li*

 

123
637
THS

AUSTENI‘NC GRAIN SIZE IN
MARTENSITIC STEBLS UNDER
POLARlZED LIGHT

Thesis for the Degree of M. S.
MICHIGAN STATE COLLEGE

Ralph Wenzel Rogers, 11'.
I942

THESE.

AUSTENITIC GRAIN SIZE IN MARTENSITIC
STEELS UNDER PJLARIZED LIGHT

by
RALPH WENZEL ROGERS, JR.
M

A THESIS
Submitted to the Graduate School of Michigan
State College of Agriculture and Applied
Science in partial fulfilment of the
requirements for the degree of

MASTER OF SCIENCE

Department of
Chemical and Metallurgical Engineering
1942

(145.9425

 

 

 

 

Ac knowledgme nts

The author wishes to express his gratitude for the
encouragement and advice given by Professor R. L. Sweet,
under whose guidance this work was accomplished, and Mr.

D. D. MCGrady.

1.12680

Table 9: Contents

ForewordCOOOOOOOO.C...00......O00.....‘OOOOOOOOCOOOOOOOi

Austenitic Grain Size in Martensitic Steels Under
POlarized LigkétOOOOOOOOOOOOOOO'.00.01.000.000.0.0.0.0000].

Methods for Revealing Martensite Grains...........5
Austenite Grains and Martensite Grains...........16
Summary..........................................28
Appendix: Electrolytic Polishing of Steel.............30
Apparatus........................................33
Experiments on Electrolytic Polishing............36
Summary and Conclusions..........................58
Suggestions for Further Investigation............62

Bibliogx‘apllyCCOO0.000000000000000000.00.00.0000000000064

E0: gwgd

The original purpose of this investigation was
to explore the possibilities of using polarized light
to reveal the previous austenite grains in fully
hardened steels.

' After performing a number of experiments toward
this end by mechanically polishing steel specimens,
it was decided to use some electrolytically polished
specimens to secure an.undistorted scratch-free
surface.

Because of the incompleteness of the literature
descriptive of electrolytic polishing of steels and
other metals, it was necessary to investigate and‘
experiment rather thoroughly before satisfactory
polishing resulting.

In the end, much of the time devoted to this
thesis was spent in the latter phase or the invest-
igation; hence the appendix, which contains the
theory, data, and results of the experiments of
electrolytic polishing, constitutes a not incons
siderable part of the thesis.

Apstenitic grain §ize In martensitic Steels Under
Eolarized Light

Throughout the many years of investigation of the
metallography of steel, a problem about which there
still remains considerable doubt is that of the nature
of martensite and the reason for its hardness. Be-
cause of the acicular or needle-like structure of
martensite and its apparent lack of grain structure,
many theories have been evolved on the basis of hypo-
thetical structure in contradiction to the prevalent
one that hardening is caused primarily by lattice
distortion in the super saturated solid solution of
carbon incxbiron, caused by rapid cooling when
austenite is severely quenched. For example, Jefferies
and Archerl formulated their "slip interference"
theory on the basis of the formation of extremely
small crystals (in martensite) oftxhirons-the hardness
being caused entirely by the sub-microscopic size of
the grains and their resulting resistance to slip.

Heindlhofer and Being, however, by the application
and careful interpretation of x-ray analysis, showed
that the crystallograms of martensite are indicative
of comparatively large,grains. In.addition.to the
xpray, these investigators used polarized light to
show that grains or "blocks" of martensite do exist
and are of a size corresponding to that of the

parent austenite grains.

 

1‘

(2)

It is the purpose of this paper, then, to estab-
lish methods for best revealing the martensite grains
clearly, to investigate further the relationship of
austenite grains and martensite_grains over a range
of sizes, and to establish the practicability of using
polarized light to determine the grains size of the
parent austenite, since the latter is of prime import-
ance in establishing the prOperties of a steel.

Polarized light is made up of waveswhose vibrations
are in one plane only as opposed to ordinary light in
which the waves vibrate in all directions. Its use in
the examination of metals depends on the ability to
distinguish between isotropic and anisotropic crystals

under polarized light.
0516C} Diane.

 

 

 

 

 

 

 

 

 

 

 

 

 

' Opiecfve.
r
A 1
F ‘ ---—-——— V Prism has?
' A CW/
e e fete diner
”P at .......

 

 

 

Figure 1. Light Train.For Metallograph Equipped
With Polarized Light '
In the Bausch and Lamb research metallograph
used (Figure l), the light from a tungsten are is
polarized by passing through a Nicol prism in.the

(3)

vertical illuminator. After reflection.from a surface*
which does not change the polarization of the beam, it
passes back into the same prism in such a way that the
beam is totally reflected and the effect of "crossed"
Nicole is complete darkness. If the polarization of
the beam.is changed.hy reflection, the components of
the reflected light perpendicular to the original

bean are transmitted by the prism and viewed in the
eyepiece or camera.

An isotropic crystal is one whose properties are
the same in all directions. Polarized light is unp
changed on reflection from such a surface so that under
"crossed Nicola" total extinction of light occurs in
all positions of the crystal.

An anisotropic crystal has different properties
in different directions and changes the plane of
polarization on reflecting a polarized beam. Hence
when viewed under "crossed Nicole", anisotropic crystals
'will very in intensity from.total extinction to maximum
intensity, depending upon the angle formed between
each crystal's reflecting plane and the plane of polar-
ization of the beam. (

'Austenite, which.has a face-centered cubic structure,
transforms to martensite upon quenching, the latter
belonging to the body-centered cubic type. An inter-
mediate product of tetragonal type lattice, which trans-
forms to the body-centered cubic at low temperatures,

is believed to form first. If the transformation takes

(4)

place along the crystal planes of the austenite, the
' resulting martensite should show a grain structure
similar to the parent austenite.

As the martensite lattice varies distinctly from
an exact body-centered crystal because of the carbon
held in super-saturated solution causing the axes to
have a ratio as high as 1.0S, the martensite grains
show anisotropy. If it is composed of crystals of
relatively parallel planes, martensite will appear
under polarized light to be made up of grains which
vary in intensity according to their orientation.
Moreover, on revolving the martensite specimen, the
intensities will change, each grain.going through two
maxima and two minima in a 360° revolution.

Some of the practical difficulties associated
‘with polarized light make its use limited.to a greater
extent than is at first apparent. For example, the
low intensity and low contrast of the reflected light
make photography difficult. Also much that can be seen
visually by rotating the specimen and.changing the
intensities of the grains cannot be retained on a
photomicrograph. The amorphous or "Beilby" layer of
flowed metal formed on polishing is isotropic and must
be removed to form.a reflecting surface of the aniso-
trOpio martensite only. This is usually accomplished
by etching, but the resulting roughening of the sur-
face makes the intensity of the reflected beam con-
siderably less.

(5)
W ESE. hassling ism ens__i_te grains

The technique used to secure a surface for invest-
igation under polarized light is important because of the
rather exacting limits set by its use. various factors
were investigated in order to determine the optimum
conditions for revealing the grain structure of marten-
site under polarized light. Particular attention.was
devoted to the smallsgrained specimens since the large-
grained martensite can be revealed with comparative
ease by any methods satisfactory for the former.

The factors investigated were the methods of re-
moving the amorphous layer of disturbed metal and the
effect of tempering. The methods used to remove the
surface metal were etching and electrolytic polishing.

As mentioned.before, it is necessary to remove the
disturbed layer of metal, which is amorphous and iso-
tropic in nature, in order to obtain reflection from
the anisotropic metal beneath. If this is done by
etching, it is necessary to avoid over-etching since
the already low intensity of reflected polarizedlight
is rendered considerably lower by roughness caused by
too deep an attack.

The ordinary etching solutions used did not seem
to improve the revelation of a small grain size except
for a special reagent for austenite grain boundaries

 

(6)

in martensite. This reagent, proposed by J.R. Vilellaa,
consists of one gram of picric acid in five parts of
hydrochloric acid and 95 parts of ethyl alcohol. In”.
large-grained martensite, the previous austenite grains
show up under ordinary light with this reagent (figure 2),
but small grained specimens show little or nothing.
With small grains under polarized light, fair results '
can'be obtained with this reagent. much.of its advan-
tage lies in its slow rate of attack so that the degree
of etch can be closely controlled and over-etching
avoided.

The use of electrolytic polishing to remove the
surface layer offers the advantages of a strainpfree,
undisturbed surface secured by electropolishing instead
of mechanical_polishing to remove scratches left by
the fine emery paper. Here the danger of roughening
the surface and diffusing the light is again present
and conditions for obtaining a good polish were invest-
igated to considerable length as the literature on
this process is quite incomplete and misleading. The
theory, results and data of this part of the invest-

, igation are contained in an appendix to this thesis.

. In general, as might be expected, it was found
that with electrOpolishing no further etching is

needed. Sometimes, however, a light attack by Vilella's

reagent, 2% picral, or 2% nital was found to increase

(7)

the clarity of the picture. It was found best to ex-
amine the specimen unetched and then to etch if nec-
essary as no definite procedure could be specified.

Figure 3 is a specimen of S.A.E. 1090 steel
quenched from 15500 F., electropolished and not etched,
while Figure 4 is the same steel etched with Vilella's
reagent. Figure 5 is S.A.E. 1090 steel, quenched from
17000 F., electropolished and not etched. Figure 6
is S.A.E. 1090 steel quenched from 1900° F. and etched
with nital.

Figures 7,8,9 and 10 are S.A.E. 1040 steel in.the
annealed, normalized, water-quenched, and oil-quenched
conditions under polarized light after electropolishing
and etching with nital.

The purpose of tempering martensite steels for
observation under polarized light is to complete the
transformation fromttetragonal to more nearly cubic
body centered crystals. Probably because the latter is
in a more strainpfree state, the crystal planes are
more perfect and so the light is reflected in a more
orderly manner with resulting clarity in the pictures.

Figures 11,12,13 are specimens of S.A.E. 1090
steel quenched from 14500 F. Figure 11 shows the
resulting structure while figures 13 and 12 were tem-
pered at 450° F. for 15 minutes and 30 minutes respect-
-ively.

(8)

Figures 14 and 15 are photomicrographs of a
specimen of S.AtE. 1090 steel quenched from 1900° F.,
before and after tempering at 475° F. for 15 minutes.

From these photomicrographs it is seen that
tempering is only a slight aid in improving the clarity
of small grained martensite under polarized light, but
it improves the observation of large grained steel to

a greater degree.

Figure 2
SAE 1090 .
quenched from 1900° F.
etched with Vilella'a
reagent
100x

 

Figure 3
SAE 1090
quenched from 1550° F.
tempered at 450° F.
electrolytically pol-
ished, no etch
100x
polarized light

 

(10)

Figure 4

SAE 1090
quenched from 15500 F.
tempered at 450° F.
electnolytically polisl-
ed and etched with
Vilella's reagent

100x

 

polarized light

Figure 5
quenched from 17000 F.
tempered at 450° F.
electrolytically y01-.

ished, no etch
lth

polarized light

 

(11)

Figure 6
SAE 1090
quenched from 19000 F.
electrolytically pol-
ished and etched with
nital
375x
polarized light

 

Figure 7
SAE 1040
annealed

electrolytically pol-

  
 

ished and etched.with ‘
nital
375x
polarized light

(12)

Figure 8

SAE 1040

normalized

-
l
0
D.
V.
l
l
a
C
m
l
0
r
t
C
e
l
e

ished and etched with

nital
750x

polarized light

 

Figure 9

SAE 1040

water quenched

-
l
0
P
V.
l
l
a
c
.1
n.
l
0
r
t
C
e
l
e

ished and etcned with

Vilella's reagent

100x

 

polarized light

(13)

Figure 10
SAE 1040
oil quenched
electrolytically pol-
ished and etched with
picral
100x
polarized light

 

Figure 11
SAE 1090
quenched from 14500 F.
etched with
Vilella's reagent
100x
polarized light

 

(l4)

2
l
m.
.1
F

 

F
O
O
5
4
l
mu 1
SH
00
lr
f
ma
9
a
n
e
U
0.

tempered for 30 minutes

at 475° F.

etched with

Vilella's reagent

100x

polarized light

Figure 13

SAE 1090
quenched from 14500 F.

S
e
m
n
.1
m
5
l
.r
O
f
d
e
P
e
M;
H
e
t

at 475° F.

etched With

Vilella's reagent

100x

 

polarized light

(15)

Figure 14
SAE 1090
quenched from 19000 F.
etched with
picral and nital
100x
polarized light

 

Figure 15

SAE 1090
quenched from 19000 F.
tempered 15 minutes

at 475° F.

etched with

Vilella's reagent
100x

 

polarized light

(16)

austenitg gagins 5nd gartegsite grgins

There are a number of methods used by the metal-
lurgist to determine the previous austenitic grain size
in.martensitic steels. Among these are the arrested
quench, gradient quench, slow cooling, preferential
oxidation, heat tinting and case carburization methods
and the use of special reagents. All of these are
limited to steels of certain carbon ranges or by the
necessity of using unusual equipment, or because of
. the inordinate length of time required for the test.
martensitic steels of approximately eutectoid compos-
ition are those in.which it is most difficult to es-
tablish the austenitic grain size so that for these
experiments S.A.E. 1090 steel was used. Eutectoid
steel also has the advantage of consisting of onxy one
constituent-martensite- in a quenched structure,aso
that there is no other structure present to complicate
the observations. '

To establish the relationship between austenite
grains and martensite grains, samples of steel were
quenched from 18500 F. and 1450° F. into cold water.
Samples from the same bars were submitted to a gradient
quench from the same temperatures. This consists of
quenching one end of the specimen only--so that at a
certain point along the axis of the specimen, the

cooling rate is such to cause the formation of troostite

(I

~
V v
, .
_ ‘ ‘ y .
.4 no ' . '
.
, . .
a . . u 'e
a . - - .7
' A
I A
a - x . . . o .
.
a
. , .-
.
a
‘ . |
O is v .
' ,.’ ' ~ .

 

(17)

in the grain boundaries of the austenite as rapid
cooling occurs. At room temperature this zone con-
sists of martensite outlined by thin lines of dark-
etching troostite, thus showing the location of the
previous austenite grain boundaries.

Figures 16 and 17 were quenched from 1450° F.,
the former a gradient quench under ordinary light and
the latter an identical steel after regular quenching
and photographed under polarized light.

Figures 18 and 19 represent a similar treatment
from.1850° F., the former being the gradient quench.

From these comparisons, it is readily seen that
the grains in martensite under polarized light are of
the same order of magnitude as the parent austenite
grains. Also, the difference in clarity of the photo-
micrographs by both methods between the large and
small grains is illustrated. The temperaturesof
1450° F. and 1850° F. were chosen to secure austenite
grains before and after the coarsening temperature of
this steel was reached, and so a considerable diff-
erence in the austenitic grain size obtains upon quenching.

Figures 20-27 together with Figures 17 and 19
show the structure of martensite under polarized light
secured by quenching steel from temperatures which
varied by increments of 50° F. between 1450° F. and
1900° F. From a comparison of these pictures, it is

 

illlllllll||l|l|lll1llll|lllvll

(18)

soon that the specimens quenched from 1700° F. and
below are small grained and lacking in distinctness,
while those quenched from above this temperature are
both large grained and well defined. The border case,
Figure 24, shows evidence of duplex structure but lacks
the distinctness of Figures 25 and 26. The coarsening
temperature of this steel may be established at about
1750° F. by these photomicrographs. That more than
merely increased.grain.growth occurs at this temper-
ature seems likely since such a considerable difference
in appearance and distinctness results from quenching
from above and below this temperature.

The general appearance of the grain revealed by
polarized light is typical of a face-centered crystal
such as austenite-~the grain being equi~axed and roughly
hexagonal, especially in the large—grained specimens.
In examining martensite under polarized light, care must
be taken in determining the size of grains. For ex,
ample in Figure 28, several orientations within one
grain.are visible as shown by the same picture under
ordinary light, Figure 29. By rotating the specimens,
the areas of a grain showing different orientations
will display a maxima or minima at approximately or
exactly the same angle. Figure 30 illustrates this,
as it was taken at an angle of 33 degrees from.Figure 5.

This is the same way that "twinned" austenite
grains show different orientations under polarized

light and for the same reason; stresses are relieved

 

 

 

F.

(19)

in a grain by a "twisting" of a part of the structure
so that the orientation varies in different parts of
the original grain. As the light reflected varies in
intensity according to the angle made with the plane
of polarization, the differently oriented parts vary
in intensity. Twinning is especially evident in Figures
27 and 6.

magnifications beyond 100 diameters showed no add-
itional detail under polarized light and are at a dis-
advantage as the number ef grains under observation is
correspondingly smaller. Since the percentage of well-
outlined grains is low, to determine the size of the
grains, as large a field as is possible should be used.
Figures 6, 7, and 8 and Figure 31, which is of the same
area shown in Figure 24, were taken at magnifications
other than 100 diameters.

 

II‘AI

(20)

Figure 16

SAE 1080

m
0
n
h
c
n
m
0..
m
e
.1
d
a
m.

1450° F.

100x

 

Figure 17

SAE 1090

r..
0
no
%
l
m
m
pl
.a
e
a
n
e
u
o.

tempered at 4000 F.

100x

 

polarized light

(21)

Figure 18

SAE 1090

m
0
r
f
.L
C
n
m
q
t
n
e
.1
d
m

n“
O
o
%
1

 

x
me
C
1

Op
1
M»
.1.
F

O
.(C
C
l

.n
o
m
l
m
o
r
f
d
e
a.
n
e
u
C.

tempered at 4000 F.

100x
polarized light

 

(22)

Figure 20

SAE 1090

F
O
0
w
l
m
0
P
f
d
6.
qr.
C
n
e
u
0..

tempered at 4500 F.

100x
polarized light

 

1..
2
m
.1
F

n“

O
0
a
l

0.
am
if
Md
e
a
n
e
u
0.

tempered at 450° F.

’100x
polarized light

 

(23)

Figure 22
SAE 1090
quenched froml600o F.
tempered at 4500 F.
100x
polarized light

 

Figure 23
SAE 1090
quenched from 16-500 F.
tempered at 450° F.
100x
polarized light

 

(24)

Figure 24
SAE 1090
quenched from 1700° F.
tempered at 450° F.
100x
polarized light

 

Figure 25
SAE 1090
quenched from 1750° F.
tempered at 450° F.
100x
polarized light

 

(25)

Figure 26
SAE 1090
quenched from 1800°F.
tempered at 450° F.
100x
polarized light

 

Figure 27
SAE~1090
quenched from 1900° F.
tempered at 450° F.
100x
polarized light

 

 

 

(26)

Figure 28
SAE 1090
Quenched from 1900° F.
etched with
Vilella's reagent
100x
polarized light

 

Figure 29
SAE 1090
quenched from 1900° F
etched'with
Vilella‘s reagent
100x

 

(27)

Figure 30
SAE 1090
quenched from 1900° F.
etched with
Vilella's reagent
100x
polarized light

 

Figure 31
SAE 1090
quenched from 1700° F.

electrolytically
polished

500x
polarized light

 

(28)

§EEE§£I

From the series of pictures taken under polarized
light, it may be seen that the size of grains revealed
is of the same order of magnitude as the parent aus-
tenite. According to Heindlhofer and Bainz, these are
actual grains of martensite, and not merely traces of
the austenite.

For the purposes of making an estimate of the grain
size of the parent austenite and for a determination of
the coarsening temperature of a steel, polarized light
can be used successfully. For steels near the eutec-
toid composition, this method is especially valuable
and is the only simple method available. In other
steels, it offers advantages in time saving as no slow
annealing is required; quenching and a fifteen minute
tempering is the only heat treatment required. In
steels which are already hardened, this is the only
method for ascertaining the grain size and structure,
except for special martensitic etches, the successful

use of which is frequently impossible.

For further investigation, it is suggested that
the alloy and plain carbon steels over a range of
compositions be studied to determine the limits of
polarized light for as the composition of martensite
approaches alpha-ferrite, the anisotropy disappears

 

.oe.‘

(29)

at some point. This was borne out by attempts to
observe 1010 steel under polarized light which were

unsuccessful.

Apps ndix

(30)
APPENDIX
Electrolytic Eolishigg 9; Steel

 

The electrolytic polishing of metals was first
attempted by Jacquet who successfully polished coppers’5
and tin7L by this method. Briefly, it consists of the
reverse of e1ectro~p1ating; the specimen is placed at
the anode of an electro-plating bath and conditions
are so controlled as to secure the removal of a layer
of metal at the anode rather than an even deposit at
the cathode.

Conditions for successful electrolytic polishing
must be controlled very carefully to avoid a rough
surface caused by pitting, etching, or uneven attack.
The variables which have an influence on the quality of
polish are the electrolyte, current density, voltage,
time, agitation, temperature, and the shape, area and
location of the cathode.

These variables have been the subject of consid-
erable investigation and while results have been ob-
tained for many metals, there is still lacking enough
information to be able to duplicate results in most
instances.

It was the object of this investigation to deter-
mine the conditions necessary and to construct the
apparatus with which to electrolytically polish commer-
cial steels quickly and so secure a strainpfree, scratch-

less surface.

(31)

A procedure for systematic investigation was also
devised as a guide for future investigations of this
type of work. i

The most important factor in either polishing
or plating is the proper choice of electrolyte. In-
vestigators agree that a viscous, nonsconducting film
is necessary to establish a concentration.gradient
between.the elevations and depressions of the surface to
be polished. Uhlig4 points out that according to the-
oretical electricity, high gradients of potential exist-
in the vicinity of points of great curvature-- even
though the total gradient be small--hence at the elevations
a better anode efficiency exists and the removal of
Ametal proceeds at‘a greater rate than in the depressions.

This film.man.be an oxide film, a layer of gas,
insoluble products of electrolyte and metal, a static
liquid film. or any combination of these, and it is
especially necessary to secure a good film for the pol-
ishing of nonphomogeneous alloys to avoid uneven attack
on different phasesa.

The choice of electrolyte then should be one in
which.such a film is formed. The formation of metal
salts which are but slightly soluble in the electrolyte
satisfies the requirements. The addition of organic
liquids to reduce the diffusion of the film and to

lower the conductivity of the electrolyte is recommended

by Uhlig4.

(32)

To determine the proper current density and voltage,
and whether the electrolyte will result in satisfactory
polishing, a curve of current versus voltage should be
made, using the metal and electrolyte in question. The
curve for a successful combination will have a shape
similar to one of the following, which were secured by 1

various investigators.

(6
co. A (a) co. A s 2 f

 

 

 

 

(x1 " B as

 

V

Figure 32. Voltage-Current Density Curves

For curves b and c of Figure 32,_the range for
successful polishing occurs in the flat part between
A and B9’l°’°. In the portion of the curve above B,
gas is evolved and an uneven surface results while
below A, etching occurss. For curve a, the critical
point A-is the lower limit at which polishing proceeds
successfully. Below this, the rate of diffusion of
the film exceeds its rate of formation and hence the film
is net maintained4.

Experimental polishes at the voltage values so
determined will ascertain their correctness, and by

varying conditions such as temperature, cathode distance,

(33)

and the like, the optimum conditions for polishing
the metal can be determined.

W

The apparatus used for electrolytic polishing
was based upon the use of 220 volt direct current
which was available in the laboratory. Since the
voltages to be applied were not that high, the fol-

lowing setup was used to secure the desired voltage

and current.

 

 

 

 

 

R2.

 

 

 

 

 

 

 

 

 

 

Figure 330.. Electrical Circuit for Electrolytic

Polishing
The switch 81 permits the use of high or low
voltage while 32 throws in either the ammeter or

: uw‘

(34)

milliammeter. The rheostats R1 and.R2 control the current,
the latter being used for close control.

The cell consisted of a battery jar, usually set
in cooling bathe“ of ice water, with a motor driven
stirrer. The cathode was placed parallel to the bot-
tom while the specimen was clamped to the holder.

This holder was built from an old microscope frame

and makes possible close regulation of the depth of
the specimen in the electrolyte, and easy removal of
the specimen from both, because of a hinged frame.
Figure 33bis a photograph of the cell, including the ;
holder and stirrer, but without the external cooling
bath. Figure 33cshows the control panel as diagrammed
in Figure 33a.

Airflfln—‘u

‘ . h—
, 1:th _

 

 

 

 

 

 

 

 

 

 

 

Figure 33c. Control Board

(36)
Experiments 9. Electrolytic Polishing

The experiments on electrolytic polishing of steel
were primarily concerned with S.A.E. 1090 steel since
that is what is used throughout the polarized light
exPeriments. Other steels were also polished electro-
lytically to verify the conclusions reached with 1090.
These included S.A.E. 1040, 1010, and 10120.

For the first attempts to polish electrolytically,
a storage battery was used for the current source.
After a number of trials, the voltage secured proved to
be too low to polish steel so that the 220 volt D.C.
setup previously described was devised and proved
successful.

The electrolyte used in all of the experiments
was one recommended for steel by a number of writers
and consisted of 765 parts of acetic anhydride, 185
parts of perchloric acid, and 50 parts of water. The
film formed in this electrolyte with steel is said to
be [FeaAc6 (OH);ICIO4°4H20 by Jacquet and Rocquetll.
The solution is made up by chilling the chemicals
before and during mixing, as considerable heat is
evolved and the danger of a violent reaction is present.
The solution is allowed to stand for 24 hours before
using.

Most reports on electrolytic polishing emphasize

'the current density as the prime factor in securing a

 

(37)

successful polish, so that in the first of these ex-
periments the current density was the primary subject
of investigation. Hewever, this did not prove to be
very successful in controlling the quality of polish
because of the difficulty encountered in keeping a
constant area in contact with the electrolyte. As.
the apparatus was designed to be a simple and rapid
means of polishing metals, the sample to be polished
was simply suspended in the electrolyte, the total
area immersed being subject to polishing action. Hence,
when the solution was agitated, the wave motion caused
a variation in area. Also, the depth of immersion
was difficult to ascertain because of the low surface
tension of the solution showing in the high degree of
adhesion to the metal. Coating the sides of the spec-
imen with paraffin.was tried in order to maintain a
constant area but the paraffin was attacked by the
electrolyte and failed to protect the specimen. For
these reasons, control by current density was not con-
sistent and results were not always successful.
Controlling the polishing by means of the applied
voltage was next attempted. Since the voltage does
not vary with the area in contact with the electrolyte,
the above difficulties were eliminated. To determine
‘what voltage values to use, the voltage-current curves

*were secured for each metal. The technique employed

{1+
I v. Ll_

LOI‘

11'.

‘lli

 

'.I" M! l‘l‘ III-“I '1‘.

'l'vl 'L “al I I

LHbll Y l"\l\.'l\§ I \Inl..a

 

 

finelII o IIIIIEIOIIIMIII. — fl HI vIel II
. . I .
u l . . l u N
—. n f . . . . * o u n .* o H w r
. . . . . _
— . a a . q .
a a n . a
. o . .
"IIIIIOI iIIJ-.IIIO OI+IIIIIIVI OI AIIIIIIIIOIOI IIIIII OeI III
— . .
m a . _ _ u
_ . . _ . ._
u e o e . u o ,
— _ _ . u e
- .
_ _ ” u
l 31.. _ n + :1 1%:
. . _
. * ._ n _
a g o .
-. I . f .. 3
u . o .
. _ . . If
_ o .
. -
IIeIIIIO OIIIOoIII fIO'IIIIo-IIIII -IIITIIOII o-II I. .I-Iv w
u e p o
. _ »
g . n H M V
. . .e a
_ . fl _ _
. _
. x . n _ 01
O
.

—-—'———_‘.—-— ""—"1"'——'_‘_1'— —.—.- 'n.
v
I
I
.f

7

. I
I
. m... ,wr- -_.._-.

O
O
I
I
I
e
— —-—o-—- o»- T—o.c-- ~ - 0-- -

 

I
I

l
I

I..1|.-I.III.II.X!:III
_/
Ty

I45! .7

 

III II

 

!
I

a /

 

 

 

 

 

‘v

I
l
T

 

 

 

 

 

 

 

 

 

.-./’0..

 

 

 

 

 

 

Q

b 35 4'

O

Q o

/5..2b_:25,3

I
0

I

I
9-0-4

I

I

SIRE {RAM-0%? "lOl'Q'Stegl -_-_

0

450

I
m-——
I

:34

O
I
I
I
I
I
I
.- —..—.-oq—. .— .—-——-.
I

33"

I
I
i

I

i

fCurrént'vaIue

I

I

ritiona‘

.—- ‘4». o

iarea

I
I

lto

s are "p'tropo

I
-— - coo—O
l

a..-

- l

-..__. . .-.-,—..
I
I

I
0.4— .5.— -o.---
I

 

 

'
I I a”: I Cl O I I ‘- 0 O I" e- ‘ I
. _ .
. a . c
M . r
l l
H n
n. I ' D I“ I ”I. ‘V I
u
-. I I - . ..... v 1: o I w:
. . V
A a
4 .......... - -q lilflO.
.
C e
e
. . _ _
m i C . . i 5
I .101 :5 I I IO :0! I Iuv I I 0.. Id Ilrt a
. h _ . . 3
_ - . e .
. .
a
g .
- h
. _ .9 .
a , m . no
I e I I If 0 l I I I I. I II. on
H . . . a
_ .
. _ _
a o . . .
_ o . .
u _ I N —n o
n _ 1
In- 0‘ 00 II .90 'I F O I I e IIV.
_ u o t 2
. . o
.
n e o
e o e.
. e . .
o _
w _ AV
0 I. I I» OOOOOOO O In- allllLv. e.
l . o l
_ _ H a
. _ e . .
- .i. .7 ,2 n - - . ,. 1.2.0.-
. a /
a . H
. . w
0
o -
.
.
I I :9 Lbe
. _
:1 L; l . .
_ i H
- o
h M II I III I d o A] I. e‘lll .- ICIIIIIV50I
“ . O —
0 . OI: a a
_ V .. . .
u v .1; . A ..
- , H l... a _ g
4 4 AN 4 1 < . n e.v‘
MW.“ mw no no no no
[T m. I7 2 6 _ v _
IT . .
o . _ .
. . . .
IIIIIII o... I e sous-u O .I e I 04-5 0 n . u .0 .II o.
_ . . .
e u _ s I
u . . .
H a . . . _
_ _. m u M . .
_ .. . .
. . . .

Q .
u

I

res-Current: Cur“ bf - 1°49 " 5‘31““

O

I

3.352;... V0 '5

o w

4

.-.— ——-—-o<o

. .
.’...--.---
I

I .
...-——.o1 —-¢ -‘

 

 

FRG'IY E‘UU":="‘ ""‘3 "' ‘ "' ' - I"‘-'Ivrltrl l v \n lulw'wlrt‘lr.

 

    

 

' -__ v —I_ _- I --— 'w. _ I--' — 1 -—-—-—-- I ~—— ‘ -- - -—x-r - '-- --- ' - - “I. - I - - ' '_. II- — 1 C I U —C -- —' -~- :
Y . 1 I ll 0 I I I I I I :
' ; " ' (39) 3 z t 2 I I 1
I . i ' : ' : i
I . . . I I 1 3
o. I ‘ I 00.-.. v D I‘ ’ - :u -‘--!I. -. I; z
' ' ’ I ' I
I I e I Q
I I I o
I I I l ‘
I I I I I
0- I E- O—o 0 U:
f ; I . a
I I I ' I 3
I : : I I
Gee--- .I- 0—. T- 0| *H u-—:- h 1
l I I I
r' : i ; I
i ' I 3 I
Dan. : -—-0— e— -: ‘O— - l ..... I
I .
I f I

 

 

I
O a
O
- I
I ‘ i I
’ O
' C- ‘- ——-—-*— —J——‘—- .

 

 

 

 

 

Q 55% f0 15' 20 25 so 55 4o vo/rs ‘_ _
Figure 36. VoltageeCurrent Curve p: 1090 Steel : . I i

Current values are preportional to area ? 3
i . . ; ; g
I
/20I ;
Ina} - : i
/00- 3 3
80 f a
' 60 ;
4Qf7
20? , I :

 

 

. I
5 /_0 /5 2'0 25 3o 35 4'0 -_ vo/fs
Figure 3!. VoltagefCurrent Curve sf 10120 steel

 

(40)

was to shut off the stirrer while a reading of the
voltmeter and ammeter was taken, restoring agitation
between readings to approximate conditions maintained
while polishing. Plotting voltage against current, the
curves for the steels studied are drawn in Figures 34-
37. For two determinations, different areas of contact
will manifest themselves in two current values for the
sale voltage, but the breaks in the curve in both cases
will occur at the same voltage. Since these breaks

are what determine the range of polishing, this method
is successful for securing the information desired.

The curves so secured show the typical flat portion
where, according to theory, successful electrolytic
polishing occurs. To test this, photomicrographs were
made of SAE 1040, 1090, 1010, and 10120 steels electro--
polished at the voltages indicated by the curves. These
photomicrographs are those of Figures 38-41.

From these illustrations it can.be seen that
scratches from grinding are removed and a flat even
surface secured. The black spots appearing in all of
these photomicrographs are not pits as would appear at
first observation but are inclusions which appear above
the surface. This canibe shown in two ways under the
:microecope--by the use of polarized light and by care-
fully measuring the height of the inclusions by the

calibrated depth of focus screw of the microscope.

 

 

rfi

(41)

Under polarized light, the inclusions appear bright
against a dark.background as they reflect light which
is unpolarized and therefore which passes the crossed
Nicol prisms, while the light is reflected from the
metal matrix still polarized. This is complicated in
the examination of hardened specimens since the mar-
tensite shows anisotropy and the background metal is
no longer entirely black, but enough difference in ins
tensity remains to distinguish the inclusions.

The inclusions are not necessarily of the size
of the spots, in fact are probably in all cases smaller

than indicated. The inclusions distort the surrounding

surface in such a way that light is reflected away from
the eyepiece, and both inclusion and surrounding sur~
face appear as a black spot. This effect may occur to
a greater or lesser degree, (as will be shown.later)
depending upon the part of the curve to which the Volt-
ageused corresponds. The time of polishing also plays
a part in this effect, for the amount of metal removed
is proportionalto the time of polishing and the height
of the inclusion depends on the metal removed. The
inclusion effect appears especially bad on photomicro-
graphs such as these of the surface alone, but fer the
more widely used cases where the specimen is etched
to study the microstructure, the effect is masked by

the etched surface until almost unnoticeable.

('42 )

Figure 38
SAE 1010
hardened
electrolytically pol-
ished at 18 volts
100x

i...—
- \e-

 

Figure 39
SAE 1040
hardened

electrolytically polished
at 16 volts
100x

 

 

(43)

Figure 40
SAE 1090
hardened
electrolytically pol-
ished at 16 volts
100x

 

Figure 41
SAE 10120
hardened
electrolytically pol-
ished at 12 volts
100x

 

 

(44)

To illustrate the differences in polish obtained
at voltages corresponding to different points along
the curve, photomicrographs were made of SAE 1090 steel
which was polished at 2,5,9,12,l6,20,24,28, and 40 volts.
These correspond to the points indicated in Figure 36
as listed.below.

Figgge Point Voltage
42 A 2
43 B 5
44 C 9
45 D 12
40 E 16
46 F 20
47 G 24
48 H 28
49 I 40

Figures 42 and.43 show the result of polishing at
voltages corresponding to points below the flat portion
of the voltage-current curve where a viscous film is
not maintained and uneven attack.results. Figures 44
and 45 show the inclusion effect to a marhed degree
although the surface is flat and even except for the
inclusions. Upon close examination, the actual inclusion
may be seen in the center of the black area in many
cases. Figure 50 is a photomicrograph at 750 diameters
of one of the inclusions and the surrounding distorted
surface of Figure 45. ~The presence of many very small
inclusions is shown in this photograph. This shows
one of the disadvantages of polishing electrolytically

(45)

as apposed to mechanical polishing, for in the latter
inclusions are removed with the metal instead of being
left on the surface.

Figures 40,46,and.47 illustrate the degree of
polish obtainable at the upper portion of the current-
voltage curve and in the higher values of the flat
portion. Here the inclusion effect is minimized and a
good polish obtained.

At higher voltages, surfaces as shown in Figures
48 and 49 were obtained. An etching effect is be-
ginning to show while at 40 voltp, the inclusions are
prominent, probably because of the larger amount of
metal removed. Under polarized light, metals polished
at high voltage values show the anisotropy of martensite
without further etching better than polishes at low
voltages. According to Jacquet5, at voltages corres-
ponding to points above the flat portion of the current-
voltage curve, gas is evolved.which causes uneven attack.
As no gas bubbles are observable in this electrolyte
at any voltage, it is presumed that this complication
does not occur in this case and polishing may proceed
at higher voltages.

*****************

(46)

Figure 42
SAE 1090
hardened
electrolytically pol-
ished at 2 volts
100x

 

Figure 43
SEE 109C
hardened
electrolytically pol-
ished at 5 volts
lCOx

 

(47 )

Figure 44

SAE 1020
hardened

-
l
0
D.
h
l
a
c
.l
t
H
O
r
t
c
e
1
e

ished at 9 volts

 

100x

Figure 45

SAE 1090

hardened

a
t
l

0

V
t

a
d

e
h

8
.1

100x

 

(48)

Figure 46
SAE lOeO

hardened
electrolytically pol-
ished at 20 volts
1001

 

Figure 47
SAE 1090
hardened
electrolytically pol-
ished at 24 volts
100x

 

(49)

Figure 48
SAE 1090
hardened
electrolytically pol-
ished at 28 volts
100x

 

Figure 49
SAE 1090
hardened
electrolytically pol-
ished at 40 volts
lOOx

 

P.

(50)

Figure 50
SAE 1090
hardened
electrolytically pol-
ished at 12 volts
750x

 

Figure 51
SAE 10120
spheroidized
electrolytically pol-
ished at 12 volts
750x

 

 

(51)

The composition of a steel is not the only factor
affecting the voltage range, for the phases present af-
fect the voltage values to be used and the quality of
the resulting polish. Figure 51 is a photomicrograph
of the same SAE 10120 steel as shown in Figure 41, but
in the former the cementite is in a spheroidized condition
while the latter shows a fully hardened structure. Both
_were polished at approximately the same voltage, but
this is often not the case. The etching effect of Figure
51 is caused by the different rate of attack on the two
phases, leaving the carbide in relief. The following
table gives the voltage ranges for two steels with two

different structures for each.

SAE 1090 annealed 2-10 volts
1090 hardened 6-20

SAE 1040 annealed 8-22 volts
1040 hardened 10-26

While the differences in range may be small as in
the second case, they may also be of considerable sig-

nificance as in the first.

***#***#******

The degree of polish obtained depends to a great
extent upon the freshness of the acetic anhydride-per-
chloric acid electrolyte, for as it ages, the surface
obtained is more or less strongly etched. This deter-
ioration may be caused by oxidation of the organic com-

pounds, absorption of water, or both. The solution

(52)

darkens with age which supports the first explanation,
while the addition of water was observed to decrease
the quality of the polish obtained thereafter.

The next series of pictures were made from spec-
imens of SAE 1040 steel which were polished in an
electrolyte which had been used for some time and resulted
in etched surfaces. The specimens were different struc-
tures and.hence vary considerably in the surface revealed.
In addition, the same surfaces were lightly etched with
2% nital to show the variation of the electrolytically
etched to the fully etched surfaces. The decrease in
prominence of surface irregularities is also, upon

etching, brought out in these photomicrographs of the
surfaces after etching with nital. Figures 52 and 53
are photomicrographs of hardened structure, Figures 54
and 55 of annealed structure, and 56 and 57 of normal-
ized structure. The "wavy" appearance of ferrite grains
in the annealed and normalized structures and the etched
needles of the hardened structure are apparently in-
dications that the electrolyte has deteriorated and its
usefulness impaired. Staining of the surface when.the
specimen is withdrawn from the electrolyte also is not-
iceable when the latter has deteriorated but does not
result when using fresh electrolyte.

Figures 58 and 59 are photomicrographs at ICC
and 375 diameters of SAE 1090 steel which was quenched
at a rate less than the critical rate for full hardening

(53)

so that a structure consisting of both martensite and
troostite_resulted. The polish was accomplished in a
fresh solution of electrolyte and shows the etching_of3
the troostite while the martensite was not etched by the
electrolytic polishing. The rate-of attack is the same
for both phases as neither is in relief-and a smooth

but sharply contrasting surface is obtained.

4)

Figure 52

SAE 1040

hardened

-
l
m.
h.
l
a
c
.1
n.
l
O
r
it
C
e
l
e

8
t
l

O

V
6
rd
in

a
d

9..
.ra

15

 

 

X
5
7
3

Figure 53

SAE 1040

hardened

.
l
O
p
h
l
a
c
.1
t
u.
0
r
t
c
e
l
e

ished at 26 volts

nital etch

375x

 

(b5)

Figure 54

I \‘ I
a _.,l//_/, R. _
‘ \ . {Li/12,... 7'3‘.

annealed ‘:-.6 "i511:‘ ,f' a

[film] W) . I

electrolytically pol-

iShed at 20 VOltB '3‘: _ fi'f I 'l" "'t . swam 5,;
' I 41 ‘ w’t ' "5‘
I/j/Qm‘ ’i .: . ‘ :-‘ .- ‘ 4:1".-
375x 2),, 7 "" g 2
(g I f '

   

' I -"I V‘ m‘ .
'V J./.". y‘",

5' “a" a: 3..» ‘ Figure 55
Q 1 Y ‘ SAE 1040
annealed
electrolytically pol-
ished at 20 volts
nital etch

375x

 

(56)

Figure 56
SAE 1040
normalized
electrolytically pol-
ished at 18 volts
750x

 

Figure 57
SAE 1040
normalized
electrolytically pol-
ished at 18 volts
nital etch
750x

 

 

(57)

Figure 58
SAE 1090

slow‘quenched
electrolytically pol-
ished at 24 volts
lCOx

 

Figure 59
SAE 1090
slow quenched
electrolytlcally pol-
ished at 24 volts
nital etch
350x

 

 

(58)
ggmmggy and Conclusions

The variables investigated and the results obtained
or observations made will be discussed individually for
each variable.

Electrolyte- The acetic anhydride-perchloric acid sol-
ution, which was the only one used extensively, was

found to be successful when used fresh, but after sev-

 

eral weeks its effectiveness diminishes. Then it is no 5

longer possible to secure a smooth surface for etching

 

and staining occur.

gurrent Density- Because of the setup of the laboratory
apparatus used, the current density was not controlled.
It was found to decrease when the voltage was held con-
stant after the polishing had proceeded for a few seconds,
indicating the formation of a film which retards the
passage of current.

Voltage- The following table gives the kinds of steel
polished electrolytically, the range of voltages for
successful polishing according to current-voltage curves,
and the voltages which were successfully employed in

obtaining a good polish.

 

 

 

Steel Veltage Range _ Voltage Used
SAE 1010 a 16-30 volts 20 volts
SAE 1040 annealed 8-22 volts 16 volts

1040 hardened 10-26 16
SAE 1090 annealed 2-10 volts

1090 hardened 6-20 16 volts
SAE 10120 normalized 2-12 volts 12 volts

10120 hardened 2-12 12

f (

 

(59)

In general, it was found that the method of con-
trolling the polish by means of the voltage applied was
quite successful. Determination of the voltage to be
used by means of the current-voltage curves was verified,
as a satisfactory polish was secured in the ranges in»
dicated in all cases. Because of the inclusion effect,

it was observed that the higher values indicated min-

-- zinc-‘7'!”
r

imized this effect and gave a better surface.
ime- The time factor was not investigated extensively,

for it is dependent upon the current density which in

 

turn depends upon the voltage. Enough time must be al-
lowed to remove sufficient metal to secure a smooth sur-
face below any scratches present, so that the time is
also dependent upon the depth of the scratches. As the
specimens were all polished from 4/0 emery paper, the
depth of scratches was the same in each case. Five
minutes was found to be long enough in most cases, but
where a low current density exists and the surface is
large, this time may be doubled. Over polishing results
in an increase in the inclusion effect, as more inclusions
are left in relief, and in increased etching of the metal
itself. The Optimum time for polishing would be just
enough to remove all scratches. For a given metal,
electrolyte, and degree of previous polish, it should

be possible to construct a table from experiments that

would enable one to choose the proper length of time

'(60)

according to the current density existing and the

area of the surface.

Agitationp The agitation throughout these experiments
‘was provided by a motor-driven glass stirrer of the
propellor type. A stirring rate of about 150 r.p.m.
was used in all cases. Stirring was found to be nec-
essary to avoid the appearance of prominent ridges or
undulations instead of a flat surface.

Temperature- While no definite data was collected on
the effect of temperature, the best polishes were ob-
tained when the electrolyte was cooled by an external
ice bath to below 10°C. The electrolyte was observed
to resist deterioration better when polishing was done
in an ice bath instead of at room temperature. The
effect of temperature may be attributed to better film
maintenance because of a lower rate of diffusion at

low temperatures. This would be especially so during
polishing as considerable heat is evolved by the pas-
sage of the current which could conceivably have a strong
influence on the film if the electrolyte were not cooled.
Cathode- The cathode used throughout the experiments
was of stainless steel and about two inches in diameter.
Its position was held constant: parallel to the surface

being polished and at a distance of one-half to one inch.

gonclusions- While electrolytic polishing is not a fool-

'M

 

(61)

proof substitute for mechanical polishing. it does
offer a rapid method of securing a scratchless, «strain-
free surface after more or less extensive preliminary
experiments are made to determine the correct Operating
conditions. After this is done, the saving in time of
polishing and the quality of surface obtained make the
method most useful.

I," 'v“; ‘a-J

 

m .

(62)

Sgggestions for Further Investigation

The most troublesome problem of electrolytic pol-
ishing is the inclusion effect, which in commercial
steels containing considerable impurities greatly low-
ers the quality of surface obtained. It may be possible
by more extensive investigation with this object in
mind to greatly lessen this effect. The time factor
is especially important, for the optimum length of
polishing should result in a minimum of inclusions
left in relief. Together with the use of voltages in-
dicated in this thesis to be most beneficial in reducing
the inclusion effect, the latter may be lessened to
such a degree as to be unimportant.

Other electrolytes may be investigated as a sub-
stitute for the acetic anhydride-perchloric acid sol-
ution which deteridrates with use and must be frequently
replaced. From the literature, mixtures of sulfuric
and citric acids, phosphoric acid and glycerine, and
methanol and nitric acid have been reported as success-
ful and may make satisfactory substitutes.

A suggested improvement in the design of the
electroepolishing apparatus is a better means for rap-
idly removing the specimen from the electrolyte and
washing the latter from its surface in order to avoid

staining.

 

A
O“
(A)

V

The extension of this type of investigation to
other metals opens up an almost limitless field of
work, for with the differences in compositions of the
' alloys and multiplicity of electrolytes possible as
well-as the other variable factors of electrolytic
polishing, a host of experiments must be performed be-
fore the Optimum conditions for each metal are deter-

mined. It is hoped that the system of investigation

\ II'L JD. Kin-Ir. “’3..— Agl .

outlined and fellowed in this thesis for steel will
simplfy this type of investigation.

 

10.

11.

(64)

Bibliography

 

Z. Jeffries and R. Archer:"The Science of metals", Ch.
XII, MCGraw-Hill, 1924, New York

K. Heindlhofer and E. C. Bain: Study of the Grain Struct-
ure of martensite, Transactions A.S.S.T., ;§, 70, (1930)

P. A. Jacquet: Electrolytic Method for Obtaining Bright

Surfaces, Transactions Electrochem. Soc., g2, 629,(1936)

H. H. Uhlig: Electrolytic Polishing of Stainless Steel,
Transactions Electrochem. Soc., 28, 265, (1940)

P. A. Jacquet: On the Anodic Behaviour of COpper...
Nature, 135, 1076, (1935)

J. R. Vilella: Etching of Martensite, Metals Handbook,
1939 edition

P. A. Jacquet: Electrolytic Polishing of Tin...., Int-
ernational Tin Council Bulletin 90, February, 1939

H. Pray and C. L. Faust: Comments on.E1ectrolytic pol-
ishing of Metals, Iron.Age, 145, 33, (1940)

H. lewery, H. Wilkinson and D. L. Smare: Influence of
Polished Surface on Optical Constants of COpper,
Phile Itlage, 22, 769’ (1936)

W. H. J. Vernon and E. G. Strand: Electrolytic Polish-
ing of Zinc, Nature, 142, 477, 1161, (1938)

Electrolytic Poliahing of Iron and Steel, The Engineer,
167, 39, (1939)

 

v

T

‘ ‘ ' " j I. I
ww‘e it
1‘ , ' y ' .' , .u

3"". a.
b. .

I. ‘ —‘
) .‘- 1"
\,fi\ "ts-.37,
"""¢"

«:1. a:

L. *4, ”1"...“ -.
(rt! 'v-:°
\‘t‘.’

'3??sz

o“_ .I.
‘ \
t .
«'4,

.
‘ .-

v, 's'fl
‘1‘" “

a
n
n a_y_

‘ u..
a. c a

 

t r

 

 

I- .\

 

S,-

 

| .

 

 

“'11 \fi\\1flxj\flfijujn [1W jfgifiufljgfigfijm?“