close

Вход

Забыли?

вход по аккаунту

?

EFFECT OF 1/2 PERCENT OR LESS OF SOME ALLOYING ELEMENTS ON THE HARDNESS AND MICOSTRUCTURE OF PURE 1.1 PERCENT CARBON STEELS

код для вставкиСкачать
THE PENNSYLVANIA STATE COLLEGE
The Graduate School
Department of Metallurgy
Effect of 1/2 Per Cent, or Less, of Some Alloying Elements
on the Hardness and Microstructure of Pure 1.1 Per Cent
Carbon Steels
A Dissertation
by
Bertrand Saunders Norris
Submitted in partial fulfillment
of the requirements
for the degree of
Doctor of Philosophy
August, 194-0
*
Approved:
fgui,
1940 -
ftJUj IV 1940 -
C G G G j
,
, Dept, of Metallurgy
Head of the Dept.
Acknowl edgment
.
%
Acknowledgment is gratefully made to the Research Grant for
Fundamental Studies in. Metallurgy made by a group of steel companies.
For invaluable aid and guidance I wish to express ny apprecia­
tion to Dr. G.
r.
Austin.
I also wish to express my appreciation to
Dr. M. C« Fetzer and Dr. G. H- Samans for their helpful criticisms and
suggestions.
I am appreciative of the assistance rendered by Battelle
Memorial Institute in the preparation of the alloys; by the Latrobe
Electric Steel Company in the fabrication of the alloys; and by Bethlehem
Steel Company in the chemical analysis of the alloys.
Much assistance has been rendered by the National Youth Admin­
istration in the preparation of samples for heat treatment and metallograph
ic examination.
For this aid I am very thankful.
Many thanks are given to my fellow graduate students for their
helpful criticisms, suggestions, and assistance in the conduction of this
research.
The author also wishes to acknowledge the aid given by the
stenographers who labored so diligently in preparing the typewritten copies
of this report.
Table of Contents
Page Number
Chapter I.
Introduction................................
1
Chapter II.
Review of Literature.........................
3
Chapter III. Preparation of Steels.......................
Chapter IV.
Chapter V.
Chapter VI.
12
A.
Manufacture............................
12
B.
Chemical composition............
13
Effect of Tempering on Hardness and
Microstructure.................
15
A.
Experimentalprocedure..................
15
B.
Experimental
17
results.................
1. Austenitic grain size...........
17
2. Hardness.............................
17
C.
Analysis of hardness data..
21
D.
Micro structure.
27
E.
Discussion of results..................
28
F.
General conclusions.................... *
31
..................
Formation of Graphite at Sub—Eutectoid
Temperatures..........
34-
A.
Experimental procedure..................
34
B.
Experimental results.......
35
C.
Discussion of results.......
52
D.
Conclusions.........
61
Microstructure of Annealed Steels......
63
A.
Experimental Procedure..................
63
B.
Experimental Results...................
63
C.
Conclusions
65
.......................
Bibliography..............................................
^
List of Tables
Page Number
Table I.
Chemical Composition................... .
14
Table II.
Austenitic Grain Size after
1 Hour at 1000°C.....................
18
Table III.
Rockwell nB " Hardness...................
18
Table IV.
Slope and Intercept Values for
Brinell-log Time Curves....... ..... .
23
Chemical Composition of Carbides in
Low Chromium and ManganeseSteels
30
Graphite Analysis.......................
39
Table V.
Table VI.
List of Graphs
Page Number
Figure 1
Effect of Time of Tempering at 550° and 710°C on
Hardness of Carbon and High Alloy Steels, contain­
ing approximately 1/2 per cent of added element.
19
Figure 2
Semilogarithmic Plot Showing Relation Between Hard­
ness and Time of Tempering at 550° and 710°C for
Carbon and High Alloy Steels, containing approximate­
ly 1/2 per cent of added element.
20
Figure 3
Semilogarithmic Plot Showing Relation Between
Hardness and Time of Tempering at 550°, 590°,
630°, 670°, and 710°C for the Carbon, Sulphur,
Tin, Manganese, and Silicon Steels.
21
Figure 4
Semilogarithmic Plot Showing Relation Between Hard­
ness and Time of Tempering at 550°, 590°, 630°,
670°, and 710°C for the Nickel, Chromium, and
Copper Steels.
21
Figure 5
Semilogarithmic Plot Showing Relation Between
Hardness and Time of Tempering at 550°, 590°,
630°, 670°, and 710°C for the Aluminum Steels.
22
Figure 6
Effect of Temperature on Intercept C (Brinell Hard­
ness after tempering for 1 hour) for all steels.
23
Figure 7
Effect of Low and High Amounts of Alloy on Inter­
cept C (Brinell Hardness after tempering for 1
hour) at 550° and 710°C.
24
Figure 8
Chart Showing Rate of Change of Softening as a
Function of Hardness for All Steels when
Tempered at 710°C.
25
Figure 9
Chart Showing Comparison of Effects of 0.07 to 0.08
per cent, and approximately l/2 per cent Added
Element on the Rate of Change of Softening of
Alloys, as a Function of Hardness when Tempered
at 710°C.
26
Figure 10.
Chart Showing Comparison of Effects of Approximately
1/2 per cent Chromium and Nickel on the Rate of
Change of Softening as a Function of Hardness for
all Tempering Temperatures Investigated.
27
Figure 11.
Chart Showing Comparison of Effects of approximately
1/2 per cent Nickel and Chromium on the Rate of
Change of Softening as a Function of Temperature
at Selected Hardness Values.
27
Figure 16.
Semilogarithmic Plot Showing Relation Between
Hardness and Time of Tempering at 550°, 590°, 630°,
670°, and 710°C for Carbon Manganese, and Aluminum
Steels.
37
Figure 17.
Relationship Between Tempering Temperature and
Hardness for Carbon, Manganese, Silicon, and
Aluminum Steels Tempered 125 Hours.
38
Figure IS.
Relationship Between Tempering Temperature and
Quantity of Graphite Formed in Carbon and High.
Alloy Steels, containing approximately l/2 per
cent Added Element.
39
Figure 25.
Relationship Between Alloy Content and Hardness
after Various Vacuum Treatments for Carbon, Aluminum,
Manganese, and Silicon Steels.
47
Figure 26.
Relationship Between Alloy Content and Hardness
After Tempering in Lead or a Vacuum for 125 Hours
at 630° or 710°C.
49
Figure 30.
Relationship Between Alloy Content and Hardness
After Tempering for 125 Hours at 630° or 710°C in
Lead or a Vacuum.
51
CHAPTER I
Introduction
Investigators have been striving for a number of years to
understand the factors that influence the behavior of steels during heat
treatment.
It was particularly baffling when steels of "similar” chemical
composition would behave very differently from each other.
It has only
been within recent years that attention has been focused on the im­
purities that are normally omitted from the regular chemical analysis.
Attention was recently called to the fact that graphite may
form in some commercial carbon tool steels when they are subjected to
prolonged tempering treatments.
Yet, graphite may be absent in other
steels that have a chemical composition quite similar to those in which
the graphite occurred.
Since commercial steels contain so many chemical elements it
is impossible to determine the role played by each element in controlling
the mechanical properties of a steel.
In an endeavor to evaluate the
individual effect of some of the chemical elements commonly found in
commercial high carbon steels, the present investigation was undertaken.
Twenty—four alloys containing approximately 1.1% carbon and
I/256 or less of a third element were prepared under carefully controlled
conditions.
The third elements employed were manganese, silicon, nickel,
chromium, aluminum, sulphur, phosphorous and tin.
- 2 -
Since previous investigations indicate that commercial carbon
tool steels of "similar1' chemical composition may behave very dissimilar­
ly when tempered in the temperature range 550° to 710°C for prolonged
periods of time, similar treatments were given the twenty—four alloys.
While the behavior of the alloys to the heat treatments was followed by
means of hardness measurements and microscopic examinations.
The hardness and me tall ograph! c data resulting from the temper­
ing treatments are divided into two chapters.
In Chapter IV is presented
the effect of the chemical elements on the rate of softening resulting
from the coalescence of carbide particles.
While the effect of chemical
elements on the stability of iron carbide, graphitization, is discussed
in Chapter V.
Some of the alloys exhibited peculiar microstructures after
annealing.
These peculiarities are discussed in Chapter VI.
- 3 CHAPTER II
Review of Literature
Information dealing with the effect of small amounts of alloying
elements on the behavior of steels during heat treatment is not very
plentiful.
While the small amount that is available chiefly relates to
steels containing low or medium quantities of carbon.
Due to the scarcity
of data for high carbon steels, steels with approximately 1% carbon, it
has been necessary to review the literature of the lower carbon steels in
order to obtain an estimate of the probable effect
that small amounts
of the various chemical elements may exert on the mechanical properties
of high carbon steels.
Particular attention was paid to the hardness
data since that property was used in the present investigation to follow
the changes that occurred in the steels during heat treatment.
Ishen (15) subjected a one per cent carbon steel to heat treat­
ments which were designed to produce the carbides in various shapes and
sizes.
He showed that for the same tensil strength, laminated carbides
gave less ductility than spheroidal carbides, and that the carbides
assumed an equiaxed condition more rapidly during tempering when the
steel had been previously oil-quenched from above the lower critical
than when slowly cooled from the.same temperature.
Thus, finely dis­
persed carbides became more rapidly equiaxed than coarse lamillar carbides.
Other investigators have observed that the condition of the
steel prior to heating at temperatures below the lower critical point
has a profound influence upon the rate at which the carbides spheroidize.
- 4- -
Bailey and Roberts (16) found that the cast structure of a 0.3% carbon
steel required ten times as much time to produce a given spheroidized
state, when tempered at 675°C, than when normalized.
While cold working
reduced the time to one fiftieth of that required for the normalized con­
dition.
Taylor (17) found that when a 0.13% carbon steel was cold worked
by an elongation of 14$ in a tensil machine it would reach a chosen
degree of spheroidization, vrhen heated at 650°-700°C,/one fourth of the
time required by a normalized structure.
While an annealed structure was
three times longer than a normalized structure.
Engel (33) studied the softening characteristics of a typical
carbon steel, containing 0 .94% carbon and 0 .4-0% manganese, when heat treat­
ed to four different initial structural states and thereafter tempered
for a period, ranging from 2 seconds to 22 hours at five different
temperatures: 315, 4-80, 565, 650, and 700°C.
were
The four initial states
martensite, bainite, fine nodular pearlite, and coarse pearlite.
He found that, in general, the larger the size of the carbide particles
in the initial structure, the lower is the rate of softening during
tempering.
At any given temperature, the coarse lamillar structure soften­
ed most slowly, and after 22 hours at the higher temperatures this
structure was even slightly harder than the martensitic structure after
a similar temper.
At the lower temperatures, the difference in hardness,
in terms of Brinell number, between martensite and bainite was nearly
independent of tempering time after a brief initial period.
While at
the higher temperatures, the four tempered structures approached a common
hardness, which gradually decreased with increased tempering time.
- 5 Higher carbon steels appear to take more time to reach a standard
degree of spheroidization than steels containing lower quantities of
carbon.
When plain carbon steels containing 0.13, 0.45, and 0.85%G were
investigated by Taylor (17) it was discovered that a low carbon steel
would spheroidize more rapidly than a high carbon steel when heating was
conducted in the temperature range 600—700°C.
Duma (18) reports that sorbitizing, lead quenching, liquid
hardening and cold work treatments increase the rate of carbide spheroidization.
While steels of eutectoid composition are the most sluggish
and unresponsive to this reaction.
Alloys seem to promote rather than
hinder spheroidization by virtue of their refining or sorbitizing in­
fluence upon the structure.
The author, however, failed to mention the
alloying elements which he had observed to behave in this manner.
Various reviewers have conducted extensive surveys of the
literature pertaining to the effect of various alloying elements on the
mechanical properties of steels.
The results of these surveys have been
published by the Engineering Foundation in the form of monographs (6, 7,
8, 19, 20, 21).
A very good treatise has also been written by Bain (5)
on the ^Functions of the Alloying Elements in Steel”.
Since his data
show in a clearer and more condensed form the same general trends that
are exhibited by the data listed in the monographs, only his results,
pertaining to the alloying elements used in the present investigation,
will be reviewed here.
Bain (5) has used Rockwell "C” hardness-tempering temperature
charts to illustrate the effect of silicon, manganese, nickel, and
- 6 -
chromium, in quantities ranging from approximately 1/2 to 5 per cent,
on the hardness of tempered steels.
It is quite evident from these
charts that chromium, silicon and manganese are able to impart greater
resistance to softening to a medium carbon steel during tempering than
nickel.
In fact, the presence of 3 l/2% nickel only increases the
hardness about 4 Rockwell "C" hardness units over that of a plain carbon
steel when tempering is conducted for 1 hour at 500-700°C.
Whereas a
similar increase occurs in a manganese steel when the manganese content
is only increased from 0.75 to 1.75$.
Silicon appears to behave similar
to manganese while chromium imparts a slightly greater resistance against
softening.
It has been shown by Breuil (22) that copper increases the
tempered hardness of low and medium carbon steels.
When steels contain­
ing 0.4$C were tempered at 300°C, the Brinell hardness was increased 40
units by 1/2% copper and approximately 80 units by an addition of 1% of
this element.
Smith and Palmer (23) and Lorig and MacLaren (8) have clearly
shown that the hardness of copper steels may be increased by aging
treatments.
Copper contents in excess of 1% appear to produce the most
marked effect.
While the magnitude of the change decreases quite rapid­
ly when the copper content drops below 1$, 0.8% Cu only producing about
one half as much effect as 1$ of this element.
Graphitization has long been known to occur in high carbon
steels during heat treatment, but the amount of information present in the
literature is small.
Most of the available literature deals with the
- 7
formation of graphite during heating at sub-critical temperatures.
How­
ever, several articles give information pertaining to graphitization in
the critical range.
Green (24) observed that a black area appeared in some hypereutectoid steel bars when they were annealed just under the lower critical.
One of the crucible made steels had a chemical composition of 1.16%C,
0.30% Mn, 0.25%Si, 0.024%P, and 0.02%S.
steel during an anneal at 704° to 724°C.
A black area developed in this
A light area., present at the
center of the bar, was surrounded by the black area which in turn was
surrounded by a circumferential light zone.
Microscopic examination
revealed that carbides were present in the light areas but were absent
in the dark zone where complete graphitization had been effected by the
heat treatment.
Emmons (25) has observed the presence of graphite in tool
steels when they were subjected to temperatures below the Ar-^ point.
He noticed that the graphite had formed from the cementite "in situ".
While Portevin and Chevenard (26) have reported graphitization in a
1.6%C, 0.28%Si, and 0.16% Mn steel when it was heated below the
point.
Kinzel and Moore (27) found graphite even in a 0.15%C, 0.20%Si,
0.49%Mn, 0.023%S and 0.015%F Steel after it had been used in an oil still
for three years at a temperature well below the eutectoid.
Mathews (28) has stated that he found graphitization to be a
maximum in a steel between about 500°^ and the lower critical.
_ 8 -
It has not been until recent years that serious effort has been
expanded toward the determination of the factors that promote graphitiza­
tion in steels.
When six commercial carbon tool steels, containing
approximately 1%C, were subjected to prolonged tempering treatments in
a "covered” lead bath (up to 125 hours) between 550° and 710°C, Austin
and Norris (1 , 2 ) found that graphite formed more readily in three of
the steels than it did in the remaining three steels.
The only chemical
element that appeared to be affecting the behavior of these steels was
metallic aluminum, graphite forming more readily in the steels containing
the higher amounts of this element.
exert
Temperature was also observed to
a profound influence on the graphitization.
Maximum quantity of
graphite occurred when tempering was conducted at approximately 670°C
which was approximately 60°C below the critical range of the steels.
At temperatures where graphite readily formed, a softening
effect due to graphitization was usually not evident in the hardness data
until the tempering time was between 5 snd 25 hours.
Specimens, that
were heated in a vacuum and water quenched from the tempering temperature
six times during a 125 hour treatment, in order to facilitate the measure­
ment of electrical resistance and coercive force after various time in­
tervals, showed more graphite than specimens tempered continuously in a
lead bath for 125 hours.
It was suggested by the authors that the greater
quantity of graphite present in the vacuum treated specimens was due to the
repeated quenching operations.
Fetzer (29) has more recently found that
repeated quenching from a tempering temperature exerts little effect on
graphitization.
However, he has observed that graphitization occurs more
readily when tempering is conducted in a vacuum than when it is performed
- 9 in a lead bath.
Austin and Fetzer (11) studied the behavior of the same six
steels used by Austin and Norris (1, 2) at temperatures in the critical
range.
They found that the chemical elements that accelerate graph­
itization in white iron during annealing also play; a similar part in
steels, when heating is performed in the critical range.
Tims the
greatest amount of graphite would occur in the steels containing the
largest amount of silicon, aluminum, nickel, and copper.
It is quite
evident from their work and that of Austin and Norris (1, 2) that the
factors which promote graphitization at temperatures below the critical
range may be different from those that operate in the critical tempera­
ture range.
As a result of investigations conducted on eighteen addition­
al commercial carbon tool steels, Austin and Fetzer (39) conclude that
the presence of aluminum oxide promotes graphitization at sub—eutectoid
temperatures.
It is shown that aluminum oxide, formed either during the
deoxidation of the molten steel or during a tempering treatment where
oxygen is permitted to diffuse into the steel and combine with the
aluminum, may be effective in promoting the carbide decomposition reaction.
As far as it is known, Wells (30) was the first investigator
to study graphitization in highly purified iron-carbon alloys.
He was
successful in producing graphite in alloys containing 0.13 to 2.98J&C
by heating them in the critical range.
The behavior of these alloys was
studied by dilatometric and microscopic methods.
It was shown that the
rate of graphitization is greatly increased by the presence of graphite
nuclei, and the rate of decomposition of carbide into graphite and
- 10 -
austenlte is considerably increased as the temperature is increased.
The
author attempts to show that graphite may form directly from austenite as
well as from carbide.
He also presents evidence to illustrate that
graphite may form on crystallographic planes in carbide masses and that
the carbide may change in composition during graphitization.
It is also
demonstrated that graphitization will take place more slowly in an alloy
containing 2.98%C when the alloy is heated in a hydrogen-hydrocarbon
atmosphere than when heating is conducted in a vacuum.
The presence of graphite in steel was formerly considered to be
an undesired condition? however, during the past several years, one steel
company has purposely produced graphite in certain die steels in order
to impart greater wear resistance to dies.
Schwartz (12, 31) has summarized the effect of various chemical
elements upon the rate of graphitization in malleable iron.
He states
that amal1 amounts of aluminum, silicon, nickel, titanium, cobalt, and
copper accelerate graphitization, the effects of cobalt and copper
being small, while chromium, manganese, molybdenum, and tin retard it.
Boyles (13) has clearly shown that hydrogen retards the formation
of graphite during the solidification of a cast iron.
While Boegehold
(14.) suggests that hydrogen, through the addition of moisture during
melting, decreases the rate of formation of graphite in cast iron during
a malleabilizing anneal.
In a review of the literature pertaining to graphitization,
Schwartz (12) concludes that the early attempts to determine the thermo­
dynamic stability of iron carbide failed because of a lack of realiable
data for the calculation of the free energy of the graphitizing reaction.
- 11 -
In an attempt to eliminate the confusion existing in connection
with the correct value of the entropy of iron carbide, Seitz, McDonald,
and Wells (32) redetermined the heat capacity of iron carbide between
68° and 29&°K using the highly purified iron-carbon alloys of Wells (30)
and then by means of the heat capacity results calculated the entropy.
The values of entropy determined by previous investigators fell into two
groups, namely, 23 to 25, and 30 to 32.
The work of the above three
investigators produced the following value for the entropy at 29S°K:
25.7 e.u.
From the solubility curves of Wells (30) for carbon and iron
carbide in austenite, they calculated the value of the free energy and
heat of formation of the carbide.
AF° = 4.700 cal.
Thus, at 29S°K, A H = 5^00 cal and
Thus the positive sign of the free energy value indicates
that iron carbide is unstable at room temperature.
- 12 CHAPTER III
Preparation of Steels
Manufacture
Through the kind cooperation of Battelle Memorial Institute 24
"pure*1 lypereutectoid steels were prepared.
It was desired that these
steels contain approximately 1 .1% carbon and 0 .05, 0.10, or 0.50% of a
third chemical element siich as aluminum, silicon, manganese, nickel,
chromium, copper, sulphur, phosphorous, or tin.
Since the presence of
oxygen is believed by many investigators to markedly affect the physical
properties of heat treated steels, the present series of steels were
melted and cast in a hydrogen atmosphere.
The author had the advantage
of the complete supervision of these preparations by Dr. C. H. Lorig.
In order to facilitate the production of a similar amount of
carbon in all of the steels, a carbon-iron alloy, containing approximate­
ly 1% carbon, was prepared from electrolytic iron and carbon.
The
electrolytic iron analyzed 0.01%Si, 0.006%P, 0.009%S, and a nil amount of
Mn.
While No. 296
Acheson graphite, containing less than 1% ash,
furnished the carbon.
Melting was conducted in an induction furnace
under an air atmosphere.
As soon as melting was completed, the molten
metal was cast into 50 pound ingots.
Each ingot was forged to a diameter
of 2 1/4. inches and then cut transversely into sections of such a length
as to furnish the desired amount of metal for one charge in the hydrogen
atmosphbre — melting furnace.
Each section was analyzed for carbon to
determine the magnitude of the carbon addition required in the final
preparation of the alloy ingots.
- 13 The final melting operation was conducted in an induction
furnace in a hydrogen atmosphere.
The carbon-iron alloy previously pre­
pared was charged into the furnace.
When melting was complete, carbon
and the other alloying elements were added, to bring the composition of
the melt to the desired value, and the resulting molten metal cast into
a 7 1/2 pound ingot in a big-end-up mo. d co-.Gaining a hot top.
Each ingot
was permitted to solidify in the hydrogen atmosphere and after it had
cooled to room temperature it contained a bright silvery color.
The ingots, with a maximum diameter of 3 inches, were reduced
to a diameter of 3/4- inch by forging.
Heating, prior to the forging
operation, was conducted in a large pipe partially filled with charcoal.
With the exception of the sulphur steel, all of the steels were forged
at a temperature of 1095-H20°C.
While a lower temperature range of 870
-900°C was used for the sulphur steel.
Through the assistance of the Latrobe Electric Steel Company,
the forged bars were reduced by hot rolling to a diameter slightly greater
than 3/8 inch.
rolling.
All steels were heated to a temperature of 980°C prior to
The rolled bars were annealed at 790°C in a highly reducing
atmosphere and subsequently ’’centreless" ground to a diameter of 3/8 inch,
approximately 0.03 inch being removed from the diameter by the grinding
operation.
7/hile the annealing treatment placed the carbides in a spheroid—
ized condition.
The author is indebted to Dr. M. K. Smith for the assist­
ance which he rendered in the final preparation of these steels.
Chemical Composition
The chemical composition of the 2/+ bypereutectoid steels is
tabulated in Table I.
The quantity of all of the chemical elements, with
- u
-
the exception of copper, aluminum and tin, was determined by the usual
chemical methods.
spectroscopically.
While copper, aluminum and tin were determined
The spectroscopic method was also used for manganese,
silicon, nickel, and chromium when they occurred as impurities.
The
analyses were performed by the Bethlehem Steel Company through the kind
assistance of Mr. A. C. Chamberlin.
An examination of Table I reveals that the desired compositions
were approximated fairly closely and the quantity of impurities is much
lower than that commonly found in commercial steels of similar carbon
content.
Steel 76 was originally to have been a phosphorus steel.
However, because of the accidental presence of 0.24$ Mn, it has been
grouped with the manganese steels.
Some small variations occurred in
the carbon content but only four steels show a difference greater than
0.05$ from the initially desired value of 1 .1$.
It should be noted that there are two plain carbon steels
(87 and 89) of identical analysis.
Steel 87 was prepared as a pure
|
iron—carbon steel without any addition of other chemical elements whereas
number 89 was similarly prepared, but v/ith the addition of 0.025$ aluminum.
Since analysis shows that aluminum was less than. 0.005$ in steel 89, this
steel has been classed with carbon steel 87.
0.01$ aluminum steel.
Steel 29 was prepared as a
Although this aluminum content was approximated,
the accidental pickup of 0 .12$ silicon renders it necessary to classify
it as a silicon steel since the regular silicon steels also contained
this amount of aluminum.
Tt=.ble I
Chemical ComMOsltion Per Cent
Steel
Alloy
C
tin
Si
Ni
C r
87
89
None
None
1.08
1.09
0.006
.005
o.on
.023
0.017
.018
0.002
.003
26
?7
28
A1
1.12
1.15
1.14
.005
.006
.006
.03
.01
.015
.015
.018
.015
.00*
.00f
.002
72
73
76
83
Mn
1.14
1.11
1.10
1.14
.04
.08
.24
.30
.01
.01
.015
.042
.014
.018
.013
.015
1.08
1.20
1.08
1.15
.006
.005
.006
.006
.08
.12
.14
.48
ii
tl
it
" (p>
ii
Cu
A1
0.008 <0.005
ir
.006
.009
.008
S
0.015
.016
P
8
0 .01? <03
n
.012
.011
.01?
.073
.37
.015
.015
.014
.014
.014
.014
.003
.00c!
.002
.005
.015
.009
.004
.006
.006
.006
.007
.003
.015
.014
.017
.019
.014
.014
.051
.014
.014
.015
.012
.010
.005
.005
.005
.002
.006
.007
.005
.004
.006
.007
.008
.006
.016
.015
.017
.015
.01?
.012
.012
.014
it
it
ti
n
ii
n
ii
ti
n
77
29
78
79
Si
66
67
74
N i
it
ti
1.10
1.14
1.04
.009
.005
.007
.028
.022
.024
,08
.13
.5c
.002
.00*
.003
.007
.006
.008
<.005
.007
.006
.014
-.014
.017
.012
.01?
.012
ti
82
84
85
Cr
it
1.07
1.14
1.16
.004
.004
.006
.033
.07
.13
.46
.004
.004
..007
.003
.003
<.005
.018
.016
.015
.014
.014
.01?
it
it
.024
.015
.017
.018
69
70
71
Cu
n
n
1.11
1.10
1.16
.006
.005
.005
.01
.01
.01
.018
.015
.015
.005
.00?
.002
.068
.104
.35
<.005
<.005
.007
.014
.015
.014
.015
.014
.014
ti
it
80
81
S
Sn
1.15
1.08
.006
.004
.01
.024
.017
.016
.003
.003
.005
.007
.007
.003
.048
.018
.012
.012
ii
it
H
It
n
.O o O
ii
ti
ii
n
O .i
- 15 -
CHAPTER IV
Effect of Tempering on Hardness and Microstructure
Experimental Procedure
In order to study the effect of the various alloying elements
on hardness and microstructure, the 24- hypereutectoid steels were
hardened and then tempered at different temperatures for various time
intervals.
The hardening treatment consisted of heating specimens, 1/2
inch long and 3/8 inch diameter, at 1000°C for 1 hour, in a vacuum, and
subsequently water quenching.
A temperature of 1000°C was chosen in order
to effect chemical homogenation and to obtain a uniform austenitic grain
Size over the complete series of steels.
All specimens that were to
receive the same tempering treatment were hardened simultaneously to
insure a uniform hardening treatment for all steels.
Heating during the
hardening operation was conducted in an electrically heated silica tube.
The temperature was measured by means of a platinum-platinum rhodium
thermocouple and its maximum variation was +5°C.
The vacuum ranged from
0.10 to 0.15 nun. Hg. and was sufficient to prevent oxidation.
Following the hardening treatment, the steels were tempered
in a lead bath which was covered with Vlead pot11 carbon to prevent
oxidation, and water quenched from the tempering temperature.
The temper­
ing temperatures ranged from 550 to 710°C and the time of tempering
varied from l/2 to 125 hours.
The lead bath temperature was maintained
- 16 -
to within + 3°C of the indicated temperature by means of a T,Tagn electric
control.
alumel
All indicating thermocouples were composed of chromel and
wires and were regularly checked against the freezing point of
pure aluminum.
Separate specimens were used for each tempering treatment
and hardness studies were made on a transverse cut of each specimen.
- 17 -
Experimental Results
Austenitic Grain Size
The austenitic grain size of these steels has been previouslydetermined by Steigerwalt (3).
His results for a heating of 1 hour at
1000°C, which is the same heating period that is used in the present
hardening treatment, are given in Table II.
It is quite evident that
the grain size is quite similar for most of the steels, being between
1 and 3.
While the grain size of the steels that fail to fall within
this range is only slightly different.
Hardness
The complete hardness test data obtained on the quenched and
tempered samples are presented in Table III.
Where it will be noted that
five different tempering temperatures were employed, namely, 550°, 590°,
630°, 670°, and 710°C, and six different periods at temperature were
maintained - l/2 , 1 , 5, 25, 75, and 125 hours.
All steels were completely martensitic after hardening and
exhibited a Rockwell "C" hardness of approximately 65.
In order to permit an analysis of the data listed in Table III,
it is necessary to present a series of selected plots to demonstrate the
significance of the data.
Considering first the steels containing the highest amount of
alloying element, the relation between Rockwell "B" hardness and tempering
time is shown in Fig» 1 for the lowest and highest tempering temperature^,
- IS 550° and 710°C, respectively.
One of the pure carbon steels, 87, is in­
cluded as a basis for comparison.
The most striking characteristic of
the curves is the fact that the position of the curves with respect to the
hardness coordinate is profoundly altered by the alloying elements.
At
550°C the chromium steel is markedly higher in the chart than the other
alloys while the nickel and aluminum steels follow almost exactly the
behavior of the pure carbon steel.
The sudden drop in the hardness of
the aluminum steel between 25 and 75 hours is due to graphitization.
The silicon, copper, and manganese steels lie between the two extremes,
although the manganese steel appears to be approaching the lower group
of steels.
At 710°C some changes occur in the relative position of the
curves.
When the time is less than 75 hours, the hardness of the chromium
steel is much greater than that of the other steels.
However, after 125
hours the silicon steel possesses the highest hardness.
The erratic
hardness, values at 75 and 125 hours of the copper steel makes it im­
possible to accurately determine the time hardness—time curve.
Particu­
larly interesting is the switch in the position of the curves represent­
ing the behavior of the nickel and manganese steels.
At 550°C, the
nickel steel closely follows the carbon steel while the curve of the
manganese steel lies above these two steels.
However, when the
temperature is raised to 710°C, the manganese steel is below the carbon
and nickel steels while the nickel steel lies considerably above the
carbon steel.
It has been previously pointed out (l) that an approximate
linear relationship exists between Rockwell "B" hardness and log temper-
j9 A
A.S.T. M.
Grain Size
79
0.48 Si
2
66
0.08 N-i
3
Percent
Alloy
A.S.T.M.
Grain Size
87
None
3
89
None
Steel
Steel
oo
Percent
ALloy
■P-ro
Table II - Austenitic Grain Size After 1 Hour at 1000°G.
8
26
0.012 Al
2
67
0 J. 2 «
3
27
0.073 "
3
74
0.53 "
1
28
0.37
82
0.07 Cr
2
72
0.04 Mn
2 (60%)
4 (40%).
4
84
0.13 "
3
73
0.08 '•
1
85
0.46 ,l
76
0.24 " (P)
1
69
0.068 Cu
2 1/2 (60%)
5
(40%)
A
S3
0.30 Mn
2
70
0.104 11
1
77
0.08 Si
3 1/2
71
0.35
1 1/2
2
80
0.048 S
4
3
81
0.05
1
29
78
"
0.12
0.14 "
"
Sn
-
- 19 -
1------------------
1-------
J J ------J
E£L™__
----
Per Cent
Alloy
Quenched
Rock. "C"
87
89
None
None
65.7
6 5.4
104.5
104.8
104.5
104.3
102.3
101.8
99.8
100.4
98.2
9 9.5
97.9
98.6
103.5
103.0
10J
10(
£•6
27
28
0.012 A1
0.073 "
0 .37 "
65.4
65.1
63.6
104.6
105.6
109.1
103.6
104.9
107.1
101.9
101.5
103.9
100.2
97.8
100.4
97.3
77.4
51.4
97.0
50.2
43.7
103.2
1' 103.8
106.2
io:
10?
10'
72
73
76
83
Ok04
0.08
0 .2 4
0 .3 0
Mn
"
"(P)
"
6 4.4
6 4.4
6 5 .0
64.5
105.2
105.5
108.8
108.7
104.1
104.6
107.7
1 06.0
102.4
102.4
104.3
103 9
99.8
100.5
103.3
102.1
98.6
98.1
100.3
100.7
97.8
97.1
101.1
98.7
103.9
103.6
106.3
105.8
102
102
104
104
77
29
78
79
0.08
0.12
0 .1 4
0.48
Si
"
"
"
65.8
63.8
65 .1
64.8
106.4
107.0
106.4
109.5
105,6
105.5
105.4
108.4
102.0
102.5
101.9
105.3
100.8
101.8
100.4
103.5
99.3
100.4
9 9.3
102.3
98.8
99.7
97.8
101.1
103.4
104.5
104.3
106.4
102
102
102
104
66
67
74
0.08 Ni
0.12 "
0 .53 «
66.1
64.8
64.8
104.9
104.9
105.6
104.7
104.5
104.4
102.3
102.2
101.3
100.5
100.2
99.1
99.4
98.9
99.8
98.0
98.7
97.8
103.1
102.9
103.5
102
102
102
82
84
85
0.07 Cr
0.13 » ■
0.46 "
66.0
64.7
63.0
106.0
107.9
111.7
105.5
106.8
111.5
103.1
104.6
108.9
101.6
103.1
107.3
99.7
101.3
105.6
99.2
100.2
104.9
104.4
106.3
110.8
103
104
109
69
70
71
0.07 Cu
0.10 "
0.35 ''
64.2
65.4
64.3
106.3
105.2
105.6
104.2
103.9
105.9
101.7
101.7
103.7
100.7
100.1
102.2
99.4
98.8
101.1
98.9
98.2
102.0
103.6
102.6
104.5
102
102
103
80
81
0.048 S
0.05 Sn
64.6
65.3
105.3
106.6
104.5
106.1
101.6
103.4
9 9.4
101.0
98.6
9 9.3
97.4
98.9
103.5
103.6
101
102
S teel
' 1/2
on tem pering a t 550u , 590w and b'j0“ C.
— Time a t 550 Degrees C ent.-H ours — ■.
1
5
25
75
125 '
Tfte Denavior oi -one suxprnu- «*uu
t i n s t e e l s i n d i c a t e s t h a t th e y only d i f f e r s l i g h t l y from t h e carbon
s t e e ls .
Timi
' 1/2
T able I I I - Rockwell "B" Hardness
Time a t 590 D egrees C e n t.-H o u rs — v /
1
5
25
75
125
1 /2
5 '
1 /2
.9
.6
1 0 3 .5
1 0 3 .0
1 02.3
1 0 0.9
1 0 0 .3
1 0 0 .0
9 8 .6
9 7.5
9 7 .5
9 5 .8
9 6.4
9 5 .0
1 0 1 .4
1 0 0.4
.0
.2
.7
103.2
' 1 03.8
106.2
101.2
102.2
1 0 4.9
9 9 .4
9 9 .5
1 01.8
9 6 .6
9 5 .0
8 3 .9
94.8
8 2 .9
4 4 .8
9 4 .0
60.8
4 2 .5
1 0 0.0
100.8
1 0 3 .0
.8
.1
.1
.7
1 0 3 .9
1 0 3 .6
106.3
105.8
1 0 2 .0
1 0 2 .8
1 0 4 .9
1 0 4 .8
9 9 .7
99 .8
1 0 3 .0
1 01.7
9 8 .0
9 7 .4
9 9 .8
9 9 .4
9 5 .7
9 5 .9
9 8.3
9 6.8
9 4 .7
9 3 .6
9 7 .0
9 4 .7
.8
.7
.8
.1
1 0 3 .4
104.5
1 0 4 .3
1 0 6.4
1 0 2.5
1 0 3 .1
1 0 2 .9
1 0 4 .9
9 9 .8
100.8
1 0 0.3
1 0 3.4
9 8 .7
9 9.2
9 8.7
1 0 1 .6
9 7.1
9 7 .5
9 7 .5
9 9 .9
D
7
3
1 0 3 .1
1 0 2 .9
1 03.5
102.2
102.7
1 0 2 .3
1 0 0 .0
100.3
100.3
9 8 .3
9 8 .4
9 8 .0
.2
.2
.9
1 04.4
106.3
110.8
103.8
1 0 4 .9
109.8
1 01 .2
1 0 2 .9
1 0 7 .4
.9
,2
,0
103.6
1 02.6
104.5
1 02.4
1 0 2 .1
1 0 3 .4
,4
,9
103.5
103.6
1 01.5
1 0 2 .6
Time a t 630 Degrees C en t.-H o u rs
1
25
75
5
125
99.2
9 9 .4
9 8 .G
95.8
9 5.7
9 4 .7
9 5 .0
9 2 .6
9 5 .9
9 1.8
99
99
9 8.5
95.9
9 8 .6
96.1
1 0 2 .1 j.099.2
93.2
9 1 .9
8 9 .3
91 .1
8 9 .0
48.3
9 0.3
8 8 .6
4 1 .9
98
99
101
1 00.9
100.7
1 0 2.7
1 0 2.9
9 8 .9
99.2
1 01.7
100.8
97.5
97 .0
100.0
98.3
93 .8
9 3 .6
9 6 .5
9 4 .2
9 0.1
90.7
94.5
9 1 .6
8 9 .0
8 9.2
9 5.9
8 7 .5
90
90
101
101
9 6 .7
9 7 .2
9 4 .9
9 9 .1
1 0 1.6
1 0 1.1
101.3
1 03 .9
1 00.4
1 0 0.0
101.1
102.8
98 .4
98.5
98.3
101.6
9 6 .0
9 5 .7
9 5 .4
97 .1
94 .6
95.2
93 .3
9 6 .9
9 3 .7
9 4 .9
9 2 .6
9 5 .3
100,
100,
99.
102.
9 7 .1
9 7 .6
9 6 .6
9 4 .4
9 5 .7
9 8 .1
100.8
1 0 0.9
1 00.7
100.2
1 00.4
99.2
97.8
9 8 .4
9 7 .0
95.2
9 5 .3
9 5 .3
9 2 .3
9 3 .7
93.2
9 1 .0
9 4 .0
9 0.9
98.
100.
98,
9 9 .5
101.2
1 0 5 .4
9 7 .4
9 9 .3
103.2
9 6 .7
9 8 .1
101.2
1 02 .0
104.2
108.5
101.5
1 02.4
108.3
98.5
101.3
105.7
95 .7
9 8 .4
1 02.4
9 4.7
9 5 .9
9 9.3
93.5
95.2
98.7
100.
102.
108.
1 0 0 .0
9 9 .9
1 02 .0
9 8 .8
9 7 .7
1 0 3.1
96.2
9 6 .5
1 0 0.0
9 8 .5
9 5 .8
9 7 .7
1 00.7
1 01.5
102.4
99.6
99.6
101.2
9 8.1
97.8
99.3
9 5 .9
9 5 .2
9 7 .6
93.6
95.9
96.7
9 3 .3
95 .4
9 5.5
99.
99.
100.
9 9 .9
1 00 .1
9 8 .5
9 8 .7
9 6 .3
9 7.8
9 4 .9
9 5 .6
1 0 2 .0
1 0 1 .0
100.1
100.4
96.7
97.3
9 4 .9
9 4 .8
92.8
9 3.6
9 2 .9
9 1.3
99.
98.
- Time a t 670 D eg rees Cent .-H o u rs — A
1 /2
1
5
25
75
125
f
1 /2
Time a t 710 D eg rees C e n t.- H o u r s —
1
5
25
75
125
9 9 .2
9 9 ,9
9 8 .9
9 9 .2
9 5 .8
9 5 .5
9? .3
9 1 .4
89.9
89.5
0 9 .8
8 8 .1
9 6 .1
9 5 .4
9 4 .7
9 5 .7
9 2 .0
9 1 .6
8 8 .9
8 8 .3
8 6 .3
8 6 .r
8 5 .3
8 5 .6
96 3
9 9 ,8
1 0 1 .6
9 5 .8
9 6 .6
9 9 ,5
9 3 .2
9 5 .1
9 5 .8
9 0 .7
8 8 .9
9 2 .2
86.5
83.7
8 8.4
8 6 .8
0 4 .5
8 7 .4
9 5 .2
9 6.2
9 8 .0
9 3 .9
9 4 .5
9 6 .9
8 9 .6
8 9 .0
9 3 .1
8 7 .8
8 4 .6
8 8 .7
8 4 .7
8 5 .6
8 7 .5
8 3 .6
8 5 .1
8 5 ,5
9 9 .4
9 9 .2
1 0 1 .5
1 0 1 .5
9 6 .9
9 7 .8
9 9 .1
9 9 .4
9 5 .4
9 4 .0
9 5 .9
9 6 .3
8 3 .4
8 9 .6
9 1 .9
9 2 .1
87.2
8 8.7
9 1.2
8 9 .6
8 6 .4
8 7 .5
9 0 .5
8 8 .5
9 6 .3
9 6 .3
9 8 .9
9 7 .6
9 4 .5
9 3 .9
9 7 .0
9 7 .1
9 0 .7
9 0 .1
9 4 .0
9 2 .6
8 5 .9
8 7 .2
9 0 .0
8 8 .8
8 3 .9
8 3 .0
8 7 .2
8 5 .0
8 5 .5
8 2 .8
8 6 .7
8 4 .6
1 0 0 .3
LOO. 5
9 9 .8
1 0 2 .9
9 8 .9
9 9 .7
9 7 .7
1 0 1 .9
9 5 .2
9 7 .2
9 5 .9
9 8 .8
9 3 .7
9 4 .8
9 2 .0
9 5 .0
91.7
92.2
91 .6
93 .9
9 1 .4
9 3 .0
9 0 .9
9 3 .8
9 6 .7
9 7 .6
9 7 .1
1 0 0 .9
9 6 .6
■''5.9
9 6 .5
9 9 .0
9 2 .3
9 3 .1
92 .8
9 5 .9
8 8 .2
9 1 .5
8 9 .7
9 2 .3
8 7 .3
9 2 .0
8 7 .6
9 2 .2
8 6 .0
8 9 .2
8 7 .4
9 2 .8
9 8 .8
1 0 0 .4
9 8 .4
9 8 .5
9913
9 6 .8
9 4 .8
9 5 .8
9 5 .8
9 1 .8
9 2 .0
9 2 .2
8 9 ,9
9 1 .2
8 9 .7
8 8 .8
9 0 .5
9 0 .2
9 6 .1
9 7 .6
9 7 .3
9 4 .8
9 6 .2
9 5 .6
9 1 .5
9 2 .4
9 1 .6
8 7 .0
9 1 .2
9 0 .4
8 4 .8
8 8 .7
8 8 .6
8 4 .4
8 5 .7
8 7 .7
1 0 0 .2
1 0 2 .9
1 0 8 .0
9 8 .5
1 0 1 .5
1 0 6 .6
9 6 .4
9 8 .2
1 0 2 .8
9 3 .4
9 4 .8
9 8 .7
8 9.8
9 3 .7
9 5 .4
9 0 .0
9 2 .4
9 4 .7
9 6 .9
9 9 .8
1 0 4 .6
9 7 .3
9 9 .0
1 0 4 .2
9 2 .5
9 5 .4
9 9 .0
9 0 .0
9 1 .8
9 5 .9
8 6.2
8 9 .3
9 2 .7
8 5 .2
8 6 .6
9 1 .6
9 9 .2
9 9 .2
1 0 0 .8
9 8 .3
9 7 .5
9 9 .6
9 5 .8
9 5 .3
9 7 .4
9 2 .6
9 1 .8
9 5 .9
9 1 .2
9 0 .4
9 3.2
8 9 .1
8 9 .3
9 2 .5
9 7 .0
9 6 .4
9 8 .2
9 6 .1
9 5 .7
9 7 .3
9 2 .0
9 1 .1
9 4 .1
9 0 .0
8 8 .5
9 2 .1
8 7 .0
8 6 .5
9 3 .4
8 5 .1
8 3 .5
8 7 .6
9 9 .0
9 8 .2
9 7 .5
9 8 .5
9 4 .4
9 4 .8
9 1 .5
9 1 .5
8 8 .3
9 0 .1
9 0 .6
9 0 .2
9 4 .7
9 5 .4
9 0 .7
9 1.8
8 9 .0
9 0 .1
8 8 .2
8 8 .5
8 5 .8
8 6 .8
9 5.5
9 6 .2
- 19 -
ing time for commercial hypereutectoid carbon steels when considered for
time intervals between 1/2 and 125 hours.
This type of plot is presented
in Fig. 2 for the steels containing the maximum amount of alloying element.
These are the same steels that were shown in Fig. 1.
It may be noted that
with the exception of a few points, the data generally fall very close to
the straight lines.
Several peculiarities, however, may be noted.
At
550°G the nickel steel fails to soften as much at 75 and 125 hours as
would be expected from its behavior at the shorter times.
While the drop
in the hardness of the aluminum steel after 25 hours has been previously
explained, in connection with Fig. 1, on the basis of graphitization.
At 710°C, the silicon and copper steels show marked irregularity in the
behavior at 75 and 125 hours.
The relative position of the curves will
be discussed in a later section.
Semi-log plots of Rockwell MB M hardness and tempering time for
al 1 of the 24 steels studied in the present investigation are assembled
in Figs. 3, Ay and 5.
It should be noted that the hardness scale for each
temperature group is displaced 10 hardness units from the scale of the
preceding group.
silicon steels.
Fig. 3 includes the carbon, sulphur, tin, manganese, and
The nickel, chromium, and copper steels are located in
Fig. .4 while F±g. 5 contains the aluminum steels.
One carbon steel, 87,
is included in all plots as a basis for reference.
The behavior of the carbon, sulphur, and tin steels is shown in
Fig. 3A.
The two carbon steels react alike except for a slight difference
on tempering at 550°, 590° and 630°C.
The behavior of the sulphur and
tin steels indicates that they only differ slightly from the carbon
steels.
L
CD
_]
LD IOO
x:
0.37 AL
o
o
cr
0.48
sf~~
— ^ 3.5_ CO
Q
- -Q-
cr
<
X
80
O.
40
TEM PERING
Fig.
80
120
TIM E-H O U R S
1 - Eff e c t of Time of Tempering at 550° and 710°C
on Hardness of C a r b o n and H i g h A l l o y Steels
containing a p p r o x i m a t e l y 1/2 percent of ad d e d
element.
- 20 -
In the manganese steels, Fig. 3B, it is evident that the
resistance to softening is a function of the manganese content.
The
phosphorous steel, 76, which is now grouped with the manganese steels
due to the accidental addition of 0.24-% manganese, exhibits a greater
hardness after any given treatment than any of the other three manganese
steels.
While the carbon steel has a greater hardness than these same
three steels at all temperatures above 550°C when the time of tempering
is greater than 75 hours.
Silicon, like manganese, also increases the resistance to
softening (Fig. 3C). However, the silicon steels are usually harder
than the carbon steel after a given treatment at all temperatures, ex­
cept 630°C, regardless of the length of the tempering whereas the manganese
steels were generally softer than the carbon steel after a prolonged
temper.
The high silicon steel exhibits a peculiar break in its curve
at 710°C.
The hardness remains almost constant between 25 and 125 hours
and it is interesting to note that this hardness is quite similar to
that produced by 125 hours at 670°C.
In Figs. 4A, B, and C are illustrated the manner in which
nickel, chromium, and copper steels react to tempering.
Nickel produces
little effect on the resistance to softening, the nickel steels being
very similar to the carbon steel (Fig. 4-6.)•
However, chromium, Fig. 4B>
exerts a profound effect on softening, it increasing the resistance to
softening more than any of the other alloying elements studied.
copper, Fig.
chromium.
While
exerts an effect intermediate between nickel and
550°C
■< 110
%
HARDNESS - ,ROCKWELL
CD
100
—
*k
;
0.37 AL
0,30 MN
0 .5
I
LOG
5
25
7 5 125
TE M PER IN G T I M E - HOURS
Fig. 2 - Semilog a r i t h m i c Plot Showing Relation Bet w e e n
Hardness and Time of Tempering at 550° and 710°C
for C a r b o n and H i g h Alloy Steels, containing
a p p r o x i m a t e l y 1/2 percent of added element.
%
- 21 -
A study of the relationship between spftening and time at
temperature for the aluminum steels (Fig. 5) shows characteristics
different from those already described for the other steels.
In the
early stages of tempering, all the aluminum steels show a linearity
such as was exhibited by the other steels.
However, when
the time of
heating is increased beyond 25 hours, the hardness of the higher
aluminum steels suffers a rapid decrease at 550°, 590° and 630°C.
This
graphitization tendency will be discussed more fully in the next chapter.
Analysis of Hardness Data
Since the Rockwell hardness test is more rapid and better
suited for small specimens than the more fundamental Brinell hardness
test, the Rockwell test was used in the present investigation.
However,
since a linear relationship does not exist between the scales of these
two hardness tests for the hardness range encountered in the present
research, all Rockwell "B” hardness values were converted into Brinell
hardness (4-) before further analysis was made of the behavior of the
alloying elements.
When the resulting Brinell values were plotted against
log tempering time, the straight lines were quite similar to the Rockwell
plots shown in Figs. 3, 4, and 5.
slope of the lines.
A difference, hoYrever, occurred in the
In the Brinell plots the slope was practically the
same at el 1 temperatures whereas, it increased with increase in temperature
in the Rockwell graphs.
In the semi—log plots previously
m e n t i o n e d it
was quite evident
that some alloying elements shift the curves to higher hardness levels.
Chromium was an outstanding example of this behavior.
Since an intercept
B
c
SILICON
M^GtfKfESE
i
0 - - - 0 0.04
0
□ 008
□----- 0 0 0 5 S
h----- A 0.05 SN
1 C-87
o --- -o C -89
a
<
x C
•— ♦024.
A A0.30
x c
0----0 0j08
D------0 0.14
• ----- *0.12
4-----4 0 .4 8
110
100
no ioo
590°
590°
CD
i
100
HARDNESS - ROCKWELL
100
L A
630°
!00
95
670°
100 90
670°
90
9 0 100
710°
710°
90
90
80
0.5
25
75 125
05
I
5
25
75 125
05
I
6
25
75 125
LOG TEMPERING TIME - HOURS
Fig. 3 - Semilogarithmic Plot Showing delation Between Hardness
and Time of Tempering at 550°, 590°, 630°, 670°, and
710°C for the Carhon, Sulphur, Tin, Manganese, and
Silicon Steels.
B
A
NICKEL
X
O
XC
O 0.08
C
COPPER
CHROMIUM
G--- 0 O.G------------ *---- * C
0--- 0
A
A 0.53
o--- -o 0.07 A----A 0.40
^
-0 0.07
0
^
^
A--- A 0.35
f-4■ k^
550°C
>*-. ‘-HL
]■—
£3
k
J Tr
rk?
.
590° T
k
^-4.
100
J!>■ ^
1
HARDNESS - ROCKWELL
590°
L.^
A
590°
OD
100
ICO
630°
630'
630'
i■'4
100
1
-<
670°
100
X
90
s
670°
A"*
^
1
. 1n
0 0 100
r k -a
't^
90
710°
\
710°
'C
710°
s
k
90
,
4
90
E
<
80
0.5
75 125
LOG
TEM PERING
I
5
25
75 125
T IM E -H O U R S
Fig. 4 - Semilogarithmic Plot Shov/ing Relation Between Hardness and
Time of Tempering at 550°, 590°, 650°, 670°, and 710°C for
the Nickel, Chromium, and Copper Steels.
- 22 on the hardness coordinate can be used to designate this effect the
hardness intercept
where log t = 0 (t = 1 hour) has been arbitrarily-
chosen for this function.
This intercept is the value of the constant
"C" present in the equation of the straight line: B = a log t + C
where B is Brinell hardness "A” is the slope and "t" is time in hours.
Thus, nC" corresponds to the hardness after a tempering of 1 hour.
The values of the intercept "C", that is, hardness after 1
hour of tempering, are tabulated in Table IV.
TiVhile in Fig. 6 they
are shown plotted in seven charts for carbon, sulphur, tin and for each
of the six other elements under consideration.
The eighth chart shows
the data for the steels containing the maximum amount of alloying element.
The intercept "C” of the nickel, sulphur, and tin steels and also the
3teels containing the lowest amount of the other alloying elements was
very similar to the intercept of the carbon steels.
However, larger
quantities of manganese, silicon, aluminum, and copper increased the
hardness to some extent.
Thus, in amounts up to about 1/2 per cent,
these alloying elements will cause a hypereutectoid steel, after a temper
of 1 hour, to show a hardness that is 10—30 Brinell hardness units great­
er than a carbon steel after a similar heat treatment.
However, v/hen
the alloying element is chromium, the difference is very marked, a steel
containing approximately 1/2 per cent chromium increasing the hardness
by 70 to 100 units.
The relative behavior of these higher alloy steels
is clearly illustrated by the chart at the lower right hand corner of
Fig. 6.
The relative effect of the various alloying elements on the
hardness after 1 hour of tempering is furtherillustrated by Fig. 7.
Here,
ALUMIUM
x---------* C
o--------o 0.0i
0— — Q 0,07
A
------------a 0.37
MO
100
100
95
100
CD 100 90
90
6 70°
90 100
90
710°
i
90
85
80
40
i
40
0,5
I
5
75 125
LOG TEMPERING TIME- HOURS
Fig. 5 - Semilogarithmic Plot Showing Relation Between
Hardness and Time of Tempering at 550°, 590°, 650°,
670°, and 710°C for the Aluminum Steels.
- 23 the alloying elements are listed in decreasing order of effectiveness
at the lowest and the highest tempering temperatures.
The solid lines
join the hardness values plotted for the steels with the highest amount
of alloy while the steels containing an approximately constant small
amount, 0.07 to 0.08 per cent, of alloy are connected by the broken lines.
It is quite evident that l/2^S chromium is much more effective than a
similar quantity of silicon.
exerted no effect.
While an even greater quantity of nickel
It is interesting to note that small quantities of
the alloying elements produced little effect.
Since the rate of softening of a steel is dependent on the
kind and quantity of alloying elements present it would be of interest
to know the relative effect of the various alloying elements on this rate.
This information can be obtained from the hardness-time plots by com­
puting the slope of a tangent to the curve at any given hardness or
time value.
However, more accurate values may be obtained from the in­
formation furnished by the semi—log plots of hardness and time.
Since
the equation, (l) B = a I°g^Q t + G is the equation of the hardness-time
curve as well as the straight line of the semi-log plot, its first
the
derivative will be the slope of/hardness—time curve at any given hard­
ness or time value.
Changing the logarithm to the base e, equation (1)
becomes (2) B = 0.434- a log
9
t + C.
Its first derivative is (3)
Thus, by obtaining the slope na" of the straight line
of the semi—log plot it is possible to calculate the rate at which the
Brinell hardness is changing at any given time, i.e., rate of softening.
Values of ”an have been determined for the various steels from
Table IV.
Slope and Intercept Values for brinell-Log Time Curves
Tempering
550°C
590°C
630 °C
670°C
710°C
p
p
•
p
p
p
cn
a
ca
cn
TO
CO
a
.rH
0.
£0
CL
<3,
3j—1
<
1
3
1
—1
t
o
<
1
3
r
H
0) rH
O
r
H
O
r
H
O
r
H
O
ia
H)
O -CHD
<D >
a) >
(H 03
0
)
>
(
h a>
P
s
a)
i
n
<
D
>
.
1
Oh ho
TO
:
j C!
3 "SM
3
a.
1O
O -P *H jj 0
tm 113 *cPo -c
rH erf C
0L. hO 0 -aP
-H erf O 00 O -<Po -a
rH erf O
CJ -0P
ho 1
1—1 O
1
H
0
3
f
l
u
r
H
C
D
r
H
O
C
P
n
1
3
P
n
<
H
1
1
3
a
p
h
w a
M TO
K1 a
M TO
W S3
•H TO
to £
•H CQ
MG
•H TO
268
1 9 .0
1 6 .0
2 3 .0
2 28
2 55
2 0 .0
1 7 .5
239
207
278
238
2 2 .0
2b.0
256
2 2 .5
2 2 .5
233
1 9 .0
204
-P
J-H
(
>>
r0H
O1) O
)
1
N
1
1)
j■J
O H-h
C:.i
CO
.-4
67
39
iione
.u
26
27
28
0 .0 1 2 *1
0 .0 7 3 11
0 .3 7 11
72
73
76
63
0 .0 4
0 .0 8
0 .2 4
0 .3 0
Mn
ti
77
29
78
79
0 .0 8
0 .1 2
0 .1 4
0.4&
66
67
74
2 2 .0
3 7 .0
4<:. 0
2 70
279
301
2 6 .0
3 6 .5
4*>. 5
256
257
282
2 4 .0
2 7 .0
2 5 .5
234
232
252
2 4 .5
2 7 .0
3 1 .5
218
218
237
2 1 .0
2 5 .0
2 7 .0
202
206
220
2 3 .5
2 6 .5
" (P) 3 5 . 0
ir
2 9 .0
275
280
31 2
296
2 3 .5
2 4 .0
2 9 .5
3 3 .0
256
256
285
282
2 6 .5
2 6 .5
2 3 .5
3 1 .5
236
23 7
250
250
2 7 .0
2 6 .5
2 6 .0
3 0 .5
22 4
227
234
238
2 5 .0
2 5 .0
2 8 .5
3 1 .0
206
206
225
220
Si
>1
2 7 .5
2 7 .0
2 9 .5
3 4 .0
289
292
288
315
1 7 .5
1 9 .5
2 1 .0
2 4 .5
256
266
265
287
2 0 .0
1 8 .5
2 3 .0
2 3 .0
242
241
242
262
2 4 .0
2 2 .0
2 2 .5
2 7 .0
235
238
231
252
2 3 .0
1 7 .0
2 3 .0
2 5 .5
214
2 16
216
235
0 . 0 8 Ni
0 . 1 2 it
0 . 5 3 ti
2 4 .0
2 4 .0
2 1 .5
2 $0
277
273
1 7 .0
2 1 .0
2 1 .0
254
257
25 7
2 5 .0
2 1 .0
20. 5
242
242
236
2 5 .0
2 4 .5
2 0 .5
230
235
224
2 2 .5
•23.0
2 1 .0
205
216
215
82
84
85
0 . 0 7 Cr.
0 . 1 3 ti
0 . 4 6 it
2 6 .0
2 8 .0
4 2 .0
287
301
370
2 3 .0
2 8 .0
4 8 ,0
270
286
352
2 1 .5
2 4 .0
4 2 .5
245
260
320
2 4 .0
2 6 .5
4 7 .0
233
249
300
2 7 .0
3 0 .0
3 9 .0
220
233
268
69
70
71
0 .0 7
0 .1 0
0 .3 5
ti
n
1 7 .5
2 1 .0
2 0 .5
272
273
288
lb.5
1 8 .0
1 9 .5
258
252
268
1 9 .0
2 0 .5
1 7 .0
23 b
241
299
2 5 .5
2 3 .0
21. 5
232
22b
238
2 4 .5
2 4 .0
2 4 .0
215
209
224
80
81
0 .04s S
0 . 0 5 Sn
2 5 .0
2 8 .5
278
290
22. 5
2 3 .0
256
25 7
2 2 .5
2 3 .0
238
242
2 7 .0
2 2 .5
226
227
1 9 .0
1 9 .0
204
210
ti
11
Cu
HARDNESS
M
BRiNELL
I1
"‘
fots*
O•
Aj
CR
1
cu
.
550
600
650
O.S3 - ^
700 550
600
650
700
TEM PERING
■'i.
— - . 0.07
'T
3
“6
X
°/o"
550
600
650
700
550
600
1
350
L
300
H IG H 1
ALLOY
650
\
250
n
700
200
TEMPERATURE- °C
Fig. 6 - Effect of Temperature on Intercept C (Brinell
Hardness after Tempering for 1 Hour) for all
Steels.
-
24
-
their Brinell-log time curves and the results are tabulated in Table IV.
A comparison of the values obtained for a temperature such as 710°C will
show that the value of "a" for the 0./+6% Cr steel is much greater than
that of any of the other steels.
Hence it is quite obvious from eq. 3
that for a constant time of tempering this high chromium steel will have
a greater softening rate than any of the other steels.
Yet, in spite of
this apparent high rate the chromium steel has the maximum hardness.
It
is believed that a truer picture of the role played by the alloying elements
is obtained by calculating the rate of softening at a constant hardness
instead of at a constant time.
In order to make this calculation it is
necessary merely to determine from the semilog plot the time required to
produce a given hardness and when this value is substituted for "t" in
eq. 3 the rate of softening at that particular hardness is obtained.
The rate of softening at 710°C was calculated for all the steels
and the results have been plotted in Fig. 8.
In this figure, the rate of
softening is plotted against the Brinell hardness where the rate was
determined.
The carbon, sulphur, and tin steels are grouped in one of
the seven charts while each of the remaining charts include one of the
six other remaining alloying elements.
It is quite evident that all of
the elements exert some effect on the position of the curves.
chromium exerts the greatest influence.
Although
With increase in alloying
element, the curves are shifted tp the left.
Thus, at a given hardness
the rate of softening is decreased by increase in alloy content.
exceptions to this relationship should be noted.
Several
The phosphorous bearing
manganese steel (0.245& manganese) has a lower rate of softening for a given
hardness value than the steel containing the highest amount of manganese.
I HOUR
AFTER
350
5 50°C
0.46
0.37
300
BRINELL
HARDNESS
C
INTERCEPT
x0.46
0.30
0.35
XO.46
250
\0-48
0.37
X
- 'o _9;07- o.o <9
200
0.53
71 0 ° C
~ o-
CR
0,30
0.35
-
AL
MN
CU
Nl
ALLOYING
ELEMENT
*
Fig. 7
Effect of Low and High Amounts of Alloy
on Intercept C (Brinell Hardness after
Tempering for 1 Hour) at 550° and 710°C.
- 25 The intermediate silicon steels containing 0.12 and 0.1.4 per cent of the
alloy have quite different rates at low hardness in spite of the similar­
ity in the silicon contents.
The main difference in the chemical composi­
tion of these steels is the carbon contents.
The 0.12% silicon steel,
which has the lowest rate of softening, has 0,12% more carbon than the
higher silicon steel.
The low copper steel softens at a rate intermediate
between the medium and high copper steels.
Except for the difference in
copper content the low and medium copper steels have identical chemical
analysis.
The curves in Fig. 3 also show that the rate of softening
becomes smaller as the hardness decreases.
This is what would be expect­
ed to happen since there is less tendency to soften when the hardness
gets close to its limiting value.
Some of the curves in Fig. 8 have been consolidated into the
two graphs shown in Fig. 9*
^he graph on the left illustrates the effect
of 0.07-0.08% of alloy while the right hand graph shows the effect produced
by approximately l/2 per cent.
When the amount of alloying element is
small, there is a slight division between the steels.
The aluminum,
manganese, and nickel steels appear to have a slightly greater rate of
softening than the chromium, 'silicon and copper steels while the rate of
the carbon steel is intermediate.
At low hardness values the carbon
steel reacts like the chromium, silicon, and copper group whereas at high­
er hardnesses it is similar to the other group.
It is interestingto note
that in the higher alloy steels, right hand chart, the steels are
arranged in the same order from left to right as prevailed in the low
alloy steels; however, the degree of separation of the curves is much
greater.
It should also be noted that the carbon steel lies to the right
-10
MN
PER
HOUR
-10
0 .4 8 '
-6
0.2 4 )
0 .3 0 -
-4
-4
■=2
-2
OF CHANGE
IN
BRINELL
-0
-10
-10
cu
CR
-a
-0
-6
-4
RATE
AL
-4
0.13
-2
250
-2
210
170
BRINELL HARDNESS
Fig. 8 - Chart Showing Rate of Change of Softening
as a Function of Hardness for all Steels
W h e n Tempered at 710°C.
^
- 26 -
of the high alloy steels.
Aluminum, manganese, and nickel react quite
Similarly even when present in the larger amounts.
While chromium,
silicon, and copper produce lower rates of softening, at any given hard­
ness, than the other alloying elements, it is quite evident that chromium
reduces the rate much more effectively than either silicon or copper.
In order to illustrate the effect of change in temperature on
the rate of softening with respect to hardness two steels, exhibiting
widely different rates at a given temperature, namely, the 0,53% nickel
and the 0.4-6% chromium alloys, have been compared.
Incidentally the curves
representing the nickel steel are not markedly different from those for
the plain carbon steel.
Isothermal curves for these steels are shown in
Fig. 10 where it is evident that at a given hardness, the rate of
softening of both steels is increased as the tempering temperature is
raised.
The important point, however, is revealed by the grouping of
the alloys with respect to the Brinell hardness coordinate.
Thus in
making comparison for a common rate of softening of 10 Brinell hardness
units per hour at 550°C, the nickel alloy shows this softening rate at
275 Brinell while under similar conditions the chromium alloy exhibits
this rate at about 360 Brinell.
On the other hand, when both have been
given the time necessary to attain a common hardness, say 275 Brinell,
while the nickel steel is still dropping in hardness at a rate of 10
units per hour the chromium alloy is decreasing in hardness at a rate
much less than 1 unit per hour.
Probably this factor is better revealed by the plot prepared
in Fig. 11, where the rate of softening is plotted against tempering
IN BRINELL/HOUR
HIGH
ALLOY
ALLOY
/1
0 .0 7 - 0 .0 8 2
i
-10
-8
CR
i
cu
iTfh
-6
MN
i
CHANGE
,CR
i
-4
RATE
OF
-2
250
170 2 5 0
BRINELL
210
170
HARDNESS
*
Fig. 9 - C h a r t Shewing C o m p a r i s o n of Effects of 0.07
to 0.08 Percent, and A p p r o x i m a t e l y 1/2 Percent
a d d e d E l e m e n t on the Rate of Change of Softening
of Alloys, as a F u n c t i o n of Hardness when Tempered
at 7 1 0 ° C .
- 27 -
temperature.
It is quite evident from these curves that at a given hard­
ness such as 24.0 Brinell, the rate of softening of the nickel steel in­
creases much more rapidly with temperature than it does in the chromium
steel.
Microstructure
The micro structure of the steels containing the maximum amount
of alloying element, and also the carbon steel 8 7 , after a tempering
of 125 hours at 5 5 0 ° , 6 3 0 ° , and 710°C is shown in Figs. 1 2 , 1 3 , 14..
The structures obtained at 550°C are shown in Fig. 1 2 .
Except for the
chromium and aluminum steels, the various alloys have produced results
quite similar to the plain carbon steel.
The carbides in the chromium
steel are very much smaller than those in the other steels.
While no
carbides exist in the aluminum steel since they have been completely
decomposed into graphite and ferrite.
When
tempering is conducted at
630°C , Fig. 1 3 , larger carbides are produced in all the steels.
However,
the carbides of the nickel and carbon steels appear to have grown to a
slightly larger size than they did in the other steels.
steel continues to show the smallest carbides.
absent in the aluminum steel.
The chromium
While carbides are again
In Fig. 1-4» 710°C, it is evident that the
carbon, nickel, and copper steels contain the largest carbides, while
the smallest carbides are again present in the chromium steel.
The
manganese, silicon, and aluminum steels lie between the two extremes.
It
should be noted that the aluminum steel- has retained all of its carbon
in the fonm of carbide at this temperature instead of in the form of
graphite such as occurred at the two lower temperatures.
At 550° s-ud
630°C , the carbides of all steels were usually arranged in patterns that
0.46 °/0 Cr
--©. 53 /0 Ni
I -10
8
670°
■710°
710° \
,550°
UJ
-4
,630°
>70°
u_ -2
UJ
0
370
Fig.
330
200
250
BRINELL JHAP£ME'Sl'S
210
10 - Ch art showing C o m p a r i s o n of Effects
of Appr o x i m a t e l y 1/2 Percent Chromium
and Nickel on the Rate of Change of
Softening as a Function of Hardness
for all Tempering Temperatures
Investigated.
;
IN BRINELL./HOUR
OF CHANGE
RATE
0.53 °/0 NI
0,46 J c R
-10
-8
260
BH
240
BH
220
BH
-6
-6
-4
320 / 300 /
BH , BH/ 260
V BH
260
BH,
240
-2
BHy
-2
200
220
BH
550
600
650
BH^.
700
TE M PE R IN G
550
600
650
700
T E M P E R A T U R E - °C
Fig. 11 - Chart showing Comparison of Effects
of Approximately 1/2 Percent Nickel
and Chromium on the Rate of Change
of Softening as a Function of Tempera­
ture at Selected Hardness Values.
- 28 -
resembled to some extent the martenistic structure from which the carbides
had precipitated during tempering.
However, when the temperature becomes
sufficiently high, the carbides become so few in number that the patterns
largely disappear.
In ^ig. 14-, the chromium and manganese steels are the
only steels that tend to show much resemblance to an ordered arrangement
of the carbides.
The change in the size of carbide particles with increasing
times of tempering at 710°C is shown in Fig. 15 for two steels.
The
nickel steel represents the steels in which the carbides have a tendency
to grow rapidly whereas the chromium steel illustrates the case wheie the
carbides grow more slowly.
Little difference exists between the 4&nd
5 hour structures of the same steel.
However, it is very evident that
the carbides are smaller in the chromium steel than they are in the nickel
steel after the same treatment.
When the time of tempering is increased
from 5 to 125 hours, a marked increase occurs in the size of the carbides.
The interesting thing about this increase is the fact that the size in­
creased approximately 2 to 3 times in both steels.
This tends to in­
dicate that the carbides in the chromium steel will never be as large
as the carbides in the nickel steel.
Discussion of Results
In order to explain the behavior of the alloying elements on
hardness and micro structure, it is necessary to know the distribution
of the alloying elements between the phases present.
alloy steels the phases are ferrite and carbide.
In tempered low
A survey of the
literature reveals that only a small amount of quantitative work has
- 17,
/
D
G,
< » , § > ?
Fig. 15 - Photomicrographs of Carbon ana High Alio; Steels
Temper
(approximately 1/2 percent added element
ed 125 Hours at 650°C.
A. - Carbon B. - 0.50 M n
C. - 0.48 Si D. - 0 55 Ni
E. - 0.46 Cr F. - 0.55 C u G. - 0.57 A1
X2000
Picric Acid -Etch for A - f , Nitric Acid Etch for i
A
L
"(So
° ®
fl.
f?-.
o
B
£
o
4? o
°
j
'Cv ci
r) ■»
o 0 ° ’^
■7 J / o ^ Qrv> Q «=> 0°
*< 9 ^ £ °
S*0.!.Zfcs
< ?rfl <?
L V ^
°
^
'
c^f<A °*»
F
C
0M
j:j9? 0 .
0
V
c»
o
o
o
Fig. 14 - Photomicrographs of Carbon anc^ High Alloy
Steels (approximately 1/2 percent added element).
Tempered 125 Hours at 710°C.
A. - Carbon B. - 0.30 Mn C. - 0.48 Si D. - 0.53 Ni
E. - 0.46 Cr F. - 0.35 Cu G. - 0.37 A1 X2000
Picric Acid Etch for all Steels.
- 29 -
been conducted on the composition of the carbides present in various alloy
steels and most of these investigators failed to show the distribution
of the al.loying element between ferrite and carbide.
More analytical work has been conducted on the carbides of low
chromium steels than for steels containing small amounts of a n y other
alloying element.
The results of four investigators are shown in Table
V for annealed steels containing 0.97 to 2.24# Cr and O.84. to 1.05#C.
The distribution of the chromium between the ferrite and carbide was
calculated from the steel and carbide analyses and the results are shown
in the last two columns of the table.
of the chromium is in the carbide.
some annealed manganese steel3.
These values show that over SO#
Results are also shown in Table V for
The manganese varies from 0.4- to 2.2#
while the carbon is approximately 0.8#.
It is
quite evident from the
last two columns of the table that only about 60# of the manganese is
present in the carbide.
These results for the manganese and chromium
steels substantiate the general belief that chromium is a stronger
carbide forming element than manganese.
However, these data, naturally,
do not indicate whether the distribution ratio previously mentioned
holds for steels that have been tempered.
No data could be obtained for the composition of the carbides
present in silicon, nickel, copper, and aluminum steels.
However, various
writers (5, 6, 7, 8, 9, 19) suggest that these chemical elements are
chiefly located in the ferrite of annealed low alloy steels.
There are two possible ways in which an alloying element can
influence the hardness of a steel.
First, it may dissolve in the ferrite
and make the ferrite more resistant to deformation.
In other words make
A
D
fS
V
^
-x^r^S ,7r' *^ v ^ v ,^ .; «?
^
0
co
c S i'P
rf* * o
3
^cy o
■
x
£
*
9°"a3?K^JO”
X?c=Cr*ir
-*<
d ° & g oJ
, * j c^
3 f ^
.
k
;;■ •:/
O ' *o ;
n
/> J) ° ^rv
° (r*
^ ^ !
*<r ,on . c? i
>
t
a
f
'
i
S
F
M
C
^
-
^
O,
cx * " - r y
.^V.
o<
V-*5**CsV^'^
Fig. 15 - Photomicrographs of 0.53 Ni and 0.46 Cr Steels,
Tempered at 710°C for 1, 5, and 125 Hours.
0.53 N i Steel A. - 1
Hour B. - 5 Hours C. - 125 Hours
0.46 Cr Steel D. - 1
Hour E. - 5 Hours F. - 125 Hours
Picric Acid, ^tch X2000.
- 30 the ferrite harder.
Second, the growth of carbide particles may be retard­
ed by the presence of the alloying element in the carbide or ferrite
phases.
Since small carbide particles promote higher hardness than large
carbide particles, the retarding action of the alloying element on the
coalescence of the carbide would leadto a higher hardness.
It is believed that the relatively strong action of chromium
on the hardness of tempered steels is principally due to a strong retard­
ing action of the chromium on the coalescence of the carbide particles.
While the slight hardening effect produced by manganese, silicon, aluminum,
and copper is probably due to the same reason.
The resistance of the high silicon alloy, containing 0./S%
silicon, to further softening after a
hardness of approximately 93 Rock­
well B is reached, see Fig. 2, may be
due to a hardening action of the
silicon on the ferrite.
Since the carbide particles become very large
when they are heated at high temperatures, it is possible that the hardness
of a steel containing large carbides may be primarily dependent on the
hardness of the ferrite.
The behavior of the silicon steel cannot be ex­
plained on the basis of a lack of carbide particle growth since it was
very evident from the microstructures that the carbides continued to grow
in size while the hardness remained practically constant.
TABLE V - CTTPMTCAL COMPOSITION OF CARBIDES IN LOW CHROMIUM M D MANGANESE STEELS
Chemical Composition - Percent
Steel
Carbide
Source of
Information
C
Cr
C
Calculated Distribiitioii’~6£j.AHoy
Cr
% Cr
in Ferrite
% Cr
in Carbide
Arnold and Read (34)
0.84
0.99
7.20
6.99
17
83
Maurer (35)
1.02
0.97
7.10
6.90
1
99
Westgren (36)
1.03
1.67
6.83
9.62
13
87
Campbell (37)
0.85
2.24
6.73
15.01
15
85
1.05
2.24
6.81
11.44
21
79
C
Mn
% Mn
in Ferrite
ii
C
Mh
% Mn
in Carbide
0.78
0.41
6.80
2.21
39
61
tt
0.78
0.83
6.86
4.61
37
63
it
0.85
1.16
7.17
6.31
35
65
it
0.36
2.21
7.50
11.68
39
61
Arnold and Read (38)
m
- 31 -
General Conclusions
The following conclusions may be drawn concerning the hardness
and micro structure of pure 1.1% carbon steels to which have been added
one-half of a per cent or less of a third element, when the hardened
steels are tempered in the temperature range 550° to 710°C for time
intervals of 1/2 hour up to 125 hours.
1. Approximate linear relationship exists between Rockwell "B”,
or Brinell hardness and logarithm of the tempering time from 1/2 to 125
hours for all alloys at all temperatures investigated, so long as
graph!tization does not occur.
With the highest aluminum steel it has
been recorded that graph!tization may be complete at 550° to 630°C after
125 hours.
In the initial stages of tempering the reaction is complex
involving both the release of strain and the decomposition of martensite.
While the tempering reaction from 1/2 hour onwards is essentially a process
of carbide spheroidization.
2. In the alloys containing approximately 1/2 per cent added
element, data obtained at the end of the short tempering period of one
hour at all tempering temperatures, indicate that manganese, silicon,
aluminum, and copper show a hardness number from 10 to 30 Brinell units
greater than that of the plain carbon steels.
Nickel shows no apparent
effect, whereas chromium exhibits a resistance to softening which is
profoundly
investigated.
different from that of any of the other alloying elements
Thus after 1 hour at 550°C the chromium steel is about
100 Brinell units harder than the carbon steel and even after 1 hour at
710°C this difference is 70 Brinell units.
- 32 -
3. For the alloys containing less than 0.15% added element, the
data obtained after 1 hour treatment at each of the five tempering tempera­
tures demonstrates that with the exception of chromium little resistance
to softening is obtained.
4. In an effort to provide an analysis of the relative effective­
ness of the various alloying elements on resistance to softening, curves
have been presented which show the variation in rate of softening (ex­
pressed as rate of change in Brinell hardness per hour) at constant
temperature (710°C), with change in hardness.
This analysis showed the
relatively minor effects of added element up to 0.08%, but it clearly
revealed that for approximately 1/2 per cent of element present, manganese,
nickel, and aluminum cause only a small increase in resistance to softening
wh i l e
copper, silicon, and chromium, show increasingly marked effect in the
order listed.
5. Similarly, data have been derived to permit presentation of
curves to illustrate the effect of temperature on the variation in soften­
ing rate, with change in hardness on both the chromium and nickel alloys
containing approximately l/2 per cent of these elements.
At all tempera­
tures investigated the relative rate of softening at any selected Brinell
hardness was very much greater for the nickel than for the chromium steel.
Indeed the conditions obtaining with the nickel alloy were not very
different from those for the plain carbon steels.
6. Plots for the variation in rates of softening at selected
hardness values with increase in tempering temperature, for the high nickel
and chromium steels further illustrate and emphasize the difference
in behavior of these two elements.
- 33 7.
Selected photomicrographg6btained
from several samples o
the tempered alloys have been Included to illustrate the correlation
between change in hardness and the degree of carbide spheroidization of
the steels.
The microstructure of a completely graphitized aluminum
steel has also been included.
- 34 -
CHAPTER V
Formation of Graphite at Sub-Euteotoid Temperatures
It has been previously mentioned in the preceding chapter that
pronounced graph!tization occurred in the high aluminum steels which
produced a marked effect on the hardness.
During microscopic examination,
the presence of graphite was also observed in some of the other alloys.
Although its quantity was much less than that present in the alnm-T m im
steels and was apparently insufficient to produce a noticeable effect on
the hardness.
In order to study further the factors that were causing graphite
to form in these "pure" steels during tempering, additional heat treat­
ments were performed.
The results of these treatments as well as those
obtained from the earlier treatments are presented in this chapter.
Since the early work revealed that hardness was not an accurate indicator
of the presence, of small amounts of graphite, the effect^roduced by
the heat treatments have been followed with microscopic examinations as
well as with hardness measurements.
Experimental Procedure
In order to study the effect of the various alloying elements
on graphitization, combinations of the following four heat treatments
were employed:
1. Vacuum - harden: - 1 hour at 1000°C in a vacuum followed by a
water quench.
2. Vacuum — normalize: — 20 hours at 1025°C plus 11 1/2 hours at
1125°C followed by an oil quench from 1025°C, all heating being
conducted in a vacuum.
- 35 3- Lead bath temper: - l/2 to 125 hours at 510°-710°C followed by a
water quench, with all heating being conducted in a lead bath covered
with nlead pot" carbon.
4. Vacuum tempers _ 125 hours at 630° or 710°C in a vacuum followed
by a water quench.
The vacuum harden and lead bath temper treatments were the same
as those described in Chapter IV.
The vacuum normalize was conducted in a sillimanite refractory
tube which was heated by a globar furnace.
The specimens rested on a
refractory boat that had been constructed of alundum cement and fired
at a high temperature.
The vacuum pressure was approximately 0.08 mm.
of Hg at 1025°C and 0.10 mm. of Hg at 1125°C.
'While the temperature
fluctuations were +10°C on a platinum-platinum rhodium thermocouple.
Preliminary work indicated that an oil quench from a high temperature
was necessary in order to prevent the formation of graphite during cool­
ing through the critical range.
The vacuum temper was performed in an electrically heated
silica tube.
An electric control maintained the temperature within
j^3°C of the desired tempera'ture.
The vacuum pressure ranged between
0.05 and. 0.10 mm. of Hg., it improving with increase in time.
Specimens l/2 inch long and 3/S inch diameter were used for
hardness and microscopic examinations.
While larger specimens, 1 1/2
inches long and 3/8 inch diameter, were used for graphite analysis.
Experimental Results
It has been previously pointed out in Chapter IV that graphit—
ization occurred in some of the aluminum steels and its presence exerted
a marked effect on the hardness.
This effect was clearly shown in
Fig. 5 "by the semilog plots of Rockwell "B" hardness and tempering time.
Metallographic examination of other steels showed that graphite was also
present in some of the manganese steels.
Semilogarithmic Rockwell ’’B" hardness-tempering time curves of
some of the aluminum and manganese steels have been collected in Fig. 16
to illustrate the magnitude of the effect produced by grapbitization on
the hardness.
All of these steels were heated in a vacuum for one hour
at 1000°C and water quenched prior to tempering in a lead bath.
The
carbon and low aluminum steel contained very little graphite and have
been included for comparison purposes.
It should be borne in mind that
graph!tization tends to destroy the linear relationship between hardness
and log tempering time.
The two higher aluminum steels exhibit a marked tendency to
soften at the lower temperature during the longer periods of tempering.
At 550°C, these steels show a profound drop in hardness when the time
exceeds 25 hours.
"When the temperature is increased to 590°C, the
downward break in the curve of the 0.07% aluminum steel continues to
take place beyond 25 hours.
However, in the 0.37% aluminum steel, the
break occurs more rapidly and is now found between 5 and 25 hours.
A
further increase in temperature to 630°C causes the 0.07% aluminum steel
to have a linear curve for the entire time interval of 1/2 to 125 hours.
While a discontinuity is still evident in the curve of the 0.37% aluminum
steel between 5 and 25 hours although its magnitude is somewhat less
than that occurring at 590°C.
However, when tempering is conducted at
- 37 -
670° and 710°C, a break does nob occur in the curves of either of these
steels, both steels exhibiting linearity up to the maximum time of 125
hours.
It should be noted that the low aluminum steel containing 0.01%
aluminum has linear curves at all five of the tempering temperatures.
Of the two manganese steels, only the steel containing 0.30%
manganese shows any discontinuities in its curves.
At 590° and 630°C,
the departure from linearity occurs between 75 and 125 hours.
However,
the magnitude of these discontinuities is very small in comparison to those
that occur in the 0.37% aluminum steel at these same temperatures.
Al­
though the lower manganese steel contains almost as much manganese as the
0.30% manganese steel its curves are always linear at all of the
temperatures.
This behavior may be due to the presence of 0.031%
phosphorus in the 0.24-% manganese steel.
It Is quite evident from
Fig. 16 that the temperature at which
tempering Is conducted may have a profound effect on the magnitude of
the softening produced by graph!tization.
The effect of temperdture on
these same steels is more clearly shown in Fig. 17 where the Rockwell "B"
hardness, after a tempering of 125 hours, is plotted against the tempering
temperature.
The hardness data of a 0.4-&% silicon steel has been includ­
ed to illustrate the fact that the presence of such a large quantity of
this so-called graphitizing element has exerted little effect toward the
production of low hardness.
It will be shown later that this lack of
softening was due to the absence of graphite.
It is quite evident from Fig. 17 that the 0.07% aluminum steel
has a greater tendency to soften at 550°C than at any higher temperature.
•------• C
, .
X----- X 0 . 2 4 M N ( P )
£>
a 0.30 M N
W ---- V 0.01 A L
c------ c 0.07 A L
o----- o 0 . 3 7 A L
IIO
550° C
110 100
100
590°
100
95
630°
CD
100
_J
_l
UJ
£
^
U
o
cr
l
^
to
UJ
2:
Q
cr
<
lOO 9 0
670°
90
fl. .^
9 0 100
90
710°
90
x
85
80
45
Vo
40
25
LOG
T E MP E R I N G
75
125
TIME - H O U R S
Fig. 1 6 — Semilogarithmic Plot Showing Relation Between Hardness
and Time of Tempering at 550°> 5 9 0 ° , 6 3 0 ° , 6 7 0 ° , and
710°C for Carton, Manganese, and Aluminum Steels.
- 38 -
While in the 0.37% aluminum steel the hardness is a minimum at the three
lowest temperatures, namely, 550°, 590°, and 630°C.
The only other steel
to show an abnormally large temperature effect in its curve was the 0.30%
manganese steel, which showed less resistance to softening at 630°C than
at either higher or lower temperatures.
In order to determine quantitatively the relative amount of
graphite present in the various alloy steels, a set of large specimens,
1 l/2 inches long and 3/8 inch diameter, were subjected to selected
tempering treatments and analyzed chemically for graphite.
The results
obtained for the steels containing aluminum, manganese, silicon, nickel,
chromium, copper, sulphur, phosphorus and tin are listed in Table VI and
plotted against the tempering temperature in Fig. IB.
The time of temper­
ing at all temperatures was 125 hours although data for a temper of
54 hours are also included for the high aluminum steel in order to
illustrate more clearly the temperature region where graphitization is
a ma-xinnim for this steel.
It should be noted that the two aluminum
steels were treated at 510°C as well as at temperatures ranging from
550° to 670°C.
After a treatment of 125 hours at 630°C, the quantity of
graphite was less than 0.1% in the carbon steel and in the alloy
steels
containing approximately 1/2 per cent of silicon, nickel, chromium,
and copper.
While the 0.07% aluminum, 0.04-8% sulphur and 0.05% tin steels
only contained slightly more graphite, about 0.15%.
The 0.24% and 0.30%
manganese steels have four to five times as much graphite as the plain
carbon steel while the high aluminum steel was completely graphitized and
100
-a
90
0.30
MN"
OQ
_1
N.
CO
CO
Ld
Q
cn
<
x
50
-o
40
550
590
TE M P E R IN G
630
670
710
T E M P E R A T U R E - °C
Fig. 17 - Relationship Between Tempering Temperature and Hardness
for Carbon, Manganese, Silicon, and Aluminum Steels
Tempered 125 Hour s.
- 39 -
contained eleven times as much graphite as the carbon steel.
It is
interesting to note that this high aluminum steel possessed practically
as much graphite after a temper of 54 hours as it did after 125 hours at
630°C.
When the high aluminum carbon and silicon steels are tempered
at 550°G for 125 hours the quantity of graphite is little different from
that obtained at 630°C, although the aluminum steel is completely
graphitized whereas only a small amount of graphite is present in the
carbon and silicon steels.
In fact, the silicon steel has practically
the same amount of graphite throughout the temperature interval 550°
to 670°C.
In changing the tempering temperature from 630°C to 550°C
a noticeable change, however, occurs in the graphite content of the
0.0755 aluminum steel.
At the former temperature only a small amount of
graphite is famed whereas at 550 °C it is practically completely graphit­
ized.
It is evident from the curve of this steel that graphitization
takes place most rapidly at 550°C.
A larger quantity of aluminum appears
to raise the temperature of maximum graphitization tendency.
Thus, for
a steel containing 0.37% aluminum, graphite forms most readily in the
temperature region of 590° to 630°C.
In the case of the high manganese
steel, 0.3055 manganese, graphite forms most easily at 630°C which is
quite similar to the maximum graphitizing temperature of the high aluminum
steel.
Although it must be born? in mind that much more graphite is
-farmed in the aluminum steel during an equal period of heating.
The unetched micro structures of some of the steels are shown
in Figs. 19 and 20 at a magnification of 100 diameters.
All structures
are the result of a 125 hour heating at the indicated tempering
TABLE VI - GRAPHITE ANALYSIS
Steel
Percent Tempering Time at
Percent
Alloy
Temp. °C Temp.Era. Graphite
Steel
87
none
550
125
0.06
76
87
27
27
27
27
28
28
28
28
28
28
28
28
u
0.07
0.07
0.07
0.07
0.57
0.57
0.57
0.57
0.57
0.57
0.57
0.57
630
510
550
590
650
550
590
650
510
550
590
650
670
■ 125
125
125
125
125
54
54
54
125
125
125
125
125
0.07
0.49
1.06
0.57
0.14
0.75
1.10
1.06
0.47
1.12
1.08
1.09
0.23
85
83
83
79
79
79
79
79
74
85
71
80
81
A1
A1
A1
A1
A1
A1
Al
Al
Al
Al
Al
Al
Percent Tempering Time at
Alloy
Temp. °C Temp, ffrs.
0.24
Mn (P)
0.50 Mn
0.30 Mn
0.50 Mn
0.48 Si
0.48 Si
0.48 Si
0.48 Si
0.48 Si
0.53 Ni
0.46 Cr
0.55 Cu
0.048 S
0.05 Sn
Percent
Graphite
650
125
0.32
590
630
670
550
590
630
670
710
630
630
630
630
650
125
125
125
125
125
125
125
125
125
125
125
125
125
0.16
0.39
0.13
0.05
0.07
0.07
0.09
0.09
0.07
0.09
0.03
0.13
0.15
TEMPERED
125 H OURS
1.0
0.8
0.8
0.6
0.6
i
1
Lul
0.4
0.2
0.2
0.048 S
0.0
510
550
TE M P E R IN G
590
630
670
0.0
T E M P E R A T U R E -°C
Fig. 18 - Relationship Between Tempering Temperature and
Quantity of Graphite Formed in Carhon and High Alloy
Steels, Containing Approximately 1/2 Per Cent Added
Element.
temperature.
550°G.
The carbon steel is shown in Fig. 19A after treatment at
Only a small amount of graphite
vailed in the
is present. This condition pre­
temperature range 550° to ZL0°C.
shown in Fig.19E, appears to contain more
The 0.4.8% silicon steel,
graphitethan the plain carbon
steel although chemical analysis indicated that the amount of graphite
was similar.
This apparent discrepancy may be due to the difficulty
encountered in accurately determining small amounts of graphite by
chemical analysis.
The structure of the silicon steel only changed
slightly with temperature.
Many large graphite particles occur in the 0.07% aluminum steel
when it is tempered at $50° and 590°C, Figs. 19B and 190.
As the
temperature increases the graphite particles become larger and less numer­
ous.
At 630°C, very little graphite forms in the interior.
some is present at the edge as shown in Fig. 19D.
However,
The 0.37% aluminum
steel was completely graphitized when tempered for 125 hours at 550°,
590°, and 630°C.
The size and number of the graphite particles result­
ing from these treatments are illustrated by Figs. 19 (F. G. and H.)
It is evident that the size increases while the number of particles
decreases with increase in temperature.
When this steel is treated at
670°C, graphite is practically absent from the interior.
appreciable quantity forms at the edge, Fig. 20H.
Although an
With further increase
in tempering temperature, graphite fails to appear in any part of the
steel during a heating period of 125 hours.
It should be noted that
when the 0.07 and 0.37% aluminum steels are tempered at 550° and 590°C,
the graphite particles are larger in the steel containing the smallest
amount of aluminum, compare Fig. 19 (B and C) with Fig. 19 (F and G).
- a
-
a
amount
Figs. 20A and 20B illustrate that/small/of graphite forms in
the 0.08% manganese steel during tempering at 630°C and 670°C, with a
maximum amount being produced at the lower temperature.
The 0.30%
manganese steel also shows maximum graphitization at 630°G, Fig. 20C,
although much more graphite is formed in this steel than in the lower
manganese steel after treatment at this temperature.
However, when
these two steels are tempered at 670°C, the quantity of graphite is quite
similar, compare Fig. 20D with Fig. 20B.
Whenever graphite was present in a steel after tempering in
a lead bath, the graphite particles were always smaller at the edge
than in the interior.
Since this behavior prevailed on the transverse
edge cut prior to heat treatment, as well as on the peripheral edge of
a specimen, the difference in size of the graphite particles is the result
of the heat treatment rather than of a condition inherited from the
fabrication processes.
A comparison of the interior structures of the
0.37% aluminum steel, Tig. 19 (F, G, H) with the structures occurring
at the edge, Fig. 20 (E, F, G) clearly illustrates the difference in the
size of the graphite.
The etched structures resulting from a tempering of 125 hours
at 630°C are shown in Fig. 21 for the high silicon, aluminum, and
manganese steels, as well as for a low manganese steel.
The structures
shown in Figs. 21A and 21B were magnified 1000 diameters whereas the
remaining photomicrographs in this figure were taken at a magnification
of 250 diameters.
When the structure of the silicon steel is highly magnified,
Fig. 21A, the small dark spots are observed .to. he located adjacent to the
Fig. 19 - Photomicrographs of Carbon, Aluminum, and Silicon Steels
as Tempered 125 Hours.
At 550°C: A. - Carbon
E. - O .48 Si
B. - 0.07A1
F. - 0.27 Al
At 590°C: C. - 0.07A1
G. - 0.27 Al
At 620°C: D. -0.07A1 (Edge) H. - 0.27 Al
Unetched. 3EL00
-
carbide particles.
Fig. 21C.
42
-
A lower magnification of this structure is shown in
Figs. 21E and 21F illustrate the fact that a graphite nucleus
grows at the expense of the carbide immediately surrounding it.
Fig.
21D shows the completely graphitized structure of the high aluminum alloy
at a magnification of 250 diameters, while the structure of a graphite
particle is shown in Fig. 21B at a higher magnification.
The effect of tempering time on the progress of graphitization
at 630°C is illustrated in Fig. 22 by unetched and etched structures of
the 0.37% aluminum steel.
The structures were produced during tempering
periods of 5, 25, and 125 hours.
The unetched structures are presented
in Figs. 22A to 22C at a magnification of 100 diameters while Figs. 22D
to 22F illustrate the etched structures when magnified 250 times.
Schwartz (12) has stated that graphitization is markedly
accelerated during the malleabilizing anneal of a white iron by a rapid
cool from a high temperature prior to the anneal.
To determine the effect
of the hardening treatment on the formation of graphite in the present
steels, the 0.07% and 0.37% aluminum, 0.30% manganese, and the 0.48%
silicon steels were heated for 125 hours at 630°C in a lead bath without
any previous hardening treatment.
Prior to the heating in the lead bath,
the carbides in these four steels were in a spheroidized condition as a
result of the commercial anneal that had been given to these steels
subsequent to hot rolling.
Microscopic examination revealed that practic­
ally no graphite was present in the steels after the treatment in the
lead bath.
This result is illustrated at magnifications of 250 and 1000
diameters by the 0.37% aluminum steel in Fig. 23 (A, E) and by the 0.3Q/t>
manganese steel in Fig. 23 (B, F).
A comparison of Figs. 23A and 23B
*
l & T ^ l r ' - l ^ I i / V - I ^ L . I*
JV r f ? - s S Z' I ' X T .rJ
Fig. 21 - Photomicrographs of Carbon, Aluminum, Silicon, and
Manganese Steels Tempered 125 Hours at 630°C.
X1000: A. - O.4.8 Si.
B. - 0.37 Al,
X250:
C. - 0.4-8 Si.
D. - 0.37 Al.
E. - 0.08 Mn.
F. - 0.30 Mu.
- 43 -
with Figs. 21D and 21F, respectively, clearly indicates that the harden­
ing treatment has promoted graphitization.
Several factors may be introduced by the hardening treatment
which might promote graphitization.
They are:
(1) The small size of the carbides present during the early part
of the tempering operation.
(2) Stressed condition of the metal.
In order to obtain some information pertaining to the first
factor, size of carbides, two specimens of the 0.37$ aluminum steal were
hardened in the usual manner by heating at 1000°G for 1 hour and water
quenching.
One specimen was then tempered for 1/2 hour at 710°C in a
lead bath while the other was heated for 125 hours at the same tempera­
ture.
These treatments produced carbides of two different sizes without
the formation of graphite.
The two specimens were then heated together
for 125 hours at 630°C in a lead bath.
Only a small amount of graphite
resulted from these treatments and only a slight difference could be
noted in the quantity present in the two specimens, Fig. 23 (C, D, G, H).
Thus, it appears that if the size of carbide particles effects graphitiza­
tion, the carbides must be much smaller than the size produced by a
treatment of l/2 hour
at 710°C.
The effect of stress on graphitization, second factor, was
studied by annealing specimens that had been previously subjected to vari­
ous amounts of cold work.
A cylindrical specimen of the 0.37$ aluminum
steel (commercially annealed condition) was deformed on the periphery and
one end by means of a Brinell hardness penetrator.
The spheroidized
*
- 44 -
carbides in this specimen were similar in size to those produced by a
temper of 125 hours at 710°C.
Following the indentation by the hardness
penetrator, the specimen was heated in a lead bath for 125 hours at 630°C
and sectioned longitudinally.
Specimen
The unetched structure of one half of the
is shown in Fig. 24A at a magnification of 7 diameters.
The
indentation made on the periphery is shown at the top of the picture while
the depression at the left is located at the end of the specimen.
A great
deal of graphite formed in the areas adjacent to these impressions.
How­
ever, the quantity of graphite decreased as the distance from the indenta­
tion increased, while complete absence of graphite was observed in the other
half of the specimen.
It is evident in Fig. 24A, that graphite is scarce
in certain areas such as on right side of picture.
The magnitude of the deformation has a profound effect on the
size and quantity of the graphite particles.
It can be seen in Fig. 24A
that the graphite particles are large near the edge whee maximum deforma­
tion took place.
Whereas a little farther inward they are smaller.
While
at greater distances, where deformation was small, the particles again
become large and finally disappear.
etched structures
Photomicrographs illustrating the
present at the edge, a short distance inward, and at a
greater distance are presented in Fig. 24- (C, D, E) at a magnification of
250 diameters.
In Fig. 24-C, it is evident that the large graphite particles
are located at a short distance from the edge while much carbide lies
between the edge and the graphite.
At a short distance farther inward,
Fig. 241, the structure is completely graphitized and the graphite particles
are smaller than those located near the edge.
In Fig. 24-E, is shown an
area free of graphite which was located at a great distance from the edge
>
in
>3)
~
"ox
/
j° ^ o O ^ i o
!. ^°
. o <
■
be
0
•* IC-b/b „ o
S
4
1
^
’• -d .VJ?
0
o
/o
^
e<) “• C b
o , ‘to
o>
,» ■>
y ^ 'r . V r ; .
V
p 9 5 D ^'/JS=rf " V
»
f
d
?
-
4
i
Y
^
k
<* C? ? 0^0
<=— J "v> 3
■ " .o °
0
B
'
F
0°
c'rvv ^
U *
^
: . . — e -.
i]b> C^o
0
V
) •
.,
-
*J£l\ u°'vp
M a % « ! • * « <* £ & o ? s 0
4 J © 5 S ^ "■ * « $
m
S
C
§
f
f
e
4
~
^
^^r.-uZ-^o
v “* «.u .o
o'*: i
. c ^ _ . ^
..
& *
'//' ' W-L “
^„ ,.• „ „*
• .yo 0.°o0 o
>~A
ji “ oV ir^cvr* ^ s > s>u >
rt
*i
fk
"f.
r.
c v :
U b <***»'>
.
-
CJ
G
O
_
m
>/ • fc-C*° &>
I
° '.G e -7 ^
<i>". ""*:^l?
'T - V.
F
D
v.-SOTo^
xo4<^L
fifu®».:\Vn «4
.%*'•
^ -\-v=^f^)Y
Fig. 23 - Photomicrographs of High Aluminum and Manganese Steels
Annealed 125 Hours at 630°G After Different Initial
Heat Treatments.
As Received. 0.37 Al: - A. — X250
E. - X1000
0.30 Ufa: - B. - X250
F. - X1000
Tempered 1/2 Horn- at 710°C.
0.37 Al: G. - X250. G. - X1000
Tempered 125 Hours at 710°C
0.37 Al: D. - X250. H. - X1000
Picric Acid Etch
where the metal had been practically undeformed during the production of
the indentation.
Thus, it is quite evident from this series of photomicro­
graphs that a stressed condition in the metal may have a profound
accelerating effect on graphitization.
Further evidence illustrating
the effect of deformation on graphitization is shown in Fig. 24B, X7,
in a sample of the 0.07% aluminum steel.
When this sample was heated for
125 hours at 630°C in a lead bath without a previous hardening treatment,
graphite formed near
^ • two stencil
marks, one is not shown in picture,
that had been stamped on the end of the specimen prior to the heat treat­
ment.
Boyles (13) and. Boegehold (14-) have shown that the presence
of hydrogen in the furnace atmosphere during the melting of cast iron may
retard the formation of graphite during solidification and also during a
malleabilizing anneal.
Since the present steels were melted and solidified
in a hydrogen atmosphere, it was thought that enough hydrogen might have
been retained during the fabrication process to prevent some of the alloy­
ing elements such as silicon, nickel, and copper from promoting graphitiza­
tion during the tempering treatment.
These elements are known
promote
graphitization in malleable iron.
In an attempt to evaluate the effect of hydrogen on graphit­
ization, a preliminary survey was conducted in order to determine the type
of vacuum treatment most suitable for the purpose.
Three different
vacuum treatments were employed, as well as the lead bath tempering
treatment previously used, on silicon, aluminum, manganese, and plain
carbon steels.
These four treatments are listed as follows.
1. Harden in vacuum and temper in lead bath
2. Normalize in vacuum, harden in vacuum, and temper in lead bath
v .i;
i; Jf t ’t•'■" *■•£. *'* „
K !*r*.;;rI'"';
■;-•;/•■,■•t./’J**.*>.*•.*-*•
f i S f t f c j * V< v?S--r •:.-.
A
>. \
I**ss’
iUf*«■*'••-***••.*V.
.?*r. s .« v * ' . .
^
I--'-..-
^
*.*•’ •
■
• ••
c
••*• "^i:
..
**
,s
'
v . . -•
v'.**>;•.•.st'i'’"U:
* 1%
D
B
E
Fig. 24- - Photomicrographs (Longitudinal Section) of 0.37 and 0.07%
Aluminum Steels Annealed 125 Hours at 630°C After
Cold Work.
0.37 Al: A. - One Half of Sample
C. - Edge
D. - Short Distance from Edge.
E. - Great Distance from Edge
0.07A1: B. - End of Specimen ContainingStencil Marks.
A - B, Unetched. X7
C - E. Picric Acid Etch. X250
- 46 -
3. Harden in vacuum, and temper in vacuum
4. Normalize in vacuum, harden in vacuum, and temper in vacuum
The hardening treatment consisted of a heating at 1000°C for 1 hour in a
vacuum followed by a water quench.
Y/hile tempering was conducted for 125
hours at 630°C in the furnace atmosphere indicated, with a water quench
following the heating period.
The normalize was conducted in a vacuum.
Its heating schedule consisted of 20 hours at 1025°C plus 11 1/2 hours
at 1125°C.
quench.
While subsequent cooling from 1025°C was effected by an oil
The vacuum pressure in the furnace during the hardening and
normalizing treatments ranged from 0.10 to 0.15 nun of Hg. while during the
tempering operations it fell within the limits 0.05 to 0.10 mm of Hg.
The 8 steels subjected to these heat treatments were the
plain carbon (#87), 0.01, 0.08, and 0.37% aluminum, 0.04, 0.08, and 0.30%
manganese, and the 0.48% silicon.
While the hardness of these steels
resulting from each of the four treatments is plotted in Fig. 25.
The hardness of the plain carbon steel was approximately
90 Rockwell B, after all treatments.
The hardness data are shown on the
left side of the graph containing the aluminum steels, Fig. 25All of the treatments were effective in producing a low
Rockwell nB ” hardness of approximately 45 in the high aluminum steel.
How­
ever, only the treatments that involved a temper in a vacuum were able to
produce a low hardness in the low aluminum steels.
Y/hile the vacuum-
normalizing treatment only exerted a slight effect on the final hardness
of these steels.
When the manganese steels are normalized, hardened, and
- 4.7 -
tempered in a vacuum a low hardness results.
However, when they are
hardened in a vacuum and tempered in a lead bath a high hardness is obtained.
A prolonged treatment in a vacuum either at high or low temperatures
appears to exert a profound softening effect on the low manganese steel.
While the higher manganese steels only exhibit a low hardness after a
heating in a vacuum conducted at both high and low temperatures.
Only one treatment was effective in promoting softening in
the silicon steel.
temper in a vacuum.
This treatment consisted of a normalize, harden, and
Although most of the steel was fairly soft, a hard
area was present in the center.
A survey of the hardness data reveals that the lowest hardness
is produced in all the steels when they are subjected to the following
series of heating operations: normalize in vacuum (20 hours at 1025°C plus
11 1/2 hours at 1125°C, oil quench), harden in vacuum (1 hour at 1000°G,
water quench), and temper in vacuum (125 hours at 630°C, water quench).
The extremely low hardness produced by this treatment was due to graph­
itization.
In order to determine the effect of this heat treatment on
the hardness of the steels containing nickel, copper, chromium, sulphur,
tin and small amounts of silicon, samples of these steels were subjected
to the treatment while specimens of the plain carbon ($S7), high aluminum,
and high silicon steels were also treated to check the reproducibility
of the heat treatment
A comparison of the hardness data obtained after
the second heat treatment of the latter three steels with the data secured
following the first treatment indicates that approximately similar results
TEMPERED
125
HOURS
AT
630°C
100o
\j H , T W
100
H J 1.VAC)
L
\
"AN
\
NH,
T(VAC)
!
ii
4i
/
60
*
A
\
O
6
\\
f
1
rr=:'
0.1
80
X
\
40
L
t►CENTER
\\
\
60
\
f
\\
V-
HARDNESS - ROCKWELL
m 80
W h T"
It (p b )
If
40
0.2
0.3
AL
ALLOYING
0.4 O
0.1
ELEMENT
0.2
MN1
-
0.3
0.48
SI
°/
Fig. 25. - Relationship Between Alloy Content and Hardness After
Various Vacuum Treatments for Carbon, Aluminum,
Manganese, and Silicon Steels.
- 43 -
were obtained for the carbon and aluminum steels.
While a more uniform
hardness was obtained in the silicon steels after the "check" heat treat­
ment since no hard spot was present in the center of the sample such as
occurred in the first sample treated.
The "check" sample also had a lower
average hardness because of the absence of hard spots.
The hardness data for all of the alloy steels, as well as the
check data of the high aluminum and silicon steels, have been plotted
against the percentage of alloying element.
The resulting curves are
shown in the seven charts located in the upper half of Fig. 26 by means
of dashed lines.
The hardness value of the plain carbon steel #87 is
located at the left end of each curve.
While the hardness of the "check"
sample of the high aluminum and high silicon steels is the lowest point
shown as a dot where the abscissa reads 0.37% aluminum and 0.4.8% silicon.
The hardness values of all the steels resulting from a temper of 125
hours at 630°C in a lead bath have also been plotted in these charts as a
basis for comparison.
The latter data, indicated by circles, are
connected by a continuous line.
It is quite evident from the curves that
the new, all vacuum, heat treatment produces a marked softening in the
aluminum, manganese, and high silicon steels whereas its effect on the
hardness of the plain carbon, nickel, copper, chromium, low silicon,
sulphur, and tin steels was only slightly different from the result
produced by a temper in a lead bath.
Microscopic examination of the specimens tempered for 125 hours
at 630°C in a lead bath revealed that graphite only occurred in appreciable
quantity in the 0.07 and 0.37% aluminum, 0.08, 0.24, and 0.30% manganese
steels.
The unetched micro structure of some of the aluminum, manganese, and/
*
- 49 -
steels is shown in Fig. 27 at a magnification of 100 diameters.
Only a
very small amount of graphite occurred in the interior of the 0.07% aluminum,
and the 0.4-S% silicon steels.
This is illustrated by Figs. 27A and 27F.
However, it is evident from Fig. 27A that appreciable graphite formed at the
edge of the aluminum steel.
The entire structure of the plain carbon
steels was quite similar to the interior of these two steels.
Since the
0.31% aluminum steel was completely graphitized by this treatment, a large
amount of graphite is present in its structure, Fig. 27B.
Although
graphite was present in the 0.08 and 0.30% manganese steels, Figs. 27D
and 27E, its quantity was much less than that occurring in the high
aluminum steel.
However, it should be noted that the high manganese
steel has mox-e graphite than the lower manganese steel.
When the steels were normalized, hardened, and tempered 125
hours at 630°C in a vacuum, large quantities of graphite were produced
throughout the interior of the aluminum, manganese, and high silicon steels.
The steels containing 0.07 and 0.37% aluminum, 0.04 and 0.08% manganese
were completely graphitized while the 0.01% aluminum, 0.24- and 0.30%
manganese and 0.4-3% silicon steels only contained a small amount of
carbide.
Although the plain carbon, low silicon, nickel, copper, chromium,
sulphur, and tin steels contained practically no graphite in the interior,
some graphite occurred in a zone at the edge.
The widest zone occurred in
the low silicon steels while the chromium steels had the smallest zone
which 7/as less than half as wide as the zone exhibited by the plain carbon
steels.
The zone occurring in the nickel, copper, sulphur and tin steels
was similar to that present in the plain carbon steels.
In Fig. 28 are shown the unetched structures of some of the
steels after the vacuum treatment.
It is quite evident from the photo-
-------------• ----------^
100
H A R D E N E D IN V A C U U M , T E M P E R E D
IN L E A D
N O RM ALI2ED, HARDENED jT E M P E R E D
(A L L IN
TEMPERED
125 HOURS AT 6 3 0 °C
AL
VACUUM)
MN
SN
eo
60
"V
40
0.1
m
too
100
L- -
LU
CR
CU
80
01
0.2
03
0.4
0.5 0
TEMPERED
If) 100
0.1
0.2
125
0.3
HOURS
0.4 0
0.1
0.2
0.3
0.5
AT
100
MN
AL
CO
LU
X
O
CC
<X
0.4
80'
80
60
80
40
40
0
100
■p ■■■
N 1
1
00
n
-
0.1
CR
CU
---
=-=«
1
1
1
~r^=r’
80
ALLOYING ELEMENT — %
F i e . 26 - R e l a t i o n s h i p B e tw e e n A llo y C o n te n t a n d H a rd n e s s A f t e r
T e m p e rin g i n L e a d o r a Vacuum f o r 125 H o u rs a t 630 C
or 710°C.
_
—
o«n
- 50 -
micrographs that all of the steels contain a large amount of graphite.
A
comparison of Fig. 28 with Fig. 27 shows the difference in the amount of
graphite produced by the vacuum and lead bath tempering treatments.
It
should be noted that the size of the graphite particles in the high
aluminum steel is quite different after the two treatments, Figs. 27B and
28B, even though graphitization was complete in both cases.
The largest
graphite particles occurred after the lead bath treatment. Fig. 27B.
It has been previously mentioned that steels may contain
appreciable quantities of graphite after a prolonged temper at 630°C in
a lead bath but when they are tempered at 710°C for the same length of
time, in the same medium, no graphite forms.
In order to determine whether
the same result occurs when heating is conducted in a vacuum, samples of
all the steels were normalized, hardened, and tempered for 125 hours at
710°C, with all heating operations conducted in a vacuum.
The hardness
data resulting from this treatment are shown in the seven graphs located
in the lower half of Fig. 26 and are connected by the dashed lines.
While
the continuous lines join together the hardness values produced by a temper
of 125 hours at 710°C in a lead bath.
The hardness values of #87 plain
carbon steel are again located at the left end of each curve.
It is
quite evident from the curves that the vacuum treatment failed to produce
as much softening in some of the steels as it did wnen tempering was con­
ducted at 630°C.
However, the aluminum, manganese, and silicon steels
continued to be the only steels that exhibited any appreciable degree of
softening.
An interesting discovery was made when the vacuum treated
specimens were examined under a microscope.
Some of the steels were com­
pletely graphitized in spite of the fact that they had high hardness.
s.
*.
•
••
•
**
•
-
. .
.9 •
•' i . .
.
.
»
••
••
4
*
.
'.
* *f #
•- **<v'
•.
I
A
‘A
*•
D
J
1 »
*• ” •.
*«% ; J• •**
•* j *„•« •**’ .•.
•
•
*■
<i*
*•
•
•
n
f
,»* #
.
t .) **
•,
**•
.4 • • ’■* - <•» •
V
».
—Wi..—••• -4""
%•
T
..**■ ..».
• ■ • -.T»—
.^ • ':.<**•.- '•■
'^ t f r •' ••• 1 .* » - .
x*
.
*v-• **<
\•1J
* *,
**
* j-1
4^«
**• %
..
'* ~ ■.
*• *
*
i
'
*
*
'
\*f’
.* •
•
•
•
.
#
»
•
•«.
•
•
*
'
•
•
•
. . . »
.•
*
*
.* # r •
*
• •• ••*■
i3-* .* ;/
, 1
J
x
"I
♦j
F
m M
:
•
*
*
..
•
' * .
*
,r
.
*
i
1
28 - Photomicrographs of Soeels Tempered 125 Hours at
630°C in a Vacuum.
A. — 0.07 Al.
B. — 0.37 Al.
C. — 0.0-+ Mn.
D. - 0.08 Mn.
E. - 0.30 Mn.
F. - 0.4-3 Si.
TJnetched. X100
The 0.01, 0.07, and 0.37% aluminum, 0.04 and 0.08% manganese and the 0.4S%
silicon steels exhibited complete graphitization while a large amount of
graphite was present in the 0.24 and 0.30% manganese and the 0.08% silicon
steels.
Although graphite was absent from the interior of the plain carbon,
0.12 and 0 .14% silicon, 0.12 and 0.53% nickel, and the 0.35% copper steels
some graphite occurred at the edge of the these steels.
The unetched structures, X100, of some of the steels that con­
tained large quantities of graphite are shown in Fig. 29, all of the steels
shown, with the exception of the 0.30% manganese steel, were completely
graphitized.
While some carbide was still present in the latter steel.
In order to illustrate more clearly the effect of temperature
on hardness, the data of Fig. 26 have been replotted in the form shown in
Fig. 30.
The hardness data presented in the upper half of the figure were
produced by the tempering treatments conducted in a lead bath while the
data located in the lower half were produced by the vacuum treatment.
The
results produced by a temper at 630°G are joined by a continuous line
while the broken line is used to connect the hardness points derived from
the 710°C heating.
The marked effect of tempering temperature on the
hardness of the aluminum, manganese, and silicon steels, when they are given
the vacuum treatment, is very clearly shown.
The etched microstructures of some of the steels after a
vacuum treatment at 630° and 710°G are shown in Figs. 31 and 32, respective—
shown
ly, at a magnification of 250 diameters. All of the steels/are complete­
ly graphitized except the plain carbon and high manganese steels.
No graph­
ite is present in the plain carbon steel while only a small amount of carbide
Is present in the manganese steel.
ferrite in the silicon steel.
Note the peculiar structure of the
TEMPERED
125
TEMPERED
125
H A R D E N E D IN V A C U U M ,
HOURS AT
HOURS AT
TEMPERED
630°C
7I0°C
IN LE A D
IOC
100
AL
MN
00
80
60
60
40
40
CD
ROCKWELL
100
100
CU
CR
—.—
°
80
0.1
0.2
0.3
0.4
0.5 O
01
0.2
0.3
0.4 0
0.1
02
0.3
0.4
0.5
-
0
HARDNESS
NORM ALIZED j H A RDENED , T E M P E R E D
(AL L IN VACUUM)
100
AL
MN
80
80
_\
/
60
60
40
0.1
100
100
CU
80
80
0.1
0 .2
03
0 .4
0.5 0
0.1
02
0.3
0.4 0
ALLOYING ELEMENT -
0.1
0.2
03
04
0.5
’
(
*
Fig. 30 - Relationship Between Alloy Content and Hardness After
Tempering for i25 Hours at 630°C or 710°C in Lead or a
Vacuum .
- 52 -
D is c u s s io n o f R e s u lts
I t is
b e l i e v e d t h a t a t l e a s t tw o c o n d i t i o n s m u s t e x i s t i n a
s t e e l i n o r d e r f o r g r a p h i t e t o f o rm d u r i n g t e m p e r in g i n a l e a d b a t h :
p re s e n c e o f c e r t a i n c h e m ic a l e le m e n ts , an d o f s t r e s s o r s t r a i n .
T h e p r e s e n c e o f s t r e s s a l o n e d o e s n o t a p p e a r t o e x e r t much
in flu e n c e on th e s t a b i l i t y
in a le a d b a th .
c arb o n s t e e l s .
o f i r o n c a r b i d e w hen te m p e r in g i s c o n d u c te d
T h i s i s q u i t e e v i d e n t fro m t h e b e h a v i o r o f t h e p l a i n
When t h e y w e re p l a c e d i n a h i g h l y s t r e s s e d c o n d i t i o n b y
t h e h a r d e n i n g t r e a t m e n t , g r a p h i t e f a i l e d t o fo rm d u r i n g t h e s u b s e q u e n t
te m p e r in g o p e r a t i o n .
T h u s , w hen t h e q u a n t i t y o f t h e i m p u r i t i e s a r e low
i n a n i r o n —c a r b o n a l l o y ,
s t r e s s d o e s n o t a c c e l e r a t e t h e d e c o m p o s itio n o f
th e c a r b id e to a n o tic e a b le
e x t e n t d u r i n g a h e a t i n g o f 125 h o u r s i n t h e
t e m p e r a t u r e r a n g e 5 5 0 ° —7 1 0 °C .
T he e x p e r i m e n t a l d a t a a l s o show t h a t t h e p r e s e n c e o f a p p r e c i a b l e
q u a n t i t i e s o f a lu m in u m a n d m a n g a n e se f a i l t o p ro m o te g r a p h i t i z a t i o n w h ere
t h e s t e e l i s r e l a t i v e l y f r e e fro m s t r e s s p r i o r t o t h e te m p e r in g o p e r a t i o n .
F o r w hen s t e e l s c o n t a i n i n g a p p r o x i m a te ly 0.35% o f t h e s e e le m e n ts a r e
h e a t e d i n a l e a d b a t h f o r 12 5 h o u r s a t 6 3 0 °C , a f t e r t h e y h a v e b e e n e i t h e r
p r e v i o u s l y te m p e r e d a t 710°C o r g iv e n n o p r i o r h e a t t r e a t m e n t , v e r y l i t t l e
g ra p h ite i s
fo rm e d .
H o w e v e r, w hen t h e s e s t e e l s a r e p la c e d , i n a s t r e s s e d
c o n d i t i o n p r i o r t o t h e t e m p e r in g t r e a t m e n t , much g r a p h i t e i s p r o d u c e d
d u r i n g t h e t e m p e r in g o p e r a t i o n .
T h u s , t h e s im u l t a n e o u s p r e s e n c e o f s t r e s s
a n d c e r t a i n c h e m ic a l e l e m e n t s a p p e a r s t o b e a p r e r e q u i s i t e f o r r a p i d
g ra p h itiz a tio n .
The a b s e n c e o f g r a p h i t e i n t h e alu m in u m a n d m an g an ese s t e e l s
31 - Photomicrographs of Steels Tempered 125 Hours at
630°C in a Vacuum.
A. — Carbon.
B. - 0.07 Al.
C. - 0.37 Al.
D„ - 0.03 Mn.
E. - 0.30 I/In.
F. - 0.4-8 Si
Uitric Acid Etch. X250
Fig. 32 — Photomicrographs of Steels Tempered 125 Hours at 710°C
in a Vacuum.
A. — Carbon.
S. — 0.07 Al.
C. — 0.37 Al.
D. - 0.0S Mn.
E. - 0.30 Mn.
F. - 0.4S Si.
Nitric Acid Etch. X250
- 53 -
after tempering in a lead bath, at the high temperatures is possibly due
to th^emperature being too high for a stressed condition to exist
enough to influence the stability of the carbide.
long
While the low rate of
graphitization at the low temperatures, where a stressed condition could
exist more easily, is probably due to slow diffusion of the atoms involved
in the reaction.
Thus the optimum condition between stress and rate of
diffusion appears to occur in the neighborhood of 630°C for the aluminum
and manganese steels containing approximately 0.35% of these alloying
elements.
However, lower quantities of aluminum cause the maximum
graphitization to take place at lower temperatures.
taining 0.07% aluminum, it occurs at 550°C.
Thus in a steel con­
This difference in the be­
havior of the two aluminum steels may be due to the higher aluminum con­
tent imparting more resistance to the removal of the stressed condition.
This would mean that the high aluminum steel would have to be heated to a
higher temperature in order for the stresses to be removed.
Although the
quantity of graphite was small in the low manganese steels, there is some
evidence that the temperature of maximum graphitization does not change
appreciably with increase in manganese.
The maximum quantity of graphite
appeared to form at 630°C regardless of whether there was 0.08 or 0.30%
manganese present in the steel.
In the aluminum steels it is believed that metallic aluminum
and possibly aluminum oxide are acting with stress to produce graphitiza—
tion.
Since aluminum oxidizes very easily and hydrogen is unable to re­
duce the resulting oxide at room temperature, it is impossible to add
aluminum to a steel without some aluminum oxide being added.
Austin and
Fetzer (39) have clearly demonstrated that aluminum oxide will promote
- 54- -
graphiti zati on.
The factor functioning with stress to produce graphite in the
manganese steels appears to be manganese sulphide although no definite
proof is available to show that metallic manganese should not be classed
with the sulphide as an accelerator.
However it is believed that strong
evidence is available for the support of a theory based on the effect of
manganese sulphide.
direction of rolling
The sulphide particles occur in streaks parallel to the
and very often graphite particles are located in
these streaks after a tempering treatment.
Fig. 33B shows an area in the
0.3056 manganese steel where graphite is present near manganese sulphide
inclusions.
Graphite 7/ill also form in the sulphide streaks during slow
cooling from a temperature above the critical range.
in Fig* 33A.
This is illustrated
It is interesting to note that the sulphur steel, #S0, con­
tained from 2 to 3 times as much sulphur as the manganese steels and yet
it did not graphitize.
Thus, iron sulphide does not appear to promote
graphitization.
The presence of a greater quantity of graphit^in the high
manganese steels than in the steels containing small amounts of this
element may be due to the presence of a greater quantity of manganese
sulphide in the former.
Although the sulphur content is approximately
constant in all the alloys, it is possible that the low manganese steels
do not contain enough of the element to ensure that all of the sulphur
is combined with manganese.
If the quantity of manganese is low some
of the sulphur will combine with the iron.
Appreciable graphite may form in a steel during tempering be­
fore its presence will be indicated by the hardness data.
Examination
Fig. 33 - Photomicrographs Illustrating Location of Graphite
with Reference to Manganese Sulphide and Carbide.
0.37 Mn: A. - Slowly Cooled from 1125°C.
B. — Tempered 125 Hours at 630°C.
O.OS Mn: C and D. — Tempered 5 Hours at 630°C.
Picric Acid Etch for A, C, D. Nitric Acid Etch for B. X1000
- 55 -
of the hardness curve of the 0.24$ manganese steel in Fig. 16 obtained
for a temperature of 630°C reveals no discontinuities, the plots of all
data lying close to the straight line.
Yet chemical analysis of a
sample of this steel treated for 125 hours at this temperature revealed
the presence of 0.32$ graphite, Table VI.
Turning now to a consideration of the profound difference in
the stability of the carbide when tempering is conducted in a lead bath
as compared with the results obtained for a tempering treatment in vacuo,
it is quite evident from the photomicrographs in Figs. 27 and 28 that
much more graphite isproduced when tempering is conducted in a vacuum.
The supposition that hydrogen left in the steels after the melting and
casting processes is removed by a vacuum treatment appears to agree with
the present experimental results.
However, articles published by Zapffe
and Sims (4-2) and Schwartz and Guiler (43) since the present research was
completed indicate that hydrogen is very easily removed during a heating
at elevated temperatures.
A piece of white iron containing 0.0005$
hydrogen was melted by Schwartz and Guiler (4-3) > held molten in the
presence of hydrogen, at atmospheric pressure, for several hours and
then solidified in the same atmosphere.
and found to be 0.0015$.
The hydrogen was redetermined
After heating in air for 8 l/2 hours at 800°C,
the hydrogen content fell to 0.0007$.
Since the "pure" steels considered
in the present investigation were heated at high temperatures during the
forging, rolling, and annealing operations, it is probable that most of
the hydrogen retained during solidification was lost during these
fabrication operations.
Consequently, little hydrogen would be in the
steels prior to the tempering treatments.
Thus, the recent literature
- 56 -
presents a serious objection to any hydrogen theory that may be presented
to explain the behavior of the steels during a vacuum temper.
Fetzer (29) has recently found that commercial carbon tool
steels will graphitize more easily during tempering in a vacuum than when
they are heated in an oxygen-free lead bath.
Since the commercial steels
were melted under conditions that were much less favorable for solution
of hydrogen in the molten metal than the conditions prevailing during
the preparation of the ’’pure" steels, it is difficult to believe that the
increased amount of graphite formed during a heating in a vacuum is due
to the removal of hydrogen.
An explanation that appears to be more consistent with the
facts is based on the diffusion of oxygen into the steel during the
vacuum tempering treatment.
Austin and Fetzer (39) have shown that a
steel will graphitize to a greater extent when it is tempered in an
oxygen-bearing 'lead bath than it will in an oxygen—free lead bath.
These data also indicate that a steel high in metallic aluminum will
graphitize to a greater extent than a steel containing a small amount of
this element when they are tempered in an oxidized lead bath.
A
relationship similar to the above has also been obtained by Fetzer (29)
when tempering was conducted in a vacuum although a greater quantity of
graphite was produced when heating was performed in the oxygen-bearing
lead bath.
When a few of the ’'pure” steels were tempered for 125 hours at
670°C in an oxidized lead bath it was found that more graphite occurred
than when they were heated in oxygen—free lead.
The plain carbon, 0.01%
- 57 -
aluminum, 0.12 and 0.14$ silicon, 0.24 and 0.30$ manganese, 0.12$ nickel,
0.13$ chromium, 0.35$ copper and 0.04&$ sulphur steels were used in the
comparison.
Examination of their microstructures revealed that graphite
only occurred in the aluminum and manganese steels and it was more
abundant in the samples treated in the oxidized lead bath.
However the
quantity of graphite formed was much less than that produced in the
samples tempered in a vacuum for 125 hours at either 630° or 710°C.
This
looks like a contradiction to the oxygen theory suggested as an explana­
tion for the behavior of the steels in a vacuum since a heavy oxide scale
formed on the samples treated in oxygen-bearing lead while scale was
absent from the specimens tempered in a vacuum.
Although heating was
conducted for the same length of time during both tempering operations
it must be remembered that the vacuum specimens had been heated for
approximately 30 hours in a "vacuum11 in the temperature range 1025° to
1125°C prior to hardening and tempering.
Thus this high temperature
treatment might easily add more oxygen to a steel than a heating at
670°C for 125 hours in oxygen-bearing lead.
The presence of small graphite particles near the edge, the
particles in the interior being larger, of the steels that graphitized
during tempering in a "covered" lead bath may have been due to oxygen
penetration since there was a slight evidence to indicate that a thin
oxide scale formed on the samples during the sojourn in the lead.
The
formation of graphite adjacent to the edge of some of the steels
tempered in a vacuum is also possibly due to oxygen penetration.
The rapidity with which graphite forms in the aluminum,
manganese, and silicon steels when they are tempered in a vacuum at 710 C
- 58 -
tends to indicate that oxide particles are probably playing a greater
part in promoting graphitization than stress since it is difficult to
see how stress could be exerting an effect at such a high temperature.
It is interesting to note that during a vacuum—temper the most
graphite forms in the steels containing alloying elements that have a
relatively strong affinity for oxygen, namely, aluminum, silicon and
manganese.
The resistance to graphitization exhibited by the chromium
steels may be due to the chromium having a stronger affinity for carbon
than for oxygen.
It has been previously mention in connection with Fig. 30 that
a greater hardness will result when a steel is water quenched from the
tempering temperature 710°C, following complete graphitization at this
temperature, than when it is similarly cooled from 630°C after complete
graphitization has been obtained.
Whiteley (4-4-) and others have shown
that the solubility of carbon in ferrite increases with increase in
temperature and it reaches a maximum at the eutectoid temperature,
approximately 0.035$ carbon being dissolved at this temperature and
only 0.01$ at room temperature.
Since some delay occurred between the
heat treatment and the determination of the hardness of the steels
quenched from 630° and 710°C, the difference in the hardness of the two
sets of specimens may have been caused by different amounts of carbon
being dissolved in the ferrite.
Davenport and B a m (4-5) have demonstrated that a relatively
high hardness may result when a low carbon steel is quenched from a
- 59 -
temperature close to "the A^ critical point and. aged at room temperature
for several days.
Vfhen they rapidly cooled a 0.06% carbon steel from 720°C
its hardness v/as 70.
However, after aging at room temperature for 6 days,
the hardness was 85.
In the present investigation 6 days elapsed before the hardness
was determined on the specimens that were water quenched from the 710°C
tempering temperature.
The hardness of the completely graph!tized
specimens ranged from 65 to 85 Rockwell ,,B 11, with the steels containing
the highest percentage of aluminum and manganese exhibiting the higher
hardness values.
Thus, it is quite possible that the high hardnesses
are due to incipient precipitation from the ferrite.
Up to the present time no mention has been made of the possible
effect of the alloying elements on the solubility of carbon in ferrite.
Information could not be located in the literature that would pertain to
aluminum and manganese.
Theoretical considerations would lead to the
conclusions that these elements would alter the carbon solubility in
ferrite but since data are lacking it is impossible to estimate the
magnitude.
Although it is known (4.6) that the hardness of carbon-free
iron may be increased by the solution of chemical elements, it is not
believed that this explains the great hardness difference occurring
when quenching is conducted from high and low tempering temperatures.
Since it is not likely that any appreciable quantity of the alloying
elements is soluble in the graphite, the alloying elements must be com­
pletely dissolved in the iron when carbide is absent.
Temperature should
- 60 -
not change the solubility of the chemical elements, excepting carbon, in
the ferrite.
However, the alloying elements might exert an indirect
effect on the hardness by altering the carbon solubility of the ferrite.
When graphite forms in a steel its nucleus seems to form
adjacent to a carbide particle.
While subsequent growth appears to take
place at the expense of the neighboring carbide.
Photomicrographs
illustrating the presence of graphite adjacent to carbide particles are
presented in pigs. 33^ and 33D.
The small narrow dark areas in Fig. 33D,
that resemble cracks, may be locations where graphite formed in crevices
which were produced during the hardening operation.
- 61 -
Conclusions
The conclusions that will be made concerning graphitization
apply to 'Jpure" iron alloys (steels) containing approximately 1.1% carbon
and 1/2% or less of a third element when they are tempered in the
temperature range 550° to 710°C for time intervals of 1/2 hour up to
125 hours.
When
a hardened steel is tempered in a lead bath covered v/ith
nlead pot” carbon, the following conclusions can be made concerning the
factors influencing graphitization:
1. Very little graphite forms when a stressed condition,
isimparted by hardening,
such as
is absent from the steel prior to a heating at
sub-eutectoid temperatures.
2. Stress alone does not cause graphite to form.
3. Appreciable graphite forms wrhen stress is present in steels con­
taining
4.
metallic aluminum or manganese sulphide.
Steels containing up to approximately 1/2% of silicon, nickel,
chromium, and copper or about 0.05% sulphur and tin do not graphitize to
any appreciable extent when stress is present.
5. In aluminum and manganese steels, graphitization is a maximum
yd.thin the temperature range 550°-630°C.
An increase in aluminum from
0.07 to 0.37% raises the temperature of maximum reaction from 550°C to
a temperature lying in the range 590° "to 630°C.
While an increase in
manganese up to 0.30% does not appear to shift the temperature from 630°C.
6. Appreciable graphite (approximately 0.30%) may form in a steel
before its presence is indicated by the hardness.
- 62 -
The following conclusions may be made concerning the effects produced in
a steel when it is normalized, hardened, and tempered in a vacuum:
7.
Much graphite forms in steels containing aluminum, manganese,
and silicon.
While the quantity of graphite produced is much greater than
the quantity obtained when the steels are hardened in a vacuum and
tempered in lead.
8.
Only a small amount of graphite (occurring principally at the
edge) is formed in carbon and alloy steels containing up to approximately
l/2 per cent of nickel, chromium, and copper.
While a similar result
obtains for about 0 .05 % sulphur or tin.
9.
Graphitization appears to take place as rapidly at 710°C as it
does at 630°C.
10. The magnitude of the graphitization is markedly lowered in the
manganese and silicon steels when the vacuum normalizing treatment is
omitted.
When these steels are hardened and tempered in a vacuum at 630°C,
the quantity of graphite is little different from that formed during a temper
in lead at the same temperature.
11. The increased rate of graphitization resulting from a prolonged
heating in a vacuum appears to be due to the penetration of oxygen during
the treatment.
12. In steels containing aluminum, manganese, or silicon, the ferrite
has a greater solubility for carbon at 710°C than at 630°C.
- 63 -
CHAPTER VI
Microstructure of Annealed Steels
When the "pure" steels are heated above the critical range and
slowly cooled to room temperature only the structure of the higher alloys
shows much departure from the structure of the plain carbon steels.
Since
the low alloy steels are fairly similar to the carbon steels, only the
structures of the steels containing approximately 1/2 per cent of added
third element will be discussed in this chapter.
Experimental Procedure
Annealed structures were prepared by heating samples in a vacuum
(0.10-0.15 m.m. Hg) for 1 hour at 1000°C and slowly cooling at a rate of
approximately 5°C per minute.
Experimental Results
The microstructures of the carbon, #87, and high alloy steels
are reproduced at a magnification of 200 diameters in Fig. 34..
In Fig.
34A it is quite evident that appreciable ferrite has formed in the plain
carbon steel between the grain boundary carbide and the pearlite.
other words, the plain carbon steel is abnormal.
In
When approximately 1/2
per cent of manganese, silicon, nickel, chromium, and aluminum is added
to the plain carbon steel the structure becomes less abnormal.
This is
clearly demonstrated by a comparison of Figs. 34B> 34-C, 34P> 34-E, and
34G with Fig. 34A.
Copper was the only chemical element, studied in
large quantities, that failed to produce a structure that was less abnormal
Fig. 34 - Photomicrographs of Carbon and High Alloy Steels
(Containing Approximately 1/2 Per Cent of Added
Element) Annealed 1 Hour at 1000°C.
A. - Carbon.
B. - 0.30 Ivin. C. - 0.48 Si. D. - 0.53 Ni.
E. - 0.46 Cr.
F. - 0.35 Cu.
G. - 0.37 Al.
PicricAcid Etch. X200
- 64 -
than the structure of the plain carbon steel.
In fact, 0.35% copper
appears to make the structure slightly more abnormal.
with Fig. 34A.
Compare Fig. 34F
It is interesting to note that chromium was the only
element that produced a structure that was normal, that is, no free ferrite
located adjacent to grain boundary carbide.
Some of the alloying elements produce peculiar effects on the
grain boundary carbide.
Striking differences can be noted in the silicon,
chromium, and aluminum steels, Fig. 34 (C, E and G).
contain approximately 1.15% carbon.
All of these steels
The network carbide of the chromium
steel, Fig. 34E, is generally fairly uniform in width and has a’light
color.
The carbide of the silicon steel, Fig. 34C, has a dark appearance
and its width is non-uniform.
While it is practically impossible to
identify any grain boundary carbide in the aluminum steel, Fig. 34G, at
a magnification of 200 diameters.
A higher magnification, X1000, of
the structures of these three steels as well as that of the plain carbon
steel are illustrated in Fig. 35-
Fig. 35A represents the plain carbon
steel while the silicon bearing steel is shown in Fig. 35B.
It is quite
evident from the latter figure that the presence of silicon causes the
grain boundary carbide to form in a "feathery-like" mass.
This may ex­
plain the dark appearance of the carbide at a low magnification.
Fig.
35C demonstrates that grain boundary carbide is present in the aluminum
steel but due to its narrow width and discontinuous nature it is difficult
to distinguish it from the carbide present in the pearlite.
Fig. 35D
illustrates the structure of the chromium steel when it is viewed at a
high magnification.
Fig. 35 - Photomicrographs of Carbon and High Silicon, Aluminum,
and chromium Steels (Containing Approximately 1/2 Per
Cent Added Element) Annealed 1 Hour at 1000°C.
A. — Carbon.
B. - 0.4-S Si.
C. - 0.37 Al.
D. - 0.4-6 Cr.
Picric Acid Etch. X1000
Conclusions
1. When a ’’pure” iron-carbon alloy, containing approximately 1.1$
carbon is prepared in the manner described in Chapter ill and subsequent­
ly annealed at 1000°C for 1 hour, its structure is abnormal.
2. The addition of approximately 1/2 per cent of manganese, silicon,
nickel, chromium, and aluminum makes the structure of the iron-carbon
alloy less abnormal, chromium having a very strong effect.
While copper
makes it slightly more abnormal.
3. Silicon promotes the formation of a '’feathery-like" carbidic mass
at the grain boundary.
4. Aluminum promotes the formation of a thin discontinuous carbide
network.
- 66 -
Bibliography
1.
2
.
C. R. Austin and B. S. Norris, “Effect of Tempering Quenched Hypereutectoid Steels on the Physical Properties and Microstructure”,
Transactions, A. S. M., Vol. 26, 1933, p. 788.
C. R. Austin and B. S. Norris, "Temperature-Gradient Studies on
Tempering Reactions of Quenched High-Carbon Steels",
Metals Technology, A. I. M. E., June 1938, p. 1.
3.
R. D. Steigerwalt, "Austenitic Grain Size of Pure Low-Alloy Steels",
Unpublished Senior Thesis, The Pennsylvania State College, 1939.
4-
Metals Handbook, 1939» p. 127, American Society for Metals.
5.
E. C. Bain, "Functions of the Alloying Elements in Steel",
American Society for Metals.
6.
E. S. Greiner, J. S. Marsh, and B. Stoughton, "The Alloys of Iron
and Silicon", McGraw-Hill Publishing Company.
7.
J. S. Marsh, "The Alloys of Iron and Nickel", Vol. 1, McGraw—HillPublishing Company.
8 . J. L. Gregg and B. N. Daniloff, "The Alloys of Iron and Copper",
McGraw-Hill Publishing Company.
9.
D. K. Bullens and Battelle Memorial Institute, "Steel and Its Heat
Treatment", Vol. 2, Wiley Publishing Company.
10. F. R. Bonte and M. Fleishann, "Developments in Graphitic Steel for
Tools and Dies", Metal Progress, April 1937, p. 4-09.
11. C. R. Austin and M. C. Fetzer, "Reactions to Annealing above the
Eutectoid Temperature of Quenched Rypereuteetoid Steels",
Transactions, A. S. M., V27, 1939* p. 18.
- 67 -
12.
H. A. Schwartz, "The Conversion of Solid Cementite Into Iron and
Graphite", Journal, Iron and Steel Institute, vol II, 1938,
p. 205P.
13.
A. Boyles, "Freezing of Cast Iron", Transactions, A. I. M. E.,
Vol. 125, 1937, p. 141.
14.
A. L. Boegehold, "Some Unusual Aspects of Malleable Iron Melting",
Transactions, A. S. M., Vol. 26, 1938, p. IO84.
15.
H. C. Ishen, "The Spheroidization of Cementite", Heat Treating and
Forging, July 1922, p. 300.
16. R. W. Bailey and A. M. Roberts, "Testing of Materials for Service
in High-Temperature Steam Plant", Institute of Mechanical
Engineers, Vol. 122, 1932, p. 209.
17.
B. J. H. Taylor, "The Spheroidization of Steels at Sub—Critical
Temperatures", Doctor Dissertation, University of Michigan,
1935.
18.
J. A. Duma, "Spheroidizing Steels", Heat Treating and Forging,
Dec. 1934, P* 585.
19.
A. B. Kinzel and W. Crafts, "The Alloys of Iron and Chromium",
Vol. I, McGraw-Hill Publishing Company.
.
J. L. Gregg, "The Alloys of Iron and Molybdenum", McGraw-Hill
Publishing Company.
.
J. L. Gregg, "The Alloys of iron and Tungsten",. McGraw-Hill
Publishing Company.
20
21
22 .
P. Breuil, "Copper Steels", Journal, iron and Steel Institute,
Vol. 74, 1907, p. 1.
23.
C. S. Smith and E. W. Palmer, "The Precipitation Hardening of Copper
Steels", Transactions, A. I. M. E., Vol. 105, 1933, p. 133.
- 68 -
24.
A. W. P. Green, "Black Fractures in Carbon Tool Steels", Chemical
and Metallurgical Engineering, Vol. 27, 1922, p. 265.
25.
R. S. Archer, "Graphitization of White Cast Irons", Discussion byEmmons, Transactions, A. I. M. E., Vol. 67, 1922, p. 445.
26.
A. Portevin and P. Chevenard, "La Graphitization des Acier a Basse
Temperature", C. r . Academy Science, Paris, Vol. 193, 1931, p. 169.
27.
A. B. Kinzel and R. W. Moore, "Graphite in Low-Carbon Steel",
Transactions, A. I. M. E. Vol. 116, 1935, p. 318.
28.
S. Epstein, "The Alloys of Iron and Carbon", Vol. 1, p. 14-3,
McGraw-Hill Publishing Company.
29.
M. C. Fetzer, Private Communication.
30.
C. Wells, "Graphitization in High Purity Iron-Carbon Alloys",
Transactions, A. S. M., Vol 26, 1938, p. 289.
31.
H. A. Schwartz, "Malleable Iron - Recent Progress", Metals and
Alloys, Vol. 2, 1931, p. 14-3.
32.
H. Seitz, H. J. McDonald, and C. VJells, "Heat Capacity of Iron
Carbide from 68° to 298°K. and the Thermodynamic Properties
of Iron Carbide", Metals Technology, Dec. 1939.
33.
E. H. Engel, "The Softening Rate of a Steel When Tempered from
Different Initial Structures", Transactions, A. S. M., Vol. 27:
1939, p. 1.
34-.
J. 0. Arnold and A. A. Read, "The Chemical and Mechanical Relations
of Iron, Chromium, and Carbon", Journal, Iron and Steel
Institute, Vol. I, 1911 , P* 24-9.
35.
E. Maurer, T. Doring, and H. Buttig, "The Composition of Carbide in
Chromium Steels", Archiv. fur das Eisenhuttenwesen, Vol. 7,
1933-34-, P. 247.
i
- 69 -
36.
. Westgren, G. Phragmin, and T. Negresco, "On the Structure of the
Iron-Chroraium-Carbon System", Journal, Iron and Steel Institute,
1928, p. 383.
37.
. D. Campbell and J. F. Ross, "The Chromium-Iron Equilibrium in
Carbides Removed from Annealed 2.23 Peir Cent Chrome Steels",
Journal, Iron and Steel Institute, Vol. II, 1925, p. 255.
38.
. 0. Arnold and A. A. Read, "The Chemical and Mechanical Relations
of Iron, Manganese, and Carbon", Journal, Iron and Steel
Institute, Vol. I, 1910, p. 169-
39.
C. R. Austin and M. C. Fetzer, "Effect of Composition and Steel Making
Practice on Graphitization Below the Af,of Eighteen 1 Per Cent
Carbon Steels", Presented at February Meeting, 1940, of the
A. I. M. E.
40.
J. H. Kelley, "A Study of Spheroidization of Bypereutectoid Steels
and Thermal Gradients", Unpublished Senior Thesis, The Pennsyl­
vania State College, 1935.
41.
H. A. Schwartz, "Nucleization of Graphite by Manganese Sulphide",
Metals and Alloys, Vol. 6, Nov. 1935*
42.
C. A. Zapffe and C. e . Sims, "Hydrogen, Flakes and Shatter Cracks",
Metals and Alloys, Vol. 11, June 1940, p. 177.
43.
H. A. Schwartz and G. M. Guiler, "Hydrogen in Solid White Cast Iron",
Transactions, A. F. A., Vol. XLVII, March 1940, p. 742.
44*
J. H. Whiteley, "The solution of Cementite in Alpha Iron and Its
Precipitation", Journal, Iron and Steel Institute, Vol. CXVI,
1927, II, p. 293.
45.
E
46.
R. M. Allen, Private Communication.
S. Davenport and e . C. Bain, "The Aging of Steel", Transactions,
A. S. M., Vol. 23, 1935, P- 1047.
Документ
Категория
Без категории
Просмотров
0
Размер файла
5 540 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа