
Theta, 51 tive merits of Squirrel-cage
Total Works Cost, 111 and, 195-201
Two-circuit Armature Winding, 197 Squirrel-cage Design, 173-178
Fractional Pitch Windings, 16-18 Synchronous Motors versus, 202-212
Friction Losses in Inherent Regulation, 44; 54; 55; 66;
Induction Motors, Data for esti- 71; 80-82
mating, 166 Institution of Electrical Engineers,
Synchronous Generators, 93-96; at Bibliography of papers on Poly-
high speeds, 99, 100 phase Generators, 246-251 on
Full Load Power-factor, Estimation Polyphase Motors, 252-255
of, in Squirrel-cage Induction Insulation (of)
Motors, 155 Field Spools, 88, 89
Full Pitch Windings, 14-18; 120, 197 Lamination of Induction Motors,
126
Generators Slot (in)
Induction, see Induction Genera- Slip-ring Induction Motors, 196;
tors 200
Synchronous, see Synchronous Squirrel-cage Induction Motors,
Generators 114; 128-130
Synchronous versus Induction, 209, Synchronous Generators, 10, 11
210, 213
Goldschmidt, Dr. Rudolf, on Power- Kierstead's Formula for Circle Ratio,
factors of Induction Motors, 202 152; 153
Gross Core Length in
Induction Generators, 218 Lap Winding, 18; 19; 197
Squirrel-cage Induction Motors, 108 Leakage Factor, 22; 23; 42. See
Synchronous Generators, 6 also Circle Ratio
262 INDEX
Losses (in) Mean Length of
Squirrel-cage Induction Motors, Magnetic Circuit in Induction
155 et seq.; 166; 178 et seq. Motors, 121 in Synchronous
Synchronous Generator, 93-98 Generators, 31
Effect of High Speed on, 99-101 Turn of Winding, 21
in Slip-ring Induction Motors,
Magnet Core, Material and Shape 199
suitable for 2500 kva. Synchro- in Squirrel-cage Induction Mo-
nous Generator, 23-25 tors, 155-157
Magnet Yoke, Calculations for Syn- Metric Wire Table, Appendix, 257
chronous Generator, 29 Motors, Induction, see Induction
Motors
Magnetic Motor is Transformer of Energy, 160
Circuit
Squirrel-cage Induction Motors Net Core Length in Synchronous
Design, 120 et seq. Generators, 26
Magnetomotive Force, Esti- No-load
mation of, 124; 134 Current of Induction Motor, 140
Mean Length of, 121 Saturation Curves, 37-39 In-
Sketch, 137; 138 fluence of Modifications in, 78-
Synchronous Generators 83
Design, 22 et seq.

Magnetomotive Force, Esti- Open Protected Type of Squirrel-cage

mation of, 32 et seq. Induction Motor, 186
Mean Length of, 31 Output Coefficient
Data for Teeth and Air-gap in In- Formula, Discussion of Significance
duction Motors, 134 of, 6; 7
End Rings, Use of in Squirrel-cage Squirrel-cage Induction Motors,
Motors, 206; 207 Values for, 108; 109
Flux, 12 et seq. Synchronous Generator, Table of
Distribution in Induction Mo- Values for, 5; 6
tors, 122-125 Output from Rotor Conductors in
Estimation of Flux per Pole, for Induction Motors, 167
Synchronous Generators, 18-
22 Partly Distributed Windings, In-
Materials, Saturation Data of va- ductance Calculations, 48, 49
rious, 32; 33 Peripheral Loading, Appropriate Val-
Reluctance of Sheet Steel, 137 ues f.or Squirrel-cage Induction
Magnetizing Current of Induction Motors, 113-115 for Synchro-
Motor, 139; 140 nous Generators, 8
Magnetomotive Force Peripheral Speed of
Axes of Field and Armature, 45; Squirrel-cage Induction Motors,
46 113
Induction Generator Calculations, Synchronous Generators, 25
222 Pitch
Induction Motor Calculations, 134 Polar, see Polar Pitch
per Cm. for various Materials, Rotor Slot in
Table, 33 Slip-ring Induction Motors, 196
Synchronous Generator Calcula- Squirrel-cage Induction Motors,
tions, 32 et seq, 54 174; 196
Tabulated data of, 32; 34; 38; 42; Slot, see Tooth Pitch
58; 61; 64; 65; 67; 69; 73; 74; Tooth, see Tooth Pitch
137 (of) Windings, 16-18; 120; 197
 INDEX 263

Polar Pitch, Suitable Values for Rotor (of)

Squirrel-cage Induction Motors, Induction Generators, 223; 224
106; 107 C&nductors in, 214 et seq
Synchronous Generators, 4; 5 Slots in, 217
Poles, Data for number of in Slip-ring Induction Motor,
Squirrel-cage Induction Motors, Slot Pitch, 196
106 Slots, 195-197
Synchronous Generators, 3 Windings for, 195; 197-199
Power-factor Squirrel-cage Induction Motor
Curves, 192; 193 Conductors,
Efficiency, Dependence of on P.F. Eddy Current Losses as affect-
of Load, 97; 101; 102 ing the Torque, 168; 169;
Induction Motors versus Synchron- 177; 192; 194
ous Motors, 202-212 Output from, 167
Saturation Curves, Estimation of Core
for various, 55 et seq Densities in, 135
Squirrel-cage Induction Motors, Loss in, 158; 159; 164-166;
Estimation of, 115; 116; 155; 193; 194
191-193 Material for, Choice of; 158
Pressure Resistance, 167; 168
Formula Slots
Discussion leading up to deri- Design of, 132
vation of, 13; 14 Number, 173
(for) Squirrel-cage Induction Pitch, 174; 196
Motors, 119 Squirrel-cage, Design of, 173-
Winding Pitch Factor in, 16; 120 178
Regulation, Method of Estimating,
39 et seq Salient Pole Generator, Calculations
Total Internal of Synchronous for 2500 kva., 3 et seq
Generator, 53 Derivation of Design for Induction
Generator from; 216-222
Radial Depth of Air-gap (for) Specification of, 40
Induction Motors, 133 see also Synchronous Generator
Synchronous Generators, 36; 37 Saturation
Radiating Surface at Air-gap, Data Curves for Synchronous Generator,
for Estimating, 181 55 et seq; 74
Ratio of Transformation in No-load Curves, 37-39 Influ-
Induction Generators, 217; 223 ence of Modifications of, 78-83
Slip-ring Induction Motors, 199 Data of various Magnetic Mate-
Squirrel-cage Induction Motors, rials, 32; 33
174 Sheet Steel, Magnetic Reluctance of
Reactance of Windings of Synchro- 137
nous Generators, 48 Short-circuit Curve for Synchronous
Reactance Voltage, 53 Determina- Generators, 76-78
tion of Value for Synchronous Sines, Table of, Appendix, 256
Generators, 48-51 Single-layer Windings, 18
Regulation, Excitation for Synchro- Skin Effect to improve Starting
nous Generators, 71-73 Torque of Synchronous Motors,
Resistance of Squirrel-cage Rotor, 207-209
167; 168; 194 Slip, 106; 163; 164; 192; 194; 215
Robinson, L. T., Skin Effect Investi- Slip-ring Induction Motor, Discus-
gation on Machine-steel Bars, sion of relative merits of Squirrel-
208; 209 cage and, 195-201
264 INDEX
Slot-embedded Windings, Inductance Stator Continued
and Reactance of, 48-50 Core of Continued
Slot Synchronous Generators, 27; 28
Insulation in Weight, 92
InductionMotors, 114; 128- Current Density suitable for Induc-
130; 196; 200 tion Motors, 127; 128
Synchronous Generator, 10; 11 PR Loss in
Pitch, see Tooth Pitch Induction Motors, 155-157
Space Factor, 11; 131 Synchronous Generators, 93-96
Tolerance, 127; 196 Slot Pitch, Values for Induction
Slots Motors, 118
Rotor, for Slots in
Induction Generators, 217 Induction Generators, 217
Slip-ring Induction Motors, 195- Induction Motors, 119; 127;
197 131
Squirrel-cage Induction Motors, Synchronous Generators, 11
132; 141; 173; 174; 196 Teeth, Data for Induction Motors,
Stator, for 121-127
Induction Generators, 217 Steam-turbine Driven Sets, Rotors
Induction Motors, 117-119; 127; with Distributed Field Windings
131; 141 for, 99 Circulating Air Calcu-
Synchronous Generators, 9; 11 lations, 102-104
Space Factor (of) Step-down Transformers for small
Field Spools of Synchronous Gen- Motors, 114
erators, 88; 89 Step-up Transformer for Induction
Slot, 11; 131 Generator, 223
Specification of 2500 kva. Synchro- Synchronous Generators, 1
nous Generator, 40; 41 Distributed Field Winding Type,
Speed 99 et seq
Control, Methods of providing, Efficiency, Dependence of on
195 Power-factor of Load, 97; 101;
Curves, 194 102
High-speed Sets, Characteristics Induction Generator versus, 209;
of, 99 210; 213
Synchronous versus Induction Mo- Salient Pole Type
tors for Low and High, 202-212 Calculations for 2500 kva., 3 et
Spiral Windings, 18-20 seq
Spread of Winding, 14-16; 120 Derivation of Design for Induc-
Spreading Coefficients, 36 tion Generator from, 216-
Squirrel-cage Induction. Motor, see 222
Induction Motor Specification, 40; 41
Starting Torque of Squirrel-cage Synchronous Motors, 1 Induction
Motor, 169-172 Motors versus, 202-212
Stator
Conductors, Determination of Di- Tabulated Data of Magnetomotive
mensions for Induction Motors, Force Calculations, 32; 34; 38;
127 42; 58; 61; 64; 65; 67; 69; 73;
Core of 74; 137
Induction Motors Tabulation of
Density, 135 Losses and Efficiencies in Squirrel-
Loss, 158; 159 cage Induction Motors, 180
Material preferable for, 158 Squirrel-cage Induction Motor Di-
Weight of, Estimation, 159 ameters, 136
 INDEX 265

Tabulation of Continued Variable Losses in Synchronous Gen-

Synchronous Generator Diameters, erators, 94-96
28, 31: 37 Ventilating Ducts for
Tangents, Table of, Appendix, 256 Induction Motors, 125; 126
Teeth, Stator, in Induction Motors, Synchronous Generators, 25; 26
121-127 Ventilation of
Temperature Rise, Data for estima- Induction Generators, Methods
ting, 90; 181-183 suitable for, 220; 221
Theta and its Significance, 51 et seq Synchronous Generators with Dis-
Thoroughly Distributed Windings, tributed Field Windings, 102-104
Inductance Calculations, 48; 49 Vibration, Critical Speed of, 224
Tooth Volt-ampere Curves, 75, 83
Densities in Induction Motors, Voltage
121-125 Formula, see Pressure Formula
Pitch 26; 118; 196 Regulation, Method of Estimating,
Torque, 162; 195 39 et seq
Eddy Current Losses in Rotor
Conductors as affecting, 168; Watts per Ton for Squirrel-cage
169; 177; 192; 194 Induction Motor, 183-186
Starting (of) Weight of
Squirrel-cage Induction Motor, Induction Motors, 110; 111; 183-
169-172 185
Synchronous Motors appropri- Stator Core of Induction Motors,
ate for high, 204 et seq 159
Torque Factor, 167 Synchronous Generators, 7
Total Net Weight of Whole-coiled Windings, 18-20
Induction Motors, 110; 111 Data Winding
for estimating, 185 Pitch, 16-18; 120; 197
Synchronous Generators, 7 Pitch Factor, 16; 120
Total Works Cost of Induction Spreads, 14-16; 120
Motors, 158 Methods of esti- Inductance and Reactance of, 45-
mating, 111-113; 118 49
Transformers Types of, 16-20; 41-43
Auto, for starting up Induction Distributed Field, Polyphase
Motors, 170 Generators with, 99 et seq
Step-down, for small Motors, 114 (for) Slip-ring Induction Motors,
Step-up, for Induction Generators, 197-199
223 Wire Table, Metric, Appendix, 257
Two-layer Winding, 18; 197
 THIS BOOK IS DUE ON THE LAST BATE
STAMPED BELOW
AN INITIAL FINE OF 25 CENTS
WILL BE ASSESSED FOR FAILURE TO RETURN
THIS BOOK ON THE DATE DUE. THE PENALTY
WILL INCREASE TO SO CENTS ON THE FOURTH
DAY AND TO $1.OO ON THE SEVENTH DAY
OVERDUE.

JUL 2 195T-P

SEP 5

J.I) 21-20m-6,'32
 YC 33501

257829