«(y 1

. ‘ .
’V . n
v )

. 7'.
‘ 5 fl ‘ h
fiukfifigx
‘7’ 163‘: m.
nix-km W

E ,'h
A -

5 92.83"
I

‘a«uu.. z.

.. g. -. -,

z . 5.- vi”;

2 T." 5-.“

KM»
'31,

”9 {9.61m

‘.~r‘-
r
.n ._
A.--..

‘7»ng

. nu»- sm-V
v.‘ .
r.

.1..—
““-~A‘. -
3“..A;s.t..k“~

d

‘Qs

u

1.4 r
" n» .a) ‘-

‘7
I. ”n. ..

un- Iz‘

. , .
”mm

‘4‘."

SI

do“.
an“.

d:
' I'm" 'f Y}
‘ “(0‘
\‘I’ ’1': A"
. } ‘ “

 

AEEAS

D‘llr

MICROSTRUCTURES OF QUENCHED CARBON STEEL IN THE
. . CRITICAL RANGE

by
IILLIAI ALFRED §§EE,

A IRESIS
submitted to the Gradnete School or lichizan
State College of Agriculture and Applied

Science in partial fulfilhent of the
requirements for the-degree of

HASIER or BCIENGI

Department of Chemical‘zncineering
1941

 

 

 

 

 

1.4.3515

 

 

 

 

 

 

, a?“

 

 

AOKNOILIDGHEBTS
The author Iishes to express his gratitude for
the cooperation and guidance received from the late
Professor H. E. Publow.

 

He also vishee to express his thanks to Professor

R. L. Sweet under whose guidance the tort was continued. !

V

TABLE OF CONTENTS

IDtl‘OduCtianeeeeeeeeeeeeeeeeeeeeeeeeeeee P880 1
Experimental Procedure and Results...... Page 3

Conclusions and Suggestions for
Further 'Orkeeeeeeeeeeeeeeeeeeeeeee Page 85

Bibliography....uuu.................. Page 87

*****§*

 

-1-
INTRODUCTION

The hardening of steel is a subJect of great
‘inOrtance in the netallurgical field. It is a subject
which has been investigated cuite extensively in.past
years, and it seelm safe to say that it will continue
to he investigated in the future. The exact manner in
which austenite transforns when rapidly cooled through
the critical range is still a matter of controversy
along letallurgists.
.. is the title indicates, this writing is concern-
ed with the nicrostructures of duenched carbon steels
in the critical range. All work was done with the Chev-
enard Thermal Analyser, a direct type of dilatoneter
available in the research laboratory here. The study
was limited to specimens of four different steels, SAE
1020, 1045, 1060 and 1090, with the enphasis on the
latter. _ y _

The primary purpose of this investigation was to
attempt to check the accuracy of the dilatouetric method
of determining the progress of the critical transformation
of the steels studied. This was done by heating the spec-
ilens in the dilatcleter to a desired point on the dila-
tation curve, and then removing then duickly from the

instrument to he quenched in cold water. The tine requir-

 

 

-2-
ed to remove each sample from the instrument and immerse
it in the quenching medium was usually about one second,
or at the most, not more than two seconds. Hence by neg. ‘
lecting the cooling of the specimen in air during the
quenching operation it could be assumed that the micro-
structure of the quenched specimen corresponded to the
actual microstructure at the temperature indicated by
the dilatometer. This will be explained in more detail
in subsequent discussion.

The quenching medium used throughout in this »
work was cold tap water. The etching of the polished
metallographio specimens was done with hital, a two per
cent solution of concentrated nitric acid in ethyl al-

cohol.

‘I'D'Ifi'i‘i

 

-3-

EXPERIMENTAL PROCEDURE AND RESULTS

Figure 1 shows the portion of the Iron-Carbon
equilibrium diagram which enters into this discussion.
This was reproduced from Rosenholts and Oesterle1.Above
the upper critical line, the A3 line for hypoeutectoid
iron-carbon alloys or the Ac, line for hypereutectoid
steels, a steel specimen would consist entirely of aunt.
emits. Austenite is defined by the letals Handbookz as a
solid solution in which gamma iron is the solvent. The
maximum solubility of carbon in gamma iron is l.7$ at
2066 deg. P., or 1130 deg. c. This decreases with temp-
erature to about 0.87% at 1333 deg. 3., or 725 deg. 0..

Let us consider as an example, alloy 'A' of fig-
ure 1 which contains 0.4; carbon. Above the A3 line this
alloy would consist of austenite, a solid solution of
0.4% carbon in gamma iron. If this specimen is cooled,
as indicated by the dotted line on the diagram, ferrite
begins to separate out at the A3 line according to well-
known metallographic principles. As the temperature is
further lowered--s1owly to maintain equilibriume-more
ferrite separates out of the solid solution and raises
the carbon content of the austenite remaining. At the
lower critical, or A1 line, the remaining austenite

transforms to a eutectoid aggregate of ferrite and com-

 

cuisine .2 (vs, 2.0sz DEANGMZ no. 20. u+m

 

.-5-
entite known as pearlite. For the sake of clearness, these
last three terms will be defined.

The term “ferrite” is applied to solid solutions
in which alpha iron is the solventz. The maximum solubility
of carbon in alpha iron occurs at 725 deg. c. and is app-

roximately 0-035$- This decreases as the temperature falls

 

to loss than 0.01% at room temperature.

Alpha iron, of course, is one of the three allo-
trOpic forms of iron. In pure iron, the alpha, or body-
centered cubic form is stable below 906 dog. 0. Gamma iron
which is the solvent in austenite has a face-centered cu-
bic crystal structure and is stable from 906 to 1400 dog.
0. From 1400 to 1535 deg. c. the crystal structure is
body-centered cubic, and this form is known as delta iron.

Iron and carbon form a hard, brittle, crystalline
compound, the composition of which is represented by the
formula 3e30. This compound is referred to as iron carbide
or cementitez. It contains 6.68% carbon by weight.

Pearlite is the lamellar aggregate of ferrite and
cementite resulting from the direct transformation of aust-
enite at the A1 temperature. By annealing steel at a temp-

1,2-- All sources of information used are listed in a
bibliography at the end of the paper...

-6-
erature below but near the lower critical point, it is
possible to cause the cementite to spheroidize in a ferr-
ite matrix. This aggregate is known as “granular“ or
“globular“ pearlitez.
As mentioned before, in plain carbon steel. pear-

lite forms when austenite is slowly cooled past the lower

critical line. If this rate of cooling is increased var- 4
ions structures may be formed instead of pearlite, dep-
ending upon the rate of cooling. A very rapid rate of coolaw;
ing such as quenching in cold water results in the austen-
ite being converted to 'martensite'. i

As defined in the Metals Handbookz, martensite is a
microconstituent or structure in quenched steel character-
ized by an accicular or needle-like pattern on the surface
of polish. It has the maximum hardness of any of the de-
composition products of austenite.

A slower rate of cooling than that necessary to
produce martensite, may result in a structure termed "troo-
stiteg. It etches very rapidly and as a result appears
quite dark under the microscope. It probably consists of a
very fine aggregate of ferrite and cementite.

The-different types of microstructures which have

been defined briefly above will be expanded upon in more,
detail as they come up in the discussion of the experimen-

tal work.

Figure 2 is a photograph of the dilatometer used
in all the experimental work. This is the Chevenard therm-
al analyser, a direct type of dilatometer, which records
expansion-time-temperature data. The expansion and con-
traction of a specimen is recorded by means of a lever
system and a stylus which records on a revolving drum giv-
ing a continuous record of expansion and contraction.

To determine the temperature at any given time,

a standard pyros specimen having a nearly constant coef-

ficient of expansion over a working range is placed beside

 

the test specimen, and is connected by a similar lever
system to another recording stylus. Two curves are thus
plotted simultaneously, one showing time versus dilatation
of the experimental Specimen and the other showing time
versus dilatation of the known standard. The temperature of
the specimen can then be found at any given time by using a
scale calibrated from the standard pyros specimen, and a
dilatation curve can be plotted showing temperature versus
~dilatatien in units x 10‘3 per unit.

Figure 3(a) is a typical curve showing the dilata-
tion of a specimen of 1090 steel plotted against temperat-

ure. Ihile figure 3(b) is an exact duplicate of the origin-
al dilatometer curve from which 3(a) was plotted. From A

 

-8-

 

 

 

 

Figure 2-- The Chevenard Dilatometer

 

-9.
to B normal expansion of ferrite and cementite occurs. At
'3 the expansion stops and contmaction begins denoting the
beginning of the critical range transformation, or Ac
transformation. From B to 0 the loss in volume due to the
alphapgamma transformation continues as shown by the con-
traction, while from 0 to D nearly constant volume obtains
until the solution of the carbides is complete}. From D to
x, at which point the furnace current was shut off, expan-
sion of austenite takes place.

Cooling at the normal furnace rate takes place and
there is a period of austenite contraction until a point is
reached where expansion again starts, marking the beginning
of the Ar transformation. The increase in volume accompany-
ing the gamma to alpha formation continues until the trans-
formation is completed and then the normal contraction
takes place.

Ihen heating through the critical range the trans-
formation is referred to as the Ac transformation; on cool-
ing it is termed the Ar transformation. Thus on cooling
the lines on the equilibrium diagram are designated as the
Arl line, Ar3 line, and so on. While on heating, the lines
become the Acl and A33 lines.

Having considered a typical dilatation-temperature

curve, we now come to the actual experimental work of this

 

 

. 1;.qrt .2 L vb nilnr...‘ )_nqun: 3) r.) JPN

-11-

 

 

 

 

 

 

 

 

 

“ox-6m. stab

 

 

 

 

 

Figure 3(b)-- Duplicate of an original dilatation

curve e

-12-
problem.

To aid 1n identifying the samples, the 1090 spec-
imens are termed as the 'A' samples, the 1060 steels as
the 'B' samples, 1045 as '0', and 1020 as the "D" samples.
Since only that portion of the dilatation curve which cov-
ers the critical range is of interest in this discussion,
the origin has been suppressed in each of the following
curves so as to enlarge the transformation range. The be-
ginning of each curve is marked '8', and the end of the
curve or the point where the specimen was removed from
the instrument and quenched is marked “Q”.

-SAB 1090 8tec1s-- V

Figure 4 is the dilatation curve of specimen A-l.
As the curve indicates this specimen of 1090 steel was
quenched when it had cooled just to the upper critical
or Ar} line. Accordingly the photomicrographs of this
sample would be expected to show a typical martensitic
structure. These microstructures are shown in the photo-
graphs of figure 5, and they are seen to be martensitic.

The hardness of specimen A91 was found to be 620
on the Rockwell instrument. lartensite is the hardest
known structure in steel. It is also extremely brittle.

Figure 6 is the dilatation curve of specimen ‘92.

An attempt was made to quench this sample when it had

 

 

5(a) 8001

 

5(b) 25001

4
\ l. v.
, 4:“, ,‘

 

Figure 5-- Photomicrographs of specimen A91.
Rockwell Hardness--62c

 

 

 

7(a) 1001

 

 

Figure 7-- Photomicrographs of specimen A92.
Rockwell Hardness--520

8(a) zsoox

 

8(b) 2500i

 

J'igure 8-- Photomicrographs of specimen A-2.

 

. r - \.,x\\\\
o :2 OV\
.-. $9

\

o\ .
.e

s

J.-
, . cydm.
. o .3 .
140‘. u. res-1‘1! at 173.1,“.-

 

 

 

2500X
Figure 9-- A 9" x 7” plate of Specimen Ar2.

 

-19-
cooled to a point about halfway through the critical
range. However it must be noted that for a steel of
this composition (0.9% carbon) the entire critical temp-
erature range is very narrow. For this reason the loss
in temperature during the time the specimen is lifted
through the air from the dilatometer to the quenching
bath may have a real effect on the results.

The photomicrographs for specimen A-2 are shown
in figures 7, 8 and 9. The structures observed here are
light patches of martensite, dark masses of troostite
and definite areas of lamellar pearlite. This indicates
that the specimen was quenched just as the austenite
was in the process or transforming to pearlite. In
other words the austenite apparently goes to martensite
to troostite and then to pearlite. Or it may transform
directly to troostite and then to pearlite.

Figure 10 is the dilatation curve of specimen
A—3, which was quenched when the curve traced by the
dilatometer indicated that the specimen was Just past
the lower critical or Arl line. The photomicrographs in
figures 11, 12 and 13 show the transformation to pear-
lite to be practically completed. However figure 11(a)

at 8001 shows some rather ill-defined areas indicating
that the final stage of aggregation has not yet been

-20-
reached. These areas might be called sorbite, or sorbit-
1c areas, which is the name usually given to a transfor-
mation stage in between troostite and pearlite.

Figure 13 is a print from a large plate taken at
a magnification of 50001. This print is included merely
to show the possibilities of such a magnification. It is
evident that nothing is gained in clearness or increase
in detail.

It was decided to repeat the work Just described
with the quenching being done froma heating curve instead
of a cooling curve. That is, the specimens were heated
through the critical range, allowed to cool down through
it, and then heated up to the desired quenching point in
the critical range. The main difference here is that the
quenching is done from a relatively higher temperature
because of the hysteresis effect. As may be readily not-
iced from all the dilatation curves, the critical trans-
formation takes place at definitely higher temperatures
when heating than when cooling.

Specimen A94 was heated beyond the upper crit-
ical or Ac3 line and quenched. As expected, the struct-
ure turned out to be martensitic. The photomicrographs
show the martensitic matrix with some small white par-

ticles of cementite scattered through it. These cemen-

 

11(a) soox

 

11(b) 25001

 

Figure 11-- Photomicrographs of specimen A-3.
Rockwell Hardness--95.'B

 

 

- I _ o
\ u \ V")
2- - \ .,,‘
1 ’ \ .h 1 I}, \ fl
. q , v
-. \\ (x )1’ '1' |'._
, gt ‘ 1 ‘,f,iffivfi
1 , ‘ .c« a“ , :“J4;u
‘ - 1‘. .‘7 , ‘_ ‘7’ l
k , :’ I ’0‘..1 /"7.
‘t x ‘1 ' I 1 "Q i '
, :4“ r“, ’ ° m
- . 1‘ ’ " ;“)‘- E ‘
) 1 ‘3’; ‘x x a .
,‘ ~‘~\ ‘ _'.‘ ’ c. /‘
\‘ ’2 ‘ '~ (v ’v 5,, ‘ L‘ 9 ‘
‘\, C 5 , ,
, k ’K- r‘
. - _ .- s A ~ 3’.
.g-r ex; / k
‘X ' (

v, 7' 1 ’p ’ 5L fl «1’
mil/4 ’1 vhf/07 ”$3“
/I I 1 '

 

 

 

 

 

) ',

1 4 fl‘ \ \

.. w «m.

r . ‘;n l—"“ \ r‘\.

‘ , - U m , ‘ \ r
. I 1" \ a
(”x A A ‘~'\ \3 .
. t )K- “i “
\' I- "\-'\

‘ \u \ " I \

‘ C l‘ ‘

i.’\ V. \ [I ‘\\-\

‘ /" H ‘ ‘c

'
,r \ \ ..
2500X

Figure 12- A 9' x 7' plate of specimen A93.

-24-

 

 

 

 

 

50001

Figure 13-- A 9" x 7" plate of specimen

A-3.

-25-

14(a) soox

n.a..r a:

.2 e

. . c... .. .0: . .J
.. .. .r‘ u: fit.-

0

n 01-. .
cm. s.

. mWWVrML .

 

 

X
0
0
5
2

14(b)

; as received.

Figure 14-- structure of 1090 steel

15(a) 800x

 

15(b ) 25001

 

Figure 15-- Structure of 1090 steel; as received.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

I q 4 .1 I I- - t _ a .IIII
, . l _ H _ .
. h .- . . . _ n ..
- .V _ _ n _ ~- »
I I - I4 ...... I . I ..... III I e . II I I I1 I- c e I 4 WI v I I I I ......... I. . I . I e I II; e I II I 9 t I II.
_ u " fl . H - V + . _
fl _ . ¢ - . _ . a y _
- . _ . . ~ _ - , ‘
. I m _ - V " fl W
VI I III. [Ii + 4 I 4
_ ,- V - . _ _ .
- m - fl - -
- . _ . a -
l .- _ f .- . - _ _
........ e 4 e Iv . 4 I . a . . .. .~ I; Io 4 Av II e Ii II II ..I* . 4 I . I I .e I I ... 4
- . . - . a . .
- - - fi . - - . fl .
. _ _
._ n w H h . 4 . - _ _
. . . _ . - , n
L . I I k
a. - . 4 H I N H H _ . ‘1 I
- . _ - . .. . a _ _
. - - - l . . -
V - . ~ ~ . . . _
. I ..... i I n ..... a I ..... I I k I . I I“ III e I I I . - I I. r I I IT PI p I o . I I . I
. - _
_ - _ I .- . u , H -
m, .- _ ,_. 4 H
TIII I I I I II. I n I II ,. I, . I III; II III.‘ I III 4 I II IIIIO IIT I'II. .OIIIIII.I .AI I [LTIIIII I PleIIIIIl . 7.1], IIQII.IIIII fiIII III JIIIIIOI’wIlI {oz-Ii .II. VI rII‘IIIIJIIIAIllI'. I III! -1 .IIIIIII III III+_II I IIIé'IIiII IIIII‘III IIIII ._flII .I '1 91.111.11.‘ IIIIIIIJI'II
. _ _ , ,. . _ 4 _ M a
_ I , _ M . I ,4 .
.w ._. I l - I H .I.IM .I.. V. In: I I . I
- m p n I l _ H A W
. . . . - - . -
_ . . a o _ .
I I I ll w .0 . In I
- w - a i n a _ , .-
. 5 _ x l .- - , e .
_ . . I
_ ._ w - w I m u _
. . ..... I ...... . . ..... I I. I I III III II I .,
- _ - m u “1 H
I - I .
_, W M V . - l 0 V l.
TYOII II’IIII‘IIHI'IIIIVII4III mYII ‘1 I III! .I'I.l9IO.tII|I .Ia IIIIII IIIIIIIVII AI 4“ I'4III I IIIL? I, IIIIJIIII III A II'IIII OII IJIV. ‘I III e I I I I IIIIIII I I‘IIIIIIIIJIIe, IIAY 'IiIIIIII-I‘I.IILIIII714IIIIII
, , . . .
_ _ h I . H p .,
_ M - V m H H _ B l
i, w .1... H... ..I I. .I. II I.HH. loaIOI .IrI I I ..I v ....... -..
- a w _ . .
_ . I .
e m M I w 6 an m
I. -I I. , _I _ 5 v . k
. - - a V ’ -‘ ~
. . - x I
. _ - n . e _
I V .. F. I a v II; I . . III A I III I I II I h I
u fl _ - .N m
_ . _- ,. W _ . U I m
. . - . _ . ,-
vI I ‘I I. II”? III +9 I II I L r II.II4I|IIIIFII I Ilil.‘ II Av II III] IlII‘Ie- 'IIIII‘IIIIII. v.. .III I III III e III: I I . I .1 fl! #ill‘cvllu‘ rfitlIIlIllId.IIIIIIIIIOII .lIllleI.|I .I.
2 a. m u I H . .C n
. . a . . A . a
u v _ 4 . _
. -I . . . . A .. -I .IIIweIie I1? .I I LI I
W m DI m -
. - I . I . -
H w ll. ”I.“ J M-
" _ w . H .7. “8 MW l
. - 4 . -
_ .+ I h ..... H 7, --..tI;TII.I .M ........
. , _ - ..
m x H V H H a M .
m , , . . .Q - .
v I I I .v IIII i I .III IIIIII IIIII¢I.!III.I I I III. II +I I IIeI. II IIeIIIIlIIIII IIIIIIIIII ILIII III IIIIIIIII 7|. IIIIIIIII] III. I I I II I I. I IIIIPIIIOllullI.e . lLl,IIIIII IIII
. a n - «A _
. - . h
, _ l - . _ D _
I I H . o I I o ‘- II I III I . .I I. I P . HI -
. V - _ . . .
_ A . . _ . - l
_ — — -. .
. I _ . H _ a u I
w I, I ,I I. . _II I IIIILI I ,, 6 I-
_ . x . . c ._ _ .
_ - . ._ _ . m 1 .
_ - . . .
4 I “- I e I a I I I I . ........ F I v I f I IlIelI + .. I I . IIII . I I H I
m . I ” . n v N m , . e
- . - . . . -
, H T ,_ H u . r
I. III. . I I IIIII I III I II . . IIIIII..IIIIIII IIIIIIIIII III . III II II II. I I IIT III .I. , III .4 III III IIIIII IJIIII .IIIIIIII I v I IIIIIIIIIII> IIIIII I III I r I IIl.¢IIIIII|r.II IIIIIII. Ifrl-YIIIIIIIIIII .I.+.IIIIIIII|IIIIIII IIII.IIIIIII.[I IIflI If- I . II IAVIIIIIII A A l
. m _ . u H use
V . . _ i
. — d J o , n
. ..... I m o ....... . a e I l YI I I Io . I I7 I III e I I.
u . _ . 4 - m .
- . - .
l _ _ M . N _
- . a . -
I. ‘1 g I. . L I a .I _ w I
_ .I. a A. .
. I -
H _ n w . n I! I M H . II M _
- _ . - - - q .
. v ...... II o I II. I I Ii I If 4 . . I
H H - I
u i . . snowman: _ _ _ I _
- _ - I - - - l __
- - _ .
v IIIII II .III II Ill IIIIIIIII IIIIOIIJIII IAVIIIIIII III- IIIIIL III III .I ' 4|I IIIIIIIIIIIlTIIII IYIIIII. J- 7! II + I II I,Ii.ILII I II IIIquIIIIvII.iIII.I.I.III .ZIIIT. VIII I I'I - IIII .IIlI JIIIII|III.IlILrIIIIIIIIIIII‘IIHI.IIIIIII IIII
. _ . - . . n -
n , I H V k _ _ H H
- .- ,- . - . - - _.
..... I . - .. - . I v . .I I + I I F I - I . I . .
l _ , _ -. l - 1
- , . , _ h .. H
i V _ ,

 

 

17(a) aoox

 

17(b) zsoox

 

Figure 17-- Photomicrographs of specimen A-4.
Rockwell Hardness--660

-29-
tits particles might indicate that the sample was slight-
ly below the Ac} line at the moment of quenching. Figure
16 is the dilatation curve for this specimen. The presence
of cementite particles may also mean that not enough
time was allowed in heating to take all of the carbides
into solution. The microstructures for this specimen are
shown in figure l7(a) and (b). is received photomicro-
graphs for the sample are shown in figure 14.

The microstructures of Specimens A95 and Ash are
very similar although they were quenched from definitely
different places along the dilatation curve. As figure
18 shows, specimen A95 was quenched when the curve ind-
icated that it was just entering the critical range,
whereas specimen Ar6, from figure 20, was almost
through the critical range when quenched. The pearlite
in both Specimens shows patches of a sorbitic appear-
ance. It is probable that these specimens, OSpecially
A96, were allowed to cool too long in air during the
quenching procedure, and hence do not indicate the
true structures at the points indicated by the curves.

As received photographs for A95 and Ar6 are shown in
figure 15.
All of the work described so far indicates that

quenching a steel of near-eutectoid composition from a

-30-

dilatation curve is a "hit or miss“ proposition. A dil-
atation curve will indicate approximately where the
critical range starts and ends. However quenching a
sample so as to show the true structure at any point on
the curve is practically impossible with the Chevenard
instrument because of the narrow temperature range and
definite cooling of the specimen in air as it is lifted
out to the quenching bath.

It was decided to run several more specimens of
1090 steel on the dilatometer in an attempt to stOp the
austenite-a>pearlite transformation just at its begin;
ning. Sauveur‘ makes the statement that ”when steel
containing 0.87% carbon is slowly cooled through its
critical range (Arl line) at some 760 deg. 0., it is
believed by many authorities that the gamma-alpha trans-
formation of the solvent starts at the boundaries and
along some of the crystallographic planes of the poly-
hedral austenite grains. Were it possible to retain the
steel but partially transformed into pearlite, a Wid-
manstatten structure would be obtained with pearlite
(the alpha phase) at the crystallographic planes em-
bedded in undecomposed austenite."

The first Specimen, A97, was quenched when the

dilatation curve (figure 22) indicated that it had

 

l9(a) 800x

 

19(b) 2500K

 

Figure 19» Photomicrographs of specimen A-S.
Rockwell Hardness-~94B

 

 

21(a) soox

 

21(b) zsoox

 

Figure 21-- Photomicrographs of specimen A-6.
Rockwell Hardness--9SB

 

 

‘
§

smith

‘0
Qu

23—min“

-

T

-37-
cooled just to the Ar3 line. The microstructures of fig-
ure 23 show some small,white particles of cementite sep-
arated out. These carbide particles were probably never
in solution because of the low temperature and short
time in the austenite range. The martensitic matrix
shows that the austenite to pearlite transformation (Arl
line or range) had not begun.

The next specimen, A98, was quenched when it had
cooled further down into the critical range, as shown by
figure 24. The microstructures for this sample (figure 25)
are almost identical with those for Ar2 (figure 7). The
transformation of austenite to pearlite was about half
completed when the specimen was quenched.

The microstrutures for specimen A~9 (figure 27)
show the transformation to pearlite to be practically
complete, although judging by the curves of figures 24
and 26, the transformation should not be as far along
as that in A—B. This again is probably due to the error
caused by fumbling on the part of the Operator in per-
forming the quenching Operation, which, in turn, results
in the specimen being cooled too much in air.

The last attempt, Specimen A910, resulted in the

‘transformation being halted at the point desired. The

etched microstructure consisted almost entirely of a

~38-
martensitic matrix with small patches of troostite and
pearlite scattered through it. Figure 29 shows a view of
a typical portion of this microstructure. A patch of im-
perfectly formed pearlite witha border of troostite is
shown at various magnifications in figures 29(a) and (b)
and figure 30(a). Figure 30(b) is another photomicrograph
showing a mass consisting entirely of troostite separating
out of the martensite which was austenite before the spec-
imen was quenched.

It seems apparent that the transformation of aust-
enite goes either to martensite to troostite to pearlite,-
or from rIausmeni’te‘: directly to troostite and then to pear-
lite. The transformation seems to start? at nuclei or
cooling centers, and to spread out from there radially
in the irregular patches shown. There seems to be no

evidence of a Widmanstatten (needle-like) structure

being formed by the pearlite and troostite in the marten-
sitic matrix.

According to Honda5 the A1 transformation takes
place in two stepped changes. Gamma iron with carbon in
solution changes to alpha iron with carbon in solution
and the latter then changes into an aggregate of alpha

iron and cementite:

xustenite-—+»uartensite-e»Troostite, sorbite or
pearlite

 

25(a) soox

 

25(b) esoox

 

Figure 25-- Photomicrographs of specimen A-8.

..
.

II .I
..

 

-4 2-

 

 

I
O
O
5
2

27(b)

9.

Figure 27—- Photomicrographs of spec-imen A-

 

29(a)

 

‘ . r I -
a‘civ-I“%‘\ _‘ '

.I ‘ " . - .
. '.'. ‘4‘,’ ’4 \)‘\;,i'

\

29(b) coo-1.;

5001

 

Figure 29-- Photomicrographs of specimen A-lO.

2500!

\I e
a

II\

0

3

 

 

Figure 30-- Photomicrographs of specimen A-lO.

-45-

In the case of slow cooling, the steps above take
place one after the other immediately, so that the result
is: ‘

Austenite————*Pearlite

In the case of rapid cooling, the first step
(Austenite-—a>lartensite) is greatly retarded and begins
to take place at 200 to 300 deg. o. By the time the first
change is completed, the specimen is at room temperature,
and hence the second change or step (martensite—aspear-
lite) is arrested because this change involves the diff?
usion of carbon atoms through iron, and this is arrested
by the great viscosity of the specimen at low temper-
ature. As a result martensite is obtained at room temp-
erature.

Pearlite, sorbite and troostite, from the phys-
ical point of View, are all the same mixture of phase,
consisting of a mixture of ferrite and cementite, the
difference being the degree of fineness of cementite
particles or lamellae.

Dr. Sauveur, however, eXpresses the View that
martensite, troostite and sorbite are aggregates repre-
senting different degrees of agglomeration of carbide
particlesG.

Many authorities hold the Opinion that marten-

site is a solid solution of carbon or of the carbide in

-47-

alpha iron . Troostite and sorbite are fine aggregates

7
of the carbon and alpha iron (ferrite), and pearlite,

an aggregate of the same constituents in the form of
well-defined lamellae.

If a steel of eutectoid composition (0.87% car-
bon) is heated above its critical range and plunged into
cold water, austenite is converted to martensite. Accord-
ing to Sauveur, the presence in alpha iron of 0.87% car-
bon in solid solution would result in so great a degree
of supersaturation as to be inconceivable that the sol-
ution could survives.

He believes that having created by quenching so
excessive a degree of supersaturation, precipitation of
the solute in particles of submicroscopic dimensions
would take place at once in the quenching bath. Marten-
site, therefore, is a completely, or nearly completely,
aged constituent consisting of submicrosc0pic particles
dispersed in an alpha iron matrix (ferrite). This is in
line with the precipitation, or age-hardening theory,
and eXplains the great hardness of martensite.

The structuresgenerally known as troostite and

sorbite represent, according to thiS‘view, various

stages of agglomeration of the carbide particles. As

~48-
the particles increase in size, the hardness and strength
decrease while the ductility increases.

No matter which view of the nature of martensite
is accepted, it seems that according to most authorities
austenite transforms first to martensite, on slow cool-
ing, and then to troostite and finally to pearlite.

i'fi-ii'li-l».
—-SAE 1060 steels--

Specimens B-l, B~2 and 3-3 are samples of 1060
steel which were quenched from different points as they
cooled through the critical range. 3-1 (See figure 31)
was quenched at about the Ar} line. It shows a marten-
sitic structure in the photomicrographs of figure 32,
although some ferrite has begun to separate out.

Specimen B-2 (See figure 33) was quenched about
halfway through the critical range according to the dil-
atation curve. The photomicrographs (figure 34) show
some troostite and pearlite indicating that the Arl
transformation had begun at the moment of quenching.

Specimen 3-3 was quenched when Just past the Ari
line and shows a structure (figures 36 and 37) of ferrite
and pearlite, as was eXpected.

Two samples of 1050 steel (B-4 and B~3) were

next quenched from a heating curve. Specimen 3-4 was

-49-
quenched when it was heated to Just above the A03 line.
The structure should have been martensitic, but a nar-
row darkened band was noticed across the polished and
etched surface of the specimen. Upon examination with
the microscOpe, this turned out to be a zone of the
specimen which had evidently cooled at a slower rate
than the rest of the Specimen during the quenching op-
eration.

The microstructures shown in figure 39 show the
region along the edge of the dark, quick-etching strip.
The light-colored martensitic matrix can be seen merging
into the darker area, which is made up of dark needles.
Figure 40(a) is a photomicrograph of the martensitic
portion of the specimen, while 40(b) is a photograph on
the center of the dark area, which probably is acicular
troostite.

All other specimens obtained in this investiga-
tion were mounted in bakelite, an Operation which am-
ounts to heating the specimen up to about 130 deg. C.
for a period of about ten or fifteen minutes. The bake-
lite mould was used to facilitate the polishing of the
specimens for microscOpic examination. This particular

specimen, 3-4, was polished and photographed without

-50-
mounting in bakelite by the use of a special clamping
holder. Figures 39 and 40 are photomicrographs of the
unmounted specimen.

Specimen B-4 was next mounted in bakelite, re-
polished and etched, and photographed again. The photo-
micrographs of the mounted specimen are shown in figure
41. It was noted that the dark colored band disappeared
leaving an even, etched surface. A comparison of the
microstructures in figures 39 and 40, and 41 seems to
show definite differences. It might be said that the deg-
ree of agglomeration of the carbide particles has been
carried a little further in the mounted and heated Spec-
imen giving a more definite structure. However it is

more likely that the martensite c£:ths of the mounted
specimen has been etched deeper than that shown in fig-
ure 40(a) so that actually there is very little diff-
erence. More investigation should be carried out along
this line.

Specimen 3-5 was quenched when it hadheated
Just to the Acl line as shown in figure 42. The micro-
structure in figure 43 turned out to be pearlitic as

BIPOCtheee

*******

 

32(a) soox

 

32(b) 25001

 

Figure 32-- Photomicrographs of specimen B-l.
Rockwell Hardness--550

 

34(a) soox

 

34(b) 25001:

 

Figure 34-- Photomicrographs of specimen 3-2.
Rockwell Hardness-470

 

-55-

8001

)
a
II\
6
1)

 

 

X
0
O
5
2

36(b)

Figure 36-- Photomicrographs of specimen 3-3.

Rockwell Hardness-~13C

 

\ t

q.

,\
V‘-

w.
.. 9,.

 

 

 

2 500x

 

Figure 37-- A 9" x 7" plate of specimen 13-3.

 

39(a) 8001

 

39(b) 25001

 

Figure 39-- Photomicrographs of specimen 3-4.
Rockwell Hardness-~64C

40(a) 25001

 

40(b) 25001

 

Figure 40-- Photomicrographs of specimen 3-4.

41(a) 8001

 

41(b) 25001:

 

Figure 41-- Photomicrographs of specimen 3-4 after
it was mounted in bakelite.

 

)

 

 

"“‘T’M‘

43(a) soox

 

43(b) 25001

 

   
  

    

Neg?)

    

‘ w
— ‘. 5
h.,;

 
 

Figure 43-- Photomicrographs of specimen 3-5.
Rockwell Hardness--87B

-54-
--SAE 1045 Steels--

A series of SAE 1045 steels were next run off.
The results are very similar to those obtained with the
1060 steels. Figures 44, 46 and 48 are the dilatation
curves for samples 0—1, 0-2 and 0-3 which were quenched
from a cooling curve; 0-1 from about the Ar} line, 0-2
about halfway through the range, and 0-3 from just above
the Ari line. The microstructure of 0-1 (figure 45)
shows martensite; that of 0-2 (figure 47) shows marten-
site and ferrite (white patches) and 0-3 (figure 49)
shows ferrite with troostite breaking up into pearlite.
There is, of course, more ferrite in the 0 specimens
than the B specimens, because of the lower carbon con-
tent of the 0 specimens.

Figures 50 and 52 are dilatation curves of
specimens 0-4 and 0-5 respectively. 0-4 was quenched
*when heated just above the Ac} line, and the micro-
structure in figure 51 shows small patches of ferrite
Iin a martensitic matrix. This indicates that the spec-
:1men was slightly below the Ac} line at the moment of
tluenching. Specimen 0-5 was quenched just above the Acl
Iline, and shows ferrite and martensite in its micro-

structure (Figure 53)...

{iii-I

 

-66-

 

 

X
0
0
5
2

45(b)

Figure 45-- Photomicrographs of specimen 0-1.

Rockwell Hardness--630

u - - m
. . _ .
. h u . a

v

 

 

47(a) aoox

 

47(b) 25001

 

Figure 47-- Photomicrographs of specimen 0-2.
Rockwell Hardness-—460

 

49(a) soox

 

49(b) 2500!

 

Figure 49-- Photomicrographs of specimen 0-3.
Rockwell Hardness--BSB

 

51(3) 8001

 

51(b) 25001

Figure 51—- Photomicrographs of specimen 0-4.
Rockwell Hardness-660

 

-
-
-
477
-
.

 

-75-
--SAE 1020 Steels--

The last three samples for this work were sam-
ples of SAE 1620 steel. These were all quenched froma
cooling curve. Specimen D-l was quenched at Just about
the Ar} line, as indicated by figure 54. The micro-
structure (figures 55 and 56), however, shows that the
specimen had cooled below the Ar} line at the moment of
quenching, as patches of ferrite may be seen in the
martensite. It must be remembered that the martensite of
this specimen is of a much lower carbon content than
that of other specimens investigated, as the austenite
only held 0.2% carbon in solution.

Specimens D-2 and D-3 were quenched when they had
cooled a little further into the critical range than D-l.
The microstructures are shown in figures 589 59, 61 and
62 and the dilatation curves in figures 57 and 60. It was
found extremely hard to locate the lower end of the Arl
transformation for this steel by using the dilatometer.
However the Ar} line can apparently be accurately placed.

These Specimens were the last of the experimental

work of this investigation...

***********

 

.4.-

-77-

              

     

:wmw.:
\, f, p . I
1:3. . WNW}

.\ . x .
.3wihv-.
ex, A... A» .m, 1W
.7 use\ ,mmwwt

 

     

 

    

 

I
0
O
5
2

55(b)

Rockwell Hardness-—39C

Figure 55-- Photomicrographs of specimen D-l.

 

 

 

 

2500X

 

Figure 56-- A 9" x 7" plate of specimen D-l.

 

-80-

58(a) 8001

 

58(b) 2500x

 

Figure 58-- Photomicrographs of specimen D—2.
Rockwell Hardness--9SB

-81-

 

O
\

i;

\
¢

.\

\-

 

‘ ‘

 

y“ . r :.
. ‘ i ‘ -.:\\\\\ M ~
, ~ ' ._ \Ns-i} - 3i s.

 

2500X

Figure 59-- A 9' x 7' plate of specimen D-2.

 

 

61(a) soox

 

61(b) 25001

Figure 61-- Photomicrographs of specimen D—3
Rockwell Hardness--97B

 

 

 

 

 

 

Figure 62-- A 9' x 7" plate of specimen D-3.

~85-

--CONCLUSIONS AND SUGGESTIONS FOR FURTHER WORK--

The dilatometer can be used to indicate the be-
ginning and the end of the critical transformation in
steeld quite accurately, as verified from the micro-
structures of samples quenched from this region. The
one exception was the SAE 1020 steel, since it was
found that there was no definite indication of the end
of the Arl transformation for this steel. Further in-
vestigation of this is undoubtedly in order.

The method of quenching used in this investi-
gation is not the best possible. The hot samples were
exposed to the room air for a time of between one and
two seconds while they were being lifted out of the in-
strument to be quenched. The error introduced was espec-
ially noticed when steels of near-eutectoid composition
(1090) were being used. A dilatometer might be designed
and built which would facilitate faster quenching.

All the microstructures observed seemed to show
that austenite does not transform directly to pearlite,
but goes through the transformation stages of martensite
and troostite first. Sorbite may be looked upon as im-

perfectly deve10ped pearlite, or an advanced stage of

troostite. The Widmanstatten structure of Sauveur4 was

-86-
not observed in any of the transformation structures.

If this work was repeated with all the specimens
being run at the same standard length (55 mm), the per-
manent length changes in the quenched specimens could be
observed and compared. By the use of special clamps, the
Specimens could be polished and photographed without
first mounting in bakelite, so that the true, untempered

martensitic structures could be observed...

*‘Ii-i-i-X'i'i

5--

-87-
*BIBLIOGRAPHY*

Elements of Ferrous Metallurgy ‘9 Page 133
Rosenholtz and Oesterle 2nd Edition

Metals Handbook--l939 Edition
Definitions of Heat Treating Terms

Dilatometric Analysis of Steel and Some Results

of Dilatometric Heat Treatment.

R.W.Woodward and S.P.R0ckwell Vol 13 My,l928
American Society for Steel Treating. Transactions

Metallography and Heat Treatment of Iron and Steel

A. Sauveur 4th Edition Page 282
Austenite, Martensite, Pearlite; the Classical
View Kotaro Honda pages 34-35
Metal Progress V01 28 August, 1935
Views on the Microstructure of Steel

A. Sauveur Pages 65-66

Metal Progress Vol 28 Nov., 1935

Same as 4 Page 251

Same as 7 and 4 Page 274

****

A

m

$10033) USE 0‘

 

 

 

 

342 09

|\\\\\|\|\\\|\\

 

828