- Email: [email protected]
Changes in residual stress during the tension fatigue of normalized and peened SAE 1040 steel
Materials Science and Engineering, 56 (1982) 259 - 263
259
Changes in Residual Stress during the Tension Fatigue of Normalized and Peened SAE 1040 Steel M. McCLINTON* and J. B. COHEN
Department o f Materials Science and Engineering, The Technological Institute, Northwestern University, Evanston, IL 60201 (U.S.A.) (Received February 27, 1982; in revised form May 6, 1982}
SUMMARY
During the tension-tension fatigue o f normalized SAE 1040 steel under applied stress control, compressive residual stresses develop, but only in or near deformation markings. However, if this steel is shot peened prior to fatigue, the compressive residual stresses produced by peening are eliminated during fatigue and replaced by tensile residual stresses, when the applied (fatigue) stress is above the endurance limit.
reported to be tensile (+) after impact fatigue [9], b u t compressive (--) stresses develop in reverse bending [10, 11]. More information is needed concerning changes in residual stresses during fatigue, particularly under modes of loading other than bending, for high cycles, and in annealed and shot-peened conditions. In this paper, preliminary information is presented on the high cycle tension-tension fatigue of both normalized and peened plain carbon steel.
2. EXPERIMENTALPROCEDURES 1. INTRODUCTION
The mechanical working of the near-surface regions of a manufactured item is often carried out, e.g. by shot peening, because it is c o m m o n l y accepted that this processing increases the fatigue limit [ 1 ]. One result of this deformation is the formation of a compressive residual stress in the surface, because the material in this vicinity is restrained by the interior of the part. For soft steels the plastic deformation due to peening, and not the residual stresses, appears to produce most of this improvement in fatigue limit, whereas for hardened steels the residual stresses are most important [2]. Nearly all the research on this topic has been carried o u t by fatiguing in bending or torsion. The t y p e of loading may be important in assessing the effect of a surface treatment. Furthermore, stresses can fade during fatigue [ 3 - 8]. It is also established that residual stresses develop during the high cycle fatigue o f annealed metals and alloys. These stresses are *Present address: Inland Steel Company Research Laboratories, East Chicago, IN 46312, U.S.A. 0025-5416]82/0000-0000/$02.75
2.1. Specimens Hot-rolled SAE 1040 steel plate was normalized by heating at 1145 K for 30 min and air cooling. Strips were cut with the length in the rolling direction of the original plate. Fatigue specimens were then machined with a gauge section 9 mm long, 6 mm wide and 6.43 mm thick, with an arc of 6.35 mm radius between the grip and gauge sections. All specimens were milled to remove oxide and the decarburized zone (detected metallographically) and were subsequently ground through 600 grit paper. After an anneal for 1 h at 973 K in argon, followed by furnace cooling, the stress was typically 2 0 - 30 MPa. This small value is similar to the reproducibility of the X-ray measurements used to obtain the stresses {this method will be described below}. Therefore it was assumed that this anneal was sufficient to provide essentially stress-free specimens. Some of these specimens were shot peened for 45 s (on both faces and on the edges} with 230 grit steel shot, yielding an Almen A intensity (the middle Almen range with center deflections of the Almen test strip of 0.1 © Elsevier Sequoia/Printed in The Netherlands
260 0
I
I
I
I
I
-100
t
~1 _
// ///
~-200
//
TABLE 1 The experimentally determined mechanical properties of the normalized and normalized and shot-peened SAE 1040 steel (two specimens, each state) a SAE 1040 steel
//
oy(0.2%) (MPa)
-300
Ultimate tensile strength
Elongation (%)
(MPa)
Normalized Shot peened
-~-400
408 426
614 596
27 27
-500 I .05
I .15
I .25
Thickness
I .3S
I :45
I ,55
Removed (ram)
Fig. 1. Stress profile or shot-peened SAE 1040 steel: , uncorrected;- - -, corrected for layer removal (the layers were removed by chemical etching (the etchant consisted of 100 parts 30% H202 and 10 parts 48% HF)) and stress gradients. 0.5 mm) o f 0.343 mm. The profile of residual stress f r o m the surface t o the interior is shown in Fig. 1, as det er m i ned by X-ray diffraction. Because the corrections [12] for layer removal and the stress gradient were small, no corrections were made to all subseq u en t measurements of a profile. It should be n o t e d t h a t there is no indication of tensile stresses just below the surface as might occur because o f overpeening.* T he mechanical properties of th e specimens are summarized in Table 1. Shot peening causes only small changes in these mechanical properties. The endurance limit is available only for fully reversed loading (216 MPa [13] ). For t h e t e n s i o n - t e n s i o n fatigue e m p l o y e d in this study, this limit was estimated to be approximately twice this value. 2.2. Fatigue testing Cyclic loading was p e r f o r m e d on an MTS servohydraulic machine, in load control, at 50 Hz. Th e waveform was sinusoidal, with R = 0, and a m a x i m u m stress ranging f r om 207 to 412 MPa. Observations of the surface were made at 100× in white light, with a travelling microscope. Specimens were removed f r o m the machine f r om time to time *The stresses were measured in o n l y the ferrite and are balanced through the cross section by stresses in the carbides, and by a small tensile stress at depths greater than those e x a m i n e d here.
aExperiments were performed using an Instron tensile-testing machine. The specimens were of the same shape and size as those used for the fatigue tests.
to measure the residual stress at various numbers of cycles and t hen reinserted in the grips. (A test was not carried t o failure, e x c e p t where indicated in the text.) 2.3. X-ray m e a s u r e m e n t s An a u t o m a t e d Picker X-ray d i f f r a c t o m e t e r was e m p l o y e d with (line-focused) Co K radiation. The position of the 310 diffraction peak was used t o determine residual stresses. T he beam divergence was 0.4°(20 ) for the annealed specimens, and the beam was masked t o provide an irradiated area on the specimen of 2.3 m m 2. For the shot-peened specimens, a divergence of 1 ° was employed. The irradiated area was 14.8 m m 2. The halfwidth was 3.3 ° com pared with a value of 0.6 ° for the annealed specimens. This broader less intense peak necessitated the wider divergence. (The receiving slit was 0.2 ° wide in all cases.) Stresses were obtained via an on-line minic o m p u t e r control system [14] with t he socalled stationary slit method. A five-point parabolic fit was e m p l o y e d t o the t op 15% of a peak, t o establish its 20 position, typical of m o d e r n practice in the U.S.A. In stress measurements using X-ray diffraction the stress is obtained from the slope of interplanar spacing d (calculated from the peak position) versus sin2~ where ~ is the tilt of the specimen surface normal from its usual position (in which position the normal t o the surface bisects the incident and diffracted beams). Six ~ tilts t o 45 ° were em pl oyed, approxi m at el y equally spaced. The surface residual stress was obtained in the direction of the load, i.e. along the long direction o f a specimen's sur-
261
face. There was only slight curvature in d versus sin2@, which indicates that there was no appreciable change in stress over the penetration depth of the X-ray beam [15]. (The correlation coefficient for a least-squares fit of such data to a straight line was typically 0.99 or better.) Measurements were repeated on several specimens, and the results indicated a precision of 20 - 30 MPa or less (3% 10% of the stress value), as was indicated above.
- 240
-20C o n
'~ -16c
~ o
I/) (D hi
n-- -12C Io:I d
<
-80
o
{D o o
~: -4C
o -o-o-
j
3. R E S U L T S
UNCYCLED
VALUE i
3.1. As-normalized specimens After 106 cycles, a plastic strain of 3 X 10- s was detected at an applied stress of 296 MPa (from the cyclic stress-strain hysteresis loop). This value increased to 5 X 10- s at the highest stress. No plastic strain was detected below 296 MPa, and the changes in residual stress were within experimental error, even after 106 cycles (at or below this applied stress). At 376 MPa and between 0 and 5000 cycles, a specimen remained smooth in appearance under the optical microscope. At a higher number of cycles, deformation bands were noticed in some areas, similar to Liider's bands. These increased in size and frequency with further cycling, covering the entire gauge length at approximately 20 000 cycles. Up to this range, significant changes in residual stress were detected (Fig. 2(a)) but only when the X-ray beam was placed over the deformation bands. {This is why the X-ray beam was masked, to reduce its area to a size near that of the bands.} After the bands covered the entire specimen, further changes in stress occurred, and the value was independent of the location of the X-ray beam. A similar sequence was observed at a higher applied stress, 412 MPa (Fig. 2(b)), but with deformation bands first occurring at a much smaller number of cycles (600 cycles). This specimen failed at 1.8 X l 0 s cycles. After cycling so that the residual stress was uniform across the gauge length, and the deformation bands were throughout the gauge area, layers were removed by etching and it was found that the compressive stresses extended to at least 0.8 m m below the surface (at which point the etching was stopped).
4CI0°
i
I01
i
i
102 iO3 NUMBER OF
i
i
I06
i
10z
i
l0s
CYCLES
(a) -Z40, -200 A O Q.
o# ii•
-16c !
-12c -8C -40 S
0
UNCYCLED
VALUE
4o I(
i
i01
i
i
i
i
i
I0s 102 103 i04 105 NUMBER OF CYCLES
i
107
i
I0 s
(b) Fig. 2. (a) R e s i d u a l stress v s . n u m b e r o f cycles for n o r m a l i z e d S A E 1 0 4 0 steel at a n a p p l i e d stress o f 3 7 6 MPa. T h e t w o s y m b o l s (e, o) r e p r e s e n t t w o different specimens cycled under identical conditions. P o i n t s w i t h a h o r i z o n t a l line i n d i c a t e a r e p e a t e d residual stress m e a s u r e m e n t at a d i f f e r e n t l o c a t i o n o n t h e s a m e s p e c i m e n , w h e r e a s p o i n t s w i t h a vertical line r e p r e s e n t t h e r a n g e o f r e p e a t e d m e a s u r e m e n t s at t h e s a m e l o c a t i o n . (b) R e s i d u a l stress v s . n u m b e r o f cycles for n o r m a l i z e d S A E 1 0 4 0 steel a t a n applied stress o f 4 1 2 MPa. P o i n t s w i t h a vertical line i n d i c a t e t h e r a n g e o f r e p e a t e d m e a s u r e m e n t s o f residual stress a t t h e same l o c a t i o n .
3.2. Shot-peened condition In this case the residual stress was the same from region to region in the gauge area and, because of the surface roughening caused by the peening, no deformation bands due to fatigue were apparent. The changes in residual stress during fatigue at various loads are
262
-600
-4O(
-6OC
,,..,. UNCYCLED VALUE
a. -50C
•
•
-50C
UNCYCLED VALUE
-40C
~-30C (/) bU (/') u,.l t..,
,'," -300
-2OC
I-(h
-IOC
-.I
-2OC ee
_~ ~ lOG (D tlJ
r,-
C
W
I OC
20C
30 ¢
i
i
I0'
Io z
i
i
t0 3
Io 4
i
i
1o5
1o6
i
20C0 o
Io T
'
I 2
i0 s
NUMBER
NUMBER OF CYCLES
(a)
104
OF
i0 s
i0 •
107
CYCLES
(b)
-6OO UNCYCLED
'~ -50C
VALUE
I"1.
-40C (,/) nr
-soo
I--
(n _ 20C .J
~ - I00 W
a:
C
O
I00
O
•
O~
0
200
30O,oO
i
I01
i 10 2
i 10 3
NUMBER
i 10 4
i I0 ~
i 10 6
i 10 7
OF CYCLES
Fig. 3. Residual stress v s . number of cycles for a shot peened specimen at applied stresses of (a) 263 MPa, (b) 336 MPa and (c) 400 MPa.
(e) shown in Fig. 3. There was some fading even well below the estimated endurance limit, i.e. at an applied stress of 263 MPa. Significant hysteresis (plastic strain) was detected even at this stress. Fading was more pronounced at 336 MPa. At an applied stress of 408 MPa, close to the static yield stress, the compressive residual stress induced by peening was replaced by a tensile residual stress between 0 and 100 cycles. This specimen had a life of only 115 000 cycles. The profile of residual stress was obtained on a second specimen, after 800 cycles at this last applied stress, and is shown in Fig. 4, which should be compared with the profile just after peening (Fig. 1). It is clear that the original profile has been completely altered by fatigue; tensile residual stresses extend to a depth of about 0.1 mm.
200 o
0.
IOO
,,_I Q -IOO n.-
- 200
.o's
.,'~ THICKNESS
.2's
.sb
REMOVED
.,;'s
.s's
(ram)
Fig. 4. Stress profile of a shot-peened specimen after 800 cycles at 408 MPa.
263 4. DISCUSSION We consider first the u n p e e n e d steel, in which compressive stress develops in the d e f o r m a t i o n bands. Elongated grains in such bands would be restrained by t he surrounding material and this is m os t like the cause of the compressive stress in t he bands. Such stresses would impede fatigue crack propagation. After shot peening, t he near-surface regions are severely cold worked. In t e n s i o n - t e n s i o n fatigue, the observed change of sign in the residual stress near the surface (from compressive to tensile) implies that the bulk is deforming (elongating) m o r e than t he surface regions and putting the near-surface regions into tension. (The fact t h a t plastic deformation occurred in this case is indicated by the o p en hysteresis loops.) It is clear f r o m these results t hat the shot peening o f steels is n o t always beneficial for fatigue life. Under t e n s i o n - t e n s i o n fatigue, or if the fatigue history is largely tensile, and if the applied stresses are near the fatigue limit, deleterious stresses m a y develop subsequent to peening, b u t not in the absence of peening.
ACKNOWLEDGMENTS Mr. W. P. Evans, Caterpillar T r a c t o r Company, graciously p e r f o r m e d the shot peening. This research was s uppor t ed by t he Office of Naval Research. T he X-ray measurements were carried o u t in the X-ray Facility at N o r t h w e s t e r n University, supported in part by the National Science F o u n d a t i o n - M a t e r i a l s Research L a b o r a t o r y Program unde r Grant
DMR-76-80847. Portions of this study were presented (by M. M.) in partial fulfillment o f the requirements for the M.S. degree in the Materials Science and Engineering Department, N o r t h w e s t e r n University.
REFERENCES
1 Metals Handbook, Vol. 2, American Society for
Metals, Metals Park, OH, 1964, p. 398. 2 W. P. Evans, R. E. Ricklefs and J. F. Millan, in J. B. Cohen and J. E. Hilliard (eds.), Local A tomic Arrangements Studied by X-ray Diffraction,
Gordon and Breach, New York, 1966, p. 351. 3 S. Taira and K. Hayashi, Proc. 9th Jpn. Congr. on Testing Materials, 1966, Kyoto University, Kyoto, 1966, p. 1. 4 S. Kodama, 3rd Int. Conf. on Fracture, Mi~nchen, 1973, Paper V]4418. 5 T. Hayama and H. Yoshitake, Bull. JSME, 14 (1971) 1272. 6 E. J. Pattison and D. S. Dugdale, Metallurgica, 66 (1962) 228. 7 B. Jaensson, Proc. Semin. on Fatigue, Saabgarden, Rimforsa, 1977.
8 G. R. Leverant, B. S. Langer, A. Yuen and S. W. Hopkins, Metall. Trans. A, 10 (1979) 251. 9 G. Koves, Adv. X-ray Anal., 9 (1965) 468. 10 S. Taira and Y. Murakami, Proc. 4th Jpn. Congr. on Testing Materials, 1961, Kyoto University, Kyoto, 1961, p. 5. 11 S. Taira and K. Honda, Proc. 5th Jpn. Congr. on Testing Materials, 1962, Kyoto University, Kyoto, 1962, p. 8. 12 M. E. Hilley, J. A. Larson, G. F. Jatczak and R. E. Ricklefs (eds.), SAE Publ. J784a, 1971 (Society of Automotive Engineers, Warrendale, PA). 13 N. E. Frost, K. J. Marsh and A. L. P. Pook, Metal Fatigue, Clarendon, Oxford, 1964, p. 58. 14 M. R. James and J. B. Cohen, Adv. X-ray Anal., 20 (1977) 291. 15 J. B. Cohen, H. Dolle and M. R. James, NBS Spec. Publ. 567, 1980, p. 453 (National Bureau of Standards, U.S. Department of Commerce).
259
Changes in Residual Stress during the Tension Fatigue of Normalized and Peened SAE 1040 Steel M. McCLINTON* and J. B. COHEN
Department o f Materials Science and Engineering, The Technological Institute, Northwestern University, Evanston, IL 60201 (U.S.A.) (Received February 27, 1982; in revised form May 6, 1982}
SUMMARY
During the tension-tension fatigue o f normalized SAE 1040 steel under applied stress control, compressive residual stresses develop, but only in or near deformation markings. However, if this steel is shot peened prior to fatigue, the compressive residual stresses produced by peening are eliminated during fatigue and replaced by tensile residual stresses, when the applied (fatigue) stress is above the endurance limit.
reported to be tensile (+) after impact fatigue [9], b u t compressive (--) stresses develop in reverse bending [10, 11]. More information is needed concerning changes in residual stresses during fatigue, particularly under modes of loading other than bending, for high cycles, and in annealed and shot-peened conditions. In this paper, preliminary information is presented on the high cycle tension-tension fatigue of both normalized and peened plain carbon steel.
2. EXPERIMENTALPROCEDURES 1. INTRODUCTION
The mechanical working of the near-surface regions of a manufactured item is often carried out, e.g. by shot peening, because it is c o m m o n l y accepted that this processing increases the fatigue limit [ 1 ]. One result of this deformation is the formation of a compressive residual stress in the surface, because the material in this vicinity is restrained by the interior of the part. For soft steels the plastic deformation due to peening, and not the residual stresses, appears to produce most of this improvement in fatigue limit, whereas for hardened steels the residual stresses are most important [2]. Nearly all the research on this topic has been carried o u t by fatiguing in bending or torsion. The t y p e of loading may be important in assessing the effect of a surface treatment. Furthermore, stresses can fade during fatigue [ 3 - 8]. It is also established that residual stresses develop during the high cycle fatigue o f annealed metals and alloys. These stresses are *Present address: Inland Steel Company Research Laboratories, East Chicago, IN 46312, U.S.A. 0025-5416]82/0000-0000/$02.75
2.1. Specimens Hot-rolled SAE 1040 steel plate was normalized by heating at 1145 K for 30 min and air cooling. Strips were cut with the length in the rolling direction of the original plate. Fatigue specimens were then machined with a gauge section 9 mm long, 6 mm wide and 6.43 mm thick, with an arc of 6.35 mm radius between the grip and gauge sections. All specimens were milled to remove oxide and the decarburized zone (detected metallographically) and were subsequently ground through 600 grit paper. After an anneal for 1 h at 973 K in argon, followed by furnace cooling, the stress was typically 2 0 - 30 MPa. This small value is similar to the reproducibility of the X-ray measurements used to obtain the stresses {this method will be described below}. Therefore it was assumed that this anneal was sufficient to provide essentially stress-free specimens. Some of these specimens were shot peened for 45 s (on both faces and on the edges} with 230 grit steel shot, yielding an Almen A intensity (the middle Almen range with center deflections of the Almen test strip of 0.1 © Elsevier Sequoia/Printed in The Netherlands
260 0
I
I
I
I
I
-100
t
~1 _
// ///
~-200
//
TABLE 1 The experimentally determined mechanical properties of the normalized and normalized and shot-peened SAE 1040 steel (two specimens, each state) a SAE 1040 steel
//
oy(0.2%) (MPa)
-300
Ultimate tensile strength
Elongation (%)
(MPa)
Normalized Shot peened
-~-400
408 426
614 596
27 27
-500 I .05
I .15
I .25
Thickness
I .3S
I :45
I ,55
Removed (ram)
Fig. 1. Stress profile or shot-peened SAE 1040 steel: , uncorrected;- - -, corrected for layer removal (the layers were removed by chemical etching (the etchant consisted of 100 parts 30% H202 and 10 parts 48% HF)) and stress gradients. 0.5 mm) o f 0.343 mm. The profile of residual stress f r o m the surface t o the interior is shown in Fig. 1, as det er m i ned by X-ray diffraction. Because the corrections [12] for layer removal and the stress gradient were small, no corrections were made to all subseq u en t measurements of a profile. It should be n o t e d t h a t there is no indication of tensile stresses just below the surface as might occur because o f overpeening.* T he mechanical properties of th e specimens are summarized in Table 1. Shot peening causes only small changes in these mechanical properties. The endurance limit is available only for fully reversed loading (216 MPa [13] ). For t h e t e n s i o n - t e n s i o n fatigue e m p l o y e d in this study, this limit was estimated to be approximately twice this value. 2.2. Fatigue testing Cyclic loading was p e r f o r m e d on an MTS servohydraulic machine, in load control, at 50 Hz. Th e waveform was sinusoidal, with R = 0, and a m a x i m u m stress ranging f r om 207 to 412 MPa. Observations of the surface were made at 100× in white light, with a travelling microscope. Specimens were removed f r o m the machine f r om time to time *The stresses were measured in o n l y the ferrite and are balanced through the cross section by stresses in the carbides, and by a small tensile stress at depths greater than those e x a m i n e d here.
aExperiments were performed using an Instron tensile-testing machine. The specimens were of the same shape and size as those used for the fatigue tests.
to measure the residual stress at various numbers of cycles and t hen reinserted in the grips. (A test was not carried t o failure, e x c e p t where indicated in the text.) 2.3. X-ray m e a s u r e m e n t s An a u t o m a t e d Picker X-ray d i f f r a c t o m e t e r was e m p l o y e d with (line-focused) Co K radiation. The position of the 310 diffraction peak was used t o determine residual stresses. T he beam divergence was 0.4°(20 ) for the annealed specimens, and the beam was masked t o provide an irradiated area on the specimen of 2.3 m m 2. For the shot-peened specimens, a divergence of 1 ° was employed. The irradiated area was 14.8 m m 2. The halfwidth was 3.3 ° com pared with a value of 0.6 ° for the annealed specimens. This broader less intense peak necessitated the wider divergence. (The receiving slit was 0.2 ° wide in all cases.) Stresses were obtained via an on-line minic o m p u t e r control system [14] with t he socalled stationary slit method. A five-point parabolic fit was e m p l o y e d t o the t op 15% of a peak, t o establish its 20 position, typical of m o d e r n practice in the U.S.A. In stress measurements using X-ray diffraction the stress is obtained from the slope of interplanar spacing d (calculated from the peak position) versus sin2~ where ~ is the tilt of the specimen surface normal from its usual position (in which position the normal t o the surface bisects the incident and diffracted beams). Six ~ tilts t o 45 ° were em pl oyed, approxi m at el y equally spaced. The surface residual stress was obtained in the direction of the load, i.e. along the long direction o f a specimen's sur-
261
face. There was only slight curvature in d versus sin2@, which indicates that there was no appreciable change in stress over the penetration depth of the X-ray beam [15]. (The correlation coefficient for a least-squares fit of such data to a straight line was typically 0.99 or better.) Measurements were repeated on several specimens, and the results indicated a precision of 20 - 30 MPa or less (3% 10% of the stress value), as was indicated above.
- 240
-20C o n
'~ -16c
~ o
I/) (D hi
n-- -12C Io:I d
<
-80
o
{D o o
~: -4C
o -o-o-
j
3. R E S U L T S
UNCYCLED
VALUE i
3.1. As-normalized specimens After 106 cycles, a plastic strain of 3 X 10- s was detected at an applied stress of 296 MPa (from the cyclic stress-strain hysteresis loop). This value increased to 5 X 10- s at the highest stress. No plastic strain was detected below 296 MPa, and the changes in residual stress were within experimental error, even after 106 cycles (at or below this applied stress). At 376 MPa and between 0 and 5000 cycles, a specimen remained smooth in appearance under the optical microscope. At a higher number of cycles, deformation bands were noticed in some areas, similar to Liider's bands. These increased in size and frequency with further cycling, covering the entire gauge length at approximately 20 000 cycles. Up to this range, significant changes in residual stress were detected (Fig. 2(a)) but only when the X-ray beam was placed over the deformation bands. {This is why the X-ray beam was masked, to reduce its area to a size near that of the bands.} After the bands covered the entire specimen, further changes in stress occurred, and the value was independent of the location of the X-ray beam. A similar sequence was observed at a higher applied stress, 412 MPa (Fig. 2(b)), but with deformation bands first occurring at a much smaller number of cycles (600 cycles). This specimen failed at 1.8 X l 0 s cycles. After cycling so that the residual stress was uniform across the gauge length, and the deformation bands were throughout the gauge area, layers were removed by etching and it was found that the compressive stresses extended to at least 0.8 m m below the surface (at which point the etching was stopped).
4CI0°
i
I01
i
i
102 iO3 NUMBER OF
i
i
I06
i
10z
i
l0s
CYCLES
(a) -Z40, -200 A O Q.
o# ii•
-16c !
-12c -8C -40 S
0
UNCYCLED
VALUE
4o I(
i
i01
i
i
i
i
i
I0s 102 103 i04 105 NUMBER OF CYCLES
i
107
i
I0 s
(b) Fig. 2. (a) R e s i d u a l stress v s . n u m b e r o f cycles for n o r m a l i z e d S A E 1 0 4 0 steel at a n a p p l i e d stress o f 3 7 6 MPa. T h e t w o s y m b o l s (e, o) r e p r e s e n t t w o different specimens cycled under identical conditions. P o i n t s w i t h a h o r i z o n t a l line i n d i c a t e a r e p e a t e d residual stress m e a s u r e m e n t at a d i f f e r e n t l o c a t i o n o n t h e s a m e s p e c i m e n , w h e r e a s p o i n t s w i t h a vertical line r e p r e s e n t t h e r a n g e o f r e p e a t e d m e a s u r e m e n t s at t h e s a m e l o c a t i o n . (b) R e s i d u a l stress v s . n u m b e r o f cycles for n o r m a l i z e d S A E 1 0 4 0 steel a t a n applied stress o f 4 1 2 MPa. P o i n t s w i t h a vertical line i n d i c a t e t h e r a n g e o f r e p e a t e d m e a s u r e m e n t s o f residual stress a t t h e same l o c a t i o n .
3.2. Shot-peened condition In this case the residual stress was the same from region to region in the gauge area and, because of the surface roughening caused by the peening, no deformation bands due to fatigue were apparent. The changes in residual stress during fatigue at various loads are
262
-600
-4O(
-6OC
,,..,. UNCYCLED VALUE
a. -50C
•
•
-50C
UNCYCLED VALUE
-40C
~-30C (/) bU (/') u,.l t..,
,'," -300
-2OC
I-(h
-IOC
-.I
-2OC ee
_~ ~ lOG (D tlJ
r,-
C
W
I OC
20C
30 ¢
i
i
I0'
Io z
i
i
t0 3
Io 4
i
i
1o5
1o6
i
20C0 o
Io T
'
I 2
i0 s
NUMBER
NUMBER OF CYCLES
(a)
104
OF
i0 s
i0 •
107
CYCLES
(b)
-6OO UNCYCLED
'~ -50C
VALUE
I"1.
-40C (,/) nr
-soo
I--
(n _ 20C .J
~ - I00 W
a:
C
O
I00
O
•
O~
0
200
30O,oO
i
I01
i 10 2
i 10 3
NUMBER
i 10 4
i I0 ~
i 10 6
i 10 7
OF CYCLES
Fig. 3. Residual stress v s . number of cycles for a shot peened specimen at applied stresses of (a) 263 MPa, (b) 336 MPa and (c) 400 MPa.
(e) shown in Fig. 3. There was some fading even well below the estimated endurance limit, i.e. at an applied stress of 263 MPa. Significant hysteresis (plastic strain) was detected even at this stress. Fading was more pronounced at 336 MPa. At an applied stress of 408 MPa, close to the static yield stress, the compressive residual stress induced by peening was replaced by a tensile residual stress between 0 and 100 cycles. This specimen had a life of only 115 000 cycles. The profile of residual stress was obtained on a second specimen, after 800 cycles at this last applied stress, and is shown in Fig. 4, which should be compared with the profile just after peening (Fig. 1). It is clear that the original profile has been completely altered by fatigue; tensile residual stresses extend to a depth of about 0.1 mm.
200 o
0.
IOO
,,_I Q -IOO n.-
- 200
.o's
.,'~ THICKNESS
.2's
.sb
REMOVED
.,;'s
.s's
(ram)
Fig. 4. Stress profile of a shot-peened specimen after 800 cycles at 408 MPa.
263 4. DISCUSSION We consider first the u n p e e n e d steel, in which compressive stress develops in the d e f o r m a t i o n bands. Elongated grains in such bands would be restrained by t he surrounding material and this is m os t like the cause of the compressive stress in t he bands. Such stresses would impede fatigue crack propagation. After shot peening, t he near-surface regions are severely cold worked. In t e n s i o n - t e n s i o n fatigue, the observed change of sign in the residual stress near the surface (from compressive to tensile) implies that the bulk is deforming (elongating) m o r e than t he surface regions and putting the near-surface regions into tension. (The fact t h a t plastic deformation occurred in this case is indicated by the o p en hysteresis loops.) It is clear f r o m these results t hat the shot peening o f steels is n o t always beneficial for fatigue life. Under t e n s i o n - t e n s i o n fatigue, or if the fatigue history is largely tensile, and if the applied stresses are near the fatigue limit, deleterious stresses m a y develop subsequent to peening, b u t not in the absence of peening.
ACKNOWLEDGMENTS Mr. W. P. Evans, Caterpillar T r a c t o r Company, graciously p e r f o r m e d the shot peening. This research was s uppor t ed by t he Office of Naval Research. T he X-ray measurements were carried o u t in the X-ray Facility at N o r t h w e s t e r n University, supported in part by the National Science F o u n d a t i o n - M a t e r i a l s Research L a b o r a t o r y Program unde r Grant
DMR-76-80847. Portions of this study were presented (by M. M.) in partial fulfillment o f the requirements for the M.S. degree in the Materials Science and Engineering Department, N o r t h w e s t e r n University.
REFERENCES
1 Metals Handbook, Vol. 2, American Society for
Metals, Metals Park, OH, 1964, p. 398. 2 W. P. Evans, R. E. Ricklefs and J. F. Millan, in J. B. Cohen and J. E. Hilliard (eds.), Local A tomic Arrangements Studied by X-ray Diffraction,
Gordon and Breach, New York, 1966, p. 351. 3 S. Taira and K. Hayashi, Proc. 9th Jpn. Congr. on Testing Materials, 1966, Kyoto University, Kyoto, 1966, p. 1. 4 S. Kodama, 3rd Int. Conf. on Fracture, Mi~nchen, 1973, Paper V]4418. 5 T. Hayama and H. Yoshitake, Bull. JSME, 14 (1971) 1272. 6 E. J. Pattison and D. S. Dugdale, Metallurgica, 66 (1962) 228. 7 B. Jaensson, Proc. Semin. on Fatigue, Saabgarden, Rimforsa, 1977.
8 G. R. Leverant, B. S. Langer, A. Yuen and S. W. Hopkins, Metall. Trans. A, 10 (1979) 251. 9 G. Koves, Adv. X-ray Anal., 9 (1965) 468. 10 S. Taira and Y. Murakami, Proc. 4th Jpn. Congr. on Testing Materials, 1961, Kyoto University, Kyoto, 1961, p. 5. 11 S. Taira and K. Honda, Proc. 5th Jpn. Congr. on Testing Materials, 1962, Kyoto University, Kyoto, 1962, p. 8. 12 M. E. Hilley, J. A. Larson, G. F. Jatczak and R. E. Ricklefs (eds.), SAE Publ. J784a, 1971 (Society of Automotive Engineers, Warrendale, PA). 13 N. E. Frost, K. J. Marsh and A. L. P. Pook, Metal Fatigue, Clarendon, Oxford, 1964, p. 58. 14 M. R. James and J. B. Cohen, Adv. X-ray Anal., 20 (1977) 291. 15 J. B. Cohen, H. Dolle and M. R. James, NBS Spec. Publ. 567, 1980, p. 453 (National Bureau of Standards, U.S. Department of Commerce).