A Text Book of Electrical Machinery

Qlarnell
Hmoerattg
ffiihratg
Stifuta,
l^tm Unrfc
THE
ALEXANDER GRAY MEMORIAL

LIBRARY
ELECTRICAL ENGINEERING
Tii e ai rx
ar
Cornell University Library
TK 21S2.R98
A text-book of electrical machinery .v.
1
3 1924 004 407 536
Cornell University Library
The
tine
original of
tiiis
book
is in
Cornell University Library.
There are no known copyright
restrictions in
text.
the United States on the use of the
http://www.archive.org/details/cu31924004407536
A TEXT-BOOK
OF
Electrical Machinery.
VOLUME
I.
ELECTRIC, MAGNETIC, AND ELECTROSTATIC CIRCUITS.
HARRIS
Member
the
J.
RYAN,
M.E.,
of
Professor of Electrical Engineering, Leland Stanford Jr. University;

of the
American Institute of Electrical Engineers and American Society of Mechanical Engineers, etc.
HENRY
H.
NORRIS, M.E.,
Professor of Electrical Engineering, Sib leyCollege, Cornell University; Member of the American Institute of Electrical Engineers and of the Society for the Promotion of Engineering Education, etc.
GEORGE
Associate
L.
HOXIE, M.M.E.,
Ph.D.,
Consulting Electrical Engineer;
Member of the American Institute of Electrical Engineers and Member of the American Society of Mechanical Engineers,
FIRST EDITION.
SECOND THOUSAND.
NEW YORK: JOHN WILEY & SONS. London: CHAPMAN & HALL, Limited.
Copyright, igo3,
BY
HARRIS
J.
RYAK.
ROBERT DDUMMOriD, PRJNTER, NEW YORK
PREFACE.
The
student of electrical engineering comes to the technical

of his course well grounded in the prinin the rela-
and professional part

ciples of elementary
and applied mathematics and

of
tions
and
is
characteristics
physical
phenomena.
His
next
task
to
leam
to apply this training to the
working principles
of engineering, both those underlying the design and operation of electrical machinery and those
upon which general engineering
methods are based.
With
these facts in
mind
it
has been found
desirable to produce a
text for the
purpose of communicating to
the student the working principles mentioned above and to pre-
pare him for reading profitably the literature of his profession.
The book has been

time
it
designed as a distinctively engineering
text,
not as a work on physics or apphed mathematics.
At the same
has been found desirable to restate in engineering terms
the elementary laws and principles of those sciences which bear

directly
upon the subject
in hand.
As a
it
result of experience in teaching electrical engineering
has been found most satisfactory, both in maintaining the
interest of the student to found the treatment

circuit,
and
in economizing his time
and energy,
upon
the laws of the alternating-current
from which the

follows
treatment of continuous-current phe-
nomena
naturally.
The
application
It
of these
laws
is
illustrated
by means of a few problems.
has not, however.
IV
PREFACE.
1;o
been the purpose
make
this
a problem book, and the teacher

for class
and student should prepare additional problems
and
home
trical
use.
In this part of the work, books of the nature of " Elec-
Problems," by Hooper and Wells, will be found of service.

I covers the laws of the electric, magnetic,
Volume
tural
and
elec-
trostatic circuits in
such a manner that the analyses of the struccharacteristics
and performance
of electrical machinery,
treated in
Volume
II,
may be
easily followed.
all
The
authors express obhgations to
contributors to the
literature of the profession.
Whi
e the
method of no one author

all
has been followed, the aim has been to profit by the work of
and
to provide
an
in'.roductory text
wherewith
to
prepare the
pro-
student to secure most profitably

fessional text, laboratories, lectures,
further training from
and the unhmited sources

literature
of electrical-engineering personnel
and
The
material of
Volume
when amphfied with
additional
thor-
problems, preferably taken from practical cases,
may be
oughly covered in
fifty
recitations, the
two volumes being de-
signed to provide profitable work for approximately one hundred

class exercises.
The form
of the material in this

its
volume
is
the result of several
years of experience in
in Cornell University.
Ithaca,
use as a text for the instruction of classes
New
York,
Sept.
i,
1903.
COMTEISITS.
PAGB
Chapter
I.
Electricity and Magnetism.
Electricity
and
Electrical
Energy
Electromotive Force
z 5
Magnetism
Chapter
II.
Fundamental and Derived Units.

12
15
in Electric Circuits
19.
Fundamental Units
Derived Units
Power Consumption Problems in the Use
of Electrical Units
20
Chapter
III.
Periodic Curves.
25 31 42 45
Properties of Sine Curves
Combinations of Sine Curves The Fourier's Series for an Alternating Quantity Analysis of a General Periodic Curve
Chapter IV.
Complex Quantities.
60
65
Vectors Alternating Quantities and Vectors
Chapter V.
Laws of the Electric
Circuit.
Consumption of e.m.f. in Single Circuits Problems in Single Series Circuits Problems in Simple Multiple Circuits Consumption of e.m.f. in Series-multiple Circuits
68
76
80
82
Chapter VI.
Electric Power.
92
Function of the Electric Circuit Power with Current and e.m.f. in phase Power with Current and e.m.f. in quadrature Power with Current and e.m.l neither in phase nor in quadrature Average Power with non-sine form e.m.f. and Current
93
95
96
99 100
The
Equivalent Sine
Wave
V
VI
CONTENTS.
PAGE
Chapter VII.
Circuit.
Magnetomotive Force and the Laws of the Magnetic

105 112
122
M.m.f. and the Magnetization Curve Matters Affecting Permeability Reluctance of the Magnetic Circuit Magnetic Hysteresis Ewing's Theory of Magnetism Illustrative Problems in Magnetic Circuits
125
132
13S
Chapter
VIII.
Rotating Magnetic Fields.

142
145 146
of the Rotating Pivot Field
Polyphase c.m.f .s, Currents and Fields
Components
Production of a Rotating Pivot Field Components of the Rotating Cylinder Field
Production of Rotating Cylinder Fields
150 162
Chapter IX.
The
Electrostatic Field.
Corona
173 180
General Characteristics
The
Electrostatic
Dielectric Thickness to avoid Dielectric Hysteresis Dielectric Conduction
Corona
189
ipS 196
Problems on the Electrostatic Field
198
Chapter X.
Losses in Electric Circuits.

199 200
207 213 216 217 223 230
Sources of Circuit Losses

Resistance
Inductance Skin Effect in Conductors Eddy Losses in Conductors Eddy Losses in Magnetic Circuits Capacity of Transmission Lines and Cables
Magnetic and Dielectric Hysteresis and Dielectric Conduction
Appendix
iiJ
TABLE OF IMPORTANT SYMBOLS AND

ABBREVIATIONS.
B
C
density of magnetic flux or induction.

electrostatic capacity.
e.m.f... electromotive force in volts.
E
e
e.m.f. effective value.
e.m.f. instantaneous value.
F
f.
mechanical
force.
frequency in cycles per second.
H
/
i
magnetomotive force
in gilberts.
current in amperes, effective value.

current, instantaneous value.
/
L
pL
-t/"^^-
inductance in henrys.
m.m.f. .magnetomotive force.
magnetic permeability.
electric
. . . .
P
$
Q.
,
power.
flux.
total
magnetic induction or
quantity of electricity.
electric resistance.
r
(R t
magnetic reluctance.
time in seconds. angle of phase difference.
electric
W
X
z
energy or work.
reactance.
impedance.
used.
This table contains only those symbols and abbreviations which are frequently Those which are used locally only are explained when used.
ELECTRICAL MACHINERY.
CHAPTER
I.
ELECTRICITY AND MAGNETISM.
1.
Electricity
and
electrical energy.
2.
Electromotive force.
Three methods
for maintaining
an e,m.
Measure-
ment
3.
of e.m.f.
Magnetism: a. Magnetomotive
b.
force.
4.
Magnetic flux. Water flow, t. Tension of the magnetic field. d. Other hydraulic analogies to magnetism. Magnetic tension and flux density.
I.
Electricity and Electrical Energy.
Electrical
pheare,
nomena
are manifestations of molecular action.
There
unfortunately,
no means available
Electricity
for
observing the exact
character of the
molecular mechanisms upon which these

must,
for
phenomena depend.
this
reason,
be
studied like heat, light, chemical energy, and other forms of
molecular energy, that
is,
by
its
effects.
By
observation of
the' results of the operation of
molecular forces, as manifested
in
mass motion or in chemical action, some idea of the forces
can be gained.
As
implied in the preceding statement, electricity
is
a form
of energy.
This energy has the same character as has com-
[2
plete
mass motion, which
is
shown by the
fact that
it
readily transformed into
any other kind of energy.
may be It may
suitable
also be transferred from point to point
by the use of
molecular kinematic connection, and
it is
this ability to transfer

it'
power without mass motion which makes

carrier of
electrical
is
the only successful

transformation of
this action
energy oyer long distances.

energy
is
The
electrical
work, and the rate of
electrical
power, just as in the case of mass motion.
Electrical energy
may be
stored in the production of

is
mag-
netism just as mechanical energy

velocity of a mass.
It
stored in accelerating the

its
may
be recovered from
stored form.
Likewise electrical energy becomes potential energy

electric
when an
This
charge
is
is
taken up by capacity, just as mechanical

elasticity in
energy
stored
is
by
compressing a spring.
phenomenon
also reversible.
These
illustrations point to
the identity of electrical and mechanical energies, and the
important points
in the
study of electrical engineering are:
(i) the conservation

electricity.
of all energy; (2) the energy character of
2.
initial
Electromotive Force.
Electromotive
force*
is
the
cause of the electric current and of electrostatic attrac-
tion ^
This e.m.f.
follows
a.
i>'.
may
be maintained by one of three methods as
Thermo-electric.
Chemico-electric.
c.
Dynamo-electric.
the junction of two metals
is
a.
When
heated, an e.m.f.
is
produced, the value of which depends on the metals and the

temperature to which their junction
is
heated.
This e.m.f.
may
be used to cause a flow of
electric current
by connecting
* Usually. written e.m.f.

the unheated terminals and thus part of the applied heat
energy may be
utilized electrically.
is
But practically the portion

this lack
is
of heat so transformed
very small, and because of

very
little
of economy the thermal couple
used.
Fig.
Fig.
I.
Cox Thermo-electric Generator,

It
represents a commercial form of the electric' thermo-pile.

is
known
b.
as the
Cox
generator.
in metallic contact, are placed
If
two unlike metals, not
in a bath of
some
liquid
is
which attacks one of them more than
up between the metals, and by suitable connection outside the liquid an electric current may be
the other, an e.m.f.
set
produced.
This chemicail generation of current has
its
prac-
tical application in the primary battery which has an important
place in small work.
Fig. 2 shows a form of primary cell
which
c.
is
in
common
third
use.
for the generation of
The
is
method
an electromotive
it
force
the all-important one to the engineer, and
consists
in
in the application of the principle that a wire
moved
magnetic
field in
such a direction as to cut across the magnetic
flux
b
an electromotive
its
of the field will have
produced
in
it
force the value of
which
will
depend on the length of wire,
Fig. 2
A Typical
Primary
Cell.
Fig. 3
Simple Dynamo-electric Machine,

field
velocity,
trates
and the strength of the
cut
for
by
it.
Fig. 3 illus-
experimentally this method
electro-mechanically
developing an e.m.f.
This principle
is
used in the construction
3]
ELECTRICITY
all
AND MAGNETISM.
of
dynamo-electric machines which form the main means
for the conversion of

vice versa.
mechanical into electrical energy, and
Measurement of
cated and
its
e.in.f.
is'
The
indi-
presence
of an e.m.f.,
or
difference of potential,
amount may be
elec-
measured by means of an
trostatic voltmeter.
This instru-
ment, which
is
also
is
known
in
as
an electrometer,
commercial form
shown
a
It
is
in Fig. 4.
utilizes the facts that
there
repulsion or attraction between
two
electrically
charged bodies
and that two bodies may be

charged by connection to the
terminals of a circuit in which
a difference of potential exists. In

the
measurement of low
electrical pressures * the elec-
trometer plates constituting the
charged bodies are numerous

in order that the loss of attrac-
tion
due to the low pressure

be made up.
is
may
In this form
Fig.
4..
the instrument
known
as a
-Multicellular Electrostatic Voltmeter.
multicellular voltmeter.
3.
net.
Magnetism.
It
is
In Fig.
5,
NS
is
a permanent bar
mag-
made
of hardened crucible steel
and has been

has
magnetized through some natural means, such as contact with

another bar magnet, or with a piece of loadstone, or
it
been placed
*
in a solenoid carrying
an
is
electric current.
often used for e.m.f.
When
The
expression " electric pressure "
this
bar
is
remote from other magnetic substances and the

it is
immediate region about

a magnetic
field is
field,
examined with a small compass,

This,
flux,
such as that illustrated, will be found.
due to magnetism or magnetic
which emanates,
from
lines
one end of the bar and returns to the other as the

show.
Extensive experimental researches conducted
by
physicists
have led to the following conclusions
in
this;
connection
KiG.
5.
Magnetic Field surrounding a Bar Magnet.
Fig. 6.
Hydraulic Model of the MagBar Magnet.
netic Field of a
a.
The magnetism about

Such m.m.f.
is
the bar
magnet
is
due to a mag-
netomotive force (m.m.f.) that resides in the molecules 6f the

bar magnet.
proportional to the length of the
of molecules
bar, and, therefore, to the
number
which consti-
tute a single filament of the bar.

d.
m.m.f. sets up a difference of magnetic pressure
3]
between the two ends of the bar, which causes magnetism or magnetic flux to be established from one end of the bar to the
other.
In Fig. 6
is
an
illustration of
an hydraulic model of a
of the
permanent bar magnet.
The model
represents
all
magnet proper
is
M,
located in a vessel of comparatively large size containing
water.
The
part
the
bar
metal tube perforated on
sides with
magnet. It is a numerous small holes.

furnishes
rotating
shaft
carrying screw
propellers,' _/5!5!7")
the model with within the tube.
a water-motive force directed from
to
This force corresponds to the magneto-
motive force that resides in the molecules of the bar magnet.
When
the propellers, ffff, are set in motion, the water will
enter the tube at
S and go
out at N.
The
flux of this
water
represents the magnetic flux.
The
lines
drawn
all
in the figure
represent the direction of the water flux at
points in the
immediate region of the tube, while the space between these

lines
is
a measure of the cross-section over which a definite
rate of water flux occurs.
In the same manner, then, as in this model the lines in
show at once the direction of magnetic amount at any point in the region of the magnet.
Fig.
5
c.
flux
and
its
At
is
all
points within
the field of flux there exists
mechanical force related to the magnet.

force
The
nature of this
as follows
its
tension along
The magnetic flux own direction and

is
possesses a mechanical
a mechanical pressure
everywhere at right angles to the direction of the magnetic

flux.
The
entire field of flux

it
rigidly attached to the

it is
mag-
netic body from which

It
emanates or by which
established.
has been found by experimental

is
means, to be described
later, that this tension
proportional to the square of the rate

point.
of magnetic
flux
is
at
any
The
it
lateral
pressure
of
magnetic flux
exists along
so intimately associated with the tension that
its
own
direction that
is
hardly necessary to
It is
[4
distinguish between the two.

their difference, however, to keep in
necessary to keep in mind necessary in meclianics
much
as
it is
mind the
difference
between action and reaction.
All engineering problems in magnetism are solved in terms of

the tension of the magnetic flux.
d.
Physical experience has
shown
further
that magnetic
flux
is
established in a closed circuit just as the water currents
in the hydraulic
model flow
in closed circuits; i.e.,
whatever
amount of water current

passes out at the
enters the
5 end
of the tube also
N end.
for the
An
entirely analogous property has
been found to exist
magnetism about a magnetic body.

is
Whatever amount
opposite pole.
of magnetic flux
emitted from one side or

re-enters at the
in
end, called pole, precisely that
same amount
is
This amount
everywhere
existence en
route from one pole to the other.

flux within the
Of
is
the nature of magnetic
magnetic body
little
known.
Researches,
however, have long since proven that within the magnetic
body there
flux
exists a state of things corresponding to a con-
tinuity of the
is
magnetic flux whereby a complete

is
circuit of
such
always established, just as
the case with the electric
current in the closed electric circuit.
Because our knowledge
of the state of things within the magnetic

point, that
body ceases
at this
which completes the
circuit of
magnetic flux within
and through the magnetic body is called induction. In Fig. 7, 4. Magnetic Tension and Flux Density. AAA is a bar of soft wrought iron formed as shown, and
the bar magnet

of
as
NS
is
mounted
with
in front
it
AAA
shown
and
in
in contact
much
of
the figure.
Under these
evidences
circumstances
Fig.
almost
all
7. Magnet and
^^^^'^^
magnetic flux will be found to have disappeared

in
the region that

the
surrounds
^5.
This
is
due to the
fact
that
magnetic flux has
4J

It
disappeared as induction in the soft wrought iron.
has
been found that but a very small amount of magnetic pressure

is
consumed
in
maintaining the induction in the iron bar, while
the balance of the magnetic pressure generated by the m.m.f.

that resides in the steel bar
is
used up point by point, simul-
taneously with
the steel bar
its
origin, in maintaining the induction
through
NS.
Thus
results
it is
that practically no difference of

for
magnetic pressure
along NS, which accounts

flux.
It is true, too,
the
disappearance of the magnetic

total
that the
amount of induction which

circuit
exists at
any
cross-section of
as that
the
NSAAA
is
practically
the
same
which
exists at
any other
cross-section.
In Fig. 8 the magnet

S,
NS
is
mounted on a knife-edge
at
and the end
is
suspended from the stirrup of a scale-beam
Fig.
8.Apparatus
for Studying the
Tension due to Magnetic Induction.
The mounting ofiV5is with another knife-edge as shown. air-gap, ag, is formed separating so adjusted that a small
from
The
following facts
may be
observed experimentally
with this apparatus
10
1,
[+
The
existence
of induction
in
NS
and
AAA,
and a
corresponding amount of magnetic flux across the air-gaps at
N and S,
2
will at
once be shown by the tension registered on

proportional to the square of the rate at
distributed over a given cross-section.
the scale-beam.
The tension
is is
which magnetic flux
An
indication of the
amount of
flux
from the pole
N into
given
the face of the soft iron armature or keeper,
AAA,
is
by the throw of the galvanometer needle when the turn of wire, T, is drawn away from the position showji in the figure
so as to cut the magnetic flux.
flux
is
The
tension of the lines of
weighed on the scale-beam and found to be P.
Another
bar
is
now
used in the place of the one upon which the
Fig.
9.
Fig. 10.
Illustrating the Effect of
Change of Air-gap Area.
measurement has
just
been made.
The second bar has the

first,
same
its
cross-section and length as the

its
but instead of being

is
square in cross-section
horizontal thickness
one-half and
vertical thickness is twice the thickness of the bar that
has
in
just
been removed.
That
is, it
is
set
on edge
is,
as
shown
Fig. 10,
and the area of the air-gaps
therefore, one-half
that found in the preceding case, which
is
illustrated in Fig. 9.
4j
ELECTRICITY yiND MAGNETISM.
ir
The m.m.f.
same
of the bar has been adjusted so as to set up the
flux as before.
This
may be shown by
the kick given
by the galvanometer when the turn of wire,

flux at the air-gap. find
it
T, cuts across the
Upon weighing
the tension of the flux
we
to be 2F.
Note the significance of these experiments.
First experiment:
Total flux, ^.
Cross-section at air-gap,
A.
Flux
density,
B =
,
-r.
Observed magnetic pull, F.
Second experiment:
Total flux, $.
Cross-section at air-gap,
\A.
Flux
density,
B"
-r-r-
= 2B'
A
instead oiÂ,.s.nd
,
Observed magnetic pull, 2F.
Had
the area in this experiment been
had the magnetic density been maintained at B" = 2B' the magnetic pull would evidently have been 4F. Thus we find
when
remains constant,
Flux density = B' ; the magnetic pull " " " " =: 2B' ; "
If
= F. = 4.F.
'
the flux density had been increased in a third experi-
ment to 3^' we should have found the magnetic pull to be gF. Thus we learn experimentally that the contractile tension of
the magnetic fluk
density.
is
proportional to the square
of the flux
pull,'
The numerical
value of this magnetic

Its
F,
is
XB'\ where
is
constant arbitrarily chosen.

is
numerical
value in the centimeter-gram-second system
-^
07t
CHAPTER
II.
FUNDAMENTAL AND DERIVED

SYNOPSIS.
5.
UNITS.
Fundamental
a.
i.
c.
units.
The unit of magnetic flux. The unit of current. The unit of electromotive force.
Through dynamo and simple conductor. Through dynamo, conductor, and condenser. Through dynamo, conductor, and electrolytic
units.
6.
The
electric circuit.
u.
b.
c.
cells,
7.
Derived
a.
b.
t
.
d.
c.
f.
8.
The The The The The The
unit of resistance.
unit of inductance.
unit quantity of electricity.
unit of capacity. unit of power.
unit of energy.
Power consumption in electric circuits. 1 Power consumed by resistance. 2. Power consumed by counter e.m.f.
Problems in the use of electrical
units.
9.
5.
Fundamental Units.
action
is
The
much
dynamic character of elecin
tromagnetic
so
accord with
common
mechanical experience that the electromagnetic actions have
been chosen to form the basis

absolute electrical units,
for the definition of
a system of
often called the

is
centimeter-gram-
second
{c.g.s.) system.
This system
electrical
universally adopted
in electrophysics
and
in
engineering.
The
c.g.s.
units are usually of inconvenient magnitude and a system of
practical units
is
necessary.
_
The
practical units are arbitrary
multiples of the c.g.s. units.

12
5] a.
UNITS.
13
The unit of magnetic
flux.
unit,
is
Practical unit, the Maxwell, equal to the c.g.s.
the flux which will produce a tension in

I
its
own
directio7i
of
-H Stt dynes
when
distributed uniformly over one sq. cm. of
cross-section.
The name of this unit of magnetic flux The density of magnetic flux is the number
unit cross-section.
b.
is
the maxwell.
of maxwells per
The unit of current.

Ampere:
one-tenth of the c.g.s. unit.
it
Practical. unit, the
From experimental
fleld of
research
has been found

in
that
straight conductor carrying an electric current
a uniform
magnetic flux will be acted upon by a mechanical force

it
tending to move
It
at right angles to the direction of the flux.

is
has been found that this force
proportional to the length
of the conductor in the flux, the current strength, and the flux
density.
These
facts
form the basis
for defining the unit of
current strength as follows
One
unit of current in a wire located in
and at right angles
to
a uniform field of unit flux density will cause a mechanical force

af one dyne
This
is
to
be applied to each centimeter length of the conto the
ductor, at right angles both
flux and
to the
conductor
the c.g.s. unit.
One-tenth of
it
has been adopted
as a convenient unit for practical purposes.

practical unit of current
c.
The name
of this
is
the ampere.
unit of electromotive force. Practical unit, the Volt: one hundred million
generated in a conductor moved across a
that
is
The
c.g.s. units.
is
Experimental research has determined that an e.m.f.

field
of magnetic flux
proportional to the velocity, flux density, and length of

flux.
is
the conductor moving through the

facts the value of the unit of e.m.f.
On
the basis of these
determined by definition
as follows:
14
[6-
One unit of electromotive force
is
generated in each centimeter-
length of a conductor moving at a velocity of one centimeter
per
second through a uniform field of unit flux density angles both

to
and
at right
the flux
and
to
the length
of the conductor.
This
unit
tical
is
is is
a convenient definition, but the size of the resulting
inconveniently small for practical purposes.
The
pracsize,
unit that has been adopted, because of its convenient
one hundred million (lO^) times the

definition.
size of the c.g.s. unit as
determined by
the volt.
The name
of this practical unit
is
There are other
electric
and magnetic
units.
They
are
derived from the above fundamental units,
among which must

These derived
be included the unit of time, or the second.
units will be discussed in connection with the control of the

electric current 6.
and the magnetic

Circuit.
is
flux.
The
Electric
In
no case
is
electric action
possible unless there
a complete circuit over which a transfer

occur.
of an electric charge
may
This circuit must be com-
plete through the source of e.m.f. as well as

route.
a.
by an external
In Fig.
circuit
is
1 1
the
dynamo forms
by
a part of the conducting

e.-m.f.
is
electric
in
which the actuating

this e.m.f.
generated.
The
current
established
through the complete
circuit of the
dynamo and conductor
as indicated in the figure.
Fig.
1 1.
Simple Conductor
Fig.
12
Circuit.
Fig. 12.
Circuit containing Condenser.

be closed through the
CÎ3>
b.
In
the circuit
may
.dynamo, the conductor, and a condenser.

cumstances a rush of current, as indicated
at
Under these cirby the instrument

closed.
A,
will occur at the instant the circuit
is
As soon
as the strain in the dielectric of the condenser will no longer
7]
FUNDAMENTAL AND
DERIIÊD UNITS.
all
15
increase under the pressure of the actuating e.m.f.,
current
through the condenser

terminals are reversed
Fig-.
1
will
cease.
Now
if
the
conductor
3,
by some such means as illustrated in a momentary current will be shown on the instrument
and applying
it
while the process of relieving the dielectric strain in one

in the other
is
direction
going on.
Continued
reversal of the condenser terminals
by revolving the comimpulses
mutator
in Fig.
3 will cause a succession of current
to be established in the circuit.
^
Fig.
'I'I'I'I
1
13. Circuit containing Condenser with Commutator.

in Fig.
14,
Fig. 14. Circuit containing Electrolytic Cells.
c.
Again, as
the circuit
may
be established
through the dynamo, an

electrolytic cells.
electric conductor,
and one or more
In general the cells will transmit the current
with more or less irregularity through processes of internal
molecular changes
or
transfers.
Electrolytic
cells
do not
merely close an
electric circuit as
does a conductor or a con-
denser; they are generally sources of e.m.f. and thus modify

the current by subtracting from or adding to the source of
e.m.f.
7.
Derived Units.
in
Current
Control.
The
current
that
will
be set up
source
any
electric circuit
depends upon the value of

resistance,
the
e.m.f.,
and upon the values of the

circuit.
capacity,
and inductance of the
Capacity and induc-
tance present phenomena like resilience and mass motion in
mechanics.
a.
The unit of
resistance.
Practical unit, the

It is
Ohm:
10^ c.g.s. tinits.
found experimentally that
when
a current
is
established
in a
is
conductor at any constant temperature, electric pressure

in
consumed
direct proportion to the value of the current.
ELECTRICAL M/ICHINERY.
is
[7
This property of a conductor
called
electric
resistance.
Based on the above

is
fact,
the deiinition of the unit of resistance
as follows
One unit of resistance will consume one unit of pressure per

unit of current.
The
ohm
is
practical unit of resistance

volt
is
called the
ohm.
It
con-
sumes one
per ampere.
On
account of the ampere the

it
ten times and on account of the volt
is
a hundred
million (lo^) times, or a total of one thousand million (lo')

times, the magnitude of the c.g.s. unit.
b.
The unit of inductance.

sets
Practical unit, the Henry: lo^ c.g.s. units.
Every conductor carrying current

about
itself
up magnetic
flux
This has been established by experiment.

it
In
the same
way
is
learned that such flux cannot be brought
into existence
by the current without cutting across the conit is
ductor about which
established.
is
In doing so an e.m.f
is
generated in the conductor that

the magnetic flux cuts across
equal to the rate at which
it.
As
is
long as the current
is
changing, an e.m.f
direction
is
self-generated in the conductor.
The
oppose
of this self-induced e.m.f.
always such as
to
the corresponding current change.
The
process of setting up
it
magnetic
flux
about a conductor by the current
carries
is
called self-induction,
is
and the
ability to self-generate
an e.m.f.
given the
name
inductance.
The above
facts constitute the
basis for defining the unit of inductance as follows
circuit possesses one unit
of inductance when a unit rate
of change of current
Since this
is
in the circuit generates one unit
a derived unit, the

is
of e.in.f magnitude of the corre-
sponding practical unit
at
once determined by reference to

the the
the ampere and the volt.

practical unit of time.
The second always remains as The ampere is one-tenth of, and
volt one hundred million (lo') times the corresponding absolute
7]
UNITS:
i7
unit.
This would make the practical unit of inductance one

(lo') times the absolute unit
thousand million
definition.
determined
byis
The name
of this practical unit of inductance
the henry.
c.
The unit quantity of

a dielectric
is
electricity.
Practical unit, the Coulomb, one-tenth c.g.s. unit.
When
strain
is
subjected to electric pressure a definite
produced.
In the production of this strain a quantity
of electricity must be applied by transfer through the circuit.
A unit for quantity of electricity

One unit quantity of
transferred by one
-unit
is,
therefore, necessary.
is
electricity
equal
to
the quantity
of current
in one unit
of time.
Its corre-
This unit
is
derived from the fundamental units.

is
sponding
practical value
it is
the ampere-second.
On
account
of the ampere
one-tenth of the value of the absolute unit.

is
The
practical unit quantity of electricity
called the coulomb.
d.
The unit of
capacity.
^
Practical units, the Farad, IQ-
c.g.s. unit,
and
the Micro-
farad,
If
10-'^^ c.g.s. unit.
an
electric circuit
be closed through a condenser, some

circuit will
or
all
of the impressed pressure of the electric

dielectric.
be
taken up by the capacity of the
Experiment
reveals in this connection the following fact:

constituting the capacity will take up a
The
dielectric
quantity of electricity,
or electric charge, in proportion to the electric pressure applied
between
electric
its
faces.
As
this applied pressure is
changed, the
charge accepted by the
dielectric
is
correspondingly
is
changed.
The
rate of transfer of electric charge

is
the value
of the electric current by which the change
accomplished.
The
unit of capacity
is,
therefore, defined in terms of the units
of current, pressure, and time, thus A dielectric in an electric circuit has a capacity of unity
l8
the transfer through
it
[7
when
of unit current requires a unit

of
tlie
rate of change of the applied pressure.
The magnitude
of the
corresponding
practical
unit
capacity, therefore, becomes onq tenth on account of
ampere, and one hundred-millionth (lO"*) on account of the

volt,
making the
is
practical unit
one thousand-millionth (io~')

of
for
of the absolute unit.
capacity
the farad.
The name of this practical unit The farad is inconveniently large

so that condensers are
most
practical
purposes,
is
ordinarily
rated in a unit that
one millionth of the farad.

microfarad.
This sub-
stitute for the farad is called the

is,
The microfarad
therefore,
one million-thousand-millionth (lO~^') of the
absolute unit.
e.
The unit of power.

has been found by experiment that the power in any
is
Practical unit, the Watt: id' e.g. s. units.

It
part of an electric circuit

e.m.f. at
its
proportional to the product of the
terminals and the current present.
The
unit of
power
is
applied zvhen a unit of current
is
established by a unit of pressure.
The corresponding ampere. The name of
practical
unit of
power
is
the volt-
this unit is the

it
watt.
Being derived
one-tenth on
from the ampere and the volt
is,
therefore,
account of the ampere and one hundred million (lo^) times on account of the
unit.
i.e.,
volt,
or ten million
(lo') times the absolute
It
has a mechanical equivalent of .ooijôj horse-power;
there are j/fS watts in one horse-power,
f.
The unit of energy.

is
Practical unit, .the Joule : id' e.g. s. units.
Power
the rate of transformation of energy from one
form to another.
The amount
of energy thus transformed is
the product of the power and the time.
8]
FUNDAMENT/I L AND DERIVED UNITS.

T/ie Unit
I9>
of energy
is
transformed by unit power
in unit
time.
The
8.
corresponding
practical
unit
is
is
the
watt-second
(lo'' c.g.s.').
The name
in
of this unit in
the joule.
Circuits.
Power Consumption
Electric
By
the
consumption of power
transformation
1.
an electric
circuit is
meant the transThis
formation of electrical energy into some other form.
may
occur in two ways

Resistance.
Power Consumed by consumed by the resistance

lishing the current.
in
The
is
power that
IE, where /
isis-
of a conductor
the current in the conductor and
the pressure used in estab-
By
definition the value of the resistance
ohms means
the
number of the
volts used per
ampere
in
setting
up the current
in the conductor.
It follows, then,
that
E^ Ir.
The power consumed by the
resistance of the wire
is,
therefore,
W= 1E=
This power
is
Pr.
lost
from
the wire as heat.
The
electric
power
changed
to
heat in a conductor due
to its resistance is
propor-
tional to the square of the current.
The
relation
E=
Ir, as
above determined,
may be
transr
posed so as to stand
^='
when
it
(O
states
becomes Ohm's law, which

electric circuit the
In a closed
current equals the ratio of the
electromotive force to the resistance of such circuit.

2
Power Consumed by Counter-Electromotive

electrical
is
Force.
When
circuit or
energy
is
stored in or about an electric
transformed into mechanical or chemical energy.
20
is
[9
an e.m.f.
produced
in the circuit with
a direction opposed
to the current.
Such an
e.m.f.
is
called a counter-electro-
motive
forjned
force.
electrical
The rate at which the

is
energy
is
stored or trans-
equal
to the
product of the counter-electromotive force
and the current. The energy of a

dielectric
circuit.
may be transformed into heat by and magnetic naolecular action in and about the
circuit
Such processes are substantially the same as the
dissipation of heat in resistance.

9.
Problems in the Use of Electrical Units.
Prob.
A
in.,
i.
If
the space between two poles,
and
find
//
B, Fig. 15,
is
a magnetic field and the

is
area of each pole
10 sq.
the pull between the poles in pounds
when
the density of magnetism in the
field is
10,000 maxwells per

5/6. 9 lbs.
sq.
cm.
One pound
Prob.
is
is
445,000 dynes.
Ans.
16 there
sq.
2.
In the magnetic circuit
shown
area
lbs..'
in Fig.
a density of magnetism of 10,000 maxwells per
cm
{B
10.000).
()
What
should
be, the
of each pole in
((5)
sq. ins.:
To produce
a pull of 100
To produce
in.
a pull of 50 lbs..'
(a)
.8667 sq.
A7is.
Prob.
3.
If the
{b) .4333 " " armature of the magnet shown
in
Fig. 16
lbs.,
exposes 10
sq. in. surface to
each pole and weighs 10

sq.
what density of induction per

weight of
(a)
cm.
50
lbs..'
(b)
100
lbs..'
needed to support a (c) 200 lbs..'

is
(a)
(b)
(c)
2280 maxwells per " << 3088

"
11
sq.
'< <i
cm. <
11
Ans.
<i
Prob. 4.
The
wire
4266 shown in
<(
Fig. 17 carries
10 amperes.
9]
FUNDAMENT/IL /iND DERIVED UNITS.

total
The
magnetism, uniformly distributed over the poles,
is
1,000,000 maxwells.
force,
The
poles
are 3
in.
square.
What
1
measured
in
pounds, urges the wire across the

.29s
lb.
field
Ans.
Fig. 16.
Fig- i7-
Prob.
5.
wire in a uniform magnetic
field
makes an
When 10 angle of 45 with a plane normal to the field. amperes flow through the wire and the induction density is
50,000 maxwells per
sq. in.,
what
force,
measured
it
in
pounds
?
per inch of length of the conductor, urges
across the field

lb.
.0313
Ans.
Fig. 18
Prob.
6.
How much work
in
ft.
-lbs. will
be done
in passing
magnetic a wire carrying 100 amperes through the
field
shown
in Fig. 19,
[9
when 10,000,000 maxwells
of flux exist
square.'
between
the poles: (a) the poles are
When the
square
poles are 10
.'
in.
{b)
When
5 in.
()
{l>)
7.373
fl.-lbs.
Ans.
Prob.
7.
What
e.m.f. in volts
is
7-171 produced when the con-
"
ductor shown in Fig. 20 cuts across the uniform magnetic
sT"
^X
[Fig. 20.
N i^' s
/
field at
a uniform velocity of 5000
ft.
per minute
field.
.?
There are
i
,000,000 maxwells of magnetism in the
(^=
,000,000.)
3.33 volts.
Ans.
Fig. 21.
Prob.
8.
The
resistance of the circuit in
Fig. 21
is
10
ohms.
If a
par* of the circuit passes at constant velocity
9]
FUNDAMENTAL AND DERiyED

field
UNITS.
25
through a uniform
netism
is
3 in.
square in which the total

()
magflow
?
725,000 maxwells:
What
current
will
through the wire when the

(3)
circuit cuts the field in .01
second
In .005 second
(c)
How many joules

?
of energy will be
lost in heating the wire in (a)
(d) In
{i>) ?
() .0725 ampere.
(b)
{c)
Ans.
.1450
"
joule.
" "
" "
is
.0005256
{d) .0010512
Prob.
9.
circuit of 2
A battery furnishing a pressure of ohms resistance. A part of the
o volts
in
a
a
circuit is in
is
uniform
field 3 in. square, in
which the magnetism
1,000,000
Fig. 22.
maxwells.
With what
force,
field
.'
measured
in
pounds, will the
wire be urged across the
(See Fig. 22.)
.1474
Prob.
10.
lb.
Ans.
condenser
The
23
is
e.m.f.
at the terminals of the
shown
in Fig.
changed from 1000 volts
positive to
1000
The capacity of the condenser volts negative in ^^ second. > {a) What average current flows for the time is 7 microfarads,
{b)
How much
energy
is
stored in the condenser at 1000 volts

'
pressure.?
()
.84 ampere.
(^) 3-5 joules.
Ans. "
24
ELECTRIC/tL MACHINERY.
[9 in Fig. If .01
Prob. II.
The
3
current in the circuit

to
shown
24
is
changed from
henrys
10 amperes in .2 second.
is
volt
opposes the change, what

?
the value of the inductance in
.000285 henry.
Ans.
-E
o
Fig. 23.
^wsm^
^-o
Fig. 24.
Prob. 12.
The
current in a circuit similar to that of Fig.

in
24
rises
is
from 10 to 20 amperes
3 henrys.
two seconds.
The
inducif
tance
What
average volts are consumed

is
the
is
resistance of the circuit

rising
?
three
(6)
When
current has
ohms {a) while current become uniform ?

(a)
(<5)
,
60
volts.
Ans.
60
"
CHAPTER
III.
PERIODIC CURVES.
SYNOPSIS.
10. Properties of the sine curve.
a.
i.
c.
d.
c.
The alternating quantity. The sine curve in rectangular co-ordinates. Time as abscissae for sine curves. The sine curve in polar co-ordinates. The average value of a sine curve.
1.
Graphic determination.
Analytic determination. Analytic determination.
2.
f.
The
effective value of a sine curve.
1.
2.
Graphic determination.
11.
Combination of sine curves.

a.
i.
c.
Phase relations of sine curves.

Addition of sine curves of the same frequency,
Sum
of sine curves in quadrature.
d.
Product of sine curves of the same frequency.

I.
.i.
When When
the sine curves'are in phase.

the sine curves are in quadrature.
c.
Rate of change of sine values.

Addition of sine curves of different frequencies.
f.
12. Fourier's series.
13. Analysis of the general periodic curve.
Terminology of alternating
quantities.
10. Properties
of the
Sine Curve.
a.
An
alternating
quantity
tive
is
a quantity the values of which are alternately posi-
and negative.
The
successive values of such a quantity

is
may
plex.
follow a simple law, or one which
more or
less
com-
The
simplest possible law
is
am X
25
(2)
26
In this equation
;ir
[lO
is
a continually increasing angle, and
is,
therefore, alternately positive

is
and negative.
Its
maximum
is
value in either direction
A, which occurs when x
90, or
90 plus any multiple of 180.
plane curve drawn in any system of co-ordinates and
giving graphically the relation between
x and y
in the
above
equation
is
called
a curve of
sines, or often in
electrical
literature a sine curve.

b.
The Sine Curve
in
Rectangular Co-ordinates.
If suc-
cessive values of the angle
are plotted as abscissae, and the
corresponding values o{ y are taken as ordinates, the resulting

sine curve
is
of the form
shown
in Fig. 25.
convenient
Fig. 25.
method of plotting the curve is there given. If ^ be a constant radius vector which starts from a horizontal position and
rotates in a positive direction, counter-clockwise,
its
projection
on a
is
vertical line at
any instant
is
equal to
sin x,
where
ir
the angle between

c.
and the
origin.
Time as Abscisses for Sine Curves.

It is
Sine curves
will
be
used in this text to represent graphically alternating quantities
which vary with time.
often convenient, therefore, to
use time as the independent variable in the equation

jc
yi sin X.
This
may be done by
expressing
it
in
terms of time.
As x
and
is
a uniformly increasing angle,
varies directly as time,
its
'
lo]
PERIODIC CURkÊS.
is
27
value at any instant
determined by the time which has

left
elapsed since the radius vector
the
initial
position,
and
go
by the angular
velocity of the radius vector.
Thus when
its
equals the angular velocity of the radius vector,

position with respect to
its initial
angular
(,
position
oot.
is,
at the time
X
Substituting in (2),
y
This
is
=.
\T\ oat
(3)
the equation of the sine curve in terms of angular

If 7"
velocity and time.
be the time of a complete revolution
of the radius vector,
or
00
2n = -^.
becomes
(4)
By
substitution, equation (3) then
J/
= A smî
Polar Co-ordinates.
The Sine Curve same values of x and y,

d.
circle,
in
When
curve
the
as determined in the preceding secis
tion, are plotted in polar co-ordinates, the resulting
Fig. 26.
The
negative sine values are found for values

In this part of the cycle the
of
X between
is
e.
180 and 360.
traced over by The Average Value of a Sine Curve. By "value of a is meant the value of an ordinate to the curve curve of sines
curve
the negative end of the radius vector.
'
'
as
drawn
in rectangular co-ordinates; i.e., a

for a
jj/
value.
The
The
average value
plete
complete period, corresponding to a comvector

is
revolution
of the
A, Fig.
25,
is
zero.
average value for a half period
obtained by dividing the area

i.e.,
of a loop of the curve by
its
length,
by
ti
when
the angles
28
fia
in rectangular
are measured in radians.

co-ordinates,
this
if
For an actual curve,
area
may be
obtained by means of the
planimeter, or,
plotted on cross-section paper,
by counting:
Fig. 26,
the enclosed squares.
Thus,
in Fig. 25, there are
33.7 squares
enclosed by one loop of the curve, while the length of the loop
is
10.
The average
which
is:
ordinate
is,
therefore,
33.7 divided by
10, or 3.37.
This corresponds exactly with the value found
analytically,
Average ordinate
=X
It
71
maximum
3-37-
ordinate
= -5-3 =
Analytically the same result
tegration, as follows:
is
obtained by a simple in-
y= A
Area
z=
sm
X, dx.
y dx ^ A sm X
y dx
=Aj
sin
x dx
lOj
PERIODIC CURVES.
= A f COS
average ordinate
;ir
2A,
area
t
2A =
TAe average value of a

value.
sine curve is
times
its
maximum
f.
The
'
'
Effective
Value of a Sine Curve.
By
"effective
value
meant the square root of the mean of the squares of This is always positive and has the the instantaneous values.
is
same value whether taken for an In Fig. 27 is shown the period.
entire period or for a half
sine curve b
and the curve
Fig. 27.
of squares
c.
Each
ordinate of the latter curve
is
the square
of the corresponding ordinate of the former. value of c may be obtained by the methods used for determin-
The average
ing the average value of the sine curve.
The
effective value
of ^
=
I.
V average
c.
Analytical Determination from the Rectangular Curve.
For curve b, j ' " c,y
= A sin x. = A^s\r?x
(S)
30
[lO
y dx = A^J
sin*
x dx
A^ r'
/
(i
cos 2x)dx cos 2X dx
dx
Average value ofc

Effective value
A^
*
of b
A
V2
2.
Graphical Determination from the Polar Curve.

tliat
It
has
been pointed out

co-ordinates,
is
a sine curve,
when
plotted in polar
circle.
The
area of the circle
may
be con-
sidered as
made up
of infinitesimal triangles, with apexes at
the centre of revolution of the vector and having an angular
width of
Ja
and altitude
b,
as in Fig. 28.
The
areas of the
Fig. 28.
triangles will be proportional to the squares of their altitudes,
as the angle
Aa
is
constant, and therefore the total area
is
proportional to the
triangles.
mean
of the squares of the altitudes of the

circle
is
The complete
generated while the radius
vector, b, sweeps through an angle of 180.

If a semicircle
be drawn, as shown
circle, its
in Fig. 28, with the
diameter tangent to the
area will be the
same
as that

ii]
PERIODIC CURVES.
circle, if the
3^
of the
square of
its
radius, r,
is
equal to the
mean
of
tri-
of the squares of
b.
This
is
evident from the facts that the
area of the semicircle
may be
considered as
made up
angles as in the preceding case and therefore proportional to

r^,
and that the
radius, r,
sweeps through the same angle swept through by b

the average
in in
generating the semicircle as was
generating the
circle.
That
is,
of the squared
is
Tadius vectors of the sine curve in polar co-ordinates
the
square of a radius vector of constant length which would sweep over the same area in passing through an angular distance of
j8o.
As
the areas of the circle and the semicircle are equal,
A
''=
V2'
But
f^
where A i^, and
is
the diameter of the circle.
the
mean of
r
^
Vmean of IP'
effective b
A = j^Phase Relations
oj
II.
Combinations of Sine Curves.
a.
Sine Curves.
curve
is
A more general form

y
=.
for the equation of a sine
^vci
{x
-\is
a)
(6)
where a
0
is
a constant angle, and
x
is
an angle increasing from
at a uniform rate.
The
value of the ordinate to the curve
on
starting, or at the time zero,
y^=^
sin a.
The
and
curve
y'
Â
sin {x
-\- oi)
its
generating radius vector. A, are drawn in Fig. 29.
[11
second curve,
b,
is
drawn
is
in the
same
figure
The
equation of the second curve
where
/S'
is
a constant angle.
The two
angle
(/?
curves are said to differ in phase position

a), as that is the
b,
by the
angular difference between corre-
sponding points on a and
say between the points at which
+Y
Fig. 29.
they cross the

vectors
{/?
X
B
axis in
the
in
same
direction.
and
also differ
phase position
is
The radius by the angle

and
a).
Since counter-clockwise rotation
taken as positive,
the radius vector
is
ahead of
in angular position,
the sine curve b
is
ahead of the sine curve
a, or it is said to
is
lead a
by an angle of (/?
(/S
a).
Conversely, a
said to lag
an angle of
a) behind b.
is
When the When

are then
ft
angle between two sine curves
90, the curves
are said to be in quadrature (Fig. 30).
a, the
radius vectors
coincide,
1).
and the sine
curves are said to be in phase see (Fig. 3
Their equations
= A sin {x y" = B sin (x -fy'
-\-
a),
a),
Ill
PERIODIC CURVES.
ratio
33
ciarves,
and the
between ordinates to the two

is
drawn
through the same point, x,
A B
One
curve
sin
(x-\-a)
A'
=i ^=r=.
sin {x-\- a)
constant.
may
therefore be derived from the other

,
by
multi-
plying each ordinate of that other by a constant.
Fig.
30.
Fig. 31.
The
in
ratio between corresponding ordinates

is
of two sine curves
phase
b.
constant.
Addition of Sine Curves of the Same Frequency.

are
Sine
their
curves
said to have the same frequency when
generating radius vectors revolve with the

velocity.
same
angular
Sine curves
ordinates,
i.e.,
may be added by the

ordinates lying in the
addition of corresponding
same
vertical line.
Each
such addition gives the corresponding ordinate to the curve of
34
["
by the determina-
sums.
The
curve of sums
may be
plotted
tion of a large
number of such
points.
Fig. 32.
In Fig. 32, the sine curves a and
i>
have the same
fre-
quency, and are generated by the radius vectors

It is
and B.
is
required to find the form of th curve
s,
which
their
sum.
Construct the parallelogram

ordinate to the sine curve a
OASB.
is
The
at
any
instant,
x^^
OA
The
sin
xOA xOB
A-^x^.
is,
ordinate to the sine curve b
at the
same
instant,
OB
The ordinate
sin
B^x^.
s, is,
to the curve of sums,

SjjTj
at the
same
instant,
= AjX^-{- B^x^
SÂ^
where
=
sin
B'^x^,
AS sin AÂS = OB
and
xOB = B^x^=
xOA
SÂ^,
OS
Therefore
sin
xOS = OA
is
sin
+ ^5 sin A^S
OS
on the ordinate
S^x^^
the projection of
through
x^.
"]
Since this
is
PERIODIC CURVES.
true for
35
any position of the parallelogram
OASB,
it
follows that
sine curve
The sum of two sine curves having' the same frequency is the which is generated by the diagonal of the parallelo-
gram formed on the radius vectors of the component curves. c. Sum of Sine Curves in Quadrature. The equation
y"
= B cos = B sin
may
be written
y"
when
it is
[x
+ J)
is
seen to be a sine curve that
90 ahead of the curve
y ^^ A
The sum
of the curves
jj/'
sin X.
=^
sin
and
is
y"
= B cos x
A
is
therefore a sine curve which
generated by a radius vector
forming the diagonal of the rectangle produced by drawing
making the angle
x,
and B, making the angle x
-\
with
Fig. 33.
the zero position.
(See Fig. 33.)
The
value of this diagonal
must be
S=
VA^
+ ^.
36
[X
The magnitude
of
of the
its
sum
is
dependent on the magnitude'

it
and B, while
is
phase position, or the angle a which

ratio
makes with A, Thus
governed by the
between
and B.
A
zero,
= tan a.
or
tan
-r.
a may therefore vary between the and A and B are positive, S
limits o
lies in
and 90.
first
If
x be
the
quadrant.
With
"
A A
negative and
positive,
lies in
the second quadrant.
ii]
PERIODIC CURVES.
37
express a sine curve as the

cosine curves, thus:
sum
of the corresponding sine
and
J
instead of
=A =
S
sm X
sin(;ir
-\-
cos
(/)
y
nT
+ "d
d.
Product of Sine Curves of the

curves
is
Same
Frequency.
By
product of sine
meant the curves

Let
b' ,
of products of
instantaneous values.
I
.
When
the sine curves are in phase.
it
be required
to find the product of the sine curves b
and
Fig. 35.
Fig. 35.
Their equations are
= ^ sin X, y = A' sin X,

j>/
and the product
of their ordinates at
any
instant
is
yy'
= y^ = AA' Ax? X
(8)
Expanding
this, it
becomes
cos 2x)
,
y^
A A' = ^ (I
2
(9>
38
[11
Transposing the axis of this curve from

of
,
XX to X'X' a distance
its
equation becomes
^2=^1
AA'
AA'
J-=
(^2jr
^cos2x,
,
.
.
(10)
^2= - -y-sm
which
axis
is
+ -j, ....
i>',
(11)
a sine curve of twice the frequency of ^ or
with
its
AA'
at a
distance
2
above the axis of
l>
and
^'.
The
average value of
this
curve referred to tne axis
XX
AA'
is
.
This
may
be written
A A' = =
.
when
b'.
it is
seen to be the prod-
\/2
j/2
uct of the effective values of b and
The average value of

phase
is
tlie
product of two sine values in

This
is
the product of their effective values.
one-half
the product of their

2.
maximum
values.
When
the sine curves are in
quadrature.
Let
b' ,
it
be
required to find the product of the sine values b and

36, which are
in quadrature.
Fig.
Their equations are
^^
sin X,
y = A' sin
yy'
\x
-|
A' cos
x,
{A
sin
x){A' cos x)
= AA'
sin 2x.
(12)
This
is
a sine curve of double frequency, the axis of which

b'.
coincides with the axis of b and

for a
Its
average value taken
complete period
is
zero,
"J
PERIODIC CURVES.
39
sine curves
in
The average value of the product of two

quadrature
is zero.
Fig. 36.
When
Let
the sine curves are not in

it
phase and not
in
quad-
rature.
be required to find the product of the sine

Fig. 37.
curves b and
b' ,
One
with
of these sine curves, say

is
b'
may be
the
divided into components, one of which

in
in phase,
and
other
quadrature,
b.
These components are
generated by the radius vectors
OB^
OB,^
= OB cos OB sin 6,
the
e,
where
is
the angle of phase difference between

of
and B.
The product
(11)
and
B is
sum
of the products of
with
each of the components of B.
These products are given by

is
and
(12).
The average
quadrature product
zero.
The
maximum
value of the component of
B
6.
which
is
in
phase with
is
OB^
= OB cos
40
ELECTRIG/IL MJCHINERY.
["
ll]
PERIODIC CURVES.
average product of the sine values
41
The
and
B is
therefore
OA
V2
OB
COS
d.
V2
sine curves is the
The average value of the product of two

product of their
effective values, times the cosine
of their angle
of phase difference. e. The Rate of Change Curve of Sine Values. The rate at which the ordinates to a sine curve change their values is
variable,
and
if
plotted in rectangular co-ordinates produces a
Fig. 38.
sine curve in quadrature with the one from which
it is
derived.
(See Fig. 38.)
The
first
curve
is
jf/
sin X,
it is
or plotted with time as abscissae
y
where
a sm
27r -=,^,
T is
the time of one complete period.
Then
df :=
-i^fra
cos -i^tdt,
42
[i*
and the
rate of
change
^ = asm[-+-^t)
'The rate of change curve
is
(I3>
therefore
its
sine curve go in
advance of the original curve, and

great.
values are -~r times as

sine curve is
Conversely:
The integral of a
a second
sine curve po behind the original curve,
and
its
values are
times as great.
f
The
tive
The Addition of Sine Curves of Different Frequencies.

is
curve resulting from the addition of sine curves of different

non-sine in form.
It is
frequencies
always irregular
posi-
and negative portions taken with reference to the

or
X axis
by
may
may
not be alike.
of this fact led
to
The
investigation
the
discovery
Fourier that periodic sine curves of different frequencies
may
always be found
that their
in value,
phase positions, and frequencies such
sum
will
produce a periodic alternating curve of any
desired form.
Fourier determined a systematic method for obtaining the
which when added would reproduce any given periodic curve however irregular
periodic sine curves of different frequencies
it
might be.
12.
Such a
an
series
is
known
for
as a Fourier's Series.
The Fourier's Series
Any
be
Periodic Alternating
in
is
Curve.
If
irregular
curve
plotted
rectangular
co-ordinates and the length of a complete period

to 2n, or 360,
it
made
equal
by the addition of sine waves properly selected as to amplitude and phase position, one of which has the same frequency, or wave length, as the
reconstructed
irregular wave,
may be
and others having twice, three times,

/^ that of the irregular wave.
four
times,
n times that frequency, and having

wave lengths
of 1/2,-1/3, i/4'
12]
PERIODIC CURyES.
43
The component sine curves which combine to produce the total curve are known as the first, second, third, etc. harmonics. The first harmonic, or fundamental, has the fi-e.
. .
quency of the
firequency,
total curve
the second harmonic
twice that
and so on.
,
This convention
differs slightly
from
that adopted in music
where a tone having twice the frequency

called the
first
of the fundamental
in convention
is
is
harmonic.
The change
used for convenience only.
The
general equation. for a curve

is
made up
of harmonic sine
components
jf
A'
sin {x
-{- o-j) -|-
A"
sin
(2x
.
-j-
a^y
+
ylj,
A'"
sin (3;r
+ 3) +
. .
+ A" sin
(nx
+ ).
(14)
A^, A^,
A^&re
a^, a^,
the amplitudes of the component

.
sine waves, and
a.^,
a are constant angles which
specify the position of the radius vectors generating the
ponent curves when x
is
zero or
is
comany multiple of 360.

is
These
relations are
shown
in Fig. 39.
In this figure the irregular curve
made up
of three
twice,
harmonic sine components, having frequencies of once,

and
three
times
that
of the
irregular
curve.
The
radius
T'ectors
A',
A",
and
and A'" which generate the component
curves rotate with different angular velocities, and are
numIn the
bered
I,
2,
3 to correspond to their velocities.
figure these radius vectors are
drawn
in
heavy
lines for the
position in
;r
which x
45".
is
zero,
and
in dotted lines for the position
jf J
The
equations
of the fundamental, second,
and third harmonic
sine
components of 5 are
y = A' sin {x + 30), y = A" sin (2x + 60), y" = A'" sm + 120),
(3;f
44
'ELECTRICAL MACHINERY.
[iz
13]
PERIODIC CURVES.
45
and the equation of the curve
5 is
{2x
A' sin{x-\-
30)
+ A" sin
+ 60)
120).
4-^'"sin (3^:+
Note that
the
this
(15)
curve does not have Hke loops above and below
X axis.
From
the
statements already
made
it
follows
that
by
properly selecting the values A', A", A'", etc., and the constant angles of phase displacement; a^, a^, a^, etc., the above
equation
may
represent any finite, continuous, and periodic
curve whatever, no matter
how
irregular
it
may
be.
13. Analysis of a General Periodic Curve

tion of
The
is
separa-
an irregular periodic curve
into
its
components
first
always
possible,
and usually
it
will
be found that the
five
comtotal
ponents give a sufficiently close approximation to the

curve to answer
all
practical purposes.
The
actual analysis
may
1
be made graphically or analytically.

is
In either case the
method
based on the following laws
other harnuinic is zero
The average of the product of any harmonic with any when taken for a complete cycle of the
irregular curve.
The average of the product of two sine curves in phase, having the same frequency, is half the product of their and
2.
amplitudes.
The average of the product of two sine curves in quadrature and having the same frequency is zero. The discussion of the first law is beyond the scope of the
3
present text.*
The meaning
of the law
may
be illustrated
sine curve
a.
graphically as in Figs. 40 and 41.
In Fig.
40 the
b has twice the frequency of the fundamental curve
Their
product taken over a complete period of a gives a curve
having loops of equal areas above and below the

* See " Fourier's Series and Spherical Harmonics"
X axis.
The
Byerly.
46
10.
in
13]
PERIODIC CURVES.
is
47
Similarly
a,
average ordinate of the product

in Fig.
therefore zero.
41 the product of
b,
having twice the frequency of

a,
and
c,
having three times the frequency of
gives a product
curve which has the average ordinate zero.
38,
For the proof of the second law see Section 11, and for the third law see Section 1 1, d, 2, page
d,
page
39.
The product of one quantity by another, is identical with the sum of the products of that quantity and each of the components of the other.
Thus,
-\- b)
x{a
= xa
-\-
xb
(16)
It follows, therefore, that the
average ordinate, taken over a
complete period, of the product of an irregular curve and a

sine curve having n times
its
frequency,
is
equal to the average
ordinate of the product of
this sine
curve and the component

is
of the
nth.
harmonic of the irregular curve which

ordinate, taken
in
phase
limits,
with
it.
The average
between the same
of the product of this sine curve and any other harmonic com-
ponent of the
components,
total curve, or the
sum
of
all
the other harmonic
is
zero.
The
general equation
J
may
= A'
sin {x -\- Wj) sin (3;r
+
f
A'"
+ A" sin {2x + a^ + ^3) + + A" sin {nx +

.
. .
ar
be rewritten by Section
11, c, p. 37, eq. (7), as follows:
A^sin X
-\-
A^
sin 'SV
+ A^ sin
-\-
ix
-\-
sin nx,
+ ^j cos x-\- j cos 2x

Let
it
-(-
^3 cos 2^-\-
(17)
. .
+ B cos nx.
coefficient,
be required to
,
find
any
say A^, and the

is
phase angle, a^
of this equation,
when
the curve itself
given.
From
Section 11, pages 35 and 2^,
(A'"f
A,^
+ B,\
48
[i3
and
"3
Multiply graphically a complete wave of the total curve by
the curve
jj/3'
sin 2,x.
The average
curve j^
ordinate
of this product,
E,
is
equal to the
factor
average ordinate of the product of the
latter
by the
2,x
A^
sin
ix,
since the product of
y^
= sin
with
each of the other components of

ordinate of the product of
j/3'
y
IX
is
zero.
But the average
= =
+.,,
sin
and
j/3
^3
sin 2>x
is
and
this
may
be placed equal to the average ordinate of the
graphic multiplication, or
A,
Similarly, as
total curve
=.
2E.
is
the average ordinate of the product of the
and the curve

^3"
cos ^x,
sine
it
follows that B^
2F.
The
this
harmonic, A'", have thus been determined.
and cosine components of It was shown

this
above that the amount and phase position of

therefore, immediately
harmonic
are,
known.
The
fied
analysis of a general curve
may
If
be often much simplii.e.,
by an inspection
first,
of
its
form.
only odd harmonics,
the
third,
fifth,
etc.,
are present, the phase relations
between components
will
be similar at the end of each half
I3j
PERIODIC CURyES.
(See Fig. 45.)
49
wave-length of the fundamental.

will
Such a curve
have
like loops
above and below the
axis, and, con-
versely,
Curves having
tain only the
like loops
above and below the
X axis
con-
odd hartnonics
In this case
the average
This
is
the case with nearly all curves which represent the
behavior of electrical machinery.
ordinate of the graphical multiplications

half
may
be taken over a
-wave instead of over a complete wave, since the product

itself
curve repeats
every half period of the fundamental, where
only odd harmonics are present.
Where
the even harmonics are present,
i.e.,
the second,
fourth, sixth, etc., the phase positions of the
components are
similar only at the close of each complete period of the general
periodic curve.
The
result
is
that all such general curves must

;
be unsymmetrical about their axes

tions of
the plus and minus por-
each period must
differ in
form.
Curves having unlike loops above and below the

general contain both odd and even harmonics.
X axis in
periodic
The
following problem
is
given to illustrate fully the applianalysis
cation of the above
method of
by which any
its
alternating curve
may
be resolved into
harmonic com-
ponents.
Problem.
It is
required to determine the harmonic
comin
ponents of the irregular curve drawn with the heavy line

Fig. 42.
This curve
is
symmetrical about
This fact
is
its
axis and should contain
no even harmonics.
tested in the diagram of Fig.
42, in the case of the second harmonic.
The
unit sine curve of double
frequency drawn in this
diagram constitutes the analyzer

second harmonic.
original curve
It
is
for the sine
component
of the
drawn 40
:
to a
larger scale than

original curve
is
the
by the
ratio
i.
The
multi-
in
^3]
PERIODIC CURVES.
5'
plied
laid
by
this analyzer, the result
being the product curve as
down
in the
diagram.
its
An
inspection of this product curve shows that
average
value must be zero.

tive areas for a
From symmetry
the positive and nega-
complete period that the curve encloses with
the axis must be equal, their
sum must be
zero,
and the
be zero.
phase
average value of the product curve

This result
is
rnust, therefore,
due evidently to the
fact that the relative
positions of the second harmonic with reference to the plus
and
minus portions of the original curve are such as to give
alter-
nately plus and minus signs to corresponding product values.
The
relative
phase positions of the original curve and the

to give
second harmonic sine analyzer are such as

reference to the axis.
product
values throughout a complete cycle symmetrically located with
For every plus value there
is
a corre-
sponding negative value causing the average value of the product curve to be zero.
A trial of
Further
a cosine second harmonic analyzer will produce
the same result, and for the same reasons.

trial
of any even harmonic analyzer will produce

for similar reasons.
an average zero product

It
follows from this test, therefore, that this symmetrical
irregular curve contains
no even harmonics, otherwise the
product of
it
and some of the even harmonic analyzers would
give positive average values.
The
curve being quite irregular and symmetrical will be

in the
found to be rich
odd harmonics.
In Figs. 43 and 44
sine,
are given diagrams showing the application of the
and
In
cosine third harmonic analyzers to determine the value and
phase position of the third harmonic of the original curve.
Fig. 43 the unit sine analyzer for the third harmonic analyzer
is
drawn and the product curve determined
as shown.
The
52
[13
^3]
PERIODIC CURVES.
53
gggggggg
3
n
hj
54
positive
eiECTRICAL MACHINERY.
[13
and negative areas of the product curve are then
determined by means of a planimeter.

divided
The
final
net area
is
by the length taken

is
to actual scale of a
complete
period, producing thereby the average value of the product.
This value
drawn
in Fig. 43,
is
and so labeled.
In doing this
the numerical work
as follows
Positive area
5140
Negative area
454^
594
Net
positive area
Length of one period
360
594 7:=
-\- 1. 65
Average product
The maximum
value of the sine component of the third

is
harmonic of the original curve
therefore
^3
Fig.
= 2(+
i-6s)
= + 3-30.
44 gives graphically a corresponding analysis to determine the cosine cohiponent of the third harmonic of the The numerical work for this is as follows: original curve.
Positive area
13]
PERIODIC CURVES.
55
is
The
total third
harmonic component
therefore
= VAi+Bi, A'" = ^3-3^ 6.82^

A"'
^'"
+ tan
is
.-6.82 *
'
3-3
7.56,
-64
6'.
This third harmonic component
drawn
to scale in
amount
and phase position
in Fig. 45.
The fundamental and the fifth harmonics have been determined by this method. The only difference in detail met with
consists in the use of sine
and cosine unit analyzers having and

fifth
periodicities corresponding to those of the first
har-
monics which they are

ing equation:
to determine.
The numerical values

in the follow-
and phase positions of these harmonics are given
=
*
66.8 sin
-\-
4.4 cos
-\-
3.3 sin
^x
6.82 cos 2^
(18)
The accompanying
7.24 sin
5;if -j-
1-99 cos 5x.
figure will assist the reader to see that the ratio of the
-X
-Y
cosine to the sine components gives the tangent of the angle of phase difference, not the cotangent as might at first be supposed.
and
56
[13
The
equation
may
conveniently and obviously be written
in the following
manner:
1-995).*
f
or,
it
(66.8^
+ 3.33 - 7.24J +y(4.4i - 6.823 +

54')
by combining the quadrature components

be written
of the harmonics,
may
> = 66.9 sin (x + 3

70
+ 7.56 sin (sx - 64 + 7-5 sin (5;p+ 164
6')
36').
13]
PERIODIC CURlÊS.
Evidently
if
57
harmonics.
the original curve had
more proFig. 45.

ninth,
nounced minor
difference
irregularities there
it
would be more of a residual

in
between
and the summation curve
Such residual could then be expressed by the seventh,

etc.,
harmonics.
of the
By means
ponents.
the
method
illustrated in the
above problem
any periodic curve may be broken up
into
its
harmonic combe expressed as
Conversely, any periodic curve

of
its
may
sum
harmonic components.
All problems that arise in the treatment of general periodic

or general alternating quantities
may
be solved by the treat-
ment of
their
harmonic components by the methods and laws
given heretofore for the treatment of simple alternating or sine-
wave problems.* Problerns numbered 47 and 48 illustrate this method of solving problems that arise in dealing with alternating quantities.
Effective
Value of a Non-sine Alternating Curve.

is
The
effective value of an alternating curve
the square root of the
mean
square of
is
its
values.
The
equation of the general alter-
nating curve
y=
A^svc\.
j5j
X cos X
Ar
-\-
A^
2x B^ cos 2x
sin
-\-
A^s'm
'ix
T,x
-f
-{-
-j-
+ B^ cos
(19)
Squaring,
f A^ sin^ X + A^ sin^ 2x + A^ sin^ B^ cos^ X + B^ cos2 2x + B^ cos^

-j-
2,^
-f
(20)
ix -^
terms of the form
{A^
(A
sin sin
mx){A
sin nx) -{-,..-{
nx)(B cos nx)
+
-\-
+
.
{B^ cos mx)(B cos nx)

* From
this it is seen that
methods employed
all
for the treatment of sine
are necessary for the convenient solution of
non-sine waves.
waves Such methods
have, therefore, an entirely general application.
58
All of these product terms have a
[13
mean
value of zero.
See Section
page 45
law
i.
From
Section 10,/,
(Mean ^
sin^
;i:)
= ^^
(21)
From which
(Mean ^)
it
follows that
= J(^^ + ^/+^/+
effective value
j/
is,
+ B,'+B,' + B,^+
.).
(22)
The
therefore,
Effective
= -^VA,^+Ai+Ai+.
Since Â^ and iB^ are the
.B,^
+ Bi+B,' +
.).
(23)
mean
squares of their correspond-
ing curves, the result

square of any harmonic
may
is
be further reduced.
The mean
^J =
where
KA^
+ B^),
(see Sec.
1 1, <:,)
(24)
is
the effective value of the harmonic.
Substituting,
Effective^
= VH^ + H^
-\-
H^ ^
effective
(25)
The
effective value
of a general periodic alternating curve is

values
the square root of the
sum of the squares of the
of its harmonic components Terminology of Alternating Quantities.

equation of the sine curve,
In
the general
y=
if
sin {x -|- ),
is
the angle
increases at a uniform rate, the curve
said to
be periodic.
As
the curve represents an alternating quantity,

is
the expression alternation
used to designate a half
wave
which
is
generated while the ordinate varies in value from
13]
PERIODIC CURVES.
in
59
in the other direction.
maximum
one direction to maximum
Two
consecutive alternations constitute a complete cycle, and

is
the time occupied in passing through a complete cycle

period.
This
is
the time during which
increases
by 360".
the
The number
second
is
of cycles per second or the
number
of periods per
called
the
periodicity,
or
more commonly
frequency.
In the treatment of alternating currents the term
wave
is
generally used in lieu of the expression periodic alter-
nating curve. or quantity.
CHAPTER
IV.
COMPLEX QUANTITIES.
SYNOPSIS.
14. Vectors.
a.
b.
c.
Definitions.
Graphical conventions.
Addition of complex quantities.
Multiplication and division of complex quantities.
d.
15. Alternating quantities

a.
b.
and
vectors.
Alternating quantities represented
by
vectors.
Operations upon vectors representing alternating quantities.
14. Vectors.
a. Definitions.
A vector or directed quanA

lies in
tity
is
a quantity having both magnitude and direction.

is
vector
represented graphically by a line which has a length
representing the magnitude of the quantity, and which

the direction of the vector.
The
line is usually
drawn with
an arrow-head to indicate
its
direction.
vector
can only be
represented analytically
its
by an
its
expression which takes account of
direction as Well as
magnitude.
b.
Such an expression
is
a complex quantity.
position
Graphical Conventions.
The
of a vector
is
specified
its
by reference
end.
to a horizontal line
which passes through

horizontal
in the zero position,
If the vector
initial
If the vector coincides with this

left
line,
and extends from
to right,
it is
and
is
may be
represented by a positive numeral.

left, it is
horizontal, but extends from right to
negative, and
may
be represented by a negative numeral.
If
a vector be
60
multiplied
by the
factor
i),
it
is
reversed in direction.
14]
COMPLEX QUANTITIES.
process of reversal
is
61
in
The
assumed to consist
initial
a counter-
clockwise rotation of 180 about the

as a centre.
If a vector
end of the vector
be multiplied twice by
rotated 180.
i, i.e.,
by (|/
if, or
by (
i), it is
It is
logical to
assume that each factor produces half of

multiplying a vector by V^
90.
i
this rotation, or that
gives to
it
a positive rotation of
For convenience
in algebraic expression, the letter

(/
J is
used as the equivalent of
Then
I,
J= Vj^= The
cation.
I.
-y= -Voperation of division

is
I,
the reverse of that of multipli-
Thus the
it
division of a vector
by /
is
is
equivalent to
giving
factor
a negative rotation of 90, and
represented by the
j.
Thus,
^
^J _
The
The
quantity a
-\-jl>,
Fig. 46,
is
a vector which
is
the
sum
of a positive horizontal line, a, and a positive vertical line, y^.
quantity a
jb
i.
is
a vector which
is
the
horizontal line, a, and a negative vertical hne,
sum of a jb.
positive
Summary.
is
vector which
its
is
horizontal and positive

If horizontal
denoted by a quantity giving
length only.
and negative, the quantity is preceded by the minus sign. 2. A vector which is vertical and positive is denoted by a
quantity giving
its
length,
is
and by the factor/
If vertical
and
negative, this product
preceded by the minus sign.
62
3.
[14
vector which
is
neither
horizontal
nor vertical
is
expressed as the sum of
its
horizontal and vertical components.

is
real. A A quantity containing the factor/ imaginary. A quantity made up of real and imaginary components
quantity not containing the factor/

is
is
a complex quantity.
c.
Addition of Complex Quantities.
Vectors may be added
conveniently by the addition of the corresponding complex
Fig. 46.
Diagram of Complex Quantities.

For example, the complex
of the
quantities,
from which the absolute value as well as the phase

readily
position
is
obtained.
quantity representing the

{c
sum
vectors [a -\-jb)
and
^jd)
is
{a^jb-)^{c^jd)
The
= {a^d)^j{b+d).
is
(26)
absolute length of the vector
/= ^{a^cf + (bJ^d)\
This occupies a phase position with respect to the
(27)
X
:
axis
which
is
found as follows
sm
S/(a
b-\-d
+ cf ^
(^
+ df
(28)
14]
COMPLEX QUANTITIES.
Numerical
Illustrations.
63
Prob.
3 ^rJ'7, 5
13.
Draw
the vectors
S.
S.ys. 3 Prob. 14.
+74>
-76,
+^6.
Add
the following vectors, and draw the vector
representing their
sum
also
draw the vectors separately and
add them graphically.

()....
(2+78) {b).... (3-75) ic)... .(- 4 +71)

Multiplication
quantities are
It is
(4-73) (9+76) (- 6 -78)

multipled and
(i+y6).
(3-72).
(-
9 +J7)Quantities.
d.
and Division of Complex
be
Complex
binomials.
divided like ordinary
only necessary to bear in mind the meaning
of the factor 7'.

substituted.
Whenever 7^
{a -{-jUf
occurs, the minus sign
may
Thus,
(a -\-jb)(r -\-jx)
= a" ^j2ab -\-fb'' = 2 .j-j2al> ^, = ar -\-jbr -\-jax \-j^x = ar -}-j'{br ax) bx.
-\-
(29)
This product
may
be written
{ar
Its absolute value
bx)
-fj{br
-\-
ax)
(30)
is,
therefore.
V{ar
After expanding
bx)^-{-{br + axY.
(31)
we may
Va'
reduce this value to the form

Vr^
+ blx
+ x'.
....
(32)
From
product.
this result
it
follows that the product of the absolute

is
values of two vectors
equal to the absolute value of their
Let
6',
6",
and
6,
respectively,
be the angles which
a +jb, r +jx,
and
their product,
(ar
bx)
-\-j\br
+ ax),
64
with the horizontal.
cos
0'
[14
make
Then
=
(^2
+
r
b
^3)4'
cos 6"
=
{a'
sin d'
+ b^f
X
sin 6"
=
ar
cos
cos
Substituting,
(6"
=
('
(9")
bx
X
(^
^)^
+ ^)*'
^'
= cos
^'
cos e"
- sin
sin 0".
we
have, after reducing,
cos
{e'
+ e") =
wherefore
+ ^ = +
(^
^'
b^Y-
ar-bx X (^* + ^)*'
(33)
(9".
From
angles.
this it follows that the

is
phase angle of the product of

their individual
two vectors
equal to the
is
sum of
phase
Phase angle
the angle which the vector makes.
with the line of reference, usually-
taken as the horizontal.
The
complex
quantity resulting
from
the multiplication or division of

quantities
may be
laid
down
to
graphically.
Thus, in Fig.
are-
47, the vectors
oA and oB
vectors
be
multiplied.
Expressed
are
analytically,
these
7
is
+y2 and 3 +73. Their product 2 +76 +72 1-6=15 +727.

1
Fig. 47
The
The numerical
values of
1/3^
or
absolute
are
Multiplication of Vectors.
oA
32
oB, and
oR
V72-f
2^
7.27,
=4.24,
15]
COMPLEX QUANTITIES.
65
and
ViS^
+ 27^ =
30.8
7.27
4-24.
The absolute value of the product of two or more vectors is the product of the absolute values of the vectors. The angle
which this product makes with the
angles
initial line is the
sum of the
made by
the several vectors.

is
As
the operation of division
the reverse of that of multi-
plication, a similar
law
may be
stated for the division of
com-
plex quantities.
The absolute value of the

by another,
is
quotient,
when one
vector is divided
obtained by dividing the absolute value of the
dividend by that of the divisor.
The angle which
this quotient
makes with the
initial line is the difference

line.
of the angles made
by the two vectors with the initial
As
vector
a corollary
is
it
may
be stated that the reciprocal of a

its
the reciprocal of
absolute value with

is
its
angle
reversed.
Thus,
in Fig. 48,
OA'
the reciprocal of
OA.
Fig. 48.
Division of Vectors.
and Vectors.
15. Alternating Quantities
a.
Alternating
Quantities Represented by
Vectors.
The
actual value of an
66
is
[i5
alternating quantity
if
changing from moment to moment, and
the positive
value for
and negative alternations are equal the average any whole number of alternations is zero. The
produced by such a quantity
is
average
effect
not,
however,
necessarily zero.
An
instance
is
found where an alternating
electric current is present in
a conductor.
is
The average
cur-
rent
is
zero, but the effect of the current
to heat the con-
ductor continuously.
A new definition of effective value may

be shown later that
this is the
now be
quantity
given.
It will
same
quantity which was defined on page 29

is
when
the alternating
electric current or e.m.f.
The
effective
value of an alternating quantity

effect
is its
value
measured
in
terms of an average
which
it
may produce
continuously.
Such a quantity, being an average value, may be expressed
by a
positive numeral or represented
by a
straight line.
It is It is
not in the strictest sense a directed quantity or vector.

nevertheless
a very great convenience
in
many
cases
to
arbitrarily assign to
an effective value a direction which shall

itself.
indicate the phase position of the alternating quantity
The
alternating quantity
is
then represented graphically by a

the effective value of the
its
is
vector having a length equal to

quantity,
and a direction representing

is
phase position.
The
phase position
a relative item, and
specified in terms of
the phase position of
some other same frequency, which is chosen

This standard
position.
b.
tities.
is
alternating quantity of the

as a standard of reference.
in the zero
always represented by a vector
Operations on Vectors Representing Alternating Quan-
Alternating
is
quantities
may be added
or subtracted
by
adding or subtracting their respective vectors.

such an operation
The
result of
represented by the resulting vector.
Independent alternating quantities
may
not be multiplied
is]
complex quantities.
67
or divided
ing their
by performing these operations on vectors representeffective values. Such an operation is meaningless

it
and the vector resulting from
does not represent the result
of the multiplication or division of the alternating quantities.
A vector
or divided
representing an effective value
may be
multiplied
by a vector representing a constant quantity having magnitude and direction, i.e., by a simple vector. The result
of such an operation
is
a vector which represents the result of
a similar operation on the quantities themselves.

It is often
very convenient to determine the result of opera-
tions performed
on alternating quantities by performing the
same operations on vectors representing their effective values. This method is used for convenience only, and its limitations,
as given above, must always be kept in mind.
Illustrations
of the statements of this section will be given in the course of
the discussions of alternating currents.
CHAPTER
V,
CIRCUIT.
LAWS OF THE ELECTRIC
16.
17. 18. 19.
Consumption of e.m.f. in single circuits. a. E.m.f. consumed in resistance. b. E.m.f. consumed in inductance. c. E.m.f. consumed in capacity. d. E.m.f. consumed in (o), (b\ and (<) combined. Problems in single series circuits. Problems in single multiple circuits. Consumption of e.m.f. in single and multiple circuits in
series.
16.
cuits.
Consumption of Electromotive Force in Single Cira. Impressed E.m.f Consumed in Resistance,

.
E=Ir
(34)
Resistance consumes e.m.f.

in direct proportion to the value
of the current.
It follows
that
where the current has the sine

form,
the
is
e.m.f.
consumed,
to
which
proportional
all instants,
the
current at
must also
plotted.
give a sine
wave when
F1G.49Showing E.m.f. Consumed

Resistance.
in
^hus,
in
Fig. 49, let the sine

Current,
/,
wave of
which when plotted
be estabat
lished through a resistance of three
ohms.
in
The e.m.f
any
instant
is
ir
3/,
the same manner
gives a sine wave, E,
having three times the height of the

68
i6j
LAIVS OF THE ELECTRIC CIRCUIT.
69
current wave.
Under these circumstances the

is
e.m.f.
and cur-
rent are in phase.
When an
alternating current
it is
it.
established through a circuit
containing resistance only,
in
phase with the impressed
pressure which establishes
A
it
if
current which
is
alternating has the value of unity
when
at
liberates heat at the
same
rate as
would a
unit direct current
passed through the same resistance.

is
Since the rate
which heat
liberated at
any
instant
is
proportional to the
it
square of the instantaneous value of the current,
follows that
the average square of the instantaneous values of an alternating

current
is
equal to the square of a continuous current which
has the same heating value.

the
Therefore the "square root of
mean
square,
' '
or the effective value of an alternating cur_/), is
rent (see Section 10,

alternating
the value
commonly
used.
For an
only,
current in
a circuit containing resistance
Ohm's law
sumed by
holds for effective values of current and e.m.f., as
well as for instantaneous values.

resistance
is
The
effective e.m.f.
con-
E=
Prob. 15.
Ir.
An
}
electric heater
has a resistance of 18 ohms.

in
it
How much
e.m.f. will be
consumed
18
when
it
carries a cur-
rent of 6 amperes
Solution:
b.
E=
Ir
=6X
108 volts.
Ans.
Impressed E.m.f. Consuined in Inductance.
E J2nfLI.
From
the definition for the unit of inductance, the e.m.f.
due to an inductance of
of current in the
units
is
times the rate of change
circuit, or
'--^Tt
(35)
di
[i6
In this instance e and
-j-
have opposite
signs, since the
e.m.
of inductance always opposes a change in the current.
Assuming
i to
be a sine wave of current,
=A
sm
-j^t,
(36)
whence
^^=A-^cosî
and
e
27r = LA^p- cos
,
(37)
2it -Tpt
(38)
That
wave.
is,
is
a sine wave of pressure 90 behind the current

is
An
equal and opposite wave of e.m.f.
consumed by
written
the inductance.
Letting e represent the consumed e.m.f.,
it
may be
LA-jT cos
-jrt,
(39)
which
is
a pressure wave 180 ahead of the self-induced e.m.f.,
or 90 ahead of the current.
The component of impressed e.m.f. which is consumed by the inductance of a circuit is go ahead of the current flowing in
the circuit.
The maximum
But the
value of this e.m.f.
is <?,
maximum
value of the current
is
= LA-7=. /_ = A.
Therefore
i6]
LyifVS
OF THE ELECTRIC
CIRCUIT.
Dividing this equation by V2, and writing
E for
^-^^ and
V2
I for -^^,
V2
this
becomes
E=
where
tively.
^rLI,
and current respec-
and / are
effective values of e.m.f.
Or, writing
/=
-^,
E=
where /"is the frequency.
2TtfLI,
(40)
As
by the
the vector representing the pressure
wave
is
90 in
advance of that representing the current wave,

factor/, thus:
this is indicated
E =j2nfLI.
The
SO.
relations given in this equation are illustrated in Fig.
Fig. 50.
Showing E.m.f. Consumed in Inductance.

by
unit current in an electric circuit in
E.m.f. developed
quadrature with
it is
called reactance.
is
Thus, in a
circuit containing inductance, the reactance
equal to the volts per ampere developed or consumed by the

inductance.
Its
value in such a circuit
is
2nfL.
f2
Prob, i6.
cuit of .2
[i6
cir-
What
wijl
be the consumption of e.m.f. in a
henry inductance and in which the current has an
effective value of 5
amperes and a frequency of 60 periods per
second
Solution: Substituting in the expression
E ^j'ZTt/LT,
volts,
=72 X
TT
60
.2
=73 77
where/ indicates
of the current.
c.
that the e.m.f. of 377 volts
is
^0 in advance
Impressed E.m.f. Consumed
in Capacity
When
the electric circuit

is
is
closed through a condenser

is
whose capacity
C farads
the developed e.m.f.
'=- C c
Q_
where
Q is
the charge in coulombs present in the condenser.
i6]
,-

circuit.
73
duces due to the reactance of the

of this e.m.f.
will, therefore,
The maximum
value
be 90 in advance of the current
and
its
value will be
-^^
-f ^^^^-
^^"
\2
"^
TV
Sec.
1 1
<?,
eq. (i 3)). (43)
Substituting,
27 n
f=-^max. ^1"
In
\7 H~
27t
4^^,
27t
?^ ^1" "r '*'
The corresponding
effective values are
Let -Tph^f,
Then
J--^'
^=J^''
e.m.f.
....
(44)
The impressed
consumed by the capacity reactance
is opposite to the reactance e.m.f., or is
The component of impressed e.m.f. which

capacity reactance of a circuit
in the circuit.
is
is
consumed by the
go behind the current flowing
These
Prob.
circuit
relations are
17.
shown
in Fig. 5
What
will
be the consumption of e.m.f.
in a
of
it.?
.5
which has a capacity of 2 microfarads, when a current ampere with a frequency of 60 p.p.s. is passed through
74
Solution: Substituting in the expression
[i6
E=-J 2
E=~J:2
=
The
d.
71
fC
I,
;r
-60 -.000002
volts.
7663
Ans.
is
factor
/indicates that the e.m.f. of 663 volts

^
90
behind the current.
Impressed E.m.f. Consumed by Resistance and the Reactances of Inductance and Capacity in Series. By combining the
results of Sections a, b,
and
c,
the e.m.f. consumed in a circuit
Fig. 51.
Showing E.m.f. Consumed in Capacity.

inductance,
containing any combination of resistance,

capacity in series
and
may be
obtained.
Thus,
for the
general case,
E=
To
rl -\-j2nfLI j
27tfC,
is
(45)
simplify this expression the letter x'

I
substituted for 2nfL^
and x"
is
substituted for
2nfC'
Then
(46)
E=
I{r
+jx' -jx").
6]
75
These
relations are
shown
in Fig. 52.
Ei=l'
>]e3-=I( Ja^^
C"
2t/C
rmmhtr^smLRESISTANCE
INDUCXANCE
M^
SOURCE OF S.M.F^
'
Fig. 52.
If
is
the effective current in amperes,

that the
it is
seen from the

' '
equation
quantity in
brackets
is
the
volts
is
per
ampere
'
'
consumed
in the circuit.
This quantity
termed
the impedance of the circuit.

If the
impedance of a
circuit
be denoted by
!,
this equation
may be
or
written
f^,
(47)
-|.
which
is
(48)
analogous to Ohm's law.
76
17.
[i/
It will
Problems in Single Series Circuits

is
be noted
that in single series circuits the current
the
same throughout
and the e.m.f.s are added.
The
various e.m.f.s
may be
con-
veniently located in phase position
by reference
to the current.
All of the quantities in this equation
may be treated
as vectors.
Graphical solutions of the following problems will ordinarily
be most instructive, and

the results obtained
ciently accurate.
if
the construction be carefully made,

will
is
by the graphical method

the algebraic solution
be
suffi-
When
attempted,
the student should carry along an approximate graphical solution at the
same time by means
of a free-hand sketch.
In this
manner
e.m.f
errors in determining the phase relation of current
and
may be
largely eliminated.
circuits,
All problems in simple electric

rent, e.m.f., or
where
either cur-
impedance
is
to be determined,
may
be solved
by means of the general equation
or,
for the
particular case of a continuous
current,
by the
equation
r
In the general case, the effective values of / and
dealt with,
are
and these may be represented by vectors.

is
Since
follows
the impedance, z,
that
tion
a vector which relates
/ and E,
it
any of the operations necessary to the solution of equa(45) may be performed upon the vectors representing the
quantities,
and the
result will be
a vector representing the
true result of the operation.
(See Section 15.)
This fact makes possible the ready solution of problems
which would otherwise

difficulties.
present
considerable
mathematical
in
The
solution
may
either
be graphical,
which
I'/J
L/iH^S
OF THE ELECTRIC
CIRCUIT.
77
case the vectors are drawn to scale, or analytical, in which

case the results of the geometrical operations are obtained by
making use
methods are
Prob. i8.
of the
algebra
of
complex
quantities.
Both
illustrated in the solutions that follow.
When
is
the e.m.f applied at the terminals of an

i
incandescent lamp
is
lo volts and the resistance of the

will flow
?
lamp
200 ohms, what current

Solution:
200 ohms.
no /= 200 =
Prob. 19.
.55 ampere. '
Ans.
furnished
current of 105 amperes
is
by the
.
armature of a railway generator that has a resistance of
ohm.
What
e.m.f.
is
consumed
in the
armature
.'
Solution:
E=
Prob. 20.
si
.15
105
15.75 volts.
Ans.
What
e.m.f. to set
must be furnished
at a frequency of
60 cycles per second

a resistance of 175
Solution:
up a current
of 8
amperes through,
.5
ohms and an inductance

z
of
henry
r -\-jx'.
E = Iz = I-ir +jx')= S-{i7i> +J2n.6o.S) = 14004-/1508 _ /v^îppa = 2059, / tan-' ^^'

'
1400
volts
=
The symbol
2059, /47"
8', volts.
Ans.
8' is
indicates that the
angle of 47
positive, or that the e.m.f.,
2059
volts, leads the current
by
78
8'.
[17
an angle of 47
The same symbol
inverted,
is
used
to denote a negative angle.
The
significance of the results
obtained above
53.
is
shown
of
graphically in Fig.
the e.m.f.s
The sum
in
OE^ consumed
consumed
in resistance
and
OE^,
inductance
gives the e.m.f.
OE, occupying a time
position 47" 8' ahead of the e.m.f.
OE^,
which
is
in
phase with the current.
The
alternating pressure applied to the circuit is numerically equal to
2059 volts and leads the current which it establishes
by 47
8'.
The phase
position of the current in this diagram

It is
is
given
to a
by the vector having the closed arrow-head.

larger scale than the e.m.f. values.
drawn
Prob. 21.
At what
pressure and at
what angle of lag
will
the current in the preceding problem be established
when
the
frequency
is
(a) 125 cycles
.'
{b)
25 cycles
.'
{a)
(b)
3442,
1534,
/6i 58\
volts.
Ans.
/24
7' , volts.
Ans.
Prob, 22.
An
alternator develops a terminal pressure of
2000
if
volts at a frequency of
60
cycles.
What current will

ohms and a
flow
the circuit contains a resistance of 100

.'
capacity
of 50 microfarads
Solution:
i=^ = r -jx100
2000
J 27r.60.5O' io~*
1
2000
2000
100-^53
II2.7\2{
7-7 5,
/s 8,
amperes.
Ans.
'7]
79
The elements
diagram Fig.
of this problem are given graphically in the
54.
The
alternat-
ing current generator establishes
through the resistance of 100 ohms
and the condenser having a capacity

of 50 microfarads, or 53
ohms of
reactance at 60 cycles, a current
of 17.75 amperes in a phase position of 28 in advance of the
generator pressure.
The
resistance
rnjwT
1=17.75,;
consumes 1775 volts, in phase and the condenser 940 volts in lagging quadrature with the current. The quadrature sum of these two presProb. 23.
2000 VOLTS
Fig. 54.
sures equals the generator pressure of 2000 volts.
What
are the
corresponding solutions of the

is
preceding problem when the dynamo frequency

cycles
.'
[a)
125
(1^)
25 cycles
()
.'
19.4, 12.3,
X^4-25 amperes. Xsi-Ss", amperes.

,
Ans.
Ans.
Prob. 24.
In Fig. 55 the dynamo. A, generates a total

pressure of 2200 volts at a frequency
of 125 cycles.
R -WNAA/
L
-Mumsj
Fig. 55-
C
I
The
resistance of
its
its
armature
is
is
ohm, and
inductance
.01 henry.
L=
is
.05 henry, 7?
50
ohms, and
current
.'
C=
20 microfarads,
What
the value of the
Solution: Total impedance
=
./
;r.
,,.,.,,. + 50)+X-05 +
= =
51 SI
lOOOOOO
\ )
.01)2.
125
-j[ ô.2rr.i2s
+747-124
-yi6.S37
-763. 661
8o
[iS
= S3.6i4,\tan-".3243 = 53.6i4,\i7S8', ohms;

current
2200
=41.034, /ly 58
,
,
amperes. Ans.. ^
53-614X1758'
^
18.
Problems in Simple Multiple Circuits.
In
simple
nNiltiple circuits the various branches of the combination are
subjected to the
same
e.m.f., but the separate currents
and
their phase positions with respect to the e.m.f.
depend upon
the impedances of the individual branch circuits.
Hence the
total cur-
various currents are referred to the e.m.f. as a phase standard,,

for the e.m.f. is the
same
for all the circuits.
The
rent
is
the geometrical
sum
of the separate currents.
In determining the total impedance of a combination of this
kind the reciprocals of the impedances of the branch circuits

are added, with due regard to their relative phase positions.
These
reciprocals of the
impedance are known as the conducset
tivities of the circuits, hence the measure of the conductivity
of a circuit
is
the
number of amperes
This
is is
up
in
it
by one volt of
applied pressure.
circuit,
in contrast to the
impedance of the
to
the measure of which
the
number of volts required
set lip one
ampere.
When the combined

its
conductivity of several
branch
circuits
is
has been
obtained, the
corresponding total
is
impedance
simply
reciprocal.
This
illustrated in the
following problems.
Prob. 25.
Two
is
incandescent lamps, having resistances of

in
100 ohms and 200 ohms, are

mains.
is
multiple
across
no-volt
What
the impedance of the combination, and

."
what
the strength of the total current

Solution:
The
I
conductivity of branch
(i") ^ '
is
100
3
.
that of
branch
(2)
^ '
200
and the
total conductivity
^
is
200
The
ioint '
8]
l/liys
OF THE ELECTRIC
200 -
CIRCUIT.
8i;
resistance
is,
therefore,
or 66.66 ohms, and the total cur-
rent
is
,, ,^
" =1.65
amperes.
Ans.
is
Prob. 26.
A pressure
R=
C=
and
of 1000 volts at 60 cycles
applied
at each of the terminals of
R, L,
and C, Fig. 56.
120 ohms, a^
2 5 microF"^- S^.
henry,
1!
farads.
The
joint
impedance and
total strength of established current are required.
Two
solu-
tions will be given.
First Solution:
T
1
1000
"^ "120
"^ ^-33
amperes,
1000
^^
^j-27f6o..s
-yS-SOS amperes,
J.
/j
1000
=79.425 amperes,
'^27r-6o-25- iQ-*
2/j,
I^, /j
= 8.33 +J4..12 = 9.30,

1000
"^ 107-37.
/26ig' amperes.
,
Ans.
^ 9.30 y26i9
\26i9', ohms.
Ans.
Second Solution: The impedances of the several branches

are, respectively,
120 ohms,
j'2n-6o.^
=ji8g ohms,
B
and
J
'
2;r'6o-2S.io-
= ;'i05
-^
'
ohms.
The
conductivities are the reciprocals of these values, or
.0083 mhos,
/ 00529
""
mhos,*
'-ee
For
(iffiniti'in
ui mho,
page
83.
82
[19
and
^.00943 mhos, respectively.
The
tivities,
total conductivity is the
sum
of the separate conduc-
or
is
00833
^00529 +700943, = .009, ^26
ig' ,
mhos,
or amperes per volt.
The impedance
therefore,
of the branched portion of the circuit
is,
.0093 ^26
19'
7=
107.5,
\26
19', volts
per ampere.
The
current
is
E = = 9.30, /26 19', amperes.
Ans.
The dynamo armature in the preceding problem has a resistance of 5 ohms and an inductance of .05 henry. What is the total impedance of the circuit and what e.m.f. is developed by the dynamo ? The impedance of the armature is 5 +718.9 ohms, and the
Prob. 27.
total
impedance
is 5
+ 96.2 +7(18.9 47.9) =

=
104.3,
101.2
729
\iS
59'.
is
ohms.
Ans.
9-3
The
e.m.f. developed
by the dynamo
104.3
= 969,
19.
\i5
59', volts.
Ans.
Consumption
of Electromotive Force in Single
and
Multiple Circuits in Series.

going problems
a.
b.
it is
From
the solution of the fore-
seen that:
Single circuits are treated
Multiple circuits are treated
by means of their impedances. by means of their conduc-
tivities; i.e.,
the reciprocals of the individual impedances that
are in multiple.
=
19]
c.
LyiH^S
OF THE ELECTRIC
CIRCUIT.
83
Single
{b~).
and
multiple circuits in series are treated as in

circuit
()
and
The impedance of a single

is
which
is
the
final
equivalent of the multiple circuit

solution
is,
then determined.
The
therefore, that of
an equivalent single
circuit
by
means
of impedances in series.
practical convenience the
For
term ohms
in
is
denote the volts per ampere consumed

dance, although the unit
a circuit
now used to by impe-
ohm
is
defined as the resistance which
consumes one volt per ampere.

Likewise the term mhos
is
used to denote the amperes per

circuit,
volt established in an electric
although the unit
mho
is
defined as the unit of conductivity equal to the reciprocal of

the ohm.
Prob. 28. In the circuit of Fig.
57
the
.
constants
are
as
follows:
Z=
henry,
R=
20 ohms, R'
=
JGENERATOR
30 ohms,
resistance
$0 microfarads.
is
The
cir-
of the line
total
negligible.
R'
Find the
cuit
impedance of the
to the
external
dynamo
at
i-V\/V\AJ
Fig. 57,
frequency of 50 cycles per second.

Solution:
Impedance of L " R
" R' "
= =
"
j2nfLt
=y3i. 416 ohms. " 20

30
"
C =
77='
763.69
"
Since the two branches R' and

joint
C"
are in multiple, their
impedance must be determined.
This
is
the reciprocal
of the joint conductivity.
Conductivity oi R'
.0333 mhos.
"
"
c =y.ois7
"
84
[19
Joint conductivity of R' and
C=
^tan-i-^^, mhos
/25i5^
0333+/0157,
= =
.0368,
.0368;
Joint impedance of i?' and

I
C=
o-7 .0368/25 IS
27.17,
\2S
15',
ohms.
In order to add this quantity to the impedances of
and
it is
best reduced to components.
27.17X25
15'
is
24.56 -yii. 58.
The
total
impedance
the
sum of
all
the impedances in
series, or is
.3-^=y3i.4i6
+ 20 + 24.s6 -yii.58
/tan-'
i',
= 44.56 +719.836 = 48.8,

=
Prob. 29. If the
^P^^
ohms.
Ans.
48.8, /24
in Fig.
ohms.
dynamo
57 delivers a terminal
pressure of 2000 volts, what current will flow in the circuit,
and what
is its
phase position relative to the terminal pressure

current in amperes
is
Solution:
The
equal to the impressed
pressure divided
by the impedance.
Taking the phase of the
impressed pressure as the phase of reference,
/=
2000
48.8 tan-i(^^^
g-
40.9 tan->|^-
^,
amperes,
=
That
is,
40.9,
\24
I',
amperes.
Ans.
the current lags behind the impressed pressure

i'.
by an
angle of 24
Prob. 30. Find the e.m.f at the ends of L, Fig. 57, and
find its
phase position both with respect to the current flowing

to the e.m.f. at the
and with respect
dynamo
terminals.

19]

Solution:
85
Its
The
current flowing
is
40.9 amperes.
pressure
is
phase
position with respect to the
dynamo
lagging by the
angle
19-8 tan-i-^
1
44.6
24
i'.
The
pressure used up in sending this current through / is
El
= 40.9
X73I-4I6 =yi284.9
is
volts.
The
factor
indicates that this pressure
90 ahead of the
i'
current flowing.
Since
it
the current
is
24
behind the
dynamo
volts, is
pressure,
follows that the pressure E^, of 1284.9
(90
24
I')
65 59'
ahead of the dynamo pressure.

Prob. 31. Find the e.m.f. with
its
Ans.
phase position relative
to the current, (a) at terminals of r ; {b) for multiple portion of

circuit.
Solution: (a)
Ey= 2QX
40.9
{b)
=818 volts,
in
phase with the current.
Ans.
E^
T.
J -0157^ = 40.9 X 27.17; tan->(^- -^j = II. 25, \25 IS', volts. Ans.
1 1
Check
is
Since the total pressure at the

in
/,
dynamo
terminals
used up
r,
and the multiple branch, the sum of the
separate e.m.f.s found for these portions should give 2000

volts, leading the current 24 \'.
The sum
^1284.9
or 2000;
is
+ 818 +
1004.S
-^473-6
1822.5 +y8ii.3,
tan-* -5
1022.5
= 2000, /'24
i',
volts.
Ans.
S6
Prob. 32. If the generator armature, Fig.
[19
57, has resistis its
ance of 4 ohms and an inductance of .008 henry, what
impedance
Solution:
s,
= =
A +y2;r.50'.oo8
or
4-72,
4 4-/2.51 ohms, /32" 7', ohms. Ans.
Prob. 33. Using the answers to problems 28 to 32, find
the impedance of the complete closed

generator, and find the ôâ/
Solution:
circuit,
including the
dynamo
is
e.m.f.
The
total
impedance
Si = 44. 56 +719.84 + 4 +72.5 = 48.56+722.35, or 53.45, / 24 40',
ohms.
Ans.
The
40.9
total
dynamo
pressure
is
53-45 /'24''4o'
Prob. 34.
^/24^j^ volts. Ans. Analternator is delivering 250 K.W. of electrical

lamps over a transmission
line at
= 2186,
power
to incandescent
12,000
volts pressure, 50
p.p.s.,
and the current lags 25"
50' behind
the e.m.f
<()
Neglecting the impedance of the generator, find
the current and (d) the impedance of the circuit.

(a) 23.15,
(^)
518.4,
\25 /25
50',
amperes.
50',
ohms.
Ans.
(c)
Prob. 35. Find (a) the resistance; (d) the reactance;
the inductance of the circuit in the preceding problem, assuming the capacity of the line to be zero.
{a)
(<5)
466.56 ohms.
Ans.
"
7225.89 ohms.
.719 henry.
(c)
"
Prob. 36.
The "charging
current"
current " * on one of the underis
ground cables of Cornell University

*
.2
ampere, at a pressure
into the cable
By "charging
the circuit
is
is
when
open and subjected
meant the current which flows to normal working pressure.
19]
LyifVS
1
OF THE ELECTRIC
CIRCUIT.
p.p.s.
87
of
100 volts and a frequency of 125

?
What
5
is
the
capacity of the cable
2314 microfarad.
Ans.
amperes
50 micro-
Prob. 37. Find the voltage necessary to send
through
this circuit (Fig. 58).
R=
10 ohms,
r-^mm^
-A/WWh
Xalternator
-^(ny
Fig. 58.
A/wv\^
Fig. 59.
farads,
.5
frequency
its
lOO p.p.s.
is
The dynamo
resistance
is
ohm, and
inductance
.05 henry.
52.535,
\2
15', volts.
Ans.
Prob. 38. Find the e.m.f. which must be generated by the

alternator in Fig. 59
when C
microfarads,
R=
ohms,
Z=
.5
henry, frequency
resistance
=
is
100 p.p.s., current

.5
amperes.
is
The dynamo
henry.
ohm, and
137-4.
its
inductance
.
.05
/82"
40' volts.
Ans.
fmw-^^wffp^
M
Fig. 60.
TRANSFORMER
Prob. 39.
At what
frequency would the reactance due to

just neutralize that
an inductance of .4 henry
due to a capacity
of
5
[19
microfarads
The
capacity and the inductance are in
series.
ii2.5p.p.s.
Ans.
Prob. 40, (Fig. 60.)
Given: Frequency
=
i?
112.5 p.p.s.,
e.m.f at transformer terminals
100 volts,
10
ohms,
{d)
Z=
.4 henry,
{a)
C=
{e)
^ microfarads.
Find the following quan{c)
tities:
Impedance,
c.
{b) Current,
d.
E.m.f. a to
b.
c.
E.m.f. b to
E.m.f. c to
(/) E.m.f. a to
{)
E.m.f. b to d.
Neglect the impedance of the transformer.

(a)
Ans.
19]
LAiyS OF THE ELECTRIC CIRCUIT.

Current in
89
circuit y2.55S Prob. 43. Find the voltage which would send 10 amperes
" "
" "
C L
=j.62Si amperes. Ans. " = J3.182 " " " =
through the
acteristics:
circuit in Fig. 63,
which has the following char-
L" = ohms
tance.
.03
= 60 p.p.s., L' = .05 henry inductance, henry inductance, i2' = 5 ohms resistance, R" = 7
Frequency
Neglect the armature resistance and induc105.46,
resistance.
/24
10', volts.
Ans.
(Wmj\
AVWV^ h:
<a>
Fig. 63. Fig. 64.
Prob. 44. Find the impedance of the circuit shown in Fig.
R = $ = 20 ohms, L = .02 henry, C = 20 microfarads, frequency = 120 p.p.s. Neglect armature impedance.
64,
and the phase relations of e.m.f. and current.
ohms, R'
24.72, /'2" 47', ohms.
Ans.
Prob. 45.
65) R' L'
In this circuit (Fig.
=
I
50 ohms,
henry,
R"
= =6 C
L"
= =
35 ohms,
.5
henry,
10 mi-
rNm-mmh
c"
microfarads,
C"
50.
crofarads, p.p.s.
total
Find the
impedance, neglecting arma-
AWW<n>
Fig. 65.
ture.
356.2,
\7242', ohms.
is
Ans.
a diagram of an
Prob. 46. Fig. 66
electric circuit
made
go
[19.
up of several multiple branches and single
circuits in series.
The
constants of this circuit are as follows:

Circuit
"
No. "
" " "

Frequency
i?
= = " 3. L' = "4a. 1= 4b. C" =

i
.
i?
100 ohms,
2.
R'
50
"
"
Z = 2 henrys. C = 100 m.f,
.05 henry,
.1
50 m.f.
60
p.p.s.
{a)
Generator
is
= 20 " r = 20 ohms. / = 30 " circuit, Z. = .01 henry,

C'
is
?
o. 5
ohm.
What
?
the impedance of the circuit
external to the generator
{b)
What
the total impedance

{c)
of the circuit including the generator
When
(d)
?
the gen-
erator delivers 2000 volts to the line terminals of the external

circuit,
what
will
be the current
in
amperes
what
e.m.f. in volts
must the generator develop
To do this (1?) What
phase position
will the current

?
occupy with reference to the

will
external or line pressure

tion
(/) What
be the phase posi-
of the
current with reference to the internal or total

?
generator pressure
Fig. 66.
{a)
{b)
{c)
29.575,
29.678,
68.32,
{d) 2027.5,
ie)
\9 \i /9 \i
31*'.
32', 1' 31 ,
32'.
ohms.
Ans,

"
amperes.
volts.
"
'9 31^'
ahead.
(/)
i32'
1 9]
LAiyS OF THE ELECTRIC CIRCUIT.

is
Prob. 47. In an electric circuit the impressed e.m.f.
E=
(190,
+ 6O3 + 30,) +>(- 36j + 73 + 3sj

is
and the impedance
Resistance
10
S,
ohms
ohms
"
Induction reactance
IS3
5^
ohms
Condenser reactance
33
I.
"
?
What
is
the effective value of the current

:
Solution
I :z
^^^^
+ ^' + 30.) +y( 361 + 73 + 355)

lo^;
16.73; 26.0,
'
I=Vi9' + 2,-S^+
i-iS'
+ 3-6^+ -419' +1-35'-.
/=
19.7s amperes;
Ans,
CHAPTER
VI.
ELECTRIC POWER.
20. Function of the electric circuit.
Average power with sine-form e.m.f. and current in phase. Average power with sine-form e.m.f. and current in quadrature. 23. Average power with sine-form e.m.f. and current not in phase and not in
21. 22.
quadrature.
24.
25.
Average power with non-sine-form
e.m.f.
and
current.
The
equivalent sine wave.
20. Function of the Electric Circuit.

circuit
is
The function of an
exactly as a
electric
to convey
electric
power
kinematic chain of connections conveys power from a source

to a point of application, except that in the former case the
action
is
molecular, while in the latter there
is
mass motion.
but must
Electric
power cannot be used
in a circuit as such,
state,
be transformed into some other

is
and
this, transformation
always accompanied or indicated by the consumption of
e.m.f. in the circuit.

in
The sum
of
all
of the
consumed e.m.f.s
is,
a circuit
is
equal to the impressed e.m.f., that

the- source of
the e.m.f.
furnished
by
power.
As an
illustration,
take the
case of a circuit supplying power to an electric motor.
The
impressed e.m.f.
is
consumed
in the resistance of the circuit,
dynamo armatures, and in overcoming the counter e.m.f. accompanying the transformation from electrical into mechanical power in the motor, thus
including that of the motor and
accounting for
all
of the e.m.f. furnished
by the generator.
92
; ;
21]
ELECTRIC POWER.
93
elec-
The
and
electric
power which
is
being transformed in an
to the
tric circuit is, at
any instant, equal
product of the e.m.f
the current.
This product
tinguish
it
may be
called the instantaneous

is
power to
dis-
from the average power, which
more commonly
met
any
with.
circuit,
In order to determine what this average will be in

the law of the variation of the power must be
the average can be determined
known and then
by
a simple
integration, either graphical or analytical.
In the following
paragraphs are discussed a number of special cases, leading
up to the general case with which the treatment

21. Case I.
closes.
Average Power with Sine-form E.m.f. and

(Fig. ^j.")
Current in Phase.
Fig. 67.
Power.
Current and E.m.f. in Phase.
Let
eax..
= maximum value of e.m.f. maximum value of current 4iax. w = instantaneous .value of power = angle between the generating {)
of the
vector of the sine
curve and the reference axis.
The
instantaneous values
power wave
will
be the
94
[21
products of the corresponding instantaneous values of current
and
e.m.f.
w = (^max. sin )(w.

sm^Q
Substituting (50) in (49),
^
sin ^)
I
= ^. 4, sin*
(49)
,
.
cos 2^
(SO)
cos 2^
;

^max. ^max.
(5^)
By
reduction,
^ = ^max. ^^max.
The power
is
^max. ^max.
'ÔS 2x}
r~is
/,_\ (52)
curve, from equation (52),
a cosme curve of
it is
twice the frequency of e.m.f. and current, and
symmetrical
Z
"'"''
about an axis which
located a distance of
""'
above
ordi-
the axis of the e.m.f and current curves.
The average
nate of this power curve, measured from the axis of the e.m.f.
and current
axes.
is
curves,
is
equal to the distance between the two

it
By
separating this average ordinate into two factors
found to be
made up
of the effective values of e.m.f.
and
current.
fe)(%)=-^^'
(53) Fig. ^j
and / are
effective values of
e.m.f and current.
shows the
relations of current, e.m.f.,
and power curves.
The
and
follows
I.
conclusions reached for the power of sine-form e.m.f.s
currents,
when
these are in phase,
may be
stated as
The average power
is
the prodtcct
of the
effective values
of e.m.f.
and current.
2 2] 2.
ELECTRIC POWER.
95
The power curve

it
is
one of twice the frequency of e.mf.

axis parallel to that of
to
and current and having an

current, located above
e.mf and
.
a distance equal
.
one half the product
of the maximuni values of e.mf and current. 22. Case II. Average Power with Sine-form E.m.f. and
Current in Quadrature.
(Fig. 68.)
As
in
Case
the instan-
FiG. 68.
Power.
Current and
E.in.f. in
Quadrature.
taneous values of the power curve are the products of the

instantaneous values of current and e.m.f.
w (e^^^,
sin &)(4iax. cos ^)
= e^^^_
4,
sin
cos &.
(54)
sin
cos
&
sin 2-9
(SS)
substituting
w=
The power
'^max. "max,
sin
29
(56)
curve from equation (56)
is
a sine curve of twice
the frequency of e.m.f. and current and
it is
symmetrical about
Its
the axis of the e.m.f. and current curves.

nate, referred to the axis of the e.m.f.
zero.
average ordiis
and current curves,
Fig. 68
shows
the relative
phase positions of the comConclusion:

in
ponents and of the product curves.
The average power transformed
circuit in
is zero.
which sine-
form
e.
m.f.
and current
are in quadrature,
96
23. Case III. Average
.
[23
Current neither in
Power with Sine-form E.m.f and Phase nor in Quadrature. (Fig. 69.)As
Fig. 69.
Power.
Current and E.m.f. not in Phase and not in Quadrature.

.
^>
before, the instantaneous values of the power are the products
of the corresponding e.m.f. and current values
^ = (^ma;t.
where
ff
sin
i&)(w.
sin {&
e))
(57)
is
the phase difference between e.m.f. and current.

less
As
III
is
is"
an angle always
than 90,
I
it is
evident that Case
intermediate between Cases
and
II,
and hence the
average product will be between the values found in these
extreme cases.
As
there
is
a phase difference of 6 degrees

evident that the current
in
between e.m.f and current,

and one
in quadrature with
it is
may
be resolved into two components, one

it,
phase with the e.m.f
and that the sum of the average products of these components with the e.m.f will give the
23]
ELECTRIC POIVER.
97
average product of the original quantities.

the current in phase with the e.m.f
^''
The component
of
is
= W.
cos
i9
sin
i9
......
(58)
and the component of the current

e.m.f.
is
in
quadrature with the
i"
= W.
sin
19
sin
(59)
As the quadrature product will be zero, as is shown in Case II, it may be neglected and the average power may be deduced
from Case
I.
For
this
component of
'
current,
-8
^max. sin
i^^_ cos 6 sin

18
= ^max.
but
4ax. sin^
COS
(9
(60)
sm3
^ 0=
""
cos 2i9
^^
(61)
I
'
^max. ^max.
COS 2^ '
COS ^
^^
. .
.
(62)
w =: EI COS
(see case I)
(63)
Stated at length, the formula just deduced becomes:
The power transformed in any circuit in which sine-form e.m.f. and current are present is the product of the effective values of e.m.f. and current and the cosine of the angle of phase
difference between them.
To
the product of e.m.f. and current in such a circuit
is
name apparent power to distinguish it from the real power given in Case III. The ratio of the real to the apparent power is given the name power factor, which is, therefore equal
given the
to the cosine of the angle of phase difference
between current
and e.m.f.
The apparent power may be
considered as com-
;; ;
98
[23
posed of two factors known as the real power and the wattless component, sometimes called the wattless power.
The
former of these
is
equal to
EI cos
the latter being
0,
EI sin
The term
9.
cos 6 in the power equation
Power
is
z=.
w = EI cos
d
sin
in
called the
power
factor,
and the term
the equation
wattless component
is
= EI sin
called the wattless factor.
In circuits containing only reactance and resistance, and

therefore those in which all of the
heat, a
power
is
transformed into
number of interesting
relations hold, as follows
Let r = resistance
X = reactance s = impedance
J = symbol
by
it.
indicating the quadrature relation of the

it
term following
to those terms not preceded
^=7
Expressing current
equation (57) in components with respect to e.m.f,,
in
its
(^4)
two quadrature
Multiplying through
by
e.m.f.,
EI=EI^^jEA
(66)
24]
ELECTRIC POIVER.
99
Since
EI
is
in
quadrature with the e.m.f.
it
represents no
circuit
power, so that the average power transformed in the

into heat
is,
W
value of I5S7 volts
z=
EI = EI cos
{6y)
Illustrative Problem.
A sine-form of
What
e.m.f. of a
maximum
when impressed upon a
certain circuit sets

14.
i
up therein a current of a maximum value of

wattmeter reading average watts placed
amperes.
in the circuit
shows a
Ans.
power consumption of 10 k.w.

the circuit
.^
is
the power factor of

91. ij^.
n^
24. Case IV. Average Power with Non-sine-form e.m.f. and Current. The harmonic component method of analysis
finds
an application
in
determining the power transformed in a
circuit in
which a non-sine-form e.m.f. establishes a non-sineSince

alternating values with
different
fre-
form current.
quencies produce an average product of zero, and since the
power
in
an alternating current
circuit is the
average of the
it
products of the instantaneous values of e.m.f. and current,

follows that the power
is
equal to the
sum
of the products of
the effective values of the harmonic e.m.f. and current com-
ponents and their respective power
factors,
W = e^^ cos + e^^ cos +

^j
9^
ej,^
cos
^5 -j-
(68)
In a circuit where
all
of the
power
is
lost as heat this equation
may be
written
^=^/l7+^3^3f +^5^5^ .^4fAs a

covers
(69)
general conclusion
it
should be stated that Case III
all
the requirements of practice except where an
100
[25
desired.
analysis of the form of the
power curve
in
is
As
all
measuring instruments are calibrated

rent and e.m.f.,
it
terms of effective cur-
follows that irregular current and e.m.f.
curves can be considered as replaced by equivalent sine curves,

in
which case the conclusions drawn

application.
It
in
Case III are capable
of general
should be noted also that the

differ
common- form of
25.
e.m.f.
and current curves does not
greatly from that of the sine curve.
The Equivalent Sine Wave.
An
inspection of prob-
lem 47 shows that where the impressed e.m.f. in a circuit does not have a simple sine-form and is therefore irregular the
current form
e.m.f.
is
likewise irregular.
will
The
effective values of the
and current
be
(70)
E = VE,^ + Ei-\-E,^+
and
/=V/i-f//H-^' +
These values are indicated
ment.
directly
(71)
by any properly con-
structed and calibrated alternating pressure or current instru-
Likewise- any corresponding wattmeter placed in the
circuit indicates directly the
power
in
watts.
Thus
for the
general alternating current circuit in practice three out of the

four quantities in
power equation,
W = EI {Power Factor) ....

are easily determined, and the fourth, power factor, equals
(72)
Power Factor =
In the general case
-^
(73)
-vry does not actually equal cos d.
2S]
ELECTRIC POIVER.
is
lor
In fact there
no
definite value for
with which to express,
the angle of phase difference of two irregular waves.

irregular
Such
of
waves are made up of a number of components

different angles
wherein corresponding components have

phase differences and frequencies. phase difference can,
See Section 13.
No
one
strictly speaking,
be applied to a pair of
aggregations of such components.

Practical requirements
demand, however, the convenience

difficult
of an equivalent angle of phase difference in lieu of a multi-
valued or composite angle
of expression or even of
comprehension.
Thus the custom has been formed of making
in the geneirJ case
Power Factor
cos
d.
From
this
it
follows that the
power equation
for the general
case wherein the e.m.f. and current
may have any

e,
form must
be written
Wz= EI cos
an equ-zîon that must be interpreted only as follows:

cos
ft
^ = ^,
where E,
I,
and
W have
been determined by alternating cur*
rent, pressure,
and power instruments giving the values
E=VE^^ + Ei + E^^ +
W= IE.
With
this conventional meaning of cos 6 a corresponding
conventional meaning must be attached to effective values of

current and e.m.f.,
/,
and E.
This convention consists
in
considering / and
to be equivalent sine
waves
of current
and
sine
e.m.f. having effective values equal to the effective values
of the actual irregular waves.
The
clearer
significance of equivalent
wave may now be made
by
reference to Fig. 70^
I02
illustrates
bs
which
irregular
waves of e.m.f. and current and
their equivalent sine waves.
In Fig. 70 the irregular wave of e.m.f., E, established the

irregular
It
is
wave
of current,
/,
differing in
form from that of E.
required that the equivalent sine waves of
E
E
and / be
EQUIVALENT SINE WAVE OP
^ EQUIVALENT SINE WAVE OF I
EFFECTIVE
E.
M.
F.
EFFECTIVE CURRENT AVERAGE VÂTTS

2962
= = =
8S. 2
VOLTS.
55.6 AMPERES,
2962.
Fig. 70.
Illustrating the Meaning and Application of Equivalent Sine Waves.

in
determined
amount and phase
position and that such
waves
be drawn
to scale in this
diagram so that they
may be com-
pared with their corresponding irregular waves.
The
instantaneous product of these curves produces the

as drawn.
power curve having a form and average value
The
25]
effective values of
ELECTRIC POfVER.
1O3
in the
and / are also drawn waves
diagram.
p. 57.
They are determined by the method

For equivalent
sine
given in Section 13,
Power Factor
cos B =.
are
EI
-=-^.
The
equivalent
sine
waves
now determined
is
after
assuming that the phase position of the current I

referred to that of E.
to be
The
values from point to point of the
equivalent sine waves then are

e
z
= V2E sin 0. = V2I sin {<p +
0).
The corresponding
and then drawn
to
curves have been platted point by point
in in full in Fig. 70.
If the irregular current wave had been chosen as the one which the phase of the wave of e.m.f. is to be referred the equivalent sine wave of current would pass through zero in a
positive direction at the
same
instant that the irregular
wave
of current passes through zero in the
same
direction.
From
the diagram
it
is
seen that this would result in shifting the
phases of the equivalent sine waves of current and e.m.f. back-
ward 16 degrees with respect to their corresponding irregular waves and with respect to the position at present determined
for
them in the diagram. Thus the precise meaning and significance of the equivalent In most routine practice E, I, and sine wave are made clear.
W as
read from the instruments and cos 6 as deduced are
interpreted for convenience as equivalent sine waves and the
equivalent angle of phase difference wherever the actual waves
E and / are
sine
irregular.
In all accurate testing, research,
wave
leads to error
and study and confusion.
the equivalent
104
[25
The
following problem will illustrate the

in irregular
solution
of a
power problem
wave forms by means
of harmonic
component
waves
EI products,
eliminating the use of equivalent sine
Prob. 48.
In an electric circuit transmitting power the
generator produces an impressed e.m.f. of
E'
(190O1
+ 60O3 + 300,) +y(20O3+
360^
+ 7O3 + 350^
this
and the motor which disposes of the power transmitted by

circuit
produces a counter-e.m.f. of
-f (i6oOj
100,)
E"
-^J{is
60, 4-
10O3+
SO5).
The impedance
of the total circuit
Resistance
3
2, '1
ohms
Induction reactance
6,
Required the
electric
power developed by the generator.
E'-E"
^ /=
(23. 1^
(30O1
+ 80O3 + 20O5) +7(-300i3+y(2i,

63,
10,).
30.,
+ 300.0
+ 49.33+33.05) -7(115.41+ 108.73+ 10. g.

73^3'
.
Power equals
W= + + w= [(1900, + 6003 + 300,) +y(- 36O1 + 703 + 350,)] X

I,E,'
. .
[(23-ii
+ 49-33+ 335) -7("S-4i+ 108.73+

Ans.
lo.i^)!
IV=
113,680 watts.
CHAPTER
VII.
MAGNETOMOTIVE FORCE AND THE LAWS OF THE

MAGNETIC CIRCUIT,
a6. Magnetomotive force and the magnetization curve.

u.
i.
c.
The unit of m.m.f. The m.m.f. of the ampere-turn. The conTluctor-turn, helix, and solenoid.
Magnetic reluctance.
d.
c.
/".
Magnetic permeability,
The magnetic
Saturation.
circuit.
g.
A.
Magnetization curves.
37, Matters affecting permeability.

a.
i.
c.
Permeability.
Effect of temperature
Effect of physical treatment
on permeability. on permeability.
d. Effect of impurities
on permeability.
sS, Reluctance of the magnetic circuit.

u.
i.
c.
Reluctance.
circuits,
Comparison of magnetic and electric Amount of magnetic leakage. 29. Magnetic hysteresis.
a.
6.
c.
Hysteresis.
Energy expended
in
overcoming hysteresis.
Derivation of mathematical expression for above.

Effect of physical treatment
Effect of impurities
d.
e. /".
upon
hysteresis,
upon
hysteresis.
Comparison of hysteresis curves.
30. Evfing's theory of magnetism.

31. Illustrative problems.
->-
26.
a.
Magnetomotive Force and the Magnetization Curve.
The Unit of Magnetomotive Force, M.m.f., the Gilbert. Symbol H. Magnetomotive force is the cause which results
105
io6
ELECTRIC/IL MACHINERY.
[26
in the production of
magnetic
flux.
By
definition one unit
of
magnetomotive force will establish in air a unit density of magnetic

is
flux through a distance of unity.
Tlae
name
of this unit
the gilbert.*
The
practical
and
c.g.s. units in this case are
identical.
The
definition of the gilbert is illustrated in Fig.

air
71
Over the cross-section of one cube of

THE MEOrUM
loi^s!/
ISAtR.
and throughout
-^^
Fig. 71.
H AT UNITY IS APPLIED UNIFORMLY THROUGHOUT THE CUBE. IT ESTABLISHES THROUGH THE CUBE ONE UNIT OF MAGNETIC FLUX
AT A DENSITY OF UNITY.
Graphical Illustration of Definition of Gilbert.

m.m.f
of one gilbert will estab-
its
length the application of a
lish through the cube one maxwell of flux uniformly distributed
over the cross-section of the cube' at a density of unity.

definition of
b.
See
maxwell, Section
5,
page
13.
The Magnetomotive Force of the
Ampere-turn.
In
Fig. 72 the straight conductor
CC
carries a current which,
when measured
in c.g.s. units, equals /.

it
researches
straight
has been found that
From experimental when such a

a
field
conductor carries current,

is
n
.:
of
-.^^
magnetic flux
conductor
flux
is
set
up about
it.
When
the
j^-j
is
remote from magnetic bodies, the
everywhere at right angles to the conits
ductor,
density varies inversely as the normal
distance from the conductor, and closed circuits
JC
FiG. 72.
of equal flux density are circular about the con-
ductor and,
therefore,
everywhere equidistant
is
from
it.
Since the flux density at any point
inversely pro-
portional to the distance of the point from the conductor, the

*
Names
of 'units.
Action of the American Institute of Electrical Engineers
Transactions, Vol. XI, 1894, p. 48.
26]
MAGNETOMOTIVE FORCE.
this
107
m.m.f. per cm. which sets up the flux at

the length of any circular flux circuit
its
point
is
also
inversely proportional to the distance from the conductor.

is
As
directly proportional to
distance from the conductor, and as the m.m.f. per cm. is

it
inversely proportional to this distance,
follows that the total

is
m.m.f. established in any complete flux circuit

a given current flowing in the conductor.
constant for
The m.m.f. around a straight conductor of

.
infinite
length
which carries a current of one ampere
47r
ts
gilberts.
The
value of
10
is
derived in the following manner:
In any space not actually occupied
bj'
magnetic bodies the

its
mechanical tension of magnetic flux along
own
direction
is
dynes per sq. cm.
(See Sec. 4.)
No
tension existed in the
molecular magnetic mechanism before the magnetic flux came The conservation of energy requires, then, that into existence.
work be done when magnetic flux is established. It requires that the amount of this work shall be equal to the product of the
tension into the distance through which the tension
is
produced.
In the field the mechanical expression for the work done in establishing a
about a conductor carrying current, as in Fig. 72,
small filament of magnetic flux through any concentric circuit
about the conductor
will, therefore,
be
(;4)
w= 2n.r--^-A
where
8ar
io8
is
it
[26
the intensity of flux tension, 27tr'the distance througli which
has been established, and
is
the area of the cross-section
of the filament.
establishing
this
An
electrical expression for the
work done
in
same
flux
filament
may
also
be written.
During the process of establishing the current

the flux about the conductor
in the conductor,
was being formed by the m.m.f.
of the current.
in
This caused a counter-e.m.f. to be generated
the conductor which was equal in
amount
to the rate of
increase of the magnetic flux in the filament.
The
rate at
which work
is
done
in
estabHshing the magnetic
field is, at
any
instant, the product of the current
and the counter-e.m.f.

the conductor
It follows, then, that
when
is
the formation of a magnetic field

in
is
about a conductor by the current

progress, electrical power
in
absorbed from the
circuit equal at
each instant to the product of the current into the countere.m.f.
The energy
of this
power disappears as potential

magnetic
flux.
is
energy
in the formation of the field of
V/hen
brought
doing
the current whose
m.m.f maintains
is
the field of flux

to zero,
to zero, the field of flux likewise

its
comes
and
in so
potential energy
given back to the electric circuit by the

circuit.
formation of an e.m.f. in the

as the current, and
it
This has the same sign
exerts itself to maintain the current for a
time
when the supply pressure is cut off". The energy thus taken from the circuit
field is
or restored to
it
again by the magnetic

bringing the magnetic
a measure of the work done in
field into existence.
At any
instant
dw
Eldt
(75)
Since flux density varies inversely with distance from the conductor,
2Tir
2 6]
MAGNETOMOTiyS FORCE.
109
or
where k
ductor.
is
the m.m.f. about a unit current in a straight concounter-e.m.f.

,
The
due to the changing

is
flux in the
filament under consideration,
^=f
where ^
circuit.
(7^)
BA,
or the maxwells in the filamentary magnetic
Substituting these quantities in the equation
dw
(ve
lEdt,
{j-j)
have
dw
2nr
2'Kr
/2
Q.nr
r-.B-A-dB,
(78)
BdB=-^-A-\B\
7Tf
(79)
Equating (74) and
(79),
B^
and solving
for k,
4;r.
Therefore 47r equals the number of gilberts of m.m.f. exerted!

in
any concentric
circuit
about a conductor carrying one c.g.s.

is
unit of the current.
Since the value of one ampere

it
one tenth
of the value of the c.g.s. unit

in
follows that ike m.m.f., exerted
any complete
concentric circuit about a

is
conductor carrying
one ampere of current
or i-S^'j gilberts.
This statement
is
true
whether the path be
in a
plane
normal to the conductor or not.
Proof: If the path does not
the conductor,
it
[26
lie in
a path perpendicular to
may be
divided into elemental steps, alterit
nately in this plane, and perpendicular to
or parallel to the
is
is
conductor.
Along the
latter steps the
m.m.f.
zero.
in
An
equivalent path
may
therefore be chosen which
a plane
perpendicular to the conductor.
Thig path in the limit would

in
be made up of indefinitely small segments lying alternately

arcs of circles
and
in their
corresponding
radii.
The m.m.f.
Let pqstp,
i.
applied along such radii would evidently be zero.

Fig- 73
1
be such a path, linking with the conductor
Let
aboda be a circular path in the same plane, and concentric

with the conductor.
The path pqstp may be

and
is
divided into steps
composed of
circular arcs
radial distances.
Along
the
latter steps the
m.m.f.
is
zero.
Along each of the former the

an arc of
radii.
m.m.f. (see p. 107)
equal to the m.m.f. along

is
the circle aboda which
included between the
same
Fig. 73.
In going completely around the path pqstp the
sum
of the
m.m.f.s found
is
therefore equal to the
sum
of
all
the elemental
m.m.f.s around the circumference of the

lonce, or
if
circle,
each taken
more than once, the excess being half of one sign
26]
M^GNETOMOTiyE FORCE.
The m.m.f. around
Ill
and half of the opposite sign.

path
is
the irregulaf
therefore equal to the m.m.f. around the circle,

47r
and
is
lO /,
where
/'is
the current in amperes.

to
TAe m.m.f. due

Proof:
any current, taken
com.pletely
around a
closed path which does not link with the current,
is zero.
In Fig. 74, the m.m.f. taken completely around
the path ^^5;' includes components making a m.m.f. equal to

that from a to b.
But each component
is
included an even
number
total
c.
of times, half positive and half negative, thus giving a of zero.
sum
The Conductor-turn, Helix, and Solenoid.
If the
con-
ductor be bent so as to form a single circular turn, as in Fig.

75, the m.m.f. due to the current in the turn of conductor will
be present everywhere around the conductor.

circumstances the
sented in Fig.
field
Under these
is
of flux that will be established

lines
repre-
"]&.
The
about the conductor show the
direction of the flux at all points about the turn, while their
proximity
is
proportional
to the
flux
density,
or B.
The
Fig. 76.
Fig. 75.
is
m.m.f. applied by the current in the conductor-turn
every-
where
alike for
any closed
circuit
passing through the turn
and, therefore, around the conductor.
The
variable nature of
112
[26
the flux as to
amount and
direction
is
due to the varying cross-
section of the space about the conductor which at the
minimum
turn,
portion
at
is
limited to the cross-section enclosed
by the
and
of
the
maximum
portion
is
the unlimited cross-section

facts
indefinite space.
These two
account entirely for the

'/6.
varying
field of flux as
given in Fig.
If the
conductor be coiled into a helix, as

field established
is
in Fig. yy,
it is
found that the character of

ductor by the current in
it
around the con-
given by the lines in Fig. 78.

the same for any closed route
In this case, as in that of a single turn of the electric circuit,

it
is
found that the m.m.f.
is
through the helix and, therefore, around the conductor, and

that the
variable
amount and direction of the flux is determined by the length and character of the magnetic flux circuit.
difference,
There
is this
however, to be observed.
In the case
Fig. 78.
Fig. 77.
of a single turn any closed circuit through the turn passes
around the conductor once^ while for the helix the closed flux
circuit
through the helix passes around the conductor as many times as there are turns in the helix. The m.m.f. exerted by
it is
a helix along any closed circuit through
greater than the
corresponding m.m.f of a single conductor by as

as there are turns in the helix.
many
times
When
of turns
practice
a conductor spiral
is
made up
The
of more than one layer
it is
technically called a solenoid, though generally in

called simply a
coil.
it is
solenoid or coil has the
same m.m.f.
properties as those of the helix.
26]
MAGNETOMOTIl/E FORCE.
113
In electrical engineering, the chief and practically the only

source of m.m.f.
is
the current-conducting solenoid or
coil.
Its
m.m.f. equals the produt of its turns and current, times 1.2^7.
a solenoid having n turns carrying I amperes, the The product nl in practice is called m.m.f. 1.2^-jnI.
in
Thus
ampere-turns.
d.
Magnetic Reluctance.
Its
Unit.
Magnetic reluctance,
It is
or
netizable region
lished
more generally reluctance, is the property of every magwhereby the amount of magnetic flux estabby a given amount
of m.m.f.
is
limited.
analogous
to resistance in the electric circuit.
One unit of reluctance
in
a magnetic circuit will require one
gilbert to establish one maxwell.
The
reluctance across a cube, one centimetre on a side,
is
H
p
is
therefore the
e.
symbol
for specific reluctance.
Magnetic Permeability.
is
Magnetic permeability, or more

to
specific
generally permeability,
the reciprocal of specific reluctance conductivity in
and
is,
therefore, analogous
Its
the
electric circuit.
value
is
B
f.
nickel,
The Magnetic Circuit. With, the exception of iron, and cobalt, no substance or material used in engineerinto the
in
ing
when brought
the other hand,
magnetic
circuit will
modify the
appreciably.
is
character of the flux
amount and
direction
On
when
iron, nickel, or cobalt
placed in
the magnetic circuit a powerful modification in the resulting

flux occurs,
soft iron
both
in
amount and
direction.
Thus
if
a bar of
be placed in a solenoid carrying a strong current, the
bar
at
once becomes a powerful magnet.
As soon
as the
current in the solenoid ceases the bar of soft iron will cease to
114
ELECTRIC/I L MACHINERY.
[26
be a magnet.
In Fig. 79,
soft
if
the bar
NS as
well as the keeper
AAA
AAA,
he made of
wrought iron and a solenoid surrounds
the effects produced
by introducing such materials
into
the magnetic circuit of a solenoid
may
be accurately observed.
The arrangement
circuit
is
of the solenoid through which passes the.
occupied by the soft wrought-iron bars

in Fig.
AAA
and JVS
shown
79.
By weighing
the tension of the flux

J
THIS FORCE TO BE WEIGHED UPON A BALANCE
a.a.
^r<i.g.
Fig. 79.
established across ag; data will

erties of the soft
be obtained showing the propmagnetic

circuit.
wrought iron
is
in the
Assume
a balance.
that sufficient current, /,
passed through the solenoid to pro-
duce a flux tension at ^ which
may be weighed on
in closed circuits
The
flux
produced
is
set
up
which pass
through the solenoid.
It
has already been shown that the

circuit,
m.m.f. due to the solenoid, taken around any particular

is
the same as that around any other.

this
The
flux set
up around
to the
any path by
constant m.m.f.
is
therefore inversely propor-
tional to the length
of the path, and directly proportional
area of its cross-Section.
These
quantities remaining the same,
the flux depends upon the character of the material through
which
it
passes,
i.e.,
upon
its
permeability.
Example: In
Fig.
79 the permeability of the iron parts

it
AAA
and
NS
is
so high that
is
quite accurate to
assume
that all the flux follows the iron path where possible,
straight from iron to iron at their gaps.
and goes
The flux
crossing the
26]
M/iGNETOMOTiyE FORCE.
is
115
air-gaps
then equal to the induction through the iron, and

is
the area over which this flux
distributed
is
sensibly equal to
the area of the end of the bar
AAA,
in
the distance across the
air-gap being
made very
in sq.
small.
Let
be the observed weight
dynes,
/
the area of
cross-section
cms. at ag, and
the distance in cms.
through both air-gaps, ag; then
P^_
B>
~A~
and
'3^
/STtPxi
^=[-A-)
The
portion of the
(8)
is
m.m.f
of the solenoid that
used in putting
the flux across the air-gaps equals
H^ Bl
If the
gilberts.
number
of turns in the solenoid be n, then the m.m.f.

in
this
of
the
solenoid
experiment
if
is
nl ampere-turns,
or 1.257%/ gilberts.
Now
the value of B, as determined
by the balance, were found

maxwells per
i.257/.
densities
sq. cm., the
to be
between zero and 12,000

at the air-gaps will
m.m.f used
be
found to be a large part of the total m.m.f of the solenoid,
This means that the iron
piart
of the circuit, at flux of the
below 12,000, consumes but

at the air-gaps.
/,
little
m.m.f
of
is
the solenoid, while the remainder, or most of this m.m.f.,
consumed
Should the current,
be increased to
I^ in
the solenoid,
If in this
B at
way
ag
/^
will
be increased to a larger value,
B^.
has been
made
large
enough
to cause B^ to attain a
value above
14,000 maxwells per sq. cm., the m.m.f coniron part of the magnetic circuit,
sumed by the
m.m.fj= \.2t,TnI^
will
IB^,
....
14,000 the
(81)
be much greater
less, or
in proportion
than when the flux density
was
B.
In
fact, for
values of
B above
soft
'
ii6
will
[26
wrought iron
suddenly increase
its
reluctance from a very
low value
materials.
to a value nearly as high as that of air
and ordinary
All those materials which exhibit this low reluctance for
magnetic flux at moderate

materials.
g. Magnetization Curves.
densities
are
called
magnetic
To
make
soft
a close study of the
properties of a magnetic material, the magnetic balance
may
the
be used.
The experiments with

in the
wrought iron
in
magnetic balance
carefully
preceding sub-section
f should
be
made.
Small air-gaps at ag and ag are used to
eliminate the error due to the fringes of flux over the edges.
Frequent weighings of the flux tension and current readings

are
made
at intervals corresponding to
approximately uniform
values of
increments of current.
The corresponding
and
for the iron part of the circuit are
deduced from equations (80)
and
(81).
In (81) the values of /f are divided
by the length
circuit,
in centimetres of the iron part of the
magnetic
thus
giving the m.m.f.s
consumed
in the iron per
centimetre length.
are platted with
'
These values
for soft electrical sheet steel
rectangular coordinates in Fig. 80, forming the
'
sheet steel
'
curve there drawn.*

soft
similar curve for the best qualities of
wrought iron
electrical
is
almost identical with the one here drawn
for
steel.
Such a graphic representation of the

is
properties of a magnetic material
called a
B-ff
curve.
It
shows clearly that the
steel
has but small reluctance to the
establishment of magnetic flux at densities below

*
B=
80,000
The B-ff
curves in Fig. 80 are platted in the (maxwells per square inch',

to use ampere-turns per inch
(ampere-turns per inch) system.

it is
now customary
For engineering purposes in the United States, and maxwells per square inch
for
B-H
values
in lieu of gilberts per centimetre
and maxwells per square

c.g.s.
centimetre.
i.e.,
The
B-H values
in Fig.
80
may
be reduced to the
system,
system of values, by dividing the valuesby 6.45, the number of square centimetres in a square inch, and the .^values by
to the gilbert-maxwell-centimetre
(2.54-!- 1.257), the centimetres in
an inch and the
gilberts in
an ampere-turn.
26]
M/IGNETOMOTiyE FORCE.
MAXWELLS PER SQUARE
INCH.
117
0=
'S
Ii8
in.
[26
per sq.
per inch
Above 80,000 per sq. in. the m.m.f. consumed length increases much more rapidly in proportion to
in.
the increase in the flux density, showing that the reluctance in

the neighborhood of 80,000 maxwells- per sq.
no longer
remains
uniformly
small.
For values
up
above
80,000 the
105,000,
reluctance rapidly increases

after
to the value
B ==
which there
is
but slight farther increase in the reluctance.
The
B-H curve
above
this point
approaches more and more

line
nearly to a straight
line.
This
ultimately becomes
approximately parallel to a straight line which
may
be drawn
through the origin, representing the relation between the
magnetic
flux, B^^^, in
maxwells per
in
sq. in., established in air,
and the magnetizing force

which
is
ampere turns per inch length,
necessary to set up the flux, B, in the steel or iron.

is
This relation
//"=
B^^^,
from definition, c.g.s., and the equais
tion of this part of the curve
approximately
B^l.^gH^C,
where
is
a constant representing the distance this part of the

lies
B-H
curve
above the
H-B^j,^ curve, or
it
represents the
it
increased induction due to the presence of the iron before
becomes saturated.
h.
Saturation.
magnetic material
is
said
to be
saturated when
it
already
carries such a dense induction that a further increase in the
magnetizing force, H, produces no m.ore increase in the induction

than would be produced in air by a similar increase in
netising force.
mag-
To
reach this limit absolutely requires in iron an extremely
high magnetizing force, between 2000 and 5000 gilberts per

centimetre length, or from 4000 to 10,000 ampere-turns per
inch of length.
Iron
may be
very nearly saturated, however,

so-called
'
'
by much smaller
forces.
The
knee
' '
of the
B-H
27]
is
119
curve
of
' '
found in Fig. 80 at about

'
H ^ 40.
for
This
is
a region
approaching saturation,
'
and
for practical
purposes the
large
iron
may
be considered saturated above
this point, as
increases in
H are thereafter necessary

It is
;
comparatively small
increases in B.
usual to speak of the iron
when
in this
condition as saturated
meaning thereby not that the actual

in
point of saturation as defined has been reached or passed, but

that
an abrupt change
its
magnetic
behavior
is
susceptibility
has
if it
occurred, and that thereafter
its
very
much
as
were
in fact saturated.
27. Matters Affecting Permeability.

jj.
a.
Permeability.
The symbol is used for permeability. Then B = piH, or density of magnetic

Evidently
curve
the
is if
_/u
induction
is
the
product of permeability and magnetizing force.

is
constant for any material, the
B-H
if
a straight line for that material.
Conversely, also,
line,
B-H
curve for any material

is
is
not a straight
air
for
that material
not constant.
For
>" is
and
for all materials
except
iron, nickel,
and
cobalt,
relations
for the
between
/^
and
for the
"sheet
very nearly unity. The steel " curve, and

in
" cast iron " curve of Fig. 80, are given

3000
Fig. 81.
S2OOO
o
Siooo
>
rapidly to a value from which
nitely.
it
[27
gradually diminishes indefiis
In Fig. 8i the /<-/f relation

,2000
given for wrought iron
5,000
10,000
15,000
20,000
H
Fig.
8ifl.
IN
GILBERTS PER CENTIMETER,
Permeability Curve for Wrought Iron

by Bidwell
Strongly Magnetized.
as obtained
at very high flux densities established
by corresponding high values of

length in the iron.
b.
applied per
centimetre
Effect of Temperature on Permeability.
The
permefound at
ability of iron at ordinary temperatures varies but little. with
change
in temperature.
Very remarkable
effects are
high temperatures.
The
curve in Fig. 82, given by Hopkin-
10,000
9,000
'
^7]
MAGNETOMOTiyE FORCE.
force of
ff =
=^ a, and curve C with curve .5 with //"= gilberts per centimetre length of the 4$, where wrought iron. In each case the permeability falls to about
=.:
.3,
unity at a temperature of 785.
With
the two larger magnetizis
ing forces the change in permeability
rather abrupt, but with

is
the smaller value of 11 the suddenness of this change
much
more marked, and immediately before

rises
falling the permeability

1 1
very rapidly to the extremely high value of

place
is
,000.
The
rod
it
point at which this change takes
one where other

If a
changes
of iron
is
is
in the physical properties of the iron occur.
heated to a bright red and allowed to cool slowly,

'
noticed that at this
'
critical
temperature
;
' '
there
is
a sudden
check, for a brief period, to the cooling

this
is
in
some forms of iron

its
so
marked
as to produce a noticeable brightening in
redness.
This phenomenon has been given the name
of
" recaiescence."
At
the same time the rod, which has pre-
viously been contracting, lengthens for an instant before continuing to shorten.

c.
Effect of Physical
Treatment on
is
Permeability.
The
permeability of iron or steel

physical treatment which
ceived.
it
very largely dependent on the

1,500
has re-
In general the softer grades
are
more permeable, and any process

toward
softening,
tending
such
as
a.
1,000
annealing, increases the permeability.
Any
process of hardening
decreases
the permeability.
The
curves
and
B, Fig. 83, show the effect of annealing on permeability.
Curve
is
taken
Fig.
from an ordinary sample of

,
electrical
4.1
^. z? r, steel castmg, and curve B, from the

, 1
83. The Effect

^^^^^
^^^^
of
Anneal
.^^^
Permeability of
same sample
d.
after
thorough annealing.
Electrical Steel Casting.
Effect of Impurities on Permeability.
Almost
all
im-
122
purities in iron
[28
its
have a detrimental
effect
upon
permeability.
Some
are of a
much more marked

is
character
than others.
Combined carbon
present in
injurious.
particularly
silicon,
bad.
Manganese, nickel,
chromium, tungsten,
phosphorus, and sulphur,

small
when
also
more than
ordinarily
amounts,
are
So many things operate

it is
to affect the permeability
of a magnetic substance that
difficult to
assign the proper
weight to each.
It
has been found that
many
substances
may
be present
in iijon or steel in small

its
amounts as impurities with-
out appreciably affecting

ties will often offset
permeability.
Again two impuri-
the detrimental characteristics imparted
by
either.
also
The mechanical character of irons and steels is improved or harmed by the presence of impurities. The
is
presence of certain impurities

the iron or
.steel
often necessary to impart to
the best mechanical-physical characteristics.
The
as to
engineer, therefore,
who
undertakes to produce electrical
steels
must make a judicious use of the impurities so produce a good foundry and machine-shop material and and
irons
it
yet have
suffer least in permeability.

'
Modern manufacturers
'
are able to produce grades of
'
electrical
'
iron
and
It is
steel
which are
possible,
fairly
uniform and highly permeable.
not
however, to predict the permeability of a
steel, or
iron, exactly, in
advance of a
test,
and
different
samples from
the same lot of iron, or even from different parts of the

sheet,
same
may show
appreciable differences in magnetic quahty.*
28. Reluctance of the Magnetic Circuit.
a.
Reluctance.
The
symbol
counter action of any region to the setting up of
magits
netic flux
through
is
it
has been termed
its
reluctance.
Specific
reluctance
is
essentially the reciprocal of permeability
and
p
*
.
is to
An
excellent
paper on
this subject
by Parshall and Hobart
be found in
Engineering, London, January 21, 1898.
28]
MAGNETOMOTiyE FORCE.
specific
123
is
The
reluctance
for
any material
the
magnetic
reluctance of a cube of that material one cm. on a side.
We
is
may
write as an expression for the magnetic circuit which
similar to
Ohm's law
for the electric circuit
Flux
M.m.f.
Reluctance
The
tional to
actual reluctance along

its
any path
is
inversely proporits
area and directly proportional to

/
length.
Thua
is
the reluctance of a bar of length Reluctance
and cross-section
(82)
and the expression
for the
magnetic
circuit
becomes
Flux
whicfi
is
=z
=^
m.m.f.J
ixA
....
"X.
it is
(83)
a practical working expression.
In the solution of an actual problem,
necessary to
know, or
to assume, an approximate value for the flux density,

/
as the value of
b.
depends upon
this.
netic circuit to
differs
Comparison of Magnetic and Electric Circuits. A magwhich the relations just given may be applied,
electric circuit, for
from the
which a similar expression
has been used, in one important particular.

cuit
is is
confined to a metal path or conductor.
The electric cirThe conductor

is
surrounded by insulating materials which prevent the escape
of current.
for
On
flux.
the other hand, no substance
an
insulator-
magnetic
The
poorest conductor of magnetism has;
a permeability
little less
than unity.
It is, therefore,
impossi-
ble to confine the flux to any definite path.
We
may, indeed,
provide a path of such low reluctance that the greater part of

the flux will be found in
intensity,
this path.
it,
but a
field
of greater or less
and extending
to
an
indefinite distance, will
surround
124
In Fig. 84
is
[28
familiar type of two-pole
shown the magnetic circuit of an early and dynamo. The useful magnetic flux
I.
^'/ \\\ \ \Ni/ 1 / V' tîA^y>-Jy^^>y-^P-^yA-A'y'^>.>

1 r /
:^^
I
Fig. 84.
lies
^^
in
the
path,
A AAA,
but surrounding this path
is
leakage" or "stray"
field,
which
in this type of
machine
often amounts to
c.
40 per cent of the whole.
Amount of Magnetic Leakage. It is often necessary to estimate the amount of leakage which will occur in air between iron surfaces, when a m.m.f. acts across the intervening space.
It is
only possible to do this approximately at best, and
it is,
therefore, not desirable to enter into an elaborate discussion of

this matter.
few simple cases will serve as

is
illustrations.
The
reluctance across any path in air
its
average length in
its
centimetres divided
^^5?
by the average area of
cross-section,
IJ
Fig. 85.
<
A,
a
Fig. 86.
h
Fig. 87.
measured
in
square centimetres
In Fig. 85 this
is
simply
A'
in Fig.
86
A,+A^.
it is
In more complex cases, such as in
29]
I2S
Fig. 87, ab, and cd, the approximate path of the flux may be sketched to scale and the average length and area of the path
estimated, f29. Magnetic Hysteresis. . Hysteresis. li a ring of iron is magnetized by a steadily increasing force, and this force
is
then steadily diminished,

the two
the
cases do
it
is
found that the
B-H curves
curve taken
in
not coincide.
force
is
The
B-H
while
magnetizing
lies
decreasing
entirely above the
similar curve traced for an increas-
ing magnetizing force.

88, curves a
(See Fig.
and
3.)
is
When
the
magnetizing force
induction
removed the
fall
does not
to zero.
in
In other words, the
changes
induction
do
not correspond
to
Fig. 88.
the changes in magnetizing force,
but lag behind them.
This la^-
ging of the magnetism has been given the name hysteresis.

If after
.bringing
the force
H to
zero
it
is
again steadily
with
increased, the
either
resulting
B-H
curve does not coincide
of the others, and again the lagging change of

is
magalter-
netism
are
noted to change as the curve

the relations between
c.
Fig. 88.
In Fig. 89
shown
and H, when
is
nately increased and diminished.
It is to
be noted that these

is
curves form closed loops.
When
the ring of iron
magnetized
by an alternating current the magnetizing force is not only The increased and decreased but is reversed in direction.
cyclic relation between
B and
ior a
bar magnetized in this
manner
oi H.
b.
is
shown
in Fig. 90.
Three separate curves are there
shown, giving
this relation for three different alternating values
Each of these is a closed loop. Energy Expended in Overcoming
Hysteresis.
It
has
126
in a
[29,
been shown that energy exists

set
magnetic
field.
field
up while a current
is
increasing contains stored energy,
which helps to maintain the current
when
it
decreases.
If
the
+B
Fig. 89.
field is entirely in air it
completely disappears with the current.

is
In this case precisely the same amount of electrical energy
given back to the circuit by the destruction of the

previously required from the circuit to establish
is
it.
field as
was
does
If the field
it
wholly or partly
in iron or other
magnetic material,
not completely disappear

ceases.
field is
(See Fig. 88.)
when the current that established it The energy required to establish the
circuit.
not entirely restored to the
To demagnetize
in
the iron completely requires the expenditure of energy
the
opposite sense, or a magnetizing force, oA', Fig. 90, must be

applied.
sitely
If the
then demagnetized iron be magnetized oppois
and demagnetized, the curve, A' DA,

is
traced and
additional electric energy

that to
expended.
It is obvious, then,
carry the iron through a complete miagnetic cycle,

90,
tracing one of the loops in Fig.
definite quantity of

29]
electrical
127
energy must be expended in overcoming magnetic
hysteresis.
c.
Derivation of Expression for
Energy Dissipated
in
/,
Hysteresis.
Suppose
by a
a bar of cross-section A, and length

coil
to be magnetized
of n turns uniformly
coil
wound upon
it.
Let the length of the bar and

their diameters, so that
be great
in
comparison with
H may be uniform
throughout.
Let
dB.
the current increase slightly, causing a corresponding increase

in B.
The
total flux in the bar is increased

in the coil
an amount
This induces
dB
a counter-e.m.f., nAr-.
dt
is
If
ihe the
value of the current during this time, the work done
dB dW= nAirrdt = nAidB. dt

Since the volume of the bar
.
is IA,
the work done per cu. cm.

VH, per
nidB
-.
IS
.The magnetizmg
r force,
4'f
cm.,
is
-j,
when
is in
c.g.s. units; substituting the
work becomes
dW= HdB.
W = ^fHdB = ergs
as the energy expended per
tion.
cu.
(84)
cm. per cycle of magnetiza-
The power expended
in watts per cu.
cm.
is
~
where
fndB,
(85)
is
the time in seconds occupied
by one complete
cycle of magnetization.
The
value of /
HdB
taken
around one complete loop

is,
is
the area of the loop.
This area
therefore, proportional to
128
[29
the energy dissipated in carrying the iron through a complete

cycle of magnetization.
d.
Effect of Physical Treatment upon Hysteresis.
In Fig.
91 are shown two hysteresis loops from the same sample of
soft iron wire.
Curve a
is for
the sample after being hardened
by
stretching,
and curve b
after
is
thorough annealing
greater.
Note that
the area of the former loop
This shows that the
work done
cycle
is
in carrying the iron
through a complete magnetic

each case.
is
greater for the hardened wire, as the magnetizing force
.^varies between the
same
limits in
Note
also that
the flux density at similar values of

iron.
greater in the annealed
These
facts actually
go together.
Hysteresis losses are affected
by temperature.
little'
For
all
ordinary temperatures up to 200 C. but

place.
change takes
less,
At higher temperatures
fifth
these losses
grow
reach-
ing from one
to
one tenth of
their original values in the
29]
Above
800''
129
irons lose their
neighborhood of 700" C.
magnetic properties.
C.
These temperature-hysteresis properties
have served no
useful purpose in engineering.
In most irons and electrical steels the hysteresis quality

deteriorates with age,
i.e.,
is
the hysteresis losses increase slowly

called aging.
with time.
This process
Whenever
electrical
steel sheets are used in alternating current
machinery the
is
highest practicable quality in regard to hysteresis
sought.
for
As
these
losses
increase with age in different degrees
different treatments applied in the manufacture of the sheets
much
sheets
attention has been given to the conditions under which
may
be produced that display an average minimum

It
hysteresis loss.
has been found that the chief factor in the
treatment
in
of the steel sheets to secure
minimum aging
Above
consists
annealing the sheets froni the highest practicable tempera-
ture.
This temperature
is
about 900 C.
that tem-
perature the sheets tend to stick together and form excessive

scale.
Solid irons and steels might be annealed at higher
temperatures, a useless advantage in this respect on account

of the fact that magnetic
circuits for alternating flux
must
always be
built
up out of sheets
to avoid
eddy current
losses.
The aging
of actual use.
process further depends upon the temperature
of the sheets in the electrical machinery under the conditions

It
has been found through extensive tests that
high permeability electrical steel sheets annealed from the highest temperature do not age much when not subjected to a
higher operating temperature in the machinery or apparatus
than 60
C,
while serious aging will result

raised to 90 C.
table
tests
if
the operating
temperature
is
The
following
of results
given
by
Parshall
and
Hobart from actual

sheets
show the
effects of
annealing and
aging on the hysteresis quality of a particular sample of steel
13
[29
Time
in
Hours.
29]
hysteresis.
M/tGNETOMOTIVE FORCE.
'31
In recent years great attention has been paid to

.
the production of electrical steels showing small hysteresis

losses
and great permeability.

curve, Fig. 92,
The
is
left-hand curve, Fig. 92,
shows the characteristics of one of the best commercial grades.
The right-hand
B
sa.oM.
16,000
from a sample not quite so
132
rf
[S"
where
units
steel.
cu. in.,
is
the quality-constant determined
by the system
of
employed, and the hysteresis character of the iron or
Using maxwells per
sq.
77
in.
and ergs per cycle per
the following values of
are found to occur in
The
Poor
highest quality
001
Excellent commercial qualities

qualities
0015
002 and higher.
The
law:
following example illustrates the use of Steinmetz's
Required the watts
lost
through hysteresis in 1000
cti.
in.
of best quality commercial electrical steel sheets operated at a
maximum
alternating induction of 30,000 maxwells per square
inch and at 100 cycles per second.

Solution
Watts
30.
lost
-""^S
30.0'-10,
1000
100
^^^
^^^_
Swing's Theory
of
Magnetism.
Weber
suggested
that the individual
molecules of a magnetic substance are
themselves magnets and that these magnets are more or less

free to turn about,
under the influence of outside
forces.
This
theory evidently accounts for such a substance becoming a
magnet under the influence of such forces. Weber, Maxwell, and others have suggested theories, based on this one as
fundamental,
etc.
to
account
for
hysteresis,
the
B-H
relation,
These
theories involve various directing forces, frictional

etc.,
resistance,
effects of stress,
which complicate matters,
Ewing has simphfied

more than the
the whole matter
by assuming nothing
frfee
original idea of small magnets,
to turn.
He
has shown, both theoretically and by means of working
models, that the mutual action of the molecular magnets upon
each other
is
quite sufficient to account for
most of the observed
phenomena.
3]
M/IGNETOMOTIVE FORCE.
133
The theory
(See Fig. 93.)
is
as follows
The
small magnets,
when not
is
under the influence of outside
forces,
form stable groupings.

applied
When
a small magnetizing force
-- \
-^ -> /
H=0
\\
/ /
^^
a
^X
b
f \
Fig. 93.
Ewing's Magnet Model.

If the force is
to a group the individual molecules turn slightly in the direction of that force..
(See Fig. gib.)
removed,
the old groupings are again formed.
This corresponds to the

reached where the
lower part of the
B-H curve.
be applied, a point
is
If a stronger force
old groups begin to break up, the magnets swinging violently
around, forming
new groupings which have

direction
a general aligning
tendency
individual
in
the
of the applied force, though no

in that direction.
magnet may be exactly
This
condition of groups, just on the point of breaking up, corre-
sponds to the second stage of magnetization indicated by the

steepest part of the typical
B-H
curve,
where a small
in-
crease in
H causes
is
a large change in the induction.

this stage or later, the
If the
outside force
removed during
magnets
do not return
to their original groupings, but the bar remains
more or
If,
less strongly
magnetized.
(See curve
b,
Fig. 88, and
Fig. 94c.)
after the original
groups have been thoroughly broken

is still
up, the magnetizing force
further increased, the

for the
newly
formed groups are not broken up,
magnets composing
134
are already turned
L30
them
force.
more
or less into the direction of the
The magnets, however, swing more and more

is
nearly
into line as the force
indefinitely increased,
an
infinite force
being theoretically required to
make them
exactly parallel with
each other.
This corresponds to the upper part of the
B-H
H=0
\
t
t
t f
t
t
/
\
/
\
t
t
\
I
t
!
31]
MAGNETOMOriyE FORCE.
135
in iron
magnets represent now the average molecular condition

for the steepest portion of the
B-II curve.
On
bringing the
bar
in
still
nearer, all of the
little
magnets arranged themselves
the direction of the externally applied m.m.f. (see Fig.
94^),
corresponding to the condition of saturation
in
mag-
netized soft iron.
The
bar was then removed entirely from

little
the neighborhood of the model and the
magnets grouped
themselves as shown in Fig. 94c, which corresponds to the lower end of the curve
i>
in Fig. 88.
if
Thus
any angle
energy,
iron
it is
seen that
a magnetic body like iron

free to
is
made
up of molecules which are actual magnets

in their
swing through
own
positions,
and
in this
swinging dissipate
all
the essential
are accounted for
phenomena by Ewing's
of magnetic induction in
theory.
When
most of
the molecular magnets are arranged ap-
proximately
in a single direction, the entire
mass of iron becomes a powerful magnet,

establishing through itself a large flux due
to the influence of a comparatively small
actuating or impressed m.m.f. externally

applied.
Professor Ewing's working model con-
tained twenty-four pivoted magnets.

c
By
,1 means of a delicate instrument he measJ

1
.
.
Fig. 95. Hysteresis Card obtained
by Ewing from Magnet

^""^^l-
ured the B-If relation for the flux set up
through
this
model and a small actuating m.m.f. externally
applied in a cycle of gradual change, and obtained thereby the

hysteresis card given in Fig. 95.
31. Illustrative Problems.
Prob.
49.
Required the re-
luctance of the magnetic circuit shown in Fig. 79

total
when
the
induction
is
290,000 maxwells.
The dimensions and
materials are as follows:
Bar NS, 16" by 3" by
3", cast steel.
136
[31
Bar
AAA,
30" by 3" by 3", cast
iron.
Air-gaps, ^" each.

Solution:
The
reluctance of each part of the circuit
is
given
by the formula
and the permeability

80,
is
obtained from the
B-H
curves of Fig.
by
reducing the values of
B and B
'
H to
the c.g.s. system
and then substituting
in the expression
After reducing the
dimensions to centimeters and
square
centimeters the reluctances are as follows:
40.6
61,-
=
=
is
210
58
.006256; ^
'
^35
î
"Tq5^
the
-010948.
of these, or .017966.
The
total reluctance
sum
Ans.
Prob. 50.
How many
.''
ampere-turns must be applied to the
magnetic
tion
circuit of
Prob. 49 in order that the required induc-
may
be set up
Solution:
M.m.f.
.017966
290,000
are
5210
gilberts.
The corresponding ampere-turns

It
M.m.f.
.AfTt
5210
I-2S7
414s ampere-turns. ^
Ans.
Prob. 51.
How many
ampere-turns must be applied in
order to set up twice this induction.'
3i]
-.^"
10,000 maxwells per
sq. inch.
in sq.
137
Solution:
cm. correspond to
64,500 per
From
the curves 15 ampere-turns per
inch are required
the steel and 290 in the iron.
Total
ampere-turns
in steel are
16
15
= =
240.
Total ampere-turns in iron are
30
290
8700.
In the air-gaps the m.m.f.

m.m.f.
in gilberts is
.01094
580,000
6345,
which reduces
to
6345
5047 ampere-turns.
The
total
m.m.f. then becomes

-\-
240
8700
+ 5047 =
is
13,987 ampere-turns.
Ans.
The
cast iron consumes the largest part of this.
Prob. 52.
What
the total pull in pounds as measured

in Fig. 79,
\>y the apparatus
shown
when
is
5000 maxwells
per sq.
cm..''
Solution:
Pull in dynes
AB =-=
=
57,600,000.
,,
Pull
m
.
pounds ^
57,600,000
445. 000
129.4
lbs.
Ans.
Prob.
53.
In Fig. 84 the
"leakage coefficient"
in the
is
1.4.
This means that 1.4 times the induction desired
armature
must be produced, the remainder forming the leakage field. -How many ampere-turns must be wound on the vertical cores.
138
in order that
[31
A A,
wells
an armature induction of 7,000,000 max?
may
be produced
The
constants of the circuits are as follows
Horizontal yoke. A, 11" by 11" by 28".

Vertical cylindrical cores,
A A,
diameter 10.5", length 16".
Pole-pieces, height 12", width perpendicular to paper 12".
Armature
sq. in.
core, diameter 10", length 11".
Air-gap, double
length
8 in.
approximate area 152
Solution:
The approximate
length and area of the path of
the flux must be found for each part of the circuit.

follows
This
is
as
Yoke, length 28
in.,
area 121 sq.
in.
Two Two Two

The
1.4)
is
cores, length 32 in., area 86.6 sq. in.
pole-pieces, length 10 in., area 144 sq.
in.
air-gaps, length 1.18 in., area 152 sq. in.
Armature
(actual iron), length 11.8 in., area 80 sq. in.
total induction of
9,800,000 maxwells (7,000,000
assumed to pass through the yoke when 7,000,000

set
maxwells are
in the cores
up
in the air-gaps
and armature core.
That
may
will
be estimated at 9,000,000 maxwells and the

receive
pole-pieces
about 8,000,000 maxwells.
The
induction densities are as follows:
= = cores = .5 in poles = 5 in air-gaps B in armature core =

B in 5 in
yoke
81,000 maxwells per " " 104,000

55,500
sq. in.
" "
46,000
87,500
" " "
"
" "
"
"
The corresponding
length of each part are
values of ampere-turns as determined
by multiplying the m.m.f. per inch (from Fig. 80) by the
3i]
MAGNETOMOTIVE
FORCE.
139
In cast-steel yoke, 25
28
In cast-steel cores, 115

In cast-steel poles,
12
X 32 X lO
18
In sheet-steel armature core, 30 In air-gap,

^
.
= = = X ii-8 =
=
3,680
120
700 ampere-turns. "

" "
354
46,000
5-^
The
1.
2.54
^^^
=17^
21,854.
Total ampere-turns
Prob. 54.
is
Ans.
permeability of a certain specimen of iron
1000 at an induction density of 12,000 maxwells per square
centimetre.
How much
m.m.f., in gilberts,
is
required to
?
maintain this density in a rod of this iron one foot in length

365.75 gilberts.
Prob. 55.
Ans.
A cast-steel ring,
one foot
is
in
mean diameter and
of one square inch cross-section,

force of 1000 ampere-turns.
in Fig.
subjected to a magnetizing
By means
of the curves shown
80* find the total induction produced.
81,000 maxwells.
Ans.
Prob. 56. Calculate the reluctance of the cast-steel and air
magnetic
density
is
circuit represented in Fig. 96,
when
the induction
100,000 maxwells per square inch.
Determine also
Fig. 96.
Fig. 97.
the m.m.f. in ampere-turns necessary to rhaintain this density.
Reluctance .17089.
10,678 ampere-turns.
* These curves
will
Ans. "
be used in subsequent problems in this chapter.
140
131
Prob. 57. 8000 ampere-turns surround the magnetic circuit
shown
will
in Fig. 97.
The
ring
is
."
of cast iron.
How much
ilux
be established in the ring

Prob. 58.
220,800 maxwells.
Ans.
is
The
transformer core shown in Fig. 98
made
of electrical steel laminations.
How many
ampere-turns must
?/Rounii
Rod
*<2^
3'J
141
Prob. 60. Fig. 100 shows a magnetic circuit of cast steel

consisting of four parts
:
a U-shaped piece, two air-gaps, and

of 3000 gilberts applied
a keeper or armature.
With a m.m.f.
CHAPTER
VIII.
ROTATING MAGNETIC FIELDS.
32. Polyphase e.m.f.s, currents,

a.
b.
and
field.
fields.
Polyphase quantities.
The
1.
rotating magnetic
Pivot
fields.
fields.
fields.
2.
Cylinder
33.
Components of the rotating pivot

a.
b.
I.
34. Production of a rotating pivot field.
By means
of two-phase currents,
of three-phase currents.
By means
Irregular rotating pivot fields.
d.
Limited use of pivot
fields.
field.
35.
Components of the rotating cylinder magnetic

a.
b.
c.
36. Practical production of rotating cylinder
magnetic
fields.
Limited number of phases employed. The cylinder fields formed by two- and three-phase currents fluctuate. Rotating cylinder fields produced by two-phase currents.
Rotating cylinder
fields
d.
produced by three-phase currents.

a.
32.
Polyphase E.m.f.s,* Currents, and Fields

electrical engineering frequent use
is
In
modern
made
of two,
three, four, or six
and sometimes more associated alternating

machinery
This
designated
pressures, currents, and magnetic fields in electrical
and
its
connecting circuits and auxiliary apparatus.

is
multiple arrangement of alternating quantities

as polyphase.
Polyphase
is
the general term used to desigis
nate that more than one alternating quantity
in use.
The
and
corresponding terms for actual cases are two-phase, three-
phase, four-phase,
etc.
Polyphase e.m.fs,
currents,
magnetic
fields are
equal in value and differ equally in phase.

142
32]
ROT/iTlNG MAGNETIC FIELDS.
143
In machinery the coils forming the several circuits carrying

the
several currents
and the magnetic
fields
they establish
are spaced equally with respect to one another.
Polyphase magnetic' magnetic

fields,
fields
give rise to uniform rotating
and are very
useful for a variety of purposes
in engineering.
Their significance will be better understood
by means of the following illustrations that rotating magnetic fields and their components afford. b. The Hotating Magnetic Field. Rotating magnetic fields as applied in engineering occur in two classes, which may be
called pivot and cylinder fields.

1
pivot field
may
be any single
the
field established in air

is
or other mobile
medium wherein
field
made
to revolve
about a definite axis located approximately at the centre of the

field
and
at right angles to the flux.

1
In Fig.
02 the circular
coil
CC, carrying a continuous
current, establishes a field of magnetic flux distributed as the

lines in the figure
show.
When
the coil
is
stationary and the
current constant the field of flux

is
correspondingly
stationary
coil
and constant.
through
its
If the
be
rotated about an axis passing
own
field
plane
and
centre of the
of flux, so
long as the current is maintained

in the coil
the flux will rotate

it
likewise precisely as though
were a rigid body attached to

the
coil.
Such a
field, in
what-
y\q.
102.
Model
field.
of
Rotating Pivot
ever manner produced, for the
Field.
sake of clearness will be called a pivot

2.
cylinder field consists of coordinated sets of positive
144
fields
[3*
and of negative
of flux established through a cylindrical
air-gap rotated about the axis of the air-gap cylinder.

Fig. 103 illustrates the production of such a field in the
Fig. 103.
Model of Rotating Cylinder Field.

This model employs tw6 magneti^-^
simplest possible manner.
cores built of electrical sheet steel.

cylindrical shell;
it
The
is
is
stationary and
outer core forms a mounted concentric
with an inner cylinder built up in the same manner.

inner core rotates.
The
Upon
it
in
notches or grooves are mounted
coils carrying continuous current introduced
through slip-rings
is
as indicated.
Between the inner and outer cylinders there
a clearance space or air-gap.

establish magnetic
fields
The
coils while carrying current
through and about themselves, as

the core upon which these coils
indicated in Fig. 103.

are
When
mounted
rotates, the coils
and the
fields of flux
they pro-
duce rotate with them.
These rotating cylinder
fields are
used extensively in mod-
ern electrical machinery where they are invariably produced
by means of alternating currents, as will be shown later. The particular model above used has but little application. It has
been introduced here
precisely
for
the express purpose of making clear

field.
what
is
meant by rotating cylinder
33]
ROTATING MAGNETIC
33.
FIELDS.
US
Components
of the
field
Rotating Pivot Field.
Any
uniform stationary single

Fig. 102,
of flux, such as illustrated in
may
be broken up into components along any set

such
field
of axes.
When
rotates
its
components remain
If the
field
stationary in position
rotates uniformly,
it
and alternate
in value.
will
be seen that the components
alter-
nate as sine values.

Algebraically,
it is
more convenient
to study
components
taken at right angles than along axes at other angles.

rotating
field
may. however, be resolved
into
components
of. its
along any axis by writing each component in terms

Graphically, one set of axes are as convenient as another.

d.
Rectangular Components of a Pivot Field.
In
Fig. 104
let
(5
be a uniform
field
of magnetic flux parallel to the plane
Fig. 104.
Rectangular Components of the Pivot Field.
of this paper and making the angle B with the horizontal or

axis.
It
may
be assumed
that this uniform field of flux
was
established over a limited space at the centre of the coil

in Fig. 102.
CC
The
field at that
point
is
quite uniform for small

coil.
distances in any direction from the centre of the
The
horizontal
component
cos B
ox
of this field
is
sin
(^+3.
146
vertical
[34
and the
component
is
#
Suppose the
0'
sin d.
iield
$ be
rotated so as to
with the horizontal.
The
horizontal
new angle and vertical com-
make
ponents will
now be
^
<?
sin [d' sin

6'.
+ J),
It is
evident that a continuation of the angular displacement

v/ill
of the flux, ^, so as to pass through a complete revolution

at the
same time cause the horizontal and

the
field rotates
vertical
components
to vary as curves of sines throughdut one complete cycle.
When
and
uniformly, therefore, the horizontal

field constitute
vertical
components of such
simple fixed
alternating fields of flux that are equal and in quadrature phase.
From this it follows that Two equal simple alternating

a uniform rotating magnetic field.
fields
offlux located at right
angles, displaced in phase by one quarter of a cycle, will produce
34. Production of a Rotating Pivot Field.
a.
By means
carrying
of Two-phase Currents.
current which sets up a
tion
is
In Fig.
field
102
is
shown a
coil
of flux that in density and direclines.
given by the curving
When
the current alter-
nates, the field of flux alternates in the
same manner but without

the centre of this coil
is
change of
there
relative density
and
direction, i.e., the character of
flux distribution remains the

is
same.
At
a small region in which the field
uniform in amount
field set
and
direction.
Keeping these
and
facts in
mind, the
up
at the centre of the coils in Fig. 105
may be
studied.
These
coils
2 are circular in
form and mounted at
right angles to one another with their centres at a

point.
common
i
.
An
alternating current
is
set
up through
coil
In
34]
ROTATING MAGNETIC
FIELDS.
147
in Strength
coil 2 there is set
up an alternating current equal

i
to
and
in
quadrature with that in
These
coils will establish at their centres alternating flux
densities given in
maximum
and
amount, direction, and phase by

If the currents
the vectors B^ and B^ respectively.

sines, the horizontal
vary as
will
vertical flux densities B^
and B^
vary as sines and be exactly equal to the horizontal and ver00000 00000 00000 OOOOO
00000 00000 00000 00000 OOQOQ
148
b.
[34
By Means of Three-phase
Currents.
In Fig.
;
106 three
coils, Nos. I, 2, and 3, are mounted about a common centre and spaced with symmetry 120 apart. Assume that an alter-
nating current passes through each of these coils
that such
currents are equal and that they differ in phase from each other
by
120.
Each
coil will establish a
uniformly distributed alterits
nating
field
of flux in a small region about
centre.
The
direction of this flux will be at right angles to the plane of the

coil.
The
three coils will establish three components of a rotatis
ing flux that
uniform in density and angular velocity.
This
follows from the fact that
when any uniform
field
rotating at
uniform velocity
at 120, such
is
broken up into components along axes taken
components constitute an arrangement of threecoils in Fig. 106,
phase alternating fluxes identical as to character with three

fluxes
formed at the centres of the
ticular value of constant, uniformly rotating field
Some parwhen broken

coils.
up
in this
way
will resolve into

field
components
identical in char-
acter
and amount with the
formed at the centre of the
The combination of these fields must produce that same value

of uniformly rotating field.
The
value of the rotating field formed by a three-phase
field is easily
determined, since
at
it is
only necessary to add the
three fields
any
instant.
This can be accomplished by
resolving the fields into rectangular components.

field will
The
resultant
then be obtained at any instant by taking the square
root of the
sum
of the squares of the
sums of vertical and
hori-
zontal components respectively.
Thus the values

time are
of the three-phase fields as functions ot
B^ sin
e,
BjSin
B^
(^+
120),
sin (9 4- 240),
where
& varies uniformly with time.
34]
ROTATING MAGNETIC
Coil
FIELDS.
it
149
will then
No.
is
located in the vertical axis;
establish flux in the horizontal axis
which does not have to be
resolved into components.

flux densities
and 3 will establish and B^, 120 and 240 remote from B^, and B^
coils 2
The
as
shown
in Fig. 106.
The
values of the vertical components oi B^ and B^ will
therefore be
sin 120 sin
B^
sin {0
sin {d
-\.
120)
240 ^3
240)
= + .866^^ sin {d + 120); = .866^3 sin {6 + 240)

will
and the horizontal components

cos 120 ^2 sin ((9+ 120)
cos 240 ^3 sin
(6^
be
(i9
240)
= ^B^sm + 120); = ^^3 sin (B 240).

-\-
The sum
of the vertical components will be
.866^2 sin [0
120)
.866^3 sin {6
+ 240),
and of the horizontal components

B^ sin d
^B^
sin [6
+ 120) - ^B^ sin {B + 240).

field will
The
value of the rotating
be the combination.of these
Rotating
B=
{.S66\B^
+
B^
[5jSin(9-
+ 120) - B^ sin {6 + 240)]^ ^\B^sm{d+ 120) + ^jsin ((9+ 240)|]^[*.

sin {d
On
substituting
any
particular value of 6
and denoting
=
c.
B^
B^ by ^max. we have
Rotating
B=
i.S-^âx.-
Irregular Rotating Pivot Fields.
Practical
application
of the pivot field is made occasionally where the components The forming the same differ in amount and phase position.
above treatment applies to these
fields also
when note
is
made
of the fact that the values of B^, B^, etc., are different in
15
that they have their
[SS
amount and
^j. ^2, etc.
own
different
phase positions,
Irregular components produce rotating fields that fluctuate

in value
d.
and angular velocity.
Limited Use of Pivot Fields.
But
in
a limited analysis
air
needs to be made of the pivot
field
fi-ee
or a
homo-
geneous medium.
There
is
little
need
in
engineering for
determining values of rotating

the
air.
fields established
completely in
for
Such
fields
are only used in instruments and
demonstrational experiments.
The
rotating cylinder fields are far
more
useful in engineer-
ing on account of their application in the induction motor,
frequency converters, phase transformers, etc., where they are
formed by multiple fixed cylindrical alternating

35.
fields.
Components
103
in a
of the Rotating Cylinder

coils
Magnetic Field.
and
In
each
coils
Fig.
assume that the
are spread out
mounted
number of small grooves instead of being bunched
in a single set of grooves.
Assume
will
further that these
have been spread out
in
such a fashion as to cause them
to
establish fields of flux
which
have a circumferential
density distribution through the cylindrical air-gap correspond-
ing to the curve of sines.
In the air-gap over the surface of

fall
the magnetic cores the flux densities will rise and

tive
in posi-
and negative values as
sine waves.
After assuming also
that the inner core, coils, and field are revolving at a uniform
angular velocity there will exist in the air-gap cylinder of this
model a rotating cylinder

It is
field
having the form applied in
engineering and for which the components are to be determined.
important that a clear idea be gained of the character

field
of the rotating
this the
here specified.
To
assist in
accomplishing
cyHnder
field
has been developed in Fig. 107.
The
arrow-heads show by their length and position the flux density

as
it
occurs from point to point in the
field.
Assume
that the
3S]
ROTATING MAGNETIC
is
FIELDS.
in Fig.
151
direction of rotation
such that the waves move

is
107
from
left
to right.
Since this diagram
the development of
a cylinder, the points

off at
and
are identical.
As
wave runs
field is
B it
simultaneously appears at
This cylinder
here clearly seen to be one rotating at a constant velocity and
having a circumferential sine-wave flux density
distribution.
Fig. 107.
Development of Sine-form Cylinder
Field.
The components
rotating field
of a cylinder field of this character will

in
be best understood by noting the manner
which such a
may
by
be produced by means of a number of sets

polyphase currents,
side
of
coils
carrying
and
mounted
sym-
metrically side
on one of the cores
in Fig. 103 in lieu
of the coils there shown.
To make
this clear
assume the
actual case where nine circuits form nine sets of coils, each
carrying one of nine polyphase currents.

core be stationary.
are
In Fig. 103 let each
On
the inner core the nine sets of coils

to one another
mounted symmetrically with respect

Fig. 108 gives a diagram
these' circuits or coils are
and
the cylinder.
in
showing the manner
which
mounted.
AA'BB' is
a por-
tion of the surface of the core developed whereon these circuits
are mounted.
its
As
this
diagram
indicates, each circuit threads
way
across the surface of the cylinder as
many
times as
152
L35
there are to be poles or plus and minus regions of flux in the
cylinder
field.
The beginning and

numbered
6', 7',
8',
i,
terminal ends
8,
of these
circuits
3', 4',
are
S',
2,
3, 4,
5, 6, 7,
and
9,
and
i',
2',
and
9'.
In electrical machinery the portion of a
circuit that crosses a core in
each instance in the manner here
applied
is
called an inductor.
Assuming that there
are to be
1^
2
3
35]
ROTATING MAGNETIC
FIELDS.
153
in the reverse order
have a negative position.

in
This distinc-
tion
is
made
solely for the sake of convenience in distinguishing
between the two directions

the core upon which
it
which the
circuit is led across
is
mounted.
It is also
necessary to
have a convenient means
for referring to the positions
occupied
Fig. 109. Formation of a Uniform Sine- wave Rotating Cylinder Field by Means of a Polyphase Arrangement of Circuits and Currents.
by the inductors on the
surface of the core.
To do
this the
circumferential distance on the surface
of the core between
neigboring inductors or sets of inductors of the same sign
belonging to the same
circuit
is
called a polar interval.

all
As
stated above these fifty-four inductors occupy
of the
surface of the core.

interval
is
developed section
for a single polar
given in the lower portion of Fig.
109.
AB
is
taken along the surface of the outer core, and
CD
along the
154
[35
surface of the inner core.
The
air-gap separating the
two
cores
is
^C
or
BD.
It is
seen that each inductor occupies a
circumferential surface of one eighteenth of a polar interval or
cycle of inductor-circuit connection.
It is
customary to refer
to such a space in degrees, counting a polar interval 360.
The
interval
between inductor centres

field
is,
therefore, 20.
To
form the rotating
a separate sine-wave alternating current

circuits.
must be passed through each of the nine

rents
These cur-
must be
alike in periodicity
and amounts, while they must

Neighboring currents
differ in
phase by an amount that corresponds to the circumdisplacement of their

differ in
ferential
circuits.
must, therefore,
phase by
20, corresponding to the
position displacement of the circuits in
which they
exist.
A comprehensive diagram
circuits,
of this arrangement of inductors,
and currents
is i^,
given in Fig.
no.
The
nine sine
waves of current-time
of one current
i^, i^, i^, i^, i^, i^, i^, i^,

,
have been
length
established through the corresponding circuits.
If the
wave
in this
diagram be allowed to represent

at;
also a polar interval, then the position
wljich the currents
pass through zero
may
represent the location of the inductors
of the corresponding circuits.
Thus the numbers immediately

give the circuits
to
under the heading
circuits
which the
distinguish
inductors in this part of the cylinder belong.
To
negative from positive inductors their corresponding inductors

are
marked prime.
Frqm
in
is
this
diagram one
may obtain
in
the value of the current
any inductor and therefore

possible to
any
circuit at
any
instant.
It
know from
instant to instant the values
and signs
be deter-
of all the currents in all the inductors, from
which the resulting
m.m.f. and therefore the
fields
they establish
may
mined.
This has been done for ten different instants, and the
values of the currents and their direction across the core have
been recorded
in the table given in Fig.
10.
3S]
ROTATING MAGNETIC
FIELDS.
it
155
is
In order to note the m.m.f. of a circuit the value of the current and
It is
its
necessary that
direction in the circuit be
known.
necessary in this instance, therefore, to

100
know whether
156
tors
[35
when
it
will pass
through a positive inductor from the rear
to the front of the cylindrical core.
In the table positively directed currents occur in light-
bodied type and those which are negatively directed occur in
heavy-bodied type.
light value of current, therefore, passes
through
its
circuit over the core
from rear to
front, while the
black values do the opposite.

tion will be positive
The
sign of the current direcits
are alike,
when the signs of and negative when not alike.

^
inductor and
itself
Let the instant

first
in the
diagram be selected
for the
determination of the m.m.f.s that are produced

circuits.
by these
polyphase
The
vertical line at this instant will cut
the nine current waves, and the distances from these intercepts
to the axis of the
waves measure the corresponding instan-
taneous currents, which at this
moment
are
all
negative.
The
current direction through the nine positive inductors at this

instant
is
therefore minus
and the values are recorded
in black.
The next series of nine inductors are negative, through which when negatively directed currents are passed a positive direcThe currents in these nine tion across the core results.
negative inductors have the same values as in their correspond-
ing positive inductors, sincj they are in the same corresponding

circuits.
The same corresponding values

signs.
of current are there-
fore
always given
for the negative inductors as for their positive
companions except with opposite
With
this
understanding the table can

time
for
3'
made up
1
for the other instants of

/
now be quickly selected. The values

interval were:
6' 7'
3'
of the currents at the instant

2 3
=
9
o
1'
any polar
4'
5'
8'
9'
00
34
64
t
87
I
98
98
87
64
34
00
34
64
87
98
98
87
64
i
34
When
^circuits
the value of the current in circuit No.

-|-
has
increased from o to
34, while the currents in all of the other
have changed, remaining negative.
The
currents
and
35]
ROTATING MAGNETIC
FIELDS.
157
their directions in inductors of each polar interval
have there-
fore
1
become
2
3
34
00
34
4 64
1'
3'
3'
4'
5'
6'
7'
8'
9'
87
98
98
87
64
34
00
34
64
87
98
98
87
64
In precisely this
tions
same manner the currents and

for
their direc-
may be
determined
any other instant whatsoever..
The
table
table carries out the results for instants taken at con^
venient intervals up to
10 or for angles up to 200.

for instants
The
results-
might have been made up
taken at any
shorter or longer, regular or irregular interval.
The
would have been the same
as in this case,
i.e.,
a regular for-
ward movement
of the positive
and negative m.m.f s.
The
field
of flux established by these moving

in
4.
m.m.f s has
6, 7, 8, 9,
been determined
to the instant
1',
connection with Fig. 109, corresponding

It is
seen that inductors
5,
2',
3',
and
4'
convey current across the core

in
in a negative
direction
and are drawn
black section in the diagram.
The
remaining inductors in this polar interval, 1,2,

8', 9',
3, 4, 5', 6', 7',
have current
in
them traversing the core

section.
in a positive
direction
and are drawn with clear
The
current-
inductor status in the other polar intervals of this

cisely the
field is
pre-
same
as in the one given in this diagram.
Over the
These
circumference of the cylinder there are, therefore, alternate

regions of nine positive and nine negative inductors.
present about themselves in closed circuits through the air-gaps
and cores ampere-turns of m.m.f as indicated by the closed

curves in Fig. 109.
It is
seen that the
maximum amperediminish in closed
turns occur in closed circuits about the whole set of positive
or negative inductors.
circuits within the
The ampere-turns
space occupied by each positive or negative

values of zero are reached at their
set until
the
minimum
centres.
The
values of the currents in the inductors have
been
iS8
[35
plotted in rectangular coordinates in the upper portion of Fig.

109.
This curve
Its
is
a sine
wave and
is
is
labelled ampere-in-
ductors.
maximum
value
100 ampere-turns, while the

i'
ampere-turns of inductors 9 and

of the
which
lie
on either side
maximum have
'
the values each of
100
in
sin (4
20)
98.48,
or 98 as put
down
the table.
the
The m.m.f. exerted about the inductors will be given by sum of the inductor-currents included within each closed
circuit
magnetic
taken symmetrically with respect to positive
and negative sets.*

table
These sums
.are
given in the following
35]
ROTATING MAGNETIC
FIELDS.
159
for
Corresponding values of flux density have been determined
the other closed circuits and their ampere-turns given in the
above
table,
and the
results
have been platted to locate the
wave
/
labelled Rotating Flux, in Fig. 109.
The wave
is
drawn
here in amount and position corresponding, to a single instant
4.
As
time progresses the m.m.f.s that establish the flux

in
wave do not change
value nor distribution, but
move
pro1
gressively forward with time as
shown
in
the table of Fig.
10.
As
the m.m.f.s of each polar interval

establish
move forward
the flux
waves they
move
with them simultaneously.
This
progression circumferentially with time keeps up uniformly

until the entire circumference of the cylinder has
been traversed
so
by each wave.
inductors.
Then
the process repeats
itself indefinitely
long as the nine-phased currents are maintained through the
Thus
it is
seen that by means of a

field
static structure
a rotat-
ing magnetic
having six poles
is
formed having the same
form and character as the rotating field established in the mechanically operated model as described in connection with
Figs. 103 and 107.
A
wave
lishes
study of this wave shows that
it
obeys the sine law and

of m.m.f. that estab-
occupies a position in lagging quadrature with respect to the

of inductor-currents.
it
The wave
has for
its
values the sums of the corresponding

in the
ampere-inductors taken
the characteristics
manner shown above. From of the rotating field formed by polyphase
currents as above considered the following properties of the
general case wherein the rotating cylinder field is formed by any number of polyphase currents having a symmetrical phase and inductor displacement may now be determined.
Let
/max.
be the
maximum
value of each of the polyphase
currents
n, their
number;
l6o
a, the
[35
number
of times each circuit traverses the core
circumferentially
/, the frequency of the currents
/,
the single thickness of the cylindrical air-gap; and

field
, the flux density of the rotating

air-gap.
formed
in
the
In the above example each circuit ended after traversing the circumference of the core once.
These
circuits
might have
been made
to traverse the core
two or more times, each time

first
going over the same circumferential route as the

displaced
when
of
by one inductor
space.
Denoting the number

in a positive or
times the circuit has thus been applied by a as stated above,

it
follows that the
number of inductors
negative
collection of inductor-currents
where Ij
is
the number of inductors.

value of the wave of m.m.f. applied per
The maximum
double air-gap, neglecting the reluctance of the iron, will be
H=
lo
an
y_
7t
""
In this expression, by Sec. lO^,
max.
is
the average value of the current in any set of positive or
negative inductor-currents.
The sum
of such must be
7t
and
TMs
turns
.
is
also the value of the m.m.f. of each set in
ampere-
3S]
ROTATING MAGNETIC
It follows, therefore, that the
FIELDS.
161
maximum
m.m.f. in gilberts
is
of positive or negative sets of inductor-currents

4.71
The maximum
will
density of the established rotating flux wave
be
Rotating ^.
=H 2l^
Ti^nl^^,
(88)
Since the flux has a sine wave distribution, by Sec. 10^,

the average value of the flux density established will be
Rotating
b.
B.^^,
It
(rotating B^^^).
(8g)
Terminology of Polyphase
Currefits
Applied for the
Production of Rotating Cylinder Fields.

nine
In the above example

a rotating
field.
phases were used to produce
The
inductors were spaced apart one eighteenth of a polar interval or flux
wave
length.
The
currents had a neighboring phase

It
difference of one eighteenth of a period, or 20".
appears,
then, that there were applied here nine phases having neigh-
boring phase differences
of one
eighteenth
period.
It
is
customary to
call
such a polyphase aggregation of currents a
nine-phase current or an eighteenth-phased current.
The term "nine-phase current" simply

circuits,
implies that nine

differing in
' '
each having an alternating current
phase
eighteenthposition from the others, are employed. The term phased current" states that the associated currents have an interval phase difference of one eighteenth period, leaving one
to infer from general
knowledge that nine phases only are
used.
The reason why
nine phases are used in lieu of eighteen

is
phases with an eighteenth period and polar interval spacing
i62
ELBCTRIC/1L MACHINERY.
fact
:
[3^
is
due to the following

lent of a
Any wave
of current
the equiva-
wave of equal magnitude one

it
half period or
is
i8o
sent
remote from
through the
direction.
in
phase position, provided such current
circuit
under consideration in a negative or reverse

easily accomplished
is
This
is
by reversing the
ter-
minals of that part of the circuit which

in a reverse direction.
to receive the current
In the above rotating cylinder-field
exaniple this was done by running the circuits through their

inductor portions in
directions alternately positive
and then
negative, thus in effect
applying an eighteen-phase current by
means
of a nine-phase current.
will
As
be seen
in
a later chapter the term n-phase current
does not always mean that the currents have the one2nth period
of phase displacement.
Often, as where
3,
the term three-
phase
a one
is
apt to refer to three equal alternating quantities having period of phase displacement.
jd
The term
n-phase,
therefore, in the terminology of the present time simply implies
that a polyphase circuit has n circuits wherein the currents
may have
period.
a neighboring phase difference of two nth or one nth
On
etc.,
the other hand the term " nth-phased current,'' e.m.f.,

It
has a perfectly definite meaning.

circuits
means that
one-
half n
and alternating quantities are employed having
a phase and position displacement of one nth period or polar

interval.
36. Practical Production of Rotating
Fields.
Magnetic Cylinder
a.
Limited Number of Phases Em-ployed.

for
The
of
requirements
simplicity in practice
limit
the
number
phases employed to two and three, with occasional applications
oifour and
six.
While the higher number of phases are more
desirable for the production of rotating cylinder fields in alter-
nating current motors and similar machinery, yet the extra
36]
ROTATING MAGNETIC
FIELDS.
163
complication due to the use of so
many circuits
has limited the
number
b.
of phases
employed
as stated above.
The Cylinder Fields Formed by Two- and Three-phase Currents Fluctuate. In the above illustration of the formation
of
a.
rotating cylinder field
by means
of nine phases, so large
a number was chosen in order that the changes in m.m.f. from

inductor to inductor in each set of positive or negative inductor-currents might not be excessive.
Under such circumstances
the
sum of
the inductor-currents and, therefore, ampere-turns
of each set remains approximately constant from instant to

instant.
The
rotating flux that
is
thus
established remains
larger the
is
hkewise approximately constant.
The
number
of
phases employed the more exactly constant

rotating field.
the established
Where
in
but two and three phases are employed the change

is
m.m.f.
excessive from inductor to inductor.
The
total
m.m.f. of each set of inductor-currents fluctuates from instant

to instant
and the wave of rotating
flux does not
have the sine
form
c.
it
has a variable form and a fluctuating value.
Rotating Cylinder Field Produced by Two-phase Cur-
rents.
In the upper portion of Fig.
1 1 1
is
given a diagram of
the circuits of a two-phase arrangement of inductors for the
production of a rotating cylinder
field.
Under
all
circum-
stances a similar arrangement of inductors and their connecting

circuits
used in any class of electrical machinery
is
called a
winding.
In the winding of Fig.
1 1 1
each of the two-phase
circuits
traverse the circumference of the core three times.
Thus
the
inductors of each phase occur
in alternate
first
groups of three each

in
on the surface of the core.

series through inductors
I,
The
phased current passes
2, 3, 7, 8, 9,
13,
14,
IS, etc.,
164
circuit passes
[36
and the second phased

inductors,
through the remaining
4, 5, 6, 10, II, 12, 16, 17, 18, etc.
As
the diagram indicates, the currents and inductors occupy
the quarter-phased relation similar to the eighteenth-phaseu

l"\
36]
ROTATING MAGNETIC
FIELDS.
165
great a transition of m.m.f. values to occur from inductor to

inductor.
To
der
are
bring out the
maximum
variation in the rotating cyHn-
field
produced by a two-phase current where the inductors

in a
mounted
winding
like that
given in Fig.
1 1
1,
the wave
of established magnetic flux

limits of such variation.
may
be determined at the extreme

will occur as follows
field flux
These
The maximum
when one
is
value of the
density occurs
when
the two-phase currents are equal,

zero and the other
and the minimum value
maximum.
been plotted from the values
zero and the other maxi-
In Fig. Ill the curve
B has
efflux established the curve B'
when
the two-phase currents are equal, and
when one has become
mum.
The
reader will have no difficulty in repeating this
operation for his
own
satisfaction, using the
method employed
for
in connection with Figs. 109
and no, and assuming

is
con-
venience that the length of the double air-gap

values of
equal.
such that the

will
and the m.m.f.s establishing the same
be
The
values of the two-phase currents are
âx. sin
sin
(90)
22=^.ax.
(;ir
--^j
(91)
The
flux
wave
B was
traced for
'
>
-max. ^^
^=
when
i^
135,
sin
135=
+ -707,
|)
2;=sin (135
= + .707,
8
and the corresponding inductor-current values are

]
3
.707
3
.707
4
.707
5
.707
6
.707
10
11
18 etc.
.707
.707
.707
.707
.707
.707
.707 "
66
[36
The wave B' was

45 and
/j
traced
when
all
values had
moved forward
= =
sin 180
0.0,
sin
180
-j
=+
i.o,
123456789 0001 IIOOOI

The heavy
cross-section.
and the corresponding inductor-current values had changed to

10
11
I
12
I
etc.
angular lines in these curves locate the flux that

if
would be established
the inductors did not occupy appreciable

to the fact that their size
is
Owing
always con-
siderable with reference to the depth of the air-gap
distances between themselves, the resulting
and the wave of flux is
always
less
angular in outline and more like the broken line

in
waves that are drawn
over the angular waves.
By
in
locating the flux waves at other instants they will be
found to conform to values intermediate between those given
and B'.
values of
The maximum
and B' are
Rotating B^^^_
Rotating B'^^,
Since
Sin 45
their ratio
is
= 6 X 707 = 4-242, = 3X1 = 3.000. =

I
.707
=-^-,
In the general case for the two-phase circuit this ratio

^max.
is
W = ~:^ = ^max. aV2
^2.
36]
ROTATING MAGNETIC
FIELDS.
167
In the rotating cylinder field established by two-phase currents the
maximum and minimum values of the rotating m..m.f.s

densities
and flux
It is
have the ratio V2.
proper at this point to note a comparison of the value

rotating m.m.f. of a two-phase current set
of the
maximum
up
through the above cyHnder winding, as given by the general

formula developed
in Sec. 35,
with those values given by
and B' which are

turns.
identical with their corresponding
ampere-
The
value of
H in equation (87) H =^ 10
is
ani. max.*
71
and
in
ampere-turns
i68
I36
by
B and
B' and equation (89)
may
also be
made.
and ^'
are averaged from the curves as follows:

^t.av.
^, -4't.av.
=^t.av.
^t.
av.
2. 121) \ Ratio .Qi, ^ ^'t.av.= 2.333) 12 2 by equation (89) 2.430. X
From
these values
it is
seen that
The average rotating cylinder flux established by a twophase current fluctuates over a range of ten per cent during each
eighth
cycle.
The average rotating cylinder flux thus established

indicates should be the case. *
d.
is
slightly less than the equation for the general polyphase circuit
Cylinder Fields Produced by Three-phase Currents.

is
In
the upper portion of Fig. 112
given a diagram of a three-
phase winding that corresponds to the two-phase winding of

Fig. III.
The
value of
is
2; thus each set of positive or
negative inductors will number
I^
just the
= an=
6,
same
as applied in the
above two-phase case.
Thus
the winding
may employ the same
inductors, inductor spacing,
and air-gap depth which permits a comparison of the chief

features of the rotating cylinder fields
three-phase currents.
formed by the two- and These are the two forms of polyphase
current chiefly used in practice for the production of rotating

fields.
must separate the less as the^numStated in another way, the difference is due to ber of phases in use is increased. the rectangular irregular deviations from the sine form wherein only the outer corners are to be found on the edges of the sine wave to which they belong. With a large number of phases this difference must, therefore, entirely disappear.
is
* This difference
due
to the appreciable distances that
inductors of a two- and a three-phase winding.
The
effect
grows
36]
ROTATING MAGNETIC
For the three-phase
FIELDS.
169
circuits the highest of the
maximum
of the cur-
values of the rotating flux wave occurs

rents
is
when one
at its
PHAfiP
1
maximum and
_^
the other two have a value of half
170
ELECTRICAL MACHJNERY.
This
is
[3^
not only a three-phase current but a third-phased

is
current as well, and as such

rotating cylinder fields.
not generally used to produce
The
Thus
current i'"
its
is
shifted in
phase by n, or i8o, by simply

it
reversing
i'"
connection in the circuit where
is
applied.
becomes
i'" z=i^= /^.

=
and by writing
J'm^x.
sin [x
-j-j,
]
; .
(95)
sin
(.ir
(96)
i"
the
three-phase, one-third-phased
current becomes
a three-
phase, one-sixth-phased current, as given
by the equations
(97)
4,ax. = âx. =
sin
jr,
h
4
sin (jr
-],
(98)
4ax. sin (^
y)
(99)
In the winding diagram of Fig.

phases, circuits,
No.
I
I, 7.
112 the corresponding
and inductors belong together as follows

Inductors.
i3>
etc.; etc.; etc.;
2,
8,
14, 16,
etc. etc. etc.
3, 5.
9.
IS.
'7-
4,
6,
10,
12,
".
18,
An
half
inspection of equations (97), (98), and (99) shows that

will
one of the values
be
maximum and
X
the other two each a
maximum when
210,
36]
ROTATING MAGNETIC
FIELDS.
171
'which will locate
in
a convenient part of Fig. 112.

-'max.
Assume
>
when
the values of the three currents at this instant will be

/j
= = =
sin
210
.5,
sin (210 in
-^) = +.S,
j
sin (^210"
3
-f- i-O,
and the inductor-currents

Inductors....
1
will
5
I
be
6
I
3-4
.5
7
.5
8
.5
9
.$
10
-S
11
I
13, etc.
I,
Currents
.5.5
etc.
and the wave

values.
in
Fig. 112
is
laid
down by means
(97), (98),
of these
Further inspection of equations

that the next instant at
and
(99)
shows
which one of the currents
will
be zero
and the other two equal occurs when
240.
The
values of the currents will then be

j
= = =
sin
240
i
?2
sm
= .866, A (240 ^ 0.0,

'
27r\
j
sin \^240
= +
.866,
and the corresponding inductor currents

Inductors
1334 56789
of these values the rotating flux
will
be
10
Currents .866 ,.866
0.0 0.0 .866 .866 .866 .866 0.0 0.0
11 .866
13, etc.
.866
By means
down.
wave has
b.een laid
The
air-gap in this instance has a double depth such
that values of
identical.
and corresponding values of ampere-turns

is
ai-e
This
done
for
convenience only.
From
these
172
[36
curves and the general formulse applicable to this case the fol-
lowing results
may now be deduced:

^'_.
3.464
r^"^^^
2
3.732.
12
A^ max. by
âv.
B'.,
eq. (87)
anl^^_
1^
A^
av.
by
eq. (89)
it
3-8i5= 2 333. ^ 2.309, = 2.430.
From
the above result
follows that
TAe average rotating cylinder flux established by a threephase current fliectiiates over a range of about one per cent during each twelfth cycle. The average rotating cylinder flux thus established
indicates should be the
is
slightly less than the equation for the general polyphase circuit
case*
* See foot-note on page 168.
CHAPTER
IX.
THE ELECTROSTATIC
FIELD.
37. General characteristics of the electrostatic field.

u.
b.
c.
(/.
The
field
of electrostatic
flux.
Specific inductive capacity, J^.
The dielectric flux constant, i. The electric pressure gradient,

Experiments that
illustrate
G.
38.
The
electrostatic corona.
tt.
and define corona phenomena.
i.
c.
Testing dielectrics for break-down gradient.
Break-down
test of a
lead-covered cable.
39. Dielectric thickness required to avoid corona.
40. Dielectric hysteresis.

,
41. Dielectric conduction.
37. General Characteristics of the Electrostatic Field.
The
use
electrostatic field exists

It
everywhere about the

direct practical value
electricits
current circuit.
in
has
little
beyond
without
electrostatic
indicating
it is
or
measuring instruments.
applied;
its
Indirectly,
however,
universally
service
definite
no
electric current could
be confined and limited to a
path along a narrow conductor.

actual physical properties of the electrostatic field were
field.
The
has,
determined as early as those of the .magnetic

however, been much
less
There
working experience with the
electrostatic field than with the electric current
and the mag-
netic field.
fact that
The cause of this lack of experience is due to the we make so little direct use of electrostatic actions.
practical concern that the electrostatic field usually
173
The most
174
gives
is
L^7
the failure of a dielectric under electric pressure to In the event

is
withstand the strains that are thereby produced.

of such failure the dielectric in which the field
formed
is
ruptured and a current follows, forming the
common phenom-
enon of arcing.
Prior to the use of high-pressure alternating currents for the
economical long-distance transmission of power,

culty
little
real diffi-
was met with
in
providing insulations in machinery and
on the
lines to withstand pressures
up to 2000
volts.
Later,
when
the attempt was
made
to use
alternating pressures of
difficulties
10,000 volts and upwards, great insulation

found and had to be overcome.
experience with the electrostatic
e.m.f.s has been gained.
a.
In overcoming
field
were them much
produced by alternating
The Field of Electrostatic Flux.

of e.m.f.
In Fig.
11 2,
and
are metal electrodes that are connected to the terminals of
some source
These electrodes are rods that terminate
Fig. 113.
in discs
between the faces of which
is
mounted a
plate of glass
as shown.
On
applying a moderate electric pressure to

field of flux will
is
and
an electrostatic
be established
figure
in density
lines.
and direction that
shown in the The conventional meaning of these

field
by the curved
lines thus
used to represent
the electrostatic
of flux
is
exactly the same as the corre-
37]
THE ELECTROSTATIC
FIELD.
175
to represent
spending meaning of similar

a magnetic
field
lines that are
drawn
of flux.
In any locality the lines have the

their
is
direction of the flux,

i.e., their
and
number
in a unit cross-section,
rate of occurrence,
proportional to the electrostatic
flux density.
Every medium or material through which an

current
is
electric pres-
sure establishes an electric molecular strain in lieu of an electric

called a dielectric.
All non-conductors or insulators
are dielectrics.
The
electrostatic field
is
strain
produced
in the dielectric
that everywhere surrounds conductors between which an e.m.f.
has produced a difference of potential.
The
strain is distributed
its
everywhere
in
such a manner as to require for
application
the same potential difference between the two conductors by
any route whatsoever. 'Such a strain about conductors between which there
tial
in the dielectric
is
media
a difference of poten-
constitutes the electrostatic field of flux.

is
The
quantitaas that
tive character of its establishment

for
precisely the
same
magnetic flux about a permanent magnet or other source

In
of m.m.f.
the production of the strain that constitutes
amount of current must pass for a In other and more conventional phraseology, definite time. quantity or charge of 'electricity must be applied to a definite
electrostatic flux, a definite
the dielectric.
Simultaneously with the flow of current through a
dielectric,
a counter-electromotive force
to
is
formed
in it that is
i.e.,
proportional
the
time-integral of the current,
the product of the
current strength
and
the time that elapses while such current is
passing.
So small
is
the time-integral of current that

in
is
ordinarily
necessary for the production
the dielectric of a counter-e.m.f.
that is equal and opposite to the impressed or actuating e.m.f. that this process for most practical purposes may be assumed to
176
ELECTRlCy4L M/iCHINERY.
[37
occur instantaneously.
On
the formation of a counter-e.m.f.

,
which
tric
is
equal and opposite to the impressed e.m.f.
the dielec-
ceases completely to carry current.

insignificant
is this
Of so
in
little
importance
and so
phenomenon
connection with alL
ordinary electric circuits at ordinary e.m.f.s that the convention
from the beginning of electrical science has been to say that no

current passes through a dielectric
electric pressure.
when
it
is
subjected to an
it is
In dealing with condensers
the general
practice even to-day to adopt the convention that electricity is
some iking
instead of a form of energy.
For example,
we
say a condenser has been charged, and in speaking further of

the process of charging a condenser
we say
that a positive
charge was taken to one face of the
dielectric of the
condenser
and from the other
face
an equal negative charge was removed.

is
The
real fact to appreciate
that a dielectric subjected to
difference of potential passes
no current only
after a time-inte-
gral of current has passed sufficient to cause the dielectric to
produce a counter-e.m.f equal to the e.m.f. which gives rise

to the difference of potential.
Informing
produced
this counter-e.m.f.
a strain or electrostatic flux

in
is
in the dielectric.
This
it
any case may eventually be made high enough

In
establishing electrostatic flux,
to cause
to rupture.
energy
is
used in an
amount equal
faces of
to the integral product of the current
through
the dielectric, the counter-e.m.f., and the time.

tlie
By uniting the
dielectric with
a conductor this energy will be
returned by the dielectric and will be dissipated as heat in the
conductor by the current that
is
established.
The amount
of
energy thus returned

stored,
is
never quite equal to the original amount
owing
to lack of perfect elasticity in the dielectric.
The
follows
chief characteristics of the electrostatic field are as
An
e.m.f. will
pass current through a
dielectric until there
37]
is
THE ELECTROSTATIC
in the dielectric
FIELD.
177
e.in.f.
formed
an equal and opposite
This
results in the formation of

dielectric.
a field of electrostatic flux within the

analogous
is
There exists an electrostatic-flux
circuit
to
the
magnetic-flux circuit while the impressed e.mf.
changing.
electrostatic
While the impressed e.mf.

circuit
is
is
changing the
always closed
in a
manner
entirely analogous to the

circuit is
closed magnetic circuit.
The magnetic
is
made up
of
magnetic flux and magnetic induction through magnetic bodies.
The
electrostatic
circuit
is
completed through dielectrics by

then, that an electric current
electrostatic flux that
changing, and by an electric current
through a conductor.
in
It follows,
one part of an
electrostatic
circuit is the equivalent of
definite rate of
change of
electrostatic flux in another part of
the circuit.
The analogy between

is
the electrostatic circuit and
the magnetic circuit

It
not complete in one important respect.

is
does not hold when the electrostatic flux

for this failure
not changing.
is
The reason
changing
is
when
the electrostatic flux
not
due to the
fact that in
magnetism there
in
is
nothing
to take the place of the
conductor
electrostatics.
in
That
which would correspond to a conductor
magnetism would
respects.
is
have to be a medium possessing nearly perfect permeability.*
The analogy
example,
in the
is
not complete in
circuit
many minor
For
magnetic
when
the m.m.f.
removed
the field of magnetic flux simultaneously ceases to exist.
The
trans-
energy stored
in the field in the

in
form of magnetic flux
is
formed to some other form

m.m.f.
the process of removing the
In the electrostatic circuit
when
the e.m.f.
is
removed
without removing the conducting portion of the circuit in which

*
The
electrostatic circuit tliat
is
analogous to the magnetic circuit
is
considered
density
is
here as being an electrostatic

changing.
field
wherein the electrostatic
flux
The
present state of the science does not clearly define a further

In most respects the electrostatic circuit should be underits
analogy between them.
stood to possess characteristics distinctly
own.
178
[37
the e.m.f. was produced, the electrostatic field of flux will dis-
appear simultaneously and the energy of which

will
it is
constituted
be transformed just as in the disappearance of the magnetic

If,
field.
however, we remove both the conductor and the

will
e.m.f at the same time the electrostatic flux

from the
dielectric.
It
not disappear
closed
It will
remain as a
circuit, partially
is
upon
itself
cannot disappear, as there

its
no operating means
or vehicle for the transformation of
energy.
The
electrostatic field
is,
is
everywhere manifested through a
dielectric, that
a non-conducting medium.
The
air is
the
universal dielectric within
which
all
conductors are ordinarily
immersed.
The
air is often, for
mechanical or other reasons,
displaced in the immediate neighborhood of the conductors

soH,d or hquid dielectrics.
by
Solid
and
liquid dielectrics are

is
more permeable
for electro-
static flux
is,
than
air,
or than are the gases generally.
That
at a given e.m.f.
impressed upon a given thickness and

is
volume a greater amount of energy

dielectrics than in
b.
stored in solid and liquid
gaseous dielectrics.
Specific Inductive Capacity,
K,
is
the ratio of the electroa

dielectric
static
flux that
is
established through
air
to
that
established
e.m.f.s.
through
at
corresponding
dimensions and
For convenience, hereafter dielectric " will be referred to as

c.
' '
electrostatic flux
through a
dielectric flux.
k,
is
The
Dielectric
Flux Constant,
the
number
of
coulombs that are passed through an inch cube of the

at a pressure of
dielectric
one volt between two opposite
faces.
The energy taken up by a one-inch cube of air at one

impressed between two opposite faces
Since a joule
is
volt
is
1.122
lO"^^ joules.
one coulomb-volt, and the average pressure

is
applied in establishing an electrostatic field of flux
one half
the
final
pressure,
it
follows that
the
number of coulombs
37]
THE ELECTROSTATIC
air
FIELD.
179
applied per inch cube of
per volt
is
twice the corresponding^
amount
The
is
of energy in joules.
Dielectric
This makes
k,
Flux Constant,
io~*^ coulombs.
for air as defined above
equal to 2.244
In any case the joules of energy, w, stored per cubic inch

of dielectric will be
zu
^kE^,
(100)
where
is
the pressure applied per inch of thickness of the
dielectric.
The corresponding coulombs
of dielectric flux will be
D = kE
d.
(loi)
is
The
Electric Pressure Gradient, G,
the e.m.f. applied
per unit thickness and, therefore, per inch of the dielectric.
Based on the above
definitions,
the following
tables of
dielectric properties will be found to be useful:
K
Name
of Dielectric.
Specific Inductive Capacity.
Dielectric Flux ConJoules of Energy stant for Inch Cube stored per Inch Cube Volt, to be multiat one Volt per inch, to be multiplied plied by 10 '3.
by
10
".
Vacuum
Air
hard, old hard, new hard, 13 extra dense flint lowest value Porcelain Shellac
Glass, Glass, Glass, Glass, Glass,
I
.9996
6.96
2.236 2.244
15.6
1. 1.
118 122
7,8
3"
3-31 9-9 2.8
6,98 7-44 22,25 6.30

9.9
3-49 3-72 11,12

3-
IS
4.4 2.74
3 2.12
Sulphur Rubber, pure Rubber, vulcanized Rubber, hard

Paraffin
Wax
Olive
oil
Turpentine
Sperm
oil
Petroleum, crude headlight " Petroleum, Vaseline oil

' '
2.69 2.25 2 1.86 3.08 2.2j 3-9 2.07

. .
6.14 6.75 4-75 6.04 5-05
4.50 4.17 6-93 S-OS

6.95
4-95 3-07 3-37 2-37 3,02 2,52 2,25 2,08 3-46 2.52
'
2. II
2.17
4-52 4-75 4.87
3-47 2,26 2.37 2-43
i8o
ELECTRIC/1 L M/tCHlNERY.
[38
Name
of Dielectric.
38]
THE ELECTROSTATIC
FIELD.
l8l
electrode and streamers are drawn in solid black, and the~ homogeneous corona, wherever it is seen to develop, is drawn by means of the broken-line shading. In looking at this phenomenon the eye sees many more streamers at any one
instant than really exist.
The
The
streamers form and die out with

Persistence of vision
each alternation of the electric pressure.

causes one to see
actually occur.
many more
In Fig.
streamers at any one instant than
114 the streamers and corona are
Fig. 114.
drawn much
as they are seen to exist at
any one
instant.
The
size of the plate is
about 8
10 inches.
Occasionally a pair
of streamers, one on either side, will build out to the edge of the plate.
circuit, in
They
will
unite,
forming for the instant a short
a thick, very bright straight streamer that emits a
loud report.
volts
On
raising the pressure to 30,000 alternating
many
of these longer streamers short-circuit over the
edges of the plate, emitting a series of loud reports.
Finally
some one streamer
will
get over the edge at a sufficiently
early stage in an alternation so as to produce heat enough in
the short-circuit path to form an alternating-current arc.
Such
an arc places a continuous short-circuit between the electrodes This is promptly stopped by and, therefore, upon the source.
the circuit-breaker in the primary of the high-pressure trans-
82
[3S
former.
The corona and
the non-short-circuiting streamers

to the
also emit sound,
complex note that corresponds
alternating periodicity.
If the
pressure between the electrodes

for
is
maintained at
20,000 or 25,000 volts

five
a few minutes, generally less than
minutes, the glass plate will puncture.

it
On examining
The
the
puncture critically
rather than
is
observed to be a small hole, melted

glass.
smashed through the
glass has been
heated very much.
The
points on the glass in the immediate
neighborhood of the electrodes upon which streamers strike

continuously become highly heated in a short time.
One
is
brought to the conclusion that these streamers are very hot.
By
operating at a lower pressure where the

is
homogeneous
be seen that
It is
is
corona
formed without the streamers,
it
will
such corona develops also a great deal of heat.
easily
discernible, however, that the heat of the streamers
enor-
mously localized as compared with that of the uniform corona;

If
an
effort is
made
to apply a pressure
between the
elec-
trodes that will be high enough to break through the glass

before appreciable heating has taken place, one
at the
is
astonished
enormous pressure that can be applied

plate
for
a mere instant
it
to
of glass of this thickness.
In
fact,
is
quite
impossible, without the use of a very large plate, to apply

sufficient
pressure
to
break
down
this
thickness
of
glass
immediately.
That the corona streamers, when the pressure

enough
to start them,
will travel
is
great
long distances, and carry
thereby the electric pressure over large areas of the glass or

other dielectric,
in Fig.
foil
is
well brought out in the experiment illustrated

is
115.
There a .i-inch glass plate

side,
coated with tinis
on one
and on the other a thin wire
wound back
and occa-
and
forth at intervals that average ij inches.
Here and there
the intervals
amount
to double this, or 3 inches,
38]
THE ELECTROSTATIC
is
FIELD.
183
sionally the interval
but ^ inch.
An
alternating pressure of
17,500 effective volts

tin-foil
coating.
applied between the wire and the Corona and streamers will form over the
is
now
entire surface of the plate in a
most
prolific fashion.
loud
complex note
about
is
emitted, and the
variegated fashion great brilliancy.

.
phenomenon presents in The current used will be

In
ampere
at a
power
factor of about 50 per cent.
GLASS PLATE.
.ÎRE
CEMENTED TO FRONT SIDE
Fig.
115.
this
form of apparatus the streamers, though very intense,

from one point and then from another.
will
strike first
is
The
plate
not excessively heated at one spot and the operation can be
continued for a half-hour without danger of puncturing.
Thus
one
is
enabled to determine the curves of e.m.f and current.
Particularly from a study of the curves of current that flow to
supply the streamers and the corona, one finds that these
streamers form
a capacity and air-rupturing phenomenon.
When
the dielectric flux density at the edges of the electrode,

little strata
through the
of air that exist there, attains a value
that will rupture the
air,
such rupture promptly occurs and

in the air is re-
the alternating dielectric flux at this point
placed by an electric current conducted by the arcing phe-
84
ELECTRICAL M/tCHINERY.
[38
is
nomenon.
tended.
density
is
Thus
in effect the
edge of the electrode
ex-
On
the edge of this arcing current the dielectric flux
raised also to the rupturing point,

fall
and the corona

it
extends further until the
of potential through
is
so lowers
the flux density that the air beyond
no longer ruptured.
The temperature coefficient of the corona as a conductor is As the corona broadens, the least resistance for the passage of current to its outlying districts will occur by the concentration of current along narrow paths that become
negative.
highly heated and very conductive.

In the light of this behavior one should study the phe-
nomena of corona and streamers on the plate in Fig. 115. The bright streamers are drawn in black and the non-streamer
or
even corona
is
drawn
in
broken-line
hatching.
is
It
is
observed that where the space to be covered

half inch or thereabouts, no streamers form.
is
narrow, onethe space
Where
somewhat wider even corona forms along the wires outward
to a certain distance.
The space along
the middle
is
supplied
with a streamer that starts from one point on the wire on one
side of the space.
This streamer darts to the center of the
space and there divides and extends either

entire length of the middle space.
way along
the
At
the pressure of 17,500
volts here used these streamers will easily extend a foot or
more.
At
a space that
will
is
still
wider, about three inches, the

is
middle portion
be supplied with even corona that

side
fed
by
two streamers, one from the wire on one

from the wire on the other
they are much
side.
and the other

it is
In
all
cases
clear that
better conducting avenues of current for supplyitself.
ing the even corona than the corona
They
obtain this
better conductivity through the heat generated by the current
they conduct, owing to the negative temperature coefficient of

the highly heated air as a conductor.
That the streamers form
easily
when they have
their origin
38J
THE ELECTROSTATIC
FIELD.
185
from the surface of a metallic electrode, and that they form

with great difficulty
dielectric,
is
when such electrode is covered with a brought out by the experiments with the wine-
glasses in Figs. 116 and 117 and the glass. tubes in Figs. 118
and
19.
The
shown
wine-glasses are plain in pattern and
differ in
size so
that one
may
be set into the other with bottoms touching as

Pressure between the electrodes
is
in Fig. 116.
turned
Fig. 116.
Fig.
117.
on gradually up
to 30,000 volts.
First the violet blue
even
corona appears, then after that broadens, the streamers appear and develop in length and intensity until at 30,000 volts they
a short circuit, and leap over the edges of the glasses, forming The opening of the circuit-breaker. all action ceases on the
corona and streamers are drawn, using the same convention as

before.
The
glasses are
now remounted,
as
shown
in
Fig.
117,
86
[38
and the pressure of 30,000 volts again turned on.

corona only
to
is
Even
now
observed.
The
pressure
may be
elevated
40,000 and 50,000 volts without the formation of streamers.

is
There
no change
in thickness of the glass dielectric in this
experiment.
The amount
in the
of even
corona
in
the
second
arrangement
immediate neighborhood of the electrodes

first
remains the same as in the
arrangement.
âsja
A
Fig. 118.
In Fig. 118 a 12-inch glass tube having a bore of
|-
inch
and a thickness of aluminum wire that

it.
walls of about y^ ii^ch

is
is
slipped over an
just large
enough
to pass easily through

is
At
the middle of this tube and at right angles thereto
laid
an aluminum wire, the dimensions of which are the same

first.
as for the
pressure between the wires
is
now
turned
on and slowly increased to 20,000 or 25,000

tube the same as the
volts.
A brilliant
display of easily formed streamers will occur.

first is
Another glass
applied between
is
now shpped
over the bare aluminum

is
wire and a pressure of 40,000 to 50,000 volts

the wires.
A brilliant broad corona forms
that
free
from
all
regularly occurring streamers.
Occasionally a faint streamer
38]
will
THE ELECTROST/ITIC
FIELD.
187
run over the surface of one tube or another, in no wise presenting the rich density that the streamers possess in the
earlier experiments.
Fig. 119.
In the next experiment the tubes of Fig.
19,
which have
of
been described above toward the close of the discussion
Fig. 118, are separated ij inches and a pressure of 60,000

is applied. There is now witnessed a most beautiful homogeneous corona of strong ultra-violet light. It emits a loud note and rapidly but uniformly heats up the tubes. b. Testing Dielectrics for Break-down Gradient. Owing
volts
to the intensely localized heating effects of corona streamers
and the consequent

correct
local
destruction
of the
dielectric,
no
break-down
test
can be made when they are permitted

heats so rapidly that a
to form.
test
is
The homogeneous corona

in a
must be concluded
very few seconds when such corona

form.
present even in a
weak
Every
effort
must be made
to eliminate the corona as completely as possible
when making
these
tests.
When
the samples are of such a character that they
may
Fig. 120.
be melted and cast into any desired shape, the form of testpiece given in Fig. 120
is
excellent.
rectangular slab-
of.
the dielectric
is
[38
cast about a metal ball as shown.
The
ball
the electrodes.
and the metal plate on the under side of the slab will provide The distance between the face of the ball
nearest the plate and the plate will be the thickness of dielectric
punctured.
The
lateral thick construction of the
sample
cuts off all corona.
When
the dielectric comes in the forrh of thin sheets, such
as paraffined
and linseed-oiled paper, the
test
samples
may
be
Fig. 123.
made up
mounted
in pairs of tubes
and slipped over thin metal rods

Glass in the form of tubes
at right angles
1 1
and broken quickly by pressure applied

8.
as described in Fig.
tested properly
may
be
by
this
is
method.
easily arranged.
liquid
sample
The
to
electrode rods,
terminated with balls i inch in diameter, are separated by

a distance through which the liquid
is
be broken.
to surround
In a
convenient vessel the liquid
is
then
made
is
the
electrodes and the break-down pressure
applied.
39]
THE ELECTROSTATIC
FIELD.
189
Where
a strong dielectric sample can be obtained only in

it
the form of a homogeneous sheet

glass shields as
oil is
is
placed between two
shown
in Fig.
121.
Paraffin or other suitable
applied and retained in place

air
by
capillarity for the pur-
pose of displacing the This

is
and completely avoiding corona.
fairly satisfactory.
If the in the
sample must be tested as a plate and cannot be tested

in
manner described
Fig. 121, the
method shown
is
in
Fig. 122
may
be used.
The
inevitable corona
kept from
concentrating into streamers by the application of the metal
guard-rings connected in series by means of condensers as
shown.
The
last
testing of samples
by
this
method
is
tedious.
The
part in
method
is
of practical use either in whole or in

is
many
other ways where dry insulation
used for
building high-pressure apparatus.
TAe guard-ring or conductor variously applied

means for modifying
objectionable features.
c.
is
a reliable
the corona
and remo-ving many of its most
tion
Break-down Test of a Lead-covered Cable. The insulaand lead cover are removed as shown in Fig. 123. To
, /<8BEST0BN
"''"iSSfcli;;^
Lead COVERING
Fig. 123.
avoid corona destruction, asbestos wool
is
applied in a bunch
over the ends as shown.
Asbestos slightly conducts; thus

the asbestos will
in
corona
carry.
is
displaced by the current which

this
In
at
arrangement cables
will,
general,
break
some point under the lead covering instead of at the ends, as will always occur when the corona is permitted. In 39. Dielectric Thickness Required to Avoid Corona.
down
designing machinery and engineering appliances
in
which high
19 pressures
are
employed,
insulators
[39
must
be
selected
and
arranged in such a manner as to avoid corona almost entirely.
The corona must be avoided

{a)
It is
for
two reasons.
highly destructive.
When
occurring even to a
very moderate degree the insulators or dielectrics over which

it
is
allowed to play
suffer rapid deterioration
and soon give
way.
Even
the most durable insulators will not last indefinitely
under the action of a moderately strong corona.

{b)
The corona when allowed
to form constitutes a serious

its
waste of power.
formation in
Through economical requirements alone
connection with high-pressure
machinery and
be better
appliances must largely be eliminated.
The
best
method
for eliminating the
corona
will
understood
after
the corona as
some study has been made of the character of brought out by the above experiments and their
media or materials as given
seen that thin
significance in connection therewith of the varying dielectric
properties of the various insulating

in the tables
on pages 179 and 180.

it
From
the table of rupturing gradients
is
films of air are ruptured only with
much
greater pressures pro-
portionally than
are
greater thickne.sses.
For example,
it
requires the application of pressure at the rate of 250,000 volts
per inch to rupture an air-film .05 inch thick and but 15,000
volts per inch to rupture air-thickness greater than
5
inches.
This property of a rupturing gradient that varies with the

thickness belongs in varying degree to the gases only.
It is
supposed to be due to the
fact that metals or
conducting elec-
trodes condense on their surfaces dense layers of the gas in
which they are immersed.
It
has been found by experiment
that the greater the density and, therefore, the greater the
pressure under which a gas exists,

pressure per unit thickness that
this
it
the greater the electric

it.
is
required to rupture
From
air in
follows that
if
electric
conductors do condense the
39]
THE ELECTROSTATIC
FIELD.
191
dense films over
their surfaces, such films will
be
much more
difficult to rupture
electrically than will
be the case relatively
for
much
greater thicknesses of air in bulk.
study of the rupturing gradients of other substances

this table
given in
shows that
oils
rupture at from about 50,000
to 150,000 volts per inch, and solid dielectrics usually from
200,000 to 500,000 volts per inch, while linseed-oil paper has

the exceptionally high rupturing gradient of 1,000,000 volts per inch.
It
appears, therefore, that even the thinnest films
of
air are
ruptured about as easily as the poorer solid dielecair
trics,
while ordinary thicknesses of
such as are usually em-
ployed between conductors upon which high differences of

pressure are applied, are grossly inferior in rupturing quality to
the average solid
dielectric.
For
air in bulk
and
for glass the

I
rupturing gradients are 15 and 300 kilo-volts, a ratio of 20 to

in favor of glass;
for air
in
bulk and linseed-oil paper the
rupturing gradients are 15 and looo kilo-volts, a ratio of 66

to
I
in favor of the linseed-oil paper.
Some
additional light
is
thrown upon the practical properspecific inductive capacity of
ties of dielectrics applied to withstand high electric pressures
backed up with power when the
^ dielectric and By Section 37^ it was

its
rupturing gradient are
studied together.
seen that dielectric flux equals
D= kE,
from which
it
follows, that since
k and
E at the
point of rup-
ture are constant for

dielectric flux
any given
is
dielectric of unit thickness, the
D at
which a given
dielectric will rupture
is
also
constant.
The
value of k
/&= 2.544
i.e.,
io-"-isr.
(102)
and
the
specific
inductive
capacity are
the
same
expressed
in different units.
'
192
[Sg^
As
the specific inductive capacities and rupturing pressure-
gradients vary widely and independently, a study of the values
of dielectric flux should also be made.
Obviously such a study

different dielec-
must be
trics
practically useful
'
where two or more

'
have to be mounted
'
in series
'
to withstand the pressure
between two conductors.

others in the series.
In such a case each dielectric of
necessity must be subjected to the
same
dielectric flux as the
Under such circumstances the dielectric
insulator that breaks at the lowest dielectric flux density will
be the
first
to give
way even though
individually
it
may have
the highest rupturing pressure-gradient.
The
following table gives the rupturing values of dielectric
flux for those substances that occur in both of the
above tables
::
TABLE GIVING RUPTURING VALUES OF DIELECTRIC FLUX DENSITIES IN THE COULOMB-VOLT-INCH SYSTEM OF UNITS.
Air Air Air
less
than .05 inch

I
560,000
X
'
io~^*
.5 to
I
inch
63,000 "
56,000
to 2 inches
5
Air 2 to
inches...
Air
to 10 inches
42,000 " 34,000 "

3,000,000 "
. .
Glass
Linseed-oil paper,* assumed.
Paraffin
5,000;000 "
Petroleum
Turpentine
900,000 " 200,000 "

800,000 "
air as
A
* It
study of this table reveals the fact that while
dielectric has
many excellent
properties,
it is
unfortunately very
is
unfortunate that the specific inductive capacities of
many
of the most
important insulators used in practice are as yet
unknown
which
in the literature.
In
assuming 5,000,000
10-"
as the dielectric flux at
linseed-oil
paper breaks
its
down
the rupturing pressure-gradient
was taken
at 1,000,000 volts
and
specific
inductive capacity was assumed to be 2.25, for which the value of k is 5.05
x io-^'
39]
THE ELECTROSTATIC
in this respect.
FIELD.
it
193
weak
at
Except
far
for the thinnest films
breaks
dielectric
flux
densities
below those
at
which other
can
common
insulators break.
Except where a
air
liquid insulator
be used to surround high-pressure conductors and thus completely exclude the

air,
always
occurs
as an
insulator
dielectrics
inserted in series with other and
more powerful
arranged between the conductors of a high-pressure electric

circuit.
It is
this unfortunate characteristic of air that gives rise to
the
phenomenon
called corona.
is
In Fig. 113 the flux set up

is
through the glass plate
most dense over the area that
under and directly between the electrodes which face the plate
and are pressed against the same.

trodes facing the glass the flux
their surfaces.
is
At
the edges of the elecit is
nearly as great as
within
Under the surfaces of the electrodes, between them and the glass plate, a very thin film of air exists, broken here and there where the electrodes actually touch the glass.
Surrounding the immediate edges of the electrodes and separating them and the neighboring surface of the glass plate there
are air-films also that are
somewhat
thicker.
Beyond the edges

grows
less.
of the electrodes the flux density rapidly
In a case like this the flux diminishes with the

electrode
air as
distance from the edges of the
more
its
rapidly than
does the dielectric strength of the

increases.
bulk-thickness
is
The
result in the case of Fig. 113
that as the
dielectric flux
between the electrodes increases with increase

is
of applied pressure a point

to
reached where the flux amounts

is
560,000
10-'',
which
great enough to rupture the
air-films
under and immediately surrounding the edges of
the electrodes and the

occurs.
commencement
of corona
formation
The limits of this text do not permit analysis of the corona phenomena further than that which is necessary to account
194
quantitatively for
its
[39
origin.
In the above example of the
quantitative formation of corona given in connection with Fig.
113
it
was stated that the
dielectric
flux
diminishes more
rapidly through the air-space beyond the edges of the elec-
trodes than does the dielectric-flux-density rupturing value for
the corresponding air-thicknesses encountered.

for the fact that
This accounts
while air-films that reside on the faces of the
electrodes
separated
by powerful
dielectrics
have a much
first
higher rupturing gradient than air in bulk, yet they are the
portions of the air to rupture and they are always the seat of
initial
corona formation.
all
These same circumstances occur

is
in
almost
cases where a powerful dielectric

is
used to insulate
pressure.
two conductors between which

be mounted near or
in effect
applied
a high
This frequently occurs in practice when the conductors must

near together in the high-pressure
expression for the thickness of
machinery or apparatus.
a powerful dielectric that
An
is
required to withstand a given presis,
sure without the formation of corona

follows
therefore, derived as
The
dielectric
flux
at
no point between the conductors
considered as electrodes must exceed the value above which

thin air will be broken, causing the formation of corona.
It
was seen above that pronounced corona would appear

560,000
at
lO"''' units
of dielectric flux density in the coulombthirds of this value the corona

safe dielectric flux density at
volt-inch system.
will
be practically absent.
will
At about two The
which corona
be absent and the qualities of the insulating

its
material will not be impaired on

io~''-
account
is
400,000
This value in any given case must be equal to
kE
-J-
400,000
lo-is,
where
/ is the thickness of
the dielectric separating the con-
4]
THE ELECTROSTATIC
FIELD.
195
ductor electrodes, and k and
are the dielectric flux constant of the insulator employed and the applied pressure respectively.
Tke thickness that a powerful insulator must have for pracin withstanding a high pressure backed up with power is, therefore,
tical duty
kE
400,000
X
It
10-'*
Ci3j
40. Dielectric Hysteresis

that energy in definite amounts
is
was found
in
Section ijc
required to establish a field

is
of dielectric flux.
When
the field of flux
destroyed the
energy
energy
is
is
restored for the most part to the electric energy

originally.
source that estabhshed the dielectric flux

is
This
it
never completely returned to the source; some of
invariably lost as heat in the

is
medium where
the field of
dielectric flux
formed.
The
has
character of this loss bears a
striking analogy to that sustained through magnetic hysteresis,
for
which reason
yet very httle
it
been called dielectric hys-
teresis.
As
ence.
is
definitely
known
of the character of
dielectric hysteresis
beyond the mere

certain,
fact of its actual exist-
One
thing
is
however, that the quantities of

insignifi-
this loss ordinarily to
be met with are comparatively

In
all
cant and usually negligible.
practice,
as
shown
be
in
Section 39, insulators must be so selected and proportioned

that the dielectric flux set up through
cient
to
them
flux
will not
suffi-
cause
corona.
When
the
and therefore the
molecular electric strain of the insulating medium are limited

to that extent,
is
it is
found that quantitatively dielectric hysteresis
practically negligible.
For e'kample, Steinmetz * has found

Charles Proteus Steinmetz.
p. 1065,
* Some Notes on Dielectric Losses. World and Engineer, Volume XXXVII,
Electrical
June
22, 1901.
196
[41
that the efficiency of commercial tin-foil-paraffitied-paper condensers, manufactured
by the General
Electric
Company,
is
99.5 per cent.

these condensers
at
This means that of the energy supplied to
when operated on
alternating-current circuits
normal pressure 99. 5 per cent is returned to the circuit by the condenser. A condenser is simply an aggregation of
dielectric
sheets faced with
electrodes
that
deliver electric
pressure which causes dielectric flux to be established through
the sheets or plates.

of the
Mr. Steinmetz, from a close observation

this dielectric
it
phenomena connected with

is
loss of
per
cent,
led to believe that
most of
is
due to the formation
of a
weak corona,
as
may
be seen from the closing para-
graph of
follows
'
'
his paper here referred to,
and which
is
quoted as
As
I
regards the existence of a true dielectric hysteresis,

believe such a
while
phenomenon
is
exists,
am
air
under the
impression that at least a very large part of the observed loss

is
not due to hysteresis, but
due to traces of
which are
still
occluded in the dielectric of the condenser; and caused

of the air-molecules under the influ-
by the mechanical motion

ence of the alternating
It is
electric stress."
important, therefore, to keep in mind the existence of
dielectric hysteresis,
though under most circumstances that

energy occasioned through
this
arise in practice the loss of
phenomenon
tors.
is
of no consequence.
41. Dielectric Conduction.
There
are no perfect insula-
This
is
the
same thing
as saying that all dielectrics are
conductors to some appreciable extent.
This conduction of dielectrics has a twofold character

()
Current
may
flow over the surfaces of the dielectric

to
which invariably absorb moisture
some extent and
collect
dust consisting of conducting bodies.
4i]
{b) It is
THE ELECTROSTATIC
FIELD.
197
found to be a fact that electric pressure will always
pass small amounts of current through the bulk cross-section
of any insulator or dielectric.
The
physical characteristics of the
phenomena
of dielectric
surface current leakage occurring as described above in (a) are

easily understood.
At
the present time the nature of actual

is
dielectric conduction as described in {b)
not well understood
beyond the quantitative

the dielectric as such
facts.
It
has long since been

{b), is
that true dielectric conduction, class

is
apt to
known be least when

reason for
most powerful.
For
this
many
a
years past
it
has been customary to rate the strength of

its
dielectric in
terms of
insulatioii resistance corresponding
to given dimensions of the dielectric.
In
modern times
is
it
has
come
to be understood that insulation resistance
not a reliable
dielectrics as
factor with
insulators.
k,
which to rate the practical worth of

It
has been found that the
dielectric flux constant,
and the rupturing gradient, G, are much more rational and trustworthy factors with which to gauge the insulating strength
of a given dielectric.
To
familiarize the reader with the extent to
which ordinary
insulators are conductive the following table of insulators
and
* their corresponding insulation specific resistances are given :

I"="l^t''-
per c m^cube.
Temperature.
Mica
Gutta percha
84
449
28 34
Hard rubber
Paraffin
Glass,
flint
16,700
Porcelain
54
X X X X X X
iQi^
lo*^
20
24
lo'^
lo'^ 10'' ioi=
46 46"
o" 0
Kennelly.
* These values were compiled from the original sources by Houston and A more complete table may be found in Foster's "Electrical Engineer's
Pocket-book."
198
42.
[42
Problems on the
Electrostatic Field.
Prob.
62.
Determine the capacity,
in microfarads,
is
i
.
of a glass plate con-
denser wherein the plate

of the opposite sides with
Solution:
inch thick and coated on each
square foot of tin -foil.
By
definition, Sec. jd,
and by Sees. i6c and
37<:,
capacity in farads equals
r
where
^
through the
dielectric.
<P^ is the total dielectric flux
By
equation (loi), which
is
written for unit thickness,
where
section.
/ is
the thickness of the dielectric and
its
cross-
It follows that the capacity in microfarads equals
k--j-ioK
Substituting the dimensions given in the problem and the
value of ^ for old hard glass given in the table on page 179,
each in the inch-coulomb system,
<;=
15.6
X IO-15- .io,
Ans.
C
Prob. 63. It
is
.002246 microfarad.
desired to use vulcanized rubber insulated^
lead-covered cable for the electrical connections between a

transmission line and the
transformers,
is
switches,
volts.
etc.,
of a
power
the
service.
The
line pressure
20,000
minimum
Solution:
permissible thickness that the rubber
What is. may have

?
Assume
by
that a single thickness of the rubber
may have
ness
is
to endure the full 20,000 volts.
The
required thick-
obtained
substitution in equation (103):

Ti 10-"
/=
kE
400,000
20,000
400,000
-3
inch.
Ans.
Prob. 64.
What
thickness of vulcanized rubber must

if
be
used in the above problem
the total pressure

.5
is
raised ta
33,000 volts
.?
inch.
Ans.
CHAPTER
X.
LOSSES IN ELECTRIC CIRCUITS.

SYNOPSIS.
43. Sources of circuit losses. 44. Resistance.
a.
b.
c.
Composition of resistance.
The square mil and

Specific resistance.
the circular mil.
d.
e.
Temperature and resistance. Convenient formulse.
f. Problems. 45. Inductance.

a.
6.
c.
Elements of inductance.
Inductance of transmission
lines.
Inductance of electric circuit linked with magnetic
circuit.
46. Skin effect in conductors.

47.
48.
Eddy Eddy
current losses in conductors.

current losses in magnetic circuits.
u.
b.
c.
Nature of eddy current
losses in
magnetic
circuits.
Eddy Eddy
current losses in iron wires. current losses in iron sheets.
49. Capacity of transmission lines

a.
b.
t.
and
cables.
Composition of capacity.
Capacity of overhead transmission

Capacity of underground cables.
dielectric hysteresis
lines,
50.
Magnetic and
and
dielectric conduction.
43. Sources of Circuit Losses.
The
transfer
of energy-
through an
electric circuit involves certain losses

,
which may
be those of current, of e.m.f.

combinations of these.
or of power, or they
may
be
For example,
in the case of a circuit
containing inductance and capacity reactance and negligible

resistance the loss
is
one of pressure.
is,
In a circuit which conthe capacity due to the
tains only distributed capacity, that
presence of the conductors and the insulating medium separating them, there
is
the loss of current used in charging and

199
200
[44
discharging the dielectric which separates the conductors.

all
In
practical cases the transfer of electric
energy
is
accom-
panied by loss of power in one or more of the following ways

1
In heat produced by the passage of the current along
the circuit against resistance.

2.
In heat due to the circulation of eddy currents in the
conductors.
3.
In heat produced
by
dielectric
and magnetic hysteresis
in the 4.
media about the
circuit.
In leakage of current through the insulating
medium
wind-
between the conductors.
These
losses cover those
which are met with
in the
ings of electrical machinery,
such as dynamos, motors, and
transformers, as well as in circuits intended for the transmission of electric
power
to greater or less distances.
The
sources of electrical losses (resistance, inductance, and
capacity) are properties of circuits which have already been
studied from the standpoint of current control.
These must
now be analyzed
affect their
in order to
determine the quantities which

properties, resistance
magnitude.
Of these
may be
called that
which
is
inherent in the conductors, while induct-
ance and distributed capacity are due to the relations of the

wires forming the circuit to each other, depending also
upon
In
the properties of the media separating the
conductors.
its
case capacity
is
inserted in series with the circuit,
reactance
causes a loss of pressure.

44. Resistance.
a.
Composition of Resistance.
The
re-
sistance of a circuit depends
upon the dimensions of the conresistances,

specific
ductors and upon their
specific
being
directly
proportional
to
the
length
and
resistance,
and
is,
inversely proportional to the cross-sectional area.
That
^==~f
(104)
44]

of manufacture, wires are
2ol
For convenience
by the Brown
use.
made
is
of certain
standard sectional areas.
In this country the plan originated

in
&
Sharpe Manufacturing Company
general
In this system,
in
known
as
the B.
&
S.
gauge, wires
range
diameter from No. oooo, having a diameter of .460
inch, to No. 40, having a diameter of .003144 inch.
The
B.
&
S. wire
gauge
is
so constructed that successive diam-
eters of wire form very nearly a geometrical progression.

ratio is
The
about 1.123 and the diameters double almost exactly

size.
very sixth increase in
The
cross-sectional areas of the

size.
wires are almost exactly double every third increase in
No. 10 wire
in this
gauge has a diameter of nearly one-tenth
inch and a resistance of almost exactly one

feet at ordinary temperature.
ohm
per thousand
As
other wire gauges are in
use,
it is
usually necessary to Specify the gauge in connection
Tvith the
number of the wire. The properties of both copper and aluminum wire in accordance with the B. & S. gauge
are given in the appendix.
d.
The Square Mil and the Circular Mil.

it
In specifying
the area of a circular conductor
has become customary to
use a conventional unit area based on the linear unit oi one
thousandth of an inch.
There
is
used as an abbreviation for
one thousandth of an inch the word mil.

Conductors having rectangular cross-sections are estimated
and
specified in terms of a unit square mil.
One
square mil
.001''
square inch.
Conductors having circular cross-sections are estimated
and
specified in terms of a unit circular area
having a diameter
is
of one thousandth of an inch.

called the circular mil.
This circular unit of area
One
circular mil
.001 '
.000,000,785,4 square inch.
202
[44
its
The area of any

diameter whatever
circle is
proportional to the square of
may
be the unit of measurement.

is,
circular
area measured in circular mils
therefore, simply the
square
of the diameter
in mils.
In estimating the sizes of round conductors, the use of the

circular mil eliminates the repeated use of n.
As most
conis
ductors are round rather than i-ectangular, the circular mil
much more
c.
frequently used than the square mil.
Specific Resistance.
is
The
specific resistance
of a conIn
ducting material
the resistance of a unit of volume.

it is
conformity with the c.g.s. system
the resistance between

side.
two opposite faces of a cube one centimetre on a

the c.g.s. unit
is is
While
used to some extent, a more convenient unit

is,
the resistance of one mil- foot, that

in diameter, or
of a cylindrical wire
in area,
one mil
long.
one circular mil
and one foot

of
the resistance
As one
ohms
is
mil-foot
of soft copper has a r'esistance

,
exactly ten
at a temperature of 50.4 F.
of this unit
one easily remembered and applied.
Illustrative Problem.
What
is
the resistance of a
in length at
,000, 000-
circular-mil copper cable
one mile
a temperature
of 50.4
F..?
:
Solution
R=
d.
10
cm.
10
X
,
5280
^ 1,000,000
XI = .0528 ^
ohm.
Ans.
Temperature and Resistance.
Electrical
resistance
varies considerably with temperature, each conducting material
having
its
own
temperature-coefficient, or percentage increase
in resistance per
degree
rise in
temperature.
For many conductors

exactly constant.
perature.
It
this
temperature coefficient
slightly with
is
not
usually varies
the tem-
44]
203
is
For any
particular specimen the variation with temperature
given by the expression

R,
Avhere
i?^ is
R,{i-{-at+6^)
(105)
the resistance of such specimen at the standard
"temperature, and R^ the corresponding resistance at a rise of
temperature of
degrees, and a and d are coefficients.
The
present.
resistance-temperature coefficients of metals and alloys

their
depend upon
physical
condition
and
the
impurities
is
As
all
the term d^ in the above resistance formula

for
always small
tance for
a given specimen,
it is
of
no practical impor-
ordinary engineering requirements.
The
varia-
tions that occur in degree of purity
standard products
practical
make
this
and physical condition of second term useless. The usual

for the temperature-resistance
form of the expression
variation of metals
and alloys
R,
is,
therefore,
R,{i
+ ai)
in
electrical
(106)
The
metals
for
following table gives the resistance data for the three
and two alloys most used

electric currents.
engineering
conducting
Copper and aluminum are

electric pressure
used where conduction with minimum loss of

is
desired,
and
iron,
German
is
silver,
and manganin are used

J'^P?'"^'"i;\
'
where great resistance

Material.
required.
Resistance, Ohms per Mil-foot at Zero Centigrade.
see e"^"^ "'
( '
1061"
'
Copper,
soft,
pure
9.612
16.02
60.
127.
0.004284*
0.0039
0. 0045 3
Aluminum, pure
Iron, soft, pure
German
silver
9.00044
Manganin
* This coefficient
is
291.
important owing to
its
0.00001
extensive
application for the
'determination of the working temperatures in electrical machinery.

-authoritative single-valued coefficient
is
The most
p.
here given.
See Report of Committee,

169
;
Editorial, Electrical
yournal British Institution of Electrical Engineers, January, 1900, World and En^neer, March 17, 1900, Vol. XXXV. Weight in pounds of one cir, mil-foot of copper, .000003029.
p. 389.
204
-;
ELECTRICAL MACHINEHY.
Iron in addition to
its
144-
use as a resistance metal
is
used
extensively as a conductor for one or both sides of the electric

circuit in electric fraction.
German- silver
high
It
is
is
an
alioj
having 60 per cent of copper, 26

It
per cent of zinc, and 14 per cent of nickel.

relatively, .a
possesses
specific
resistance
and a low temperature

for
coefficient.
extensively used
the
construction
of
resistance appliances for current control in electrical engineering.
Manganin
cent nickel.
is
an alloy, invented by Edward Weston, having

zero at 45
84 per cent copper, 12 per cent manganese, and about 4 per

Its
temperature coefficient
coefficient
is
is
C.
;.
below that temperature the

negative; in either case
it
is
positive,
very small.
and above Manganin is used
extensively in the construction of electrical measuring instru-
ments and to some extent

electric pressure
for resistances i-equired to take
up
and adjust currents.*

resistance of the field circuit of
in
Illustrative
Prob lent. -^Ths.
an
electric generator not
engaged
duty
is
100 ohms at the

This generator
field
temperature of the surrounding

after
air,
18 C.
is
being operated for a given time
found to have a
resistance of 109
ohms.
The
field is
wound with copper
wire.
What
ance
?
rise
of temperature causes this increase of field resist-
Solution: a.
The
general case
Rtx
is
^0(1
+ 0.
Combining and reducing,
Many
other metals and alloys are occasionally used in electrical engineering.
Extensive data relating to the conducting properties of such metals and alloys are.
to be found in Foster's Electrical Engineer's
Handbook.
;;
44]
205
Temperature increase
b.
t^
t^.
The
particular case in the
above problem.
Substituting
in equation (107),
t.
2
100 109 5 .004284 X 100
18
109
=:
'
100
^^ ô 40.0 c ^
Temperature increase
z=
= 40.6
t^ t^
18
22.6 C.
Ans.
e.
Convenient Formulm.
These
formulee apply to circular
conductors, for which the following symbols will be used:
area in circular mils
Af, circular mils per ampere;

d,
diameter of conductor over insulation in inches
, e.m.f. consumed by resistance;

/,
/,
current in the conductor
length of conductor in feet

resistance of conductor;
specific resistance in
r,
r^^
,
ohms per
circular mil-foot
w, weight of total conductor in pounds

W(,
,
weight
in
pounds of a
circular mil-foot of conductor^
V, volume of a coil in cubic inches.
From
equation (104),
Ohms
of resistance, r
'^.
.
Circular mils, cross-section,
A = /=
r-A
.
(io8)>
Length of conductor
in feet,
(109);
K.m.f consumed by
resistance, .
r -l-I
-^^^
. .
(iioy

3o6
[44
Length of conductor
in a coil,
/=
V
-r^.
, .
(m)
(112)
Resistance of a
coil,
r=
/=
-7
^^,
3-
Since
A = A.-L
and
w
I-w^.Ai
-r,
by
substitution
Watts lost in a conductor by resistance, Pr^z

/.
az
(^ '3)
Problems
in the Resistance
of Electric Circuits.
Prob.
With
65.
What
size of wire (B.
&
S.
gauge) would be selected to
transmit 1000 H.P. at 6000 volts terminal pressure a dirtance

of 2.5 miles, with 5.5
per cent loss in the line

?
{c)
copper wire
Solution
{F)
With aluminum wire
1000 ^
X X
746
24. 3 amperes.
6000
.055
=
2
330
volts.
() Substituting in equation
(no) and reducing,
_ ~ = =
10.382
2.5
5280
124.3
330
103,300 circular mils.
B.e^^^^No.o
,^.
^ '

17-303
^-r-
^ ^ ~
2.5
5280
124.3
330
B.
a.Tîu^N..
000
= =

Ans.
is
Prob. 66.
What weight
of copper wire
required to trans.'
mit 10 amperes a distance of 250 feet with 2 volts drop
No. 6 being the nearest gauge
size,
the weight
lbs.
is
39.7
Ans.
4S]

Prob. 6^.
207
What
?
is
the loss in watts per mile in a circuit of

in pressure is 15
No. o B.
&
S.
aluminum wire when the drop
volts per mile
130 watts.
Ans.
Prob. 68. If a three-phase circuit requires 75 per cent of the copper needed for a single-phase line, at what pressure can
1000 H.P. be
deliiêred
50 miles from the generator by a'

loss,
three-phase current at 16 per cent
with the use of 31,576

.''
pounds of copper, neglecting reactance
31,200 volts.
netic circuit a m.m.f. of 1000 ampere-turns
coil
Ans.
Prob. 69. In order to produce a given induction in a

is
mag-
required in the
which
is
the source of the m.m.f.

is
The mean
length of a
turn on the coil
2 feet.
What
size of
copper wire must be

coil
?
used
if
100 volts are supplied at the terminals of the

if
What
size
200
volts are supplied
.''
45.
Inductance.
Nos. 27 and 30 respectively.

a.
Ans.
Elements of Inductance.
circuit
it
Inductance
which
produced by
has been defined as that property of a

a changing current produces within
nature as to resist the change.
by
virtue of
an e.m.f. of such a
is
This e.m.f.
field
the increase or decrease of the magnetic

circuit.
enclosed by the
The
process
of calculating
the inductance of any

in induction
circuit consists in
determining the total change

in the current will
which a given change

circuit.
produce within the
From
the total induction per unit current and the
number of
unit rate
turns in the circuit, the voltage corresponding to a
of change of the
current can be
found.
is
If the
permeability of the
rate of
medium
within the circuit

is
constant, the
change of the induction

is
proportional to that of the
current and the inductance
consequently constant.
This
is
the case where the wires are separated by a non-magnetic
medium,
as in a transmission line suspended in
air.
If the
2o8
circuit
is
[45
produces induction in a magnetic body, the inductance
not constant, as the permeability varies irregularly with the
induction.
b.
Inductance of Transmission Lines.
power-transmis-
number of wires suspended in air and separated by a distance great enough to insure good elecsion line usually consists of a
'trical
insulation.
This separation involves the enclosure of

is
such an area that the inductive drop
serious
if
the line
is
number of miles
in length.
The
density of induction about a conductor has been
shown
in c.g.s. units to
be
B=
(see
Sect.
(114)
26,
page
108).
in
The
constant k represents the

is
m.m.f. about a unit current

to 47r.
a straight conductor and
equal
Equation
(i
14) m.ay be rewritten in the form
^'=7.
where ^
is
("5)
a distance not less than the radius of the conductor,
for within the
wire the density of flux decreases from a maxi-
mum
value at the surface to zero at the axis.
From
the
density of flux thus found the total flux which exists about
each centimetre length of conductor

mined.
may be
readily deter-
The
flux
between two normal planes, a centimetre
apart, will cut the conductor included
between the planes when
the current disappears from the circuit and will thus generate
in
it
an e.m.f.
Over an area dr
(see Fig. 124),

is
measured
in
a plane of the conductor axis, the flux

21 dr
d^
(116)
45]

F
209
1cm
dt
CONDUCTOR
rNORMAU
J=LANE8
'
Fig. 124.
The
the
total flux
which
is
included between the conductor and
another forming a part of the same circuit and separated from

first
by a
distance
d is
2/log, -,
^2l
where
r^ is
("7)
the radius of the conductor

125 that
first
It is
evident from Fig.
the flux produced by the
conis,
ductor beyond the second, that

a greater distance from
it
than d,
has no e.m.f.-producing power, for

it
cuts
both
conductors
or destroyed.
when
the
Fig. 125.
field is
made
From
(117)
we may
write
d$
di
d di
(118)
but
d^
dt
=^=4
(119)
2IO
[45
As
a
inductance has been defined as the e.m.f. produced by
unit rate of
change of current
in th
circuit,
if
di/dt in
formula (119) be
made
unity, the expression
becomes
(120)
L=^2
The
d
log,
flux within the conductor can
be readily calculated by
is
the same method used before, remembering that the flux

less as the conductor axis
is
approached.
Assuming a uniform
ductor section,
it
distribution of current over the con-
is
evident that at any cylindrical surface of

is
radius r the density of flux produced
that due to the current

total current in the
enclosed by the cyhnder.

conductor.
Let /be the
Then
r^ ^= X
;
2lr 2/ = T
U-2i\
By
integration,
<?
= 2/
2
f"''
r di
I.
(122)
By reducing to
find that
the effect of unit rate of change of current
we
varies from a value of unity at the axis of the con-
ductor to one of zero at
be taken at one half

ductor
is
The average value may its surface. The complete inductance of the conthe values within and without
it,
obtained
by adding
which gives
in c.g.s. units
^
The
inconvenient
2log.
d +j
I
(123)
is
value of the inductance as derived from this formula

size,
of
so that the mile
is
usually employed as the
measure of length and the millihenry (one-thousandth henry)

as that of inductance.
Inserting the constants for this trans-
formation, (123) becomes
L
where
^
.160931J2
log,
-+
^j,
(124)
is
the inductance in millihenrys per conductor-mile.
45]
The
inductance of each of the two parallel conductors of a

is is
transmission circuit
the same, hence the total inductance of
a single-phase line
Prob. 70.
twice that of a single conductor.

is
What
the
inductance
of
single-phase
transmission line
150 miles long (300 miles of conductor),

?
using wires of .75 inch diameter, suspended 42 inches apart

.479 henry.
Atis.
Formula (124)
is
reliable only
when non-magnetic
is
materials
are used for alternating-current conductors.
If iron or steel
be substituted, the inductance
greatly
increased.
The
average inductance within the conductor must be multiplied by

the permeability of the material to the circular magnetic flux in
the wire, and the constant, which
is
.
5 for
copper or aluminum,
steel.
will be about 150 times as great for iron and in
For
this,
common
c.
with other reasons, magnetic materials are not

lines.
used for alternating-current transmission
Inductance of an Electric Circuit Linked with a Magnetic
Circuit.
^When
the
induction
as
in
loUows a
circuit
definite path,
like that
shown
in Fig.
126,
the
inductance,
with any
particular
current flowing,
may
DC readily calculated by the use

of permeability or magnetization
curves.
By
definition of inductance,
Fig. 126.
Ê (per E
dv
dt
(125)
turn)
d<P
-^,
(total)
= n-y-,
d^
212
[45
or
" dt
dt
d^
di
..
'
di
An
inspection of the magnetization curves shows that

;
/< is
not constant
any given current the inductance must be obtained by determining the corresponding flux from
therefore for
the curve.
Then
(126)
may
be written
L
From
= nj
(127)
equation (83), page 123, in c.g.s. units,
^
Substituting in (127),
^j~Y./i,nnI.
(128)
^=
where
/^
^^7
(129)
corresponds to the particular current at which the

is
inductance
desired.
Prob. 71. Find the inductance of two coils, of 500 turns

each, which surround the closed magnetic circuit
Fig. 126,
when
ampere flows
shown in through the coils and the
constants of the circuit are as follows
Length
of circuit
circuit
= 4 feet
square inches.
Area of
=16
The
curve
is
circuit is
of electrical sheet steel for
which the
B-H
in
given in Fig. 80, page 117.

:
Solution
The m.m.f
of lOOO ampere-turns
is
consumed
feet
of circuit,
the corresponding
ampere-turns per inch
being
1000
X
12
20.8.
46]
213,
The corresponding
65,000.
density in naaxwells per square inch
is
B
ff
Q-^ c.g.s. units)
978.7.
^=
\7t
1000^
X 16 X 6.45 X 10^x48x2.54
978.7
'-4'3 ^^"y^-
Ans.
When
the magnetic circuit linked with a coil

air
is
oir'
is
made up
wholly or partly of
other non-magnetic material, the
reluctance of the path

furnish the necessary
not definite and experience alone will

for calculating the inductance.
judgment
The same
general principles which give satisfactory results in
straight conductors
and
in closed
magnetic
circuits will yield
results in the other cases to the accuracy with
which
will
be accurate in proportion
which the leakage of magnetism between

This
turns and the reluctance of the air-circuit are estimated.
matter
stants.
is
one of design, requiring the use of empirical con-
46. Skin Effect in Conductors.
The
calculations of in-
ductance have been based on the assumption that the current

is
uniformly distributed in the conductors.
This assumption
gives satisfactory results in determining the inductance of the

circuit as a
whole, but the current
is
really not so distributed.
This uneven distribution, while not affecting the inductance of

the circuit seriously, does increase the resistance-drop in
cases to a considerable extent, and the cause of this
distribution of current, as well as
its
some
uneven
effect,
must now be conof concentric cylin-
sidered.
Suppose that the conductor

drical
layers.
is
made up
will
The
external
in the
flux
cut all layers alike,
producing no difference
inductance of the different layers;

flux set
The
itself
outer layer will be cut
by none of the
up within
about the axis of the conductor, while the axis will be cut
by
all flux set
up within the outer layer or surface of the con-
214
ductor.
[46
The inductance
of the
outer layer
of the
con-
ductor will be due to the cutting of the external flux only,

while the inductance at the conductor axis will be due to the
cutting of the flux within and without the conductor.
is
Thus
it
seen that the inductance of a conductor

its
is
greater at
is
its
centre than at
surface.
The
practical result of this
to
cause less current to flow at the centre of the conductor and
more toward
its
surface.
The
reactive e.m.f. at the axis
is
E'
2nfIL
is
271/7(2 log^
h Af
(130)
while that at the surface
^"=2;r//(2log,^J
(131)
is
The
resulting difference of e.m.f.
E'
E" =
2nfl.
(132)
The
and
face.
current in the conductor will so distribute itself that
the e.m.f. consumed in each cylindrical layer will be the same

therefore the current density will be greatest in the layers
,
with the least counter-e.m.f
that
is,
in those nearest the sur-
non.
The name skin effect has been given to this phenomeThe effect is practically to increase the resistance of the
It is
wire by making part of the copper ineffective.
possible
to have such a great counter-e.m.f at the axis of the con-
ductor as to produce a flow of current in the reverse direction

to that at the surface.
actually be less
if
In this case the resistance drop would
the centre of the wire were removed, for the
reason that the circuit of this reverse current must be completed

at the surface of the conductor,
is
and the current density there

of the
skin effect
greater with the useless current present.
The mathematical
calculation
is
too
elaborate to be discussed here, but the following formula will
give an idea of the quantities which affect the increase of
46J
resistance
215
met by an alternating current over that met by a
direct current:
12
U,.iov
78o l:^Ti^9J +'
(133)
where
i2 is
R^. is
the alternating-current resistance
the direct-current resistance
/ is the
length of the conductor in cms.
As
is
a quantity proportional to the area of the conductor,

is
the skin effect
seen to depend upon the area of the conductor

in
it.
and the frequency of the current

skin effect in iron
This formula applies to non-magnetic conductors.
The
and
steel
is
so great as to preclude the use
of these metals for economical alternating-current conductors
except at low periodicities or small sections or both.

should be noted that the increased drop due to skin
It
effect is
an ohmic or resistance one, and

reactance, although the latter
it is is
not a loss of pressure in
present also.
is
The
effect of
the reactance for the present discussion

tion of the current
to alter the distribu-
and hence to increase the alternating-curskin effect
rent resistance.
The
may
be reduced by changing
the form of the conductor to that of a band, or by making use

of several lighter circuits having the
same combined
current-
carrying capacity.
ductors of moderate
At
size,
ordinary frequencies
and with con-
such as are used for alternating-current

is
transmission, the increased drop

in the following
not serious, as
is
illustrated
problem.
Illustrative
Example.
How
much
greater resistance will
be met by an alternating current of a frequency of 130 p.p.s. than by a direct current in a No. 0000 copper wire
.'
Solution
From
the wire table
we
find that
feet.
No. opoo has
a resistance of .04906
ohm
per 1000
After changing
* Formula from Gerard's Lejons sur
I'Electricitd.
2l6
this
[47
length to centimetres, substitution in the formula gives as
the increase
2.14 per cent.
Ans.
With
increases,
larger sizes at this frequency the skin effect rapidly
becoming 17.5 per cent
for
round conductors that
are .7622 inch in diameter and carrying current at 133 cycles
per second.
47. Eddy-current Losses in Conductors.
In
order that
conductors
electrical
may
perform their proper function of transferring

loss, all possible useless
energy without unnecessary

be eliminated.
currents must
Such useless currents heat the

its
conductor, waste energy, and reduce
useful carrying capacity.
When
field,
a conductor
or
is
is
moving through a non-ufiiform magnetic

is
moving
into or out of a iield, there
a difference
in the e.m.f. generated in its different parts.
In Fig. 127 the
'
:.
-.'A.
48]

it
217
as an eddy-
of the machinery in which
occurs
it is
known
Current.
Eddy-current losses
may
be reduced and practically elimiit,
nated by stranding the conductor or by laminating

motion.
the
plane of the sheets being perpendicular to the direction of

(See Fig. 128.)
in small conductors
Eddy-current losses are not important

in
dynamos and motors, but may reach

if
serious proportions in
the large, bar-wound armatures of generators or motors of

large capacity
proper precautions are not taken to prevent
their occurrence.
48. Eddy-current Losses in Magnetic Circuits. a. Nature
of Eddy -current Losses

magnetic flux
will
in
Magnetic
is
Circuits.
Steel,
iron, or
other magnetic material that
permeated by an alternating
flux
form about and at right angles to such

be established.
closed conducting circuits in which e.m.f.s are produced and

in
which eddy currents
will
Such currents are

in circulat-
generally entirely useless, and the
PR
in
power used
heating
the
ing them
material.
is
completely wasted
magnetic
The important
losses of this character occur

it
where
a magnetic circuit has established through

flux
an alternating
by means
of an alternating-current circuit
wound about
the magnetic circuit, or in electrical machinery where a part
of the magnetic circuit

sating flux.
is
subjected to alternating flux or pul-
In the
first
case the electric power in the alternating-current
circuit furnishes the
energy required to circulate the eddies
in
the magnetic thus
circuit with
which
it is
linked
these eddy losses
become a
part of the losses of pressure and

circuit.
is
power
losses
that occur in the electric
In the second case mechanical power
the source from
Mechanical power which the eddy-current losses are taken. thus applied is a portion of the total mechanical power supplied
2l8
ELECTRIC/1L MACHINERY.
[48
to the generator, or in the

ical
motor
it is
a portion of the mechan-
power developed.
Under
these circumstances the eddy-
current losses do not form a part of the losses that occur in a

particular electric circuit.
The magnitude
source.
It
of these losses
is
independent of their supply

fre-
depends upon the alternating flux density,
quency, form of the magnetic material

tion of the flux,
in relation to the direc-
and upon the resistance of the same.

is
No
attempt
made
in engineering to use solid
magnetic
circuits to
accommodate alternating
eddy
losses
flux
because of the mag-
nitude of
occurring therein.
Such
circuits
are
invariably built
are coated in
up of
thin sheets or thin wires. as to insulate
The
surfaces
some manner so
them from one

are formed in
another.
Such insulation
may
be a coating of black oxide,,
japan, or asphaltum.
Thus the only eddies that
magnetic
circuits
under practical conditions are those due to
the thicknesses of the wires or sheets employed in constructing
such magnetic
b.
circuits.
Eddy-current Losses in Iron Wires.

is
The
circular cross-
section of an iron wire
given in Fig. 129.
The eddy-current
first
loss in a cylindrical differential portion
dA
of unit length will

;
be deter-
mined
the total loss per unit length
may
then be obtained by integration,

is
^max.
in
the flux density in the wire

p.
max.
sq.
in.
is
the radius of the wire

is
R^j,
the resistance of an inch-
Fig. 129.
Section of Iron Wire.

e.m.f.
in
cube of the wire.
The
the
circular portion
cj^
dA, Fig.
29,
from
equation (40) equals
E=
(27r/)(-^.^-r5.^..io-),*
* See note in
Appendix,
sec. 3.
4^]
219
wherein,
by
Sec. 7d,
Watts
lost per unit
length of
dA
dw
d{PR)
= d^.
Resistance per unit length of
dA
R=
dw
Watts
27rr,R,
V^^'fB^^^rno-'fdri
^
2nriR^p
lost in wire per unit length:
-'=-i;^
Volume
of wire per unit length
034)
Length of wire per cubic inch of
iron:
Watts
lost per unit-cube of iron wire
w' ^=^'
^=^^5^
Rî,
0^5T
per inch-cube reduced from the mil-foot value given on

is
page 203
'^
60
12
TT
^ 4 microhms at 0.0 C. 3.927

10'
220
[48
The
watts lost through eddies due to alternating magnetic
flux in soft iron wire of
round cross-section
in.
at zero
degrees
C,
neglecting skin
efifect,
per cu.
are therefore
w=
where r
is
.eipBi^^f X
10-",
(136)
the radius of the iron in inches,
B is
the
maximum
Example.
flux density per square inch and.
_/is the frequency in cycles per second.
Illustrative
Find how many watts

I
will
be
lost
through eddies per cubic inch of iron wire where the data are
as follows:
Radius, of wire
inch
Flux density, m.
Frequency
p. sq. in
10,000
100
Substituting in equation (136),
W= .63
When
loo^
10,000^
is
.1'
io"i'*= .63 watt.
Arts.
its
iron wire
used for alternating magnetic cores
diameter should be
made
small enough to limit the loss through
eddy currents
to a certain per cent of the hysteresis loss.

ratio of
Let e be the
eddy to hysteresis
losses thus allowed.
Then, from equations (86) and(i3'6),
r"
=
63/'^Lx.
10-
V
Illustrative
590e77
JB^
y -^max.
(137)
Example.
What must the radius of the wire be

?
to keep the
eddy
losses within 25 per cent of the hysteresis loss
of excellent commercial electrical steel wire operated at a flux
density of 30,000 maxwells per -square inch and 100 cycles
48]
LOSSES
m ELECTRIC
6=
.25,
CIRCUITS.
221
V= .0015, ^w^x. = 30,000,
/=
100.
R^ assumed
the
same
as for pure soft iron.
Substituting these values in equation (137),
=/
r
c.
590 X .25 X .0015 100 X 30,000"

in.,
.0098
or a diameter of .0196
in.
Ans.
Eddy-current Losses in Iron Sheets.

;
Let
be the thick-
ness of the sheet in inches

^
b the width of the sheet in inches.
-222
49]

t,
223
(i
accomplished by dividing 2r by
as given
by equations
37)
and
(140),
2r
UX
1590
^
. .
7=V^^85~='-^33.
49.
(141)
Capacity of Transmission Lines and Cables
(;
Composition of Capacity.
ized, as in
cuit
Capacity
it
in
a circuit
may
be local-
a condenser, or
may be
distributed along a cirIf
and be due to the medium between the conductors.

is
condenser
in
series
is
with
a circuit, e.m.f.
is
consumed.
When
a condenser
in
parallel with a circuit, or
when
the
dielectric
medium between
is
the
conductors
provides a dis-
tributed capacity in parallel with the circuit, a leading current
of low-power factor
drawn from the generator and a

In long insulated
certain
amount of energy
is
wasted
in the resistance of the circuit
due
to such charging current.
and armored
underground cables
nitude.
this current
it
may be
of considerable
mag-
In long air lines
may
be great enough to have an
influence upon the operation of the line circuit.
As
the curit
rent
will
is
practically in leading quadrature with the pressure
have a tendency to neutralize the magnetizing current of

motors, and
similar
transformers, induction
apparatus, and
hence
b.
in
some cases prove
a benefit, rather
than a detriment.
Capacity of Overhead Transmission Lines.
In Fig.
131
C and
which
C"
are two long conductors
mounted
in the air
remote
from other conductors.

of e.m.f.
They
are connected to
some source them

up at
that produces an electric pressure between
sets
up a
field
of dielectric flux.
The
field is set
right angles to the conductors.
Its distribution is
shown by
the curved lines drawn in the figure.
From
point to point
throughout the
field
the flux
is
ever varying in amount and
direction, due to the fact that the cross-section of the space
through which the flux
is
established constantly varies.
The
224
[49
distribution of flux that actually occurs
must be one by which

flux taken
the counter-pressure formed
by the presence of the

is
over any route from conductor to conductor
equal to the
Fig. 131.
applied pressure.
By
definition, Sec. yd,
the capacity between
any
electric
conductors equals
C=
coulombs per
volt.
^=X
where
175^j, is the total electrostatic flux as specified
<'42)
on page
By
equation (loi), which
is
written for a unit cube,
$D=DA
where
section.
kEA
d
is
the thickness of the dielectric and'
its
cross-
Substituting in equation (142),
kA
d(143)
The
values of
and
are composite for the case given in
Fig. 131, so this expression can be applied only to a differen-
49]
tial

portion
225
of the
dielectric
in
which the conductors are
mounted.
The
vertical line
C and
this
C"
locates a neutral plane.
drawn midway between the conductors Between no two points of
plane can there be a difference of pressure.
A conducting
One
be
sheet oi great area
maybe mounted
in the field in this position
without disturbing the flux in amount and distribution.

half of the e.m.f. applied between the conductors,
may
applied between
C and the plane,

,
and the other half between
the plane and

direction.
It is
C" without
altering the field in
amount and
evident that one of these conductors

its
may now
altering
be removed with the result that

the other half, and as
corresponding half of the
electrostatic field will disappear without in
any way
shown
in Fig. 132.
Fig. 132.
Fro. 133-
Mathematical research has shown that where a medium is permeated by a field of flux emanating from a long round
wire
to
a
the
body presenting
effective
it
a plane of the
surface
of indefinite
flux
is
extent,
length
path
of such
double what
is
for the
case in which the same wire sets up
26
'
[49
a flux through the above
medium
to the walls of a cylinder of

is
indefinite length, the radius of
which
the same as the dis-
tance between the wire and plane, with the restriction that the
distance of the wire from the plane must be twenty or
more
times
its
own
diameter.
This has been found to be true for
any character of flux through any medium wherein the pressure taken up is directly proportional to the flux density and the
distance in the direction of the flux.
It follows,
therefore,
that the distance of flow of dielectric flux, magnetic flux, or

electric
current in the
it
medium surrounding
in Fig.
132
and the plane is double that for the same medium and the walls of the cylinder, d and r in Fig. 133 between being the same in both cases.
between
The conductor
are
C in
Fig. 133
and the conducting cylinder
mounted
in the free air
and replace the conductor
and
air,
the conducting plane in Fig. 132, also mounted in the free
with
d and
r the same in each case, for the purpose of deter-
mining the capacity of the conductor

applied between the conductors.
<Pj, is
in
coulombs per volt
the total dielectric flux in coulombs per inch length
of conductor.
D
df
is
is
any point within the field. any portion of d measured from the center of C,
dielectric flux density at
Then
L^
likewise
Therefore
49]
227
Integrating,
r Since capacity in farads
is
C=
-^,
n,
C
For
2nk
-r
(145)
practical purposes
it is
convenient to use the
common
logarithm, the microfarad in lieu of the farad, and the mile for
the unit of length.
For
air
2.244
io~'=-
Reduction factor for miles " " microfarads.... " " " " logjD
63,360
1,000,000
2.3026
Making the above
substitutions
and reductions, the capacity
of a single wire becomes
^=-^'
0388
(H6)*
Tvhere
d is
the distance of the centre of the conductor from the
neutral plane,
and r
is
the radius of the wire.
The
circuit
capacity of two parallel wires belonging to the same

in the free air will
suspended
be ^C, since the capacities

in series.
of each wire to the neutral plane are
This
may
be
seen from another point of view.
current which flows through a

is
conductor to establish
dielectric flux
* The
called a charging current.

is
By
equation (44)
effective length of flux patli
2d.
See page 226.
'
;:
2 28
[49
the charging current due to the capacity of a conductor with

reference to
its
neutral plane
r
is
^=-75^
From symmetry
lish
27rfCE
(^47)
the electromotive force required to estab-
the charging current
/ from
wire to wire
is
2E, and
(148)
27[fC'{2E)
io
where
is
the capacity from one wire to the other.
27t/(2y
from (147),
^
therefore
=
to
10/
27r/E
C'
iC
(149)
Tke capacity from wire

is
wire of a single transmission line .0388/

2
log^-)
where
C is
/
is
the capacity of the line in microfarads the length of the line in miles
half the distance between the lines in inches
d is
r
is
the radius of the line conductor in inches.
Illustrative
Example.
Find
the
charging current in a
transmission line having the following dimensions
Radius of conductor
Distance between centres
1625 inch
24 inches
50 miles
Length of line
E.m.f.
50,000 volts
Frequency
60 cycles
49]

Substituting in equations (147) and (150),
229
=
c.
27r. 60-
50,000
.0388
,
50
\
2A
'^^
amperes.
Ans.
Capacity of Underground Cables.
For
underground
It
service the conductor usually has a circular cross-section.

js
insulated with a cylindrical sheath of strong dielectric over

is
which a lead covering

the lead there
is
drawn
to
keep out moisture.
Over
for
generally applied
an extra sheathing
mechanical protection.
The
value of the capacity given by equation (145) evidently
applies also to this case.
This value reduces to
10* "'",;/" c= 32-772HX
(ISO
where
C
k
is
is
the capacity in microfarads the dielectric flux constant in coulombs per inch-
cube per volt;

/ is the
length of the cable in feet
d is
r
is
inner radius of the lead sheathing
the radius of the conductor.
Illustrative
Example.
What
miles
is
the capacity of a vulcan-
ized-rubber lead-covered cable wherein the radius of the con-
ductor
is
.25 in.
and the radius of the lead sheath .55

is 3
.'
in.
and
the length of the cable
The
value of k for vulcanized rubber

at 6.04
is
given in the table
on page 179
equation (151),
io~^'.
Substituting these values in
6.04
X 10-" X
32.77
io
5280
,
C=
.915 microfarad.
Ans.
23
50.
[SQ
Magnetic and Dielectric Hysteresis and Dielectric
Conduction.
In alternating-current circuits electrostatic and magnetic
fields are
formed about the conductors as the pressure and cur-
rent attain positive values, and
when such
pressure
and current
and mag-
reverse
by passing through zero
their electrostatic
netic fields are destroyed
and then re-formed with an opposite

current.
sign corresponding to the change of sign that has taken place:

in the electric pressure
static or
and
In forming an electrois
a magnetic
circuit.
field
a certain amount of energy

for the
removed
field
from the
Such energy
it
most part
is
returned ta
the circuit from which passes through zero.
was derived each time that the

never
all
It is
returned to the
circuit, as.
some energy
itself
is
always
lost in hysteresis.
This loss impresses,
upon the
electric circuit in the
loss
due to resistance.
same manner as does a The magnitude of magnetic and elec-have been treated
in
trostatic hysteresis losses
Chapters VII;
and IX.
tical
Magnetic hysteresis losses are large enough in prac-
operations to require that they be kept in

as possible under
mind and
limited
as
much
most circumstances.
In engineering
Electrostatic hysteresis from a practical point of view appropriates insignificant
amounts of energy.
it
is
not considered as a factor in the
economical operation
of
machinery and
circuits.
Where
by
powerful dielectrics are used to insulate conductors

is
delivering high-pressure current there
some
loss of
energy
(See
is
direct conduction of current through the dielectric.
Sec. 41.)
The magnitude
of the energy lost in this
manner
so small as not to be a factor in electrical-engineering economies, particularly where the dielectric

respects.
is
satisfactory in other
The corona
is
a gaseous dielectric conduction that
is
quite
it is
wasteful of electrical energy.
As
stated in Chapter IX,
so]
23
not allowed to form over conductors in electrical machinery

that
must be brought near together and insulated by powerful on account of the rapid destruction of the insulation
causes.
dielectrics
that
less
it
The
destructive character of the corona has

itself
opportunity to manifest
lines,
on high-pressure long-dis-
tance transmission
which are composed of round bare

on glass or porcelain insulators
than in the insulation of highetc.
conductors supported
at intervals of about
in the air
1
00
feet,
pressure machinery, transformers,

stances, however, the extent to
to form
is
Under such circumis
which the corona
allowed
closely limited also on account of the loss of energy
that
it
involves.
loss of
by the corona class of dielectric conduction has been examined experimentally and reported upon by eminent engineers.*
power
in high-pressure transmission circuits
The
In the diagram of Fig.
34 are plotted the losses observed

line of
by measurements made by R. D. Mershon upon a conductors, .162 inch in diameter and 11,720 feet in
located at Telluride, Colorado, separated 15, 22, inches and
subjected to
volts.
length,
35,
and 52
in.,
alternating pressure varying from
20,000 to 60,000
in.,
The
curves labelled 15 in., 22
35
and 52 in. show the relation between the watts lost on the line and the alternating pressures when the wires were separated by the corresponding distances 15, 22, 35, and 52
inches.
The form
of the applied pressure
wave approximated
facts
the sine wave.

Careful experimental
that the losses recorded
work by Mershon developed the
by the portions of these curves below
the sudden upward bend are due to a general current leakage over the insulators, and to a very small extent through the at-
* High-voltage Power Transmission, by Chas. F.

Elec. Eng'rs, Vol.
Scott.
Trans.
Am.
Inst.
XV,
p. 531, 1898.
232
[50
mosphere; and that the great losses recorded above the bend
in the curves are
due entirely to corona conduction.

results
A study of these
sion line
is
shows that the corona conduction
of the capacity-charging current of a high-pressure transmis-
very wasteful of electric power.
The
results
show
that wide separation of the conductors
somewhat
increases the
pressure at which the loss begins, but does not alter the rate
of increase of the loss with increase of pressure.
The
critical
curves bring out clearly the fact that there exists a

pressure for each given set of line conditions, at which
a further increase in pressure causes a very

portional loss of power.
much
greater pro-
The
cause
is
obviously as follows
When
allel
an
electric pressure is applied
between a pair of pardielectric
round conductors mounted
in the
atmosphere a
flux will
in
be established
in the
atmosphere about the conductors
amount and
distribution as
shown
in Fig. 131.
is
tric flux
density under these circumstances
The dielecmuch greater in

It
atmospheric zones next to the surfaces of the conductors than

it is
at a greater distance
from the conductors.
follows that
is
when
the electric pressure applied between the conductors
increased, a point will be reached at
which the
dielectric flux
density in the air next to the conductor surface will exceed the
rupturing point and will break through a certain radial distance
in the air.
When
this occurs loss of
power
at
once ensues.
Such
loss occurs
through heat generated by the gaseous con-
ductivity that constitutes the corona.
On
further increase of
is
the pressure, the ruptured zone through which corona
dis-
played rapidly increases in depth with correspondingly rapid

increase in
I.
power consumed.
air
This
is
due to two causes

dielectric rupture or
When
becomes conductive by
of an arc
by the formation
to
by the separation of the connecting

it
terminals of a circuit carrying current,
presents a resistance
the
flow of the
current that diminishes as such current
5]
233
increases.
Thus,
in the case
under consideration, while the

.
diminution of the dielectric flux density with the radial dis-
tance from the centre of the conductor tends to limit the thickness of the corona, the inverse resistance-variation character
of the corona current-conducting zone tends to extend the thickness of such zone and thus counteract the limiting action
of the radial lessening of the flux density.
next to the metallic surface of the conductor, the capacity of the conductor has been increased
2.
After rupture of the
air
by the
virtual increase in the diameter of the conductor that

it
has occurred through the existence about
of the conducting
is
atmospheric zone.
An
increase in capacity
followed
by an
increase in current flow,
which, as stated above, reacts to increase the thickness of the corona or conducting zone.
SOOOh
16
) 25 30 36 40 ALTEEUSlATlNe IjirUOVOUXS.
46
so
GS
60
Fig. 134.
Curves showing watts lost in corona on a 2j-mile experimental highpressure power-transmission line at various pressures and distances between wires of .162 inch diameter. The points marked on the curves near the base were located by equation 154.*
Thus the character of the
variation of pressure applied be-
tween the conductors with respect to the power lost as found by Mershon and given in Fig. 1 34 is accounted for. When
* See foot-note on p. 236.
334
LS'*
the dielectric flux density, D^, at which sufficient rupturing has
occurred in the atmospheric zone next to the conductor surface

to
cause appreciable power-loss
is
known, the following
rela-
tion
must exist between the applied pressure, the diameter of
the conductors and their distances apart:
The
capacity of one wire,
i.e., its
capacity with reference
to the neutral plane (see equations loi, 142, 145,
and
146),
is
C=
The
and the
2;r/^
.0. l^)circumference of the wire
dielectric flux density
immediately outside the surface

is
of the round conductor or wire
^"^ 2nr
Substituting,
^0,
l2dV
('52)
where D^
arated
is
the average flux density about the surface of the
conductor.
As
the conductors on transmission lines are sepforty diameters, the dielectric flux
all sides
is
more than
on
prac-
tically uniform
next to the surface of the conductor,

all sides
and the corona

pressure.
will start
on
at
at practically the
same
The
pressure
which the corona or
dielectric
conduction-loss occurs on a line of conductors of given diameters
and distance apart
is,
therefore,
'"ids
E=
where
> is
y^-^
which appreciable power
(153)
the lowest dielectric flux density in the atmosphere

is lost.
at given barometric pressure at
There
is
no need
for
analyzing this phenomenon further, so
49]
235
as to obtain the subsequent relation between the magnitude of
the power lost and applied pressure, radius of conductors and

distance apart, for the reason that such power increases so
rapidly with the applied pressure that
it
will not
pay under any
circumstances to elevate the pressure to or beyond the point
where
this power-loss begins.
The law given by

pressure that
equation (153) determines the
maximum
more than
may
be applied between a conductor having a

neutral plane, in order that not
given radius and
its
an appreciable
occur.
loss of
power from
dielectric conduction
may
In practice a lower value should be used.

is
For convenience, equation (153)

"Where
'^
rewritten thus:
Emax.
= maximum pressure applied between the conductors = radius of the conductors in inches; s = separation between centres of conductors in irtches; Df^ = lowest dielectric flux at which appreciable power
r
is
wasted, in the coulomb-inch-volt system of
units.
Mershon's
results as given
by the curves
in Fig.
134
may
be
used to test this law as follows:
The 52"
curve shows that appreciable power-waste was
evident at a pressure of 52,000 volts, alternating; the diameter
of the conductor
was .162
inch,
and
their distance apart 15
inches.
Substituting these values in equation 154, the value
of
Dq
is
found to be
A=
2.055
1.71
X
X
39.200
.081
lo&ioi.-^g'i)
'*'
=
* 2.055 + 1.71
190,500
10-^^.
X
is
10"
=
factor
the form
of the pressure
wave
in lieu of 4/2; see Scott,
A.
I.
E. E. Tr.ms., Vol.
XV,
p. 550.
236
LS o
When
15, 22,
this value of
D^
is
applied in equation (154) and the
various distances between the conductors given in Fig. 134,
and 35
inches, together with the values for the radius,
and diameters of the conductors, .081 and .162 inch, are there substituted, the following values of E at which the power waste
begins are obtained:
d
.
^
A. C. e.m.f. at which corona loss begins ^^ determined by equation (154).
mcneo.
IS
42,100
45-IOO
22
35 52
48,900
52,000
The
tion
points on the curves of Fig. 134 corresponding to
these e.m.f.s
occupy positions
rise in
in
each instance
in that por-
where the sudden
the power loss occurs.
The
results of Mershon's measurements, therefore, prove the law
given by equation (154) in regard to the relation of the electric

pressure to the distance between the conductors at which initial
corona loss occurs.*

in conclusion
it
should be noted that, in dealing with very
high pressures
for the
economical transmission of power over

all
long distances, care should be exercised at
points to see
that the outer diameters of the conductors shall be large
enough so
as to avoid all corona-loss.
* Mershon's measurements that determine the curves in Fig. 134 were
made
in
1897-1898 mercury.
at
Ill
high altitude where the barometric pressure is about 20 inches of a recent, 1903, investigation made by one of the authors it was
-Co, at
found that the value of
which corona or atmospheric condaction

;
loss begins,
varies directly as the density of the atmosphere
it
varies, therefore, as barometric

It
pressure and temperature vary the atmospheric density.
was
also found that
Da
less
varies irregularly with variation in the diameter of the conductor for diameters
than one-quarter of an inch, while
it
is
constant for larger diameters.
The
results of this investigation are given in section
4 of the appendix.
appendix. f See section 4 of the
APPENDIX.
238
APPENDIX.
ALUMINUM WIRE TABLE.*

Am.
APPENDIX.
2.
239
The
student should keep clearly in mind the relation of the
ampere-inductor
and
the
ampere-turn.
The
m.m.f.
that
an
ampere-inductor will exert through a magnetic
circuit that sur-
rounds
it is
precisely the
is
same as
that of
an ampere-turn.
The
in-
inductor in any case

circuit
but a part of a complete electric-current

circuit
which surrounds the magnetic

is
once for each
ductor that
enveloped by such magnetic
circuit.
The
student can check the above statement of this fact for

field
himself by noting in connection with any of the cylinder
windings and their magnetic

cludes
circuits that
when
the circuit incircuit,
but one
inductor per individual

circuit
magnetic
the
corresponding
that
electric
has made but one turn around
particular magnetic
circuit,
no matter how many other

taken fhrough in so doing;
inductors the circuit

likewise he
may have been
may
note that where a current circuit passes through

that are surrounded
two or more inductors

circuit,
it
by a
single
magnetic
makes
two or more complete turns around such mag-
netic circuit regardless of the
number
of other inductors sur-
rounded by. other magnetic

circuit
3.
circuits
through which the current
may have been

.E^ve.
completed
= 4 X / X B^^nr,^ X lo^^
total flux cuts the circuit
The
factor
4 indicates that the K

2
times per cycle.

EefiE.
= -Eave. X ~" X
V/ 2
2
-Eeff.
=4X/X5,ax.'rn'X-
X -7- X io-, V2
= 2;r/(^^ri^5âx.Xio-').
4.
In an investigation conducted by one of the authors the

density of the atmosphere surrounding the conductors
following facts were found regarding the variation of the dielectric flux
240
APPENDIX.
of a high alternating pressure hne at which corona or atmospheric
conduction occurs:
The atmosphere about

strength at
that
is
the conductor has the least dielectric
a radial distance from the center of the conductor
a small amount greater than the radius of the conductor.

forces are such as to cause the zone of
The molecular
atmosphere
in direct contact with the conductor to possess greater dielectric
strength than the zones beyond.

diate contact with the conductor
tric flux density, it is
Thus, while the zone in immeis
subjected to a greater dielec-
not the
first to
rupture owing to
its
superior
dielectric strength.
For a conductor having a diameter of one-quarter inch or

greater,
the atmospheric zone that

is
first
ruptures and becomes,
conductive
located at a constant distance of .07 inch from the
surface qf the conductor.
For such conductor,

first
therefore,
is
the
radius of the atmospheric zone that

0} the conductor plus .07 inch.
ruptures
the radius^
For
this
ckss of conductors the
dielectric flux density at which,
the weakest atmospheric zone ruptures remains constant.
For diameters smaller than one-quarter inch the distance from the conductor surface at which the weakest atmospheric zone
occurs
grows
less
as the conductor diameter diminishes;
the
in-
dielectric flux density at
which the weakest zone ruptures
creases as the conductor diameter decreases.
The
table
below gives the corresponding values of
radii of the
reis.
weakest atmospheric zones and the dielectric flux densities

quired to rupture such zones, wherein the following notation
used
r,
radius of conductor
iii
inches;
/, radius of weakest atmospheric zone in inches;
r'r, thickness of weakest atmospheric zone in inches;

Z?o', dielectric- flux
density in coulombs per square inch at
which
the weakest atmospheric zone ruptures;
APPENDIX.
241
Poj dielectric flux density at surface of conductor corresponding

to
Diameter
D'
242
>()'
APPENDIX.
= 170,000X10"^'
at 70 F.
and
29.5 inches barometer,
and
for
which equation (154^) becomes
-Emax.=35o.ooo(''+-o7)logi/yj
(iS4&)
at 70 F.
and
29.5 inches barometer.
Since the value of > varies directly as the density of the
atmosphere, a corresponding factor must be introduced in equations (154)
and (154a) where values
of max. are to be deter-
mined
for barometric pressures
and temperatures other than

F.,
29.5" B.
and 70
at
which the values of
Z>o given in the
above table were deter-
mined.
The
weight of a cubic foot of atmosphere at any temperature

is
and barometric pressure
1^^3253X6
459-2 +
/'
^
wherein
6= height of barometer in inches; in degrees Fahrenheit; i = temperature

responding absolute temperature;
lbs. of
459.2 -Hi, cor-
1.3253= weight in
459.2 cu.
ft.
o air at 0 F. and at
one inch of barometric pressure.
Reduction factor for atmospheric density due to any temperature and barometric pressure:
1-3253^
I-32S3X 29.5
17^946
459.2 +
i!
459.2+70
4S9-2 + <
* Kent's Mechanical Engineer's Pocket-book: Properties of Air.
APPENDIX.
It follows, therefore, that
243;
when equations
(154a) and (154&)
are written so as to include the factor that accounts for changes

in temperature
and barometric pressure they become
I7-94&
'-'mav.
459.2 + i
realize
(s; ,, ,,_ X350jOoo(r+.o7)logio(^yj.
(154&')
In order to
something of the practical meaning of
these laws proper substitutions were
made
in equations
(154)
and (1546)
hne
to determine the diameters of conductors required
to operate without atmospheric conduction loss

at various pressures ranging
on a high-pressure
to
from 50,000
250,000
voltS'
at a separation of 48 inches, a temperature of 70 F., and a baro-
metric pressure of 29.5 inches of mercury.
For diameters
less
than
.25 inch
equation (154) was used and
the corresponding values of r

table
and D^ were taken from the above

to
and substituted therein

loss
determine values of rnax. ^t
which atmospheric
would occur.
The
values of
E^^
for
diameters larger than .25 inch were determined by substitution

in equation (154&), the corresponding values of
r.
In
this
manner a
series of values of Emax.
was determined
radii.
a-S
corresponding to the series of assumed values of conductor
curve was then located using the values of r and max.

the
relation
coordinates, showing
between
the
size
of
the
conductors and
of
maximums
of
pressure-waves at which loss
power by atmospheric conduction would occur.

table
From
this
curve values of r and their corresponding values of Emax. were
taken and placed

given
also
in
the
given below, where there are
the
effective
pressures having
90%
of the values
of the foregoing pressures at which the atmospheric conductive

lo"~
would occur.
The
latter pressures are
about the highest
244
APPENDIX.
that could be used with their corresponding sizes of conductors
without
encountering
the
corona
loss.
The
pressures were
selected so as to cover a range of 50,000 to 250,000 effective sine-
wave
volts.
Maximum Volts
at which
Atmos-
pheric Conduction Loss Occurs.
APPENDIX.
scale
245
amounts
to
rise of
about
7%
at
maximum
of the pressure
wave
and which means that the which corona will first form
by
for a given set of conditions will be lowered
7%
when
the
temperature
is
elevated from 70 to 100 F.

at these very high pressures,
Power can probably be transmitted

250,000
volts,
only on a scale sufficiently large to warrant the expense of building a cover over the Hne so as to keep snow, rain,
and dust away from
and then only after the electrical manufacturing knowledge and art can produce practical transformers
it
for the production of electric
power
in large
amounts
at such
high pressures.
Where
the diameter of the solid or stranded conductor reis
quired to avoid corona
larger than that required to conduct
the current economically the conductor
may be made
tubular.
The
electric pressures considered
above have been
maximum
instantaneous pressures or effective values of sine-wave pressures.
Equations (154), (1540), (1546),

written so as to give
,(iS4a')>
maximum
will
instantaneous values.
and (154^) were These
values divided by
V2
render sine-wave effective values.

line pressure
For
all cases,
however, where the wave form of the

sine
differs
from the
wave the maximum-to-effective
factor
for
the particular
wave under consideration must be known and

in lieu of the
must be used
only
when the form of the
V2, which is practically applicable wave is known to be a very close approxi-
mation
to the sine wave.
The
constants of a high-pressure circuit relating to the in-
ductance and capacity of the hne and transformers and the
armature reaction of the generator are such as

sine pressure
to
modify the
in general
re-
wave
in
most cases materially, thus causing
the maximum-to-effective factor to be different from \/2,
quiring
correspondingly
conductors
is
having
larger
or
smaller
diameters, according as this factor
greater or- less than the

all
V2.
From
equation (154^')
it
is
obvious that in
cases where
^46
electrical
APPENDIX.
circuits
are to be insulated for high pressures
and
where the atmosphere or other gases cannot be entirely displaced
irom the immediate region of such

.dielectrics,
circuits
by sohd or Uquid
as
is
generally the case, the practical value of the
total or composite insulation
must be greatly increased by en-
closing such electric circuit in tight
and strong compartments
capable of sustaining a high pneumatic or hydrostatic pressure according as dry or hquid insulation is employed. Such high
pressure should be applied for the purpose of greatly increasing
the strength of the atmosphere or other gas that
in the presence of the solid
may
be confined
or hquid
dielectric
which constitutes
the
main
insulation.
The compression
s.ilation
of the air or gas in the presence of the in-
will
greatly
improve the practical
dielectric
strength
carried
to resist
of the insulation of such high-pressure circuits until

to a point
it is
where the
dielectric strength of the air

is
and gas
rupture and corona formation

dielectric strength of the
equal to or greater than the

or liquid insulation.
main
solid
At
orJinary temperatures and barometric pressures the alternating

dielectric flux that
an insulation
air or gas
will stand is limited to that at

is
which the included
broken; the heating thereby
ciused, rapidly effects the destruction of the

(See Sec. 39.)
main
insulation.
Mechanical pressure
will increase the dielectric
strength of the air or gas

tical
and correspondingly increase the pracinsulation..
working strength of the main
INDEX.
Addition of Complex Quantities Addition of Sine Curves

Aging, Magnetic
Alternating Current
PACE 62
33> 34
129
Power Quantities and Vectors
gj
65
Quantity, Effective Value of

Quantities, Terminology of
Alternation, Definition of
66
58 58
201 13 106 45 128
American Wire Gauge Ampere, Definition of

Ampere-turn, m.m.f.
of.
Analysis of Periodic Curves, Harmonic Annealing, Effect of upon Hysteresis
Analyzer,
Harmonic
Apparent Power Average Value of Sine Curve

Battery, Nature of
49 97 27
3,
Break-dovm Gradient, Testing
Dielectrics for
187
Test of Lead-covered Cable
189
201
Brown and Sharpe Wire Gauge

Cable, Break-down Test of Lead-covered
Cables, Capacity of
1.
189 223-229
223
72 226
229
17
Capacity, Composition of
e.m.f.
Consumed
in
Capacity of a Condiictor
Cables, Problem in
Dielectric
Line, Problem in
228
Transmission Line and Cables Underground Cables Capacity, Problems Illustrating Unit of
Specific Inductive
223 229
23,
24 17
179
Unit of
247
248
INDEX.
PAGE 117 117
12
Cast Iron, Magnetic Properties of Steel, Magnetic Properties of

C.g.s.
System
Characteristics of Electrostatic Field
Charging a Condenser Current Circuit, Condenser Conductor

Electric Electrolytic
173 176 227
14
14 14
15
.,
Laws
of
Magnetic
114-123
199 113
177
Circuit Losses, Sources of
Magnetic Circuit, Nature of Electrostatic Circuits, Problems in Resistances of Circular Mil Coefl&cients of Hysteresis Loss CoefEcient, Temperature of Conductors
Coil, Definition of
206 201
132 202 112
m.m.f. of
Coils,
112
211
Inductance of
Combination of Harmonic Components Combinations of Sine Curves. . : Comparison of Electric and Magnetic Circuits
. .
56
31
Complex
Quantities, Addition of
123 62
63
Division of
Multiplication of
Complex Quantity, Definition of Components of Irregular Alternating Quantities

Composition of Capacity Inductance
Resistance
63 60-62 43 223 207 200
Condenser, Charging a
Condenser Circuit
Conduction, Dielectric
Conductivity, Definition of
176 14 196-230
80
226
'
Conductor, Capacity of a
Conductor Circuit
Conductor, m.m.f. about a Conductors, Eddy Losses in Formula for Resistance of
^4 i6
216
20S 213 202
Skin Effect in
Specific Resistance of.
."
Table of
Specific Resistances of
203
Conductor-turn, Definition of
Conductors, Problems in Resistance of
202, 204, 206
Corona, Curves of Power Loss in Dielectric Thickness to Avoid
233 189
INDEX.
Corona, Electrostatic Experiments Illustrating Electrostatic
249
PAGE
180-193 180
236 233
3 17 19
Corona Forming Pressure Corona Power Loss, Problem
in
Cox Generator
Coulomb, Definition of Counter e.m.f ., Power consumed by
Constant, Dielectric Flux. Control of Current
178
15
Consumption of e.m.f. Critical Temperature Current, Charging

Current Control Current, Polyphase
in Single Circuits
68
121
227
15
142
of.
Problems Illustrating Unit

Unit of
Curves, Hysteresis
21-23
13
Magnetization Curves of Power Loss in Corona

Cycle, Definition of
125-131 116
233 59 151 144
Cylinder Field, Development of
Model
of
of.
Production
152-163
150 162
142
Cylinder Fields, Applications of Practical Production of Cylinder Magnetic Field Cylinder Rotating Field, Components of
Dielectric, Capacity of
150
17
Dielectric Conduction
Dielectric, Definition of
196^230
Term
Dielectric
Dielectric
Energy Absorbed in Flux Constant Flux Constants. Table of

Hysteresis
175 178
178
179
195-230
175
Nature of Table of Energy Stored in Table of Specific Resistances of Testing for Break-down Gradient. Dielectric Thickness to Avoid Corona Difierence of Phase
Dielectric,
Dielectrics,
179 197 187

189
32 63 58
13 112
Division of
Complex Quantities
Definition of Alternation
Ampere
Coil
Complex Quantity
Conductivity
the Conductor
60-62 80
iii
17
Turn
Coulomb
250
INDEX.
PAGE
Definition of Cycle
59
17
Farad Frequency
Gilbert
Helix.
59 105 112
,
Henry Impedance
Inductor
Joule
16 75 152
19
Maxwell
Microfarad
113 18
16
Ohm
Period
Periodic
Periodicity
59 58 59 153
71 112
Polar Interval Reactance Solenoid Polyphase Vector Volt
142 60
14 18
Watt
Wave
Winding
Density, Magnetic Flux
59 163
13
Density of Magnetic Flux Derived Units
8
i
15
3,
Dynamo, Nature
of
Eddy Loss and Hysteresis Curves Eddy Losses in Conductors

Iron Sheets
Iron^heets, Problems in
Iron Wires Iron Wires, Problems in
Effective Value
131
216
221
222
218 220
217
Magnetic Circuits from Polar Coordinates

of Alternating Quantity
30 66
-.
non-sine Alternating Quantity

Sine Curve
Electrical
:
57 29 18
Horse Power Electrical Steel, Magnetic Qualities of Electrical Units, Problems in use of Electric and Magnetic Circuits, Comparison of
Electric Circuit
116
20
123 14
'.
Electric Circuit, Function of the Electric Circuits, Electric
p2
ig
Power consumed
in
Power
q3
INDEX.
25
PAGE
Electric Pressure Gradient..
Electricity,
179
i
Nature of Elenients of Inductance

Electrolytic Circuit
207
15
2
,
Electro-motive Force, Nature of

Electrostatic Circuit,
Nature of
177
Corona
Field, Characteristics of
180-193
173 198
Problems on Uses of
Flux, Field of
173 174
5
E.m.f.
Consumed
Voltmeter in Capacity
72 74 69 68 82 68
2
Impedance
Inductance
Resistance
E.m.f.Coiistmiption in Multiple Series Circuits
Single Circuits
E.m.f
.,
Means
for
Maintaining
Polyphase Problems Illustrating Unit of

,
142 22
5
E.m.f.
Measurement
E.m.f., Unit of
-
13
2
Energy, Electrical Energy Loss due to Hysteresis Energy, Problem Illustrating Unit of
125
23
Unit of Equivalent Sine
i&
Wave
100.
132-
Ewing's Theory of Magnetism Experiments Illustrating Electrostatic Corona

Factor,
i8a.
Power
Wattless
97 98
17
Farad, Definition of
Field of Electrostatic Flux Flux Densities, Table of Rupturing Dielectric
174 192
13
Flux Density, Magnetic

Flux, Field of Electrostatic Flux, Magnetic Field of Foot, Mil
i74
7
202 205
Formulse for Resistances of Conductors

Fourier's Series
42
33-59"
Frequency, Definition of Function of the Electric Circuit
92
Fundamental Component of Alternating Quantity

Units
43
12
201;
Gauge, Wire
252
INDEX.
PAGE
Gilbert, Definition of
105 179 187 180
Gradient, Electric Pressure Testing Dielectrics for Break-down

Gradients, Table of Dielectric Rupturing
Graphical Conventions for Vectors Representation of Sine Curve Guard-ring, Function of

Hardening, Effect of upon Hysteresis
60 26
1S9
128
49
45 45 49
Harmonic Analyses
Laws
of
of Periodic Curves
Problem Illustrating Harmonic Components of Alternating Quantities

Helix, Definition of
43
112
112
16 18
7
m.m.f. of
Henry, Definition of
Horse Power, Electrical

Hydraulic Model of Magnet
Hysteresis, Coefi&cients
132
Curves
Dielectric
125-131
195-230
128 128 130 128
Effect of Annealing
Effect of Effect of Effect of
upon Hardening upon Impurities upon Physical Treatment upon
Energy Loss due to Magnetic Nature of Magnetic
125
230
125
Impedance, Definition of
e.m.f
.
75
in
Consumed
Consumed
Inductance, Elements of
e.m.f.
in
74 207
6c)
of Coils
of Transmission LineSi
211
208
207
Nature of Problems Illustrating Problems Illustrating Unit of

Unit of
Inductive Capacity Specific
Induction, Magnetic
Inductor, Definition of
211-213
24 16
lyg
g 1^2
,y ,,
In Phase, Product of Sine Curves

Relation
Insulation Resistance
Iron,
jq-
Magnetic Properties of Cast
ny
INDEX.
53
PAGE
Iron Sheets, Iron Wires,
Eddy Losses in Problem in Eddy Losses Eddy Losses in
221
in
22*
218.
Joule, Definition of
19.
Law
of
Ohm
Harmonic
Analjfsis
19.
45.
Laws
of
Magnetit Circuit Magnetic Tension Leakage, Magnetic Problem Illustrating Magnetic Lines, Capacity of Transmission
Inductance of
Losses in Conductors, Iron Sheets,
114-123.
10.
124 137
223. 208.
Eddy Eddy
Eddy
216
22r
Iron Wires, Eddy.
aiS
217
199 233
Magnetic
Circuits,
Sources of Circuit
Loss of Power in Corona, Curves of
Magnet, Field about

Hydraulic Model of
6
7
Magnetic Aging Magnetic and Electric Magnetic Circuit
129
Circuits,
Comparison of
123
113
Laws of Magnetic Circuits, Eddy Losses

Problems in Magnetic
Field, Cylinder
114-123
in
217
135
142-
Pivot
Polyphase Magnetic Fields, Rotating Work done on
142 142
142,
145 107
Magnetic Flux Magnetic Flux Density Magnetic Flux, Problems

Unit of
7
8, 13;
Illustrating Unit of
20
13
Magnetic Hysteresis
230
125
Nature of Magnetic Induction Leakage Problems Illustrating

Magnetic Permeability
Properties of Cast Iron
&
124.
137 113-119 117

117 116
Properties of Cast Steel Properties of Electrical Steel
254
INDEX.
PAGB 113 118 8
9,
Magnetic Reluctance of Unit

Saturation
Tension
Measurement
M.m.f. of Coil M.m.f about a Conductor M.m.f. of Helix the Ampere-turn
.
10
112
106
112
106
112
Solenoid
Magneto-motive Force, Nature of

Unit of
6
105 132 6
Magnetism, Ewing's Theory of Nature of Magnetization Curves Magnets, Permanent Problems in Strength of
Maxwell,' Definition of
116 130 20
13
S
9,
Measurement of e.m.f
Magnetic Tension Mershon's Power Loss Curves
Microfarad, Definition of
Mil, Circular Mil-foot
10
18
233
201
202
201
Mil, Square
Multiple Circuits, Problems in
80
in
consumed Problems in Multiplication of Complex quantities

Multiple-series Circuits, e.m.f.
82
83 63
57
Non-sine Alternating Quantity, Effective value of
Ohm,
Definition of
16
19
Ohm's Law
Operations on Vectors
TPeriod, Definition of
66 59 58
59 130
Periodic, Definition of.'

Periodicity, Definition of
Permanent Magnets
Permeabilityi Effect of Impurities
upon upon
121
121
Physical Treatment
Temperature upon.
Magnetic. .. Matters Affecting
120
113, 119
Problems Illustrating Magnetic
119 135-139
phase Angle
Difference.
;
64
32
Relations of Sine Curves
31
INDEX.
255
PAGB
Pivot Field, Components
of.
Model
of
Production of Irregular Rotating Limited Use of

Pivot Magnet Field Polar Interval, Definition of Polyphase, Definition of
e.m.f.s.
145 143 146

149 150
143
153 142
142
Term
Currents and Fields
Power,
a. c
93 97
e.m.f 19 19
in Resistance
Apparent
Consumed by Counter
Consumption
Electric
in Electric Circuits
19
Factor
Problem Illustrating Problem Illustrating Calculation
of Electric
93 97 99 104 97
iS
Real Unit of With Non-sine e.m.f. and Current Problem in Capacity of Cables Transmission Line Corona Power Loss Eddy Losses in Iron Sheets
Electric
99 229 228
236 222
104
215
Power
Harmonic Analysis Power Factor
the Unit of Energy
in Iron
Skin Effect in Conductors

Illustrating
49 99
23
in
Eddy Losses
Wires
220
135
Problems in Magnetic Circuits

Multiple Series Circuits
Circuits
Series Circuits
83
80
76 202-206 20
198 211-213
20
'.
Resistance of Conductors and Circuits Use of Electrical Units
on
Electrostatic Field
Illustrating
Inductance Magnetic Tension The Unit of Capacity Current

e.m.f
23, 24
21-23
22
Inductance Magnetic Flux

Resistance
24 20
22-24
3'r
Product of Sine Curves

in
Phase
37
256
INDEX.
PAGE
Product of Sine Curves in Quadrature not in Phase nor Quadrature Properties of the Sine Curve
Quadrature, Product of Sine Curves in
38
39 25
38
32
35.
^ .
Quadrature Relation of Sine Curves Sum of the Sine Curves in Quantity of Electricity, Unit of
36
17
Rate of Change of Sine Values

Reactance, Definition of
41
71
Real Power
Quantities
97 62
121 135
113, r22
Recalescence
Reluctance, Problems Illustrating Magnetic
Specific
Unit of Magnetic
Resistance, Composition of
e.m.f.
Consumed
in
113 200 68
197
Insulation
Resistance of Conductors, Problems in

Specific
202-206 202
19 22-23, 24
Resistance,
Power Consumed
Problems
in
Illustrating Unit of.
Resistances of Conductors, Table of Specific
Wires, Formulae for

Resistances, Table of Dielectric Specific
203 205
197
15
Resistance, Unit of
Rotating Cylinder Fields, Components of Field Poljrphase Formulse
Three-phase Formulae Production of Field, Two-phase Formulae Magnetic Fields

Fields, Practical
150 159-161 168 162

163
142, 143
Pivot. Field,
Components
of.
.
r45
of.
Fields, Irregular
Rupturing Dielectric Flux Densities, Table Gradients, Table of Dielectric

Saturation, Magnetic
Series Circuits, Problems in
Series, Fourier's
149 192 180
rrS
Series-multiple Circuits, e.m.f.

Sheets,
Consumed
in
i
76 42 82
22r 27
Eddy Losses
in Iron
Sine Curve, Average Value,

Effective Value of
in Polar Coordinates
29
27
Rectangular Coordinates
Properties of the
26
25
INDEX.
257
PAGE
bine Curves, Addition of
33,
34
î
Combinations of in Quadrature, Product of
Sumof
not in Phase and not in Quadrature, Product of in Phase, Product of
38 35,36
39
37 31
Phase Relations of Product of
Time
as abscissae for
37 26
41 100
for
in
'
Sine Values, Rate of Change of Wave, Equivalent Skin Effect in Conductors
Formula Problem
Solenoid, Definition of
213 215 215
n2
.....--..-.--......-,...-......,,........,,,, 112
ipg 178
m.m.f of
.
Sources of Circuit Losses

Specific Inductive Capacity
Capacities,
Table of
179
113, 122
Reluctance Resistance of Conductors Resistances of Conductors, Table of

Dielectrics,
202
203
'
Table of
Square Mil Steel, Magnetic Qualities of Electrical

Properties of Cast
197 201
116
117 124
33>
Stray Field
Sum
of Sine Curves
in Quadrature
35,
34 36
56
Synthesis of
Harmonic Components
Table of Corona-forming Pressures Dielectric Flux Constants Rupturing Gradients Energy Stored in Dielectrics Rupturing Dielectric Flux Densities Specific Resistances of Conductors
Inductive Capacities
Resistances of Dielectrics
236
179 180
179 192
203 179
'.
Temperature
Coefficient of Conductors
197 202
121
Temperature, Critical
Tension, Magnetic
Problem
Illustrating
8 20 20 S8 132
2
Problems Illustrating Magnetic Terminology of Alternating Quantities Theory of Magnetism

Thermo-pile, Electric Three-phase Cylinder Fields
163-168
2S8
INDEX.
PAGE 14S 26
211
Three-phase Pivot Field
Time
as Abscissae for Sine Curves
Open Circuit Inductance of Transmission Lines, Capacity of Inductance of Problem in Capacity of

Transformer,
225 20S 228

163
Two-phase Cylinder Fields

Pivot Field
146
of
<
Underground Cables, Capacity

Unit of Capacity
229
17
23, 24
Problems Illustrating
Current
13
Problems
e.m.f
e.m.f.,
Illustrating
zi-23
13 22 i&
Problems Illustrating Energy Problem Illustrating Inductance Problems Illustrating Unit of Magnetic Flux Problems Illustrating
Reluctance
23 16 24
13
20
113
105
18.
M.mi
Power
Quantity of Electricity
Resistance
17
Problems Illustrating
Units, Derived
15 22-24
15
Fundamental Probleins in Use of

Vector, Definition of
I2
Electrical
20
60
and
65
Vectors, Alternating Quantities
Graphical Conventions Operations on

Volt, Definition of
,
60 66
14
Voltmeter, Electrostatic
S
18
Watt, Definition of
Wattless Factor
98
59 132
163 201
Wave, Definition of Weber's Hypothesis Winding, Definition of Wire Gauge, American or B. & S Wires, Eddy Losses in Iron Formulae for Resistance of Work done on a Magnetic Field
220
205,
107
Work,
Electrical

A Text Book of Electrical Machinery

Enviado por

Dados do documento

Descrição original:

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

A Text Book of Electrical Machinery

Enviado por

Direitos autorais:

Formatos disponíveis

Qlarnell

ALEXANDER GRAY MEMORIAL

Cornell University Library

3 1924 004 407 536

Cornell University Library

Cornell University Library.

There are no known copyright

the United States on the use of the

ELECTRIC, MAGNETIC, AND ELECTROSTATIC CIRCUITS.

Professor of Electrical Engineering, Leland Stanford Jr. University;

Consulting Electrical Engineer;

ROBERT DDUMMOriD, PRJNTER, NEW YORK

student of electrical engineering comes to the technical

and professional part

and applied mathematics and

to apply this training to the

upon which general engineering

methods are based.

has been found

text for the

the student the working principles mentioned above and to pre-

pare him for reading profitably the literature of his profession.

The book has been

designed as a distinctively engineering

not as a work on physics or apphed mathematics.

has been found desirable to restate in engineering terms

the elementary laws and principles of those sciences which bear

upon the subject

result of experience in teaching electrical engineering

has been found most satisfactory, both in maintaining the

interest of the student to found the treatment

in economizing his time

the laws of the alternating-current

from which the

treatment of continuous-current phe-

by means of a few problems.

has not, however.

been the purpose

a problem book, and the teacher

and student should prepare additional problems

In this part of the work, books of the nature of " Elec-

Problems," by Hooper and Wells, will be found of service.

such a manner that the analyses of the struccharacteristics

authors express obhgations to

literature of the profession.

method of no one author

student to secure most profitably

further training from

and the unhmited sources

when amphfied with

problems, preferably taken from practical cases,

two volumes being de-

signed to provide profitable work for approximately one hundred

of the material in this

the result of several

use as a text for the instruction of classes

Electricity and Magnetism.

Fundamental and Derived Units.

Power Consumption Problems in the Use

Properties of Sine Curves

Vectors Alternating Quantities and Vectors

Laws of the Electric

Magnetomotive Force and the Laws of the Magnetic