STEEL
AND ITS HEAT TREATMENT
BY
DENISON K. BULLENS
*
Consulting Metallurgist
FIRST EDITION
FIRST THOUSAND
NEW YORK
JOHN WILEY & SONS, INC.
LONDON: CHAPMAN & HALL, LIMITED
1916
Copyright, 1916
BY
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.
v
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.
X
Lbs. per Sq. In.
On Basic Steel.
y
Lbs. per Sq. In.
0.05
not
0.10
80*
130
0.15
120
150
0.20
160
170
25
200
190
0.30
240
210
35.
280
230
0.40
320
250
0.45.
360
0.50
400
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
d
o>
ta
fl
o
,0
>*:
/
/
/
/
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
19
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.)
20
STEEL AND ITS HEAT TREATMENT
IG. 4. 0.32 per cent. Carbon. Ferrite (White) and Pearlite (Dark).
(Ordnance Dept.)
X75.
JP IG 5, 0.49 per cent. Carbon. Ferrite (White) and Pearlite (Dark). X75.
(Ordnance Dept.)
THE STRUCTURE OF STEEL
21
0.57 per cent. Carbon. Ferrite and Pearlite. X75. (Ordnance Dept.)
^
M
FIG. 7. 0.71 per cent. Carbon. Ferrite and Pearlite. X75. (Ordnance Dept.)
22
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.
24
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
25
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> ^^"
4
Y/
'/
/
fJNl
<:::: ^^
<<
x
*Y\ '
f/
Jl
/
\
/
/
/
/
>
/
^
V
>
/
/
/
/
/
/
/
X
\
'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
26
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
/
1
g 1600
\
.
/
\
J*
/*
^3
Vf
i
^<
5
s
\
o
b i r >oo
V
>
a
\
/
g
\
4
3
A2
\
s
1
2
^
^.
j/
A 1-2-3
1300
la
- v
lb
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
28
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
Y
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
29
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."
30
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
35
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.)
36
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
37
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.
39
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.
42
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
43
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
45
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
47
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.
48
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.
52
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
I 5
o
I
300
X
/
4"
8"
Size of Section.
FIG. 41. Curves Derived from Rate of Heating.
thrup Co.)
(Courtesy of Leeds & Nor-
F.
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
55
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.
56
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
65
66
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
ii
in
IV
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 -'> '.
v4 v r ><
*^ *
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
70
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
73
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
F.
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
79
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
83
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-
II
r
a i
IV
II
.
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
VJ
1 N
1 1
1 D )
^1
L^
( C II
1 r
V 1
r\
1 1
i p )
/I - .
' - [y
1 S i 1
i r
M
r\
A u
' ?J
(t- n-
\ j
L
\
\j ....
_ L
s
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 procedurethis 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.
86
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
89
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
2.
99,750
60,000
15.0
35.3
3.
104,500
65,000
17.5
35.7
4.
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
93
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
94
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
06
TEMPERING AND TOUGHENING
97
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-
Temper-
ature, De-
frees
Color.
ature, De-
frees
Color.
ahr.
*
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.
14
8
420
28
8
490
15
8
430
38
8
510
16
8
440
60
8
530
17
8
450
96
8
550
18 5
8
460
200
8
560
20
8
470
Melted lead
610 to 630
24
8
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
.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
27
.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 VI 1
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.
1 Cuts 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 < C0 2 + 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.' ...dr. .."-*.,'-. ?- * V."^"iV. >ty
FIG. 69. Carburization of 0.11 per
cent. Carbon Steel at a Temper-
ature under the Upper Critical
Range with BaCO 3 and Char-
coal, for One Hour. (Nolly
and Veyret.)
>' ** * '!A** V * ' ^ "* V * V*'^
! K"*\ > * " '^ f 'C^i^
^ *" * V; "'' ^*SJ y *' 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
*& - - - &/.: t -"^: *^
FIG. 74. Carburization with Char-
coal for One Hour at 1560F.,
0.11 per cent. Carbon Steel.
(Nolly and Veyret.)
?-i/vv:
y.vtj": r r ..,, - -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
14
16
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-
o
b
I
I
I
1
. . .
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
25
65
to
45
Auto, lever
0.18
0.40
1650
800
70,030
45,400
32
64
Pressed auto, frame.
0.22
0.40
1530
800
71,950
43,400
29
56
Engine forging
26
0.28
1650
1025
77,210
52,200
28
65
Old rolled Y in. plate
0.24
0.60
1525
900
93,300
65,250
20.5
51
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
30
64,960
34,050
38.0
61.0
1500
20
65,180
32,930
38.3
58.3
1500
105
64,060
33,150
39.1
62.3
1830
15
62,940
31,810
35.7
56.3
2120
5
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.
C.
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
under
0.045
70,000
to
85,000
38,000
to
50,000
28
to
20
55
to
40
Forged gun jacket
acid steel
0.35
0.25
0.038
0.019
77,080
39,500
27
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
28
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 85,00050,0002245. 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 ^1 D 1 ^
(g) aequinx saonp-rcH nauug
1 1 1 1 1
Annealed
i
5
1
1
i
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.
/
\
1
/
\
ffl
1
\
/
\
/! 1
\
/
/
\
/
\
\
/
i
\
\
/
I
\
/
A
\
\
/
/
/
\
/
^
\
/
\
/
/
^
V
/
'
/
X
\
]!:ml.md
1 i 1 1 1
1 1 3 ' 1
tpni 9Jimbg aod spunoj '(z) iitniq o^stJig PUB '(x) m3naj?s ^Ijanax
"= 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 500 o 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,
wa ter/1200 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 1337 C
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.
20
Austenltic
Martensitic
10
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*$*1S
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
F.
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.
2%
3%
5%
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.
2
1.50
5
0.95
3
1.30
6
0.85
4
1.10
7
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
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,
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
40
39
39.37
Cemented pieces, quenched from 1830 F.
84
69
78.05
Cemented pieces, quenched from 1380 F.
88
85
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 . 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
g
'(l)qiST:
'(5) uaay jo noponpaH pau
NICKEL STEELS
283
(9) aaqtunjj ssanpjwjj adoaso-ispg
8 g
7
\/
TV
lull.
li
"S-S
I
!!
J35
\
bg asd spunoj
fuaO 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 8U 8UO X
2
s
NICKEL STEELS
285
(9) jaqtnnjj ssanpavH adoosojspg aaoqg
(g) jaq
7
\
/
7
iot$
1 -2 of a
1400
n as
\
Is
^
1600^
De^r
*aJj
W,,,.,. 12001
e*
Fahr
III
K
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
!
5
I
J
o
I
g
f
i
o
O*
|
g
' Tl ' V
g
"S
o
M
e3
"H
3
g
i
1?
3
. a>
"'
. 9
O
CQ
i
S
|
g Sill
c
0)
H
1
S-
BB
"3
V
a
d
i
H
08
G
IO
o
*
5
13 a> "3 *" "3
(-. ; +j a>
a c cs c!
15
o>
C
~
cd
I
a
5
S
3
jt?T
w
5
q)
^
o
O
w
W
<
*
^
H
S
<5 O < W
1
|
1
g
1
3
CO
^
4) 0) S
&? ? : : S
e3 cS . -
i
<
i
09
00
u
a
a
SB
OB
i
OB
0)
OB
OB
d
3
0?
^ d o O
1 i. * ! i
1
ft
8
e
H
H
H
H
H
m
~f.
S
9
*
a
w M w w o
K
H
o>
^
J
J
j-
j
O
O
a
Tr
g -^
o
$
^
3
p
00 ^5
^
^
H
od
QQ
OQ
OQ
QQ
^
<
S
*-
fc
O O JOQ ad i5
ai
of
g
g'
.C
'C
c
C
x x f
1
O
d
s
o
'C
o
d
S
d
N
d
I
I
O
o
1
o
CO
i
CO
S S o d o
ic c o
IfiflJ
fl!Hi
mill
!
lul
S19 | *S8
oo ^ 5c>o
d /
/
/
\
/
/
\
/
1
/
1
/
/
i
1
Hardened 1
-1
01
i
1
1
1 1
gill
qoaj anmbg i^d spanoj *(j) is^n 3 JW B (3: POB '([) qi3a3ng 8(;saaj,
S* S s S
;aao wj (0 Baiy )o aoninpaa pao (?;) s0 (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
03
o o o o o o
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 C5 C5
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.
13
79,070
41,660
31.1
58.75
180
283
143
16
14
77,060
34,720
36.6
68.23
180
273
143
18
15
83,220
38,640
32.2
57.5
180
215
156
18
16
87,580
45,140
22.2
35.5
171
108
170
20
17
92,290
48,830
16.1
24.0
85
66
187
22
Heat Treated (Hardened in oil, reheated to 1025 F.)
13
82,080
53,760
30.9
65.6
180
239
163
15
14
105,350
75,450
25.3
54.4
180
226
351
30
15
127,400
77,800
21.1
49.2
180
122
444
39
16
247,300
216,830
7.74
23.2
34
33
512
42
17
4
24
512
46
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.
I.
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
35-0 15
Phosphorus
Sulphur . . .
0.015-0
0.02-0
0.015-0
0.02-0
0.02-0
0.02-0
0.02-0
025-0
Silicon
0.40-0 10
0.40-0.10
0.40-0 10
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.
. 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
M
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>OOiOiO ^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
o
88
1> t^
Jt
%m
~~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 ;p d'"O 3 jS'd 3 p
6 'a'a3'a3'a3 t a3 p 73cro H cicrx
^ ^ ^ ^ -^ - 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
0025
0.00025
0.0025
i
0.0005
0.001
2
0.003
0.003
1
i
0.0015
0.002
3
0.0035
O.Q035
u
0.002
4
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 saw 7 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 . 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
47
370
181
600
136,440
46
50.5
388
60
835
131,017
43.5
49.5
351
86
1025
96,852
34.5
39
268
152
1230
105,400
34
37
282
167
Series B, Quenched in oil from 1650 F.
450
130,922
46
52.5
394
90
625
134,232
43
57
371
82
820
141,147
46
56
389
104
1025
126,320
42
50
371
108 .
1210
83,457
31
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 WIRE 1
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
|
us
J
00
CO
^|fi
00
1
s 8
00
2
P
if
!
*
&
2
- <
IN
CO
i
(N
HO
IN
BE
5
IN
CN
CO
fe fi
CO
1, g
R
fc <
O
00
IN
z
CO
Bfc
o
|2
1C
00
1C
-IN
fefl
*
H
1
(N
l
IQ
CO
1
CO
EQ
g
PROCESS
AN-
NEALED
o
o
CO
O
CO
(N
4M
00
o
IS
CO
Ho
IN
Eg
^
w* ?,
o
IN
og ^
CO
CONDITION
MATERIAL
PER CENT.
REDUCTION
BY DRAWING
TENSILE
STRENGTH
POUNDS PER SQ. IN.
PER CENT.
ELONGATION
IN TEN INCHES
^
ll
~ o
S?X3
31
'
o- 5
|
w
'S
Ll
H
H
CO
218 000
C5
II
O
CO
184 000
IN
Ij
156 000
O
ES
OQ
CO
208 000
00
ii
8"
-, CU
O-i
Ofi
H Q
CO
CO
1
05
. ts
o
co
156 000
(N
Ii
- 5jl
Q
o
128 000
00
t".
^MntHZ^tT 1 ^ f he ^ ffe ^ f cold drawin e 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<
is
CO
176 000
CO
(N
Q H
H Q
I
163 000
IN
1
>-i W
CO
00
(N
4
o
115000
(N
00
P
s
146 000
00
IN
IS
s
00
IN
122 000
os
O
CONDITION
OF
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
/
/\
/
/
/
/
\
/
/
1
I
I
J
/
7
s"
r
^
'-"""
-'
x
s*
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
]
/
J\
\
f
1
/
/
^
**
^
<<
"*'
^
i * 1
^
x*
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.
10
20
30
40 50
60
70
80
90
-200
-100
-
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
32
50
68
86
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
10
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
18
.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.
10
20
30
40
50
60
70
80
90
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
B
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
C
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
E
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
G
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
K
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
M
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
N
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
P
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
Q
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
S
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
U
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
W
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