Full text of "Steel and its heat treatment"

See other formats

STEEL

AND ITS HEAT TREATMENT

DENISON K. BULLENS

* »

Consulting Metallurgist

FIRST EDITION

FIRST THOUSAND

NEW YORK

JOHN WILEY & SONS, INC.

LONDON: CHAPMAN & HALL, LIMITED
1916

DENISON K. BULLENS

.•1/Z rights r?serocd

THE SCIENTIFIC PRESS
ROBERT DRUMMOND AND COMPANY

IN MEMORY OF MY FATHER

331168

PREFACE

MODERN Heat Treatment should be considered as an art or trade,
since it certainly requires knowledge, skill and judgment for its
proper performance. These, in turn, necessitate at least some knowl-
edge of heat, of steel, and of the effect of heat upon steel. And all
three factors are linked together by the " human element." The
author has therefore endeavored to bring together the theoretical and
practical sides of the general subject of steel and its heat treatment
in such a manner as will, he hopes, be understandable by that
" human element."

It has been the author's attempt to make the chapters dealing
with the heating problem more of a " heat talk " than of a " furnace
talk"; of heat application rather than details of construction; of the
importance of the human element and scientific efficiency rather than
the elimination of the human element through scientific management;
and finally, of viewing the heating problem as an engineering prop-
osition, adapting each fuel to proper furnace design and operation
to meet the requirements of the problem in hand, and by so doing
aim for the adoption of the standard heating unit in terms of finished
product — " the cost of a unit of quantity of given quality."

He has attempted to make as practical as possible those chapters
relating to steel and the effect of heat upon steel. Theories have been
advanced only so far as has been thought necessary for a clear
understanding of principles. Wherever possible, illustrations in the
form of photomicrographs and charts have been given. The data
given under the various types of heat-treated steels have been
checked as far as possible and every effort has been made to be
correct.

vi PREFACE

To the many friends who have aided him in the preparation of
this book the author would express his sincere appreciation. Effort
has been made to give due credit for cuts and data at the proper
place, and for such as may not have been made, acknowledgment is
hereby rendered.

DENISON K. BTTLLENS.
PHILADELPHIA,

October 1, 1915.

CONTENTS

CHAPTER PAGE

I. THE TESTING OF STEEL 'I'-
ll. THE STRUCTURE OF STEEL 16

III. ANNEALING 39

IV. HARDENING 65

V. TEMPERING AND TOUGHENING 96

VI. CASE CARBURIZING . 112

VII. CASE HARDENING: THERMAL TREATMENT 154

VIII. HEAT GENERATION 173

IX. HEAT APPLICATION 192

X. CARBON STEELS 229

XI. NICKEL STEELS 257

XII. CHROME STEELS 295

XIII. CHROME NICKEL STEELS 306

XIV. VANADIUM STEELS 335

XV. MANGANESE, SILICON AND OTHER ALLOY STEELS 344

XVI. TOOL STEEL AND TOOLS 357

XVII. MISCELLANEOUS TREATMENTS 386

XVIII. PYROMETERS AND CRITICAL RANGE DETERMINATIONS. . . 408

vii

STEEL AND ITS HEAT TREATMENT

CHAPTER I
THE TESTING OF STEEL

Growth of Heat Treatment. — Probably no one division in the
metallurgy of steel has taken such wonderful strides in recent years
as has the art of heat treatment. Twenty years ago the scientific
knowledge and technical application of heat treatment were but
very limited. Such as it was, it usually consisted in " heating to a
red heat " for annealing, or perhaps the instructions called for
" harden at a bright red and temper to a straw color." Then it
was an art guarded with much secrecy and confined for the most
part to makers of tools and a certain few specialties.

Practically all alloy steels require treatment of one sort or another.
In the " natural " state very few steels present their full value, so
that heat treatment is not only advisable but often mandatory.

Necessity for Heat Treatment. — Take for example the steels
used in the automobile industry. The frame requires resistance to
vibratory stresses occasioned by rough roads, as well as strength and
toughness. Rear axles must have great tortional resistance; front
axles must withstand vibrations. The steering parts must be
strong, tough and without brittleness; the springs must neither sag
nor break. Crank-shafts must be able to resist impact, besides being
stiff. Gears are subject to wear and must be capable of withstanding
this action if a smoothly running transmission is to be had. And
so each separate part might be named, all having a more or less
severe duty to perform and requiring steels possessing various
degrees of strength, toughness, resilience, endurance, shock-resisting
and wearing qualities.

Testing. — These various combinations of static and dynamic
strength} are obtained by adjusting and correlating both the chemical

2 STEEL AND ITS HEAT TREATMENT

composition and heat treatment of the steel. Certain chemical
components intensify the static properties of the material; others
may affect the dynamic qualities. Thus by coupling with a steel of
suitable chemical combination the proper heat treatment, there
arises a product with physical properties most adapted for the work
in hand. Similarly, having once produced a suitable article, it then
remains to duplicate it. To this end all rational heat treatment
must be aligned and standardization of results be obtained. In
order to accomplish these specific requirements, the influence of
definite chemical composition and definite treatment must be known,
as will be described in later chapters. The guide to this work is
frequent and constant testing, and a definite knowledge of the vari-
ous components should be possessed by every heat-treatment man.
Thus we may say that the purpose of practical testing is (1) to sup-
ply information as to suitable material and its qualities for different
purposes, both for the manufacturer of the material and for the
designer or user, and (2) to test the specified uniform quality of the
material.

Stresses and Strains. — Testing resolves itself into a determination
of the strength of the material, which in turn is measured by the
application of a force and its resultant effect. The force put upon
a body is termed the stress, and the deformation resulting from that
force is the strain. Upon the method of applying that force depends
the nature of the test. Thus we may conveniently classify such
stresses under the following headings:

A. Steady or constant loads — static stresses;

B. Repeated static stresses and accelerated stresses — fatigue

stresses;

C. Suddenly applied loads — impact stresses;

D. Repeated impact or vibratory stresses — dynamic stresses;

E. Miscellaneous tests such as resistance to penetration, wear,

etc.

Tensile Strength. — The most common test for static strength,
that is, the strength of the steel under constant load and without
shock or vibration, is the tensile test. Thus we may call the tensile
strength the absolute strength of the metal under tension, i.e., the
force actually required to pull the metal asunder. A standard test
piece is gripped between the upper and lower jaws of a testing machine
and the total resistance to rupture is measured. Knowing the area
of the cross-section of the test piece and the load required to break

THE TESTING OF STEEL 3

it, the strength per square unit may then be calculated. The tensile
strength is usually given in pounds per square inch (American),
tons per square inch (British), or kilograms per square millimeter
(metric system; 1 kg. per mm.2 =1422. 32 Ibs. per square inch). The
accuracy of the tensile test is dependent not only upon the conditions
under which the test is made, such as the rate of pulling, alignment
of test piece in the machine, etc., and which are more or less influenced
by the human element, but also upon the metal itself. The higher
the tensile strength and brittleness of the steel, the greater the
possibility of error; differences of several thousand pounds per
square inch are often encountered in the same piece of high-tensile,
heat-treated steel, even in the absence of brittleness.

Test pieces are generally taken half way between the center and
the outside of the piece, and longitudinally or " with the grain."
Occasionally it is necessary to take tests transversely or " across the
grain"; in this case the results will be lower than in the longitudinal
test, the exact amount depending upon the composition and treat-
ment of the steel.

The static strength increases in direct proportion to the carbon
content of the steel. For the ordinary basic and acid open-hearth
steels, without heat treatment, Campbell gives the following formulae
by which the tensile strength of such steels may be roughly deter-
mined. These results apply for steel " in the natural."

Acid open-hearth steel :

Tensile strength = 40,000 +1000C + 1000P +xMn.
Basic open hearth steel:

Tensile strength = 41,500+7700 + 1000P+?/Mn.

In these formulae, C equals each one point (0.01 per cent.) of
carbon as determined by combustion, P equals each 0.01 per cent,
of phosphorus, Mn equals each 0.01 per cent, of manganese, and x
and y are given in the table on page 4.

Elastic Limit (Tension). — The term " elastic limit " has probably
been more ill-used than any other common technical testing name,
with the possible exception of " hardness." Among its many
definitions the two which stand out pre-eminently are (1) the least
stress at which the material retains a permanent deformation or
" set " after the removal of the stress; and (2) the least stress under

STEEL AND ITS HEAT TREATMENT

Percentage of Carbon.

On Acid Steel.

Lbs. per Sq. In.

On Basic Steel.

Lbs. per Sq. In.

0.05

not

0.10

80*

130

0.15

120

150

0.20

160

170

0 25

200

190

0.30

240

210

0 35.

280

230

0.40

320

250

0.45.

360

0.50

400

0 55

440

0.60

480

* Beginning only with 0.4 per cent, manganese,
t Beginning only with 0.3 per cent, manganese.

which ductile material exhibits a marked yielding — sometimes
denoted as the " yield point."

The determination of the true elastic limit should always be
taken from a curve plotted, using an extensometer, from a series
of careful observations, as otherwise sets caused by non-homo-
geneity and initial stress might be obtained which do not repre-
sent the plasticity of the material. This method of determining
the elastic limit is but little used commercially, as the amount of
labor involved is too great.

The yield point, or commercial elastic limit, is obtained by
noting the stress at which the test piece first begins to " give " or
elongate. This may be obtained by means of two prick-punch marks
and observing the first signs of elongating by means of dividers held
on these points; or by noting the drop of the weighing beam or halt
in the load indicator (" jockey ") ; or by means of the general appear-
ance of the test piece.

In its practical application the elastic limit may be called the
working strength of the material, for in most cases the steel or
machine part becomes useless when strained beyond its elastic
limit. This is particularly true of automobile construction, in
which the value of a car is dependent upon the correct adjustment
and alignment of its several working parts, such as in transmissions
and transmission suspensions. All tests given in this book, unless
otherwise noted, refer to the commercial elastic limit or yield
point.

THE TESTING OF STEEL 5

The relation existing between the elastic limit and the tensile
strength is too broad a subject for discussion here, as the varying
chemical compositions and heat treatments exert such a tremen-
dous influence; a study of the results given in following chapters
will show a proportionality of 40 per cent, and upward.

Elongation. — The elongation is measured in per cent, of the
original test section and is commonly the amount of stretch which
will occur in the material when pulled apart by tension. It is
usually measured in relation to an initial distance of 2 or 8 in.,
or 100 mm. when the metric system is used, but other specifica-
tions as used in Europe give a definite relation of original gauge-
length to the thickness or diameter of the specimen.

Reduction (or Contraction) of Area. — The reduction of area
refers to the area at the point of rupture, usually reported in per
cent, reduction of the original area — that is, the original area of the
test piece minus the area of the smallest cross-section after frac-
ture; this divided by the original area is the percentage reduction of
area.

Ductility. — The percentage elongation and percentage reduction
of area are a measure of the " ductility " of the material, usually
varying inversely with the tensile strength. The true measure of the
ductility of the steel cannot be taken alone from either the elonga-
tion or reduction of area, as the results obtained in either case will
depend in a large measure upon the size of the test piece, the method
of testing, etc. Many engineers regard the reduction of area as the
more reliable; this is offset by the fact that many steel specifications
make no mention of the reduction of area, but particularly specify
the percentage elongation. Ductility may also be defined as the
amount of distortion of the material before final rupture.

Compressive Strength. — The compressive strength of material
is its resistance to crushing. The test is generally carried out upon
a small cylinder or 1-in. cube of the metal, using the same machine
as for the tensile test. Care must be used to see that the line of
strain passes exactly through the axis of the specimen, and that the
plates above and below the piece have a greater resistance to pene-
tration than the metal to be tested. The application of the term
elastic limit is similar to that in the tensile test.

Torsional Strength. — As its name implies, the torsion test is used
to determine the resistance to twisting. This test is very largely
used to-day for automobile steel and is measured in inch-pounds
with the amount of distortion given in degrees. The elastic limit is

6 STEEL AND ITS HEAT TREATMENT

obtained as in a tension test, using either a tropometer or an auto-
graphic attachment. The usual comparison is by calculating the
shearing stress in pounds per square inch.

Endurance. — The computation and understanding of such static
stresses as have been previously outlined are comparatively simple.
The requirement to be fulfilled in designing is that the working stress
shall not exceed the elastic limit of the material, whether it be
in tension, compression or torsion. Numerous every-day failures,
however, which cannot be accounted for by the limited information
given by such tests, have forced investigators to probe more deeply
into the complicated kinematic forces which seem to have such a
great influence upon the " life " of the metal. It is now a well-
known fact that, if a stress is applied a great number of times, i.e.,
repeated, each application being made before the material has had
time to recover from the preceding stress, the material will event-
ually break even though the stress is below the elastic limit of the
material. These repeated stresses upon steel cause a gradual dis-
turbance of the structure and its component particles, which greatly
weakens the material, and is called fatigue. The resistance to
fatigue and its numerical test value may be termed the endurance
of the steel. The stresses embodied under the heading of fatigue
may be broadly classed as repeated static stresses and acceleration
stresses ranging from zero to maximum or from a negative maximum
to a positive maximum — alternating— stresses.

Fatigue Stresses. — These stresses are produced in a machine
part by an external force or forces of varying strength and direction
acting upon the part. When the force is produced by a continu-
ously varying acceleration or retardation of masses taking part in
the movement of the machine part, they may be conveniently termed
acceleration stresses.1 Typical stresses of this category are the
revolving shaft stress on a loaded wheel or machine axle, the piston
pressure and the acceleration pressure of the movable parts in the
piston rod and crank-shaft of high-speed steam and oil engines.
These perpetual stresses or so-called fatigue stresses are the essential
ones' in the movable parts of most high-speed machines, and a knowl-
edge of the capacity of the material to resist them should serve as
a basis for the selection of the material and design.

Rotary Bending. — Such static endurance tests may be carried
out in a machine of the rotary bending type, such as the Wohler
or the White-Souther machines. From a study of a large number of
1 J. O. Roos af Hjelmsaeter, Int. Assoc. Test. Mat., 1912, Vol. II, No. 9.

THE TESTING OF STEEL 7

experiments made on a rotary bending machine of the Wohler
design, Foos concludes that such endurance tests are not suitable as
specification tests, but are of great value in the selection of material
and the heat treatment for various purposes.

On the other hand, the real value of the rotary bending test as a
criterion of the brittleness-fatigue endurance has been of late greatly
questioned. That the results usually obtained are largely indicative
of the elastic limit alone is probably more in accord with our present-
day knowledge. The results from a series of tests conducted by
Foos upon a Wohler type rotary bending machine with steels of
0.11, 0.40 and 0.65 per cent, carbon, given in the following table,
would tend to support the latter theory, as one would naturally
expect from past experience that the 0.40 per cent, carbon steel would
have a greater fatigue-resisting strength than the 0.65 carbon steel.

ROTARY BENDING TESTS, WOHLER MACHINE

Chemical.

Static Properties.

Endurance
Limit.

Fiber

Stress

<p
1

c
o

Tensile
Strength
Lbs. per
Sq. In.

Elastic
Limit
Lbs. per
Sq. In.

Elonga-
tion,
Per Cent
in
3.94 Ins.

in Kg.
giving
I racture
after
1 Million

•H

Revolu-

tions.

0.11

0.33

0.019

0.013

0.01

49,770

32,990

34.1

0.40

0.51

0.027lo.011

0.20

82,760

50,770

23.8

O.T.

0.40

0.51

0.027

0.011

0.20

109,780

70,820

15.6

0.65

0.49

0.023

0.007

0.20

116,040

50,900

14.3

O.T.

0.65

0.49

0.023

0.007

0.20

151,020

94,560

11.0

Treatment. " A," heated at 1560° F. for 30 minutes, and air-cooled.

" O.T.," heated at 1560° F. for 30 minutes, quenched in mineral oil, and
re-heated to 1025° F.

Suddenly Applied Loads. — Machine parts at one time or another
may be exposed to abnormal working loads such as may result from
a single accidental blow or a sudden retardation of masses in motion,
and which in any case cannot be supposed to be frequently repeated.
Such abnormal stresses are therefore in the nature of suddenly applied
loads or impact stresses, and constitute a different group from those
previously discussed under static stresses. It is evident that such
.stresses demand that the material be able to sustain a great work of
deformation for a single or a few impacts without rupture — that is,

8 STEEL AND ITS HEAT TREATMENT

the machine part shall sustain as little damage or deformation as
possible. In general the " ductility " (elongation and reduction of
area) has for a long time been used as a measure of the work of rup-
ture. But, although such tests are of comparative value, they do
not measure either the ductility under impact or the strength or
resistance under impact.

Drop Test. — The drop test as commercially applied may be
described as an aggravated bend test on a large scale. It is a relative
or qualitative test only, usually made on a full-size forging, to deter-
mine roughly the homogeneity of the metal and its ductility under
shock. We have arbitrarily separated this test from the " impact "
tests, reserving the latter as applied to the specific measurement of
the impact strength upon a more or less standardized test bar.
The most common application of the drop test is that of locomotive
axles, in which it is required that the axle shall stand a specified,
number of blows at a given height without rupture and without
exceeding, as a result of the first blow, a certain deflection.

Impact Strength. — The impact test is used to determine the
ability of the metal to withstand a suddenly applied load in the
nature of an impact or shock, thus detecting brittleness or lack of
toughness. This function is called resilience by the foreign technical
world, referring to fragility or the converse of brittleness, and is
stated in terms of the specific work of rupture under impact. It
should not be confused with the English word " resilience," which
is interpreted in this country as " springiness." This fragility is
not defined by the tensile test, although an experienced steel man,
from an examination of the size and aspect of the grain and other
conditions of the fracture of the test piece, can usually express an
opinion as to the fragility, but he cannot assess any definite value.
Although the drop test specifies the fragility in a qualitative manner,
it does not measure the actual resistance to rupture, and is therefore
but an imperfect test. In order to overcome such objections and
to arrive at a definite value, machines have been devised which
break a special notched bar by a blow — the force required to rupture
the metal being measured in kilogram-meters or foot-pounds.
Notched test bars are used in order to localize the deformations.
The blow must be delivered with sufficient velocity to bring out the
desired brittleness functions. This blow or impact may be obtained
by a falling weight (the Fremont machine), by a falling pendulum
(the Charpy principle), or by a revolving fly-wheel bearing a releas-
able knife (the Guillery machine), these three representing the most

THE TESTING OF STEEL 9

common types of impact machines in use abroad, as well as the
Olsen pendulum type (using a test specimen in the form of a canti-
lever) in this country.

Impact Tests. — Fremont * recommends as the ideal conditions
to be realized in the application of the impact test: (1) A minimum
drop of four meters, or proportional to the impact speed; (2) the
weight of the anvil block to be equal to at least forty times the
weight of the tup; (3) sufficient ease and rapidity of adjustment of
the machine.

On account of the many factors entering into the problem and
the numerous designs of impact testing machines, the majority of
the testing associations have abstained from prescribing any special
type of apparatus for performing the test. The Copenhagen Con-
gress (1909) of the International Society for Testing Materials has,
however, recommended the use of a standard notched test bar of
30X30X160 mm., or a smaller bar of 10X10 mm. cross-section
when the larger size is not available. On the contrary, there are
many who believe that a smaller, rectangular test bar reveals more
clearly the local defects which form the nucleus of future cracks, etc.

Use of Impact Tests.— The fragility test has a double purpose —
to point out steel which is defective, either inherently or by incorrect
heat treatment; and to act as a valuable aid for the adjustment of
a proper heat treatment. It is evident that steels burdened with
sulphur and phosphorus, or rotten with piping and segregation, will
always remain brittle whatever one may do. But ordinary, sound
stock, properly treated, is nearly always non-brittle. The degree
of brittleness will of course vary according to the composition,
treatment and use of the different steels. The effect of heat treat-
ment upon the impact strength is very great, so that due care should
be used in so adjusting the chemical composition and treatment of
the steel as to give the best combination for the work in hand.
The impact test, in conjunction with other tests, gives a quick
method of determining a quality the importance of which is yearly
becoming more prominent scientifically; commercially, however,
the impact tests are so unreliable, or vary so greatly, that they can
hardly be used with any degree of accuracy.

Fatigue Impacts. — An impact or shock has a considerably greater
effect than a stress slowly applied, and if repeated a sufficient num-
ber of times will eventually result in the rupture of the specimen.
When such frequently repeated stresses are comparatively small —
1 Ch. Fremont, Proc. Int. Assoc. Test. Mat,, Vol. II, No. 9, 1912.

10 STEEL AXD ITS HEAT TREATMENT

that is, are well below the elastic limit of the material — they may be
termed fatigue impacts. Their measurement, as determined by the
energy or amount of work they represent, is a principal component
of the dynamic strength of the material. Apparatus for thus test-
ing the material is developed upon the principle of alternating
impacts. As practical examples of such stresses in commercial
application there may be mentioned the stresses to which the axles
of locomotives or railway cars are subjected every time the wheel
passes over a rail joint, or the impact stresses sustained by different
parts of a motor car when passing over bad roads, or in changing to
different speeds, etc.

Alternating Impact Machines. — Various machines for establishing
comparative numerical values for this dynamic strength have been
designed, in the endeavor to produce an alternating flexure and at
the same time deliver a blow or impact. This has been accomplished
by applying blows to the upper part of a test piece with the aid of
two pendulum balls which are made to fall alternately from opposite
directions from a certain height against the test piece. Other
machines have been patterned along the lines of the Upton-Lewis
machine, keeping the fiber stress well within the elastic limit.

Alternating Impact Test Results. — The study of a great number of
alternating impact tests of a comparative nature, the views of other
engineers, and the study of steel parts broken in service, would lead
the author to the opinion that the dynamic strength (using the term
in its broadest meaning) of straight carbon steels reaches a maxi-
mum at 0.25 to 0.35 per cent, carbon, with perhaps even narrower
limits of 0.25 to 0.30 per cent, carbon. Further, the maximum
endurance is obtained when the steels have been properly hardened
from a temperature slightly over the upper critical range and tough-
ened at a temperature of 1200° to 1250° F. The maximum com-
bination of static working strength, ductility, resistance to shock
and vibration, and endurance probably is obtained in straight car-
bon steels with 0.35 to 0.45 per cent, carbon when subjected to the
above treatment.

Relation of Various Tests. — There does not seem to be any simple
relation between the elastic limit at steady tensile stress and the
limit of endurance in the rotary bending test, although some engineers
consider it as a " reflection of the elastic limit." The rotary bending
is undoubtedly less than the former, and according to the investiga-
tions of some engineers, the limit of endurance (rotary bending) will
generally amount to about 50 to 80 per cent, of the elastic limit.

THE TESTING OF STEEL 11

According to experiments made by Foos on straight carbon steels,
the limit of endurance for rotary bending and alternating impact
within the elastic limit agree fairly closely.

A high value in the work of rupture in the impact test may be
considered to give comparative security in the case of occasional
abnormal over-loads.

Thus, the requirements for a high-quality steel for machine parts
are a high limit of endurance for the normal stresses and a high figure
of rupture for the abnormal stresses. As a rule, these qualities are
opposed to each other in ordinary materials, and it must rest upon
experience as to which to give the preference; in parts which suffer
through vibration and other fatigue stresses, it will probably be
wiser to give preference to the endurance properties. It must be
remembered that heat treatment and the various alloys may
entirely change the different properties.

Hardness. — " Hardness l may be defined as the property of
resisting penetration, and conversely, a hard body is one which,
under suitable conditions, readily penetrates a softer material.
There are, however, in metals various kinds or manifestations of
hardness according to the form of stress to which the metal may be
subjected. These include tensile hardness, cutting hardness, abra-
sion hardness, and elastic hardness; doubtless other varieties could
also be recognized when the experimental conditions are modified
so as to bring into operation properties of the material in addition
to that of simple, or what may be conveniently called mineralogical
hardness. This has been defined by Dana as ' the resistance offered
by a smooth surface to abrasion.' The usual quantitative tests for
hardness are static in character, but the conditions are profoundly
modified when the penetrating body is moving with greater or less
velocity. The resistance to the action of running water; to the
effect of a sandblast; or to the result of the pounding of a heavy
locomotive on a steel rail, afford examples of what might perhaps for
purposes of distinction be called dynamic hardness, which is a
branch of the subject which has received little quantitative
examination."

Brinell Hardness.— The Brinell test consists in the pressing of a
hardened steel ball into the surface of the object under test by
means of a fixed load. The dimensions of the impression thus ob-
tained form the basis for calculating the hardness. If the number
of kilograms making up the load is divided by the spherical surface
1 Thomas Turner, Inst. Journ., May, 1909.

12 STEEL AND ITS HEAT TREATMENT

of the impression, expressed in square millimeters, a number is
obtained, expressing the pressure exerted per square millimeter of
ball impression. This number is now accepted as a measure of
hardness, and it is hence called the Brinell hardness number. It is
generally sufficient to utilize the diameter of the ball impression
itself as a measure of the hardness. In order to make tests exo-
cuted at different works directly comparable, a standard ball of 10
mm. and a load of 3000 kilograms are used.

Brinell Transference Number. — It is a well-established fact that
the Brinell hardness numbers follow very closely the tensile strength
of the same types of steel, whether the steel be " in the natural,"
or whether it has been subjected to some heat-treatment process.
For this reason it is particularly applicable to the rapid testing of
steel from which it would be impracticable to take regular tensile
tests. A few comparisons between the actual tensile strength as
obtained by pulling tests and the hardness number obtained from
the test pieces will serve as a basis for future calculations. By ob-
taining such a " transference number " the probable tensile strength
of the steel in question may be easily computed by multiplying the
hardness number by the " transference " number. This transference
number will vary with the chemical composition of the steel, and
to a small extent with the manner of testing (whether with or across
the grain) and between tempered and annealed steels. On the whole,
however, the test is comparatively accurate for steels purchased
or made under the same general chemical specification. For straight
carbon steels this transference number may be said to be about
500 to 520. The Brinell method has a great advantage over the
scleroscope in that it does not require an extremely smooth or
polished . surf ace for the test; the removal of scale by filing is prac-
tically the only requirement.

Many companies are standardizing their heat treatment product
by taking the hardness of each piece treated, thus ensuring a close
range of the desired tensile properties. On account of the influence
of the size of the original section upon the physical results as obtained
by the pull-test, it is much easier to determine the transference
number for the specific grade of steel being treated, and then vary
the toughening temperature so as to give the desired Brinell hardness
number. It is the author's experience that this method is fairly
accurate, and that the Brinell number will give a close approximation
of the true tensile strength regardless of whether the treated bar is
1 in. or 5 ins. thick. The Brinell method is simple and com-

THE TESTING OF STEEL

mercially efficient, with the exception of either very thin or highly
tempered material.

For reference convenience, the relation between the diameter of
the impression and the hardness number is given in the following
table:

BRINELL'S HARDNESS-NUMBERS
Diameter of Steel Ball = 10 mm.

Diameter
of Ball
Impression
mm.

"°&

ll-§«

•o 3^§

J§£ (SCO

Diameter
of Ball
Impression
mm.

Hardness
Number for
a Load of
3000 Kgr.

Diameter
of Ball
Impression
mm.

.*l&

rcXi^HH
§ S 00
fg*l

S^ cdec

Diameter
of Ball
Impression
mm.

Hardness
Number for
a Load of
3000 Kgr.

Diameter
of Ball
Impression
mm.

Hardness
Number for
a Load of
3000 Kgr.

946

418

228

143

2.05

898

3.05

402

4.05

223

5.05

140

6.05

2.10

857

3.10

387

4.10

217

5.10

137

6.10

2.15

817

3.15

375

4.15

212

5.15

134

6.15

2.20

782

3.20

364

4.20

207

5.20

131

6.20

2.25

744

3.25

351

4.25

202

5.25

128

6.25

2.30

713

3.30

340

4.30

196

5.30

126

6.30

2.35

683

3.35

332

4.35

192

5.35

124

6.35

2.40

652

3.40

321

4.40

187

5.40

121

6.40

2.45

627

3.45

311

4.45

183

5.45

118

6.45

2.50

600

3.50

302

4.50

179

5.50

116

6.50

2.55

578

3.55

293

4.55

174

5.55

114

6.55

2.60

555

3.60

286

4.60

170

5.60

112

6.60

2.65

532

3.65

277

4.65

166

5.65

109

6.65

2.70

512

3.70

269

4.70

163

5.70

107

6.70

2.75

495

3.75

262

4.75

159

5.75

105

6.75

2.80

477

3.80

255

4.80

156

5.80

103

6.80

71.5

2.85

460

3.85

248

4.85

153

5.85

101

6.85

2.90

444

3.90

241

4.90

149

5.90

6.90

2.95

430

3.95

235

4.95

146

5.95

6.95

Shore Scleroscope. — The principle of the Shore scleroscope hard-
ness test is based upon the height of rebound of a diamond-faced ham-
mer when dropped from a standard height upon the surface of the
material to be tested. One of the great disadvantages of the sclero-
scope, in the author's experience, is that it requires a highly polished
surface in order to obtain anywhere near accurate and comparative
results. The scleroscope readings are probably more indicative of
the elastic limit than of the tensile strength. The following experi-
ments made with the scleroscope, illustrating the results to be ob-
tained with different methods of polishing, may afford some explana-
tion of the variations often characteristic of this instrument:

STEEL AND ITS HEAT TREATMENT

Specimen rough filed readings, 25 to 30

smooth filed readings, 27 to 32

rubbed with emery cloth No. 1 readings, 32
rubbed with emery paper No. 00 readings, 32
polished with diamantine readings, 33

Ballistic Test. — The ballistic test is distinct from the static hard-
ness tests above described in that it is a measure of the dynamic
hardness by resistance to penetration under violent impact. From
the author's experience with protective-deck and bullet-proof steel,
the Brinell hardness is only a measure of the tensile strength and does
not give a comprehensive idea of the ballistic qualities of the plate or
sheet. Similarly, tests made by the Italian Government show that
none of the tests mentioned in this chapter, either static or dynamic,
nor their ensemble, gives a sufficient indication of the resistance
which such plates will oppose in firing tests.

Wear.— Wear, or the hardness of material as indicated by its
resistance to abrasive action, has demanded considerable attention
of late on account of the increased development of high-power
machines and engines. The increased weight put upon locomotive
axles and rails, the higher speeding of rotating parts, and a similar
tendency to wear in other machine parts have all necessitated
further study of this important property of steel.

For resistance to abrasion, Robin l has arrived at the following
values, these being obtained upon annealed steel with carbon con-
tents as given, manganese — 0.25 per cent, to 0.30 per cent.; phos-
phorus— 0.015 per cent, to 0.40 per cent.; silicon — about 0.20 per
cent. :

WEAR BY ABRASION. ANNEALED STEEL

Carbon
Content
Per Cent.

Abrasive
Figure.

Carbon
Content
Per Cent.

Abrasive
Figure.

0.07

295

0.65

308

0.12

293

0.69

280

0.25

312

0.83

258

0.38

350

1.00

252

0.60

312

1.03

252

J. Robin, Inst. Journ., II, 1910.

THE TESTING OF STEEL 15

This would tend to show that the wear is not proportional to the
carbon content in annealed carbon steels, and that the maximum
wear might be expected with steels of approximately 0.40 per cent,
carbon. He further concludes from other experiments that the
abrasive wear increases with the percentage of phosphorus and
diminishes with the amount of manganese and silicon.

CHAPTER .II
THE STRUCTURE OF STEEL

Steel. — Steel is an alloy, the principal and essential chemical
constituents of which are iron and carbon. With these there are
usually certain impurities, such as phosphorus, sulphur, and silicon,
which have not been entirely eliminated during the process of
manufacture, as well as manganese — and perhaps other alloys such
as nickel and chrome — which may have been intentionally added
for a definite purpose. Of the elements which go to make up ordinary
steel, the manganese, phosphorus, sulphur and silicon — the impurities
— generally total to about 1 per cent. ; the carbon will vary from a
few hundredths of 1 per cent, to about 2 per cent.; and the
balance will be iron.

Furthermore, steel is not a simple substance like copper or gold,
but isjnore like granite,1 in that it is made up of a number of individual
grains (let us say) or " minerals," corresponding to the quartz, mica
and feldspar of the granite. Thus, in steel which has cooled slowly
from a high temperature, we have " ferrite," " cementite " and
" pearlite." And just as the relative amounts of the quartz, mica
and feldspar may vary in the rocks of the granite class, so will the
relative proportions of ferrite, cementite and pearlite vary in dif-
ferent steels according to the specific chemical composition of the
steel as a whole.

Cementite. — As we have stated above, the carbon and iron are
the essential, as well as controlling, elements in the steel — and this
is particularly true of the carbon. In steels which have cooled slowly
from a high temperature, the carbon is first and always combined
with a ^definite amount of iron to form a " carbide of iron," corre-
sponding to the chemical symbol FesC. This compound consists of
6.6 per cent, carbon and 93.4 per cent, iron, and is micrographically
known as " cementite." The balance of the iron, is practically
carbon-free and is known as " ferrite."

Pearlite. — Now during the process of cooling at a moderate rate
from a red heat, this cementite will form a mechanical mixture with a

THE STRUCTURE OF STEEL 17

definite amount of ferrite, so that the resultant will contain approx-
imately 0.9 per cent, carbon. This new constituent is called " Pearl-
ite " and usually consists of interst ratified layers or bands of. ferrite
and cementite. Pearlite is regarded as a separate and distinct
constituent of steel, as it forms distinct " grains " when present in
any appreciable quantity, always contains this definite percentage
of carbon, and — as will be explained later — is always born at a definite
range of temperatures.

Eutectoid Steels. — From this it will be seen that a steel contain-
ing 0.9 per cent, carbon will consist entirely of pearlite. Such steels
are known as " eutectoid " steels, and that ratio of -carbon as the
eutectoid ratio.

Hypo-eutectoid Steels. — Steels containing less than this eutectoid
ratio of carbon will consist of a definite amount of pearlite, varying
according to the carbon content of the steel proper, and the balance
in " free " or " excess " ferrite. These steels are called " hypo-
eutectoid " steels, as they contain less than 0.9 per cent, carbon.

Hyper-eutectoid Steels. — Similarly, if the carbon content exceeds
0.9 per cent, carbon, there will not be sufficient ferrite to inter-
stratify with all of the cementite, so that these steels will consist of
pearlite plus free cementite. Such steels are called " hyper-eutec-
toid " steels.

Expressing this in a different way, we may say that very low
carbon steels are made up of ferrite with a little pearlite. With
increase in the carbon content of the steel, the amount of pearlite
will likewise increase, with a corresponding diminution in the amount
of free ferrite, until at 0.9 per cent, carbon the steel will be wholly
pearlitic. Beyond this point the amount of pearlite will decrease,
with a corresponding increase in the amount of free cementite.

Structure of Slowly Cooled Steels. — In slowly cooled steels we
may therefore tell with great accuracy the approximate structural
composition of the steel. And, vice versa, knowing the relative pro-
portions of pearlite and ferrite or cementite, as determined micro-
scopically, we may determine the approximate carbon content of the
steel.

This is represented graphically in Sauveur's diagram as shown in
Fig. 1, in which the percentage carbon is represented by the ab-
scissae and the percentage constituents by the ordinates.

These .facts are also illustrated by the photomicrographs in Figs.
2 to 9, representing the structure of slowly cooled steel of 0.06, 0.18,
0.32, 0.49, 0.57, 0.71, 0.83 and 1.46 per cent, carbon respectively. In

STEEL AND ITS HEAT TREATMENT

Figs. 3 to 7 it will be seen that the pearlite (dark; structure not
brought out by the etching) gradually increases in amount, while the
ferrite (light) proportionally diminishes. Fig. 8 shows a steel of the
eutectoid composition in which the ferrite (dark) and the cementite
(light) are interstratified with each other, there being practically no
" free " ferrite such as characterized the lower carbon steels. Fig. 9
shows the structure of a 1.46 per cent, carbon steel in which the free
cementite (white) occurs as a network between the grains of pearlite
(dark). In Fig. 10, showing a steel of just above the eutectoid

KX)

Ferrite

7T;

^~^OM

uentite

>•>*:

Pea

rlite

0.2 0.4 0.6 0.8 1.0 1,2 1.'
Per Cent Carbon

FIG. 1. — Ferrite-pearlite-cementite Diagram. (Sauveur.)

ratio, we see the first appearance of the free cementite between the
pearlite crystals. In Fig. 11 we have this whole range represented
by means of case carburizing a " dead soft " steel.

Physical Properties Dependent upon Constituents. — Upon the
relative proportions of these constituents will depend the physical
properties of the slowly cooled steel, neglecting for the time being
their relative arrangement. Each of these components — ferrite,
pearlite and cementite — has certain physical characteristics with
which we must be familiar in order to gain some idea of the proper-
ties of such steels.

THE STRUCTURE OF STEEL

FIG. 2. — 0.06 per cent. Carbon. Approximately Pure Ferrite. X75. (Ord-
nance Dept.)

v > *-'V

i - .V

FIG. 3.— 0.18 per cent, Carbon. Ferrite (White) and Pearlite (Dark). X75.

(Ordnance Dept.)

STEEL AND ITS HEAT TREATMENT

IG. 4.— 0.32 per cent. Carbon. Ferrite (White) and Pearlite (Dark).

(Ordnance Dept.)

X75.

JPIG 5, — 0.49 per cent. Carbon. Ferrite (White) and Pearlite (Dark). X75.

(Ordnance Dept.)

THE STRUCTURE OF STEEL

0.57 per cent. Carbon. Ferrite and Pearlite. X75. (Ordnance Dept.)

•

FIG. 7. — 0.71 per cent. Carbon. Ferrite and Pearlite. X75. (Ordnance Dept.)

STEEL AND ITS HEAT TREATMENT

FIG. 8.— 0.83 per cent. Carbon. Pearlite. X485. (Ordnance Dept.)

FIG. 9. — 1.46 per cent. Carbon.

Pearlite and Cementite (White),
nance Dept.)

X75. (Ord-

THE STRUCTURE OF STEEL 23

Ferrite.— Ferrite is soft, ductile and relatively weak. It has a
Tensile strength of approximately 40,000 to 50,000 Ibs. per square
inch, with an elongation of about 40 per cent, in 2 ins. Ferrite
in itself has no hardening power as applicable to industrial purposes.
It is magnetic and has a high electric conductivity. Its appear-
ance under the microscope has been shown in the photomicro-
graphs previously mentioned — that is, as polyhedral crystals in the
low carbon steels.

Pearlite. — As previously mentioned, the common occurrence of
pearlite in slowly cooled steels is in the lamellar formation, is being

FIG. 10. — Laminated Pearlite and First Appearance (as Veins between Grains)
of the Excess Cementite. XlOO. (Titanium Alloys Mfg. Co.)

composed of alternate plates of ferrite (showing dark under the
microscope) and cementite (showing white under the microscope).
As will be shown later, under different rates of cooling pearlite may
exist in other formations and dependent upon the relative arrange-
ment of the ferrite and cementite of which it is composed; some
of these various modifications are shown in Figs. 8 and 10. Normal
pearlite, that is, interstratified bands of ferrite and cementite such
as shown in Fig. 8, has a tensile strength of approximately 125,000 to
130,000 Ibs. per square inch, with an elongation of about 10 per cent,
in 2 ins.

STEEL AND ITS HEAT TREATMENT

Cementite. — The properties of cementite are very little known
with the exception of its great hardness and brittleness, which are a
maximum. Free cementite, that is, unassociated with ferrite to
form pearlite, probably does not have a tensile strength much
greater than 5,000 Ibs. per square inch. Its ordinary occurrence
in slowly cooled steels (carbon greater than 0.9 per cent.) is either
as a network, such as we have seen, or as spines and needles.

Pearlite and
Cementite

Pearlite

Pearlite and
Ferrite

®3$.

> *•

FIG. 11. — Case-carburized Steel, Showing nearly Carbonless Steel (Bottom)
Gradating into High-carbon Steel (Top). (Weyl.)

Static Strength. — We may now sum up these facts in their relation
to the static strength of slowly cooled steel as follows: Free ferrite
has a minimum tensile strength with maximum ductility; pearlite
has a maximum tensile strength with low ductility; free cementite
confers added hardness and brittleness, with a consequent lowering
of the tensile strength. In other words, by increasing the amount
of pearlite in the steel, we increase the static strength but with a

THE STRUCTURE OF STEEL

corresponding decrease in the ductility. And as an increase in the
amount of pearlite necessarily means an increase in the amount of
carbon, the effect of increased carbon will give the same results.
This is shown graphically in the diagram of Fig. 12.

Heat Treatment. — Heat treatment in general consists in chang-
ing or regulating the structure of the steel by various methods of

Per Cent. Elongation.
8 8 S
Tensile Strength, Lbs. per Sq. Inch

]S £ 8 8 8 i

jt> ^^"

'/
/

fJNl

<::::^^

*Y\ '

/
/
/

/
/

'Carbon 0.2 0.4 0.6 0.8 1.0 1.2 1.
Per Cent. Carbon.

FIG. 12. — Approximate Influence of Carbon upon the Strength and Ductility

of Steel.

heating and cooling. By the term " structure " is meant (1) the
metallographic constituents, among which are those just described;
(2) the size of the grain; (3) the net-work. In order to understand
the nature of these changes and their application it will be necessary
to have a clear understanding of the mechanism by which these
changes are brought about.

Critical Points. — The nature of steel, as explained before, is
complex. The structure of any particular steel may be modified

STEEL AND ITS HEAT TREATMENT

or entirely changed by various degrees of heating, and all of which
take place in the steel while it is in the solid condition. These
structural changes take place at temperatures known as the " critical
points " or " critical ranges " of the steel. These critical ranges are
denoted by the letter "A," followed by the letter "c" (abbreviation
for the French word " chauffage," signifying " heating ") or the

1800

1700

4»

g 1600

5 •

b ir>oo

s
1

•^.

A 1-2-3

1300

-v

1200

1100

) 0.20 0.40 0.60 0.80 1.00 1.20 1.40

Per Cent. Carbon.

FIG. 13. — Critical Range or Carbon-Iron Diagram.

letter " r " (" refroidissement " or " cooling ")• These signs, Ac or
Ar, are further modified by the numerals 1, 2 or 3, indicating the
particular point referred to. Thus Acl would mean the first critical
range passed upon heating the steel beyond a certain temperature,
and so forth. These critical points or ranges are indicated graphic-
ally in Fig. 13.

THE STRUCTURE OF STEEL 27

In considering this diagram let us devote our attention to a
certain specific case, such as a low-carbon steel with about 0.2 per
cent, carbon. We will also assume that the steel is in the normal
condition resulting from slow cooling, in that it consists of about
25 per cent, pearlite and about 75 per cent, free ferrite. We will
also first consider what is the influence which these changes occurring
during the critical ranges have upon the constituents of the steel.

In the first place, practically no change in the constituents occurs
during heating until a temperature corresponding to the lower
critical range, Acl, is reached, which is equivalent to about 1330° F.

In passing through this critical range there is a complete change
in the nature and structure of the pearlite, it being converted into an
entirely new constituent with new characteristics. This is tech-
nically known under the generic term of a " solid solution," micro-
graphically called " Austenite." The excess ferrite remains
unchanged.

Solid Solutions. — To understand better the nature of this new
component let us consider the interaction between salt and ice.
When these two substances are placed in contact with each other, we
know that under suitable conditions of temperature the salt and
ice merge into one another and so pass from the state of two separate
substances or mechanical mixture into that of one separate substance
or brine solution. A similar process takes place in the case of the
pearlite, except that the resultant solution is solid instead of being
a liquid. The individual plates of ferrite and cementite which have
characterized the pearlite grains now merge into one another, form-
ing this new substance or constituent, known as a solid solution.
This new constituent, save that it is a solid and not a liquid, has all
the properties of a liquid solution. Its original components are
merged into a single entity, giving a complete indefiniteness of com-
position, and with entirely new characteristics.

Absorptive Power of Austenite. — Just as the brine solution can
dissolve more salt or ice with increased temperature, so this solid
solution of iron and iron carbide possesses the power of absorbing
more free ferrite or free cementite. Therefore, as the temperature
is raised above that of the lower critical range (Acl), and there being
an excess of ferrite in this particular steel (0.2 per cent, carbon),
the solid solution or austenite begins to absorb this ferrite. This
continues progressively with increased temperature until the upper
critical range, Ac3, is reached, or, for this particular steel, a tempera-
ture of about 1525° F. At this temperature the last of the remaining

STEEL AND ITS HEAT TREATMENT

excess of ferrite is absorbed by the austenite, so that above the upper
critical range of the steel the steel is composed entirely of austenite —
the solid solution.

These changes are illustrated graphically in Fig. 14. It will be
seen that the initial pearlite, comprising about 25 per cent, of the
normal steel, changes into austenite (the solid solution) at a tempera-
ture corresponding to that of the lower critical range, Acl, and then
progressively absorbs the free ferrite until at a temperature corre-
sponding to that of the upper critical range, Ac3, the whole steel con-
sist^ of austenite.

AcS

Free Fcrrite

FIG. 14. — Change of Pearlite and Free Ferrite into Austenite during Heating.
Carbon about 0.2 per cent.

These same changes are shown microscopically in Figs. 15, 16,
17 and 18. The first photomicrograph shows the normal condition
of the steel, being made up of a small proportion of pearlite (dark),
and a large amount of free ferrite (light) . The three other structures
were obtained by heating this same steel to temperatures above the
lower critical range and then " fixing " the structure obtained at
those temperatures by " quenching." Fig. 16 shows the structure
representative of a temperature between that of the Acl and Ac2,
the solid solution l (dark) having increased considerably in amount

1 Strictly speaking, the dark areas thus referred to as the " solid solution "
are not austenite, but its transitional stage, martensite. In the ordinary carbon
steels austenite as such cannot be retained by the ordinary methods of quenching

THE STRUCTURE OF STEEL

over that of the original pearlite as in the previous figure. Fig. 17
represents the structure obtained at a temperature somewhat under
that of the upper critical range, Ac3; in this case it will be noted
that the solid solution covers nearly the whole field, there being but a
small amount of the free ferrite (white). The structure representa-
tive of heating to slightly above the upper critical range is shown in
Fig. 18; it will be seen that the free ferrite has now been entirely
absorbed by the solid solution. Also note the extremely refined

FIG. 15. — Normal Low-carbon Steel as Rolled. X60. (Bullens.)

structure, as we shall have occasion to refer to this particular feature
a little later.

Allotropic Modifications of Ferrite. — Associated with these
critical ranges there is also a change in the allotropic 1 form of the
ferrite (iron). Thus pure ferrite (as distinguished from the ferrite

(as will be explained under the chapter on Hardening), but changes into martens-
ite. Martensite, however, is also a solid solution, and for the purposes of explana-
tion in this chapter — in order not to complicate matters — we will consider it
permissible to use the term as indicated.

1 Sauveur defines allotropy as " suggesting marked and sudden changes in
some of the properties of a substance occurring at certain critical temperatures,
without any change of state or of chemical composition."

STEEL AND ITS HEAT TREATMENT

FIG. 16. — Low-carbon Steel Quenched between Acl and Ac2. X60. (Bullens.).

FIG. 17.— Low-carbon Steel Quenched a Little below Ac3. X60, (Bullens.)

THE STRUCTURE OF STEEL 31

associated with cementite to form pearlite) in its normal condition
is called " alpha "-ferrite or " alpha "-iron, and is characterized
by extreme ductility and magnetic properties. Upon heating this
alpha-ferrite to a little over 1400° F., corresponding to the critical
range Ac2, the iron becomes practically non-magnetic and is then
known as " beta "-ferrite or " beta "-iron. Upon further heating to
a temperature above the upper critical range, Ac3, there is still
another change in the allotropic modification of the iron, it being
known as " gamma "-ferrite; this gamma-iron is slightly softer than
the beta modification. Gamma-iron has the property of being able

FIG. 18.— Low-carbon Steel Quenched above A3. X60. (Bullens.)

to dissolve carbon or iron carbide, a characteristic which is not held
by alpha-iron.

Merging of the Critical Points. — Now by referring to the carbon-
iron diagram in Fig. 13 it will be noted that at the eutectoid ratio of
carbon, that is, at about 0.9 per cent, carbon,1 the three critical ranges
Al, A2, and A3, merge into one. That is, steels consisting of pearlite
alone, when heated to a temperature beyond this point, will change
directly into the solid solution austenite, which will consist of a
solution of carbide (or carbon, according to some authorities) in

1 The eutectoid ratio on the chart is given as 0.85 per cent, carbon. Accord-
ing to the authority selected this ratio will vary between 0.8 and 0.9 per cent,
carbon; but the more recent tendency is to adopt 0.90 per cent.

32 STEEL AND ITS HEAT TREATMENT

gamma-iron. Similarly, as normal pearlite always represents this
eutectoid ratio, the same change of pearlite into a solid solution of
carbide in gamma-iron will always occur at this temperature in ordi-
nary carbon steels irrespective of the carbon content of the steel as
a whole.

Changes in Heating Different Steels. — With this explanation
clearly in mind, we may now refer back to the example of the 0.2
per cent, carbon steel and more fully explain the changes which take
place in the constituents. Under normal conditions, this steel will
consist of pearlite plus alpha-ferrite. Upon heating through the
Acl range, the pearlite will change into austenite, the iron of which
will be in the gamma modification; the free ferrite will still remain
in the alpha condition. Upon further heating through the zone
marked " 2 " on the diagram Fig. 13, the austenite will begin to
absorb the free ferrite. Upon passing through the Ac 2 range the
balance of the free ferrite will pass from the alpha modification into
that of beta-f errite ; the steel as a whole will be hard and non-mag-
netic. Upon further heating (zone 3) the remnant of the beta free
ferrite will be gradually absorbed, so that on passing through the
critical range, Ac3, the whole steel will be in the condition of austen-
ite (zone 5), or a solution of iron carbide (or carbon) in gamma-iron.

In a similar manner we might explain the changes in constituents
which take place upon heating normal steel with any carbon up
to that of the eutectoid ratio. With a carbon content somewhere
between 0.3 and 0.4 per cent, (varying according to different author-
ities) it will be noted that the A2 and A3 ranges merge into one,
known as A2-3.

In a manner analogous to the absorption of free ferrite by the
solid solution in the hypo-eutectoid steels, the free cementite will be
absorbed in the case of the hyper-eutectoid steels, the final solution
taking place at a temperature range indicated by the line Acm.
The only difference, and that a practical one, is that the solution of
the free cementite takes place more sluggishly than the solution of
the free ferrite of the lower carbon steels.

The Ar Ranges. — Corresponding critical changes take place
upon cooling slowly from above the upper critical range, except that
they occur in the reverse order and with opposite effect. On account
of the molecular inertia, however, we find that these critical ranges
(of cooling, Ar3, Ar2, Arl, etc.) are a number of degrees below the
temperatures at which they appeared on heating. This difference
is dependent upon length of exposure and the temperature to which

THE STRUCTURE OF STEEL 33

the steel was subjected, the rate of cooling, and, more particularly,
upon the influence of the alloying elements which may have been
added to the steel. Some of the alloys, if present in sufficient amount,
will cause the recalescent points to fall below normal temperatures,
and are the basis of air-hardening steels and similar compositions.

Changes on Slow Cooling. — Upon slow cooling from above the
upper critical range, the solid solution will commence to reject the
excess ferrite (or, of course, the excess cementite in the case of hyper-
eutectoid steels) as the temperature decreases from Ar3 to Arl.
The reverse changes in the physical nature and properties of the
iron occur at the critical ranges during cooling as those previously
noted under heating. When the lower critical range is reached, the
excess ferrite or cementite will have been entirely rejected, and as
the steel passes downward through this range (or point), the solid
solution — now containing 0.9 per cent, carbon — will change into
pearlite. Under similar conditions of cooling, the original steel and
the present heated and cooled steel will have the same structure.

Refinement. — Before leaving the subject of the influence which
heating through these various critical ranges has upon the structure
of the steel, there are a few points which we wish to mention briefly
concerning refinement. Again assuming that the steel is in the
normal condition, no change will take place in the structure until the
temperature has been raised at least to that of the lower critical
range. At this temperature the original pearlite grains are com-
pletely changed and will possess that maximum refinement which
the formation of the austenite can impart — that is, complete refine-
ment. If the steel has a carbon content other than that of the eutec-
toid ratio (i.e., contains free ferrite or free cementite), the steel as a
whole will not be refined; the excess ferrite or cementite will remain
unaltered and the steel will retain its original grain-size. This is
brought out by a comparison of Figs. 15 and 16. Complete refine-
ment of the steel as a whole will not result until the steel has been
heated to a temperature slightly over that of the upper critical range,
as a comparison of Figs. 17 and 18 will prove, and as is evident from
previous discussion. A clear understanding of these principles must
be had, as they form the basis of many of the heat treatment pro-
cesses which will be later developed.

Grain-Size Beyond Ac3. — As the temperature is progressively
raised above the critical range, a gradual coarsening of the aus-
tenite grains occurs. This increase size is not only a function of the
temperature, but also of the length of time at which the high tern-

34 STEEL AND ITS HEAT TREATMENT

perature is maintained. The practical application of the principles
noted in this and the previous sections will be considered in the chap-
ter on Annealing.

Network. — The third factor in the structural changes taking
place upon heating is the effect of temperature upon the network.
All hypo-eutectoid steels in the normal condition are made up of
pearlite with a varying amount of excess ferrite, the latter decreasing
with the increase in carbon content. From our study of the inter-

FIG. 19. — Microstructure of Cast-steel Ingot as Cast. X75. (Ordnance Dept.)
Tensile Strength, 77,000. Elastic Limit, 39,000. Elongation, 10.5.
Red. of Area, 16.9.

nal mechanism by which the constituents of the steel are formed
by slow cooling, we know that the pearlite forms the basis of the
structure, the ferrite being rejected by the solid solution (pre-pearlite).
Being thrown out to the boundaries of these austenitic grains, the
excess ferrite forms a network around these grains. Upon reheat-
ing, this network is gradually absorbed, its final absorption taking
place upon passing the upper critical range. This change is similar
to that explained previously under the description of the action of
the excess ferrite.

THE STRUCTURE OF STEEL

FIG. 20.— Microstructure of Cast Steel Ingot Forged to 1450° F. X75. (Ord-
nance Dept.) Tensile Strength, 83,500. Elastic Limit, 50,500.
Elongation, 27.5. Red. of Area, 43.3.

FIG. 21. — Microstructure of Steel Subjected to Cold Work, and Showing Dis-
tortion of Grain. X50. (Ordnance Dept.)

STEEL AND ITS HEAT TREATMENT

FIG. 22. — Hammer-hardened Steel, 0.46 per cent. Carbon. X300. (Savoia.)

FIG. 23.— Effect of Cold Rolling on 0.20 per cent. Carbon Steel. X60.

(Bullens.)

THE STRUCTURE OF STEEL

FIG. 24. — Effect of Punching upon Structure of &-in. Chrome-nickel Steel
Plate. Hole Downwards and at Right. X50. (Bullens.)

FIG. 25. — Machining Strains on Surface of Mild Steel. (Brearley.)

38 STEEL AND ITS HEAT TREATMENT

The Effect of Work on Grain-Size.1 — Steel cooled slowly and
undisturbed from a high temperature will show a coarsely granular
or crystalline structure, and the size of the grain is a function of the
temperature and time during which the material is held at the maxi-
mum temperature, and the rate at which the material is cooled. In
large masses of material the structure will be coarser in the center
than at the surface, due to the difference in rate of cooling. In
order to overcome this difference and at the same time produce a
homogeneous, uniform material, the steel is worked during the period
at which grain growth would ordinarily take place. Steel which has
been hot- worked down to the Arl point will show a finer grain, and
will be stronger than the same steel slowly cooled without work,
and will at the same time show high ductility. Examples of steel
worked and unworked are shown in the photomicrographs of Figs.
19 and 20.

Steel which has been worked below the Arl range — that is, cold-
worked — will show considerable distortion of grain, as is illustrated
by Fig. 21, and may even become hardened, Fig. 22. Cold rolling
frequently develops a weak, laminated structure, as is shown in Fig.
23. Even punching or machining operations may greatly affect
the structure, examples of which are given in Figs. 24 and 25.

1 In part from Bulletin 1961, Ordnance Dept.

CHAPTER III
ANNEALING

Annealing. — Annealing, in its commercial application, may have
for its purpose any or all of the following aims: (1) to " soften " the
steel and thus put it in condition for machining or to meet certain
physical specifications; (2) to relieve any internal stresses or strains
caused by previous hardening or elaborating operations; (3) to obtain
the maximum refinement of the grain in combination with large
ductility.

Thus, depending upon the results desired, commercial annealing
will consist of a heating operation carried to some predetermined
temperature — although not necessarily over the critical range — to
produce the results desired in items 1 and 2 previously noted, and
followed by a moderately slow cooling of the metal from that tem-
perature. True or full annealing requires a heating to above the
upper critical range of the steel.

Elemental Considerations. — In the abstract, annealing would
appear to be but a suitable correlation of the following elements:

1. Rate of heating;

2. Temperature of heating;

3. Length of heating;

4. Rate of cooling.

But in actual practice the success to be attained in annealing (or in
any heat treatment process, for that matter) must depend upon
the judgment and skill of the furnace operator in applying the basic
principle's which may be derived from a consideration of the above
factors. Thus it is the man who determines the manner of placing
the charge in the furnace, of regulating the flow and composition of
the hot gases, of determining when the steel has been uniformly
and thoroughly heated, and similar fundamentals. For while it is
advisable and perhaps necessary to understand the theory behind
the actual work itself — the " wherefore " — the " wherewithal " is
largely a personal equation and should be borne in mind throughout
every theoretic discussion of principles or practice.

40 STEEL AND ITS HEAT TREATMENT

Heat Application. — The manner in which the steel is placed in
the furnace is a factor of supreme importance. It may even be
said that three different kinds of annealing may be produced in the
same furnace operating at the same indicated pyrometer reading,
dependent simply upon the method of placing the sto\;k in the furnace.
Particular stress should be laid upon the necessity for getting heat
to the center and bottom of the charge, not only for the sake of
uniform annealing, but also to shorten the time of absorption and
lessen the time of exposing the top and outside edges of the charge
to the heat and influences of the chamber atmosphere. Thus it has
been shown that a uniform chamber temperature does not necessa-
rily mean a uniformly annealed product; that a circulation of heat
through the mass is more desirable than the mere application from
the outside; that, with the same chamber uniformity, it is possible
to vary the quality of the anneal by the manner in which the stock is
placed in the annealing zone. It is advisable to raise the charge
above the furnace floor or hearth upon suitable blocks or supports,
to separate each piece from the other, and to avoid localized heating
through over-loading. It is only by such means that there will be
provided an opportunity for the circulation of the hot gases through
the charge.

Pre-Heating. — Slow, careful and uniform heating is always
advisable regardless of the chemical composition or physical condi-
tion of the steel. Heating to such temperatures as are common in
general annealing practice necessarily results in more or less change
of physical condition or molecular readjustment, and the greater
the hardness, brittleness and amount of internal strain in the metal,
the greater will be the deleterious effect of such heating. Thus
objects of intricate design, or with varying cross-sections, or steel
in a hard, brittle condition, should be given the greatest care in heat-
ing in order that the release of any strains shall not cause warpage
or otherwise injure the metal. Such pieces should never be placed
directly in a hot furnace, but should be given a careful pre-heating.

ANNEALING HYPO-EUTECTOID STEELS

Microscopic Changes. — In the previous chapter we have explained
that, in the ordinary cast, rolled or forged sections (pearlitic in
character), there is virtually no change in grain size or in constit-
uents during the heating to a temperature below that of the lower
critical range Acl. That is, there is no refinement of the steel.

ANNEALING 41

As the temperature passes the Acl range there occurs the complete
change of the pearlite to the solid solution, giving the maximum
refinement to the austenite.

Passing through zone 2 (Refer to Fig. 13) the excess ferrite is
progressively absorbed by the solid solution, causing an apparent
decrease in the grain-size of the steel as a whole. This absorption
is the slower the greater the carbon until the carbon nears 0.85
per cent., but is offset by the fact that the amount of free ferrite
decreases as the eutectoid ratio is approached.

Upon passing through the critical range Ac2 we have the forma-
tion of beta iron with no apparent change between the relative
grain-size of the alpha ferrite and beta ferrite grains. The same
absorption of the excess ferrite continues progressively, but with
increased sluggishness (due to the supposed properties of beta ferrite) .
This applies to steels with say 0.12 to 0.30 per cent, carbon. In
the very low carbon steels Howe 1 sums up the probable changes
during this period in a provisional proposition that t»>) if initially
fine-grained the steel coarsens, though only very slowly; (b) if
initially coarse-grained it refines slowly; (c) to coarsen again upon
long exposure to these temperatures.

The changes taking place through zone 2 continue through
zone 3, although more slowly. If the rate of heating through this
range of temperatures is comparatively slow, there will be a complete
absorption of the remaining ferrite just before Ac3 is reached.
Under ordinary circumstances final absorption will occur on passing
through the Ac3 range.

The Upper Critical Range. — As the steel passes the upper critical
range there is the complete refining of the grain, it becoming very
fine and almost amorphous. As the temperature increases beyond
this range the grain-size coarsens, causing a diminution in the
strength of the steel. The effect upon the physical properties of
the steel is great. The tensile strength is increased somewhat as
the temperature advances. The elastic limit rises until a point is
reached about 175° to 200° F. over the upper critical range, after
which it then decreases. The elongation and reduction of area
decrease very rapidly. These changes in the physical properties
are shown graphically in Fig. 26 in which the results obtained by
heating a 0.40 per cent, carbon, basic open-hearth steel to a definite
temperature and then slow-cooling with the furnace are plotted

1 H. M. Howe, "Life History of Network and Ferrite Grains in Carbon Steel,"
Proc, A, S, T, M,, Vol. XI, 1911.

STEEL AND ITS HEAT TREATMENT

against the temperatures. It will be noted that the softest and
most ductile steel is obtained at approximately 1475° F., which is
about 50° over the upper critical range.

Heating Over the Upper Critical Range. — The effect of heating
beyond the critical range is well developed by the series of photo-
graphs (by Howe) shown in Figs. 27, 28, 29, 30 and 31. The steel
(0.40 per cent, carbon, 0.16 per cent, manganese) was heated to the

1,200 1,300 1,400 1,500 1,600 1,700 1,800

Degrees Fahrenheit

FIG. 26. — Effect of Annealing Temperature on Physical Properties.

temperatures indicated, held at those temperatures for ten minutes,
and then cooled in air. There is a difference in grain-size between
that cooled from 1472° F. and from 1652° F., showing that anneal-
ing should never be carried very far beyond the upper critical range
or Ac3 point unless for special reasons. As the high temperatures
were successively raised to 1832° and 2012° the grain-size becomes
noticeably larger, until at 2192° the steel is " burnt." These
photomicrographs also exhibit the effect of air cooling upon the
structure, in that it develops a distinct net-work or cellular structure.
The effect of heating beyond the upper critical range is also brought

ANNEALING

EFFECT OF HEATING BEYOND Ac3,

0.40 per cent. Carbon Steel Heated at Temperatures Indicated for Ten Minutes and

AIR COOLED. •

FIG. 27.— 1472° F. X40. (Howe.) FIG. 28.— 1652° F. X40. (Howe.)

FIG, 29,— 1832° F. X40. (Howe.) FIG. 30.— 2012° F. X40, (Howe.)

FIG. 31.— 2192° F. X40. (Howe.)

44 STEEL AND ITS HEAT TREATMENT

out in an analogous manner by Figs. 32, 33, 34, 35 and 36, except
that in this case the steel has been cooled very slowly (furnace cooled)
from the specific temperatures.

Use of the Microscope for Checking Structural Changes. — From
previous theoretical discussion, it is evident that in order to fulfill the
true or full annealing operation, it is necessary to heat the metal to
over the upper critical range of the steel in order to obtain the com-
plete change of structure with the smallest grain-size possible. The
microscopic changes which take place during such heating of a 0.28
per cent, carbon steel are given as follows:

Temperature. Structure.

1325° F. Very coarse ferrite and pearlite similar to

the original bar.
1375° F. Laminae of ferrite strong, ground-mass

refined.

1425° F. About 25 per cent, ferrite laminae left.
1475° F. Trace of coarse ferrite unabsorbed.
1500° F. Complete refining. No coarse ferrite.
1550° F. Structure similar.
1650° F. Refined but grain-size coarsening.

These experiments l were carried out with a view to discover the
cause of failure of an eye-bar (carbon 0.28 per cent.) when placed
in service. The original steel had been annealed several times at
temperatures under the upper critical range, but a microscopic study
showed that these heatings had simply refined the pearlitic ground-
mass. In other words, it was found that the proper annealing
temperature necessary to obtain a completely refined steel was
beyond the upper critical range. In this steel it would seem to be
about 1500° F. The lower critical range is shown by the refining
of the ground-mass which occurred between 1325° and 1375°.

Diffusion. — We have repeatedly stated that complete absorption
of the excess ferrite takes place at the upper critical range of the steel.
Although this statement is true, there is another phase of this
absorption to be considered, and a full understanding of which will
probably clear up many of the questions which have perplexed
those unfamiliar with the theory of annealing. This phenomenon
may be called " diffusion." Let us hark back to our former simile

i Wm. Campbell, " Further Notes on the Annealing of Steel," Proc. A. S. T.
M., Vol. X, 1910.

ANNEALING

EFFECT OF HEATING BEYOND Ac3.

0.40per cent.'Carbon Steel Heated at Temperatures Indicated for Ten Minutes and

FURNACE COOLED.

FIG. 32.— 1472° F. X40. (Howe.) FIG. 33.— 1652° F. X 40. . (Howe.)

FIG. 34.— 1832° F. X40. (Howe.) FIG. 35.— 2012° F. X40, (Howe.)

FIG. 36.— 2192° F. X40. (Howe.)

^46 STEEL AND ITS HEAT TREATMENT

of the salt and brine solution. When a grain of salt is dissolved by
the brine, it is the solution in the immediate neighborhood of the
salt crystal which acts as the solvent and not the entire volume of
the brine solution. In time, however, the dissolved salt will eventu-
ally diffuse through the whole body of brine and the brine will then
be of equal composition throughout. Now a similar process is going
on in the steel when the solid solution (austenite) is absorbing the
excess ferrite, and it will be found that complete absorption may not
mean complete diffusion or equalization. The process of equalization
goes on with the rise in temperature. If the passage through tem-
peratures under that of the upper critical range is only slow enough,
a large part of the diffusion will have occurred by the time Ac3 is
reached. In order that there may be complete diffusion, and there-
fore complete grain-refining, the sojourn at a temperature approxi-
mating Ac3 must be long enough for this complete diffusion of the
absorbed excess ferrite and therefore of the solid solution. Although
exposure to a higher temperature would naturally hasten this diffu-
sion, it would be at a cost of coarsening the austenite grains. The
effects of non-equalization will be discussed in a later part of the
chapter.

Rate of Heating. — Studying the rate of heating from the practical
aspect there is also another factor to be considered — that of bringing
the whole mass of the steel to the proper temperature evenly. It
is self-evident that the center of a large mass of steel, such as loco-
motive axles or steel blooms, will lag in temperature behind the
exterior. In other words, it is the tendency of the core to be con-
siderably lower in temperature than the shell or outside of the steel.
It is then a common procedure to raise the temperature of the fur-
nace beyond the proper annealing heat in order to drive the heat
to the center of the piece to be annealed. This is a great mistake.
It is far better to take the extra time required to heat more slowly
as the proper temperature is neared, thus bringing the steel to an
even temperature throughout. If this were not done, the exterior
of the piece might be carried beyond the proper temperature — and, in
general, a needlessly high temperature is injurious and tends to
recoarsen the grain.

Expressing this question in a different way, we may say that the
furnace in which the metal is being heated for annealing should in
no case be run at a higher indicated temperature than the maximum
temperature to which the metal itself is to be heated. To illustrate:
a piece of steel heats, cools and decarbonizes on the corners first.

ANNEALING

The life of the entire piece of steel is no greater than the life of the
corners. If the steel is placed in a hot furnace, the corners are apt
to be heated long before the major part of the mass. If the temper-
ature is high, the corners are overheated before the center of the
mass is saturated. From this commonplace example there should
be indicated the necessity for slow, soaking heats in order to prevent
overheating the corners of the metal, and further, the necessity of
soft hazy heats to prevent oxidation or decarburization of the
exposed edges.

Temperature of Heating. — Assuming that the proper degree of
care has been used in heating the steel, the next question is the degree
of heat necessary. Reduced to lowest terms, the true or full anneal-
ing operation requires the production of an entirely new crystalline
structure, the constituents of which shall be of the smallest grain-
size attainable; this operation should also eliminate all internal
strains and stresses. As previously described, this new structure
is given birth at a temperature known as the " upper critical range "
of the steel. The exact temperature 1 will depend upon the chemical
composition of the steel, and, more particularly, upon the carbon
content. As this transformation does not occur suddenly, but
usually covers a range of some 25° to 50° it is customary to adopt
a temperature of about 50° over the upper critical range as the
proper annealing heat. For straight-carbon steels these may be
roughly given as shown in the chart in Fig. 37. The upper critical
range is approximately located by the dash line on the chart.

The temperatures recommended by the American Society for
Testing Materials 2 are as follows :

Range of Carbon Content.

Range of Annealing
Temperature.

Less than 0.12 per cent.
0.12 to 0.29
0.30 to 0.49
0.50 to 1.00

1607° to 1697° F.
1544° to 1598° F.
1499° to 1544° F.
1454° to 1499° F.

1 Methods for determining the critical ranges are described in Chapter
XVIII.

2 It will be noticed that the temperatures recommended by A. S. T. M. are
distinctly higher — especially for the tool-steel grades — than those advised by
the author. In the light of my own experience, and that of others, I believe
that the lower the temperature which can be used to give the desired results,
the greater will be the maximum efficiency of the annealed steel.

STEEL AND ITS HEAT TREATMENT

Length of Heating. — Ordinarily the underlying practice of this
part of the operation is to heat the steel until the whole mass has
been heated uniformly throughout at the proper temperature. This
will of course depend upon the size of the object. This full heating is
generally sufficient to give birth to the new grain structure and relieve
all internal stresses. The proper rate of cooling should then maintain
the steel in that condition. If the steel should be quenched in some
hardening bath such as oil or water, this new grain-size and rear-

iroo

1200

o.i

0.3 0.4 0.5 0.6

Carbon Content, Per Cent.

0.9

FIG. 37. — Annealing Range for Carbon Steels,

rangement of the structure would be kept. The annealing operation
should theoretically bring about approximately the same results as
to grain-size, neglecting for the moment the effect of slow cooling
through the transformation ranges. From the standpoint of prac-
tice, however, much difficulty is experienced in this regard, par-
ticularly in cases where the mechanical work upon the steel has been
severe, and also in alloy steels.

It seems that the greater the internal stress upon the steel the
greater is the amount of intermolecular lag or final release of this

ANNEALING 49

stress behind the actual change of constituents. That is, even though
a totally new structure may be set up by the annealing temperature,
there remains for a considerable length of time a tendency of the new

FIG. 38.— Frame Steel as Rolled. X60. (Bullens.)

structure to return, upon slow cooling, to the stressed condition of
the original, even though the constituents themselves may be those
born at the new temperature.

FIG. 39.— Frame Steel Partly Annealed. X60. (Bullens.)

This point is illustrated in Figs. 38, 39 and 40. These are
photomicrographs taken from tests made upon chrome nickel steel
plates for automobile frames : Fig. 38 shows the structure of the steel

50 STEEL AND ITS HEAT TREATMENT

as rolled; Fig. 39 shows the steel after a short annealing at a tem-
perature above the upper critical range; and Fig. 40 shows the same
steel after a long anneal at the same temperature. It will be noticed
that the steel in Fig. 39 has taken on approximately the same struc-
tural constituents as in the fully annealed piece as shown in Fig. 40,
but that it still remains in the "stressed condition of Fig. 38, even
though the annealing temperatures were the same in both cases.
It is important, therefore, if a soft steel, free from all internal strains
and stresses is desired, that a sufficient length of time be allowed for
the permanent elimination of these intermolecular strains, before and
after cooling. In the case of the steel plate just referred to it re-

FIG. 40.— Frame Steel Fully Annealed. X60. (Bullens.)

quired some twelve hours for the complete change or equalization to
take place!

" Milky-Ways." — We have previously explained this same
phenomenon under the heading of " Diffusion," as this is the scientific
principle underlying it. The reoccurrence or reformation of these
laminations or other stressed structures is due to the fact that the
complete effacement by equalization had not taken place. In other
words, it means that where these stressed areas occur the carbon
content as a whole is less than in the rest of the mass. Where ferrite
predominates, as in the lower carbon steels, there will the mass more
easily coalesce into what may be termed " milky-ways " (Howe).
In order to equalize the steel as a whole the length of time of the
sojourn at or slightly above Ac3 should be inversely proportional
to the time occupied in reaching that temperature.

ANNEALING 51

Alloy Steel. — Alloy steel is particularly an example of the re-
tarded transformation as described above, although the author has
repeatedly found it in carbon steels cold worked. Most notable of
the alloy steels exhibiting this peculiarity are chrome, chrome-
nickel and chrome-vanadium steels. Many users and even manu-
facturers of these steels contend that annealing will not give entirely
satisfactory results. The oft-encountered " hard-spots " would
seem to bear out this dispute. From the experience of the author
the proposition develops into a simple question of time. The alloy-
ing metals add to the density of the grain, so that a longer time is
needed to complete the change in entirety. It was found that a
certain 3-inch rolled-round approximating 0.50 per cent, carbon,
1.50 per cent, nickel and 0.50 per cent, chrome required sixteen hours
for this complete change, together with the elimination of hard-
spots, to take place ; high-carbon high-chrome steel often takes days
for a complete anneal.

Time of Heating. — Assuming a proper rate of heating, it therefore
remains to determine the required length of heating by means of
experimentation, taking into consideration such points as have been
mentioned above. Along these lines some interesting experiments
have been carried out by Mr. M. E. Leeds l for the determination
of the variations in rates of heating of specimens of different sizes
to various furnace temperatures and which in some degree answer
the oft- repeated question " How long shall we heat this piece of
steel? " The experiments were made with round specimens of nor-
mal open-hearth carbon steel approximating 0.5 per cent, carbon,
and ranging in size from 2 ins. to 12 ins. in diameter, by 24 ins. long.
Each specimen was heated to four temperatures, namely, 1000°,
1200°, 1400°, and 1600° F. During the time of heating a continuous
record was kept of furnace temperatures, the temperature of the sur-
face of the specimen, and of one to three points in its interior. While
the results obtained are necessarily of relative value only on account
of the varying furnace conditions which might be found elsewhere,
there are, nevertheless, several interesting conclusions of value
which were drawn from these experiments:

1. "Variation in Time of Heating with Size. — As would be
expected, the smaller specimens heat more rapidly than the larger.
In curves (Fig. 41) the relation between the size of specimen and
time of heating to various temperatures are brought out. Except
in a very general way, this information could not be used as a guide
1 M. E, Leeds, A. S. T. M., June Meeting, 1915.

STEEL AND ITS HEAT TREATMENT

to heating practice, as the rates would vary with the size of furnace
and probably with other conditions.

2. " Relation between Time of Heating and Furnace Tempera-
ture.— The time of heating for a specimen of any size is less when it
is brought up to 1600° F. than when brought up to 1200° F., and less

300

8"
Size of Section.

FIG. 41. Curves Derived from Rate of Heating.

thrup Co.)

(Courtesy of Leeds & Nor-

for 1200° F. than for 1000° F., although it is greater for 1400
than for any other temperature.

" It is more difficult to account for the fact that the higher
temperatures are attained more rapidly than the lower ones. This
fact, however, appears to be clearly demonstrated. It may be that

ANNEALING 53

the specimens received a large amount of their heat by radiation from
the furnace walls. The heat transfer by radiation between two bodies
at different temperatures is proportional to the difference between the
fourth powers of their absolute temperatures, and so for a 100° dif-
ference in temperature between furnace wall and test specimens,
at 1600° F., the heat transfer would be at a higher rate than for the
same temperature difference at lower temperatures.

3. " Relation between Surface and Interior Temperatures.—
From all of the curves, it is deduced that there is no large difference
in temperature of the points inside of the specimen. This was quite
surprising, as it was expected that the 12-in. specimen would show
considerable differences of temperature between a point 2 ins. from
the surface and the center.

" All of the runs show that the contact couple is at a higher
temperature than any of the interior couples until the specimen
has attained the temperature of the furnace. It cannot properly be
assumed that the temperature shown by the contact couple is exactly
that of the surface of the specimen.

" When the contact couple attains the furnace temperature, all
parts of the specimen have also attained that temperature. This
suggests a practical method of using contact couples in conjunction'
with furnace couples, namely, by means of the furnace couple the
furnace should be held at the temperature at which it is desired to
treat the specimen, and the contact couple should then be used to
determine when the specimen has assumed the desired temperature.

4. " Contact Couple Shows Time of Transformation. — The
curves (not given here) showing the heating of the 12-in. and 8-in.
specimens to 1400° and 1600° F. show that the transformation point
is clearly shown by the couples inside of the specimen, and that it
is also shown by the contact couple. The interior couples show,
with approximate correctness, the temperature at which the trans-
formation takes place. The contact couple shows a corresponding
flexure in its curvature, at the same time as the interior couples,
though not at the same temperature." The close correspondence
in time between the flexures of the contact couple and the interior
couples points to what Mr. Leeds believes is an important new
method of determining when a piece of steel has been heated
through its transformation point.

Rate of Cooling. — We know from our study of the previous
chapters that in hypo-eutectoid steels the solid solution rejects the
excess ferrite upon cooling through the transformation range. This

54 STEEL AND ITS HEAT TREATMENT

ferrite will form either a network around the grains of solid solution
or pearlite, or will coalesce into irregular masses, the same being
dependent upon the rate of cooling. A moderately slow cooling will
develop the cellular or network structure without breaking it up.
A very slow cooling will break up the network structure, giving
ample time for the ferrite to coalesce into large masses. The slower
the cooling through the transformation ranges, the greater also will
be the size of the grains.

Effect of Cooling. — The effect of the varied rate of cooling is
illustrated in the photomicrographs of a 0.45 per cent, carbon steel

FIG. 42. — Network Structure, 0.45 per cent. Carbon Steel. X100. (Bullens.)

shown in Figs. 42, 43 and 44, all taken at the same magnification.
All three pieces were heated to a temperature somewhat in excess of
the full annealing temperature. The steel of Fig. 42 was cooled quite
rapidly (air-cooled) ; that of Fig. 43 was cooled rapidly through the
upper part of the transformation range, but slowly through the lower
critical range ; that of Fig. 44 was cooled with the furnace. Thus we
have the network structure in the first case, showing a comparatively
small grain-size. In the second instance the network is coarse
and the pearlite is fairly well developed. A very slow cooling, as in
the third case, has resulted in a coalescence of the ferrite into large
grains, intermingling with the coarse pearlite. The ferrite in all
three photomicrographs is represented by the white constituent.

ANNEALING

Some very interesting facts might be drawn from a study of these
photomicrographs in comparison with those previously mentioned
in the series of Figs. 27 to 31, and Figs. 32 to 36.

FIG. 43. — Coarse Network Structure, 0.45 per cent. Carbon Steel.

(Bullens.)

X100.

FIG. 44. — Coalesced Ferrite and Pearlite, 0.45 per cent. Carbon Steel.

(Bullens.)

xioo.

STEEL AND ITS HEAT TREATMENT

Effect of the Rate of Cooling upon the Pearlite. — Not only does
the rate of cooling from the annealing temperature have a very great
effect upon the network and grain structure, but also upon the char-
acteristics of the pearlitic constituents of the steel. The rate of

MlCROSTRUCTURE.

SEGREGATION STAGES.

I. Sorbite or " sorbitic pearl-
ite." Cementite emulsi-
fied.

MECHANICAL PROPERTIES.

Tensile strength about 150,000

Ibs. per sq. in.
Elongation about 10% in 2 ins.

II. Sorbite passing into nor- Tensile strength about 125,000
malpearlite. Semi-segre- Ibs. per sq. in.
gated cementite. Elongation about 15% in 2 ins.

III. Finely laminated pearlite.

IV. Laminated pearlite. Com-
pletely segregated ce-
mentite.

Tensile strength about 100,000

Ibs. per sq. in.
Elongation about 10% in 2 ina.

Tensile strength about 85,000

Ibs. per sq. In.
Elongation about 8% in 2 ina.

Cementite — white
Ferrite — black.

V. Laminated pearlite pass- Tensile strength about 75,000

ing into massive pearl- Ibs. per sq. in.

i t e . Cementite and Elongation about 5% in 2 ins.
ferrite each coagulating

FIG. 45. — Pearlitic Development.

cooling through the lower critical range, at which the transformation
of the solid solution into pearlite is effected, will so change the
arrangement of the ferrite and cementite constituents of the pearlite
that widely varying physical results may be obtained in this manner.

ANNEALING 57

As we will explain later, the austenUe does not directly change into
pearlite, but passes through a series of transition constituents with
varying physical properties. The majority of these, however, are
not retained in the steel through methods of cooling other than
quenching (which may or not be followed by a reheating), so that
we need consider only the very last transition, sorbite. This com-
ponent sorbite represents the last stage of the transition austenite to
pearlite, and in which the individual particles of ferrite and cementite
are just on the verge of coalescing. Sorbite, or sorbitic-pearlite,
is noted for its combination of high tensile strength (i.e., in comparison
with the later phases of pearlite) and ductility. Sorbite is generally
formed by air cooling through the lower critical range, and is shown
in Fig. 45. This figure also illustrates the different phases of the
pearlite, together with their approximate physical characteristics.
From this it will be evident that the rate of cooling must necessarily
have a great influence upon the physical properties of the slowly
cooled or annealed steel, and that the operation must be adjusted
accordingly.

Definite Cooling. — Thus we see that having obtained the per-
manent release of all internal strains and stresses and brought about
the formation of an entirely new grain-size and structure by means of
proper heating, it now remains to adjust the physical properties by
means of regulating the rate of cooling. In general, there are three
methods of cooling as used in the annealing process. These are:
(1) Cooling in and with the furnace, (2) removing the steel from
the furnace and covering with some blanketing substance such as
lime, sand, ashes, etc., (3) cooling in air. Cooling by means of
quenching is not a true annealing operation, and will therefore be
considered under the subject of " Hardening."

Furnace Cooling. — Cooling in and with the furnace will generally
give the slowest cooling of the steel which is possible if the furnace
is of heavy construction and can be tightly closed. Furnace cooling
will give a maximum " softness " and ductility — that is, the tensile
strength and elastic limit will be at a minimum, and the elongation
and reduction of area will be large. Steel in this condition will be
in a suitable condition for ordinary machining, and will also have
the quality of resisting a small number of severe distortions.

Slow Cooling. — In cases where the objects are of large size, an
approximation of furnace cooling may be obtained by removing the
steel from the furnace and covering with some blanketing substance
and slow conductor of heat such as lime, sand or ashes. This will

58 STEEL AND ITS HEAT TREATMENT

also permit the recharging of the furnace for another heat without
loss of time.

Pit Annealing. — Where a large tonnage of steel must be annealed,
a pit lined with brick or concrete and suitably fitted with cover
plates is sometimes made. The hot steel is immediately delivered
from the annealing furance to the pit and covered with ashes. Cool-
ing by this method of pit-annealing is often slower than cooling in
the furnace itself if the latter is not properly constructed so that no
cold air can find its way in.

Size, of Object. — It is readily realized that the size of the object
has a great bearing upon the rate of cooling. Under the same con-
ditions a smaller object will cool much more rapidly and will there-
fore be harder and less ductile than a piece of considerable size.
The rate of cooling must therefore be proportioned to the size of
the object.

Air Cooling. — If the steel is removed from the furnace and allowed
to cool in air the physical properties will be proportional to the
dimensions of the piece and also dependent upon the carbon content.
Thin objects and those with high carbon content cannot stand so
rapid a cooling as thick and low carbon ones, lest their ductility be
too greatly sacrificed. In this regard the American Society for
Testing Materials recommends the following: " Thick objects with
less than 0.50 per cent, of carbon may be cooled completely in air,
of course protected from rain or snow. Objects with 0.50 per cent,
of carbon or more, and thin objects with from 0.30 to 0.50 per cent,
of carbon may be cooled in air if their cooling is somewhat retarded,
as, for instance, by massing them together, as happens in the case
of rails." This more rapid cooling will give great strength and
high elastic limit, but less ductility.

Combination Air and Furnace Cooling. — Besides the regular air
or furnace cooling there are a number of different combinations of the
two which have given great success in innumerable cases. We will
give them briefly as follows:

1. Heat to slightly over Ac3, air cool to just over Arl, return to

a furnace which is held at that temperature (about 1350° F.), heat

until uniform, and then cool slowly. The latter heating should not

be any longer than is possible. This method will tend to prevent the

formation of large amounts of free ferrite, but will affect the pearlite,

as there will be slow cooling through the Arl range.

(f*1*^. Heat to slightly over the Ac3 range, air cool to just under the

i Arl range, return to a furnace and heat at 1350° F. and slow cool.

ANNEALING 59

This method will effect a greater " toughening " if the temperature
has not been prolonged too greatly at the second heating.
& 3. Heat to slightly above Ac3, air cool to below Arl, return to a
furnace heated at a temperature slightly below Arl (about 1200° to
1250° F.), hold at this temperature until uniformly heated, and slow
cool. In fact, the last cooling may be made in the air if desired, as
there will be little or no change in cooling from under the lower
Critical range.

Fine Grain Annealing. — Of the three special methods given, the
third is the preferable, as well as the most uniform and certain in
results. By permitting the steel to air cool to a temperature below
the lowest transformation, advantage is taken of any " hardening
effect " or retardation in the transformation of austenite into a
conglomerate of pearlite and ferrite. This effect will increase
with the percentage of carbon and the smaller the size of the piece.
The reheating to a temperature below the lower critical range, if
not prolonged, will neither change the grain size nor allow, of the
coalescing of the excess ferrite or of the individual constituents of
the pearlite, but will form a mass of irresolvable and intermixed
pearlite and ferrite known as " sorbite." At the same time, however,
it will give the maximum combination of large ductility, good strength
and excellent machining properties. This method is of particular
value in the annealing of tool steels, in which it has given most
excellent results.

The main objection to both of the other two methods is that a
varying duration of heating above the lower critical range will cause
corresponding changes in the results, so that no absolutely definite
result capable of commercial duplication can be obtained. The
methods, however, find many applications in a general way, particu-
larly in steels of medium carbon content which have been severely
stressed by previous mechanical elaboration. The second method
especially will give the advantage of having had a double heating
through the lower critical range (besides the minimum grain-size
conferred by fairly rapid cooling from the upper critical range), and
thus breaking up the previous structure.

Double Annealing. — The next variable which may be used is
that of heating to a temperature considerably in excess of the upper
critical range, air cooling to under the lower critical range, and
reheating to slightly above the lower or upper critical range. As;
an example of this the author will cite a case which was success-
fully solved by this method. Certain medium-carbon steel plates

60 STEEL AND ITS HEAT TREATMENT

had been finished at a temperature considerably under the proper
temperature for hot-rolling and thus had been considerably stressed
— in fact, the ordinary annealing method would not relieve this con-
dition. The plates were put in a furnace with a car-bottom, heated
thoroughly at about 1700° F. (that is, considerably over the Ac3
range), air cooled until black, and then reheated and slow cooled
from a temperature slightly over the lower critical range. The
laminations occasioned by the rolling were entirely eliminated by
the high temperature, and their reformation prevented by the rapid
cooling in air. The second anneal then thoroughly softened the
steel and put it in good condition for the following forming opera-
tions. This steel might, of course, have been reannealed at the
Ac3 range (instead of the Acl range) and an even better product
obtained.

Tool Steel Annealing. — The annealing of hypo-eutectoid tool
steel may be broadly grouped under two headings, dependent upon
the initial condition of the steel and upon the results desired. Tool
steel which has been carefully hammered is undoubtedly strength-
ened by this mechanical elaboration; a full annealing — that is, heat-
ing at a temperature over the critical range — will entirely destroy
the results of the forging operation. If it is therefore desired simply
to anneal the steel in order to put it in suitable condition for machine
work — that is, to soften it and at the same time to retain the bene-
ficial effects of the forging — the annealing operation should be
carried out at a temperature less than that of the critical range, or in
the neighborhood of 1200° to 1250° F. On the other hand, if it
is desired to obtain the finest grain size possible, the maximum
softness, and to entirely eliminate any previous heating or forging
work, the annealing should be carried out at a temperature slightly
over that of the critical range, or in the neighborhood of 1400° F.,
dependent upon the composition of the steel in question.
k*. Protection of Steel. — One of the vital points in obtaining a satis-
factory steel after annealing is the protection of its surface. Steel
when heated beyond a low-red heat exhibits a great tendency to
oxidize or scale, this action increasing, in the presence of oxygen,
with the temperature and the length of time involved. This con-
dition will exist in furnaces operated so as to produce sharp heats,
instead of soft, slightly hazy, reducing atmospheres. Decarburiza-
tion to a depth of J to \ inch is not a rare occurrence where improper
combustion and heat application is the rule. If, due to poor furnace
design and worse operation, such conditions do exist •

ANNEALING 61

produce a clean surface it will be necessary to protect the exposed
surface of the steel in some manner. Tool steel is often annealed
by placing in a tube, packing carefully with charcoal, and then clos-
ing the ends of the tube with caps or luting with clay.

On the other hand, the prevention of oxidation or the scaling of
the metal during the heating process is a simple thing with the
proper furnace design and operation. Assuming such a design, if the
furnace is operated so as to produce soft, hazy heats such as we have
previously mentioned, there should be no occasion for packing the
steel in charcoal or other such substances. This statement is made
not as one of theory, but as one of actual practice. Under-fired fur-
naces are being run to-day on brass cartridge case work in which there
is less oxidation and decolorization of the metal than in other fur-
naces in which the metal is packed in charcoal; not only is a better
product being obtained, but at less operating copt. \~ „

Box-Annealing. — For the protection of larger masses or a number
of smaller pieces, " box-annealing " is often resorted to. This par-
ticularly applies to cases where a finished surface must not be
injured. The steel is placed in a rectangular pot or box made of
cast iron or of plates riveted together. This box may or not be
lined with some refractory substance such as silica brick. The metal
is then carefully packed with some material such as ground mica,
sand, charcoal, charred bone or leather, lime, etc. If the steel is
low in carbon a carbonaceous or carbon monoxide generating sub-
stance must not be used, for a slight case-hardening action would
take place. In the case of higher carbon steels, and especially of
tool steels, reducing agents may be used, although it is better to mix
the charcoal with clean ashes. Sand and ground mica are probably
the most satisfactory of the simple non-reducing, refractory materials.
The cover is then placed on the box and' the box with its contents is
charged into the furnace and given the proper degree and duration of
heating.. The box should be raised from the floor of the furnace so
that the hot gases may have opportunity for circulation around it.
When properly heated throughout, the box may be removed from the
furnace and allowed to cool to atmospheric temperature.

Stead's Brittleness. — We have previously stated that practically
no change occurs below the Acl range if no previous hardening of the
steel has taken place. The one exception is that of very low-carbon
steels and is due to the fact that steels very low in carbon behave
more like pure or carbonless iron, there being but small percentages
of cementite (and therefore pearlite) to influence the grain-size.

62 STEEL AND ITS HEAT TREATMENT

Upon heating such steels through the upper part of zone la (refer
to diagram in Fig. 13), a distinct coarsening of the ferrite grains
occurs, this being a function of time as well as of temperature.
Steels of such carbon held at say 1100° F. for a considerable length of
time will develop such coarsening of grain-size as to make the steel
unfit for commercial use if any degree of strength is required. This
phenomenon is known as " Stead's Brittleness." With steels of
greater carbon content the increased pearlite so operates upon the
molecular structure of the steel that practically no change occurs
until the Acl range is reached.

ANNEALING HYPER-EUTECTOID STEELS

Critical Ranges. — Strictly speaking, hyper-eutectoid steels have
two critical ranges: the A 1.2.3, at which — on heating — the pearlite
changes into the solid solution; and the A cm range, at which — on
heating — there is the final solution of the excess cementite— just as
in hypo-eutectoid steels the Ac3 range represents the solution of the
last of the excess ferrite. However, on account of the relatively
small proportion of free cementite in the ordinary hyper-eutectoid
steels, and also because there is a large increase in grain-size upon
heating to the Ac. cm range — the temperature position of the latter
increasing very rapidly with increase in the carbon content — the
Ac. cm range requires .but little practical consideration and the
majority of the annealing operations are more intimately connected
with the principal critical range Ac 1.2.3.

Commercial Annealing. — Similarly to hypo-eutectoid steels, the
annealing of high-carbon steels may have for its object any or all of
the following factors: (1) the release of internal strains and stresses
set up by previous operations, (2) the softening of the steel to place
it in a suitable condition for machining, (3) the entire change of
structure.

The first item may be accomplished by a simple reheating at
temperatures below those of the critical range. The second and
third items are more complex in their solution, as the form in which
the excess cementite may exist is one of the governing factors.

If the mass of the steel is in the sorbitic state, as may generally
be expected in the usual tool steel, satisfactory results (the softening
of the steel for machining, and relieving the internal strains) may be
obtained by an annealing at a temperature slightly under that of the
principal critical range, or at about 1250° to 1300° F. This heating

ANNEALING 63

should not be prolonged for such length of time as may cause the
excess cementite to coagulate, but only until the steel has been thor-
oughly and uniformly heated throughout.

On the other hand, if it is desired to obtain the complete change
of structure, and to refine the grain (previously coarse), it will be
necessary to heat to a temperature at least in excess of the Ac 1.2.3
range (about 1340° F.). For steels with a carbon content approx-
imating 0.9 per cent., such heating will accomplish the complete
change of structure and give the finest grain-size obtainable through
annealing. For steels with a carbon content considerably in excess
of the eutectoid ratio the annealing may be done at similar tempera-
tures, provided, however, that the excess cementite is more or less
in solution in the sorbite.

Incidentally, if the condition stated under (3) is desired, and it
will warrant the expense, the best method is first to oil quench from
a temperature somewhat over the Ac 1 .2.3 range, and subsequently
anneal at a temperature just below that range.

Normalizing. — If the steel to be annealed has the free cementite
existing as network or spines, which would make the steel difficult of
machining, annealing at the usual temperatures (Ac 1.2.3) will not
affect this cementite: it will simply refine the ground-mass. In
order to eliminate this free cementite, it will be necessary first to
normalize or quench the steel from a temperature above that of the
Ac. cm range. That is, air cooling from a temperature of say 1750°
or 1800° F. will not permit of the reformation of coagulated cement-
ite. The second annealing may then be carried out at a temperature
of 1375° with the refining of the grain size and complete softening of
the steel as a whole; this second heating should be just as short
as possible in order to prevent the reformation of the free
cementite.

Spheroidizing the Cementite. — The above method may be further
modified by reheating to a temperature slightly under the lower
critical range instead of over it. The objection to this is that the
steel will not be refined, but will possess the large grain size charac-
teristic of the high temperature. On the other hand, the lower
annealing temperature will entirely prevent the formation of the free
cementite as either spines or as a network. Instead, it will be found
that the excess cementite will be thrown out, under these conditions,
as little nodules or " spheroids " if the reheating temperature is just
about at the end of the lower critical range; or, under certain con-
ditions, the whole mass of the steel may be called " granular," if such

64 STEEL AND ITS HEAT TREATMENT

a term is permissible. Further reference to this spheroidal forma-
tion of cementite, as obtained by a double " quenching," is given
under Chapter VII. Spheroidal cementite in annealed steels may
also be obtained by very slow cooling through the end of the Arl
transformation: cementite in this condition is a great help in the
machining of high-carbon steels.

CHAPTER IV
HARDENING

Hardening. — Fundamentally, the operation of hardening in-
volves two operations of change in temperature : heating and cooling.
The function of the heating is (1) to obtain the best refinement, and
(2) to obtain the formation of the " hard " constituents of the steel.
Having done this, the steel must then be held in this condition by
very rapid cooling — that is, by quenching in some medium such as
water or oil. Associated with both the heating and rapid cooling
there must be as great a degree of uniformity as is possible.

Changes on Heating. — Steel, when properly hardened, should
show no trace of the original structure, such as coarse grain size,
network, unabsorbed ferrite (in hypo-eutectoid steels), or any other
peculiarities of untreated steel. If such are present in the hardened
steel it goes to prove that the operation was not properly carried out.
Further, if the structure of the steel has not been suitably changed or
developed by the heating operation, it most assuredly will not be
altered for the better by subsequent quenching. The.most that such
quenching can do is to retain the characteristics which the heating
has developed.

An attempt has been made graphically to illustrate these facts
in the chart in Fig. 46. Column 1 (at the left) represents a normal,
0.4 per cent, carbon, pearlitic steel (at the bottom of the column),
and the structural changes taking place in that steel as it is pro-
gressively heated through and beyond the critical ranges. For the
present ,it is assumed that the structure and micrographic constit-
uents obtained by heating to various temperatures, such as A to E,
may be retained by quenching, as illustrated by columns II to VI.

Thus heating to a temperature A, under that of the lower critical
range, will produce no change in the original steel, which consists of
pearlite (the cross-hatched circles) and ferrite (the black area).
The quenching likewise will produce no change, as is illustrated by
Column II.

Heating to a temperature B, slightly over the lower critical range,
will change the pearlite to the solid solution (represented by

STEEL AND ITS HEAT TREATMENT

the dotted area), but without affecting the free ferrite. Quenching,
column III, will therefore produce a semi-hardened steel — since the
solid solution is the " hard " constituent — with a refinement of the
" ground-mass " (the original pearlite) only.

Heating to a temperature C, between the lower and upper critical
ranges, will effect a progressive absorption by the solid solution of

FIG. 46. — Changes in a 0.4 per cent. Carbon Steel on Heating and Quenching.

the remaining free ferrite. Quenching, column IV, will therefore
produce a " harder " steel than in case III, but nevertheless without
complete refinement of the steel as a whole.

Heating to a temperature D, slightly over the upper critical
range, if prolonged for a length of time sufficient to effect complete
diffusion and equalization, will entirely refine the steel, giving it the
smallest grain size possible. Quenching, column V, will retain this
condition and give the maximum hardness possible.

HARDENING 67

Heating to a temperature E, considerably over that of the upper
critical range, will tend to increase the grain size; and quenching,
column VI, will retain this condition, giving a more brittle steel.

Relation of Hardening to Annealing. — Thus it will be seen that
during the heating operation the changes taking place in the micro-
scopic constituents and the structure as a whole are similar in both
hardening and annealing. The main difference in the final results of
the two processes is due to the rate of cooling through the critical
ranges, and, therefore, upon the nature of the micro-constituents
which are thereby retained in the steel when cold.

The effect of slow cooling through the critical ranges, which is
characteristic of true annealing, has been discussed; in brief it may
be said that the austenite or solid solution shifts its carbon content
through generating pro-eutectoid ferrite (or cementite) to the eutec-
toid ratio of about 0.85 to 0.9 per cent, carbon, and then transforms
with increase of volume at Arl into pearlite, with which the ejected
ferrite (or cementite) remains mixed. This change or decomposition
of the austenite, however, does not take place suddenly or spas-
modically, but develops by stages; and that these intermediary
stages between austenite and its final constituents may be recognized
and identified under the microscope as martensite, troostite, osmond-
ite and sorbite is generally accepted. Hardening is but the result
of obstructing this transition, thereby retaining in the steel the
" hard " austenite or its early decomposition products martensite
or troostite.

Austenite. — Austenite is only obtained with difficulty in the
ordinary carbon steels, and even then is usually decomposed in part
into martensite. The two agents 1 — rapid cooling and carbon —
tending to obstruct this transition must be grouped in suitable pro-
portions— that is, the carbon content must be high, and the cooling
take place with extreme rapidity. With about 1.5 per cent, carbon
steel, such as is generally used in corrugating and roll-turning tools,
when quenched in brine or very cold water from about 1400° F.,
about one-half of the austenite will remain unaltered. When the
carbon is about 1.1 per cent. — which may be regarded as about
the minimum limit, although the author has succeeded in obtaining
some austenite with water-quenched 0.9 per cent, carbon steel —
the cooling must be done in iced solutions from a temperature of
1800° F. or more.

1 Alloys are also obstructing agents in the sense that, if present in the proper
amount, they lower the temperature at which the transition will commence.

68 STEEL AND ITS HEAT TREATMENT

The hardness of austenite, as preserved in hardened high-carbon
steels, does not fall very far short of that of the accompanying mar-
tensite, probably because the austenite is partly transformed into
martensite in cooling. On the whole, however, austenite may be
regarded as being considerably softer than martensite, and also much
tougher; the austenite as obtained in high manganese and high nickel
steels is but moderately hard.

Martensite. — Martensite is the chief characteristic constituent
of hardened carbon steels when cooled rapidly in water from a tem-
perature above the A3 range. In very high-carbon steels, rapidly

FIG. 47. — Martensite. X75. (Ordnance Dept.)

cooled, the martensite is associated with austenite. In the lower
carbon steels hardened in water, in high-carbon steels hardened in
oil, or in thick pieces of high-carbon steel hardened in water, mar-
tensite is associated with troostite and with some pro-eutectoid
ferrite or cementite.

Of the transition constituents — austenite to pearlite — martensite
is the hardest and also the most brittle, having extremely high tensile
strength with little or no ductility. Microscopically martensite is
characterized by a needle-like structure as is shown in Fig. 47.

Troostite. — Troostite is obtained by cooling through the trans-
formation range at an intermediate rate, as in small pieces of steel

HARDENING 69

when quenched in oil, or quenched in water from the middle of the
transformation range, or in the center of larger pieces quenched in
water from above the critical range. The early appearance of troos-
tite in tool steel is shown in Fig. 48.

The hardness of troostite is intermediate between that of the
martensitic and pearlitic state corresponding to the carbon content
of the specimen. In general, the hardness increases, the elastic limit
rises, and the ductility decreases, as the carbon content increases.

» H '" j -•'> '.

v4vr ><

*^ *

NW >

••••••H

v-v N 7'.^

FIG. 48.— Troostite (Dark) in Hardened Carbon Tool Steel. X100. (Bullens.)

Sorbite. — Sorbite, when obtained by hardening, is ill defined and
almost amorphous; it is 'softer than troostite for a given carbon
content. Dependent upon the carbon content, sorbite may be
obtained by quenching small pieces of steel in oil or in molten
lead, or even by air-cooling them; or it may be obtained by quench-
ing in water from just above the bottom of the Arl range. Sorbite,
and to some extent, troostite, are more characteristic of tempered

STEEL AND ITS HEAT TREATMENT

steels than of hardened steels. The transformation of troostite into
troosto-sorbite is shown in Fig. 49.

Temperature for Hardening. — As a general rule, hardening is
carried out from a temperature of about 50° F. above the line A3-A2.3
-Al.2.3 in Fig. 13. This is done to obtain the best refinement
of the steel as well as the maximum hardening effect. Rapid cooling
of medium and low-carbon steels from a temperature just above the

FIG. 49,— -Trooste-Sorbite (Dark), X100. (Bullens.)

bottom of the critical range Al, will not bring out the maximum
hardening effect. The general temperatures most applicable for
individual steels are given in subsequent chapters.

When the maximum results, both as affecting the structure and
also the physical results, are to be obtained, experiments should be
made to determine exactly how far over the critical range the steel
should be taken. In some steels it will be found that approximately
50° F. will accomplish this purpose; in others it may be necessary

HARDENING 71

to raise the temperature to even 150° F. or more. This effect will
also be influenced by the size of the piece to be hardened, as will
be shown under " Tool Steel."

Heating for Hardening. — The general rules for heating for
hardening may be simply stated, but their fullest comprehension and
application may be obtained only in the light of experience. This
heating requires much more care than heating for annealing (if such
be possible) , on account of the diametrically opposite functions which
are indicative of the two operations. Heating for annealing is
followed by slow cooling and the gradual release of all stresses
and strains: heating for hardening is followed by the most severe
test to which steel can be put — very rapid cooling, accompanied
by the setting up of a condition of stress and strain. In general,
we may say that the heating for hardening should be slow, uniform,
and thorough, and to the lowest temperature which will give the
desired results.

Non-uniformity in heating must of necessity result in lack of
uniformity in cooling, which in turn is the genesis of most of the
troubles in the hardening process. Hardening cracks are more
often the result of uneven heating than of a defect in the steel.
Heating requires time and care. The peculiarities of each steel and
article must be thoughtfully studied; experiments must be made;
and the clear judgment of experience applied to each individual
case. It has been well said that " Steel is mercurial and delicately
responsive to heat; its records appear in its own structure."

Lowest Quenching Temperature the Best. — The lowest heat
which will give the results desired should always be used in hardening.
This point can be brought home in no better way than to give the
results of two tests made which illustrate exactly this principle.
Two automobile gears were made from the same bar, by the same
man, and in all other ways as nearly alike as possible. The tests
were made by a disinterested third party.

Number 1 was quenched in oil from 1450° F., annealed at 1400°,
hardened in oil from 1450°, and tempered at 475° in oil. It gave a
sclerescope hardness of 76 to 78. It withstood 48 blows of a 10-lb.
hammer dropping 30 ins. before a tooth could be broken out, or 8
blows of a 10-lb. hammer dropping 48 ins.

Number 2 was quenched in oil from 1400°, which was just
over the critical range, and determined by when the magnet " let
go." It was annealed at just under that temperature, followed
by hardening in oil from 1400° and tempering in oil at 475°. The

72 STEEL AND ITS HEAT TREATMENT

hardness was the same as with No. 1. In this case, however, it
required 200 blows of the 10-lb. hammer falling 30 ins., or 78 blows
with a fall of 48 ins.

The effect of the increase of only 50° in the hardening temperature
is self-evident: it meant a difference in efficiency in the ratio of 48
to 200.

Temperature of Quenching. — It is not only the uniformity of
heating of the steel object which is necessary for uniform and proper
hardening, but also — and equally important — the uniformity of
temperature of the piece at the moment of quenching. A piece of
steel may be properly heated at the moment previous to its with-
drawal from the furnace; but that same piece may have wide dif-
ferences of temperatures in different parts of the mass at the moment
of quenching. Non-uniformity of heat saturation at the latter
instant must inevitably result in non-uniformity of hardening —
with the attendant possibility of warping, cracking and similar fea-
tures. Whether or not the indirect cause of such a condition is
due to the shape and size of the object or to the method of handling
the stock between the furnace and bath is immaterial as far as the
basic principle outlined above is concerned. These are but a mani-
festation of the ever-present "personal element."

Overheating. — Overheating is probably one of the most com-
mon sins of the hardening shop. Unfortunately, many "practical "
men still believe in the efficacy of high temperatures for greater
hardening effect. Although this may — to a very limited extent —
be true, the weakening of the steel by the increase in grain size and
greater hardening strains as obtained by high temperatures more
than offset the questionable production of greater hardness. Fully
80 per cent, of the complaints of " bad steel " which have been
brought to the author's attention have been the direct result of over-
heating. Both theory and practice support the old rule that " the
lowest heat which will give the desired results is the best heat."

The Magnet in Hardening. — It will be recalled that steel becomes
non-magnetic (for all practical purposes) in passing through and
beyond those temperatures represented by the heavy black line in
Fig. 50. For steels with about 0.35 per cent, carbon and upwards
this temperature line also corresponds to the best refinement of the
steel in heating. We therefore have a very simple and practical
means of determining the proper temperatures for hardening such
steels. All that is necessary is to apply an ordinary horse-shoe
magnet, suspended from a suitable rod, against the hot steel. When

HARDENING

the correct hardening temperature has been reached there will be
no attraction between the magnet and the steel. For steels with less
than about 0.3 per cent, carbon the rise in temperature between the
Ac2 and Ac3 ranges may be estimated and the steel hardened when
it has reached the latter temperature. It will be found that the use

1800
1700
1600
1500
1400
1300
1209

0.2

O.i

0.6

0.8

1.0

1.2

FIG. 60. — Carbon-iron Diagram Showing Temperatures at which Ordinary-
Carbon Steel Loses Its Magnetism on Heating.

of the magnet will be of great value to those not having the proper
pyrometer control over their heating operations and who have to
depend entirely upon the eye for gauging such. An instrument
more convenient for this purpose than the ordinary magnet is made
by magnetizing a small, elongated, diamond-shaped piece of steel,

74 STEEL AND ITS HEAT TREATMENT

and supported between two pins in the end of a forked rod, as is
shown in Fig. 51.

Motion during Hardening. — When a piece of steel is quenched,
movement should be given to the steel, to the quenching medium,
or to both. This is for two reasons: (1) in order to cool the steel as
rapidly as possible, and (2) to break up any tendency toward the
formation of a distinct line between the hardened and unhardened
parts of a differentially hardened forging. The first factor should be
self-evident: agitation of the bath or movement of the hot steel
will lower the temperature of the oil or water which is cooling the
steel, will prevent the formation of vapor or steam around the steel,
and in other ways more rapidly cool the metal. In the second place,
if a piece of hot steel, such as a chisel or die-block, were to be immersed
to a certain point and held there quietly, the rapid cooling would
harden the steel up to the point of immersion and no further; in other
words, there would be a sharp line of demarkation between the hard-

FIG. 51. — A Magnet Used in Hardening.

ened and unhardened parts, and which in turn would be a source
of great weakness and possible fracture. So, in quenching, move
the piece up and down in the bath; or if it is to be only partly
immersed, agitate the bath so that there will be no distinct line of
hardening: avoid straight-line hardening.

Furnace Equipment. — The general principles of heat application
will be discussed elsewhere. Some of the main points to be con-
sidered are uniformity of heated product, quality, and economic
efficiency, and to conduct all operations with these in view. The
material to be heated should be so arranged in the furnace that there
is ample room for the circulation of the heat through the mass. The
furnace should be so designed as to suit the class of work to be heated,
and so operated that the heat shall be evenly distributed and of the
same temperature from floor to roof and side to side. Only under
special circumstances, such as in connection with automatic furnaces,
is it desirable to operate the furnace at a temperature higher than the
maximum temperature desired in the steel.

The Human Element arid Basic Heat Treatment Conditions. —
What is it or who is it that determines when the charge in the cham-

HARDENING 75

her is saturated with heat to the temperature indicated by the
pyrometer?

What determines the manner in which the charge is placed in the
furnace and the room for circulation throughout the mass?'

What determines when the bottom and center of the mass are
at the same temperature as the top and outside?

What regulates the flow and composition of gases in the chamber
around the stock and the discharge of heat from the chamber?

What determines if each piece is heated like every other piece
and is uniform throughout, and whether each piece goes into the
quenching bath at the same temperature as all the others and at the
temperature indicated by the pyrometer?

What is it that controls the flow of air into the furnace or the flow
of gases from the furnace, and by so doing determines whether the
atmosphere surrounding the stock is oxidizing or neutral, and
whether the fuel is conserved or wasted?

These are some of the elements that affect proper heat treat-
ment, and they are determined, not by a pryometer nor a furnace nor
by similar apparatus nor by any mechanical means, but by the same
methods that govern the quality of the products of the kitchen —
the judgment and skill of the operator.

Heating Baths. — Despite the efficiency of design of many furnaces,
— and not mentioning those of poor design — their often inefficient
operation has a general tendency towards non-uniformity in heating
and oxidation. In the effort to solve these two problems at once —
that is, to surround the object to be heated with a constant and uni-
form heat on all sides, and to avoid contact with the air — the appli-
cation of various molten baths has come about. Chief of these heat-
ing mediums are molten lead and certain salts. On account of the
operating cost and necessarily small capacity, however, their use
is largely confined to the heating of tools and other articles requir-
ing particular care, uniformity, and freedom from oxidation. The
principaKuse for such baths, in the author's opinion, should be for
the retention of a bright surface on the metal after hardening,
and not for uniformity in heating; any furnace, properly designed
for the work in hand, heated with the right fuel, and correctly oper-
ated, should give entire uniformity of heating.

Heating in Lead. — Before the advent of the modern heat treat-
ment furnace, heating in molten lead represented the most practical
method of obtaining uniform heating. With a reasonable amount
of care and attention the typical lead bath may be maintained at

76 STEEL AND ITS HEAT TREATMENT

such temperatures as the ordinary hardening operation requires and
with a satisfactory degree of uniformity. Its use, however, presents
many difficulties. The bath must be frequently agitated to preserve
a uniform temperature. When heated to over 1200° F. lead begins
to volatilize, giving off fumes which are both offensive and poisonous ;
suitable ventilation, such as may be obtained with a properly de-
signed hood, should be provided to remove these fumes. Further,
the bath must be covered with powdered charcoal to reduce the oxides
or dross which are formed in the molten lead. Many plants will not
use lead baths if temperatures greater than 1475° or 1500° F. are
necessary. On account of its high specific gravity, heating in lead
requires some method of holding the steel beneath the surface, as
otherwise the tool would float on the surface of the bath and thus be
unevenly heated. One of the most troublesome difficulties with
lead baths is the tendency of the lead to stick in the holes, threads,
or even to the surface of the tool when it is removed for quenching,
so that uniformity of cooling is sometimes materially affected. Al-
though this particular difficulty has been largely eliminated by the
use of a paste, the trouble may simply be aggravated in case this
coating has not been carefully and properly applied.

Salt Baths. — Many of the difficulties encountered in the use of
lead for heating may be overcome by the substitution of different
salts. Their lower specific gravity permits of a more uniform
circulation and there is no tendency of the tool to float on the sur-
face of the bath. At the usual temperatures used for hardening
there is little or no vaporization. Although lead may prevent the
steel from oxidation while the steel is being heated, as soon as the
tool comes in contact with the air on removal from the lead bath,
a thin film of oxide is formed; with the salt bath, on the contrary,
the steel receives a thin and uniform coating of molten salt, which
protects the surface of the metal.

The minimum temperature of the salt bath may be very closely
estimated without the use of a pyrometer. Common table salt has
a freezing-point of 1472° F., and if it should be melted with potassium
chloride (freezing-point 1325° F.) or other salts, the freezing-point
of the melt may be quite accurately adjusted over a wide range of
temperatures. By keeping the bath very near its freezing-point by
a suitable regulation of the heat, overheating of the steel may be
entirely overcome. Further, if the composition of the salt bath has
been so adjusted as to approximate the proper hardening temper-
ature, when the steel is removed from the bath it may be quenched

HARDENING 77

just at the time when the salt film begins to solidify — or at exactly
the correct temperature.

BATHS FOR QUENCHING

General Properties of Quenching Media. — The main thought in
selecting a proper bath for quenching is the rapidity with which the
heat is removed from the hot steel. This property of transference or
withdrawal of heat from the solid by and to the liquid, will depend
upon the specific heat of the liquid, its conductivity, viscosity and
volatility. That is, the specific heat will indicate the heat-absorptive
power of the liquid; the conductivity will measure its capacity for
transferring the heat thus absorbed to the cooler part of the bath ;
the viscosity affects the motion of the liquid and thus influences the
uniformity of cooling; and the volatility indicates the temperature
at which the liquid will become gaseous, thus forming a vapor around
the steel and preventing the quick removal of the heat from the
steel. By obtaining a suitable combination of these various prop-
erties a bath giving the desired effect may be obtained.

Temperature of the Bath. — The continuous use of any bath for
quenching will gradually and progressively raise the temperature of
the liquid used. As a general rule, differences in the temperature of
the bath will give rise to varying results in the actual hardening taking
place — the higher the temperature of the bath, the less its cooling
efficiency. This is especially noticeable with water; a change of
50° or 100° F. will often entirely alter the physical properties of the
quenched steel. The effect is less marked with oils, and with some
oils may be almost negligible for certain classes of work. On the
whole it is decidedly better practice to maintain as nearly as possible
a standard temperature in the quenching bath.

Quenching Speed of Different Media. — It is evident that the
cooling medium used, its temperature and condition will affect the
rate of cooling. Matthews 1 and Stagg have devoted considerable
time to investigating numerous commercial media which are in use
in typical hardening plants of the country at the present time.
Their method was as follows : A suitable test piece was machined from
a solid bar, and a hole drilled through one end to within an equal dis-
tance from each side and bottom of the test piece. Into this hole
a calibrated, platinum-rhodium couple was inserted and the leads
connected to a calibrated galvanometer. The test piece was then

i Matthews and Stagg, " Factors in Hardening Tool Steel," A. S. M. E., 1915.

78 STEEL AND ITS HEAT TREATMENT

immersed in a lead pot, and the lead pot was maintained at a tem-
perature of 1200° F. When the couple inside the test piece was at
1200° F., and the couple in the lead pot also read 1200° F., the test
piece was removed and quenched in 25 gals, of the quenching medium
under consideration. At the start the quenching medium was at
room temperature. The time in seconds that it took the test piece
to fall from a temperature of 1200° F. to a temperature of 700° F.,
was noted by the aid of a stop-watch. It is clear that immersing the
test piece in the quenching medium raised the temperature of the
medium. The test piece was then replaced in the lead, heated to
1200° F., quenched into the medium at this higher temperature and
the time again taken with the stop-watch. These operations were
continued until the quenching medium, in the case of oils, had
attained a temperature of about 250° F. The results obtained, time
in seconds, for a fall from 1200° F. to 700° F., were plotted against
the temperature of the quenching medium and a series of curves as
shown in Fig. 52 l were obtained.

The various curves represent the following quenching media :

W. Syracuse city water.

B. Brine.

Sec.

1. New fish oil; average of readings from 80° to 250° F. . 85

2. No. 2 lard oil 87

3. Prime lard oil in use two years 99

4. Boiled linseed oil 101

5. Raw linseed oil 102

6. New extra-bleached fish oil 106

7. New yellow cottonseed oil 107

8. New tempering oil; 60% cottonseed, 40% mineral. . . . 122.6

9. New mineral tempering oil 130

10. No. 1 dark tempering oil 157 . 3

11. Special " C " oil 164.7

A consideration of the results is interesting. Pure water (curve
W) has a fairly constant quenching rate up to a temperature of 100°
F., where it begins to fall off. At 125° the slope is very marked.
Brine solutions (curve B) have both a quicker rate of cooling and
are more effective at higher temperatures than water. The curve
does not begin to fall off seriously until a temperature in the
neighborhood of 150° is reached. Where water at 70° cooled the
test piece in 60 sec., the brine solutions cooled it in 55 sec.

1 For the sake of brevity and clearness the numerous curves as given by Mat-
thews and Stagg have here been grouped under one plot.

HARDENING

As is well known, the oils are slower in their quenching powers
than water or brine solutions, but the majority of them have a
much more constant rate of cooling at higher temperatures than water
or brine. The curves shown in 10 and 11 are for thick viscous oils
similar to cylinder oils. These curves are particularly interesting
in that they have slower quenching abilities at low temperatures than
at higher temperatures. A comparison of curves 2 and 3 shows the
variation in quenching power of the same oil due to continued ser-

100 150

Time jaSeconds

FIG. 52, — Quenching Power of Liquids.

vice. The differences in quenching rates may well account for
different results from the same steel in different shops, or in the same
shop due to change of oil used.

Water Spray for Hardening. — Water sprayed under pressure is
the quickest agent for rapid cooling in common use, exceeding in its
hardening qualities either brine or water baths. The main point to be
noted is that there shall be sufficient volume and pressure to prevent
the formation of a blanket of steam between the hot steel and the

80 STEEL AND ITS HEAT TREATMENT^

spray. Its most common use is for such tools as sledges and others
requiring a differential hardening, and for armor plate.

Brine. — Brine is used only in certain particular lines, such as file-
hardening, for which an extremely hard surface is required. Unless
the steel has been most carefully heated, and is of a proper chemical
composition, quenching in brine is almost certain to crack the steel.
This is particularly true of large sections, for in these the very sudden
cooling of the outer surface, while the center is still hot, will set up
stresses and strains which will not be relieved or equalized in the
short time allowed, and with the inevitable results.

Water Quenching. — The author is a firm believer in the use of
oil for quenching, rather than water, and would recommend its use
whenever conditions permit. Water cools the steel more rapidly, but
its more drastic action increases the internal strain and consequent
liability to fracture. For the low-carbon steels, and for small
and comparatively simple sections of the higher carbons, water
quenching may be used without much danger. Of course in cases
where it is required that the surface shall be glass-hard, or that
the maximum tensile strength be obtained, water quenching is man-
datory. On the other hand, if the steel is to be given a full heat
treatment (i.e., quenching and toughening), the difference in hard-
ness as obtained by the two baths may usually be nearly equalized
by using a lower drawing temperature for the oil-quenched piece;
that is, if a 0.40 per cent, carbon steel forging is quenched in water
and toughened at say 1200° F., approximately the same static prop-
erties may be obtained by oil quenching and a subsequent reheating
to say 1050° F. The principal objections to the latter method are
that the lower drawing temperature is not so easily recognized by its
color, nor will the dynamic properties probably be quite as high —
though this last point is questionable. Generally speaking, however,
oil quenching is more desirable than water quenching.

Oil Tempering. — The term " oil tempering," referring to the
quenching in oil, is one which has become current in the trade, so
that the term, " hardening " often refers to quenching in water only,
or in some medium which will give an equivalent or greater hardness.
Strictly speaking, the use of " tempering " in this sense is a mis-
nomer, for it should be used as indicative of a slight reheating or
" softening " of the quenched steel.

Special Quenching Methods. — It often happens that especially
high tensile results are desired in certain large forgings of such size
and chemical composition that direct quenching in water is deemed

HARDENING 81

unwise, and yet in which it is desired to obtain as near the maximum
effect of water cooling as possible. A method which has proven in a
large measure successful is to use a bath of oil resting upon an equiva-
lent or greater volume of cold water. The forgings, when heated
to the proper temperature, are lowered into the oil for a few seconds
and thence into the water. The oil forms a film on the surface of the
steel, so that the sudden effect of the water is somewhat diminished
or retarded. The rapidity of cooling may be controlled by the dura-
tion of the oil quenching. It is obvious that in using this method
there must be a sufficient volume of water under the oil to prevent
the formation of steam and its consequent danger.

For small tools or thin instruments such as saws, the above
method may be so modified as simply to have a film of oil upon the
surface of the water, the oil in this case consisting of some animal
or vegetable oil. The heated tool is plunged directly and evenly
through the oil film so that it enters the water with a thin coating
of burnt oil which protects it from the direct action of the water and
lessens the risk of fracture. The amount of oil may of course be
increased as desired. The main objection to these methods is the
lack of uniformity in hardening unless the operator has had more or
less experience.

A method which is extensively used in some tool works is that of
using a combination of water and oil quenching, that is, first plunging
the tool into water until a certain amount of heat has been removed,
and then transfer to the oil, where it remains until cold.

Molten lead is sometimes used as a quenching medium for small
sections in which great toughness and only a moderate degree of hard-
ness is desired. Although dependent upon the carbon content, steel
subjected to this process will generally be sorbitic. Such treatment
will require no further reheating.

Other Aqueous Quenching Media. — Hardeners, at one time or
another, have tried about everything under the sun in the attempt
to discover some new and wonderful quenching medium which would
accomplish the phenomenal. The results, for the most part, do not
warrant the addition of expensive chemicals; and if the experi-
menters do claim the marvelous, the " gold-brick " scheme is gen-
erally revealed by thorough investigation.

Some substances, such as lime, soap, etc., may be added to form
a protective coating around the steel. Calcium chloride will raise
the boiling-point of water to a considerable degree, so that the solu-
tion may be used at a temperature up to 150° or 175° F. without

82 STEEL AND ITS HEAT TREATMENT

danger, and at the same time give many of the advantages which
oil hardening possesses. Some salts increase the hardening effect of
water; others purify the water or soften it. One of the most inter-
esting (and wonderful?) combinations which has come to the author's
attention contained — by addition— ammonia, glycerine, sal-ammo-
niac, spirits of nitre, ammonium sulphate, alum and zinc sulphate!

Differential Hardening. — In certain tools, such as anvil faces,
die blocks, edge tools, and the like, it is desired to obtain a very hard
outer part, surface or edge, to be " backed " by a less hard and
tougher steel. That is, the steel is gradually and progressively to
change from extreme hardness to the opposite, or what we may
term differential hardening. This phase may be obtained either
by heating the whole mass of the steel — as in die blocks, or by heating
only part of the article — as in chisels; in either case that part which
is to have the greatest hardness is immersed or quenched. By this
method the heat is gradually withdrawn from the part not immersed
through that part which is being subjected to the cooling bath, so
that the mass of steel as a whole will become progressively softer
or tougher from the hardened face or edge to the opposite side.
Precautions must be taken to avoid straight-line hardening,

Cooling the Water Bath. — Where water is used as the quenching
medium it is customary to maintain a flow of fresh, cold water into
the quenching tank so as to keep a uniform temperature and purity.
Water which has been used for any length of time without renewal
goes " stale " with a corresponding loss in cooling efficiency. If the
cost of water is such that it is inadvisable to dispose of the overflow
from the tank, the hot water may be cooled by spraying, cooling
towers, etc., aerated, and then returned to the tank.

Cooling the Oil Bath. — The common methods for cooling the oil-
quenching bath may be broadly classified as follows: (1) The cir-
culation of cold water around, or through coils in the bath; (2) the
circulation of the oil itself; (3) by the use of compressed air.

One of the simplest methods for cooling the oil when in small
tanks and not too constantly used, is to place the oil tank within a
larger tank, with a space of say 2 to 6 ins. between the two tanks.
This space is kept filled with cold water. As in all these systems,
the intake should be at the bottom of the tank, with the outlet or
overflow at the top. The main objection to this method is the fact
that the heat in the oil must penetrate through the walls of the tank
before it can be conducted away by the water.

The next type of cooling makes use of coils or radiators placed

HARDENING

within the oil tank and the circulation of cold water through these
pipes. These water lines are placed close against the side of the
tank so that they may not interfere with the work being treated.
From his own experience, the author does not feel that the radiator

J L

a! ' rj-

r
a i

01 I n-

J-n

ft-

FIG. 53. — Radiator Type of Cooling System.

type as shown in Fig. 53 gives as great efficiency as the simple coil
system of Fig. 54. With the difference of temperature of the oil
in the bottom of the tank, as contrasted with the hotter oil at the
top, it is difficult to obtain a thorough circulation of the cold water

1 N

1 1

1 D )

„ L^

( C II

1 r

V 1

1 1

i p )

/I - .

' - [y

1 S i 1

i r

A u

' ?J

(t- n-

\ j

\j — ....

_ L

FIG. 54. — Coil Type of Cooling System.

through all sections of the radiator. Further, this same difference in
temperature has the tendency towards unequal expansion of the top
and bottom pipes, which may cause a leakage of water into the oil
and its attendant dangers. In the coil system there is of necessity
a complete circulation, together with the elimination of expansion

84 STEEL AND ITS HEAT TREATMENT

dangers. These pipes vary in size from about 1 J ins. to 3 ins. diam-
eter; the latter size has given excellent satisfaction in a tank
approximately 8 ft. wide by 16 ft. long and with a working
capacity of about 8000 gals, of oil. Guide strips should be placed
at intervals along the coils — from top to bottom — to prevent any
articles from catching against the pipes while the quenched material
is being raised out of the tank.

Circulation and Cooling of the Oil Itself. — The best results for
keeping down the temperature of the oil bath are undoubtedly to be
had when the oil itself is circulated. The circulation is continually
bringing cold oil into the vicinity of the hot metal, removing the hot
oil from the tank, as well as giving a more uniform temperature to
the bath as a whole. In the previous systems the heat must be
taken away by gradual and progressive transference from the region
of the hot steel towards the sides of the tank, and at the best is a
slow procedure—this is assuming that the oil is not kept in motion
by compressed air. In the present system, the heat is taken away
from the quenching bath by the actual removal of the hot oil itself.

The usual methods are to pump the hot oil from the tank and then
through coils which are cooled by suitable means; or by maintaining
large supply tanks in which the oil will have sufficient time to cool
before being returned to the quenching tank. In the former pro-
cedure the coils containing the hot oil may be cooled by refrigerating
— such as the ammonia process, etc. — or by placing the coils in a
water tank, or by cooling the coils with a continual stream or spray of
water. Where the size of plant will permit the installation of a
refrigerating system, such a method is by far the most satisfactory;
the heat may be removed very quickly, and the temperature of the
oil controlled at any desired temperature by the regulation of its
flow through the cooling coils.

As an example of the water-bath method, one steel company
pumps the oil from the quenching tank — holding some 12,000 gals. —
through 3-in. pipes and thence through coils placed in a large water
tank used for the mill supply. The cold oil then returns to the
quenching tank by gravity.

For smaller plants the coils may be most conveniently cooled
by the use of tiny streams of water trickling over the coils. On the
whole, this is probably the most satisfactory system of all for small
plants. In one case (in which the question of the cost of water
was important) this method was found to be both cheaper and to give
a higher cooling efficiency than could be obtained by setting the coils

HARDENING 85

in a small water tank. In the latter case the heat is removed by
transference from one part of the water to that further removed
from the coils, so that unless a very good flow is maintained, the
cooling will be comparatively slow. Further, the water removed
from the tank is, on the whole, but lukewarm, and therefore but
imperfectly accomplishes its mission. On the other hand, in the
drip system a small amount of cold water is always in contact with
the coil, giving a maximum cooling efficiency with a minimum
expense.

A recent heat treatment installation l attacks the problem of
keeping the quenching medium at a uniform and low temperature by
the maintenance of a large and separate supply of oil. The hardening
is done in special quenching tank cars, as shown in Fig. 55, and
which are wheeled to any furnace desired. Just before quenching
commences the valve in pipe K is turned on and a 2-in. stream of cold
oil is kept flowing into the tank. The hot oil passes out through the
overflow pipe L, through the hole in the floor and into a pipe that
conducts it into an underground tank. This underground pipe is
made very large, so that there will be no danger of its clogging,
which would necessitate tearing up the floor. Each furnace through-
out the 400-ft. length of the shop is provided with a similar inlet
pipe and floor hole connection to the pipe which carries away the
overflow. From the underground tank the oil is pumped to
upright tanks close to the outside of the building; from these
tanks the oil flows by gravity to the tank cars.

Use of Compressed Air. — The advisability of using compressed
air in the -quenching tank is a much debated point. If applied
intelligently, however, it undoubtedly renders great assistance in
the hardening and cooling operations. In systems in which the oil
is kept in constant and fairly rapid circulation, it is neither required
nor advised.2 But if the oil is cooled by the circulation of water
in pipes, the use of compressed air is often mandatory in order to
obtain the maximum, as well as uniform, cooling efficiency of both
oil and water. In any case, the air must not be allowed to come in
contact with the hot steel, as soft spots would result; neither should it
be used in too great quantities nor pressure, especially with the
heavier and low-grade oils, as it may cause the precipitation of

1 " A Modern Heat-Treatment Plant,'7 Machinery, Sept., 1914.

2 The cold oil forced into the quenching tank may be distributed under pres-
sure to different parts of the tank, thus providing excellent circulation, and
accomplishing the same results as compressed air.

STEEL AND ITS HEAT TREATMENT

certain constituents of the oil, or cause the formation of a scum or
foam on the surface of the oil. When the air sets up a fairly efficient

circulation of the oil (or water, if water is the quenching bath), it
has accomplished its mission. Compressed air should rarely be
used with animal or vegetable oils on account of oxidation.

HARDENING 87

Size of Quenching Tank. — The volume of the quenching medium
to be used, and hence the size of the tank, depends principally upon
the size and number of the pieces to be hardened, and also upon the
method used for cooling the quenching bath. The tank should
always be of sufficient size to take with ease the maximum size stock
to be treated, besides a generous allowance on all sides for a suffi-
cient body of oil or water, for rapidity in handling the material, and
for circulation. Further, the size of the tank should be proportioned
to the degree to which the solution can be kept cooled when the hard-
ening department is operating at maximum capacity; the more
efficient the cooling system, the smaller the size of tank necessary.
On the whole, it is decidedly preferable to have the tank too large
than too small.

CRACKING AND WARPING

Influence of Non-uniformity of Section on Cracking. — One of the
main causes of steel breaking in hardening is from the unequal con-
traction and expansion in different parts of the steel. If it were pos-
sible to get every particle of the steel cold at the same moment there
would be an end to danger of this sort. But as this is a physical
impossibility, we must approach such a condition as near as we can.
This danger of cracking is particularly emphasized in forgings or
tools of unequal thickness. If the thinner part should be first
immersed in the quenching bath (e.g., water), it would become cool
much sooner than the heavier sections; that is, the thin part would
be cold or " fixed " while the thicker part of the article was still
contracting from loss of heat. Hence the thin part in its then hard
and brittle state cannot " give " and will consequently break; or,
if it does not break at the time of hardening, the steel is held in such
a state of stress that it is ready to break when applied to the work,
or even when being tempered. These influences are the more marked
with the greater the rapidity of cooling and hardening effect of the
bath, as well as with the increase in carbon content and alloys.

Influence of Bulk of Section on Cracking. — Further, the danger of
cracking is dependent upon the bulk of the article, even though it
be of uniform section. Its effect is repeatedly illustrated by large
forgings such as locomotive axles, crank-pins, etc., of rather high
carbons quenched in water. This point is illustrated by the case
of a locomotive crank-pin which had been hardened in water and
then toughened. A thorough examination of the forging before
shipment to the railroad company revealed no external evidences

88 STEEL AND ITS HEAT TREATMENT

of any crack; but when it had been in service but a very short time
it fractured badly. Examination then showed that it had evidently
been in a state of stress within its center, with the development of
an embryo crack; the dynamic stresses to which it had been sub-
jected in service were sufficient to raise the tension beyond what the
steel would stand, with the resultant internal fracture and its pro-
gressive development into complete rupture.

Expansion and Contraction. — In view of recent research work
this phenomenon of cracking may be explained in a theoretical
manner along the following lines. We know that when a piece
of steel is heated through the critical range the formation of
austenite takes place with a decrease in volume; and a somewhat
corresponding and opposite increase in volume occurs when it is
cooled through the same critical range. Now if a large forging of
considerable diameter is quenched rapidly, the outer sections will
be held in the hardened condition, and therefore rigid and stressed.
Meanwhile the interior of the steel, being cooled much less rapidly,
will in all probability actually pass from the austenitic-martensitic
condition into that of pearlite, accompanied by the increase of
volume noted above. If the outer portion or surface of the steel
is unable to withstand this expansive force, rupture must necessarity
occur. Illustrative of this, the author has seen heavy locomotive
axle forgings, after removal from the oil-hardening bath, actually
break open with a tremendous report. However, if the forging
has not been hardened too drastically, and is removed from the
quenching bath before entirely cold, an immediate reheating or
toughening process will generally relieve these stresses before any
actual damage takes place.

Hollow Boring. — In order to avoid such dangers, there appears to
be a decided tendency toward requiring the drilling of axles, shafts
and heavy forgings of large diameters to provide for heat treatment
and to remove defective material. It is undoubtedly the fact that
heat treatment will not attain its full effects in the core of a large
section. With a solid axle, the heat, upon quenching, is removed
by a flowing from the center to the outside and thence to the hard-
ening bath; the amount of heat is so great, however, that at the best
the core will be but semi-hardened, and in most cases will but be
grain-refined, or annealed. This point was well illustrated by one
company in its experiments: it split open a large, heat-treated
driving axle; the fracture showed that the heat treatment had
penetrated the ends to a depth of about 6 or 8 ins., and on the

HARDENING

sides to a depth of about one-half the radius; the fracture of the
core was similar to that of annealed steel. Again, the loss of duc-
tility and failure of the heat-treatment process thoroughly to pene-
trate the core of semi-hardened steel are shown by the following
results obtained from a 12-in. axle, heat treated, and taken at regular
intervals from the center to the outside:

Tensile Strength.
Lbs. per Sq. In.

Elastic Limit,
Lbs. per Sq. In.

Elongation,
per Cent in 2 Ins.

Reduction
of Area, per Cent.

1. (Center)

95,000

60,000

7.5

9.6

99,750

60,000

15.0

35.3

104,500

65,000

17.5

35.7

104,500

65,500

19.0

40.3

5. (Outside)

106,500

70,000

21.5

47.7

Treatment: Quenched in water from 1580° F.; toughened at 1100° F.
Analysis: Carbon, 0.35; manganese, 0.56; phosphorus, 0.020; sulphur, 0.024; nickel,
1.19; chrome, 0.31.

By means of drilling a hole through the axle, the quenching solu-
tion is able to remove the heat from both the inner and outer part
of the axle at the same time. Hollow-bored axles should be quenched
vertically whenever possible, and a constant flow of the oil or water
through the bore be supplied.

The American Railway Master Mechanics' Association in its
proposed specifications for alloy steel locomotive forgings (June, 1914)
calls for " drilling forgings over 7 ins. in diameter, unless otherwise
specified by the purchaser. The committee has found a great tend-
ency among users of quenched and tempered steel to require drilling
of parts over 7 ins., and this practice is advocated by steel-makers.
In the case of axles and crank-pins particularly, drilling takes away
practically nothing from the strength of the part; it removes the
material from the center where defective material is most likely to
exist and where it is least subject to the beneficial effects of heat
treatment, and it allows the forging to adapt itself to expansion and
contraction due to heating and cooling."

Warping. — Warping is but another manifestation of the effect
of unequal contraction and expansion, originating mainly in incorrect
heating or neglect in the manner of quenching, rather than in the
more drastic effect of the bath itself. Non-uniform heating must
inevitably result in warping, for if some parts are hotter than others
when the steel is quenched, it is evident that the rate of cooling over
the entire length of the piece cannot be the same. The general

90 STEEL AND ITS HEAT TREATMENT

tendency will be for bars to buckle or twist, due to unequal contrac-
tion during hardening. Take, for example, a bar which has been
placed upon the relatively cold floor of a heating furnace in which
the main heat application comes from above. Under these condi-
tions the tendency will be for the bar to become more heated along
the upper surface than in that in contact with the cold floor. If
the bar should now be quenched, the under part — being lower in
temperature — would contract first (provided it were heated and
quenched from a temperature over the critical range) and thus
become bowed. But if the temperature in the cooler part of the bar
were under the critical range, the tendency would be to bend in the
opposite direction. Other variations in heating might give a double
bend; certain localized heating might even cause twisting or tor-
sional strains.

Manner of Quenching. — Uniformity of quenching is requisite
to good hardening work. As a general rule, objects should be
quenched vertically in the direction of their greatest length. Like
all rules, there are certain exceptions which must be made to this
general statement — such as in the case of half-rounds and articles of
a corresponding design, as well as in such cases where economic
handling requires other methods, as with shafts, small axles, plates,
etc. But where no special facilities have been designed for uniform
quenching, the above rule will be found worthy of adoption for
symmetrical sections, and especially with unskilled workmen.

The reasons for this may be best explained by taking small
automobile drive-shafts as an example. In pulling the piece out of
the furnace with the tongs, the tendency is to grasp it nearer the
end than at the middle; consequently, in the general haste to get the
steel into the quenching bath as soon as possible, the average work-
man is very apt to drop or plunge it into the oil or water at an angle —
that is, one end of the piece strikes the quenching solution before the
remainder of the steel. Hence, initial hardening strains are set up
which usually result in a bent shaft when it is removed from the
tank. It is very difficult, in the space of a second or two, to get
hold of the bar exactly at the middle and also to lower it into the
water or oil so that both ends are immersed at the same identical
moment — which this method of quenching demands. Now if the
workman was to aim at immersing the piece end foremost, as in Fig.
56, grasping it near the end (as usual) with his tongs, the weight of
the shaft would automatically tend to bring the shaft to the normal,
and the quenching would be more nearly uniform, Axles and

HARDENING 91

forgings of a similar nature should be quenched vertically whenever
possible, as less strains are set up in the axle by this mode of quench-
ing. Extensive investigations by one locomotive builder would tend
to show that axles quenched horizontally (as is customary) develop
a series of stresses which, when plotted, appear as an oval around the
axis of the axle instead of as a circle.

FIG. 56. — Proper Method of Quenching Small Round Bars.

Hollow forgings, such as guns, hollow tools, etc., should always
be quenched vertically, so that the quenching medium may have a
free flow through the bore, and also to prevent the pocketing of any
steam or vapor which may be formed by the contact of the hot steel
and the solution.

Round Sections. — The hardening of round sections without
cracking or bending, and without undue labor cost, presents a problem

92 STEEL AND ITS HEAT TREATMENT

which has attracted much study. The danger of fracture, especially
of internal origin — whether actual or potential — is always greatest
in the circular section. This is largely due to the fact that all the
stresses and their subsequent strains are grouped symmetrically and
converge upon the central axis. Both the square bar with its corners,
and the plate or sheet with its larger surface exposure, can more
easily yield to the internal stresses and afford relief — either in cooling
during hardening or in the reheating for tempering or toughening —
than can the circular section. Further, there is greater danger of
bending and twisting due to non-uniform cooling in the long, round
bar than in almost any other common section. As has been noted,
short lengths of rounds of small diameter should always be quenched
vertically. But when it comes to the handling of large numbers of
larger bars, either of greater length or diameter, this method is
obviously at a disadvantage. Yet if the bars are simply dropped into
the bath by hand, even if every effort is made to have the axis of the
bar parallel to the surface of the quenching medium, general unsatis-
factory results are obtained, due to non-uniform cooling.

One satisfactory method for quenching such bars is shown in
Fig. 55, in which automobile shafts are handled. The bars, after
careful heating, are pulled out with long rods which have a hooked
end, across the inclined steel fore-hearth J, whence they drop on to a
jointed rack in the oil tank and are quenched. By starting the bars
with their axes parallel to the surface of the oil, they must neces-
sarily be held in the same relative position as they pass down the
rack into the oil. The rolling also effects a more uniform cooling
of the shaft in relation to its central axis. Fig. 57 shows how the
traveling crane lifts one side of this jointed rack to raise the shafts
out of the oil and dumps them on to the truck at the side.
i v An improvement on this method to give further uniformity in
cooling, and which has been used on finished shafts with almost the
entire elimination of bending, is illustrated in principle in Fig. 58.
The apparatus consists of a number of inclined planes or racks
(similar to that shown in Fig. 57), made from small bars or old rails
which are held in position by suitable cross-pieces. The hot shaft
is started down the first plane and passes into the oil or water;
thence it drops to the next, and so on until it reaches the bottom,
and is removed by suitable methods. Notice that the change from
one plane to the next causes a reversal in the direction of rolling, so
that any stresses set up by one plane are practically counteracted
by the next plane, giving a maximum uniformity in cooling. The

HARDENING

angle of incline has a great deal to do with the practical working out
of the procedure, and should be varied according to the diameter of

the bar, its chemical composition, and the nature of the quenching
medium. The rate of travel down the incline should not be too
rapid, but should nevertheless be sufficient to allow the reversing

STEEL AND ITS HEAT TREATMENT

action of the several planes to take its effect before the steel is to'd
cold. The angle of the planes may be increased as greater depth in
the solution is reached. The bar should be cold when it reaches the
bottom of the tank. The angle of the first incline is the most
important, and should be determined by experiment; it will gen-
erally be in the vicinity of 10° or 15°. In one plant in which this
method was used the number of shafts requiring straightening was
reduced from a very high percentage to less than 1 per cent, of
the total number treated.

i FIG. 58. — Rough Sketch of Inclined Racks for Quenching Rounds.

Double Quenching. — The effect of a double quench is, as a general
rule, to raise the elastic limit and tensile strength without diminishing
the ductility. This is for the most part due to the higher degree of
refinement which this double quenching makes possible, thus putting
the steel in the best possible condition. If the steel is in good con-
dition (i.e., refinement) before the first quenching, the influence of
the second quenching will be the less in proportion. It is often cus-
tomary first to quench from a temperature 100° or 200° F. over the
critical range, and then, for the second quenching, to heat just enough
over the critical range to obtain the degree of hardness desired.

HARDENING 95

For high-carbon steels the double quenching is not to be recom-
mended except under unusual conditions — such, for example, when
the steel has been greatly overheated in some previous operation.
The hardening of high-carbon steels is at best a difficult operation,
and the less heating to which such steel is subjected the better.

Manganese on Hardening. — As we have previously mentioned,
the presence of manganese causes a greater hardening effect, due to
its obstructing the austenite transition. This increase in hardness —
in ordinary carbon steels with less than 1.75 per cent, manganese —
is commonly thought to be associated with an increase in brittle-
ness,1 and with the danger of cracking during or immediately sub-
sequent to . quenching. Forethought must therefore be used in
obtaining the proper combination of manganese, carbon, and rate of
cooling to avoid the latter difficulty. The general limits of safety
for practical work may be broadly (but not invariably) set somewhat
as follows: water quenching is always dangerous when the mangan-
ese content runs up around 1.50 per cent., even in low-carbon steels;
with approximately 1.00 per cent, manganese water quenching may
be used — although not advised — with mild forging steels; with the
progressive increase in carbon the manganese content should be
rapidly lowered, so that in tool steels for water hardening the mangan-
ese is under 0.40 per cent., and with very high-carbon tools is not over
0.25 per cent. Dependent upon the size and general shape (design)
of the piece, as well as the condition (refinement) of the steel, oil
quenching is generally safe up to 1.75 per cent, manganese with
0.60 per cent, carbon — in fact, one well-known oil-hardening tool
steel analyzes about 0.90 per cent, carbon with 1.60 per cent, man-
ganese. The subject of high manganese steels will be considered
under a separate chapter.

1 Refer to Chapter XV for a further discussion of this point.

CHAPTER V
TEMPERING AND TOUGHENING

TEMPERING

Tempering. — When a piece of carbon tool steel is heated to a
red heat and quenched in water (i.e., hardened), the steel becomes
hard, brittle, and is held in such a state of stress that its use —
except in a few particular cases — would be highly inadvisable.
This hardening operation has arrested the austenitic transition at the
martensitic stage, and prevented it from advancing further, as into
troostite, etc. Under these circumstances, the application of heat
will now accomplish two results: (1) it will relieve the hardening
strains, and (2) permit the transition to proceed. By properly
adjusting the temperature of this reheating process, any desired
stage in the martensite-troostite transition may be obtained. And
by permitting just the right amount of the hard, brittle martensite
to go over into the softer and tougher troostite, any desired combina-
tion of physical properties within the capacity of that particular
steel may be realized. This process of " letting down " or softening
is called tempering.

Troostite. — If the steel has been fully hardened so that it consists
entirely of martensite, troostite will begin to form at somewhere in
the vicinity of 400° F., or possibly lower. As the tempering tempera-
ture is progressively raised, the troostite increases in amount until
at about 750° F. it begins to change into sorbite. Thus steel in the
tempered condition is usually characterized by the presence of more
or less troostite, dependent upon the degree of hardening and upon
the tempering. Just as martensite may be said to represent the
condition of hardened steel, or pearlite that of annealed steel, so
troostite is indicative of a tempered steel — whether it be obtained
by water quenching and reheating, or by quenching in some less
drastic medium such as oil but with no reheating. The question of
whether troostite represents a complete step in the transformation
is not definitely known, and as far as practical heat-treatment work

TEMPERING AND TOUGHENING

is concerned is but a question of scientific value; the value of troostite
in its influence upon the hardness and allied properties of tempered
steel is, however, definitely recognized.

Hardening Strains. — It should be always remembered that
tempering not only softens the steel through the influence of troostite,
but also relieves the strains set up in hardening. This last factor
should not be lost sight of, for although the proper degree of hard-
ness is requisite for specific work, no tool will eventually prove of
much value if it retains the state of strain occasioned by rapid
cooling. This statement applies not only to water quenching, but
also to oil quenching (or oil tempering) . Even the influence of boil-
ing water is often sufficient to relieve more or less of these strains,
if it is not desired to further soften the steel by higher reheating.
Naturally, however, the higher the softening temperature the better
will be the condition of the steel in this regard.

Temper Colors. — Nature has provided a useful and more or less
empirical indication of the degree to which tempering has affected
the steel through the formation of a surface film of oxide colors (oxide
of iron). If a piece of hardened steel is brightened with emery
paper or other suitable means, and is then slowly heated with expo-
sure to the air, the brightened surface will take on characteristic
" temper colors." These commence with a very faint yellow and
progressively change with increase of temperature through varying
degrees of yellow, brown, purple and blue. That these colors bear
a definite relation to, and are closely indicative of, a known tempera-
ture, under certain conditions, is now a generally accepted fact.
Although a difference in distinguishing the various shades of color
is bound to occur on account of the " personal equation," the follow-
ing table is fairly representative :

Temper-

ature, De-
frees

Color.

ature, De-
frees

Color.

ahr.

420

Very faint yellow

510

Brown

430

Yellowish-white or light straw

520

Brown purple (peacock)

440

Light yellow

530

Light purple

[450

Pale yellow straw

540

Purple

'460

Straw

550

Dark purple

470

Dark Yellow

560

Light blue

480

Deep straw

570

Blue

490

Yellow brown

600

Dark blue

500

Brown yellow ,

625

Blue tinged with green

98 STEEL AND ITS HEAT TREATMENT

Limitation of Color Method. — The previous statement regarding
the relation of tempering colors to temperature is true in its entirety
only under certain definite conditions of heating, and which are
largely dependent upon the time element. So long as the heat of
the steel is being progressively raised — that is, so long as the temper-
ature of the fire, furnace or tempering plate is greater than the
temperature of the steel — the temper colors indicate the temperature
of that part of the steel most affected — the surface. But when the
steel is being kept at a definite tempering temperature for any length
of time, the colors do not represent the actual temperature. This
point is readily illustrated by heating a small piece of hardened steel
at a constant temperature for a considerable period of time. Thus,
in one instance, a straw color was produced in about a minute, but
changed to a brown in about ten minutes, and to a purple in about
forty minutes; and yet the temperature of the steel was never
higher than 460° F., representative of the straw color. In other
words, the time element has developed a new set of conditions w.hich
may greatly affect the depth of oxidation or color.

On the other hand, it is a debatable point as to whether or not
these temper colors represent the actual condition (not the temper-
ature) of the steel itself. Some tool makers maintain that the
efficiency of the tool — both in hardness and in other properties — is
the same whether the color has been obtained by a short heating at a
high temperature, or a longer heating at a lower temperature. That
is, the ultimate results are indicated by the temper color, independent
of the method of obtaining it. Others aver that such is not the case.

Tempering for Depth. — It is obvious that the temper color is
at the best but a surface indication. For some tools or articles which
require a specific superficial hardness only, and in which the condition
of the center of the tool is of little consequence, it probably does
not matter a great deal in the ultimate results whether the temper
color — a straw color for example — has been obtained by a few min-
utes' heating at 460° F., or by heating for a longer period at say 360°
F. Contrariwise, if the tool or part is to be subjected to stresses of
such nature as demand the best that the steel is capable of, the
greatest degree of uniformity and release of hardening strains is
requisite. Such may only be obtained by a thorough heating at a
specified temperature, and which may be entirely independent of the
color indication. In such cases, to use the above temperatures, the
thorough heating at 360° — it more uniformly affecting the whole
mass of the steel — might prove immeasurably better than the

TEMPERING AND TOUGHENING 99

incidental surface heating to 460°. And as will be mentioned later,
a continued heating at 460° would again be an improvement over
either color method.

Quenching after Tempering. — The method of tempering by color
indication inherently requires immersion when the specified color
is reached to prevent any further rise in temperature, or in the
blacksmith's phrase, to " set the grain." Although it is possible
so carefully to heat the steel that the maximum effect is just to
develop the color desired — and no further, such methods take so
much time and patience that they are rarely carried out in practice.
The necessity of such immersion or quenching, even in the hands
of an experienced hardener, is the source of many troubles. Not only
does the quenching probably induce further strains into the steel,
but it is also entirely inconsistent with uniformity of results. If the
object is of considerable size, or varies greatly in dimension of
adjoining sections to be similarly tempered, or is of intricate design,
the difficulty in obtaining the same temper throughout even on the
surface (to say nothing of the interior of the steel), will be greatly
magnified. If the proper color is reached on one part before another,
there will be a corresponding difference in hardness. And thus the
difficulties multiply ad infinitum.

Use of Liquid Baths. — Later methods involving the use of liquid
baths for heating overcome the difficulties in color tempering,
eliminate — as a general rule — the necessity for quenching, and
further give complete uniformity of heating throughout the whole
mass of the steel and the maximum elimination of hardening strains
as can be obtained at the temperature used. By maintaining the
bath at the proper temperature there can be no overheating, the
heat must penetrate all parts of the steel alike, and the " personal
equation " is as nearly eliminated as is possible. This method has
the further advantage of cutting down labor costs and increasing
the output, since a number of pieces may be heated at the same
time, and while one lot is being tempered another bath may be
charged or discharged.

Comparison of Physical Properties Obtained. — An excellent
example of the efficiency of bath tempering is illustrated in auto-
mobile gears. On account of the relatively thin section of the teeth
as compared with the mass of the gear, exact tempering by the ordi-
nary temper-color practice is rather difficult. The teeth, which
should be the hardest, take the temper first, and are therefore the
softest part of the gear as a whole. If the gears were to be tern-

100 STEEL AND ITS HEAT TREATMENT

pered by revolving on a hot bar much better results would be ob-
tained than by ordinary tempering, but the time and cost elements
would prove excessive where hundreds of pieces were to be handled.
By the use of a suitable liquid tempering bath thorough uniformity
could be obtained throughout. Where by the color method, the core
of the gear would have the tendency to be too hard, the teeth per-
haps too brittle or soft in places, and only the surface of the gear
as a whole affected by the temper-color representing say 475° F.,
by the more modern method the whole mass of the gear would have
the physical properties as characterized the drawing temperature
of the 475° F.

Exact Temperatures. — Too much attention cannot be given to
the necessity of obtaining exact temperatures in the tempering opera-
tion. For the average run of carbon tools the tempering range is
very narrow, probably within a hundred degrees for the great
majority. The tempering action takes place extremely rapidly and
often a difference of 15° or 20° may cause much trouble. Trying
to temper tools over an open fire may be all right in isolated
cases, but it spells failure if made a general practice.

Tempering Methods. — The procedure to be employed in temper-
ing must necessarily depend upon the nature of the tool or part.
Methods must be developed to satisfy the individual requirements
and are too numerous to discuss here. Briefly, however, the more
common practices may be covered by the tempering plate, the sand
bath, and such liquid baths as oil, lead and alloys, and molten
salts.

Tempering Plate. — The tempering plate generally consists of
an iron casting planed smooth on top, and heated from beneath by
suitable means, such as gas, oil, or even a coal or coke fire. The
steel articles are placed on the plate and moved about until they
have attained the proper temper color and then quenched. Fig. 59
shows a characteristic equipment for heating, hardening and temper-
ing dies; Q represents the discharge end of the heating furnace,
R the quenching tank, and T the tempering plate, the latter being-
heated by oil burners from beneath.

Sand Bath. — In order to effect more uniform tempering of small
tools, a pan of clean, well-dried sand may be placed on a suitable
hot-plate, or in a furnace. The sand is held at the desired tem-
perature, which may be determined by the insertion of a ther-
mometer or pyrometer couple, and may be protected by covering
with a suitable hood. The oxide colors on the steel may also be

TEMPERING AND TOUGHENING

101

used as a measure of the tempering, as there is of course free
access of air between the particles of the sand.

Oil Baths. — For much of the ordinary tempering work an oil
bath will probably prove as satisfactory as any method for temper-
atures up to about 500° F. or even higher. The chief requisites are
a tank holding an ample supply of oil, a suitable furnace or method
of heating by which accurate and constant temperatures may be
obtained, and a mercury thermometer for determining the tempera-
ture of the oil. Mineral oil with a flash-point of some 600° F. is

FIG, 59,— Quenching and Tempering Dies, (" Machinery.")

generally used for the bath ; certain of the animal and vegetable oils
are also occasionally used.

Handling the Material. — Oil baths, and similarly the salt
baths, are provided with a wire basket in which the pieces to
be tempered are placed and which is then lowered into the oil.
By this method a number of pieces may be tempered at once,
besides preventing the steel from coming in contact with the sides
or bottom of the tank, which is apt to be hotter than the oil.
It is advisable, whenever possible, to allow the hardened steel to

102 STEEL AND ITS HEAT TREATMENT

come up gradually to the desired temperature, and not to immerse
in the oil when the latter is already at the highest heat. Rather
put the steel in the oil when the latter is about 200° to 300° F. and
let the two heat up together. The reason for this is that the pre-
heating— if it may be thus termed— allows the heat to penetrate
more gradually, softening the outer portion of the steel in such a
way that the inner and stressed part may be more gradually relieved
and thus avoiding the danger of fracture. Sudden heating has the
tendency to set up new stresses which must in turn be overcome.
The length of time allowed for the tempering to take place will
depend upon the size and nature of the piece under treatment;
fifteen minutes or so after the maximum temperature has been
reached will generally be sufficient for the average run of small
tools, gears, etc., while larger parts require more time in proportion.
If large and small parts are tempered at the same time it will do no
harm to the small pieces if they are not removed until the larger
pieces are ready, although on general principles long-continued
heating is never desirable after the steel has responded to the desired
heating. When the full effect of the tempering has been attained,
the pieces may then be removed from the oil and allowed to cool off
in the air, for if the steel has been thoroughly heated at the maximum
temperature of the tempering operation, no further change will
take place in the ordinary steels; each phase of the transition is
represented by a definite temperature for each steel, so that no
further step in the transition will occur unless the temperature is
raised — with the possible theoretic exception of very long-continued
heating. For some large work, such as die blocks, large cutters,
etc., the steel is allowed to cool off in the oil in order to procure the
greatest elimination of strains.

Salt Baths. — If higher drawing temperatures than those possible
with oil are desired, a bath of salts may be used. A combination of
two parts of potassium nitrate and three parts of sodium nitrate
melts at about 450° F. and may be used up to about 1000° F.
Methods of heating and using are similar to those with oil baths, and
described under Hardening Baths. The use of nitrate salts instead
of the chloride salts is necessary on account of the lower tempera-
ture desired.

Lead Baths; Alloys. — Lead, having a melting-point of about
610° to 630° F., may also be used for tempering where temperatures
higher than its melting-point are required. The disadvantages are
similar to those noted under its use for heating for hardening. The

TEMPERING AND TOUGHENING

103

melting-point may be lowered by alloying the lead with tin, and
temperatures suitable for ordinary tempering may be obtained
approximately as follows: l

Lead
Parts.

Tin
Parts.

Approx. Melt-
ing Temp. ° F.

Lead
Parts.

Tin
Parts.

Approx. Melt-
ing Temp. ° F.

420

490

430

510

440

530

450

550

18 5

460

200

560

470

Melted lead

610 to 630

480

The use of these various alloys of predetermined melting-points
for tempering is similar to that previously explained when selecting
a combination of salts with certain melting-point in the hardening
operation.

TOUGHENING

Sorbite. — As the reheating or drawing temperature is increased
still further beyond the tempering range we find that another stage
in the austenitic transition commences — the change of troostite
into sorbite. Like the change from martensite to troostite, the
formation of sorbite does not take place spontaneously throughout
the whole steel, but increases gradually and progressively. Most
writers believe that sorbite is essentially an uncoagulated conglomer-
ate of irresoluble pearlite with ferrite in hypo-eutectoid (less than
about 0.85 per cent, carbon), and cementite in hyper-eutectoid steels
respectively, but that it often contains some incompletely trans-
formed matter. Its components at all times tend to coagulate
into pearlite. On higher heating, sorbite changes into sorbitic
pearlite, then slowly into granular pearlite, and probably indirectly
into lameliar pearlite. Sorbite differs from troostite in that it is
softer for a given carbon content, and in usually being associated
with pearlite instead of martensite, and from pearlite in being
irresoluble into separate particles of ferrite and cementite.

Importance of Sorbite. — The main importance of sorbite is due
to its physical properties. Although slightly less ductile than pearl-
itic steel for a given carbon content, its tenacity and elastic limit are
so high that a higher combination of these three properties can be

1 Table by O. M. Becker, using melting-point of lead as 610° F.

104 STEEL AND ITS HEAT TREATMENT

had in sorbitic than in pearlitic steels. Steels which are so treated
as to contain sorbite are often called " toughened " steel.

Toughening Range.— The transition of troostite— the chief
characteristic of tempered steel, into sorbite — characteristic of
toughened steel, is gradual, and progresses with the increase and
duration of the reheating. At some point, depending upon the
composition of the steel and the degree to which the steel has been
affected by the hardening process, sorbite is formed. If we accept
sorbite as the characteristic constituent of toughened steel (and
which it undoubtedly is), we may then consider as the lower limit
of the toughening range that temperature which will produce sor-
bite. In fully hardened steel of the medium forging and higher
carbon analyses, characteristic sorbite begins to form at about
750° F. At about 1250° to 1300° F. the sorbite coagulates into
pearlite, which is distinctive of annealed steel. With these facts
in view we may then consider, in a general way, that the toughening
range lays approximately between 750° and 1250° F. It must be
remembered, nevertheless, that these temperatures are in no sense
definite, but are arbitrarily taken as representative of a class of heat-
treatment work: differences in chemical composition, the degree of
hardening, the size of work, etc., all play their part.

Influence of Toughening. — When a piece of hardened steel is
reheated for toughening, each specific temperature has a certain
definite influence upon the steel. The results of this toughening
process are interpreted by the ability of the steel to do certain work,
to withstand the application of stated loads, or as measured by
standard methods of testing. On account of the almost universal
use of the last named for purposes of comparison, we will deal briefly
with (1) the static strength, as measured by the tensile strength
and elastic limit, (2) the ductility, as measured by the percentage
elongation and reduction of area, and (3) the dynamic strength, as
measured by the alternating impact test.

Effect of Increased Temperature. — Each increase in the toughen-
ing temperature lowers the tensile strength and elastic limit, but with
a corresponding increase in the ductility and dynamic endurance.
With the majority of ordinary carbon, nickel, chrome and vanadium
steels the ratio of the elastic limit to the tensile strength remains
very nearly constant throughout the sorbitic range (which we
assumed to be approximately from 750° to about 1250° F.). Be-
yond these temperatures, and coincident with the formation of pearl-
ite, the values for the elastic limit and tensile strength — of each par-

TEMPERING AND TOUGHENING

105

106 STEEL AND ITS HEAT TREATMENT

ticular steel — begin noticeably to diverge until they reach their
smallest ratio in fully annealed steel. Up to near the end of the
sorbitic range the graphs obtained by plotting the elastic limit and
tensile strength against the drawing temperatures are, for general
purposes, straight lines, but beyond this range curve towards the
horizontal, as is represented in Fig. 60.

Effect on Ductility. — These changes are accompanied by reverse
changes in the ductility, as measured by the reduction of area
and elongation. As interpreted by the research work of others,
and from his own experimental work, the author is inclined to the
belief that these two factors differ from each other in that the
reduction of area generally reaches a maximum at about the end of
the sorbitic range and then decreases, while the elongation does not
attain its maximum until the steel is fully annealed or in the pearlitic
condition. Be this as it may, through the sorbitic stage at least,
each increment of decrease in tensile strength and elastic limit is
associated with, and counterbalanced by, an increase in the reduction
of area and elongation. This combination of static strength and
ductility is further almost directly proportional to the toughening
temperature.

Impact Strength. — The effect of toughening upon other properties
and especially in relation to the impact strength, is shown in Fig.
60, rearranged from the work of Grard. The steels, approximating
0.15, 0.40 and 0.50 per cent, carbon, were hardened and then re-
heated to temperatures varying from no tempering up to 2200° F.
The impact strength curves present some extremely interesting facts.
We find that the greatest resistance to shock to be obtained from
a toughening, after hardening, at a temperature about 100° F.
below the upper critical range (Ac3); annealing at a temperature
superior to the Ac3 range gives a lower impact strength. Further,
as the temperature is raised more and more and overheating results,
there is a marked diminution in the impact strength. Increase in
the carbon content, assuming the same heat treatment, diminishes
the impact strength. Tempering (reheating up to say 600° F.)
has little or no effect upon the impact strength. As a general prop-
osition we may sum up by stating that it is preferable, in order
to obtain the greatest impact strength, to keep the 'carbon
content as low as possible and to have a high drawing tempe-
rature.

Capacity of the Steel. — Thus it will be seen that by changing the
drawing temperature the grouping of these factors may be varied

TEMPERING AND TOUGHENING 107

through a considerable range and limited only by what we may call,
for want of a better phrase, the " capacity of the steel." This
quantity is defined largely by the chemical composition, the method
of manufacture, the size of the piece to be treated, and by other
subordinate factors. With these qualifying conditions in mind, we
may further define the capacity of the steel as the limiting ratio of
strength to ductility. Each steel, as qualified above, has certain
definite limits within which the physical properties may be varied.
At one end of the see-saw, as in hardened steel, there is a maximum
tensile strength with minimum ductility; and at the other extreme,
as in fully annealed or sorbitic-pearlitic steel, there will be a mini-
mum tensile strength with maximum ductility. Following out the
simile of the see-saw, we may place tenacity on one end and ductility
on the other; when one is up, the other must be down; both cannot
be up nor both down at one and the same time; raise one and the
other must fall. The heat-treatment man now stands on the middle
of the board and by means of his reheating temperature can adjust
the opposing factors to that position which he desires; but he cannot
change the maximum and minimum of either, because they are fixed
by the limitations previously mentioned at the beginning of the
paragraph, and over these he has no control as far as the individual
stael is concerned.

Duplication of Results. — Happily for the heat-treatment man,
each grouping is distinctive of a definite toughening temperature,
other conditions being the same. When he has once determined the
relation existing between static strength, ductility, and temperature,
for a given size piece of work made from a steel of specific analysis,
he knows that he can approximately duplicate his results under like
conditions at any time. Not that he can absolutely and ultra-
scientifically obtain results within a few pounds elastic limit or hun-
dredths of a per cent, elongation — for such are neither necessary
nor expected — but that he can reasonably expect to get a com-
mercially acceptable duplication. It is with this thought in mind
that the subsequent chapters have been developed, giving under
each steel many of the results and details which have been obtained
in practice and experiment, and which should prove advantageous
to the average heat-treatment man as a time-saver.

Slow Cooling and Stresses and Strains. — It is one of the incon-
trovertible facts of heat-treatment work that slow cooling predicates
the release of internal stresses and strains. Not only is this true
of the full-annealing process — as indicative of slow cooling from a

108 STEEL AND ITS HEAT TREATMENT

temperature above that of the critical range, but also of the toughen-
ing operation. In fact, the very nature of the usefulness of tough-
ened steel depends upon the absence of a state of strain just as much
as upon specific static or dynamic properties. Strange as it may
seem, some of the failures in locomotive forgings may be traced
back to the lack of slow cooling after toughening; and this trouble
is coming to be recognized in many specifications by the require-
ment of cooling in the furnace after toughening. Just as the dangers
in hardening increase with the rapidity of cooling, carbon content
and size of section, so are they likewise magnified in cooling after
toughening — although on a smaller scale. If these factors become
noticeably important, cooling in air from the toughening tempera-
ture may set up such a new series of cooling strains that many of
the real advantages of toughening may be invalidated.

Use of Furnace Cooling. — The greater part of hardened and
toughened work, such as automobile and other small forgings, may
not require furnace cooling, besides being economically impracticable.
But even with these it is desirable that the pieces should be piled
together after removal from the furnace so that the cooling will be
retarded. For forgings of section greater than 3 or 4 ins., such as
heavy machine parts, ordnance, etc., cooling in the furnace is
always desirable. It may be said that such slow cooling never did
any harm, and it may do a world of good in relieving strains.

Effect of Furnace Cooling on Physical Properties. — Contrary to
the opinion held by some, the author does not believe that slow
cooling in the furnace has any noticeable tendency to further
" soften " the usual straight carbon or alloy steels to which the
toughening process is generally applied. That is, for similar pieces
of the same steel treated alike, equivalent physical test results would
be obtained in the forging which had been furnace cooled as in the
one which had been allowed to cool in the air — the tests being taken
from the same relative position. In making this statement there
is, however, one other necessary qualification: it is assumed that
the whole mass of the steel has been thoroughly heated at the tough-
ening temperature. Otherwise the effect of the toughening would
not be so great in the air-cooled piece as in the slowly cooled piece,
for the latter would have greater opportunity to be affected by the
heat of the furnace during the furnace cooling. During the tough-
ening range the effect of the heat upon the transition, except for
very large pieces, practically ceases as soon as the source of heat is
removed — as by air cooling.

TEMPERING AND TOUGHENING 109

High vs. Low Toughening Temperatures. — On the hypothesis
that either of two specified analyses would prove equally satisfactory,
under suitable treatment, for the same piece of work, but that on
account of the difference in chemical composition one steel would
require toughening at say 1200° F. and the other at say 800° or
900° F., the selection of the higher drawing-point steel should be
ma/le. Such conditions often arise in heat-treatment plants handling
a variety of commercial work and it may be well to sum up briefly
the reasons for the above conclusion.

The more stable the state of equilibrium which exists between
the transition constituents the more lasting and effectual will be the
treatment. Further, the smaller the amount of internal strains
which may remain in the steel from the previous hardening operation
the better. Both of these conditions are more nearly brought about
by the higher drawing temperature.

As there is also a decided tendency for the dynamic strength
to reach a maximum at about 1200° to 1300° F. it is probable that
the higher drawing temperature steel will have a greater dynamic
strength than the other steel, provided that there is not too much
difference between the chemical compositions of the two steels.

From the furnace man's point of view the temperatures around
1200°, being of characteristic visible reds, are decidedly more easily
recognizable than those temperatures around 800° to 1000°, since
with these lower temperatures there is very little visible heat color.
The higher drawing temperatures therefore aid in the efficiency of
judging the heating operation and lead to greater uniformity of
control and of results.

Quenching Medium vs. Toughening Temperature. — There is
another phase of the high or low toughening temperature proposition
which cannot be solved by any general rule, but only after due
consideration of all the circumstances involved; this relates to the
condition of affairs when there is no opportunity for the choice of
steel, but depends more upon the selection of the quenching medium
in relation to the toughening temperature. As we have noted,
water quenching gives a harder steel than oil quenching. It natu-
rally follows that, in order to obtain approximately the same physi-
cal results, the oil-quenched piece must be drawn at a lower temper-
ature than the water-quenched piece. The arguments regarding
water vs. oil quenching, and low vs. high drawing temperatures have
been previously discussed. If the solution were to be developed
entirely along these lines it is orobable that in the majority of cases;

110 STEEL AND ITS HEAT TREATMENT

the oil quenching (giving less hardening strains) and lower drawing
temperature would be employed. In other words, the difficulties
to be encountered with water quenching — the hardening operation
being the more drastic of the two — would more than outweigh the
the disadvantages of the lower toughening temperature. This is a
question in which the personal element and experience of the heat-
treatment man would be paramount.

Influence of the Carbon Content. — In respect to the selection of
the steel in relation to the treatment there remains the consideration
of the influence of the carbon content. Carbon not only intensifies
the effect of the rapid cooling (hardening), but it also directly
augments the brittleness of the steel. Or, to put it in other words,
the greater the carbon content the greater the hardening strains,
and the lower the ductility which can be obtained with a stated
tensile strength. It is therefore usually desirable to provide a steel
with as low a carbon content as will give the desired results.

Toughening vs. Annealing. — It is only within comparatively
recent years that the toughening process with its attendant sorbitic
structure has been used and understood. Previously, annealing
was generally the cure-all for brittleness and a strained condition
of the steel. Pearlite — produced by annealing — on account of its
entangled structure, gives a large measure of ductility; but also
gives a minimum tenacity. The appearance of sorbite, however, is
even more entangled than pearlite; sorbite is far superior to pearlite
in tensile strength and especially in elastic limit. Thus by obtaining
a sorbitic steel by suitable treatment, almost as much ductility,
greater working strength, greater dynamic strength, and — by being
able to use a lower carbon steel — less brittleness may be obtained
than in a pearlitic or annealed steel.

Standardization of Results. — With the same degree of hardening,
and if the reheating has been uniiorm and thorough at a given tem-
perature, the physical results will be comparatively the same for
material of equivalent section and the same composition. That
is, the product will be standard for standardized treatment. Fur-
ther, in order to get standard results with steel purchased under the
same general specifications (i.e., each chemical constituent within
certain limits) , the toughening temperatures may be varied according
to the chemical composition. To illustrate: the following heats of
steel of varying chemical composition and made by several steel
companies were manufactured into a certain product which, when
heat treated, required an elastic limit of 85,000 to 95,000 Ibs. per

TEMPERING AND TOUGHENING

111

square inch, and an elongation of not less than 16 per cent, in
2 ins. In spite of the varying carbon, manganese, chrome and
nickel contents, the toughening temperatures (maintained within
5° F. under or over) were so adjusted as to give the desired results.
Thousands of pieces, some weighing as much as 200 Ibs., all ful-
filled, by actual test, the standard physical specifications.

Carbon.

Manga-
nese.

Phosphorus.

Sulphur.

Chrome.

Nickel.

Toughen-
ing Temp.
Deg. Fahr.

0.16

0.43

0.015

0.017

0.62

1.82

1050

.185

.44

.010

.015

.57

1.56

975

.20

.43

.009

.016

.64

1.74

1025

.20

.46

.011

.017

.40

1.56

950

.21

.48

.015

.60

1.77

1050

.21

.50

.017

.014

.67

1.84

1075

.23

.50

.016

.018

.65

1.73

1075

.245

.53

.015

.020

.62

1.79

1120

.25

.50

.011

.019

.64

1.40

1140

.26

.43

.010

.018

.60

1.65

1100

.49

.015

.021

.63

1.79

1150

.28

.51

.008

.011

.41

1.57

1120

Quench-Toughening.— A process which has been used consider-
ably for the treatment of large forgings of uniform section, such as
heavy axles, is that of heating as usual for hardening and then
quenching in oil for a specified number of seconds, followed by air
cooling. The oil quenching affects the steel to a certain depth, but
still leaves a considerable amount of heat in the forging when
removed from the bath. As the forging cools in the air this heat
from within will toughen or " soften " the steel affected by the
quenching. In order to obtain equivalent results under varying
conditions the number of seconds required for immersion in the oil
of a piece of given size must be determined by experiment and strictly
adhered to. Forgings treated by this process are characterized by
a soft or annealed core, with a progressively toughened outer part.

Physical Results. — In subsequent chapters will be given results
obtained in actual practice by the use of various toughening tem-
peratures for different grades of steel,

CHAPTER VI1
CASE CARBURIZING

Object of Case Hardening. — The object of case hardening or
partial cementation is the production of a hard wearing surface (the
" case ") on low carbon steel, and at the same time the retention or
increase of the toughness of the " core " of the metal. The process
may be roughly divided into two distinct periods. First, the car-
burization or impregnation of the surface by which the carbon con-
tent is sufficiently raised — dependent upon the demands of the work
— so as to give a steel capable of taking on very great surface hardness.
Second, suitable heat treatment which shall develop the properties
of both case and core. The complete operation should not only
result in the obtaining of a very hard case, but also and simultaneously
in the realization of special mechanical properties in the core — more
especially that of non-brittleness. Briefly, the aim is to have a
piece of steel which shall possess a minimum fragility and a maximum
surface hardness.

Requirements for Case Carburizing. — In order to obtain a case
rich in carbon, the metal is heated in the presence of a body which is
capable of delivering this carbon, by more or less complex reactions,
which is then dissolved by the steel. Aside from the use of gases
in the newer processes involving such factors as pressure, quantity,
etc., there are four main factors which must be considered in the
carburizing operation:

1. The solvent: that is, the steel;

2. The product to be dissolved, or more exactly, the compound

capable of delivering the carbon, i.e., the cement;

3. The temperature;

4. The time of contact between the steel and the carburizing

agent.

1Cuts by Giolitti from " The Cementation of Iron and Steel," by courtesy of
McGraw-Hill Book Co.; references made in this chapter to investigations by
Giolitti are also from the above.

112

CASE CARBURIZING 113

THE STEEL

The Steel. — The character of the initial steel used for case car-
burizing depends largely upon the fact that one of the main desires
is to eliminate brittleness in the core. We have seen that any
increase of carbon, other conditions being equal, will increase the
brittleness, particularly when the carbon content is raised to over
about 0.25 per cent. Further, as practically all commercial car-
burizing processes involving case hardening are followed by one or
more hardening operations, it follows that the use of a steel with a
higher carbon content will also increase the brittleness through
quenching. For these reasons it is therefore necessary to keep
the carbon content of the steel to be carburized quite low, prefer-
ably under 0.25 per cent, for straight carbon steels. In fact, the best
French practice is to demand a carbon content of not over 0.12 per
cent., further qualified by the specifications that the core after
quenching shall give a tensile strength of about 54,000 Ibs. per square
inch and not to exceed 85,000 Ibs., together with an elongation of
30 per cent, in 100 mm. (3.94 ins.).

However, one of the important and often unsatisfactory results
of using an extra-soft steel is the difficulty encountered in machining
(before carburizing) . If the carbon is extremely low the steel is
very apt to tear, and thus increasing the amount of grinding after
hardening — in order to obtain a perfectly smooth surface. For
this reason, the general American practice is to adopt a carbon
content about midway between the extreme upper and lower limits
and specify a steel with about 0.16 to 0.22 per cent, carbon. The
higher carbons also give increased stiffness to the core which, in
some cases, is necessary.

It is generally recognized that the carbon content, at least
up to some 0.50 per cent., has no influence upon the velocity of
penetration of the carburization, i.e., the depth of carburization
which will be obtained for a given length of exposure.

On the other hand, the initial carbon content of the steel will
have a decided influence upon the maximum carbon content which
will be obtained in the case; the higher the initial carbon, the higher
the maximum carbon concentration in the case.

Manganese. — It is considered the best practice, in general, to
require a low manganese content with about 0.30 to 0.35 per cent,
as the maximum. It should be remembered that the case which
will be formed during the carburization will be characteristic of

114 STEEL AND ITS HEAT TREATMENT

a high-duty tool steel and will have the properties of such. Thus
manganese will increase the hardness of the case (and also of the
core) and will make the steel as a whole more sensitive to rapid cool-
ing. In spite of this, it is often customary, especially in British
practice, to use a manganese content of about 0.70 per cent. — and
in some cases even up to 0.90 per cent. — in order to obtain greater
stiffness in the core. Manganese at such percentages also increases
the brittleness produced by long heating during carburization, and
diminishes the efficacy of the regenerative quenching. These last
named points are also true when the silicon is much over 0.30 per
cent.

Other Impurities. — It is self-evident that the content of phos-
phorus and sulphur in the initial steel should be just as low as is
possible. Slag, blow-holes, segregation, and all other impurities
and imperfections should be entirely absent from steels for case
hardening.

THE CEMENT

Direct Action of Carbon. — Carburization by its very nature
requires the presence of free carbon in some form or other, either as
a solid body, or as some gas which will produce free carbon by its
decomposition. The mere presence of free carbon in contact with
iron, however, will not satisfy the conditions necessary for commercial
carburization. Although it has been shown scientifically that car-
bon alone, without the intervention of any gas, will carburize iron
if it is kept in contact with it for a sufficiently long time and at a
sufficiently high temperature, this direct action, as far as industrial
results are concerned, is negligible. That is, the ordinary forms
of solid carbon, such as wood charcoal, sugar charcoal, etc., exercise
directly on iron but a very slight carburizing action in the absence of
gases.

Action of Gases. — It will be noted that emphasis has been laid
upon the " direct action " in the " absence of gases." This at once
leads to the question as to what is meant by the action of gases, and
which, in turn, involves the mechanism of cementation itself. It is
a well-known fact that when steel is heated, the " pores of the steel
are opened " — to use the vernacular expression — it becomes easily
permeable to gases, and the surrounding gases diffuse into the steel.
This is true whether the steel is heated in the ordinary atmosphere,
when the gases consist of nitrogen and oxygen, or whether it is heated
in some specially prepared atmosphere, such as carbon monoxide,

CASE CARBURIZING 115

illuminating gas, etc. The main fact to be realized is that the gases
do penetrate into the steel, although the effect of the gases will
depend upon the composition of the gas, besides such other factors
as pressure, temperature, and so forth. Thus, recognizing that the
direct action of carbon — that is, the carburizing results obtained by
mere contact of carbon with iron — is commercially negligible in the
absence of gases, it is evident that carburization must be intimately
related to the presence of gases. In other words, the gases (or, more
exactly, certain gases) must in themselves act as the carrier or
vehicle for the carbon. That this carrier action, or transporting of
the carbon, has not been definitely recognized or determined until
recently has been due to the fact that practically all of the solid
cements generate the necessary gases through their own decomposi-
tion and interaction with the occluded air. Further, the intense
and critical study of this action has been developed only by the
research work in connection with the newer processes of case
carburizing by means of gases alone.

Action of Oxygen. — As a typic 1 example of this diffusion and its
effect we may consider any ordinary carburization process in which
wood charcoal is used as the base cement. When the carburizing
material and articles are packed in the carburization boxes there is
necessarily a considerable quantity of air also occluded with the
particles of the cement. Under the influence of heat the oxygen of
the occluded air will react with the carbon or charcoal to form car-
bon monoxide gas, which has the symbol CO. Then, as the tem-
perature of the box and contents increases to the temperature of the
carburization proper, these gases of carbon monoxide permeate or
diffuse through the surface and outer section of the steel. At the
same time, by catalytic action, the carbon monoxide gas decom-
poses when it comes in contact with the steel and sets free a part
of the carbon it contains. This decomposition may be represented
by the reversible reaction

2CO <± C02 + C

carbon monoxide^carbon dioxide (gas) -+- carbon (solid).

Thus, as the gas diffuses into the mass of the steel it continues to
decompose, setting free new quantities of carbon within the steel.
This carbon, at the proper temperatures of carburization, passes
directly into solution in the steel and forms a true steel proper. The
reaction above, being reversible — as might be shown — will continue
indefinitely under suitable conditions, the charcoal regenerating the

116 STEEL AND ITS HEAT TREATMENT

supply of carbon monoxide. Further, while it is a well-known fact
that carbon monoxide, acting alone on iron, will deposit free carbon
on the surface of the iron, this action takes place only at temperatures
lower than those ordinarily used for commercial cementation. In
other words, the carburizing action of charcoal as used in practice
is not due to the direct action of the carbon, but is due (under the
conditions named, which of course may be modified by the presence
of other gases or components of the cement) entirely to the specific
action of carbon monoxide as a gas.

Nitrogen. — The action of the oxygen of the occluded air being
accounted for, the accompanying constituent nitrogen must be con-
sidered. Although it has been shown that during carburization the
nitrogen may and will diffuse in small amounts into the steel, it is
now certain that the presence of pure nitrogen does not increase,
except to a minimum extent, the carburizing action of free carbon.
In fact, instead of nitrogen being requisite — as many still believe —
it may even exert a pernicious effect. LeChatelier has suggested that
the increase in brittleness sometimes observed in those parts of the
steel subjected to cementation, but which the carburization has not
even reached, may be due to this nitrogen. It might be added that
this deleterious nitrogenizing theory is further supported by experi-
ments along other lines — particularly in the apparent cleansing effect
for nitrogen of the titanium additions to steel during manufacture.
Another general effect of nitrogen gas is to reduce the cementing
action of the carbon monoxide mentioned by its diluting the car-
burizing gas. For practical purposes of carburization, however,
the action of nitrogen in the presence of free carbon is too slight to
influence commercially the results obtained with a given cement,
unless actually added (in gaseous cementation) as a diluent.

Carbonates. — The ash of the carbonaceous matter may also
contain carbonates of the alkali or alkaline-earth metals. Or these
carbonates, such as barium carbonate, may be added directly to the
cement. In the light of the most reliable and recent researches it
would appear, contrary to previously accepted theories, that the
activity of these carbonates is not due to the formation of volatile
cyanides by the action of the nitrogen of the occluded air, but exclu-
sively to the formation of carbon monoxide produced by the action
of the hot carbon on the carbon dioxide produced through the dis-
sociation of the carbonates. Thus the effect of such carbonates
is similar to that produced by carbon monoxide under similar con-
ditions.

CASE CARBURIZING 117

Cyanides. — The most maligned constituents of cements are the
cyanogen group. In the past it has been thought that the deriva-
tives of this group played the chief part in carburization processes.
This, however, has been strongly disproved by Giolitti, who ad-
mirably explains the matter as follows: That cyanogen and the
more or less volatile cyanides can cement iron intensely is beyond
doubt. Moreover, it is well known that fused potassium and potas-
sium ferrocyanide are used in the pure state to obtain thin and
strongly carburized zones (as in superficial carburization or cyanide
hardening) . In industrial practice the cyanides do not exist already
formed, but may be formed in very small quantity by the action of
the nitrogen of the air (occluded in the cement) on the carbon used
and on the small quantities of alkali constituting a part of the ashes
of this carbon. Although the formation of small quantities of alkali
cements cannot therefore be wholly avoided in industrial car-
burization with carbon as a base, the part which is played by these
traces of volatile cyanides is certainly negligible in comparison with
that of the carbon monoxide formed by the action of the air on the
carbon used as cement. — He then submits conclusive proofs to
substantiate these statements.

Carbon Monoxide Gas. — Carburization carried out by the use of
carbon monoxide gas alone will give a mild or gradual carburization
in which the maximum carbon content is comparatively low —
not usually reaching the eutectoid ratio even at the periphery — and
which diminishes progressively and in a uniform and slow manner
passing from the surface of the case toward the interior of the car-
burized piece. Carburized zones of this type correspond always
and only to carburization carried on with pure carbon monoxide,
a concentration-depth diagram of which is shown in Fig. 61. On
account of its definite chemical composition and simplicity of action,
the general behavior of carbon monoxide is known within almost
exact limits. The carburizing action is easily regulated, and the
case may be obtained with certainty with any kind of steel in com-
mercial use.

When working under suitable conditions, carbon monoxide —
either alone or with a mixture in which the carbon monoxide can exer-
cise its maximum carburizing action — will give the greatest velocity
of carburization, i.e., the depth reached in a given time by the car-
burized zone. This depth is also a direct function of the time or
length of exposure.

All other conditions being equal, the higher the temperature of

118

STEEL AND ITS HEAT TREATMENT

carburization using carbon monoxide, the smaller will be the maxi-
mum carbon content of the case. Similarly, the lower the pressure
of the carbon monoxide, the smaller the maximum carbon content;
and the greater the quantity of pure carbon monoxide gas coming
in contact with a unit of surface, the greater the carbon concentra-
tion.

Under suitable conditions, carbon monoxide gas will deposit
no carbon on the surface of the steel being carburized, so that there
is little difficulty in keeping the surface bright. Further, the use of
carbon monoxide reduces to a minimum the deformations and varia-
tions in volume due to the carburizing processes. Carbon monoxide
also lends itself in obtaining a good protection of the parts of the steel
which it is not desired to carburize.

i.o

0.8
0.6
0.4
0.2

0.5

1.5

2.5 3 MM.

FIG. 61. — Carburization at 2010° F. for Ten Hours with Carbon Monoxide.

(Giolitti.)

Hydrocarbons. — Most of the forms of solid carbon used in prac-
tical carburization are not pure, but may contain organic residues not
wholly decomposed, or considerable proportions of ash rich in cer-
tain carbonates. Thus charred bone, charred leather and similar
organic products often used, will, under the influence of heat, evolve
hydrocarbons. These hydrocarbons, by more or less complex
reactions, deposit the excess of finely divided carbon which they con-
tain on the surface of the metal; and this, in turn, being in perfect
contact with the metal, at high temperatures may cause a direct
carburization by contact. But further and vastly more important
than this direct action of the carbon deposit on the surface of the
metal, is the carburization by means of the specific action of the gas
itself, although of course depending more specifically upon the exact
conditions of carburization. In a manner somewhat analogous to

CASE CARBURIZING

119

that of the decomposition of the carbon monoxide within the steel,
yielding carbon directly to the steel, the hydrocarbon gases will also
diffuse into the steel and there yield carbon. Hydrocarbons there-
fore also act as carriers for the carbon and effect a carburization due
to the specific action of the gas.

Carburization with pure hydrocarbon gases give cases of a type
corresponding to Figs. 62 and 63, and to Fig. 64. These are
characterized on slow cooling by (1) a layer or zone of hyper-eutectoid
steel consisting of free cementite and pear lite ; (2) by a layer of eutec-
toid steel, generally quite thin; and (3) by an internal layer of
hypo-eutectoid steel. The main points to be noticed are, that

5<C.

1.4

1.2
1.0
0.8
0.6
0.4
0.2

0 0.5 1 1.5 2 2.5 3 MM.
FIG. 62. — Carburization at 1830° F. for Five Hours with Ethylene. (Giolitti.)

the case contains a structure with greater than 0.9 per cent, carbon,
and more emphatically, that the concentration of the carbon often
diminishes in a markedly non-uniform manner or discontinuity.
Zones of this type are always found in carburizations carried out
with hydrocarbons; they also are typical of carburizations obtained
with many of the solid carburizing compounds used in commercial
work in which the action of the hydrocarbons greatly predominates,
or in the presence of cyanides (superficial cementation).

Of the specific action of the gaseous hydrocarbons, we may make
the following remarks. The depth or velocity of penetration in-
creases, similarly to carbon monoxide, with the time of exposure.
In the case of carburization with ethylene and methane, the cemented
zones obtained in a definite time, although likewise increasing

120 STEEL AND ITS HEAT TREATMENT

markedly in thickness with rise in temperature, other things remaining
constant, maintain about the same concentration and the same dis-
tribution of the carbon in the three zones — thus differing widely
from carbon monoxide. In contrast with the use of these pure
gases, the use of hydrocarbons in practice presents a different aspect,
especially when compared with the use of carbon monoxide in prac-
tice. Contrary to the simplicity of the reactions which always
characterize the cementation by carbon monoxide, the complexity
of the reactions with hydrocarbons increases enormously in industrial
work. The gas in such instances does not consist of a single, chem-
ically definite hydrocarbon, but of a mixture of various hydrocarbons.
If we work at a comparatively low temperature, such as at, or slightly

Hyper-eutectoid ^ . f

Eutectoid

Hypo-eutectoid

FIG. 63. — Carburization with Hydro-Carbons. X25. (Bullens.)

under, the upper critical range, the process is slow and non-uniform.
At the high temperatures generally used, cemented zones of exces-
sively high carbon are always produced. The same complexity of
reactions make it difficult, in practice, to work with a cement having
hydrocarbons as a base, either as a mixture of solids in the carburizing
box, or as gases in the newer processes, in such a way as to obtain
well-defined results. Thus the use of such hydrocarbons is not
advantageous where a certain value of maximum concentration,
combined with a definite distribution of that carbon, is necessary
in carburized steels in which a considerable depth is desired.

Enfoliation. — All those who have had much to do with case
hardening and its products are familiar with the flaking, chipping, or
even peeling off of parts of the case from the remainder of the steel.

CASE CARBURIZING

121

These fractures are entirely different from those occurring in homo-
geneous high-carbon hardened steels. While in the latter the frac-
tures always have a characteristic conchoidal form, in case-hardened
steels the chipping or enfoliation always takes place along a line
corresponding to the separation of two zones exhibiting markedly
different structure or " grain." A microscopic and chemical investi-
gation brings out the fact that this line or plane of weakness charac-
terizes the separation of the hyper-eutectoid zone from that of the
hypo-eutectoid zone, or at a carbon content corresponding to that of
about 0.90 per cent. Further, this plane of weakness corresponds
to a discontinuity in the concentration or distribution of the carbon

1.4
1.2
1.0
0.3
0.6
0.4
0.2

0.5 1 1.5 2 2.5 3 MM.
FIG. 64.— €arburization at 1920° F. for Four Hours with Ethylene. (Giolitti.)

which is characteristic of carburized zones of the hydrocarbon type
previously described.

It is now evident that, in order to eliminate the possibility and
dangers of this enfoliation, we must

(1) obtain a gradual and progressive change in the distribution
of the carbon so that it will vary from the minimum of the core to
the maximum at the surface of the case, and in no place exhibit the
phenomenon of discontinuity; .

(2) eliminate the possibility of discontinuity at the eutectoid by
keeping the maximum concentration of the carbon at or below 0.90
per cent, carbon (thus eliminating the hyper-eutectoid zone);

(3) and in any case, modify by suitable heat treatments the
structure obtained by carburization.

122 STEEL AND ITS HEAT TREATMENT

Maximum Carbon Concentration. — Now while we have under
(2) advised the eutectoid carbon ratio as a means of preventing
enfoliation, it must not be at once concluded that enfoliation is the
direct sequence of increasing the carbon concentration maximum
to over 0.9 per cent. Such is not the case if the proper heat treatment
methods are employed. Unfortunately, however, the majority of
commercial plants employing case hardening do not either under-
stand, or are unable to put into practice, the methods which are
necessary when the carbon content of the case runs beyond 0.9 per
cent, carbon. That such high carbon contents are undeniably
advantageous in many instances where it has been generally thought
that their use was impossible will also be shown, as will the so-called
" secret " processes of treating the steel. But for plants which are
unable to employ the necessary metallurgical skill and appliances,
it will be far better to adopt such case-hardening processes as will
turn out a good product having a maximum carbon concentration in
the case of about 0.9 per cent. The further advantages of this will
be brought out under the discussion of heat-treatment methods in
Chapter VII.

Intermediary Type of Carburized Zone. — Recognizing under these
conditions the validity of not exceeding the eutectoid limit, and the
obvious advantages of preventing discontinuity between the core and
the surface of the case regardless of the maximum carbon content,
it is evident that we must obtain a cemented zone intermediary
between those of the two general types, previously described. In
other words, the type of case must have the principal character-
istics of the carbon monoxide type, but which are modified — by
increasing the carbon content — by cements typical of the hydro-
carbons, or other suitable procedure.

Carbon Monoxide Plus Hydrocarbons. — From the results of
experiments carried out with carbon monoxide plus specific amounts
of hydrocarbons, Giolitti shows that the additive effect of the latter,
as compared with those carried out with pure carbon monoxide, may
be summed up as follows: " The addition of small quantities of
volatile hydrocarbons to carbon monoxide merely raises the con-
centration of the carbon in the external layers of the cemented zones
above the value which would result from the use of pure carbon
monoxide under identical experimental conditions. This increase
is greater the larger the proportion of the hydrocarbon contained in
the gaseous mixture, as long as this proportion does not reach
a value such that the velocity with which the free carbon is formed

CASE CARBURIZIXG

123

by the decomposition of the hydrocarbon does not surpass the veloc-
ity with which this carbon passes through the stage of carbon mon-
oxide into solution in the iron. From this limit the excess of carbon
which is liberated begins to deposit on the steel and the concentration
of the carbon in the external layers of the cemented zone reaches the
maximum value corresponding to that which is obtained by cement-
ing with solid cements, or with cements which behave as such, and
from this point on, the concentration and the distribution of the
carbon in the cemented zones no longer vary markedly, even if the
proportion of the hydrocarbon increases greatly."

" From what precedes it is evidently possible to obtain, by means
of mixtures of carbon monoxide and vapors of volatile hydrocarbons,
cemented zones in which the maximum concentration of the carbon

0.8
0.6
0.4
0.2

.2 .4

1.2 1.4 1.6 1.8

MM.

FIG. 65. — Cemented Zone, Intermediate Type, Carburized with Carbon Monox-
ide Plus 3.1 per cent. Ethylene. (Giolitti.)

in the external layers has a definite value, lying between a minimum
corresponding to that which would be obtained by working under
the given conditions with pure carbon monoxide, and a maximum
which would be obtained by working with vapors of the hydrocarbon
alone. This is achieved simply by using gaseous mixtures containing
a proper proportion of hydrocarbon varying with the conditions
under which cementation is to be effected, such as temperature,
pressure, relation between the velocity of the gaseous current and
the surface of the steel to be cemented, etc."

An example of this is shown by the concentration-depth diagram
in Fig. 65, the results of which were obtained experimentally by
cementing 0.26 per cent, carbon steel cylinders for four hours at a
temperature of 1830° F. in a mixture of carbon monoxide with 3.1
per cent, of ethylene. It will be noted that the carbon decreases
progressively and in a slow and uniform manner, but that the addi-

124 STEEL AND ITS HEAT TREATMENT

tion of the hydrocarbon has raised the maximum carbon content up
to nearly the eutectoid ratio. Thus, in this case, there has been pro-
duced a carburized zone of an intermediary type which fulfils the
requirements stated for the avoidance of enfoliation.

Following along these lines of using a gaseous mixture consisting
of certain proportions of carbon monoxide gas and the volatile
hydrocarbons, several industrial methods have been worked out, and
which have given excellent satisfaction. The application of the
same theory is also applicable to the commercial solid cements in
which the necessary gases are evolved during the heating operation,
but on account of the greater lack of control the variations to be
obtained are necessarily of considerable extent.

Carbon Plus Carbon Monoxide. — As we have stated, the carburiz-
ing action of solid carbon in the absence of all gases is commercially
negligible. But by introducing oxygen which will form the gaseous
vehicle (carbon monoxide), or by adding carbon monoxide directly,
the presence of solid carbon greatly intensifies the carburization.
Thus, similarly to definite mixtures of carbon monoxide plus hydro-
carbons, the desirable form of the intermediary type of carburized
zone may be obtained by carbon monoxide in the presence of solid
cements.

By varying the various factors of temperature, time of exposure,
pressure of gas, etc., the use of a mixed cement may be varied within
wide limits, and with the production of a hyper-eutectic zone if so
desired. This latter comes into great practical use when it is desired
to produce zones of considerable width. Co-ordinated with this is
the use of carbon monoxide as an " equalizer," that is, by first carry-
ing out the carburization process in the usual way (with mixed
cements), the maximum concentration of the carbon may be made
quite high; this is then followed by the use of carbon monoxide alone
(without the presence of granular carbon). By these means the
concentration of the carbon may be lowered — by the distributive
action of the carbon monoxide, to such maximum concentration as
may be desired. This is graphically shown in Fig. 66, by Giolitti.
The steel used was of the composition :

Per cent.

Carbon 0.12

Manganese 0 . 47

Phosphorus.. . . . 0.03

Sulphur 0.02

Silicon.. ..0.06

CASE CARBURIZING

125

Curve a shows the concentration depth after carburization for ten
hours at 2010° F. with mixed cement. Curve b represents the
results after heating the preceding for five hours at the same tem-
perature in " isolated " carbon monoxide. Curve c gives the results
after another five hours heating at the same temperature in "isolated"
carbon monoxide. Thus we see that the curves have undergone
a gradual change in form and position due to the action of carbon
monoxide alone. Such methods as these will permit of the elimina-
tion of the dangerous hyper-eutectoid zone, and at the same time give

jto.

1.3-

1.2-

1.1

1.0

0.9-

0.8-

0.7

0.6-

0.5-

0.4-

0.3-

0.2

o.i H

9 MM.

FIG. 66. — Distributive Action of Carbon Monoxide. (Giolitti.)

all the benefits to be obtained from a carburized zone of the inter-
mediate type previously described.

TEMPERATURE AND TIME FACTORS

Solution of the Carbon. — We have seen how certain gases, by
diffusing into the steel, precipitate free carbon within the steel.
Now this carbon, under suitable conditions, may be dissolved at once
by the iron, forming a true steel. It is evident that the solubility of
this carbon (or carbide) must depend upon the allotropic condition of
the iron, which, in turn, will depend upon the temperature. As we
have explained in previous chapters, iron may be held in the alpha,
beta or gamma state. Thus, if a piece of normal 0.2 per cent, carbon

126 STEEL AND ITS HEAT TREATMENT

steel is heated, none of the cementite which is mechanically mixed
with ferrite (iron) to make up the mechanical mixture pearlite is
affected until the lower critical temperature of about 1350° F. is
reached. At this temperature the iron of the pearlite, previously
in the alpha condition, changes into gamma iron and dissolves the
cementite, the two forming a solid solution or austenite. In other
words, it is necessary for the iron to be in a higher allotropic condition
than that of the alpha stage in order to dissolve carbon (or carbide) .
As the temperature of the steel is progressively raised, more and more
of the excess iron is dissolved by the austenite, until at the Ac3 range
the whole mass of the steel consists of austenite and has all the
iron in the gamma state.

Thus we see that while it is possible for carburization to take
place at temperatures varying between the Acl and Ac3 ranges,
the carburization must necessarily be not only slow, but also irreg-
ular and non-uniform. In other words, the minimum temperature
which should be used in industrial carburization should not be
lower than the upper critical range of the initial steel to be
carburized.

Depth of Penetration. — All commercial carburizing processes
must provide a depth of case which will satisfy the requirements of
the specific use to which the steel will be put in practice. Thus
many parts will only require a depth of case of say -^ or ^ of an
inch, or if grinding is necessary, this may be increased to y^ of an
inch; other parts, such as armor plate, may require a considerable
thickness. With a definite depth of case in view, economic consider-
ations require that the velocity of penetration shall be definitely
known in relation to the factors of time and temperature, the nature
of the carburizing agent, and — in the case of gases — such other factors
as pressure, quantity of gas, etc.

The penetration of carbon (differentiating this depth of penetra-
tion from the distribution of the carbon) increases with the temper-
ature and with the time of exposure, but not always in direct pro-
portion to these two factors. Given a definite temperature and
carburizing compound, it may be said in general that the carburiza-
tion commences and continues at a comparatively high rate of speed
until the outer layers are saturated with carbon — dependent, of
course, upon the nature of the cement; there is then a drop in the
rate of carburization, varying according to the temperature, and this
in turn is followed by a velocity of penetration which seems to be
more nearly proportional to the length of exposure.

CASE CARBURIZING

127

These facts are shown graphically in Figs. 67 and 68, the former
illustrating this velocity of penetration particularly for short expos-
ures, while the latter emphasizes the effect of long continued heating.
In each case the experiments were carried out under conditions usu-
ally adopted in industrial establishments. The bars of soft steel
were allowed to soak at the stated temperatures for definite lengths
of time and the penetration was then measured. The results from
which these graphs were plotted were reported, in the first instance,

1234567

FIG. 67. — Velocity of Penetration, Short Exposures.

•

by a company using a cement of the barium carbonate-carbon-
aceous type, and in the second case by Giolitti, using a common
commercial cement consisting of ground-wood charcoal treated
with 5 per cent, potassium ferrocyanide and mixed with an equal
weight of barium carbonate.

In the case of the first compound it is interesting to note that
there is a decided decrease in the relative rate of penetration when
the depth of case reaches approximately 0.05 in. This cycle
appears to take place in the same order at all temperatures used,
with the difference that the relative speeds of penetration are higher

128

STEEL AND ITS HEAT TREATMENT

at higher temperatures, although not proportionally so. Thus we
may coin a phrase and call this depth of 0.05 in. the " critical
penetration" of this individual compound. Considering the depth
of penetration only, and disregarding other economic and technical
factors which may enter into consideration, it is evident that this
particular compound may be used to good advantage when a case
corresponding to about 0.05 in. is desired. To this depth the
steel will be carburized at a maximum speed or velocity, and there-

.14

.12

.10

.08

12 24 36 48 60 72 81 % 103 120 Hours
FIG. 68. — Velocity of Penetration, Long Exposures. (Giolitti.)

fore at the minimum furnace or heating cost. With this particular
compound simply as an illustration, any other commercial carburizing
compound might be studied from the practical side for the determina-
tion of this critical depth of penetration, and use made of the results
to good advantage in reducing operation costs.

The influence of temperatures of carburization under the upper
critical range upon the depth of penetration and the maximum car-
bon content is also brought out by a comparison of Figs. 69 and 70
These photomicrographs represent the effect of cementation upon a
0.11 per cent, carbon steel carburized for one hour in a mixture of

CASE CARBURIZING

129

charcoal and barium carbonate, but at different temperatures.
Fig. 69 shows the results of a temperature under the A3 range;
Fig. 70 the cementation at a temperature considerably over the A3
range. The practical value is self-evident.

Liquation. — Sudden variations in the concentration of the carbon
in the cemented zone may be manifested when intense carburiza-
tion is effected at high temperatures, and the carburized pieces are
allowed to cool slowly through a more or less wide interval of tem-

, 3~'"A- £ • . -^*" 7 - .. '•*&*.'•£? - - Vv

're'vZ^~ »- '^' * '''^^w^-i;-^

./>.' ...dr. ..«"-••*.,'-. •?- * V«."^"iV. » >ty

FIG. 69. — Carburization of 0.11 per
cent. Carbon Steel at a Temper-
ature under the Upper Critical
Range with BaCO3 and Char-
coal, for One Hour. (Nolly
and Veyret.)

•>' ** * '!A** • V * ' ^ ••"*• V •* •V*'^

! K"*\ > * " '^ f '••C^i^

^ *" * » V; "'' »^*SJy*'4^*^'"''»^^f

FIG. 70. — Carburization of 0.11 per
cent. Carbon Steel at a Temper-
ature Considerably Over the
Upper Critical Range with
BaCOa and Charcoal, for One
Hour. (Nolly and Veyret.)

perature before being quenched. This variation consists of a true
liquation of the cement ite (and of the ferrite) during their segrega-
tion front the solid solution. Take, for example, the diagrams in
Figs. 71 and 72, representing the results obtained by carburizing a
0.26 per cent, carbon steel for four hours at 1830° F. in ethylene, with
the difference, however, that the carburized steel represented by
Fig. 71 was cooled — during 32 minutes — to a temperature of 1380° F.
and then quenched, while that of Fig. 72 was quenched immediately
following Carburization from that temperature (1830°).

Comparing the two diagrams we see that, while the concentration
of the carbon in Fig. 72 decreases continuously and uniformly as we

130

STEEL AND ITS HEAT TREATMENT

proceed from the surface towards the core, that in Fig. 71 shows a
marked increase, followed by a very rapid decrease, before it exhibits

*c.

1.6

1.4
1.2
1.0
0.8
0.6
0.4
0.2

-— _,___

""""-

•*N^^

1.4 1.0 1.8

2.2 MM.

FIG. 71. — Liquation of Cementite through Slow Cooling. (Giolitti.)

that gradual decrease which characterized the first case. As the
carburization proper was identical in both cases, it is evident that

JfC.
JL6

±2

1.0
0.8
0.6
<U
03

.* .4 .6 .8 1 1.2 1.4 1.6 1.8 MM.

FIG. 72. — Prevention of Liquation by Quenching, (Giolitti.)

the discontinuity in the carbon distribution must be due to the dif-
ference in the rate of cooling following the carburization proper.

CASE CARBURIZING

131

Thus we see that the liquation, or accumulation, of the cementite
in the external layers will tend to emphasize the line of demarkation
between the hyper- and hypo-eutectoid zones. And this, in turn, will
magnify the dangers of enfoliation. Fig. 73 is a photomicrograph
from a piece of carburized steel which failed in service; the cause
of enfoliation in this case is undoubtedly due to this phenomenon
of liquation.

This is another reason why those processes of carburization should
be used which will avoid the formation of the hyper-eutectoid zone,

x*.

'^j^a&£''^P

^^ati^**

FIG. 73. — Failure Due to Liquation. (Giolitti.)

and thus eliminate the possibility of enfoliation through differences
in carbon distribution caused by accumulation of the free cementite.
However, if cements which give carburized zones of the hydro-
carbon type are used, it is apparent that this accumulation may be
avoided by quenching from the temperature of carburization.
Although this is possible in the newer processes of using either a
gaseous mixture or a mixed cement, it is manifestly impossible under
the older methods of using solid cements. We will shortly describe
methods of heat treatment by which the effect of the cementite
accumulation may be corrected in the majority of instances, as well

132 STEEL AND ITS HEAT TREATMENT

as giving reasons why quenching from the temperature of carburiza-
tion — when it is 1750° F. or more — is technically bad.

Oscillating Temperatures. — There is still another effect of tem-
perature which should be mentioned in this connection on account
of its action in many industrial carburizing processes. This is the
effect of non-uniform or oscillating temperatures during carburiza-
tion, so often met with in practice. Without going into the theo-
retical explanations involved, it has been shown by Giolitti and
Scavia that, where under normal conditions the formation of free
cementite cannot take place, by carrying out the carburization under
identical conditions but with variable temperatures oscillating within
definite intervals, the occurrence of free cementite will result. The
industrial importance of this is evident. In the first place it explains
the abnormal increase in free cementite which is sometimes met with
in practice in which the normal nature of the carburization should
produce cemented zones of the intermediate type and the absence of
free cementite, besides demonstrating the necessity of maintaining
uniform temperatures throughout the whole carburization within
a very close range. And in the second place it furnishes a means of
carrying out the heating during carburization in such a way as
purposely to cause, with certainty, the formation of free cementite
when it is desired to obtain cemented zones capable of taking an
exceedingly high degree of hardness by quenching without its being
necessary that their brittleness be reduced to a minimum.

Temperature Factors. — Although much has been said and
written concerning the effect of temperature upon carburization, the
majority have confused the velocity or depth of the carbon concen-
tration with the maximum concentration and distribution of the car-
bon. While it has been shown that the depth of penetration is, in
general, a direct function of the temperature with all commercial
carburizing mixtures, the same is not entirely true of the maximum
concentration of the carbon. In fact, through the study of pure
gases such as we have indicated, it has been shown that in the case
of carbon monoxide alone the maximum concentration of the carbon
actually decreases, other things being equal, with increase in the
temperature of carburization.

In the cases of the newer processes involving the uses of gases or
a mixed cement, the effect of such temperature on the maximum
concentration and distribution of the carbon may be practically
varied at will by a change in the other factors previously mentioned.
In other words, on account of the almost absolute control with which

CASE CARBURIZING 133

such processes of carburization may be regulated, the actual effect of
the temperature is minimized and does not play the important part
which is manifested in the older processes involving the use of solids
alone.

Influence of Temperature on Different Cements. — The com-
plexity and lack of control of the reactions involved in the use of
solid cements make the factor of temperature extremely important.
Thus some cements, such as charcoal plus barium carbonate, which
at the lower temperatures give " gradual " or " mild " cases, at the
higher temperatures may act as " sudden " or " quick " cements.
A consideration of the common solid cements (of which we are now
speaking) as used in general commercial work would tend towards
the conclusion that their action, on the whole, is more gradual at
the lower temperatures than at the higher temperatures. Further,
it might even be said that, other things being equal, the higher the
temperature of carburization the higher will be the maximum carbon
concentration in the case. Although these statements may not hold
good for all instances in which carbon is used as the base, their prac-
tical working out is generally evidenced in the general average of
thin cemented zones found in commercial case hardening.

Low Temperature Carburization. — Now as the majority of case-
hardened products require the elimination of brittleness, both in
case and core, and as the use of " gradual " cements at the lower
temperatures of carburization will advance that condition of affairs
(i.e., the formation of cemented zones of the intermediate type,
showing the absence of the hyper-eutectoid-zone, and a gradual
distribution of the carbon concentration from the external surface
of the case to the core) the use of the lower temperatures of carbur-
ization is coming more and more into vogue. That is, the tendency
is to use moderate heats and maintain them for a length of time
sufficient to obtain a reasonable depth of case. These heats may be
said, in a general way, to correspond with temperatures of about
100° F. over the upper critical range of the steel to be carburized.
Although the depth of case is largely dependent upon the temper-
ature, as well as upon the time of carburization, under the above con-
ditions it should be considered poor practice to raise the temperature
to the high limit simply for the purpose of reducing the time element;
the repair items on the furnace will increase, the fuel cost will be
greater, and— above all — the maximum carbon concentration and
the relation of the various zones to each other so changed that
the whole character of the finished product may be altered.

134

STEP:L AND ITS HEAT TREATMENT

Relation of Temperature to Grain-Size. — Another important
feature in determining whether or not to use a high temperature
for case carburizing is the relation of such temperature to the grain
size. As has been explained in previous chapters, upon passing the
Acl range on heating the grain size begins to coarsen, and most
noticeably so after passing the upper critical range or the low carbur-
izing temperatures. Thus from about 1550° F. and so on up to
1850° or 1900°- F. (which are about the maximum temperatures
used) the grain increases in size most markedly. This increase
in grain size has a direct bearing upon the impact strength of the
steel, as is shown by the following results (by Guillet) obtained by
annealing case-hardening steel bars at the temperatures given for
eight hours, and which under the conditions of a normal annealing
gave an impact test of 28 kilogram-meters.

Temperature,

Impact Test,
Kilogram-metres.

1470

1560

1650

1740

1830

1920

2010

Thus it will be seen that any heating of long duration at temperatures
above the upper critical range (the low case carburizing temperature)
greatly lowers the resistance of the steel to impact or shock. When
high temperature carburizing is necessary, however, suitable treat-
ment after carburizing may be used to " regenerate " the core.
Even then, however, the effect of long-continued heating at high
temperatures is often manifest.

High-temperature Carburization. — Although the advisability
of using the lower temperatures for case carburizing has been
emphasized, it must not be thought that the higher temperatures
of 1800° F., or even higher, are never to be used. Contrarily, the
latter are often mandatory in certain classes of work where speed
of penetration is the first requisite, or where low cost of production
is necessitated and the absence of brittleness is not a prime factor.
If overheating and burning are suitably guarded against, and the
methods of packing are such as will keep warping and distortion at a
minimum, and the carburizing process is followed by a technically

CASE CARBURIZING 135

adjusted series of heat treatment operations, most excellent results
can be obtained. In fact, it is only within comparatively recent
years that the use of the lower temperatures has been practiced
to any great extent; a decade or so ago if 1550° F. or thereabouts
had been suggested as giving greater efficiency, the proposal would
have been laughed at by the majority of " practical " hardeners.

COMMERCIAL DATA

Simple Solid Cements. — In the foregoing sections we have
attempted to give in brief the underlying principles which govern all
processes of partial carburization. By the use of these principles
we have further shown that, in carburizations with gaseous or certain
mixed cements, it is possible to so select the conditions of carburiza-
tion as to obtain with certainty cemented zones of predetermined
and definite form. Similarly, the same principles are applicable
to carburization with solid cements, although (as we have shown)
it is not possible to even approximate the same accuracy on account
of the lack of control of such cementations. Nevertheless, a study of
the preceding pages should convince one of the importance of using
carburizing compounds, the composition and manner of acting of
which are definitely known. With such principles and ideas in mind,
it should be comparatively easy for each one to prepare for himself those
cements which are much more simple, effective, and of less cost than
many of those purchased from dealers at high prices, under fancy names,
and of unknown composition. With these thoughts in mind, we will
briefly discuss some of the more simple compounds used in commercial
work.

Wood Charcoal. — Finely divided carbon is the simplest of the
solid cements, the purest form in commercial practice being wood
charcoal. As we have seen, the carburizing activity of powdered
wood charcoal is dependent upon the formation and action of carbon
monoxide, and which is further diluted by the nitrogen of the occluded
air. It is * evident that the use of this charcoal in the ordinary
short carburizations for the production of cases of ^j to ^ in. in
thickness will have the tendency to give cemented zones of low
and irregular carbon content. Its use for the deeper carburizations,
however, may be distinctly advantageous, as it opposes the formation
of zones too high in carbon. Figs. 74 and 75 show the effect of tem-
perature upon carburizations with wood charcoal, the first figure
representing carburizing at 1560° F., and the second photo-
micrograph at 1925° F,

136

STEEL AND ITS HEAT TREATMENT

Animal Charcoal. — Thus, for thin cases, wood charcoal is gen-
erally mixed with certain proportions of the less pure charcoals,
such as those produced by the carbonization or charring of leather,
bones, hoofs, horns, hair and other animal refuse, etc. In these
last cements we therefore have the action of pure carbon or char-
coal greatly intensified through the generation of volatile hydro-
carbons. We will later mention the influence of the phosphorus
and sulphur which these cements may contain. By mixing wood

^p^fs^^p

- .v.-.-^v--*. i*?^jap-vT4'

«-:''^i^-^.^»t

"• r / .->*&««• -•• - - &/•.: t-"^: *^

FIG. 74. — Carburization with Char-
coal for One Hour at 1560°F.,
0.11 per cent. Carbon Steel.
(Nolly and Veyret.)

?-i/vv:

y.vt«j":» rr..,, - -v.

^'Kisi

'^'tit/rfv:^^^^^

FIG. 75. — Carburization with Char-
coal for One Hour at 1925° F.,
0.11 per cent. Carbon Steel.
(Nolly and Veyret.)

charcoal with definite proportions of these animal charcoals, the
carburizing action may be roughly adjusted between the minimum
value to be obtained with wood charcoal and the maximum value
of the animal charcoals. Of the more common " mild " cements
thus obtained we may mention the following:

Parts

a. Powdered oak charcoal 5

Powdered leather charcoal 2

Lampblack 3

b. Wood charcoal 7

Animal charcoal .3

CASE CARBURIZING 137

Parte

c. Powdered beech charcoal 3

Powdered horn charcoal 2

Powdered animal charcoal 2

Common Salt. — Common salt (sodium chloride) is used in many
works in addition to charcoal, it seeming to give better results than
wood charcoal alone. Exactly what is its specific action is not
thoroughly understood. Thus we have the mixture:

Parts

Wood charcoal 7 to 9

Common salt 3 to 1

Barium Carbonate. — One of the best solid cements for general use
is that consisting of:

Parts

Barium carbonate . . 40

Powdered wood charcoal 60

Its action is well known and is as we have previously described.
For cases of small depths it gives carburized zones markedly more
homogeneous than those furnished by other solid cements. Giolitti
sums up its advantages as follows: " In general, the maximum con-
centration of the carbon in the cemented zones obtained with carbon
and barium carbonate at temperatures between 1650° and 2010° F.
varies from a minimum of about 0.7 per cent., for the very thin zones
obtained near 1650°, to a maximum of about 1.3 per cent, for the
zones thicker than 1 mm. (0.04 in.) obtained near 2010° F.

" Another advantage of this cement lies in its property of being
' regenerated ' easily and spontaneously when it is left exposed in a
thin layer to the air, after having been used in the usual manner.
This process of regeneration is due to the fact that the barium oxide
formed during cementation by the dissociation of the barium car-
bonate, absorbs carbon dioxide from the air, again forming barium
carbonate. After a certain number of alternating cementations and
regenerations it is necessary to add some wood charcoal to the cement
to replace that burned during the cementation and during the dis-
charging of the boxes.

" The preparation of this cement consists simply in finely grind-
ing and intimately mixing the wood charcoal and barium carbonate.
If the natural barium carbonate (witherite) is used, it is necessary
to powder it carefully before adding it to the carbon; the finely

138

STEEL AND ITS HEAT TREATMENT

divided precipitated barium carbonate, on the contrary, can be
mixed directly with the granulated carbon and the one operation of
grinding the carbon can be used for preparing the mixture."

It is, of course, not always necessary to use the above mixture
ratio of 40-60, although this combination has been shown to give
about as good results as may be obtained. The conditions of heat-
ing, temperature, size of the pieces, type of carburization box and
method of packing, etc., will alter each individual carburization,
and experiments should be made to determine as exactly as possible
the proper combination of the different factors of carburization which

24 6 8 10

Duration of the Heating (Hours)

FIG. 76. — Carburization with Common Carburizing Compounds.

(Scott.)

enter into consideration. One of the governing factors which is often
overlooked is the action of the charcoal, dependent upon its composi-
tion. Thus much of the ordinary commercial charcoal still contains
considerable volatile or organic matter (hydrocarbons) which may
distinctly alter the effect of the carburizing. In order to reduce
the intensifying action of such constituents, and to reduce the forma-
tion of the hyper-eutectoid zone, it is always advisable first to
calcine the charcoal before using.

The general relation existing between the depth of penetration
due to charcoal, charred leather, and the usual 40-60 barium car-
bonate-charcoal mixture is graphically shown in Fig. 76, obtained by
Scott in the carburization of soft steel bars at 1650° F.

CASE CARBURIZING 139

Gradual Cements. — The cements which we have just enumerated
are generally classed as " gradual," for reasons previously given.
Yet on the other hand, these same cements, under different con-
ditions of carrying out the carburization, act as " sudden " or
" quick " cements. Thus the barium carbonate mixture when used
at low temperatures, or for the carburization of pieces of large
dimension which heat up slowly, may furnish cemented zones in
which the maximum carbon concentration may not be over the
eutectoid ratio (0.9 per cent.). The same mixture, on the con-
trary, may become a sudden cement at the very high temperatures
and in carburizing objects of small dimensions.

Other Solid Cements. — In addition to the use of charcoal plus
the animal charcoals, barium carbonate and common salt, the other
agents which may be added are innumerable. To give a list of them
would occupy several pages, besides leading to the inevitable con-
clusion that the efficacy of the majority of them is small, or might
be even detrimental. For it should be stated with emphasis that
the more simply and more chemically definite a cement can be made,
the greater will be the industrial advantages.

Sudden Cements. — The nature of the most important of these
additions is to make the mixture act quickly, giving rise to a thin
cemented zone of high carbon in a very short interval. Thus we
have the use of coke saturated with mineral oil, of the saturation of
the charcoal in solutions of cyanides or ferrocyanides, and of the
presence in greater or less quantities of the concentrated salts of
cyanogen as specific additions. Of those used in practice, the follow-
ing example is extremely interesting;

11 Ibs. prussiate of potash,
30 Ibs. sal soda,
20 Ibs. coarse salt,

6 .bushels powdered hickory charcoal,
30 quarts water.

Grenet recommends the following cements which have given good
results in practice:

Parts.

a. Powdered wood charcoal 1

Salt i

Sawdust 1

140 STEEL AND ITS HEAT TREATMENT

Parts

6. Coal with 30 per cent, volatile matter 5

Charred leather 5

Salt 1

Sawdust 15

c. Charred leather 10

Yellow prussiate 2

Sawdust : 10

The velocity of carburization increases gradually from the first to
the third of these cements. The sawdust, by making the mass
more porous, increases the activity of the gases.

Size of the Carburizing Box. — The selection or general design of
the box or container for carburizing is worthy of more attention than
is frequently given to it. In the attempt to get a uniform case, much
thought and research has been given to the selection of the steel,
the carburizing mixture and the degree and duration of heating; and
yet in many instances it has all proven unavailing. It must be
remembered that the inside of a small box takes quite a while to
come up to the temperature of the furnace; and that if a large box
is used, the material in the center may, and does, lag behind the
indicated furnace temperature several hours or its time equivalent —
several hundred degrees. The greater the size of the box, the larger
will be this error, and the greater the actual difference in the thick-
ness of case taken on by steel near the sides of the box as compared
with that near the center of the box. No manipulation of the furnace
can change this effect; it can only be remedied by altering the
dimensions of the box itself. Here, then, lies one explanation of
many unexplained failures.

The box should not be larger than is absolutely necessary, even
where large quantities are to be carburized in it. It should be narrow
in at least one dimension so that the heat has a chance to penetrate
quickly at least from two sides and reach all the contents at about
the same time. Further, the boxes should not be made too deep in
proportion to their other dimensions, as it makes it more difficult
to pack the parts into them if so made. Whenever possible, the
design of the box should follow the outline of the piece to be carbur-
ized, allowing about 1 to 2 ins. all around for clearance and packing,
so that the surfaces may be uniformly heated and carburized alike.

Material for Boxes. — Malleable iron probably gives as good
satisfaction as any of the materials used in making the boxes. Cast-

CASE CARBURIZING 141

iron boxes, although of comparatively small initial cost, will not stand
reheating very many times, and have the further objectionable
feature of being somewhat porous. Soft-steel plates and wrought
iron may also be made up into good boxes.

The thickness of the wall forms an important feature of the box,
for if it is too thin it easily burns through, and if too thick it offers
too much resistance to the penetration of the heat to the interior.
For the ordinary size boxes, wall thicknesses of J to J in. are common
practice. The boxes should be provided with feet so that the heat
may circulate all around them. The cover should be as close fitting
as is practicable, and should also be provided with ribs along the top
to prevent excessive warping. Ribs along the side also add to the
service of the box, besides making handling with grappling irons
more easy. The sides of the boxes should taper slightly towards
the bottom so that the contents can be the more quickly dumped out.

Packing. — Carefulness in packing is fundamental to good practice
and uniformity of results, just as much as carefulness in heating or
treatment. The method of packing should be such as will insure
as nearly as possible the even heating and uniform carburizing of all
pieces in the same box.

The method of packing must necessarily vaiy with each type
of article to be handled. Heavy pieces, or pieces of regular shape,
do not require the care and patience which should be used with pieces
of intricate design or with those which on account of their size and
shape may be readily influenced by high temperatures. The packing
of such pieces must be individualized. For example, long, slender
pieces should always be packed vertically, so that the pieces will
be held in position by the carburizing material and cannot sag under
the influence of the high temperatures. Again, gears and similar
pieces may be most suitably packed in tubes, so that the same
amount of carburizing material and the same degree and length of
heating may influence all parts of the periphery in equal proportion.
In carburizing screws and bolts it is well to distribute them in
the box in two opposite rows, each row having the head of the screw
towards the side of the box and the stem towards the center. New
compound may be used at the sides and old compound in the center.
Owing to the difference in heat and the difference in the carburizing
power of the compounds, this will cause a much deeper carburization
of the heads than of the stems — which is exactly what is desired.
Again, should a narrow or low box not be available in connection
with small work, carburizing compound which has been used once

142 STEEL AND ITS HEAT TREATMENT

before may be put next to the sides of the box while the new com-
pound is placed in the center. In this way the difference in car-
burizing which might result from the different temperatures in
various parts of the box may be offset.

The first step in the general operation of packing is to cover the
bottom of the box with the compound to a depth of 1J to 2 ins.,
tamping it solidly into place. The parts to be carburized are
then placed firmly upon this bed so that the compound and work
are in close contact with each other. The pieces should in no case
touch the sides of the box, but should be placed about 1 to 1J ins.
away from it. Further, the articles should be separated from each
other by at least \ to 1 in., dependent upon their size and the depth
of case desired. If the articles should touch one another, it is evident
that the carburizing action will have less influence at that particular
point — with resulting soft spots. Non -uniformity of case may also
result if there is not sufficient carburizing material in the box; it is
better to err by using too much than too little.

After the first layer of work has been placed in the box, it is
entirely covered with the carburizing compound. This should
be packed and tamped down around and over the pieces so as to
have the particles of cement in close contact with the steel, but yet
not so tightly as to prevent the free circulation of the carburizing
gases which are generated during the heating process. When the
first layer has thus been suitably packed and covered, the same
procedure is repeated until the box is nearly filled. The point to be
kept in mind is that each and every piece should be surrounded on
all sides by a suitable amount of the carburizing compound.

At least 2 ins. of the compound should form the top blanket over
the last layer of work. Some shops adopt the following, with the
aim of further preventing the escape of the gases : about 2 ins. from
the top of the box sheet-steel strips about -^ in. thick are laid over
the last layer of the carburizing material and these, in turn, are
covered with about 1 in. of powdered charcoal. When the box is
finally packed, the cover is placed on the box and the edges are care-
fully sealed with fire-clay or asbestos cement. The box is now
ready for the heating operation.

Type of Furnace. — It is not our intention to recommend any
particular type of furnace for carburizing work, but rather to
emphasize the necessity of designing the furnace to suit the work.
And, as is evident, the conditions will vary greatly from one plant
to another.

CASE CARBURIZING 143

There are three main points which should be taken into account
for case-hardening furnaces. (1) The furnace shall be capable of
attaining easily the maximum temperature which shall be necessary
for the carburizing work, and which temperature may be as high as
2000° F. (2) It must be possible to obtain a thoroughly uniform
heat application at any of the intervening temperatures, and of
maintaining that uniform heating with little or no variation hour
after hour. (The effect of oscillating temperatures has been de-
scribed.) (3) The atmosphere in the furnace shall be non-oxidizing,
in order to protect the carburizing boxes from the intense
oxidation which would otherwise occur at the high temperatures
necessary.

The Heating. — The two principal points to be mentioned under
this heading are: the heating, at least up to 1300° F., should be
gradual; (2) the heating beyond this temperature should be uniform
over all parts of the carburizing box. It has been shown by several
experimenters that the energetic liberation of gases commences
very strongly at temperatures somewhat under 1300° F. for the
majority of solid cements, and it is advisable to diminish this factor
as much as possible in order to obtain a more gradual cementation.
Furthermore, it gives more opportunity for the steel to adjust itself
to the effect of heating. The second point made is self-evident:
non-uniformity of heating must necessarily result in non-uniformity
of product.

Sulphur Diffusion. — The influence of sulphur contained in the
cements is an extremely important factor in carburization carried
on with solid cements. Grayson l has produced uncontrovertible
evidence that sulphur will diffuse into iron at the temperatures
ordinarily used for carburization with such substances as charred
leather (which, under the conditions of his case-hardening experi-
ments, contained 0.55 per cent, total sulphur), and that this sulphur
combines with the manganese and iron to form manganese and iron
sulphides.

Thus in Fig. 77, which is a photomicrograph of a piece of 0.17
per cent, carbon steel carburized for six hours at 1650° to 1750° F.
with charred leather, it will be noticed that on the edge are present,
in large quantities, sulphide of manganese, also sulphide of iron with
ferrite crystals intermingled. That this is sulphide was later proven
by means of silver prints and by analysis — which showed 2.10 per
cent, of sulphur increase in the first 0.0025 inch.

!S. A. Grayson, Inst. Journ., No. 1, 1910.

144

STEEL AND ITS HEAT TREATMENT

This sulphur diffusion is a very serious matter, because when the
surface is saturated, as in this figure, it tends to produce a soft skin,

FIG. 77.— Soft Case Due to Sulphur
Diffusion. (Grayson.)

•

FIG. 78. — Sulphides Diffusing
Further into Case with Higher
Temperatures of Carburization.
(Grayson.)

FIG. 79. — Sulphide Globules in Carburized Steel after Hardening. (Grayson.)

and even if present in smaller proportions it will weaken the structure
considerably, thus making it very " chippy," consequently causing

CASE CARBURIZING

145

two effects which must essentially be avoided in any case-hardened
work.

In Fig. 78, being a similar steel carburized at 1750° to 1830° F.,
the sulphide is again present, but not in such a large proportion;
thus the higher temperature has volatilized still more of the sul-

FIG. 80. — American Gas Furnace Co.'s Carburizing Machine.

phur from the carburizing material. Fig. 79 shows the same car-
burized piece as in Fig. 77, but afterwards reheated and quenched
in water from 1380° F. In this reheating the sulphide tends to
" ball " itself up, and, if anything, diffuse further in.

Thus it may be seen that, for proper carburizing, the solid
cements should be as free from sulphur as is possible. On the other

146

STEEL AND ITS HEAT TREATMENT

*b f

CASE CARBURIZING

147

hand, the barium carbonate mixtures generally used do not contain
sulphur, and this sulphur diffusion cannot take place.

American Gas Furnace Process. — The apparatus for carburizing

with gas, as devised by the American Gas Furnace Company, is
shown in Fig. 80. The carburizing machine consists of a carburizing
retort enclosed by a cylindrical furnace body in which it rotates,

148 STEEL AND ITS HEAT TREATMENT

together with suitable arrangements for charging and discharging
the work, burners for securing a proper distribution of the fuel, and
supply pipes for gas and air. The machine shown has a space
available for work of 30 ins. in length by 7 ins. in diameter. It is
suitable for work not over 6 ins. in diameter or 20 ins. in length; for
shafts, tubes, mandrels and bars of nearly equal thickness through-
out of not over 24 ins. in length or 5 ins. in diameter; or for small
pieces such as screws, washers, discs, etc., of a charge of about 100
pounds. The machine uses ordinary illuminating gas for both heat-
and carburizing.

The vertical section, Fig. 81, through the center lengthwise,
shows the heavy wrought-iron retort A, which is slowly rotated on
the rollers BB by the gear C, in contact with worm D, propelled by
a sprocket and chain belt. The reference letters EE show air spaces
in the retort formed by the two pistons /, between which the work
is confined, to the properly heated central section of the retort.
Letters FF indicate the heating space surrounding the retort, into
which the fuel gas and air are injected under pressure, from two rows
of burners indicated in the upper half of the casing by the letter G.
The cover H, closing the retort, is connected with the piston-like
disc marked /, by the pipe J, which is the vent of the retort. The
cover H and disc I are withdrawn to charge the retort and replaced
after the work is inserted.

Carburizing machines connected with an automatic quenching
bath are shown in Fig. 82.

SUPERFICIAL HARDENING

Superficial hardening differs from case carburizing in that in
the former method the outer and higher carbon section constitutes
a " skin " of only a few thousandths of an inch in thickness, while
in the case-carburizing process the carburized zone forms a case
of noticeable thickness. Exactly the same principles apply, however,
in both instances, and which have been previously explained.

Processes. — The superficial hardening processes may be grouped
under the headings of " cyanide hardening " and " pack hardening. "

The cyanide hardening processes are essentially used for the pur-
pose of obtaining an extreme degree of surface hardness (wear) on
low-carbon or machinery steel, and in which it is not necessary to
obtain high resistance to shock, etc.

On the other hand, pack hardening is essentially a method of
heating used particularly for fine threaded tools and other tool-

CASE CARBURIZING 149

steel work. The process, when correctly carried out, permits of
uniform heating with the entire elimination of oxidation by surround-
ing the steel with a carbonaceous packing. But further, by prolong-
ing the duration of heating at the hardening temperature, a very
thin skin of higher carbon content may be formed, so that pack
hardening may develop, either intentionally or otherwise, into a
superficial hardening process.

Cyanide Hardening. — In cyanide hardening the superficial car-
burizing and hardening may be effected by one of two general
methods: (1) immersion of the object in a bath of liquid potassium
cyanide or other mixture with cyanogen as the base, followed by
quenching; (2) coating or sprinkling the surface of the object with an
adhesive mixture of finely pulverized carburizing cyanogenous salt
or " varnish," heating the steel to the proper hardening tempera-
ture— and thus melting the cyanide — and hardening as usual.
The first or " immersion " process is by far the most efficient, both
as to unifcrmity of the carburized zone and simplicity and uniformity
of operation. Further, this first method has the tendency to reduce
deformation and oxidation during heating and quenching, since,
as previously explained, heating in any molten bath has this effect.

The Immersion Method. — The method of cyanide hardening by
immersion is quite simple. The salt, usually potassium cyanide
(KCN), is melted in a suitable pot-furnace, and is maintained at a
temperature a little over the upper critical range of the steel to be
carburized and hardened. This temperature, for ordinary machin-
ery steel, is about 1550° to 1600° F. The steel is then immersed
in the molten cyanide and kept there until it has been uniformly
heated; or this heating may be somewhat prolonged in order to
obtain a greater depth of skin. In general, however, it is not advis-
able to heat for a length of time much greater than ten or fifteen
minutes, or at temperatures much over the critical range, since such
heating will tend to give non-uniform and high-carbon zones which,
after quenching, are intensely brittle and may chip off in service.
Quenching is usually done in lime water in order to neutralize the
cyanide remaining on the steel. Some concerns adopt the method
of immersing the steel in the cyanide as soon as it has become molten,
permitting the steel to heat up with the bath, and then quenching
as soon as the desired temperature of say 1550° to 1575° F. has been
attained.

It is absolutely necessary to remember that cyanogen compounds
are deadly poisonous, and every precaution should be adopted when

150

STEEL AND ITS HEAT TREATMENT

using them. Furnaces should be supplied with hoods which have
strong draft. Gloves should be used in handling all work, for if
cyanide gets into a fresh cut or scratch it will prove deadly. In
some cases, when working at the furnaces, it is even advisable to
use face masks and to cover up any exposed parts of the body.

Cyanide Hardening Plant. — A battery of twenty cyanide fur-
naces is shown in Figs. 83 and 84. l In front of the first pair of

FIG. 83. — Battery of Cyanide Furnaces' — Special Quenching Machines for
Clutch Rings in Foreground. ("Machinery.")

furnaces in Fig. 83 two special machines are shown which suddenly
cool or quench the work as fast as it can be heated and removed
from the furnace. They are used for hardening the steel ring discs
shown at U. These alternate with brass discs in a multiple-disc
clutch on the engine of an automobile. Each pair of furnaces shown
in these two figures is covered with a hood to convey the poisonous
fumes to the outer atmosphere through pipes extending through the
roof. In addition to this, sheet-metal shields are located in front of

1 E. F. Lake, in " Machinery," Sept., 1914.

CASE CARBURIZING

151

the furnace openings shown at V to carry away from the workmen
any fumes that might come through these openings. (These shields
were removed for photographing.) At the end of the cyanide fur-

b£

I
I
1

naces shown in Fig. 84 is a stationary tank of lime water in which some
of the work is quenched. On the floor is shown a tray loaded with
bevel differential gears and having a long rod for a handle. This is
lowered into the cyanide bath to heat the gears, then lifted out and

152 STEEL AND ITS HEAT TREATMENT

lowered into the quenching tank, and when cool the gears are dumped
into boxes to take to the tempering furnaces. Other parts that are
being hardened are shown in the metal boxes beside the tray of gears.

Other Cyanide Methods. — The second general process of cyanide
hardening, in its simplest form, is to heat the steel to about 1550° F.
or so; sprinkle upon it, or plunge it into, potassium cyanide or
potassium ferrocyanide; again heat to 1550° to 1600° F. until the
cyanide is melted; and then quench in water. In case the amount
of cyanide obtained the first time is insufficient, the operation previ-
ous to quenching may be repeated until a layer of the required thick-
ness is obtained. It is, of course, necessary to have a clean surface,
free from scale and oxidation, so that the carburizing reactions may
take place readily.

Other more elaborate processes based upon the above are in use,
and involve the application of special carburizing varnishes. On the
whole, however, the simpler the process or carburizing compound,
the more efficacious will it be in actual and everyday practice.

Pack Hardening. — Pack hardening, as a superficial carburizing
process, so raises the carbon content in the surface of the steel that
the tools may be hardened in oil — instead of in water — and still
obtain the requisite degree of either cutting or wearing hardness.
For certain work requiring almost perfect hardening results this
method cannot be overestimated. In cases in which the required
degree of hardness may usually be obtained only by the use of water
quenching, oil quenching may now be used; and with it will be
associated the toughness of core inherent with the use of oil as a
quenching medium. Further, on account of the uniformity in heat-
ing and the use of oil quenching, the tendency to crack or warp is
largely eliminated.

The method is, in fact, a case-carburizing process. The packing
in boxes is carried out in exactly the same manner as is carburization
with solid cements, and similar precautions should be used to prevent
the tools being jarred out of position or touching each other. In
pack hardening, however, to each tool or piece of steel should be
attached a wire, so that the tool may be removed promptly to the
quenching bath when the requisite degree and duration of heat has
been attained.

The temperature to be used should be but slightly over the
critical range of the steel, thus differing from the higher temper-
atures which are customary in case-carburizing processes. As the
pack-hardening process is usually used for steels of tool-steel analysis,

CASE CARBURIZING 153

this temperature will be about 1375° to 1400° F. The length of
time required for the heating will, of course, depend upon the size
and number of the pieces to be treated in one box, and the depth of
skin desired; for ordinary small tools this will generally be about
two hours after the proper temperature has been attained by the
steel itself.

Packing material which would be harmful to the steel should not
be used. Bone, for example, usually contains phosphorus, which is
apt to make the steel brittle — although burnt bone is not as high in
this element as is raw bone. Sulphur must also be guarded against.
If the initial steel does not contain more than 1.20 or 1.25 per cent,
carbon, charred leather makes a very good packing material. If
the carbon content exceeds these values, charred hoofs, or a mixture
of charred hoofs and horns is better than charred leather, since the
latter will under such conditions have the tendency to give a too
highly carburized and brittle zone.

The temperature of the steel in the box may be gauged by means
of test rods the same size as the tools, or by test wires, or by suitable
pyrometer equipment.

CHAPTER VII
CASE HARDENING: THERMAL TREATMENT

Heat-treatment Requirements.— It may be said that practically
all objects which have undergone the case-car burizing processes
previously described require a subsequent heat treatment of some
nature. As one of the essential aims of the case-hardening process
is to produce a hard-wearing surface, and as carburized steels through
their slow cooling from high temperatures will be more or less lack-
ing in this necessary hardness, it is evident that a hardening process
is necessary. In dealing with the subject of case hardening we will
therefore assume that the carburized steel must undergo some
hardening process or processes which will bring about this desired
condition of affairs.

Secondly, in order that we may at once differentiate the ultimate
aims of such hardening, and simplify our discussion, we will assume
that we also desire to obtain a minimum brittleness in both case and
core. Previous explanations prove that this condition requires that
the " grain size " be reduced to a minimum, that is, that the steel
as a whole must be refined.

To sum up, the specific aims which we have in view require that
the heat treatment shall combine hardening and grain refinement.

Comparison with Homogeneous Steels. — The heat treatment of a
carburized steel differs from that of a homogeneous steel only in the
fact that, instead of considering the influence of such treatment
upon one steel, we really have to do with two, or even three main
classes of steels at once. That is, the carburized steel consists of
(1) the core, or low-carbon steel; (2) the carburized zones of the case
with about 0.9 per cent, carbon as the maximum; and, in many
instances, (3) the carburized zones of the case which, under conditions
of slow cooling from the temperature of carburization, contain an
excess of free cementite, i.e., greater than 0.9 per cent, carbon. Our
heat treatment must therefore be adjusted so as to superimpose
the effect of one class upon the other.

Now as the present tendency of case carburizing in industrial
practice is to preclude the formation of zones containing free cement-

154

CASE HARDENING: THERMAL TREATMENT 155

ite (class 3), we will postpone the discussion of the heat treatment
which involves that class, and thus further simplify matters. We
now, therefore, have but to consider the related heat treatment of
steels of very low-carbon content and those containing the eutectoid
ratio of carbon as a maximum.

Effect of the Temperature of Carburization. — Into this heat
treatment there now enters the factor of the temperature of car-
burization, and its specific influence upon the size of grain in the
two classes of steels. In the first place, it is axiomatic that the
effect of any heating at, or slightly above, the Ac3 range for the soft
steels, and the Acl.2.3 range for the hard steels, is to produce the
maximum grain refinement (unless such heating is extremely pro-
longed) . And further, that the effect of any heating at temperatures
considerably above these temperatures is to produce a size of grain of
proportionally greater size for the respective steels. (In this chapter
we are using the phrase " grain size " in its general colloquial mean-
ing.) Thus, if carburization were carried out at 1600° F. for a case
carburizing steel of 0.15 per cent, carbon — that is, at a temperature
but slightly over that of the upper critical range of the initial steel —
it would produce the minimum grain size in the steel of the core,
and a certain and proportionally greater grain size in the steel of the
case. Steels carburized at the lower temperatures we will call
Group A. Steels carburized at the higher temperatures or about
1800° F. — that is, considerably over that of the upper critical range —
we will call Group B, the grain size of both core and case being
proportionally greater than the minimum.

Classification. — We now have a further means of classifying our
heat-treatment processes according to the temperature which was
used in carburization because of the effect of such temperatures on
the refinement of the steel. That is, with carburizations of Group
A, the steel of the core will already be refined, and we need only
consider the refining of the case; while in Group B the steel of both
core and case will require refinement. Both groups will, of course,
require the hardening of the case.

Treatment of Group A. — Assuming the conditions as in Group A
(carburization at a temperature slightly over the upper critical range
of the initial steel), it is evident that the complete heat treatment
following carburizing will only require the hardening and refining
of the case, in which, by previous assumption, the maximum car-
bon content is about 0.9 per cent. A consideration of the principles
of heat treatment at once shows us that a single quenching at about

156 STEEL AND ITS HEAT TREATMENT

1375° F., that is, slightly over the Al.2.3 range, will bring about the
fulfillment of both of these conditions. Further, the quenching at
this temperature, and under existing conditions, will not affect the
present refinement of the core, nor — if the carbon content is low —
will it in crease the brittleness due to the changing of the pearlite
of the core into martensite (in fact, it has the opposite effect of in-
creasing the toughness in the very low carbon steels) . Thus, by this
single quenching, we have completed the requirements originally
demanded.

Treatment of Group B. — Turning now to case carburizations at
the higher temperatures, Group B, it is evident that in addition to
the refining and hardening of the case we must also refine or regener-
ate the core. This, we know, may be best accomplished by quench-
ing the previously cooled steel at a temperature slightly above the
Ac3 range of the steel of the core. This quenching will put the entire
steel, both case and core, in the martensitic condition, refine the
core, but not refine the case. By following this first quenching by
the quenching at the lower temperature described in the previous
paragraph we accomplish the following : the hardness and refinement
of the case reaches a maximum; the refinement of the core produced
by the first or regenerative quenching is not changed; the strains
or brittleness which may have been produced in the core through
the first quenching are relieved. In other words, by superimposing
one quenching upon the other we have attained the desired prop-
erties.

Effect of Hyper-Eutectoid Zone. — The next variable is that due
to a hyper-eutectoid zone in the case, or carburized steels which con-
tain free cementite upon slow cooling from the temperature of car-
burization. If we apply the treatment previously described under
Group A the condition of this free cementite will not be affected.
This is due to the fact that, upon heating, this free cementite is not
dissolved by the solid solution austenite until a temperature corre-
sponding to the Acm range (see Fig. 13) of the maximum carbon
content of the case is attained, and which is obviously higher than
1375° F.

Influence of Free Cementite. — Now it has been repeatedly
demonstrated in practice that the presence of free cementite existing,
as it usually does, in the form of films between the grains (i.e., as a
network), or even as spines, increases the brittleness of the case,
interposes lines of weakness, and often results in the chipping off
of parts of the case. (Incidentally it might be mentioned that this

CASE HARDENING: THERMAL TREATMENT 157

is another reason for desiring a maximum carbon content in the case
of about 0.9 per cent, when suitably adjusted and controlled methods
of heat treatment are not used.) To eliminate this source of dan-
ger — the free cementite — it will be necessary to heat the steel above
the Acm range of the steel of the case in order to get this cementite
" into solution," and to then " fix " it in that condition by quench-
ing from that temperature. Now, provided that the maximum carbon
content is not sufficiently high so as to raise this Acm range above the
Ac3 range of the steel of the core, it is apparent that the treatment
of Group B previously noted (the double quenching) will also serve
in this instance. This treatment will likewise be applicable, with the
above proviso, regardless of the temperature of carburization.

•' *

V •

* "

7^-' * ( ,«t

FIG. 85.— Core of Steel Carburized at 1830° F. and Slow Cooled. (Bullens.)

Photomicrographic Study. — The principles brought out by this
series of treatments and their individual effect on the case and core
may be more graphically illustrated by means of the series of photo-
micrographs shown in Figs. 85 et seq. The steel in these photo-
micrographs represents an ordinary low-carbon steel which has been
cemented at 1830° F. in such a manner as to produce a carburized
zone containing greater than 0.9 per cent, carbon. In all cases the
steel has been allowed to cool slowly from the temperature of cemen-
tation.

Structure after Slow Cooling. — The micro-structure of the core
upon slow cooling is shown in Fig. 85, it consisting of coarse ferrite

158

STEEL AND ITS HEAT TREATMENT

(light) and a small amount of coarse pearlite (dark). Similarly, the
micro-structure of the external layers of the case is illustrated in

FIG. 86.— Case of Steel Carburized at 1830° F., and Slow Cooled. (Bullens.)

FIG. 87. — Core of Steel Carburized at 1830° F., Slow Cooled, and Quenched from

1375° F. (Bullens.)

Fig. 86, in which it is seen that the laige grains of sorbitic pearlite
are surrounded by the characteristic network structure of free
cementite. In other words, the steel as a whole exhibits the non-

CASE HARDENING: THERMAL TREATMENT

159

refinement characteristic of the high temperature of carburization,
and the case is further weakened by the presence of free cementite.

Effect of Lower Quenching on the Core. — If we should now
quench the steel from about 1375° F., we see from Fig. 87 that the
effect upon the core is to change the pearlite into martensite plus
osmondite (the darker areas), to slightly increase it in amount (on
account of the fact that this quenching temperature is somewhat
above the Al range), but does not give any great amount of grain
refinement to the core as a whole.

FIG. 88. — Case of Steel Carburized at 1830° F., Slow Cooled, and Quenched from

1375° F. (Bullens.)

Effect of Lower Quenching on the Case. Similarly we see from
Fig. 88, representing the micro-structure of the hardened high-carbon
case, that, although the initial pearlite itself (compare with Fig. 86)
has been refined, as well as changed into hard martensite, the cement-
ite network has remained unaffected. This last consequently
causes the original coarse structure of the case as a whole to be
retained (that is, unrefined), as well as the inherent brittleness due
to this free cementite.

Regeneration of the Core. — Now by quenching the steel from a
temperature just above the upper critical range of the steel of the

160 STEEL AND ITS HEAT TREATMENT

core, or at about 1650° F., we see from Fig. 89 that the core consists
entirely of homogeneous martensite, and further, that the former
coarse grain has been entirely obliterated. In other words, we have
" regenerated " the core.

Effect of Regenerative Quenching on the Case. — The effect of
this same regenerative quenching upon the high-carbon case is
shown in Fig. 90. From this photomicrograph it is evident
that, although we have effected a rearrangement and entangling
of the cementite (white) and thus largely reduced the weaken-
ing and embrittling effect of the free cementite (as in Fig, 86), the

FIG. 89.— Core of Steel Carburized at 1830° F., Slow Cooled, and Quenched from

1650° F. (Bullens.)

heating and quenching temperature of 1650° F. has not been suffi-
cient to dissolve entirely and " fix " the cementite in the martensite.
It is at once apparent that in this particular steel we have ex-
ceeded the proviso regarding the maximum carbon content which
we enunciated in a previous paragraph.

Treatment of High-carbon Case. — This leads to a consideration
of what method we shall apply when the maximum carbon con-
tent of the case exceeds that percentage which will cause the Acm
range to be above the Ac3 range of the steel of the core. So in this
particular instance we have two procedures open to us: we may either

CASE HARDENING: THERMAL TREATMENT 161

proceed with the quenching at 1650° F., obtain the best possible
refinement of the core, and accept with as good grace as we can the
presence of free cementite in the final case — granting that it is better
distributed by this quenching than by the treatment as in Fig. 88;
or, we may raise the temperature of the initial quenching to such
a temperature as will completely dissolve and fix the excess cementite,
even though it does increase the grain size (and, therefore, the brittle-
ness) of the core. In either procedure it will of course be necessary
to follow the initial quenching by the hardening quenching. The
proposition then comes to the point as to which is the more impor-

FIG. 90.— Case of Steel Carburized at 1830° F., Slow Cooled, and Quenched from

1650° F. (Bullens.)

tant, (1) the greatest refinement of the core, or (2) the most advan-
tageous treatment for the case, such as will give minimum brittleness,
minimum possibility for enfoliation to occur, and the best wearing
surface.

Effect of Very High Quenching Temperature on the Core. —
Assuming that the higher temperatures, which will be necessary if
the second item is to predominate, can be used without the oxidation
of the steel (such as by the use of salt baths) , excessive warping, and
so forth, a study of the photomicrograph of Fig. 91 will aid in solving
the problem. This figure represents the structure of the core after
quenching at 1830° F., followed by a second — the " hardening " —

162 STEEL AND ITS HEAT TREATMENT

quenching from 1375° F. It is at once manifest that the high quench-
ing heat has not greatly increased the grain size of the core. And
further, as the ferrite and sorbite have been distributed over the whole
section in fine particles, the core should prove very tough on this
account. That is, such treatment will generally prove satisfactory
for the core so long as the initial carbon content is not too high,
and we may proceed along the lines which shall produce the best
case.

Treatment for " Best Case." — What, now, is the best case —
in other words, the best wearing surface; and how may it be obtained?
From previous discussion, and from a study of the photomicro-

Fio. 91.— Core of Steel Carburized at 1830° F., Slow Cooled, and Double
Quenched from 1830° and 1375° F. (Bullens.)

graphs of this and the preceding chapter, it must be evident that the
best wearing surface, all things considered, is not characterized by
the presence of free cementite as a network or as spines. Granting
this, the way by which this condition may be avoided, assuming that
the carburized steel contains a hyper-eutectoid zone, is first to
eliminate the free cementite by quenching above the Acm range of
the case, and to be -followed by a treatment — for purposes of grain
refinement and hardening of the case — such as will not reproduce
the original network condition of the free cementite. Of necessity,
for reasons previously given, this second operation must consist of
a quenching at about 1375° F. for straight carbon steels.

CASE HARDENING: THERMAL TREATMENT 163

Effect of Double Quenching on the High-carbon Case. — The

effect of these two quenching operations is shown in the photo-
micrograph of Fig. 92. The steel contained about 1.40 per cent,
carbon and was quenched from about 1850° F., followed by another
quenching at 1375° F. In this instance it will be noted that the
free cementite appears as white dots or " spheroids " upon the darker
martensitic groundmass, and that there is not the slightest appear-
ance of the originally characteristic network structure of free
cementite. If the temperature of the initial quenching has not
been sufficiently high so as to dissolve all the original network of
free cementite, a structure will be obtained showing both sphe-
roidal and network cementite.

FIG. 92. — 1.40 per cent. Carbon Steel, Double Quenched from 1850° and

1375° F, X60.

Spheroidal Cementite. — The importance of this spheroidal-
cementite type of hyper-eutectoid structure as a wearing surface
cannot be over-emphasized. When such a steel is first placed in
service, the tendency will be for the martensite gradually to wear
away, leaving the extremely hard spheroids of free cementite to take
the wear. We then have the ideal conditions for a wearing or bear-
ing surface: the innumerable " points " of cementite, imbedded in
the softer and tougher martensite, act as the bearing-points; and this
wearing surface may be ideally lubricated through the circulation
of oil in the free zone representing the difference between the surface
of the cementite and that of the martensite.

164 STEEL AND ITS HEAT TREATMENT

Spheroidal Ferrite. — This same treatment will effect the " spher-
oidalizing " of the free ferrite of the hypo-eutectoid zone and of the
core in a like manner. In Fig. 93 there is shown the core of a case-
hardened steel of 0.22 per cent, carbon which had been double
quenched at 1850° and 1375° F. A treatment of this nature will
therefore put the whole steel, both case and core, in an analogous
condition, and the effect of any " liquation " of either the ferrite or
the cementite, caused either by slow cooling or— with cementite —
by oscillating temperatures of carburization, will, for the most part,
be overcome by such treatment.

FIG. 93. — Core Containing 0.22 per cent. Carbon, Double Quenched from 1850°
F. and 1375° F. (Bullens.)

Avoiding High Quenching Temperatures. — Referring again to the
effect of high quenching temperatures upon the refinement of the
core, it may be said that such temperatures will necessarily not bring
out the fullest elimination of brittleness in the core. For this, as
well as for other practical reasons involved in the obtaining of such
high temperatures, it is advisable to avoid their use. This may be
accomplished in two ways: by avoiding such carburizing methods
as will necessitate their use, as will be evident from the definition
which follows; or by a preliminary quenching directly subsequent
to carburization which we will discuss a little later.

CASE HARDENING: THERMAL TREATMENT 165

Maximum Efficiency in Case-hardened Steels. — Gathering
together some of the facts previously discussed, we will state that the
best wearing surface, in combination with minimum brittleness of case
(as shown by the absence of enfoliation) and of core (as shown by
shock tests), as well as with minimum difficulties of treatment, will
be had under the following conditions :

(1) When the maximum carbon concentration in the case is
greater than 0.9 per cent., but does not exceed that amount which will
cause the temperature of the Acm range of the case to exceed the
temperature of the A3 range of the steel of the core; and (2) when
the following conditions subsequent to case carburizing are rigorously
observed and their effect, with the exception of a, is at a maximum:
(a) Slow cooling from the temperature of carburization (which we
will shortly discuss); (6) quenching from a temperature slightly
above the A3 range of the initial steel; followed by (c), a quenching
from a temperature slightly above the Ac 1.2. 3 range of the steel.

Maximum Carbon Content. — Although the first statement
regarding the maximum carbon content which we have recom-
mended, that is, over the eutectoid ratio, is in direct contradiction
to the opinion of many metallurgists, it is nevertheless strongly
supported by industrial results as well as by theory. But it should
also be distinctly noted that the conditions of the specific heat
treatments under the second statement are strongly qualified, in
that the best technical methods — involving accuracy and uniformity
of heating and heat control — shall be instituted, and that the effect
of each quenching shall be at a maximum. If such conditions can-
not be complied with, it will be decidedly preferable to adopt such
methods of carburization as will produce a maximum carbon content
in the case of not much exceeding 0.9 per cent.

Maximum Effect for Cementite Solution. — For reasons which
will shortly be evident, it will be advisable further to amplify the
proviso that " the effect of each quenching shall be at a maximum."
First, then, in regard to the effect of the initial quenching upon the
solution of the cement ite. It is well known that the solution of the
free cementite in hyper-eutectoid steels takes place very slowly.
Due to this sluggish action, it will often be found that a heating
of short duration slightly above the Acm range will not entirely
dissolve the excess cementite. And further, that it is often necessary,
in order to avoid a prolonged heating at the apparent Acm temper-
ature, to increase this temperature to a considerable extent. Now
when the maximum carbon content of the case is such that the

166 STEEL AND ITS HEAT TREATMENT

theoretic Acm temperature is considerably below the Ac3 range of the
steel of the core, it is evident that there will be little difficulty in
satisfactorily obtaining the full solution of this cementite. But, on
the other hand, if the two temperatures named almost coincide, it is
manifest that the maximum effect of the initial heating and quench-
ing relative to the solution of the free cementite will not always be
obtained unless such heating is prolonged. To increase the duration
of this heating is also inadvisable, because this would tend towards
the diffusion or equalization of the carbon content in the various
external layers; and this, in turn, would be contrary to the purpose
for which the high carbon was originally obtained through carburiza-
tion. Again, quenching from a higher temperature than that origi-
nally set would obviously exceed the provisions previously named as
those necessary to obtain the best product, and for the present may
be eliminated.

Double Initial Quenching for Solution of Cementite. — It will
therefore be inadvisable to raise the maximum carbon content of
the case sufficiently high (through carburization) so as to bring about
the condition of affairs which we have just been discussing. If we
abide by the arbitrary rules which we have laid down, the only way
out of such difficulty, if it should exist, is to double quench from the
initial temperature.

Relation of Initial Carbon to Maximum Carbon. — Another
variable which should also be noted under this subject is the
influence of the carbon content of the steel of the core upon the Ac3
range. A study of the chart, Fig. 13, will show that between the
minimum carbon content used for case-hardening steels, or about
0.05 per cent., and the maximum carbon, or about 0.25 per cent.,
there is a difference of about 125° F. This will mean a corresponding
difference in the possible initial quenching temperature, and will, in
turn, influence the factor of the maximum carbon content in the case
which it is possible for us to use under these rules. This factor
must therefore be taken into account in the method of carburizing.
We may then say that the lower the carbon content of the steel
to be carburized, the greater may be the maximum carbon content
in the case — again assuming the previous conditions to hold.

Double Regenerative Quenching. — Let us now consider the
effect of the initial quenching temperature on the core. In the
chapter dealing with case carburizing it was stated that the higher
the temperature of carburization, and the greater the length of expo-
sure at that temperature, the greater would be the grain size and its

CASE HARDENING: THERMAL TREATMENT 167

influence upon subsequent regeneration. Under such conditions
it will not always be possible to obtain, by a single initial quenching,
the full refinement of the core. The only alternative, in order to
satisfy the set conditions, will be to double quench at the initial
temperature.

Slow Cooling after Carburization. — In a previous section we men-
tioned that the first step subsequent to carburization was to allow
the steel to cool slowly from the temperature of such carburization.
When solid cements are used, the method involving the immediate
removal of the cemented pieces from the carburizing boxes and
throwing them into the quenching bath cannot be too strongly con-
demned, especially if there is to be no regenerative quenching.
In the first place, it is a practical impossibility to remove all the
pieces from the box and to so quench them that the results will be
identical. This statement and its logical conclusions hardly need
further explanation.

In the second place, in order to obtain a full refinement of the
steel, it is absolutely necessary that the material shall be reheated
from a temperature below the lowest critical range to a temperature
beyond the upper critical range, for otherwise full regeneration will
not take place. If the objects have been immediately quenched
from a temperature near that of the carburization (i.e., without hav-
ing been previously slow-cooled), the grain size retained by this
quenching will be that characteristic of the highest temperature
reached during the carburization. The grain size thus given to the
core will be large, because the temperature of carburization must
obviously be high if quenching is to take place before the temperature
of the steel, during removal from the box, falls below that of the
hardening point. If the steel should be put into service in the con-
dition just mentioned, it would not be capable of withstanding any
great amount of shock on account of its inherent brittleness. And
even if the first haphazard quenching should be followed by a re-
heating and quenching from slightly above the lowest critical range
it is evident from previous discussion that the steel of the core as a
whole will not be regenerated.

In other words, if the carburization has given the proper maximum
carbon content in the case, previously stated, such a quenching will
be of little economic importance because it must always be followed
by the double quenching (regenerative and hardening) necessary to
produce maximum efficiency. Under such conditions, and for both
theoretic and practical reasons, it is advisable to permit the car-

168 STEEL AND ITS HEAT TREATMENT

burized steel to cool in the boxes to a temperature at least lower than
that of the Arl range.

Benefits from Preliminary Quenching. — Leaving aside the
consideration of those steels which require only a surface hardness,
there are only two benefits which can accrue from quenching directly
after carburization. First, there is the prevention of the " liquation "
of the excess cementite during slow cooling, with the possible resulting
disadvantages through enfoliation, or similarly, the liquation of the
ferrite. The author believes that the effect of this phenomenon of
liquation, although strongly emphasized by Giolitti, may be largely
counteracted by the results of the effective double quenching and its
consequent " spheroidalizing " action. The use of the preliminary
quenching, assuming the proper maximum carbon, may be regarded
as of indirect benefit in this first proposition.

Second, and of particular and direct importance, is when the
maximum carbon content of the case exceeds that amount at which
the temperature of the Acm range is equal to, or greater than, the
temperature of the Ac3 range of the steel of the core. Under these
conditions the preliminary quenching — as we call it — will prevent
the precipitation and coagulation of the excess cementite into the
network and spines which are so difficult to redissolve during regen-
erative heating. Consequently, this preliminary quenching will
permit the direct use of the regenerative quenching at its proper
temperature, even though the carbon content of the case is higher
than the governing ratio between Acm and Ac3 and which, under
conditions of slow cooling, would demand the use of a higher regenera-
tive quenching. It is manifest, however, that such preliminary
quenching, to be effective, must take place at a temperature higher
than the specific Acm temperature, or at about that of the cementa-
tion proper.

Use of Salt-bath Heating. — Before summing up the treatments
given in the foregoing pages, there are three points of practical inter-
est which should be noted. The first of these has to do with the
method of heating the steel for quenching. It is obvious that oxida-
tion, even of very slight amount, must be entirely prevented. The
best and surest method of attaining this is by the use of molten baths.
Of these, the salt baths are to be preferred to the use of lead, at least
for temperatures over 1500° F., on account of the poisonous fumes of
the latter at the high temperatures.

Interrupted Regenerative Quenching. — The second item refers
to the regenerative quenching. On account of the tendency of the

CASE HARDENING: THERMAL TREATMENT 169

high-carbon steels to check or crack when high quenching tempera-
tures are used, it is advisable to remove the steel from the water
bath when its red color is seen to disappear. As the steel " loses its
color " at a temperature under that of the lowest critical range-
that is, below that temperature at which the transformation in cool-
ing is totally effected, it is evident that this interrupted cooling will
in no wise affect the regeneration of the core. Its influence upon the
structure of the case will also have little practical importance,
primarily because it is not desired through this quenching to obtain
a maximum hardness; and further, because there will be little or no
tendency for any excess cementite to precipitate as a network struc-
ture. If any of the excess cementite should be thrown out of solu-
tion, it is more apt to be of the spheroidal type. Whether or not this
cooling is interrupted at about 900° F., it is always advisable to
remove the steel from the bath before it has become entirely cold.

Coagulation of Cementite. — In the third place, we would refer
briefly to the " hardening " or second quenching. If the case con-
tains greater than the eutectoid ratio of carbon, the duration of the
heating at this lower temperature should not be prolonged over a
greater period than is necessary thoroughly and uniformly to heat
the case to the proper hardening temperature. A prolonged heating
would have the tendency to coagulate the cementite which is ordi-
narily precipitated at this temperature, thus opposing the realiza-
tion of the conditions of maximum effectiveness.

Summary. — We may sum up the general situation, and give to
each class of steel the treatment which we recommend to obtain the
" best wearing surface," combined with minimum brittleness of case
and core.

Classification of Case-carburized Steels

Group A. Steels case carburized at temperatures approximating

that of the upper critical range of the initial steel.
Group B: Steels case carburized at temperatures considerably exceed-
ing that of the upper critical range of the initial steel.
Class 1. Maximum carbon content of the case does not exceed

0.9 per cent.

Class 2. Maximum carbon content of the case greater than
0.9 per cent., but is less than when Acm of the case
equals Ac3 of the core.

Class 3. Maximum carbon content of the case greater than
that specified under (2).

170 STEEL AND ITS HEAT TREATMENT

Classification of Treatments for Specific Steels

Group A. Class 1. Treatment I.

2. II.

3. III. or IV.

Group B. Class 1. Treatment II.

2. II.

3. III. or IV.
Treatment I.

a. Cool slowly.

b. Quench from slightly over the Acl.2.3, or about 1375° F.

Treatment II.

a. Cool slowly.

b. Quench from slightly over Ac3 of the core. Dependent upon

the carbon content, this will vary from 1650° to about
1525° F.

c. Quench from slightly over Acl.2.3, or about 1375° F.

Treatment III.

a. Quench directly subsequent to carburization, without slow

cooling, from at or near the temperature of carburization
but not lower than Acm. Dependent upon the carbon
content of the case, Acm will vary from about 1650° F. for
1.20 per cent, carbon (or thereabouts), to about 1800° for
1.45 per cent, carbon. It is not advisable to quench at a
temperature higher than 1800° F.

b. Treatment as in II, 6 and c.

Treatment IV.
a. Cool slowly.

6. Quench from a temperature over the Acm, dependent upon the
carbon content of the case. (See III, a.)

c. Quench from slightly over Acl.2.3, or about 1375° F.

NOTE : This treatment requires a slight sacrifice in the minimum
brittleness of core in order to obtain " best wearing surface."

Mechanical Effects of Treatments. — The effect upon the mechan-
ical properties of the case and core of various treatments is given in
the following table taken from Guillet. The steel used was of the
ordinary type for case hardening, classed as " extra soft."

CASE HARDENING: THERMAL TREATMENT

171

Treatment.

Resistance of
the Core to
Shock in kg.m.

Surface Hard-
ness of the ,
Case, Shore
Method.

Non-cemented steel, heated at 1700° F
in air

. and cooled

20.6

Non-cemented steel, quenched at 1700C
Steel cemented at 1830° F. for 0.047 in
slowly . .

F. in water
. and cooled

23.8
13.5

38.5

Same cementation; quenched at 1830°
Same cementation; quenched twice i
1830° and 1375° F

F. in water .
n water, at

23.2
25.5

79.8
84.0

Further Treatments not giving Maximum Efficiency. — Case-
hardened objects having a comparatively thin cemented zone
(yg- in. or less) may broadly be divided into those articles which
require only surface hardness and work under fairly uniform pressure
without shock, and those articles which must withstand shock,
bending strains, etc. We have discussed at some length both the
carburization and the heat treatment which are required by those
of the latter class. The heat treatment of those articles of the first
class we have previously referred to, but for purposes of summarizing
we may divide it as follows;

Treatment V.

a. Quench directly after carburization (without slow cooling),
but at a temperature not less than 1350° F., or that of Arl.
The results may be varied over a wide range according to the
temperature of quenching.

NOTE: This treatment is for those articles which merely demand
a hard surface, and in which brittleness and enfoliation may not be
considered.

Treatment VI.

a. Cool slowly.

b. Quench from slightly over Ac3 of the initial steel, varying from

1650° to 1525° according to the carbon.

NOTE: This treatment is for those articles which demand a
tough core and a comparatively hard surface — that is, the elimination
of brittleness in the core is of more importance than maximum
surface hardness.

172 STEEL AND ITS HEAT TREATMENT

Treatment VII. (Similar to Treatment I.)

a. Cool slowly.

b. Quench from about 1375° F., or slightly over Acl.

NOTE : This treatment is for articles which demand a maximum
surface hardness, or as much as can be obtained from a single quench-
ing, without reference to the brittleness of core or to the dangers
of enfoliation through the presence of free cementite. With low
temperatures of carburization and with a carbon maximum of 0.9
per cent, this classification would of course correspond to Group A,
Class 1.

Alloy Steels. — The treatment of alloy steels will be considered
under their respective chapters. In the main, however, the theory
of treatment does not vary, although the actual temperatures may
be changed on account of the influence of certain alloys upon the
position of the critical ranges.

CHAPTER VIII
HEAT GENERATION

Distinctive Conditions. — In industrial heating, and particularly
the sequence of operations applied to the heat treatment of steel,
it is but hard common sense to state that there is no general solution
applicable to the heating element or furnace. The application of
heat to these various operations, with the accompanying design of
furnace equipment, is an engineering problem, and it must be con-
sidered as such, and in the broadest manner, if the greatest efficiency
is to be obtained. No single type of furnace, fuel nor " system "
of burning can be applied as a " cure-all." Each case must be dealt
with on its merits and the furnace and the fuel and their application
to the work in hand must, in the final analysis, be based upon the
results obtained, measured commercially. Consequently, as no two
problems are exactly alike, it necessarily follows that the furnace
equipment must be designed to suit the individual plant with its
distinctive conditions. The average heat-treatment shop needs a
drastic awakening from the lethargy of " cut-and-dried " systems,
poorly designed and " home-made " furnaces, inefficient treatment
and handling of products.

Quality of Product vs. Cost. — Quality of product and cost of
manufacture are the basis of heat-treating operations. Quality of
product covers the proper heating of the material to meet the met-
allurgical requirements, and its physical condition to meet the
mechanical requirements. Cost of manufacture includes the cost
of fuel, power, labor, special equipment and material, such as boxes,
tools, quenching fluids, etc., as well as fixed charges on the equip-
ment, floor space, etc. Many of the mistakes that have been made in
heat-treatment installations are due to the fact that the problem has
been considered from the standpoint of fuel alone or of the first cost
of installation. Such a view is short-sighted, for the cost of fuel
alone makes up but a comparatively small part of the total produc-
tion cost. But when these items are considered in their proper place
with the other items of operating cost and with the proper inter-

173

174 STEEL AND ITS HEAT TREATMENT

pretation of the relation of these to the cost of the finished product,
it will generally be found that the cost of fuel and the cost of installa-
tion become of secondary importance in measuring or setting a stand-
ard of excellence to which the product must conform. In other
words, the ultimate aim of any heat-treating process, from the
economic standpoint, is to obtain the best heating of the product at
the least total cost.

The Standard Heating Unit. — There is no definite standard
employed for the measurement of production cost in industrial heat-
ing, as with power or light, because the conditions are continually
varying and there is no one definite point or method to determine
the cost. In power the test is the cost per brake horse-power hour
at the shaft of the machine, irrespective of the purpose of the appara-
tus. In light the test is the cost per candle power hour, irrespective
of the fuel employed or the means of utilizing or applying it. In
electricity it is the kilowatt hour measured at some definite point.
The nearest approach we can make to a standard for the commercial
determination in industrial heating is to suggest — " the cost per unit
of quantity of given quality." This is somewhat indefinite and dif-
ficult of location, owing to the many different standards for quality
and the great latitude in furnace design which affects the elements
entering into the cost of production above noted.

Such a standard means the abandonment of the technique of
combustion and other thermal considerations that are usually fol-
lowed. It means that the cost of finished product is paramount,
regardless of fuel cost. And it makes a point of considering first
of all the application of heat to the stock, and then an efficient
method of handling that material. Working along these lines has
produced real results in heating efficiency (if such a term is per-
missible) with oil fuel; the gas fraternity are beginning to recognize
the basic truth of the statement that fuel cost does not determine
heating cost; and the electrical men are slowly falling in line. Each
fuel has limits within which it can be used, and these are determined
by the nature of operations regardless of fuel cost.

Heating. — Any talk upon industrial heating must necessarily
take into account the right fuel and its proper application, a suitable
furnace design and construction, and a proper layout and efficient
handling of materials. Although each of these propositions is, in
a sense, distinct, it is obvious that each involves and must be
co-ordinated with the others. Similarly, the broad subject of
heating must deal with:

HEAT GENERATION 175

(1) The generation of the heat — the fuel;

(2) A system for applying the heat — the furnace;

(3) The utilization of the heat — the uniform heating of the
stock;

(4) The conservation of the heat — guarding against losses.
Comparative Fuel Costs. — The comparison of initial fuel costs

is always an interesting subject, but unless it is carefully amplified
and taken only in " small doses," it is apt to prove the truth of the
old saying that " a little learning is a dangerous thing." All of the
factors given in the previous paragraph must be considered in the
selection of any fuel, for, after all is said and done, it is the cost
of heating as shown by the finished product, and not the B.T.U. cost
of fuel, which counts the most.

The chart given on page 176 illustrates graphically the relation-
ship between commercial fuels, based on their heat unit cost.

This chart facilitates the determination of the cost per million
B.T.U. 's of commercial fuels at different prices; the relative prices
for different fuels at a definite price for one fuel or per million B.T.U. 's
and so forth.

To illustrate — the cost per million B.T.U.'s would be: at 3 cents
per gallon for fuel oil— 21.5 cents; at 20 cents per M for 1000
B.T.U. natural gas— 20 cents; and at $5.00 per ton for 12,000 B.T.U.
coal — 20.8 cents.

Again, at $5.00 per ton for 12,000 B.T.U. coal, the relative prices
for the other fuels, keeping the same B.T.U. cost, would be — $5.80
per ton for 14,000 B.T.U. coal; 21 cents per M for 1000 B.T.U.
natural gas; 12.6 cents per M for city gas; 2.94 cents per gal.
for oil; 2.6 cents per M for 125 B.T.U. producer gas, etc.

Further — at an assumed cost of 30 cents per million B.T.U.'s,
the relative prices for various fuels would be — 4.2 cents per gal. for
oil; 5 cents per M for 165 B.T.U. producer gas; 9 cents per M for
water gas; 18 cents per M for city gas; $6.00 per ton for 10,000
B.T.U. coal, etc.

The B.T.U. Value. — The fanacy of selecting a fuel merely by
its B.T.U. value alone, however, should be self-evident. To illus-
trate, take the case of anthracite coal: when broken into pieces the
size of a man's fist it is used in the ordinary hot-air " heater " or
house furnace; when crushed to a smaller size it is used in the kitchen
stove; crushed to rice-size it may be used for forced-draft boilers;
pulverize it to a dust and it may be used in the " powdered coal "
systems. But would the furnace size be suitable for the kitchen

176

STEEL AND ITS HEAT TREATMENT

S 8 8 S
^sqiooog

'HO jo 'I^O J9d Jo SB o jo "jj -no 0001 aed sixtiao ut aou^

HEAT GENERATION 177

stove, or would stove coal satisfy the conditions necessary for the
latter systems? Assuredly not; and yet the B.T.U. value of the
coal remains absolutely unchanged. In other words, it is not the
cost of the B.T.U. 's in coal, but all in the manner of its us.e.

The Combustible Mixture. — Again, the so-called B.T.U. com-
parison of fuels, which has long been generally accepted as standard,
is misleading. In fact, it is not the B.T.U. value of the fuel, but
the B.T.U. value of the combustible mixture that counts. There is
not as much difference in the heating value of the combustible mix-
tures of the various gases as there is between the heating value of
the gases themselves without considering the combustible mixture.
For example, if city gas were assumed to be 600 B.T.U. and producer
gas 120 B.T.U., the heating value of the gases would generally be
considered in the proportion of five to one; and many comparisons
are made upon this basis. In practice, however, whether it be in
an internal combustion engine or in a furnace, the actual values of
the fuels are not anywhere near the ratio of five to one; if they were,
then an engine of given size cylinder which would develop 100 H.P.
with city gas would only give 20 H.P. with producer gas. This
comparison most people would agree is ridiculous and unreasonable
on its face, in the light of practice in power; and yet the same people
do not hesitate to draw such a comparison when considering heat,
although the determinative conditions are just as true in one case as
in another.

The Right Fuel. — The generation of heat for heat-treatment
purposes involves the proper application of the right fuel. Super-
ficially the choice of fuel appears to be a simple one. Too many
persons, however, are prone to select off-hand some fuel such as
coal because the initial cost is low, or maybe natural gas if it is near
at hand, or even electricity just because it sounds attractive and is
easily controlled. Although each of these in its proper sphere would
be the logical source of heat, as general proposition no one fuel is
appropriate to every case. Neither low initial cost, nor local supply,
nor ready control sums up the situation. These are but factors in the
case as a whole, and the value of one fuel cannot be measured by
its use in the way another fuel would be used. There is always a
right fuel for the particular work in hand, so that each problem
must be thoroughly studied and understood if the best solution is
to be had.

Fuel Efficiency. — There is, or at least should be, no argument
on the fuel question. The relationship of the various fuels is fixed

178 STEEL AND ITS HEAT TREATMENT

by physical law and commercial conditions. The term " fuel
efficiency " is not used in power nor in domestic house-heating work;
and neither does it exist in industrial heating. Generally speaking,
it may be assumed that any difference in results between oil, or gas,
or electricity for heating furnaces should be attributed to the manner
in which the heat is applied to the stock and not to any inherent
advantage in one form of energy over the other. It is not proper to
say that oil is cheaper than electricity or coal or gas, or vice versa,
or that a higher-price fuel can displace another because the former
is " more efficient." There is no such thing as that one will do more
than another. It is the nature of the operation and the manner of
applying heat that counts, and not the fuel. Thus, the manner of
its use may be efficient or the means of employing or utilizing it,
but certainly not the fuel itself. It can be said that electricity in a
given type of furnace will produce a better result than oil in a given
type of furnace, but it will be noticed in so doing that it is the dif-
ference in the manner of applying the heat, which is equivalent to a
difference in furnace design, that determines the result, and not to
any advantage in one form of energy over another. The term " fuel
efficiency/' therefore, is misleading in that it is employed to express
a condition that does not exist.

Fuel vs. Operations. — It may be said that the extent of the
heating operation more or less determines the fuel. It would be good
practice to employ city gas for the annealing of a small quantity of
wire in a loft building ; and with a larger quantity and other working
conditions the fuel might be oil; but if the operation be conducted
on a still larger scale, as in a rolling mill where the size and type
of furnace will so peimit, then the fuel might be producer gas or coal.
Thus, city gas at $1.00 a thousand for many operations might be
recommended in preference to oil at 1 cent a gallon, and electricity
at any price in preference to oil or gas at any price. In annealing,
for instance, we might use oil for certain results and in same room use
coal for annealing the same metal, but for different results. Thus
the nature of the operation more or less determines the working
limits, regardless of fuel cost.

The " Fluid Fuel." — We can even go a step further and state
that no one fuel has a monopoly on uniform heating, or control, or
economy. At the present time there are many furnaces of good
design, operating on cheap coal, which produce a better quality of
product at less cost and maintain more uniform temperatures,
with more accurate control, than other furnaces of poor design built

HEAT GENERATION 179

for the same purpose, using either oil or clean- washed producer gas.
All other things being equal, a " fluid fuel " (as distinguished from
a solid fuel) is generally to be preferred, as it lends itself more readily
to accurate control. Ordinarily it would be assumed that a fluid
fuel would permit of greater flexibility of operation than a solid fuel,
and it invariably will when all other conditions are equal. However,
as previously noted, there are many cases where the advantages of
the more flexible fuel are lost with inefficient means of utilizing it.
But it is misleading to couple any fuel with the term Uniform
Heating, or Control, or Economy, without specifying or qualifying
the manner in which it is to be used. It is the manner of applying
the heat and not the fuel which thus determines the success or failure
of the operation.

Selection of Fluid Fuel. — Even the selection of a fluid fuel is
dependent upon conditions other than heating value or composition
of the fuel itself. For instance, we might consider a very small
furnace for annealing, tempering or hardening small pieces of stock.
There would seem to be no question but that gas would be the fuel
most generally preferred. But even the selection of the gas itself
would be dependent upon conditions other than that of temperature
control. To illustrate: while one might use either producer gas,
water gas, city gas or natural gas in the case just mentioned, it
would not follow that each of the fuels could be considered if the
operation was one requiring a very high temperature, such as welding.
The same conditions would hold good if the operation was one requir-
ing a very low temperature, such as japanning. If, however, the
operation were to be conducted on a large scale, and one involving
the use of a large furnace in which there would be ample room for
combustion, there would be brought into competition with these
gases a liquid fuel in the form of oil, which, by reason of the latitude
afforded in furnace design, could compete from the standpoint of
temperature control; and, depending upon the market, it might com-
pete in the matter of price.

Influence of Working Conditions. — There is generally a confusion
between the terms Uniform Heat Distribution and Uniform Fuel
Distribution; the two do not necessarily go together, although at
times they do. For instance, in the case of japanning, requiring a
low temperature, particularly in small and medium-size ovens, a fluid
fuel in the form of gas is generally preferred. The reason for this
is that the fuel may be applied on all sides of the oven and burned in
small quantities at different points. This could not be as readily

180 STEEL AND ITS HEAT TREATMENT

done with a liquid fuel like oil, which by reason of its great calorific
power and lack of control when burned in very small quantities in a
very limited space, would prohibit a distribution of the fuel in the
manner usually provided for gas. If the even were large and there
was plenty of room for combustion and distribution of heat through
flues, it would be possible to approach the same conditions of uniform
heating, as far as the stock is concerned, without uniform distribution
of fuel. In one case the result of heating the material is accomplished
by distributing the points of heat generation, while in the other it is
produced by localizing the point of heat generation and distributing
the heat after it is generated. This goes to show why working con-
ditions are, as a rule, determinative not only of the fuel, but of the
manner in which it may be employed. Also, to show that the con-
clusions formed in one case may be reversed in another, when a change
in working conditions makes it either possible or desirable, even
though there be no change in the composition or price of the fuels
themselves.

Regeneration. — The question of regeneration or recuperation
is usually misunderstood. Recuperation is generally commercially
desirable on general principles, but there are times when it is phy-
sically necessary, irrespective of the economy of the operation. For
instance, neither hot nor cold producer gas is suitable for high forging
or welding heats in furnaces without regeneration or recuperation,
because the heating value of the fuel is so low. Because of their
greater heat value, water gas, natural gas, city gas or fuel oil might
be used for operations with which producer gas could not be con-
sidered, for the reasons given above. But if the furnaces were large
and the principles of recuperation could be employed, then the
producer gas could compete, and the extent to which it could com-
pete would be determined by fuel cost coupled with the manner in
which the fuels were employed. Very often a comparison is made
with producer gas used in this form against the other fuels to show
that the gas would be the cheaper. But this in itself is not conclusive
unless the determination is based upon the results secured by the
use of the other fuels in a manner which would involve recuperation,
and in this way take advantage of every possible saving. Producer
gas, as a rule, would show the highest efficiency from the standpoint
of recuperation, because the volume of the inert or non-combustible
gases is greater; but it does not follow that this efficiency of recupera-
tion is in itself determinative of the fuel.

Cost of Delivering Fluid Fuels. — Even though fuels were com-

HEAT GENERATION 181

pared in the manner above indicated, it would not follow that the
result would be as conclusive from the standpoint of industrial heating
engineering as it would be from that of fuel. The reason for this is
that some fuels by their very nature are better adapted to the
manipulation of conditions governing a manufacturing process
than others. For instance, fuel oil, or any cold-washed gas that
could be delivered to a furnace through pipes, and thus distributed
to scattered units throughout a plant, could be more easily handled
than producer gas delivered in a hot state through a flue.

Cost Factors with Producer Gas.— The conditions that hold good
with an open-hearth furnace do not obtain in a small forging or heat-
treating installation made up of scattered units. There is no question
but that the B.T.U. cost at the producer is less with hot gas than with
cold washed gas, because of the heat loss in the latter in the processes
of scrubbing, drying and cleaning. On this comparison of B.T.U.
cost at the producer much has been written and many claims have
been made for superiority one way and another. It must be borne
in mind, however, that in industrial heating it is not the cost at the
producer or at the point of fuel supply that counts, but rather the cost
of the finished product at the delivery end of the furnace. This
heating cost, as it might be termed, takes into account not only the
fuel at the furnace, but the utilization of the fuel in the furnace; and
on top of this, the labor and any other charges, such as interest and
depreciation, that may be legitimately charged against the equip-
ment itself. The fuel, at best, is secondary and not of the para-
mount importance which many have erroneously tried to make it.

Fuel Equipment. — In the cases of such artificial gases as pro-
ducer gas, water gas, etc., the cost of the necessary plant equipment
for their manufacture, besides interest and depreciation, must also
be figured in with the cost of the gas. It is evident that the heat-
treatment plant must be of considerable size to warrant a large
initial plant investment for the manufacture of such gas. Similarly,
if oil is to be used, there must also be an allowance for storage tanks
for protection against poor deliveries and unusual service demands.
The cost of piping for fuel distribution is lower in the case of oil
than with gas, as smaller pipes are used.

Fuel Supply.- — The selection of a fuel involves not only its heat-
unit value, but a consideration of local conditions of constancy of
supply and of sufficient quantity without undue fluctuation in price.
The source of fuel supply must be absolutely dependable, both in
regularity of deliveries and uniformity of fuel. This item is largely

182 STEEL AND ITS HEAT TREATMENT

connected with the use of oil and is a phase of the fuel question well
worth studying. Thus, in considering the availability of any fuel,
the locality, natural resources and freight should be taken into
account.

Uniformity of Fuel. — High quality of product requires uniformity
of heating, and which in turn requires uniformity of fuel. Such a
product may be obtained only by having an absolute control of the
volume, temperature and composition of the gases to and from all
points of the chamber. Any changes taking place in the composition
of the fuel not only makes the work of the furnace man more difficult
and exacting, but also must inevitably result in intermittent heating
and unsteadiness of operation. That is, variable supply or non-
uniform fuel is not conducive to good product. In this connection
it is advisable to note that the presence of sulphur — as in producer
gas or coal — is more or less injurious to the metal when at high
temperatures, as it has been shown that hot steel is capable of being
sulphurized as well as carburized.

Fuel Oil. — There is probably no one fuel that has been more
abused than fuel oil. Its great concentrated heat value and flexibility
of application and control have been, commercially, its greatest draw-
back, for the reason that these advantages have permitted the applica-
tion of the fuel in a haphazard manner by people who seem to be
satisfied so long as it was burned and made heat in some form or
other. Even many of the manufacturers of oil-burning equipment
are not entirely free from this criticism. Much of the competition
which has been held against fuel oil, and comparative statements of
operating costs that have been made up, are not based on an efficient
use of the fuel at all.

Air Control with Fuel Oil. — The majority of heating equipment
installed with oil are lacking in some of the very elementary essen-
tials necessary for good combustion, and in this respect are not
anywhere near as efficient as an ordinary kitchen stove. There seems
to be an absolute disregard of the fact that it is just as necessary
to control all of the air entering into a furnace with oil fuel as it is
with gas or coal. Most city gas equipment has this provision, as
have boilers or ordinary stoves, but it seems to be entirely lacking
with most oil-burning equipment, although the reasons for it are just
as important in one case as in another. The common practice is
to inject oil in a hole through the side of a furnace; and many people
think that because there is a valve on the oil and air lines that they
consequently control the air. This, however, is not true, for the

HEAT GENERATION 183

reason that in many furnaces by far the greater proportion of air
required for the combustion of the fuel is " induced " by the force of
the blast and does not pass through the burner itself. It has been
by reason of such conditions that oil has been abandoned and given
a bad name in many places where the conditions would be reversed
if it were properly handled.

The Human Element. — Thus the average man, when he sees a
kerosene lamp smoke and blacken the chimney, or a gas mantle puff
and impair the light, will ordinarily recognize that something is
wrong, and immediately make the adjustments necessary to over-
come the difficulty. Both of these are the effects of a common cause
— the improper mixture of the fuel and air necessary for proper
combustion; and in making such adjustments, whether he knows
it or not, he is merely establishing the proper relationship between
the fuel and air necessary for good combustion.

Yet we will find this selfsame man, day after day and year after
year, operate or permit others to operate, at great expense, furnaces
with oil or gas which smoke and puff and pollute the atmosphere
with hot and obnoxious gases, but never think of making the ad-
justments necessary, which are the same as those required in case of
the lamp. Such a man is either not " on the job " or his furnace is
lacking in the essentials for good combustion common to every house-
hold lamp or stove, whether it burn oil or gas or coal. Yet the
majority of men employed in heat-treating work, as well as a large
percentage of furnaces, particularly those fired with oil or coal, are
open to this criticism, which is evidence of the necessity for improve-
ment in the personal element, at least to the extent of either making
the adjustments if provision for such is on the furnaces, or at least to
insist that the furnaces be designed on the A.B.C. principles of heat
generation. And when such adjustments are made with proper
furnaces, the operator benefits himself by decreasing the heat and
gases affecting his health and comfort, and benefits his employer in
turning out better product and saving fuel and power and conserving
the life of his furnace. Such conditions, which actually exist in the
majority of shops in the country, are a sad commentary on the work
of efficiency, safety first and industrial betterment ideas so prevalent
at this time, and sustain the point that we often look at and do not see
opportunities for improvement that can be made in a simple way.
The average furnace operator appears to act on the principle that he
is not making a good showing unless he has plenty of smoke and flame
belching out of every crevice of the furnace — probably for the same

184 STEEL AND ITS HEAT TREATMENT

reason, or lack of reason, responsible for the blacksmith striking two
blows on the anvil to one on the horseshoe.

It would appear only reasonable to assume that the existence of
such conditions in the shop does not permit the owner of such shop
to make the statement that one of his most important manufacturing
operations — i.e., the heat treatment of good steel — is conducted under
the best possible methods with the best possible furnace equipment
by the highest grade men, or that it is even on a par with the average
machine-shop practice. And all this notwithstanding the evidence
that may be offered in the way of " Temperature Records " or
" Fuel-burning Equipment " in an attempt to sustain the point.

It is invariably cheaper to do it right. The extra expense in solved
in furnaces of heavy construction with proper provision for applying
heat to the chamber and for removing it from the chamber, is insig-
nificant when compared with the saving effected.

The Value of the Furnace Operator. — If, as generally conceded,
men are paid in proportion to their skill and the part that such skill
plays in the make-up of a finished product, then a good annealer is
worth more than a roller in a rolling mill, and a good man in charge
of heat-treatment work is worth more than an automatic machine
operator in an automobile or machine shop. In one case the man
operates a machine and it is a machine that more or less determines
the result, and at any rate they are a mechanical check on the opera-
tion. In the other case it is purely a question of skill, experience,
and judgment, with no mechanical check upon the major part of the
operation. The furnace and all auxiliary appliances are but tools,
and while it is necessary that they should be of the best, they are
nevertheless but tools in effect and bear about the same relation to
the result that a good tool in the hands of a good operator bears to
any other result. Furthermore, it is the judgment of the furnace
operator that determines if the work of all that have preceded him
shall be spoiled or improved upon, and if the time, labor, and money
spent to produce the results sought for are capitalized or wasted.

Purchasing Brains. — It has been stated that the average manu-
facturer will unhesitatingly invest money in anything involved in
his processes of manufacture outside of the human element, which is
virtually paying a premium on everything but brains. It is common
practice to install an expensive machine, costing thousands of dollars,
and employ a cheap, inefficient man to run it, notwithstanding that
it is the man who controls the output and cost of operating that
machine. When the relationship of the human element to the result

HEAT GENERATION 185

is of less importance, as, for instance, with a machine press — as it
surely is when compared with a furnace — then there will be mani-
fested the false economy effected by the employment of unskilled,
inefficient operators. In the final analysis the machine or the fur-
nace is nothing more than a tool in the hands of the operator, and
the value of the human element depends upon the amount of skill
that must be exercised in the use of a tool, whether it be a hammer,
a chisel, a press or a furnace. There are many cases when it is per-
missible to employ unskilled labor in connection with a machine,
where the operation is more or less automatic and the operator is
required to do nothing more than start or stop the movement and
feed material. Such a practice, however, is foolhardy with a furnace,
because of the paramount importance of the human element in the
operation. It is a waste of money to install efficient types of fur-
naces, which are necessarily expensive, without intelligent super-
vision over the operation in the form of at least one efficient man who
can either operate it himself or direct its operation by others. The
practice of employing at least one skilled man for such purpose is
gaining headway and will undoubtedly continue to do so, as his labor
is usually more than paid for by the savings effected in the cost of
operation, to say nothing of the betterment of the product. A good
furnace coupled with a poor operator does not make the proper
combination, and when both are of inferior caliber, as they so often
are, then it is unreasonable to suppose that the all-important heating
operations are conducted as they should be, even though there may
be some, but nevertheless weak, evidence in the form of pyrometer
records to the contrary.

Effect of Operation. — The operation of the furnace in the shop
should be regarded in the same light as the stove in the kitchen.
The furnace operators should be taught that furnaces operate on the
same principles as an ordinary house-heating coal stove; both must
be given the proper attention in the matter of regulating the dampers
for the air supply and flue gases in order to accomplish the same
results. The importance of regulating the amount of air which is
used for the combustion of the fuel cannot be too strongly emphasized.
In this respect the ordinary kitchen stove is in many ways superior
to many so-called heart-treatment furnaces. Just study the cook-
stove in your own home and see how many different ways there are
for adjusting the air supply and draft. To illustrate: under the fire
there are usually one or possibly two dampers ; directly over the fire
there is another damper for checking the fire; another damper will

186 STEEL AND ITS HEAT TREATMENT

permit of the carrying of the heat currents around the oven (conser-
vation) ; and still another damper or two will shut off the draft up
the chimney. As a general question, how many heat-treatment
furnaces have an equal number of devices for controlling the combus-
tion or heating? Apart from the question of recuperation and its
direct saving in fuel, there is involved the method of operation,
and question of furnace design which in itself is modified by the
method of operation. While in some cases the application of the
fuel is poor, in others the construction of the furnace may be at
fault or not adapted to the work it has to do.

Furnace Design. — Poorly designed furnaces are the cause of much
of the difficulty and time spent in trying to account for poor results
which could be better spent in insuring better conditions. Many
furnaces to-day are operated with elaborate pyrometer systems, but
with a complete disregard for the conditions that make good heat-
ing possible. There is no control of the air for combustion entering
the furnace, no control of the waste gases leaving the furnace, and no
knowledge of their composition or action while in the furnace. Under
such conditions good combustion, with soft heats and uniformly
heated stock, is out of the question. The object is not merely to
burn fuel, or to make heat, but to apply the heat economically and
effectively.

Oil Burners. — There is altogether too much importance attached
to oil burners, both by manufacturers of such appliances as well as
by users. Too many people have the idea that all that is necessary
to do with oil is to buy an " efficient " burner and build a furnace
around it. But there is really no such thing as an oil burner in the
sense usually taken for a gas burner. The very term is a misnomer,
as the oil burner is nothing more than a valve and its efficiency is
mechanical and not thermal. In fact, the majority of oil burners
are not, properly speaking, mixing valves, as most gas burners are,
for the reason that the most successful burners are nothing more
than valves which introduce fuel and air into the furnace in pro-
portions fixed by the operator, the actual mixing taking place in the
furnace and not in the burner itself. It does not matter as much
how the fuel is delivered at the furnace as what is done with it
after it is delivered in the furnace.

Burner vs. Furnace. — This influence which the design of the fur-
nace— that is, the method of using the fuel in the furnace — has upon
the economy of heating, regardless of the burner, is well brought
out by the following chart, to accompany Figs. 94, 95 and 96.

PROGRESS IN ROTARY FURNACE CONSTRUCTION. 187

Three different types doing the same work.

FIG. 94. — Externally Fired Cast-iron
Tumbling-barrel Furnace (1898).

FIG. 95. — Externally Fired Cast-iron Helical-cylinder Furnace (1906).

Pyrometer Connection

Steam ||| / ,__-, Discharge Hood

or Air-

Charging Drum

FIG. 96. — Internally Fired Tile-lined Helical-cylinder Furnace (1909).

188

STEEL AND ITS HEAT TREATMENT

The figures are taken from actual operation and weight of material
and fuel.

Design of Furnace.

Externally Fired
Cast-iron Tumbling-
barrel Furnace,
1898.

Externally Fired
Cast-iron Helical-
cylinder Furnace,
1906.

Internally Fired
Tile-lined Helical-
cylinder Furnace,
1909.

Metal heated
per hour

287 pounds

576 pounds

1450 pounds

Increase of metal
heated per hour

100 per cent.

405 per cent.

Fuel oil burned
per hour

6.41 U. S. gals.

3.64 U. S. gals.

3.48 U. S. gals.

Decrease of fuel
oil burned per hour

43.3 per cent.

45.7 per cent.

Metal heated per
gallon of fuel oil

45 pounds

158 pounds

416 pounds

In each case the pieces heated were of the same size and
material and the lightest in individual weight of their kind. The
figures represent three different types of rotary heating furnaces
doing the same work. The advantage in favor of the internally
fired helical furnace — Fig. 96 — was still more marked when heating
pieces of greater individual weight. The material was also of better
quality, being freer from oxidation. The cost of repairs is also very
much less for the internally fired helical furnace than either of the
others.

Now these three furnaces were operated with the same burner,
the same fuel oil, the same steam pressure for atomizing, the same
air for combustion, the same material, the same temperature, at the
same time, and by the same men. This then illustrates the point
that it is the furnace and not the burner alone which produces the
desired results.

" Quality of Heat." — Much has been said about " the quality
of heat"; but as heat can differ only in the degree of temperature,
such a statement must therefore refer to the atmosphere in the fur-
nace. One of the most recent books on the subject of heat treatment
of steel has the following statements :

" Until recently, the only known way of producing heat of the
required intensity was by combustion — the burning of some fuel.
The attendant disadvantages of this are well known. The crude

HEAT GENERATION 189

open coal forge is capable of heating the steel, but leaves much to be
desired as regards the quality of the heat, its uniformity, and the
temperature control. In order to produce heat at all, the carbon
in the coal must be combined with the oxygen of the air, and a
strongly oxidizing flame is unavoidable. The steel exposed to this
action, or to the inevitable results of it, suffers accordingly. The
coke-burning furnace offered some improvements, but only in detail.
Now there are highly perfected furnaces for burning oil and gas,
and some of these offer still further advances, but the principle at
the basis of all of these is the same — there must be a ' burning '
process to produce the heat; oxidation must be present with all fuel-
combustion furnaces.

" Through what means, then, may we obtain the proper quality
of heat, uniformly applied, and of the right degree? The electric
furnace for the heating of steel brings the answer. It overcomes
most of the objections to the ' combustion process ' by introducing
a new principle."

Further on the statements are made: "The atmosphere in
the heating chamber of the electric furnace is inherently l reducing '
in its nature, due to the fact that the hot carbon plates absorb all
of the atmospheric oxygen. By raising the door slightly, and open-
ing the draft-hole at the rear, a slight current of air may be admitted
which will counteract this tendency. Leaving the door open slightly
more would allow an excess of air to enter, so that an oxidizing
atmosphere could be produced. Between the extreme points fine
shades of atmospheric conditions can be obtained. Thus the qual-
ity of the heat can be absolutely and easily regulated."

Furnace Atmospheres from Combustion. — Commenting on the
above, the question of atmosphere in the heating chamber is one of
operation of the furnace, assuming the proper furnace design, and
simply comes down to the relation between the fuel and the
air supply. In the design and operation of the best types of
heating furnaces it is the aim to produce an atmosphere, under
pressure, which virtually contains no oxygen and only a very
slight amount of reducing elements. Take for example the
under-fired type of furnace, properly designed and operated.
The actual combustion of the fuel takes place in the separate
chamber under the hearth where the amount of air taken into
the furnace is just sufficient to produce perfect combustion of
the fuel. By the time the hot gases actually reach the heating
chamber they are thoroughly mixed and, on account of the design

190 STEEL AND ITS HEAT TREATMENT

of the furnace, a pressure is built up. This pressure stops any inflow
of free oxygen through the doors, and otherwise surrounds the steel
with the neutral atmosphere of hot gases. That these hot gases
may actually contain no oxygen and very little reducing vapors is
well shown by the following analysis :

Per Cent.

Carbon dioxide 12.5

Oxygen 0

Oil vapors 0

Carbon monoxide 2.1

Nitrogen 85.4

From a study of. this analysis, which was taken under ordinary
operating conditions of a well designed furnace and without the
operator knowing that anything unusual was expected of him, it cer-
tainly cannot be said that the " combustion process " produces,
under proper conditions, anything but a proper neutral atmos-
phere.

Furnace Atmospheres with Electricity. — In contrast with this,
consider the effect taking place in the electric furnace previously
referred to. In this case the heat is virtually supplied outside of,
and through, the walls of the chamber. Free access of unaltered
atmospheric air to the inside of the furnace exists, and coming in
contact with the heated steel results in a very rapid oxidation of the
metal. An attempt is sometimes made to remedy this condition
by introducing charcoal into the chamber with which the oxygen
is supposed to combine to form an inert atmosphere of carbon
dioxide, or even by exposing the carbon electrodes of the electric
element and allowing the oxygen present in the chamber to attack
them and thus consume the free oxygen. It must be realized that
if this elimination of the free oxygen thus occurs it must take place in
the furnace itself and in the chamber where the stock is being heated
and the steel is therefore more or less exposed to oxidation.

The argument previously advanced that the atmosphere in the
electric furnace may be controlled by slightly raising the door and
opening a vent in the back of the furnace is entirely contrary to heat-
ing principles. The cost of the heating itself advances when this
takes place because cold air is entering through the front and the
heat is allowed to flow out through the vent. Such a proposition
is analogous to the effort of trying to heat a house in the winter
with the doors open.

HEAT GENERATION 191

Electricity for Heating. — The advantages of electricity as a
source of heat compared with oil or vice versa are determined
only by the nature of the operation regardless of fuel cost. Each
form of energy has its own field of use. When it is considered that
fuel oil under proper application is continually being used within
temperature limits of 10° F., it must be evident that electricity must
offer some powerful indirect advantages for a great deal of heat
treatment work in order to be given a chance. On the other hand,
there are, however, many cases like the open-hearth, reverberatory
and other forms of heating where the limit has been about reached
with fuel. It is in such operations where electricity, by reason of a
better method of applying heat to the stock, can overcome the
disadvantage of higher fuel cost with less actual energy for the opera-
tion. Thus we might also designate certain forms of chain welding
as good examples where electricity might be advantageous.

Fuel vs. Product. — Reduced to lowest terms, heat generation is
an economic problem. Commercial heat treatment requires the
production of the best attainable results at the lowest cost. We
may sum up the fuel question by stating that in connection with
appropriate furnace design and furnace operation, that fuel should be
used which will cost the least viewed from the standpoint of finished
product. This cost not only applies to increased output, but also to
the quality of the product. The former results in lower manufactur-
ing costs; the latter in greater efficiency and higher selling prices.

CHAPTER IX
HEAT APPLICATION

Furnace Equipment in General.— The usual consideration of
furnace equipment for heat-treatment operations is based on require-
ments of " accurate temperature/7 " heat control," " temperature
variation," etc., with the result that the trade literature — which is
the catechism of many, if not of the majority interested in the work —
has developed a standard of heating which is altogether too low and
must be superseded by one based on a broader view of the problem
in order to effect greater progress.

The existence of such a standard is probably due to the reasoning
— " that a uniformly heated piece naturally involves a uniform
temperature within the furnace, and the less the variation in tem-
perature the better the results will be; therefore, in order to produce
a uniformly heated product it is necessary to employ a furnace that
will produce a uniform heat with minimum variation of tempera-
ture." The natural result of such reasoning has been a development
of pyrometers to indicate the variations in temperature, and the
development of many different designs of furnaces which have been
vigorously and extensively exploited on the strength of claims for
" accurate temperature control," " minimum temperature varia-
tion," " neutral or reducing flame as desired," " no oxidation," etc.

That phase of the work covering pyrometry has been of ines-
timable value and is more highly developed than nearly any other
branch of heat-treatment work. It is, however, unfortunately too
often considered as an end when in reality it is but a means to
an end — a means which if properly employed will indicate the simple
principles back of " heat application," as in comparison with " heat
generation " or " heat utilization."

The standard of heating requirements, with a means in the form
of pyrometry to check them, has developed much discussion of the
relative merits of the different designs of furnaces offered to meet
them. Much of this has been directed towards the claims made for
the furnaces supported with evidence in the form of pyrometer charts,

192

HEAT APPLICATION 193

heat logs, and other data incident to the indication of heat, tending
to show the ability of the various designs to meet the heating require-
ments above outlined. This is all right as far as it goes, but it does
not go far enough. It may be considered as " evidence " of a heat
condition in some part of a furnace chamber, but not necessarily
as " proof " of a uniformly heated product within that furnace
chamber. If it were otherwise, then we would have not to deal so
often with variations in the finished product without any apparent
variation in the indicated temperature. Heating a furnace chamber
uniformly, and uniformly heating a product within that furnace
chamber, are two distinct operations. The former must accompany
the latter, but the mere indication of the former does not by any
means prove the existence of the latter. It does not follow that the
temperature variation indicated in any two points in a chamber
when empty will be the same as that indicated when the chamber is
partially filled, or more particularly, when it is filled to full normal
capacity. The real test of a furnace for a given operation from the
standpoint of uniformly heated product, is not the temperature
variation when the chamber is empty or partially filled, but the
temperature variation around the mass to be heated when the chamber
is loaded to full capacity.

This is a simple fundamental rule which, when considered with
reference to a given operation, illustrates why it is possible to secure
better results in finished product with one type of furnace over
another without any apparent difference in the indicated variation
in temperature at any given point.

To illustrate: Let us consider two gas-fired bake-ovens of the
same size, each designed to accommodate six cakes of a given
size, the essential difference between the two being that one
is heated by gas jets at the bottom, as in Fig. 97 (a), and the
other by gas jets at the top, as in (6), the problem being the
uniform heating or baking of the six cakes to be placed on the
tray x.

If a thermo-couple should be introduced at any two points on the
tray x in the empty ovens, it will be found that the heat will be
fairly uniform with either design, even though with one design the
actual oven temperature may be different from the other. It is
reasonable to suppose that one small cake, as in (c), would compare
favorably with (d) , as there is ample room for circulation of the heat
from one side to the other. We have in one case (a or b) an indi-
cation of a uniform chamber temperature, and in the other (c-d)

194

STEEL AND ITS HEAT TREATMENT

an indication of a uniform chamber temperature and of a uniformly
heated product.

Now suppose that each oven is filled to normal capacity, as in
(e) or (/). If the space between the pieces is small, it is very likely
that, in a given time, a given temperature will produce a color on the
side of the cake nearest the fire that will be different from the color

(a)

nnnnnn

FIG. 97.

on the opposite side. It is also likely that the variation between
any two points on the same side of the tray will be very slight, but
that the variation between the under and upper sides of the tray
will be considerable. This indicates the possibility of turning out
a product not uniformly heated from a chamber that may be uni-
formly heated when empty, or but partially filled, and at the same
time constantly indicating a uniform temperature on any lateral
plane.

HEAT APPLICATION 195

This illustrates how weak and irrelevant such claims as previously
noted are to the real question — the uniformity of the heated product.
The very manner of placing the stock in the chamber may affect the
final result. Uniform temperature in the chamber is a part but not
all of the process. It is the peculiar conditions or requirements of
each case that determine the type of furnace to use and the manner
of heating the steel in it. There is no such thing as an ideal furnace
for heat treatment any more than there is an ideal engine or fuel for
power, for in one case as in the other, the point is determined by the
working conditions.

Heat Application. — Heating a piece of steel, boiling a cup of water
or baking a potato are alike heat-treating operations, and in so far
as each leads to an absorption of heat, they are comparable. A cup
of water will boil much sooner if the heat is applied from the bottom
rather than from the top downwards. In the ordinary gas stove
the oven is heated from jets below, while these same jets deflected
downwards heat the broiler from above; a potato in the oven is
heated evenly through to the center without burning the outside,
but a potato placed on the broiler is burned to a crisp on the top
while the center remains hard and uncooked.

Many carburizing boxes give evidence of having been " broiled "
rather than " baked." Boys still roast potatoes in open bonfires;
men still heat steel in open smith-fires; and the results are about the
same. An inefficient plant means that the product seldom reaches
and never maintains the standard to which it is entitled, and also
that the cost of up-keep and labor far outweigh any difference in first
cost of plant.

To make results harmonize with the requirements in each case
requires furnaces of the best possible construction, fuel that would
cost the least viewed from the standpoint of finished product,
and it further requires that these furnaces be so arranged and such
methods devised that the material may be heated and handled with
the least labor and loss of time.

Only when this harmonious combination of suitable furnaces,
right fuel, proper furnace layout and efficient material handling
conditions has been secured can there be accomplished the best
heating of a product at the least total cost.

The " One " Furnace. — One often reads of the claims of furnace
manufacturers that this or that furnace is the " only one " for heat
treatment. This is all wrong, as there is no one type of furnace for
heating any more than there is one type of building for machine

196 STEEL AND ITS HEAT TREATMENT

operations. There are certain principles of construction, heat
generation, application or utilization that, when properly combined,
make up the right furnace; but it is always the local shop conditions
that determine how these combinations should be effected. There
is as much more experience and skill required for the determination
of a furnace design than in building it, as there is between an architect
and builder or carpenter in the design and erection of a building.

Furnace Guarantees. — One would not think of asking a stove
manufacturer to guarantee good cooking with his stove, and it is as
unreasonable to expect, as it is foolish to offer, a furnace guaranteed
for good heat treatment without the proper handling. The right
furnace, like the right stove, only makes it possible; and it is the man,
like the cook, that determines the final result. What is most needed
at this time is a better appreciation of heat application to useful
work, as removed from heat generation, combustion or utilization.
There should be more study given to the absorption of heat by the
product and the manner of placing and handling to secure the best
heating results.

Uniform Heating. — The problem of applying heat to industrial
work is but a cooking operation, like the baking of bread; the bread
must be heated uniformly throughout. Similarly the charge in a
heat-treatment furnace may be best heated when it has opportunity
to absorb heat uniformly from all sides. And as far as the heat
absorption is concerned it does not matter whether this heat is
supplied as radiating heat from the lining of the furnace or through
the direct application of hot gases. Except in the case when elec-
tricity is used as the source of heat energy, it may be said that the
majority of commercial heat-treatment furnaces involve the applica-
tion to the work of hot gases obtained through a combustion process.
If it were then possible to apply this heat so that the charge would
be equally and simultaneously heated on all sides the general ideal
condition would be met.

Underfiring. — In principle,1 it is best to apply the heat first to
the bottom of the charge. It is natural law that heat or hot gases
tend to rise — a very simple and important fact and yet one so fre-
quently overlooked in industrial heat application. Further, since
in practical, every-day work the height of suspension of the charge
(in order to provide for circulation around the entire mass) is more
or less reduced to a minimum, and which necessarily results in the

1 We say this, because we will show later that certain conditions may entirely
reverse this in practice. — AUTHOR.

HEAT APPLICATION

197

major part of the charge to be heated being near the hearth or even
laying directly upon it, we may consider that the best construction
in general to adopt is to place the initial heat where it is most needed
— which is under the charge. As much advantage as possible is then
taken of the natural law of hot gases rising in effecting a further
application of that heat to the sides and top of the charge. This,
in effect, is underfiring.

Simple Under-fired Furnace. — A common type of heat treatment
furnace is illustrated in Fig. 98. The heat is generated in the com-
bustion chamber under the hearth, supplying heat to the hearth.
The hot gases then rise upon either side of the hearth to the roof,
where they become mixed or equalized, and then are forced down
upon the floor of the chamber by the pressure which has been built

FIG. 98.

FIG. 99.

up. In the type of furnace shown it will be noted that the furnace
door opening is the same height as the roof arch, and that there is
also a vent located in the roof. Such a design is permissible for
small furnaces built upon legs to lighten the weight for transpor-
tation, and where first cost is an important feature.

Side Ledges. — One of the objections to this common type of
furnace is the inherent tendency to overheat the sides of the hearth.
As the Tiot gases sweep upwards from the combustion chamber
towards the roof the sides of the hearth will become hotter than the
central area, with the consequent superheating of any material
placed at the sides of the furnace near the ports. To overcome this
tendency to localized heating, the sides of the hearth are frequently
protected as illustrated in Fig. 99. Such an arrangement affords
plenty of room for circulation at the sides, tends to prevent cutting
action near the floor line and automatically stops any overloading
at the sides of the furnace.

198

STEEL AND ITS HEAT TREATMENT

High Ledges. — It was originally thought that the flow and heat
transfer of the hot gas currents would be in an underfired furnace,
as in Figs. 98 or 99, from the combustion chamber to the hearth and
the metal on the hearth, and thence to the roof. It was then thought
that if the hot gases, after leaving the combustion chamber, could
be made first to travel to the roof, be thoroughly equalized, and
then be brought down to the hearth, that a more uniform heating
of the charge would result. The furnace in Fig. 100 shows an early
development of the underfired furnace in a misguided attempt to
do this. The side ledges were made to reach nearly to the roof so
that the gases had to go there directly after leaving the combustion
chamber. Experience with furnaces of the type shown in Fig. 100
soon showed, however, that the hot gases — following natural law —

FIG. 100.

would first go to the roof whether these high side walls were there or
not. In other words, they were proven to be not needed.

The second advantage hoped for through the high ledges in Fig.
100 was entirely to prevent any localized heating at the side of the
charge. But it was then discovered that the small opening between
the top of the ledge and the roof arch tended greatly to increase the
velocity of the gases as they entered the heating chamber — that is,
to form a blast action at the top of the furnace. Thus, if the charge
were anywhere near the height of the door opening, a localized heat-
ing at the edges and top of the charge would at once result. For
this and the previously mentioned reasons this type of furnace
(Fig. 100) is not generally used now.

Roof Vents. — It will be noted that in the furnaces in Figs. 98
and 99 there is a vent in the rocf arch, and while this is general prac-
tice, it is not good. One writer, in discussing this point, states that
" as only approximately 20 per cent, of the air for combustion is

HEAT APPLICATION

199

oxygen, the balance is inert gases which unfortunately must be heated
to the temperature of the furnace and expelled as quickly as possible.
In a scientifically designed furnace, this is readily done by the aid of
the burner. If allowed to pocket or remain stationary in any
portion of the furnace, the inert gases cause uneven temperatures.
" . . . The vents for the escape of ... the consumed and inert
gases should always be located in the oven roof or arch."

In the first place, that writer evidently loses sight of the fact
that, with perfect combustion, all the gases become " inert " (assum-
ing that he means a non-supporter of combustion); and that the
furnace and stock are principally heated by the blanket action of hot
gases. He admits that these gases are hot, but then reasons that
because they are hot and inert they must be " expelled as quickly
as possible," But why throw heat away? In other words, he

FIG. 101.

believes in trying to heat his house in winter time by throwing open the
skylights or windows! He fails to perceive that with an open vent
in the roof the hot gas currents will tend to short-circuit directly
from the combustion chamber to the roof and discharge at the max-
imum chamber temperature. (Incidentally, exactly what part the
burner — which is merely a valve for injecting oil into a furnace-
plays in this ejection is not mentioned.) But by eliminating the
vent in the roof the gases are given ample opportunity thoroughly to
mix in the upper part of the chamber and are then forced down as
a blanket at a reduced velocity upon the stock. This same pres-
sure, as we have explained in the previous chapter, will prevent cold
air from the outside from finding its way into the furnace. If
" pocketing " is feared this may easily be overcome by providing
some flue outlet on a level with the hearth, as in Fig. 101.

Vents and Cold Streaks. — The question of roof vents may also
be approached from a different angle — that they will tend to set up a

200

STEEL AND ITS HEAT TREATMENT

current of cold air through the heating chamber. The natural
tendency of roof vents is, as we have previously said, to short-cir-
cuit the hot gases through the roof. This means that there will be
a pull or suction around the door towards the inside of the furnace,
across the charge and thence to and out through the roof. Cold
air, with free oxygen, will, therefore, be sucked into the furnace
and will cause scaling, non-uniform heating and widely variant
results. The further result of these vents is a loss of heat and destruc-
tion of the lining due to quick cooling after the burner has been shut
off. This condition is emphasized by an unfortunately common
type of heating furnace shown in Fig. 102; in this instance the heat
is supplied from above the hearth, and by looking into a furnace of
this type a dark streak of cold air will be seen to lay directly over the
steel1 on the hearth.

FIG. 102.

Door Heights.— It will also be noted that in Figs. 98 and 99 the
height of the door opening is the same as the height of the roof arch ;
this is also objectionable practice under most circumstances. It is
reasonable to assume that the height of the door opening is governed
by the maximum height of the charge desired, plus a reasonable
allowance to facilitate the handling of the material. If this is
true, then it may be expected that the height of the charge
may frequently reach almost to the roof. This condition will lead
to localized heating at the top of the charge; and if there is an open
vent in the roof such a condition will be magnified for reasons previ-
ously mentioned. It is, therefore, advisable to have the roof higher,
as illustrated in Fig. 103, so that even with the maximum charge
there will be ample space for the gases to become thoroughly equalized
and to prevent the charge from encroaching upon the hotter zone
close to the roof.

1 This particular example is taken from a well-known spring shop.

HEAT APPLICATION

201

The Heat Reservoir. — Further, when the door is opened in the
furnace in Fig. 103, there is always a reservoir of heat left in the upper
part of the furnace to aid in heating the next charge and maintaining
the temperature of the furnace during discharging or recharging

FIG. 103.

FIG. 104.

FIG. 105.

operations. With the furnaces hi Figs. 98 and 99 this is impossible,
for when the door is opened the furnace will be almost entirely
emptied of its hot gases because the door opening is the same height
as the roof. This is perhaps more clearly illustrated by the furnaces
in Figs. 104 and 105. It is better practice to have the roof higher

202

STEEL AND ITS HEAT TREATMENT

than the door opening, and always to keep the door height as low
as possible, as is illustrated by the furnace in Fig. 106.

Height of Chamber vs. Height of Charge. — In order to make
quick use of the heat which is thus retained in the furnace, it is a
part of the furnace man's job to provide for ample space for circula-
tion throughout the mass. Thus with the charge arranged as in
Fig. 104 there is ample room for the circulation of the hot gases
around the charge, and the heat application would be much better
and more uniform than in Fig. 105, where the top of the charge is
near the roof. From this it will be seen that the size and arrange-
ment of the charge is an indication of the heat application value
of a furnace. Better results will be obtained in a high chamber
with a low charge than if the charge is higher.

FIG. 106.

It is better practice to have the chamber high enough to afford
plenty of space for circulation, as is illustrated in Fig. 106. This
diagram also illustrates the points previously made concerning roof
vents and door heights.

Influence of Mass. — The furnaces shown in Figs. 107 to 115 incl.
illustrate different commercial methods for the pot annealing of
wire, and are particularly a propos as examples of heat application
on account of the size of the mass to be heated.

Thus in Fig. 107, we have a furnace with a combustion chamber
under the hearth, and with a large pot resting directly upon the
hearth. No matter how uniform the temperature may be in the
chamber there will always be a tendency for a cold zone to form at
the bottom of the pot — as is represented by the shaded portion in the
drawing. This is due to the fact that the floor of the furnace is of a
refractory nature, and will not transmit the heat as fast as the pot

HEAT APPLICATION

203

will take it away. The placing of the pots in the furnaces in
Figs. 109, 110 and 111 is better in this respect, inasmuch as it pro-
vides for the circulation of the hot gases under the pot.

FIG. 107.

FIG. 108.

With a short pot, as in Fig. 107, it is sometimes permissible to
use a furnace in which the heat is applied from the top of the charge
downwards; but there is a limit to this because when the height of

iip^l t^y

V///////////////A

FIG. 109.

FIG. 110.

the pot increases, as with Figs. 108, 110 and 111, there will be a
tendency to overheat the top of the pot before the bottom is at the
right temperature. For this reason the height of the pot (or charge)
in itself determines the type of furnace.

204

STEEL AND ITS HEAT TREATMENT

In some plants the heat is applied to the bottom (but without a
separate combustion chamber) and taken off at the top, as in Fig.
110; but even with this construction it is difficult to secure a uni-
formly heated product, as the temperature of the top and bottom of
the furnace rarely would be the same. In this respect the design of
Fig. Ill is better. This latter provides for underfiring, for circula-
tion under the pot, and further takes the gases off at the hearth
(not shown) , inasmuch as the heat must rise to the roof and return
to the floor before it can escape.

When the construction in Fig. 110 (with a heavy roof -door)
takes the form of that in Fig. 112 (with a cast-iron top), as it fre-

FIG. ill.

FIG. 112.

quently does in practice, the method is open to still more criticism.
In either case the removal of the door, which constitutes the roof
of the furnace, permits the escape of a maximum amount of heat
and cools off the furnace. But in Fig. 112 the roof, being of metal,
radiates a great amount of heat even with the top on the furnace,
and is severe on the men. It is better practice to employ a type of
furnace in which the charge is introduced through a door or opening
in the side or end instead of through the top of the furnace.

Fig. 113 illustrates a construction of pot that is employed very
successfully in annealing wire, and it gives very good results, inas-
much as the heat is applied to the center of the coils as well as the

HEAT APPLICATION

205

outside, and to the bottom as well as the top. It goes to prove that
the type of charge has much to do with uniform heating.

Figs. 114 and 115 illustrate the relative merits of heating high
pots from the bottom up (Fig. 114) and from the top down (Fig.
115). In the underfired furnace, also provided with a movable
ball type carriage or charging device, the heat is again first applied
where it is most needed — at the bottom — and the hot gases rising
will naturally take care of the heat application at the top. In the
overfired furnace (Fig. 115) the gases must descend and even though

FIG. 113.

FIG. 114,

the heat js forced to circulate under the hearth, that part of the hearth
under the charge will always be the cold spot, due to the fact that the
absorptive powers of the iron pot is greater than the heat input of the
hearth by the hot gases thereunder. The relative advantages of the
two methods of heat application will be more fully discussed in sub-
sequent sections, but it is here evident that the height of charge
again determines the method of applying the heat.

Influence of Character of the Charge. — A charge in a furnace
heats up in about the same manner as a plate of ice cream while
melting— that is, from the top and outside edges towards the center —

206

STEEL AND ITS HEAT TREATMENT

particularly when the heat is applied from above. The difference
between uniform chamber temperature and uniformly heated product
is illustrated by Figs. 116 and 117, in which it is assumed that the

FIG. 115.

charge consists of small pieces — such as lock-washers or cartridge
shells — laid on the chamber floor. It is immaterial in this case
whether the furnace is underfired (Fig. 116) or overfired (Fig. 117),

FIG. 116.

or how uniform the actual temperature in the chamber may be,
because there will always be a tendency for a cool section in the
center, as shown in Fig. 118. It is practically impossible to get a

HEAT APPLICATION

207

uniformly heated product under such conditions unless the charge
is so split up as to permit free circulation of heat through the mass.
The real test in annealing, hardening or tempering is in the uni-
formity of cooling — that is to say, it is not only necessary that a piece
should be uniformly heated, but it must also be uniformly cooled.
For this reason, instead of heating small pieces as shown in Figs.
116 and 117, it is better practice to employ an automatic type of

FIG. 117.

furnace like Figs. 119 or 121. With these designs each surface
of each piece is exposed to the action of the heat, and it is heated
and cooled uniformly. In other words, the material to be heated is
indicative of the method of heat application and furnace design.

Muffle Furnaces. — Muffles, as ordinarily constructed for heat
treatment work, do not necessarily prevent oxidation, other claims

FIG. 118.

to the contrary. The following are statements taken from recent
publications on this subject: (a) " \Vhenoxidation or the formation
of scale is particularly objectionable, furnaces of the muffle type
are often used, having a refractory retort in which the steel is
placed so as to exclude the products of combustion." (6) " The
metal does not become saturated with any of the products of
combustion . . .", referring to the furnace illustrated in Fig. 123.

208

STEEL AND ITS HEAT TREATMENT

The first statement, referring to the merits of the muffle versus
the open chamber, must lead to the conclusion that muffles are synon-
ymous with improper open chamber work. We have previously
explained that, with proper furnace design and correct operation,
an atmosphere may be produced which will contain no free oxygen,

FIG. 119.

FIG. 120.

no oil vapors, and just enough carbon monoxide to take care of any
air which may possibly find its way into the heating chamber through
unforeseen causes ; that this slightly hazy atmosphere results from an
absolute control of the air supply, in combination with the right
furnace design; and that a furnace operated in such a manner will

FIG. 121.

FIG. 122.

give a product which will often be better than that heated under
charcoal. Thus when it is found necessary to exclude the products
of combustion from contact with the hot steel, it means that such
gases contain free oxygen, and which is identical with improper
furnace operation or design.

HEAT APPLICATION

209

Again, if the products of combustion are excluded from the muffle,
the question reverts to the fact that there is no method of keeping
the outside air or oxygen from finding its way into the muffle. As
there is no pressure of gases from within, since the pressure which
should be caused by the products of combustion is absent, the free
oxygen must inevitably find its way into the muffle. For these
reasons, therefore, muffles do not prevent oxidation.

Semi-Muffle Furnaces. — On the other hand, if a pressure is built
up from within the muffle to prevent air from entering, there must
be some opening between the muffle and the hot gas chamber sur-
rounding it. In such a case it is no longer a true muffle and does not
exclude the products of combustion. Thus in Fig. 123 there are

FIG. 123,

shown openings in the roof of the muffle. If the course of the hot
gases is such that the gases come downwards through these openings,
then why have any roof at all on the muffle? — for the products of
combustion enter the heating chamber and the furnace approximates
open chamber construction. Or, if the flow is upwards through these
openings, then outside air will be sucked into and through the muffle,
and oxidation will be set up in that manner. In other words, whether
or not the furnace, as, and for the purpose designed, is properly
operated, there is no occasion for the remaining solid part of the
muffle roof, and the statement contradicts itself.

Influence of Nature of Fuel on Furnace Design. — The overfired,
perforated-arch type of furnace — under which Fig. 123, previously
discussed, may be classed, and which general type is common to the
construction shown in Fig. 124 — is a development originally intended

210

STEEL AND ITS HEAT TREATMENT

to meet certain conditions with oil fuel, but which is neither necessary
nor desirable with gas or coal. The development also illustrates the
difference between uniform application of heat and uniform applica-
tion of fuel which was discussed in the last chapter (q.v.).

With a type of rotary annealing furnace like that in Figs. 121
and 122, using gas as a fuel, the burners are usually placed on
both sides, so that a small amount of fuel is injected through each
burner. This distributes the application of the fuel. With oil,
however, the consumption under a similar arrangement of burners

FIG. 124.

would be so low that the burners could not be kept going steadily,
and for this reason it was desired to employ but one burner. This,
in turn, was open to the objection that there would be a hot streak
directly in front of the burner and which would react unfavorably on
the charge. To overcome this the perforated arch was employed so
as to lessen the streaking effect of the flame and thus distribute the
heat, as is shown in Figs. 119 and 120.

It is therefore evident that the nature of the fuel in these cases
determines the design of the furnace. In the first case (Figs. 121
and 122) the uniformity of heating, aside from the nature of the

HEAT APPLICATION

charge, is secured by a uniform burn-
ing of the fuel — gas — throughout the
length of the furnace. In the second
case (Figs. 119 and 120) the fuel-
oil — input is concentrated and the
perforated arch construction is em-
ployed to secure the heat distribution.
The perforated arch, which was found
advisable in the case of oil, for the
purpose in view, is entirely unnecessary
with gas fuel.

Another illustration of this per-
forated-arch type as applied to a
long, low hearth with concentrated

FIG. 125.

fuel supply, and which followed the
above development, is shown in Figs.
125 and 126. In this case also the
fuel input is from one side of the fur-
nace and above the hearth; the per-
forated , arch distributes the heat to
the chamber beneath and from which
the gases pass underneath the hearth.
But it should be remembered that
while the perforated arch construc-
tion may be perfectly proper under
certain conditions, it may be en-
tirely out of place under other
conditions, and yet using the same
fuel.

212

STEEL AND ITS HEAT TREATMENT

Perforated-arch Furnaces. — From the type of furnace of Figs.
125 and 126 it is but a short step to the overfired, perforated-arch
furnace shown in Fig. 124 (previously alluded to), and which has
been somewhat widely employed for general heat treatment work —
often regardless of distinctive shop and heat application conditions.
It will be noted that there are burners on both sides of the furnace
(generally staggered), that the combustion takes place in a cham-
ber above the main wyorking chamber, and that the gases then pass
down through the perforated arch into that chamber and are taken
out from under the floor.

FIG. 127. — De-carburization of Steel by High-velocity Gases.
X60. (Bullens.)

This design provides, in effect, for the application of heat from
above through a perforated arch, and is permissible with low charges,
but when the charges are high there is a tendency to overheat the
top. If a thorough study is made of the perforated arch itself,
it will be found that the actual openings only total about twenty-
five or thirty per cent, of the total chamber area. With a continual
input of fuel and air into the hot combustion chamber above, and
with but a small exit for the hot gases, these hot gases must enter
the working chamber at a high velocity. If the furnace is charged
to anywhere near the height of the working opening or arch, which

HEAT APPLICATION 213

is a common procedure, the hot gases will impinge upon the top of
the charge at high velocity. This inevitably results in severe cut-
ting action and oxidation, the zone of which is shown in the charge
in Fig. 124. This is further illustrated by the photomicrograph
of Fig. 127, taken from the edge of an annealing charge of chrome
nickel steel plates piled as illustrated in Fig. 124. It will be seen
that the steel has been entirely decarburized along one edge, even
though the actual indicated temperature of the furnace was only
about 1350° F. In the plant from which this example was taken it
was no uncommon occurrence to lose as much as J or even J in. of
metal on each side of a pile 10 or 12 in. wide.

Aside from changing the type design of the furnace, the only
method of overcoming this particular trouble without decreasing
the production is to increase the height of the working chamber.
In this manner the gases are given an opportunity to expand before
reaching the metal, and thus reduce the high velocity caused by the
perforated arch. In other words, an overfired, perforated-arch
furnace is permissible with low charges, in comparison with cham-
ber height, such as is shown by the dotted line charge in Fig. 124.

Overfired Furnaces. — In order to overcome the necessity for
comparatively high-working chambers, as occasioned by the con-
ditions above referred to, the perforated-arch construction may be
eliminated. This results in a type of overfired furnace illustrated
in Fig. 128. In this case the charge is heated from the top down-
wards, and the gases pass out from under the floor. Aside from
the fact that the hot gases have to pass downwards, the question
then arises as to the manner in which the charge is heated. Since
the heat is applied from above there is no question but that the
top of the charge will be heated ; but how about the bottom of the
charge?

It should be borne in mind that this construction (with flues
under the hearth) does not necessarily result in a hot floor, even
though it be granted, for sake of argument, that there is a consider-
able volume of gases under the floor and that the temperature of
these gases is the same as above the floor. The specific heat of the
charge is such that it absorbs heat from the floor, and unless the rate
of input is greater than the rate of absorption the floor will cool under
the charge in proportion to the manner in which it is packed. This
cold spot, in turn, means a cold zone through the center and bottom
of the charge, and until this cold zone is removed the charge is not
uniformly heated. But in order to remove it, it is necessary to

214

STEEL AND ITS HEAT TREATMENT

lengthen the time of exposure, which results in a tendency to expose
the outside edges of the charge to the action of the heat and gases
longer than is necessary with other construction. In practice this
cold spot is never totally eliminated in this type of furnace.

Even with the I-bar floor construction illustrated in Fig. 128,
which is the best in use for this type of furnace, the floor is not as.
hot as it should be. The reason for this is that, irrespective of
the temperature or volume of heat under the floor, the rate of trans-
mission of heat to the under side of the charge is no greater than
that possible through the vertical section of the I-bar. The rate
of transmission through the tiles separating the bars is still less than

FIG, 128.

through the bars themselves on account of the low conductivity
of the material. When the construction is made without the
I-bars, as it sometimes is to lower the cost, the conditions are still
worse.

From this it will be seen, as has been repeatedly proven in
practice, that there is a disadvantage in any construction which
does make possible a floor temperature equal to that above the
charge; or a circulation of heat under the charge to make up the loss
in transmission and decrease the time of exposure by decreasing
the area of the cold zone.

Influence of Arrangement of Charge. — Such a condition just
described illustrates very forcibly the difference in heat application
obtained when a furnace is full and when it is empty. That a

HEAT APPLICATION 215

furnace will give uniform temperatures without a charge is no cri-
terion that the heat application to a charge will be uniform. The
type of furnace illustrated in Figs. 129 to 134 is used extensively
in the annealing or wire and tool steel, and many people wonder
why it is that with almost perfect pyrometer records they do not get
a uniform product.

The furnace is fired with coal from a fire-box at one end, the flame
and heat passing over a bridge wall to the heating chamber; at the
other end of the hearth the hot gases pass down and under the
hearth through a series of flues, and from thence to the chimney.
The hottest part of the hearth is near the bridge wall.

In the case of the charge of wire in Figs. 129 and 130 the non-
uniformity of product is partly due to the fact that the first piece
in is the last piece out and vice versa, so that the first piece is exposed
to the highest temperature for the longest time, and the last piece
to the lowest temperature for the shortest time.

With the method of charging tool steel in Figs. 131 and 132 the
non-uniformity is partly due to the lack of circulation through the
charge; the tubes rest directly upon the hearth and are packed
tightly together. This can be somewhat overcome by rearranging
the tubes as in Figs. 133 and 134, separating them and raising them
up from the hearth.

All of these (Figs. 129-134) are open to the objection that the
heat is not uniform throughout the length of the charge, and while
it is possible to vary the quality of the product by rearranging the
charge without affecting the pyrometer readings, as above illus-
trated, there is still the fact that the heat should be uniformly
applied throughout the entire length and through the mass in order
to secure a uniformly heated product.

This latter point is also illustrated by Figs. 135 and 136. In
carburizing work the practice of Fig. 135 is often followed, filling the
furnace to its maximum capacity by packing the boxes up to the side
walls as well as in front. It is much better practice to maintain
circulating space on the sides and ends, as shown by Fig. 136, and
not to place the boxes beyond a certain imaginary line such as is
illustrated by the dotted line of the drawing. The method of
handling or arranging the charge in the furnace is equally important
with correct furnace design and proper operation or heat distribution.

Other Furnace Designs. — The designs of furnaces intermediate
between the two principal types — underfired and overfired — are
legion. One characteristic furnace in common use is that illustrated

216

STEEL AND ITS HEAT TREATMENT

HEAT APPLICATION

217

by Fig. 137. In this it will be noted that the underfiring principle
has been used, locating the combustion chamber under the hearth;
that the hot gases pass upwards to the heating chamber on one

FIG. 135.

FIG. 136.

FIG. 137.

side of the hearth; and that a roof vent is located on the opposite
side of the chamber. In this design the benefit of underfiring is
largely negatived by the poor heat application to that part of the
charge and hearth directly under the vent and farthest removed

218

STEEL AND ITS HEAT TREATMENT

from the heating chamber intake ports; the hot gas currents will
short-circuit from the ports to the vent. A good overtired furnace
would be better practice.

Coal Furnaces. — Fig. 138 illustrates a common type of heat treat-
ment furnace using hard coal as the fuel. The heat is generated
from coal placed on the grate at the left of the furnace illustrated,
passes over the bridge wall into the heating chamber, and then under
the hearth to the flues and to the chimney. From points previously
raised upon other furnaces there will be noted the tendency to
localized heating near the bridge wall, the fact that the height of

FIG. 138.

the door opening is virtually the same as that of the roof arch,
and the tendency towards a cold hearth. In such coal furnaces as
generally designed and operated there is a decided lack of control
of the volume, temperature and composition of the gases to and from
all points in the chamber.

Again, when local conditions advocate the use of a cheap fuel,
such as coal (and why use hard coal at $7 or so a ton when soft
coal at $2 a ton could be made to do the same work?), the gener 1
method of burning it, as here illustrated, is found to be inefficient
from the standpoint of fuel application and unsatisfactory from
the heat application viewpoint. The only proper way to attack
such a problem is first to gasify the coal and properly to utilize that
gas; not necessarily to generate the gas in a separate producer

HEAT APPLICATION

219

and carry it by expensive flues to the furnace, but to combine the
two operations in one efficient, self-contained unit. Such is being
done, and such furnaces are today producing a better heated
product, at less operating cost, than many furnaces using oil or gas.

FIG. 189.

FIG. 140.

Car-bottoms.— Figs. 139, 140 and 141 represent three designs of
the car type furnace — the open-chamber, perforated-arch and coal-
fired furnaces respectively.

Furnaces with the movable car-bottom may be mechanically

220

STEEL AND ITS HEAT TREATMENT

efficient, but they are thermally inefficient. In cases where there is
much work to be handled of a large and variable size, the use of
such furnaces may be advisable. But it is the same proposition of
cold hearths versus hot hearths which has been previously discussed,
but carries the matter one step farther. In this case, each time
the hot car is removed from the furnace a large amount of heat
is lost, and the furnace must be refired with a cold hearth. Simi-
larly, the radiation losses during the time between the removal of
one charge and the placement of the new car are very great. How-
ever, the saving effected in both labor and time may be consider-
able under certain conditions as above noted, but in any event

FIG. 141

the charge should be raised from off the hearth lo give the best
circulation possible. Furnaces with car-bottoms should only be
used when there are no others which will prove as commercially
effective.

Underfired Furnaces. — The aim of any uniform heating opera-
tion should be to supply heat to all sides of the charge. Under
ordinary conditions, this is probably most nearly accomplished
in the underfired type of furnace with a perforated floor, and an
effort has been made to overcome the disadvantages previously
referred to in the discussion of other types of furnaces. Fig. 142
illustrates a recent patented type of underfired furnace, and com-
parison should be made with the overfired furnaces previously
described, such as in Fig. 128.

HEAT APPLICATION

221

The heating is done from the bottom upwards instead of from
the top downwards. Heat naturally rises, and with such construc-
tion as in Fig. 142, if the floor is hot the roof is hot, although it is
possible to obtain the reverse in an overfired furnace. Even
in an underfired furnace the bottom can never be heated more than
the top. In the construction outlined in the drawing, the gases
are passed through large combustion chambers and compelled to
circulate through several hundred feet of ports on both sides, as
well as through the floor, each of which exposes considerable area

FIG. 142.

against which the gases are wiped. In this way there is a minute
subdivision of the volume and a thorough mixture.

Such, a furnace design is also in accord with the fact that correct
heating is a function of pressure. Thus the heat, in going to the
roof, naturally stratifies and builds up a natural pressure, which
will be spread over the entire area before it will come down. The
result is a pressure always on the floor and the elimination of streams
of gases of unequal temperature and composition. The hot gases
surround the steel as a blanket, and have a minimum velocity.
With such methods of applying the heat, the area of the cold zone is
quickly decreased, and this, in turn, lessens the length of exposure.

It will also be noted that provision is made for circulation on

222 STEEL AND ITS HEAT TREATMENT

both sides of the charge, independent of the manner in which it is
packed. It is impossible to overload the furnace and cut off the
circulation, and even though the charge were the full width and
height of the working opening there would still be room on each
side for circulation. In the particular furnace illustrated, this
extra room costs about three feet in chamber width; the area for
circulation on the sides is about 30 per cent, of the width of door,
and it is by the room afforded with such greater width that the
velocity of the gases is cut down.

On the other hand, and in the case of the overfired furnace of
Fig. 128, it will be noted that if the chamber is the same width
as the working opening, it will be necessary to employ compara-
tively small charges in order to get circulation through the restricted
areas on the sides. To gain time, with this practice there is a dan-
ger of overheating for the reason that, as the area is decreased, the
pressure must be increased for a given B.T.U. input, which works
out in practice, as a rule, to the detriment of the top and exposed
edges of the charge.

Flue Construction. — For furnaces of any considerable size, flue
construction is absolutely necessary, not only to provide an escape
for the waste gases, but also to direct the hot gas currents during
their passage through the furnace, and to conserve the heat in those
gases after they have left the heating chamber proper. Primarily,
the practice is to circulate the gases around the stock to be heated,
to heat the chamber as a whole, and then to pass them out at the
coldest part of the furnace. Some of the furnace drawings given
have shown in some degree such provisions, but in order not to com-
plicate the discussion of heat application, the subject of heat con-
servation has been little dwelt upon. The latter is, in fact, a prob-
lem which must be studied out for each particular design. In any
case the flues should be arranged so as to prevent short-circuiting
of the hot gas cycle.

Conservation of Heat. — Additional flue construction and thicker
walls both tend towards the conservation of waste heat. The dis-
cussion thus far has had to do with single furnaces, and the extent
to which the thickness of the walls might be increased is obviously
restricted within narrow limits by the cost of construction. Since
losses by radiation are largely preventive, any arrangement or
grouping together of furnaces of a similar type which will tend to
unite them, thus eliminating exposed walls, should be made a sub-
ject of study.

HEAT APPLICATION

223

Variety of Furnace Plans. — The diagrams in Figs. 143, 144 and 145
are intended to illustrate as floor plans a few of the many different

furnace designs which are employed in practice. All of these, which
are more or less empirical, as well as hundreds of others not shown,

224

STEEL AND ITS HEAT TREATMENT

have been built in a variety of sizes for oil, gas, coal, coke and wood,
with different methods of applying the heat to suit different opera-
tions ranging from small needles to eighty tons of steel at a
charge.

In designing furnace equipment it is not only necessary to
consider combustion and the more important points of heat applica-
tion as well as the fuel suited to both, but likewise the method of
handling material to and from the furnace, together with the floor

WA w/,

y////////////////,

V77A

V//7A W/,

FIG. 144. — Unit Furnace System Development.

space available, which are no small factors in the cost of production
and installation.

The purpose should be to keep the material, the men, the fur-
naces and machinery in continuous operation, or as near it as possi-
ble, because each is more or less dependent upon the others and
all must be properly linked together to secure the best all-around
results. It is just as necessary to adapt the furnace design to manu-
facturing conditions as it is with machine tools, but the latitude
for variation is much greater.

Owing to the great variety of heating operations and shop
conditions, it is rarely found that the same identical furnace can be

HEAT APPLICATION

225

properly employed in two shops, or even in separate departments
of the same shop, for similar operations.

Thus these sketches will serve to illustrate some of the develop-
ment that has been made, as well as the latitude possible in design-
ing furnace equipment, and further shows that furnaces cannot
well be standardized owing to the great variety of conditions which
must be met.

Unit Furnace System. — Every large factory employing heat
treatment methods is more or less confined to a general type of
product which requires some particular and standardized treat-
ment. Such conditions would seem to be naturally and readily
adapted to the small-furnace unit practice, having elasticity of pro-
duction as its general aim. Thus it is no unusual sight to see tien
or fifteen small furnaces, often on legs, of similar type and size,
strung out in a row. And yet if the manager of that plant were to

\//////A

W7777A

W J

K////

'/////////////

FIG. 145.

visit a boiler room where each of ten or more boilers were each set
off by itself, how caustic would be his comments. The same prin-
ciples of heat generation, utilization and conservation are applica-
ble to both. Not only does the unit system of furnace arrangement
require, as a general rule, more care and attention, involve more
steps for the furnace man, longer time for heating and general
inefficient handling, but it also tends to raise the cost of fuel out of
all proportion to the product turned out. Radiation losses are
largely responsible for this last cost factor.

Experience has shown that three, or perhaps four, small single
furnaces of the same type for the same work, are about the general
economic limit of the small-furnace unit system.

To explain some of the diagrams in Figs. 143, 144 and 145,

226 STEEL AND ITS HEAT TREATMENT

previously referred to, and to develop the growth of the multiple
furnace, we might commence with a, b and c of Fig. 143.

In (a) the door opening is the full width of the chamber, giving
opportunity for packing the charge to the full width of the open-
ing; such a condition, as previously explained, is usually not advis-
able, as it tends to cut off circulation. It may be remedied as in
(6), by placing jambs on each side of the front.

In (c) there is an opening at each end of the furnace, allowing
for charging at one end and discharging at the other, or for work-
ing the same operation from both ends simultaneously. Such
construction is thermally bad in that it permits a direct draft from
one door to the other, with consequent loss of heat. If it were a
question of doing the same work from each side at the same time,
the layout might be changed as in (d) or (e) to prevent the cold
air draft. Or if it were a matter of charging and discharging simul-
taneously without interference, the design might better take the
form of (f), having the two openings at one end, and widening the
furnace to make up for loss in length. In the latter construction the
door openings could be smaller, and one side could be worked out
while the other side was being charged and heated.

Twin Chambers. — The next logical step in avoiding the small unit
system is to use twin chamber construction. For each exposed
wall of the unit system which is removed, the more can we increase
the sturdiness of construction, and there diminish the heat radia-
tion loss of the remaining walls without additional construction
cost. Thus the development of two furnaces of the (a), (b), (c)
or (d) type, Fig. 143, will be that as shown in (oa), (66), (cc) and
(dd). Such furnaces will have the advantages of better heat applica-
tion and conservation, and perhaps of handling material, but will
have the disadvantage that if one side " goes down/' the other will
usually have to go out of commission also.

In some cases the dividing wall may be omitted as in (g) and (h) ,
giving one large hearth. But in this case only one temperature
work can be done at one time in place of the two temperatures
possible in the twin chamber construction.

Furnace Batteries. — There are many plants which uso small,
light furnaces, built on legs which might advantageously combine
the smaller units into batteries and obtain both better heat applica-
tion and handling methods. Other things being equal, the construc-
tion and arrangement of Fig. 144 (6) and (c) would be much better
than that of (a).

HEAT APPLICATION 227

Other designs as necessitated by definite conditions of material
handling or shop efficiency are illustrated in Fig. 145, and may be
varied, of course, ad libitum.

General Furnace Considerations.- — Full data should be gathered
as to the sizes, shapes and approximate output, both maximum and
minimum, of the material to be handled, besides the specific treat-
ment desired.

Careful consideration should then be given as to whether or not
the operations can be made mechanical, i.e., automatic or semi-
automatic, or continuous. From this information the specific
dimensions of the furnace can be deduced. As a general principle,
it may be stated that the heating chamber should be large, but
with a minimum area of exposure or door opening.

The furnace should be of the best possible design to suit the par-
ticular work in hand and under the certain existing factory con-
ditions. The specific purpose for which the furnace is to be used
should be definitely decided — whether for annealing, hardening,
toughening, carburizing, etc., or for a combination of such opera-
tions. In this connection it may be said that the maximum effi-
ciency of any heat treatment operation or furnace is obtained by
using a furnace operating continuously on one class of work at one
temperature and especially designed for that purpose.

Much thought should be given to the layout efficiency. The
furnaces should be so arranged and such methods devised, that
the material may be heated and handled with the least labor and
loss of time.

Practical Notes. — In concluding the discussion of heat applica-
tion, the author would ask for a thoughtful consideration of the
following additional suggestions on heating:

(1) The rate of heat absorption by the' object, under the regular
methods of packing and furnace operation, should be obtained. The
furnace, man should know exactly the length of time it takes
thoroughly to saturate that piece of steel under definite working
conditions, and with the furnace maintained at the desired tem-
perature.

(2) A furnace should always be maintained at one tempera-
ture, not permitting the indicated temperature in the furnace to
go higher than that desired in the finished product. Forcing the
furnace after each discharge is bad practice, and tends to place too
much confidence in the weak link — the human element.

(3) The pyrometer and the time clock should go together.

228 STEEL AND ITS HEAT TREATMENT

(4) Heating and handling methods should tend to the prin-
ciple of putting a cold piece in when a hot piece is taken out; of
heating small units at one time (not meaning small furnaces, but
units of charge) rather than increasing the mass; of giving the last
piece in the same amount of heat as the first piece.

(5) There is no one fuel, furnace, or any other
will satisfy every condition.

CHAPTER X

CARBON STEELS

Foreword. — In this and in the following chapters on various
steels it is the author's intention to give the physical results which
are representative of the different steels and their treatment. Such
results have been gathered from practical work and experiment,
and although the results of various treatments will vary according
to the individual steel and the personal equation of the operator,
they may be considered as fairly representative of the steel and treat-
ments given.

Further, it must be remembered that the size of section or mass
of the steel has a very important influence upon the physical test
results. The same results will not be obtained in a steel bar of
4 ins. diameter as in a bar of the same steel with similar treatment
and of only 1J ins. diameter. Similarly, different results will be
obtained near the outer surface of a large forging in comparison
with a test taken near the center.

As an example of the effect of the size of piece upon the tensile
strength, under the same treatment, we may cite the following
examples :

Diameter
of Bar
Inches.

Tensile Strength.
Lbs. per Sq. In.

137,000

132,000

127,000

122,000

113,000

105,000

100,000

Hardness vs. Maximum Strength. — The following equations
connecting maximum strength, Brinell hardness number and sclero-
scope hardness number have been computed l from several hundred

1R. R. Abbott, A. S. T. M., Vol. XV, Part II, 1915, p. 43 et seq.

229

230

STEEL AND ITS HEAT TREATMENT

tests made with carbon steels of different carbon content and heat
treated to bring out all possible physical properties:

(1) M = 0.735-28.

(2) M=4A £-28.

(3) £ = 5.6 S+14.

M = maximum strength in units of 1000 Ibs. per sq. in.
# = the Brinell hardness number.
S = the scleroscope hardness number.

The maximum strength corresponding to different Brinell
values as determined by equation (1) for carbon steels is as follows:

Brinell.

Maximum Strength,
Lbs. per Sq. In.

Brinell.

Maximum Strength,
Lbs. per Sq. In.

ICO

45,000

350

227,000

150

81,000

400

264,000

200

118,000

450

300,000

250

154,000

500

337,000

300

191,000

550

373,000

The maximum strength corresponding to different scleroscope
values as determined by equation (2), and the corresponding Brin-
ell numbers as determined by equation (3), for carbon steels, are
as follows:

Scleroscope.

Maximum Strength,
Lbs. per Sq. In.

Brinell.

60,000

126

104,000

182

148,000

238

192,000

294

236,000

350

280,000

406

324,000

462

368,000

518

100

412,000

574

VERY LOW CAKBON STEELS: UNDER 0.15 CARBON

The " dead soft " steels, about 0.10 per cent, carbon, find but
little application to heat-treatment purposes. Contrary to general
opinion, these steels do respond to heat treatment, although, of
course, to a very limited extent. The best treatment to which these
steels can be subjected is a quenching from about 1550° to 1600° F.,

CARBON STEELS

231

with or without reheating, the reheating being omitted when con-
ditions (of strain caused by quenching) will permit. Such a
treatment will refine the grain and remove any strains set up by
previous working; will confer added toughness; and will put the
steel in the best condition for machining. This last is an important
point, for this very low carbon steel, without high manganese and
phosphorus, and in either the annealed or toughened condition, often
does not machine freely, but is apt to tear badly in threading and
turning operations. The heat treatment of very low carbon steel
as applied to the wire industry is discussed in a subsequent chapter.

Annealed. — For annealing 0.1 per cent, steel, heat as rapidly as
consistent with the size and shape of the piece to a temperature
slightly above the upper critical range of the steel, approximately
1600° F. In these very low carbon steels the change into austenite
takes place very rapidly, so that it is only necessary to allow the heat
to penetrate the steel at the temperature noted above. As the
grain begins to coarsen rapidly with increase in temperature and
length of time held there, care should be taken in not overheating
nor maintaining the annealing temperature for too long a time.
Cooling may be carried out comparatively rapidly without danger
of hardening the steel. The annealing of these very low carbon
steels is usually for the purpose of relieving such strains as may be
incurred by cold crystallization or by previous heating at a low-
red heat for any great length of time.

Heat Treated. — The results obtained from the treatment of test
bars | ins. in thickness of acid open-hearth steel of the composition:

Per Cent.

Carbon 0.10

Manganese 0.32

Phosphorus 0.028

Sulphur 0.024

Silicon 0.019

are given in the following table:

Steel.

Tensile
Strength.
Lbs. per
Sq. In;

Elastic
Limit .
Lbs. per
Sq. In.

Elongation.
Per Cent,
in 8 Ins.

Reduction
of Area.
Per Cent.

As rolled

51 625

35500

34 5

65 3

Annealed

48,800

31,770

37.5

67.5

Water quenched from 1575° F. .

64,720

45,550

22.8

61.15

1575° F. Water/13000 F

53,055

36,500

35.35

66.05

232

STEEL AND ITS HEAT TREATMENT

GENERAL SPECIFICATION, ANNEALED

Tensile Strength.
Lbs. per Sq. In.

Elastic Limit.
Lbs. per Sq. In.

Elongation.
Per Cent, in 2 Ins.

Reduction,
of Area.
Per Cent.

45,000
to

28,000
to

40
to

65
to

55,000

36,000

0.15-0.25 CARBON STEEL

This grade of straight carbon steel is generally known to the trade
as " machinery steel," and as such has innumerable uses where
strength is not an all-important factor. The steel forges and ma-
chines well. The lower carbons find their greatest application in
the case-hardening processes which have been previously described.
The higher carbons are used considerably in certain engine forg-
ings such as tie rods, valve stems, nuts, flanges, pins, levers, etc.;
for machine work of various description; for structural purposes
in automobile construction, etc.

Heat Treated. — Heat treatment of the lower carbons of this
range confers but little additional strength except in thin sections,
but does have a most desirable influence in the refinement of grain
after forging or other elaboration. Hardening should be done
from a temperature exceeding the upper critical range — which is
about 1550° F. for 0.15 per cent, carbon, and about 1525° F. for
0.20 per cent, carbon — in order to effect the full absorption and
diffusion of the excess ferrite. Some engineers recommend quench-
ing at 1650° F. or even higher, but the author believes that such
high temperatures are not only detrimental on account of a greater
tendency to warping, oxidation and higher cost of treatment, but
are also unnecessary metallurgically. In other words, those tem-
peratures should be used which will produce the most efficient com-
bination of physical properties, refinement of grain and low cost of
production. From the results of extensive research work upon
0.18 to 0.28 carbon stock used for automobile purposes, and from a
study of its working out in practice, the author recommends a
quenching temperature of about 1500° to 1525° F. for these steels.
Temperatures lower than 1500° do not bring out the full effect
of the treatment, as is shown by the following average results (from
a large number of tests) upon the same steel;

CARBON STEELS

233

Quenched in Oil
from — ° F.; Re-
heated to 800° F.

Tensile Strength.
Lbs. per Sq. In.

Elastic Limit.
Lbs. per Sq. In.

Elongation.
Per Cent in
2 Ins.

Reduction
of Area.
Per Cent.

1450
1500

70,220
79,590

43,460
52,500

24.1
25.6

48.4
52.6

With hardening temperatures higher than 1550° F. there is prac-
tically no increase in the physical properties worthy of mention,
and, moreover, the structure then begins to coarsen rapidly. The
microscope x shows little or none of the original structure when the
steel has been quenched from about 1500° to 1525° F.

With carbons greater than 0.18 or 0.20 per cent., and particu-
larly if the section is small, or the manganese content is more than
0.60 per cent., the necessity of reheating or toughening after quench-
ing becomes apparent. Hardening small sections, such as are used
in automobile construction, from about 1525° F. without subse-
quent drawing — especially if water has been used as the cooling
medium — will produce an inherently brittle steel. The physical
characteristics under these conditions will be approximately as
follows :

Tensile strength, Ibs. per sq. in 90,000 to 110,000

Elastic limit, Ibs. per sq. in 60,000 to 75,000

Elongation, per cent, in 2 ins 17 to 12

Reduction of area, per cent 30 to 15

By reheating to 800° or 900° F. a considerable increase in tougn-
ness and ductility is obtained, approximating:

Tensile strength, Ibs. per sq. in. . . .

Elastic limit, Ibs. per sq. in .

Elongation, per cent, in 2 ins

Reduction of area, per cent

70,000 to 85,000
45,000 to 60,000
35 to 20
65 to 45

Cold-rolled material, subsequently given the same heat treatment
as hot-rolled material of the same chemical composition, will
usually show about 8000 to 10,000 Ibs. per square inch higher in
elastic limit and tensile strength.

Characteristic results from commercial work are given in the
following table:

1 See also page 44.

234

STEEL AND ITS HEAT TREATMENT

Material.

Carbon.

be
c
3
$

Quenched
in Oil from
°F.

Re-
heated
to °F.

Tensile
Strength.
Lbs. per
Sq. In.

Elastic
Limit.
Lbs. per
Sq. In.

Elon-
gation.
Per
Cent
in 2 In.

Reduc-
tion of
Area.
Per
Cent.

General char- ..> . . .
acteristics

0.18
to
0.25

0.40
to
0.80

1500
to
1550

800
to
900

70,000
to
85,000

45,000
to
60,000

35
to

65
to
45

Auto, lever

0.18

0.40

1650

800

70,030

45,400

Pressed auto, frame.

0.22

0.40

1530

800

71,950

43,400

Engine forging

0 26

0.28

1650

1025

77,210

52,200

Old rolled Y± in. plate

0.24

0.60

1525

900

93,300

65,250

20.5

The above remarks apply mainly to the smaller sections up
to 2 ins. in thickness, but are nevertheless applicable in part to
heavy work. With the increase in sectional area, the effect of hard-
ening decreases, and for particularly heavy work may result only
in a refinement of grain. Thus, for heavy, oil-treated forgings,
toughening may not be considered a necessity; such reheating will,
however, relieve the strains which are always inherent to quenched
steels. Large forgings thus treated will show an elastic limit of
30,000 to 50,000 Ibs. per square inch, with an elongation of 35 to
25 per cent, in 2 ins.

Annealed. — There is probably more disagreement and argument
as to the proper annealing temperatures for this range of carbon
steel than for any other. Opinion and practice are divided over the
use of a comparatively high temperature — 50° to 100° over the
upper critical range — or a lower temperature laying somewhere
between the Acl and Ac3 ranges. In this group the Acl and Ac3
ranges are widely separated and the influence of the carbon-mangan-
ese content is rapidly increasing. The high annealing temperature,
1550° to 1600° F. or more, will give ample opportunity for the
absorption of the excess ferrite, for diffusion and for equalization.
On the other hand, there is according to some authorities a marked
increase in grain size from 1350° or 1375° F. and upwards.

The whole question really depends upon the condition of the steel
before annealing. If the " breaking-down " during elaboration —
either rolling or forging — has been severe, if high temperatures have
been used, and if the finishing temperature has not been just right,
a high annealing temperature may be necessary to entirely relieve
the strains and equalize the steel. On the other hand, if the steel

CARBON STEELS

235

has been carefully worked and the micrographic structure is fairly
good, the lower temperatures will probably be entirely satisfactory.
Much must be left to the operator and his own particular problem.
The main point to bear in mind is that the lowest temperature
should be used which will produce the desired results.

If we assume as average figures for annealed steel of this cer-
bon range:

Tensile strength, Ibs. per sq. in 58,000 to 65,000

Elastic limit, Ibs. per sq. in 28,000 to 35,000

Elongation in 2 ins., per cent over 30

and compare these with the results of a tensile test taken from
the steel to be annealed, a very good idea of the degree and length
of heating may be obtained. For example, the following results
from If -in. rounds for gun barrels show that a high annealing tem-
perature was not necessary in this case, inasmuch as the original
steel was in excellent condition.

Gun barrel steel, If -in. rounds.
Carbon, 0.18 per cent.
Manganese, 0.50 per cent.
Phosphorus, 0.070 per cent.
Sulphur, 0.055 per cent.
Silicon, 0.055 per cent.

Treatment.

Tensile
Strength.
Lbs. per Sq. In.

Elastic
Limit.
Lbs. per Sq. In.

Elongation.
Per Cent.
In 3 Ins.

Reduction
of Area.
Per Cent.

As Rolled

66,750

33,820

33.3

57.6

Annealed at

degrees F. for minutes

1360-1400

64,960

34,050

38.0

61.0

1500

65,180

32,930

38.3

58.3

1500

105

64,060

33,150

39.1

62.3

1830

62,940

31,810

35.7

56.3

2120

61,150

31,580

33.8

53.1

On the other hand, the following cold-rolled automobile-frame steel
was particularly " hard " before annealing and required a tempera-
ture of 1550° F. to relieve thoroughly the effect of the cold work:

Carbon, 0.24 per cent.

Manganese, 0.38 per cent.

Phosphorus. 0.028 per cent. •

Sulphur, 0.038 per cent.

236

STEEL AND ITS HEAT TREATMENT

Tensile
Strength.
Lbs. per Sq. In.

Elastic
Limit.
Lbs. per Sq. In.

Elongation.
Per Cent,
in 2 Ins.

Before annealing

100 400

68,500

18.6

After annealing at 1550° F

66,000

38,100

37.0

For the average run of annealing work for this range of carbon,
a temperature of about 1500° F. will be found to give satisfactory
results; individual cases must be treated as such.

0.25-0.35 CARBON STEEL

Steel containing from 0.25 to 0.35 per cent, carbon is known
as soft-forging steel and is used principally for structural ptirposes
in infinite variety. It responds in a most satisfactory manner to
welding, forging and machining, and may be vastly improved by
proper heat treatment. Under skillful treatment, the variety of
combinations of -strength and ductility are to be had in probably
no other range of carbons.

Relative to static strength, some really wonderful results— for
straight carbon steels — in the way of high tensile strength with high
ductility have been obtained from heat-treated (oil quenched and
toughened) forgings of 0.30 to 0.35 per cent, carbon. The follow-
ing results, obtained from the center of a 5-in. electric car, heat-
treated axle, the #xle being selected at random from a group of
about one hundred forgings, give an idea of the extent to which
proper heat treatment may develop the physical properties;

Electric Car Axle, 0.32 Carbon, Acid Steel

Tensile strength, Ibs. per sq. in 91,700

Elastic limit, Ibs. per sq. in 61,620

Elongation, per cent, in 2 ins 33.5

Reduction of area, per cent 48.1

In the hardened condition — without subsequent tempering —
these steels may be used for gears. In the toughened condition
these steels present the maximum resistance to fatigue and other
dynamic stresses, as represented by alternating impact and other
tests, over any of the straight carbon steels; the dynamic strength
probably apexes at about 0.30 per cent, carbon, as far as the author
can judge from his own researches and from the work and conclusions
of others.

CARBON STEELS 237

Untreated. — In the untreated condition, with standard man-
ganese, phosphorus and sulphur, the average tensile strength of
these steels will be about as follows:

Carbon. Acid Steel. Basic Steel.

0.25 to 0.30 67,000 to 78,000 63,000 to 72,000

0.30 to 0.35 69,000 to 83,000 65,000 to 74,000

Rolled plates, from 2 to 4 ins. thick, made of basic steel with 0.25
to 0.35 per cent, carbon and about 0.40 per cent, manganese, will
usually fulfill the following specifications :

[Tensile strength, Ibs. per sq. in 65,000 to 75,000

Elastic limit, Ibs. per sq. in 33,000 to 37,000

Elongation, per cent, in 2 ins 30 to 25

Reduction of area, per cent 50 to 36

These results may also be considered as generally applicable to
untreated steel of this analysis, but which has had more or less
elaboration or working.

Heat Treated. — The upper critical range decreases from about
1500° F. for 0.25 per cent, carbon, to about 1425° F. for the 0.35
per cent, carbon steel. Practical experience has shown that a
quenching temperature of 1500° to 1525° F. for the lower carbons of
this range, and 1450° to 1500° F. for the higher carbons will give
satisfactory results under ordinary conditions. If the heating has
been conducted uniformly and not too rapidly — especially when
approaching the maximum temperature — the original structure of
the steel should be entirely eliminated, as the temperatures recom-
mended are distinctly above the upper critical range. Never-
theless, some metallurgists prefer to quench these steels from a
higher temperature, say 1575° to 1600° F., in order to make cer-
tain of the complete change in structure and to obtain a maximum
hardening effect. In either case, intelligent furnace operation and
heat control will probably be the governing factor rather than the
indicated furnace temperature or mere theorizing.

lor forgings in which especially high qualities are desired, double
quenching will produce a refinement of grain and correspondingly
higher elastic limit and ductility than are usually obtained by the
single treatment. The temperatures recommended for this range
of carbons are :

1. Jirst quenching 'from 1600° F., or from 1500° to 1550° F.
if the higher quenching should prove too drastic.

238 STEEL AND ITS HEAT TREATMENT

2. Second quenching from 1425° to 1450° F., followed by

3. Suitable toughening according to the size of piece and

physical properties desired.

The results to be obtained from heat treatment will vary largely
for this range of carbon in particular, due to such influence as the
increase of a few points in the carbon content (particularly noticeable
in these mild steels), the size of the section, the quenching medium,
and so forth. The results given under the 0.15 to 0.25 carbon range,
and under the 0.35 to 0.45 carbon range to follow, may be used as a
general measure of the carbons under discussion. Stated roughly,
these carbons will give elastic limits ranging from 35,000 to 80,000

FIG. 146. — 0.28 per cent. Carbon Steel. X39. (Campbell.)

Ibs. per square inch, with corresponding elongations of 30 to 10 per
cent, in 2 ins.

Annealed. — As has been previously explained, heating for anneal-
ing to just above the Acl (lower) critical range will refine the ground-
mass only, while complete refinement is shown by the disappearance
of the ferrite and network beyond the upper critical range (Ac3).
As an example of this, examine the photomicrographs of a basic
open-hearth steel containing 0.28 per cent, carbon and 0.52 per cent,
manganese, as shown in Figs. 146, 147 and 148. The first photo-
graph shows the original steel with its coarse, weak structure. Fig.
147 shows the same steel annealed at 1425° F., or just over the Acl
range; the pearlitic ground-mass has been entirely refined, but there
still remains the unabsorbed and undiffused excess ferrite. Fig.

CARBON STEELS

239

148 shows the same still heated to 1520° F. and slow cooled in the
same manner; but in this case the structure has been entirely changed
and refined by heating to a temperature over the upper critical range.

. .

FIG. 147.— 0.28 per cent. Carbon Steel Annealed at 1425° F.
X39. (Campbell.)

FIG. 148. — 0.28 per cent. Carbon Steel Annealed at 1520° F.
X39. (Campbell.)

Practical experience has shown that a temperature of 1500° to
1525° F. will give excellent results for the full annealing of steels
within this range of carbons. On account of the hardening effect
of air cooling steeis with over 0.20 per cent, carbon when in small

240 STEEL AND ITS HEAT TREATMENT

sections, these steels should be slow cooled, either in the furnace,
in lime or in ashes.

In regard to the physical properties to be obtained from the
annealing of these steels, the lower carbons of this range should
always meet tthe U. S. Government specification of:

Tensile strength, Ibs. per sq. in 60,000

Elastic limit, Ibs. per sq. in 30,000

Elongation, per cent, in 2 ins 30

while the higher carbons will usually give :

Elastic limit, Ibs. per sq. in 35,000 to 45,000

Elongation, per cent, in 2 ins 22 to 32

Reduction of area, per cent 30 to 60

0.35-0.45 CARBON STEEL

Straight carbon steels with 0.35 to 0.45 per cent, carbon are
particularly suited to medium and heavy forgings for which the
lower carbons would not give sufficient strength, and for which it
is also not desirable to use water quenching on account of the possi-
bility of starting incipient cracks or strains. This steel is commonly
used for high-duty and moving machine parts; for axles, side bars,
crankpins and other locomotive forgings ; for guns and gun forgings ;
for crank shafts, driving shafts and similar automobile parts; and
for general structural purposes requiring the combination of maxi-
mum strength with minimum brittleness. It has excellent dynamic
strength, although probably not quite so much as the previous
class of 0.25-0.35 carbon. Steel with 0.40 carbon according to
Robin x presents the greatest resistance to abrasive action (wear) .
These steels are easy to machine when in the annealed or soft-
toughened condition, but should not be used for screw machine stock.

The upper critical range temperature of this steel is about 1425° F.
to 1400° F.

Untreated. — The average untreated American open-hearth steel
with standard manganese, phosphorus and sulphur will average
about as follows in tensile strength:

Carbon. Acid Steel. Basic Steel.

0.35 to 0.40 78,000 to 92,000 70,000 to 78,000

0.40 to 0.45 87,000 to 100,000 76,000 to 89,000

1 J. Robin, Inst. Journ., II, 1910.

CARBON STEELS

241

Annealed.

Remarks.

Mn.

Phos.

Sul.

Tensile
Strength.
Lbs. per
Sq. In.

Elastic
Lir it.
Lbs. per
Sq. In.

Elong-
ation.

%in
2 Ins.

Red. of
Area.
Per

Cent.

General limits. . . .

0.35
to
0.45

not
over
0.70

under
0.045

70,000
to
85,000

38,000
to
50,000

28
to
20

to
40

Forged gun jacket
acid steel

0.35

0.25

0.038

0.019

77,080

39,500

Forged gun jacket
basic steel

0.43

0.22

tr.

0.023

78,180

43,100

25.5

8-in. axle acid
steel annealed
at 1400° F

0.42

0.51

78,420

47,460

54.5

Heat-treated. — Large sections, when quenched in good mineral oil
from 1400° to 1500° F., and toughened at 900° to 1200° F. (according
to the carbon content and largest section), should always meet the
specification of 85,000—50,000—22—45. The following tests taken
from large forgings show the variety of combinations of strength
and ductility which may be obtained:

Forging.

Carbon.

Treatment.

Tensile
Strength.
Lbs. per
Sq. In.

Elastic
Limit.
Lbs. per
Sq. In.

Elonga-
tino.
Per Cent,
in 2 Ins.

Red. of
Area.
Per Cent.

Gun jacket
Axle

0.35
0.41

1500-0/1200
1450-w/lOOO

109,560
90,250

65,090
54,575

16.5
25 4

52 4

Gun jacket
Shaft

0.43
0.42

1500-0/1200
1525- 0/1300

111,100
82,040

69,700
57,060

17.0
29.0

55.0

o = oil. w = water.

It is always advisable to keep the drawing temperature as near
1200° to 1250° F. as possible, not only because it is easier for the
furnace operator to obtain more accurate temperature control at
these more readily distinguished " reds," but also on account of the
greater dynamic strength which is obtained by the use of the higher
drawing temperatures.

The results obtained from the water quenching from 1450° F.
and subsequent toughening of small rounds of 0.40 per cent, carbon
steel are given in the chart in Fig. 149.

242

STEEL AND ITS HEAT TREATMENT

§ ""g*"* OOS0^1D1 ^

(g) aequinx saonp-rcH nauug

1 1 1 1 1

Annealed

••

I
i
§

i
l

Chemical Analysis: Critical Ranges: Size of Section Treatment:
C. 0.40 Acl 1330° 1 inch round Quenched in
Mn. 0.60 Ac2 water from
P. 0.02 Ac3 1410° 1450,° and
S. 0.03 drawn as
-. given.

FIG. 149. — Normal Characteristics of 0.40 Carbon Steel, Heat Treated.

ffl

/! 1

«i

]!:ml.md

1 i 1 1 1

1 1 3 ' 1 §

•tpni 9Jimbg aod spunoj '(z) iitniq o^stJig PUB '(x) m3naj?s ^Ijanax

"= 0 o

^ § s «' s s

CARBON STEELS 243

LOCOMOTIVE AXLES

Locomotive axles and other heavy forgings used in locomotive
construction are illustrative of the treatment of pieces of large
section of the above carbon range.

Heat-treated Axles. — For heat-treated axles the carbon content
will range between 0.35 and 0.50 per cent., and with such steel
the treatment may generally be adjusted to meet the standard
specification of:

Tensile strength 85,000 Ibs. per sq. in.

Elastic limit 50,000 Ibs. per sq. in.

Elongation 22 per cent, in 2 ins.

Reduction of area. . . . • 45 per cent.

The quenching of the axles, usually from temperatures of 1400°
to 1500° F., is mainly a proposition of correct heat application and
efficient handling. Both oil and water are used extensively for
hardening axles. Water will bring out the full effect of heat treat-
ment by giving the highest tensile test properties of which the steel
is capable; with the same ductility, oil quenching will give lower
tensile values than water quenching. A steel with lower carbon
content may more logically be used with water quenching than with
oil. Water requires no expensive cooling nor circulation system,
and has practically no cost of upkeep or replenishment, as all
these may be regulated by the intake of fresh, cold water. On
the other hand, many engineers severely condemn the use of water
in that it is too harsh in its action upon large masses of steel — as in
axles, that cracks are more liable to develop, and that internal
strains are set up which often are not always entirely relieved by
the reheating or toughening. Oil is much the safer quenching
medium to use for axles, will give more uniform results, and should
only be replaced by water for economic reasons.

The hardened axles are charged while still warm into the re-
heating furnace, which is maintained at such temperature as will
relieve all strains set up in the hardening and at the same time
give the physical properties desired. This temperature will vary
between 900° and 1200° F., depending upon the chemical composi-
tion of the steel. The higher the drawing temperature, the more
ductile the steel and apparent coarseness of the grain, due to the
transformation of the transition constituents troostite and sor-
bite into pearlite plus free ferrite. Straight carbon steels quenched
in water and drawn at 1000° or 1100° F. will be entirely sorbitic, but

244 STEEL AND ITS HEAT TREATMENT

at 1200° may show considerable free ferrite. Too much emphasis
cannot be given to the necessity of keeping a uniform temperature
and allowing sufficient time for the heat thoroughly to penetrate the
axle. It is then preferable for the axles to cool with the furnace,
rather than to remove them while still at the toughening temperature.
The temperatures used by one manufacturer of acid steel axles
and other large forgings to meet the standard A. S. T. M. specifica-
tions are as follows: after quenching in oil from 1450° F., reheat as
below and air-cool or cool in ashes.

0.42 to 0.45 per cent, carbon 1175° F.

0.38 to 0.42 " " " 1125° F.

0.33 to 0.38 " " " 1075° F.

0.28 to 0.33 " " " 1000° F.

The following are characteristic tests from Open-hearth steel
locomotive axles (Penna. R.R.) :

0.41 Carbon

•t?nrtrfM* Treated,

orged . j 500o water/ ! 000° F.

Tensile strength 73,627 90,250

Elastic limit 31,505 54,575

Elongation 31.6 25.4

Reduction of area. . 43.6 52.4

0.50 Carbon

Treated,
water/12000 F.

Tensile strength .............. 83,430 90,092

Elastic limit ................. 34,370 53,655

Elongation ................... 22.6 27.1

Reduction of area ............. 30.0 52.8

" Tempered " Axles. — For want of a better name, under this
heading might be included such as are treated by the " Coffin " or
similar processes. The main principle consists in heating the axle as
usual for hardening, then immersing in the quenching bath for a cal-
culated number of seconds, immediately withdrawing, and allowing
the heat from the interior of the axle to " temper " the part which
had been hardened by the short immersion in the oil or water.
This process has been developed to such a nicety that surprisingly
uniform results may be obtained if the test is always taken from the
same relative place, such as half-way between the center and the
outside, The main arguments for the process are that it is simple

CARBON STEELS 245

and that the axle will have a tough (annealed) core, and, at the same
time, a hard wearing surface. Non-uniformity of structure is the
principal argument of those condemning the process, and on account
of the inevitable " human equation " which enters into it, would
seem to be not without justice in many instances.

Annealed Axles. — The axles are heated up slowly and uniformly
to a temperature slightly in excess of the upper critical range, main-
tained at this temperature for sufficient time for the steel to respond
to the heat, and then cooled with the furnace. If the working
and the finishing temperature during forging have been adjusted
so as to give a fine grain to the steel, besides good physical test re-
sults, it will be found that heating to a temperature over the critical
range may not be necessary in many instances. Where it is neces-
sary to anneal large numbers of heavy axles, the hot axles may be
removed quickly to a pit and covered with lime or ashes. Annealing
alone will generally overcome the strains set up in the previous proc-
esses of manufacture, but it does not bring out the higher physical
properties of which the steel is capable. Annealed axles will show
pearlite and free ferrite, the apparent size of the ferrite grains de-
pending upon the rate of cooling and the time thus given for the
ferrite to separate out from the matrix.

In order to obtain the required tensile strength upon anneal-
ing it is necessary to use a steel of higher carbon content than that
used for full heat treatment. A tensile strength of 80,000 Ibs. per
square inch will require a 0.50 carbon steel or higher. A temperature
of 1500° F. is generally recommended (A. S. T. M.) for annealing
0.40 to 0.60 per cent, carbon steel, but since the critical range of
this steel is about 1400° F. or a little under, an annealing temperature
of 1400° to 1450° F. will give a better fracture, together with a better
combination of tensile strength and ductility.

The author has had much better results with the lower temper-
ature, although the time required for the annealing generally is longer.
The following results from acid open-hearth steel will show the
effect of the lower temperature anneal :

Heat 4261

Annealed at 1400°

Carbon, 0.42 per cent
Manganese, 0.51 per cent . . .
Phosphorus, 0.034 per cent . .
Sulphur 0.028 per cent

Tensile strength
Elastic limit
Elongation in 2 ins ...
Reduction of area. . . .

78,420 Ibs. per sq. in.
47,460 Ibs. per sq. in.
28 per cent.
54.6 per cent.

246 STEEL AND ITS HEAT TREATMENT

Failures of Heat-treated Axles. — Aside from piping, segrega-
tion, and other impurities in the steel, improper heat treatment is
the active cause of failure of both heat-treated carbon and alloy
steel axles. Unequal or insufficient heating in either the hardening
or toughening processes will produce unequal stresses, which in
turn will sooner or later result in failures. These failures are
always transverse, and never longitudinal. Water quenching large
sections has a strong tendency to produce cracks, often not appear-
ing on the surface, and which may open up when subjected to the
heavy duty and " pounding " when placed in service. Such defects
may be sometimes discovered by the drop test, but its expense
prevents many railroads from using it for the test of every axle.
Heat-treated axles, when given ample reduction in the forging
operation, carefully and uniformly heated to the proper tempera-
tures, and held at those temperatures for a time sufficient for the
steel to respond throughout, should prove vastly superior to un-
treated or annealed axles.

On the other hand, the engineering departments of many rail-
roads have become considerably alarmed over the frequent failures
of so-called " heat-treated " axles, and many have absolutely refused
to have anything to do with axles which have been oil or water
quenched. This really serious phase of the axle question has led
to the investigation of the possibilities of hardening such forgings
in air or steam. Surprisingly good results have been obtained by
methods based upon this system — and from the technical stand-
point are indeed remarkable, since a considerable toughening heat is
necessary with even 0.40 per cent, carbon steel.

0.45-0.60 CARBON STEEL

Treatment of Large Sections. — As the carbon content is pro-
gressively increased beyond 0.45 per cent., its effect becomes quite
noticeable in the added brittleness of the steel. This is strongly
illustrated by the fact that a general study of heat-treatment prac-
tice will show that there is very little quenching of large sections
when the carbon content exceeds the 0.50 per cent. mark. The
dangers to be encountered, both in the treatment itself and by
possible fracture in service, almost prohibit such treatment of large
sections. Any increase in static strength which can be obtained by
quenching and toughening is most certainly acquired with the
ever-present danger of cracking, or of starting incipient cracks.
For these reasons it is, therefore, apparent that the full heat treat-

CARBON STEELS

247

ment of large sections, even though it may bring out higher physical
characteristics in the steel — as is shown by the subsequent figures
obtained from the treatment of a 0.50 per cent, carbon axle — is
becoming less and less of a factor in steels of these carbons.

0.50 Per Cent. Carbon Axle

Forged.

Quenched in Water
from 1400° F.
Toughened at 1200°

Tensile strength, Ibs. per sq. in
Elastic limit so in

83,430
34370

90,090
53655

Elongation per cent, in 2 ins

22 6

27 1

Reduction of area, per cent

30.0

52.8

Tempering, and Small Sections. — On the other hand, the harden-
ing and tempering (as distinguished from toughening) of the smaller
sections, such as gears, dies, etc., begins to take an important place
in heat-treatment work with these carbons. In such cases the
increased-carbon content brings about an inherently possible wearing
hardness which is developed by hardening and tempering. The
medium and smaller size sections may be satisfactorily hardened
in water with but a small proportion of the danger which would
inevitably result from the water quenching (or even oil quenching)
of larger sections. And, by varying the reheating temperatures,
the following approximate physical results may be obtained:

Elastic limit, Ibs. per sq. in.... . .

Elongation, per cent, in 2 ins. ..
Reduction of area, per cent

. 50,000 to 110,000
20 to 5
50 to 15

Annealing. — The commrecial annealing of steel of say 0.50 to
0.60 per cent, carbon will give a variety of results which in them-
selves have proven a stumbling block for many a heat treater.
This is largely due to the prominent part and effect of different rates
of cooling in relation to the size or mass of the steel. To illus-
trate: 6X6 in. billets of 0.50 to 0.55 carbon which have been heated
to 1400° F. and furnace cooled will, in general, meet the specifica-
tions of

Tensile strength 80,000

Elastic limit 40,000

Elongation 22

Reduction of area . 35

248 STEEL AND ITS HEAT TREATMENT

On the other hand, smaller sections, annealed in the same manner
and in the same furnace, will according to their size, give physical
results varying anywhere between

Elastic limit 45,000 to 60,000

Elongation 20 to 15

Reduction of i area 40 to 30

In other words, the extreme variability in the rate of cooling,
as dependent upon the size of section and mass of the steel, its rela-
tion to the size of the furnace, the degree to which the cooling of
the furnace may be controlled, and numerous other related factors
make the commercial annealing of these steels an individual problem
as far as actual physical results are concerned.

It is therefore always advisable, if specific physical results must
be obtained by annealing (used in the broad interpretation of the
term) , to take first a preliminary test of the steel in the condition as
received. From such results it will then be evident how much the
steel must be " let down," and the proper reheating temperature
may be judged from previous experience or by experiment. Al-
though annealing at a temperature under the critical range will
not change the general structure of a pearlitic steel, it will relieve
the strains and stresses, and thereby improve the steel. But fur-
ther, the previous elaboration, such as rolling or forging, which the
steel has undergone, will, in a majority of cases in actual practice,
leave the steel in more or less of a sorbitic state. Under such
conditions, a reheating — or commercial annealing — will actually
change the physical results, even though the annealing temperature
is under the critical range.

Such commercial annealing or reheating temperatures may vary
from 900° F. and upwards through the upper critical range. In the
author's experience there is little or no change in the physical test
results through the annealing of such steel at temperatures under
900° F. or thereabouts. But from this temperature upwards
the sorbitic constituents will gradually coagulate into the pearlite
and ferrite, with a corresponding lowering of the static strength
and increase in the ductility. Consequently, by regulating the
commercial annealing temperature, the physical results may be
" let down " to the desired limits.

If it is desired to change entirely the structure of the steel and
to obtain the finest grain size possible, with maximum ductility, it

CARBON STEELS 249

will be necessary to anneal the steel at a temperature slightly in
excess of the upper critical range, followed by slow cooling.

The influence of the rate of cooling, as exerted by air cooling,
is manifested in the peculiar statement that the tensile strength of
these hard forging steels may be actually raised by annealing (as
distinguished from quenching). It is a well-known fact that steel
of such carbon content when cooled in air at a more or less rapid
rate through the critical range will take on a noticeable degree of
hardness. The author has found that this fundamental principle
may be applied to great advantage in the treatment of axles — with,
of course, certain modifications — and that it is even necessary to
reheat or toughen in order to lower the tensile strength and obtain
the proper ratio of static strength to ductility. Such a process is
now being developed by a large manufacturer of axles, and will in
all probability have an influence upon the heat treatment of axles
and other forgings of large section.

SHRAPNEL

Shrapnel are illustrative of this range of carbon and of pieces
of medium section.

The current specifications for foreign shrapnel cover a wide range
of physical properties, varying between 80,000 and 140,000 Ibs.
per square inch in tensile strength, with 20 to 8 per cent, elongation.
The chemical composition of the steel used will be approximately
between 0.50 and 0.60 per cent, carbon, although the extreme limits
are 0.35 to 0.8 per cent., 0.4 to 1.0 per cent, manganese, phosphorus,
sulphur and silicon about normal, and with or without the addition
of chrome or nickel. Thus, according to the physical or chemical
specifications worked under, some shrapnel manufacturers have been
able to meet their particular specifications without any treatment
except perhaps cooling the cases in lime after forming; others have
had to anneal, or harden and temper; while still others have add
to carry out all three heating operations.

There is nothing unusual in the heat treatment required. The
proposition in short is merely one of proper heat application in fur-
naces of correct design and construction; and yet one may see
almost any and every kind of a furnace being operated in almost
any and every kind of a way except the right one, with the result
of large rejections.

The latest and a very efficient type of furnace for this work is

250

STEEL AND ITS HEAT TREATMENT

designed on underfired principles, with a continuous and auto-
matic charging and discharging of the shrapnel. The cycle for the
complete hardening and drawing is as follows: one man places
the rough-formed shells on the charging platform of the hardening
furnace, as shown in Fig. 149a; an automatic device takes the
shrapnel into and through the heating chamber at a specified and
predetermined rate; the heated shrapnel are then discharged con-
tinuously from the furnace into an oil-quenching bath, as shown in

FIG. 149a. — Automatic Hardening Furnaces, Charging End, Working on
3-in. Shrapnel.

Fig. 1496, from which they are removed by a conveyor system and
delivered to a table in front of the second or drawing furnace. Here
another man takes them from the table and places them on the
charging table, as shown in Fig. 149c; the shrapnel then follow
the same course as in the first furnace, except that they are
received from the furnace into wheelbarrows or similar means of
taking them away, as shown in Fig. 149d. The actual temperature
of the shrapnel in the case of one plant was 1500° F. for hardening
and 850° F. for the tempering or drawing operation.

CARBON STEELS

251

FIG. 1496. — Hardening Furnaces, Discharge End; Oil Quenching Tanks at the

Left.

FIG. 149c. — Tempering Furnaces — Charging End — Conveyer from Quenching

Tank.

252

STEEL AND ITS HEAT TREATMENT

In connection with the above equipment a comparison between
the old hand method and the new automatic appliances is extremely
interesting. The following are actual tests made in the same plant
on 3.3-in. shrapnel:

Old Practice

Two furnaces, 6 men to each furnace:

One furnace, 6 men, 910 shells in 10 hours
One furnace, 6 men, 1207 " " "
Total, two furnaces, 12 men, 2117 " " "

Average of 176 shells per man per 10 hours.

FIG. 149d. — Tempering Furnaces — Discharge End.

New Practice
Two furnaces, with 3 men total:

One furnace, 2508 shells in 10 hours

One furnace, 2603 " " "

Total, two furnaces, 3 men, 5111 " " " "
Average of 1704 shells per man per 10 hours.

In addition to obtaining some 1000 per cent, increased output per man
with the new underfired, automatic furnace, it has been shown that
in the old method the average rejections were running around 15 to
20 per cent., and were only about 3 per cent, for the new practice.

CARBON STEELS

253

CARBON STEELS WITH OVER 0.60 CARBON

Treatment in General. — The treatment of high-carbon steel
develops into the two propositions of hardening and annealing.
Toughening, as referring to high reheating temperatures subse-
quent to hardening, is but very little used, due to the fact that these
steels are too brittle for ordinary structural purposes. Similarly,
tempering is governed entirely by the degree of hardness required
by the tool and is dependent, not only upon the chemical analysis,
but in a larger measure upon the result of the hardening operation.

Hardening. — The precautions to be adopted in hardening may be
repeated in the following general summary:

(1) Use the lowest temperature which will give the desired

results.

(2) Heat slowly and uniformly.

(3) The higher the carbon content, the greater is the degree of

care which must be used, and, in general, the more
narrow the hardening temperature limits.

The temperatures to be used in hardening are largely governed
by the carbon content, and which, in turn, influences the position
of the critical range. We may sum up these factors as follows:

Carbon Content. Per Cent.

Critical Range. °F.

Hardening Temperature. °F.

0.60

1340-1380

1400-1460

0.70

1340-1375

1400-1450

0.80

1340-1365

1390-1450

0.90

1340-1360

1375-1450

1.00

1340-1360

1375-1450

1.10

1340-1360

1375-1430

1.20

1340-1360

1375-1430

1.30

1340-1360

1375-1420

1.40

1340-1360

1375-1420

In giving the above hardening temperatures we have assumed
that the previous mechanical and heating operations have left the
free cementite (in hyper-eutectoid steels — greater than 0.9 per cent,
carbon) well distributed, or emulsified, throughout the steel. This
will generally be true when the proper finishing temperatures,
either in rolling or in forging, have been used. In such cases,
therefore, it will not be necessary to heat to above the Ac. cm range
in order to emulsify the free cementite, and following with a sub-
sequent quenching from slightly above the principal critical range.

254

STEEL AND ITS HEAT TREATMENT

If, however, the previous heating operations have left the free
cementite in the form of spines or network, it will be mandatory to
use the double-quenching method in order to spheroidalize this
free cementite and thus obtain the maximum wearing and cutting
hardness; for details of such procedure the subject matter in Chap-
ter VII should be studied.

Annealing. — The general subject of annealing hyper-eutectoid
steels has been discussed in Chapter III. A series of physical test
results of experiments carried out by Fdbry 1 upon the annealing of
steels with carbon contents of 0.58 to 1.36 per cent., the size of bar
being 1.18 ins. square, and the selected annealing temperatures
being maintained for three hours, are given in the following tables:

0.58 PER CENT. CARBON STEEL — ANNEALED

Treat-
ment.

Tests.

Hard-
ness.

Microscopic.

Annealed
Deg.
Fahr.

Tensile
Strength,
Lbs. per
sq. In.

Elastic
Limit,
Lbs.
per
sq. In.

Elonga-
tion,
per cent
In 3.15
Ins.

Red. of
Area,
per cent

Brlnell
No.

Structure.

Notes.

1110

99,540

—

15.8

43.4

196

Free ferrlte and
pearllte.

Ferrlte reticulated, meshes
filled with grainy pearl-
lte.

1200

98,420

45,510

17.7

49.0

183

Ferrlte begins to change
Into pearllte.

1290

84,200

39,820

20.7

59.2

174

Smaller ferrlte crys-
tals and pearllte.

Structure essentially dif-
fering from other speci-
mens, because the ferrlte
Is uniformly distributed.

1380

93,860

36,980 18.6

43.4

176

1470

96,860

39,820

19.1

36.8 .

187

Free ferrlte and
pearllte.

Ferrlte forms a network;
pearlite partly grainy,
partly lamellar.

1560

96,710

39,820

17.9

35.6

183

1650

98,130

39,820

18.6

36.8

185

Network of large ferrlte
crystals filled with pre-
dominantly grainy pearl-
lte.

1740

93,860

36,980

16.7

33.6

187

1830

100,700

39.820

13.1

25.2

196

Critical range Ac commences at 1337°, maximum at 1355°.

1 Zs. Fabry, " The Variation in the Mechanical Properties and Structures of a
Few Special Tool Steels Annealed between 600° and 1000° C. " Int. Soc. Tes.
Mat., 1912.

CARBON STEELS
0.81 PER CENT. CARBON STEEL — ANNEALED

255

Treat-
ment.

Tests.

Hard-
ness.

Microscopic.

Annealed
Deg.
Fahr.

Tensile
Strength,
Lbs. per
sq. In.

Elastic
Limit,
Lbs.
per
sq. In.

Elonga-
tion,
per cent
In 3. 15
ins.

Red. of
Area,
per cent

Brlnell

No.

1110

102,950

—

13.1

37.6

212

1200

106,400

42,670

14.0

35.6

207

1290

99,540

39,820

17.8

43.4

187

1380

100,700

31,290

14.6

29.4

183

1470

102,840

31,290

12.5

14.8

196

1560
1650

105,250

36,980

13.1

23.0

187

100,400

31,290

13.1

19.4

203

1740

98,980

31,290

10.2

14.0

207

Critical range Ac commences at 1328°, maximum at 1337C

0.92 PER CENT. CARBON STEEL — ANNEALED

Treat-
ment.

Tests.

Hard-
ness.

M Icroscopic.

Annealed
Deg.
Fahr.

Tensile
Strength,
Lbs. per
sq. In.

Elastic
Limit,
Lbs.
per
sq. In.

Elonga-
tion,
per cent
In 3.15
Ins.

Red. of
Area,
per cent

Brlnell
No.

Structure.

Notes.

1110

122,900

—

12.0

23.0

228

Euctectlc.

Grainy pearllte with larger
grains.

1200

120,600

42,670

13.0

25.2

217

1290

98,420

36,980

11.5

33.6

163

Structure perfectly homo-
geneous and essentially
differing from those of
other specimens.

1380

91,030

34,130

17.8

43.4

174

1470

113,500

31,290

10.5

14.8

212

Grainy pearllte.

1560

112,100

—

9.1

14.0

207

Lamellar pearllte.

1650

112,500

31,290

8.7

14.8

216

1740

105,800

31,290

9.0

11.6

214

1830

123,450

36,980

6.8

9.2

228

Indications of overheated
structure.

Critical range Ac begins at 1346°, maximum at 1355°.

256

STEEL AND ITS HEAT TREATMENT

1.11 PER CENT. CARBON STEEL — ANNEALED

Treat-
ment.

Tes

ts.

Hard-
ness.

Microscopic.

Annealed
Deg.
Fahr.

Tensile
Strength,
Lbs. per
sq. in.

Elastic
Limit,
Lbs.
per
SQ. In.

Elonga-
tion,
per cent
in 3.15
Ins.

Red. of
Area,
per cent

Brinell
No.

1110

12°, 550

—

9.7

20.8

248

1200

126,300

51,200

12.6

23.0

235

1290

108,650

54,040

10.3

23.0

185

1380

88,180

39,820

19.6

49.0

170

1470

91,590

36,980

16.7

36.8

178

1560

96,700

25,600

10.3

18.6

196

1650

105,100

29,580

6.1

10.0

207

1740

100,700

25,600

6.6

9.2

202

1830

116,050

36,980

6.0

6.8

228

Critical range Ac begins at 1337°, maximum at 1346°.

1.36 PER CENT. CARBON STEEL — ANNEALED

Treat-
ment.

Tests.

Hard-
ness.

Microscopic.

Annealed
Deg.
Fahr.

Tensile
Strength,
Lbs. per
sq. in.

Elastic
Limit,
Lbs.
per
sq. in.

Elonga-
tion,
per cent
in 3.15
Ins.

Red. of
Area,
per cent

Brinell
No.

Structure.

Notes.

1110

132,550

—

6.2

11.6

262

Free cementlte and
pearl ite.

Cementite reticulated,
meshes filled partly with
lamellar, partly with
grainy pearllte.

1200

129,700

67,980

8.5

14.0

255

Free cementlte begins to
change Into pear lite.

1290

123,450

48,355

9.6

16.2

288

Fine-grained
cementite.

Structure appears uniform.

1380

93,300

45,510

14.6

37.6

192

As before, grains finer.

1470

90,310

47,500

17.3

36.8

187

Free cementlte and
grainy pearllte.

Cementite concentrated
again into small crystals.

1560

93,300

42,670

13.7

27.4

187

1650

102,100

32,710

4.5

5.0

209

Free cementlte,
grainy and lamel-
lar pearllte.

Cementlte crystals larger,
pearlite partly in lamellae.

1740

95,580

28,446

4.6

6.8

196

Free cementite and
and lamellar
pearllte.

Structure essentially al-
tered. Cementite form;?
a network with large
meshes. Steel Is over-
heated.

1830

101,830

—

2.6

4.4

223

Critical range Ac begins at 1345°, maximum at 1355°.

CHAPTER XI
NICKEL STEELS

Nickel Steel. — Nickel may well be said to have been the pioneer
among the common alloys now used in steel manufacture. Origi-
nally added merely to give increased strength and toughness over
that obtained in the ordinary rolled structural steel, the development
and possibilities of heat treatment have greatly enhanced its value,
so that nickel steel holds a premier position in alloy steel metallurgy.

The chief difficulties attendant upon its use have been its tend-
ency to develop a laminated structure and its liability to, seams.
But when care is used in its manufacture and rolling, and it is not
made in too large heats or ingots, and when piping and segregation
are avoided by confining the finished product to that produced from
the bottom two-thirds of the ingot, an admirable product for many
purposes is obtained.

Nickel steel has remarkably good mechanical qualities when
subjected to suitable heat treatment and is an excellent steel for case
hardening. In machining qualities it usually takes first place
among the alloy steels.

Strength and Ductility. — Nickel primarily influences the strength
and ductility of steel in that the nickel is dissolved directly in the
iron or ferrite, in contradistinction to such elements as chrome and
manganese which unite with, and emphasize the characteristics of
the cement it ic component. Thus, for the forging grades of ordinary
nickel steer in the natural condition, the addition of each 1 per cent,
of nickel up to about 5 per cent, will cause an approximate increase
of 4000 to 6000 Ibs. per square inch in the tensile strength and
elastic limit over that of the corresponding straight carbon steel,
without any decrease in the ductility. This influence of nickel
upon the static strength of steel also increases to some degree with
the percentage of carbon. To illustrate the effect of nickel upon
steel in the natural condition: a steel with 0.25 per cent, of carbon
and 3.5 per cent, nickel will have a tensile strength equivalent to
that of a straight carbon steel with 0.45 per cent, carbon, a propor-

257

258 STEEL AND ITS HEAT TREATMENT

tionately greater elastic limit, and the advantageous ductility of the
lower carbon grade.

Necessity for Heat Treatment. — On the other hand, and in con-
nection with the use of alloy steels in general, it should be borne in
mind that such steels should be used in the heat-treated condition
only — that is, not in either an annealed or natural condition. In
the latter conditions there is a benefit, as compared with straight car-
bon steels and as illustrated above, but often is not commensurate
with the increased cost. In the heat-treated condition, however,
there is a very marked improvement in physical characteristics.
And closely associated with this is the finely divided state of both
ferrite and pear lite which characterizes heat-treated nickel steel.

Nickel vs. the Critical Ranges. — One of the most interesting
phenomena connected with nickel steel is the effect of nickel upon
the position of the critical ranges. Nickel, like carbon, has the
property of lowering the points of the allot ropic transformations of
iron, but in a more marked degree. Just as we have seen, in the
chapter on Hardening, how " rapid cooling " and " carbon " are
" obstructing agents " in preventing the transformation of the
austenite into martensite into pearlite, so likewise does nickel act
as an obstructing agent. The effect of nickel is obtained through
a lowering of the Ar ranges, so that the temperatures of the critical
ranges on cooling may even be brought below atmospheric temper-
atures. Thus we may have a steel which, without quenching, may
be pearlitic, troostitic, martensitic or austenitic, dependent upon
the relative percentages of nickel and carbon. Hence, such steels
containing nickel may be classified according to their microscopic
constituents which are obtained upon slow cooling from a high
temperature.

Classification of Nickel Steels. — In Fig. 150 there is shown
graphically the influence of the nickel-carbon ratio upon the struc-
ture of nickel steels as cast, or as moderately cooled from a high
temperature.

The " Pearlitic " nickel steels are those in which the critical
ranges are all above the ordinary temperatures, so that such steels
as slow cooled from a high temperature will consist of pearlite
plus ferrite (or cementite). These are the ordinary commercial
nickel steels, and are represented by the lower triangle of Fig. 150.

" Martensitic " nickel steels contain that percentage of nickel
and carbon which will so lower the position of the critical ranges
on cooling that only the partial transformation may proceed. That

NICKEL STEELS

259

is, the austenite is transformed into martensite, but no further —
the steel being too rigid to allow a more complete transformation at
the low temperatures involved. These steels correspond to the mid-
dle triangle in Fig. 150. Nickel steels martensitic throughout
have no practical value, as it is impossible to work or machine them.
On the other hand, great importance is attached to the use of cer-
tain pearlitic nickel steels which can become martensitic upon case
carburizing — due to the increased carbon content.

Austenltic

Martensitic

Pearlitic

0.40

0.80

1.20

1.60

FIG. 150. — Influence of the Nickel-carbon Content upon the Structure of Nickel

Steels as Cast.

A still further increase in the nickel or carbon content will cause
the critical range on cooling to fall below atmospheric temper-
atures, so that such steels will be characterized by an " austenitic "
or " polyhedral " structure, and are known under these names.

Micrographic Structure.— These changes in structure are illus-
trated by the series of photomicrographs (by Savoia) given in Figs.
151 to 158, all the steels being in the natural condition, having
approximately the same carbon content (0.25 per cent.), but with
increasing percentages of nickel.

260

STEEL AND ITS HEAT TREATMENT

FIG. 151.— Steel with 0.25 per cent. Carbon, 2 per cent. Nickel. X650.

(Savoia.)

»*^^^w

aft*

^5 *6s —

L*$*1«»S

FK;. 152. — Carbon, 0.25 per cent. Nickel, 5 per cent. X650.

(Savoia.)

NICKEL STEELS

261

FIG. 153.— Carbon, 0.25 per cent. Nickel, 7 per cent. X650.
(Savoia.)

FIG. 154.— Carbon, 0.25 per cent. Nickel, 10 per cent. X650.
(Savoia.)

262

STEEL AND ITS HEAT TREATMENT

FIG. 155. — Carbon, 0,25 per cent. Nickel, 12 per cent. X650.

(Savoia.)

FIG. 156. — Carbon, 0.25 per cent. Nickel, 15 per cent. X650,
(Savoia.)

NICKEL STEELS

263

FIG. 157. — Carbon, 0.25 per cent. Nickel, 20 per cent. X650.

(Savoia.)

FIG. 158.— Carbon, 0.25 per cent. Nickel, 25 per cent. X650.
(Savoia.)

264

STEEL AND ITS HEAT TREATMENT

It will be seen that the structure of the 2 per cent, nickel steel
(Fig. 151) is similar to that of a corresponding straight carbon steel,
but is finer and more homogeneous.

The 5 per cent, nickel steel (Fig. 152) shows a still finer and
denser structure, in that the pearlite is more divided and distributed.

With the 7 per cent, nickel (Fig. 153) the ferrite and pearlite
are still seen, but they are distributed in a special manner as if
disturbed by the approach of a transformation. A tendency to
orientiate, somewhat like martensite, is also noticeable.

Pearlitic

Martensitic

Austenilic

200,000

150,000

100,000

FIG. 159. — Comparative Physical Properties of Nickel Steels with 0.25 per cent.

Carbon,

The steels with 10 and 12 per cent, nickel (Figs. 154 and 155)
are both wholly martensitic.

With the 15 per cent, nickel (Fig. 156) intensely white con-
stituents appear amidst the martensite and probably represent the
first appearance of austenite. The latter increases quite noticeably
in the steel with 20 per cent, nickel (Fig. 157), taking on its poly-
hedral form.

At 25 per cent, nickel (Fig. 158) the whole steel is characterized
by large polyhedra of gamma-iron.

Physical Properties with Increasing Nickel. — The physical
properties of these same cast nickel steels are plotted graphically in

NICKEL STEELS

265

the chart in Fig. 159. It will be noted that the abrupt changes in
the curves correspond very closely with the theoretic structure given
by the upper abscissae ; and that these same physical properties are
indicative of the essential characteristics of pearlite, martensite and

1600

1500

1400

1300

1200

£C. 0.20 0.40 0.60 0.80

FIG. 160. — Critical Changes on Heating 3 per cent. Nickel Steel.

1.0

austenite, as denoted by the tensile strength, ductility (elongation)
and resistance to shock.

Critical Range of Pearlitic Steels. — For nickel steels with less
than 5 to 7 per cent, nickel, each 1 per cent, nickel lowers the crit-
ical range (Acl) on heating by about 15° to 20° F., and also lowers
the Arl range (cooling) by about 30° to 40° F., below those of the

266

STEEL AND ITS HEAT TREATMENT

corresponding ranges for straight carbon steels of the same carbon
and manganese contents. Similarly, there is a corresponding lower-
ing of the other critical ranges.

This is graphically shown in Fig. 160, which the author has care-
fully plotted from a series of observations obtained with 3 per cent,
nickel steels of various carbon contents. It will be seen from this
curve that the critical ranges on heating are about 60° F. below the
corresponding straight carbon steels. With the very low carbons
there appears to be a tendency for the Ac3 curve to flatten out; this
is further substantiated by results with steels containing 5 to 7 per
cent, nickel. Beyond the eutectoid ratio of carbon it was found
that the Ar range would begin to drop quite rapidly (not shown in the
diagram) below its normal value, as might be expected from the
fact that an increase in carbon in these steels act in an analogous
manner to an increase in nickel.

The approximate temperatures of the Acl and Ar ranges for
nickel steels are as follows;

Per Cent. Nickel. Acl, ° F.

Ar, ° F.

1340

1280

1.0

1320

1240

2.0

1300

1200

2.5

1290

1180

3.0

1280

1160

3.5

1270

1140

4.0

1260

1120

4.5

1250

1100

5.0

1240

1080

6.0

1220

1040

7.0

1200

1000

It must be borne in mind, however, that the Acl temperatures may
vary considerably from steel to steel — but those given above will
probably be about the average of those obtained in practice, and
will in any event be within ±25° F. On the other hand, the Arl
temperatures given are liable to an even greater variation, as the
maximum temperature attained in heating, the length of time
occupied in both heating and cooling, the effect of the higher car-
bon contents, and many other experimental factors, all tend to
change the position of the Arl range.

From these figures, and from the critical range diagram given
for 3 per cent, nickel steels, it will be observed that the hardening of

NICKEL STEELS 267

nickel steels may be carried out at temperatures considerably lower
than those required by the corresponding straight carbon steels.

The Eutectoid for Nickel Steels.— The effect of additional
nickel, or at least up to 7 per cent., is to reduce the eutectoid carbon
ratio below that of the 0.9 value for straight carbon steels. That is,
a nickel steel with 3 per cent, nickel will be saturated, having neither
excess ferrite nor excess cementite (on slow cooling), at about 0.75
to 0.8 per cent, carbon; while in 7 per cent, nickel steel the eutec-
toid ratio appears to be about 0.6 per cent, carbon. This fact is of
great importance in case-hardening work, in that it not only permits
of a shorter duration of the carburization in order to obtain the
maximum carbon concentration necessary in the case, but also
reduces the carbon content over which it is likely that enfoliation
may occur.

Heat Treatment of Pearlitic Nickel Steels. — The heat treatment
of pearlitic nickel steels presents some very interesting phenomena
which are quite distinctive from ordinary straight carbon steels.
One would naturally assume that the treatment of pearlitic nickel
steels would be carried out in an analogous manner to that of the
ordinary carbon steels — that is, the quenching should be done at a
temperature slightly in excess of the upper critical range, provided
that the duration of heating at the maximum temperature has been
sufficient to effect the entire solution of the previous components,
together with their diffusion and the equalization of the steel as a
whole. Similarly, as in carbon steels, we would assume that we
might replace the length of heating at the proper quenching temper-
ature by a higher temperature in order more quickly to effect the
equalization of the steel; provided, however, that this new and
more elevated temperature shall not produce too great a deteriora-
tion in the metal through increase in grain-size, etc. — or that this
higher quenching is followed by a quenching at the proper tempera-
ture. In straight carbon steels the change of structure by heating
slightly above the upper critical range takes place quickly as a general
rule; and the coarsening or embrittling of the steel also occurs
rapidly when higher temperatures are used. The influence of
nickel in the steel, however, often necessitates a modification, or
permits a simplification, of these general principles, both in regard
to the temperature of quenching and the length of heating.

In the first place, the addition of nickel appears to make the
solution of the ferrite or cementite and the equalization of the steel
as a whole take place more slowly than in the ordinary carbon

268 STEEL AND ITS HEAT TREATMENT

steels. Thus, if we take a steel containing some 4 or 5 per cent,
nickel, and a mild or medium carbon content, and quench it
after a normal heating at a temperature some 50° F. over the
critical range, the transformation is often incomplete and the
martensite not uniformly distributed nor equalized.

In such an event, which is usually characteristic of nickel steel
which has either undergone a more or less severe elaboration or work-
ing, or has been finished at too low a temperature, or has been sub-
jected to a prolonged heating at some high temperature, there are
then four methods of procedure available:

(1) Prolonged heating at the proper quenching temperature

to effect the necessary transformation, followed by
quenching;

(2) Heating to a higher temperature than in (1), and quench-

ing;

(3) Heating to the higher temperature, cooling to a temper-

perature a little above the Ar temperature, and then
quenching;

(4) Quenching, or air cooling, from the higher temperature,

followed by a normal reheating to a temperature slightly
hi excess of the Ac range, and quenching.

These propositions at once evoke a discussion of further char-
acteristics which the presence of nickel involves.

If an ordinary carbon steel be heated for a considerable duration
of time at a temperature even slightly over the critical range, the
grain-size will begin to increase, with a corresponding decrease in
both the static and dynamic strength of the material. On the
other hand, if a nickel steel be subjected to a length and temperature
of heating equivalent to that of the carbon steel, the pearlite and
ferrite grains will remain (after slow cooling) considerably finer,
more uniformly distributed, and much more subdivided than the
carbon steel. This characteristic permits the greater duration of
heating as required under the first proposition, without any per-
ceptible deterioration such as would be noticeable in a straight car-
bon steel with the same prolonged heating. However, such treat-
ment— when required — is disadvantageous from the commercial
standpoint, as it decreases the capacity of the heat treatment plant,
with a corresponding increase in the cost of production.

Again, the increased brittleness due to a more or less prolonged
heating at temperatures in excess of the upper critical range is con-
siderably less for nickel steels than for ordinary carbon steels. This

NICKEL STEELS 269

fact is well illustrated by the following results upon a straight carbon
steel in comparison with a 2 per cent, nickel steel of the same carbon
content, taken from a memoire l by Guillet:

Length of Heating

Resistance

to Shock.

at 1830° F.

Ordinary extra-soft steel.

Extra-soft steel with
2 per cent nickel.

Normal heating

20 kgms.

60 kgms. (not broken)

Four hours

4 . 5 kgms.

60 kgms. (not broken)

Six hours

4.0 kgms.

60 kgms. (not broken)

If, in order to obtain the full equalization of the steel and also to
avoid a prolonged heating at the lower and theoretic temperature,
it shall be necessary to heat and quench from a higher temperature,
such operation may be undertaken without that fear of greatly
increasing the brittleness which would most probably occur in a
straight carbon steel.

Although it is granted that a heating to this high temperature
may be necessary, a quenching from this same high temperature
would not be entirely logical if this were to be the only quenching,
and also if viewed from the standpoint of the best product. In
such high temperature quenchings there is the ever-present danger
of cracking and warping. Further, it is a generally admitted fact
that no change in the molecular arrangement of the steel occurs
in cooling such a steel until the upper critical range on cooling (Ar3)
is reached. Assuming this to be true, we may then modify the previ-
ous treatment (proposition 2) by first cooling the steel — after heat-
ing to the high temperature — to a temperature slightly above that
of the Ar3 range, and then quench, as stated under proposition 3.
This treatment will retain all the benefits which may accrue from
the original high-temperature heating, but at the same time will
diminish to a considerable degree the dangers of cracking and
warping. And as the critical ranges on cooling in nickel steels are
even further below those of the Ac ranges in comparison with
straight carbon steels, this quenching temperature will be reason-
ably low.

Objections which may be offered to this method are that the
quenching from just over the Ar range may not give the maximum

*M. L. Guillet, " Traitements thermiques des aciers speciaux," Rev. de
Met., July, 1910.

270 STEEL AND ITS HEAT TREATMENT

hardening effect unless the quenching temperature has been gauged
just rightly, or if the carbon content is low. The first objection may
be overcome by suitable temperature control; if the quenching
temperature should fall too low, the difference in the hardening
effect, for forgings or full-heat treated work, may be later corrected
by using a lower toughening or drawing temperature. By the use
of exact methods, such as one furnace maintained at the high tem-
perature, and then another furnace (into which the steel may be
subsequently placed) maintained at the temperature a little over the
Ar range, the first objection may be entirely cleared away. The
second objection may also be at once overruled by the fact that
the treatment of the low-carbon steels is generally limited in com-
mercial work to the obtaining of a suitable toughness and absence
of brittleness (regeneration), and that it is usually not desired to
obtain the maximum hardness.

In brief, it does not matter whether the same mechanical prop-
erties in a pull test are obtained by a quenching made at a very
high temperature, or by a quenching at a lower temperature follow-
ing the return. As these results in the mechanical properties are
practically the same, the treatment under proposition 3, as compared
with Number 2, is always more advantageous from the point of
view of non-brittleness and probably also from the point of view of
the strength of the piece.

The most serious objection to the treatment in either (2) or (3),
however, is the increase in brittleness which is liable to occur if
the high temperature heating is unduly prolonged. Although the
presence of nickel tends to diminish such a condition, the effect
of high heating is always towards the increase in grain-size, and
coarse martensite generally corresponds to a diminution in the
strength of the steel. Assuming that a temperature considerably
in excess of the upper critical range is mandatory, any ill effects
resulting therefrom may be entirely overcome by a double heating
and cooling, and yet also retaining the benefits of such high temper-
ature heating. That is, by cooling the steel — but not quenching,
unless the original structure is very bad indeed; or unless the most
perfect structure is desired — from the high temperature to a tem-
perature under that of the Al range, in order to impress the effect
of the high temperature upon the steel, followed by a reheating
to a temperature slightly in excess of the upper critical range, and
then quenching. Such a hardening treatment, either with air cool-
ing and a subsequent single quenching, or with a double quenching,

NICKEL STEELS 271

is the best, although the most expensive. As this method has been
discussed in its relation to carbon steels, and as its influence is
approximately the same with pearlitic nickel steels, it will not be
necessary to dwell further upon it.

In general, the treatments (for the best quenching effect) given
under (1), (2) and (3) will suffice for ordinary commercial practice,
but with the preference given to (1) or (3). That under (4) is best
if the higher cost is permissible.

Moreover, it should be borne in mind that, in perhaps even a
majority of cases, the regular and normal quenching from a tem-
perature slightly in excess of the upper critical range (Ac3), after
a thorough and uniform heating at that temperature, will generally
suffice — and especially for small work. But in order more fully to
explain the difficulties which are sometimes met with in the treat-
ment of nickel steels, the author has entered into the foregoing
explanations. As a safe and general fundamental principle, re-
peatedly urged, it is always advisable to quench from the lowest
temperature which will give the desired results.

The tempering and toughening of pearlitic nickel steels is carried
out exactly as with straight carbon steels, previously explained, and
is dealt with in more detail later on in the chapter.

CARBURIZATION OF NICKEL STEELS

The general principles of the carburization of nickel steels are
similar to those which apply to straight carbon steels, and should
not require repetition. There are, however, certain peculiarities,
presenting both advantages and disadvantages, which should be
mentioned.

(1) Nickel steels show less susceptibility to brittleness due to
prolonged heating at the high temperatures often used in carburiza-
tion than do the corresponding carbon steels. This important fact
not only gives a steel better able to withstand shock, but also gives
a well-defined means of simplifying the subsequent heat treatment
if desired. Such advantages may be readily obtained by the addi-
tion of even 2 per cent, of nickel, and largely compensate for the
slightly higher cost.

(2) The variations in the concentration of the carbon in the
carburized zones are more gradual and uniform in nickel steels than
in carbon steels. This better distribution of the carbon therefore
tends towards the prevention of a distinct line of demarkation be-
tween the different zones, and thus to eliminate the chipping and

272

STEEL AND ITS HEAT TREATMENT

flaking off of the case. Similarly, the phenomenon of " liquation, "-
a principal factor in such enfoliation — is less marked, under equal
conditions, in nickel steels than in carbon steels.

(3) Although it is true that carburization proceeds with greater
slowness with some solid carburizing compounds, referring to their
use with nickel steels with less than say 3.5 per cent, nickel, the use
of a mixed cement (carbon monoxide plus carbon) will effect a car-
burization with a rapidity equal to that with ordinary carbon steels.

(4) Under the same conditions, the depth of penetration of the
carburized zone for a given time, using a mixed cement, is even
slightly higher for nickel steels than for carbon steels.

(5) The lesser hardness which, with the same treatment, is
possessed by the carburized zones in nickel steel as compared with
the carburized zones in carbon steels under identical conditions, is
due not only to the different effects of different quenchings, but
also to the smaller concentration (especially for less than 3 per cent,
nickel) of carbon in the carburized zones. This disadvantage may
be eliminated by raising the carbon in the carburized zone by a
suitable change in the conduct of the carburization.

(6) When the nickel content exceeds 3 per cent, the maximum
concentration of the carbon in the carburized zones decreases with
an increase in the percentage of nickel contained in the steel. The
following table, from Giolitti,1 contains data relative to the maximum
concentration reached by the carbon in the carburized zones when
carburizing, under various conditions, steels with varying nickel
contents :

Condition of Carburization.

Nickel Content.

25%

30%

Carbon monoxide:
5 hours at 1740° F

0.38
0.35

1.53

0.23
0.35

0.93

1.28

0.70
0.80
0.73
0.74
0.83

0.15,
0.17

0.39
0.63

0.67
0.40

5 hours at 1920° F
Ethylene:
5 hours at 1740° F

1.12

5 hours at 1920° F

0.84

0.64
0.59

0^73

Mixed cement:
2 hours at 1830° F

0.70
1.12
0.83
0.92
1.07

5 hours at 1830° F.

5 hours at 1920° F

2 hours at 2010° F
5 hours at 2010° F

1 F. Giolitti, " The Cementation of Iron and Steel.'

NICKEL STEELS

273

(7) By employing a nickel steel of the proper nickel content, and
carburizing in such a manner as to attain a definite maximum
carbon concentration in the case, a steel characterized by a tough
core and a martensitic structure in the case may be obtained with-
out subsequent quenching. The approximate maximum carbon con-
centration in the case which it is necessary to obtain for steels with
definite percentages of nickel in order to produce a martensitic struc-
ture without quenching, may be given about as follows:

Per cent. Nickel.

Per cent. Carbon.

Per cent. Nickel.

Per cent. Carbon.

1.50

0.95

1.30

0.85

1.10

0.75

Such methods eliminate the necessity for subsequent heat treat-
ment, if so desired, and effect corresponding reductions in the cost
of the process, besides obviating, in a large measure, such important
factors as warping, grinding, etc. Further, by extending the car-
burization so as to reach a maximum of 1.5 per cent, carbon at the
periphery, for steels containing 5 to 7 per cent, nickel, there can
also be obtained a superficial layer, superimposed upon the mar-
tensitic zone, containing austenite, which easily admits of polishing
without loss.

(8) The lower temperatures at which the critical ranges are
located, in the pearlitic nickel steels, permit a lower temperature
to be used in case carburizing, which is an important factor in
intricate or exact work.

THERMAL TREATMENT AFTER CARBURIZATION

In general, the thermal treatment of nickel steels, subsequent
to case carburizing, may be classified according to the structure of
the case after slow cooling from the temperature of carburization —
that is, whether it is pearlitic or martensitic. Although this struc-
ture depends primarily upon the conduct of the carburization and
the maximum carbon concentration thus obtained in the case, the
procedure as practically carried out in commercial work will usually
give (upon slow cooling after carburization) (1) a pearlitic structure
in the case for steels with nickel contents under 4 per cent, and (2) a
more or less martensitic case for steels with 4 to 7 per cent, nickel.

274

STEEL AND ITS HEAT TREATMENT

As explained in Chapter VII, the best treatment which can be
given any case-carburized pearlitic steel is that involving a double
quenching. Each quenching — for regeneration and for hardening —
should be carried out at the most suitable temperature, and which is
fixed by the transformation points of the core and case respectively.
These treatments for nickel steels with 2 to 2J and 3 to 3J per cent,
nickel are approximately as follows:

Carburization. — Carburize at the desired temperature, usually
1600° to 1750° F. Cool slowly in the carburizing material
(assuming solid cements).

Thermal Treatment. —

Nickel Content, Per cent.

2 to '*

l|.

3 t(

) 3*.

Carbon Content, Per cent.

0.10 to 0.15

0.15 to 0.20

0.10 to 0.15

0.15 to 0.20

Regenerative
quenching

(a) 1550-1600°

1500-1550°

1475-1525°

1450-1500°

Hardening
quenching

or (6) 1600°
1325-1375°

1600°
1325-1375°

1600°
1300-1350°

The steel may be removed from the quenching bath as soon as
it loses its red color during the regenerative quenching, and imme-
diately reheated for the second quenching. Practice differs as to
the temperature to be used for the regenerative quenching, some
preferring to quench from a temperature slightly above the Ac3
range, as under (a), while others prefer to use the higher temper-
ature (6). The reasons for these have been discussed in previous
sections.

On the other hand, for 2 per cent, nickel steels, Guillet recom-
mends temperatures distinctly higher than those given above —
which probably coincide with the best American practice — for the
regenerative quenching, and which he gives as follows:

Regenerative quenching 1760° to 1800°

Hardening quenching 1365° to 1420°

The hardening quenching should be conducted at the lowest pos-
sible temperature at which the metal of the case will become glass-
hard. In many instances it will be found that temperatures some-
what lower than those given in the above table can be used. For

NICKEL STEELS 275

example, the critical curve given in Fig. 161 for a steel with 0.13 per
cent, carbon, 0.49 per cent, manganese and 3.35 per cent, nickel,
shows the Acl range to be about 1250° F., so that a temperature
under 1300° to 1350° might be used effectively for the hardening
quenching.

The effect of different treatments upon Quillet's 2 per cent,
nickel steel in its resistance to shock is shown in the following table:

Treatment. Resistance to Shock.

Steel with 2 per cent, nickel and 0.1 per cent, carbon: •

Heated to 1700° F. and air cooled 33.4 kgms.

Quenched in water from 1700° F 34 . 5

Same steel cemented at 1830° F. for 1.2 mm.:

Slow cooled 31 .0

Quenched in water from 1830° F 33.5

Double-quenched in water, 1830° and 1380° 36.0

Quenched in water from 1380° F , , , , , , 32, 0

FIG. 161, — Critical Range Diagram.

Simplified Thermal Treatments after Carburization. — On account
of the fact that the brittleness of the core (with nickel steels) is not
greatly increased by the heating during carburization if the tempera-
ture of that operation is not too high, and as the Ac3 range of the
ordinary nickel steels is considerably lower than that of the corre-
sponding straight carbon steel, it makes it possible, as we have seen,
to effect a regenerative quenching at a temperature in the neighbor-
hood of 1500° -1550° F, Further, as the nickel steel case can be

276 STEEL AND ITS HEAT TREATMENT

hardened at a temperature considerably above the normal Acl
without losing too much of its hardness or increasing too largely in
brittleness, it follows that, in many instances, the regenerative
quenching may also serve as a hardening quenching. This permits
the simplification of the treatment to a single quenching for nickel
steels, unless the piece is to be subjected to exceptional stress. The
practical usefulness of this method is obvious.

It is evident, however, that the double quenching will always
give considerably better results for both core and case. This is
particularly shown in the depth and degree of hardness obtained by
the lower quenching over the higher quenching temperature by the
following experiments by Guillet on 2 per cent, nickel steels:

Shore Hardness Numbers.

Treatment.

Maximum.

Minimum.

Mean.

Cemented pieces, not quenched

39.37

Cemented pieces, quenched from 1830° F.

78.05

Cemented pieces, quenched from 1380° F.

86.56

Case Hardening by Air Cooling. — Again, the case-hardening
process may be even further simplified by the use of nickel steels with
3.5 per cent, of nickel, or more, and conducting the carburization in
such a manner as will produce a maximum carbon content in the case,
which will give a martensitic structure on air cooling from the tem-
perature of carburization. An example of this is shown in Figs. 162
and 163, representing a case-carburized steel with an initial carbon
content of 0.176 per cent., with 3.44 per cent, nickel; the steel was
then air cooled directly after carburization. The thickness of the
martensitic zone is about 0.5 mm. Under the lower magnification
(Fig. 163) a solid troostitic band is seen to separate the martensite
and the sorbito-pearlite portions. The principal advantage which
this method presents consists of its great simplicity, and also in the
fact that it permits the avoidance of deformation which so often
accompanies any quenching operation. Nickel steels which are
martensitic after air cooling may be troostitic, sorbitic, or even pearl-
itic after very slow cooling in the furnace, while they may be austen-
itic on water quenching.

Case Hardening 5 to 7 Per Cent. Nickel Steels.— Advancing
another step and using nickel steels with 5 to 7 per cent, nickel, we
find that the ordinary carburization and subsequent slow cooling

NICKEL STEELS

277

FIG. 162.— Nickel Steel. Nickel, 3.44 per cent. Carbon, 0.176 per cent.
Case Carburized and Air Cooled. X 100. (Sauveur and Reinhardt.)

FIG. 163. — Same Steel and Treatment as in Fig. 162.
(Sauveur and Reinhardt.)

X50.

278

STEEL AND ITS HEAT TREATMENT

will produce a case with characteristics varying over a wide range,
dependent upon the nickel-carbon ratio in the case. In other words,
the transformation range of the metal of the case on cooling may be
even further reduced below that of the previous example, giving
either a martensitic or martenso-austenitic structure upon slow cool-
ing. Therefore, when it is not required to produce an extremely
tough core, nor to obtain extreme hardness in the case, the carbur-
ized pieces with 5 to 7 per cent, nickel may simply be allowed to
cool slowly in the carburizing mixture after carburization.

The use of the method just indicated, however, has its dis-
advantages. The following table shows the results of scleroscope
hardness tests made by Guillet on a steel containing 7 per cent,
nickel and 0.12 per cent, carbon, carburized to a depth of 0.1 mm.,
but not quenched:

Treatment and Tests.

Shore Hardness Numbers.

Mean.

Maximum.

Minimum.

Test made on the surface.

18.5
26.5
24.5
20.2

21
27
25
20

17
26
24
21

Test made at a depth of 0.2 mm

Test made at a depth of 0.4 mm .......
Test made at a depth of 0.6 mm

From this particular instance it will be seen that the surface zone
is partly austenitic, so that a very great hardness is not obtained.

In the second place, it is evident that the core of the piece thus
treated has not been regenerated, although, as we have said before,
nickel steel does not have the maximum brittleness which a straight
carbon steel would have under the same conditions of cooling.

The structure of a steel containing 4.86 per cent, nickel and 0.115
per cent, carbon, intensely carburized, and air cooled, is shown in
Fig. 164. This photomicrograph shows that the effect of such
treatment is to produce a case which is largely austenitic.

The best practice, however, both American and foreign, specifies
the use of a double-quench treatment subsequent to a mild carburiza-
tion, and using a soft steel with 4.5 to 6 per cent, nickel. Such steels
have many peculiar advantages: the carburization may be con-
ducted at a moderate temperature; a maximum carbon content in
the carburized zone of only about 0.45 to 0.6 per cent, is necessary
to produce a glass-hard surface on oil quenching; and the core
becomes exceedingly strong, as well as tough and non-brittle. From

NICKEL STEELS 279

these facts it is evident that the lowering of the maximum carbon
concentration to a percentage not exceeding that of the eutectoid
ratio will almost entirely eliminate the danger of chipping and
flaking. The use of moderate temperatures for carburization and
of oil for quenching decreases the liability to warping and fracture.
The physical characteristics of the carburized zone after the second
oil quenching, the steel of the case having an approximate chemical

FIG. 164. — Nickel Steel. Nickel, 4.86 per cent. Carbon, 0.115 per cent.
Case Carburized and Air Cooled. XlOO, (Sauveur and Reinhardt.)

composition of 0.45 per cent, carbon and 5.0 per cent, nickel, will
be approximately:

Tensile strength, Ibs. per sq. in 260,000

Elastic limit, Ibs. per sq. in 250,000

Elongation in 2 ins., per cent 2

Reduction of area, per cent 5

Brinell hardness 490

Scleroscope hardness 74

280 STEEL AND ITS HEAT TREATMENT

The following physical results taken from the core of a double-
quenched steel analyzing:

Carbon.. 0.105

Manganese 0 . 43

Phosphorus 0. 014

Sulphur 0.030

Silicon 0.11

Nickel 5.0

show that the core will have great strength, high ductility (from the
reduction of area) , and very slight brittleness (as shown by the shock
test) :

Tensile strength, Ibs. per sq. in 200,000

Elastic limit, Ibs. per sq. in 170,000

Elongation in 2 ins., per cent 12

Reduction of area, per cent 54

Resistance to shock 75

Brinell hardness 295

The same steel, having an upper critical range of 1425° F., and
annealed at 1475° F., gave:

Tensile strength, pounds per square inch. . . . 90,600

Elastic limit, pounds per square inch 60,160

Elongation in 2 ins., per cent 20

Reduction of area per cut 60. 5

Resistance to shock 116

Brinell hardness . 179

The regenerative quenching for these steels may be carried out
either at a temperature slightly in excess of the upper critical range —
- or at about 1475° F., or, in order to more fully equalize and effect
the regeneration of the core, at some higher temperature, such as
1600° F. The hardening quenching temperature should be slightly
over the Ac range of the case, or approximately 1275° to 1325° F.
Oil may be used for both quenchings. For the characteristic French
steel containing about 6 per cent, nickel Guillet recommends the
temperatures of 1560° and 1250° F. respectively for the double
quenching.

3.5 PER CENT. NICKEL STEEL

We have previously discussed some of the factors which enter
into the quenching of nickel steels in general. Whether or not it

NICKEL STEELS

281

(9) jaqtnn^ asaap-rn jj adoosoaapg a-ioqg
g §

(g) jaqtn

Previ
to 18
coole
Quenc
fro
draw

33S

•ipai aaunbg aad spanoj '(z) »iniri

§ £ 8

«(f ) wajv J° aotaonpaa put! •(£) saqDnj j; a;

282

STEEL AND ITS HEAT TREATMENT

(g) Jaqmnjf ssanpanH edoosoaspg oa

8 g g

(g) jaqmnfj

13335

o o o' o o

a • a • ••-•-

» D g PP 02 tn ^-,

aiwnbg jod epnnoj '(%) li

'(l)qiST:

'(5) uaay jo noponpaH pau

NICKEL STEELS

283

(9) aaqtunjj ssanpjwjj adoaso-ispg

8 g §

lull.

"S-S

!!§
J35

bg asd spunoj

•f»uaO aaj '(t) weay jo

«(g) saqoni g u;

284

STEEL AND ITS HEAT TREATMENT

(9) jaqranji ssaapJUR adoosoaops 3-iou.s
§ S §

(g) aaqran^

ical Ranges : Size
Ac 1240° 1 in
Ar 1060°

ical Analysis
C. 0.39
Mn. 0.48
P. 0.009
S. 0.023
Si. 0.10
Ni. 3.36

I 1 !

•uoui eavnbg jad spunoj *(z) »iniri D!?B«13 P"« '(i) qVaea^g 8U8UOX

NICKEL STEELS

285

(9) jaqtnnjj ssanpavH adoosojspg aaoqg

(g) jaq

iot$
1 -2 of a

1400°
n as

<!o

•qoni wtrabg jad spunoj '(g) »iran oi^n«ia ptre '(i) qiS

•?nao Jaj '(f ) way jo no;;onpaH pn« '(g) saqoui g ni

STEEL AND ITS HEAT TREATMENT

may be necessary to use a temperature considerably in excess of the
upper critical range for hardening will depend upon the condition of
the steel as it comes to the heat-treatment plant; in many cases the
normal heating and quenching will suffice for general purposes.
The normal hardening temperatures for 3 to 3J per cent, nickel steel
may be approximately determined by reference to the critical range
diagram in Fig, 160, and by adding 50 to 100 degrees to the upper

500

IS 400

300

20U

Effect of Mass
C. 0.25 P. 0.01

Si. 0.09 S. 0.012

Mn. 0.67 Ni. 3.47

*'/*

Size in Inches

FIG. 170.— Effect of Mass upon the Hardness of Nickel Steel, Oil Treated.
(Matthews & Stagg.)

critical range for the given carbon content. It should be noted,
however, that the best hardening temperature should be determined
experimentally, whenever possible, for the particular stock to be
treated, since the method of manufacture, elaboration, size of sec-
tion, and various other factors all influence such temperature.

The normal characteristics obtained by the heat treatment of 1-in.
nickel steel rounds are shown in the charts of Figs. 165 to 169. It

NICKEL STEELS

287

should be remembered that these figures, although they have been
carefully checked up with other results as far as is possible, are
experimental figures, and should be used with discretion. In other
words, ordinary commercial heat-treatment practice involves so many
variables, and especially the " personal equation," that it should
not be expected that these results will be duplicated in every instance.

Brincll Hardness

Effect of Mass
C. 0.25 P. 0.01
Si. U.09 S. 0.012
Mn. 0.07 Ni. 3.47

•*,

10CO

••» —

FaJ

Jter

-^-L

1— -£.

-I t

f11

K)]-H

Ir-Fl

hr— VV'-ate

r-60(

Deg

'-x->

1600^

De^r

*aJj

• W,,,.,. 12001

Fahr

III

1 l/2 2 2'/i 3

Size in Inches

FIG. 171. — Effect of Mass upon the Hardness of Nickel Steel, Water Quenched.

(Matthews & Stagg.)

Again, these results represent the treatment of l-in. round sections —
so that while such results might be duplicated in practice with
sections up to 1J ins. in diameter, further increase in the size of
section will inevitably lower the physical test results to be obtained
under the same treatment. The influence of mass upon the Brinell
hardness is shown in Figs. 170 and 171.

On the other hand, the normal characteristics for annealed 3J per
cent, nickel steel, as given in the diagram in Fig. 172, represent the

288

STEEL AND ITS HEAT TREATMENT

average results which are, and should be, obtained in commercial
practice in the annealing of the more common and larger sections of
nickel steel. They represent, moreover, the minimum requirements
which are characteristic of many existing steel specifications for
3J per cent, nickel steel, annealed, for such uses as engine forgings,

100,000

0.35 0.40

Per Cent. Carbon

FIG. 172. — Normal Characteristics of Annealed 3.5 per cent. Nickel Steel.
Large-size Sections of Forgings. Manganese Approx. 0.6 per cent.

ordnance forgings, rolled slabs and billets, etc., both for Govern-
ment and commercial uses.

Similarly, the following physical results for heat-treated work
(quenched and toughened) have been taken from various specifica-
tions in order to show the minimum results which may be expected
in commercial practice. The manganese requirements are approx-
imately 0.50 to 0.70 per cent., and the nickel content not less than
3.25 per cent.

NICKEL STEELS

289

I
<o

C
V

o
19

»C 1C

7 7 o

^ *o ^

iC iO

ItJ

HN
1-4

1 I I « 2 ! 1

»c

C<l

»C ^H ^-c 00 ^ N '-|

01 (N (N -H

Elastic Limit.

35,000-45,000

d
d

40,000-50,000

50,000-125,000

60,000-130,000

43,000- 51,000

82,000-109,000

»o

120,000

§o ^ d §
5 ic ic o
d ic i V ic

^C 1C

* 55,000- 70,000

70,000-200,000

I 1 ; : §

»— '

: : o

g£

• 1C

• ac

•d

c
3

•

•g

' Tl ' V

•o

"H
3

•g

3
. a>

"§'

. 9

O
CQ

•g Sill

•c

S-

13 a> "3 *" "3
(-. ; +j a>

a c cs c!

15
o>

jt?T

«5

•^

<5 O •< W <!

**!

ffi

;

; ; ; g,

tJ-

•

—

: ::":f

^_.

M: ai : : °

• ,

^r1

IM*

«s

b£) bC *

_§

•3

.S .S ;' ; g

•&'

,fl

•?

o o : : M

o>
1

•^

4) 0) • • S

&? £? : : S

e3 cS . - 0

<—

u
a

a
SB

^ — d o O

1 i. * ! i

•8

~f.

w M w w o

—

•g -^ • •

•

00 ^5

*-•

O O JOQ ad i5

x x ° ® f

S S o d o

ic »c o <i o

CO CO CO CO ^

d
•<t

o o o o o

290

STEEL AND ITS HEAT TREATMENT

The relation which the maximum size of section bears to the
physical properties of steel is well illustrated by the following speci-
fication for Railway Motor Shafts — the steel to contain 3 to 4 per
cent, nickel:

Max. Dia.f
Inches.

Tensile Strength,
Lbs. per Sq. In.

Elastic Limit,
Lbs. per Sq. In.

Elongation, Per
Cent, in 2 Ins.

Reduction of
Area, PerCent.

95,000

65,000

90,000

60,000

85,000

55,000

80,000

45,000

over

80,000

45,000

The following equations connecting maximum strength, Brin-
ell hardness number and scleroscope hardness number have been
computed 1 from several hundred tests made with nickel steels
of different carbon content and heat treated to bring out all pos-
sible physical properties:

(1) M = 0.71 5-32.

(2) M = 3.5 S-6.

(3) £ = 5.0 £+48.

M = maximum strength in units of 1000 Ibs. per sq. in.
B = the Brinell hardness number.
S = the scleroscope hardness number.

The maximum strength corresponding to different Brinell val-
ues as determined by equation (1) for these steels is as follows:

Brinell.

Maximum Strength,
Lbs. per Sq. In.

Brinell.

Maximum Strength,
Lbs. per Sq. In.

100

39,000

350

216,000

150

74,000

400

252,000

200

110,000

450

287,000

250

145,000

500

323,000

300

181,000

550

358,000

The maximum strength corresponding to different scleroscope
values as determined by equation (2), and the corresponding Brin-

R. R. Abbott, A. S. T. M., Vol. XV, Part II, 1915, p. 43 et seq.

NICKEL STEELS

291

ell numbers as determined by equation (3), for these steels, are as
follows :

Scleroscope.

Maximum Strength,
Lbs. per Sq. In.

Brinell.

64,000

148

99,000

198

134,000

248

169,000

298

204,000

348

239,000

398

274,000

448

309,000

498

100

344,000

548

5 PER CENT. NICKEL STEEL

The use of nickel steel with the higher nickel content is now
largely limited to case-hardening purposes, which we have previously
described.

The physical results obtained from the treatment of 1-in.
sections, containing 5 per cent, nickel and 0.33 and 0.43 per cent,
carbon, are shown in the charts in Figs. 173 and 174 respectively.

HIGH-NICKEL STEELS

The high-nickel steels of 25 to 35 per cent, nickel are used prin-
cipally for gas-engine valves and spindles, ignition and boiler tubes,
and for other similar purposes. These nickel steels are extremely
tough, dense, have a high resistance to shock, a low coefficient of
expansion, and — in particular — are little subject to corrosion.

Their physical properties in the natural condition may be given
.as follows:

25 to 28% Nickel 30 to 35% Nickel
(0.3 to 0.5% Carbon) (average)
Tensile strength, Ibs. per sq. in. . . 85,000 to 92,000 95,000

Elastic limit, Ibs. per sq. in 35,000 to 50,000 50,000

Elongation, per cent, in 2 ins. .. 30 to 35 40

Reduction of area, per cent .... 50 to 60 58

These steels do not respond to heat treatment, but may be an-
nealed at about 1450° F. to facilitate machining, after which the

292

STEEL AND ITS HEAT TREATMENT

(9) jaqnmx sssnpjwE odoosojapg ajoqg

IfiflJ

fl!Hi

mill

lul

•S19 | *S8

oo ^ 5c>o

d <J| P« • • B

§ 1

i a

ojtmbg jad sptmoj '(g) *i

NICKEL STEELS

293

294 STEEL AND ITS HEAT TREATMENT

physical properties of the 30 to 35 per cent, nickel steels given above
will average:

Tensile strength, Ibs. per sq. in 85,000

Elongation, per cent, in 2 ins 30

Reduction of area, per cent 40

Nickel steels with 35 to 38 per cent, nickel and 0.3 to 0.5 per cent,
carbon have a coefficient of expansion which is less than any metal
known and amounts practically to zero. These alloys are used for
various clock, geodetic and similar instruments for precise measure-
ments. The physical properties of these steels are approximately:

Tensile strength, Ibs. per sq. in 100,000 to 115,000

Elastic limit, Ibs. per sq. in 64,000 to 78,000

Elongation, per cent, in 2 ins 35 to 25

Reduction of area, per cent 50

CHAPTER XII
CHROME STEELS

CHROME in steel has the characteristic function of opposing both
the disintegration and reconstitution of cementite.1 This is made
noticeable by the changes in the critical ranges of the steel, as it
makes them take place more slowly, that is, it has the tendency to
raise the Ac range and to lower the Ar range. Chrome steels are
therefore capable of greater hardness because rapid cooling is able
more completely to prevent the decomposition of the austenite.2
The greater hardness of chrome steels is also due to the presence of
double carbides of chrome and iron in the steel in the hardened or
slightly tempered condition. This additional mineral hardness is ob-
tained without raising the brittleness to such a degree as does carbon.
The degree of hardness is within certain limits dependent upon the
carbon content, as chrome alone will not harden iron. Harbord 3
states that carbonless, or nearly carbonless, chrome steel does not
harden when water quenched. Toughness is also conferred by the
degree of fineness of the structure, which is a characteristic of chrome
steels (similar to that of nickel), thus increasing the tensile strength
and elastic limit without a noticeable loss of ductility. Thus we
have that condition of " tough-hardness " which makes chrome
steel so valuable in parts requiring great resistance to wear (abrasive
action4). In regard to corrosion, Chappell5 states that " in neu-
tral corroding media the resistance offered to corrosion apparently
rises with the percentage of chromium. This is particularly the case
for salt water, and the employment of chromium steels in the con-
struction of ships would appear to be fully justified on this ground
alone. — The corrosion factor does not appear to be a purely addi-
tive quantity."

1 Savcra, " Metallography " (trans.), p. 315.

2 Stoughton, " Metallurgy of Steel," p. 407.

3 Harbord, " Metallurgy of Steel," Vol. I, p. 390.

4 See F. Robin, Journal. Iron and Steel Inst., Vol. II, 1910.
6 Chappell, Journ. Iron and Steel Inst., 1912.

295

296

STEEL AND ITS HEAT TREATMENT

The influence of chrome towards increasing the brittleness of
steel, especially upon prolonged heating at high temperatures in the
case-hardening process, is shown in the following results of tests by
Quillet:1

Resistanc

3 to Shock.

Treatment.

Chrome 0.70%.
Carbon 0.05%.

Chrome 1.20%.
Carbon 0.05%.

Annealed
Quenched .

32 kgms.
22

25 kgms.
15

Heated for 4 hours at 1830° F.
After double quenching

5
26

5
20

0.5 PER CENT. CHROME STEELS

Low-chrome steels find many valuable uses and at only a slightly
increased cost, since the usual charge by open-hearth steel manu-
facturers for 0.5 per cent, chrome steel is only about one or two
dollars a ton over the " base price." One-half of one per cent,
chrome raises the critical range on heating by about 25° to 35° F.
over that of the corresponding straight carbon steel. As a leeway
of at least 50° F. over the critical range is usually allowed for in
hardening, these chrome steels may in general be hardened at the
same temperature as for straight carbon steels of the same carbon
content.

0.5 CHROME, LOW CARBON

For carbons up to about 0.35 per cent, the addition of this small
amount of chrome confers practically no additional physical prop-
erties other than those which might be obtained by the use of a
slightly higher carbon content in a straight carbon steel. On the
other hand, for case-hardening purposes this small amount of chrome
confers homogeneity, greater strength and wearing qualities, due to
the much finer grain throughout after the double quenching, and to
the presence of double carbides in the case. It should be remembered
that chrome strengthens the cementitic element of the structure
of steel, which in turn must depend upon the amount of pear lite;
nickel, on the other hand, influences the ferrite constituent. Al-
though both elements will tend to make the structure much finer,

1 L. Guillet, " Trempe, recuit, revenu — "

CHROME STEELS 297

it is evident that while nickel will have its greatest effect upon the
lower carbon steels (those containing large amounts of ferrite),
chrome will be of the most importance in the high carbons in which
there is considerable cementite. Thus it is that although chrome
in small amounts will be of little direct importance in ordinary heat-
treated low-carbon steels, it will be of tremendous importance in
any case-hardening operation which will produce a high-carbon case.

The addition of chrome probably increases the velocity of pene-
tration of the carburization, under identical conditions, over that
of the corresponding straight carbon steels. With greater amounts
of chrome there is also the tendency toward a higher maximum
carbon concentration than that obtained with carbon steels similarly
carburized. The tendency to surface oxidation during carburiza-
tion as a characteristic of chrome steels has been noted by several
investigators; methods involving the use of mixed cements may be
used, however, which will modify or even eliminate this action.

As chrome emphasizes the harmful effects of prolonged heating
(opposite to the action of nickel), it is always necessary to double
quench the steel; that is, a regenerative quenching as well as the
usual hardening quenching. The greater surface hardness obtained
by the use of chrome steels permits the use of oil for both quenchings,
if desired, and thus tends to avoid deformation of the steel.

0.50 CHROME — 0.35 TO 0.50 CARBON

It has been the author's experience that with a carbon content
up to say 0.50 per cent, carbon, the addition of a half per cent,
chrome will give, after heat treatment, about 15 per cent, increase
in tensile strength, 10 per cent, increase in elastic limit, with prac-
tically no loss in ductility, over straight carbon steels of the same
carbon content. In the hardened condition it gives excellent ser-
vice for wearing surfaces, such as the jaws of wrenches, small gears,
etc. The tables on page 298 give test results obtained from open-
hearth steels.

0.50 CHROME, OVER 0.50 CARBON

With the increase in carbon the influence of chrome becomes
even more marked, due to the increasing amounts of double carbide.
The hardness increase is greater proportionally than the carbon
increase. For well-bits and jars in the hardened condition this steel
has no equal among the low-priced alloys or straight carbon steels.

298

STEEL AND ITS HEAT TREATMENT

For die blocks used in drop-forging work it does not seem quite to
" hit the mark," apparently not having the requisite toughness
to offset the brittleness, especially in the larger sections. With
0.70 to 0.80 per cent, carbon it makes an excellent chisel, while with
0.90 to 1.00 per cent, carbon and about 0.60 per cent manganese it
gives even better results for pneumatic chipping chisels than do
many varieties of high-speed steel. The Germans in particular
have made great use of 0.5 per cent, chrome steels for tools, such
as drills, saw-blades, knives, razors, files, and similar tools requir-
ing a keen cutting edge. Further increase in hardness may also
be obtained by the addition of silicon and manganese. The mini-
mum hardening temperatures for the higher carbons should always
be used to obtain the maximum effect of the chrome.

Treatment.

Mn.

Si. Cr.

1500° F. oil/12000 F.*

0.36

0.44

0.008

0.021

0.05

0.57

1500° F. oil/13000 F.*

1500° F. oil/13000 F *

As rolled, 4"X4" billet...

0.40

0.50

0.015

0.025

0.55

1460° F. water/1000* . .

0.47

0.60

0.006

0.025

0.108

0.51

1460° F. water/1100*. .

1460° F. water/1200*. .

1460° F. water/1300*. .

8" round as hammered . .

0.47

0 . 55

0.015

0.029

0.57

Treatment.

Tensile
Strength,
Lbs. per
Sq. in.

Elastic
Limit,
Lbs. per
Sq. In.

Elon-
gation,
Per cent,
in 2 Ins.

Reduc-
tion of
Area.
Per cent.

Brinell
Hard-
ness.

Sclero-
scope
Hard-
ness.

1500° F. oil/12000 F.*

103,200

76,000

27.0

62.0

1500° F. oil/13000 F.*

99,450

71,240

22.0

51.4

1500° F. oil/13000 F.*

94,100

67,460

28.0

69.0

As rolled, 4"x4" billet. .

93,000

72,000

26.0

50.0

1460° F. water/1000* . .

168,420

153,720

19.0

52.6

311

1460° F. water/1100*. .

143,060

128,860

19.0

60.2

277

1460° F. water/1200*. .

112,930

120,750

22.0

60.5

262

1460° F. water/ 1300*. .

120,550

109,540

25.0

67.6

212

8" round as hammered . .

95,000

50,000

22.0

50.0

* Tests from 1-in. rounds.

Critical range diagrams are shown in Figs. 175 and 176.

Results obtained upon oil quenching from 1400° F. and subse-
quent toughening of 1-in. rounds of the approximate composition
of 0.70 per cent, carbon, 0.60 manganese and 0.50 chrome are
given in the following table; note especially the high proportion

CHROME STEELS

299

FIG. 175. — Critical Range Diagram of Heat 2101. Carbon, 0.47 per cent.;
Manganese, 0.60 per cent.; Phosphorus, 0.006 percent.; Sulphur,
0.025 per cent.; Silicon, 0.108 per cent.; Chrome, 0.51 per cent.

FIG. 176.— Critical Ranges of Basic Open-hearth Steel, Heat 8148. Carbon,
0.50 per cent.; Manganese, 0.49 per cent.; Phosphorus, 0.010 per
cent.; Sulphur, 0.026 per cent.; Silicon, 0.05 per cent.; Chrome,
0.57 per cent.

300

STEEL AND ITS HEAT TREATMENT

of the elastic limit to the tensile strength, combined with good duc-
tility:

Toughening
Temperature, °F.

Tensile Strength,
Lbs. per Sq. In.

Elastic Limit,
Lbs. per Sq. In.

Elongation, Per
Cent, in 2 Ins.

Reduction of
Area, Per Cent.

900

199,500

179,050

6.0

25.6

1000

168,900

143,500

12.5

33.8

1100

145,700

119,400

15.0

42.2

1200

120,000

105,000

17.0

47.5

1300

107,100

91,400

22.0

58.1

The critical range diagram is shown in Fig. 177

FIG. 177. — Critical Range Diagram of Chrome Carbon Steel. Carbon, Approx.
0.70 per cent.; Manganese, Approx. 0.60 per cent.; Chrome,
Approx. 0.50 per cent.

1.00 PER CENT. CHROME STEELS

Chrome steels with about 1.00 per cent, chrome, with high
carbon, find their greatest use in balls, ball-races, cones, roller
bearings, crushing machinery, safe steel, tools, and other parts
requiring a very hard surface. The use of about 1 per cent, each
of carbon and chrome appears to give the highest combination
of " tough-hardness " and plasticity. Such steel requires care in
forging, which must be done at a good red heat and with powerful
blows. As forged, the steel is much too hard for ordinary ma-
chine work and must therefore be thoroughly annealed.

Annealing. — Annealing at the usual annealing temperatures for
an ordinary length of time will not generally suffice, due to the slow-

CHROME STEELS 301

ness with which the cementite is taken into solid solution by the aus-
tenite, and which is well illustrated by the following case: A num-
ber of 3-in. forged rounds were annealed at 1400° F. for four hours
and slow cooled with the furnace, but were then too hard for machin-
ing; they were reannealed for sixteen hours in a similar manner,
and although the grain was refined, they could be sawed only with
difficulty. The most expeditious method for annealing this steel
is to normalize and then anneal, as follows: first thoroughly heat
the steel at a temperature above the Acm point for the " solution "
of the cementite, air cool to a temperature beneath that of the Ar
point to prevent disintegration, reheat to a temperature slightly
over the Ac 1.2. 3 point to refine the grain, and slow cool in the
furnace or in lime to obtain the maximum degree of ductility (soft-
ness). After such a treatment the steel is easily machinable and
will have a Brinell hardness of about 130 to 170. For a steel con-
taining 1.00 to 1.40 per cent, carbon, under 0.50 per cent, man-
ganese, and about 1.00 per cent, chrome, the following temperatures
may be used to advantage:

1. Heat to 1700° to 1750° F.

2. Air cool to about 800° F.

3. Heat to 1400° F.

4. Slow cool in furnace or in lime.

Note. — Add 35° to the temperatures (1) and (3) if the chrome
is up to 1.50 per cent.

If, however, it is desired to anneal the steel by the straight anneal
only (i.e., using but one temperature and one heating), -this may be
done by a prolonged length of heating, followed by a very gradual
and extremely slow cooling. Thus the 3-in rounds previously
referred to were satisfactorily annealed by heating to a tem-
perature of about 1400° F., maintaining this temperature for about
sixty hours, and then cooling with extreme slowness through the
critical range. This is the more common method used by manu-
facturers of chrome steel for roller bearings (about 1 per cent, car-
bon and 1.25 to 1.50 per cent, chrome); the temperatures vary
between 1400° and 1475° F.; the length of time varies upon the
mass of the charge, and usually takes several days. Such steel
in the full annealed condition should have a Brinell hardness of not
over 170.

Hardening. — These steels take on great hardness, both on the
surface and at depth, when hardened, for which either water or
oil may be used. In the hardened condition the Shore scleroscope

302

STEEL AND ITS HEAT TREATMENT

gives a hardness figure of about 100. The critical ranges for these
steels with over 0.90 per cent, carbon will vary from 1330° to 1375°
F. for 0.5 per cent, chrome, to 1400° to 1450° for 1.5 per cent,
chrome. The results obtained from the heat treatment of 1-in.
rounds of a 0.64 per cent, carbon chrome steel are as follows:

0.64 CARBON; 0.28 MANGANESE; 0.17 SILICON; 1.04 CHROME.

Quenched in Oil
from 1600° F.
and Toughened
at - Deg. F.

Tensile
Strength,
Lbs. per
Sq. In.

Elastic
Limit,
Lbs. per
Sq. In.

Elongation,
Per Cent,
in 2 Ins.

Reduction
of Area,
Per Cent.

Brinell
Hardness.

750

227,500

170,000

5.0

13.5

477

930

212,000

155,000

8.0

19.5

444

1110

186,000

127,500

10.0

22.5

387

2.00 PER CENT. CHROME STEELS

The tables on page 303, taken from the work of McWilliams and
Barnes,1 and rearranged, show the physical properties of 2 per
cent, chrome steels of ascending carbons as rolled, heat treated and
annealed. Chrome steels with about 2 per cent, of chrome are
largely used in the manufacture of armor-piercing projectiles, besides
in such objects which require an extremely hard-wearing surface
such as in crushers, cold rolls, drawing dies, special files, etc.

HIGH-CHROME CARBON STEELS

For general practical purposes and heat-treatment work the
chrome content is limited to that percentage below which the steel
as cast will be pearlitic — that is, the critical temperatures on cool-
ing are all above atmospheric temperatures and the steel structure
is composed of pearlite plus either ferrite or cementite. The hard-
ness increases with the chrome content, and, according to Arnold
and Read, appears to be independent of the carbon content, whilst
the brittleness is far less than in carbon steels of the same carbon
content.

Martensitic Steels. — When the chrome content reaches from
5 to 7 per cent., dependent upon the carbon, the change point, Ar,
will fall below normal temperatures and the structure will become
troostitic or martensitic. That is, the structure will be comparable
with that of a straight carbon steel in the hardened condition. These

1 "Iron and Steel Inst. Journ."

CHROME STEELS

303

2.00 CHROME — 0.20 CARBON.

CRITICAL RANGE Ac3 = 1512° F.

Tensile

Elastic

Elon-

Reduc-

Alterna-

Treatment.

Strength.

Limit,

gation,

tion of

tions

Lbs. per
Sq. In.

Per Cent,
in 2 Ins.

Area,
Per Cent.

(Ar-
nold's).

1. As rolled

70,400

45,600

30.5

71.2

331

2. 1475° F. in water/ 750° F.

137,200

134,000

12.5

40.6

3. 1475° F. in water/10250 F.

116,000

110,000

16.0

50.7

144

4. 1475° F. in water/13000 F.

82,400

64,000

28.0

70.2

234

5. Annealed ....

66,000

32,000

40.5

77.9

410

2.00 CHROME— 0.25 CARBON.

CRITICAL RANGE Ac3 = 1490° F.

As rolled.

67,200

48,800

30.0

68.4

312

1475° F. in water/ 750° F.

176,400

156,600

12.0

42.5

103

1475° F. in water/10250 F.

144,000

136,000

14.5

51.5

1475° F. in water/13000 F.

96,000

82,000

25.0

68.6

204

Annealed

70.000

32.000

39 . 5

73 8

437

2.00 CHROME— 0.32 CARBON. CRITICAL RANGE Ac3 = 1445° F.

As rolled.

92,600

60,000

26.0

65.4

355

1475° F. in water/ 750° F.

200,000

184,000

9.5

37^0

1475° F. in water/10250 F.

159,200

151,800

15.0

52.2

141

1475° F. in water/13000 F.

109,600

94,000

22.5

67.2

197

Annealed

60.800

28.800

37 0

70 7

482

2.00 CHROME— 0.50 CARBON. CRITICAL RANGE Ac3 = 1432° F.

As rolled.

107,600

64,000

20.5

65.8

378

1475° F. in water/ 750° F.

228,200

224,000

9.0

30.3

1475° F. in water/ 1025° F.

179,200

170,200

13.0

42.5

111

1475° F. in water/13000 F.

124,800

114,000

21.0

61.5

169

Annealed

75.200

25.400

28.0 !

55.4

440

2.00 CHROME — 0.65 CARBON. CRITICAL RANGE Ac3 = 1440° F.

As rolled. . . . . .

142,600

116,000

14.5

41.1

292

1475° F. in water/ 750° F.

1475°-F. in water/10250 F.

193,000

188,200

10.0

32.4

1475° F. in water/13000 F.

125,200

113,600

21.0

55.6

133

Annealed

97.800

64.000

21 5

62 2

214-

2.00 CHROME — 0.85 CARBON. CRITICAL RANGE Ac3 = 1430° F.

As rolled

151,800

104,000

10.0

18.3

178

1475° F. in water/ 750° F.

1475° F. in water/10250 F.

191,400

185,000

8.5

28.2

1475° F. in water/13000 F.

126,000

115,000

20.0

51.7

155

Annealed . . .

80,200

37,600

32.0

63.5

316

304

STEEL AND ITS HEAT TREATMENT

steels have high tensile strength and elastic limit, low ductility,
great hardness and medium brittleness. Heat treatment has little
or no influence, except, perhaps, to refine the grain. On account of
their physical characteristics these steels are but little used, except
as applied to special tool steel. As the chrome content is again
increased to about 12 to 15 per cent., intensely white grains of the
double carbide of chrome and iron form within the martensite, and

Double Carbide

Martensite

Mart.

Pearlite

Double
Carbide

1.65

2.20

FIG. 178. — Microscopic Constituents of Chrome Carbon Steels.

gradually occupy the whole field with further increase of chrome.
These structural changes for varying percentages of chrome, and with
0.2 and 0.8 per cent, carbon respectively, are given by Guillet as
follows:

Structure With 0.2 Carbon

Pearlitic 0 to 7 per cent. Cr.

Troostitic 7 to 8 per cent. Cr.

Martensitic 8 to 13 per cent. Cr.

Martensiteplusdouble carbide 13 to 20 per cent. Cr. j lg and oyer cent
Double carbide ... over 20 per cent. Cr. J

With 0.8 Carbon
0 to 5 per cent. Cr.

5 to 18 per cent. Cr.

per cent.
These changes are shown graphically in Fig. 178.

CHROME STEELS

305

The effect of annealing and heat treatment upon high-chrome
carbon steels, with approximately 0.4 per cent, carbon is given in
the following table:1

HEAT TREATMENT OF HIGH-CHROME STEELS, 0.4 CARBON

Chrome,
Per Cent

Treatment.

Tensile
Strength,
Lbs. per
Sq. In.

Elastic
Limit
Lbs. per
Sq. In.

Elon-
gation
Per cent,
in 2 Ins.

Reduction
of Area,
Per Cent.

Annealed

53,760

39,872

24.0

5
10
15

Hardened & tempered
Annealed
Hardened & tempered
Annealed
Hardened & tempered
Annealed

123,648
94,080
121,632
101,472
130,144
80,864

109,312
51,296
94,976
56,896
109,312
47,488

12
21.5
12
18.5
11.5
21 5

37.0
44.0
53.6
50.0
54.6
46 5

20
25

Hardened & tempered
Annealed
Hardened

90,272
94,526
90,496

61,824
66,752
61,824

19.5
18
20

51.5
62.1
50 0

Annealed .

93,184

71,232

62 0

Hardened

87360

64518

65 0

Further data on high-chrome steels may be obtained from the
researches of Guillet,2 Portevin,3 Arnold and Read,4 Becker,5
Mars,6 and others.

1 J. Holtzer & Cie., Loire, France, from Harbord's " Metallurgy," I, 391.

2 Guillet, " Les Aciers Speciaux."

3 A. Portevin, " Revue de Metallurgie," 1909, No. 12, p. 1264, " Metal-
lurgie," 1910, Heft 6, s. 177.

4 Arnold and Read, " Iron and Steel Inst. Journ."
* O. M. Becker, " High-speed Steel."

6 Mars, " Spezialstahle." 1912.

CHAPTER XIII
CHROME NICKEL STEELS

Chrome Nickel vs. Chrome Vanadium Steels. — Chrome nickel
steels, as a type composition, probably represent the best all-around
alloy steels in commercial use for general purposes. By this it is
not to be inferred that chrome nickel should always be used in pref-
erence to other alloys; as a matter of fact, each type is more or less
peculiarly adapted to work of a distinctive nature. On the other
hand, chrome nickel steel of suitable composition will satisfy nearly
every conditipn for structural and similar purposes. Much has been
said and done with chrome vanadium steels, and while the latter
undoubtedly do fill a long-felt want along certain lines, it should not
be said that chrome vanadium steels are superior to chrome nickel
steels. In fact, with a few exceptions, chrome nickel steels of suitable
composition will generally measure up to any standards set by the
ordinary vanadium alloys and at equal or at even less cost. Neither
chrome vanadium, nor chrome nickel, nor any one type of steel is a
general prescription for the every ill of the steel user: each steel has its
distinctive characteristics and applications. And notwithstanding
the mass of advertising " literature " to the contrary, it would also
be decidedly improper to state, as a general rule, that either is
superior to the other.

Influence of Chrome and Nickel. — Chrome nickel steels of suit-
able composition appear to have the beneficial effects of both the
chrome and nickel, but without the disadvantages which are inherent
in the use of either one separately. Moreover, the presence of both
chrome and nickel seems to intensify certain physical characteristics.
To the increased ductility and toughness conferred by nickel on the
ferrite there is added the mineral hardness given to the cement ite
and pear lite by the chrome, but with a greater resultant effect.
Again, while the addition of nickel alone serves to diminish the
susceptibility to brittleness in the steel upon prolonged heating or
sudden cooling — in comparison with the corresponding straight
carbon steels — and, on the other hand, the presence of chrome

306

CHROME NICKEL STEELS 307

alone tends to the opposite effect, a suitable combination of the two
alloying elements tends to neutralize the harmful effects and also to
magnify the good points. This is not only brought out in the
static strength and ductility, but also in the dynamic strength or
fatigue resistance.

Statements have frequently appeared in print to the effect that
nickel " poisons " the steel dynamically; that chrome has little
influence one way or the other upon the fatigue resistance; and that
chrome nickel steels are inferior along these lines to certain other
specific alloy steels. In considering these broad statements there
are three things in particular which should be noted. Firstly, that
in the present state of the art of dynamic strength testing, the re-
sults so obtained are often widely divergent for the same steel, not
to mention any comparison of results upon different steels of dis-
similar type. Secondly, even assuming that concordant and strictly
comparative results could be thus obtained by means of the testing
machines now in use, the majority of the experimental results pub
forth to prove the general inferiority (dynamically) of chrome
nickel steels in relation to certain other types (e.g., chrome vanadium)
are oftentimes not really comparative at all, since the two distinct-
ive types of steel have been heat treated alike. That is, while it
may be perfectly good practice to quench a chrome vanadium steel
from say 1650° F., it might be distinctively poor practice to quench
a chrome nickel steel from the same temperature. And yet many
" comparative " results have been obtained in just such a manner,
to the detriment of either one steel or the other. Rather, then,
should each steel be treated in that impersonal and strictly scientific
manner which will tend to bring out the maximum qualities of each;
and then should the tests be made upon the same machine under like
conditions. Thirdly, whatever may be the influence of chrome
or nickel alone upon the dynamic strength of steel, it has been re-
peatedly demonstrated that the proper combination of the two
alloys undoubtedly produces a type of metal with vastly improved
capacity for resistance to fatigue.

Commercial Ratio of Chrome and Nickel Content. — From the
author's experience in both the manufacture and use of chrome
nickel steels it would appear that there is some ratio existing between
the proportion of the chrome and nickel which will give the most
efficient combination of physical characteristics. In other words,
by combining the chrome and nickel in some such ratio, the less
susceptibility to brittleness upon prolonged heating which is char-

308 STEEL AND ITS HEAT TREATMENT

acteristic of nickel additions will modify the greater susceptibility
to brittleness which is given by chrome alone, giving a stronger
and better steel than may be obtained when this ratio is not ob-
served. Again, it will be observed that if the chrome content greatly
exceeds a certain proportion in respect to the nickel, the steel will be
more difficult to heat treat successfully, the temperature limits are
more narrow, and the possibility of poor results is greatly increased.
This best ratio is probably about 2J parts of nickel to about 1 part
of chrome. Thus we have the principal standard types of 3.5 nickel

FIG. 179.— Protective Deck Steel. (Bullens.)

and 1.5 chrome, 1.5 nickel and 0.6 chrome, and various intermediate
types.

Carburization. — The carburization of chrome nickel steels does
not differ in principle from that previously described. These steels
generally carburize more rapidly and better than straight carbon
steels, and, in particular, give the characteristic gradual cemented
zone which should always be aimed for. The presence of suitable
proportions of chrome and nickel, as previously mentioned, also
gives that low brittleness of core which is so desirable; this fact
even permits the use of steels up to some 0.3 per cent, carbon without

CHROME NICKEL STEELS 309

great danger. The use of chrome nickel steel in case-hardening
work covers a wide range — from small gears subject to great shock
and wear to the heaviest grades of armor plate.

Heat Treatment.— The heat treatment of these steels does not
present any new problems. In the main the discussion under the
chapters on Carbon Steels and Nickel Steels will apply equally well
to chrome nickel steels. Similarly to nickel steels, these steels are
less susceptible to the deleterious influence of high temperatures,
and which will be subsequently mentioned. Suitable heat treatment
will develop a very fine micro-structure,^ is shown in Fig. 179,
representing the structure of specially treated chrome nickel steel
used for protective deck plate on battleships; 4he physical proper-
ies on this particular steel were: A

Tensile strength, Ibs. per sq. in 132,000

Elastic limit, Ibs. per sq. in 116,700

Elongation, per cent, in 2 ins 23

Reduction of area, per cent 64

Proper annealing will likewise develop a good micro-structure
in the steel, as is shown in Fig. 40. The critical ranges of chrome
nickel steels are somewhat lower than those of the corresponding
straight carbon steels, so that lower temperatures may be used for
quenching.

In general, the best treatments which can be given to these alloy
steels after forging are as follows:

a. Quench in oil from about 175° to 200° F. over the critical

range.
6. Quench in oil from about 50° over the critical range.

c. Anneal at about 75° under the critical range (also see II).

Machine.

d. Quench in the proper medium from about 50° over the

range.

e. Draw the temper to suit the work in hand.

For shafts and other structural parts in which the desired physical
properties may be obtained by a drawing temperature of about 600°

310 STEEL AND ITS HEAT TREATMENT

F. or over, and which will leave the steel in a machinable condition,
Treatment I may be modified at (c) as thus noted, and no further
treatment will be required. But if the drawing temperature must
be much lower, as for gears, the full treatment as in (I) is advisable.

a. Quench in oil from about 175° to 200° F. over the critical

range.

b. Quench in oil from about 50° over the critical range.

c. Draw at 900° or more, as the work may require. Machine.

Ill

The full treatment 'as given under (I) may be modified, if
desired, to the following, for parts to be drawn below 900° or
1000° F.:

a. Quench in oil from about 175° to 200° over the critical

range.

b. Reheat to about 25° to 50° F. over the critical range and

cool slowly. Machine.

c. Quench in oil from about 50° over the critical range.

d. Draw to the temperature required by the work.

LOW CHROME NICKEL STEELS

The low chrome nickel steels, containing approximately 0.5
per cent, chrome and 1.5 per cent, nickel, are the most used of all
the chrome nickel alloys. After forging or rolling, this grade of
synthetic steel may be heat treated to develop physical character-
istics nearly equivalent to a 3.5 per cent, nickel steel of similar car-
bon content. It does not have the objectionable tendency to
laminate which may characterize the latter steel, and on account
of the less cost of alloys, this chrome nickel steel is sold at a price
considerably lower than that of 3.5 per cent, nickel steel. This
grade of chrome nickel steel forges well and machines easily, does
not require the more narrow temperature limits in heat treatment as
do some steels containing a larger, although not as well proportioned,
amount of chrome and nickel.

The physical tests obtained from heat-treated steel of 1-in.
sections, containing approximately 0.5 per cent, chrome and 1.5
per cent, nickel, are given in the charts in Figs. 180 to 183.

CHROME NICKEL STEELS

311

312

STEEL AND ITS HEAT TREATMENT

(9) jaquing 963upann ndoMoaaps ajoqg
§ 8 § V?

(S) jaqnznx.«9an

i i

Section : Treatment:
round Quenched in
water from
1475, and
drawn as

given-

itical Range
Acl 13°
Ac3 1
Arl

nalysi
0.28
0.49
0.015

.
0.63
1.79

mc
C.
Mn
P.

Si.
Cr.

Ni.

jad spanoj <(%) lira 1*1

3 S

•jnao 43 j '(f ) tiojy ;o

5 «

pnw '(g) s-jtpni g u;

CHROME NICKEL STEELS

313

314

STEEL AND ITS HEAT TREATMENT

(9) aaqiaiifj ss^npavu edoosoaapg eaoijs
§ 12 S 'S

(g) jaquinN ssanpatJH Iia«U3

1 1 1 1 §

Annealed

„

§
i
i
i
§
i
i

ilus.

Illli!

Chemical Analysis: Critical Ranges: Size of Section:
C. 0.545 Ao T300°-1320° 1 inch round
Mn. 0.50 Ar U80°

lllss

d d o' e5 J

pi on S 5 iS

1 1 1 11

1 1 S I s"

qonj ejtinbg jad spunoj «(g) ?tm;i opsBia P"B '(i) q^Suaa^s ^Iisnax

10 0 «» 0 U5

JUSQ J»j '(f ) «ajy jo noijonpa^j pu« '(g) aaqanj % nt uonB8noi3

CHROME NICKEL STEELS

315

Small automobile forgings specifying 0.20 to 0.27 per cent,
carbon (with the above amount of chrome and nickel) may readi]y
be heat treated to fulfill the requirements of:

Tensile strength, Ibs. per sq. in 100,000

Elastic limit, Ibs. per sq. in 85,000

Elongation, per cent, in 2 ins 16

The critical range of steels of this analysis is about 1425-1450° F.;
a quenching temperature of 1475° to 1500° will generally give the
best results in small sections which only require a single quenching.

An interesting comparison between this grade of chrome nickel
steel and 3J per cent, nickel steel with the same carbon content is
shown by the following requirements for automobile axles of a diam-
eter of If ins. as specified by one manufacturer;

SPECIFICATION FOR HEAT-TREATED AUTOMOBILE AXLES, 1^-m- DIAM.

Chrome Nickel
Steel.

Nickel Steel.

Carbon

0.30 to 0.40

Manganese

0.50 to 0.70

Phos. and sul..

under 0 . 04

under 0.04

Silicon

0.20

Nickel

1.50

3.50

Chrome

0.50

Tensile strength, Ibs. per sq. in
Elastic limit, Ibs. per sq. in.

120,000
110,000

135,000
120,000

Elongation per cent in 2 ins

16.5

Reduction of Area, per cent
Bend test, flat around.

180°

The following results were obtained from test pieces taken from
full-size forgings of approximately 8 to 10 ins. in diameter, and
having an analysis of carbon 0.38 per cent., manganese 0.55 per
cent., chrome 0.31 per cent., and nickel 1.20 per cent. Many
interesting points were noticed in the treatment of this grade of
steel approximating the analysis given, and especially the seeming
contradiction of the annealing and hardening temperatures. The
critical range of this steel is approximately 1340° to 1360° F. Forg-
ings annealed at temperatures slightly above these show perfect
annealing. If the annealing temperature should be raised, it is at
the expense of ductility, and the fracture becomes coarsely crystalline

316

STEEL AND ITS HEAT TREATMENT

and shows "fire." The physical properties thus obtained are
shown in the following table :

Treatment.

Tensile
Strength,
Lbs. per
Sq. In.

Elastic
Limit,
Lbs. per
Sq. In.

Elonga-
tion
Per Cent.
in 2 Ins.

Reduction
of Area,
Per Cent.

As forged . .

86,500

45,000

34.6

Annealed at 1360°-1380°

75,250

35,500

57.9

74,300

35,000

50.1

Annealed at 1550°-1575°

77,750

42,000

26.0

Unless previously normalized (as by air cooling from a temper-
ature such as used in the high-temperature anneal above) , or double
quenched as outlined in the treatments previously given, the harden-
ing of sections of say 3 ins. diameter or more demands the use of
a temperature some 200° over the critical range to bring out
the full effects of this combination of chrome and nickel. This
is true of both oil and water quenching, but more noticeably so
in the case of oil baths. Parenthetically, it is interesting to compare
such treatment with that which has been previously described under
nickel steels; the technical reasons will then be clear. It is the
author's experience, as well as that of numerous others, that a
hardening temperature of 1550° to 1580° F. is required if the steel
has not been previously treated; take for example the following
tests on an 8-in round bar:

Quenched in

Drawn

Tensile

Elastic

Elongation,

Reduction

Oil from

at Deg.

Strength,

Limit,

Per Cent.

of Area,

Deg. F.

Lbs. per Sq. In.

in 2 Ins.

Per Cent.

1450

900

79,750

46,750

28.0

58.9

1500

900

88,300

53,000

23.5

55.4

1580

1050

99,000

71,500

23.5

61.9

1580

1050

92,200

63,600

27.0

62.4

1580

1050 95,880

70,700

24.0

62.0

It will be noticed that the quenching heat of 1580° F. not only gives
higher tensile strength, elastic limit and ductility, but also permits
of a drawing temperature some 150° higher. Microscopically the
structure obtained by the high-quenching temperature is excellent,
as is shown in Fig. 184. The structure of the same piece after
forging, and before treatment, is shown in Fig. 185.

Steel with approximately 0.50 per cent, chrome, 1.50 per cent,
nickel and about 0.40 carbon may be readily heat treated to fulfill

CHROME NICKEL STEELS

317

FIG. 184.— Chrome Nickel Steel Axled, Oil Quenched from 1580° F., Drawn at
1050° F. X100. (Bullens.)

FIG. 185.— Chrome Nickel Steel Axle as Forged. X100. (Bullens.)

318

STEEL AND ITS HEAT TREATMENT

the specification, in large sections up to 12 ins. diameter, and with
proportionally higher tensile results in small sections, of:

Tensile strength, Ibs. per sq. in 90,000

Elastic limit, Ibs. per sq. in 60,000

Elongation, per cent, in 2 ins 22

Reduction of area, per cent 50

The following equations connecting maximum strength, Brin-
ell hardness number and scleroscope hardness number have been
computed l from several hundred tests made with low chrome
nickel steel (1.5 per cent, nickel and 0.5 per cent, chrome) of different
carbon content and heat treated to bring out all possible physical
properties:

(1) M = 0.68 B -22.

(2) M = 3.7 S-l.

(3) B = 5A S+33.

M= maximum strength in units of 1000 Ibs. per sq. in.
# = the Brinell hardness number.
S=the scleroscope hardness number.

The maximum strength corresponding to different Brinell val-
ues as determined by equation (1) for these steels is as follows:

Brinell.

Maximum Strength,
Lbs. per Sq. In.

Brinell.

Maximum Strength,
Lbs. per Sq. In.

100

46,000

350

216,000

150

80,000

400

250,000

200

114,000

450

284,000

250

148,000

500

318,000

300

182,000

550

352,000

The maximum strength corresponding to different scleroscope
values as determined by equation (2), and the corresponding Brin-
ell numbers as determined by equation (3), for these steels, are as
follows:

1R. R. Abbott, A. S. T, M., Vol. XV, Part II, 1915, p. 43 et seq.

CHROME NICKEL STEELS

319

Scleroscope.

Maximum Strength,
Lbs. per Sq. In.

Brinell.

73,000

141

110,000

195

147,000

249

184,000

303

221,000

357

258,000

411

295,000

465

332,000

519

100

369,000

573

HIGH CHROME NICKEL STEELS

Chrome nickel steels containing approximately 3.5 per cent,
nickel and 1.5 per cent, chrome comprise a type of steel with dis-
tinctive physical characteristics, but which obviously are not shown
by the results of ordinary pull test values when taken in comparison
with the low chrome nickel steels. The following figures, giving the
ordinary physical properties, illustrate the latter point. Dependent
upon the section, treatment, and carbon content (0.2 to 0.5 per
cent.), they may be given as follows:

Composition.

Tensile Strength.

Elastic Limit.

Elongation.

Reduction of Area.

3.5 Nickel
1 . 5 Chrome

85,000 to
275,000

55,000 to
265,000

26 to 10

65 to 35

1 . 5 Nickel
0 . 5 Chrome

80,000 to
264,000

56,000 to
240,000

30 to 8

70 to 27. 5

It is evident, since the above results show but little difference,
that the superiority of the high chrome nickel steel does not appear
in the static properties. On the other hand, there is a tremendous
difference between the two types (in favor of the higher alloy) in
the dynamic and endurance strength, such as freedom from brittle-
ness and resistance to shock. This is illustrated by certain specific
uses, as examples, to which these steels are put and which demand
the highest attainable combination of dynamic strength, resistance
to shock, and high static strength. Thus with about 0.2 to 0.3
per cent, carbon these steels are used in protective deck plate,
requiring that peculiar combination of properties which comprise
ballistic strength; with a slightly higher carbon content, and cer-
tain other modifications, we have a typical Krupp armor plate; and

320

STEEL AND ITS HEAT TREATMENT

with 0.45 to 0.50 per cent, carbon these steels are used in high-duty
gears, and in which it is possible to hammer one tooth against its
neighbor without breaking it off.

Or, as it has been expressed in e very-day terms, the effect of the
larger amounts of alloys in suitable combination is like a comparison
between a trained athlete and the amateur. Each man may be
able to lift a maximum weight of say 200 Ibs. But when it comes
to repeating that same feat a number of times in succession, the
trained man, with his developed powers of endurance, will win
every time. And thus it is with the high alloy steel.

Typical results for a steel of this type are given in the chart in
Fig. 186.

The following equations connecting maximum strength, Brin-
ell hardness number and scleroscope hardness number have been
computed 1 from several hundred tests made with high chrome-
nickel steel (3.5 per cent, nickel and 1 per cent, chrome) of different
carbon content and heat treated to bring out all possible physical
properties :

(1) M = 0.7lB-33.
" (2) M = 3.7 S-3

(3) £=4.8 ,S+58.

M = maximum strength in units of 1000 Ibs. per sq. in.
# = the Brinell hardness number.
$ = the scleroscope hardness number.

The maximum strength corresponding to different Brinell val-
ues as determined by equation (1) for these steels is as follows:

Brinell .

Maximum Strength,
Lbs. per Sq. In.

Brinell.

Maximum Strength,
Lbs. per Sq. In.

100

38,000

350

215,000

150

73,000

400

251,000

200

109,000

450

286,000

250

144,000

500

322,000

300

180,000

550

357,000

The maximum strength corresponding to different scleroscope
values as determined by equation (2), and the corresponding Brin-

»R. R. Abbott, A. S. T. M., Vol. XV, Part II, 1915, p. 43 et seq.

CHROME NICKEL STEELS

321

wwiipjirn IpuiJQ

a.-,

Illliill

1 §

I 1

ipnj tuwnbg jod spunoj

J3J '

a 2

i 7, «« nop«Snn|5f

322

STEEL AND ITS HEAT TREATMENT

ell numbers as determined by equation (3), for these steels, are as
follows:

Scleroscope.

Maximum Strength,
Lbs. per Sq. In.

Brinell.

71,000

154

108,000

202

145,000

250

182,000

298

219,000

346

256,000

394

293,000

442

330,000

490

100

367,000

538

INTERMEDIATE TYPES OF CHROME NICKEL STEELS

Between the high and low composition types of chrome nickel
steels previously given there are a great variety of combinations of
the two alloying elements. Thus, in the chart in Fig. 187, we have
the results obtained from the heat treatment of 1-in. rounds with
1.0 per cent, chrome and 1.75 per cent, nickel. Other results,
from similar compositions, taken from representative practice in the
automobile world are given as follows :

CARBON 0.26 TO 0.35.

Treatment.

Tensile
Strength,
Lbs. per Sq. In.

Elastic
Limit,
Lbs. per Sq. In.

Elongation,
Per Cent,
in 2 Ins.

Reduction
of Area,
Per Cent.

Brinell
Hardness.

Hardened. . . .
Toughened. . .
Untreated ....

197,000
110,000
106,000

135,000
90,000
70,000

9
25

37
55
45

460 to 480
235 to 250

CARBON 0.46 TO 0.55.

Tempered ....

305,000

265,000

480 to 525

Toughened. . .

130,000

114,000

300 to 335

Annealed

95,000

68,000

180 to 200

The effect of mass upon the latter type of chrome nickel steel is
shown in Fig. 188.

The chart in Fig. 189 gives the results of tests upon steel con-
taining 0.75 per cent, chrome and 3.0 per cent, nickel, while Fig. 190
illustrates a characteristic French steel containing 0.50 per cent,
chrome and 2.50 per cent, nickel.

CHROME NICKEL STEELS

323

(9) Jaqmn^ asanpjcjj Pdoosoaopg ajoqg

1 s § s

(g) aaqmnx: ssonpaBH ipuua

S § S S §

^ ^ CO C^ r^

Annealed

l-H

-2

Chemical Analysis: Critical R^ges: Size of Section: Treatment:
C. 0.45 Acl.2.3 1390° 1 Inch round Previously double
Mn. 0.45 Ar3.2.1 1250° quenched iu Oil

»• fg t0800' Qu±:dTnIeOdil

Si: $ K^

'/.-

— . i

/ :
/

/ .

' f

Hardened

1 1 1 1 i"

•qonj sjunbg jad spnnoj '(%) 11011-7 oi-js^ia pa« '(i) qiSnaaig eitsaax

i § 3 » . 3

**aaO J3J '(f) «3JV J° nojionpaji pan '(g) sdipaj g n; aon«Suoi3

324

STEEL AND ITS HEAT TREATMENT

Steels with 0.60 per cent, chrome, 3.5 per cent, nickel, and
0.2 to 0.4 per cent, carbon, in medium-size forgings, may readily
be treated to give a minimum of:

Tensile strength, Ibs. per sq. in 120,000

Elastic limit, Ibs. per sq. in 105,000

Elongation, per cent, in 2 ins 20

Illustrative of the relation of drawing temperatures to the carbon

Effect of. M

C. 0.50 P. 0.01

Si. 0.16 S. 0.011

Mn. 0.41 Cr.

Ni. 2.02

200

1 I1/* 2

Size in Inches

FIG. 188. — Effect of Mass upon the Hardness. (Matthews & Stagg.)

content for steels of this composition and with the same size of
section,1 to meet the above specification, the following may be of
interest:

Per Cent. Carbon.

Drawing Temperature.

0.24
0.27
0.36

1150° F.
1200° F.
1240° F.

1 Protective Deck Plate.

CHROME NICKEL STEELS

325

(9) aaqain^j gasapatifj sdooso-

§ 'i? §

ilfslll

! fihsa

Mlllii

£ &•

•2 g
I 2

I?
fll'

f 3 3 S S 3

0 =5 3 =- d o«

1 1

S S

^ 55

•qoui aatinbg «d spnnoj

'(f ) «aay jo noi^anpaa pas '(g) saqon

326

STEEL AND ITS HEAT TREATMENT

(9) ujqcun^ s«apieH adoiaowps a^HS

g g s s

(S) jaqmaji 3wapieH l[aaiia
11111

a
g

I
1
1
g

1
I

Chemical Analysts: Critical Range: Size of Sett ioa» Treatment!
C. 0.33 Ac 1120° Itoltfincli Quenched
Wn. 0.42 Ar 1230° in water
Si. 0.14 fromlDCOf
Cr. 0.42 and drawn
Ni. 2.43 as below.

• ,

)

__ .

•» /

Hardened1

-1

§ § 1 § 1

gill §
•qoaj anmbg i^d spanoj *(j) is^n 3JWB(3: POB '([) qi3a3ng 8(;saaj,

S* « S s S

•;aao wj «(0 Baiy )o aoninpaa pao «(?;) s»<pai j ai uouwaaoia

CHROME NICKEL STEELS

327

SPECIAL CHROME NICKEL STEELS

The following tests by Revillon on various combinations of car-
bon, chrome and nickel in chrome nickel steels will be of interest:

ANALYSES

No.

Carbon.

Man-
ganese.

Phos-
phorus.

Sulphur.

Silicon.

Chrome.

Nickel.

.22

.54

.009

.044

.36

.35

2.19

.25

.027

.043

.08

.48

2.75

.425

.27

.006

.042

.20

1.20

2.86

.17

.53

.006

.053

.16

.18

3.47

.105

.43

.014

.030

.11

.85

4.38

.86

.23

.014

.030

.11

.86

.88

.25

.52

.006

.053

. 17

1.28

3.82

.31

.70

.014

.021

.17

1.48

2.75

.42

.22

.013

.057

.11

.31

4.09

.45

.28

.014

.030

.11

.58

2.25

.52

.27

.006

.030

.39

.43

2.80

.77

.32

.014

.030

.11

.19

1.13

.10

.35

.003

.035

.31

1.75

5.36

.265

.24

.014

.030

1.27

2.33

4.40

.27

.39

.014

.030

.11

.85

4.90

.36

.37

.006

.053

.23

1.15

4.20

.39

.68

.018

.021

.35

.78

5.19

CRITICAL POINTS

Degrees Fahr.

No.

Ac.

Ar.

Ac.

Ar.

1472

1256

1418

1229

1463

1274

1472

1283

1508

1265

1508

1301

1454

1274

1400

860

1427

1139

1400

986

1418

1301

1490

986

1355

941

1418

770

1454

617

1436

482

1400

788

328

STEEL AND ITS HEAT TREATMENT

ANNEALED

No.

Annealing
Temper-
ature,
Deg. F.

Tensile
Strength,
Lbs. per
Sq. In.

Elastic
Limit,
Lbs. per
Sq. In.

Elon-
gation,
in 2 Ins.
Per Cent.

Reduc-
tion
of Area,
Per Cent.

Guillery
Shock
Test.

Brinell
Hard-
ness
Number.

1472

80,690

56,320

64.9

133.8

153

1472

87,690

61,160

55.7

170

1292

105,400

73,670

63.2

112.1

197

1652

87,760

50,200

21.5

36,. 1

168

1472

90,600

60,160

60.5

115.7

179

1382

137,960

72,960

23.8

21.7

210

1292

114,650

64,570

17.5

50.5

1112

134,410

126,440

14.5

59.2

54.2

250

1112

115,930

99,560

65.8

101.2

217

1382

122,320

71,120

45.5

43.4

220

1382

135,830

81,780

49.8

54.2

251

1112

119,610

83,910

9.5

46.7

39.8

273

1112

163,850

142,940

13.5

58.4

137.4

178

1652

123,030

75,100

5.5

9.8

68.6

232

1382

142,510

93,870

6.5

39.8

288

1112

128,290

119,610

62.4

50.6

225

1112

147,210

91,600

14.5

52.8

268

NOTE: It will be noticed that in a number of instances the temperature used in the
above annealing is under the Ac point, which will explain the high tensile results obtained.
Such cases do not represent full annealing.

HEAT TREATED

No.

Quenching
Bath and
.Temper-
ature,
Deg. F.

Draw-
ing
Tem-
pera-
ture
Deg. F.

Tensile
Strength,
Lbs. per
Sq. In.

Elastic
Limit,
Lbs. per
Sq. In.

Elon-
gation
in 2 Ins.
Per
Cent.

Reduc-
tion of
Area,
Per
Cent.

Guil-
lery
Shock
Test.

Brinell
Hard-
ness
Num-
ber.

Water, 1382

204,500

180,510

44.3

68.6

370

Oil, 1472

225,010

189,170

19.3

418

Oil, 1472

572

264,550

214,820

6.3

42.7

54.2

412

Oil, 1562

197,700

173,520

15.4

61.5

328

Water, 1382

201,970

173,520

72.3

295

Oil, 1382

932

230,410

221,880

1.5

4.3

36.1

388

Oil, 1562

208,370

183,350

9.5

343

Oil, 1472

292,280

289,440

9.5

36.3

57.8

425

Air, 1472

221,020

189,870

54.2

396

Water, 1472

932

190,580

173,520

6.5

47.8

79.5

301

Cil, 1472

572

215,530

202,250

43.4

395

Oil, 1472

932

210,500

181,930

14.1

32.5

425

Water, 1472

188,020

168,970

72.3

286

Water, 1562

183,190

160,010

52.5

57.8

298

Water, 1382

932

187,640

157,730

45.7

72.3

300

Air, 1562

235,390

225,860

24.5

54.2

402

Air, 1472

32.5

512

CHROME NICKEL STEELS 329

CHROME NICKEL STEEL IN AUTOMOBILE CONSTRUCTION

0.25 Carbon and under

Principally for case-hardening purposes, such as bevel driving
and transmission systems, steering-wheel pivot pins, cam rollers,
push rods, and similar parts which must not only have a hard
exterior surface, but possess strength as well.

0.25 to 0.35 Carbon

Axles. — Steering knuckles, bolts, pinions, steering pivots,
spindles, driving shafts, etc., gears with light case, drawn. Gears
hardened, but not drawn.

This grade of chrome nickel steel forges and machines well,
and responds to heat treatment in matter of strength as well as of
toughness.

0.35 to 0.45 Carbon

Crankshafts. — Countershafts, propeller shafts, live axles,
diving shafts.

This grade possesses under suitable heat treatment a high
degree of strength and considerable toughness. Its fatigue-resisting
(endurance) properties are extremely high.

0.45 to 0.55 Carbon

Tempered Gears. — This grade probably gives the greatest
possible hardness with the least possible brittleness (in combination)
of any steel for transmission purposes.

MAYARI CHROME NICKEL STEEL

Mayari steel is a " natural alloy " steel containing from .20 per
cent, to .70 per cent, chrome and 1.00 per cent, to 1.50 per cent,
nickel. It is made from a low-phosphorus Cuban ore containing the
alloying elements chrome and nickel. In the blast furnace the
chrome and nickel in the ore are reduced, forming a natural con-
stituent of the iron. By means of the duplex process — Bessemer
converter and open hearth — the iron is then made into steel, the
chrome and nickel pass into the steel, forming a natural alloy, with
no other addition of these elements in the furnace or ladle being
necessary. Mayari steel has given excellent satisfaction in a large
number of cases, although it undoubtedly is not equal to synthetic
chrome nickel steel where the highest quality chrome nickel steel is
required.

330

STEEL AND ITS HEAT TREATMENT

In the natural or forged condition Mayari steel has from 8000
to 10,000 Ibs. per square inch higher tensile strength and elastic
limit than a carbon steel of the same carbon content. Like all
alloy steels, it welds with more or less difficulty by the ordinary
methods, and would not be recommended for purposes where a
welded part is subject to great strains. By careful work in a Thomp-
son electric welding machine, excellent results are obtained, so that
where this method is applicable Mayari steel may be welded satis-
factorily.

The physical properties of Mayari steel, heat treated, in | in.
bars, are shown in Figs. 191, 192 and 193. The effect on the physical
properties of variation in the size of the piece treated is indicated in
the charts, Figs. 194 and 195, which show the properties of heat-
treated rounds from 1 in. to 6 ins., and J in. to 4J ins. diameter,
respectively. All of the rounds on the same chart were from the
same heat of steel. These were treated together at the same time in
exactly the same manner. The first chart is 0.28 per cent, carbon,
and the second 0.39 per cent, carbon; both grades contained 0.45
per cent, chrome with the usual nickel. On the bars over 2 ins.
in diameter the tests were taken one-half the distance from the
center to the outside, and on the smaller rounds they were taken
from the center.

The following table 1 shows the approximate difference in draw-
ing temperatures for Mayari steel of larger sizes than those given
on the charts of Figs. 191 to 193. When it is desired to obtain the
same elastic limit on a size larger than f-in. diameter, find the draw-
ing temperature on the chart, then by making the allowance given
in the table below for the size desired, the proper temperature for
this elastic limit will be determined. The other properties will
vary from those on the chart by the percentage shown in the table.

Physical Properties; Per Cent, of that given on Charts for

J-in. Rounds.

Change in

Diameter.

Drawing

Temperature.

Tensile
Strength,
Per Cent.

Elastic
Limit,
Per Cent.

Elongation,
Per Cent.

Reduction of
Area,
Per Cent.

fin.

100

2| ins.

- 90° F.

102

100

3£ ins.

-135°F.

110

100

4j ins.

-235° F.

122

100

iPenna. Steel Co.

CHROME NICKEL STEELS

331

(9) joquni^ ssa

(g) jtoqranx ssaupjBH i

1 1 1 §"

ajtmbg Jad spunoj '(g) liran 0139*13 puts '(]) T)j3nang ejisn

s s

'(f) raay 30 noijonpaa PUB «(g) saqonj g tn noij«3noia

332

STEEL AND ITS HEAT TREATMENT

1 e § s

i
1

1
1

6
%

M
o

I— 1

a*
P^

^ ^ (.Jaaqozn^^HUO-a^ ^

„

§
i
1
I
5

'Ei

?.
13

0*
I

I 1 1 1 I

1 1 1 sf . 1

•qoaj a-nmbgaad spnnoj '(3) iturri at^wq^ pun '(i) q|3u3j;g aijanox

o ;? fi "» ^

-JH30 aaj '(*) G-»JV ;o not^npan pun «(g) s^qouj 2 ni uot;nanota

CHROME NICKEL STEELS

333

334

STEEL AND ITS HEAT TREATMENT

•2<)oooo

»-i280000
£270 000
tQ 200 000
£250000
p. 240 (XX)
: 230 000
£ 2 >0 (XX)
^210000

«: 200 ooo
•g lyoooo

.5 180000
M 170000
SlGOOOO
"uo 150 000
^ 140000
W 130 000
^ 120000

3110000

£ 100 000

= 80000

5! 70000

02 CO 000

<U 50000

= 40000

C 30 000 —

S20000

10000

^^^j

05 Jt

•*

FIG. 194. — Effect of Size on Physical Properties of Mayari Steel, 0.30 Carbon.
Same Analysis and Same Treatment, (Penna. Steel Co,) ,

C 280 000
W270000

S 140000

a 130 ooo

120 000

"2 no ooo

a 100 000
5 90000
be 80000
= 70000

±l/a rouad

FIG. 195.— Effect of Size on Physical Properties of Mayari Steel, 0.40 Carbon.
Same Analysis and Treatment, (Penna, Steel Co.)

CHAPTER XIV
VANADIUM STEELS

THE author approaches the general subject of vanadium steels
with much hesitation. What has not been said about the merits
or demerits of chrome vanadium versus chrome nickel is hardly
worth mentioning; and many of the " facts " (or fancies?) put forth
in the advocation of one and the utter condemnation of the other
should not be mentioned. Having had to do with the manufacture
as well as the use of both, he feels that much may be said in favor
of each, but that neither steel is the " one and only" — each has its
specific sphere of usefulness and most advantageous application.

The principal effect of vanadium additions to steel is its
effect upon the physical characteristics of the steel. Like most
alloys, vanadium tends to give a finer and denser structure
than that ordinarily obtained in straight carbon steels. In true
vanadium steels, i.e., steels in which vanadium is present in
definite commercial quantities, the general action of vanadium is
similar to that of many of the alloys previously discussed, but it
also presents other interesting phenomena. Vanadium, in the
regular steels containing about 0.12 to 0.20 per cent, vanadium,
is probably present in both the ferrite (similar to nickel) and also as
a double carbide in the cementite (similar to chrome). In support
of the first statement that vanadium is in solid solution in the ferrite
are the results of many tests which appear to show that the duc-
tility is higher in these commercial vanadium steels than in corre-
sponding, steels which do not contain vanadium. This is based on
the assumption, generally accepted, that the nature of the ferrite
element is indicative, to a large degree, of the ductility of the steel.

The proof direct that vanadium forms a double carbide is illus-
trated by steels with higher percentages of contained vanadium.
Thus steels containing 0.2 per cent, carbon and up to 0.7 per cent,
vanadium, or 0.8 per cent, carbon and 0.5 per cent, vanadium, are
normally pearlitic; but any increase in the vanadium content over
these limits will produce a characteristic double carbide component.

335

336 STEEL AND ITS HEAT TREATMENT

From these limiting ratios of carbon and vanadium it is evident
that vanadium has a powerful influence upon the transformation
ranges — more so, indeed, than any of the common alloying elements.
This also goes to show the reason why only small quantities—
0.25 per cent, vanadium or under — are necessary to produce a
noticeable effect.

As a general proposition, any alloy which tends to form a cemen-
titic compound in steel also has the tendency to require a higher
temperature for quenching in order to bring the steel as a whole
into a state of equalization. This was found to be true in the case

FIG. 196. — Chrome Vanadium Steel, Type A, Oil Treated at the Same Tempera-
tures Used for a Corresponding Chrome Nickel Steel. X60. (Bullens.)

of chrome, and it is also true of vanadium steels. A study of the
heat-treatment data subsequently given will show that vanadium
alloy steels give the best. results with an apparently abnormally high
quenching temperature, or at about 1560° to 1600° F. for the me-
dium carbon grades. This point is also illustrated by the photo-
micrograph in Fig. 196; which illustrates the structure of a rolled
plate of " Type A " chrome vanadium steel oil treated at the tem-
peratures best suited for a chrome nickel steel of the same carbon
and manganese content. From the structure thus shown it is
evident that the steel as a whole has not been equalized at the
temperature for hardening (1500° F.) which was used, since the

VANADIUM STEELS 337

ferrite (white) is still segregated and tends to follow the lamellar
structure of the original steel.

On the other hand, if we follow out the characteristics peculiar
to most alloy steels of a carbide nature, we would expect that the
vanadium steels would be inherently more sensitive to prolonged
heating or rapid cooling. Now while it is true that steels containing
vanadium will give a greater depth of hardness upon suitable quench-
ing than will some steels of a ferritic nature (such as nickel steels),
it does not appear to be true that vanadium abnormally increases
the sensitiveness of the steel to prolonged heating. This appears
to be one of the anomalies of vanadium steels.

Viewed from the standpoint of physical test values, vanadium
requires the presence of another alloy as an " intensifier," in order
that the full effect and influence of the vanadium additions may be
felt. Just as chrome greatly intensifies the influence of nickel in
steel, so chrome also seems to bring out the latent capabilities of
vanadium, but to an even greater extent. Thus the majority of
the vanadium steels now in commercial use are of the chrome vana-
dium type.

The predominant note which is always sounded when speaking
or writing about chrome vanadium or vanadium steels is that of
increased dynamic strength. There is little doubt but that vana-
dium greatly increases the dynamic strength in comparison with
that of a corresponding straight carbon steel. Upon the relative
merits, as regards dynamic strength, of chrome vanadium and
chrome nickel steels, we have commented under the latter steels.
Extensive tests made by the author to determine dynamic strength
have led to varying results, and he deems it best to leave the
subject with the warning given in the opening paragraph of this
chapter.

The following equations connecting maximum strength, Brin-
ell hardness number and scleroscope hardness number have been
computed 1 from several hundred tests made with chrome vanadium
of different carbon content and heat treated to bring out all pos-
sible physical properties:

(1) M =

(2) M=4.2 £-21.

(3) £ = 5.5 S+27.

R. R. Abbott, A. S. T. M., Vol. XV, Part II, 1915, p. 43 et seq.

338

STEEL AND ITS HEAT TREATMENT

M = maximum strength in units of 1000 Ibs. per sq. in.
B = the Brinell hardness number.
S = the scleroscope hardness number.

The maximum strength corresponding to different Brinell val-
ues as determined by equation (1) for these steels is as follows:

Brinell.

Maximum Strength,
Lbs. per Sq. In.

Brinell.

Maximum Strength,
Lbs. per Sq. In.

100

42,000

350

219,000

150

77,000

400

255,000

200

113,000

450

290,000

250

148,000

500

326,000

300

184,000

550

361,000

Scleroscope.

Maximum Strength,
Lbs. per Sq. In.

Brinell.

63,000

137

105,000

192

147,000

247

189,000

302

231,000

357

273,000

412

315,000

467

357,000

522

100

399,000

577

Static test results 1 upon various " types " of vanadium steels
follow:

In part by the American Vanadium Co., Pittsburgh, Pa.

VANADIUM STEELS

339

TYPE " A " CHROME- VANADIUM STEEL
Tests from Small Sections

Carbon 26% Manganese

Chromium 92% Silicon

Vanadium 20%

.48%
.06%

Treatment.

Tensile
Strength.

Elastic
Limit.

Elongation
in 2 Ins.,%.

Reduction

of Area, %.

As rolled

132,000

110,000

19.0

51.5

Annealed 1475° F
Oil tempered:
1650 °-l 155° F .

83,700
133,000

61,000
99,020

34.8
30 0

66.4
69.9

1650-1110
1650 -1020
1650 - 930

137,000
141,500
162 700

112,000
123,000
146,250

20.0
18.0
15 0

61.0
63.5
57.0

1650 - 840

177,500

151,500

14.0

53.0

1650 - 750

183,500

155,000

13.0

51.0

1560-1155
1560 -1110

131,000
133,000

100,000
108,400

28.0
17.5

67.0
65.4

1560 -1020

137,500

112,750

21.0

64.5

1560-930 ......
1560 - 840
1560 - 750

156,800
171,100
173,900

138,440
147,150
149,800

16.5
15.0
13.0

59.8
61.0
57.0

Water tempered:
1650 °-l 155° F..

156,000

133,000

18 0

62 5

1650 -1110 .

160,900

149,700

16.0

60.4

1650 -1020
1650-930
1650 - 840
1560-1155
1560-1110
156C-1020
1560 - 930

167,800
183,200
204,800
153,050
156,500
166,800
176,950

151,000
166,800
176,200
136,600
146,300
149,100
165,000

12.0
12.5
12.5
27.0
17.0
14.0
14 0

53.6
56.5
54.5
60.0
61.0
58.9
59 0

1560 - 840

201,800

172,800

12.5

54.5

Tests from Medium Sections

Carbon 23% Manganese 58%

Chromium 82% Silicon 105%

Vanadium 17%

Stock.

Treatment.

Tensile
Strength.

Elastic
Limit.

Elongation
in 2 Ins., %.

Reduction
of Area, %.

Oil tempered:

2^-in

1650°-1050° F..

125,730

108,950

19.0

60 4

2i-in

1650 -1050

124,160

106,000

20.0

60.6

2f-in.

1650 -1050

122,740

104,750

19.5

57.0

2f-in.

1650 -1050

126,700

111,500

17.0

53.0

2|-in.

1650-1050

121,080

106,500

18.0

60.7

2|-in.

1650-1050

124,130

107,000

18.5

61.1

340

STEEL AND ITS HEAT TREATMENT

Test from 6-in. Tender Axle

Carbon 29% Manganese 28%

Chromium 1.00% Silicon 06%

Vanadium 20%

Treatment.

Tensile
Strength.

Elastic
Limit.

Elongation
in 2 Ins.,%.

Reduction
of Area,%.

Water tempered:
1690°-1155°F

115,000

90,000

21.0

55.0

Tests from Locomotive Driving Axles, 10 7ns. Diameter
Average Test of 287 Heat-treated Axles

Carbon 35% Manganese 50%

Chromium 90% Vanadium 22%

Elastic limit, pounds per square inch 81,600

Tensile strength, pounds per square inch 108,890

Elongation in 2 ins., per cent 21 . 75

Reduction of area, per cent 58 . 75

TYPE " D " CHROME VANADIUM STEEL
Tests on Small Sections

Carbon 50% Manganese 92%

Chromium 1 . 02% Silicon 065%

Vanadium 20%

Treatment.

Tensile
Strength.

Elastic
Limit.

Elongation
in 2 Ins.,%.

Reduction
of Area, %.

Brinell
Hardness
No.

As rolled

153,350

124,450

12.5

37.0

286

Annealed 1475° F

103,440

63,660

25.8

61.5

187

Oil tempered:

1650°-11100 F

186,800

170,000

15.5

45.2

340

1650 -1020

201,150

186,100

13.0

45.5

364

1650 - 930

209,800

192,200

12.5

42.5

364

1650 - 840

227,040

217,360

10.0

35.5

402

1650 - 750

264,500

239,700

6.5

17.0

444

1600-1110

186,100

161,200

13.5

45.5

340

1600 -1020

205,500

187,000

12.0

45.0

340

1600- 930

214,050

203,600

11.5

43.0

380

1600 - 840

237,500

221,000

10.0

29.5

418

1560-1020

197,100

187,100

12.5

45.0

340

1560 - 930

214,270

201,400

11.5

36.0

418

1560 - 840

234,150

215,850

9.0

28.5

418

1560 - 750

261,850

. 240,000

7.0

22.0

418

1520-1020

183,500

177,250

14.5

47.5

340

1520 - 930

215,450

193,100

12.0

41.5

387

1520 - 840

237,750

213,400

10.0

35.5

387

1520 - 750

260,500

240,000

8.0

24.0

444

VANADIUM STEELS

341

Fig. 197. — Effect of Mass upon the Hardness of Chrome Vanadium Steel.

(Matthews & Stagg.)

The effect of mass upon the hardness of steel of this type is
shown in Fig. 197. l

TYPE " G " CHROME VANADIUM STEEL

Carbon 60% Manganese 54%

Chromium 88% Silicon 175%

Vanadium 19%

Treatment.

Tensile
Strength.

Elastic
Limit.

Elongation
in 2 ins.,%.

Reduction
of Area,%.

Brinell
Hardness
No.

Oil tempered:

1650°-1110°F

205,190

179,300

13.0

37.0

402

1650 - 930

240,400

220,000

10.0

28.3

477

1650-750

273,000

248,660

8.0

27.3

532

From Matthews and Stagg, " Factors in Hardening Tool Steel."

342

STEEL AND ITS HEAT TREATMENT

NICKEL VANADIUM STEEL

Carbon 29%

Nickel 3.41%

Vanadium. .

Magnanese . . .
Silicon

20%

.45%
.090%

Treatment.

Tensile
Strength.

Elastic
Limit.

Elongation
in 2 ins.,%.

Reduction
of Area, %.

Annealed 800° C

107,300

73,000

23.5

55.5

Oil tempered:

1600°-1160°F

148,300

126,250

18.0

58.0

1600-1110

150,000

128,500

17.5

57.4

1600 -1020

151,500

132,500

16.0

56.9

1600-930

162,000

144,200

14.5

52.6

1600 - 840

178,200

157,210

13.0

52.7

1600- 750

193,200

163,000

12.0

50.2

1520-1160

137,700

123,000

16.0

59.0

1520 -1110

140,700

125,500

17.5

54.2

1520 -1020

148,100

126,800

16.5

55.0

1520 - 930

154,900

135,000

15.5

57.2

1520 - 840

165,800

146,500

14.0

55.2

1520 - 750

181,000

162,800

14.0

53.5

Water tempered:

1600°-! 160° F

148,000

126,700

18.5

58.1

1600-1110

153,800

133,100

15.0

58.8

1600 -1020

156,300

136,500

14.0

54.5

1600-930

161,200

146,700

14.5

56.4

1600-840

186,400

173,300

13.0

52.7

1600 - 750

195,200

176,580

12.0

52.2

1520-1160

139,800

128,570

18.5

59.7

1520 -1110

146,000

132,250

14.0

57.5

1520 -1020

154,600

133,900

15.5

56.3

1520 - 930

160,400

144,600

15.0

51.7

1520-840

184,500

176,750

13.0

53.0

1520 - 750

199,300

182,700

12.0

50.0

VANADIUM STEELS

343

W-H

ClCOi-HOOtOOil^'— iOO<MOCO

lOiOiOiOiOiOtO^O^O'OiOiO
cO cO cO cO CO CO cO cO cO CO CO CO

g O CO CO CO CO CO

o o o o o o

h h t* ' I* I* id

C 0> 03

o o o o o o

35 o p 1-1 1-1 I-H

O T-i T— I 1— I 1— I 1— I

T i I I I I

IO CO »0 C5 C5

00 N

T^ IO

i ^

IO CO

CHAPTER XV
MANGANESE, SILICON AND OTHER ALLOY STEEL

MANGANESE STEELS

THE term " manganese steel," by commercial usage, generally
refers to steels with that high percentage of manganese which will
cause the metal to become austenitic under the conditions of ordinary
cooling or suitable heat treatment. But before proceeding to a
discussion of such steels, it is desirable to amplify the remarks we
have previously made upon the subject of pearlitic manganese
steels.

PEARLITIC MANGANESE STEELS

It is a well-known fact that manganese in these steels adds con-
siderably to the tensile strength; this beneficial effect is further
dependent upon the percentage of carbon, as has been previously
shown. On the other hand, the effect of manganese in normally
pearlitic manganese steels upon the fragility of the steel is a more or
less undetermined factor. Many persons have undoubtedly con-
fused the subject of inherent brittleness or non-resistance to shock
with the results caused by the sensitiveness of these steels to certain
heat-treatment methods. They have assumed, because a piece of
steel with 1 to 2 per cent, manganese may have cracked on drastic
water quenching, that the steel was " brittle," when, as a matter of
fact, this result was probably due to reasons entirely apart from the
dynamic strength of the metal. Thus it may be said that this situa-
tion has led to the belief that manganese contributed an embrittling
effect to the steel. Even assuming that such an influence may
exist in the case of very high-carbon steels, it distinctly has not been
proven to be true of the hypo-eutectoid steels. In fact, there is
now considerable evidence which tends to show that the lower-
carbon pearlitic manganese steels, when properly made and suitably
heat treated) are not brittle. If the steel is made in small heats, has
been thoroughly refined, and with the elimination of impurities to

344

MANGANESE, SILICON AND OTHER ALLOY STEELS 345

a minimum, the author believes that a great deal may be accom-
plished with such steels.

With steel made by the ordinary open-hearth process, it should be
remembered that the presence of any considerable amount of man-
ganese, such as 1 per cent, or more, has the tendency to increase the
sensitiveness of the steel in its response either to prolonged heating
at temperatures above the critical range, or to rapid cooling from
such temperatures. Thus high temperatures of annealing will
increase the grain size very rapidly; while high carbon may cause
the steel to fracture when water quenched.

On the other hand, certain manganese steels with 1.5 to 2 per
cent, manganese and a considerable carbon content, made in the
electric furnace, have shown wonderful mechanical properties, and,
in addition, will stand a tremendous amount of abuse in their thermal
treatment without any great ill effects. Granting that the electric
furnace is capable of producing a higher grade of steel than other
processes now in use, it must nevertheless be evident that a large
proportion of the merits of these pearlitic manganese steels must
be due to the inherent influence of the manganese itself.

In treating pearlitic manganese steels it should be remembered
that each 0.1 per cent, manganese will lower the critical range on
heating by about 5° to 6° F., so that lower temperatures may, and
in most cases should, be used for their hardening or full annealing.
In general, the effect of manganese on the critical ranges is about
twice that of nickel.

HIGH-MANGANESE STEELS

In general, the requirements for producing a commercial mangan-
ese steel necessitate a manganese content of about 6 or 8 per cent,
to 20 per cent., in combination with the proper amount of carbon.
Below the lower limits given, the steel, even by the most suitable
treatment, may be characterized by the presence of weak and
brittle martensite. The upper limits are determined by the cost
of the manganese additions, and further, by the again predominating
influence of the carbon content (when the manganese rises to around
20 per cent.), which will make the steel stiff and brittle when cold.
Most manganese steels will have about 11 or 12 per cent, manganese
and about 1.0 to 1.2 per cent, carbon.

Recent research work along the lines of determining the proper
combination of carbon and manganese has greatly widened the
commercial range for the manganese content, so that the more

346 STEEL AND ITS HEAT TREATMENT

recent steels have the tendency toward a percentage of manganese
lower than that originally thought necessary. Similarly, the field
for the use of high manganese steels has also been considerably
broadened. Above all, however, the peculiar merit of these steels
lies in the resistance to abrasive wear, in combination with suffi-
cient strength and ductility. In this regard, manganese steels
appear to resist the abrasive wear characteristic of heavy impacts
of hard substances better than that caused by the sliding attrition
of hardened parts, or like that of an abrasive wheel.

Aside from the dynamic strength, the selection of a manganese
steel for any specific work depends upon the correlation of wear-
ing qualities and static properties. In general, and in connection
with a maximum wear resistance, it may be said that the most
ductile steel which will give an elastic limit sufficiently high to avoid
distortion in service will be best. And these, in turn, depend upon
the proper combination of carbon and manganese. Thus a steel
with 9 to 11 per cent, manganese and the proper amount of carbon
will have a higher elastic limit than a steel with over 11 per cent,
manganese. Again, steel with 11 per cent, manganese and 1.10
per cent, carbon will have a higher elastic limit than a steel with
15 per cent, manganese and 0.8 per cent, carbon. With high man-
ganese and low carbon, steels quenched in water from 1830° F. will
give a low elastic limit and a flow of metal which may prove excessive
for many duties. A great deal also depends upon a suitable heat
treatment of the steel.

As might be expected from our knowledge of the influence of
the rate of cooling upon the structure of high-alloy steels, the physical
properties of these high-manganese steels are greatly modified by
the method of casting, the size of the casting, and the mechanical
elaboration. The first two factors in particular have a great influence
upon the toughness of the metal. The average tests of commercial
manganese steels with about 11 or 12 per cent, manganese and a
little over 1.0 per cent, carbon will give approximately the following:

Condition of the metal.

Tensile Strength,
Lbs. per Sq. In.

Elastic Limit,
Lbs. per Sq. In.

Elongation,
% in 2 Irs.

Cast

82000

45000

Rolled

135,000-140,000

60 000-70 000

30-40

Forged

142,000

55 000

The elastic limit of some sections as rolled may even go as high as
75,000 Ibs. per square inch; the proper heating and working of the

MANGANESE, SILICON AND OTHER ALLOY STEELS 347

metal plays a very important part in the results to be obtained on
physical test.

A common specification for manganese steel rails is as follows:
Chemical :

Carbon, per cent 0. 95 to 1 . 15

Manganese, per cent 10 to 13

Silicon, per cent 0. 20 to 0.40

Phosphorus, per cent under 0. 10

Sulphur, per cent under 0. 06

Physical :

Tensile strength, Ibs. per sq. in. . . 100,000

Elastic limit, Ibs. per sq. in 55,000

Elongation in 2 ins., per cent 20

Small amounts of chrome are sometimes added to increase the
elastic limit, so that in rolled sections the elastic limit will often be
as high as 85,000. It is stated that the resistance to shock is not
apparently lowered by the addition of chrome up to 1 per cent.,
but with chrome above 0.5 per cent, the elongation is rapidly de-
creased, and with chrome above 1 per cent, the elongation falls
below 20 per cent, in 2 inches.

The heat treatment of high-manganese steels presents a most
important phase in connection with the successful application of
these steels. Incorrect treatment is responsible for many of the
failures which have been registered against manganese steels, and
usually has been caused either by a mistaken idea of the particular
structure best suited to the specific work in hand, or by a lack of
sufficient knowledge of the mechanics of the austenite transforma-
tion. The use of the microscope, and a judicious consideration
and application of the results obtained, are probably the best means
of solving, a given problem in connection with heat-treatment
adjustments.

This thermal treatment for the majority of these steels involves
two distinct, though correlated factors: (1) The change of grain
size, and (2) the relationship of austenite and carbide, with or with-
out the presence of martensite. As the principal manganese steels
now used in commercial practice do not naturally contain mar-
tensite, nor is it generally wanted, its consideration may be omitted.

The necessity for the first requirement should be obvious: com-
mercial high-manganese steel as cast is fundamentally austenitic;

348

STEEL AND ITS HEAT TREATMENT

the crystals are often excessively large, and in many instances form
a weak, columnar structure. It is evident that such a steel, for
many purposes, will be entirely unsatisfactory.

On the other hand, many high-manganese steels as forged, are
characterized by an exceedingly fine, almost chalky structure, and
yet may be very brittle. If a bar of manganese steel should be
heated to 1800° F., and one half be allowed to cool slowly and the
other half quenched in water, both ends will have a comparatively
fine structure or grain size, yet the slow-cooled end will have but
2 to 4 per cent, elongation as against 50 to 60 per cent, elongation
in the quenched end.

Fig. 198. — Commercial Manganese Steel Annealed at 1750° F.
X100. (Bullens.)

Although annealing will effect a change in the grain size, to
a greater or lesser extent, it will also have a far-reaching, vastly
more important, and injurious result — and we intentionally omit any
reference to the tendency which certain compositions might have to
become martensitic on very slow cooling. This effect of annealing
is due to the formation, on very slow cooling, of the maximum
amount of carbide — an extremely hard and brittle manganitic
cementite rejected by the austenite, and which forms as a weak
membrane around the austenite grains, as spines and needles, or
in other characteristic manner. This is shown in the photomicro-
graph in Fig. 198, taken from a rolled commercial manganese steel

MANGANESE, SILICON AND OTHER ALLOY STEELS 349

and then annealed at 1750° F. The effect of annealing upon the
physical properties is shown by the following table of results,
obtained from tests upon annealed manganese steels:

HIGH-MANGANESE STEELS, ANNEALED.

Carbon,
Per Cent.

Manganese,
Per Cent.

Tensile
Strength,
Lbs. per Sq. In.

Bias tic
Limit,
Lbs. per Sq. In.

Elongation
in 2 Ins.,
Per Cent.

Reduction
of Area.
Per Cent.

0.95

10.07

94,600

67,500

1.4

1.00

11.21

99,950

77,660

0.75

1.07

13.38

103,670

62,350

3.4

Annealing these high-carbon manganese steels is obviously
illogical, and it is the carbide which is the main source of the diffi-
culty. And yet a large proportion of the peculiar and distinctive
wearing qualities of these steels is probably due to the manganitic
carbide, although in a form other than that just described.

On heating to a high temperature the carbide membrane produced
by slow cooling, or the carbide segregations, are gradually taken
into solution by the austenite; and by rapid cooling from that same
temperature the carbide is more or less prevented from reprecip-
itating, and especially from taking on that enweakening structure
(i.e., as a membrane or segregations) previously mentioned. Since
the carbide originally formed in casting is very sluggish in its response
to heat in being absorbed by the austenite (as is also a character-
istic more or less marked in all hyper-eutectoid steels), and as the
equalization of the steel as a whole also takes place slowly, a high
temperature is necessary. Further, the temperature must also be
high, and the cooling be effected very rapidly — such as water quench-
ing— to retain the carbide in solution. Such a treatment will give
the most ductile steel. The temperature required is generally not
less thanJ830° F. Thus a water quenching from 1830° F. of the
annealed steels previously given will show a tensile strength of about
135,000 to 145,000 Ibs. per square inch, with an elongation of 50
to 60 per cent.

On the other hand, by varying the factors of temperature,
duration of heating, and rate of cooling, it is possible to obtain phys-
ical properties covering a wide range. The static strength and
ductility are largely governed by the amount of the original free car-
bide which is taken into solution and there retained by water quench-
ing. Thus the properties of the steel, looking at the results of heat

350 STEEL AND ITS HEAT TREATMENT

treatment from this point of view, may be varied from that charac-
teristic of the steel as cast, rolled, or forged, to that indicative of a
full " water toughening."

There is another possible development of scientific heat treat-
ment, however, which the author believes has not been fully noted—
not even by some manufacturers of high-manganese steel itself.
And by this is meant a treatment which will " spheroidalize " the
carbide. The application of such a process as applicable to straight
carbon steels, and the superior qualities thereby resulting in resist-
ance to wear as relative to case-hardening work, have been discussed
under Chapter VII. A microscopic examination of many properly
treated high-tungsten steels will show a similar spheroidalizing
action, and to which many of the mechanical properties of these
steels may be due — even though the exact relation of cause and effect
may not be known. And so certain processes may be worked out
(or so the author is led to believe) for these high-manganese steels,
in which resistance to wear is the main result desired. Based upon
theory, and upon partly developed experiments, the author offers
the above merely as suggestions.

In conclusion, the author would call attention to the fact that
high-manganese steel has no critical or transformation points or
ranges. Thus while in the ordinary steels the heat treatment is
more or less guided by such temperatures, in high-manganese steels
the only criterion of proper temperatures is the relation of the car-
bide to the physical properties: the absorption, with or without the
precipitation, of such carbide, is the underlying basis for heat-treat-
ment adjustment.

SILICON AND SILICO-MANGANESE STEELS

The use of silicon in commercial steels is practically confined
to two classes: (1) a medium carbon and about 1.50 per cent, silicon,
for use in tempered gears and springs — known as silico-manganese
steel; and (2) a nearly carbonless steel with up to 3.50 per cent,
silicon, for use in electrical apparatus.

The manufacture of silico-manganese steels in the open hearth
must be carefully watched on account of their great tendency to
piping and segregation, and a large discard must be made in the
cropping of the ingots to insure good steel. Silico-manganese steels
have considerable popularity among the foreign automobile manu-
facturers; their use in this country, however, is usually limited to
the lower-priced cars. Although the lower cost favors their use,

MANGANESE, SILICON AND OTHER ALLOY STEELS 351

their great sensitiveness to heat treatment and feeble resistance to
shock limits their field of usefulness. But when handled with great
care the silico-manganese (and also the silico-chrome) steels will
give good results in works well equipped for obtaining accurate
results in their heat-treatment operations. The temperature limits
for quenching are narrower than for most alloy steels, and the steel
responds altogether too quickly to variations in heating and cooling.
Although these steels will give high static test results upon suitable
treatment, the brittleness which is inherent to this type of steel
usually proves the governing factor.

A typical American analysis for silico-manganese steel for gears
and springs is as follows:

Carbon, per cent 0.43 to 0. 53

Manganese, per cent 0. 50 to 0. 70

Silicon, per cent 1 . 25 to 1 . 50

Upon suitable treatment, usually a quenching in oil from about 1550°
to 1600° F., followed by tempering to suit the requirements, the fol-
lowing results are representative :

Tensile strength, Ibs. per sq. in 195,000 to 230,000

Elastic limit, Ibs. per sq. in 175,000 to 220,000

Elongation, per cent, in 2 ins 12 to 8

The tables (taken from a work by Revillon) on page 352 give some
characteristic French and German gear steels of the silicon-man-
ganese type, together with their physical properties. Steel No. 6
(a straight carbon steel) is given for comparison, and was designed
to do away with the reheating or tempering necessary with the
high-silicon steels; it is reported that tests on the untempered gears
gave very satisfactory results.

As a side-light on the effect of certain treatments of pieces of
large section of silico-manganese steel, the following may be of
interest. The author recently made some experiments with a char-
acteristic steel, 5 ins. in diameter, and analyzing 0.44 per cent, car-
bon, 0.60 per cent, manganese, and 1.50 per cent, silicon. The
purpose was to determine the possibility of using this steel in place
of a 0.60 per cent, chrome steel with approximately the same man-
ganese and carbon content. The principal requirement to be met
was to obtain a glass-hard surface such as could be obtained with the
chrome steel. It was found that the hardness requirement could
not be met with the silico-manganese steel, and more important,

352

STEEL AND ITS HEAT TREATMENT

that the steel would often split or crack when quenched in either
the cold water or brine bath used for the chrome steel.

EXPERIMENTS WITH SILICO-MANGANESE GEAR STEELS

No.

1
2
3
4
5

Chemical Analysis.

Annealed at 1650° F.

Car-
bon.

Man-
ganese

Silicon

Tensile
Strength,
Lbs. per
Sq. In.

Elastic
Limit,
Lbs. per
Sq. In.

Elon-
gation

Reduce
tion
of
Area.

Brinell
Hard-
ness.

Shock
Test.

0.40
0.57
0.50
0.70
0.39

0.57
0.61
0.64
0.78
0.52

1.89
1.22
1.64
1.87
1.98

105,820
127,860
123,450
142,510
109,230

68,410
74,100
82,490
80,790
78,510

19
15
12.5
15.5
20

40
29
21
35
43

207
225
215
241
203

36.1
21.7
32.5
23.3
43.4

0.45

0.42

0.43

87,760

48,780

19.5

197

43.4

No.

Quenched oil from 1520° F., drawn at 930° F.*

Tensile
Strength,
Lbs. per Sq. In.

Elastic Limit,
Lbs. per Sq. In.

Elongation.

Reduction
of Area.

Brinell
Hardness.

Shock

Test.

1
2
3
4
5

178,380
219,460
156,450
210,500
135,120

159,300
204,500
136,830
197,700
.107,810

4.5
5.5
4
2.5
11

12
15
19
17
43

315
467
307
435
274

39.8
25.3
75.9
47
47

278,770

271,660

3.1

422

39.8

* Nos. 5 and 6 were quenched from 1560° F. ; No. 6 was drawn at 400° F.
The critical ranges of No. 5 were: Ac, 1560°; Ar, 1410°. For No. 6, Ac, 1420°-
Ar, 1290°.

The most important use for straight silicon steels is that for
electromagnets and for other electrical purposes demanding a igh
magnetic permeability or electrical resistance. Hadfield's silicon
steel, containing approximately 2.75 per cent, silicon, and with car-
bon, manganese and the other impurities as low as possible, is repre-
sentative of this class. His treatment for this steel consists of first
heating it to about 1950° F. and cooling quickly, and then heating
to 1380° F. and cooling very slowly, and which is sometimes fol-
lowed by a reheating to 1475° F. and cooling very slowly.

Another silicon steel, used in place of dynamo sheet iron, specifies
similar carbon, manganese, etc., but with a silicon content of about
3.25 per cent. The thermal treatment recommended for this steel
is a thorough heating at about 1430° to 1475° F., followed by very
slow cooling.

MANGANESE, SILICON AND OTHER ALLOY STEELS 353

TUNGSTEN STEELS

The pearlitic low-tungsten steels when quenched from the proper
temperature do not appear to be any more modified by this quench-
ing than are the corresponding straight carbon steels; the effect of
tungsten in such steels is, however, to increase the tensile strength,
with the degree of brittleness remaining about the same. For this
reason tungsten is sometimes used in place of silicon — which has a
feeble resistance to shock — for springs. The following table gives
the analysis and physical properties of a characteristic low-tungsten
spring steel:

Carbon 0.45%

Manganese 0.22%

Silicon 0.30%

Tungsten 0.60%

Annealed.

Quenched in Oil from
1560°, Drawn at
930° F.

Tensile strength, Ibs. per sq. in
Elastic limit, Ibs. per sq in

113,500-121,000

85 000

185,000
- 128 000

Elongation, per cent

The use of tungsten for ordinary structural purposes is mainly
limited by the fact that such steels have to be made by the crucible
process.

The other, and most important uses for tungsten, are those for
permanent magnets (the steel usually being used in the hardened
condition), and for various varieties of tool steels in both high-
speed and water- or oil-hardening types . Since these steels involve
such a multitude of analyses and treatments, and form a subject of
their own, it has been deemed best to omit any further discussion,
but to refer the reader to works already published.

MOLYBDENUM STEELS

On account of its high cost the use of molybdenum has been
largely confined to high-speed and similar steel — and even there it
has usually been superseded by tungsten. In the lower percentages,
molybdenum may be present in steel as a fairly easily decomposable
iron-molybdenum compound; with larger amounts of both molyb-
denum and carbon it is generally believed that the molybdenum
forms a double carbide in a similar manner to chrome.

In the pearlitic molybdenum steels the influence of molybdenum
is much like that of chrome, in that it increases the tendency to

354

STEEL AND ITS HEAT TREATMENT

greater hardness with proper quenching or with cold work, and like-
wise to increased brittleness upon prolonged heating at high temper-
atures. On the other hand, the molybdenum steels have a markedly
higher ductility and toughness, besides an increased dynamic
strength. The best results (disregarding its use for tools) have
been obtained with the use of 1 to 2 per cent, molybdenum, in
combination with the proper proportion of carbon — the carbon having
a marked influence upon the physical properties of molybdenum
steels. Considerable experimentation has been carried out with
pearlitic molybdenum steels for rifle barrels and large guns; it has
also been used in high-duty machine parts such as propeller-shaft
forgings. It is reported that excellent results have been obtained
in case-hardening steels with about 1 per cent, molybdenum.

The influence of molybdenum depends largely upon the heat
treatment, as is shown in the following series of tests by Giessen
with steels containing 1, 2, 4, and 8 per cent, molybdenum:

1.00 PER CENT. MOLYBDENUM STEEL

Chemical.

As Rolled.

No.

Mo.

Tensile
Strength,

Elastic
Limit,

Elongation,
Per Cent.

Reduction
of Area

Lbs. per Sq. In.

in 2 Ins.

Per Cent.

0.195

1.03

67,040

44,800

33.31

64.32

0.445

1.05

108,800

78,800

19.5

49.23

0.87

1.02

160,000

104,000

14.5

34.36

1.215

1.10

117,340

1.0

2.02

Annealed

No.

Tensile
Strength,
Lbs. per
Sq. in.

Elastic
Limit,
Lbs. per
Sq. In.

Elon-
gation
Per Cent,
in 2 Ins.

Reduc-
tion of
Area,
Per Cent.

Bend
Test,
Deg.

Alter-
nat-
ing
Str'gth

Brinell
Hard-
ness.

Sclero-
scope
Hard-
ness.

52,300

27,800

35.5

65 75

180

336

71,420

38,720

25.0

39.2

180

210

131

108,100

52,000

17.22

22.25

103

228

85,300

52,100

5.55

7.5

207

Heat Treated (Hardened in oil, reheated to 1025° F.).

90,100

47,150

27.46

68.4

180

301

241

210,560

168,600

14.08

49.2

180

137

387

240,490

193,700

9.15

25.2

418

279,000

203,900

4.92

12.0

512

MANGANESE, SILICON AND OTHER ALLOY STEELS 355
2.00 PER CENT. MOLYBDENUM STEEL

Chemical.

As Rolled.

No.

Tensile

»
Elastic

Elongation,

Reduction

Mo.

Strength,

Limit,

Per Cent.

of Area,

Lbs. per Sq. In.

Lbs perSq. In.

in 2 Ins.

Per Cent.

0.246

2.176

117,820

21.05

57.0

0.442

2.181

150,980

16.7

46.41

0.883

2.186

198,910

124,250

12.1

32.07

1.21

2.109

216,830

169,340

7.04

9.6

Annealed

No.

Tensile
Strength,
Lbs. per
Sq. In.

Elastic
Limit,
Lbs. per
Sq. In.

Elon-
gation
Per Cent,
in 2 Ins.

Reduc-
tion of
of Area,
Per Cent.

Bend
Test,
Deg.

Alter-
nat-
ing
Str'gth

Brinell
Hard-
ness.

Sclero-
scope
Hard-
ness.

65,070

31,580

33.3

62.5

180

370

116

82,300

43,400

27.7

44.3

180

259

143

107,070

54,770

18.8

27.5

100

126

207

95,200

61,710

9.4

13.5

196

Heat Treated (Hardened in oil, reheated to 1025° F.).

171,140

115,400

15.49

54.4

180

172

387

211,460

149,100

14.08

47.2

180

103

444

260,800

178,800

5.63

12.0

512

270,940

512

4.00 PER CENT. MOLYBDENUM STEEL

No.

Chemical.

As Rolled.

Mo.

Tensile
Strength,
Lbs. per Sq. In.

Elastic
Limit,
Lbs. per Sq. In.

Elongation,
Per Cent,
in 2 Ins.

Reduction
of Area,
Per Cent.

9
10
11
12

0.19
0.487
0.865
1.06

4.11
4.01
4.00
4.02

119,120
188,160
230,270
239,230

75,370
120,060

179,900

21.70
13.5
8.0
10.56

52.71
33.81
17.27
18.40

356

STEEL AND ITS HEAT TREATMENT

Annealed

No.

Tensile
Strength,
Lbs. per
Sq. In.

Elastic
Limit,
Lbs. per
Sq. In.

Elon-
gation,
Per Cent,
in 2 Ins.

Reduc-
tion of
Area,
Per Cent.

Bend
Test,
Deg.

Alter-
nat-
ing
Str'gth

Brinell
Hard-
ness.

Sclero-
scope
Hard-
ness.

63,390

31,470

42.7

72.5

180

366

116

77,060

42,220

28.3

52.0

180

247

143

94,300

45,920

20.5

34.0

180

146

179

92,960

42,560

15.5

20.5

196

Heat Treated (Hardened in oil, reheated to 1025° F.)

88,180

66,500

30.20

64.0

180

329

286

188,900

139,780

11.26

41.6

123

109

444

258,050

203,940

4.22

4.8

512

282,240

267,000

7.04

23.2

532

8.00 PER CENT. MOLYBDENUM STEEL

No.

Chemical.

As Rolled.

Mo.

Tensile
Strength,
Lbs. per Sq. In.

Elastic
Limit,
Lbs. per Sq. In.

Elongation,
Per Cent,
in 2 Ins.

Reduction
of Area,
Per Cent.

13
14
15
16

0.135
0.361
0.445
0.775
1.125

8.01

8.17
8.11
7.85
7.92

92,290
148,290
215,040
193,890
245,500

154,110
149,090
189,950

25.7
19.4
19.71
9.85
8.45

52.22
45.9
34.0
18.4
16.4

Annealed

No.

Tensile
Strength,
Lbs. j>er
Sq. In.

Elastic
Limit,
Lbs. per
Sq. In.

Elon-
gation,
Per Cent
in 2 Ins.

Reduc-
tion of
Area,
Per Cent.

Bend
Test,
Deg.

Alter-
nat-
ing
Str'gth

Brinell
Hard-
ness.

Sclero-
scope
Hard-
ness.

79,070

41,660

31.1

58.75

180

283

143

77,060

34,720

36.6

68.23

180

273

143

83,220

38,640

32.2

57.5

180

215

156

87,580

45,140

22.2

35.5

171

108

170

92,290

48,830

16.1

24.0

187

Heat Treated (Hardened in oil, reheated to 1025° F.)

82,080

53,760

30.9

65.6

180

239

163

105,350

75,450

25.3

54.4

180

226

351

127,400

77,800

21.1

49.2

180

122

444

247,300

216,830

7.74

23.2

512

CHAPTER XVI

TOOL STEEL AND TOOLS

THE problem of selecting a proper grade of steel in relation to
the work required is one hitherto met by the steel manufacturer
alone. Until recently he has recommended this or that steel for a
given requirement, depending more or less upon his general knowl-
edge of the purpose for which the tool is to be used, and upon the
experience of his customers in the past. But with the entrance of
the technical man into manufacturing concerns and the great im-
provements resulting therefrom, a fuller knowledge of various steels,
their composition, applicability and efficiency has been demanded.
This has resulted in a wider dissemination of information regarding
the physical, chemical and mechanical properties of steels manu-
factured by various steel companies, and a corresponding education
of both maker and buyer.

Grade.— For the aid and information of their customers, the steel
maker usually groups his tool-steel products into various " grades "
and " tempers." The former term refers to the " quality " of the
steel, according to the class of raw material which has been used,
together with the skill and care taken in producing the finished
material. The highest grades should be used for tools operating
under severe working conditions, demanding great endurance and
resistance to torsional or other strains, or upon which a large labor
cost has been placed. These conditions, such as are found in expen-
sive dies, milling cutters, taps, etc., would require a high-grade
steel. For such purposes as mill-picks, cheap tools, etc., it would
be folly to use any but a lower-quality steel. Wear, the cost of
redressing, regrinding and heat treatment are other factors which
must be considered in the selection of the proper and most economical
steel which will give the greatest efficiency in all senses of the word.
With this in mind, the following brief synopsis is given:

1. Finest tools and dies: expense for material the smallest item
entering into the cost and upkeep of the finished tool ;

2. Finishing tools for lathe and planer work; special taps,

357

358 STEEL AND ITS HEAT TREATMENT

reamers, milling cutters and other similar tools requiring a high-
grade steel; wood-working and corrugating tools;

3. General tool purposes;

4. Ordinary purposes, such as chisels, smith and boiler shop
work, etc.

5. For rough or heavy work.

Expressing this in a different way, we may say that the choice
of a grade of tool steel depends upon three factors :

1. The precision of the work required of the tool;

2. The relative cost of the steel in comparison with the labor
involved in the manufacture of the tool;

3. The life of the finished tool and its relation to the cost of pro-
duction.

Temper. — Carbon tool steels are further denoted by the " tem-
per." In tool-steel sales parlance this refers to the percentage of
carbon in the steel and may be denoted by figures or letters. Such
classifications generally refer to a 10-point carbon limit — thus No. 7
temper may refer to 0.65 to 0.75 per cent, carbon, or it may be
represented by whatever the individual company has arbitrarily
selected. In this connection it should be noted that this " temper "
does not refer to, and should not be confused with the word temper
as indicating the operation of " letting down " the steel after
hardening.

General recommendations for the proper carbon content to use
for various tools are given in the following table; these, however,
must not be regarded as absolute, for much will depend upon the
grade of steel and upon the exact use of the tool,

APPROXIMATE CARBON CONTENT FOR ORDINARY TOOLS

Carbon, rp i

Per Cent.

1.50 Tools requiring extreme hardness. For turning chilled-
rolls and tempered gun-forgings. Roll corrugating.
1.40 Hard lathe work generally. Chilled-roll turning. Cor-
rugating.
Graver tools.
Brass-working tools.

1 . 30 General lathe, slotter and planer tools.
Razors.
Drawing dies,

TOOL STEEL AND TOOLS 359

Mandrels, granite points, scale pivots, bush hammers,
peen-hammers.

Ball-races.

Files.

Trimming dies. Cutting dies.
1 . 20 Twist drills. Small taps.

Screw dies, threading dies.

Edge tools generally. Cutlery.

Cold stamping dies, leather-cutting dies, cloth dies, glove
dies.

Nail dies, jewelers' rolls and dies.
1 . 10 Milling cutters and circular cutters of all descriptions.

Wood-working tools, forming tools, saws, mill picks, axes.

Small punches.

Taps.

Cup and cone steel.

Small springs. Anvils.
1.00 Reamers, drifts, broaches.

Large milling cutters, saw swages.

Springs.

Mining drills, channeling drills.

Large cutting and trimming dies.
0.90 Hand chisels, punches.

Drop dies for cold work, small shear knives.

Chipping chisels.

Cutting and blanking punches and dies.
0.80 Large shear knives, chisels, hammers, sledges, track chisels.

Cold sets, forging dies, hammer dies, boiler-maker's tools.

Vise-jaws. Oil-well bits and jars. Mason's tools.
0. 70 Smith shop tools, track tools, cupping tools, hot sets.

Set screws.

0.60 Hot work and battering tools generally. Bolt and rivet
headers.

Hot drop forging dies. Rivet sets. Flatteners, fullers,

wedges.

0.50 Machinery parts. Track bolt dies where water is con-
tinually running on dies (hot work).

Navy Specifications. — The United States Navy specifies the
following straight carbon tool steel for its general requirements :

360

STEEL AND ITS HEAT TREATMENT

Class.

II.

III.

IV.

Carbon
Manganese

1.25-1.15
0.35-0.15

1.15-1.05
0.35-0.15

0.95-0.85
0.35-0.15

0.85-0.75
0 35-0 15

Phosphorus
Sulphur . . .

0.015-0
0.02-0

0.02-0
0.02-0

0.02-0
0 025-0

Silicon

0.40-0 10

0.40-0.10

0.40-0 10

0 40-0 10

Chrome and vanadium optional.

Class I. Lathe and planer tools, drills, taps, reamers, screw-cutting

dies; taps and tools requiring keen cutting edge combined with

great hardness.
Class II. Milling cutters, mandrels, trimmer dies, threading dies,

and general machine-shop tools requiring keen cutting edge

combined with hardness.
Class III. Pneumatic chisels, punches, shear-blades, etc., and in

general tools requiring hard surface with considerable tenacity.
Class IV. Rivet sets, hammers, cupping tools, smith tools, hot-drop

forge dies, etc.; tools requiring great toughness combined with

necessary hardness.

The Navy Department also maintains the requirements as to
grade by requiring a steel which will stand rehardening a specified
number of times without cracking.

General Properties. — The following table shows the relative
toughness and hardness of tool steel of the different carbon contents :

Carbon,
Per Cent.

0 . 50 Toughness only.

0.60 Great toughness with properties suitable for hardening and

tempering.

0.70 Excellent toughness, but with cutting edge.
0.80 Tough tool steel, withstanding shocks, etc.
0.90 Good cutting edge but with toughness an important

factor.

1 . 00 Toughness and cutting edge about equal.
1 . 20 Great hardness combined with some toughness.
1 30 Great hardness in cutting edge. Toughness slight factor.
1.40 Extreme hardness in cutting edge first requirement.

Toughness slight factor.

Some metallurgists consider that it is safer to select a too hard steel
and draw the temper at a higher temperature than to choose a too

TOOL STEEL AND TOOLS

361

soft steel with a view to increasing its hardness by a weaker temper-
ing. Opposed to this is the fact that the higher the carbon
content the more the care which will be required in the harden-
ing operation, since the steel becomes more sensitive to
overheating.

GENERAL TEMPERING COLORS FOR TOOLS

Faint yellow: Steel-engraving tools.

Light turning tools.

Hammer faces.

Planing tools for steel.

Ivory-cutting tools.

Planing tools for iron.

Paper-cutting knives.

Wood-engraving tools.
Light yellow: Milling and other circular cutters for metal.

Bone-cutting tools.

Scrapers for brass.

Shear blades in general.

Boring cutters.

Leather-cutting dies.

Screw dies.

Inserted saw teeth.

Taps.

Rock drills.

Chasing tools.

Penknives.
Straw: Dies and punches in general.

Moulding and planing cutters for hardwood.

Reamers.

Gouges.

Brace bits.

Plane irons.

Stone-cutting tools.
Deep straw: Twist drills.

Cup tools.

Wood borers.

Circular saws for cold metal.

Cooper's tools.

Augers.

362

STEEL AND ITS HEAT TREATMENT

Brown. Drifts.

Circular cutters for wood.

Dental and surgical instruments.

Axes and adzes.

Saws for bone and ivory.
Peacock: Cold sets for steel and cast iron.

Hand chisels for steel and iron.

Boiler-maker's tools.

Firmer chisels.

Hack saws.
Purple. Moulding and planer cutters for soft wood.

Smith tools and battering tools generally.
Blue : Screwdrivers.

Saws for wood.

Springs in general.

These colors are for general crucible steel with low manganese.
Their applicability to particular work and special steels may be taken

1460

1440

•51420

>1400

1380

1360

Diameter in Inches
FIG. 199. — Temperature-size Curve for Hardening Tools.

in a general way, but that temperature must be adopted which will
suit the special work or steel in hand.

Hardening. — We have previously discussed the fact that with any
increase in the mass of the steel there is a corresponding decrease
in both the maximum surface hardness and the depth of hardness,
when quenched from the same temperature. This difference in
hardness is due to the difference in the rate of cooling of the small

TOOL STEEL AND TOOLS

363

and large sections. In order to produce the same degree of hard-
ness in a small and large section, as applied to small tools, it is neces-
sary to heat the larger section hotter for hardening than the smaller.
To illustrate: Matthews and Stagg have worked out the relation of
mass to temperature for one particular grade of the same tool steel
in which the sizes varied from Y& m- diameter to f in. diameter, and

1200 1300 1400 1500 1600

FIG. 200. — Loss of Hardness Due to High Hardening

1700 1800

Tempera tures . (Shore . )

found that a difference of about 60° F. in heating was necessary to
produce the same degree of hardness in the two extreme sizes. Their
temperature-size curve is given in Fig. 199.

The table on page 365 gives the approximate temperatures for
handling general tool steels. Two columns are given under harden-
ing temperatures as representing the best practice of two well-known

364 STEEL AND ITS HEAT TREATMENT

steel companies. As a general proposition, the lowest temperature
should be used for hardening which will give the desired results:
the use of abnormally high temperatures will increase the grain size,
weaken the steel, and reduce the hardness. These last factors
become even the more apparent with increase in the carbon content,
as is roughly illustrated by the scleroscope readings as given in
Shore's chart in Fig. 200.

On the other hand, on account of mass action and other individual
and distinctive shop conditions, it is difficult to set the upper limit
over which hardening should not be done. Certain classes of work
often require temperatures which might prove excessive for other
work; thus one instance has come to the author's attention in which
the hardening of certain 1 -^-in. rounds of 0.9 per cent, carbon stock
are hardened at 1600° to 1620° F. and 80 per cent, more service is
being obtained than from the same steel hardened at 1460° F.
Again, another well-known company hardens 0.9 per cent, carbon
steel of approximately the same size at 1370° F. and obtains better
service than when hardened at higher temperatures. Each case,
in other words, must be handled separately and those temperatures
worked out which will give the best solution of that particular
problem.

Distortion Factors. — Slender pieces of steel, when hot, will bend
under the application of a steady, even though slight, load. The
weight of the part being heated for hardening is often sufficient to
cause noticeable distortion if the tool is placed in the furnace in
such a manner that it is not carefully supported. For this reason,
such tools are best heated when held in a vertical position, with the
point of support at the upper end of the piece, the tool being so held
that it automatically comes to the normal position as will a plumb-
bob.

Distortion may be due to the initial condition of the steel, such
as may result from forging, rolling, machining, etc. Any strains
which exist in the tool previous to heating for hardening are relieved
when the piece is heated, but the readjustment of such strains may
cause a bending or twisting of the tool. In making the tool it is
advisable to rough down to within about ^ in. of the finished size
and then anneal in some non-oxidizing material to relieve the
machining strains. If the tools are not straight after annealing,
they should be heated, straightened while hot (do not straighten in
the cold), and then reannealed. The tools are then finished and
are ready for hardening.

TOOL STEEL AND TOOLS

365

LO >O O O iO »O

a> be ' b b o o o o b o o

O _ f*> ^ O O ^ O O (**) ^ O O

c3 c3 o3 c3 c3 o3 c3 c3 c3

O Pi CO *O *O iO »O CO CO 0^ CN

fe £ M T^ Tfi

OOO>O»OiOiO ^O >O

TtHr^COCOCOCOCO CO CO

00 t^* CO cO CO CO cO CO CO

COCOCOCOCOCOCO CO CO

CO CO CO CO CO CO CO CO CO
COCOCOCOCOCOCO CO CO

1> t^

J—t

~~Z ': i g

5 : : : •§ <§ i 1

^ ^ o S ^ -,

j bC 3 "*•* tiT

• «S' B «

be c8 c§ cj ^ o '

CO CO ^ ^^ C^ C^ «*-H »»-H «^H Q

r"O'"O'd'"C;pd'"O 3 jS'd 3 p

6 'a'a3'a3'a3ta3p73croHcicrx

^ ^ ^ £ £ ^ -^ -q * ^ °

:::::! o ^

- & -s ; ; i S 1

gS| 3 ^ ; ||

§ H -« °

» f-, O

O : o •*->

|| 1| I|| | |

B- di . -fi 6 « S w S K

efl-P

= o^

O O O O O O O

t> 00 C5 O rH CM CO
O C > O r-i r-I r-i r4

366 STEEL AND ITS HEAT TREATMENT

The influence of annealing or previous hardening operations
upon the change in shape, such as increase or decrease in diameter
and length, or combinations of these, is very difficult to foretell.
As a general rule, however, such changes are the more marked with
repetitive hardenings, and with increased percentages of carbon.

Changes in Length. — Most tool steel has a tendency to contract
in length upon hardening, and especially after previous annealing
or hardenings. This change, whether it is expansion or shrinkage, is
dependent upon the chemical composition and uniformity of the
steel, the grain size and influence of the mechanical elaboration
and annealing, the uniformity and amount of heat used in hardening,
the method and rapidity of cooling, and innumerable other variables.
The commercial application of heat-treatment principles when ap-
plied to fine tools for their standardization to exact measurements
must be along the lines of standardizing each step of the process.
The steel must initially be kept uniform in composition and physical
test; the limits of treatment must be held to a minimum difference:
and having somewhat accomplished these aims, the average change
of size for specific dimensions must be studied and the tool made
accordingly. Thus in the matter of taps, by obtaining the average
contraction lengthwise for a given size tap blank, under standard
conditions, the thread may be cut upon lathes having the lead screws
so adjusted that the pitch given to the tap before hardening will
just come right after hardening.

Changes in Diameter. — Assuming that the many other variables
affecting distortion might be reduced to a constant or standard, it
is a generally accepted condition that there is a relationship existing
between the amount of distortion by swelling after hardening and
the original diameter of the piece. Most slender tools, in ordinary
commercial practice, have a tendency to expand in diameter after
hardening. In factories where a standard steel and standard methods
are in vogue the amount of increase is a very important factor, as
is shown by the following data obtained from an exhaustive study 1
of taps:

Thus this particular problem was attacked with the view of
obtaining, under standard shop conditions, the average increase in
diameter due to hardening for each size of tap. With these results
it was then possible to determine the necessary angle diameter of
the tap before hardening, so that the hardened tap would meet the
final requirements.

i Woodward.

TOOL STEEL AND TOOLS

367

Diameter of
Tap.

Average Increase
in Diameter.

Diameter of
Tap.

Average Increase
in Diameter.

yjj inch

1^ inch

0 0025

0.00025

0.0025

0.0005
0.001

0.003
0.003

0.0015
0.002

0.0035
O.Q035

0.002

0.004

It was also found, however, that experiments upon the same steel,
under apparently the same conditions, showed that there may be
very great variations in the effect of hardening upon the diameter;
and when the various other factors are taken into account, the
difficulty of prognosticating exactly what is going to happen is even
more apparent.

Heating.— Much of the difficulty experienced through distortion
or cracking may be largely diminished by the proper application of
the principles of heating such as have been discussed elsewhere.
The heating should be done slowly, carefully and at a uniform rate.
In no case should the temperature of the furnace be greater than
the maximum temperature to which the steel is to be heated. After
the steel has been thoroughly heated, a further continuance will
only tend to weaken the steel by increasing the grain size. The
furnace should be of such design, construction and operation that
it shall be of uniform temperature over its whole hearth, shall heat
at a uniform rate, shall not be greatly affected by the introduction
of fresh charges, shall have a neutral or reducing atmosphere, and
shall be under exact control. Further, it is not only the temperature
to which the steel has been heated in the furnace that counts, but
also the temperature (and uniformity of temperature) of the steel
when it goes into the quenching bath,

HEAT TREATMENT OF TOOLS

In the following pages there is given a description of practical
methods of treating certain tools. It is not intended that the data
shall be comprehensive of all methods or of all tools, but shall give
practical hints and methods which may aid the man in the small
shop who may have such work to do but occasionally. It should
also be remembered that the ideas given are representative of a

368 STEEL AND ITS HEAT TREATMENT

type of treatment or tool, and which may find application for many
uses not given. The methods given are taken from practical work,
and have given satisfaction.

Chisels. — Chisels belong to that class of tools which can be
advantageously hardened without first grinding away the " skin."
Chisels are not hardened by heating up the whole tool, but only
applying the heat to 2 ins. or so of the cutting end; too short a
distance is detrimental, as the long, unhardened shank may bend
under heavy blows. In quenching, care should be used to avoid a
distinct line of demarkation between the hardened £nd unhardened
parts, as otherwise the end may break off in service. After harden-
ing the cutting end by immersing in water and at the same time mov-
ing the chisel up and down, the tool should be removed while suffi-
cient heat for tempering still remains in the shank. The cutting
edge is brightened with emery cloth. Temper colors will soon
appear in the brightened spot, which are caused by the heat from
the shank " running down " towards the cutting edge; the temper
colors may be " spread out " by holding the end near the fire if it
is desired to have a broad band near the edge. When the proper
temper color is obtained at and near the edge, the chisel should
be immediately plunged into water until cold in order to prevent
further softening.

The degree of tempering depends upon the steel which has been
used, the duties in service and upon the general experience and judg-
ment of the hardener. In general it may be said that chisels for
metal should be tempered to about a peacock color; for stone to a
purple color, although many stone and granite chisels are drawn to
only a light straw color; and for soft material to a blue color. If
a very tough steel is desired, the chisels may be given a double
tempering by heating to the temper color twice, the first color
being rubbed off after quenching. The lower the carbon used the
lower should be the tempering; in fact, a 0.45 per cent, carbon steel
is sometimes used after hardening without subsequent tempering.

An excellent plan in treating chisels is to oil treat the blank
before forging out, quenching the steel in oil and from a temperature
sufficient to harden it well. This will stiffen the shank and keep the
head from upsetting during use.

The problem of producing a good chipping chisel is simple,
and yet has met with more difficulties than the majority of other
such tools. The correct adjustment of temperatures plays a most
important part in this work. For forging, the steel should be heated

TOOL STEEL AND TOOLS 369

up gradually and not much beyond a bright-red heat.1 The work
should be done rapidly, aiming to obtain the greatest reduction to
size while the steel is yet near its first heat; as the temperature falls,
the blows should be lighter and quicker. It has been the author's
experience that added toughness is given to the steel if these light,
quick blows are carried on until the steel neare a dull-red heat. If
reheating is necessary, adopt the same care in raising the temper-
ature as before. The chisel should be immediately hardened and
tempered after forging when possible. If the pressure of work
will not permit of this, the chisel should be allowed to cool as slowly
as possible by sticking the hot end in dry dirt; this latter will tend to
eliminate the cooling strains which might otherwise be set up. For
hardening heat slowly to the lowest temperature at which the steel
will harden (generally about 1350° to 1400° F.), allow the heat to
penetrate and harden as usual. If care is used, and sufficient
thought has been given to the design of the chisel, a hundred per
cent, improvement may be easily obtained over the ordinary " hit
or miss " method.

An open-hearth steel which has given unusual service for pneu-
matic chipping chisels is as follows: Carbon, 0.90 to 1.00 per cent.;
manganese .50; phosphorus and sulphur low; chrome, 0.50 per
cent. The hardening temperature of this steel is about 1400° F.

Die Blocks. — The problem of satisfactorily treating die blocks is
one which will try the hardener's knowledge and patience, increas-
ing in difficulty with the intricacy of design and size. Primarily
the requirements of any die are (1) a hard face, (2) sufficient depth of
hardening to prevent the impression from sinking, and (3) a tough
back or body to take up the shock or blow. Further, the hardening
must be conducted in such a manner as will prevent any change of
size — such as warping — besides producing a clean, sharp impression
free from scale, pitting or checks. These fundamental requirements,
assuming that the right kind of steel is used, involve the factors of
proper heating and the most suitable method of cooling.

1 For the convenience of the tool hardener who has to work without a pyrom-
eter and must gauge temperatures by the eye, the following table of approxi
mate heat colors in moderate diffused daylight is given :

White 2200° F. Cherry or full red 1375° F.

Light; yellow 1975 Medium cherry 1250

Lemon 1825 Dark cherry 1175

Orange 1725 Blood red 1050

Salmon 1650 Faint red 900

Bright red 1550

370 STEEL AND ITS HEAT TREATMENT

More die blocks are warped or cracked through improper heat-
ing than through any other cause. It is absolutely essential that
the entire mass of the steel of the block shall be heated carefully,
uniformly, through and through, to the proper temperature.

It is always best to allow the die to heat up with the furnace so
that any strains which may exist in the steel may be removed grad-
ually and give the mass of steel ample time to adjust itself to the rise
in temperature. If the furnace for heating for hardening is already
at or near the quenching temperature when the die is ready for
heating, it will be advisable carefully to preheat the die in a separate
furnace. Equipment is subsequently described for automatically
preheating and full-heating in the same furnace. In no case should
the temperature of the heating furnace be greater than the maximum
temperature from which the steel is to be quenched. The practice
—unfortunately common — of forcing the furnace in order more
quickly to heat the steel is to be strongly condemned ; such practice
must inevitably result in the overheating of the edges or corners
of the die and produce unsatisfactory results. The extra time
spent in careful and uniform preheating, and in all subsequent heat-
ing operations, will be well worth the expense.

The exact temperature best suited for hardening may be men-
tioned here only in a general way. The chemical composition of the
steel, the size of the die block, the depth of hardness required, the
condition of the steel before hardening, and many other factors
must be taken into consideration. In general, nickel and chrome
nickel steels may be quenched at lower temperatures than those used
for the corresponding carbon steels, while vanadium and chrome
vanadium steels usually require higher temperatures. Again, the
greater the mass of the steel, and the greater the depth of hardness
required, the higher is the temperature for quenching. Some die
hardeners even find that a temperature which will produce a slightly
coarse grain is advisable for certain classes of work, such as dies
for cold forming under a heavy drop. In other words, the proper
temperature for hardening must be determined by experiment, but
the lowest temperature which will produce the desired results is
always the best.

The next, and a vastly important factor, is the duration of heat-
ing at the predetermined temperature for hardening. The point
to be emphasized is that the mass of the steel must be uniformly
heated throughout — and this takes time. Not only must the outer
sections attain the necessary degree of heat, but also the very

TOOL STEEL AND TOOLS 371

center of the die. Disregard of this fundamental principle is the
basis of a large proportion of the failures through cracking or warp-
ing, and which are so often attributed to " the steel is no good."
If the die block lays directly on the hearth of the furnace (and this,
it might be mentioned, is not the best practice, since any work is
best heated when the hot gas currents can circulate entirely around
it), the penetration of the heat may be roughly determined by
moving the block to one side and noting the color of the space va-
cated; if this area is not of the same color as the rest of the furnace
floor the heat has not thoroughly penetrated to the center of the
die. This test, called the " drawrouagh " by old English hardeners,
should be followed even though the play of heat colors over the ex-
posed portions of the block appear uniform.

For protection of the impression from oxidation through contact
with the air, the faces of the dies may be packed in carbonaceous
material. One of the large manufacturers of silverware packs his
dies as follows: A small sheet-iron pan, about 2 ins. high and
about an inch or so wider than the die all around, is partly filled
with granulated animal charcoal or bone. The die is then pressed
firmly upon the charcoal, forming an impression, and is then care-
fully removed. This impression is sprinkled with powdered animal
charcoal, and with very fine steel filings. The die is carefully
replaced in position, surrounded with more granulated animal char-
coal to the height of the pan, and the space between the top of the
pan and the die carefully luted with fire-clay. Upon heating, the
filings and powdered charcoal fuse together upon the surface of the
steel, forming a protective coating which eliminates oxidation during
heating, but which is washed off during the quenching. A little
brushing with oil and emery powder will immediately produce a
clean, bright surface. Another method is first to paint the surface
of the die with a thick paste made of linseed or cottonseed oil and
powdered bone-black; the die is then placed in a shallow pan upon
a half-and-half mixture of fresh bone and powdered charcoal, in a
suitable pan or box, which is then filled and luted as above.

The old-time method of an immediate and total quenching of
the block until it is quite cold should be attempted only with the
simplest forms and small sizes of dies. Large blocks have a great
tendency to warp, bulge, or even crack if a total immersion is adopted,
this being caused by the unequal contraction of the metal of the
surface and of the core. It may be said, however, that this difficulty
may be largely avoided if the block is previously given a special

372 STEEL AND ITS HEAT TREATMENT

treatment consisting of oil quenching from just over the critical
range, followed by an annealing at a temperature just under the
lower critical range. Total quenching also should not be used if an
extremely hard face is desired, since the heat cannot usually be
removed quickly enough.

The best practice for hardening large die blocks consists in first
carefully preheating the die, then slowly raising it to the hardening
temperature and allowing it to soak at this temperature until it is
thoroughly heated. For this work a furnace should be used in which
accurate temperatures and uniform heating can be obtained. Large
blocks may be most easily handled by the use of a hoist on a swinging
run-way, mono-rail or overhead crane, and equipped with tongs
or "dogs." The dogs fit into holes which have previously been
drilled in opposite sides of the block about half way between the
upper and lower faces.

When the block is properly heated, it is removed to the front of
the furnace, gripped with the dogs, run over to a position above the
quenching tank, lowered face-downwards entirely into the water or
oil for a few seconds (to prevent warping), and then raised out of
the quenching bath until immersed about 1 in. deeper than the
depth of the deepest impression in the die. The surface of the bath
should be kept in motion, or else the block should be slowly raised
and lowered a little so that there will be no one line of hardening.
The hardening will be greatly increased if a stream or heavy spray
of water (assuming water to be used for hardening) is directed
against the face of the block, or into the impression. In the case of
blocks which contain a deep impression, such as are used for certain
classes of gears, etc., it will be necessary to have a stream of water
thus impinge upon the impression in order to harden it; the face
of the block will take on great hardness, and the heat from the
unsubmerged part will gradually be drawn out.

When the face of the block is entirely cold, and the majority
of the heat taken out of the other portion of the block (usually at
about a very dark red, but dependent upon the size of the block),
it is raised out of the water, reversed to face up, and brightened with
emery paper. The heat in the hot part of the block will gradually
temper the hardened face. When this approaches a good straw color
the block is immersed in water or oil until cold; in some instances
where a softer block is desired, the block may be allowed to cool in
the open without the use of water to stop the temper. In case there
is not sufficient heat left in the block after hardening to bring out

TOOL STEEL AND TOOLS

373

the desired temper color, the block may be stood in front of the fur-
nace, back to the heat, or placed on a hot bar of steel, or laid on top
of a low smith fire, until the proper temper is reached. Die blocks
will generally give more uniform service when drawn in a tempering
bath when such is possible. In case the block is of intricate design
and requires very particular tempering in weak spots, this may be
done by the local application of heat by means of hot plates, etc.

FIG. 201. — Intake End of Special Furnace for Hardening Forging Dies.
(" Machinery.")

Die blocks hardened and tempered as directed above should pro-
duce a strong, tough base and core, increasing in hardness as the
face is approached.

Semi-automatic furnaces for heating die blocks for hardening,
in which the preheating as well as the final heating are done in the
same furnace, are illustrated in Figs. 201 and 202. Four runways,
filled with 3-in. malleable-iron balls, extend throughout the length
of the heating floor of the furnace. Castings which fit over the

374

STEEL AND ITS HEAT TREATMENT

balls in two of these run-ways and which are of suitable size to carry
one of the dies, are placed in position in the end of the furnace shown
in Fig. 201. When the cold die is placed on this casting, as shown at
0, one of the pneumatic pushers P is brought into play and the cast-
ings act as a cart, carrying the die into the furnace. By following
the first casting and its die with others, the furnace is gradually
filled, the ram pushing the whole line of dies further into the furnace
with each new addition. The furnace is so designed and operated
that the temperature at the charging end is low, but gradually

FIG. 202.— Quenching and Tempering Dies. (" Machinery.")

increases up to the maximum near the other end of the furnace;
preheating and the final heating are thus obtained in the same
furnace. Each furnace is double tracked and heats two rows of
dies at once. At the end of the furnace shown in Fig. 202 the
hot dies come out on the extension of the run-way marked Q. The
faces of the dies are turned downwards so that the dies may be
picked up by the traveling crane and lowered into the quenching
tank, as shown at • R. A stream of water also plays against the
impression, as usual. The hot plate shown at T is used for the
tempering.

TOOL STEEL AND TOOLS

375

Engraved dies for spoons, forks, knives, etc., are treated at one
plant by the following method. After packing and heating as
described in a previous section, the dies are quenched face up in
water at a temperature of about 70° to 80° F., to a depth of within
about \ in. of the face. Water at this temperature seems to give the
best results in this particular instance — colder water is too harsh,
while warmer water does not sufficiently distribute the strains nor
give sufficient hardness. As soon as the cooling effect just begins
to creep towards the face of the die, and which only takes a few

FIG. 203. — Method of Hardening Engraved Die.

seconds; the die is immediately wholly immersed in a vertical posi-
tion in the water, with the impression turned toward a heavy stream
of water which impinges directly upon it. The arrangement of the
quenching bath is shown in Fig. 203: the die (a) rests upon a wire
platform (6); the water is supplied under pressure through a IJ-in.
pipe (c), flowing out through a J-in. slot (d) which extends from the
level of the die support to the top of the pipe. The die remains in
the water bath until the " singing " has stopped, about 50 to 90
seconds, and is then cooled in oil until cold. The hardened die is
later ^tempered in oil to 435° F.

376 STEEL AND ITS HEAT TREATMENT

Many alloy steels have been experimented with in recent years
for the purpose of increasing the production of forgings from a
given impression, thus avoiding the loss of time and expense incurred
in redressing the die-blocks. A chrome nickel steel containing
about 0.50 to 0.60 per cent, carbon, 0.50 per cent, chrome and 1.50
per cent, nickel has been found to give most economical results.
These die blocks are hardened and tempered in the usual way,
using a temperature of 1400° for the hardening heat. If the carbon
content runs above 0.60 or 0.65 per cent, it has been the author's
experience that cracking during or directly after hardening may
result. These blocks are greatly improved, not only in the length
of service to be obtained, but also in the elimination of warpage
during hardening, and of danger of cracking, by giving the block,
before machining, a full heat treatment and toughening or annealing ;
blocks which approximate the composition noted should be quenched
in oil from a temperature of 1400° to 1450° F., and then full annealed
at about 1250° F. Such a treatment gives excellent results, and will
also show up any defects such as pipes, seams, etc., before the expen-
sive machine work has been done.

Dies used in engraving work, and in the jewelry and optical
trades, must have a glass finish, both in smoothness and in hardness.
If subjected to the usual quenching, followed by sand-blast, acid
bath or cyanide, a large amount of stoning and polishing would be
required. This may be obviated by the use of borax or boracic acid
in the following manner. Fill the matrix with powdered boracic
acid and place near a fire until it melts, which temperature is con-
siderably below the tempering point or color of the steel. Follow
this with a second addition of boracic acid and then harden as usual.
Although the salt will generally come off in the quenching, it pro-
tects the polished surface of the die and does not interfere with the
hardening. In case the salt does not come off in quenching, it may
be easily removed by live steam or boiling water. The hardening
may be done by complete or partial submersion, depending upon the
thickness and general design of the die. Engraving dies are usually
tempered to a light straw color.

Drills. — For occasional work in hardening drills, the following
procedure may be used : If an open fire is the only available source
of heat for hardening, the points of the drill should be kept out of
the hottest part of the fire at first, drawing them in as the upper
parts become heated. The heat should extend over a considerable
portion of the drill. Quench vertically in water, and keep the drill

TOOL STEEL AND TOOLS 377

moving up and down so that there is no abrupt line of demarkation
of the hardening. If the drill is held quietly in the water, fracture
across the water line is a common occurrence when the drill is placed
in service. Allow the drill to remain in the water until the im-
mersed part is entirely cold. Remove, brighten, and allow the heat
in the shank to run into the hardened part until a dark straw color
appears on the cutting edge. The drill should then be immediately
and entirely immersed in water. If there is not sufficient heat in
the shank to bring out the temper color, use hot ashes, or similar
means. The drawing operation upon hardened drills should pref-
erably be carried out in an oil or salt bath subsequent to straight-
ening; drawing expensive tools to color is poor practice.

It is always advisable, however, if an open fire must be used for
heating, as noted above, to heat the drill in a pipe or tube to prevent
the direct contact of the fire and the steel, or with charcoal to prevent
oxidation. The heating should be done slowly, uniformly, and to as
low a temperature as is possible and consistent with the desired
results.

In cases where a large number of drills are to be hardened, it
is advisable to use a special hardening tank. The shape of the
lands of the drill is such that the steam formed by the contact of
the water and the hot metal will in many instances prevent the
water from penetrating to the flutes and properly hardening them,
besides having a similar influence on the end of the drill, which will
become the new cutting edge as the point is ground back. This
buffer or blanket of steam may be eliminated by maintaining a
constant flow of cold water into the grooves and against the end
of the drills. Perforated pipes may be placed up the sides of the
quenching tank, and through which the cold water is forced into the
grooves; similarly, a jet from the bottom strikes against the end of
the drill.

For drills for holes under J in. in diameter, the hardening heat
should be allowed to penetrate only through the cutting part. The
drill should then be quenched entirely and the temper drawn to suit
the work. The reason for not allowing the hardening heat thor-
oughly to penetrate to the core of the drill is that sudden quenching
of a small, slender piece might cause severe strains to be set up in the
steel; such drills also require a tough core to be able to withstand
the torsional effect in the actual drilling operation. Most of the small
drills are quenched in oil. The temper color is usually a dark straw.
If the tempering is accomplished by placing the drills upon a heated

378 STEEL AND ITS HEAT TREATMENT

bar, the cutting parts must be allowed to project for some distance
over the edge of the hot bar, for otherwise the heat will be too sud-
denly applied.

Milling Cutters. — Under this class are included cutters of varying
description, such as milling cutters, forming cutters, slotting cut-
ters, angle cutters, etc. This consists, in general, of a cylindrical
piece of steel with a bore through the center, and teeth on the cir-
cumference, sides, or both. The unequal forces of contraction
and expansion affect these tools to a large extent. In designing
a cutter, as large a mandrel hole as is possible should be used, as
larger holes will permit the steel to be hardened more uniformly.
If the mandrel holes are standardized, large cutters may have a
part of the sides (in the absence of side or angular teeth) dished or
paneled out at the place which would tend to garden last, that is,
half way between the two circumferences.

Great care should be used in heating milling cutters for harden-
ing. The heating atmosphere should be neutral or slightly reducing
to protect the teeth. If an open fire is used, the fuel should not be
allowed to come in contact with the cutter: this may be done
by resting the cutter on a fire-brick or plate. If a hearth furnace
is used, the cutter should not touch the floor or walls of the furnace,
but should be supported by fire-bricks or other suitable methods.
If tongs are used in handling, care must be used so that the tongs
do not touch the cutting edges; the use of wires is better practice.
If the cutter is supported on bricks, or laid on plates, it must be
turned repeatedly in heating so as not to leave any unevenly heated
spots. The cutter may be conveniently held in the quenching bath
by using a small round bar which has three prongs welded to one end,
and which extend at right angles to the axis of the bar, by slipping
the other end of the bar through the mandrel hole of the cutter ; the
latter will rest on the prongs, and then can be conveniently lowered
into the quenching bath.

Ordinary cutters are best hardened by the use of two small cir-
cular plates of a diameter slightly greater than that of the cutter,
and with holes bored through the center corresponding to the size
of the mandrel hole of the cutter. One plate is placed on each end
of the cutter, and the whole placed on the suspension tool as de-
scribed above and immersed vertically in the quenching bath.
By the use of these plates, the hardening will affect the steel along
the entire length of the teeth and at right angles to the center line
of the cutter. This will also eliminate the circular fracture or

TOOL STEEL AND TOOLS 379

flaking of the teeth which so often characterizes milling cutters
subjected to uneven cooling. While in the quenching bath, the cut-
ter should be moved up and down and not from side to side ; this will
permit the solution to pass through the center hole and give an evenly
hardened core. The combined use of water and oil (" broken hard-
ening ") in the following manner is good for hardening for large
cutters: quench in water until the " singing " caused by the water
boiling on the hot steel has stopped, and then immerse in oil until
cold; warm the cutter in boiling water to relieve the strains and
temper when convenient. Pack-hardening is also used to some extent
for milling cutters in order to prevent oxidation; in this case each
piece should be quenched separately. Salt baths and lead baths are
also used for heating. One of the main points to be observed in
quenching milling cutters is that long cutters should be plunged
vertically and thin ones edgewise.

The tempering of milling cutters is often done by the insertion
of a hot rod through the mandrel hole and revolving the cutter on
it until the proper temper color is obtained. The most satisfactory
results are to be obtained with the use of an oil bath, as an even hard-
ness can be best obtained in this manner. Small cutters are tem-
pered to a light straw color, or yellowish-white. For medium-sized
cutters a good straw color may be used. Very large cutters, on
account of the lesser effect of the hardening, may not require temper-
ing, but it is always advisable to heat them in boiling water to make
them uniform and remove the hardening strains.

For hollow mills it is not necessary to heat for hardening very
much above the teeth, as it is not required that the back should be
hard. Harden with the teeth upwards, working the piece up and
down in the quenching bath to get the solution circulating through
the hole.

T-slot milling cutters should be hardened, not only through the
cutting portion, but also through the entire length of the neck,
especially if this is of small diameter. In tempering, the cutting
portion should be drawn to a straw color and the neck to a blue color.

Files. — Before the file blanks can be ground and the teeth cut
it is necessary to anneal the steel. This is often accomplished by
packing the blanks in air-tight oblong boxes and annealing at about
1300° to 1400° F.

Lead baths continue to be most used as the heating medium.
Salt baths have been tried with varying degrees of success, but in
the main have proven unsatisfactory. This is due in a large measure

380 STEEL AND ITS HEAT TREATMENT

to the fact that oxide of iron (scale) may settle upon the teeth of the
file, causing soft spots when hardened. The method of dipping the
file into a solution of ferrocyanide and allowing the coating to dry
upon the surface of the steel before heating has been tried. The
objections to the use of this method are that a decomposition of the
ferrocyanide will yield additional iron oxide and poisonous fumes.
Other salts of a harmless character have been tried with little success.
The general procedure is to cover the file with a paste which pro-
tects the edges of the teeth in the hardening process, heating in lead
to the proper temperature (about 1400° F.) and quenching in water
in a vertical position. One file-maker uses a paste made of the
following base: ground charred leather, 2 parts; table salt, 4 parts;
and flour, 3 parts. The file is given a coat of this paste, which is
allowed to dry before heating. It is said that the melting-point of
this paste will give the proper hardening temperature. After being
hardened, and while the file is still warm, it is put through the final
straightening process.

Half-round files require particular attention on account of their
tendency to warp : before hardening, the file is bent back on a fixed
template of such form as experience has shown will bring the file to a
true line upon hardening; the file is placed again in the template
before it is quite hard, strained to the proper degree, and water is
thrown on the upper surface of the file to make it quite cold before
the strain is relieved; the file is then entirely quenched and will
usually return " to the true " after the final hardening.

After the final straightening the files are " scrubbed " to remove
the paste, and are then washed in lime water and dried by holding
them in steam. The tang is then toughened or " blued " by dipping
it into a special bath maintained at the proper temperature.

File steel will vary in carbon from 0.90 to 1.60 per cent., accord-
ing to the size, shape and use of the file; manganese under 0.40 per
cent.; low phosphorus and sulphur; and in the case of exceptionally
good files, a small percentage of chrome. Nickel is generally con-
sidered as detrimental to files.

Reichhelm 1 shows the detrimental effect of heat variations in
hardening in the microscopic photographs of two fractures of the
same file magnified 160 times. This file is one of the highest grade
produced in Europe, and Fig. 204 shows the fracture of this file as
imported, while Fig. 205 shows a fracture of the same file, a section
of which was rehardened, after the exact degree of heat required
1 " Machinery/' Dec., 1914.

TOOL STEEL AND TOOLS

381

FIG. 204. — Photomicrograph of High-grade Foreign File.
(Reichhelm.)

X160.

FIG. 205.— Photomicrograph of Same File Rehardened.
(Reichhelm.)

X160.

382 STEEL AND ITS HEAT TREATMENT

for this particular steel had been experimentally determined. Fig.
204 therefore shows the result of the best hardening practice in
Europe, aided by the pyrometer, while Fig. 205 shows the hardening
of this identical file by the correct heat automatically maintained.
That any number of files, or tools of any kind, can be hardened so
as to show uniformly the excellent fracture exhibited in Fig. 204 is
due to automatic heat control, as has been demonstrated conclusively
in daily practice for over three years.

Both of the photographs of fractures have been pronounced
excellent by competent judges, but the decidedly finer grain and
more even diffusion of the carbon shown in Fig. 205 produced a
difference in the durability of the file teeth of nearly 50 per cent., as
compared with the section of the file as originally hardened and shown
in Fig. 204.

Punches and Dies. — Similar to all round tools, punches show a
great tendency to flake off at the corners, sometimes a whole ring
breaking off. Assuming proper heating, this may be overcome to
a large extent by means of a water spray. Dies of intricate shape
and possessing sharp angles should be most carefully handled. It
is often advisable to fill these angles with a little putty or fire-clay
to lessen the hardening effect and prevent the formation of quench-
ing strains at right angles to the diagonal. A piece of binding wire
may also serve for this purpose. Dies should generally be quenched
flat, depending upon the shape of the piece. Small punches should
not be quenched in real cold water on account of the liability to
cracking under sudden cooling — an oil bath or lukewarm water
is far preferable. Dies or ,any press tools having holes near the
edge should always have these holes filled with clay in order to pre-
vent cracking or too great hardening; graphite or asbestos may also
be used for plugging the holes for stripper or guide screws. Punches
and dies are generally tempered to about a straw color, the depth of
this varying according to the thickness and hardness of the material
to be punched. The tempering may be carried out by setting the
hardened pieces in front of a hot furnace, laying on hot plates, in
oil baths or in hot sand.

Reamers. — Reamers may be heated in lead to protect the cutting
edge from the direct action of the heat and oxygen. The lead may
be prevented from sticking to the tool if the latter is brushed, in
the case of small reamers, with a little soft soap. Larger reamers
may be protected with a paste made of black lead and water or
lampblack and linseed oil, both of which should be allowed to dry

TOOL STEEL AND TOOLS 383

on the tool before heating for hardening. If the reamer has been
hardened by the use of water alone, and is larger than f in.
in diameter, it is advisable to hold it over the fire directly after
being removed from the hardening bath, or to set it in hot water
for a few moments, in order to remove — as far as possible — the
strains which have been caused by the hardening process. This
should always be done in the case of shell reamers and other special
reamers of any considerable size, whether the quenching medium
has been oil or water. Broken hardening is most excellent for tools
of this description. Large fluted reamers require to have only the
ribs heated to the proper temperature, and then quenched; temper-
ing will not then be required. Ordinary fluted reamers are tempered
to a yellowish white or very light straw color. Six-sided or eight-
sided reamers may be tempered to a light straw color. Square
reamers, triangular reamers and half-round reamers may be tem-
pered to a dark straw color, due to the fact that they take hold of
the work more deeply and might break if not tempered a trifle
softer.

Half-round reamers should not be quenched vertically, but with
the half-round side at an angle of 20 to 45 degrees to the surface of
the bath. If half-round reamers should be quenched vertically,
it will be necessary to move them in a horizontal manner in the
direction of the half-round side at the same time as immersed ver-
tically.

The shanks of reamers, taps, drills, broaches and similar tools
may be toughened by local lead tempering.

Rings. — Rings, collars arid hollow tools comprise a class which
require great hardness in the inner circumference or bore. Quench-
ing is usually done by means of allowing a full stream of water to
flow through the bore if it is quite small, or in the case of tools with
larger bores the insertion of a small pipe with a series of holes in its
circumference and through which a continuous stream of water may
be forced, forming a spray. In the first case it is advisable to set
the tool upon an asbestos-covered washer in which has been cut a
hole slightly larger than the size of the bore of the tool and then
apply the flange end of the water-supply line or pipe to the other
opening. Rings or collars requiring resistance to frictional wear
require no tempering. Eccentric rings cannot be quenched as usual,
as the relative thickness and thinness of the opposite sides would
tend to give unequal expansion and contraction and cause the hole
to become oval-shaped. This may be overcome by binding a small

384

STEEL AND ITS HEAT TREATMENT

piece of iron or steel to the thin side, heating, and quenching ver-
tically.

Rivet Sets. — Rivet sets should never be quenched directly by
immersion, as this will tend to make the edges of the cup break off,
the center to remain soft, and leave a line of great weakness between

FIG. 206.— Rough Method of Hardening a Rivet Set.

the hardened and unhardened parts. A simple and proper method
is to hold the cup under or over a stream of water so that the latter
will impinge directly upon the bottom of the cup, as shown in Fig.
206. If there are numbers of rivet sets to be hardened, an arrange-
ment of clips or holders under each tap or spigot may easily be set
up. The tempering may be carried out as in the case of chisels
(permitting the heat in the shank to temper the cup) or the shank

TOOL STEEL AND TOOLS 385

may be placed in a lead bath and the color allowed to run up into
the cup; the rivet set should then be entirely quenched to prevent
further softening.

Brearley makes the following points, which are of great interest.
Rivet sets may have a short life due to the wear on the head, which
is as often a failure as that produced by actual fracture. This is
pronounced in the case of annealed stock. He advises hardening
the head in oil before hardening the cup. Upon reheating for hard-
ening the cup, and tempering, a steel of great toughness is obtained,
which neither splits nor forms a mushroom head.

Saws. — Saws may be hardened by either of two methods — direct
immersion, or press or roll hardening. Circular saws may be heated
by enclosing in a sheet-iron case or box between layers of charcoal.
Sufficient space for expansion must be allowed to eliminate chance
for buckling. Saws may also be heated on the hearth (if level) of
any type of hardening furnace; it is advisable, however, to rest
the saw on an iron or steel plate so that the heating may be gradual
and uniform. The greater part of the secret for the successful hard-
ening of saws without buckling is a slow and careful heating. The
saws when heated to the proper temperature may be taken out
separately with tongs or a J-shaped hook. For direct quenching
they should be immersed edgewise and in a perfectly vertical position.
It is better to have a thin layer of oil on the surface of the water
bath, as the oil will ignite when the hot saw enters it, forming a
thin, protective coating on the saw7 and thus lessening the risk of
fracture. Oil alone, or oil with tallow dissolved in it will give suffi-
cient hardness for thin saws. The saws may also be placed between
lumps of tallow. The latter (tallow) is a better hardener than
oil, and therefore gives a greater and deeper hardening. Thin cir-
cular saws, and all ordinary saws such as hack saws, hand saws, etc.,
may be most satisfactorily hardened by means of a press. A com-
mon and inexpensive method is to have two cast-iron plates hinged
together, with the inner surfaces well oiled with a heavy oil. The
hot saw is placed between the plates, which are then clamped to-
gether and held until the saw is cold. Thin band saws are often
hardened by means of rolls. Circular saws for metal cutting should
be tempered to a dark purple color, or to a light blue for wood cutting.
Hack saws require tempering to a light purple color,

CHAPTER XVII
MISCELLANEOUS TREATMENTS

THE following examples and discussions of certain heat-treat-
ment methods have been selected in an arbitrary manner as repre-
sentative of distinct classes of work. Many others might just as
well have been taken, but the author feels that those selected will
perhaps illustrate in a general way some of the many problems
which arise in the course of ordinary heat-treatment work.

GEARS

Gear-steel Classification. — Automobile and similar machine
gears may be broadly grouped according to the method of heat
treatment, which, of course, is dependent upon the composition of
the steel. Thus the three classes are:

(1) The case-hardened gear, using a steel of low-carbon content
— generally less than 0.25 per cent — and depending upon the case-
carburizing process to give an outer layer of high-carbon steel and
upon the subsequent hardening processes to produce the necessary
wearing surface of sufficient hardness.

(2) The oil-hardened and tempered gear, using a steel of the alloy
type of about 0.45 to 0.55 per cent, carbon.

(3) The hardened gear (without subsequent tempering), using a
steel of an intermediary carbon content — about 0.30 per cent.

Requirements of Gears. — All high-duty gears require that the
steel shall be readily forgeable and machineable, and that after
treatment it shall have the greatest possible hardness with the least
possible brittleness. In this connection it may be said that surface
hardness is often more desirable than tensile strength, while the
question of brittleness is very important on account of shocks.

Case-hardened vs. Oil-tempered Gears. — The merits or
demerits of each type depend largely upon the point of view and the
personal experience of the user. Expert opinion may differ widely,

386

MISCELLANEOUS TREATMENTS 387

as is shown by the following excerpts from addresses by two well-
known metallurgists. One says:1

" Several years of observation and contact with the trade leads
me to prefer the case-hardened gear. The result of direct tests
upon thousands of gears of both types leads me to the following con-
clusions: (1) The static strength of a case-hardened gear is equal
to that of an oil-hardened gear, assuming in both cases that steel
of the same class and approximate analysis has been used and that
the respective heat treatments have been equally well and properly
conducted. (2) Direct experiments proved that the case-hard-
ened gear resists shock better than the oil tempered. (3) As regards
resistance to wear the same type is incomparably better, although
perhaps not as silent in action.

" One of the leading makers of gears has proved this to his own
satisfaction of late by an arrangement of shafts and gears whereby
energy is transmitted through two case-hardened gears, in mesh
with each other, to two oil-hardened gears. The gears are of the
same size. The conditions of the test were severe. Five sets
of the oil-hardened gea'rs have already been worn out, while the
original case-hardened gears are still in service and show the tools
marks.

" Upon the part of many there is a strong objection to case hard-
ening. In nine cases out of ten this is doubtless due to the fact
that the case-hardening operation has not been reduced to a science.
The depth of case, the relation of case to core, the time and tem-
perature to produce certain results and the exact control of these
conditions, together with an accurate knowledge of the material to
be treated, are factors that enter into successful case-hardening
practice. Further points in favor of this method are easier machin-
ing of the blanks, and at least equal static and dynamic properties
with kss chance of injury in hardening."

Then here is the opposing argument:2 " For machine tools,
hardened high-carbon alloy steel gears appear to be preferable to
case-hardened gears for a number of reasons :

" 1. Physically they are stronger and tougher and should there-
fore be better able to resist sudden impacts and extraordinary
loads. They do not show by file and scleroscope test the same

1 J. A. Matthews, " Alloy Steels for Motor Car Construction," Journ. Frank-
lin Inst, May, 1909.

2 From a paper by J. H. Parker, before National Machine Tool Builders'
Assoc.

388 STEEL AND ITS HEAT TREATMENT

degree of hardness as case-hardened gears, but, nevertheless, with
proper design, the dense-grained gear-tooth resists wear more satis-
factorily, as was demonstrated recently by the examination of a
motor-car transmission that had covered over 100,000 miles. The
high-carbon steel gears in this car still showed the original tool marks.
Not long ago a designer of machine tools commented on the ap-
parent softness of some hardened high-carbon gears, but found after
several months of hard service that they still showed tool marks,
thus proving hardness ample for wear.

"2. In service, especially for clash gears, the superiority of
these gears is most marked. On the clashing faces, case-hardened
gears are likely to have the hard case chipped off, thereby exposing
the soft core to the impact of clashing. The hard chips fall into the
gearing and may find their way into bearings, thus causing trouble.
High-carbon steel gears with a uniform hardness throughout do not
chip, nor do they ' dub over.'

" 3. The heat treatment of high-carbon steel gears is much
simpler than that required for proper case hardening. It is shorter,
less costly and produces a more uniform product, and as the gear
is heated but once for hardening, as compared with three times for
case hardening, the finished gear is certain to be freer from warpage.
The cost of proper case hardening is not generally appreciated, but
it has been found that a case-hardening steel must cost three to
four cents per pound less than a regular high-carbon hardening
steel, if finished gears made from both materials are to cost the
same.

" With all heat-treated gears, little points in design are impor-
tant. The gear-teeth should not be undercut, for if the section at
the root-line is smaller than at the pitch-line, greater hardness and
brittleness is produced where least desired. Great differences in
section should be avoided wherever possible, so as to do away with
excessive warpage. Sharp edges and angles, even in key-ways, are
the cause of internal hardening strains which frequently result in
failures; hence, wherever possible, a fillet should be used in place
of a sharp angle."

Case-hardened Gears— Treatment. — The steel for a case-hard-
ened gear should be low in carbon, preferably under 0.25 per cent.;
should be carburized so as to produce a case of a depth of about
^T or ^ inch and contain a maximum carbon concentration of
about 0.9 per cent.; and should then be suitably heat treated.
Since the principles of case hardening have been described elsewhere,

MISCELLANEOUS TREATMENTS 389

it will be necessary here only to outline the process, which is as
follows :

(Gear blank).

1. Anneal.

2. Rough machine to approximate size
(3. Light re-anneal.)

4. Finishing machine.

5. Carburize at about 1600°-1650° F.

6. Cool slowly in carburizing box.

7. Reheat and oil quench from 1550-1625° F.

8. Reheat and oil quench from 1350-1425° F.
(9. Temper, if desired, to not over 400° F.)

The temperatures given are only approximate, depending upon the
analysis of the steel, the mass of the steel, the results desired, and
various other factors. Nos. 3 and 9 may be omitted if desired.

Oil-hardened Gears — Treatment. — For the higher-carbon steels
used for oil-hardened gears it is always advisable to give the gear
blanks a preliminary treatment to develop the highest qualities of
the alloy steels and the greatest uniformity in their physical
properties. This treatment will also give the greatest " softness "
of which the steel is capable. This preliminary treatment (before
machining) is:

1. Quench in oil from about 150° to 200° F. over the criti-

cal range.

2. Quench in oil from about 50° F. over the critical range.

3. Anneal at a temperature about 75° F. under the criti-

cal range.

If this preliminary treatment is not given, the gears blanks should
be given a thorough annealing. The slight reanneal after rough
machining and before the final cut is optional; it always helps,
however.

The final treatment consists in an oil-hardening and tempering
process. For the majority of alloy steels this quenching is done
from a temperature about 50° F. over the critical range; in the case
of chrome vanadium steels, however, the best results are generally
obtained by the use of a higher temperature. The temperatures
generally used for the standard types of alloy steels for automobile
gears, approximating 0.45 to 0.55 per cent, carbon, are about as
follows :

390 STEEL AND ITS HEAT TREATMENT

Chrome nickel steels:

1.5 per cent, nickel, 0.5 per cent, chrome, 1400° F.
1.75 per cent, nickel, 1.0 per cent, chrome, 1425
3.0 per cent, nickel, 0.75 per cent, chrome, 1375
3 . 5 per cent, nickel, 1 . 5 per cent, chrome, 1400

Nickel steels:

3.5 per cent, nickel 1400° F.

5.0 per cent, nickel 1375

Chrome vanadium steel:

Type " D " (1 . 0 per cent, chrome, 0.8 per cent.

manganese, 0.16 per cent, vanadium) 1575° F.

Silico-manganese steel :

1.5 per cent, silicon, 0.7 per cent, manganese 1550° F.

The usual precautions should be observed such as uniform and
thorough heating, protection from oxidation, etc. Further, the
gear should be quenched in the direction of its axis so that the oil
can be made to circulate around the teeth, etc. The notes given
under " Milling Cutters " 1 might also be of interest in their bearing
upon gear treatment.

The tempering is usually done at a temperature of 400° F. or
upwards, depending upon the nature of the steel and upon the
results desired. It should again be stated that a longer tempering
at the lower temperature is preferable to a quicker and shorter tem-
pering at a higher temperature. Thus, if a gear were to have the
temper drawn quickly, the teeth, which should be the hardest, will
be softer than the hub, which will remain brittle; with a longer
heating at a lower temperature this will not be the case, since the
whole gear will have responded throughout. Similarly, for these
reasons, it is inadvisable to temper gears " by color," but to use an
oil bath or a mixture of low melting-point salts.

For gears made of alloy steel with only about 0.30 per cent,
carbon the tempering operation is usually omitted. It is always
best, however, to reheat the oil-quenched gears in boiling water for
a short time in order to remove the hardening strains; such treat-
ment will have little or no influence on the hardness and strength.
The quenching temperature for such steels will of course be higher
by some 50° or 75° than that given under the 0.45-0.55 per cent,
carbon steels.

iCf.Ch.XVI.

MISCELLANEOUS TREATMENTS 391

SPRINGS

The usual analysis for carbon steel springs is approximately:

Carbon 0.90 to 1 . 10 per cent.

Manganese under 0.40

Phosphorus under 0. 04

Sulphur under 0. 04

Silicon up to 0.25

It is dangerous to allow the percentage of carbon to run up to
1.25 per cent, (as is sometimes done), on account of the possibility
of the formation of free cementite, which is an extremely brittle
constituent. A crack might easily start in an area of cementite and
when once started would follow through the cementite to the outer
surface. Lower carbons would preclude the presence of free cement-
ite. Finely divided cementite would also be less dangerous,
and this could be obtained by hardening at a lower temperature
(about 1400° F.), since crystallized and granular cementite can
only be obtained by heating for a prolonged time at a high temper-
ature.

Aside from improper analysis, the majority of spring failures and
troubles may be laid to abnormally high temperatures for heating
for fitting followed directly by quenching from whatever temper-
ature the steel may happen to be at; and then, as if this were not
bad enough, to temper by " flashing." From general knowledge it
appears that the maker of springs has not kept pace with improve-
ments in spring steel and with the increased severity of the duty
expected of springs.

The old practice of high temperatures and of forming and
hardening springs with a single heating cannot be persisted in if
maximum quality and service are to be secured. The " practical "
spring-fitter generally heats the steel to about as high a temperature
as it will take without burning. Its effect upon the structure of
of steel has been explained in preceding chapters, and also above in
its relation to very high-carbon spring steel.

But even assuming that the proper temperatures have been used
in fitting, the time taken to go through the forming operation is
sufficient to give the steel a chance to cool down to a temperature
which will not give the most satisfactory results in hardening.
The steel is not of uniform temperature over its length so that, if it
be quenched directly after forming, it will probably lock up internal

392 STEEL AND ITS HEAT TREATMENT

strains of uncertain magnitude — to say nothing of the insufficient
hardening if the temperature be under that of the critical range.
In other words, the spring should be put back in the furnace again (it
being generally preferable that the maximum temperature for forming
shall be the same as that required for hardening) and reheated for
a few minutes so that it will be heated uniformly throughout at the
right temperature. If high temperatures have been used for form-
ing it will be advisable to allow the steel to cool to a temperature
under that of the Ar range before reheating for hardening; if this
is not done the steel will retain the coarse grain-structure character-
istic of the high heat for forming. If it is found that the steel
departs from its shape at all during this reheating, it may be put
through the rolls again previous to quenching, the time occupied
being small compared with that for the original bending. The
spring should then be quenched in some good, heavy tempering oil.

For drawing the temper it is never advisable to use the process
known as " flashing." The practice of replacing the steel, after
quenching, in a high temperature furnace until the outside of the
steel reaches the desired temperature is one which cannot be too
strongly denounced, because of the impossibility of uniform treat-
ment. No time is allowed for the heat to soak to the center, with
the result that the hardness increases from the outside — a most
undesirable condition. All spring steel should be drawn back in a
suitable low-temperature furnace maintained at the proper temper-
ature. The steel should be kept in the furnace for a time sufficient
to allow of a uniform heating throughout. Lead baths and salt
baths are also used considerably for this work.

The proper temperatures for treating carbon spring steel have
been given considerable attention by the American Society for
Testing Materials. Their experiments were made with test speci-
mens If by f by 14 ins. long with straight edges, and analyzing
about 1.10 per cent, carbon. The results of these tests (1911)
are given in the tables on page 394.

It is apparent that at a quenching temperature of 1500° F. the
maximum results are obtained with a drawing temperature of about
600° F., while with a quenching temperature of 1650° F. the maxi-
mum elastic limit was found with a drawing temperature of about
800° F. In the former group, Series A, 1500°-600° F., it was
found that the angle of bend at rupture showed an average of
slightly over 59°, there being considerable variation between the
specimens; while in the second group, Series B, 1650-800° F., the

MISCELLANEOUS TREATMENTS

393

average angle was slightly over 103°, without any specimen going
below 76°. These results are particularly interesting in view of the
fact that the critical range of these steels is about 1350° F., and
that one would naturally expect that a temperature of about 1400°
F., i.e., slightly over the critical range, would give the best
results. Whether or not such would show up in vibratory tests is
a question which should be given attention.

TRANSVERSE, HARDNESS AND BENDING TESTS OF CARBON SPRING STEEL
Series A, Quenched in oil from 1500° F.

Hardness.

Bend Test,

Temper
Drawn to

Elastic Limit,

(transverse)

Scleroscope.

Angle Bent
through at

Deg. F.

Lbs. per Sq. In.

T)f*{«"t All

Rupture,

On Flat.

On Edge.

-tsrineii.

Deg.

425

129,137

48.5

370

181

600

136,440

50.5

388

835

131,017

43.5

49.5

351

1025

96,852

34.5

268

152

1230

105,400

282

167

Series B, Quenched in oil from 1650° F.

450

130,922

52.5

394

625

134,232

371

820

141,147

389

104

1025

126,320

371

108 .

1210

83,457

36.5

260

180

ALLOY STEEL SPRINGS

The service conditions to which automobile springs are sub-
jected are extremely severe, for they have to sustain the shocks at
speed of the irregularities of the ordinary highway, built for slow-
moving, horse-drawn vehicles. The necessity for high elastic limit,
combined with great toughness and anti-fatigue qualities, make the
use of alloy steel almost mandatory.

The alloy steels in use are of the same analysis of those previously
given under the heading of " Oil-hardened and tempered Gears "
(q.v.). The quenching temperatures are likewise the same as
there given, but the drawing temperatures are higher — generally
from 850° to 1025° F. As far as static strength is concerned, the
majority of the now common alloy compositions will give about
the same test values, approximately :

394 STEEL AND ITS HEAT TREATMENT

Tensile strength, Ibs. per sq. in. ... 190,000 to 250,000

Elastic limit, Ibs. per sq. in 170,000 to 225,000

Elongation, per cent, in 2 ins 15 to 6

Reduction of area, per cent . 45 to 20

Some of the alloy steels, and particularly the chrome vanadium
type, require annealing before shearing. The chrome vanadium
steels used for springs are readily susceptible to " temper," and it
is likely that the rapid air cooling of small flats after they leave the
rolls will cause them to be brittle, thus giving a great amount of
trouble in shearing. The annealing of this chrome vanadium steel
is done by bringing the steel up to a full cherry-red heat in the
furnace (about 1475° F.) and allowing it to cool slowly after being
maintained at this temperature for a sufficient time to allow of uni-
form heating.

The new steels cannot be handled just like the old carbon steel
springs and still obtain from them the maximum development of
their powers. However, the new steels, being in general lower in
carbon, will stand much abuse in heat treatment and still pro-
duce springs of quality undreamed of a decade ago. While as a
class spring-makers have been driven to the use of alloy steels, they
have not as a class been forced to handle them scientifically.

Alloy steels especially should not be heated any higher for form-
ing than is absolutely necessary. Then they should always be
reheated to the proper temperature for quenching in order to make
sure that the entire steel is uniformly heated throughout to that
temperature, which must be exact. The same remarks about tem-
pering as given under carbon steel springs likewise apply here, and
with added emphasis.

OIL-WELL BITS

Bits used for drilling oil wells, gas wells, etc., represent that
class of large implements requiring " end heats." The hardening of
these bits is necessarily an operation to be carried out in the field,
since the bits require a more or less frequent dressing and must be
rehardened after each heating. An extremely hard end and face,
together with a strong, tough core and shank are the principal
requirements for this work.

About 6 or 8 ins. of the bit is carefully heated in the fire (usually
a common blacksmith forge), to the proper temperature — usually
about 1500° F. Higher temperatures should not be used unless
absolutely required by the nature of the steel. Any scale should be

MICELLANEOUS TREATMENTS

395

carefully and quickly brushed off before quenching. The bit is
then removed from the fire and allowed to rest in a bucket of coarse
salt for a second or two. This salt treatment may be omitted, but
it undoubtedly gives better results; the direct use of brine is gen-
erally too severe for most bit steels.

A box or trough should previously be fitted with a wooden grating
made of slats, the top of which will be about 3 or 4 ins. under the
surface of the water in the box. Some drillers add vitriol to the
water quenching bath to obtain a greater hardness. The bit should
then be quickly lowered vertically into the cold water until it rests
upon the wooden grating, and should be allowed to remain there
until cold.

The precautions to be observed are: (1) Lower vertically, in
order to obtain an equal hardness on both faces of the bit; (2) do
not quench to a greater depth than 3 or 4 ins. ; (3) do not move the
bit nor splash the heated part of the shank with water; (4) allow
the steel to remain in the water until cold, generally over night.
Although the surface of the water bath may steam, it will generally
be found that directly beneath the surface the water is cold,
and likewise the end of the bit. Splashing the heated part of the
bit with water has a tendency to draw the temper of the faces.
Immersion to a greater depth than 3 or 4 ins. is apt to give a soft
bit. If these precautions are carefully observed, and the steel is
of the right analysis, a bit with a glass-hard surface and a strong,
tough core will be obtained. Such bits require no tempering, and
should not chip off.

Oil-well bit steel will vary between 0.50 and 0.80 per cent, car-
bon and manganese, low phosphorus and sulphur, up to 0.25 per cent,
silicon, and the addition of about 0.5 per cent, chrome for the lower
carbons. The chrome bit steel, if of the proper carbon-manganese-
chrome composition, will undoubtedly give the best service. The
following analyses are characteristic of American oil-well bits used
and giving good service :

Carbon.

Manganese.

Phosphorus.

Sulphur.

Silicon.

Chrome.

0.73

0.61

0.017

0.030

0.14

0.59

0.17

0.010

0.015

0.13

0.83

0.65

0.010

0.021

0.13

0.60

0.51

0.012

0.019

0.01

0.56

0.54

0.53

0.007

0.021

0.006

0.51

0.49

0.56

0.010

0.016

0.008

0.52

396 STEEL AND ITS HEAT TREATMENT

SAFE AND VAULT STEEL

Safe and vault steel may be taken as representative of that class
of material involving different steels welded together, but for which
the proper treatment of one analysis will be sufficient for both.
Steel for safes and vaults consists of alternate layers of soft and hard
steel, and is known to the trade as " three-ply," " five-ply," etc.
By having these alternate layers there is obtained, under suitable
treatment, a metal which will have sufficient ductility (due to the
soft layers) to resist explosive forces, and at the same time be im-
penetrable to drilling, sawing or other machine operations (due to
the " hard center"). The soft layers are made of ordinary low-
carbon or " soft ." steel, while the hard centers will analyze about
0.85 to 1.05 per cent, carbon and manganese, with or without the
addition of chrome.

The plate is first machined or ground to size and the necessary
holes drilled, threaded, and plugged with fire-clay for protection.
The plate is then placed in a suitable heat-treatment furnace, and
thoroughly heated to 1400° to 1500° F., depending upon the compo-
sition of the hard layer. It is extremely important that ample
time be allowed for the heat to penetrate and thoroughly heat
the high-carbon steel, for it is upon the hardness of these layers
that the full value of the finished plate will depend. The major-
ity of the cases in which the necessary hardness was not obtained
which the author has investigated have been due to an insufficient
length of heating rather than to any fault in the analysis of the
steel.

The plate is then quickly removed from the furnace by a crane
or hoist and quenched in cold water. As the hardness is largely
dependent upon the rapidity with which the steel is cooled through
the critical range, arrangements should be made to obtain a constant
supply of cold water in contact with the steel during the quenching
operation. If the quenching is done in a tank, the inlet supply should
be large enough always to keep the water cold — the warm water
being taken away from near the top of the tank. In this case the
plate is quenched vertically; particular care should be used in getting
the whole plate into the water as quickly as possible, and in an ab-
solutely vertical position, if warpage is to be avoided. As soon as
the initial immersion is accomplished the plate may be swung to
and fro in the tank to aid in the heat removal. Other plants quench
by means of water sprays, the plate being supported on a horizontal

MISCELLANEOUS TREATMENTS 397

rack; with this method of cooling the water supply should be suffi-
cient to remove the steam as soon as it is formed.

The plates are not tempered or drawn. Specifications require
that the best high-speed steel drill shall not penetrate the hard-
center layers.

STEEL CASTINGS

In the mad rush for alloy steels and their heat treatment but
little attention has been given to the treatment of steel castings.
And yet there is an opportunity for as great, if not greater, improve-
ment in these parts as in forged or rolled sections. All steel castings
should be annealed or oil treated, not only to remove the casting
strains, but also to get the metal into the best possible condition.
Due to the method of fabrication, the rapid cooling of thin sections
and the slower cooling of adjacent thicker sections must inevitably
produce casting strains of a more or less intense nature. Similarly
and coincidently , the structure of the metal must inherently be poor :
the grain will be coarse instead of fine and " silky," the metal will
tend to have low ductility and brittleness, and the physical proper-
ties of the steel as a whole will vary considerably. Unlike forgings
and rolled sections, castings are not generally subjected to any
reheating and elaboration, so that the metal must have those prop-
erties characteristic of moderate cooling from high temperatures.

Thus the usual specifications for steel castings, in which the low
ductility will be apparent, will call for:

Tensile strength, Ibs. per sq. in 85,000

Elastic limit, Ibs. per sq. in 45,000

Elongation, per cent, in 2 ins 12

Reduction of area, per cent 18

Even the now common addition of titanium or vanadium will not
serve to eliminate entirely the necessity for subsequent treatment.
Annealing, or better still, a full heat treatment, is mandatory.

Contrary to the ideas held by many " practical " hardeners, the
principles of treating steel castings in no wise differ from those of
steel forgings of the same section and analysis. The main difficulty
encounterad is that caused by the length of time required for the
diffusion of the ferrite and the equalization of the metal as a whole.
Castings usually require considerable time for this to take place
because of the tendency of the metal to return to its original
molecular arrangement and structure during slow cooling. Thus

398 STEEL AND ITS HEAT TREATMENT

much of the unsatisfactory annealing is, technically speaking, due
to the segregation of the ferrite.

It is therefore necessary, in annealing steel castings, to (1) heat
well over the upper critical range, (2) for a length of time sufficient
to obliterate entirely the previous structure and crystallization,
and followed by (3) slow cooling. The proper annealing temperature
for the ordinary machinery castings will be between 1500° and 1600°
F., depending upon the carbon content.

If the annealing is preceded by normalizing, i.e., air cooling from
a temperature considerably above the upper critical range — say
1800° F. — the length of time required for the subsequent anneal
will be considerably shortened, besides improving the steel.

For castings with the carbon on the lower side of 0.25 or 0.30
per cent., or for castings of considerable size, air cooling from about
1600° F. will usually produce good results.

The best method, however, is that of oil quenching and annealing
or toughening — either with or without a previous normalizing.
The castings should be heated as directed under annealing, quenched
in the proper manner in oil, and then reheated to the temperature
which will give the combination of strength and ductility desired. A
drawing temperature of 1250° F. will produce the most ductile
steel.

STEEL WIRE1

The principal heat treatments used in the manufacture of wire
are: 1, annealing; 2, patenting; 3, hardening and tempering.

Annealing serves to accomplish three important functions:
1. To remove the effects of hardening due to cold work in wire
drawing or cold rolling, thus making the steel ductile and soft.
Annealing for this purpose covers principally the low-carbon wires,
those with carbon 0.25 per cent, and under. 2. To refine grain-
applied principally to the higher-carbon rods and wires, those with
carbon 0.30 per cent, and over. 3. To obtain definite structure in
the finished material — applied principally to the higher-carbon wires,
those with carbon 0.30 per cent, and over.

When a steel wire rod of the structure shown in Fig. 207 is sub-
jected to the wire-drawing process, a marked change in the grain
structure takes place. With each successive draft, the grains stretch
out in the direction of drafting until a point is reached when the

1 From a paper by J. F. Tinsley, American Iron and Steel Inst., 1914, and
The Iron Age, May 28, 1914.

MISCELLANEOUS TREATMENTS 399

grains have been elongated to the limit of their ductility. If sub-
jected to further strain by further drafting they will part and the
wire will break. Before this brittle condition is reached, therefore,
it is necessary to heat treat the wire by subjecting it to what is
known in the wire business as a " process annealing."

The effect of wire drawing in elongating the structural grain
of the steel may be seen by comparing Figs. 207, 208 and 209. Fig.
207 shows the structure of the rod before drawing; Fig. 208 shows the
structure after a 15 per cent, reduction from the rod; and Fig. 209,
the structure after a 60 per cent, reduction from the rod. All of
these micrographs represent sections taken from a plane parallel
to the axis of tho rod or wire, not cross-sections. The reason for the
marked difference in grain shown in Figs. 207 and 209 may be grasped
more clearly when it is appreciated that Fig. 209 represents a wire
reduced in the wire-drawing process to such a degree that it has
become elongated 2J times the original length of the rod.

Process or " works " annealing consists in heating the wire to a
certain temperature, maintaining that temperature until the entire
mass of steel is thoroughly heated through, and finally cooling down.
In the most common of all annealing — that to remove the effects
of cold work such as drawing — it is not necessary to reach the
critical temperature, which is 1300° F., or higher, depending on the
carbon content. A temperature of 1100° F. is entirely sufficient to
relieve the strained condition of the grain shown in Fig. 209. Fig.
210 shows the same wire that is depicted in Fig. 209 after annealing
at a temperature below the critical range.

In the annealing process the strained and elongated grains
shown in Fig. 209 break up and rearrange themselves to form a
new grain structure as shown in the micrograph. The annealed
steel of the structure shown is now in excellent condition to with-
stand further cold work in reducing it to finer sizes; or, if already
at finished size, is in good condition to meet the demands of annealed
wire service.

The effect of reduction of section incident to wire drawing on
the tensile strength and ductility of steel wire, and the marked
change brought about in these characteristics by annealing, as just
outlined, is shown in Fig. 216. This table is based on drafting and
annealing practice in reducing a low-carbon steel rod — in this case
0.10 per cent, carbon — to a fine size of wire. It will be noted that
between 80 per cent, and 90 per cent, reduction from the rod or
annealed wire can be taken before annealing is necessary.

400

STEEL AND ITS HEAT TREATMENT

It is found in practice that in cold drawing from a soft rod or
annealed wire, the increase in tensile strength is a direct function
of the amount of cold work, almost independent of other conditions.

FIG. 207.— Annealed (0.08 Carbon) Steel. (Tinsley.)

Annealing practically brings the rod or wire, regardless of size,
back to its original condition with regard to tensile strength and
ductility. It will be noted that the final annealing does not bring

FIG. 208.— Steel Wire (O.OS Carbon)
Given One Draft; 15 per cent.
Reduction from Rod. (Tinsley.)

FIG. 209.— Steel Wire (0.08 Carbon)
Given Several Drafts; 60 per
cent. Reduction from Rod.
(Tinsley.)

the tensile strength as low as previous annealing. This is due
simply to the fact that in annealing the fine sizes it is usual, in order
to avoid the mechanical sticking of the wire in coils, to anneal at
slightly lower temperatures than in ordinary process annealing.

MISCELLANEOUS TREATMENTS 401

The second important function of annealing is that of refining
grain, and its practical application in the wire mill covers principally
the medium- and higher-carbon steels. The structure of wire rods
with regard to size of grain is dependent upon the temperature at
which the rods are finished in the hot rolling mill and upon the rate
of cooling through the critical temperature of the steel. In steel of
low carbon this is not of as much importance as in the higher-carbon
steels, for the reason that the ordinary finishing temperature varia-
tions of good rolling-mill practice have less effect on grain structure
of soft rods, and therefore less effect on their physical properties.
In higher-carbon steels a fine grain is important, for it is this struc-
ture that makes for such steels their field of usefulness, where high
strength, high elastic limit and toughness are required.

Theoretically, the ideal structure would be obtained if the entire
rod could be finished at about the critical temperature. But this
is, of course, impracticable, for the reason that it is impossible
to regulate the finishing temperatures so closely, and for the addi-
tional reason that there is, necessarily, particularly in rolling very
long lengths of very small sections, a marked difference between the
finishing temperatures of the first and last end of a rod. The higher
the finishing temperatures above the critical range the coarser the
grain, and the coarser the grain the more does the steel lack the
qualities that give it value. In order to destroy the coarse or uneven
structure that may be created as just described, it is necessary to
anneal the steel by heating it just above its critical temperature and
slowly cooling it down.

The effect of overheating in coarsening the grain structure of
a 0.45 per cent, carbon steel and the refining influence of this type
of annealing is shown in Figs. 211 and 212.

The third and last class of annealing to be described— that to
obtain definite structure — is one of comparatively recent develop-
ment in the steel-wire industry and one which promises to be of con-
siderable value. Annealing of this type is applied principally to
the higher carbon wires. Since the structure of such wires can be
varied considerably within a small range of annealing temperatures,
it covers specific products and not general classes, as would be the
case in regard to the two previously described types of annealing,
Figs. 213 and 214 illustrate excellently this special type of annealing.
These photomicrographs show the structures of two annealed pieces
of the same coil of high-carbon wire, in which the annealing temper-
ature of the one specimen was 130fr° F., and of the other 1250° F.

402 STEEL AND ITS HEAT TREATMENT

It is impossible to identify the structure by a simple observation of
the fracture, which is the ordinary rough-and-ready method; nor
is it possible to regulate annealing temperatures so closely without
the use of pyrometers.

In passing to the next great class of heat treatment applied to
steel wire, patenting, it is interesting to note that we likewise pass
to another class of wire as regards grading by carbon content. It
naturally covers the medium-carbon steels, being employed chiefly on
carbons between 0.35 and 0.85 per cent. In the medium-carbon
steel wires strength and toughness are required for both process and
finished wire. Patenting makes possible this combination of strength
and toughness, and to this process is due in large measure a broad
field of application for steel wire.

The high strength and toughness of patented wire are due to
its carbon condition and to its peculiar structure. The first step
in the patenting process is to heat the wire to a temperature above
its critical range. The degree of heating is regulated according to the
carbon content of the steel, the size of rod or wire, and the time the
material is subjected to the heat. After sufficient heating, the next
step is to cool the material rapidly below its critical range, the
structure obtained depending upon the rate of cooling. In practice,
patenting is usually conducted as a continuous operation, the wire
being led through the heated tubes of a furnace and cooled by being
brought into the air or into a bath of molten lead comparatively
cool but seldom under 700° F.

A better understanding of the structure of a patented wire may
be had by a comparison of the structure obtained by slow and by
rapid cooling. If the steel after being heated is allowed to cool
slowly through the critical temperature range, the homogeneous pre-
existing solid solution of iron and iron carbide separates into a hetero-
geneous mixture of two constituents, resulting in the plate-like struc-
ture called " pear lite." In a patented wire, part of the carbide of
iron is in solid solution and the remainder, while not in solid solution,
has not had time to form into plates. The difference in structure
between slow and rapid cooling is seen in Figs. 213 and 215. The
photomicrograph of the patented wire shows, as a result of the rapid
cooling, a structure that might be termed nondescript. Metallo-
graphists will recognize the structure as " sorbite," which, in the
cooling of the higher-carbon steels from above the critical tempera-
ture, is that stage of transition just preceding the pearlitic, the final
condition of annealed steel as shown in Fig. 213. The patented

MISCELLANEOUS TREATMENTS

403

FIG. 210.— Steel Wire (0.08 Carbon)
Hard Drawn and then Annealed
below the Critical Temperature.
(Tinsley.)

FIG. 211.— Steel (0.45 Carbon) Over
heated. (Tinsley.)

FIG. 212.— Steel (0.45 Carbon)
Annealed. (Tinsley.)

FIG. 213.— Annealed (0.85 Carbon)
Steel. (Tinsley.)

FIG. 214.— Specially Annealed (0.85
Carbon) Steel for Globular
Structure. (Tinsley.)

FIG. 215.— Patented (0.85 Carbon)
Steel. (Tinsley.)

404

STEEL AND ITS HEAT TREATMENT

fcf

00
CO

^|fi

s«8

& 0

- <

5£

fe fi

1, g

° fc <

00
IN

£ z

Bfc

-IN

fefl

l§

PROCESS
AN-
NEALED

O
CO

•4M

£§

Ho
IN

•^

w* •?,

og ^

CONDITION
MATERIAL

PER CENT.
REDUCTION

BY DRAWING

TENSILE
STRENGTH

POUNDS PER SQ. IN.

PER CENT.
ELONGATION

IN TEN INCHES

~ o

S?X3

o-5

§£
H

218 000

184 000

0
IN

156 000

208 000

»-, CU

O-i
Ofi

H Q

CO
CO

. ts

o
co

156 000

0
(N

-— 5jl

128 000

00
t".

^MntHZ^tT1^ fhe ^ffe^ °f cold drawine and Patenting on the physical properties of highe
5 — At the first patenting stage, process annealing gave a tensile strength of 70,000 Ibs. p<

176 000

CO
(N

Q H
H Q

163 000

§1

>-i W

00
(N

115000

146 000

00
IN

s§

00
IN

122 000

ils§

CONDITION

MATERIAL

PER CENT.
REDUCTION

BY DRAWING

TENSILE
STRENGTH

POUNDS PER SQ. IN.

PER CENT.
ELONGATION

IN TEN INCHES

MISCELLANEOUS TREATMENTS 405

wire, therefore, represents an unsegregated condition as against the
segregated or coarsely laminated structure of annealed wire. The
high tensile strength of patented wire is due to the amount of carbon
in solution, and its toughness to the fineness of the grain structure.

Patenting serves two important functions in the wire business:
1. In the process of manufacture, the removal of the effects of cold
work, such as drawing. 2. In the finished wire to give, in conjunc-
tion with cold drawing, the required combination of strength and
toughness. Strictly speaking, patenting is not necessary simply
to relieve strain, for annealing would serve that purpose, but the
structure obtained by patenting permits much further cold drawing
than does the structure obtained by annealing. This is due primarily
to the increased ductility and toughness of the patented wire. The
effect of patenting as just described is shown in Fig. 217.

In wire making, hardening and tempering should be conducted
usually as a continuous process. In the making of tempered wire
the material is first run through the heated tubes of a furnace,
then quenched quickly in a bath of oil or water, then run into the
tempering bath of, say, molten lead, each wire being in continuous
motion from the time it enters the heating furnace until it is wound
on a reel. Hardening and tempering apply to the higher carbon steel
wires — those in which the carbon range is from 0.65 per cent= to
1.00 per cent. With varying tempering temperatures between 500°
and 1100° F., the tensile strength runs from about 340,000 Ibs. per
square inch to 150,000 Ibs. per square inch. At the lower temper-
ature the decrease in tensile strength is, as we should expect, much
greater per 100° F. range than at the higher temperatures. From
500° to 600° F. there is a drop of 60,000 Ibs. per square inch, while
between 1000° F. and 1100° F. the drop in tensile strength amounts
to only about 10,000 Ibs. per square inch.

FORGING

No small percentage of the difficulty encountered in heat-treat-
ment operations is due to improper forging methods, and ofttimes
the heat-treatment operation is nothing more than a useless effort
or attempt to get something out of a forged piece of steel that is not
actually in it. Thus, the steel man is often blamed for the absence
of quality in his steel that he actually put in it; and the heat-treat-
ment man is blamed for his lack of ability to locate such qualities,
which he properly assumes to exist, but which, nevertheless, the forge
man took out by poor heating, unknown to himself or the other two.

406 STEEL AND ITS HEAT TREATMENT

The strongest language that could be employed in an attempt
to describe the general average heat-treatment equipment, the
methods of heating, and personnel, as they are actually known to
exist, would be altogether too mild and ineffective for a proper
description of the heating methods and equipment in the majority
of forge shops in the country. As in the case of machine work, the
design of the hammers, presses and other machine equipment has
made rapid strides forward, but the two most important factors
of the operation from the metallurgical end — namely, the man and
the furnace — have either stood still or gone backwards. Many
well-informed and experienced men claim that the caliber of forge
men to-day is not what it was years ago, and that better quality of
work was produced with the old-fashioned coal or coke furnaces,
though at a higher cost, than at present with furnaces burning oil
or gas. There appears to be something in this statement, particu-
larly in view of the high quality work turned out in Europe, where
the use of high-speed machines, oil or gas fuel, and efficiency pro-
duction methods, are not as prevalent as here. If such a difference
actually exists, it can invariably be traced to the personnel of the
plant, because, as in most operations involving the skill of the oper-
ator as against the fixed movement of a machine, quality reflects
the man and his knowledge of the work. But even so, we can and
should be able to do better with fuel so closely linked with uni-
formity of temperature, steadiness of operation, and ease of control.
If we do not, then it is up to the man or the furnace and not to the
hammer or the fuel, which is in itself a good argument for improve-
ment of the heating and human equations in the operation.

Two of the weak links in forging practice, from the metallurg-
ical end, are the lack of uniformity and temperature of the heats
and the method of handling stock in and out of the furnaces.

As a rule, the heats are altogether too high, with the result that,
while the surface is apparently hot, there may be actually a " bone "
on the inside. It is common practice to see a bar drawn from a
furnace that will actually drip, and yet when placed under the ham-
mer there will be indications of lack of heating on the inside. It is
the inside of the bar that determines the physical properties of the
final forging and not the outside; and there is nothing gained in
these quick " wash " or surface heats. Slow, soft, soaking heats,
affording plenty of time to heat up, are more desirable than the
higher quick heats. The idea should be to maintain the temperature
of the furnace as near as possible to that actually required to soften

MISCELLANEOUS TREATMENTS 407

the steel to the extent necessary for its proper shaping, and to give
it plenty of time in the furnace thoroughly to soak at this temper-
ature without overheating or oxidizing the outside. The fire should
be soft and a little high in carbon, in order to reduce oxidation. The
modern alloy steels do not require high, sharp, dripping heats,
and the proper handling of them demands the slow, soft, non-
oxidizing heats above referred to.

The general design of forge furnaces is far below the standard
of heat-treating furnaces and is a point usually left to the forge man
or to a bricklayer. It is common practice to see furnaces hot on
one side and cold on the other. Also, to hear complaints of lack of
ability to heat steel properly in a furnace in which the burners blast
directly against the stock, which naturally keeps the stock nearest
the burner cool and heats the pieces farther away. There are
hundreds of such designs in use that have been turned out by furnace
builders who ought to know better.

CHAPTER XVIII
PYROMETERS AND CRITICAL RANGE DETERMINATIONS

PYROMETERS 1

Pyrometers in General. — The pyrometer has played a basic part
in the development of intelligent heat treatment. In hardening
rooms where pyrometers are not used, a discussion of any temper-
ature treatment and instructions are given as the instructions
must have been given in the Tower of Babel. There is no dis-
tinction or mutual understanding of terms, and until a pyrometer
— and an accurate one — is in a hardening room, it is not possible
for those interested in the heat treatment in that room to talk
to each other in a mutually intelligible way. Of course, where one
old hardener has been in charge for twenty years and the manage-
ment decides to take a chance on his staying with them and living
for another twenty years, it may be all right to have everything
locked up in his head; but where matters are more extensively and
more modernly conducted, it is necessary to have some language in
which people can talk; and the pyrometer, by virtue of its tempera-
ture scale, which is a conventional scale of denned terms, affords
the means of communication in a language that is mutually under-
stood. In the same way it permits records to be kept for future
reference. Where this is not done, men will be found trying to
remember the heats at which they treated this, that or the other
lot of steel; they cannot remember, and they are sure to get into
trouble if they try to. The pyrometer has changed barbarian
methods into civilized methods in a hardening room.

There is need for a greatly extended use of pyrometers of the
best possible grade, but more especially for an intelligent use of them
that will in some measure compensate for the skill in producing them

1 It is the aim of this section to deal more with the rational use of pyrometers
rather than with a detailed explanation of the theory and construction of the
numerous instruments in commercial use for heat measurement. For a fuller
explanation of the latter subject than is subsequently given, the reader is referred
to standard reference books on the subject.

408

PYROMETERS AND CRITICAL RANGE DETERMINATIONS 409

and the money involved in their installation. The pyrometer is
not all-sufficient, nor it is the cure-all for the troubles of a hardening
plant. There should be an education of the man to look upon
pyrometers as gauges and indicators of the existence of energy,
and as an aid to him in executing his work and not as a means of
releasing him from responsibility accompanying the exercise of
judgment.

The pyrometer has been of inestimable value in affording a
means to check temperature, but — and aside from the correlation
of results — its efficiency ends largely with that indication. The
uniformity of heated product, however, depends upon the manner
of applying the heat, which — with the method and cost of operation
— is primarily a function of furnace design. It is possible to indicate
a uniform temperature and yet not produce a uniformly heated prod-
uct; and unless the heat is uniformly applied to the stock at the
temperature indicated, then a uniform pyrometer reading is mis-
leading and inconclusive. Thus an elaborate pyrometer system,
with means for signaling variations in temperature and of record-
ing these variations, is not conclusive evidence of accurate heating.
The development toward better and cheaper results will be brought
about by improved heating methods, even though the temperature
recorded from any one point in a furnace chamber may be the same
as that indicated from a similar point of another furnace less effici-
ently designed.

The time element is linked inseparably with all heating opera-
tions. A piece of steel can absorb heat only so fast and no faster.
Only by operating the furnace so that the maximum temperature
is maintained for the length of time necessary uniformly to heat
the steel throughout to that temperature, is it possible to produce
the best results. In other words, the composition and the mass
of the steel must be correlated with the time element. First deter-
mine the length of time necessary, under standard furnace conditions,
to produce the necessary results; then regulate the furnace by the
aid of the pyrometer; and finally, place a clock beside the instru-
ment and work the two together. The sooner the average heat-
treatment man (and his superiors, for that matter) can be brought
to realize that a pyrometer is almost valueless without the use of a
time clock and common-sense observation of furnace conditions, the
better.

Thermo-Couples. — For the usual operations in heat-treatment
work involving temperatures of over 600° or 700° F,, the thermo-

410 STEEL AND ITS HEAT TREATMENT

couple system is the most used. The principles upon which its use
depends are simple. Expressed briefly, if the ends of two pieces of
dissimilar metals (usually as wires) are joined together and one
of the junctions (the " hot end ") is heated, the other junction (the
" cold end ") being held at a constant temperature, a feeble electric
current is generated in the circuit. This electromotive force, aside
from being dependent upon the nature of the couple, is, for the
thermo-couples in practical use, dependent upon the difference in
temperature between the hot and cold ends.

In regard to thermo-couples, standard base-metal compositions
will generally give satisfaction between 600° or 700° F. and 1800° F.;
while above 1800° F. couples of platinum and platinum-rhodium
should be used. All base-metal couples should be readily replace-
able, and, more emphatically, interchangeable. All couples should
be suitably protected with iron pipes from oxidation and rough
handling.

Position of the Thermo-Couple. — The fact that a pyrometer
may show that some particular portion of the heating zone is at the
proper temperature is no proof that the steel is also at that temper-
ature. The hot end of the couple may be so placed that it must
inevitably be hotter than the hearth of the furnace, or hotter than
any material placed on the hearth. This will be true if the end of
the couple is exposed to the direct heat of the flame. It might
therefore be concluded that the tip should be as near the work as
is possible, so that both may attain the same temperature — and
which is without doubt advisable in many instances. On the other
hand, it has been noted in some cases in which the couples have
been placed close to the work that the readings are not in accord
with the temperature of the steel because the couples, being of
smaller mass, take up readily the high peak of the flame tempera-
ture. There are certain instances where it has been found by
experience desirable to locate the tip of the couple in a recess in
the furnace wall where it was out of the course of the flame and
thus dependent for its temperature upon radiation from the main
body of the furnace lining and radiation from the work; under
some circumstances such a position is preferable.

Millivoltmeter vs. Potentiometer. — By inserting into the thermo-
couple circuit, at the cold junction, a suitable device for measuring
the electromotive force, a reading may be obtained in millivolts;
or, by suitable calibration, a reading directly in terms of temper-
ature. This indicating instrument (the pyrometer) may be of

PYROMETERS AND CRITICAL RANGE DETERMINATIONS 411

the galvanometer or millivoltmeter type, or of the potentiometer
type.

The potentiometer in theory bears much the same relation to the
millivoltmeter that the balance-arm scales bears to the spring scales.
The constancy of both the spring scales and the millivoltmeter is
entirely dependent upon the constancy of springs or of suspensions,
and upon the absence of friction. The constancy of the potentiom-
eter and of the balance-arm scales is dependent only upon the con-
stancy of standard weights in one case and a standard electromotive
force in the other. Standard weights are added to or removed
from balance scales until a balance between known and unknown
is obtained. Similarly in the potentiometer type varying fractions
of a known and presumably standard electromotive force are opposed
to the electromotive force of the thermo-couple until it is balanced
just as a standard weight is moved along a scale arm for balance.

This is the potentiometer not as it is, but as we would like to have
it. The standard cell will not stay standard if any current is drawn
from it and, consequently, the e.m.f. of the standard cell is not
opposed to the e.m.f. of the couple in potentiometers as made for
any ordinary use. Another cell or battery is brought into use and
the e.m.f. of that is opposed to the e.m.f. of the couple. Now
this secondary cell varies in e.m.f. from week to week and day to
day, and even hour to hour under use, and it is necessary contin-
ually to check this service cell against a standard cell and then to
adjust for the differences that are creeping in all the time. The
balance scales, therefore, instead of being operative with standard
weights, have a sort of beaker of boiling water as the weight, which
is continually boiling to less mass and which has to be filled up or
adjusted every few minutes by comparing it with a standard weight,
for the standard weight itself is not trusted on the scales nor is there
any other weight, i.e., battery, that can be trusted on the scales that
will not vary.

Selection of Equipment. — The selection of one type or the other
is largely a matter for economic and technical consideration. In a
word, the purchaser should consider the relation existing between
(1) accuracy, sensitiveness and constancy, (2) ease of reading, and
(3) the cost — both initial and of up-keep. There is also a psycho-
logical consideration that goes hand in hand with the above consider-
ations and which should not be lost sight of: the millivoltmeter
is a direct-reading instrument, which means that it is easy to read;
the potentiometer requires a fair amount of manipulation and is

412

STEEL AND ITS HEAT TREATMENT

somewhat less easy to read. The question then is: Which instru-
ment will the average furnace man read more frequently? No
matter how accurate a pyrometer may be, its value is only in the
use made of it.

Cold-end Temperature. — The cold-end temperature is a source
of prolific error in some pyrometer installations. It should be
remembered that all instruments are calibrated for a definite cold-
end temperature, usually 75° F. If the cold end is in a position such
that it receives the direct or radiating heat from the furnace, and

Copper LeaJs^

groxm A- re n .i pevafiii- e

•*.V- :•; : ; '•/ r:'v&2 •

FIG. 218. — Compensating Cold End Temperatures with Auxiliary Couple.
(Wilson-Maeulen Co.)

therefore varies in temperature, the indicated temperature at the
instrument will be incorrect. For this reason the cold end should
always be kept cool, and at as near a constant temperature as is
possible. This compensation may be accomplished by having the
cold end as near the ground as possible; or by letting a small stream
of cold water flow over the cold ends; or by connecting an auxiliary
couple of the same electromotive force as the furnace couple in oppo-
sition to -the couple in the furnace, and running the auxiliary couple
to an underground point at the bottom of a pipe driven a few feet
into the earth as shown in Fig. 218. The potentiometer type
equipment frequently carries the cold end directly to the instrument,

PYROMETERS AND CRITICAL RANGE DETERMINATIONS 413

entirely eliminating the effect of fluctuating temperatures near the
furnace.

Pyrometer Standardization. — One of the most important points in
connection with pyrometers is the necessity for frequent and regular
calibration of the thermo-couples. All base-metal couples should
be standardized at least once a week, and oftener if possible. Fur-
ther, new couples should always be standardized before use, since
errors may frequently be found even in supposedly correct new
couples.

There are two general methods for standardization or calibration
of thermo-couples: (1) Checking against the melting- or freezing-
points of known salts or metals, and (2) checking against a standard
millivolt meter or pyrometer.

Standardization with Common Salt. — An easy and convenient
method 1 for standardization and not necessitating the use of an
expensive laboratory equipment is that based upon determining
the melting-point of common table salt (sodium chloride). While
theoretically salt that is chemically pure should be used (and indeed
this is neither expensive nor difficult to procure), commercial accu-
racy may be obtained by using common table-salt such as is sold by
every grocer. The salt is melted in a clean crucible of fire-clay, iron
or nickel, either in a furnace or over a forge-fire, and then further
heated until a temperature of about 1600° to 1650° F. is attained.
It is essential that this crucible be clean, because a slight admixture
of a foreign substance might noticeably change the melting-point.
The thermo-couple to be calibrated is then removed from its protect-
ing tube and its hot end is immersed in the salt bath. When this
end has reached the temperature of the bath, the crucible is removed
from the source of heat and allowed to cool, and cooling readings are
then taken every ten seconds on the millivoltmeter or pyrometer.
A curve is then plotted by using time and temperature as co-ordinates,
and the temperature of the freezing-point of salt, as indicated by
this particular thermo-couple, is noted, i.e., at the point where the
temperature of the bath remains temporarily constant while the
salt is freezing. The length of time during which the temperature
is stationary depends on the size of the bath and the rate of cooling,
and is hot a factor in the calibration. The melting-point of salt is
1472° F. and the needed correction for the instrument under obser-
vation can be readily applied. The curves in Figs. 219 and 220 illus-
trate the calibration of a correct and incorrect pyrometer.

1 Carpenter Steel Co.

414

STEEL AND ITS HEAT TREATMENT

180
160
140
120
100
80
60
40
20
0

'-"""

1050° 1600° 1550° 1500° 1450°

Degrees Fahrenheit

FIG. 219. — Diagram Showing the Calibration of a Pyrometer which Reads 45° F.
Too High. (Carpenter Steel Co.)

180
100
140
120
100
SO
GO
40
20

]

•

"*' •

•i*1

1650° 1600° 1550° 1500° 1450°
.Degrees Fahrenheit

FIG. 220. — Diagram Showing the Calibration of a Pyrometer which is Correct.

(Carpenter Steel Co,)

PYROMETERS AND CRITICAL RANGE DETERMINATIONS 415

It should not be understood from the above, however, that the
salt-bath calibration cannot be made without platting a curve : in
actual practice at least a hundred tests are made without platting
any curve to one in which it is done. The observer, if awake, may
reasonably be expected to have sufficient appreciation of the lapse
of time definitely to observe the temperature at which the falling
pointer of the instrument halts. The gradual dropping of the pointer
before freezing, unless there is a large mass of salt, takes place rapidly
enough for one to be sure that the temperature is constantly falling
and the long period of rest during freezing is quite definite. The
procedure of detecting the solidification point of the salt by the
hesitation of the pointer without platting any curve is suggested
because of its simplicity.

Complete Calibration of Pyrometers. — For the complete calibra-
tion of a thermo-couple of unknown electromotive force, the new
couple may be checked against a standard instrument, placing the
two bare couples side by side in a suitable tube and taking frequent
readings over the range of temperatures desired.

If only one instrument, such as a millivoltmeter, is available,
and there is no standard couple at hand, the new couple may be
calibrated over a wide range of temperatures by the use of the follow-
ing standards ;

Water, Boiling-point 212° F.

Tin, under charcoal, Freezing-point 450

Lead, under charcoal, Freezing-point 621

Zinc, under charcoal, Freezing-point 786

Sulphur, Boiling-point 832

Aluminum, under charcoal, Freezing-point 1216

Sodium chloride, Freezing-point 1474

Potassium sulphate, Freezing-point 1958

A good practice is to make one pyrometer a standard; calibrate
it frequently by the melting-point-of-salt method, and each morning
check up every pyrometer in the works with the standard, making
the riecessary corrections to be used for the day's work. By pur-
suing this course systematically, the improved quality of the product
will much more than compensate for the extra work.

Central Switch-boards. — For plants in which there are a number
of thermo-couples, one indicating instrument with a central switch-
board may be used. As many as sixteen couples may be wired to
one selective switch, the maximum number simply depending upon
the elasticity of the system and the convenience of the operator.

416

STEEL AND ITS HEAT TREATMENT

A wiring diagram for such an installation is shown in Fig. 221. By
throwing the switch from one contact to another the connection
is made with each individual furnace. For large heat-treatment
plants the time of one man is generally taken in attending to the
system, he signaling the individual operators by means of lights
and belts the relative temperatures in the furnace. We have
previously commented upon such systems.

The Central System. — The Chalmers Motor Company operate
their system,1 having two central switch-boards with sixteen furnaces
on a switch, as follows;

Couple 8

Couple I ,

FIG. 221. — Wiring Diagram — Pyrometer and Selective Switch. Showing Four
Couples Connected with the Switch, Openings for Four More.
(Hoskins Mfg. Co.)

" We regulate the heat of the furnaces by a series of lights —
each furnace having over it a red, blue and green light. These are
used as follows: We will say that the temperature of an empty
furnace which we are about to use is 1600° F. The loading of the
furnace with forgings necessarily reduces the heat by radiation any-
where from 100° to 250°, depending upon the number of pieces put
in the furnace. When we commence to bring the heat up again to
the proper place and it gets to about 1575°, the man at the switch-
board throws on a blue light, which means to the furnace operator
that the heat is still considerably too low. When the temperature
reaches about 1590° the blue and green light is turned on, which
signifies to the operator that the furnace is still not quite hot enough.
1 Personal Correspondence.

PYROMETERS AND CRITICAL RANGE DETERMINATIONS 417

When the 1600° point is reached the green light is turned on; this
is the O. K. light and means that the temperature is correct. The
steel is then allowed to soak at this temperature for the time neces-
sary to affect the whole mass. If the heat during the operation
gets too high we use signals in an inverse manner, the red and green
lights being thrown on. If it shows a dangerous rise in temperature
the red light is thrown on. All of these lights are accompanied by
the ringing of a loud bell in the heat-treating department, which
automatically attracts the attention of the man operating the
furnaces, who at once inspect their individual furnace lights to see
if their temperature is correct."

DETERMINATION OF THE CRITICAL RANGES

Critical Ranges. — The practical importance of knowing the
exact location of the critical ranges of steel to be treated is obvious.
Their determination by means of pyrometers is based upon the fact
that the changes taking place in the steel at those temperatures
involve an absorption of heating during heating (the decalescent
points) and a giving out of heat on passing through these ranges on
cooling (the recalescent points).

Decalescent vs. Recalescent Points. — Before discussing methods,
it should be stated that, for the majority of heat-treatment work,
it is more important to know the location of the decalescent points
than that of the recalescent points. This is for several reasons. To
effect a complete change of the original structure of the steel, it must
be at least heated slightly beyond the Ac3 range, regardless of the
position of the Ar3 range. If the steel were to be heated only to the
Ar3 range, a complete change in structure cannot take place, because
the Ar3 range is always below the temperature of the Ac3 range.
Further, the position of the Ar ranges is, experimentally at least,
dependent upon the maximum temperature to which the steel is
heated, upon the length of heating at that temperature, and upon
the rate of cooling from that temperature.

It should also be again stated that the determination of the
upper critical range is of more importance than that of the lower
critical range (Al), since the majority of hardening and annealing
work demands a complete change of structure — which is obtained
only above the upper critical range (Ac3).

Temperature Difference Instruments. — American-made instru-
ments for determining the critical ranges of steel are based either

418

STEEL AND ITS HEAT TREATMENT

upon a temperature difference basis, or upon a direct record of a
single instrument. The method used by the Leeds & Northrup
apparatus, typical of the first class, involves the following points:

Two bodies are heated together in the same furnace, the one
being the steel under test and the other being a body which will

Q&f. — .3
Phos,= .033
Mng. — .700
Sin. = .253
Sir. = .029

Heating.Curve Cooling Curve

Abscissae — Temperature Differences between Sample and Non-recalescing Body.
FIG. 222. — Transformation Curves. (Leeds & Northrup Co.)

heat uniformly without undergoing any changes. If the bodies
are in sufficiently close contact they will heat at the same rate and,
barring changes in one which do not occur in the other, will remain
equal in temperature. When, however, the steel undergoes an inter-
nal change involving absorption or liberation of heat, its temperature
changes relatively to the other body and a temperature difference is
set up between the two. Hence the apparatus for the location of

PYROMETERS AND CRITICAL RANGE DETERMINATIONS 419

critical points by this method is designed to do two things: first,
to measure the temperature of the sample; second, to indicate the
temperature relationship between the sample and the unchanging
body. A curve using temperatures as ordinates and temperature
differences as abscissae is the best way of making use of the results.

Temperature Difference Records. — Fig. 222 is a reproduction of
such a plot. From the start of the test until 1205° the temperature
difference is small and constant. When the temperature of 1369°
is reached a sudden increase in the temperature difference takes
place, the Acl range. As soon as this sudden change ceases (i.e.,
transformation is completed), the sample and the unknown begin

FIG. 223. — Critical Range Curve on a Direct-reading Apparatus. Carbon,
0.44 per cent.; Manganese, 0.53 per cent.; Phosphorus, 0.035 per
cent.; Sulphur, 0.025 per cent.; Silicon, 0.028 per cent.

to equalize in temperature and the record of their decreasing differ-
ence follows a typical cooling curve between 1380° and 1455°, except
at about 1407°, where the Ac2 transformation begins to affect the
record. At 1407-1410° this Ac2 change is completed. Again at
1450° there is a departure from a smooth curve; this is the beginning
of the third transformation, which transformation is not completed
until about 1495°. This is the Ac3 transformation. On cooling,
the reverse takes place, except that the two upper points occur
closer together and appear as one. The lowest range is clear cut.

Direct-reading Instruments. — Fig. 223 shows a record obtained
from a Bristol instrument. Leaving aside a discussion of the

420 STEEL AND ITS HEAT TREATMENT

scientific pros and cons, the three main objections to this class of
instrument are: (1) the small area covered by the record, involving
less accuracy; (2) lack of that degree of sensitiveness which is
necessary to bring out the upper critical ranges; and (3) a curve
showing direct temperatures instead of temperature difference.

Practical Method for Determining Critical Ranges. — For plants
which have to determine the critical ranges but infrequently less
costly apparatus may be used. The outfit should consist of a thermo-
couple made of small wires so as to respond quickly to any slight
variation in temperature; the necessary leads; and a sensitive
milli voltmeter or pyrometer with a finely divided scale. This
instrument may also be used as a standard, or checking instrument,
for calibration work. The specimens to be tested should be small
so as to heat uniformly and quickly. These may be either a small
cylinder, say f in. diameter by If in. long, or duplicate pieces each
1J in. long by f in. wide by J in. thick. In the former case the end
of the couple is inserted in a small hole drilled through the axis of
the cylinder to a depth of about \ in.; in the latter case the pieces
are clamped together, one on either side of the end of the thermo-
couple so as to form a tight contact. The specimen is then heated
in any convenient manner, readings being taken every few seconds
as the critical ranges are reached. When the indicated temperature
is well above the upper critical range, the specimen is removed from
the heat, allowed to cool not too rapidly, and readings taken to
obtain the Ar ranges. The temperature readings, or difference in
readings, should then be plotted against the time to obtain the
necessary curves.

PYROMETERS AND CRITICAL RANGE DETERMINATIONS 421

TEMPERATURE CONVERSION TABLE

BY DR. LEONARD WALDO
Reprint from Metallurgical and Chemical Engineering.

c. °

40 50

-200
-100
- 0

F. °
-328
-148
(- 32

-346
-166
+ 14

F. °
-364
-184
- 4

F. °
-382
-202
- 22

F. °
-400
-220
- 40

F. °
-418
-238
- 58

F. °
-436
-256
- 76

F. °
-454
-274
- 94

F. °

-292
-112

F. °

-310
-130

104

122

140

158

176

194

C. °

F. °

100
200
300

400
500
600

700
800
900

212
392
572

752

932
1112

1292
1472
1652

230
410
590

770
950
1130

1310
1490
1670

248
428
608

788
968
1148

1328
1508
1688

266
446
626

806
986
1166

1346
1526
1706

284
464
644

824
1004
1184

1364
1544
1724

302

482
662

842
1022
1202

1382
1562
1742

320
500
680

860
1040
1220

1400
1580
1760

338
518
698

878
1058
1238

1418
1598
1778

356
536
716

892
1076
1256

1436
1616
1796

374
554
734

914
1094
1274

1454
1634
1814

1
2
3

4
5
6

7
8
9

1.8
3.6
5.4

7.2
9.0
10.8

12.6
14.4
16.2

18.0

1000

1832

1850

1868

1886

1904

1922

1940

1958

1976

1994

1100
1200
1300

1400
1500
1600

1700
1800
1900

2012
2192
2372

2552
2732
2912

3092
3272
3452

2030
2210
2390

2570
2750
2930

3110
3290
3470

2048
2228
2408

2588
2768
2948

3128
3308
3488

2066
2246
2426

2606
2786
2966

3146
3326
3506

2084
2264
2444

2624
2804
2984

3164
3344
3524

2102

2282
2462

2642
2822
3002

3182
3362
3542

2120
2300
2480

2660
2840
3020

3200
3380
3560

2138
2318
2498

2678
2858
3038

3218
3398
3578

2156
2336
2516

2696
2876
3056

3236
3416
3596

2174
2354
2534

2714
2894
3074

3254
3234
3614

F. °

C. °

1
2
3

4
5
6

7
8
9

10
11
12

13
14
15

16
17

.56
1.11
1.67

2.22
2.78
3.33

3.89
4.44
5.00

5.56
6.11
6.67

7.22
7.78
8.33

8.89
9.44

10.00

2000

3632

3650

3668

3686

3704

3722

3740

3758

3776

3794

2100
2200
2300

2400
2500
2600

2700
2800
2900

3812
3992
4172

4352
4532
4712

4892
5072
5252

3830
4010
4190

4370
4550
4730

4910
5090
5270

3848
4028
4208

4388
4568

4748

4928
5108
5288

3866
4046
4226

4406
4586
4766

4946
5126
5306

3884
4064
4244

4424
4604

4784

4964
5144
5324

3902
4082
4262

4442
4622
4802

4982
5162
5342

3920
4100
4280

4460
4640
4820

5000
5180
5360

3938
4118
4298

4478
4658
4838

5018
5198
5378

3956
4136
4316

4496
4676
4856

5036
5216
5396

3974
4154
4334

4514
4694
4874

5054
5234
5414

3000

5432

5450

5468

5486

5504

5522

5540

5558

5576

5594

3100
3200
3300

3400
3500
36CO

3700
3800
39CO

5612
5792
5972

6152
6332
6512

6692
6872
7052

5630
5810
5990

6170
6350
6530

6710
6890
7070

5648
5828
6008

6188
6368
6548

6728
6908
7088

5666
5846
6026

6206
6386
6566

6746
6926
7106

5684
5864
6044

6224
6404
6584

6764
6944
7124

5702
5882
6062

6242
6422
6602

6782
6962
7142

5720
5900
6080

6260
6440
6620

6800
6980
7160

5738
5918
6098

6278
6458
6638

6818
6998

7178

5756
5936
6116

6296
6476
6656

6836
7016
7196

5774
5954
6134

6314
6494
6674

6854
7034
7214

C. °

EXAMPLES: 1347° C. 2444° F. +12°.6 F. = 2456°.6 F.: 3367° F. = 1850° C. +2°.78 C.:
1852°.78 C.

INDEX

Abrasion, resistance to, 14

Abrasive wear in Mn steels, 346

Air control, 182

Air cooling, 42, 58, 249, 276

Air hardening, 246, 249

Allotropic ferrite, 29

Alloy steel:

hard spots in, 51

necessity for heat treatment, 1, 258
Alpha ferrite, 29
Alternating impact tests, 10
American gas furnace process, 147
Animal charcoal. See Charcoal.
Annealing. See Ch. III.

air cooling after, 58

commercial, 62, 248

definition of, 39

effect of, 44

elemental considerations in, 39

furnace cooling after, 57

hyper-eutectoid steels, 254

length of, 48

pit, 58

rate of cooling after, 53, 57

rate of heating in, 46

size of object, 58

slow cooling after, 58

special methods, 58, 59

temperature for, 47, 231, 234, 238,
247

time experiments, 51

vs. toughening, 110

wire, 398
Armor plate, 319
Arrangement of charge, 214
Ar ranges, 32

Atmosphere in furnace, 189
Austenite, 27, 67, 231, 348

Automatic furnaces:

for die blocks, 373

for shrapnel, 249
Automobile steel, 1, 234, 329
Axles, 1, 8, 9, 90, 236, 241, 243, 245,
247, 315, 329, 340

Ball-bearings, 300

Ballistic tests, 14

Barium carbonate, 127, 133, 137

Baths:

heating, 75
. salt, 76

tempering, 99
Best case, 162, 165
Beta ferrite, 31
Bit steel, 394
Bolts, carburizing of, 141
Boring, hollow, 88
Box annealing, 61
Boxes for carburizing, 140
Brains, purchasing of, 184
Brine, 80
Brinell hardness. See Hardness.

of carbon steels, 229

of chrome-nickel steels, 318, 320

of chrome-vanadium steels, 337

of nickel steels, 290
Brittleness, 3, 7, 8, 9, 110, 233, 240,
246, 268, 270, 271, 275, 295, 306,
344, 353

Stead's, 61
B.T.U. values, 175
Burners, 186

Calcium chloride for quenching, 81
Calibration of pyrometers, 415
Capacity of the steel, 106

423

424

INDEX

Carbides, 296, 304, 335, 348
Carbon :

concentration of, 122, 271

direct action in carburization, 114

maximum in case, 165, 276

plus carbon monoxide, 124

solution of in carburization, 125
Carbon content:

for tools, 358

influence of, 3, 110

in manganese steels, 345

in nickel steels, 264
Carbon monoxide, 115, 117, 122
Carbon steel:

under 0.15 per cent., 230

0.15-0.25 per cent., 232

0.25-0.35 per cent., 236

0.35-0.40 per cent., 240

0.45-0.60 per cent., 246

over 0.60 per cent., 253
Carbonates, 116
Car bottoms, 219
Carburization :

boxes, 140

carbon monoxide plus hydrocar-
bons, 122

carbon plus carbon monoxide, 124

depth of penetration, 126

effect of chrome, 297

gas process, 147

heat treatment requirements, 154

object of, 112

of chrome-nickel steels, 308

of nickel steels, 267, 270, 271, 273

requirements of, 112

steel for, 113, 232

temperature of, 125, 132, 133, 134,
155

with carbon monoxide, 115, 117

with simple solid cements, 135

wood charcoal, 115
Case carburizing. See Carburizing.
Case hardening:

gears, 386-389

maximum efficiency in, 165

treatment of hyper-eutectoid steels,
156

treatment of hypo-eutectoid steels,
155

Case, the best, 162, 165
Castings, 397
Cellular structure, 42, 53
Cementite, 16, 24, 63, 114, 156, 163,
165, 169, 253, 295, 301, 335, 391,
402

Centigrade tables, 421
Central pyrometer systems, 415
Chamber, height of, 202

twin- furnaces, 226
Changes:

in diameter, 366

in heating, 32

in length, 366
Charcoal, 133, 135, 136
Charge :

height of, 202

influence in heating, 205

influence of arrangement, 205

placement of, 75
Charging, 39

Chemical composition, effect of, 1
Chipping chisels, 368
Chisels, 298, 368
Chrome:

influence in carburization, 297, 308

in manganese steels, 347

vs. silico-manganese, 351
Chrome steels:

general characteristics, 295

0.5 chrome, low carbon, 296

0.5 chrome, 0.35-0.50 carbon, 297

0.5 chrome, over 0.50 carbon, 297

1.0 chrome, 300

2.0 chrome, 302

high chiome, 302
Chrome-nickel steels:

carburization, 309

gears, 390

heat treatment, 309

low Cr, low Ni, 310

0.5 Cr, 2.5 Ni, 322

0.6 Cr, 3.5 Ni, 321

0.75 Cr, 3.0 Ni, 322J

1.0 Cr, 1.75 Ni, 322

l.SCr, 3.5 Ni, 319

Mayari, 329

special analyses, 327

vs. chrome-vanadium, 306, 335

INDEX

425

Chrome-vanadium steels. See Vana-
dium.

Circulation for cooling oil, 84
Classification of:

gear steel, 386

heat treatment after carburization,
155

nickel steel-, 258
Coal furnaces, 218
Coffin process, 244
Cold crystallization, 231
Cold-end temperatures, 410, 412
Cold rolls, 302
Cold rolling, 38

vs. strength, 6

Cold work, effect on structure, 38
Color chart, 369
Colors in tempering, 97
Combustible mixture, 177
Combustion, furnace atmospheres

from, 189

Commercial annealing, 39, 62, 248
Commercial data in carburization, 135
Commercial ratio of chrome and nickel ,

307

Compensation of pyrometers, 412
Compressed air in quenching, 85
Compressive strength, 5
Conservation of heat, 222
Contact couples in heating, 53
Continuous furnaces, 249, 252, 373
Contraction in hardening, 88
Contraction of area, 5
Conversion, temperature, 421
Cooling, after annealing, 53, 231, 346
Cooling the oil bath, 82
Cooling the water bath, 82
Corrosion, 295
Couples, 409
Cracking, 87, 246.
Crank shafts, 1, 329
Critical ranges, 25, 417

changes at, 41

effect of manganese, 345^

effect of nickel, 258, 265~

heating over the, 42

merging of, 31

of chrome steel, 295, 299

of chrome-nickel steel, 309, 327

I Critical ranges of high-carbon steel, 253
of hyper-eutectoid steel, 62
of manganese steel, 350
of tool steel, 363

Cutters, 378

Cyanide hardening, 149

Cyanides in carburization, 117, 139

Dead soft steel, 230
Decalescence, 417
Deck plate, 309, 319
Depth of penetration, 272
Design of furnace, 186
Determination of critical ranges, 417
Diameter, effect on tests, 229

changes in, 366
Die Blocks, 298, 369
Dies, 247, 382.
Differential hardening, 82
Diffusion, 44, 50, 232, 267
Distortion, 364
Distribution of carbon, 126
Door heights, 200
Double annealing, 59
Double carbide steel, 304
Double quenching, 94, 163, 237, 254,

276

Double regenerative quenching, 166
Drawing dies, 302
Drawing of wire, 398
Drilling, hollow, 88
Drills, 298, 376
Drop tests, 8

Ductility, 5, 7, 106, 257, 306
Duplex process, 329
Duplication of results, 2, 107
Dynamo sheet iron, 352
Dynamic strength, 2, 110, 236, 306,

307, 319, 337

Effect of:

chrome, 295, 306
manganese, 344
mass, 286, 322, 330, 363
nickel, 257, 306
silicon, 350
vanadium, 335

426

INDEX

Elastic limit, 3, 5, 41
Electricity:

atmospheres with, 190

for heating, 191
Electromagnets, 352
Elongation, 5, 41
Endurance, 6, 10
Enfoliation, 120, 271, 279
Engine forgings, 232, 234
Engraved, dies, 374
Equalization, 44, 48, 267, 269
Equalizing action of carbon monoxide,

124

Equipment, pyrometer, 411
Eutectoid for nickel steel, 267

steel, 17
Expansion in hardening, 88

Fahrenheit tables, 421

Failures of heat-treated axles, 246

Fatigue, 2, 6, 9, 236, 307

Ferrite, 23, 29, 164, 257, 258, 296

Ferro-cyanides, 127, 139

Files, 298, 302, 379

Fine-grain annealing, 59

Fire-ends, 409

Five-ply steel, 396

Flanges, 232

Flue construction, 222

Force, 2

Forging, 406

Forging temperatures for tool steel,

363

Fragility, 9
Frames, automobile, 1
Fuel:

cost of delivering, 180

costs, 175

efficiency, 177

equipment, 181

fluid, the, 178

oil, 182

oil, air control with, 182

selection of, 179

supply, 181

the right, 177

uniformity of, 182

Fuel:

vs. furnace design, 209

vs. operations, 178

vs. product, 191
Furnace:

atmospheres, 189

batteries, 226

cooling in toughening, 108

design, 186, 209, 215, 407

equipment, 74, 185, 192

guarantees, 196

plans, 223

temperature of in heating, 46

the one, 195
Furnaces :

automatic, 250, 373

car-bottom, 219

carburizing, 142

coal, 218

continuous, 250, 373

forge, 407

general considerations, 227

muffle, 207

overfired, 213

perforated arch, 212

practical notes, on, 227

semi-muffle, 209

shrapnel, 250

twin-chamber, 226

underfired, 196, 220, 249

unit system, 225

Gamma ferrite, 31

Gases, action of, in carburization, 114

regulating the, 39
Gears, 1, 141, 236, 247, 329, 351,

386

Grade, in tool steel, 357
Gradual cements, 133, 139
Grain size, 268

at Ac3, 41

beyond Ac3, 33, 42

by different rates of cooling, 42

effect of work on, 38

in carburization, 134
Gun barrels, 235

forgings, 91, 241,354

INDEX

427

Hardening:
cyanide, 149
definition, of, 65
differential, 82
heating for, 71
nickel steels, 267
pack, 148, 152
strains, 97
superficial, 148
temperature for, 70, 232, 237, 241,

253, 274, 286, 336, 370
tool steel, 362
vs. annealing, 67
Hardness, 11, 230, 272, 290, 295, 318,

320, 337, 360

Brinell, 11. Also see Hardness,
cutting, 254
due to chrome, 297
scleroscope, 13, 170. Also see

Hardness,
wearing, 254
Hard spots, 51
Heat, quality of, 188
Heat application, 40, 195, 249
Heat conservation, 222
Heat reservoir, 201
Heat treatment:
definition of, 25
growth of, 1
necessity for, 1
Heating:

changes on, 32, 40, 65

costs, 173

distinctive conditions in, 173

factors, 174

for carburization, 143

for forging, 406

for hardening, 71

for tools, 367

in lead, 379

in salt, 76, 168

influence of chrome in, 297

large sections, 46

length of, 268, 370

prolonged, 306

rate of, 46, 51, 52

uniform, 196

unit, standard, 174

Heating:

with electricity, 191
Height of chamber, 202
Height of charge, 202
High-carbon case, treatment of, 160
High-speed steel, 353
High temperature carburization, 134
High temperatures, effect on springs,

391

Hollow boring, 88
Hollow tools, 383
Hot-ends, 410
Hot work, effect of, 38
Human element, 39, 74, 183, 237, 245,

406, 408

Hydro-carbons, 118, 122
Hyper-eutectoid steel, 17

annealing of, 62

zones in carburization, 120
Hypo-eutectoid steel, 17

annealing of, 40

Impact strength, 2, 8, 9, 106, 134, 170

Impurities in carburization, 114

Influence. See Effect.

Intensifies, 337

Intermediary types of carburized zones,

122
Interrupted regenerative quenching,

168

Jar steel, 298

Knives, 298

Laminations, 257

Lead baths, 75, 102, 379

Ledges, 197

Length, change in, 366

of heating, 39, 370
Levers, 232
Liquation, 129, 272
Locomotive axles, 243
Low-temperature carburization, 133

428

INDEX

Machine parts, 240
Machinery steel, 232
Machining:

effect on structure, 38

quality, 257
Magnet steel, 353
Magnet, use in hardening, 72
Manganese:

in carburization, 113

on hardening, 95

on machining, 231

steels, 344
M'artensite, 68
Martensitic steels, 258, 302
Mass, influence of, 202, 229, 286, 287,

322, 341, 363
Mayari steel, 329
Mechanical mixture, 16
Mechanical work, effect in annealing,

48, 59

Microscope, use of, 44
Microstructure:

of high-carbon steels, 255

of nickel steels, 259
Milky-ways, 50
Milling cutters, 378
Millivoltmeters, 410
Mineral hardness, 306
Mixed cements, 272
Molybdenum steels, 353
Motion in hardening, 74
Muffle furnaces, 207

Natural alloy, 329

Natural, steel in the, 1, 3

Navy specifications for tool steel, 360

Network, 34, 42, 53, 63

Nickel:

effect on physical properties, 264

influence of, 267

influence on critical ranges, 265
Nickel steels:

2 per cent., 264, 274

3.5 per cent., 274, 276, 310, 315

5 per cent., 264, 267, 268, 276, 280,
291

Nickel Steels:

10 per cent., 264

25-35 per cent., 264, 291

carburization of, 270

for gears, 360
Nickel-chrome steel. See Chrome

nickel.

Nickel-vanadium steel, 342
Nitrogen in carburization, 116
Normalizing, 63
Nuts, 232

Obstructing agents, 57, 258

Oil. Also see Fuel Oil.
quenching speed of, 78
vs. water for hardening, 243

Oil baths, 82, 101

Oil burners. See Burners.

Oil tempering, 80

Oil-tempered gears, 386, 389

Oil-well bits, 394

Operators, value of, 184

Oscillating temperatures, 132

Osmondite, 67

Overfired furnaces, 231

Overheating, 72, 402

Oxidation, protection from, 371

Oxygen, action in carburization, 115

Pack hardening, 148, 152
Packing for carburization, 141
Patenting, 402

Pearlite, 16, 23, 27, 56, 110, 258, 265
Penetration, depth of, 126, 272
velocity of with chrome, 297
Perforated arch furnaces, 212
Phosphorus, 3

Physical properties at Ac3, 41
Pins, 232
Pit annealing, 58
Polyhedral steels, 259
Potentiometers, 410
Preheating, 40, 370
Process annealing, 399
Producer gas, 181

INDEX

429

Prolonged heating of nickel steels, 271
Propeller shafts, 329, 354
Protection of steel, 60, 371
Protective deck plate, 309, 319, 324
Punches, 382

Punching, effect on structure, 38
Purchasing brains, 184
Pyrometers, 408

standardization, 413

use of contact couples, 53

Quality of heat, 188
Quality of product vs. first cost, 173
Quench-toughening, 111
Quenching :

after tempering, 99

baths, 77

best temperatures, 71

double, 94

manner of, 90

media, 77, 109

special methods, 80

speed, 77

tanks, 87

water for, 80

Radiation systems for cooling oil, 84

Eate of cooling, 39, 248

Rate of heating, 39

Razors, 298

Reamers, 382

Recalescence, 417

Reduction of area, 5, 41

Refinement, 33, 39, 41, 65, 160, 231,

232, 237, 238, 248, 401
Regeneration, 159, 160, 180, 274, 280
Relation of austenite to carbide, 347
Relation of physical tests, 10
Requirements of gears, 386
Resilience, 8
Rifle barrels, 354
Rings, 383
Rivet sets, 384
Roller bearings, 300
Rotary bending, 6
Rounds, hardening of, 91, 93

Safe steel, 396
Salt, use in carburizing, 137

standardization of pyrometers, 413
Salt baths, 76, 102, 168
Sand baths, 100
Saws, 298, 385

Scleroscope, 13, 130, 290, 318, 320, 337
Screw stock, 240
Screws, carburizing of, 141
Seams, 257

Selection of pyrometer equipment, 413
Selection of tool steel, 357
Sensitiveness of manganese steels, 344
Sensitiveness of silico-manganese

steels, 351
Shafts, 90, 290
Shock, resistance to, 275
Shore-hardness. See Scleroscope.
Shrapnel, 249
Silicon steels, 350, 352
Silico-manganese steels, 351, 390
Size of section, 58, 229, 248
Slow cooling, 33, 42, 57, 67, 107, 157,

167

Soft forging steel, 236
Solid solutions, 27

Solution of carbon in carburization, 125
Sorbite, 56, 59, 69, 103, 110
Specifications for tool steel, 360
Spheroidal cementite, 63, 163, 402
Spheroidal ferrite, 164
Spheroidalizing, 63, 163, 164, 254, 355
Springs, 1, 391
Static strength, 1, 2, 24, 319
Standard heating unit, 174
Standardization of results, 12, 110
Stead's brittleness, 61
Steam hardening, 246
Steel:

burnt, 42

castings, 397

for carburization, 113

nature of, 16
Steering parts, 1

Stresses and strains, 2, 6, 39, 97, 107
Structure, definition of, 25
Structure of slowly cooled steel, 17
Sudden cements, 133, 139

430

INDEX

Suddenly applied loads, 7
Sulphur diffusion, 143
Summary for case hardening, 169
Superficial hardening, 148

Table for temperature conversion, 421
Tank, size of quenching, 87
Taps, 366
Temper, 358

colors, 97, 93
Temperature :

conversion table, 421

effect on grain size, 33

effect on network, 34

relation of surface and interior, 53
Temperature of:

annealing, 39, 47

carburization, 128, 132, 155

hardening, 70, 72, 274

pack hardening, 152

quenching bath, 77

toughening, 104
Tempered axles, 244
Tempering:

color for tools, 361

definition, of, 96

for depth, 98

gears, 389

handling material in, 101

methods, 100

oil, 80

plate, 100

quenching after, 99

springs, 392
Tensile strength, 2, 41

of cementite, 24

of ferrite, 23

of pearlite, 23
Testing:

comparative results, 229

errors in, 307

purpose of, 1
Tests from center, 229
Thermo-couples, 409
Threading, treatment for, 231
Tie rods, 232
Time of heating, 231, 409

Tool steel, annealing of, 59, 60

proper carbon for, 358

selection of, 357
Torsional strength, 1, 5
Tough-hardness, 295
Toughening, 103

high vs. low temperature, 109

range, 1,04

temperature vs. mass, 330

vs. annealing, 110

vs. ductility, 106

vs. impact strength, 106
Toughness, 231, 360
Transference numbers, 12, 230, 290,

318, 320, 337

Transition constituents, 56
Troostite, 68, 96
Tungsten steel, 350, 353
Twin chamber furnaces, 226

Underfired furnaces, 220
Underfiring, 196
Uniform heating, 196
Unit furnace system, 225
Uses of chrome nickel steel, 329

Value of furnace operator, 184
Valve stems, 232
Vanadium, effect of, 335
Vanadium steels:

gears, 390

nickel, 342

Type A, 339

Type D, 340

Type G, 341
Vault steel, 396
Velocity of penetration with chrome,

297

Vents, 198, 199
Vibratory stresses, 1

Warping, 89, 232

Water bath, cooling the, 82

Water quenching, 80, 246

INDEX

431

Water spray, 79
Water toughening, 349
Water vs. oil for hardening, 243
Wear, 1, 14, 240, 302, 346
Welding of alloy steel, 330
Welding properties of tool steel, 363
Well bits, 298, 394
Wire, 398

Wood charcoal, 115, 127
Work, effect on grain size, 38
Working conditions, influence of, 179
Working strength, 4
Works, annealing, 399

Yield point, 4

FOURTEEN DAY USE

RETURN TO DESK FROM WHICH BORROWED

This book is due on the last date stamped below, or

on the date to which renewed.
Renewed books are subject to immediate recall.

J9B95J

ilVED

JUL16'69-1

LOAN DEPT.

LD 21-100m-2,'55
(B139s22)476

General Library

University of California

Berkeley

THE UNIVERSITY OF CALIFORNIA LIBRARY

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Steel and its heat treatment"

See other formats