Lehigh University
Lehigh Preserve
Theses and Dissertations
1-1-1977
Residual stress versus depth distribution resulting
from routine production processing.
Ramon Mantz
Follow this and additional works at: http://preserve.lehigh.edu/etd
Part of the Materials Science and Engineering Commons
Recommended Citation
Mantz, Ramon, "Residual stress versus depth distribution resulting from routine production processing." (1977). Theses and
Dissertations. Paper 2249.
This Thesis is brought to you for free and open access by Lehigh Preserve. It has been accepted for inclusion in Theses and Dissertations by an
authorized administrator of Lehigh Preserve. For more information, please contact [email protected].
RESIDUAL STRESS VERSUS DEPTH DISTRIBUTION
RESULTING FROM
ROUTINE PRODUCTION PROCESSING
by
Ramon Mantz
A Thesis
Presented to the Graduate Committee
of Lehigh University
in Candidacy for the Degree of
Master of Science
in
Metallurgy and Materials Science
Lehigh University
1977
ProQuest Number: EP76525
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
uest
ProQuest EP76525
Published by ProQuest LLC (2015). Copyright of the Dissertation is held by the Author.
All rights reserved.
This work is protected against unauthorized copying under Title 17, United States Code
Microform Edition © ProQuest LLC.
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106-1346
CERTIFICATE OF APPROVAL
This thesis is accepted and approved in partial fulfillment
of the requirements for the degree of Master of Science.
4 )),Y117
PATE
Professor in Charge
Chairman of the Department
of Metallurgy and Materials
Science
n
ACKNOWLEDGEMENTS
The author wishes to thank Mr. J. A. Larson and Mr. J. F. Cantrel
for providing the basic idea for this investigation as well as helpful
direction throughout the course of this work.
Appreciation is also extended to Dr. D. Thomas for his support
and useful suggestions.
The author also acknowledges the contributions and cooperation of
members of the Ingersoll-Rand Central Materials Service Laboratory
(CMSL), whose advice and assistance were invaluable.
The author thanks Ingersoll-Rand Company for providing the funds
and the CMSL for providing the facilities to carry out the investigation.
Finally, the author thanks his family for their understanding
and patience during the course of this work.
m
TABLE OF CONTENTS
Page
Abstract
I
II
III
IV
1
Introduction
2
Procedure
4
Results and Discussion
7
Conclusions
12
Bibliography
14
Appendixes
A
31
B
32
C
34
Vita
35
"IV
LIST OF TABLES
Page
Table 1
16
Table 2
16
LIST OF FIGURES
Page
Fig. 1
Full view of piston
17
Fig. 2
Sample section being positioned for x-ray
measurement
18
Fig. 3
Sample being electrolytically etched
19
Fig. 4
Microstructure of case
20
Fig. 5
Microstructure of core
21
Fig. 6
Microstructure showing decarburized layer at
surface of sample as heat treated
22
Fig. 7
Residual stress versus depth, as measured and after
correction for material removed
23
Fig. 8
Residual stress, retained austenite and Re hardness
versus depth after heat treatment
24
Fig. 9
Residual stress and retained austenite versus depth
after grinding
25
Fig. 10
Residual stress and retained austenite versus depth
after temper
26
Fig. 11
Comparison of residual stress versus depth before and
after grinding
27
Fig. 12
Residual stress versus depth before and after shot
peening
28
Fig. 13
Rockwell-C hardness and FWHM before and after shot
peening, versus depth
29
Fig. 14
Sample of graphical integration and correction for
material removed
30
VI
Abstract
Residual stress versus depth has been investigated at four stages
in the production of a percussive rock drill piston.
effects of a surface or near-surface
The beneficial
layer of compressive residual
stress in increasing the fatigue life of steel parts is well known,
and this investigation was intended to determine not only the final
residual stress versus depth distribution, but the effects of
individual process steps on the existing distribution.
Depth profiles of residual stress were measured after heat
treating, grinding, tempering and shot peening.
Measurements were
made using the two-exposure x-ray diffraction technique, and
material removal was accomplished by electrolytic etching.
In
addition to residual stress measurements, volume percent retained
austenite was also profiled versus depth.
The residual stress versus depth profile produced by heat
treating showed the desired compressive residual stresses throughout the case and into the core before becoming tensile.
Subsequent
grinding and tempering produced no changes in the original residual
stress versus depth profile other than at the surface.
Shot peening
produced a "hook" shaped profile of high compressive residual stress
with a maximum value at a depth of 0.05 to 0.13 mm (0.002 to 0.005 in.),
and extending to a depth of approximately 0.46 mm (0.018 in.).
The
resulting residual stress versus depth distribution of the final
product is the desired condition, and should contribute to increased
fatigue life of the parts.
Introduction
Residual stresses in materials arise from "
any process -
mechanical, thermal, or chemical - that results in a permanent,
non-uniform change in shape or volume." (1)
In general, these resi-
dual stresses on, or near, the surface are beneficial to the fatigue
life of the material if compressive, and harmful if tensile.
With
the introduction of the precision x-ray diffraction method for
measuring residual stresses in hardened steels, (2) studies that had
been impractical were made possible, and considerable investigation
was undertaken to establish the origin of the residuaT stresses in
these steels, as well as their effect on the fatigue life.
Early work by Koistinen (3), and Koistinen and Marburger, (4)
and more recently by Motoyama, et. al. (5) investigated the origin of
residual stress generation during carburization.
These investigations
indicate an inversion of the start of austenite to martensite transformation to some point below the surface, rather than at the surface,
as the controlling factor.
This inversion results from a lowering
of the Ms temperature by the increased carbon content near the
surface, to the extent that during cooling the Ms temperature is
initially reached at some depth below the surface, which is determined
by the chemistry and cooling rate.
Koistinen also established that
similar residual stresses could be introduced in through-hardening
steels, without altering the carbon level, with the addition of other
alloying elements by diffusion. (6)
Liss, et. al. (7) and Nelson et. al. (8) established that residual
compressive stresses could be introduced in plain carbon and low alloy
steels by severe quenching.
The origin of these heat treat stresses
is indicated to be a combination of the difference in specific volume
through the non-uniform microstructure, and thermal straining from the
heat flow and expansion-contraction characteristics of the steels.
In addition, mechanical and cold working processes have been
investigated and shown to cause varying residual stresses under
different conditions. (9) (10) (4)
The positive correlation between compressive residual stresses
on or near the surface, and increased fatigue life has been established
in investigations by Mattson and Coleman (10) and, Nelson et. al. (8).
Results of various investigations are summarized by Morrow, et. al. (11).
The present investigation is production oriented and was prompted
by occasional failure of percussive rock drill pistons in a fatigue
mode.
While the residual stresses resulting from the various produc-
tion processes are known in theory, it is also known that changes in
process conditions can cause significant deviation from the expected
residual stress condition.
Following each of four routine process steps,
the residual stress versus depth profile
was measured to establish the
effect of each subsequent step on the previous profile, as well as to
establish the residual stress versus depth profile of the finished
product.
•3-
Procedure
The piston samples, Figure 1, were selected at random from a
routine production lot, and received no special treatment during the
course of the investigation.
Although only two were to be used, six
pieces were selected to provide, as nearly as possible, identical
samples for back-up should the need arise to repeat any portion
of the investigation, and possibly to be used in future investigations.
The residual stress versus depth was investigated following
each of four production steps; heat treating, grinding, tempering
and shot peening.
Ideally, since there are four parallel sections
of bearing surface in the area of interest, all four processes
would have been investigated on one sample, thereby guaranteeing
identical starting conditions.
However, because of the size and
weight of the overall sample, it was necessary to cut-off the portion
of interest using an Allison-Campbell cut-off machine.
To avoid
any variation from normal production grinding, it was necessary to
submit a complete piston.
This necessitated the use of a second
sample which, after being returned for grinding, was sectioned in
the same manner as the original sample.
This second sample, in the
cut-up condition, presented no problems for the remaining two production steps, and the cutting up of more samples was avoided.
In removing the area of interest, care was taken to maintain
a sufficiently large section to avoid affecting the residual stresses
in the area where the measurements were to be made.
Figure 2 shows
the section of interest in position on the diffractometer.
-4-
The residual stress measurements were made by x-ray diffraction, using the two-exposure method originally described by
Christenson and Rowland (2), and set down comprehensively by Hilley
et. al. in an SAE information report (1).
The martensite (211)
diffraction peak was measured at both the normal sample orientation,
Y= 0°, and at W= 45 , using chromium K-alpha radiation and a
Siemens diffractometer.
The measuring conditions are listed in
Table I, and a brief description of the procedure is given in
Appendix A.
The volume percent retained austenite was also deter-
mined by x-ray diffraction, using the ratio of the austenite (200)
and martensite (200) measured
diffracted lines.
The conditions
are the same as in Table I, and a brief description of the procedure is given in Appendix B.
The peak breadth was determined concurrently with the residual stress measurements by recording a scan of the (211) peak at
random depths, and graphically measuring the peak width, in degrees
20, at one-half the peak intensity after background correction.
i
To determine the uniformity of the residual stress distribution,
surface values were measured on opposite sections of the surface of the
first sample, and on all four sections of the bearing surface of the
second sample after grinding.
All residual stress and retained
austenite measurements are in the longitudinal direction, at the
approximate midpoint of the bearing surface.
Profiles of the residual
stress versus depth below the surface were obtained by alternately
removing a thin layer of material and measuring the residual stress on
-5-
the new surface.
Layers of material were removed by electrolytically
etching a small area approximately 38 mm square, which was masked off
with a lacquer coating to protect the remaining portion of the piece.
The etching conditions are listed in Table II, and Figure 3 shows the
arrangement with the piece in place, suspended through a cutout in
the plexiglass plate.
Although the etch rate varied with the age of
the solution, approximately .025 mm (.001 in.) was removed in four
minutes.
Because removal of covering material alters the original stress
value at any point below the surface, corrections were made using the
relationship established for a hollow cylinder by Moore and Evans (12).
This relationship is shown in equationp] , with 6(n.)
6(^=6^, -2j^6m(r)dr
the original
[1 ]
residual stress value at radius r-j ,6,^ is the measured value at
radius r-j, and R-j and R are the inside and outside radii of the
cylinder, respectively.
The integration was performed graphically,
and a sample correction is described in Appendix C, and illustrated
in Figure 14.
Cross sections were cut from both samples for examination of the
microstructure; one being in the quenched and tempered condition and
the other being in the final product condition, having been ground,
tempered and shot peened.
Hardness surveys were taken on both sections
through the case and into the core using a Tukon microhardness tester.
Values were obtained at intervals of 0.025 mm (0.001 in.), from 0.025 mm
(0.001 in.), to 0.25 mm (0.010 in.), using a Knoop indenter with a
•6-
500 g load.
From 0.25 mm (0.010 in.) into the core, Vickers values
were obtained at intervals of 0.25 mm (.010 in.), using a 10 kg load.
Results of these measurements and the measurements of residual
stress and retained austenite are given in the next section.
Results and Discussion
The pistons are machined from a controlled-hardenability, shallowhardening, modified W-l type tool steel.
The heat treatment consists
of a double quench from above critical temperature, followed by
tempering at 477°K (204°C).
This routinely produces a case of
approximately 3.2 mm (0.125 in.), depth, the microstructure showing
fine, dense martensite with a uniform dispersion of carbides in the
case, and fine dense pearlite in the core, as shown in Figures 4 and
5 respectively.
Figure 6 shows a thin surface layer, approximately
0.064 mm (0.0025 in.), with little or no carbides, indicating slight
decarburization.
On the basis of the prescribed heat treatment a residual stress
versus depth profile is expected to show compressive stresses through
the case, before turning tensile at some point in the core, resulting
from a combination of specific volume differences and thermal straining
as proposed by Liss et. al. (7).
Figure 7 shows the measured residual
stress versus depth profile after heat treating, as well as the profile
corrected for material removed, as per Appendix C.
It can be seen
that despite the cummulative nature of the correction, only the
magnitude of the profile is changed, the basic shape remaining unchanged,
All values from this point on will be corrected values.
-7-
Initially, surface values were measured on opposite sections of
the bearing surface.
The closeness of the two values, 630 MPa and
610 MPa (91.4 ksi and 88.5 ksi), compressive, indicates a relatively
uniform distribution of residual stress around the circumference.
This indication of uniform distribution was further substantiated by
measurements on all four sections of the bearing surface following
grinding, showing a spread of only 40 MPa (5.7 ksi), at a level of
approximately 600 MPa (87 ksi), or less than 7%.
On the basis of
these measurements the sections of bearing surface for the depth
profiles were selected at random.
The residual stress versus depth profile is repeated in
Figure 8, together with profiles of volume percent retained austenite and Rockwell C hardness.
All three profiles show slightly
reduced surface values as a result of the decarburized layer.
The residual stresses, as expected, are compressive through the
case and into the core before turning to tensile at a depth of 5.85 mm
(0.230 in.).
The hardness profile shows the desired high hardness
through the case, and establishes the effective case depth to be 3.8 mm
(0.150 in.), in accordance with the criteria of Re 50.
An unexpected phenomenon is observed in the residual stress
profile, in the existence of a shallow, thin layer of high compressive
values
as compared to the remainder of the case.
The fact that the
sample is from a routine production lot, and therefore exact conditions
of heat treatment are not known, makes explanation of the origin of
this phenomenon speculative at best.
However coincidence of this
-8-
layer of high compressive residual stresses with the maximum retained
austenite concentration provides a possible explanation of a psuedocarburized case in this near-surface layer.
It is well established
that an increase in the amount of carbon dissolved in the austenite
results in a decrease of the Ms temperature (3) (5).
If, on the basis
of the increased austenite concentration in this layer, the assumption
is made that a larger proportion of carbides were dissolved during
austenitizing in this layer, the result would be a reduced Ms temperature.
According to the findings of Koistinen (3) and Motoyama et. al. (5)
this would produce a shallow, maximum compressive residual stress
similar in profile to the near surface layer of Figure 8.
If such
a psuedo-carburized effect were superimposed on the residual stresses
produced by volume differences and thermal straining, the resulting
profile might coincide with that of Figure 8.
Of more importance, however, are the effects of the subsequent
processing steps in determining the residual stress profile of the
finished product.
Figures 9 and 10 show the residual stress and
retained austenite profiles measured after grinding and after the
subsequent tempering, respectively.
In Figure 11, the residual
stress profile after grinding is superimposed on the original profile of Figure 8 with the surface position offset by 0.38 mm (0.015 in.),
the nominal thickness removed during grinding.
With the exception of
the high compressive value on the new surface, the two profiles are
virtually identical, indicating uniformity of the original conditions
in both pistons.
Comparison of the retained austenite profiles
-9-
in Figures 8 and 9 shows that, at corresponding radii, these profiles
also coincide with the exception of a slightly reduced value at the
new surface.
The fact that the residual stress has been affected only on the
surface
indicates that the grinding conditions are gentle, at least
for the final few layer removals.
Grinding has been shown to pro-
duce residual stresses from a combination of mechanical and thermal
processes, with three possible results (4) (9).
In the case of
abusive grinding, subsurface heating results in tempering and shrinkage
of the surface layers, causing tensile residual stress on the surface
and into the case.
When grinding is less abusive, mechanical defor-
mation of the surface layer plays an equal part, resulting in a high
compressive residual stress on the surface, which reverts to tensile
immediately below and continues into the case.
Gentle grinding
conditions result in little or no subsurface heating, in which case
the only residual stresses that arise are from the mechanical deformation of the surface layer, as is shown to be the case by Figure 11.
The surface reduction of retained austenite concentration is also
attributed to this mechanical deformation.
Comparison of Figures 9 and 10, shows that the tempering had
no effect on the residual stress except to reduce the surface value
by approximately 25%, from 585 MPa to 440 MPa (85 ksi to 64 ksi).
With that exception the two profiles coincide through the case and
into the core.
■10-
This reduction of surface residual stresses from tempering,
even at low temperatures, was reported by Liss, et. al. (7).
From
these results it seems reasonable to speculate that the initial
surface value in the quenched condition was above 700 MPa (100 ksi).
The most significant effect of the tempering is the general
lowering of the retained austenite concentration, even though the
shape of the profile remains unchanged.
Figure 12 shows the residual stress versus depth profile after
shot peening superimposed on the profile after tempering.
The shot
peening produced a "hook" shaped profile of high compressive stresses
which changed the previous profile to a depth of approximately 0.46 mm
(0.018 in.), at which point the profiles again coincide.
The maximum
compressive value of 1065 MPa (155 ksi), occurs at a depth of 0.05 to
0.13 mm (0.002 in. to 0.005 in.).
Figure 13 shows a hardness profile taken on the finished product,
and a plot of FWHM for the (211) diffraction line.
A direct relation-
ship has been shown between x-ray peak broadening and hardness in
tempered steels (4).
As can be seen however, the peak breadth is de-
creased significantly over the depth affected by the peening, while
the hardness remains unchanged.
Nelson et. al. (8) reported this
sharpening of the x-ray diffraction peak resulting from shot peening,
assuming that it was accompanied by softening.
In the same work, it
was also reported that peening actually decreased existing surface
residual stresses in high hardness materials.
•11-
The results shown in
Figures 12 and 13 do not agree with these findings, and suggest instead
a decrease in the microstrain accompanying the increase in macros train.
Evans and Buenneke (13), in their investigation of x-ray line
broadening in hardened steels, relate the peak width to a combination
of grain size and microstrain; the peak becoming broader with decreasing
grain size and increasing microstrain.
Although their results do not
extend above Re 50, extrapolation of their results for shot peened
samples would indicate no change in grain size, but a reduction in
microstrain, and therefore a reduction of peak width.
The residual stress versus depth profile of the finished product
appears to be highly desirable in respect to increased fatigue life.
As shown in Figures 9, 10 and 12, the three process steps investigated
caused no detrimental changes to the original heat treatment residual
stress profile.
The final step, in fact, enhanced the near surface
residual compressive stresses significantly.
A direct correlation
has been shown between the magnitude of residual compressive stress
near the surface and fatigue life (8) (10), with Mattson and Coleman (10)
reporting an increase of fatigue life, at a peak stress equal to 95%
of yield strength, of twenty times resulting from shot peening.
Al-
though none of the residual stress versus fatigue life work cited
involves parts similar in nature to those in the present investigation,
there is no reason to feel the relationship would be any different.
Conclusions
This investigation has shown the effects of routine manufacturing
processes on the original residual stress versus depth profile obtained
-12-
during heat treatment, and has shown that these processes lead to a
finished product with a desirable residual stress versus depth profile.
It can be concluded that:
1.
The heat treatment produces a desirable residual stress
versus depth profile.
2.
Grinding conditions, at least during removal of the final
layers of material, are gentle in nature and produce
no significant subsurface heating.
3.
Subsequent tempering reduces residual compressive stresses
only on the surface.
4.
Shot peening results in a "hook" shaped profile of very
high compressive residual stresses in a shallow surface
region.
5.
The manufacturing process did not affect the original
residual stress profile other than in the near-surface
region.
6.
On the basis of established direct correlation between
magnitude of residual compressive stress near the surface
and fatigue life, the final residual stress versus depth
profile is desirable.
■13-
References
1.
SAE Information Report; Residual Stress Measurement by X-ray
Diffraction - SAE J784a, edited by M. E. Hi 1 ley, et. al., Society
of Automotive Engineers, New York, N.Y., 1971.
2.
A. Christenson and E. Rowland, "X-ray Measurements of Residual
Stress in Hardened High Carbon Steel", Transactions, American
Society for Metals, Vo. 45 (1953), p. 638.
3.
D. P. Koistinen, "The Distribution of Residual Stresses in
Carburized Cases and Their Origin", Transactions, American Society
for Metals, Vol. 50, (1958), p. 227.
4.
D. P. Koistinen and R. E. Marburger, in "Internal Stresses and
Fatigue in Metals", p.p. 110 to 119, edited by G. M. Rassweiler
and W. L. Grube, Elsevier Publishing Co., New York, 1959.
5.
M. Motoyama, R. E. Ricklefs and J. A. Larson, "The Effect of
Carburizing Variables on Residual Stresses in Hardened Chromium
Steel", SAE Transactions, Vol. 84, (1975), p. 344.
6.
D. P. Koistinen, "The Generation of Residual Compressive Stresses
in the Surface Layers of Through-Hardening Steel Components by
Heat Treatment", Transactions, American Society for Metals, Vol. 57.
(1964), p. 581.
7.
R. B. Liss, C. G. Massieon and A. S. McKloskey, "The Development
of Heat Treat Stresses and Their Effect on Fatigue Strength of
Hardened Steels",SAE Transactions, Vol. 74, (1966), p. 870.
8.
D. V. Nelson, R. E. Ricklefs, and W. P. Evans, "Residual Stresses
in Quenched and Tempered Plain Carbon Steels", SAE Transactions,
Vol. 80, (1971), p. 107.
9.
L. P. Tarasov, W. S. Hyler and H. R. Letner, "Effect of Grinding
Conditions and Resultant Residual Stresses on the Fatigue Strength
of Hardened Steel", Proceedings of the American Society for Testing
Materials, Vol. 57, (1957), p. 601.
10.
R. L. Mattson and W. S. Coleman, Jr., "Effect of Shot-Peening
Variables and Residual Stresses on the Fatigue Life of Leaf-Spring
Specimens", SAE Transactions, Vol. 62, (1954), p. 546.
11.
SAE Information Report, Influence of Residual Stress on Fatigue of
Steel - SAE J783, edited by J. Morrow and J. F. Mi 11 an. Society
of Automotive Engineers, New York, N.Y., 1962.
■14-
12.
M. G. Moore and W. P. Evans, "Mathematical Correction for Stress
in Removed Layers in X-ray Diffraction Residual Stress Analysis",
SAE Transactions, Vol. 66, (1958), p. 340.
13.
W. P. Evans and R. W. Buenneke, "X-ray Line Broadening of Hardened
and Cold Worked Steel", Transactions of the Metallurgical Society
of AIME, Vol. 227, (1963), p. 447.
14.
B. D. Cullity, Elements of X-ray Diffraction Addison-Wesley Publishing
Co., Reading, Mass., 1956.
•15-
TABLE I
X-ray Measuring Conditions
Radiation
Cr K-alpha, wave length 2.29 A
Excitation Potential
40 kV
Tube Current
20 mA
Divergent Beam Slit
1°
Receiving Slit
1.0 mm
Detector
Siemens Scintillation Counter with pulseheight analysis
'
TABLE II
Electrolytic Etching Conditions
Solution
-
60% Phosphoric/40% Sulfuric
Voltage
-
12 V
Current
-
20 A
Time
-
Increments of 15 min.*
*Larger thicknesses removed by repeated applications, suitably spaced
to avoid over-heating of solution and sample.
16-
Fig. 1
Full view of piston, with area of interest
at left. Arrow shows one of four parallel
sections of bearing surface where measurements
were made. Piston is approximately 406 mm
(16 in.) long by 101 mm (4 in.) in diameter.
-17-
Fig. 2
Sample section being positioned for x-ray measurement.
18-
Fig. 3
Sample being electrolytically etched,
19-
Fig. 4
Microstructure of case; fine dense martensite
with spherical carbides. 500X.
-20-
Fig. 5
Microstructure in core; dense pearlite and
carbides. 500X.
-21-
Fig. 6
Microstructure showing decarburized layer at
surface of sample as heat treated. 500X.
■22-
E
£
Q.
<D
Q
sOl
-»->
t<a
T3
c:
(O
-o
O)
t-
13
to
na
If)
<D
E
to
re
r>
x:
4->
Q.
Ol
T3
.
10
Z3 ■D
to QJ
>
O
> E
t-
OJ
m
to
</>
<L>
i-
CD
S-
i—
03
4J •nC/0 S<D
i—
CM
(O
3
■o
+->
(X3
E
Sto o
CD 4-
•i—
D;
C
r-»-
o
»l—
+J
Ol
s:s*
o
OJ
S-
Ol so
u_ u
•1—
002 001
a|isus»i(+)
0
ooi ooz o6e o6f oos 069 ooz
(-jsAjssaJdujoo
(DdW) 9-ssa^s IDnpisey
-23-
3
y
u>
CM
o
CM
00
in
_1_
002
oot . 6
»l!su5i(+)
ssaupjDH
CM
O
in
m
_i
1_
o
CO
«
>_
obi o6s ope 06* 06s o69~~o6z
(-) dAlssaJcluJoo
(DdH) 9-ss0jjs lonpisey
.&_
u>
CM
CO
ejiuajsnv psupjsy
-24-
CM
<*.
%'IOA
t£
a;
o
s(D
-M
<T3
+->
a.
aj
-a
u>
to
to
S-
cu
>
<D
4->
in
■r—
CD
4->
to
<:
-o
<u
£=
■^
<C
a>
a:
-a
en
ra
to
to
cu
s-
+J
CM
fO
-a -a
•r- C
tO •!-
OJ s-
CTl
OJ
cn
062 061.
si!suaj.(+)
(DdH)
eg
5
ooi 062 o6e o6* 06s 069
(_) vAissajduuoo
OO/L
Q-ssejjs |Dnpisey
o
T-f
00.
u>
CM
ajiuejsnv peupjey
-25-
%|OA
E
E
a.
Q
c
+J
Q.
<D
<
u>
to
13
10
S(L)
■o
C
ft
■u
>
o
a:
CU
4->
*i—
in
CD
CO
O)
■4->
CD
ce
c
rO
p>
to
to
CD
•
S- s_
+j (U
OO QL
E
1— til
CM
"O
•rto
CU
c£
tj>
obz 061.
S|ISliai(+)
5
o6t o6z o6e 00* oos o6s o6z
(_) 9AISS9JdUU00
(DdW) 9-sseJjs' lonpjsey
CM
o
oo
CM
C0
ejiuejsnv peuiojey %'1°A
-26-
S_
<U
+J
•+n3
-o r—
£= *
en
T5
O)
SO
H<U
?
_c
'■a
e>
'iL
Z.
ii
c_
.c
o
i»
O
o
*>
■**
09
1
<
1
o
JC 44J
Q.+J
CD <D
"O (/l
4to 43 O
CO
S- C
Q) O
> •!-
+->
1
1
S<D
•«->
(O
E
CO 'I—
to to
<u o
S- Q.
+J
OO CD
O
r— (O
«o 4rs s-
T3 =J
•i- CO
to
CD
o: •
en
4- c
O -r-
c £=
O -rto i-
•
•r- cn-o
S-
B S.
O. CD
E ■!->
O 4-
CD
>
O
E
OJ
c_> re s-
CD
sCD
o6? o6t
ooi. 002 o6e oo*
(_)dAiss9Jduioo
(DdW) 9-ssajjs pnpisay
•27-
E
E
a.
Q
i~
en
c
E
«>
t)
m
D.
■♦■»
O
CD
4->
14-
O)
'E
«>
■o
CL
l/>
JC
JC
O
t>
I/)
L.
(O
CD
so
itCD
-Q
ca
+J
Q.
i
i
i
a;
a
CO
to
CM
CD
>
to
to
a»
•
r(C
13
T3
•1to
OJ
CD
OJ
Q.
S- CD
4-> C
C/) •!-
+J
O
-C
a; to
C\J
CD
SZ3
en
0
obi o6z
OOEOS^
(-) aAissaJduuoo
(PdH)
06s 069 obi
9-SS64S lDnpisay
-28-
O6Q
006
62KHO6U
64
o
CT
63J
-o—
62
U)
(/)
Q)
C 61
T3
t_
O
X60J
Before Shot Peening
After Shot Peening
"0H5
Figure 13
050
0?45~
Depth (mm)
"oio"
Rockwell C Hardness and FWHM before and
after shot peening versus depth.
■29-
"oTs
E E
E E
N
T
O
*™
£?
£
55
= <**
n
X X
<r>
u>
v"
E E
E
E
T- T-
CM
i
CM
+->
CJ
(X)
sso
o
X
-a
c
<o
0.05
0.05
X
o
II
Z Z
£ tn
8o
<n
u>
c
o
sen
O
CD
4->
c
u •
•r- T3
-C CD
O- >
fO o
S- E
en cu
S4O i—
(O
CO -ri— SQ. CD
E4->
rtJ rO
OO E
<1J
(UUUU/DdW)
•30-
w"^
2*-2J
Appendix A
Residual Stress Measurement by X-ray Diffraction
The two-exposure method for measuring residual stresses by x-ray
diffraction is covered quite comprehensively in a recent SAE Handbook
Supplement. (1)
In general, the method measures the change in diffracted
beam angle, 29, as the sample surface is rotated through a known angle,
Y , in respect to the incident x-ray beam.
This shift in peak position
is a measure of the difference in interplanar spacing at the two
orientations, which is attributed to a residual macrostress.
Section 2 of (1), equation 31, shows the stress given by
6=
2
&(aW^W^l(^- V)
where
E is the modulus of elasticity
V is Poisson's ratio
W is the angle of rotation in the x-ray beam
9 is the angle of indicence of the x-ray beam on the atomic planes
29x and 29^.
are the diffracted beam angles before and after the sample is
rotated, respectively
When V= 45 , the constant term in brackets equals 86.3 ksi, if the
literature values of 30,000 ksi and .27 are used for Young's Modulus
and Poisson's ratio respectively.
Measurements of the time to count 2X10
5
counts were made at three
equally spaced angles bracketing the peak and above 85% of the peak
intensity.
After correction for Lorentz-Polarization and absorption
these times were fitted to a parabola to accurately determine the 29
values.
These 29 values were then used in the above equation to calculate
the residual stress.
-31-
Appendix B
Determination of Volume % Retained Austenite by X-ray Diffraction
The determination of volume % retained austenite by x-ray
diffraction is based on the ratio of the net counts in the austenite
peak to the net counts in the martensite peak.
Considering the
theoretical intensity equation for a diffracted line from a constituent of a multiphase sample, (14)
where
Ii
Ci
V
1+Cps226
Sin29Cos£
p
F ?M
e
K
is
is
is
is
the
the
the
the
net intensity of the diffracted line of i
volume fraction of i
volume of a unit cell of i
Lorentz-Polarization correction
is
is
is
is
the multiplicity factor of the diffracting planes
the structure factor
a temperature factor
constant independent of the kind and amount of i,
it can be seen that for a given diffracted line, all terms other than
I-,- and Cj are constants.
By setting these terms equal to some constant.
Rj, the equation (1) is shortened to
Ii=RjCi
[2]
Taking the ratio of an austenite (Y- iron), diffraction line to a
martensite ( Ot - iron), diffraction line yields
tyi*r RtCr/RnC*
[3]
This ratio still contains two unknowns,C<y and Cot.-
In tne case
a two component sample, however, we know that
C*+Cot=1
M
■32-
°f
By substituting the above equation, solved for CQ , into equation
[3] and solving for Oy , the necessary equation is obtained,
Cr=
-Kfv^xyi,)
with onlyCy unknown.
^]
Simple multiplication of [5] by 100 changes
volume fraction to volume % retained austenite.
■33-
Appendix C
Correction for Material Removed
The corrections were made using the relationship established
by Moore and Evans (12).
Despite the large "wall" thickness, it was
decided to use the correction for a hollow cylinder, as opposed to
that for a solid cylinder.
The correction involves a mathemetical
integral containing 6 as a function of r.
Routine integration is
impossible, since this functional relationship is unknown, so the
integral was evaluated graphically instead, by plotting the integrand
gr_jv'6m(r), versus depth, and graphically determining the cummulative area under the curve from the original surface to the radius at
which the correction is desired.
the integral /
<4\
%
§ ^r) dr
r ♦fy
The value obtained is the value of
, where R and R-i are the outside and
inside radii of the cylinder, respectively, and r-\ is the radius at
which the correction is being made.
This value is then substituted
directly into the correction equation, equation #1, to yield the
corrected residual stress 6r, at radius r-\.
graphical integration is shown in Figure 14.
■34-
An example of the
VITA
Ramon Mantz, son of Mr. & Mrs. Warren L. Mantz, was
born in Laurys, Pennsylvania, on April 21, 1940.
He graduated
from Allentown High School, now Wm Allen High, in Allentown,
Pennsylvania, in June of 1958.
From 1959 to 1964, he served
in the U.S. Air Force, after which he earned a Bachelor of
Arts degree in Physics from Kutztown State College in 1969.
From 1970 to 1975, he was employed by Western Electric Company,
Allentown, Pennsylvania, where he was involved in the analysis
of materials for the microelectronics industry.
Since 1975,
he is employed by Ingersoll-Rand Company, Phillipsburg, New
Jersey, where he is involved in materials analysis as well as
failure analysis.
•35-