eee

LIBRARY
Michigan State
University

 
PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

196 ~c/CIRC/DateDue.p65-p. 14

 
AGA T  MAGITADS

A REPORT

Submitted to the Faculty of the

Michigan agricultural Collese

By

Richard Rey Leon Schroyer

Candidates for the Degree
of

Bachelor of Science

June, 1921
MAC Of SOUT INUS.

Introduction . 2. 2. © © © «© «© © © © 6 2 ew ew
Theory of Magnetism. .« 2 « 2 © « «© © © «© ©
Tne Iiportvsice or wuslity. «©. © «© « «© «© «© «
wne Tisportance of Uniformity . . 2. 2. « «© «©
The Yualities Of VLerhanent Lagnets . ... .
the sairect of an Air Gap oetween Poles...
Aging or Maturing of Mepmets . 2. « « 2. « « -e

Materials of Ientufacture .« « « « «© « e « « «

Analysis and Veat Vreatnents of the Steels

used by some of the more Prominent Manufe
ke S. Ifaenet Steel «. © ~. «© «© «© «© © © © © @
Magneto MacnetS. .« « « « «© © © © © «© «© «© © «
Meagpnet Testin:r 2. « 2 «6 «© «© «© «© © © © © «© « ©

Testing Permanent Magnets by tne use of a

Ilgepnetometer. « « « © « © « © « «© « «© « @

research Work on verrenent Masnets . ... .

LOG424

ConclusionS .« »« « « e « e

e @ @
ctures

16

19

rae

£4

27

26

Oe

57

46
INTRODUCTION.

Although a vast amount of scientific
investigation has been done, and a great deal has beg
written on the subject of permanent megnets, it re-
mains one upon which very little practical, useful in-
formation exists. ‘The reason for this is the fact that
no laws have yet been discovered, fixing any relation
between the qualities of the permanent magnet, and the
Character of the meterials and the conditions of
manufacture, upon which the megnetic qualities of the

magnet depend.

Any one who has the means for forming, herden-
ing, and magnetizing a piece of steel, can Imke a
permanent magnet of some sort, but to renufacture
permanent magnets of uniform high magnetic quality, re-
quires something of the scientist, the chemist and the.
metallurgist, tozether with the requisite special equip-

ment and an organization experienced in this class of work.

The writers will endeavor to cOmpile the best
data available from the more prominent scientists, chemists,
and metallurgists of the day, and in addition do some
research work on the production of magnets under different

heat treatments of the steel.

 
MHEORY OD MAGHEDISI.

It was early suggested that magnetization is
@ molecular phenomenon. ‘The fact thet a magnet mey be
broken into an almost unlimited number of pieces, each
of which remains a magnet with a north and south pole
of its own, points very strongly in this direction.
This view is also confirmed by the fact that a tube of
iron filings, when placed along the poles of a strong
magnet becomes megnetized, and when removed remains

magnetized, until by shaking or jarring it is demagnetized.

More convincing still is the fact that a long
wire of soft iron may be strongly magnetized and by a

slight jar lose nearly all of its magnetism.

Miost striking of all, however, is the fact that
@ piece of fron when raised to red heat loses its magnetic
quality. (roof of the above statements can be found in

any Gen. Physics).

All of these facts lead us to think that
magnetization is not a property of the surface of a body
or of the body as a whole, but of its inermost structure.
The evidence is indeed overwhelming in favor of the view

that magnetism is a molecular or atomic phenomenon.

The second step was made by We E. Weber, who
de

explained ferromagnetism by assuming thet the molecules
Of iron are themselves small permanent magnets, arranged

in no particular order, as indicated by Fig. below.

 

Ew HA
\% OT Ne \ 7

 

 

 

If these molecular magnets were perfectly free
to turn, it is clear that any magnetic force acting upon
them would be sufficient to bring them all into line and
thus produce at once the maximum intensity of magnetization.
In order to avoid this, weber imagined each molecule in its
Original direction. The effect of this restraining force
is represented in the Fig. below. The neagnetic field does
not bring all the tiny tmgnets into perfect psrrallelisn,
but merely brings them more or less into line, as there

shown.

 

ar”
we NO ™
Nr

AA

=...
ie.
—_-—

 

 

 

The third step is to develop this theory so as to
enable one to include in his explanation the facts of
paramagnetism and diamagnetism. This extension of the
theory we owe to Rowland, J. J. Thomson, and Lagevin. A
brief outline of this theory is as follows: It is well

known thet a small circular electric current behsves like
4.

a small magnet. In 1876, 2rofessor Xowland showed by a
justly celebrated and difficult experiment that a re-
volving electric charge produces the same magnetic
effects as an electric current. In 1897, J. J. Thomson
demonstrated the existence of sriell electric charges,
having a mass about 1/1700 that of sa hydrogen aton.

These small charges are negative in sizm gand can be
Obtained from a great variety of sources. They are called
electrons. The present view (1915) cf an atom is that it
is built up, in pert ot least, of a num cr of electrons
in rapid rotation about eae central nucleus of positive
electrification. #ach of tue se electrons moving in its
orbit is equivalent to a minute megnet. This view is
supported by strong experimental evidence, such as that

of the Zeeman effect.

Assuming an atomic structure of this kind,
P. Langevin in France has pointed out the possibility that
these electronic orbits, in any one atom, may be oriented

in a great variety of ways.

In the case of diamagnetic substances, these
Orbits are supposed to be distributed in stich a@ simnetrical
way that they produce no magnetic field outside of the aton,
the diamagnetic atom has no regnetic moment of its own. If,
however, this diamagnetic substance be placed in a mégnetic

field, lines of mepnetic force will be thrust thru any of
De

these orbits which are not exactly parallel to the field,
and as one looks along the lines of force, those electrons
which are revolving in a clockwise sense will be acceler-
ated, those revolving in a counter-clockwise sense will

be retarded; so that, in either event, the magnetic

moment of the atom will be increased in @ sense opposite
to that of the field. In other words, an atom with a
perfectly symmetrical distribution of electronic orbits

when placed in a msgnetic field behaves as a diaragnetic

substance.

In the case of a paramagnetic substance, it is
supposed that the orbits of the electrons are distributed
in such a wa: that they do not exactly cancel the effects,
on one another; that is, these atons have a meegnetic
moment of their owmm, and being little megnets will fall
into alignment, more or less, when acted upon by an
external field. ‘The turning of these tiny nmegnets will
of course, always be in such a sense as to bring the § pole
of the atom to face the N pole of the field-producing magnet.
Atoms which are symmetric will therefore behave in a
paramagnetic waye But superposed upon this effect will be
the diamagnetic effect discussed in the preceeding paragraph.
In most cases however, these diamagnetic effects ere emsll
in comparison with the effect of rotating the orbits. What

sort of resistence these orbits offer to rotation is entirely

unknown.
In the case of ferromesnetic substances, the
experimentel fects of Ewing's model lead one to think
that there are mituel actions between the atome or
molecules. iwhen an external field is enplied to a
ferromagnetic substance, the natural groupings of the
molecules are changed; these elementary magnets fall into
new nositions of equilibrum, one after another and we have
the effects knovn as permanent regnetism, etc. The key
then to ferromagnetism is the attraction and repulsion of

One molecular Imgenet upon the other.

By going one step farther and assuming that dia-~
magnetism is an atomic phenomenon wrile paramagnetism is
@ molecular property, one may explain an important fact
discovered by 2. Curie, namely, that the susceptibility
of paramagnets varies inversely es the absolute temper-
ature. Cn the other hand K. T. Comton end E. A. Trousdale
offer experimental evidence, ootained by X-ray photograbhy,
for thinking that paramagnetism is not a molecular, but an

atomic, phenomenon.
Te
TH IIPCRTaNCS OF QUALI™Y.

In the majority of cases, the purpose for which
permanent magnets sre used is such thet the quality,
accuracy, efficiency, or usefulness of the heachines or
instruments of which they form a part, denend directly
upon the megnetic qualities of the permanent magnets used.
In electrical measuring and recording instruments, the
maintenance of accuracy denends absolutely upon the
retentivity of the mgnets. In the case of nagnetos,
efficiency and successful operation sre both determined by

the quality of the megnets.

The qualities of the permanent megnet, that is,
cheracteristics which determine its quality as a magnet, re-
Main unchanged indefinitely. They are e characteristic of
the particular piece of steel, in its particular state,
just as its hardness, tensile strength or elastic limit.

A great many neople confuse state of magnetization with
magnetic quality, whereas they mean totally different things.
Demarnetizing a megnet changes its state of regnetization,

but dves not alter its quality; a finished piece hes, be-

fore it is mepnetized, tre aqivelities which determine how

good a hagnet it will be when regnetized. The quality of

a fagnet can be affected only by Sone treatment which will

at tne same time affect its physical charscteristics, ©.2h

as by heeting or annealing.
THE IMPORTANCE OF UNIFPORNIDY.

The desig of a rechine or instrument in which
permanent magnets are used, is, of course, based upon
some known or assumed qualities of the nagmets to be used,
and good desig requires that the megnet, in its design,
be ade»nted to the use to be nede of it, and the conditions
under which it is to operate. iwhen this hes bee@m done, it
is necessary that the tegnets be uniform in quality, other-
wise the finished product will very in quality, accuracy

or efficiency by the same amount that the usgnets do.

This uniformity of negnetic quality is esnecially
important where two or more maenets are acting in parallel
On the sane magnetic circuit, because so placing negnets
of unequal strength and retentivity has an effect similar
to that obtained by placing batteries or generators of
different voltage in parallel. The same statenent holds
true for isgnets made up of two or more leaves, sand in
the design and renufacture of compound megnets, it is a
matter of importance to see that the several portions are
so related and of such uniformity that they will operate

properly together.

For the same reason, it follows that maximum
magnetic quality in a menet of considerable cross-section,

can only be odtained by the use of steel of an analysis,
and a treatment, which will insure uniform quality
throughout the cross-section of the megnet. It is this
feature which limits the cross-section which can be
successfully used, and makes necessary the use of steel
of different chemical content, subjected to correspond-
ingly different treatrent, if the best results are to be

obtained in the mgrets of different cross-section.
10.
THE QUALITIES OF PHRUSNHUT MAGNETS.

Generally speaxing, there are two things which
are a measure of the quality of a permanent megnet. They
are:

1. The residual induction or residual magnetic flux.

2. The retentivity or coercive force.

When materials have been riagnetized and the
ragnetizing field is reduced to zero, a certain percentage
of the initial induction, B, is retained. This is called
residual induction and is designated by the symbol B.
Different substances retain magnetism with varying degrees
Of avidity or strength. The strength of field necessary to
reduce B.. to zero, that is, to deregnetize the material, is

called the coercive field and is designeted by Hcoer. or Ho:

While these two qualities or characteristics are
somewhat related, in so far as the conditions which affect
them are concerned, the existence of rexinum residual in-
duction does not necessarily imply that the retentivity is
@ maximum residual induction does not necessarily imply
that the retentivity is a maximum, or vice versa. The
best permanent magnet is the one in which both these
qualities are at the meximum. It requires long experience
in the selection of steel and the utmost skill in the

treatment of it to produce permanent reagnets which are
ll.

uniform in both residual induction and retentivity.

Almost all kinds of steel which can be tempered
will retain a portion of their initial remetisn if
magnetized. There are, however, very few kinds of steel
which raxe relatively good permanent magnets, and the
quality of the megnets mde from any kind of steel depends
upon a great many things. They are:

1. Chemical content.

2. Conditions under which the billets are nade.

3. Annealing.

4. Rolling conditions.

5. Furnace conditions and treatment during the forming
or forging of the meegnets.

6. Conditions under which final heat treatment is done.

7. The mechanical operations performed after final heat
treatment.

8. The process of maturing.

9. Method of magnetizing.

From this, it is easily seen that from the time
the steel is melted until the megnet is finished, there are
conditions and processes which materielly affect the final
quality of the magnet. ‘The matter is still further
complicated by the fact that the effect of one condition
Or process on the quality of the megmet mey depend upon

One or more of the other variable conditions.
lie

For example, the result obtained from different
heat treatments may devend somewhat upon the chemical
analysis. The fact that a change in one may alter the
effect of a change in another condition, necessitates
the keeping of continuous records of all the conditions,
in order tO so control them as to produce a product of

maximum quality and uniformity.

In Fig. "1" are the reenetization curves of
three kinds of iron; these curves show the relation be-~
tween the regnetizing force, or the magnetic fields in
the air surrounding the iron, and the corresponding
magnetic density in the iron itself when acted upon by
the field in the air. These curves show the relative
ease with which these materials are magnetized, but have

nothing to do with their ability to retain magnetism.

They are important nevertheless, because a
permanent magnet is required to drive its magnetic flux
throughout its own length, as well as thru the msgnetic
Gircuit upon which it is operating. It follows therefore
that any condition in manufacture which increases the
magnetic resistance or reluctance of a nmegmet without a
corresponding increase in the retentivity is detrimental

to its quality.

In Fig. "2" are shown the hysteresis curves of

the same three nieces of iron. ‘Tnese curves are obtained
 

eg Nene Gee we

er

Zs =—6‘
Ee | Lee Se
ee
 

ADZTION AWYENLINDIYBOYV N¥DIHOIN

 
 
13.

by megnetizing the iron to a@ knovn maximum magnetic density,
represented by the point"a”"on the curve. The megnetizing
force is then gradi.ally redvced, the tagnetic density

falling along the curve as imicated by the errow.

At the ooint "B" where the curve crogses the
line of zero magnetizing force, the distance O-B re-
presents the strenzth of permenent megnetism. If, now,
the magnetizing force ve reversed in direction and applied
so as to reduce the residual magnetism, the magnetic
density will cont: rue to decrease until it becones zero,

where the eurve crosses the horizontal axis at "Cc".

The dere pnetizing force represented by tne velue
O-C is that force necessary to balance the rmégenetizing
Sorce of the magnet itself; it is the messure of the
difficulty of magmetizing the magnet, or of the power with
which the magnet holds its magmetism. TYnis quality is
known as the retentivity or coercive force. The vertical
distance O-8 is therefore the measure of residual megnetism,
or residual induction, and denotes the strength of the Legnet
in so far as its magnetic flux is concerned, wheress the
distance O-C represents the tenacity with which the imesgnet

holds its magnetism.

Both of these quelities are of extreme importance,
and it cannot be said that one is more importent than the

Other. High magnetic strength is of little value if that
14.

strength is not made stable by high retentivity.
Neither can a megnet ve said to be of high auality

if its retentivity be high and its strength low.
Referring to Hig. "2", it will be noted that the re-
tentivity of hardened cast iron is approximately 75%
of that of hardened tungsten steel, while the residual
density of cast iron is only about 30% of that of a

tungsten magnet.

It should be born in mind, however, that both
the magnetic strength and the retentivity of any
finished regnet depend upon the varying conditions of
manufacture which we have named above. Even with
material of the sane analysis, it is possible to maxe
magnets of the same magnetic strength, but varying .
widely in retentivity. It is also possible by verying
the conditions to take from the sane steel magnets of
the sane retentivity, but varying widely in residual
Magnetism. In fact, it is quite difficult, using the
same material, to keep from making magnets which vary

in one or the other quality.

In tig. "2" the curves shovm are the result of
carrying the magnetization and the demagnetization thru
@ complete cycle, but one-fourth of this cycle is all
that is required in order to determine the qualities of

the menet, as shown in Fig. "3". attention is cslled
\

 

 

 
15.

to the fact that in the foregoing curves the megnetic
density if expressed in the flux or ragnetic lines
per unit area, and tne coercive fvrce or the
retentivity is given per wnit of length; it follows
therefore, that the total ragnetic strength of a
magnet having given residual flux per unit area is
proportional to its cross-sectional area, and the
ability of a negmet of given retentivity to hold its
magnetism is proportional to its length; thet is, the

length of the var from which it is made.
16.
THH EEFCCT OF AN AIR GAP BETWEEN POLES.

In making tests such eas are indicated in the
results plotted in Fig. "2" md Fig. "3", the magnet
under test constitutes the entire ragnetic circuit.
The curves give the result which would be obtained
if the regnet were a closed ring with no air gap be-
tween the poles. In other words, they rerresent the
quality of steel as a magnet, and that along. The
useful flux of a permanent megnet Operating across
an air gap is always less than the flux which the
magnet would give with no gap between its poles, or
when marnetically "short-circuited". On account of
the very high reluctance or megnetic resistance of air
the residual magnetism of a magnet of a given length
is less@med as the air peth between tne poles is in-
creased, and right here comes a very important con-
sideration in the design of apparatus embcdying the

use Of permanent menets.

Referring to Pig. "S", it will be noted that as
the demagnetizing force is increased from 0, the residual
magnetism reduces very siowly at first, and that as the
demagnetizing force is increased the magnetism falls
more and more rapidly tntil finally it comes down along
am almost vertical line. This indicates that for

moderate demagnetizing forces the permanent magnetism
ee Sa SP i es
eee eee ee
Sol. wwant
rc on a =
'

 

BVSTIOD AVYENLINGIBOVY NYOIHDIN
17.

is very stable, but that if the demagnetizing action
be carried beyond a certain walue the flux becomes un-
stable and the mignet will not recover its strength
when the deragnetizing force is removed.

Since the introduction of an air gan between
the poles of a permanent neégnet has the effect of
partially demsgnetizing it, end the extent of demagnet-
ization depends upon the relstive jength of the megnet
and the air gap, it is important thet the ration of
length of ragnet to length of air gap be such that
the vagnet will not be demagnetized by the air gap
to the point where megnetization becomes unstable. In
some types of apparatus the permanent magnets used are
required to operate in Opposition to a coil of wire
carrying an electric current, as in the case of a
magneto, the denagnetizing effect of which is added to
that caused by the air gap, and in designing apparatus
of this character it is important to take into account

the effects of both the air gap and the coil.

If the magnetization in a permanent magnet be
Once reduced to the point where it becomes unsteble, the
magnet will not recover, so that it is important to bear
in mind that the characteristics of the magnet and the
magnetic circuit on which it operates mist be such that
the maximum reduction of flux under any condition of

Operation must not be to a point where the flux becomes
16.

unstavle.

Since the retentivity of a magnet is the meesure
of its stability in opposition to demagnetizing effects,
the reader will readily see that this quality is of
equal, if not greater importance, than the regnetic
strength. It is possible, by the proper selection of
steel and the use of a heat treatrent corresponding
to the cheracteristics of the steel, to produce magnets
of uniformly high retentivity without sacrificing
i.gnet strength. It is also possible by using a heat
treatment adapted to the character of the steel to
obtain maxinum strength without sacrificing its re-

tentivity.

It is essential, for the best results, that the naeg-
net be selected or designed for the magnetic circuit on
which it is to opperate, and for any htegnetic circuit
considered there is a cross-section and length .of
permanent magnet which will give the maximum flux, with
the reqiisite stability, using a minimm amoung of

INagnet steel.
19.

AGING OR TatURING OF MAGI ATS.

In some classes of apvaratus, particularly
electrical indicating, integrating snd recording
instruments, the maintenance of the initial cali-
bration and accuracy of the instrumj™mts, requires
that the regnets be brought to as stable a magnetic
condition as possible before the instrument is cali-
brated. This is called sging or maturing of the magnets.
Bringing a negnet into a condition where its flux wiil

be stable requires two things, viz;

le. A physical treatuent of the steel after quench-
ing, which makes further changes in its
structure, molecular condition or rapnetic
quality improbable.

2- A magnetic treatment of the megnet after megnetize
ation, which will make changes in its state of
magnetization, uncer the conditions of ordinary

oneration improbabl3.

Yren a menet is hardened it, of course, under-
goes very rapid structural changes, and for a time after
quenching there is a change in donditions taking place,
quite rapidly at first and gradually decreasing until
the physical properties become stable. This stability,
of structural state of the steel can be hastened by

physical treatment, which process is celled aging or
 
 

hte ia ere meee PTT a Be Sala ei roi 3) mae h 2) ) be) ) |
20 6

maturing. Cbviously the trentnjmt applied mst be
such as to only age the steel without impairing its

magnetic qualities.

AS hes been already stated, it is pcescible after
magnetizing a magnet to bring it to a state of hegretiz-
ation, in which the totel or useful flux will not be
greatly altered by such delagnetizing action as that to
which the magnet will be subjected in use, provided, of
course, it is properly designed for the magnetic
circuit on which it is used. It should be clearly
understood thet this treatment aoes not alter the
retentivity of the regnet, but simply changes the magnetic
State in such a meamner that the ratio of change in
flux to magnetiving or demagnetizing effects is re-
duced at that part of the titegnetic cycle represented

by the state of residual re gnetization.

In Fig. "3" we have given a curve showing the
relation between residual density end regnetizing and
demagnetizing forces, from the point of rmaximim magcnetia-
ation to complete demagnetization, when the change from
One state to another is gradual and uninterrupted. It
will be noted that upon the application of a demagnetiz-
ing force, the residual induction begins to drop and as
the force is incressed the reduction becomes more rapid,

until a point is reached where the curve is almost
21.

parallel to the flux ordinate.

this means thet with a megnet in this state of
macnetization the flux of the magnet will very along
this curve when the wagnet is subjected to eny exterior
magnetizing or demagnetizing force, such as the action
of a coil on stray fields, etc. In order to rake the
value of the residual flux as stable as possible, it is
neceasary to have the curve as nearly parallel as

possible to the horizontal axis.

In Pig. "5" is shown the curve of a msgnet teken
as it is put through this stabilizing process. The
arrows indicate the direction of megnetization or de-
Magnetization. after the mgnet has been fully magnetiz-
ed, a demagnetizing force is applied which removes a pore
tion of the flux. It is next subjected to a nmegnetizing
force, and then to a demagnetizing effect, which hes the
effect of bringing the curve more nesrly yvarallel to the
base line, although the residusl induction is somewhat
reduced. It is obvious that this treatment should be
applied after the regnetic circuit on which the hagnet

is to operate has been completely assembled.
Le
MATERIALS OF MANUFACTURE.

Permanent magnets can be made of cast iron, of
steel carrying from e3 to 1.5% of carbon, or of alloy
steel containing carbon, together with tungsten,
molybdemum, chromium or vanadium. Cast iron magnets
are used principally in galvonometers and similar
apparatus, where weight is not objectionable and

efficiency is not a matter of importance.
CARBON STEEL MiaGN TS.

For some classes of work, where snsce end weicht
are not of importance, magnets msde of a good grade of
onen hearth carpnon steel are used. By a careful selection
of such steel with reference to its carvon, manganese,
sulphur and phosphorus content, it is possible by proper
heat treatment to produce magnets having a flux or
magnetic strength of about 75% that of tungsten or
chromium steel magnets of the same cross-section, and a
retentivity of about 70% of that of the better grades of

steels just named.
TUNGS®T2N SPuuL.

Por many years prior to the outbreax of the
Huronean war the alloy steel most commonly, in fact

almost universally used in the menufacture of hich grede
£56

ynermanent mmagnets, was one contsining principally cer-
bon, tungsten and chromium, and in which the percent-
age of manganese, sulnhur and phosphorus were kept to
the lowest practical limits. such an alloy, if
properly made in the steel mill and intelligently
processed in the magnet-rmaking plent, will produce
Insgnets of high strength and vermanence. Due to the
very high cost and the inability to obtain tungsten
steel during the war, it was necessary to develop other

alloys for use in the manufacture of permanent megmets.
CHORIUM STL.

By the use of chromium elloys, produced under
correct conditions and treated according to processes
particulsrliy adapted to them, magnets are being produced
wnich, from the standpoint of both strength and retentiv-
ity, are the equal to the tungsten product. It is a
fact that the chromium alloys do not permit of as wide
a range in the manufacturing and the heat treating
conditions as the tungsten steel, but with proper and
uniform conditions the product shows greater uniform-
ity than was the case with the tungsten alloy. That
is, the average results with echromium steel alloys are
in every way equal to those obtained with tungsten steel,
and the prodict does not vary from the averege by nearly

so wide a margin as in the case of the tungsten alloy.
246

ANALYSES AUD HEAT TREATIENTS OF THE STEELS USED BY

SOLS OF GEE MORIS PaOlincan ? MAUUPACTURIRS -

NDIANA STi PACDUCWS COCliranY.

Analysis.
Tungsten steel Chromium steel
Carbon 0.65 Carbon 0.65
Tungsten 5.75 Chromium 2.4)

Heat Treatment.

The steel is nerdened at the correct heat, which
varies according to the size, from 1480 to 1550 degrees
Fe, and quenched in water or a gcolution so that the

material is hardened throughout the entire section.

SLLITDORE ELECTRICAL CChYVany,

Analysis-chromium steel.

Caroon 0.889
Chromium £2.03
Vanadium TraceSe

Heat Treatnent.

The steel is heated to 148) degrees I'., and

quenched in water.

Mi MSPURDIN'’ COMPANY.
Analysis-cnromium steel.
Carbon 0.89
Phos. O.017

Sulphur 0.023
BD.

Silicon eno
Mans". 0.44
Chrom. 3.03

Heat Treatment.
Mane steel is heated to 1525-1550 degrees F.,

and quenched in oil.

STROMBERG-CARLSCN TRIECHCIE MaAMPacTUnING CCLPany.

Analysis.

Tungsten steel. Chromium steel.
Carbon C.39 Carbon 0.89
Tungsten 6.0) Cnromium 2.00#2.25

liest Treatient.
Chrome regnets are herdened at a temperature of

about 1450 derrees i'., while the Tungsten magnets sre

~

hardened at 1550 to 1600 degrees F.

Til BRITISH DLCLOCU-HCUS™CH CCiaaAny, LIMITOL.

Analysis.

Tungsten steel Grade l. Grade 2.
Sungsten 5.960 5.460
Langanese 0.245 0.404
Carbon 0.616 0.698

Chromium stecl.

Chromium 2.000

Caroon about 0.750
£56
Neat Trerptment.
Tne stecl is hested to 750 degrees C, or 1360

derrees i*., and quencned in water.

TiS SULIVOMO CuvUneusiC, LTp. Ci’ COAcA, Javali.

This commany has recentiy petertea a new Legnetic
steel contsining from 0.3 to 0.2 per cent carbon an
ebout 35 per cent cobalt alloyed vith either chroniun,
molybdenum, or tungsten. It is claimed thet regnet
steel of tris composition will pive a coercive force
apprecially over 200, coupled with @ remenence of about

10,000.
ZT.

Ke Se. BPAGNET STEAL.

In June 1917, a new alloy steel wes discovered
by ik. Honda, a Japanese scientist. This steel poesess-
es a very high coercive force, and no doubt is the best

magnet steel knovm.

The composition of this steel is given eas
Carbon 0.4 - 0.8 percent, Cobalt 30 - 40 percent,
Tungsten 5 ~ 9 percent, Chromium 1.5 = 8 percent.
Tetspering is best effected by heating to 950 degrees C,
Or 174) degrees F., and quenching in heavy oil. Measure-
ments of the residuel magnetism for specimens of diff-
erent composition, give values from 92) to 620 C.G.S.
units; the coercive force for the same snecimens ranged
from 226 to 257 gauss. Artificial aging heating in
boiling water and by repeated mechanical shock reduced
the residusl magnetism only 6%. The hysteresis curve
for a magnetizing force of 1,300 gauss were teken for
annealed and tempered specimens; for the annesled
Specimen the coercive force was 30 gauss end for harden-
ed steel the coercive force was 238 gauss and the energy
loss per cycle 909,000 ergs. The hardness of ennealed
and temoered siecimens was fourd to be 444 and 652
respectively on the 3rinell scale and 38 and 55 on the

shore scale. The miscostructure of the herdened steel

Showed a finer grain than for the annealed.
28.
MAGUETO Machats.

As is well knovwm, the permenent nmegnets are
the most important part of any megneto, and unless
the nagnet be a thoroughly good one and well magnetiz-
ed, the magneto will give a very poor performance on
test. The manufacture of magneto regnets is attended
with considerable difficulties because it is essential
that the menet be finished to fine mechanical
accuracy, and at the same time, the magnetic cheracter-
istics must be extremely good. 3roadly speaking, the

essential characteristics are:

1. The magnet must be thoroughly sound throughout
its whole structure. That is, there must be
no visible or incipient cracks or flaws.

2e The grinding of the pole pieces and edges must
be carried out to the necessary fine limits of
accuracy demaided by the desig of the negnetos.

4. The magnetic characteristics mst ress a certain
mMinimm. In other words, the magnet when fitted
to the magneto mst be capable of creating in
the armature core a certain nurber of wagnetic
lines and of mintaining these during subsequent

life of the magneto.

Under working conditione, the megsneto is subjected
29-6

to considerable vibration and extrene and frequent
variations in temperature. any periranent magnet is
liable to lose its strength when operating under such
conditions, and to guard a loss of this reture, it is
necessary in hardening the magnet to aim at what is
termed a coercive force, not less than a certain
minimum. Experience has shown that a save mininum is
55, and with certain grades of regnet steel it is quite
safe to go below this fieure. The coercive force figure
is simoly an indication of the power of the magnet to
retain its mernetic field. That is, it is a measure of

the mpnetic tenacity of the steel.

It is slso desirable that the regnet should pro-
duce, when fitted in place, as large a flux as possible.
The actual working flux in the armature core veries be-
tween 20,000 and 50,000 magnetic lines, the forrer
Pisure applying to polar inductor magnetos, and the
latter figure to large rotating arreture magnetos.

Small rotating armature machines have an active flux
much nearer the lower figure snecified. liiow the value
of this flux, for a given design of mgneto, is great-
ly dependent on what is called the renanence of the
magnet. An average figure is 10,000 and it is desirable
that the remanence should be ges high as possible. The

remanence ficure is an indicetion of the flux thst would
30.

circulate around the megnet if its poles were bridged
with a heavy soft iron keeper, and it were then

magnetized in that condition.

“here has been a great deal of controversy about
the coercive force ard the rersanence. It so heppens
that it is only possible to increase one of tnese
factors at the expense of the other. Thst is, if the
coercive force be increased abnornally, as it cen be
by svecial methods of hardening, the remanence will be
lowered, and vice versa. It is contended in certain
quarters that a regnet with a very hich coercive force
(70) and a fairly low remanence (9,000 to 9,500) is
better than one having sa fairly low coercive force
(55) and a very high remanence (10,500 to 11,000).

Prom experience, it has been shown that the best re-
sults are obteined by striking a happy medium, end that
it is really a retrograde move, with the megnet steels
at present available, to increase the coercive force

beyond a range of from 60 to 65.

If an endesvor be made to vnush the coercive force
beyond the upper limit, the difficulties of ran.fecture
will be increased considerably, and the steel will be
brougnit to such a degree cf rardness that incipient
cracks are likely to develop muck more rapidly. No

compensating advantesze will be obtained, and ss a mstter
dl.

of fact, the active flux created in the regneto
arnature way tend to diminish in consequence of the
lower rerianence.e It must be pointed out thet different
steels behave in different ways, and while it way be
quite sare to work down to a coercive force of 55

with one grade of steel, another grade having slight-
ly different composition will demand s minimum figure
of at least 60. This curious difference in vehaviour
can be exynlained, but the explenation is outside the

scope of this work.
32.6

MsGNe? THs®iInG.

The testing of rmeynets to determine their
Suitability for use hes been a difficult and con-
tentious problem, rainiy owing to the disarreenent
between leading authorities as to tne trie criterion
of ragnetic quality. Guite recently is hes heen egreed
by those able to pronounce judgment on this subject,
that the vartictlar characteristics of a magnet which
are of importance depend on three distinct factors:

il. Remanence,
2e Coercive force,

oe “he shave of the remanence-coercive force curve.

This will be mde clear by reference to curve l,
Wig. 642 If we assume thst a magnet is wound over its
whole length with a magnetizing coil, and then has its
poles bridged by a substantial soft iron keeper making
a good contact with them, the flux density in the body
or the ragnet after passing a very heavy current through
the coil and again breaking the circuit, would be re-

presented by B, which is called the renenence.

Now if a reverse current be passed through the
coil, the mgnet can be slowly demagnetized by gradually
increasing the value of this qmrrent. With special

apparatus, the vslue of the flux density in the regnet
 

MICHIGAN AGRICULTURAL COLLEGE

   
   

oe
eS ees

OF MAT

     

reat alae]
JS6

can be readily measured for any value of demagnetiz-
ing current, and by plotting these readings curve l,
rize 6, i8 obtained. Oorrespondcing to sone definite
value of the demagnetizing current, the flux density
is brought to zero. This corresponds to the point
where the curve strixes the axis of abscissae. The
coercive force Cp is calculated from this denagnetiz-

ing current by taking into consideration the length of

the magnet and the number of turns of the ooil.

The important agreement on this netter thet has
now been reached, relates to the ordinate CB, which is
the actual flux density in the magnet at some intere-
mediate voint in the demagnetizing force (EH), and can
be taken as the figure of merit when comparing the
magnetic quality of magnets thet are closely alike.

It will be noted that it depends on the three factors
already enumerated. Experience shows that if the value
of the demagnetizing force (H) be fixed at approximate-=
ly 35, then the corresponding value of OB, gives an
accurate indication of how the magnet will perform in
the instrument for which it is designed. In other

words, O23) under these conditions is a figure of merit.

A good method for testing magnetos is shown in
Pig. 7. The apparatus used comprises a moving coil

type of measuring instrument, from which the ordinary
“oe

  
   

CONSTANT CURRENT
. IN THE
Mee

DEMAGNETISING COILS

PSOE A

|

late 7 A Magnet A OP aha
34.

permanent magnet has been removed, special pole pieces
being provided so that the meget to be tested can be
brought quickly into contact with them. With a constant
current flowing thru the moving coil, the deflection of
the pointer on the uniformly graduated scale gives a

direct indication of the flux in the mefnet.

In the testing outrit as shown diarrammatically
in Fig. 7, two dermgnetizing coils are fitted to the
top of the instrument case so thet these enbrace the
limbs of tre regnet when it is fitted to the poles.
The magnet is first rapnetized on a separate and dis-
tinct magnetizing outfit. With each magnet, three

readinzs are taken;

1. The first reading immediately after magnetiz-
ing the imegnet and fitting it to the instrue
ment. Call this a:

2e The circuit through the densgnetizing coils is
then completed through a lerge variable resiste
ance, and the deresnetizing current gradually
increased. Corresponding to a demagnetizing
force of approximately 35, the deflection is
again observed. Call this do.

3. The demagnetizing current is further increased
until the deflection is just zero. The demegnet-

izing current is carefully noted. Call this cal.
356

By using standard nemmets of different nagnetic
characteristics, it is possible to calibrate the in-
strument so that the actual coercive fcrce snd re-
manence figures for any magnet tested can be deduced
from the readings dj and c,. Brieily, if the readings
as set forth above be taxen with a set of standard
Isgnets of any particular make, and it is necessary to
use standards representative of each quelity of steel,

the results can be plotted to give two straight lines:

(a) whe first connecting dz and the rroduct

(b) “Whe second connecting cz and Cr.

From these calibration curves, the values of By.
and Cp Tor any mernet tested, can be deduced fron the

readings dy; and c by reterriny to the calibration

1?
curves that have been plotted from readings given by

standard pagnets of corresponding nake.

But as already pointed out, B, and Cp are not the
whole story, and the reading dy taxen by itself, is a
criterion of megnetic quality, when testing magnets in
bulk Which ere, as is usual, closely alike. Tire fallacy
Of devnending entirely on the renuanence and the coercive
force figures, and fortunately, dependence has been

placed on figures almost entirely in the pest, is made
apparent by curve © in Fig. 62. It is a hypothetical
case, but assuming tnat a megnet fave such a denagnet-
izing curve, it srkould be noted that the active flux
in the armature core of the instrument to which it is
fixed would be approximately OB,/0B, or 60 per cent of
the flux produced by the other 1#gnet when fitted to
the same instrument, despite the fact that the re-

manence and the coercive force figures are identical.

The agreehnent is that the raxilmm value of the
product of flux density and demagnetizing force reéch-
ed during the deliagnetizing process, can be taken as a
figure of merit, provided that the rerenerce and
coercive force figures are of the same order. A large
number of tests have revealed the fact that on the
average the minimum point on this new curve lies
epproxirately along a line drawn through the abscissa
whose point H equals 35. On this basis, tne ordinate
OB, is approximtely a measure of the maximum product
referred to, and the two methods of testing ghould

therefore yield substantially the same results.
 

—o —_———$<_—_———. ———— ———— —
O7.

TESTING PERUMANUNT MAGI EDS 3Y GEE USa OF A

LWAGHNSTOLBTER.

The epparatus used in this rethod of testing
is shown in Fiz. 8 and consists of;
le A magnetometer,
ce A telescove witn a scale attached,
3e An appsratus for the holding of magnets to be
tested with divisions merked off in cm. for
obtaining the distance of the negnet from the

magnetometer.

™he marnetometer consists of three small megnets
mode from 3/4" lengths of a watch spring and a smell
mirror about 1/2" square fastened upon a nonmagnetic
holder, as shovn in the right hand corner of Vig. 8.
This apysratus is suspended by a single strand of silk
fibre in a box lixe structure. The silk fibre is
attached to a pin which is adjustable so as to be able
to raise and lower the magretometer. The box enclosing
the magnetometer has a glass front facing the telescope
and scale so that the deflections of the magnetometer

can be caught from the mirror.

The deflections of the magnetometer are read

through the telescope from the scale attached.
aul, ata ee
cr

¢stment for mre bald Sommerer Pn OSE
err magre a eal
“regrre wi) i -Me  d od a Dd
 ahahae a

   

et eee Sold

Ps Nae 4

Aah L eel ae a /

| AAGHETZLES bar

ee

MLPA IAN
nm rm rm rt ri i L. nm mt n ]

 

 

 

 

 

 

ae D S Magnetomerer AT TE
358.

PO SOT Up TH APLARATUS.

The apparatus is set up as shown in the figure
witn the watch springs remets or the regnetometer
acting in the same direction as the earth's magnetic
field. ‘The telescope is set at a distance of 90 cm.
from the center of the menetormeter and the scele
set at 25. The device for holding the magnets to be
tested is placed in the same plain as the magnetoreter

and at right angles to the earth's magnetic field.

TOSTING TRE MAGNUTS.

The center of the vegnet to be tested is placed
at a distance of 100 cm. from the magnetometer with the
axis of the magnet perpendicular to the earth's magnetic
field from the center of the magnetometer. The lines of
force from the megnet attract one pole of the watch
soring magnet and repell the other thus producing an
anzle between the axis of the watch spring mgnet and the
earth's mmgnetic field. That is, the permanent magnet
deflects the watch spring nazgnet from the action of the
earth's magnetic field to the extent of the intensity
of the permanent mignet. These deflections are measured
from the deflections of the scale in the nirror thru
the telescope. These deflections are averaged for both
poles of the magnets and the mgnetic monent and the

pole strength of the msgnet is obtained from the followe-
396

ing formlae:

. 3 h 3
Yt ad*’ tan theta or K equals H adv tan theta

H a a

 

 

Pole strength equals M/l c.g.s.-

Where;

Kk equals the magnetic moment,

H equals the horizontal component of the
earth's megnetic field which was 0.173
under conditions of testing.

d equals the distance of the center of the
permanent maget from the center of the
Ihagne tometer.

tan theta equals one half the deflection
read upon the scale divided by the
distance of the telescope from the
magnetometer.

1 equsls the length of the permanent magnet.
49.
RESEARCH WORK ON PARNIANTNT MAGUS.

OBJZCT:

The object of this research work is to obtain
the comperative pole strength of permanent regnets
produced under mametization diring different heat
treatinents of the steel, end to prove certain
assumptions in connection with tne theories of

magnetism.

ASSUL@TICNS:

het is the molecular theory of regnetism is
true; then it is the oninion of the writers that upon
heating a piece of steel, the molecules will be ina
more pliable state ard therefore more easily moved.
Tnet is, the freeness of movement of the molecules in-
creases in some vroportion to the increase of temper-
ature in the steel. Now if a piece of steel is heated
to a certsin temmerature and while at this heat, the
steel is mepnetized by a suitable flux, then the
molecules will arrange themselves more easily and in a
less distorted fashion then under ordinery megnetization
of the cold steel. In fact, there should be a greater
nulber Of molecules in a better alignment due to a
greater freedom of motion of these molecules, snd upon
bringing the temperature of the steel down to thet of

the foom, the molecules should be in a less strained
41.

condition than under tre ordinsry means of magnetiz-
ation. With these conditions prevailing, a stronger
wagnet should result from the heating, and it wes
under these sssumptions that the writers experimented,

the results of which follow.

Of course, the electron theory has replaced the
molecular theory t>) e certain degree, but the correct-
ness of this theory is questioned by some of the fore-
most physicists of the day. Tnis theory is, as
exnvleined in a nrevious article, that the atoms of a
substance are built up, in pert st least, of a number
of electrons in repid rotation about e certain nucleus
of positive electrification. Hach of these electrons
moving in its orbit is equivalent to a minute msemet.

In case of diamagnetic si..bstancec, these orbits are
“supposed to be distribi.tea in such a symmetricsl way
that they produce no magnetic field outside of the atom.
Tne diamagnetic atom has no feenetic moment of its own.
In the case of paramagnetic substances, it is supposed
that the orbits of the electrons ere so distributed thet
the atoms have a megnetic moment of their own, and being
little magnets will fall into elignment when acted upon
by an external field. In the case of ferromsegnetic
substances, there seem to be mtuval actions between the
atoms or molecules. when an external field is applied to

a ferromagnetic substance, the neturel groupings of the
42.

molecules are changed; these eleuentary nme gnets fall
into new positions of equilibrium, one after the other,
and we have the effects know as permanent magnetism.
The key tnen to ferromagretism is the attraction and

the repulsion of one nolecular negnet upon another.

The theory of megnetism is that the diaragnetic
effect is probably present in all substances; but in
Many cases, it is masked by tne psramagnetic tuming
of the molectle, and there is also present a mutual
action between molecules. ‘The Same reasoning and
assumptions, as were brought forth ptnder the molecular

theory, can be applied to the above theory.

THY Le BORATORY WORK IN wHIS CONNECYVION.
The steel used in this experiment was of the

following composition:

Carbon . 2. « «© «6 « « « «© 04890
Chromium 2. « © « e « © «© 36030
ManganeS@. « 6 « « « « « 0.440
Phosphorus « « « « « « « O.O17
Silicon. .« . 2. « « « « « Oe250

sulphure «© «6 « © © © « « 04023

Yhe steel had been hardened by heating to

1525 - 1550 degrees F., and quenching in oil.
43.

CrLTVICAL POINT.

A piece of the steal was annealed at 1450
degrees I'., sO as to be able to rechine it. a 3/16"
hole was drilled in one end of a piece 1-1/2" long to e
depth of 3/4". The piece was then placed in the Brom
critical point determination apperatus. Tne resulting
graph is shovm on the following page from which the

Critical point was founac to be at 1250 degrees F.

HMAT TRHAYMIONY AND MACHSTIZATION.

There were seversl plans considered for the hest
treatment and repnetizetion bit the writers finslly
come to the conelusion that tne best method was to heat
the steel to the right temmerature, pass a flux thru the
bar by means of a lerge electro-tagnet f.r several
seconds, and then to quencn the bar while it was still
being magnetized. To build an apparatus to do these
three things, heat and steel quench and mgmetize, in
one unit would be a very difficult and unsatisfectory
plan, so it was decided to régnetize end quench the

Steel outside of the furnace.

The steel (3/8" x 3/4" cross section) wes cut in
6" lengths and numbered 1,2,3,4,5,6,7,8,1]1 and 12. The

bars were heated to the following temperatures:
me se

¥ it
var a

66

se. 3

ie
2 RE
ae

Sted.

=,

= J Md
SES KS OT a.

408 ih
om By
' : rt RS bapa a
/ ia my
Pere rae eee bet tL
a a Aus iW

 
44a.

Bar WOe. 1 « «© «© «© © «© «© ¢ HKHOOM Vempereture

" NB 0 eo © ow ew ew he he TOD degrees }.
" MB ww ww wee ew 6 BUD " "
" WA 6 ww ew we ew GOD " "
" " Bow ew ew ew ew ew ew 2 LOOYD " "
" "6 6 6 6 6 6 e e 0 L100 " "

" " 7 e @ e e e e e eL200 " "

" " 8 e e e e e e e ~1 300 " "

Bars ll and 12 were heated to 1200 and 1300 de-

grees ., resvectively for checx bars.

“he bars were heated to their respective tenper-
atures and held at this heet for rive minutes to insure
a uniform heat throughout the bar. The bars were in-
Stantly placed between the poles of a large electro-
magnet and magnetized for 1) seccnds while they were at-
heat. ‘his process should have helped the formetion of
the elignment of the molectlar regnets. The bsrs were
then quenched witn w:ter while they were still being
magnetized keeping tre flux flowing thru bar for 50

seconds.

TASTING.

the pole strengths of the vegnets were found by
the use of the megnetonueter anparatus which is fully
described in the preceding erticie on testing pertenent

Magnets. The bars were placed at a distunce of 1V0 cm.

 
456

from the magnetometer end an sverage deflection wes
taken for both ooles of the naeznet. Fron these de-
flections, the mwgnetic moment and the pole density
were figured. Yhe formulae used for conputing the

Inagcne tic moment end the pole density were;

Ji; de tan theta sna pole strensth equals i
H - Tem)

Bar No. Pemnerature Deflection of Fkole strength

of the bar Magnetometer in c.g.8.
L room temp. 729935 249.8
2 70) 7.5200 240.0
3 890 700589 £058
4 909 5.4630 174.5
5 LOVO 6.8760 219.8
6 1100 65620 209.6
7 1200 6.4950 207.0
8 1300 5.1000 165.0
il 1200 6.9346 e115

12 13509 4.8962 156.3

 

a=x@s tne Wwe Abe =

 
 

BZOSINIOS AWUNLINDI¥SOV NYSIHDIN

 
re
46.

OCH OLUSICNS
eRe PeTene) atWe

The results as shown in the table and grenph
prove to be quite the opposite from wnet was
expected. According to the assumptions, the feasi-
bility of which has not been denied by any of the
phzrrsicists with which the writers have consulted,
the steel should increase in magnetic density, to a
certain extent at least, witn a specified magnetiz-
ation under a rise in temperature. This, however,
was not the case bit quite tne onvnosite. Tne greeter
the temperature at which the steel was rmegnetized
the lower the pole density obtained. The question
arrises as to tne effect of heat upon tre ratnetism
of a piece of steel from the well known fact thet
heat destrovs magnetism. This fact, however, dves not
tend to tesr dovn the sssumptions made, but auite the

contrary.

The sinple exnerirrent of heating a regnet hus
been shown to everyone wi:o has tuken ea course in physics
and it not only substantiates the molecular theory but
shows that the tmnolecules hsve been distorted from their
normal position during the process of megnetization and
that heat does have an effect upon the rovement of these

molecules. What more proof does anyone want thet the

assumptions wer logical?Is it not reasonable to suppose
47.

tnat if this exneriment were reversed ana the heet
used to mexe the molecules anpliable, ard a hagnetize
ing force be tren avplied, and the material quenched,
thet it would produce stronrer magnets with greater
retentivity? There remsins but one other conclusion
to the results and that is that the fundsmental idea
or the moleculer theory is incorrect. ‘To decisively
prove tnis would be a very laborious and difficult
problem but the writers believe that they khaeve one of
the most important ste»ping stones in tne process cf

destroying the molecular theory.
48.

Che writers wish to thank the following men for

their interest and assistance in tniS work.

Henry Lantz sublow, B.S., Chen. ng.
Assistant rrofessor of Chemistry,

Liichigean agricultural College.

ndwin Jiorrison, Ji.5S.
Assistant vrofessor of rhysics,

Wichigean Agricultural Collese.

Charles Willis Chepman, A.3B., Bs.
2rofessor of physics,

Lichigan agricultural College.

7A

Arthur xodney Sawyer, 3-Se, eile
2rofessor of Electrical Hnrineering,

Michigen Agricultural College.

Ae Ge iianover,

Stromberg Carlson Telenhone ianufecturing Co.

He. ike Curran,

Indiana Steel Products Co.

H. BH. Lrause
9

Splitdorf Hledtrical Co.
Je Se Shunk,

“he Shunk Manuracturing Co.

Je we “SSterline,

The osterline Co.

49.
DO

BIBLICGRALHY.

Weaenetos - - - A. P. Young,

ilistory or the Yheories of nether ana LMlectricity -

Wnittaizer.

Manufecture of liagneto lagnets - - - p. lk. Neldt,

Keb. 3, 1921, Autonuobile Industries.

Mlections - - - ‘xn. Inst. Jan. 1Y16&,

Theories of lagnetism - - - §. 3ushren, G.h.R.,

©>+

liayy, L9lo.

wv Y

Theory of Magnetism - - - Stradling, Journsl Fr.

Inst. A J Trust ’ 1915.

oteel for Verionent Magnets - - - S. B. Thompson,

Journal Inst. wu, Vol. 50, pe 80, 1913.

rerieanent Nagnets - - «= J. A. Mathews, iroce A.S.TeLi.

Vol. 15, p. 50, 1914.

~

Permanent Marnets - - - F. C. xelley, Gu. Kev.,

Vole 20, pe 7, July, 1917.

Pertkanent Magnets in neory and rypractice - - -

me bvershed. Journal Inst. 3. os. Serte. 1920.

liagsnet Steel - - - Uiataro Honda. U. S.~ Late 1, 256,

1352, 133, 154, Ape 27, Llc.
Sl.

Stability of Nesidusl liagnetism - - - Ne. H. Willisus,

rphnyse ieve Vay 19135. pe 36d.

Criterion for Testing Iissneto Vagnets - - -

Je De WOorsan, London flect. Jan. 2, 1920. Pe 4.

Method of Testing - - - McLachlan. wroc. Phys.

Part 3, Apr. 15, 1920.

U. Se 3ureau of Standards, Cir. 17.

“ng ineering Vol. 107, April 25, 1919, p. 525.

Dlectricul world. Vol. 15, ®Peb. &, 1Y19. p. 267.

olectrician - - - tov. 17, 1916.

SOC. 32,
PP isa
cn