Professional Documents
Culture Documents
Fatigue Curves
Edited by
Howard E. Boyer
Senior Technical Editor
American Society for Metals
The Materials
Information Society
ASM lnternatlonal"
Materials Park, Ohio 44073-0002
www.asminternational.org
Preface
This Atlas was developed to serve engineers who are looking for fatigue
data on a particular metal or alloy. In the past, the first step to locating this
data was an expensive and time-consuming search through the technical literature. Now, many ofthe important and frequently referenced curves are presented together in this one volume. They are arranged by standard alloy designationsand are accompanied by a textual explanation offatigue testing and
interpretation of test results. In each case, the individual curve is thoroughly
referenced to the original source.
Having these important curves compiled in a single book will also facilitate the computerization of these data. Plans are currently under way also to
make the data presented in this book available in ASCII files for analysis by
computer programs.
The Atlas of Fatigue Curves is obviously not complete, in that many
more curves could be included. Persons wishing to contribute curves to this
compilation for inclusion in future revisions should contact the Editors,
Technical Books, American Society for Metals, Metals Park, Ohio 44073.
Contents
Fatigue Testing
Introduction
I
Fatigue Crack Initiation 4
Fatigue Crack Propagation
12
27
43
Contents
VI
71
83
Contents
4-38.
4-39.
4-40.
4-41.
4-42.
4-43.
4-44.
4-45.
4-46.
4-47.
4-48.
4-49.
4-50.
4-51.
4-52.
4-53.
4-54.
4-55.
4-56.
4-57.
4-58.
4-59.
4-60.
4-61.
4-62.
4-63.
4-64.
5-16.
5-17.
5-18.
147
VII
viii
Contents
5-19. Effect of R-Ratio and Test Temperature on Crack Propagation of HSLA Steel
Grade I
165
5-20. Effect of Test Temperature on Fatigue Crack Propagation Behavior for Two
HSLA Steel Grades
166
5-21. Stress-Cycle Curves for Weldments of Different HSLA Steel Grades
167
5-22. Weldments (FCA W): SAE 980 X Steel vs 1006 168
5-23. Weldments (TIG): DOMEX 640 XP Steel Welded Joints vs Parent Metal
169
5-24. Weldments (FCAW Dressed by TIG): Fatigue Life Estimates Compared With
170
Experimental Data for SAE 980 X Steel
5-25. SAE 980 X Steel Weldment (FCAW): Smooth Specimen vs TIG-Dressed vs
As-Welded
171
5-26. SAE 980 X Steel Weldment (FCAW): Lap-Shear Joints
172
5-27. Microalloyed HSLA Steels: Properties of Fusion Welds
173
5-28. Microalloyed HSLA Steels: Properties of Spot Welds
174
176
183
7-1. 0.5%Mo Steel: Effect of Hold Time in Air and Vacuum at Different
Temperatures
183
7-2. DIN 14 Steel (1.5 Cr, 0.90 Mo, 0.25 V): Effect of Liquid Nitriding
184
7-3. 2.25Cr-1.0Mo Steel: Influence of Cyclic Strain Range on Endurance Limit in
Various Environments
185
7-4. 2.25Cr-1.0Mo Steel: Effect of Elevated Temperature
186
7-5. 2.25Cr-I.OMo Steel: Effect of Elevated Temperature and Strain Rate
187
7-6. 2.25Cr-1.0Mo Steel: Effect of Temperature on Fatigue Crack Growth Rate
188
7-7. 2.25Cr-1.0Mo Steel: Effect of Cyclic Frequency on Fatigue Crack Growth
Rate
189
7-8. 2.25Cr-1.0Mo Steel: Fatigue Crack Growth Rates in Air and Hydrogen
190
7-9. 2.25Cr-1.0Mo Steel: Effect of Holding Time
191
7-10. Cast 2.25Cr-1.0Mo Steel, Centrifugally Cast: Fatigue Properties at 540C
(1000 OF) 192
7-11. HII Steel: Crack Growth Rate in Water and in Water Vapor
193
7-12. 9.0Cr-1.0Mo Steel: Creep-Fatigue Characteristics
194
7-13. 9.0Cr-1.0Mo Modified Steel: Stress Amplitudes Developed in Cycling
195
7-14. 9.0Cr-1.0Mo Modified Steel: Effect of Deformation
196
197
8-1. Type 301 Stainless Steel: Scatter Band for Fatigue Crack Growth Rates
197
8-2. Type 301 Stainless Steel: Effects of Temperature and Environment on Fatigue
Crack Growth Rate
198
8-3. Type 304 Stainless Steel: Effect of Temperature on Frequency-Modified
Strains
199
8-4. Type 304 Stainless Steel: Fatigue Crack Growth Rate-Annealed and Cold
Worked
200
8-5. Type 304 Stainless Steel: Effect of Humidity on Fatigue Crack Growth Rate
201
8-6. Type 304 Stainless Steel: Effect of Aging on Fatigue Crack Growth Rate
202
8-7. Type 304 Stainless Steel: Effect of Temperature on Fatigue Crack Growth
Rate
203
8-8. Type 304 Stainless Steel: Damage Relation at 650C (1200 OF) 204
Contents
8-9. Type 304 Stainless Steel: Fatigue Crack Growth Rate at Room and Subzero
Temperatures
205
8-10. Types 304 and 304L Stainless Steel: Effect of Cryogenic Temperatures on Fatigue
Crack Growth Rate 206
8-11. Type 304 Stainless Steel: Fatigue Crack Growth Rate in Air With Variation in
Waveforms
207
8-12. Type 304 Stainless Steel: Effect of Hold Time on Cycles to Failure 208
8-13. Type 304 Stainless Steel: Effect of Hold Time and Continuous Cycling on Fatigue
Crack Growth Rates
209
8-14. Type 304 Stainless Steel: Effect of Cyclic Frequency on Fatigue Crack Growth
Rate 210
8-15. Type 304 Stainless Steel: Effect of Frequency on Fatigue Crack Growth
Behavior 211
8-16. Type 304 Stainless Steel Welded With Type 308: Fatigue Crack Growth Rates
212
8-17. Types 304 and 310 Stainless Steel: Effect of Direction on S-N 213
8-18. Types 304, 316, 321, and 348 Stainless Steel: Effects of Temperature on Fatigue
Crack Growth Rates
214
8-19. Type 309S Stainless Steel: Effect of Grain Size on Fatigue Crack Growth
Rate 215
8-20. Type 310S Stainless Steel: Effect of Temperature on Fatigue Crack Growth
Rate
216
8-21. Type 316 Stainless Steel: Growth Rate of Fatigue Cracks in Weldments
217
8-22. Type 316 Stainless Steel: Fatigue Crack Growth Rates-Aged vs Unaged
218
8-23. Type 316 Stainless Steel: Fatigue Crack Growth Rates-Effect of Aging
219
8-24. Type 316 Stainless Steel: Effect of Temperature on Fatigue Crack Growth
Rate
220
8-25. Type 316 Stainless Steel: Effect of Cyclic Frequency on Fatigue Crack Growth
Rate 221
8-26. Type 316 Stainless Steel: Fatigue Crack Growth Rate in the Annealed
Condition 222
8-27. Type 316 Stainless Steel: Effect of Environment (Sodium, Helium, and Air) on
Cycles to Failure
223
8-28. Types 316 and 321 Stainless Steel: Effects of Gaseous Environments on Fatigue
Crack Growth Rates
224
8-29. Type 32I Stainless Steel: Effect of Hold Time on Fatigue Crack Growth Rates
225
8-30. Type 403 Stainless Steel: Effect of Environment on Fatigue Crack Growth
Rate
226
8-3I. Type 403 Modified Stainless Steel: Scatter of Fatigue Crack Growth Rates
227
8-32. Type 422 Stainless Steel: Fatigue Crack Growth Rates in Precracked
Specimens
228
8-33. Type 422 Stainless Steel: Fatigue Strength-Longitudinal vs Transverse
229
8-34. Type 422 Stainless Steel: Effect of Temperature on Fatigue Strength
230
8-35. Type 422 Stainless Steel: Effects of Delta Ferrite on Fatigue Strength 231
8-36. 17-4 PH Stainless Steel: Fatigue Crack Growth Rates in Airvs Salt Solution 232
8-37. 15-5 PH Stainless Steel: Fatigue Crack Growth Rates in Air vs Salt Solution 233
8-38. PH 13-8 Mo Stainless Steel: Fatigue Crack Growth Rates at Room
Temperature 234
8-39. PH 13-8 Mo Stainless Steel: Fatigue Crack Growth Rates in Air and Sump Tank
Water 235
8-40. PH 13-8 Mo Stainless Steel: Fatigue Crack Growth Rates at Subzero
Temperatures
236
8-41. PH 13-8 Mo Stainless Steel: Constant-Life Fatigue Diagram
237
8-42. Types 600 and 329 Stainless Steel: S-NCurves for Two Processing Methods
238
8-43. Grade 21-6-9 Stainless Steel: Effect of Temperature on Fatigue Crack Growth
Rates
239
8-44. Kromarc 58 Stainless Steel: Effect of Cryogenic Temperatures on Weldments
240
8-45. Pyromet 538 Stainless Steel: Effects of Welding Methods on Fatigue Crack
Growth Rates
241
8-46. Duplex Stainless Steel KCR 171: Corrosion Fatigue
242
ix
Contents
243
9-1. Grades 200, 250, and 300 Maraging Steel: S-N Curves for Smooth and 'Notched
Specimens
243
9-2. Grade 300 Maraging Steel: Fatigue Life in Terms of Total Strain
244
245
272
Contents
11-15.
I 1-16.
11-17.
II-18.
11-19.
1I-20.
I 1-21.
1I-22.
11-23.
11-24.
I1-25.
11-26.
11-27.
11-28.
II-29.
11-30.
1I-31.
I 1-32.
I 1-33.
11-34.
11-35.
11-36.
11-37.
1I-38.
11-39.
11-40.
11-41.
I 1-42.
11-43.
11-44.
1I-45.
I 1-46.
11-47.
319
xi
xii
Contents
Contents
12-51.
12-52.
12-53.
12-54.
14-55.
12-56.
12-57.
12-58.
12-59.
12-60.
12-61.
12-62.
12-63.
12-64.
12-65.
12-66.
12-67.
12-68.
12-69.
12-70.
12-71.
12-72.
12-73.
12-74.
12-75.
12-76.
Alloy 7075 (High Purity): Effect of Iron and Silicon on Cycles to Failure
370
Alloy X-7075: Effect of Grain Size on Cycles to Failure
371
Alloy X-7075: Effect of Grain Size on Stress-Life Behavior
372
Alloy X-7075: Effect of Environment; Air vs Vacuum
373
Alloy X-7075: Effect of Environment on Two Different Grain Sizes 374
Alloy X-7075: Effect of Grain-Boundary Ledges on Cycles to Failure 375
Alloys X-7075 and 7075: Effects of Chromium Inclusions on Fatigue Crack
Propagation
376
Alloy 7475-T6: S-N Diagram for a Superplastic Fine-Grain Alloy
377
Alloy 7475: Effect of Alignment of Grain Boundaries on Cycles to Failure
378
Alloy 7475-T6: Superplastic vs Nonsuperplastic, as Related to Fatigue Crack
Growth
379
Alloys X-7075 and 7075: Effect of Chromium-Containing Inclusions on Cycles to
Failure
380
Aluminum Forging Alloys: Stress Amplitude vs Reversals to Failure
381
AI-5Mg-0.5Ag: Effect of Condition on Fatigue Characteristics
382
AI-Zn-Mg and AI-Zn-Mg-Zr: Effect of Grain Size on Strain-Life Behavior
383
AI-Zn-Mg: Strain-Life Curves of a Large-Grained Alloy
384
Aluminum With a Copper Overlay: Stress Amplitude vs Cycles to Failure
385
P/M Alloys 7090 and 7091 vs Extruded 2024 386
P / M Alloys 7090 and 709I vs 1/ M 7050 and 7075 Products
387
P/M Aluminum Alloys: Typical Fatigue Behavior
388
P / M Aluminum Alloys: Comparison With Specimens Made by Ingot
Metallurgy
389
P/M Aluminum Alloys: Comparison With Forged 7175 for Cycles to
Failure
390
Various Aluminum Alloys: Comparison of Grades for Corrosion-Fatigue Crack
Growth Rates; Air vs Salt Water 391
Various Aluminum Alloys: Comparison of Grades for Corrosion-Fatigue Crack
Growth Rates in Salt Water
392
Various Aluminum Alloys: Wrought vs Cast, and Influence of Casting Method on
Fatigue Life 393
Aluminum Casting Alloy AL-195: Interrelationship of Fatigue Properties With
Degree of Porosity
394
Aluminum Casting Alloy LM25-T6: Squeeze Formed vs Chill Cast; Effect on
Reversals to Failure
395
396
407
xiii
xiv
Contents
410
410
411
414
17-1. Unalloyed Titanium, Grade 3: S-N Curves for Annealed vs Cold Rolled
414
17-2. Unalloyed Titanium, Grade 4: S-N Curves for Three Testing
Temperatures
415
17-3. Ti-24V and Ti-32V: Stress Amplitude vs Cycles to Failure 416
17-4. Ti-5AI-2.5Sn: Effects of Notches and Types of Surface Finish
417
17-5. Ti-5AI-2.5Sn and Ti-6AI-4V: Fatigue Crack Growth Rates
418
17-6. Ti-6AI-6V-2Sn: Effects of Machining and Grinding 419
17-7. Ti-6AI-6V-2Sn (HIP): S-N Curves for Titanium Alloy Powder Consolidated
by HIP
420
17-8. Ti-6AI-6V-2Sn (HIP): S-N Curves for Annealed Plate vs HIP
421
17-9. Ti-6AI-2Sn-4Zr-2Mo: Bar Chart Presentation on Effects of Machining and
Grinding 422
17-10. Ti-6AI-2Sn-4Zr-2Mo: Constant-Life Fatigue Diagram
423
17-11. Ti-6AI-2Sn-4Zr-6Mo: Low-Cycle Axial Fatigue Curves
424
17-12. Ti-8Mo-2Fe-3AI: S-NCurves; 'Solution Treated and Aged Condition
425
17-13. Ti-IOV-2Fe-3AI: S-N Curves; Notched vs Unnotched Specimens in Axial
Fatigue 426
17-14. Ti-IOV-2Fe-3AI and Ti-6AI-4V: Comparison of Fatigue Crack Growth
Rates
427
17-15. Ti-IOV-2Fe-3AI: S-N Curve; Notched Bar Fatigue Life for a Series of Forgings
Compared With Ti-6AI-4V Plate
428
17-16. Ti-I3V-IICr-3AI: Constant-Life Fatigue Diagrams
429
17-17. Ti-6AI-4V: Effect of Condition and Notches on Fatigue Characteristics 430
17-18. Ti-6AI-4V: Effect of Direction on Endurance
431
17-19. Ti-6AI-4V: Effect of Isothermally Rolled vs Extruded Material on Cycles to
Failure 432
17-20. Ti-6AI-4V: Comparison of Wrought vs Isostatically Pressed Material for Cycles
to Failure 433
17-21. Ti-6AI-4V: Effect of Fretting and Temperature on Cycles to Failure
434
17-22. Ti-6AI-4V (Beta Rolled): Effect of Finishing Operations on Cycles to
Failure 435
17-23. Ti-6AI-4V: Effect of Yield Strength on Stress-Life Behavior 436
17-24. Ti-6AI-4V: Effect of Stress Relief on Cycles to Failure
437
17-25. Ti-6AI-4V: Interrelationship of Machining Practice and Cutting Fluids on Cycles to
Failure 438
17-26. Ti-6AI-4V: Relative Effects of Machining and Grinding Operations on Endurance
Limit
439
17-27. Ti-6AI-4V: Effects of Various Metal Removal Operations on Endurance
Limit
440
17-28. Ti-6AI-4V: Effect of Texture on Fatigue Strength 441
17-29. Ti-6AI-4V: Effect of Complex Texture on Cycles to Failure 422
17-30. Ti-6AI-4V: Effect of Texture and Environment on Cycles to Failure 443
17-31. Ti-6AI-4V: Fatigue Crack Growth Rates
444
17-32. Ti-6AI-4V: Fatigue Crack Growth Rates for ISR Tee, and Extrusions 445
17-33. Ti-6AI-4V: Fatigue Crack Growth Rates
446
17-34. Ti-6AI-4V: Effect of Final Cooling on Fatigue Crack Growth Rates
447
17-35. Ti-6AI-4V: Effect of Dwell Time on Fatigue Crack Growth Rates
448
17-36. Ti-6AI-4V: Fatigue Crack Growth Data
449
17-37. Ti-6AI-4V P / M: Comparison of HIP'd Material With Alpha-Beta Forgings for
Cycles to Failure 450
xv
Contents
17-38. Ti-6AI-4V PI M: Comparisons of HIP'd Material With Annealed Plate for Cycles
to Failure
45 I
452
17-39. Ti-6AI-4V P/M: Effect of Powder Mesh Size on Fatigue Properties
17-40. Ti-6AI-4V P/M: Comparison of Blended Elemental, Prealloyed and Wrought
Material for Effect on Cycles to Failure
453
454
17-41. Ti-6AI-4V: P/M Compacts vs 11M Specimens: Cycles to Failure
17-42. Ti-6AI-4V: Comparison of Specimens Processed by Various Fabrication Processes
for Cycles to Failure
455
456
17-43. Ti-6AI-4V: Comparison of Fatigue Crack Growth Rate, PI M vs II M
17-44. Ti-6AI-4V: Base Metal vs SSEB-Welded Material for Cycles to Failure
457
17-45. Ti-6AI-4V: Base Metal vs SSEB-Welded Material for Cycles to Failure
458
17-46. Ti-6AI-4V EB Weldments: Base Metal Compared With Flawless Weldments
459
17-47. Ti-6AI-4V EB Weldments: Effects of Porosity on Cycles to Failure
460
17-48. Ti-6AI-4V Gas Metal-Arc Weldments: Effects of Porosity on Cycles to
Failure 461
17-49. Ti-6AI-4V: Unwelded vs Electron Beam Welded Material for Cycles to
Failure
462
463
17-50. Ti-6AI-4V: S-N Diagram for Laser-Welded Sheet
464
17-51. Ti-6AI-4V (Cast): S-N Diagram for Notched Specimens
465
466
(For other data on steel castings see Sections 3,4 and 5, on carbon and
alloy steels.)
19-1. Steel Castings (General): Effect of Design and Welding Practice on Fatigue
Characteristics
466
19-2. Steel Castings (General): Effects of Discontinuities on Fatigue
Characteristics
467
468
468
469
469
470
21-2. PI M: Relation of Fatigue Limit to Tensile Strength for Sintered Steels
21-3. PI M (Nickel Steels): As-Sintered vs Quenched and Tempered for Cycles to
Failure
471
2 I-4. PI M (Nickel Steels): Relation Between Fatigue Limit and Tensile Strength for
Sintered Steels
472
21-5. P/M (Nickel Steels): Effect of Notches on Cycles to Failure for the As-Sintered
Condition 473
21-6. PI M (Nickel Steels): Effect of Notches on Cycles to Failure for the Quenched and
Tempered Condition
474
21-7. P/M (Low-Carbon, 1-5%Cu): Effects of Notches and Nitriding on Cycles to
Failure
475
2 I-8. PI M (Sintered Iron, Low-Carbon, No Copper): Effect of Density and Nitriding on
Cycles to Failure
476
21-9. P/M: Effect of Nitriding on Ductile Iron and Sintered Iron (3%Cu) for Cycles to
Failure 477
478
22-1. Brass/ Mild Steel Composite: Comparison of Brass-Clad Mild Steel With Brass and
Mild Steel for Cycles to Failure
478
22-2. Stainless Steell Mild Steel Composite: Comparison of Stainless-Clad Mild Steel
With Stainless Steel and Mild Steel for Cycles to Failure
479
xvi
Contents
480
23-1. Carbon and Alloy Steels (Seven Grades): Effects of Nitrocarburizing on Fatigue
Strength
480
23-2. Carbon and Alloy Steels (Seven Grades): Effects of Tufftriding on Fatigue
Characteristics
481
23-3. Carbon and Alloy Steels (Six Grades): Effects of Nitriding on Fatigue
Strength 482
23-4. Carbon-Manganese Steel: Effects of Nickel Coating on Fatigue Strength 483
Contents
xvii
Fatigue Testing
Introduction
Fatigue is the progressive, localized, permanent structural change that occurs in materials
subjected to fluctuating stresses and strains that
may result in cracks or fracture after a sufficient
number of fluctuations. Fatigue fractures are
caused by the simultaneous action of cyclic
stress, tensile stress and plastic strain. If anyone
of these three is not present, fatigue cracking will
not initiate and propagate. The cyclic stress
starts the crack; the tensile stress produces crack
growth (propagation). Although compressive
stress will not cause fatigue, compression load
may do so.
The process of fatigue consists of three stages:
Initial fatigue damage leading to crack nucleation and crack initiation
Progressive cyclic growth of a crack (crack
propagation) until the remaining un cracked
cross section of a part becomes too weak to
sustain the loads imposed
Final, sudden fracture of the remaining
cross section
Fatigue cracking normally results from cyclic
stresses that are well below the static yield
strength of the material. (In low-cycle fatigue,
however, or if the material has an appreciable
work-hardening rate, the stresses also may be
above the static yield strength.)
Fatigue cracks initiate and propagate in regions where the strain is most severe. Because
most engineering materials contain defects and
thus regions of stress concentration that intensify
strain, most fatigue cracks initiate and grow
from structural defects. Under the action of cyclic loading, a plastic zone (or region of deformation) develops at the defect tip. This zone of high
deformation becomes an initiation site for a fatigue crack. The crack propagates under the applied stress through the material until complete
fracture results. On the microscopic scale, the
most important feature of the fatigue process is
nucleation of one or more cracks under the influ-
Fatigue Testing
1100
.;
Smax)'
Smax I---~:------;...----r---,----
Time
Ol-\-.........,~-----Ic-I-'--~r--f--+----
Smin
L..-
--""-
......._ - ' -_ _
S-N Curves. The results offatigue crack initiation tests usually are plotted as maximum stress,
minimum stress, or stress amplitude to number
of cycles, N, to failure using a logarithmic scale
for the number of cycles. Stress is plotted on
either a linear or a logarithmic scale. The result-
I
150
2340 steel
o~ '"'
48 HRC
~ (unnotchedl'r
1000
<ll
900
:2:
800
a..
<I)"
700
Q)
600
0.
<ll
<Il
<Il
U5
500
400
300
200
125 'iii
1::-- - - -
f-
_.
'1...
.:.!.
1 --
Fatigue limit Sf
<I)"
-cr....
Fatigue limit, Sf
l
Aluminum alloy
"""Q; ~ ,}075-T6
j-a~-
75
0.
E
<ll
50
<Il
<Il
~,
U5
25
Introduction
rather than with controlled load or stress cycles.
Under controlled strain testing, fatigue life behavior is represented by a log-log plot of the total
strain range, dE, versus the number of cycles to
failure (Fig. 3).
The total strain range is separated into elastic
and plastic components. For many metals and
alloys, the elastic strain range, dE eo is equal to the
stress range divided by the modulus of elasticity.
The plastic strain range, dE p' is the difference between the total strain range and the elastic strain
range.
Stress-Concentration Factor. Stress is concentrated in a metal by structural discontinuities,
such as notches, holes, or scratches, which act as
stress raisers. The stress-concentration factor,
K" is the ratio of the area test stress in the region
of the notch (or other stress concentrators) to the
corresponding nominal stress. For determination of K" the greatest stress in the region of the
notch is calculated from the theory of elasticity,
or equivalent values are derived experimentally.
The fatigue notch factor, Kf> is the ratio of the
fatigue strength of a smooth (unnotched) specimen to the fatigue strength of a notched specimen at the same number of cycles.
Fatigue notch sensitivity, q, for a material is
determined by comparing the fatigue notch factor, K J, and the stress-concentration factor, K"
for a specimen of a given size containing a stress
concentrator of a given shape and size. A common definition of fatigue notch sensitivity is q =
(KJ - l)f(K, - 1), in which q may vary between
zero (where K J = 1) and 1 (where KJ = K,). This
value may be stated as percentage.
10- 3
L..-_ _--I.
10- 1
..J
-l..
--L.
103
10
105
Cycles to failure
Fig.3 Typical plot of strain range versus
cycles-to-failure for low-cycle fatigue
cling of notched specimens that have been precracked in fatigue. Crack length is measured as a
function of elapsed cycles, and these data are
subjected to numerical analysis to establish the
rate of crack growth, da l d N,
Crack growth rates are expressed as a function
of the crack tip stress-intensity factor range, dK.
The stress-intensity factor is calculated from expressions based on linear elastic stress analysis
and is a function of crack size, load range, and
cracked specimen geometry. Fatigue crack growth
data are typically presented in a log-log plot of
da/dNversus s (Fig. 4).
:J.K. ksivTn.
5.
10
15
20
25
30 35
1.-----r-"T"T"--r--r-~--,.,..._,_-,-,
10- 1 I---+_--+----;:--et-'~-__l
10- 3
Q)
10- 2
>-
c::
10- 4 <:"tl
10- 3
~
10 -5
10-4
I-~=---+_--+----+---__l
o~
10- 5
10
20
30
10- 6
40
J.K. MPa\ m
Fig. 4 Fatigue crack propagation rate
data in 7075-T6 aluminum alloy (R <0)
LIVE GRAPH
Click here to view
Fatigue Testing
Fatigue Cracking
Fatigue cracks normally result from cyclic
stresses that are below the yield strength of the
metal. In low-cycle fatigue, however, the cyclic
stress may be above the static yield strength, especially in a material with an appreciable workhardening rate. Generally, a fatigue crack is initiated at a highly stressed region of a component
subjected to cyclic loading of sufficient magnitude. The crack then propagates in progressive
cyclic growth through the cross section of the
part until the maximum load cannot be carried,
and complete fracture results.
Crack Nucleation. A variety of crystallographic features have been observed to nucleate
fatigue cracks. In pure metals, tubular holes that
develop in persistent slip bands, slip band extrusion-intrusion pairs at free surfaces, and twin
boundaries are common nucleation sites. Grain
boundaries in polycrystalline metals, even in the
absence of inherent grain boundary weakness,
are crack nucleation sites. At high strain rates,
this appears to be the preferred site. Nucleation
at grain boundaries appears to be a geometrical
effect, whereas nucleation at twin boundaries is
associated with active slip on crystallographic
planes immediately adjacent and parallel to the
twin boundary.
The foregoing processes also occur in alloys
and heterogeneous materials. However, alloying
and commercial production practices introduce
segregation, inclusions, second-phase particles,
and other features that disturb the structure. All
Classification of Fatigue
Testing Machines
Fatigue test specimens are primarily described
by the mode of loading:
Testing machines, however, may be universaltype machines that are capable of conducting all
of the above modes ofloading, depending on the
fixturing used.
Fatigue Testing
Ia-...
IU
r--..... ......
Q.
'"
"1
<Ii
01
10- 1
r-, <,
........ --,
~
~
C
1Il
.~
1Il
<II
a:::
10-
"- ..........
~ ~ O.n
00
~ r-,
10
.......
"
Cycles to failure, N
Fig. 6
Low-cycle fatigue curve (t1p versus N) for type 347 stainless steel
{a}
{e}
{bl
{dl
lei
Ifl
(g)
(a) Standard grip body for wedge-type grips. (b) V-grips for rounds for use in standard grip body. (c) Flat grips for
specimens for use in standard grip body. (d) Universal open-front holders. (e) Adapters for special samples (screws,
bolts, studs, etc.) for use with universal open-front holders. (f) Holders for threaded samples. (g) Snubber-type wire
grips for flexible wire or cable.
Fig. 8
Grips. Proper gripping is not simply the attachment of the test specimen in the load train.
Grip failure sometimes occurs prior to specimen
failure. Frequently, satisfactory gripping evolves
after specimen design development. Care must
be taken in grip design and specimen installation
in the grips to prevent misalignment. The grips
shown in Fig. 8 are typical of those used for axial
fatigue tests.
Axial (Direct-Stress)
Fatigue Testing Machines
The direct-stress fatigue testing machine subjects a test specimen to a uniform stress or strain
through its cross section. For the same cross section, an axial fatigue testing machine must be
able to apply a greater force than a static bending
machine to achieve the same stress.
Electromechanical systems have been developed for axial fatigue studies. Generally, these
are open-loop systems, but often have partial
closed-loop features to continuously correct
mean load.
In crank and lever machines, a cyclic load is
applied to one end of the test specimen through a
deflection-calibrated lever that is driven by a
variable-throw crank. The load is transmitted to
the specimen through a flexure system, which
provides straight-line motion to the specimen.
The other end of the specimen is connected to a
hydraulic piston that is part of an electrohydraulically controlled load-maintaining system that
senses specimen yielding. This system automatically and steplessly restores the preset load
through the hydraulic piston.
Servohydraulic closed-loop systems offer optimum control, monitoring, and versatility in fatigue testing systems. These can be obtained as
component systems and can be upgraded as required. A hydraulic actuator typically is used to
apply the load in axial fatigue testing.
Electromagnetic or magnetostrictive excitation is used for axial fatigue testing machine
drive systems, particularly when low-load amplitudes and high-cycle fatigue lives are desired in
short test durations. The high cyclicfrequency of
operation of these types of machines enables testing to long fatigue lives (> 108 cycles) within
weeks.
Bending Fatigue Machines
The most common types of fatigue machines
are small bending fatigue machines, In general,
these simple, inexpensive systems allow laboratories to conduct extensive test programs with a
low equipment investment.
Cantilever beam machines, in which the test
specimen has a tapered width, thickness, or diameter, result in a portion ofthe test area having
uniform stress with smaller load requirements
than required for uniform bending or axial fatigue of the same section size.
Rotating Beam Machines. Typical rotating
beam machine types are shown in Fig. 9. The
R. R. Moore-type machines (Fig. 9a) can operate
up to 10000 rpm. In all bending-type tests, only
the material near the surface is subjected to the
maximum stress; therefore, in a small-diameter
specimen, only a very small volume of material is
under test.
Torsional Fatigue Testing Machines
Torsional fatigue tests can be performed on
axial-type machines using the proper fixtures if
the maximum twist required is small. Specially
Fatigue Testing
~Load
(b)
(a)
(a) Four-point loading R.R. Moore testing machine. (b) Single-end rotating cantilever testing machine.
- - - - - Torque feedback - -
---I
I
I
L----
Angular
display feedback
Hydraulic
power
supply
I
I
...-----i----r--....,
I
I
Specimen
Hydraulic
service
manifold
Displacement
transducer
.,.,C:::}:;;:;;;:;;;:;:l:::;~~,.
designed torsional fatigue testing machines consist of electromechanical machines, in which linear motion is changed to rotational motion by
the use of cranks, and servo hydraulic machines,
in which rotary actuators are incorporated in a
closed-loop testing system (Fig. 10).
Special-Purpose Fatigue
Testing Machines
To perform fatigue testing of components that
are prone to fatigue failure (gears, bearings, wire,
etc.), special devices have been used, sometimes
as modifications to an existing fatigue machine.
Wire testers are a modification of rotating beam
machines, in which a length of the test wire is
( ~=====-t-$
~R
30 mm (13/16 in.)"]
50 ~m_1
-$-O~2m.)
Tapered D, 12.7 mm (0.50 in.)
Ib)
I'" 90
mm (3:6
i ~ 19
mm (% in.)
~~-R~
'\ I
12 mm (0.48 in.)
:cD
[c]
38 mm (1V2 in.)
~---~
~3-~
~
.
5"43'
38 mm (1'12 in.)
(d)
R, 75 to 250 mm (3 to 10 in.)
Ie)
(a)Torsional specimen. (bl Rotating cantilever beam specimen. (c) Rotating beam specimen. (d) Plate specimen for
cantilever reverse bending. Ie) Axial loading specimen.
Fig. 11
10
Fatigue Testing
cent ration factor, K,(see the Introduction to this
Section). In some situations, values of K,can be
calculated using the theory of elasticity, or can be
measured using photoelastic plastic models.
The effect of notches on fatigue strength is determined by comparing the S-N curves of notched
and unnotched specimens. The data for notched
specimens usually are plotted in terms of nominal stress based on the net cross section of the
specimen. The effectiveness of the notch in decreasing the fatigue limit is expressed by the
fatigue-notch factor, K p This factor is the ratio
of the fatigue limit of unnotched specimens to the
fatigue limit of notched specimens.
For materials that do not exhibit a fatigue
limit, the fatigue-notch factor is based on the fatigue strength at a specified number of cycles.
Values of KJhave been found to vary with (1)
severity of the notch, (2) type of notch, (3) material, (4) type of loading, and (5) stress level.
Fatigue limit
MPa
ksi
7.6
38
152
248
200
144
0.30
l.50
6.00
36
29
21
11
~
~
~
600
r~~;;::~~;;;;t;;~==~===~~=~===t===~==~
100 'iii
~--_t_---~2.......,_1_--~~=_---=""""
90 -""
vi
.......= _ - _ + - - _ 1 _ - - - - _ t _ - _ _ l
500
W ~
70
~
~
>
60
400
Q)
50
>
~
>
~ 300 t-----t-----t---+----+----+--"""""~--~""""""----+--___I
40
~
~
Qi
Qj
~
U;
~
200 I----t----+---+---+-----+---+---+--~.,___f-----~
10
20
50
100
200
30
500
Effect of surface conditions on the fatigue properties of steel (302 to 321 HB)
Fig. 13 represent the three most widely used empirical relationships for describing the effect of
mean stress on fatigue strength. The straight line
joining the alternating fatigue strength to the
tensile strength is the modified Goodman law.
Goodman's original law included the assumption that the fatigue limit was equal to one third
of the tensile strength; this has since been generalized to the relation shown in Fig. 13, using the
fatigue strength as determined experimentally.
Stress Amplitude. Because stress amplitude
varies widely under actual loading conditions, it
is necessary to predict fatigue life under various
stress amplitudes. The most widely used method
of estimating fatigue under complex loading is
provided by the linear damage law. This is a hypothesis first suggested by Palmgren and restated by Miner, and is sometimes known as
Miner's rule.
The assumption is made that the application
of n.cycles at a stress amplitude S;, for which the
average number of cycles to failure is N;, causes
an amount of fatigue damage that is measured by
the cumulative cycles ratio n;/N;, and that failure will occur when "'i.(n;/ N;) = 1.
This method is not applicable in all cases, and
numerous alternative theories of cumulative linear damage have been suggested. Some considerations of redistribution of stresses have been
clarified, but there is as yet no satisfactory approach for all situations.
Fatigue strength, S
CI)"
vi
~
~
OJ
C
Gerber's parabola
Mean stress, Sm
As shown by the modified Goodman line. Gerber's parabola. and Soderberg line. See text for discussion.
The effect of varying the stress amplitude (linear damage) can be evaluated experimentally by
means of a test in which a given number of stress
cycles are applied to a test piece at one stress amplitude. The test is then continued to fracture at a
different amplitude. Alternatively, the stress can
be changed from one stress amplitude to another
at regular intervals; such tests are known as
block, or interval, tests. These tests do not simulate service conditions, but may serve a useful
purpose for assessing the linear damage law and
indicating its limitations.
12
Fatigue Testing
Corrosion Fatigue
20
10
100
q
I
I
a
~
6~
t----
da
dN
~ C(~Kln
I~
dO
: W
'0
Ie?
if
sx;
I
5
~~o
I
I
I
'I
.9' r;~
i
Region 1:
slow crack
I- growth
Fatigue Crack
Propagation
fil.
50
Region 3:
rapid
unstable
crack
growlh
I
rI
r--
I
I
1
I,c
1;1
I,j>
I
Fatigue Crack
Propagation Test Methods
The general nature of fatigue crack propagation using fracture mechanics techniques is
summarized in Fig. 14. A logarithmic plot of the
crack growth per cycle, da/ dN, versus the stressintensity factor range, I:!..K, corresponding to the
load cycle applied to a specimen is illustrated.
The da/ dN versus I:!..K plot was constructed of
20
10
30
40
50 60
80 100
= C(I:!..K)"
13
1O-8(~K)2.25
102
10
10- 2
o
o
Ql
"0
>
E
E
12 Ni steel
10 Ni steel
HY-80 steel
HY-130 steel
Ql
10- 4 "0
>
~
c::
10- 3
......:
<:
<:
~
10- 5
2:!
s:
s:
10- 4
Ol
Ol
-"
0
-"
0
10- 6
~
U
co
L.
10- 5 t--t-l~tf--+--+--+----I
2 x 10- 6
'---_..I..-_..L-_..I..-_....L_--L--'
10
20
50
100
200
14
Fatigue Testing
Two holes W/3 diam
-r
W
~----+-----J_l
!I !
I
!I
I
specimen thickness.
Fig. 16 Standard center-cracked tension specimen for fatigue crack propagation testing when the width (WI of the
specimen ';;;;75 mm (3 in.)
< 0.60
~ > 0.60
W
Fatigue crack growth rate data can be calculated by several methods. The most commonly
used methods, however, are the secant and incremental polynomial methods. The secant method consists of the slope of the straight line connecting two adjacent data points. This method,
although simpler, results in more scatter in measured crack growth rate.
The incremental polynomial method fits a
secon~-order ,Polynomial expression (parabola)
to typically five to seven adjacent data points,
and the slope of this expression is the growth
rate. The incremental polynomial method eliminates some of the scatter in growth rate that is
inherent in fatigue testing.
Numerous relationships have been generated
to correlate crack growth rate and stress-intensity
data. The most widely accepted relationship is
that proposed by Paris. This is a linear relationship when plotted on log-log coordinates and
generally yields a reasonable fit to the data in
Region II (see Fig. 14) of the crack growth
regime.
Other relationships based on the Paris equation, such as the commonly used Forman equatl?n, are used to represent the variation of da/ dN
with other key variables, including load ratio, R
and the critical K value, K" at which rapid frac~
ture of the specimen occurs (Region III in Fig.
14). The Forman equation is:
da =
C(6.K)"
dN (1 - R)(K,. - 6.K)
15
0.25Wdiam
t
t
0.6W
0.275W
t
t
0.275W
0.6W
~~_1_
1 + - - - a ----;~
( - + - - - - - - w-------;~I
)
1
an
= 0.20W
~------1.25W
-------:;~I
Fig. 17 Standard compact-type specimen for fatigue crack propagation testing (see Fig.
16 for explanation of symbols)
55
2.2
(
50
E
E 45
.r=
40
~ 35
u
1.6
(J
00
20
125
Oo c
150
1.0
175
200
225 250
<iI
.><
1.2
30
25
s:
0,
1.4 Q)c
'rP0a- J
.><
1.8
",,0
(J
Cyclic crack growth rate testing in the lowgrowth regime (Region I in Fig. 14) complicates
acquisition of valid and consistent data, because
the crack growth behavior becomes more sensitive to the material, environment, and testing
procedures under this regime. Within this regime, the fatigue mechanisms of the material that
slow the crack growth rates are more significant.
The precise definition of the cyclic crack
growth rate threshold, ~K'h' varies significantly.
i~'
OJ
2.0
275
300
Cycles. 10 N
Fig. 18 Crack growth versus constantamplitude stress cycles for a Fe-10Ni8Co-1 Mo high-strength steel
...
ro
U
16
Fatigue Testing
The most accurate definition would be the stressintensity value below which fatigue crack growth
will not occur. It is extremely expensive to obtain
a true definition of IJ.K,h, and in some materials a
true threshold may be nonexistent. Generally,
designers are more interested in the near-threshold regime, such as the IJ.Kthat corres~onds to a
fatigue crack growth rate of 10-8 to 10- 0 tn] cycle
(3.9 X 10-7 to 10-9 in.j cycle). Because the duration of the tests increases greatly for each additional decade of near-threshold data (10- 8 to 10-9
to 10-10, etc., m/cycle), the precise design requirements should be determined in advance of
the test.
Short cracks
that join long
crack behavior
q/;/"
//
\',
./
\---
sr
I:!.K=
Ii V
2W
sec
tra
2a; expression
. va1'"
2a 0.95
where a = id 10r-<
I:!.P(2
+ a)
ru;
3/2 (0.886 + 4.64a - 13.32a
By W(l-a)
+ 14.Tl 3 -
5.6a )
. va liId f or-
a 0.2
where a = -a ; expression
Crack-Length
Measurement Techniques
Several different techniques have been developed to monitor the initiation, growth, and instability of cracks, including optical (visual and
photographic), electrical (eddy current and resistance), compliance, ultrasonic, and acoustic
emission monitoring techniques.
Optical Crack Measurement Techniques
Monitoring of fatigue crack length as a function of cycles is most commonly conducted visually by observing the crack at the specimen surfaces with a traveling low-power microscope at a
magnification of 20 to 50X. Crack-length measurements are made at intervals such that a
nearly even distribution of da] dN versus ~K is
achieved. The minimum amount of extension be-
17
18
Fatigue Testing
p
t
(a)
~--al~
p
(b)
(a) C(aD)
19
Displacement
gage
Load
a, a 1 a2
-+----... )
....
'Lr
x-v recorder
Specimen
Gage
condition
/P'~
J=1
Load
cell
I
10 V dc
Load cell
condition
10 V dc
Fig. 22
The required sensitivity of the systems depends on specimen geometry and size; in general,
noise-free, amplified output on the order of I V
dc per I mm (0.04 in.) of deflection is satisfactory. Similarly, for the load range applied to the
specimen, an approximately I V de change in
signal from the load cell is required for accurate
calculation of the compliance.
Electric Potential
Crack Monitoring Technique
The electrical potential, or potential drop,
technique has gained increasingly wide acceptance in fracture research as one of the most accurate and efficient methods for monitoring the
initiation and propagation of cracks. This method relies on the fact that there will be a disturbance in the electrical potential field about any
discontinuity in a current-carrying body, the
magnitude of the disturbance depending directly
on the size and shape of the discontinuity.
For the application of crack growth monitoring, the electric potential method entails passing
a constant current (maintained constant by external means) through a cracked test specimen
and measuring the change in electrical potential
across the crack as it propagates. With increasing
crack length, the uncracked cross-sectional area
of the test piece decreases, its electrical resistance
increases, and thus the potential difference between two points spanning the crack rises. By
monitoring this potential increase, Va' and comparing it with some reference potential, Vo , the
crack length to width ratio, a/ W, can be determined through the use of the relevant calibration
curve for the particular test piece geometry
concerned,
Crack Growth Studies. By far the most useful
application of the electrical potential method has
been in measurements of crack length during
crack propagation, where it has been utilized to
monitor almost all mechanisms of subcritical
crack growth and most notably to follow fatigue
crack growth. Typical crack propagation rates
derived from direct current potential measurements are shown in Fig. 23 for tests on a 2.25CrIMo steel in air, gaseous hydrogen, and hydrogen sulfide environments.
Electromechanical
Fatigue Testing Systems
The primary function of electromechanical fatigue testers is to apply millions of cycles to a test
piece at oscillating loads up to 220kN (50000 lbf)
to investigate fatigue life, or the number of cycles
to failure under controlled cyclic loading conditions. Variables associated with fatigue-life tests
are frequency of loading and unloading, amplitude of loading (maximum loads and minimum
loads), and control capabilities. The fundamental data output requirement is the number of cycles to failure, as defined by the application,
A variety of electromechanical fatigue testers
have been developed for different applications.
Forced-displacement, forced-vibration, rota-
20
Fatigue Testing
Table 3
Parameter
Forced displacement
Tension
Compression
Reverse stress
Bending
Frequency range
Load range
Forced vibration
.
Yes
Yes
.
.
Yes
.
Yes
.
Fixed
. Typically < 450 N
1001bf)
Type:
Control
Mode
Open-loop
Displacement
.
.
Maximum deflection
Advantages
Yes
Yes
Yes
Yes
Fixed, 1800 rpm
Up to 220 kN
(500001bf)
Open-loop
Load
Simple, straightforward
No load control, very
limited applications
(soft samples)
Disadvantages
'"
0
l>
Environment
Moist air
Dry hydrogen
Air
Dry hydrogen
Hydrogen sulfide
Molstair
Dry hydrogen
::
I~rt
V~
I 'I II ,
20
Frequency. Hz
50
50
5
5
5
0.05
0.05
0.1
0.1
0.1
0.75
0.75
40
~#
nO
#
00 '"
fill'
t4,f-
~
lJ
.. ..
60 10 80
I
10" ,
10-.
.
..
1)
10- 1
s:
'U
0
0
>
u
.~
- 10'
oOd9
~.'l>~ ~
rP
rPJ pO
od
!II
50
~I
I"'"
'"
i-
30
10
ern
..
Threshold :>K",
t
6
8 9 10
20
30
40
50 60 10 80 90
Fig. 23
21
Rotational bending
Resonance
Servomechanical
No
No
Yes
Yes
0-10000 rpm
Yes
Yes
Yes
Yes
40-300 Hz
Up to 180 kN
(400001bf)
Yes
Yes
Yes
Yes
0-1 Hz
Up to 90 kN
(200001bf)
Open-loop
Rotation/ bending
Closed-loop
Load
fficient, durable,
simple
.otational bending
only, limited applications
Closed-loop
Load, displacement,
strain
100 mm (4 in.)
Fully closed-loop,
high precision
Low frequency only
= Mw 2r (centrifugal) -
Ma, (inertial)
+ Kz (spring compensated)
22
Fatigue Testing
pression, bending, torsion, or reverse st~ess.
Although servo-controlled, mean-load-maintenance systems are available, the open-loop nature of the system prevents direct load measurement or control, which is characteristic of
closed-loop systems. The load applied to the
specimen is assumed to be a function of r, and a
graduated scale is provided to permit reasonably
accurate setup.
Rotational Bending Systems
Rotational bending systems effectively apply
reversed loading to the outer surface of rods or
shafts. The basic operating principle ofthe rotating beam consists of the use of a motor to rotate a
shaft of known dimensions around its longitudinal axis. By applying a known static force at the
end of the shaft, a bending moment can be applied to the test section, the outer surface of
which oscillates between tension and compression during each rotation.
The cantilevered specimen, however, is subjected to a nonuniform bending moment, which
is large at the supported end of the specimen ~nd
zero at the free end. To produce a more meanmgful uniform bending moment throughout the
tes~ piece, a specially designed tapered specimen
should be used or bending moments should be
applied to each end of the specimen. Figure 24
illustrates the rotating-beam operating mechanism and the resulting stress distribution in the
specimen.
Gage
area
Drive motor
I Tension
'----.. )
Hotation
0~:f:
Bending
mom,",
Compression
Fig.24 Schematic ofthe rotating-beam
operating mechanism and the resulting
stress distribution in the specimen
Resonance Systems
A high-speed fatigue testing system was developed by Amsler that operated at 40 to 300 Hz,
achieved high loads (up to 90 kN, or 20000 lbf),
and consumed minimal energy. It is based on a
resonant spring/mass system, in which the specimen is used, like a spring, as an integral part of
the oscillating mechanism.
The fatigue load, in the form of a sine wave, is
achieved by preloading the sample in the frame
via a complex optomechanical procedure and
dynamically loading the sample at the natural
oscillating frequency of the spring/ mass system.
The preload is maintained automatically during
the test. The dynamic load is achieved by pulsing
an electromagnet at the natural frequency of the
spring/ mass system. During resonance, the electromagnet restores any hysteresis energy lost
during the previous cycle, thereby maintaining a
constant, controllable dynamic load. Capable of
tension, compression, bending, torsional, and
reverse-stress fatigue tests, the Amsler resonant
fatigue testers were instrumental in obtaining the
vast amount of fatigue data currently available.
The resonant system is based on a similar principle, but incorporates solid-state technology to
achieve fully closed-loop control of mean and
dynamic loads. This system uses dual opposing
masses (unlike the single oscillating mass/ seismic
base of earlier systems), linked by the specimen
to achieve vibration-free resonance. A strain
gage load cell, in series with the specimen, senses
the load and automatically triggers the electromagnet to achieve self-tuning capability.
The mean load is achieved by physically moving the upper mass up or down to achieve tension
or compression, respectively; the dynamic load is
achieved by varying the width of the pulse to the
magnet beneath the lower mass. The dynamic
load, like the mean load, is electronically ma~n
tained at a preset command level through solidstate closed-loop circuitry. The remainder of the
controls and mechanisms associated with the resonant fatigue system maintain a preset air gap
between the magnet and the oscillating lower
mass, maintain preset loading conditions (S~1Ut
ting down at preset load levels or frequencies),
and power the electromagnet.
The high efficiency of resonant systems makes
them well suited to high-cycle fatigue tests, in
which closed-loop load control, high loads (up to
180 kN, or 40000 lbf), low power consumption
(around 750 W maximum for closed-loop sys-
23
1-----,---- c
(bl
(a) Typical components. (b) Transfer functions. See text for details and explanation of symbols.
Fig. 25
Servohydraulic Fatigue
Testing Systems
Servohydraulic testing machines are particularly well suited for providing the control capabilities required for fatigue testing. Extreme demands for sensitivity, resolution, stability, and
reliability are imposed by fatigue evaluations.
Displacements may have to be controlled (often
for many days) to within a few microns, and
forces can range from 100 kN to just a few newtons. This wide range of performance can be obtained with servomechanisms in general and, in
particular, with the modular concept of servohydraulic systems.
Usually, the problem of selecting the appropriate system is simply a matter of optimizing the
various components to form a system best suited
to the given testing application. In this section,
the principles underlying closed-loop servo systems are discussed briefly. In addition, the interaction between system components is illustrated,
and a brief description of their operating principles and characteristics is provided.
With any type of control system, the objective
is to obtain an output that relates as closely as
possible to the programmed input. In a fatigue
testing system, it may be desired to vary the force
on a specimen in a sinusoidal manner, at a frequency of 1 Hz over a force range of 0 to 100 kN
(0 to 22000 lbf). The only practical means to accomplish this with precision is through the use of
a negative-feedback closed-loop system. An overview of the basic principles of operation of
negative-feedback systems is provided in Fig. 25.
The blocks shown in Fig. 25(a) represent a group
oftypical components of a testing machine. The
transfer functions of each of these blocks can be
combined to produce the more simplified diagram shown in Fig. 25(b).
Placement of the switch, S I' has been added to
the diagram to permit analysis of the system
when it is open (no feedback, or an open-loop
condition) and when it is closed (providing feedback to the system). The equation governing this
simplified open-loop system is:
C= KoD
24
Fatigue Testing
resents the open-loop transfer function, and D
represents the electronic demand signal. Therefore, the output is simply proportional to the system demand if K is a constant. Unfortunately,
K is seldom a constant, because it can be influen"ced by several common system variations. The
electronic components may drift slightly, or the~r
gain may vary. The behavior of the hydraulic
components may change with tempera~ure, contamination, or wear, and the mechamcal components may vary because of thermal effects or
friction.
Servohydraulic System Components
Many commercially manufactured units are
available for each component in a typical servohydraulic testing system.
.
The programmer supplies the command signal
to the system, which is generally an analog of the
desired behavior of the controlled parameter.
For example, assume the same test conditions as
previously discussed (control the force on the
specimen in a sinusoidal manner at a frequency
of I Hz and a force range of 0 to 100 kN). In this
instance, the programmer might be set to produce an electronic signal with a sinusoidal waveform that has a frequency of I Hz and a voltage
output of 0 to 10 V. The analog is: 1 V represents
1000 N. The system can then be adjusted to produce the correct output. Any change in the programmer signal will result in a corresponding
change in the controlled parameter.
The servo-controller makes most of the adjustments necessary to optimize system performance. For example, it compares the command
signal with a signal produced by the controlled
parameter (stress or strain, for example) and relays a correction signal, if needed, to the control
device in the system (usually a flow-control
servo-valve). A servo-controller incorporates
numerous other compensatory features, such as:
must be increased to resist bending. This is generally considered necessary in the design of fatigue machines. The extra rigidity can be obtained by increasing the diameter of the support
columns or by utilizing three- or four-column
configurations.
Exceptional alignment is required of load
frames used in fatigue evaluations to minimize
undesirable bending forces. In addition, some
means is usually provided to refine the alignment
with manual adjustments when necessary. A
strain-gaged specimen can be used to make this
evaluation.
25
l-1.
27
for Steel
Schematic
S-.V curves for a material
at various
stress ratios. l!TS and IS indicate ultimate tensile
strength and yield strength, respectiveI),
in uniaxial tensile testing.
Edann.
Volume I. Proprrr~rrnnd
28
100
Fracture region
(all specimens fractured)
80
s:
rn
lii rz:
CIl
~
70
ViS:
i!
'"
CIl
0;-
:: (;
0_
.,,g
g>:;
c~
~-'
Q;
60
Finile-Iife region
(no specimens fractured)
50
II
..
40
30
0..
Infinile-lile region
Fatigue limit
20
10
O~
10
(1)
103
(1,000)
10'
(10.000)
10'
10 6
107
(100,000) (1.000.000) (10,000.000)
Source: Donald J. Wulpi, Understanding How Components Fail, American Society for Metals, Metals Park OH, 1985,P 135-137
29
30
...........
90,000
1.1
ao'tLerMcarOon.sttel,
Oil quenchedan 1 drawn'
~~
80,000
K"'!>
~,
~~e>/
10,000
l'
'~6QOOO
n
fir
J II, I,
~.
r;
....
<I)
r-,
c-J.-
<J'~
40,000
RC'-1,
'x
<Q-/'r.t>e:
" 30POO
~
~ '! ts as rolled
:l,fio..L....!. I
C
20,000 0..
~...Q'
.1
104
rr
III
Not:
<:0,"",;." ~
I v.".,c ~
i?o;C:P~
l'fo~"ea-lea
~aJfn
r- ~~n
10,000
dJ.S.Jpe,.l.
cen/-carbon 'steel
O'i''7ve",
("o$e<tand drawn
0.
.t50.000
~
'wI. 0,7quenchedolnd -
10 5
106
10'
Number of c~c1e5 for rupture.Ioq scale
108
- - - - - - - - - - - - - - - - ----,
" 0.9
"i
:--:.. 0.8
'<::
"i 0.7
o
'e 0.6
ec 0.5
e 0.4
:>
-g
w
0.3
0.2
0.1
o
Effect of steel microstructure on endurance ratio.
One of the more extensive investigations on influence of microstructure was conducted by Cazaud. The results of some of his work are summarized in the above bar
chart. His data confirm that 0.5 is a conservative number; he found ratios varying from
0.55 to 0.62 for highly tempered martensites. These data were also for steels in the
0.40% carbon range. When untempered martensite is included, the total ratio range is
from 0.26 to 0.62. Untempered 0.40% carbon martensite is about 55 HRC. Above 40
HRC, factors other than microstructure become more significant, especially nonmetallic content and residual stress.
Many believe that tempered martensite gives optimum fatigue properties. However,
much of the early work was with medium-carbon steels with intermediate hardnesses.
Only limited data are available for other structures, including low-carbon martensites.
Borik and Chapman determined the endurance limit of bainite and martensite in the
range 36 to 61 HRC. They used 5ll00, a 1.00% carbon steel. They concluded that
above 40 HRC, bainite had better fatigue properties at the same hardness than did
martensite, whereas below 40 HRC the reverse was true. They explained the results in
terms of carbide morphology and distribution. Below 40 HRC, the carbides in the
martensite are spheroidal. Above 40 HRC, the carbide associated with the bainite was
very fine and well-distributed, but below 40 HRC the carbides had a "pearlitic mode,"
which was less favorable in resisting fatigue.
Source: D. H. Breen and E. M. Wene, "Fatigue in Machines and Structures-Ground Vehicles," in Fatigue and Microstructure.
American Society for Metals, Metals Park OH, 1979,P 77
31
32
l-5.
Steel: Influence
of Derating
Factors on Fatigue
Characteristics
l-6.
Steel: Correction
Surface
Conditions
in
Factor
Bending
K, .
1.0
0.58
0.9(a)
1.0
1.0
1.0
Kd, where:
d c 0.4 in.
0.4 in. < d
5 2 in. . . .
K. . . . . . . . . .
Torsion
33
Tension
0.9
0.9
1.0
From chart above
(aI A lower value 10.06 to 0.85) may be used to
account for known or suspected undetermined
bending because of load eccentricity.
Correction
factors for surface roughness
(kJ,
type of loading (14). and poti diameter
(KJ, for
fatigue life of steel ports.
Comparative
effects of various surface
various levels of tensile strength.
Source.
Metals Handbook.9rh
Parh OH. 19% p 61
Edlrlon.
Volume
I. Properr~erand
conditions
Selcc[ia,n.
on fatigue
Ironsand
Srscls.
limit
of steels at
American
SO~ISI)
34
l-7.
Fatigue
Behavior:
Ferrous vs Nonferrous
Metals
S (stress)--N
(cycles to failure) curves. 4. ferrous
metals; B. nonferrous
metals. S, is the endurance
limit.
Traditionally.
the behavior
of a material
under conditions
of fatigue has been
studied by obtaining
the S-.lcur\es
(see above), where S is the stress and .E- is the
number of cycles to failure. For steels. in general. one obsenes a fatigue limit or endurance limit (curve A above) which represents a stress level below which the material does
not fail and can be cycled infinitely.
Such an endurance limit does not exist for nonferrous metals (curve B above). The relation between Sand ,V. it must be pointed out. is
not a single-ralue
function but serves to indicate a statistical tendency.
Up until the 196Os, almost all fatigue failures. and consequently
all the research in
the field, was confined
to moving mechanical
components
(e.g.. axles. gears, etc.).
Starting in the late 1950s. entire structures or very large structural elements (e.g.. pressure Lessels. rockets, airplane fuselages, etc.) have been studied and tested for fatigue.
This can beattributed
to the use of materials such as high-strength
alloys. together with
the advances in the fabrication
technolom.
resulting in monolithic
structures meant to
undergo high cyclic stresses in service. It is this class of materials which has shown
catastrophic
failures in fatigue. and it is for this kind of material that fracture mechanics is being applied. with considerable
success. to fatigue problems.
KumarChauls.
hlechanlcal
hle~allurg):
Prmaplesand
.Appllcaionr.
Prcm~ce-Hall.
Inc..
l-8.
Comparison
of Fatigue Characteristics:
Aluminum
Alloy
Typical
bending (R = -1) fatigue
rous and nonferrous
metals.
curves
Mild Steel vs
for fer-
Here it is noted the lack of the knee for the aluminum allo)
compared
with steel: that is. the point on the cune where the
number of cycles to failure becomes a straight line-essential11
infinity.
35
36
IOo/-----1I--+----1--+-----+~~
V>
n.
o BOI-------t---t----r-----:
o
o
Lead is often added to steels to improve machinability, although usually at the cost of a minor (usually) loss in mechanical
properties,. The interrelationship of lead additions with tensile
strength and fatigue limit is summarized in the above graph.
Source: George M. Sinclair, "Some Metallurgical Aspects of Fatigue."in Fatigue-An Interdisciplinary Approach. John J. Burke,
Norman L. Reed and Volker Weiss, Eds., Syracuse University Press, Syracuse NY, 1964. p 68
If a specimen is subjected to alternating stress (tension and compression in turn) over a range insufficient
to cause immediate fracture, gliding may occur within
some of the grains, but when the disloca tions reach a
grain-boundary they are halted, retracing their movement along the gliding-plane when the stress is reversed. If the material were ideal, it might be hoped
that the dislocations would merely move to and fro
along the plane, and that no damage would result. In
practice a large number of cycles can be withstood
without apparent damage, but in material as we know
it, slight irregularities will prevent smooth gliding indefinitely, and roughening along the original glidingplane will make movement difficult, so that gliding
will then start on another parallel plane. In the end,
bands of material will have become disorganized, and
ultimately one of two things must happen: (1) if the
stress range is low, gliding will cease altogether, the
only changes still produced by the alternating stress
being elastic, (2) if it exceeds a certain level (the fatigue
limit) the gliding will become so irregular, as to cause
separation between the moving surfaces, first locally,
producing gaps, which later will join up into cracks.
Thus above the fatigue limit (after a time which is
shorter at high stress ranges), there will be failure;
below the fatigue limit, the life, in absence of corrosion, should be indefinitely long as shown above.
In the presence of a corrosive environment the situation will be different. Disorganized atoms along a
gliding-plane may require less activation energy to
pass into a liquid than more perfectly arrayed atoms
elsewhere; certainly, while the atoms are in motion
along a gliding-plane, preferential attack may reasonably be expected even below the fatigue limit. This
means that there is no "safe stress range" within which
the life should be infinite. It is, however, convenient to
determine an endurance limit-namely, the stress
range below which the material will endure some specified number of cycles (the number must be stated).
It should be noted that, although stress-corrosion
cracking is often intergranular, corrosion-fatigue
cracks are usually transgranular, following glidingplanes inclined at such an angle as to provide high resolved shear stress. There are exceptions to both rules.
Whitwham, studying corrosion-fatigue cracks on steel,
found that, although mainly transgranular, they followed grain-boundaries for short distances, where
such boundaries chanced to run in a convenient
direction.
Source: Ulick R. Evans, An Introduction to Metallic Corrosion, 3d Edition, Edward Arnold (Publishers) Ltd and American Society
for Metals, Metals Park OH, 1982, P 160
37
38
'Copper' steel
80
:c
u
O
104
c.
100
,e
Ii;
40
'0
:>
0
:E
"'"e
e
'E
"
(/)
:c
u
105
106
107
.5
g-
lOB
ill
80
ill
40
20
104
105
106
107
lOB
:E
"'"<:
e
~
"R"."~
:>
0
60
l!!
"
104
105
106
.1
Chrome-
"
vanadium
20 steel
(/)
.,
0
104
105
106
25
104
105
106
107
\4
50
,,~
chromium
i:
107
lOB
30
104
105
106
107
"
lOB
107
35
steel
27')(, Cr.0'2')(, C
0
103
40
'E
(/)
:E
"'e"
.
~
66
c.
'0
i:1<:
:>
0
0
103
129')(, Cr,011')(, C
Tensile: 89600 Ib./sq.in
45
Ii;
,e
i:1<:
20
w
"0
e
ill
(hardened and
tempered)
014')(, C,O'9B')(, Cu
Tensile:61500 Ib./sQ.in.
g-
ii.5
,e
20
.5
60
g"c.
40
(annealed)
Tensile: 103500 Ib.lsq.in.
saline riverwater
Typical curves showing the number of cycles needed to produce fracture at different
stress ranges in absence and presence of corrosion.
Source: Ulick R. Evans, An Introduction to Metallic Corrosion, 3d Edition. Edward Arnold (Publishers) Ltd and American Society
for Metals, Metals Park OH, 1982, P 165
lOB
PR PAGAIION LIfE
I
-,---_.
I
i
1+--.,--,---,--..c--,...-,-,--.-.+---.--.---.--.,..-,-,....,.......,r+--...,.----,.---..
1
6 7 0 9 0
Source: M. Kesten and K.-F. Windgassen, "Design of Equipment to Resist Hydrogen Fatigue Service," in Current Solutions to
Hydrogen Problems in Steels, C. G. Interrante and G. M. Pressouyre, Eds., American Society for Metals, Metals Park OH, 1982,
P 390
39
40
l-1 3. Relationship
of Stress Amplitude
2Nf
Fatigue damage is caused by cyclic plastic strain, and consequently, the fatigue life
should be related to the plastic-strain amplitude. Coffin and Manson independently
proposed a relationship between the plastic-strain amplitude and the cycles to failure
of the form:
AEp/ 2 =
EJ (2NJ),
where Ej is the fatigue-ductility coefficient, 2NJ is the number of reversals to failure, and
c is the fatigue-ductility exponent. Their equation is very similar to the Basquin equation relating the elastic-strain or true-stress amplitude to the number of load reversals
to failure:
I::.E,E/2
= aa = aJ (2NJ)b
where I::.E,/2 is the elastic-strain amplitude, E is the modulus of elasticity, his the
fatigue-strength exponent, and
is the fatigue-strength coefficient. A schematic representation of these relationships and their superposition is shown in the above diagram. The summation curve is analogous to the stress-life, Wohler diagram, if the
strain amplitudes are replaced by their respective stress amplitudes. The intersection of
the Basquin and Coffin-Manson plots is normally defined as the transition between
high- and low-cycle fatigue. Consequently, the regime of low-cycle fatigue depends on
the properties (for example, the ductility) of a particular material.
a;
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure,
American Society for Metals, Metals Park OH, 1979, p 211
41
42
0.264
0.249
CD
0.232
:5!
t:I
z 0.216
0
;::: 0.200
ADDITIONAL PULSE
OBSERVED
:::>
Z
0.184
0.168
0.4 dBATIENUATION
CHANGE OBSERVED
0.152
5
10.8xl0
Typical plot of ultrasonic attenuation versus number of fatigue cycles for steel.
Joshi and Green determined the attenuation coefficient IX for longitudinal bulk
waves in aluminum and steel at 10 and 5 MHz, respectively. The measurements have
been performed in a pulse-echo mode, with the acoustic pulse reflected at the back
surface of the material. The above chart shows their results obtained on cold rolled
steel bars. The attenuation started to increase at about 6 X 105 fatigue cycles (65% of
fatigue life). At roughly 7.5 X 105 cycles (85%), an additional pulse was observed,
arriving earlier than the one reflected from the back surface.
Results are interpreted in terms of a series of microcracks being formed, probably at
the surface. As soon as the microcracks are sufficiently deep, they will change the bulk
attenuation. As soon as a macrocrack has been formed (by coalescence of microcracks), it will reflect part ofthe pulse. After that, the attenuation is primarily determined by the transmission coefficient of this single crack. Thus, the attenuation curve
(versus fatigue cycles) becomes discontinuous, as may be noticed in the above chart.
Source: O. Buck and G. A. Alers, "New Techniques for Detection and Monitoring of Fatigue Damage," in Fatigue and Microstructure, American Society for Metals. Metals Park OB, 1979. p 135
2-1. Typical S-N Curve for Low-Carbon Steel Under Axial Tension
28
26
~
1.
~ 24
I
rc)(
~ 22
't-
'-'
III
III
uJ
0::
J-
i
20
III
0
W
.J
0.
0.
18
<t:
16
14
I~
--
~
--fATIG-U-E-LlMi.:r-(REPEATEO- TEtlSION)
10
\2.
The term "fatigue" refers to the failure of metals from repetitions of stress rather than from a
single application, as occurs for example in a simple tensile test or with a brittle failure. The value
of the stress necessary to cause failure of a material from fatigue is lower than its nominal tensile
strength. For example, a sample of mild steel may have a maximum stress of 27 t.p.s.i. when
subjected to a single application ofload as in an ordinary tensile test. If, however, a stress of say
25 t.p.s.i. is applied repeatedly to the same material, failure will not take place until this has been
done a certain number of times, while at a lower stress still, the number ofload cycles required to
cause failure will be even greater. If testing is continued in this manner, a stress value will ultimately be found at which fracture will not occur, no matter how many stress repetitions are
applied. This value is known as the fatigue limit of the material. If the results from such a series of
tests are plotted, a graph such as the one above will be obtained, the curve tending to run parallel
to the abscissa after approximately IOmillion cycles (for steel), the corresponding value ofthe
stress being known as the fatigue limit. Under conditions of repeated tension the value of the
fatigue limit for the above mild steel which has a tensile strength of approximately 27 t.p.s.i.
would be of the order of 16 t.p.s.i. If the same steel was tested under conditions of reversed
bending stresses a value of the order of 12 t.p.s.i. may be found.
It must also be pointed out that where corrosive conditions operate in addition to fluctuating
stresses, failure from "corrosion-fatigue"may occur and, in these circumstances, the concept of a
fatigue limit does not apply, since if the stress applications are continued for a sufficient number
of times, ultimate failure will occur. Further, most nonferrous metals and alloys do not possess
a fatigue limit.
Source: F. R. Hutchings, "Fatigue Failure of Components of Lifting Machinery," in Failure Analysis: The British Engine Technical
Reports, F. R. Hutchings and Paul Unterweiser, Eds., American Society for Metals, Metals Park OH, 1981, P 344
43
44
..
C'l
Eoff
cP
.. 0.2
)(
..r
00.4
a 0.6
"j
:::>
f-
:::i
Do
::l!:
oct
z
a:
lii
..J
oct
f-
f-
10 3
104
loS
10 6
..
C'l
)(
Eeff
.. 0.2
00.4
..r
a 0.6
:::>
f-
:::i
Do
::l!:
oct
z
a:
f-
Vl
..J
oct
f-
f-
Runouts
r
103
104
loS
10 6
Plots in the top chart are for biaxial stretching; those in the bottom chart are for cold
rolling. Included is the data band for the undeformed material. The effect of balanced
biaxial stretching on fatigue life was as follows: at large strain amplitudes (/:;.EI/2~
~ 2.5 X 10-3 ) , the fatigue life remained approximately the same or decreased slightly
when compared to that ofthe undeformed material; in contrast, at small strain amplitudes the fatigue life increased as a result of the prior deformation.
After cold rolling, the fatigue life was approximately the same as in the undeformed
material at large strain amplitudes (short lives) but it was longer at small strain amplitudes (long lives). Thus, unlike BBS, CR appeared to cause no reduction in fatigue life
at short lives. Another difference between the two deformation modes was that the
scatter ofthe data was larger after BBS than after CR. Thus, BBS was somewhat more
detrimental to the fatigue life than CR.
Source: John M. Holt and Philippe L. Charpentier. "Effect of Cold Formingon the Strain-Controlled Fatigue Properties ofHSLA
Steel Sheets," in H'Sl.A'Stccls-c-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 217
SAE-1006
R' 0.1. t' 0.13" 13.3mml
--~--
Smooth Specimen
--<>-- TIG-Dressed
_.--fr-.- As-Welded
00 0 - - - - 0
0 roo 6_ --rrtr-lS" i:J_ _
tr
6
~D;!.~_
0
0
0
-o 0
.-._
oro
~
'-.0
0-.
o DO ~._._
._....g 0
'-''''lJ..._
'--0
10 5
..............
10 6
Source: Kon-Mei Ewing, Pei-Chung Wang. Frederick V. Lawrence, Jr., and Albert F. Houchens. "Weld Fatigue ofTlG-Dressed
SAE-98QX HSLA Steel,"in HSLA Steels-Technology & Applications. American Society for Metals, Metals Park OH. 1984,p 557
45
46
~.
or
CJ)
<J
\0 1
(a,.-32KSI
\;
ur= 32 KSI
EXPERIMENT
PREDICTION
10 5
NT
106
I
CYCLES
Total fatigue life predictions and experimental results for FCA W, AISII0061ap-shear
joints. Note that the results and predictions compare closely.
Source: Kon-Mei Ewing. Pei-Chung Wang, Frederick V. Lawrence, Jr., and Albert F. Houchens, "Weld Fatigue ofTIG-Dressed
SAE-980X HSLA Steel," in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984,p 562
47
(f)
<I
10 1
~"-'2KSI
;;;
CTr =32
a.
~
10
KSI
101
NT,
CYCLES
Total fatigue life predictions and experimental results for FCAW, AISI 1006 lap-shear
joints. Here, the prediction and actual results are very close.
Source: Kon-Mei Ewing. Pei-Chung Wang. Frederick V. Lawrence, Jr., and Albert F. Houchens, "Weld Fatigue ofTIG-Dressed
SAE-980X HSLA Steel, "in HSLA Steels-Technology &Applicalions, American Society for Metals, Metals Park OH, 1984,p 562
vi
<l
48
0.20
5.0 .
E
~
E
.x:
4.0
0.15
<,
M
E
E
<II
0-
'"
3.0
I-
:E
ui
-O.I0:J
I<l:
a:
a: 2.0
uJ
::J
'"
<l:
lLJ
::
0.05
I<l:
u,
1.0
oL--!------~---_:::_---_::::__'O
20
40
DEGREE OF COLD-WORKING,
60
Comparison of effects of cold working on wear rate and fatigue limit of fully
annealed 0.15%C mild steel. Wear was determined in sliding between the end
surfaces of cylinders at a speed of 0.56 mjs under the loads .:82 N, ():124 N
and 0:147 N in machine oil. Fatigue limit (.) was determined by reversed
bending fatigue tests of notched plate specimens 25 mm wide and 4 mm thick
having a central hole 1.5 mm in diameter.
Source: Yoshitsugu Kimura, "The Role of Fatigue in Sliding Wear," in Fundamentals of Friction and Wear of Materials, David A.
Rigney, Ed., American Society for Metals, Metals Park OH. 1981, P 215
50
10- 2
10- 4
a
<9':!
10- 3
~
E
E
~
co
...
01
00 I
1
Region 1:
slow crack
growth
I Region 3:
I rapid
I unstable
I crack
growth
I
I
I
I
I
I
10- 4
",'
1!
ie
at
eu
...e
10- 6
'"
~
.5
Z
~
co
...
ti
10- 6
l!
~
e
...u'"
l!u
...l!!
'"
:iE
10- 6
10-7
'"
:iE
10- 6
Fatigue crack growth behavior of AS33 steel. The material was ASTM
AS33 B-1 steel, with a yield strength of 470 MPa (70 ksl), Test conditions: R= 0.10; ambient room air; 24C (75 OF).
The general nature of fatigue crack growth and its description using fracture mechanics can be
briefly summarized by the example data shown in the above chart. This figure, based on the
work of Paris et al, shows a logarithmic plot of the crack growth per cycle, daj dN, versus the
stress-intensity factor range, t::.K, corresponding to the load cycle applied to a sample. The
da] dN versus t::.K plot shown is from five specimens of ASTM A533 B-1 steel tested at 24C
(75 OF). A plot of similar shape is expected with most structural alloys; the absolute values of
daj d N and t::.K are dependent on the material. Results of fatigue crack growth rate tests for
nearly all metallic structural materials have shown that the da I dN versus t::.K curves have the
following characteristics: (a) a region at low values of dald N and t::.K in which fatigue cracks
grow extremely slowly or not at all below a lower limit of t::.K called the threshold of t::.K, t::.K,,,;
(b) an intermediate region of power-law behavior described by the Paris equation:
~=C(t::.KJ'
dN
Source: J. H. Underwood and W. W. Gerberich, "Concepts of Fracture Mechanics." in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH. 1982. P 18
49
50
20
MPavm
100
100
20
10'
A514F
CON Quality
'(J '
I
/;
/,.
.,
100
10
z!.
~ 10"
<
A514F
CaT Quality
~
E
E
lS----1
u-
-::'O
15-----<
TL-- ---- .. - --- -_ :
~I=-:-:=:':::- i
l'.K. ksl\ ii1.
'00
The increased isotropy in the CaT over the CON steels is evident with the through thickness (ST, SL) orientation having the
fastest growth rate in the CON steel and showing the greatest
improvement by CaT.
Source: Alexander D. Wilson. 'The Effect of Inclusions on the Properties of Constructional Steels." in Wear and Fracture Prevention. American Society for Metals. Metals Park OH, 1981. p 196
5.10-'
CON~
A633e
CaT
A633C
CON~
A514F
caTD
1110"' 3.10-'
~~~
ICaT I
4.10'
A514F
8x10"
12110"
16.10"
20110-'
daldN, inches/cycle
Comparison of fatigue crack growth rate variation with
orientation for A633C and AS14F plates at two tJJ( levels.
These data show that the CaT improvement in FCP growth rate
takes place only at higher L1Klevels. Additionally, this figure indicates
that there is a more substantial enhancement in FCP behavior for
ASI4F. Also there generally appears to be more anisotropy in the
ASl4F steels of both quality levels. It has previously been shown that
higher strength level steels tend to be more adversely affected by inclusions associated in groups, such as present in CON steels.
Source: Alexander D. Wilson, "The Effect ofInclusions on the Properties of Constructional Steels, "in Wear and Fracture Prevention, American Society for Metals, Metals Park OH, 1981, P 197
51
52
20
:------ .....,
10. 0
I
I
100
,
,:
,
,,,
,
10"
I
I
I
I
I
~
A514F
E
E
2'/.ln(57mm)Gage
CjCON
ClCaT
10"
6 K , ksi\/fil.
Source: Alexander D. Wilson, "The Effect oflnclusions on the Properties of Constructiona ISteels,"in Wear and Fracture Prevention, American Society for Metals, Metals Park OH, 1981,p 197
MPavm
100
10-'
10-'
GI
GI
'fi 10"'
,5
4In(102mm)Gage
C=JCON
DCaT
10-'
A633C
10-'
10
,c.. K kslv'ln:
100
Source: Alexander D. Wilson,"The Effect of Inclusions of the Properties of Constructional Steels," in Wear and Fracture Prevention, American Society for Metals, Metals Park OR, 1981, P 196
53
54
80
60
r-
... 1"-1-
40
-,~
- --
---
...
uv
o VV
-in
.0/.
'."C
-
8
f-
6 fI
10'
"
data bank
.,
'flo
9-
LC?cation of failure:
o Plate or weld edge
Porosity in weld
... Slag near surface
b. Slag at midthickness
V Lack of penetration
10
r----... r-9 0
Data bank
;--I-- r-- mean
f..' curve
IA
Low-carbon steel
reinforcement off
~
ch
I,t:
. --
~-
_C
ri
if.. ...ll ...
20
on
....c
.,
OJ"
111
11
V
11
I~
1<:;1-
B 10'
10
6 8
10
Source: Metals Handbook, 9th Edition, Volume 6, Welding, Brazing, and Soldering, American Society for Metals, Metals Park
OH, 1983, P 848
-e
Reinforcement intact
Reinforcement removed
;j!.
"0'
0;
;:
"0
C
::J
(;
-----
-S
'"
40 _---""
___-~......=c-I_---~I""'-_~=__iI_---___l
----
1;;
CI>
.,
::J
.'"
'"
u,
201_----1_----1--
Ol.--
.l.-
..l-
...J......
...l-
0.2
0.4
0.6
0.8
-J
1.0
Source; Metals Handbook, 9th Edition, Volume 6, Welding, Brazing, and Soldering, American Society for Metals, Metals Park
OH, 1983, P 850
55
56
\\
80
60
\\
r-....
-,~'" <,
"
1'-,
.....
...........
.....
R . fl.
I
em orcement mtact
~I
<, ~
t--.
100
oob cycles
--.- r--r-.r------<.
.................. ~ r--
40
Reinforcement remived
.i
\/ -:
-.
-.-
o
o
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
Source: Metals Handbook, 9th Edition. Volume 6, Welding, Brazing, and Soldering, American Society for Metals, Metals Park
OH, 1983, P 849
;fi.
80
,,'
Qj
\ ~
1\
\'\
'\
III
"
<,
-g 60
'0
-S
'"
~ 40
r-,
<,
....
Reinforcement intact
I
I
'
I
- . - Reinforcement rejOved
<,
<,
"""-
" - -::::--- -~
r-._ r--
Ii:
Ql
".,'"5.
LL
--
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
Source: Metals Handbook, 9th Edition, Volume 6, Welding, Brazing, and Soldering, American Society for Metals, Metals Park
QH, 1983, P 849
57
58
80
*'
'0'
'0
5
g
60
...o
fia>
40
~~
~ \~
\ [\
<,
~ .......
I" '-,......r-
10-. ~
""'-
r-; :::::-
Cl>
:::l
t::-I-- i'---
'
r-- 1-.
a>
co
--.
"k
0.Q2
0.04
0.06
""'--.
~ t:- t--
20
a
a
';;
u..
. - Reinforcement removed
t:
. fl.
0.08
0.10
t--
00 000 cycles
0.12
0.14
0.16
Source: Metals Handbook, 9th Edition. Volume 6. Welding, Brazing, and Soldering, American Society for Metals, Metals Park
OH. 1983. P 848
'I
........
600
r ,
'1-0---
, - - - , - - r ,-,.-.-,--""--"~J.------,---r---"I---y---r-T"""T""
"-
~"""""O<UNWELDED
.0
500
ro
300
E
(/)
200
100
0 0
~<, O~
~REINFORCEMENT OFF
.~.,-.-3
~ Q....
o
__
~_
1.5
2.3 - .
* .
-~-
REINFORCEMENT
Q..
't:..
ON
3'-............
~--.
t.:
..~
j
----t--~
.........
...
h =3.8mm
r h
Tr==r?J
<,
<,
.'6~-------~ __ C
~.
~
-.r:~
CYCLES
Influence of weld reinforcement on fatigue strength (R=O) of transverse butt welds of
quenched and tempered carbonsteels, From these data it is evident that removal ofthe
reinforcement (weld dressing) improves fatigue strength and fatigue life.
Source: Drew V. Nelson, "Fatigue Considerations in Welded Structure," in Proceedings of the SAE Fatigue Conference P-109,
Society of Automotive Engineers, Inc., Warrendale PA, 1982, P 206
59
60
80
600
6Or---====:=--_
400
300
40
CT, =0
200
20
10
100~
7O:e
5O<l
vi
~ma.'2.5I,
30
OCT, = + 35 ksl
20
2
10
7
10
N1 ' Cycles
Fatigue crack initiation life predictions and experimental results for a-in. (16-mm) A36/E60S-3
butt welds.
Source: F. V. Lawrence, "The Predicted Influence of Weld Residual Stresses on Fatigue Crack Initiation," in Residual Stress for
Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981,P 114
Kfmal
31~1 DJ 0.01In.
- - . "".120kl'
---0 "" -120kli
200
200
100
-:---_-
vi
<I
---
00 0 - 0 - - - _
l.
10 ~
vi
so <I
"'----
30
20
20
10
Total fatigue life predictions and experimental results for A514F /EllO weldments with tensile and
compressive residual stresses.
Source: F. V. Lawrence, "The Predicted Influence of Weld Residual Stresses on Fatigue Crack Initiation," in Residual Stress for
Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981, P 113
61
62
8
6
100 ~
70 ::f.
50
--.,.s,
- - - ' , -0
..
s~s
K,.....
r
'"
30
20
10
NT' Cy(lu
1~'r--~--r-r""""~"'--~-~~""""~"'--~-'-"""'~~"'6oo
400
60
300
40
R'O
20
8
6
--.,-s),
- - - w,-O
~-.
s~s
K r_ .
30
20
'
10
10<
10'
NT' Cycle,
The results for the high-strength, quenched-and-tempered steels (upper chart), indicate that such
materials can sustain high residual stresses which do not relax. The total fatigue life of such materials is
strongly influenced by both residual stress (a,) and stress ratio (R). Stress relief or mechanically induced compressive residuals should be highly effective. An intermediate case is mild steel as shown in
the lower chart. Mild steels can have appreciable residual stresses; but, since the transition fatigue life
(N,,) is often very long (= 500,000 cycles), there are large amounts of plasticity at the notch root even at
long lives (106 cycles); this notch-root plasticity tends to relax rapidly the notch-root residual and mean
stresses with the result that N[is little affected forlives less than 106 cycles. The observed dependence of
N on stress ratio does, however, result in a predicted variation oftotal fatigue life with stress ratio R.
Source: F. V. Lawrence, "The Predicted Influence of Weld Residual Stresses on Fatigue Crack Initiation," in Residual Stress for
Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981, P 112
63
600
60
Zero Mean Slress
40
200
20
...;;
VI
<I
10
8
--\---====::::::::d
Mean Slress Effects With No Rela.atian
~~
100
70 2:
30
A36 Butt Weld (HAl)
KImox =3 I R-O
a r =+35 ksi
20
10
IIO~
10'
N
Cycles
Materials such as high-strength steels exhibit very little notchroot plasticity; consequently, a os may be larger than a r- The results obtained using the model agree with the experimentally
observed behavior. The above chart shows the qualitative behavior of N[ predictions.
Source: F. V. Lawrence, "The Predicted Influence of Weld Residual Stresses on Fatigue Crack Initiation," in Residual Stress for
Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981, pIli
64
10
-----
08
'--'.-,
""
~
" '"
.
- - - - f----- _._ .. _ -
"
"
i'..'..
'-'0
OB
"
"-
, .,~:,
','~
' ~~
,~
~/
..,~ ~'/~
-~/
.,~
.-:~~ /.--~-~
I~'"
V'~
%-. ~/.
..,~
'~r::-, '/8
~.,
04
f--------.--
~c
b-.,
-----_.-
A_
~.ij'/W
,-
~.
'~
~D
02
I I
I I
I I
I I
10
Fatigue strength of welded tubes: A - unwelded or welded without filler metal;
B - helical welding (700 angle); C - longitudinal or helical welding (550 or 600
angle); D - helical welding (500 angle).
Source: R. V. Salkin, "Low Cycle Fatigue of Welded Structural Steels: A Material Manufacture and Design Approach," in Proceedings of the Conference of Fatigue of Welded Structures, Vol 2, The Welding Institute, Abington Cambridge, 1971, P 193
65
.. 120
E
~
~
100
>-t--
w
o
80
~ 60
w
IX:
IX:
040
20
OL-_ _- '_ _-'--------'-_'--.l-L....L--L..J'--_ _- '_ _-'--------'-_'--L-JL..l...-'LIlk<
10'
10'
CYCLES
10'
TO FAILURE
The effects of salt concentration and temperature on the fatigue behavior of steels
have been studied. Experiments performed on mild steel specimens in distilled water
and in various concentrations of potassium chloride have shown that solutions ranging
from 2 molal to 1/40 molal have virtually identical effects on corrosion-fatigue lives,
but that at concentrations below 1/40 molal, the effect approaches that of distilled
water, although corrosion rates increase in an almost linear manner with solution ion
concentration. A similar result has been reported for deaerated 3% NaCI solution in
which corrosion rates were controlled by applied anodic currents (see above chart).
These observations indicate that a critical corrosion rate is a necessity to initiate
corrosion-fatigue failures. Additionally, increasing over-all corrosion rates over a long
range of rates has little effect on corrosion-fatigue resistance.
Source: D. J. Duquette, "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys." in
Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 344
66
55
/
I NORMAL SOLUTION
50
45
.,~ 40
Q
"'.H
=12.1
~ 35
UJ
II:
t-
'" 30
25
pH :10
20
.5
10
The effect of stress frequency on corrosion fatigue has been studied by a number of investigators but is still not completely understood. For example, an early review of corrosion fatigue
noted that it is difficult to compare the corrosion-fatigue properties of metals exposed to like
environments because data reported are usually taken at different frequencies. In general, a
given time was found to produce more damage at a higher frequency, but a given number of
cycles was found to produce greater damage at low frequencies. For low-alloy steels in fresh
water, a frequency of 1450 cycles/min produced failure in 106 cycles or II Y2 hours, but at a
frequency of 5 cycles/min, failure occurred in 0.11 X 106 cycles, or 400 hours.
To date, the effect of pH of aqueous solutions on corrosion-fatigue behavior has not received
extensive study. A study of the effect of 0.1 N HCl on the fatigue life of steels showed greater
damage in this medium than in neutral potassium chloride solutions. Tests conducted in alkaline
media, at a pH above 12.1, showed that a fatigue limit is regained, this limit improving at still
higher pH values (above chart). These investigators suggested that corrosion fatigue is a result of
differential aeration cells, which produce pits in the metal surface, and that a high pH provides
diffusion barriers (ferrous hydroxide) to oxygen on the surface. Higher fatigue limits at high pH
are explained in terms of a "better and more perfect film barrier."
Source: D. J. Duquette, "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys." in
Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 346
CJ)A
CJ)
NON-DECARBED MATERIAL
~ B ...............~
t)
~
INTRINSIC'
FOR DECARB
MATERIAL
. , ...DECARBED
-COMPOSITE
LIFE
Influence on fatigue SoN curve of soft surface
caused by decarburization.
Parts that were made from low-carbon steel, but have high-carbon surfaces resulting from carburizing, have special microstructural factors that
must be considered. From the carburizing process an intergranular oxide
network may develop. This oxide may be an alloy oxide which causes alloy
depletion in grain-boundary areas. As a rule, this condition is thought to
detract from fatigue properties. The two exceptions may be in combination
rolling and sliding contact fatigue, where the oxide network may enhance
low-cycle bending fatigue-somewhat the same as does decarburization. The
effect on high-cycle bending fatigue is deleterious, as is decarburization.
These concepts are shown schematically in the above chart.
Source: D. H. Breen and E. M. Wene, "Fatigue in Machines and Structures-Ground Vehicles," in Fatigue and Microstructure,
American Society for Metals, Metals Park OH, 1979,P 80
67
68
r------;;;:;:;;;::::::j""". 18
16
0.4
14
2
AK = 29.7 MN/m 3/
0.3
0.2
0.1
4
2
The origin of the element (such as sulfur) Onthe surface could result from
its presence in the gas phase (such as for hydrogen sulfide). It could also
originate as an enriched sulfur layer associated with a propagating crack, as
would be the case for sulfur segregated to a grain boundary. Oxygen alone on
the surface tends to drive the hydrogen-dissociation reaction rates in the opposite direction from the sulfur. The above bar chart shows how a mixture of
environments can influence the fatigue-crack growth of an alloy when all the
loading factors are kept constant.
The main influence ofthe environment is to supply the active atoms to the
vicinity of the crack tip. Subsequent interaction with the crack allows the
degradation mechanism to take place. The next step in the environmental
interaction is the transport of the active species to the location in the vicinity
of the crack tip where the degradation mechanism takes place.
Source: H. L. Marcus. "Environmental Effects 11:Fatigue-Crack Growth in Metals and Alloys," in Fatigue and Microstructure,
American Society for Metals, Metals Park OH, 1979. P 371
NONE
0.006
0.012
NONE
0.006
0.012
Source: Steel Castings Handbook. 5th Edition. Peter F. Weiser. Ed.. Steel Founders' Society of America. Rocky River OH. 1980.
P 15-29
69
70
2-28. Cast A216 (Grade WCC) Steel: Fatigue Crack Growth Rate
20
40
60 80 100
200
3r---.---.--__,_~__,_,...,.,_r_--"
lTyS : 48 ksi
2
(331 MPo)
- 4
-.I
U
-,
E
10- 6
- 8
"0
<,
0::
10- 5
0::
8
I
0::
{.?
0::
~.
'<l
I~
0
10-7{.?
0::
I<l:
"0
I-
"0
<,
"0
o.
0
<l:
0::
U
RANGE,
ksi ~
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980,
p4--16
<,
IIII
<I
--r--...,
I.O....--r--..,.....--r---r--.......
'f "''0.1
ILl
C
I-
:J
Q.
0.01
<t
Z
0:
/O,W
O\' '
0,'
'00"::0."
1)'0'
<i
PLASTIC
",
~O
0.001
"
t;
0.0001 ~_L--..l....::-~-..l...:-....l..:~....L.:=--~
7
100
10
102
106
10
REVERSALS TO FAILURE - 2N f
Low-cycle strain-control fatigue behavior of carbon steel.
A number of techniques are available for computing the lowcycle fatigue life, although a straightforward approach is simply
to compute the fatigue life from the expected cyclic plastic strain
amplitude in service. Errors in computing or estimating 6.Ep
produce a smaller change in the computed cyclic life than similar
errors in the elastic strain range. Note that there is a large difference in slopes "c" and "b" in the above diagram. Plastic strain
ranges may be computed using sophisticated finite element
techniques, estimated from simple approximations such as
Neuber's rule or experimentally measured in component or
model tests.
Source: Steel Castings Handbook, 51hEdition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980,P
4-13
71
72
50
A1MQSPHRE
NITRIDED
40
lC
'0
AT l050 F (565 C)
~
HUKBER OF CYCLES
Source: J. A. Riopelle. "Short Cycle Atmosphere Nitriding," in Source Book on Nitriding, American Society for Metals. Metals
Park OH, 1977, P 286
73
TENSILE
YIELD
STRENGTH STRENGTH ELONG. HARDNESS
ksi MPo
ksi MPo
%
BHN
CAST
WROUGHT
94 (6481
90(62~
56 (386)
56 (3B6)
25
27
187
170
350
en 50
.><:
en
J, 45
f3
g:
WROUGHT
NO
300
f3
0:
40
CAST
NOTCH
en
250 ~
~ 35
::::>
::::>
5(30
200
NOTCHED
<t
<t
~
~ 25
150
5
10
6
10
7
10
CYCLES TO FAILURE
Fatigue characteristics (S-N curves) for cast and wrought
1040 steel in the normalized and tempered condition,
both notched and unnotched. R. R. Moore rotating beam
tests, K, = 2.2.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980,
P 15-10
74
.
<l
C\l
.....
0.1
SAE 1045
or
"0
.~
Q.
0.01
C3:
c:
>-;~-
_ _ _ 1l911HB
'.:::--=----_410
..... __ - - - 3 3 0
1:1
+0
(/)
0.001
- - - - 280
2211
10
102
103
104
lOll
Reversals la Failure, 2N
10
107
Source: R. W. Landgraf. "Control of Fatigue Resistance Through Microstructure-Ferrous Alloys," in Fatigue and Microstructure, American Society for Metals, Metals Park OH. 1979.P 458
ATHOSI'HERE
NIUIDED
AT 1050 r (565 C)
40
NlHIEI OF C'l'CLlS
Source: J. A. Riopelle, "Short Cycle Atmosphere Nitriding," in Source Book on Nitriding, American Society for Metals, Metals
Park OH, 1977, P 287
75
76
\.
\.
1\ \.
\\ \
0.1
1\\'\\
600
~500~00
en
"
c;
"
t:
S
'0
200
.'\. \
Hardness, HB
<; ~ t\
.~
,,30~
-, "'-.."
0.01
n;
600
s:
~ ~ 400
<,
300<,
-200-
c:
0
0.001
0.0001
1
10
1M
1~
1~
Stress reversals to failure
1~
Predicted plots of strain versus fatigue life for typical mediumcarbon steels at the hardness levels indicated above.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH. 1978, P 673
77
60
<-,
t--
--~~p~~lf:o
~
~/lf:
1ft'! d
~O'47"
on
r--,
gj 50
o
x
lI
oil
oil
III
0:
li;
I~
~40
I"""
~
o
~
o
30
300,000
~(l'O'I7
d. r 'Y-. I
Q ...
r- u
~S-
d.2/3"~
(iQIf;
DIcJ.:2
~
{)r
D/d: 2
O/cl.ys
0+
----
I II
1.000,000
10.000,000
50,000.000
~A1
UNMODIFIED
~&j
RE-EI1TRAN, FILLET
~4
5EPARATE COLLAR
Source: G. A. Cottell, "Some Common Stress Raisers in Engineering Parts," in Failure Analysis: The British Engine Technical
Reports, F. R. Hutchings and Paul Unterweiser, Eds., American Society for Metals, Metals Park OH, 1981, P 108
78
5LEO-RUNNl:.R KEYWAY
PROFILED Kf..YWAY
70,000
PROFILED ...EV.......'(
MEDIUM-CARBON S,.EEL
(NORMALISE-D)
200.000
I ~ I
1.000.000
10,000.000
Source: G. A. Cottell, "Some Common Stress Raisers in Engineering Parts," in Failure Analysis: The British Engine Technical
Reports. F. R. Hutchings and Paul Unterweiser. Eds., American Society for Metals, Metals Park OH. 1981, p 109
+400
+200
NEAR-SURFACE
RESIDUAL
STRESS
O~-----------~-------
MPa
-200
(=
+3000 po IN/IN
TESTS STOPPED AT
10 7 CYCLES MAX.
-400
-600L---------....,...L,.-----------,J.
1.0
10
0.1
UFE CYCLES
Ie
10 6
Fully reversed fatigue tests on smooth bar specimens in medium carbon steels fully hardened show, as expected, that fatigue life increases directly with surface and near-surface residual compressive stress (see above chart). Residual stress measurements are usually made in the direction of the applied stress.
The achievement of high residual compressive stress in a part
requires a careful balance of the factors which affect this property and often involves a number of trade-offs which vary with
the application.
Source: J. Alan Burnett. "Prediction of Stresses Generated During the Heat Treating of Case Carburized Parts," in Residual Stress
for Designers and Metallurgists, Larry J. Vander Walle, Ed., American Society for Metals, Metals Park OH, 1981, P 44
79
80
211
.>:
NDR"""L.IZE ~N~
15
.4'~
:a---~
<,
-, -, ~HOT BLASi'
e;
III
i-
lJl
lil....
'4
-'
A '
Ul
ILl
NORH"LIZE -
<,
TEMP~
'"
"a...
~
~
ti
II
-5
111
4
CYCLES
III
-411 .1.--+-
+--_ _--<
--+
--+-
-+-
+--'
Ill"
CYCLES
The upper chart shows residual stresses existing on the completion of individual tests. The
similarity to S-N curves is apparent with the exception that the curve for normalized bars
(R 1 =+22 ksi) is inverted. Since the initial residual stress was known, there was a question on the
manner in which the residual stress changed during the progress oftesting. To explore this point,
two shot blasted bars were tested with applied stress levels of 40 and 55 ksi. The test on each
specimen was interrupted periodically to measure the residual stress at that time. The results are
shown in the lower chart. It is apparent that the change in residual stress is proportional to the
number of cycles when the latter is represented on a logarithmic basis. The lower chart also
points to the fact that the rate of change in residual stress is dependent on the level of applied
stress. Since the initial and final residual stress values were known for all bars, the slope for each
line could be determined.
Source: P. J. Neff. "A Quantitative Evaluation of Surface Residual Stress and Its Relation to Fatigue Performance," in Residual
Stress for Designers and Metallurgists, Larry J. Vander Walle, Ed., American Society for Metals, Metals Park OH, /981, pp 127-128
83
100
650
c,
::;;;;
600
"'"
.~ 550
u.
~ <,
500
6.
01340
.4042
_4340
05140
6. 80840
I--.I'---
'"I'--. ---
90
Vi
~
~.
:~
6.
6.
n
I--.
70
400
60
100
80
60
40
20
Martensite. %
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 676
84
160
150
140
Vi
a.
0
0
0
130
E
--'
100
Q)
120
110
90
c
~
- SAE
SAE
.a._ SAE
0 - SAE
G - SAE
Q - SAE
:J
"0
6-
c
w
50
20
30
Rockwell
40
"c"
4063
5150
4052
4140
4340
2340
50
60
Hordness
The above chart and other data can be used to show the importance of
limiting the system to low and intermediate hardnesses as well as to point out
the importance of residual stress in fatigue. These data from Garwood, Zurburg and Erickson show a very tight linear relation up to about 40 HRC.
Above that hardness, the relation deviates from linearity, seemingly depending on carbon content. Carbon, however, is in an intermediate role here,
because it affects temperability. Because response to tempering is dependent
on carbon and alloy levels, it was necessary for samples of different grades to
be tempered at different temperatures to achieve the same hardness; consequently, a variety of residual-stress conditions resulted. The tempering
temperatures were necessarily sufficiently high to obtain 40 HRC; the residual stresses were reduced to a very low level, making all samples similar in
that usually the tensile strength for small sections decreases with increasing
section size and I or decreasing hardenability to compressive values. The sequence of transformation from surface to center, together with the temperature gradients, governs the outcome.
Source: D. H. Breen and E. M. Wene, "Fatigue in Machines and Structures-Ground Vehicles," in Fatigue and Microstructure,
American Society for Metals, Metals Park OH, 1979, P 73
Steel
BOO
100
s:
::;:
.~
.>!
600
BO
5,
.~
5,
60 "g
400
u,
u,
40
200
4140
Hardness, HRC 30
X4340
32
4027
44
4063
46
4032
4B
Avg
tensile
No. of strength Hardness,
tests(a) MPa ksl HRC
Longitudinal
4027
4063
4032
Tests
11 1179 171
12 1682244
11 1627236
Transverse Tests
4027
10
4063
9
4032
10
37 to 39
47 to 48
46 to 48
It must always be considered that in rolled steels fatigue behavior is affected significantly by specimen orientation. Shown
above is the effect of orientation relative to fiber axis resulting
from hot working on the fatigue limit of low-alloy steels.
Through hardened and tempered specimens, 6.3 mm (0.250 in.)
in diameter, were taken from production billets. Specimens for
each grade were from the same heat of steel, but the tensile and
fatigue specimens were heat trated separately, thus accounting
for one discrepancy in hardness readings between the chart and
the tabulation above. Fatigue limit is for 100 million cycles.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 677
85
86
0.02
(\J
.....
<l'"
,,;
0.01
Bending Fatigue
4027
Carburlzed
b. O.OOS"Ca..
D O.OIS" Ca.e
o 0.03S"Ca..
"0
0.005
ji
.~
.. 0.002
Ul
0.001 L...-_--'-_----'--;.----'-....-_'-;;-_........_--'-;;_--'
102
103
104
lOS
10
1
Reversals 10 Follure,2N f
Bending-fatigue results for uncarburized and carburized 4027 steel.
Source: R. W. Landgraf, "Control of Fatigue Resistance Through Microstructure-Ferrous Alloys," in Fatigue and Microstructure, American Society for Metals, Metals Park OR, 1979, P 463
2006
1600
600
25 Cr '"'A 4 - annealed
.4~.= 1.2%. H1 (10MPo)
=aOO3 sec:'
Cl bose metal - mech. pot.
m bose metot - chem. pol.
surtoce treatments
In the above bar chart, the effect of surface treatment on fatigue life is summarized. The base metal was mechanically polished before surface treatment. For comparison, pot galvanizing and Ni-plating have been performed after mechanical as well
as after chemical polishing of the base metal. The results after
chemical polishing are given above in the form of dashed
columns.
The galvanizing such as Ni- and ZN-plating is by no means an
appropriate method to increase the fatigue life in hydrogen in
spite of the reduced surface roughness and protecting effect.
This is because the galvanizing produces a relatively high tensile
residual stress and the deposits possess generally poor ductility.
Source: Kyong-Tschong Rie and Werner Kohler, "Improvement of the Resistance of Metals to Cyclic Plastic Loading in High
Pressure Hydrogen Environment," in Current Solutions to Hydrogen Problems in Steels, C. G. Interrante and G. M. Presouyre,
Eds., American Society for Metals, Metals Park OH. 1982, P 380
87
88
hammered
I
/ alvanized
I
e-
300
III
I
Ia/hed
III
200
Ni- la/ed
number af cycles N
The above graph shows the cyclic strain hardening and softening curves for different surface treatments. It can be seen that the
fatigue behavior in hydrogen environment can be improved by
some surface treatments.
Source: Kyong-Tschong Rie and Werner Kohler, "Improvement of the Resistance of Metals to Cyclic Plastic Loading in High
Pressure Hydrogen Environment," in Current Solutions to Hydrogen Problems in Steels, C. G. Interrante and G. M. Pressouyre,
Eds., American Society for Metals, Metals Park OH, 1982, P 379
..e
..
2.0
..
0.
''';'.
"'l
1.0
.....
,:~
g
~
0 Ni-ptatea
0,2
60
200
500
1000 2000
cntxa! nurrtJer of cycles Na
5000
MXJ
Source: Kyong-Tschong Rie and Werner Kohler, "Improvement of the Resistance of Metals to Cyclic Plastic Loading in High
Pressure Hydrogen Environment," in Current Solutions to Hydrogen Problems in Steels, C. G. Interrante and G. M. Pressouyre,
Eds., American Society for Metals, Metals Park OH, 1982, p 380
89
90
O"ty = 1190 MN m- 2
O"tu =1310 MN m-
K =40 MN m- 3 / 2
.,
U
<Il
V>
3.4
3.8
4.2
I
4.6
5.0
liT, oK-I
Source: Herbert H. Johnson, "Keynote Lecture: Overview on Hydrogen Degradation Phenomena," in Hydrogen Embrittlement
and Stress Corrosion Cracking, R. Gibala and R. F. Hehemann, Eds., American Society for Metals, Metals Park OH, 1984, P 17
UNNOTCHED NOTCHED
75
65
<Il
.>t:
I
(f)
(f)
60
I-
D~-&
o~
............... o
"<, 80--0~
~"
06..,
....
0--
6,6..
100
80
0
.....
o,
""8<,6 . . - 6 .__
55
O-B~
UJ
a:
70
6.
(f)
(f)
50
\~\A
(f)
45
40
UJ
a:
I60
\\.\.
,,~& .........
35
~~
30
25
104
(f)
-A-A_
~.--
'.~-.-
105
106
40
107
CYCLES TO FAILURE
SoN curves of a normalized and tempered AISI 4140
Source: Steel Castings Handbook, 5th Edition. Peter F. Weiser, Ed.. Steel Founders' Society of America, Rocky River OH, 1980.
P 3-16
91
92
ELONG
%
113 (779) B7 (560)
43
110 (758) 80 (552)
61
III (7651 81 (558)
30
ksi
CAST
4135
WROUGHT 4140-L
-T
80
WROUGHT
75
CAST
(MPo)
ksi (MPol
BHN
223
217
217
_
500
70
65
<l
W
o
~
60
...........
' ......
fh . . . . . . . . .
UNNOTCHED'
-,
LONGITUDINAL
------~
<, <,
400 11.
---
en
en
en
en
w
lr
I-
en
\
\
45
30
lr
300 I-
en
""
I ',
40
35
,1__ -.
55
::i
11. 50
,~RANSVERSE
NOTCHED
LONGITUDINAL
''-......
AND TRANSVERSE
- ---NO FAILURE
=*
200
106
CYCLES
TO FAI LURE
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980,
P 4-8
4-11. 4140, 4053 and 4063 Steels: Effect of Carbon Content and
Hardness
1000,-------------,--------,--------.--------;
140
o to 2 micro-in.finish
900.1--------/-------t--------t
130
120
8001--------/-------t--,
in
Q.
sc
::;;
110
~.
E
~
.~
7001-------j-----;
100
:J
.~
.
:J
.~
u,
U.
90
6001-------:
80
70
60
30
40
50
60
70
Hardness, HAC
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, p 676
93
94
10'
.....
Q)
.....
Q)
10-'
>U
10:'
>U
.....
.....
E
E
....
....E
C121(j'
cI2 Ill"' -
'tJ'tJ
'tJ'tJ
10
10
20
304050
AK (MPaV'ffi)
a
100
10
L.L_ _-'--_---J'---'--'--'-_--,-lJ
100
AK(MPaV'ffi)
b
Fatigue crack propagation in an AISI 4140 steel: (a) longitudinal direction (parallel to rolling direction); (b) transverse direction (perpendicular to rolling direction).
The Paris power law, which describes the crack propagation rate in stage II
for a series of metals, is very useful because of its extreme simplicity. For
example, it has been observed experimentally that data points in the form of
log (do / dN) versus log ~K for a given material (constant metallurgical structure) from three different samples-edge crack in a compact tension sample,
through-thickness central crack in a plate, and plate containing a partially
through-thickness crack -all fall on the same line. Also, there is experimental
evidence that shows that the stress level by itself does not influence the fatigue
crack growth rate for stress levels below the general yielding. Thus, it can be
considered that the parameter ~K describes uniquely the crack growth rates
for many engineering applications. However, the structure of material can
influence fatigue crack growth rates drastically; the value of m can change a
lot. The above charts illustrate the directionality in the fatigue crack propagation rate in an AISI 4140 steel. The exponent m has a much higher value in
the transverse direction than in the longitudinal (rolling) direction, due to the
presence of elongated inclusions.
Source: Marc Andre Meyers and Krishan Kumar Chawla, Mechanical Metallurgy: Principles and Applications, Prentice-Hall, Inc.,
Englewood Cliffs NJ, 1984, p 714
Carras.....
LLi-O.4
I
Potential . - . _ " - - -
.~
(fj
(/) -0.6
~
>-0.8
.r
<{
~ -1.0
I-
0...
-1.2
10 5
106
CYCLES TO FAILURE
The effect of cathodic polarization on the fatigue behavior of
4140 steel (heat treated to HRC 52) in 3% NaCI solution at a
stress level below the fatigue limit in air is shown above. The use
of cathodic protection to prevent corrosion fatigue of steels depends sensitively on the hardness of the steel. For example, cathodic protection of a 4140 steel was shown to be feasible for
hardness values of Rockwell C 40. At higher hardness values, an
improvement in fatigue resistance is observed for moderate cathodic potentials, but complete protection is not possible. At potentials large enough to inhibit corrosion fatigue for softer
steels, a decrease in fatigue resistance is observed, presumably
due to hydrogen embrittlement (note above chart).
Source: D. J. Duquette, "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys," in
Fatigue and Microstructure, American Society for Metals, Metals Park OR, 1979, P 360
95
96
2.5
1.5
FUL L Y MACHINED
f===:J~
,
.... 0.2
<l:
0::
L,\
~
w
INV
CRYOLITE
HIGH
WASH
""""';;:::::====~~:.:;:=~:::..CHROM ITE
GR. S.
PROPR. WASH
_ _ _ _==--========,JS~IL~ICAGR.
COPE
DRAG
200
MACHINED
400
600
800
1000
Plate bending tests for quenched and tempered low alloy cast
4330 steel indicate that investment cast surfaces, or conventional castings produced with special mold washes, performed
equal to, or better than, fully machined and polished plate specimens. The data also suggest a tapering off of the surface effects
on the endurance ratio at 600 or more RMS surface roughness
as indicated in the above diagram.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America. Rocky River OH, 1980,
P 15-29
12
800 iir=0-
160
---'-T-'------
200
----,rr--
240
rr-r --
280
.,---,
110
7001-------1---------11--------1--------:::J 100
50% survival
e:;;
90
Ii 600 f-------t--------::;;;Io--"""'-----t------=90%
~
c
'"
.~
.""'"
1;:
99%
500
1----.....~l=____=,....",~~~k:;;o>'-=::::::=:1f===-----l
80 c
'"
.~
70
400 f - - - - - - - - ' - - - t - - - - - + - - - - - - t - - - - - - - i 60
Approximately 1000 specimens. 1 heat
50
300 '--
800
L-
1100
-----''-----
~~
_:_:_'
2000
Interrelationships of alternating stress, tensile strength and expected percent survival for heat treated 4340 steel.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 678
97
98
'.
I
Fatigue strength
exponent
slope
/
=
-0.09
Reversals to failure. 2N f
Typical data for strength versus fatigue life for annealed 4340 steel.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 672
10-11----+---"O:-"'.--I-----r- 2
= l1p +
~
E
10-4l::-
--'-:-
---lL::-
-'-::
-'--:--
-'-::--
-..L:--~"___J
100
Typical data for total strain versus fatigue life for annealed 4340 steel.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 672
99
100
4
10
( V = Uf =1200 MPa
C
Q.
3
10
~
C
b
~
~IN
10
10
Stress amplitude (!:J.aj2) versus number of reversals (2 N J) for AISI 4340 steel.
It is convenient to consider separately the elastic and the plastic components of strain. The elastic component can be readily described by means of a relation between the true stress amplitude and the number of
reversals (i.e., twice the number of cycles):
~E
a (a'J)
E
_e_=_o_= _ _ (2N)b
where ~Ee12 is the elastic strain amplitude, a o the true stress amplitude, aj-the fatigue strength coefficient
(equal to stress intercept at 2NJ = 1), NJthe number of cycles to failure, and b the fatigue strength exponent.
This relation is an empirical representation of the S- N curve above the fatigue limit. The above chart shows
an application of this relation to SAE 4340 steel. It was observed that fatigue life increased with decreasing
b. Morrow, based on energy considerations, showed that the fatigue strength exponent is given by:
n'
b=--1 + 5n'
where n' is the cyclic hardening coefficient.
Thus, the fatigue life under elastic cyclic conditions (whether stress- or strain-controlled) increases with a
reduction in n'. Of course, the higher the material coefficient aj, the better it is for fatigue. There is evidence
that ajis approximately equal to aJ' the monotonic fracture strength.
The plastic strain component is better described by the Manson-Coffin relation:
~E
-p-=
Ej( 2NJ
where ~Ep12 is the plastic strain amplitude, Ej is the ductility coefficient in fatigue and is equal to strain
intercept at 2NJ = 1, 2NJ is the number of reversals to failure, and c is the ductility exponent in fatigue.
Source: Marc Andre Meyers and Krishan Kumar Chawla, Mechanical Metallurgy: Principles and Applications. Prentice-Hall, Inc.,
Englewood Cliffs NJ, 1984. p 697
101
200
::;;
"c
'"
~
~
.>l
1200
1J,
c
1000
Vi
150
0
~
~
Vi
Periodicoverstrain
800
100
600
2
10
3
10
104
105
106
7
10
Overstrain superimposed on constant strain may have a significant effect on fatigue life. Shown above is the effect of periodic large strain cycles on fatigue life of AISI-SAE 4340 steel
hardened and tempered to a yield strength of 1100 MPa (160
ksi).
Source: Metals Handbook. 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH. 1978, P 681
102
600
--400
~200
200
.00
600
800
1000
1200
1'00
Potter has described a method for approximating a constant-lifetime fatigue diagram for unnotched specimens. Using this method, a series of points corresponding to
different lifetimes are calculated and plotted along the diagonal line on the left side
(R = -1). Each of these points is connected by a straight line to the point of the other
diagonal (R = 1.0) that corresponds to the ultimate tensile strength. A comparison
between the estimated constant-lifetime diagram and the experimentally determined
diagram is given in the above illustration. Here is presented a comparison between a
calculated constant-life fatigue diagram (solid lines) and experimentally determined
data (dashed lines). The calculated lines correspond well with the experimental lines.
Generally, the predicted lines represent lower stresses than the actual data. Estimating
fatigue parameters from the Brinell hardness number provides more conservative estimates. These results are only approximations, and the methods may not apply for
every material.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 681
103
1.0
200
c,
:;;;
~
E
150
1000
"'"
~
800
"E
'x
"1""""')
600
100
O't;..;
:9~
600
"""~~'1;1.()
..
400
10 6 cycles lifetime
50
200
o'----_---'-__---'--__--'--__-'---__ _ __''"____
~
-1200
-1000
"'1100
-tiOO
-400
-200
200
400
600
800
1000
L __
1200
_ _ '_ _____'
1400
1600
Minimumstress, MPa
Constant-lifetime fatigue diagram for AISI-SAE 4340 alloy steel (bar), hardened and
tempered to tensile strength levels of 860 MPa (125 ksi), 1035 MPa (150 ksi), 1380 MPa
(200 ksi) and 1790 MPa (260 ksi), All lines represent fatigue lifetimes of one million
cycles.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Jronsand Steels, American Society for Metals, Metals
Park OH, 1978, P 669
'x
104
160
~ 800
~
,
1;;
E
E
"~
100
~
"w
600
:!E
400
60
200
O'--_ _....L-1000
-een
' -_ _---'-
-il00
-400
-200
200
400
600
-'--_ _- ' -
800
1000
-'0
- ' -_ _
1200
1400
Minimum stress.MPa
Constant-lifetime fatigue diagram for AISI-SAE 4340alloy steel (bar), hardened and
tempered to a tensile strength of 1035 MPa (150 ksi), Solid lines represent data obtained from unnotched specimens; dashed lines represent data from specimens having
notches with K, = 3.3.
Source: Metals Handbook, 9th Edition. Volume I, Properties and Selection: Irons and Steels, American Society [or Metals, Metals
Park OH. 1978, p 667
...
Ie.
If. 1200
:;:
~
;
900
.~
~ 600
0
0
o Not decarburized
Decarburized
0
0
150""
::i
e'
:;;
0>
C
.~
0
0
100 ~
r.
<t
.....
~
300
50
>-
Source: Metals Handbook, 9th Edition. Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 674
105
106
1000
--.r--..
s::;;
~ 900
1;:
en
.~
E 800
;;:
<,
o Small inclusions
Large inclusions
<,
140
~ ~.
700
100
Fatigue life of two lots of AISI-SAE 4340H steel; one lot (lower curve) contained
abnormally large inclusions; the other lot (upper curve) contained small inclusions.
Source: Metals Handbook, 9th Edition, Volume 1, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 673
200
190
180
STRESS
RANGE.
Ib/on 2
170
tal
160
150
140
130
i20
103
SURFACE INCLUSIONS
a SUBSURFACE INCLUSIONS
DIAMETER,
on
0001-
I--~~
CYCLES
Typical initiated crack sizes are l-lO Mm. As this is an order-of-magnitude greater
than dislocation substructure sizes, such an initiated crack will behave as in a continuum. For materials with lower stacking fault energy cross-slip and PSB formation is
difficult, thus inhibiting initiation. For such materials crack initiation can occupy a
significant fraction of life. Other microstructural sites for initiation are discontinuities
such as grain and twin boundaries, the latter being particularly operative in hcp metals.
Usually, however, at ambient temperatures it is the dislocation substructures which
dominate initiation.
For stronger, more complex alloys planar slip behavior dominates, making localized slip bands the initiation sites cfthe random notch-peak topography generated by
shearing a pack of cards. The interaction of slip bands with second-phase particles
(inclusions, precipitates) can produce a local stress concentration which cracks the
interface, producing a surface crack. The above SoN curves show the results of this
process for a low alloy steel. Variations in fatigue life relate to variations in inclusion
size. As well as debonding, oxide or carbide particles can crack under concentrated
localized stresses.
Source: B. Tomkins, "Fatigue: Mechanisms," in Creep and Fatigue in High Temperature Alloys, J. Bressers, Ed., Applied Science
Publishers Ltd, London, England, 1981, p 115
107
108
Ai>
.~,
275
250
'~
225
-.
"\\
\
'00
.x:
ul
(/)
~
200
(;)
"0
.~
a.
C
175
e\
125
\
\
100
:\
150
--
"..
Uncharged
++-
Bake 24 hr -
'---
0\ \
"\
\
\.
<,
Bake 18 hr
\ t-,
75
I'\.. '"
300 ksi
Bake 12 hr
Bake 7 hr
--
Bake 3 hr
Bake 0.5 hr
-.-..
50
0.01
0.1
1
10
Fracture lime, hours
100
1000
There are many embrittling effects of hydrogen on steels: the ultimate strength of a steel may be
reduced; ductility as measured by total elongation to fracture or reduction of area may be decreased;
and crack growth may be significantly accelerated. The hydrogen responsible for these effects may be
present in the environment external to the steel or may be present internally as a reslt of steelmaking or
processing operations such as pickling or electroplating. Hydrogen may promote a transition from a
ductile to brittle fracture mode or it may reduce ductility without a change in fracture mode.
The graph above shows the effects of baking at 150C (300 OF) on the static fatigue (sustained loading) of the hydrogen-charged specimens. Increasing baking time effectively lowers hydrogen content
even in the plated specimens, and sufficient baking eventually restores the strength of charged specimens to that of uncharged specimens. The horizontal portions of the curves in the graph above are
designated as static fatigue or endurance limits, i.e., the stress level below which failure would not occur
no matter what the duration of stress application. As hydrogen content is decreased by baking, the
static fatigue limit increases.
The specimens used to obtain the above data were notched and therefore the static fatigue limits hold
for that particular notch geometry. In general, the sharper the notch, the lower the static fatigue limits,
an indication that a critical combination of hydrogen concentration and triaxial stress state is required
for crack initiation.
Source: George Krauss, Principles of Heat Treatment of Steel, American Society for Metals, Metals Park OH, 1980, p 223
400
4340
250,000 PSI
350
300
250
co
P-t
0
0
0
200
......
.....
......
r-I
Cf)
Cf)
co
150
100
RATE
UNEMB. EMBRIT.
PLATED IN LAB
1000 RPM
COMMERCIALLY PLATED
50 250 RPM
0
PLATED IN LAB
200 RPM
'V
PLATED IN LAB
.33 RPM
0
10,000
1000
10
100
CYCLES TO FAILURE
<>
Source: George Sachs, "Test Methods for Evaluating Hydrogen Ernbrittlement," in Materials Evaluation in Relation to Component Behavior (Proceedings of the Third Sagamore Ordnance Materials Research Conference). Syracuse University, Syracuse NY,
1956, P 508
109
110
----- -- ---
ATMOSPHERE
HITfUDED
------------
lAYER
5'>
HUHBER OF CYCLES
10 ~
10 5
Source: J. A. Riopelle, "Short Cycle Atmosphere Nitriding," in Source Book on Nitriding, American Society for Metals, Metals
Park OH, 1977, P 287
120
800
100
:Ii
~ 6001----+----'
Normally
heat treated
350 L -_ _--'-
----'_.L.>...
----'
105
Cvcles to failure
Comparison between fatigue limits of crankshafts (S-Nbands) and fatigue limits for separate test bars, which are indicated by plotted points
at right.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 674
111
112
:5 140
C!.
~-130
e
u;
120
110
'r-.
As ni ide
\
\.
\ "I'
ilrid ~
"1\"'r--.
nducti n har en d
DU
10 5
10 6
10 7
Cycles to Failure
As demonstrated in the above S-N curves, fatigue tests of
AISI 4340 steel in various surface hardened conditions show
that combined treatments produce endurance limits between
those developed by separate treatments.
Source: Sander A. Levy, Kenneth E. Barnes and Joseph F. Libsch, "Combining Nitriding With Induction Heating Pays a Bonus," in
Source Book on Nitriding, American Society for Metals, Metals Park OH, 1977, P 241
0---
180
----
O~---------<>6
170
Cr + SFl
<>-
Cd+
O .. ~
~Chromate
~1f8are
~160
j
ISO
o
R 0.8
140
2000 cpm
RT
Cr +
~TUngsten
Carbide
SFl
we - Shot Peened
lJOL-----,---,-.w..u.uJ.,-----,---,-.w..u.u.l.;-----'---'-.L..LlllJJ.,-----'---'-.L..Lu.u-,;---'----'-L.LJL..LUJ
Air
Stress
3.5% NaC1
Test
Condi tion
ksi
MN/m 2
Change,
%
Rotating Bending
R = -1
Bare
Cd + Chromate
Cr + Ory Film'
WC + Dry Film'
WC + Cori cone +
Dry Film'
105
105
95
90
so
724
724
655
621
621
0
-9.5
-14.3
-14.3
Bare
Cd + Chromate
Cr + SFL'
160
165
175
1103
1138
1207
WC + SFL'
140
WC + Cori cone +
SFL'
140
Axial Tension
R = 0.8
Stress
MN/m 2
Change,
%
90
13B
552
621
621
621
-81
-24
-14.3
-14.3
-14.3
+3.1
+9.4
110
165
90
758
1138
621
965
-12.5
60
414
965
-12.5
60
4.4
ks;
20
80
so
so
-31. 2
0
-43.8t
-4B. 6~
-62.5t
-57~
-62.5 t
-57f
'Shot peened
tCompared to bare alloy air value
~Compared to coated alloy air value
The above SoN curves, in conjunction with the table, contain the data obtained in air for 4340 steel, bare
and coated. Fatigue data at 107 cycles showed that the cadmium and chromium electroplates, particularly the
chromium, improved the fatigue strength. They were similar in both rotating bending and axial tension
fatigue tests. But in NaCl solution, significantly greater reductions in axial fatigue strength of the coated
alloys were observed due to environmental effects, which remains to be elucidated. Since the Cr and WC hard
(brittle) coatings have a relatively low intrinsic fatigue strength in comparison with the steel, they will become
discontinuous at a relatively low stress level owing to the development of fatigue cracks. (The Cr normally
contains internal cracks.) These cracks will permit access ofthe corrosive NaCI solution to the steel base at the
root of the fatigue crack. In the case of the axial tension test (high steady tensile load), it may be easier for the
environment to reach the crack tip.
Source; M. Levy and C. E. Swindlehurst. Jr., "Corrosion Fatigue Behavior of Coated 4340 Steel for Blade Retention Bolts of the
AH-I Helicopter.Yin Risk and Failure Analysis for Improved Performance and Reliability.John J, Burke and Volker Weiss. Eds .
Plenum Press. New York NY. 1980. P 275
113
114
150
'"
c,
:2
";;;
'"
800
e
1;;
E
E
"
"x
100
600
"1'/.
~~.,
E
:J
E
"x
"00
Q)lj...;
".9s
400
..
~~
J'J'
..
'1"..0
6'a
1:>
50
200
Constant-lifetime fatigue diagram for AISI-SAE 4340 alloy steel (bars) hardened and. tempered to a
tensile strength of 1035 MPa (150 ksi) and tested at the indicated temperatures. Solid lines represent data
obtained from unnotched specimens; dashed lines represent data from specimens having notches withK1
= 3.3. All lines represent lifetimes of ten million cycles.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 669
z
....
I
p:\
....:l
0
0
0
40
30
20
......
....:l
l=l
~4",~...
~.;~~.'::......
~....
~1ARQUENCH /-.:::
10
8
6
(400 0F OIL)'
-........:.::-~
-----.....:.....
---
4
3
2
10
10 3
10 4
CYCLES - N
Effect of quench type on fatigue of carburized differential cross.
Source: D. H. Breen and E. M. Wene, "Fatigue in Machines and Structures-Ground Vehicles," in Fatigue and Microstructure,
American Society for Metals, Metals Park OH, 1979, P 92
115
116
40
~,
30
CQ
,...J
20
0
0
0
......,
,...J
c:
u..;
if;
l:>:
~':".':'.":
~.;.......
---~
10
SHOT PEE~D
:::.:.:
lJNPEE\'ED~ __
----...:....
--
6
5
4
3
10
10 2
CYCLES - r\
Source: D. H. Breen and E. M. Wene, "Fatigue in Machines and Structures-Ground Vehicles," in Fatigue and Microstructure,
American Society for Metals, Metals Park OH, 1979, P 93
117
-- ---- -- -- --
---
r.~Y~p
Al110SPHERE
NITRIDED
HUHBER OF CYCLES
10~
1') 5
Source: J. A. Riopelle, "Short Cycle Atmosphere Nitriding," in Source Book on Nitriding, American Society for Metals, Metals
Park OH, 1977, P 286
118
100
Height strain
80
oE
60
r.-------'-----Time
LI-!
4.0
5.0
6.0
Log cycles
7.0
Axial fatigue of P 1M-forged 4620 steel as a function of height strain during forging.
In general, sensitive properties improve as the level of upsetting is increased during the forging process. The diagram above
shows the effect for fatigue resistance, although the cyclic stress
state also influences fatigue behavior. An interesting feature of
P / M-forged parts is the fact that deformation does not significantly affect through-thickness properties as it does detrimentally for wrought material. For re-pressed parts, throughthickness toughness is slightly lower than longitudinal toughness. Upsetting increases longitudinal toughness while toughness
in the through-thickness direction remains at a relatively constant level.
Source: B. Lynn Ferguson, "Part II: Fully Dense Parts and Their Applications." in Powder Metallurgy: Applications, Advantages
and Limitations, American Society for Metals, Erhard Klar, Ed., American Society for Metals, Metals Park OH, 1983, P 100
600
co
a..
:2:
.'+-----+
.
CI)
CI)
80
56 & 65%
(JrnaK
500
42%
1'------+ 30%
cil
60
400
o~
0
7.0
8.0
Log cycles
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals, Metals Park OH, 1984, p416
119
120
1/
50
Ingot material
L,o 192 h
10
1/
/
\
P/M material
L,o 5
o
100
200
500
163
1000
2000
5000
Life, h
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals. Metals Park OH, 1984, P 620
,-------.----.r------...-------.-----.------.
0.365-1
':3.
8
80 dir
3-1/4
r-
'"
"
longitudinal direction
'0
"
~
~ u~
c'i)
'0,
,,
9-7/8 R.
60
-,
,,
'------------
,,
'~
'------_Q:
21
32
Specimen
configuration
40 '--------'---------''------'-------'---'
105
10
10'
103
10'
Cycles of stress
Source: B. Lynn Ferguson, "Part II: Fully Dense Parts and Their Applications,"in Powder Metallurgy: Applications, Advantages
and Limitations, American Society for Metals, Erhard Klar, Ed., American Society for Metals, Metals Park OH, 1983, p 103
121
122
45 '"
~.
~
i------j
40
'f;
c:
35 ~
7
6
10
10
Number of cycles to failure
Spheroidite Pearlite
Tensile strength,
MPa (ksi)
641(93)
676(98)
Yield strength, MPa
(ksi)
490(71)(b) 248(36)(c)
Elongation in 2 in.,
% ..............
28.9
17.8
Reduction in area, %
57.7
25.8
Hardness, HB .....
92
89
(a) Composition 0.78 C, 0.27 Mn, 0.22 Si, 0.016 S
and 0.011 P. (b) Lower yield point. (c) 0.1% offset
yield strength.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 677
600
...
12
10
500
'"
1
2
3
4
5
6
7
8
9
10
11
.:<
12
.;
..'" 450
....
MLNG
MU\ G
MLCG
SLD G
MLNH
MU\ H
MLCH
SLD H
MLN P
MU\ P
MLCP
SLD P
~
4.14
3.79
3.45
3.10
til
..
~ 400
2. 76
'"
E
.~x
350
2.41
300
2.07
The mean predicted fatigue life is highest with a polished surface and least with a ground finish (9 versus I, etc.). Polished
surface has about 6 times and honed surface about 3 times the
fatigue life of ground finish. No interaction effect between additives and surface finish is revealed.
Source: S. Bhallacharyya, F. C. Bock, M. A. H. Howes and N. M. Parikh, "Chemical Effects of Lubrication in Contact FatiguePart II; The Statistical Analysis, Summary, and Conclusions," in Source Book on Gear Design, Technology and Performance,
Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 277
123
124
II
600
4.14
3.79
3.45
... 500
Ul
""
.
'"
rl
..
::: 450
:;
2
400
'"e
~
350
:2
300
3.10
/---e---i
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
5E
5E
5E
5E
5V
5V
5V
5V
8V
8V
8V
8V
5E
5V
8V
5E
5V
8V
GLl
GL2
PLl
PL2
GU
GL2
PLl
PL2
GLl
GL2
PLl
PL2
GH2
GH2
GH2
PH2
PH2
PH2
2. 76 ~
to
2.07
10 7
Mean Predicted Cycles to Failure
Source: S. Bhattacharyya, F. C. Bock, M. A. H. Howes and N. M. Parikh, "Chemical Effects of Lubrication in Contact FatiguePart II: The Statistical Analysis, Summary, and Conclusions," in Source Book on Gear Design, Technology and Performance,
Maurice A. H. Howes, Ed .. American Society for Metals, Metals Park OH, 1980, P 277
HLN 1.2
HL\ 1.2
HLC L2
~1.0 1.2
HI.N 112
HL\ 112
HLC 112
~LO 112
600
....
.:<
'"
4.14
J.79
J.45
'"
'"
"
J.IO
N
::l
400
2. 76
.~
x J50
2.41
"
"
:E
JOO
2.07
10
10 7
Source: S. Bhattacharyya, F. C. Bock, M. A. H. Howes and N. M. Parikh, "Chemical Effects of Lubrication in Contact FatiguePart II: The Statistical Analysis, Summary, and Conclusions," in Source Book on Gear Design, Technology and Performance,
Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 275
125
126
1
2
3
4
5
6
7
8
9
10
600
550
-;: 500
11
.Yo
12
MLN
MLN
MHN
MHN
SLD
SLD
MLN
MLN
MHN
MHN
SLD
SLD
Ll
L2
Ll
L2
Ll
L2
Hi
H2
Hi
H2
Hi
H2
4.14
j.79
) .45
'
:: 450
f---<T---i
.
.
3.10
l/}
-!:
:: 400
2.76
'"
E
~ 350
2.41
:>!
300
2.07
250 '-c--'-------'---L-...L.J...J....IJ.Lr--.1_L-LJ...L.I...l..L.L...,-_...L--..L----'-...L-L.LJL..U.,,--........J_-'----.LJ~L.LJ.J
10 5
10
10
The effects of lubricant viscosity, slip ratio, and speed on fatigue life are
shown in the diagram. The 12lines in the figure are separated in two distinct
groups, low slip (lines I to 6) and high slip (lines 7 to 12). In each group the
effects of viscosity and speed may be noted. Viscosity X speed interaction
produces complex effects on mean predicted lives which under low slip conditions are not statistically significant in their differences. Only under high
slip condition, lines 9 versus 10indicate a small statistically significant lowering in mean fatigue life in high-viscosity oil under higher speed. A comparison
of lines 11 and 12 shows that the lesser life in synthetic oil with additive is a
statistically borderline case, though the trend is similar to that with mineral
oil under the present operating conditions. The regression analysis shows
that in the present tests both speed and viscosity have nonsignificant direct
effect on life, and a few small interaction effects with steel, surface finish,
viscosity, and slip were observed.
Source: S. Bhattacharyya, F. C. Bock, M. A. H. Howes and N. M. Parikh, "Chemical Effects of Lubrication in Contact FatiguePart II: The Statistical Analysis, Summary, and Conclusions," in Source Book on Gear Design, Technology and Performance,
Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, p 276
400
2.76
300
2.07
vi
~
<,
D<
Co
"'4
::c
"
"
"
z
e
200
1.38
.}Is-,f....
:>
~o
o ()"''''
)(
ot
""'..
-,
-----
'HASiIllOAU.
'I(DI(1I0 BY MUllIPll UQRfUION ANAlYSIS
10
100
l,. LIFE. 10 6 CYClES
Comparison of stress/life data for the mineral oil with and without the
ZnDDTP additive in surface fatigue; 85 percent confidence bands for the LSD
life estimates are shown and compared with the stress/life relation predicted
from regression analysis.
The synthetic and mineral oil no-additive conditions had about the same
life. However, the life at all stress levels tested was significantly reduced for
the mineral oil with additive below that without additive, by almost a factor
of three at the L so level, further indicating a detrimental effect of the
ZnDDTP additive on life.
Both the synthetic and the mineral oil tests had lives almost two orders of
magnitude below the standard Lundberg-Palmgren calculated life. A life reduction factor is used with the Lundberg-Palmgren theory when applied to
rolling bearings having high contact angles and thus high slip; but rarely does
the slip at bearing contacts approach that level used in these tests, so it is not
surprising that the life reductions observed are much greater than the life
reduction factors normally used for bearings.
The stress/Iife plot shown above is particularly revealing. There is no
doubt that the stress / life slope for the additive oil is significantly steeper than
for the base stock, which seems to approach the Lundberg-Palmgren theory
in stress/fife slope except for the highest stresses where it is even shallower.
Source: w. E. Littmann, B. W. Kelley, W. J. Anderson, R. S. Fein, E. E. Klaus, L. B. Sibley and W. O. Winer, "Chemical Effects or
Lubrication in Contact Fatigue-Part III: Load-Life Exponent, Life Scatter, and Overall Analysis," in Source Book on Gear
Design, Technology and Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 285
127
128
0
ILl
...J
95.
90.
80.
70.
60.
50.
40.
30.
ex 20.
IL.
...
Z
ILl
10.
a:::
ILl
a.. 5.0
4.0
3.0
2.0
I. 0
L..-_-'------'-----'----'''--'--'l..J...J...L--_-'------'-----'----'''--'--'LJ...LJ
10
100
Source: C. A. Stickels and A. M. Janotik, "Controlling Residual Stresses in 52100 Bearing Steel by Heat Treatment," in Residual
Stress for Designers and Metallurgists, Larry J. Vander Walle, Ed., American Society for Metals, 1981, p 34
CASE---CORE---CARBURIZED----
10'
10
10'
REVERSALS TO FAILURE,
Summary plots of total strain amplitude: reversals-tofailure data for simulated case, simulated core, and carburized materials.
Carburized material is seen to have low-cycle fatigue resistance intermediate between the simulated case and core material,
a common intersection with simulated core material at intermediate lives; and in the long-life regime, carburized material specimens are more fatigue resistant than either simulated core or
case material specimens.
Plotting the strain-life curves for both case and core simulated
materials on a common set of axes , as shown in the above chart,
reveals an interesting feature. It has been observed that curves of
these materials intersected at a life of approximately 2.NJ = 105
reversals. This is in agreement with the results of this investigation. Intersection of the life curves for simulated case and core
materials accounts for a shift of failure location in carburized
components.
Source: J. M. Waraniak and D. F. Socie, "Cyclic Deformation and Fatigue Behavior of Carburized Steel," in Wear and Fraclure
Prevention, American Society for Metals, Metals Park OH, 1981, P 249
129
130
300
Single --.........
reheat
.........
-------~rr"~
<lID
- /""":
Direct quench ---./
100
.0
L....---L--L....L..JL...U.UJ..._...L......l-JL...L.I..LU.l...----l.---l.....L...L.L.I........ _-'--'-.L..L.J..L.LJ.J---J
103
105
107
Cycles to failure
Source: George Krauss, Principles of Heat Treatment of Steel, American Society for Metals, Metals Park OH. 1980, P 264
131
50
....
~
~
--
-- .... --
AnlOSI'IlERE
NITRIDED
:;;
iO
QlJ[HCHED AHD TD4PERD
AT 1010 F (\6\ C)
tHJKBER OF CYCUS
10
S-N curves for 8620 steel; nitrided versus not nitrided (quenched
and tempered only), showing stress versus number of cycles for
completely reversing torsional fatigue.
Source. J. A. Riopelle, "Short Cycle Atmosphere Nitriding,"in Source Book on Nitriding, American Society for Metals, 1977,p 286
132
40
~ 30
...... 20
10
o
o
o
.....
...
"':'.:.:-. - _
. ....
-.-.-.-.
8622 MATERIAL
............... .
...................
....
..... ............
........
10
10 2
10 3
10 4
10 5
CYCLES TO FAILURE
106
C/)
z
......
I
P-
o
0
0
.....
t::l
40
30
20
~~
"2'::.".::-
....-::-.:.......
10
8
6
8822 MATERIAL
.....::: ........
...... ........
........ ::: .
~
'"
4
2
GRIND AFfER H. T.
GRIND BEFORE H.T.
- . _.-
Z
......
ffi
~
10
10 2
10 3
10 4
lOS
CYCLES TO FAILURE
10 6
Source: D. H. Breen and E. M. Wene, "Fatigue in Machines and Structures-Ground Vehicles," in Fatigue and Microstructure,
American Society for Metals, Metals Park OH, 1979, pp 91-92
1.0
0.9
OoB
X-
- a-
CAViTIES
CAST STEEL - SOUND
R.R. MOORE FATIGUE SPECIMEN I UNNOTCHED I
R. R. MOORE FATIGUE SPECIMEN (NOTCHED 0.001!! in. R I
CONTAINED WITHIN THE BAND ARE THE
FOLLOWING DISCONTINUiTIES
WELD -INCOMPLETE PENETRATION
WELD - UNDERCUT
WELD - SLAG
WELD - MACHINE - SOUND
AS WELDED - SOUND
SLAG INCLUSION
HOT TEARS
0.7
0.6
0.5
l:
l-
l:
I-
e> ~ 0.4
Z
III
III
Ul
:n
0:: 0::
I-
;.:.:.:=.::.::.....:.:.:=--=-""-=~
III
III
.J
e>
~O.2
=>
lL
RANGE
0.3
III
I-
0.1
o.s
MEAN
- 0.1
STRESS
TENSILE STRENGTH
- 0.2
- 0.3
- 0.4
Goodman diagram for bending fatigue for normalized and tempered cast 8630 steel.
Data here show that severe discontinuities lower the fatigue strength of
cast steel. However, it will be observed from the Goodman diagram above
that the results of the notched [0.0015 in. (0.0381 mm) radius] R. R. Moore
fatigue specimen fall below those of the other bending fatigue values. Goodman diagrams for torsion fatigue and for a quenched and tempered heat
treatment show similar conditions with the notched fatigue values below the
surface discontinuity values. In many cases, therefore, design, based upon
notched R. R. Moore fatigue data, introduces a safety factor.
It must be remembered that the discontinuities were very severe and exceeded all ASTM classes of nondestructive inspection standards. The allowable discontinuities described in the ASTM standards are therefore expected
to exert a somewhat less damaging effect on fatigue behavior.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed.. Steel Founders' Society of America, Rocky River OH, 1980,
P 15-32
133
134
,--------,----,--------,r------,
END.
UTS
0.5
(/)
(/)
!.!l
(~)
138
(951)
138
135
137
(951)
(931)
(945)
~
0.17
0.13
0.26
0.28
g:(/)1-::c 0.4
o
WZ
::::>W
00::
i= Ii;
It
0.3
W
0-.J
I:::=0CLASS 2 SHRIN K
6.
z_(/)
~ ~ 0.2
""
t.
SHRIN~~A~~
Zl-
CLASS 6
0::
I-.J
<3:
~.
(SUB SURFACE)
flLASS 2 SHRINK
6. (EXTENg~ SURF.)
--.-~
0.1
NO FAILURE
0'--------'-----...1.--_ _---'
104
10 5
106
10 7
CYCLES
----'
TO FAILURE
Source: Steel Castings Handbook, 5th Edition. Peter F. Weiser. Ed., Steel Founders' Society of America. Rocky River OH, 1980,
P 15-30
...------,-------r-----r--------,
I
I
I
TENSILE STRENGTH
(f)
(f)
0.4 -
.84 - 91 ksi
OCr
t;;t;
oc Z
w
<1
w OC
0.3 -
rt;;
(f)
~....J
Z(f)
CLASS 6
zlOC
W
I....J
<1
NO FAILURE
....
NO. 15
NO. 13
CLASS 6 SHRINK
x-x
NO. 13
CLASS 2 SHRINK
0-0
0.1 r-
~~SHRINK
SHRINK-./
-1-..
......
0.2 I-
~~
rSOUND
x I
<.
X~
0
. . . . . . . . . ~- 9t,--:;, O'::CLASS
0 """"--
SOUND
CYCLES
TO FAI LURE
Source: Steel Castings Handbook. 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980,
P 15-31
135
136
,....------r----~----...,__---___.
T ENSILE
en
en
0.4 I-
W
lr
. . XC
o ,~?...
o Xo ..... _
0
o 0 o~--o_
<t W 0.3 ~
W lr
II-
<.
en en
X_X_
o~ 0
00 - - - 0 _
~--1
in
(917-951 MPo)
....
lr
STRENGTH
133-138 ksi
I-I
en I-
0.2 I-
j::Z
<tW
NO FAILURE
ZI-
0-0
NO. 15
SOUND
lr
W
0--<>
NO. 12
CLASS 6
SHRINK
x-x
NO. 12
CLASS
SHRINK
~
<t
0.1 I-
10 5
10 6
10 7
CYCLES
TO FAILURE
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980,
p 15-31
en
en
W
a::
I-J:
en t;
6,
SHRINK -----
0.4
( SUB SURFACE 1
~...J
Z
_en
I-Z
<l:W
Z I- 0.3
a::
STEEL
I-
...J
<l:
AXSOUND
CLASS 2
~a::
Lt
0.5
WZ
:::>W
~tn
l:i.-,-------,-----,-------.
0.6
0.2
L..-
II
14
o_"'\..
6 "'0'0
6''00,
0
UTS
~ I MPo 1
83
B4
..J.......
(5721
(5791
----'
'--6_
0_
END.
RATIO
.35
.32
.........
.....J
104
CYCLES TO FAILURE
Effect of shrinkage on plate bending fatigue of normalized and tempered cast 8630 Ni-Cr-Mo steel.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980,
P 15-31
137
138
80
Cast
8630
'"
'"
5001--------,,------=""""'ct----------t---------;
.0;
0:
60 "
::i 400 1-----------+-----2-"""=-+-~---------l
~
~
~
:;;
IJ)
300 I-----------+"""-~""'_=------t------------j
40
Normalized and
tempered to 220 HB
0.1
10
600
Ouenched and
80
tempered to
286 HB
500
.0;
c,
sc
:;;
60 ~.
400
U5
U5
300
40
200L----------'----------'-0.01
0.1
-----J
10
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, p 397
80
700
100
120
140
160
r---.-,-----..------.-.---.----.----.------y-----,
100
600
Cast steel
I---I-------Ir----t-----t-----..~!i'i~_+--___l
80
5001---1--
"'----+---t----j
60 ]
"
.~
u.
300
40
20
100
500
600
700
800
900
1000
1100
1200
The effect of notches on fatigue limit is apparent when comparing similar wrought and cast steels with regard to fatigue
limit at selected static tensile strength levels; note curves above.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 397
139
140
60
.....
..........
..... ......
.... .....
...... ......
---
AntDSPHERE
NITRIDED
50
40
QUENCIlED AND mtPERED
AT 1010
r (16' C)
MlHIER OF CYCLES
10
Source: J. A. Riopelle, "Short Cycle Atmosphere Nitriding," in Source Book on Nitriding, American Society for Metals. 1977.p 287
141
120
_110
V)
e,
':I. 100
If)
(/) 90
lJJ
~
"
\ "",
----- ;OR~
'\.'~ "
WEWDE't>+ST~~~!l
<, "--
" '.,
"
80
- - - - ~~ \~EL.\:)ED ( E.B')
.......
~eL..
i
--+ -- ---- -
i"o .. _
~--
IO~
106
107
10 8
CYCLES TO FA'LUR~
Fatigue strengths for case-hardened materials as well as
through-hardened may be satisfactorily defined using the R. R.
Moore rotating specimen test. The smooth unnotched Moore
specimen is ideally suited for studying many of the effects of
manufacturing and processing variables upon fatigue endurance. An example of the use of this testing technique in the evaluation of electron beam welding and postwelding aging effects
upon the endurance limit of basic AMS 6475 material is shown
in the above S-N diagram.
Source: Charles W. Bowen, "Review of Gear Testing Methods," in Source Book on Gear Design, Technology and Performance,
Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 346
142
600
500
'E,'! 0.75 %
400
300
w 500 .-..---r-.-n'TTTT"--'--'-rTT'1~---.-.--rTTTTrr-....,.-;".,."""""
a::
::>
--l400
g 300
(f)
w 200
--l
U
t>
100
L-..L......L--L.L..I.lllL----'--l-L..L1.JL.U.L...-..L.......L.L.L.U.Lll....---'---'L.J...J.J..WJ
10- 1
1.0
10.0
100.0
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH. 1980,
P 15-55
1= 1lef"rmal:ion ""'90-
I-
li'T
(%)
I-
II-
-~
I-
1550
O/ll---I---+-+-H--+---+-+-H--+---+----I---If-+--+---+-I-H--t---+---I-l--l
I-----I-++t--+--+--HH----I--+--t--t-+---t----t-+--HI--+---t--t-t--l
10
10~
As holds true for other ferritic steels, the effect of hold time in
compression is slightly detrimental to fatigue life.
Source: Philippe Berge, Jean-Roger Donati, Felix Pellicani and Michel Weisz, "Properties of EM 12." in Ferritic Steels for HighTemperature Applications, Ashok K. Khare, Ed., American Society for Metals, 1983, p 114
143
144
+,'
3.0 , - - - - - r - , . - , - - , - ,r-tr-'r-r-r-t-t-t-r-r-r-r-r-r-r-r-r-......-r-r-r-rr-r-r-r-r-r-r-r-r--r
o
"J
-l 1.0
REVERSE BENDING
MATERIAL A
MATERIAL B
MATERIAL C
PUSH - PULL
MATERIAL
LL
t:l
Z
<l
0::
LL
-l
<l
I 0.1
L-_----'-_..L-....l-..I......JL..LLJ...JL-_-'-_-'---...l......l.....J.-L-L..LJ--:-_-'
10 2
10 3
CYCLES TO FAILURE Nf
'f--
___
-6-----6_
.....
z
~ 0.5 % STRAIN
W
0::
-6----_6
::J
-l
100
f0-
--- A-=
~ 1.5 % STRAIN
g
(f)
.J
>U
I
1.0
10
DWELL PERIOD - h
Effect of dwell periods on fatigue characteristics of
low-alloy cast steel.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980,
pp 15-56 - 15-57
..,'
"
-l 1 0
~
u,
o
W
o
Z
a::
I '-::2--'-----'-...L..JL....L..L.L.L-'--;:---'-----'-...L..J'--'--'c...L.L'--;---'---'--'--'--'-'..........
10
CYCLES TO FAILURE - N
Source: Steel Castings Handbook. 5th Edition. Peter F. Weiser. Ed . Steel Founders' Society of America, Rocky River OR 1980,
p 15-55
145
146
100
200
300
c:---,-----r---,-----.--....--...----,
103
If)
0.6 %
W
...J
U
0.7 %
1.0 %
W
U
Z
<t
1.5 %
=>
0.5 %
0::
w
10 2
1.0 %
CONTINUOUS CYCLE
30 MIN. DWELL TIME
TEMPERATURE - OF
Influence of temperature and dwell period on the cyclic
endurance of C-O.5Mosteel at various strain amplitudes.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980
pIS-55
147
~ 0.010
....::::;
::l
o HI-FORM 50
I STRAINED
t.1006
lAND AGED
...
::E
<I:
Z
<I:
....
~
VI
.....
u
>u 0.001 '-3
10
---'10
....L.-:-
105
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention,
American Society for Metals, Metals Park OR, 1981, P 214
148
0500L
~400
~1.FORM50
o-o-~
~~
~
0.0065
"-"-~C)oO-OC>-_~0.002
gJOO
~
1006
OL.-----.L..;-----...I..::------'--;:------L-;------'
2
3
1
10
10
CYCLE NUMBER, N
10~
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention,
American Society for Metals, Metals Park OH, 1981, P 209
3=
.......o
U
cyo-~
400
0
II>
~~ 200
~:;;
DP 2
~HI-FORM50
a~ 300
>Uv;
..:
~
_000-0 DP 1
/0
'-1006
100
STRAINED-AND-AG ED
II>
OL...-_...L-_--'-_----'L-_...J..-_--'-_ _
.002
.004
.006
.008
.010
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention,
American Society for Metals, Metals Park OH, 1981, p 212
149
150
300
\,
\\.
~ -.
~
I
0: 1'-.~"" " .
~~
I
spc(o,81)
spc(O.81)
-'-0-- IAI'I'C4011.111
..06-- I APFC45It.Ol ISPcIO.81
--0-- APFC5O(1.01 SPC(O,811
IWIK
SPC(1.21)
-.......:;.r......:.~ ~.
...
~ ~.'- _
D
50
10'
--~
10'
NUMBER OF CYCLES
Source: M. Takahashi. "Criteria of High Strength Steels for Applying to Automobile Frame Components," in HSLA SteelsTechnology & Applications, American Society for Metals, Melals Park OH, 1984, P 498
0.01
t---=~~f___t--_t--_+----l
.,
"tl
.e
Q.
E
'"
e
en'i!
Proprietary HSLA
690 MPa (100 ksi] min UTS
0.001
0.0004
t---t------jt---t---t--~
L..-_~L..-_~L..-
102
103
104
_ _L -_ _.L-_
_____I
105
Cycles to failure
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 672
151
152
1.0
t
'ill-
~N
ui
0.1
/.Aj-pl= 1.871(2Nf)-o8396
::J
e,
R 2= 0.968
::iE
-e
:(
a:
Ien 0.Q1
i=
en
:5e,
BE(Cb-)
JF(Cb-V)
KF(Cb-V-SI)
0.001
10 2
10 3
10 4
10 5
10 6
REVERSALS TO FAILURE. 2Nf
10 7
For plastic strain-life relationship the statistical analysis indicates that there are no significant differences between the three
steels (F-ratio is not significant). This is further illustrated in the
above chart, where all the plastic strain data are plotted as a
function of reversals to failure. It is clear that a single straight
line can adequately describe all the data. Such a regression line is
drawn as the solid line in this chart.
Source: Shrikant P. Bhat, "Influence of Composition Within a Grade on the Fatigue Properties of HSLA Steels," in HSLA
Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 588
5-7. Comparison of HSLA Steel Grades BE, JF and KF for Total Strain
Amplitude vs Reversals to Failure '
.,.
w
::::II-
::::i
e,
:::l:
<
<
a:
I-
en
...J
-e
I-
8
12 0.1'------'----'----'----'-----'
10 2
10 3
10 4
10 5
10 6
REVERSALS TO FAILURE. 2N f
Total strain amplitude vs reversals to failure for Cb (BE),
Cb-V (JF) and Cb-V-Si (KF) steels.
Source: Shrikant P. Bhat, "Influence of Composition Within a Grade on the Fatigue Properties of HSLA Steel," in HSLA SteelsTechnology & Applications, American Society for Metals, Metals Park OH, 1984, P 587
153
154
w-
...='::::;
DUAL PHASE 1
16
AS-RECEIVED
STRAINED AND AGED
HI-FORM SO - - -
".
:e
...'"
'"
...u
u
.1--:10
__
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention,
American Society for Metals. Metals Park OH. 1981, P 213
'?
...
IJo
tJ>
Eeff
I(
0.2
0.4
00.6
.r
w
0
00
I-
::::i
e,
cta:
I-
CIl
...J
<l:
l-
e
I103
104
106
105
8
Eeff
'?
...
0
I(
.r
0.20
00.30
6
0.57
!:
...J
e,
<l:
ct
lii
a:
...J
<l:
2
0
I-
o
I103
104
105
106
Source: John M. Holt and Philippe L. Charpentier, "Effect of Cold Forming on the Strain-Controlled Fatigue Properties of HSLA
Steel Sheets," in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 218
155
156
po
)(
-r
::l
I-
:::i
a-
'lI
4
60.0
'90.06
a 0.08
00.16
1 Uniaxi81
T8nsion
881. Biaxial
Stretching
II:
In
..J
~
Runoutl
I-
Source: John M. Holtand Philippe L. Charpentier, "Effect of Cold Forming on the Strain-Controlled Fatigue Properties ofHSLA
Steel Sheets," in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, p 218
w
c
:::>
I-
::::i
0.
::l!i
2,-----------------,
0.1
:::>
0.1
I-
::::i
0.
::l!i
4(
TOTAL
.PLASTIC
AELASTIC
4(
4(
a:
:;:
I-
en
a:
I-
en
0.01
0.01
10 3
10 4
10 5
10 6
REVERSALS TO FAILURE, 2 N f -
10 7
0.001 '-----''---------'-----'-----'-----'
10 4
10 5
10 2
10 3
10 6
REVERSALS TO FAILURE, 2 N f -Strain-life curves for the Cb-V steel.
0.1
l!f
E
..J
~
4(
Z
:;:
a: 0.01
lii
.1OTAL
PLASTIC
AELASTIC
Strain-life curves: Strain-life data from constantamplitude tests for the three steels are plotted in the
three charts here respectively in the form of total
strain amplitude versus the number of complete
reversals to failure.
0.001'----'----'----'----L....:..---'
10 2
10 3
10 4
105
10 6
RI;VERSALS TO FAILURE, 2Nf
Strain-life curves for the Cb- V-Si steel.
Source: Shrikant P. Bhat, "Influence of Composition Within a Grade on the Fatigue Properties of HSLA Steels," in HSLA
Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 583
157
158
------ =........... --
600
_11
1
_
'==... - ....
.............
:E
:::
...
:-..
"'- ...:=:.-~
.
.....
~_
--;::J"
..- 0.0065
-.:-----0.003
~ .. 009--o_-.>-""""_ _ ~-0.0065
0' 2
...~boo
~
v
0---
0.0025
~200
0'-::10
'-;-
'-::-
'-::-
---'''--:-
--1
10'
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention.
American Society for Metals, Metals Park OR, 1981, p 209
..;
0.005
6
u
~ 400'r-o-OCo--o-----..,.--<-->--e..-"d""--ov-...o-O0.002
'"
~
o
~300
v
::::;
v
>-
v200
Ol.I
..L..,-
....L..:--
........-::-
--'-:-
--'
10 2
10 3
CYCLE NUMBER, N
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention.
American Society for Metals, Metals Park OH. 1981, P 208
159
160
.500
Go
:e
..__
E.. oo
=
~
'"
~-----o.--o_o(a
o
3 0 0 r - - o - - _....-
I~
....-
0.006
~hoo
.
..>
>v
Cl00
C
OL-
----''-;-
----'-:;-
--l.-:;-
--l.-:;-
---'
CYCLE NUMBER, N
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention,
American Society for Metals, Metals Park OH. 1981. P 208
162
0.010
DUAL PHASE 2
1'"
AS-RECEIVED
STRAINED AND AGED
HI-FORM SO - - - -
0.001L..:---------"---=,.-10 4
10 3
...L..,,---
..J
10 5
Source: N. Lazaridis and S. P. Bhat, "Fatigue Behavior of Cold Rolled Dual Phase Steels," in Wear and Fracture Prevention,
American Society for Metals, Metals Park OH. 1981, p 214
164
HSLA-1
G.S. =lOfLm
R =0.1
o 300K
t> 233K
Cl 173K
o 123K
Q)
!'/;f/:/.
10 ,
'/:
>.
o
"E
I:
g' i
t> 233K
Q)
I,
o~
',,
(PlI
0123K~I1!I:
, I
I
<,
01:
o 173K
o
>.
o
1/ /
1//,
o 300K
1/ t>A
I/,~
R=O.1
/?~ I
0,: t>.
HSLA- 2
G.S. = 10JLrn
Prof
, 6'
01 '
.,pI
,
I ,
/ t>,
I
Pi
/,/
P,
T(K)
300
233
173
123
3.6
5.1
6.5
10.8
T(K)
300
233
173
123
14
n
7.6
8.8
12,1
15.6
18 22 26 30 36
The only significant difference between HSLA-I and HSLA-2 is that HSLA-2 contains double the
amount of Nb that HSLA-I contains (see compositions on p 165).
The effect of temperature is seen to decrease the crack propagation rate with decreasing temperature
at low values of 6.K. However, as the stress intensity increases, a crossover occurs wherein higher
growth rates were observed, as shown in the above charts. This crossover is further reflected in the
increase in the Paris law exponent, n, where it ranged from 3.6at room temperature to 10.8at l23K for
HSLA-l. The large increase is a result ofthe change in the fracture mechanism from ductile transgranular fracture to cleavage. This behavior has also been seen in iron binary alloys where n increased from
3.5 at room temperature to 20.9 at l23K.
Source: Khlefa A. Esaklul, William W. Gerberich and James P. Lucas, "Near-Threshold Behavior of HSLA Steels." in HSLA
Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, p 569
165
HSLA-1
HSLA-1
G.S. 10pm
T-300K
G.S. 10pm
T -123K
R-O.IO
R-O.IO
R-0.35
R-0.70
10'
10
10 R-
0.35
R-0.70
e.
-,
c lO
"0
II
II
II 9 10
12 14 It III 20
2
24
II
II
II 910
1214161920
6K, MPa-m"
The effect of R-ratio on fatigue crack propagation behavior of HSLA-l at test temperature of 300
and 123K in the as-received condition.
HSLA-1
HSLA-2
..
.
Mn
Nb
SI
0.07
0.06
0.51
0.35
0.014
0.03
0.03
0.03
<0.005
0.01
Al
0.005
0.01
0.01
Ni
Cr
Fe
0.01
0.01
Rem
Rem
Source: Khlefa A. Esaklul, William W. Gerberich and James P. Lucas, "Near-Threshold Behavior of HSLA Steels," in HSLA
Steels-Technology & Applications. American Society for Metals, Metals Park OH, 1984, p 571
24
166
HSLA-1
HSLA-2
G.S. IOlJm
R-O.I
G.S. 10JLm
R' 0.1
e300K
.233 K
.173 K
.123 K
II)
u>u
.....
o
o
300K
233K
173K
123K
1I
.6K,
8 9 10
12 14 16 18 20
MPa-m"2
24
7 8 9 10 12 14 16 1820 24
Fatigue crack propagation behavior of two HSLA steels tested at temperatures of 300,233, 173 and
123K in the as-received condition.
The only significant difference between HSLA-I and HSLA-2 is that HSLA-2 contains twice
as much Nb as HSLA-I (for compositions of the steels, see p 165).
Near-threshold crack growth and threshold stress intensities for both steels in the as-received
condition are depicted in the above charts for all test temperatures. Comparison of crack growth
rates and threshold stress intensities at room temperature indicate that HSLA-2 has a higher
resistance to fatigue crack propagation than HSLA-l. The stress intensities amplitude, 11K, for
constant growth rates of IO- s and 10-9 ta] cycle are 2.0-2.5 MPa-m 1/2 higher in HSLA-2 than in
HSLA-l. The threshold stress intensityl1K'h' is also higher for HSLA-2 (8.0 MPa-m 1/ 2) compared to HSLA-I (5.5 MPa-m 1/2). The 2.0-2.5 MPa-m 1/2 difference in threshold and for the two
growth rates clearly demonstrates that there is an inherent difference in the fatigue crack propagation behavior of these two HSLA steels. This difference is also reflected at low temperatures,
where HSLA-2 showed lower crack propagation rates and higher threshold stress intensities
than HSLA-I. Furthermore, by comparison of threshold stress intensities for these two steels in
relation to the effect of decreasing temperature on increasing 11K/I" it was found that the ratios
of I1K,h (D t 11K,,, (300K) are the same for both steels.
Source: Khlefa A. Esaklul, WilliarnW. Gerberich and James P. Lucas, "Near-Threshold Behavior of HSLA Steels," in HSLA
Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 569
'II'EI.~E
""'-'~"---dE]r
~ ""
~~
~
=Jl
.~:~,:,.,
(00
~~~
lotAlERrAl
oOJtrElDl!1
I~.~
USIW 22->.....
'..po'- '.
-'ed'" , "
IoU5
I
I~
lHlER OYIW.IC
lCW> - cotI:XTJ:)N:S
OSIE 3eON
5' 52.
JO~
.=2C~
on
'<:
...... ~ ~--.....--.:.
....., ~
<,
10:
.....,~
....
~ - - - f=:-~
~-..::.
00
$~ ,,~
i
NJIIB[A
CYCLES
.q
~
I--
"
"._.-"---dEl
lHlER OYIW.IC
:;).('..,100
,.-1
""
,
, " ~.
i~!
I~
'00
LOAD - 00NXT10N9
WATEAW.
I~%
USIW 22J.E"'B
,'0;;',
.. tiS
T~
...,.,M"A.E- wac,
'-PO -,
OStEJOOn-l
-,
5152.3
...::""
)00
"""
."" F--",,~
",
~~
.",
---
"- 1--.0-
.";" c-c
0=
"'''''':
.-
sc.J
I
"".,.BR ~
CYCLES
Stress-cycles curves of welded samples of different materials under alternating bending load.
Fatigue data were derived from testing welded samples under tension and bending loads. It was surprising that under both types of load the HSLA steel and the
soft unalloyed steel hardly differed in fatigue strength; thus it can be said that the
use of HSLA steels is not justified if a component has a weld in the highest-stressed
area. An explanation for this is the loss of the thermal-mechanical effect, which is
responsible for increased strength, by the heat influence during the welding operation; and it is thought that a higher-strength manganese-alloyed steel, such as St
52.3 (according to DIN), the strength of which results from the chemical composition, would be more favorable in this respect.
Source: Klaus E. Richter, "Cold and Hot-Rolled Microalloyed Steel Sheets in Opel Cars-Experience and Applications,"in HSLA
Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 487
167
168
--6--
SAE-1006. Smooth
- - . - - SAE-980X. Smooth
- - - - SAE-980X. FCAW
- ..----.-........ -..
----.1;,.1---...- __
-----6-
-'"l:>"'&--C'IS"" 6-"tl.
10 5
106
The fatigue strengths of the smooth HSLA steel were higher than that of the low-carbon steel.
The I06-cyclefatigue limit stress of the smooth SAE 980 X steel was 469 MPa (68 ksi) and that
for the SAE 1006 steel was 283 MPa (41 ksi).
However, after welding, SAE 1006 and SAE 980 X steels exhibited similar fatigue properties
over the 104-1 06-cyclelife range studied. The I06-cycle fatigue limit stresses for FCAW SAE 1006
and SAE 980 X steels were between 114 MPa (16.5 ksi) and 117 MPa (17 ksi).
Source: Kon-Mei W. Ewing. Pei-Chung Wang. Frederick V. Lawrence, Jr., and Albert F. Houchens, "Weld Fatigue of TIGDressed SAE-980X HSLA Steel." in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH,
1984, P 556
1000
000
"00
100
r---~
__
600
Parent IIletal
sou
400
TIC-treated bull
weld
JOO
TlGtreUed fillet
weld
~ '.00
Untreated fillet
100
50 '--
.L-
.L-
-'-
105
106
.L..-........
CYClES TO FAILURE
Fatigue strength for DOMEX 640 XP. Standard-Wohlerdiagram (log-log scale) with pulsating load (R=min
stress/max stress=O). Sheet thickness 5 mm and ultimate
tensile strength 767 MPa.
.
Source: Tony Nilsson, "Formable Hot-Rolled Steel With Increased Strength," in HSLA Steels->Technology & Applications,
American Society for Metals, Metals Park OH, 1984, P 259
169
170
10
103
--.::::::::::::
CJ)
~
...
-,
(U r =-87KSI
.e'!:
en
<I
10
J!.
:E
..'=ur = 87 KSI
10 2
R = 0.1
EXPERIMENT
- - PREDICTION
I 3
10
101
105
NT,
106
108
CYCLES
Total fatigue life estimates compared to the experimental data for the FCA WITIG-dressed
SAE 980 X steel.
It should be emphasized that life estimates made on the FCAW{TlG-dressed welds were
based on geometry changes brought about by TlG-dressing. The other beneficial effects such as
removal of slag intrusions and inclusions were not considered. The close agreement between the
calculated and observed long-life fatigue properties suggested that the majority of fatigue improvement seen in TIG-dressed joints was attributable to the geometry change. The smaller
flank angle contributed significantly to the increased fatigue strengths of TIG-dressed
weldments.
Source: Kon-Mei W. Ewing, Pei-Chung Wang, Frederick V. Lawrence, Jr., and Albert F. Houchens, "Weld Fatigue of T1GDressed SAE-980X HSLA Steel," in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH,
1984, P 563
en
<I
SAE-980X
R= 0.1, t= 0.13" 13.3mml
TIG-Dressed
_.-(}-.- As-Welded
---0---
~-------4..II
~o
o '-'il... 0
__ ~
-----ZP-n---..I\
~00o
0
o
,-.~
-0....
o Jgt. '1I'o-."'"b-
'[jCD-.,_.
00
0
-c._
'-
''''"00'_
10 5
,-"",..It.0
o'-'n-
10 6
Source: Kon-Mei W. Ewing. Pei-Chung Wang, Frederick V. Lawrence. Jr., and Albert F. Houchens, "Weld Fatigue of TIGDressed SAE-980X HSLA Steel," in HSLA Steels-Technology & Applications, American Society for Metals. Metals Park OH,
1984. P 558
171
172
en
<]
10 1
SAE980 LAP-SHEAR WELDS
Kfmax =3.49 I R = 0.1
EXPERIMENT
PREDICTION
105
106
NT, CYCLES
Total fatigue life predictions and experimental results for the FCAW, SAE 980 X lap-shear joints.
Source: Kon-Mei W. Ewing, Pei-Chung Wang. Frederick V. Lawrence, Jr., and Albert F. Houchens, "Weld Fatigue of TIGDressed SAE-980X HSLA Steel,"in HSLA Steels-Technology and Applications, American Society for Metals, Metals Park OH.
1984, p 562
STRESS - LMTS
. v; ...~
:.~"",.TC'SS'O\T~ST
.~
. ':._':'0
a: WELDED SN.f'lES
"-'-dEl
10
)10.5
~~
::.
"
'0(>0'
-=
lC~
nn
'l;:
I~:
'~
oi~hOl
~~
-;.'
051E 38CN
, , ",
>l
"3
:.-........-.:..: ~
......,
'-.I,!~
~ ..... -...;:..-" ~~
.............
l:SIW n."E."~'"
-.
00
--
'---
......l[R'.l1.
",-
"00',
101"
'00
iec
COIClIJ1DN8
Lp"o -
I~""
I~
lHlER OYNAlolC
w(lO~O'""
II
lA.IM8ER(JFC'fC~t'S_
STRESS - LMTS
..
1.C.... '00
.. n
~~:
,co
lHlER OYNAlolC
a: WELDED SN.f'lES
"-"-dEi
10
Lp"o -
COIClIJ1DN8
~~~-"'S~~n
)IQ.S
I~%
'. o~
,.11'
'-~
IUTERlAl
~%
USIW 22."'.',
-.
OSIE380T'4
",
SI S23
'-="
'00
""""
'" f"........C'l>~
'00
~~
".,
--~
~--'=
=---=
r-''!!::='- ..
"=
--
-II
III
I
fUo'8(A OF CVClS-
Source: Klaus E. Richter, "Cold and Hot-Rolled Microalloyed Steel Sheets in Opel Cars-Experience and Applications,"in HSLA
Steels-s-Technology & Applications, American Society for Metals, Metals Park OH, 1984, P 487
173
174
LOAD AMPLITUDE
6P, kips
AND NUGGET ROTATION
68 N , degree
5.0
0.2
1.0
0.5
GroupC
LOAD AMPLITUDE
6P. kips
AND NUGGET ROTATION
68 N degree
5.0
2.0
1.0
0.5
LOAD
AMPLITUDE
6P, kN
5.0
2.0
1.0
0.5
Group K
Results of spot-weld fatigue tests are presented in the four plots (above and
on the facing page) for the stated conditions. Each curve shows the load
amplitude, I1P, and nugget rotation values, 116 N' for each test as a function of
cyclic life. Straight lines were fitted through the data.
1~oe""",roOlI"
LOAD AMPLITUDE
dP, kips
AND NUGGET ROTATION
d(~N' degree
5.0
175
~::S';=I~~:"
LOAD
AMPLITUDE
dP, kN
2.0
1.0
0.5
Group E
5.0
2.0
1.0
0.5
5.0
0.2
1.0
0.5
d0
dP
5.0
LOAD AMPLITUDE
dP, kips
AND NUGGET ROTATION
:>SN. degree
2.0
1.0
0.5
~
-- -
5.0
2.0
1.0
0.5
"W-IOI61M11~O"1
' 02 _ " " . '
0-66In"lf02fj1fl1
::~'
GroupN
5.0
2.0
1.0
0.5
de
dP
Group 0
5.0
2.0
1.0
0.5
1,.1 0211Yn1004..., ,
W .. l01611'WT1140n,
0 .. 66 ........,0416 ... ,
dP
de
Source: James A. Davidson, "Design-Related Methodology to Determine the Fatigue Life and Related Failure Mode of SpotWelded Sheet Steels," in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984, p 542
176
~::;;
800
<l 600
b
g 400
I-
:J
Q.
9.5mm
~ 200
6.4
3.2
~
tii
100
80
~ 60
0:
~z
1.6
0.80
40
0.40
0.20
20
2
3 4 5 6 8 10
3 4 5 6 8 10
3 4 5 6 8 10
~a,
~:::!:
4000
3000
2000
1000
~ 800
"
~
<l
p=0.20mm
600
500
400
300
200
2
3 4 5 6 8 10
3 4 56 8 I
3 4 56 8 I~
Curvature ofthe notch and I:1Kis the stress-intensity amplitude computed for an imaginary crack whose length is the same as the notch depth, a. Barsom and McNicol used this
parameter to compare N j , the cycles to fatigue-crack initiation, in HY-130 steel for
notches of constant depth but various radii of curvature. The results are shown in the
above graphs. In the upper graph, N, is plotted versus l:1a, where N, is defined as the
number of cycles to give a 50-J..!m side notch. There is a wide spread in the curves. As
expected, the sharpest notch, lowest p, gave the most rapid initiation at a given stress. The
lower graph shows I:1K/ vp plotted versus N;. A narrow-spread family of curves results;
these converge as the value of I:1K/vp is decreased to a threshold value I:1K/ vp I'h' the
minimum value to initiate fatigue cracks in notches.
Source: M. E. Fine and R. O. Ritchie, "Fatigue-Crack Initiation and Near-Threshold Crack Growth," in Fatigue and Microstructure, American Society for Metals, Metals Park OR, 1979, pp 256-257
177
300
s:::;:
1500
~.
~E
200 1;;
E
:>
E
:>
.~
.~
::;:
::;: 1000
100
Notch
500
severity
(107 cycle
IifeUmel
_~00'::0:----...,."..::-----,.L..._-_--L._------"L-_-_.L-_-_--L._-----,,------..,..L...------'25~0
-1500
-1000
--500
2000
1000
1500
Constant-lifetime fatigue diagram for 300 M alloy steel, hardened and tempered to a tensile strength of 1930
MPa (280 ksi). Solid lines represent lifetimes obtained from unnotched specimens. Dashed lines represent
lifetime of ten million cycles for specimens having the indicated notch severity.
Source: Metals Handbook. 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals. Metals
Park OH, 1978, P 670
178
IlXXl
1200
1400
MN/m2
1600
1800
2lXXl
2200
1400
200
R 0.1
TRIP
~175
1200
u
~150
.,.
SR 4340
IlXXl"/;
~125
v;
4ll
800
:J
.2'
.l!loo
600
Studies on fatigue-crack propagation (FCP) conducted under controlled stressintensity amplitude (~K) conditions indicate that deformation-induced transformation retards crack growth in lower-strength metastable austenites, particularly at low
~K, and also exerts a beneficial influence in high-strength TRIP steels, although to a
much lesser extent. This growth retardation may be due to crack-closure effects arising
from the transformation volume change, which may be particularly effective in the
fatigue-threshold regime. Smooth-bar fatigue properties appear to be dominated by
transformation hardening, which is desirable under stress-control conditions (reducing strain amplitude) but generally undesirable under strain-control conditions (increasing stress amplitude). In lower-strength austenites, transformation reduces fatigue life under conditions of controlled plastic strain amplitude; under controlled
total strain amplitude, transformation is detrimental to low-cycle fatigue life, but a
small amount of transformation may be beneficial at high cycles. Similarly, the lowcycle fatigue properties of high-strength TRIP steels are found to be degraded by transformation under controlled total strain amplitude. Under stress control, the fatigue life
of lower-strength austenites is greatly enhanced by transformation; for a stress ratio
(R= amin/ a m x ) of 0, fatigue limits in excess of the yield strength are observed. Investigation of the smooth-bar fatigue properties of high-strength TRIP steels at R= 0.1, in
which thermodynamic stability was varied by heat treatment, also revealed transfermation enhancement of fatigue life. Such enhancement allows the achievement of exceptional fatigue strength at high ultimate strength levels, as illustrated by comparison
with other high-strength steels in the above graph.
Source: G. B. Olson, "Transformation Plasticity and the Stability of Plastic Flow," in Deformation, Processing, and Structure,
George Krauss, Ed.. American Society for Metals, Metals Park OH, 1984, p 419
Q,
UI
UI
80,000
50,000
....
a:
lUI
40.000
....0
::i
lL
lL
ce
30.000
AB- IN AIR
C0 - IN ACID BRINE
EF- IN ACID BRINE WITH INHIBITOR
20,000
10.000
102
10~
t04
NUMBER OF CYCLES
10'
FAILURE
10'
to'
Effect of corrosion and corrosion inhibitors on the SoN curve for highstrength steel (sucker-rod material).
After the first brittle crack is initiated, No.2 is the slow step in the
process and electrochemical action is the slowest part of this step.
Thus, the effect of corrosion can be illustrated with curves of stress vs
logarithm-of-number-of-reversals-to-failure for sucker-rod steel. Corrosion accelerates cracks propagation, so the fatigue curve drops from
AB to CD, as shown in the graph. Deceleration of the slow stage with a
corrosion inhibitor will raise the S-Nfatigue curve from CD to EF.
Source: Joseph F. Chittum, "Corrosion Fatigue Cracking of Oil Well Sucker Rods," in Corrosion: Source Book, Seymour K.
Coburn, Ed., American Society for Metals, Metals Park, OH, 1984, P 380
179
180
.8
Ul
~.6
Ii
0
ct:
~.4
...
0
:J:
I-
I
I
~.2
UJ
II
oJ
I
I
II
0
FIRST STAGE
SECOND STAGE
!THIRD,
ISTAGEI
0
RELATIVE NUMBER OF CYCLES, PERCENT OF FAILURE
Source: Joseph F. Chittum, "Corrosion Fatigue Cracking of Oil Well Sucker Rods," in Corrosion: Source Book, Seymour K.
Coburn, Ed., American Society of Metals, Metals Park OH, 1984, P 378
181
300
o.. ~
275
i---::::::::::::
o~
250
200
(f)
(f)
Q)
...
175
( /)
"0
150
Q)
a.
a.
<t
i'...
-,
~
125
100
..1
-,.
75
\
<,
Bake
---
24 h r -
Bake 18 hr
Bake 12 hr
Bake 7 hr
Bake 3 hr
Uncharged
+-+-
,~
:---:-
0\ \
\
-\ ~
+
+
"
.-
-\
~.
~~
"\\'\
a. 225
(f)
0
0
0
i'-..
Bake 0.5 hr
--
<-
1-
50
0.01
0.1
1
10
Fracture Time, Hours
100
1000
Static fatigue curves for various hydrogen concentrations obtained by baking different times
at 150C (300 OF). Sharp-notch specimens. 230,000 psi strength level.
These are, in essence, static fatigue curves, and the lower critical stress may be considered a
static endurance limit-that is, a stress below which failure will not occur for an indefinite period
of time. This behavior is sensitive to hydrogen concentration as shown above, where it may be
seen that all delayed-failure parameters-notch strength, rupture time, and static fatigue limit
increase with decreasing hydrogen concentration. Also, even after 24 hours at 150C (300 OF),
there is still a substantial stress range, of the order of 60,000 psi, over which delayed failure will
occur. In an unnotched specimen, full recovery ofthe ductility as measured by the reduction of
area can be attained in less than 20 hours at 150C (300 OF), yet delayed failure will occur after
24 hours or longer of baking time at 150 C (300 OF).
Source: Alexander R. Troiano. "The Role of Hydrogen and Other Interstitials in the Mechanical BehaviorofMetals,"in Hydrogen
Damage Source Book, Cedric D. Beachem, Ed., American Society for Metals, Metals Park OH, 1977, P 154
182
0
0
0
225
T~.
I~~ 1\
'\
200
en
en
Q)
L..
175
( J)
"0
Q)
0..
0..
150
-a_
""'"
0
01'\.
~\
125
~\\
Notch
It-- - - -
>--
~.
75
; a1lus
J. = 2'In.
>--
>--
-\..
100
Notch
\\
i'\
.....-
50
0.01
0.1
10
100
1000
The variation of lower critical stress with notch severity is shown in this diagram. It is evident
that the static fatigue limit rises as notch severity (radius) decreases for hydrogen-charged
high-strength steels (using the same baking time).
Source: Alexander R. Troiano, "The Role of Hydrogen and Other Interstitials in the Mechanical Behavior of Metals," in Hydrogen
Damage Source Book, Cedric D. Beachem, Ed., American Society for Metals, Metals Park OH, 1977, P 155
CYCliNG
KEY
CONTINUOUS
..J
<
e
l-
....
lLL
loO
l.'
<
Z IX:
IX: I-
<
VI
VI
10 0
30 on HOLD
"
.',~""--............:..
AIR
VAC ...
..........,,:~:-:-: ....
" .""'"
".. . . ..:r.
. "=':..,:-:. . :--;; -
I::NV. TEMP.
AIR 275 K
VAC . ..
AIR 775 K
VAC. II
--.
__
.....::--........
._~.
CYCLES TO FAILURE
Effect of hold time in air and vacuum upon the fatigue endurances of a
O.5%Mo steel at 275 and 775 K.
Source: R. H. Cook and R. P. Skelton, "Environment-Dependence of the Mechanical Properties of Metal sal High Temperature,"
in Source Book on Materials for Elevated- Temperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals
Park OH, 1979, p 83
183
184
7.2 DIN 14 Steel (1.5 Cr, 0.90 Mo, 0.25 V): Effect of
Liquid Nitriding
1000
'\
900
- 140
:'\.
""-8
800
700
I'
~ 600
.s
13
IS
- 120
- 100
0A
- 80
';;;
-'"
.~
13
>-
500
400
300
- 60
\
~
105
106
- 40
107
K,=2.
Nitriding introduces residual compressive stresses at the
surface of steel parts; these residual stresses, together with the
increased strength of the nitrided layer, increase the fatigue
resistance of the part. The increase in fatigue strength that results from nitriding is illustrated in these S-N curves.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 541
KEY
3.0
2.0
z
..:
a: 1.0
fVI
0.8
-.
~XX
ENVIRONMENT
Na (30 ppm OF 21
AIR
HELIUM
. ,
x~
'~
'U ...~~~.
........~
-~
.J
u 0.6
>u
....
11K
- .. ~
.........=-
0.4
0.3
0.2
C.I
10~
CYCLES TO FAILURE
Source: R. H. Cook and R. P. Skelton, "Environment-Dependence of the Mechanical Properties of Metals at High Temperature,"
in Source Book on Materials for Elevated-Temperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals
Park OH, 1979, P 83
185
186
Testing temperature
~C
o 425
,
?f!.
OJ'
"Q
540
595
l1
FI
800,
1000
1100
.'=
a.
E
c
'iii
'"
J5
0.2
0.1 2
10
103
Cycles to failure
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 659
104
:e
B
:G
>
u
103
500
450
400
350
550
Temperature,OC
I
600
700
800
900
1000
1100
Temperature, of
Strain-controlled fatigue tests have also shown (note above) that reducing
carbon content to 0.03% results in a reduction in fatigue strength. Furthermore, because of variations in strain aging effect, specimens from one heat
with a higher carbon content ran longer at 427C (800 OF) than at 316 C
(600 OF).
Source: Metals Handbook, 91h Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, p 659
187
188
10
20
60
I--+-------+-------+--------+-----i
10.4
Testing temperature
10-31--+--oC 205
o 370
of - - - + - - - - - - - + - - - - - - - - - - - f + - - - l
400
700
1---+-6-455-850---+--------+--------,.~____J<-----,_F____I1____l
6595
1100
10-6
Q>
Q>
~
.
E 10-4
E
:i
\!
s:
~
...e
s:
~
0
l;,
.><
u
.><
eu
10.6
:e'
l;l
1;
:e'
..
:!'!
:!'!
"D
"D
10-6
10
20
60
Source: Metals Handbook. 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978. P 660
2
I 0-4
4C"7f'/
/,r:/
/ ! vr:~400C~
/
40 CfKO
2
Frequency,
cvcles/mln
o 4
40
2
2
(.1
IJJ.K. stress intensity factor, ksi
2
vrn:
2
I
1-//
~E
4C~
II;
t
III
1/
2
I
40c"m
400c:pm
2
Frequency.
cycles/min
o
4
40
2
2
5
4K. slreu intensily 'actor. MPa V;;
Ibl
189
190
10-6
v
i>.
i>.V
vi>.
Vi>.
i>.
~
-.
.Jl
i>.
10-7
-.
'"
al
i0
.dt
10-8
Ol
o~ ,
o ,
Gl
:::I
oOe
.l'
~
u
0'
oil'
't:l
.....
't:l
'"
s:
;.
Gl
1i
10-9
0.,.
l-
Ol
.;::;
'"
u..
8
10- 10
Frequency
f-
10- 11
f:
2% Cr - 1 Mo Steel
R = 0.05
V
i>.
Environment
Air
2 Hz, 50 Hz
H,
0 50
5 Hz
138 kPa H2
2 Hz
0.5 Hz
~I Threshold
,f
6 78910
20
40
60
80 100
Corrosion fatigue descriptions are further complicated by the fact that the environment may produce multiple effects. For example, Suresh et a/demonstrated that
dry hydrogen may produce afrequency-sensitiveenvironmental effect analogous to
SCC at intermediate t:J.K values and a frequency-insensitive environmental effect
near the threshold. This is illustrated in the above graph for 2.2SCr-1Mo steel tested
in air and in 138-kPa hydrogen gas. Because the sustained-load threshold for this
steel is on the order of 90 M Pa . m 112 (82 ksi . in. 112), the K""'j of about 22 MPa .
. 1/2) gives
K th :
v.
m 1/2 ( 20 k'
SI . Ill.
K ,h ({) ...?
~
lt can be seen for t:J.Kvalues greater than K ,h ({) that there is a large increase in
growth rate for the low test frequencies but not for the higher ones. Therefore, this
regime may be considered to be one where superposition might apply. In addition,
however, there is a true threshold, t:J.K,h , which appears to be frequency-insensitive
but which nevertheless decreased by about 30% to SA MPa . m 112 (4.9 ksi in. 112 )
because ofthe hydrogen environment. Such mulitple effects are poorly understood
and are clearly possible in a large number of material/environment systems.
Source: W. W. Gerberich and A. W. Gunderson, "Design, Materials Selection and Failure Analysis," in Application of Fracture
Mechanics for Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds..
American Society for Metals, Metals Park OH, 1982, P 333
191
120days
Compressive or
tensile hold
Type of strain hold
Strain
range.%
Compressive
Tensile
Both
2.0
loJ
104
Cycles to failure
1.0
C.}
0.5
:::2 } .
Cycles to failure
(b)
In these "time-to-failure/cycles-to-failure" diagrams, the lowest curve (zero hold time) indicates the corresponding time period and number of cycles to failure for continuous strain-controlled fatigue tests over the
strain range from 0.4 to 2.0% with no holding period at maximum strain. The other curves, which are
approximately parallel, are for increasing periods of holding time at maximum strain levels in either tension
or compression. The vertical curves are drawn through the number of cycles to failure for each particular
cyclic strain. For all tests at 2% strain, failure occurred in less than I000 cycles regardless of holding time or
whether the stress was tensile or compressive. The effect of reducing the strain increment and increasing the
holding time on number of cycles to failure can be determined from the appropriate curves in the figures.
Source: Metals Handbook, 9th Edition, Volume I. Properties and Selection: Irons and Steels, American Society for Metals. Metals
Park OH. 1978. pp 662-663
192
5.0
I 4.0
lLJ
o
Z
<l
cr
3.0
<r
cr
I-
(f)
2.0
...J
g
1.0
100
200
400 600
1000
2000
CYCLES TO FAILURE
4000
Source: Steel Castings Handbook, 5th Edition. Peter F. Weiser, Ed. Steel Founders' Society of America. Rocky River OH. 1980,
P 15-55
7-11. H11 Steel: Crack Growth Rate in Water and in Water Vapor
0.5 ,........----,-----,--"'"""'T--"'"""'T--~--....,
"0
0.3
c
E 0.1
.....
c
Q)
.05
H -II Steel
230 ksi Y. S.
K - 30 ksi IT'"
o
Activation Energy
9,000 cal./gm-atom
o
a:::
-...
s: .03
.Jl:
U
...
o
u .01
o Water
o Relative Humidity 100% at Test Temperature
0.005
II
II
II
II
80 F,
Crack growth rate versus temperature for an HII steel in water and
water vapor.
Source: Herbert H. Johnson. "Keynote Lecture: Overview on Hydrogen Degradation Phenomena," in Hydrogen Embrittlement
and Stress Corrosion Cracking, R. Gibala and R. F. Hehemann, Eds., American Society for Metals. Metals Park OH. 1984, P 18
193
194
LOW C 9% Cr 2% Mo AT 550C
CONTINUOUS CYCLING
(TOTAL STRAIN RANGE; FROM (2311
~
UJ
l-'
Z
10
cr
cr
~
Vl
9% Cr 2% Mo NbV
AT 550C
(PLASTIC STRAIN
RANGE; FROM (24))
~--5
MIN. TENSILE
DWELL
CYCLING
10'
CYCLES
TO FAILURE
Illustrating the elevated temperature low-cycle fatigue and creep-fatigue properties of normalized and tempered 9% Cr Mo variants.
In this chart are presented the elevated-temperature-fatigue and creepfatigue data for the 9%Cr-1%Mo steel as a single curve in terms of total strain
range against cycles to failure; also shown for direct comparison are the continuous cycling fatigue data for the low-C, 9%Cr-2%Mo variant which, although inferior at relatively high strain ranges, suggests superior endurance
may be attained in the high-cycle region. From the limited evidence, it seems
probable that normalized and tempered 9%Cr-I%Mo steel may be used in
reactor-quality sodium at service temperatures with little effect on tensile
properties and stress rupture strengths or ductility and that the short term
low-cycle fatigue endurance will be increased and fatigue crack growth rate
reduced. This behavior is a consequence of the structural stability of the
material with respect to interstitial element transfer in liquid sodium and also
the low oxygen potential of the overall system which may be expected to
preclude oxide penetration and enable partial recohesion of the crack faces
during fatigue.
Source: S. J. Sanderson, "Mechanical Properties and Metallurgy of 9%Cr I%Mo Steel," in Ferritic Steels for High-Temperature
Applications, Ashok K. Khare, Ed., American Society for Metals, Metals Park OH, 1983, P 95
-------------------------200 t
- - - - - -__
en
w
a::
------ --- ..
----Continuous Cycle
100
e
a..
::E
(/)
l-
(/)
-100
- - with 30 sec T H
---
-200
r,.,..,,"'="::=-:-:.....---......:_:'":_;:"::-=-=_,,:_::- - - _...
10
100
1000
10000
CYCLES
This chart shows stress amplitudes (tensile and compressive)
that developed in the course of cycling the modified Fe-9.0Crl.OMo steel through a total strain range of 0.5% at 649C
(1200 OF). Fatiguing was carried out in vacuum. Dotted curve
indicates continuous cycling; solid curve indicates cycling with a
30-s hold at maximum tensile strain.
Source: S. Kim, J. R. Weertman, S. Spooner, C. J. Glinka, v. Sikka and W. B. Jones, "Microstructural Evaluation ofa Ferritic
Stainless Steel by Small Angle Neutron Scattering," in Nondestructive Evaluation: Application to Materials Processing, Otto Buck
and Stanley M. Wolf, Eds., American Society for Metals, Metals Park OH, 1984, p 175
195
196
cb:
dQ
In 10 8
8
6
4
'i
Fotl(~ued
Foti~ed ( 649C, At
79
= 0.5 %,
cycles with 30 sec tension hold
Crepl
R~9~~,M~fb28
holn)
~
~
-2
-4
0.02
0.04
0.06
0
0
0.08
<I")
Source: S. Kim, J. R. Weertman, S. Spooner, C. J. Glinka, V. Sikka and W. B. Jones, "Microstructural Evaluation of a Ferritic
Stainless Steel by Small Angle Neutron Scattering," in Nondestructive Evaluation: Application to Materials Processing, Otto Buck
and Stanley M. Wolf, Eds., American Society for Metals, Metals Park OR, 1984, P 175
197
60
40
80 100
10- 3
GI
i3
~
......
10- 2
E
E
"t:l
......
I /
'"
"t:l
......
GI'
'"
/ /
..c
GI
i3
10- 4
.~
// /
i0
... 10- 3
'"
u
...u
'"
GI
:::I
'"
u..
'"
.;;
10- 4
L-.J'-
/
/
/
/
/
/
>u
'<,
"t:l
......
'"
"t:l
/
10-5
1/2 hard tvpe 301
24C (75F)
0.063 < R < 0.807
--'-_ _.1...---J'---'---'---'-...J...........L - - J
10- 6
80
Effective stress-intensity factor,
K.1f = Kmax [1 - Rl o.66 7 , MPa m 1/2
Scatter band offatigue crack growth rates of Y2hardtype 301 stainless
steel, tested at 24 0 C (75 OF), 10 Hz, andR ratios of 0.063 to 0.807 based
on effective stress-intensity factor, Kerr'
Fatigue crack growth rate data reported by Walker for Y2-hard type 301 stainless steel sheet are
summarized in the above graph. The data were obtained in air at room temperature over a series ofload
ratios (R) from 0.063 to 0.807 at a frequency of 10 Hz. These data are based on the "effective stress
intensity factor," Kerr, rather than on fj,K, to account for the effect of the range of stress ratios. Kerr is
defined as follows:
Kerr = K max (I - R)m
where m is determined empirically and R is the load ratio (minimum load/maximum load) on cyclic
loading. The crack growth rate law then becomes:
da/dN= C[Kmax(l- R)my
Results of fatigue crack growth rate tests on austenitic stainless steels have shown that the crack growth
rate tends to increase as the R ratio is increased, when compared at given values of fj,K. If tests are made
at several load ratios to determine m, then the effects of other load ratios may be estimated.
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials. James E. Campbell, William W. Gerberich and John H. Underwood, Eds.. American Society for
Metals. Metals Park OH, 1982, P 114
198
10- 5
Gl
>
c.>
I"'"
--.E
--'"
2'
"'"
"tl
"tl
10- 6
20C (68F)
L-T
Type 301
30
40
50
2 X 10- 7
60
Fatigue crack growth rates for type 301 stainless steel have been
reviewed by Pineau and Pelloux in the temperature range from - 30 to
+95 C (-22 to +203 OF).The results, summarized in this graph, were
obtained on compact specimens 7 mm (0.28 in.) thick at a cyclic frequency of 20 Hz with a sinusoidal waveform at a load ratio (R) of 0.0 1.
All specimens were tested in dry argon except one series that was tested
in laboratory air. For the annealed specimens tested in argon, fatigue
crack growth rates at a given t::.K value increased as the temperature
increased over the testing temperature range. Fatigue crack growth
rates in laboratory air at 20C (68 OF) were higher than for corresponding conditions in argon, indicating that the humidity and/ or oxygen in the air influenced the growth rates.
The warm worked specimens were reduced 65% at 450 to 500C
(840 to 930 OF), resulting in a substantial increase in strength. Fatigue
crack growth rates for the warm worked specimens (above) indicate
that the fatigue crack propagation properties of the warm worked alloy are different from those of the annealed alloy. This effect of warm
working has been observed for other austenitic stainless steels. These
differences are attributed to the extent of the strain-induced transformation at the crack tip. This transformation effect would be most noticeable in type 301, because it is less stable than the other alloys in the
UNS S3xxxx series.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OR, 1982, P 113
LIJ
.....J
<.,;)
..
430C
650C
8160C
(f)
a:
.....
(f)
Data of Berling and Slot for AISI 304stainless steel, showing frequency-modified elastic and plastic
strains at three temperatures in air,
In contrast to most other segments of our technology, interest in the fatigue problem in the power-generation industry generally involves elevated
temperature. Laboratory testing on both smooth specimens and specimens
designed for crack growth is performed with temperature and frequency or
strain rate as parameters. The importance offrequency or strain-rate effects
is shown in this chart. These data are for solution-treated AISI 304 stainless
steel subject to triangular wave shapes at equal-loading and reverse-loading
strain rates. Representation of the behavior here utilizes fatigue equations
known as frequency-modified fatigue equations. They describe the elastic
and plastic strains versus fatigue life and include the frequency or strain rate
of the test. For the present purposes they are useful in showing how increasing temperature acts to change the cyclic stress-strain response and the strainlife fatigue response of this alloy.
Source: L. F. Coffin, "Fatigue in Machines and Structures-Power Generation," in Fatigue and Microstructure, American Society
for Metals, Metals Park OH, 1979. P 13
199
200
8-4. Type 304 Stainless Steel: Fatigue Crack Growth RateAnnealed and Cold Worked
~K, ksi . in. 1/ 2
20
10
30
40
60
.,
U
I
I ~//
I
10- 3
>
u
-<;
E
E
/
/
.,
I'
'"
"t:l
....,'~
...s:
s:
IJI,'
Cl
-"u
>
u
'<,
.~
z'
Cold worked
25%, tested
at 25C
(77F)
10-5
"t:l
-..
'"
"t:l
/it'
.,u
::>
Cl
'':;
'"
10-4
y/~
z'
"t:l
'<,
U.
,1:"
I
I
Annealed, tested at
427C (800F)
-
80 100
2 X 10-4
/
10- 4
"
Annealed,
tested at
25C (77F)
Type 304
........
10
-'--_--'-_....L.._---'_....L.----lU
20
30
40
60
10- 6
80 100
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels." in Application of Fracture Mechanics for Selection of
Metallic Structural Materials. James E. Campbell, William w. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, p 120
10
20
40
60
80 100
10- 4
10- 5
~Roomair
10- 4
20
Type 304
25C mOF)
30
40
The effects of humid air environments on the room temperature fatigue crack growth rates of
specimens of annealed type 304stainless steel are shown in the above chart for specimens cycled
at 0.17 Hz with an R ratio of zero (Shahinian, Watson, and Smith). At the lower end of the t:.K
range, fatigue crack growth rates in humid air are substantially greater than crack growth rates
in dry air. However, fatigue crack growth rates of specimens oftype 304stainless steel tested in a
pressurized water reactor environment at 260 to 315 C (500 to 600 F) with R ratios of O. 2 and
0.7 were no greater than the fatigue crack growth rates in air at the same temperature with an R
ratio less than 0.1 (Bamford). However, variations in R ratios influenced the fatigue crack
growth rates in the pressurized water reactor environment.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH. 1982. p 122
201
202
10-4
10- 3
.,
>
lJ
-.~
10- 5 Z
~
'"
"t:J
10- 4
Unaged
Aged
Hold time
Zero
0.1 min
1.0 min
10- 6
Type 304
593C (1100F)
10-5
.l--_ _....L..._--'-_....I....----I_L-.L-J........J
L-..JL-.I--'-
10
20
40
60
80
100
Because austenitic stainless steels are expected to give long service life, an evaluation of
the effect oflong-time aging at service temperatures is important. Results offatigue crack
growth rate tests on specimens that were tested in the unaged and aged conditions (5000
hours at 593 C, or 1100 OF) are shown in this graph, as reported by Michel and Smith.
After aging for 5000 hours at this temperature, precipitation of M 23C 6 carbides is essentially complete. These results indicate that at 593 C (1l00 OF) there are no deleterious
effects of aging on the crack growth rates of specimens that are continuously cycled. When
a holding time of 0.1 or 1.0 minute is included in each loading cycle, there tends to be a
slight increase in the fatigue crack growth rate at a given 11Klevel.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, P 121
10
20
649O C l 1 2
.......
E
E
10-3
10-4
t<,
1/ .:
"tJi'"
..
.'"
"".
'"
100
538"C (1000"-><-/
i0
60
II>
c:;
>u
40
10- 4
f!
u
u
II>
:::l
t'I~
II>
10- 5
'"
u,
10- 5
"'"
"
.......
~~~
10-6
316C 1600F)
10
>-
u
.......
.E
z
~'
...'"
c:;
20
Type 304
40
60
100
Results offatigue crack growth rate tests on types 304 and 304L stainless
steel at room temperature and at elevated temperatures have been reported by
James and Schwenk, and by others. As shown in this graph, increasing the
exposure temperature from room temperature to 650 C (1200 F) increases
the fatigue crack growth rates at any ~Klevel within the range ofthe tests in
an air environment. These data, reported by James and Schwenk, are for
specimens of both the L-T and T-L orientations, for several different maximum alternating loads, for load ratios of 0 to 0.05, and for cyclic frequencies
from 0.033 to 6.66 Hz for the room-temperature tests and 0.067 Hz for the
elevated-temperature tests.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell. William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park DB, 1982, P 115
203
204
8-8. Type 304 Stainless Steel: Damage Relation at 650C (1200 OF)
STRAIN RATE,
IN.lIN.lSEC
4 x 10'3
o
().
~
'i7
HOLDPERIOD, MINUTES
TENSION
COMPRESSION
o
0
I
0
10
0
30
0
60
0
180
Ll
o
~
(3
0
30
3
30
30
0.1
0
0
3
30
3
0
0
0
-tI
4 x 10'3
4 x 10'4
4 X 10'5
O"rtlpN,
I3 v l3 ( k. I ) =
C = 1.158
x 10 5
13 = 0.895
k
= 0.756
The damage function was proposed by Ostergren and is based on the frequencymodified fatigue approach. A damage function is approximated by the quantity U,!:iE p ,
where u, is the maximum stress in the cycle and !:iE p is the inelastic strain range. The tensile
hysteresis energy is employed to account for the facts that low-cycle fatigue is essentially a
crack-growth process and that crack growth and damage occur only during the tensile
part of the cycle. The use of the tensile-stress quantity, in conjunction with the plasticstrain range, provides a means of accounting for loop unbalance, since, for the same
inelastic strain, a positive mean stress provides a greater hysteresis energy than does a
compressive mean stress. The method is effective in accounting for hold-time effects, as
indicated in the chart above.
Source: L. F. Coffin. "Fatigue in Machines and Structures-Power Generation, "in Fatigue and Microstructure, American Society
for Metals, Metals Park OH, 1979, P 23
8-9. Type 304 Stainless Steel: Fatigue Crack Growth Rate at Room
and Subzero Temperatures
10-4
10-4 f--+---+-------4I::L-+------=l
10
50
Stress intensity factor
range, 11K, MPa
100
vm
Fatigue crack growth rate data for type 304 austenitic stainless steel
(annealed) at room temperature and at subzero temperatures. For this
alloy, crack growth rates are nearly the same at room and cryogenic
temperatures.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Metals, American Society for Metals, Metals Park OH, 1980, p 756
205
206
20
40
60
80
Gl
~
......
E
E
10- 3
"C
......
Cll
Gl
"C
~i
Type 304L
22C (72FI
...i!=
10- 5
--
Cll
"C
tJ)
.>t!.
...u'"
~
tJ)
-.
.E
"C
...0
Gl
>
u
10- 4
Type 304L
-196, -269C
(-320, -452FI
.;;
'"
IL
10- 6
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, P 123
8-11. Type 304 Stainless Steel: Fatigue Crack Growth Rate in Air
With Variation in Waveforms
~K,
10
ksi . in.
'/
30
20
40
50
10- 4
CIl
U
~
.......
E
E
10- 3
/ /
/
/ /
I /
/ /
II
/ /
Waveforms
Z
'1::J
.......
co
'1::J
2l'
s:
i0
..
..
( (
/ /
/ /
."
u
co
u
10- 4
/ /
/ /
/
/
CIl
:J
."
.;;
co
u,
rYYY\
10- 5
20
30
40
.E
Z
} 0.067 Hz
Type 304
538C (1 0000 F)
R = 0.05
>-
~
co
'1::J
/VVV\
10
.!!
u
.......
10- 6
50
The data presented in this graph were obtained in tests with a sawtooth waveform. Changing from a sawtooth waveform to a waveform
with a short holding period at maximum load did not influence the
overall fatigue crack growth rates according to additional data reported by James and shown above.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society For
Metals, Metals Park OH, 1982,P 117
207
208
30
LABORATORY TESTS
AISI304 STAINLESS STEEL
650C
z
......
180
Z
ILl
C>
z 0.01
a:::
60
a:::
~
en
-J
60
DO 01
10
0
30 30
o
I
o NO HOLD TIME
o TENSILE HOLD TIME IN MINUTESAS INDICATED
Source: L. F. Coffin, "Fatigue in Machines and Structures-Power Generation," in Fatigue and Microstructure, American Society
for Metals, Metals Park OH, 1979, P 19
8-13. Type 304 Stainless Steel: Effect of Hold Time and Continuous
Cycling on Fatigue Crack Growth Rates
6K, ksi . in. 1/ 2
10
40
20
60
80 100
10- 1
.:E.
"
"
,r~
10- 2
1.0minhold
Type 304
593C (1100F)
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, P 118
209
210
20
30
0.0014 Hz
40
50
0.0067 Hz
10- 4
Q)
>o
--EE
z'
--'"
10-3
"tl
"tl
l!l'
Q)
...::
10- 5
s:
een
"tl
..10:
"tl
10- 4
Q)
>u
--.5
z'
--'"
;:,
en
'':;
'"
u..
Type 304
538C (lOOOF)
R = 0.05
10
20
30
40
10- 6
50
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William w. Gerberich and John H. Underwood. Eds., American Society for
Metals, Metals Park OH, 1982, P 116
~K,
kg/lmm?/2
Il
5x 101
..,
~
.l::!
s:
u
c:
z
::!2
"'
't:l
....-c.
~
co:
10-5
....
:I:
;:
0
co:
<.:l
:><:
e:(
co:
:::I
<.:l
....
-e
......
0.003 cpm
00 4cpm
10-6
40cpm
4000cpm
4
10
~K, Ib/lin)3/2
5x101
Source: L. F. Coffin, "Fatigue in Machines and Structures-Power Generation," in Fatigue and Microstructure, American Society
for Metals, Metals Park OH, 1979, P 14
211
212
8-16. Type 304 Stainless Steel Welded With Type 308: Fatigue Crack
Growth Rates
t.K. ksi . in. 1I2
t.K. ksi .
in. 1/ 2
40
20
20
40
60 80
60 80
10- 2
Ql
10- 4
Ql
c:;
10- 4
--
E
E
Shielded
metal arc
E
Z
c:;
>
u
--. 10...
"t:l
10- 3
"t:l
"t:l
co
-e
Ql
c:;
>
u
...eoi
j!l'
--.S ...::..
z '"
--.. e
i0
"t:l
10- 5
"t:l
s:
u
u
Submerged
arc
..'"
.;:;
~
.;:;
u..
10- 4
--.S
z
--..
"t:l
Ql
..~
Ql
c:;
>
u
s:
u..
Shielded
metal arc
10- 5
"t:l
10- 4
o SMAW 1
SMAW 2
24C (75F)
20
40
10-6
60 80 100
593C (l100F)
20
40
10-6
60 80 100
Fatigue crack growth rates for annealed type 304 base metal and type 308 weld metal at 24 and
593C (75 and 1100 OF), 0.17 Hz, and an R ratio of O.
Type 308 stainless steel is the alloy that is usually used for welding rod for weldments in
type 304 stainless steel when those weldments are to be exposed to room temperature or to
elevated temperatures in service. Because service experience has shown that failures are
more likely to originate in weld metal or in heat affected zones than in the base metal, it is
important to have fracture information on weldments. In general, fatigue studies at elevated temperatures on specimens from type 304 weldments have shown that the fatigue
crack growth rates in the type 308 weld metal and heat affected zones are no greater than
in comparable specimens of the base metal. Fatigue crack growth rate data obtained by
Shahinian for specimens of type 304 welded with type 308 rod by the submerged arc and
shielded metal arc processes are shown above for tests at room temperature and at 593C
(llOOF).
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels." in Application of Fracture Mechanics for Selection of
Metallic Structural Materials. James E. Campbell. William W. Gerberich and John H. Underwood. Eds.. American Society for
Metals. Metals Park OH. 1982. P 125
8-17. Types 304 and 310 Stainless Steel: Effect of Direction on S-N
400
Type 310, transverse
Type 310, longitudinal
<II
Q.
~" 300
~
~E 200
:::l
'iii
..><:
40 '"
.......'"
30 '"
","
'"...
,~
50
~~-~
E
:::l
20 ,~
x
~ 100
10
105
106
No. of stress cycles
"'
0
108
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Metals, American Society for Metals, Metals Park OH, 1980, P 32
213
214
8-18. Types 304, 316, 321, and 348 Stainless Steel: Effects of
Temperature on Fatigue Crack Growth Rates
AK, ksi . in. 1/ 2
20
10
60
100
40
10-2 ...-r----r----.------r---.---. 4 X 10-4
10- 4
Q)
>
Ll
--EE
z
--'a>"
"0
10- 3
"0
.!!!
...
s:
...3:
Ll
>
Ll
-z
-.~
"0
Cl
l'Il
..
"0
Ll
l'Il
10-5
Ll
Q)
::::I
Cl
.~
l'Il
IL.
10-4
L -_ _....L.
10
20
10-6
80 100
....I...._--l._...l.-..J.J
40
60
Fatigue crack growth rates for annealed types 304, 316, 321, and 348
stainless steel in air at room temperature and 593C (1100 OF), L-T
orientation, 0.17 Hz, and an R ratio of O.
As reported by Shahinian, Smith, and Watson, fatigue crack growth rate tests were made on singleedge-notch cantilever specimens oftypes 321 and 348 stainless steel from L-T orientation at 0.17 Hz
with an R ratio of zero at room temperature and at elevated temperatures to 593C (1100 OF). As for
types 304 and 3I6, fatigue crack growth rates in air increased with increasing testing temperature. The
curves above show that, at room temperature, the fatigue crack growth rates for types 304, 316,321,
and 348 all fall within a narrow band. For tests at 593C (1100 OF),however, specimens of type 3I6 had
the least fatigue crack propagation resistance, whereas specimens of type 348 had the highest fatigue
crack propagation resistance, over the 11Krange studied. Results of tests on specimens of types 304 and
321 were nearly the same at 593C (1100 OF) in air.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, P 138
10
20
40
in.1/2
100
r - - , - - - r - - - , - - - r - - - - - - - - - - , 10- 3
10- 2
10-4
Type 309S
Testing
frequency,
Hz
10
15
20
25
30
Grain size
45 fJm 480 fJm
-0-
-0-
10- 6
10- 7
10
20
40
100
Types 309S and 3 lOS stainless steel are the low-carbon versions of types 309 and 310. They have higher chromium and
nickel contents than type 304 and consequently have better corrosion resistance and more stable austenite than type 304. Fatigue crack growth rate data have been reported by Thompson
for tests made at room temperature on compact specimens from
plate of type 309S and the L-T orientation after heat treating to
a grain size of 45 Jlm in one set and 480 Jlm in the second set.
Specimens with the smaller grain size had substantially higher
yield and ultimate tensile strengths than the specimens with the
larger grain size. Fatigue crack growth rates were obtained on
tension-tension loading at frequencies from 10 to 30 Hz and at
an R ratio of 0.05. The results are plotted above. These data
provide further evidence that a wide variation in grain size, and
the associated variation in strength level, does not affect the results of fatigue crack growth rate tests.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds.. American Society for
Metals, Metals Park OH, 1982. P 126
215
216
20
40
.-------r---r-..---..---.......,."""T""'l~---.______.
.,
U
>
("l
10-4
10- 3
--
E
E
Base metal
-196, -269C
(-320, -452F)
--.,'
"C
co
"C
......co
.,
>
("l
10- 5
...
.c
--.5
z
-"C
een
co
"C
-"o
co
...
.,
("l
10- 4
:::l
en
'':;
co
u..
Type 3105
20
40
60
100
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds.. American Society for
Metals, Metals Park OH, 1982, P 127
10-2
20
40
60
100
4 X 10-4
Ql
10-4
13
>
u
'<,
E
E
z'
~
<0
10- 3
Ql
13
>
u
"C
'<,
fl'
E
.~
z'
"C
s:
..
'<,
<0
"C
Cl
10- 5
E
u
Ql
:::J
Cl
'.J
<0
u..
10- 4
Type 316
593C (1100F)
10- 6
20
40
60
100
Results of fatigue crack growth rate tests on weldments of type 316 stainless steel have
shown that the crack growth rates in the weld metal are generally no higher than in the
base metal and may be somewhat lower at elevated temperatures (Shahinian, Smith, and
Hawthorne). The curve shown above for unirradiated weld metal tested at 593C (1100 OF)
represents fatigue crack growth rates substantially lower than those for the unirradiated
base metal at any given I:>.K level (Shahinian). The weld was produced by the submerged
arc method using type 316 welding rod. Weldments were stress-relief annealed at 482C
(900 OF). Specimens were single-edge-notch specimens for cantilever loading and were
tested at 0.17 Hz and at an R ratio of zero. Irradiation slightly reduced the fatigue crack
growth resistance of the weld metal, but its fatigue crack growth resistance was better than
that of the unirradiated base metal.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds .. American Society for
Metals, Metals Park OH, 1982, p 134
217
218
8-22. Type 316 Stainless Steel: Fatigue Crack Growth RatesAged vs Unaged
AK. ksi . in.1/2
10
60 80
20
40
10 2 ."......---r------,;-------r--,--..,
III
EE
10- 2
z
~
"lJ
III
10- 1
10- 2
--.S
--..
>
u
J!l'
E
..c:
10- 3
e
;
"lJ
"lJ
E
u
III
:J
10-4
..
en
.;;
u..
10- 5
10- 4
0
Ii.
10- 5
v
L-_....L-
10
--l.
20
10.0 min
10- 6
80 100
...L-_--l._......L~
40
60
Results also have been reported by James for fatigue crack growth rate
tests in 20% cold worked specimens of type 316 stainless steel which were
cycled at frequencies of 0.0055 to 6.66 Hz, at 538C (lIDO OF) and at an R
ratio of 0.05. Over the 11Krange studied, the fatigue crack growth rates were
highest for the specimens subjected to the lowest cyclic frequency.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Material, James E. Campbell, William W. Gerberich and John H. Underwood, Eds.. American Society for
Metals, Metals Park OH, 1982, p 133
8-23. Type 316 Stainless Steel: Fatigue Crack Growth RatesEffect of Aging
.1K, ksi in. 112
20
40
60 80
10
10-1.......- --r-----.-------,----.--,....,
10-3
10-2
Q)
-.~
593C (1100F)
E
E
10- 4
'"
Q)
"tl
...e
U
u>'<,
Q)'
..c
10-3
.~
een
"tl
'<,
'"
"tl
u
u
10-5
Q)
:::l
en
';:;
'"
Type 316
593C (1100F)
LL.
10-4
Unaged Aged
...
Hold time
Zero
0.1 min
1.0 min
10-6
10-5
10
20
40
60
80 100
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, p 130
219
220
10
20
40
60
100
10- 4
10- 3
Ql
1)
>u
10- 5
--.5
--
2"
"'C
III
"'C
10-4
10-6
Type 316
Cold worked
10- 5
10
20
40
60
100
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, P 132
10
20
30
40
50
60
10- 4
0.0067 Hz----...
II>
10- 3
>
o
--
)7 ~0.67"'
E
E
z
~
'"
0067
"C
...eai
...;:
e
"'-y' //J
'~
II>
10- 5
s:
Cl
"'..u'""
1,1
10- 4
>
u
--.5
z
"C
<,
'"
"C
U
II>
::>
Cl
.;;
'"
u,
Type 316
538C (lOOOF)
10- 6
10
Stress-intensity factor range, dK, MPa m 1/2
Effect of variation in cyclic frequency on fatigue crack growth rate of
annealed type 316 stainless steel in air at 538C (1000 OF) and an R ratio
of 0.05.
Source: J. E. Campbell, "Fracture Properties or Wrought Stainless Steels," in Application or Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood. Eds . American Society for
Metals, Metals Park OH, 1982, P 131
221
222
8-26. Type 316 Stainless Steel: Fatigue Crack Growth Rate in the
Annealed Condition
6K, ksi . in. 1I2
10
20
30
40
60
80 100
10- 3
/1
370C (700.F)
---t!
10- 2
""1/;/
I ,
482C {900
/1
,1
p'"
//'
fi,' I
:,/
593C
.'
(1100F) /
l'
10- 4
..
>-
"I
y: /:
.5
z
~
./ ;,'I
:i
"
I'"I
/ /'1
10- 5
10- 4
316
' - -_ _'-_J.----'_L-l......J-J....I-J..J
10
20
30
40
60
10- 6
80 100
Most of the fa tigue crack growth rate testing on type 316 stainless steel has been oriented toward its use in components for nuclear reactors, but the data also are applicable
to design of equipment for fossil fuel power stations, petrochemical refineries, and chemical plants. Its improved yield strength compared with that of type 304 stainless steel is an
advantage for these applications. The austenite stability in type 316 is greater than that in
type 304, so it is advantageous to use type 316 rather than type 304 for critical cryogenic
applications. Effects of elevated temperature on crack growth rate are summarized in the
graph above.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell. William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, P 129
6.0
4,0
FATIGUE
TEST MEDIA
Na (10 ppm OF 02)
AIR
HELIUM
t'
2.0
ex:
l-
V>
1.0
...J
u 0.8
>u 0.6
0.4
EXPOSED SPECIMENS
FATIGUED IN SODIUM
FATIQJEDIN AIR
FATIGUED IN HELIUM
EXPOSURE MEDI~
0.2
Na
ue ppm OF 02)
EXPOSURE CONDITION
286 hrs AT 92S K
0.1
102
8 103
8 104
CYCLES TO FAILURE
Source: R, H. Cook and R. P, Skelton, "Environment-Dependence of the Mechanical Properties of Metals at High Temperature,"
in Source Book on Materials for Elevated-Temperature Applications, Elihu F, Bradley, Ed" American Society for Metals, Metals
Park OH, 1979, P 84
223
224
8 10
20
10-3 r----,,-..----....
40
60
100
---...,.:.-......:.=----..:..;
Type 316
649C
(1200F)
">
10-5
Roomair
E
E
z
~
."
Type 316
2SoC 177F)
10-'
Dry air
Wet nitrogen
Dry nitrogen
fi
f!
.c
"
>
-!:!
.~
12tn
.><
<0
."
f!
":tn>
.~
10-6
u,
2SoC (77F)
Room air
Wet air
10- 5
10
20
40
60
100
Effect of gas environments on fatigue crack growth rates for types 316 and 321
stainless steel at 25 and 649C (77 and 1200 OF).
Fatigue crack growth rate data at 25C (77 OF) show that crack growth
rates increased slightly with increased humidity when oxygen was present but
that high humidity in an inert gas had no significant effect. Fatigue crack
growth rates in room air at room temperature were the same for types 316
and 321 stainless steel. Furthermore, in tests at 649 C (1200 OF)in dry nitrogen, fatigue crack growth rates for types 316 and 321 also were the same. In
air, however, fatigue crack growth rates in type 316 specimens increased by a
factor of about 22 over rates in an inert environment at the same temperature.
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials. James E. Campbell. William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, p 135
10
10-'
20
40
60
80
10- 3
Type 321
593C (1100F)
10- 2
..
U
>
u
10-'
E
E
.
"
10- 3
>
.~
e
'"
""eu
..
10- 5
..
"
10-'
::I
'"
.~
u,
10- 6
Unaged Aged
10- 5
...
Hold time
Zero
0.1 min
1.0min
10-
10- 6
'--_-'-
10
..J....
20
'
.......L_ _.l----JL-..J
40
60
80 100
Fatigue crack growth rates for annealed type 321 stainless steelunaged
and aged at 593 C (1100 OF)for 5000 h and tested in air with continuous
sawtooth waveform (0.17 Hz), with 0.1 and 1.0-min hold time at anR
ratio of 0 at 593C (1100 OF).
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics ForSelection of
Metallic Structural Materials, James E. Campbell. William W. Gerberich and John H, Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, P 139
225
226
ksi in. 1/ 2
10
20
30 40
60 80
10-3 rr----,-----,--,--...........-,------,
I
/
II
/
10-4
10-5
Air
Type 403
In H20
III
"1:J
pH 7, 25C
pH 10, 25C
-- - -
pH 7, 100C
pH 10, 100C
10- 6
In 1M NaCI solution
- - - - pH2tol0, 100C
10- 5 L..-_ _---L_ _L.--L---L----I---L...J......L.J.
10
20
30 40
60 80 100
--'
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials. James E. Campbell, William W. Gerberich and John H. Underwood. Eds., American Society for
Metals. Metals Park OH, 1982. P 147
20
40
60
100
----I
10-6
200 300
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, P 145
227
228
10
20
40
60 80 100
200
71e (160F)
57e (135F)
a>
10-
.:.::::
~
"tl
Type 422
~_---l
20
---L_---l_....L.........L.
l..-
..J
10-5
200
Fatigue crack growth rates in precracked round rotating beam specimens of type 422 stainless steel in 4.5% NaCl solution at room and elevated temperatues, 10Hz, and an R ratio of-I.
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell. William W. Gerberich and John H. Underwood. Eds., American Society for
Metals, Metals Park OH. 1982, P 150
x
80-
60r-
1-----------... VIBRATING
NOTCHED
401-
Kt = 2.1
60 notch
0.030inch
root radius
CANTILEVER TESTS
FOR 108 CYCLES
x 0% delta ferrite
.x
o
5% delta ferrite
15/16% delta ferrite
A 20% delta ferrite
20 .....
I
20
40
60
80
TRANSVERSE FATIGUE STRENGTH
100 ksi
Source: J. Z. Briggs and T. D. Parker, "The Super 12%Cr Steels," in Source Book on Materials for Elevated-Temperature Applications, Elihu F. Bradley, Ed.. American Society for Metals, Metals Park OH, 1979, p 121
229
230
ksi
140
110 ,....-
IJJ
a:: 100 "-
J-
(/)
80 70 90
601()~
'.,""'-ee_
.......
room temperature
-.-
700 F(370 C)
...
106
CYCLES
101
108
SON curves for vacuum-melted type 422 stainless steel with 15% delta ferrite,
showing effect of temperature on fatigue strength.
Source: J. Z. Briggs and T. D. Parker, "The Super 12% Cr Steels," in Source Book on Materials for Elevated-Temperature
Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals Park OH, 1979, P 121
100
en
en
LLI
0:
I-
90
en
~
15~20
'~~v_
~...
00% Ferrite,Tempered 1150 F
--.. . . . . .-v..
-~
"-.. 0
(620 C)
o
0.:
15/20% Ferrite,Tempered 1200 F(650 C)
0-
80
V;;
lr-+
0-
'LONGITUDINAL ELECTROMAGNETIC
CANTILEVER-BEAM TESTS
70
lOS
107
CYCLES
S-N curves for type 422 stainless steel, which demonstrate the adverse effects of
Source: J. Z. Briggs and T. D. Parker. "The Super 12%Cr Steels," in Source Book on Materials (or Elevated-Temperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals Park OH. 1979, p 121
231
232
100
10-2
I
'---Hll00
,
R =0.05
'I
1 min hold
Salt soln
,~
I~
10-4
I
I
10- 3
il!
'"
~u
"5,
10- 4
."u..
10-6
10- 5
174 PH
'--
-'-
-'-
10
100
...1..1
10-1
Fatigue crack growth rates in WOL specimens of 17-4 PH stainless steel in the HI050 and
HllOO conditions in room temperature air and in a 3.5% NaCI solution,
Results of fatigue crack growth rate tests on specimens of 17-4PH stainless steel under
comparable conditions are presented here. Those specimens that were tested in the HI 050
condition at a stress ratio of 0.67 with a one-minute holding period at maximum load in
each cycle had the highest fatigue crack growth rates (as for 15-5PH) in the upper levels of
I:!.Kvalues. Specimens in the H II 00 condition tested in a salt solution with a one-minute
holding period, however, had fatigue crack growth rates only slightly higher than those of
comparable specimens tested in air with continuous cycling.
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels." in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals. Metals Park OH. 1982, P 156
100
)
"
Hl~50 ~I //'I
R - 0.67
I:/,
1 min hold
I 1//
I/Il
RT.ir
,,1
l,' l'
'li
10- 3
..
Hll00--.......,'
R = 0.05
~
1 min hold
"
I
I
'I
Salt soln
."
!l'
VI
'I1 ,r--HllOO
R = 0.05
.t::
e
'"
""ut!
..
..
---Hl050
R = 0.05
10Hz
Sine wave
RT air
10-'
I'
TO
1 min hold
RT air
u,
10- 6
H 1100 :----""'/'
R = 0.05
10 Hz
I
Sine wave
RTair
10-5
155 PH
1-
--L.
...1...
10
100
.J1O-
Fatigue crack growth rates in WOL specimens of 15-5 PH stainless steel in the HI050 and
HllOO conditions in room temperature air and in a 3.5% NaCI solution.
For specimens in the HI 050 condition, increasing the R ratio from 0.05 to 0.67 and incorporating a one-minute holding period at maximum load in each cXcle substantially increased the
crack growth rates at LiKvalues over 40 MPa m 1/2 (36 ksi- in. I 2). For specimens in the H 1100
condition, exposure to a salt solution environment during tests with a one-minute holding period at maximum load increased the fatigue crack growth rates over those of specimens tested in
air with one-minute holding time or with continuous cycling (see graph).
Source: J. E. Campbell. "Fracture Properties of Wrought Stainless Steels." in Application of Fracture Mechanics for Selection of
Metallic Structural Materials. James E. Campbell. William W. Gerberich and John H. Underwood. Eds . American Society for
Metals, Metals Park OH. 1982, pISS
233
234
10
20
40
60 80100
10-2
200
4 X 10-4
Gl
>
u
E
E
---
10- 4
-e
Gl
l'O
.....
"tl
>
U
---
Gl'
..
.~
l'O
.s:
10- 3
Z'
;:
"tl
l'O
"tJ
Cl
E
u
Gl
:::l
Cl
';;
l'O
u..
10-5
PH 13-8 Mo
H1100
10- 4
20
40
60 80 100
200
Source: J. E. Campbell. "Fracture Properties of Wrought Siainiess Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, P 159
8 10
20
40
60
10-2
R = 0.3
STW
L-T
10- 4
10- 3
10- 4
10- 6
PH 13-8 Mo
H1000
8 10
20
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, p 158
235
236
8 10
ksi . in. 1/ 2
20
40
60
100
10- 4
10- 5
Q)
>
u
'-:
.!:
1:1
"nl
1:1
10-6
PH 13-8 Mo
H1000
10- 7
8 10
20
40
60
100
Fatigue crack growth rate data for room temperature tests on specimens from
rolled bar and extrusions of PH 13-8 Mo (H 1000)stainless steel make up the scatter
band in the above graph. Specimens of L-T and T-L orientations were tested in
low-humidity air and in sump tank residue water at frequencies of I and 6 Hz and at
an R ratio of 0.08. Under these conditions, variations in frequency and environment
had little effect on fatigue crack growth rates. For tests at -54C (-65 OF),the rates
of fatigue crack growth were lower than those at room temperature over most of the
~Krange.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, P 157
237
-100
100
+50
150
200
1600
1400
200
1200
'"
a..
:2 1000
150
Jl
100
.S
x
'"
:2
",'
...~'"'"
E
E
800
S)<::l l'l'b
'x
'"
:2
'li~
:l
600
S)<::l
(c)
\e"""
.,-,;
'b~
~fl,
Axial fatigue
Unnotched specimens
Longitudinal and
400
50
transverse or-
200
ientations
o '--_-'--_...l-__....J...._----L_--''------''L-_-'--_-L.._--'--_----L_--'_ _-'----_...l-_-'
-1200 -1000 -800
-600
-400
-200
+200
400
600
BOO
1000
1200
1400
1600
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Metals, American Society for Metals, Metals Park OH, 1980, p 32
:l
238
8-42. Types 600 and 329 Stainless Steel: S-N Curves for Two
Processing Methods
1000
800
600
'"
o,
::2:
I
I
.~
.L
400
IV
I I I.
AISI 329 (electroslag remelted)
Vi
'"
e
Ul
0.57
100
80
'\1'
60
0.51
40
'iii
-'"
",'
'"
e
Ul
200
20
100
10'
10'
10'
10'
Cycles to failure
Steel
STAMP 600
Electroslag-remelted 329
Tensile
strength,
MPa(ksl)
Yieldstrength
(0.2% olTset),
MPa (ksi)
760
(110)
630
(91)
600
(87)
500
(73)
Mechonlcal properties
Elongationin
Reduction
50 mm (2 ln.),
%
in area, %
26
54
29
65
Impact
energy,
strength,
J (Rlb)
MPa (ksl)
Fatigue
25
(18)
35
(25)
430
(62)
320
(46)
S-N curves showing test results and mechanical properties of STAMP-processed 600
stainless steel and electroslag-remelted AISI 329 stainless steel. Fatigue ratio
(0 107/Rm) for 600 steel: 0.57. Fatigue ratio for electroslag-remelted 329 steel: 0.51.
Source: Metals Handbook, 9th Edition, Volume 7. Powder Metallurgy, American Society for Metals, Metals Park OH, 1984, p 549
10
20
40
60
21-6-9
10- 4
Q)
>
u
......
E
E
22 to -196C
/
(72 to -320F) /
-269C
(-452 F)
10- 3
2:
"t:l
......
III
"t:l
...oj..
III
s..
0
Cl
III
>
10- 5
/
/
or.
Q)
10- 4
......u
,E
2:
"t:l
......
III
"t:l
Q)
:l
Cl
'+:i
III
u.
/
10
20
10- 6
40
60
100
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels,"in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, P 140
239
240
20
40
60
100
200
10- 3
10- 4
I
I
Kromarc 58
Base metal
Weld metal
20
40
60
100
10- 6
200
For the fusion zone of a gas tungsten arc weld made with
Kromarc 58 filler metals, the KIJJ) value was 156 MPa. m l / 2
(141 ksi- in. 1/2) at -269C (-452 OF).Fatigue crack growth rate
data for the base metal at room temperature and at - 269C and
for the weld metal at -269C are shown above. The data were
obtained on compact specimens at 10 Hz and at an R ratio of
0.1. Fatigue crack growth rates for tests in liquid helium were
lower than at room temperature at the same t::.Kvalues. Therefore, if room temperature crack growth rate data are used to
estimate crack growth at cryogenic temperatures, the estimated
values will be conservative.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, P 142
10
20
30 40
>
E
E
200
10- 4
",-SMAW
24 and -269C
(75 and -452 F)
Gl
......u
60 80 100
10- 3
Z
't:l
......
Gl
'"
't:l
>
u
......
/"GTAW
24C (75F)
!l
'"
..r:
.~
'"
.~
10- 5 z
~
'"
't:l
E
u
Gl
10- 4
:l
'"
.;:;
u..
'"
20
30 40
60 80100
200
10Hz.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William w. Gerberich and John H. Underwood. Eds., American Society for
Metals, Metals Park OH, 1982, p 141
241
242
KCR 171
Whit. wat.r
400
pH 4. IS
T ISOC
300
CI
Q.
_ 0 .......
(\/
0 ...
200
l;
~
100
6 Hz
t:>
20 Hz
100 Hz
Nf
167Hz
eyel ..
Rotating bending S-Ntests were carried out in 50C (122 OF) white water
at different frequencies (6, 20, 100,and 167Hz) for samples polished with 240
grit emery paper and the results obtained are presented in the above S-N
diagram. The results thus far 0 btained for the two highest frequencies appear
to fall on the same S-N curve, and the indication is that this curve would
present a quite horizontal fatigue limit. In the short life regime (N,: 106
cycles), the results suggest that decreasing the frequency below 100 Hz displaces this portion of the S-N curve to shorter lives without significantly
changing its slope.
Source: M. Ait Bassidi, J. Masounave and J. I. Dickson, "The Corrosion Fatigue Behaviour in White Water of KCR 171," in
Duplex Stainless Steels, R. A. Lula, Ed., American Society for Metals, Metals Park OH, 1983, p 455
9-1. Grades 200, 250, and 300 Maraging Steel: S-N Curves for
Smooth and Notched Specimens
1500r - - - - - - r - - - - -.....----~----_,
CZZl
Em
css
'"
~
18Ni(300)
18Nj(250)
18Ni(200)
_ _-+-
1000
200
....=::j
150
'iii
c..
.:.l
v.
...e'"
If)
500
L.-
104
--'-
105
---L
106
Number of stress cycles
L.-
107
0
8
10
~
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH. 1978, p 451
243
244
92. Grade 300 Maraging Steel: Fatigue Life in Terms of Total Strain
-1
10
,,
\
--
/(J'f/E
-2
10
--..
- LOAD CONTROL
\
'\
- : TOTAL
~\
...
A..-
-)-Trrr~ ..c
ELASTIC
PLASTIC~\
h
-3
10
L-J....l.l..JLlJJ.U-l.....LJ..J.JJJ.U.--I....LJ.J.LWL--I...1.LI..JWJJ..--L..LUJ..LJ.ll..-l-u..uJJ.lJ
10
Source: Marc Andre Meyers and Krishan Kumar Chawla. "Mechanical Metallurgy: Principles and Applications," Prentice-Hall,
Inc.. Englewood Cliffs NJ, 1984, p 700
50
20
iii
a:
10
...J
I&.:
"B" BAR
BAR
~ S~
.
.1
10
Walter has shown that the fatigue properties of irons are highly dependent on
volume of graphite and its morphology and distribution, as well as the matrix structure. He was able to reduce these factors to some easily measurable parameters, Eo,
D, and Bhn, which gave good correlation with fatigue properties over a rather wide
range of irons (see graph). It is reasonable that these parameters relate to fatigue
performance, since they are measures offatigue-related properties. Eo, the modulus
at very small strains, is controlled mostly by the volume of free graphite and to some
degree by the graphite shape. Since the graphite present detracts from the matrix
load-carrying area, the more graphite, the higher the stress on the remaining
matrix-thus lower fatigue performance. D, the damping capacity, is controlled
mostly by the graphite morphology and to some degree by the graphite volume.
Sharp-edged flakes are greater stress raisers than rounded-edge flakes and spheroids; thus the higher the D, the poorer the fatigue performance. Bhn is largely a
measure of the matrix hardness and, to some degree, ofthe graphite volume; thus
the higher the Bhn, the better the fatigue performance. These easily measured properties are put to good use in industry as specification means and process-control
criteria.
Source: D. H. Breen and E. M. Wene, "Fatigue in Machines and Structures-Ground Vehicles," in Fatigue and Microstructure.
American Society for Metals, Metals Park OH, 1979, p 86
245
246
200
400
400
600
800
1000
I
Fatigue life
Fatigue limit
350
- 50
- 40
300
250
IV
0-
:2
::;' 200
cil
150
, ".
-,
<,
.,
'-
_____0
17
100
---- -% - "
~
<,~
50
"
20
1o
100
300
400
500
600
Temperature,oC
o Notched specimens
200
Un notched specimens
Notched specimens.
stress based on net area
Composition: 2.84 C, 1.52 Si, 1.05 Mn, 0.07 P, 0.12 S, 0.31 Cr, 0.20 Ni, 0.37 Cu. (Ref 5)
Typical fatigue life for as-cast gray iron of the above composition (left). Effect of temperature on
fatigue limit for the same gray iron (right).
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 21
247
34
Plain Iron
Alloy Iron. 1% Ni. 0.4% Cr.
0.6% Mo. 1.0% Mn
o
32
(f)
30
a.
0
0
0
..--
en
28
26
""",
'\.
230
220
210
200
190
180
<J)
.....
170
24
Knee
(f)
- /-----------1----0--
160
22
Endurance or
~
Fatigue Limit .,.
- ._--_.1. ________________
20
18
10'
en
(f)
<J)
(f)
L..
<U
o,
10 5
-L
10 6
150
140
130
10 7
Cycles To Failure
A fatigue crack starts in an area of high stress concentration after a large number
of loading cycles. It is always a brittle type of fracture even when occurring in
ductile metals. As the crack progresses it increases the stress concentration, and the
rate of propagation under the cyclic loading increases. When the cr~ss section of the
remaining metal becomes insufficient to support the maximum load, complete
failure occurs as it would under an excessive steady stress.
The number of stress applications that will induce a fatigue failure is less at higher
maximum stress values, and conversely a larger number of stress cycles can occur at
a lower maximum stress level before a fatigue crack is initiated. A plot of this
relation for a metal is called an S- N curve and relates the maximum applied stress to
the logarithm of the number of cycles for failure. When the number of cycles without failure exceeds ten million, the endurance life is considered infinite for bodycentered-cubic ferrous metals. The maximum stress that will allow this number of
cycles is established as the endurance limit, or the fatigue strength or fatigue limit.
Two typical S-N curves for a plain and alloy high-strength gray iron are presented
above.
Source: Iron Castings Handbook, Charles F. Walton, Ed.. Iron Castings Society, Inc.. 1981, p 246
.....
L..
(f)
248
21
140
20 -
'ea.n
19
0
0
0
18
en
Fatigue
Strength
18.000
130
co
Q.
PSI
1124 MPa)
120
17
en
(JJ
Q)
....
+-'
(JJ
Q)
....
16
+-'
110
(.f)
(.f)
15
14.500 DSt
(100 MPal
100
14
10
10
10 7
10
Number of Cycles
The effect of various environments and corrosion inhibitors listed in the
table below on the corrosion fatigue properties of gray iron.
Environment
Fatigue strength
psi
MPa
Air
17,920
Water
14,560
3% sodium
chloride. . . . . . . . . . . . . . . . .. 5,600
1% borax
15,680
3% "Sobenite"*
17,920
3% sodium
carbonate
17,920
3% soluble oil.
17,920
0.25% potassium
chromate
17,920
Fatigue strength
reduction factor
124
100
1.23
39
108
124
3.20
1.14
1.00
124
124
1.00
1.00
124
1.00
Source: Iron Castings Handbook. Charles F. Walton. Ed., Iron Castings Society, Inc., 1981, p 255
100
150
200
30 ~---"""T---r--""---T""""'---"""'"2
200
I--------+---~;<;;;;;;;;;;;;;;;;+.;;;;;;ijit----~
'(ij
a.
o
o
o
100
eo
CL
iii
Ul
....~
(f)
50
Cl
:il
S
0
-50
-101'-
.....L.
10
.l...-
20
...J
30
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981,P 251
249
250
20
40
60
100
Q)
o>-
Q)
o>-
~
o
~
E
s:
.~
10-'
'"
"0
2iIII
a:
s:
Load Range
(9
i3
2000
o 2200
2500
o 2500
6 3000
... 3300
10-8
'"
U
Ib (910 kg)
Ib (1000 kg)
Ib (1130 kg)
Ib (1130 kg)
Ib (1360 kg)
Ib (1500 kg)
'"
"0
10- 4 ~
III
a:
s:
(9
.:>L
10
20
30 40
60 80100
Source: Iron Castings Handbook, Charles F. Walton, Ed.. Iron Castings Society, Inc., 1981, p 250
1000 psi
60
400
50
en
en
300
40
(j)
"0
Q)
+-'
<U
Q)
li} 200
a::
'0
.~
E
Q; 100
~
..J
"0
<U
1000 psi
Q;
::J
Mean Stress
-100
Torsional fatigue strength for three levels oftensile strength with various mean
stresses.
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981, p 253
251
252
r---------------------.,
1.50
en
1.25 Q)
E 30
E
.I:
U
....OJ
.I:
1.0
.I:
0.75
-.J
....OJ
C
Q)
-.J
20
Q)
lO
L..
lO
L..
0.5
lO
....0
10
lO
I-
0.25 I-
O'--__ '---""'--_-'----_-'-----I._-'-----'-----'400
300
----l_ _----l
800 1000
500 600
--'
2000
Number of Cycles
Total length of six cracks (the first three cracks in each of two specimens of each iron)
as a function ofthe number of thermal cycles between 1100 and 400 C (590 and 200
C). Iron compositions are as follows:
Composition. %
Iron
B
C
0
E
Si
Mn
Cr
Mo
3.43
3.45
3.45
3.44
3.43
1.65
1.74
1.68
1.69
1.66
0.57
0.59
0.63
0.58
0.58
0.49
0.30
0.21
0.50
0.30
0.38
0.39
Ni
Cu
Sn
0.60
0.97
0.59
0.87
0.30
0.077
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981, pp 288, 289
Gray Iron
3% AI. 0.6% Si (65% Ferrite)
<>
2% Si
.!: .150
o 0.5% Cu. 0.5% Mo
D
o
c
A
2%
AI.
2%
Si
(4%
Ferrite)
.125
.175
+-'
.100 t---+----+---+---.,,>4F---t----+------l
Q)
....J
~
...ro
_ _--1
.075 t---+----+---.",..e..t----+-----.,.,.-=--
.050 I--------if------""--+---=-'f----+----+---'::=--'""=i
s:
g>
--+---t"7""'==------j4
10
12
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981, p 434
253
254
.15
1;..:-.,;.;;;,
..,. ,~~
~.
.10
3.7% C
3
Bridge Cracked
/~
~ ->
Unalloyed ____~
CrNi _ '?"' . ~ .
..-'
.--
.-1--.05
I ...-:.
CrNiMo0'"
CrMo(
en
.::.t.
~-.:
en
o ~
<Il
L.
'0
L.
---
3.2-3.3% C
s:
Q.
Q)
en
Q)
..c
o
c
I
Bridge Cracked
.15
..c
a.
Q)
CrNiMo /
'0
.10
'
.05
o '--__-L._ _- - '
o
100
200
.L-
300
0
600
400
500
Number of Cycles
The depth of cracks resulting from the thermal cycling of gray irons
between 860 of (460C) and room temperature.
Alloying with molybdenum and chromium provided superior thermal fatigue resistance compared with irons that contained other alloying additions or no alloying at all. In
this case, the improved thermal fatigue resistance is believed to be directly related to the
higher elevated-temperature tensile strength and better stability of the chromiummolybdenum irons. However, it must be remembered that this improvement is related to
and dependent on the temperature cycle and base iron composition, as shown above. It
has also been indicated that the development of an acicular matrix structure, by adding
relatively large quantities of molybdenum and copper, supplies a less than desirable influence on thermal fatigue cracking.
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Casings Society, Inc., 1981, p 288
100
200
300
1500
<J)
Q)
0
>-
'0
1000
Q;
.D
E
:J
500
o Lo
' - -_ _---'
10
20
--'-
-'-
30
40
'---'
50
Source: Iron Castings Handbook. Charles F. Walton, Ed . Iron Castings Society, Inc. 1981, p 285
255
256
1400
1600
1o',---,--------,r--------,---,
z
d>
c
:;;; 10'
Steel
(J
Ductile
Iron
B
<f)
<Il
>-
(J
'0
Gray
AI
; 1m",
Q> 10'
.0
E
:J
16~0.,...0--------=-.L------~-----~~
700
800
900
C
D
Si
Mn
Cr
Mo
Other
3.43
3.49
3.48
3.50
2.37
2.37
0.60
2.38
0.78
0.84
0.88
0.83
0.22
0.24
0.23
0.30
0.32
0.22
0.20
0.77
0.21 Sn
2.37 AI
1.51 Cu
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981, pp 286, 287
Ferritic Ductile
Pearlitic Ductile
Alloyed Ductile
6
10'
10'
Number of Cycles
The number of thermal cycles required to produce thermal fatigue cracking in cast irons. Compositions are tabulated below.
%Mg
Analysis
%C
%5
%Mn
%P
2.96
2.90
0.78
0.07
Ferritic Compacted
Graphite
3.52
2.61
0.25
0.05
0.015
Pearlitic Compacted
Graphite
Ferritic Ductile
3.52
2.25
0.40
0.05
0.015
3.67
3.60
2.55
2.34
4.84
0.13
0.50
0.06
Pearlitic Ductile
Alloyed Ferritic
Ductile
0.030
0.030
3.48
0.31
0.05
0.07
Alloys
0.12Cr
0.030
1.47Cu
0.54 Cu
1.02Mo
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society. Inc., 1981, pp 393, 396
257
258
~o
J.
500
700
900
I
1100
1300
I
0.5 -o\'ox>
Ferritic.~.
o
.."
.o~ "x~x:
x
o ,
'"
II:
Q)
0000
0.4
::J
'U
~
\ K
Tempered Martensite
<,
j( .....
0.3
60
100
140
160
In general the fatigue limit for ductile iron increases with tensile strength, but as with other ferrous metals, the increase is less
than proportional. The relation between the tensile strength and
the endurance ratio for the annealed, ferritic irons is different
from that of the irons with a matrix of pearlite or tempered martensite, as illustrated above.
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981, p 341
50
75
100
125
150
175
0.51-----e-..... "ki~---T___+-----__1I_-----+__-----__+--_t
..o,
~
1ie
O.4I------_+_---'''--------'''r__+-----.-+--'''o,~
......I _ - - - - - _ + _ - - - - - _ _ + - - _ _ 1
::>
"0
<:
0.31------t------+-----t--------1r---'''''''=----+----j
200
Tensile strength, MP.
The influence of tensile strength and structure on the endurance ratio of ductile
iron is indicated in this graph. Endurance ratio is defined as endurance limit divided
by tensile strength. Because the endurance ratio of ductile iron decreases as tensile
strength increases, regardless of structure, there may be little value in specifying a
higher-strength ductile iron for a structure that is prone to fatigue failure. For tempered martensite ductile iron, the improvement in fatigue strength due to an increase in tensile strength is greater than for pearlitic or ferritic structures. This is
indicated in the graph above by the shallower slope for martensite.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels. American Society for Metals, Metals
Park OH, 1978, P 45
259
260
10-16. Ductile Iron: S-N Curves for Ferritic and Pearlitic Grades,
Using V-Notched Specimens
400'.---------,-------,-------,---------,
3S01-------1-------+------+---------1
F_I6G-40-18 .......odl
~v...otchod
3ool----------+------+-------t--------I
g.
2001--------=zilJ!l
ISOI--------+----==
20
loo'f--------f-------+-------1-----------l
106
Fatiguelift. cyeles
400,-------.-----------,--------,r--------,
3S01-----------lI----------l--------l'---------l
Pllrlitlc (80-66-06 -e...t)
46 V-notched
3001-----------1r--------l--------l---------l
40
.r.
-6 2SOI -- - - - - -+----:
~e
i!'
~o
:;
il. 2001-----------1r----"
ISOI---------1I--------f--------f---------j
20
1001-----------1r-------f-------1--------l
106
Fatiguelife. cycles
Top:S-N curves, including scatter bands, for annealed ductile iron. Bottom: Similar to above except for as-cast pearlitic ductile iron. All test
specimens were V-notched (45).
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 43
10-17. Ductile Iron: S-N Curves for Ferritic and Pearlitic Grades,
Using Unnotched Specimens
.00r-------,.-------,--------,.-----------,
F.,rillc: (60.4o-18 annIId)
unnotc'*l
3501--------t------+------+--------I
3001--------t----::;
e:lE
'0 ]
2501--------t--
=
~
=
1
200
1501--------t------+------+--------I
20
1001--------+------+------1-------;
106
.OOr--------,------.----------,r--------,
3501---------f~
3001--------;------'
t:.
'0
:lE
ii 2501--------;-------+------......,1--------1
go
a
~
J200
1501-------+------+--------1------.,
20
1001-------+------+--------1------.,
5:0'"'-------'<--------';-------:--',--------'
106
10'
Faltgue lIle,cycles
Grade
Tenolle
alrenl!h
MPa
kal
Unnolched
Endurance
UmII
Endurance
MPa
kat
rallo
6040-18
80-55-06
480
680
205
275
70
99
30
40
0.43
0.40
Notched
Endurance
Umil
MPa
kat
125
165
18
24
Endul'
ance
ratio
Slreaa
coneenlrallon
lacier
0.26
0.24
1.67
1.67
Top: Similar to upper graph on the opposite page, but here the specimens were unnotched. Bottom: Similar to lower
graph on the opposite page, but here the specimens were unnotched. Data in table pertain to graphs on this and the
opposite page.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, p 44
261
262
100
600
600
80
400
60
JOO
40
::;
~ 200
c:
400
200
Q)
5 100
20
-0
c:
-0
o
-20
-100
-40
-200
~--:-f\P'A"':ld'-?-:!:---!c:---f::---,~~I:---
1001
psi
-200
200
400
600
-JO
-40 L-_-:-'-:-_-=-~_--:-'-,----_-,-:-_-=-'-MP,
100
200
JOO
400
600
o
MPa
Mean Stress
Mean Stress
Fatigue diagrams indicating endurance limits for five grades of ductile iron under bending stresses (left) and
tension-compression stresses (right). Minimum properties of the irons are given in the table below.
Iron
No.
1
2
3
4
5
1000 psi
MPa
1000 psi
MPa
Min. Elongation
Percent
55
61
72
87
102
38
42
50
60
70
36
41
51
61
72
25
28
35
42
50
17
12
7
2
2
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981, pp 344, 345 and 346
263
90 100
120 _':"';'::'_...:...r:.-....,....;.,~-, 50
350 n60r--_":,,,_-.--~......,.._--,
.........._r----r-..........
g
n8
'" 3001---+----+---i--7"'q...---f---+-----i
a..
::2:
45
40
.<:='
200 1----+---+--7"'~_+_---'
25
150
_ __'___ _-'-_ _
400
600
500
700
~
__L. _ _- - - - '
800
900
1000
1100
Tests made on 10.6-mm (O.417-in.) diameter specimens. Fully reversed stress (R = -1).
Data given in the above graph show that the endurance limit for any given
strength level of ductile iron is significantly affected by surface conditions of
unnotched specimens. The endurance limit is much higher for the polished
specimens than it is for the as-cast specimens, which have relatively rough
surfaces.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, p 45
264
60
0-
...
...
'E
::i
,.,
0)
';:; 40
<U
u,
"0
Q)
lD
30
<U
a:
360
20
100
----I-
...
'E
::i
Q)
300
::J
0)
';:;
<U
U.
.,....
0)
<U
0...
Q)
::J
------
.
..
.....
60
400
0
0
0
260
0)
.s
"0
Q)
lD
200
s0
a:
160
200
300
400
Hardness. Brinell
'iii
00
0
0
60
<U
0...
- 400
60 -
Q)
::J
0)
.~
u,
0)
40
r-
"0
Q)
lD
z-
<U
a:
30
100
...
'E
::i
A.
I
200
.~
..
...
'E
.,-
::i
360
Q)
::J
0)
';:;
<U
u,
300
0)
.s
"0
Q)
260 lD
z-
<U
300
400
600
600
700
a:
Top: Relation between Brinell hardness and fatigue limit in rotary bending for ductile iron. Bottom: Relation between rotary bending fatigue
limit and matrix hardness for ductile iron.
Source: Iron Castings Handbook, Charles F. Walton. Ed" Iron Castings Society, Inc" 1981, p 347
Rolling
Pressures
(Pounds)
75
70
500
406
580
768
65
60
400
'(ii
o,
o
0
0
en
(J)
,,
55
50
<0
n,
~
,,
"
"",
"
u:;
83
"
45
"-----------j
300
40
35
30
10'
Unrolled
10'
10'
200
10'
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc . 1981, p 348
265
266
I
36
<,
32
'iij
0.
t
0.750"
<,
2.75"R
- 250
0}17" 10.4'72"
..
225
200
0"",
~ ~()OOO
Unnotched
-"'0-
28
o
o
o
eo
n,
0
'\.
'\.0"
~
20
I'"
W
(J)
~
Ul
-i
0
o
175
"1
0.417"
0.700"
V-Notched
150
.....
125
1--0-000--
16
10'
10'
10'
100
10'
Number of Cycles
Source: Iron Castings Handbook. Charles F. Walton, Ed.. Iron Castings Society, Inc. 1981, p 341
v'rri
20
Q)
Q)
<3
<3
>-
>-
<,
:? 10-'
E
E
o
c
c
Z
"0
<,
'"
<Ii
m
a:
10-'
z
~
eo
"0
"0
'"
a:
s:
s:
o
(910-'
.:.!
o
~
U
<.9
.:.!
30 40
60 80100
Source: Iron Castings Handbook. Charles F. Walton. Ed.. Iron Castings Society. Inc.. 1981. p 349
267
268
r--------...,---------,
400
350
50
.iii
0.
<tI
c...
0
0
0
:2:
-0
300 ~
-0
--l
<tI
0
--l
40
260
30
L-
10'
-'-
10-
10'
Number of Cycles
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981,p 311
400
C-
o
0
0
<0
a..
~
60
300
enIII
40
Vi
30
en
III
Vi
200
20
10'
10'
10'
10'
Number of Cycles
The influence of as-cast surfaces, smooth machined surfaces, and machined notches on the fatigue behavior of pearlitic malleable irons. Iron 1
is grade 60003 and Iron 2 is grade 80002.
Surface finish has an important influence on fatigue properties, as shown above. Samples of malleable grades 60003 and
80002 were tested in fatigue with "as-cast" and machined surfaces. Samples of the 60003 grade were also included with a machined surface containing a sixty-degree notch that was 0.050 in.
(1.25 mm) deep. The resulting data are shown in this diagram.
Source: Iron Castings Handbook, Charles F. Walton, Ed., Iron Castings Society, Inc., 1981, p 310
269
270
1 PIECE
t----50.0
9 PIECP.S
ATHOSPIlr.RE
NITRIDED
7.600 POUNDS
1,01.--------------------------------II) 5
HUMBER OF CYCLES
Source: J. A. Riopelle, "Short Cycle Atmosphere Nitriding," in Source Book on Nitriding, American Society for Metals. Metals
Park OH, 1977, P 287
250
002
004
006
008
~ tNotch
rad.ius
o 0.13 mm or 0.005 in.
c,
<,
:;;
-5
0>
150
.~
100
~
. r--
u.
50
30
...............
t;
'"
5,
200
--
E:
20
'"
5,
.~
r--- r-s-
u.
10
o
0.5
1.0
1.5
2.0
Depth of notch, mm
Fatigue strength of unnotched ferritic malleable iron is approximately 50% of the tensile strength, or from 170 to 205 MPa (25 to 30
ksi). The graph above summarizes the effects of notches on fatigue
strength. As a rule, notch radius has little effect on fatigue strength, but
fatigue strength decreases as notch depth increases.
Source: Metals Handbook. 9th Edition. Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 65
271
272
A286
w
C)
z
:
z
c
VACUUM
o AIR 593C
VACUUM 593C
AVACUUM 20C
0.01
a:
en
~ 0.001
en
-J
Q.
Plastic-strain range versus fatigue life for A286 ferrous alloy in air and in vacuum af
593C (1095 OF). Numbers adjacent to test points indicate frequency in cycles per
minute. Note absence of frequency effects in vacuum.
Coffin has suggested that for a number of materials, virtually aU of the degradation in fatigue life at elevated temperatures can be attributed to environmental
interactions. He noted that frequency effects in the low-cycle-fatigue law could be
eliminated for a large number of metals and alloys by testing in vacuum (note
above). Additionally, it was noted that tests performed in vacuum showed transgranular crack nucleation and propagation versus intergranular nucleation and
propagation in air at elevated temperatures. These results are not unambiguous,
since Koburger has shown a frequency effect in high-cycle fatigue for directionally
solidified eutectic alloys when tested in air and in vacuum, particularly at elevated
temperatures. The primary difference in these results may be related to the lack of
intergranular cracking in eutectic alloys.
Source: D. J. Duquette, "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys," in
Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 343
/.--------- - ------
~---
l!>;:;::--::
/' --1>-;;:::::::0 /' /'
:0
-:
".0
,/
)
/
A 286-593C
Kr=3.0 .MI. =60ksi
2
--:;....--:;;.--
--
v STANDARD HT-AIR
oHT#1
-AIR
HT#I
-VACUUM
0 HT#2
-AIR
I> HT #3
-AIR
0 DS HT # 3
-AIR
~DS-STD HT
- AIR
Effect of frequency on life of notched fatigue bars of A286 at 593C (1095 OF)
in air and vacuum. As indicated, decreasing frequency has a degrading effect on
fatigue life ofsamples tested in air, with little or no effect on samples tested in a
vacuum.
Source: L. F. Coffin, "Fatigue in Machines and Structures-Power Generation," in Fatigue and Microstructure, American Society
for Metals, Metals Park OH, 1979, P 13
273
274
20
30
40
60
SO 100
A' /
10-3
53SC
(1000.F)~'
427C (SOOF) ,
,/
(i /
I"
10-4
10- 4
01
/ "'-24C (75F)
10-6
.5
2'
~
"tl
316C (600F)
I, /
ill
10-6
,
I
10-6
A2S6
I
10
20
30
40
60
SO
100
Fatigue crack growth rates for specimens of A286 stainless steel at room temperature and elevated temperatures
for tests in air at 3 Hz (RT) and 0.67 Hz (elevated temperatures), anR ratio ofO.OS, and at L-T, T-L, R-L, and R-C
orientations.
Source: J. E. Campbell, "Fracture Properties of Wrought Stainless Steels," in Application of Fracture Mechanics for Selection of
Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American Society for
Metals, Metals Park OH, 1982, p 161
'"I
2
Powder Forging
80
'
\I)
\I)
60
...
<J<:
'"
><J<:
...
40
s
<II
20
0
104
10 5
10 6
CYCLES
S-N curves for conventional and powder forgings of Astroloy (notched versus
smooth),
Testing was performed using standard methods at 705C (1300 OF) and a
combination of steady and vibratory stresses for which comparative data
were available. Cycles to first indication (crack) were comparable to conventional material. Crack propagation as judged by the number of additional
cycles from first indication to failure was slower than conventional material,
as shown above.
Source: M. M. Allen, R. L. Athey and J. B. Moore, "Application of Powder Metallurgy to Superalloy Forgings," in Source Book on
Powder Metallurgy, Samuel Bradbury, Ed., American Society for Metals, Metals Park OH, 1979, P 97
275
276
x =
Crack
o = Failure
Powder
Forging
Conventional
Forging
10 4
CYCLES
Source: M. M. Allen, R. L. Athey and J. B. Moore, "Application of Powder Metallurgy to Superalloy Forgings," in Source Book on
Powder Metallurgy, Samuel Bradbury. Ed., American Society for Metals, Metals Park OH, 1979,p 97
277
large grains
0.4
,
~
,
if
'"Z
~-LG
I I I
HI
MIN
'\
II:
t---O-I- IG
fine grain a
." 0.)
0.01
10.1~",1
HZ
AVG MAl
..
II:
~
~cb
0.01
(O.~'I
....&
......
~
~,
.,;
~
0
c
II:
E
E
II:
>
0.05
10.7621
...
c
...
II:
!:
0.0"'
11.0161
"-
...
II:
0.2
40
.0
IZO
160
ZOO
240
zeo
no
CYCLES TO CRACK
SoN curves for alloy FSX-430, showing effect of grain size on cycles to cracking,
Source: Eric Bachelet and Gerard Lesouit, "Quality of Castings of Superalloys," in Superalloys: Source Book, Matthew J. Donachie, Jr., Ed.. American Society for Metals, Metals Park OH, 1984, P 336
278
0.01
0.04
J--O-1-fG
0.4
~-LG
0.01
t-{j----I
.
'"
~
z
:ca:
...
a:
<
0.3
....
0)1
<
...%
..
!:
c
'"
C
0.03 II:
(O.TUI~
e,
II:
)~
<
Q,
,,
It'
IE
...i
O.O!
(o.~oeio
(O.Z:>41 E
E
0.04
(1.0151
large grlolinB
0.2
I
fine grains
0.5
1.0
1.5
GROWlH RAlE IN LINEAR flANGl (IO"n.lCTCLEI
2.0
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of SuperalIoys," in Superalloys: Source Book, Matthew J. Donachie, Jr.. Ed.. American Society for Metals. Metals Park OH, 1984, P 337
60
.u;
a.
50
..........
o
o
Q 40
30
20
1100
Aged 50 hr at 1350 F
""""-<
Fatigue strength
100 million cycles
"'- r---.....
(["---...:
--
:--
1400
1200
1300
Testing temperature, F
1500
Effect of testing temperature on fatigue strength of HS-31 casting alloy, after aging at 730C (1350 OF), for 100 million cycles.
Source: ASM Committee on Heat-Resistant Castings, "Heat-Resistant Alloy Castings,"in Source Book on Materials for ElevatedTemperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals Park OH, 1979, p 237
279
280
220
<,
CII
200
E
E
:z;
180
Po.
+1
Po.
......,
Ul
Ul
160
11M>
'-.<,
<,
<,
MATERIAL
<,
-"<.
"
...... ......
E-i
Ul
<,
..............
120
tI$..............
107
CYCLES TO FAILURE
SoN curves for casting alloy IN 738 LC. High-cycle fatigue properties ofnimocast alloy IN 738 LC tested at
850C (1560 oF).
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Super alloys," in Superalloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1984, P 340
240
/
220
E
E
2:
200
a..
"
STANDARD CDNDITIDN
-,
......
<,
.......
FINE GRAINS
.......
180
a..
111
111
W
0:
t-
111
160
......
.......
......
.......
140
MATERIAL
......
'
....... ~
120
e
10
10
10
CYCLES TO FAILURE
SoN curves for alloy IN 738 LC. High-cycle fatigue properties of extra-fine-grain and conventional material
tested at 850C (1560 OF).
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Superalloys," in Superalloys: Source Book. Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1984, p 340
281
282
0.4
.......-o--i ~
I I
M'N. HZ N ,
fG
-LG
0.0'
o.ZS41
MAX."2
'i
.!i
...
a:
0.02
_ O.J
IO.!lOI1
...
:...
f
....
c
e::
,
~~)-~......--<O)-----~
<,
........
....
eo
.20
160
200
CYCLES TO CRACK
oJ
:tIO
...
...a:
2:
0.04
11.0161
fine grains
2.40
::l
c
a:
c
0.03 a:
10.JUI
large grains
0.2
.;
0
320
S-N curves for alloy IN 738 Le, showing the effect of grain size on number of cycles to cracking.
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Superalloys," in Superalloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1984, P 335
283
O.GI
,
0.01
,
O.OJ
I
0,04
,
. 4
0.01
O.~~I
t;
.:
O.GZ
(0.!>08)
.0.) f-
.
......
...
!
&
..~
Ci
QOJ
(0.7611
.,
f-
........
..'"
%
0.0-
1.01'1
...
.....
IN ' ) I
t-O-i :E-Ci -
0.2 f-
0.'
1.0
FG
LG
1.5
1.0
SoN curves for alloy IN 738 Le, showing the effect of grain size on crack propagation rate.
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Superalloys," in Superalloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OB, 1984, P 336
284
11-13. IN 738 LC: Fatigue Crack Growth Rate at 850C (1560 OF)
10- 5
do
dN (m/cycle)
10- 6
100Hz
Alloy I IN 738 LC
Temperature I 850C
.0
10- 10
10- 11
fine-grained
t'<J
0.1
Waveform: Sinusoidal
10
20
30
40
-3/2
~K(MNm
)
50
60
Fatigue crack growth rate at 850C (1560 OF) in various grain sizes of alloy IN 738 LC.
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Superalloys," in Superalloys: Source Book,' Matthew J.
achie, Jr., Ed., American Society for Metals, Metals Park OH, 1984,p 341
000-
....
C>
(a)
20
'iii
0-
vi
VI
w
a:
10
50
t-
VACUUM
VI
Z
<Cl:
AIR
40
::E
(b)
30
20
10
105
10
107
LIFE. cycles
I I II lid
I I "" II
100
1000
10000
LIFE. h
Source: R. H. Cook and R. P. Skelton. "Environment-Dependence of the Mechanical Properties of Metals at High Temperature,"
in Source Book on Materials for Elevated-Temperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals
Park OH, 1979, p 81
285
286
500 427C
Joo
80
of)
,~ e:--
400
<0
a.
:;;
Ii 300
200
'-
<,
...
---
NOICh~~
specimens
(Kt~3.3)
29 Oc (85 of)
20
100
Cycles 10 failure
Source: Metals Handbook, 9th Edition. Volume 3. Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Metals, American Society for Metals, Metals Park OH, 1980, P 143
20
30 40506080 100
200 300
10- 4
-196C (-320F)
Ql
10-3
Ql
>
u
--EE
z
--
"C
"C
III
III
"C
10- 5
l!l
III
..
Ql'
s:
i
e
i
e
-269C (--452F)
Cl
"C
III
s:
--z,E
-....
Cl
10- 4
I!
u
I!
U
Ql
Ql
::J
Cl
::J
III
'':;
Cl
'':;
III
II..
II..
10-6
20
30 40 50 60 80 100
200
300
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties ofSuperalloys,"in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, p 297
287
288
"in
50
c.
g 4 a I-+-+-f-Ht--~H-Yt---+--+
o
:i 3 a 1--+-+++t--~,*h-r--+--P't"+l:>---1--+t-t-l
Q)
"-
iii
Ol
2 a I--+--+--r-f-..---r---t-+-+--i---i--t-++t---+--t-+-+i
~
o
E 10
.2?
<i
OL-..I...-...L....J.....LJ...---l...~
.............1...--'--...L....J.....LJ...---L.----L~
0.1
I
10
100
Millions of cycles to failure
Source: ASM Committee on Heat-Resistant Castings, "Heat-Resistant Alloy Castings," in Source Book on Materials for ElevatedTemperature Applications. Elihu F. Bradley, Ed.. American Society for Metals. Metals Park OH. 1979, P 235
289
Thermal fatigue
Cycles to first
crack
Material
HS-31
Inconel
ri
l-rnin cycles,
100 to 1700F
Material
avg1
avg
-1&
1713C"
,r
I
Thermal fatigue
Cycles to develop
.
8I-m,
erne k
HS-31
Inconel
Thousands of cycles
IIIII
1713C"
3 tests
each motertol
Thousands of cycles
Thermal fatigue properties of HS-31 compared with those of Inconel "713C." Left: Number of cycles required to initiate cracks. Right: Number of cycles required to develop VB-in. crack.
Source: ASM Committee on Heat-Resistant Castings. "Heat-Resistant Alloy Castings." in Source Book on Materials for ElevatedTemperature Applications. Elihu F. Bradley, Ed.. American Society for Metals. Metals Park OH. 1979, P 235
290
5 X 10-3
f-
o
2
o 5 X 10- Hz }
05 X 10-' H,
o",~/ /
5 X 10-4 -
1>
/S
/ili.
o 00
&
/;".
/&
o
5 X
10-5 I-
tL {--
a~a / I.
I
/
&
~/
20 Hz
2Hz
0.5 Hz
&&/
&
//
& ......---/
5 X 10-6
/
&1
1..----I._ _--1.._--'_...L...--1..--'---'--'-...L........L.---l
10
15
20
30
40
50 60
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Superalloys, "in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, P 294
25C (77F)
550C (1025F)
20 Hz
5 X 10-4
..
5 X 10-5
.. ,
:-
5 X 10-6
._~
: Twins
10- 6
......--L
10
No twins
....L._--'-_.J...-............L.....L.........L.-............
15
20
30
40
50 60
Dependence of FCP rate (dol dN) on stress-intensity factor range (LiK) and temperature at 20 Hz (sinusoidal
wave shape signal) for specimens of Inconel 718.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Superalloys, "in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, P 290
291
292
20
30
40
60
/'v (2)
5 X 10-2 Hz
2 Hz
_ rxr
(1)
/'v
(3)
(2)
rf
Sinusoidal ~
~
2
5 X 10- Hz
5 X
10- 4
I
(1)
'Jr
~
- / (3)
-;0
2X
I. _ -Q
r:t
I
I... /
rI
,6
.5
"tl
" tl
t6
..
Sinusoidal
2 Hz
5 X 10- 6
4l
10-6
2 X 10- 6
10
15
20
30
40
50 60
To separate out the possible effects of creep or environment from deformation mode, the authors used triangular and square wave shapes, as shown in
the top graph. The data obtained using the triangular wave at 2 Hz were the
same as the data obtained in other tests using the sine wave at the same
frequency which resulted in the lowest FCP rate. The effect of loading at the
same rate but imposing a lO-second hold time at maximum load was to increase the FCP rate only slightly, as shown in the lower graph.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Superalloys, "in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, P 295
iny2
10
20 3040 6080
5 X 10-1 1'"T'"--"""--T"-"""-T""T""1'""'I""l
Air
= He
10- 3
.,
--EE
10- 2
z'
--.,'..
"tl
Gl
>
u
......
.5
"tl
.....
10- 4
.I:
i0
..
--.
"tl
0>
...
.,
..
Z'
"tl
u
u
10-3
::;,
0>
.;::;
u..
10-5
MPa . m1/2
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Super alloys,"in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, P 287
293
294
10
20
30 40 60 80
5 X 10-1
10- 3
.,
~
E
E
--
10-2
.,
U
~
~
III
"'C
....,'
10- 4
--.5
z
-"'C
.t:
III
e
"'u..."
"'C
CI
III
U
10-3
.,
:::l
CI
'';::;
III
u..
10- 5
10- 4
He
+ 0.5%
0=
He
+ 5% S02
H 2S
L-_----I_~---L___L.....I.....L.J...L.J
10
20
~K,
40
60 80100
MPa m 112
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Super alloys,"in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, P 288
20
30 40 60 80
5 X 10- 1
Air
+ 5% S02
10- 3
II>
Q
>
u
--
10- 2
E
E
--'"
"C
II>
>
"C
i
...'"
10- 4
~
.S
'"
"C
."
~
eu
10- 3
II>
:s
."
.;;
'"
II..
10-4
'--_ _.J...._-'---'--'-...L..J.....L..L..J
10
20
40
AK, MPa
60 80100
m' /2
(The effect of air plus 5% S02 was similar to the effect of air alone.) It was observed that in
the helium atmosphere, which was used to establish a baseline, cracking was generally transgranular with well-defined striations. In the air, oxygen-bearing and sulfur-bearing environments, the crack path changed from transgranular to intergranular, indicating that an important
effect of the environment was to degrade the boundary strength by mechanisms that were not
clearly defined. It was suggested that oxygen diffusion along grain boundaries and localized
oxidation may have occurred. Another very important observation was that the effect of a given
environment on FCP could not be predicted on the basis of unstressed exposure tests. The attack
on the surfaces of unstressed specimens in aggressive S02 environments was minimal, but the
S02 environments caused substantial increases in FCP.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Superalloys, "in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, P 289
295
296
Spec. 1290" /
(CHT)
'f
~
E
III
v'
10-3
(!.
.. ,I:!> 0
/ ,
.. I
Z'
~
"~.'"
Inconel718
Tested in air at
24C (75F)
500 < f < 600 cpm,
R = 0.05
Conventional
heat treatment
40
60
80
Modified
heat treatment
I
I
(b)
20
40
60
80
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Super alloys,..in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH. 1982, p 276
-:
y "IJ./
.. ,
.I IJ./
/ ,~
Inconel718
Tested in air at
316C (600F)
f = 40 cpm, R = 0.05
t:J.
..
..I1J.4'
,0
/ IJ.I 0
.. IJ. .,
/
/
./
Conventional
heat treatment
IJ.
IJ.
(a)
Modified
heat treatment
II
(b)
II
.L..-_....L--''--J.-........--L....J-I
20
40
60 80
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Superalloys," in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, P 277
297
298
...----r--,-__r-..___..___.....-r::I:~--__r--..___.....,..____.____.___r.... T.:I
Inconel718
Tested in air at
427C (800Fl
f = 40 cpm, R = 0.05
Conventional
heat treatment
10- 2
D.
Modified
heat treatment
D.
:1
Spec. 162 (CHT1,,/1
/'1
/.~
10-3
Spec. 1291",: ~
(CHT)
lID.
0
0
.. I
II
.. I
II
10-4
.. I
.II
//
10-5 L...-
20
40
60
80
<9
----'_ _.L---L----'----''---L...............
20
40
60 80
Source: Stephen D. Antolovich and J. E. Campbell. "Fracture Properties ofSuperalloys, "in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH. 1982, p 278
1:""""----,--...-~-,.._r_1r_T"T"']r:__--__r--.___r-,.._r_1r_T"""1"":I
Inconel 718
Tested in air at
538C (1000F)
f = 40 cpm. R = 0.05
10- 2
10- 3
10- 4
10- 5
Conventional
heat treatment
Modified
heat treatment
20
40
60
80
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of'Superalloys, "in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, P 279
299
300
V'
Spec. 1289
(CHT)
10- 3
./
./
!SJ
/
10- 4
Conventional
heat treatment
b.
o
20
Modified
heat treatment
II
heat III
40
Stress-intensity factor range, l1K, MPa . m 1/2
Fatigue crack growth rate behavior of Inconel 718 tested in air at 649C (1200 OF).
CHT = conventional heat treatment. All testing was done at R = 0.05 and at a frequency of 0.67 Hz.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties ofSuperalloys."inApplication of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell. William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH. 1982, P 280
22 C (72 F)
}
-78 C (-108 F)
2.54 cm
'V -196 C (-320 F)
thickness
o -269 C (--452 F)
l>
Q)
10- 3
22 C (72 F) 0.51 cm
thickness
EE
z
<tI
"0
~f
~
.r:
e
Cl
.:.!
CJ
10- 4
Q)
:I
Cl
.;:;
<tI
u..
10
50
100
Source: Stephen D. Antolovichand J. E. Campbell, "Fracture PropertiesofSuperalloys,"in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, P 298
301
302
22C (72F)
10
50
100
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Super alloys,"in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OB, 1982,P 300
22C (72F)
-196C (-320F)
CIl
!E
10- 3
o -269C (-452F)
E
Z
~
"tl
:!l'
E
ie
en
CJ
10-4
CIl
5,
.~
'"
II..
100
Stress-intensity factor range, llK, MPa . m 1/2
Fatigue crack growth rates of Inconel X-750 at an
R ratio of 0.1 and at frequencies of 20 to 28 Hz.
Heat treatment: solution treated and double aged.
Within this frequency range, the effect of higher
temperature is to increase the FCP rate.
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Superalloys, "in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, P 299
303
304
...:::>~
'"
~
0.6
)(--------x
0.4
=ksi
100
80
0-_0______
50
i!:
...~
0"
...
0_
40
20
LONGITUOINAL SPECIMENS
ROTATING8EAM TESTS FOR \0' CYCLES
0'
200
400
600
I
50
60
0:
Ii;
1000 F
800
1
450
O!----..!:--....,-k".--""*';---=----.od;;--__=!
1200 I
I
200
300
400
600
TEST TEMPERATURE
Left: Interrelationship of prior tempering treatment and testing temperature with limiting fatigue stress, and
with fatigue ratio for Jethete M152. Right: Influence of alloy class and testing temperature on fatigue
strength for the same alloy.
Source: J. Z. Briggs and T. D. Parker, "The Super 12%CrSteels," in Source Book on Materials for Elevated-Temperature Applications, Elihu F. Bradley, Ed.. American Society for Metals, Metals Park OH, 1979, P 123
9o,------------..----,
CLASS II (Lapelloy)
90,---------------,
BO
alhersurface Irealmenls,
o Including surface lolline.
50
70
95% conlidencelimits
50
4~'='00,..----,-!;"....----:-!-::-----:*"--""IBO ksi
Left: Relation between surface hardness and mean fatigue limit for Lapelloy. Right: Relation between
tensile strength and mean fatigue limit for the same alloy.
Source: J. Z. Briggsand T. D. Parker, "The Super 12%Cr Steels," in Source Book on Materials for Elevaled-Temperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals Park OH, 1979, p 123
305
306
10
8
o
6
AIR
DRY AIR
WET AIR
VACUUM
PREOXIDIZED SPECIMEN. VACUUM TESTED
10
10
8
101
o
z
'"a:
V>
V>
a:
lV>
10
10
10 5
CYCLES TO FAILURE
Convergence of air and vacuum data was noted for AISI 3I6
steel at 1090 K, and a crossover of the air and vacuum curves
occurred for nickel, where it was suggested that oxide in cracks
could prolong life in air at low stresses. Crossovers have also
been seen in a ferritic stainless steel and a Nil Cr alloy in the
range 875-1025 K, where tests in purified argon gave shorter
endurances than those in air, impure argon, or sulfur dioxide.
Also, in single crystals of the alloy Mar-M200, air endurances
were less than those in vacuum at room temperature whilst the
reverse was true at high temperature (above). A thin oxide film,
formed during testing, suppressed surface crack initiation, but
oxide formed during pre-exposure did not.
Source: R. H. Cook and R. P. Skelton, "Environment-Dependence of the Mechanical Properties of Metals at High Temperature,"
in Source Book on Materials for Elevated-Temperature Applications, Elihu F. Bradley, Ed., American Society for Metals, Metals
Park OH, 1979, P 81
..
3,0 u>.
U
U
>.
u
4.5
N""-
E
E
'fg
'1'0
:il~ 4. 0
:
...
:il~
N""-
u
2.5 1;
or:
or:
iii
0\
:!.
0
iii0\
3.5
:!.
0
~
ij
......
;;;
;::
...
......
u
no
2.0
3,0
2.5 L-..
20
~c
.L-
40
Dendrite Arm Spacing
...r........
.....
60
1~1
Correlation between the initial crack propagation rate and the dendrite arm
spacing for Mar-MS09.
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Superalloys,' in Superalloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1984, p 338
307
308
r-----~---____r----
......---__r---_,
~
~
...
... 2000
;
c
-a......
- - - PresentStudy
u
'0
~
e
:::J
IlXXl
OL.-_.L.-_--L.
20
-'-
40
---'
-L..
60
---'
100
Source: Eric Bachelet and Gerard Lesoult, "Quality or Castings or Superalloys," in Superalloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1984. P 337
80
0
'in
70
KT
0
=
1.0
m'
fA
60
50
40
30
Life, cycles
Source: J. H. Moil, V. C. Petersen and E. J. Dulis, "Powder Metallurgy Parts for Aerospace Applications," in Powder MetallurgyApplications, Advantages and Limitations, Erhard Klar, Ed., American Society for Metals, Metals Park OH, 1983, p 275
309
310
F"t Solidification
(Condition f)
SlolI Solidlficotion
(Condillon S)
...
~...
......:>
Q
......... 100
..J
0.
...
Q
...'"
...
..J
s
~
...a:
......r
10
Q.
...
Q
...a:
CD
r
:z
:>
(;;:s;J
Creln
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Superalloys," in Superal1oys: Source Book, Matthew J, Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1984, P 339
10
20
40
60
80 100
10-2
10- 1
Stress-level
dependence
Ql
Ql
10-3
>u
-<;
E
E
z'
Z'
15-minute hold
time at maximum
tensile stress
-IV
..
"C
Ql'
-"C
IV
.....
"C
Ql'
IV
f!
10- 4
..c
Cl
>u
'<,
.~
10-2
"C
...0
10- 3
..c
~
2Cl
f!
f!
Ql
::3
Ql
Cl
::3
'';:;
Cl
'';:;
IV
10- 5
IV
U.
U.
10- 4
10
20
40
60
80
10- 6
100
The effect of environment need not always lead to more rapid crack growth. It has been
proposed that oxidation products could form in the crack tip region and prevent crack
resharpening during the unloading portion of the cycle. If the stresses are sufficiently low,
the oxidation products in the crack tip region will not be cracked and, in some systems, an
elevation of the threshold might occur. Such effects would be pronounced at high temperatures and long hold times and have actually been observed in Rene 95, as shown in the
above chart. Once the stress intensity is high enough to crack the oxides, the rate of crack
growth would be expected to increase due to the severely degraded region in the crack tip
zone.
Source: Stephen D. Antolovich and J. E. Campbell. "Fracture Properties of Superalloys. "in Application of Fracture Mechanics for
Selection of Metallic Structural Materials. James E. Campbell. William W. Gerberich and John H. Underwood. Eds.. American
Society for Metals. Melals Park OH. 1982. P 284
311
312
100
10_2 20
1000
..
2000
10-4
.5
z
-.r
....
.!
...
...
i'
10- 3
:i
l!
~
e
11'"
11'"
.."
l!
10-6
:::J
..~
:::J
'"
.~
...
'"
...'=
10- 4
o
Testing tempereture, C
bJ(
That the effect of environment can be large may be inferred from some low-cyclefatigue
studies of Rene 95 in which surface and subsurface cracking was observed at comparable
strain ranges and defect sizes. As expected, the life of the subsurface crack was much
greater than that of the surface crack, leading to the hypothesis of a strong environmental
effect. This possibility is considered in more detail in an analysis of FCP properties of
Rene 95. The FCP rate was plotted as a function of temperature for a given !:J.K range, as
shown in the above chart. It is noteworthy that there is a minimum in the FCP rate at all
!:J.K levels except 22 MPaM 1/2 (20 ksi-in.b"), where the data are at least suggestive of a
minimum. Because any environmental interaction is thermally activated, the crack
growth rate at a given !:J.K level and frequency may be written as:
da
dN = Aexp - Q(!:J.K)/ RT
Source: Stephen D. Antolovich and J. E. Campbell, "Fracture Properties of Super alloys, "in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, P 282
a..
~
vi
~
30
210
"-
"-
'Iii
'n. ...........
E
:J
.~
vi
~
20
A::;oo
0-_-0... K,
= 3.4
----__
0
r-v
--00"--
106
Number of cycles
SoN diagram for 8-816 heat-resisting alloy tested at 900C
(1650 OF), notched (broken curve) versus unnotched (solid
curve).
140
'Iii
E
:::J
E
'x
ltJ
~
313
314
11-43. Udimet 700: Fatigue Crack Growth Rates at 850C (1560 OF)
AK, ksi in. 1/ 2
10
40
20
60 80100
10- 3
10- 2
Q)
Q)
>
t.l
--EE
>
t.l
--.E
z'
"tl
'<,
'"
10- 4
"tl
...~
Q)'
.t:
01
-"
t.l
...e
Q)'
s:
'"
"tl
e
01
10-3
-"t.l
t.l
t.l
.'"
Q)
Q)
...
01
.01
u,
'"
u,
::J
::J
10- 5
Stage I
0
m = 16
KN
10- 4
Ib
c 582 1310
v 711 1600
01067 2400
l> 1244 2800
160 3600
20
10- 6
40
60 80 100
Source: Stephen D. Antolovich and J. E. Campbell. "Fracture Properties of Super alloys,"in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds.. American
Society for Metals, Metals Park OH, 1982, P 285
- - COLUMNAR GRAINE\i
AND SINGlE CRYSTAl
-- ..
-- --.
M~R.:-,~OO
10-2
-c
1700 0 F
-c
a:
Z
a:
1'-.
l-
V)
.....
-(
I-
-. -- ..
I-
w- 2
oI-
1-..
C>
- - - CONVENTIONAllY CAST
MAR.M200
I I I
---WROUGHT POIYCRYSTAlliNE
UDiMEl 700
.~
I':::--
--_.~~
r--
-.
'--.
!--.
'.
1400 0 F
,
103
CYCLES TO FAILURE
Comparison offatigue properties at 760 0 C (1400 0 F) and 925 0 C
(1700 OF) for a typical wrought nickel-base alIoy (U-700) with
conventionalIy cast, directionally solidified and mono crystal ,
Mar-M200.
Source: Francis L. Versnyder and M. E. Shank, "The Development of Columnar Grain and Single Crystal High Temperature
Materials Through Directional Solidification," in Source Book on Materials for Elevated-Temperature Applications, Elihu F.
Bradley, Ed., American Society for Metals, Metals Park OH, 1979, p 358
315
316
a. 1200
::iE
Ul
STRAIN CONTROLLED
1100
:::l
t-
:J
a.
::iE
V)
V)
Ul
0::
t-
WASPALOY
V)
10
100
N, CYCLES
1000
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure,
American Society for Metals, Metals Park OH, 1979, P 217
~ %
}
-'---}
----
Z
....
<
at 750 DC
at 650 DC
~ 0.6
1Il
L!J
0.5
0.4
z< 0.3
0::.
Small
I-
...J
-e
0.2
-....
0.1
2
100
2
1000
5
10.000
CYCLES TO FAILURE
S -N curves for X-40 showing effects of grain size and temperature on fatigue characteristics of this alloy.
Source: Eric Bachelet and Gerard Lesoult, "Quality of Castings of Superalloys," in Superalloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH. 1984, p 335
317
318
IlJ
a::
:J
='300
it
o
F-200
~150
...J
~IOO
<l
::!:
a::
ILl
~ 50
u, 40
ffi
30
z~
20
ILl
15
ILl
10
~
a::
10 15 20 30 40 50
100
200 300
CYCLES TO FAILURE OF INDIVIDUAL TESTS
The design of components that are subject to considerable temperature cycling must also
include consideration of thermal fatigue. This is particularly true ifthe temperature changes are
frequent or rapid, and nonuniform within or between casting sections. Fatigue is a condition in
which failure results from alternating load applications in shorter times, or at lower stresses,
than expected from constant-load properties. "Thermal fatigue" denotes the condition when the
stresses are primarily due to hindered expansion or contraction. Good design helps minimize the
external restraint to expansion and contraction. Rapid heating and cooling may, however, impose temperature gradients within the part causing the cooler elements of the component to
restrain the hotter elements. Finite-element computer analysis has shown that, for some industrial applications, these thermally induced stresses may exceed those resulting from the mechanicalloads.
An example of results from thermal fatigue data is presented above. This graph offers a ranking of many cast heat-resistant high-alloy grades relative to their resistance to thermal fatigue.
Such rankings are indicative of general alloy properties only because most thermal fatigue tests
are based on an arbitrary set of experimental conditions rather than on their fundamental material behavior. Nevertheless, such test results have been useful in considering alloy selection questions, and in identifying the superior thermal fatigue resistance of nickel predominating grades
and the good performance of some HH type compositions.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980,
P 19-7
500,.------typical
corrosion
-.,
fatigue strength
70
R -1 )
60
400
';:;'"""'
-.
z
300
40
J:.
0.
c:
-F
200
.~
'iii
.><
L-J
iii
,.....,
100
30
.!j
nickel alloys
(e.g. aJloys 600 and fn
r.. .-/ .
.S!'
~
20
ferritic
_ stainless
martensilic_ \ steels
iii
.~
8
1)
As shown above the corrosion-fatigue strength of bare aluminum alloys is superior only to that of magnesium alloys. Careful
surface protection may bring the corrosion-fatigue strength up
into the range of bare stainless steels or copper-nickel alloys.
Source: Markus O. Speidel, "Aluminum as a Corrosion Resistant Material," in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, R. Biloni and D. E.
Embury, Eds., American Society for Metals, Metals Park OR, 1980, P 617
319
320
250
()~ /
':>.o~'b
200
"3 /
6.
6./
r0-
ll..
~
E::::: 150
:.= (J)
.'=
0)
0)-
c i:)
moo 100
~O
"O~
C )(
UJ~
50
O~---JL...-_----I.
100
x
0
6.
200
300
400
500
600
700
Source: I. J. Polmear, Light Alloys, Edward Arnold Ltd, London, England, and American Society for Metals, Metals Park OH,
1981, P 39
500
8
in river water I N - 10
o
o
R - -1 , .. 60 Hertz I
stnooth specimens I
ambient temperature
4()()
70
60
+
50
';:;"'"'
--z
E
.--.
300
'iii
""
'--'
40
s:
s:
e;,
e;,
c
30
l!! 200
OJ
l!!
OJ
000
0CF,air ,
N-5x1Q1
-Orr-----l:I
20
100
000
I
~ ~
~~
II
~ ~~
I
~I
f2,
<0
~
I
Cl)
iij
...
Source: Markus O. Speidel, "Aluminum as a Corrosion Resistant Material," in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and D. E.
Embury, Eds., American Society for Metals, Metals Park OH, 1980, P 616
321
322
Q)
10
~
E
'----'
10-6
Z
~
<I
10- 7
10
crack orientation L - T
specimens CNP, DCB I CT
ambient temperature
environment air
R =0,0-0,1
0,1 -30 Hz
10- 9
---+--------------------4
1).-....
10
20
50
40
I
.t:.K I
60
MN m- 3f2]
Source: Markus O. Speidel, "Aluminum asa Corrosion Resistant Material, "in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloniand D. E.
Embury, Eds.. American Society for Metals. Metals Park OH, 1980, p 613
~ 30
-----------
LU
cz::
<:
:I:
20
><::
10
AI 1100 HO
AI 1100 H14
Comparison of the Knoop hardness for well-annealed (HO) and coldrolled (H14) aluminum as a function of fatigue.
Source: O. Buck and G. A. A1ers,"New Techniques for Detection and Monitoring of Fatigue Damage," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 128
323
324
16xlf
80
......
Cla
-e
zlA I
60
Vl
......
Vl
.....
z
-c'"
40 oe:
::J:
ll.
K.H.
0
0
ZO
z~
AI 1100
0
0
80
120
0
160xllY
FATIGUE CYCLES
Normalized second harmonic displacement and
Knoop hardness as a function of fatigue.
The effects of dislocation rearrangements on harmonic generation within the bulk of the material during fatigue are shown
in the above chart. Using 3D-MHz longitudinal waves, the normalized second harmonic amplitude of an initially compressiondeformed Al 1100 single crystal was monitored and found to
increase as a function of compression-compression fatigue. At
the same time, the surface hardness (Knoop) decreased. Apparently, the dislocation-loop length prior to fatigue was quite
short, since the initial amplitude of harmonic generation was
small. During fatigue softening, the cell structure that developed
(with its individual dislocations within the cells) became quite
large, so that a change of the dislocation-induced harmonic generation, A 2d , increased. Application of this technique to highstrength aluminum alloys failed, however, apparently because
of an immediate rep inning of the long loops by interstitials in
this alloy.
Source: O. Buck and G. A. Alers, "New Techniques for Detection and Monitoring of Fatigue Damage," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 131
50
~ 40
)(
N.
c:
~ 30
VI
VI
....
cr:
:n
20
10
Source: P. C. Varley, The Technology of Aluminium and Its Alloys, Butterworth & Co. Ltd., London, England, 1970, p 43
325
326
..
40
~ 35
Ii
Q)
J; 30
Cl
.- Ultrahigh Vacuum
0- Air
,
~
'"
.........
..........
<,
<,
"0
; 25
~,
<,
.........
En'durance Limit
15 s
10
I I I I III"'
2
8 10'
6 8 10
.
"~
...
J
7
2.5 x 10
l/)
~~
-3
)(
-3
'~
r""- ...
III
E
::J
, 20
r-
- 3.5 x 10
1.5 x 10- 3
Number of Cycles-to-Failure
The effects of air versus vacuum on the fatigue life
of a 2024-T3 aluminum alloy.
Source: D. J. Duquette, "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys," in
Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 337
rf.
:E
VI'
VI
...e
(/)
50
300
40 'w
...:
250
~.
30
200
150
-----2 1
100
0.01
....::.4"-3
0.1
en
20
10
For the data here, sheet 1.02 mm (0.040 in.) thick was annealed, solution heat treated and quenched, and then fatigue
tested. The sheet represented by curve I was not bent. All other
sheet was bent 90 in the annealed condition. Flattening (unbending) was done in either the annealed condition (curve 2) or
the solution heat treated and quenched condition (curves 3, 4
and 5). Details of bending and flattening were as follows: (I) Not
bent. (2) Bend radius, 3.18 mm (Ys in.); flattened in annealed
condition. (3) Bend radius, 3.18 mm (Ys in.); flattened in
quenched condition after 3 days of storage at-18 to-12C (0 to
IOOF). (4) Bend radius, 3.18 mm (Ys in.); flattened in quenched
condition after 14 days of storage at -18 to -12C (0 to 10 OF).
(5) Bend radius, 1.59 mm (1/ 16 in.); flattened in quenched condition after 3 days of storage at - 18 to - I2 C (0 to IO F).
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society
for Metals, Metals Park OH, 1979, P 35
327
328
III
'5
C
30
C::P
/
/
20
/
/
/
~a:
a:
8
A
/,
/,
10
/
/
0
0
0.5
1.0
Source: Sigmund Weissmann and William E. Mayo, "Determination of Strain Distributions and Failure Prediction by Novel X-ray
Methods," in Nondestructive Evaluation: Application to Materials Processing, Otto Buck and Stanley M. Wolf, Eds., American
Society for Metals, Metals Park OB, 1984, p 195
32
30
28
u
26
.....0
24
VI
-;'"
c::
22
I~
20
sc;
18
..,.z
.c::
e'"
16
0'
'"
>
14
,,
12 :
5,000
10,000
15,000
20,000
Here is shown that for the maximum stress of241 MPa with R
= 0.1, the 13 value increased during the first several hundred cy-
cles. This was more pronounced for the surface grains (Cr
radiation).
KO'I
Source: Sigmund Weissmann and William E. Mayo, "Determination of Strain Distributions and Failure Prediction by Novel X-ray
Methods," in Nondestructive Evaluation: Application to Materials Processing, Otto Buck and Stanley M. Wolf, Eds., American
Society for Metals, Metals Park OB, 1984, P 194
329
330
~-_:_------,O----..."....----_1
ICQ.
14
13
_ _ _ _ _ ~o (inillal ~alf.i~_~_)
12
IIL..-_ _..........
50
.L--_ _- - ' -
100
150
....l...- _ _---J
200
250
I'm
Dependence of the average rocking curve halfwidth 11 on depth distance from surface for different fractions of corrosion fatigue lives, N F' of Al
2024-T4.
X-ray rocking curve measurements were carried out as a function of depth distance from the surface, and typical results ofthe
dependence of 13 on depth distance for an alloy cycled with a =
276 MPa, corresponding to the static yield stress, are shown
above. It may be seen that the minimum 13 values at the surface
layers were larger than those in the interior. The 13 values declined up to a depth of about 50 Mm from the surface and subsequently retained a plateau value throughout the interior of the
specimen for each fraction of the life.
Source: Sigmund Weissmann and William E. Mayo, "Determination of Strain Distributions and Failure Prediction by Novel X-ray
Methods." in Nondestructive Evaluation: Application to Materials Processing, Otto Buck and Stanley M. Wolf. Eds., American
Society for Metals, Metals Park OH, 1984, P 193
260
tV
a...
:2:
CIl
CIl
220
CIl
Ol
C
.~
~ 180
<i:
140 ~
105
---'106
o ---'----'
Cycles to failure
Effect of reducing the concentration of submicron
particles in an AI-Cu-Mg alloy. X2024 is a highpurity version of the commercial alloy 2024.
The disappointing fatigue properties of age-hardened aluminum alloys are also attributed to an additional factor, which is
the metastable nature of the metallurgical structure under conditions of cyclic stressing. Localization of strain is particularly
harmful because the precipitate may be removed from certain
slip bands, which causes softening there and leads to a further
concentration of stress, so that the whole process of cracking is
accelerated.
The fatigue behavior of age-hardened aluminum alloys
would therefore be improved if fatigue deformation could be
dispersed more uniformly. Factors which prevent the formation
of coarse slip bands should assist in this regard. Thus it is to be
expected that commercial-purity alloys should perform better
than equivalent high-purity compositions because the presence
of inclusions and intermetallic compounds would tend to disperse slip. This effect has been confirmed for an Al-Cu-Mg alloy, and fatigue curves for commercial-purity and high-purity
compositions are shown in the above S-N diagram. Here the
superior fatigue behavior of the former alloy arises because slip
is more uniformly dispersed by submicron dispersoids such as
MnAI 6
Source: I. J. Po 1mear, Light Alloys, Edward Arnold Ltd, London, England, and American Society for Metals, Metals Park OH,
1981, P 40
331
332
10
15
20
10- 3
10- 2
Q)
Q)
>
u
--
E
E
10-4
10- 3
10-5
'C
10-6
10-5
..
.'"
~
0
en
10- 7
en
'';:::;
Q)
'"
'0.
c,
:l
en
'~
'C
.~
Q)
e,
UI
.. -~
'u
U.
10-6
1......
.:.:
o
Q)
Q)
:l
'"
....
...
en
U.
'C
'"
.s:
~
0
--'"
Q)"
10- 4
s:
.:.:
--.5
'C
'C
Q)"
>
e
--'"
.....
...'"
.
'".
Q)
20
Stress-intensity factor range, LiK, MPa m 1/2
Comparison of typical particle sizes in aluminum alloys
with crack advance per cycle on fatigue loading.
Source: J. G. Kaufman and J. S. Santner, "Fracture Properties of Aluminum Alloys." in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OK 1982. P 191
2.6
2.4
...
in
co
..r:::
CIl
200
fJl
"0
0>
0
300
fJl
fJl
Ql
AA
400
...
Ql
fJl
fJl
fJl
Ql
in
600
500
~
<,
700
2024 T4 oNjj
2124 T4 ANjj
2.8
co
2.2
A Ao
...J
2.0
102
103
105
..r:::
AA
"0
z
106
107
Cycles to fatigue crack initiation for specimens of aluminum alloys 2024-T4 and 2124-T4 versus stress at notch
base (computed using Neuber stress-concentration factor).
Source: M. E. Fine and R. O. Ritchie, "Fatigue-Crack Initiation and Near-Threshold Crack Growth," in Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 251
333
334
10
100
10- 1 r - - - - - - - , - - - - - -....- - - - - - - . ,
7075-T6
9 investigations
10- 4
2024-T3
8 investigations
>u
'<,
10- 5
10-
Note:
Bounds defined
by mean curves
of separate
investigations
Cl
~
E
u
~
G>
10-
.~
z
"C
......
01
"C
10- 6
Cl
.~
U.
10-7
10- 6
10
100
Source: J. G. Kaufman and J. S. Santner, "Fracture Properties. of Aluminum Alloys," in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerherich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982,P 189
'"
rn
UI
414
_ - -----+---j--+_
345
~---l~-~~
e,
- --
::E
276
207 I
a:
l-
rJ)
138
2024-T4 ---+----1
."
~.,-::'~---+---+---j
I,..... 1'I,--'~-r-''"'
". - ~.....;-"'~_ _+-_--j
~ ..
"
'"'-I' ,t'i:::::~"
J
-, <,
:.:~,.;-.~
c ~- -
<,
'q".
~.
"..
69
0
I---L.--+~-t---+-
414
I", -,,
7075-16
~----I
345 I----i---"~~.-:-l.c---+----jf----+----j
'"
c,
::E
rn
rJ)
a:
~'
.\ . f~,~~.l,.
~
276
207 1---+---''':-....'
lUI
''':".":--
--+
,,--
- - -
.~":''-1''''.1
"
t': ':', .... ~'~r, ::""II'jl
~.
,', " .,'\ ~,.
"
"e'::;"',"-j--""'-f'--..
69
o
1~
:
,
~.,"
: ~gmg
~~~TE
FORGINGS
1'"
....::.:,
EXTRUSIONS
010 NOT FAil
----L
1~
1~
1~
---~..!...:'.I,
'---->":"
-.#_.s
. _ _ . __ _ _
1~
10'
!
I.
-. --}.
"-,.J.'
108
109
CYCLES
Numerous methods have been developed to evaluate response of materials to cyclic deformation.
The earliest method was by use of S-Nplots. Typical examples are depicted above. Basic specimens
include rotating beam, axially stressed and sheet flexure. Notches have been employed to provide stress
concentration, and special specimens have been used to simulate a variety of other conditions. The S- N
response is strongly influenced by a number of conditions, including surface condition, stress ratio, and
environment. The various alloys differ widely in their response to fatigue testing-specifically, in the
number of cycles where a "level out" condition is attained. As shown in the above S- N diagrams, the
SoN response for aluminum alloys tends to level out as the number of applied cycles approaches 500
million.
Based on SoN data of smooth and sharply notched specimens and of similar tests of specimens
designed to simulate joints in structures, the following conclusions have been drawn. From fatigue
results for aluminum alloys obtained with smooth specimens . . . rather wide variations can exist
without causing appreciable differences in fatigue strengths. . . . When severely notched specimens
are used, the effects of composition and temper are even less pronounced and generally are of no
practical significance. . . . As in the case of simple notch fatigue tests, there is a lack of significant
differences in the fatigue strength of the joints of the various alloys.
Despite these laboratory data, users discovered that certain aluminum alloys performed decidedly
better than others in service when fluctuating loads were encountered. For example, airframe manufacturers determined that fatigue performance of alloy 7075-T6 was unquestionably inferior to that of
alloy 2024-T3.
Source: T. H. Sanders, Jr. and J. T. Staley. "Review of Fatigue and Fracture Research on High-Strength Aluminum Alloys." in
Fatigue and Microstructure. American Society for Metals, Metals Park OH, 1979. P 470
335
336
f
---l_+_~=1B
--F
+-__-+-_ _
400 1-_ _
<0
0..
::E
Stress ratio: R = 0
Ambient air
vi
III
-+-_ _--+
t -_ _-1 60
50
300 f----I------1I-------1I-----j----j-----t-----l
~
III
E
:J
E
'x
<0
::E
E
:J
E
'x
<0
::E
200 1 - - - - / - - - + - I f - 1 . * - - - I - - - - - t - - - - - t - - - - - + - - - - - - l
20
-:::~~~~~"'F;;;;;:;:;;:::::j::~-__jl--I
100 f -_ _ 2024-T351
rod and bar band -
~"t----I-~ 7075T73XXX
10
products band
oL -_ _-.l.
10'
103
....L
...L.
10'
105
L -_ _---'
10'
10'
....L_ _~
10'
10'
Cycles
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals, Metals Park OH, 1984,p 468
'"
70
60 ""'"
Q..
:i:
400
::I"
::I"
I;.
'" 300
E
:::J
'x 200
'"
:i:
100
103
R = 0.1
o 0 I;. Longitudinal
I
... Longtransverse
10 4
I;.
105
106
Numberof stress cycles
50 ~
'"
40 E
:::J
E
30
'x
'"
:i:
20
107
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals. American Society
for Metals, Metals Park OH, 1979, P 80
337
338
s:
:::E
400 ~
Ii:
E
300
.~
200
10
Unnotched
R
r-,
0.1
60 ]
50
24C 176of)
176C 13600F)~_
g
E
40 ~
-=-
30
:::E
20
101
Number of stress cycles
...:::E..
400
g ~
300
E
~ 200
.~
:::E 100
~ s,
0.1
~J
50 ]
- 40 g
30
_ 20
--;;;;1
"i
:::E
Source: Metals Handbook. 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society
for Metals. Metals Park OH, 1979, P 82
339
AK. kSi'in. 2
AK. kSI'in.2
10
10
LT crack oriant!tion
TL crack orianLtion
l~3.4
"
,
10-8
10- 5
{.:
.5
~.
..:!
rot"
-1i
~
10-5
.5
..:!
..:!
10-6
M-a.:
M~7.1
61
10
"iI
10- 6
61
10
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society
for Metals, Metals Park OH, 1979, p 81
340
s:
60
::;:
~
1;;
"E
'x
..
E
E
"
200
.~
::;:
::;:
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society
for Metals, Metals Park OH, 1979, P 81
0% RH
50% RH
-300
L--
....J..
10
...!-
20
.l..-
30
---I.
40
...J
50
CYCLES (x 10-3)
The dependence ofrelaxation behavior on the cyclic hardening parameter, (J. (J was varied by changing the relative humidity (RH), which affects the near surface ductility in this alloy. Values used were: (J= 6 X 10.5
for 50% RH and (J= 2 X 10-5 for 0% RH. The cyclic stress amplitude was
0.88 a yleld for both samples.
Source: M. R. James and W. L. Morris. "The Relaxation of Machining Stresses in Aluminum Alloys During Fatigue." in Residual
Stress for Designers and Metallurgists. Larry J. Vander Walle. Ed.. American Society for Metals. Metals Park OH. 1981. P 184
341
342
o
0.64 a YIELD
-300
.....
o~-----'--------l----...J-.-------l----
10
20
30
40
50
CYCLES (x 10- 3)
Surface milling produced the shallowest stress gradient and resulted in the
slowest rate of relaxation of the surface stresses. A comparison of measured
to predicted values of residual stress during fatigue is made for four "as machined" specimens in the above chart. The residual stress values were measured parallel to the external stress axis. A value of f3 = 0.0004 was used to fit
the data for all specimens. Residual stress measurements were also made in a
direction transverse to the applied stress axis. Within experimental error, the
cyclic relaxation rate was the same as in the longitudinal direction.
Source: M. R. James and W. L. Morris, "The Relaxation of Machining Stresses in Aluminum Alloys During Fatigue," in Residual
Stress for Designers and Metallurgists, Larry J. Vander Walle, Ed., American Society for Metals, Metals Park OH, 1981, P 182
-300
l--
--L
-L-
10
20
..L....-
30
----I
40
.....
50
CYCLES (x 10-3)
The relaxation behavior oftwo samples having different depth distributions of residual stress. Note the difference in the peak cyclic stress, a.
0= rolling (10%reduction); f3 = 0.012; a= 0.91 ayield'. = sand blasting;
f3 = 0.003; a = 0.71 ayield'
Source: M. R. James and W. L. Morris. "The Relaxation of Machining Stresses in Aluminum Alloys During Fatigue," in Residual
Stress for Designers and Metallurgists. Larry J. Vander Walle. Ed. American Society for Metals, Metals Park OH. 1981, P 183
343
344
The solid line in the graph represents failure; the dashed lines
indicate the percentage offatigue life expended. The exact location of these lines is highly sensitive to the material and its
microstructure as well as the influences of environment.
Source: O. Buck and G. A. Alers, "New Techniques [or Detection and Monitoring of Fatigue Damage.t' in Fatigue and Microstructure. American Society for Metals. Metals Park OH, 1979. p 104
30
If. 200
:t!
::;;
g
E
E
160
20
"
..
'x
::;;
E
E
.."
'x
::;;
100
10
50
Number of cvcles
The fatigue life of welded joints at high loads varies with the
alloy. As the load is decreased, differences disappear until, at
about one to ten million cycles of axial loading (R = 0), the
fatigue strength of an arc-welded joint is approximately the
same regardless of alloy and is 50 to 70% that of the unwelded
alloy. Typical data are given in the above graph for three aluminum alloys. Specimens were from 9.5-mm (Ys-in.) plate; weld
reinforcement removed; axial loading; R = O.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society
for Metals, Metals Park OH, 1979, P 195
345
346
10- 4
..
u
10- 3
>
~
-.
-.t.l
E
E
10-5
'C
-.
Cll
'C
.E
'C
-.
Cll
'C
10- 4
10- 6
T-S
10-5
Source: J. G. Kaufman and J. S. Santner, "Fracture Properties of Aluminum Alloys," in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, P 193
10- 2
,/
RT, moist
air~i
,,
,,,I
,
10- 3
Ql
"C
Ql
EE
z
~
10- 4
10- 5
.E
~
III
"C
10-4
10-6
Compact specimen thickness
= 46 mm (1.8 in.)
R = 1/3.
f=13&18Hz.
T-L orientation
10-5
'--
...L-_--L.
5
~K.
10
'--~
10- 7
50
MPa m1/2
Source: J. G. Kaufman and J. S. Santner, "Fracture Properties of Aluminum Alloys," in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William w. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, P 195
347
348
0.5
0.4
....
.~
0.3
a:
0.2
0.1
5086-
5086-
6061-
7075-
2024-
H34
H36
T6
T73
T3
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society
for Metals, Metals Park OH, 1979, P 220
60
400
40
300
200
.:
1II
<I
10
8
CT,=+18kSi
KI~
I 4
10
s~s
30
20
10
10~
NT' Cycles
Source: F. V. Lawrence, "The Predicted Influence of Weld Residual Stresses on Fatigue Crack Initiation," in Residual Stress for
Designers and Metallurgists, Larry J. Vander Walle, Ed., American Society for Metals, Metals Park OH, 1981, p 114
349
350
eo
40
~3-o1!l183
Bull Weld
Kt..." 3.HI.
400
300
- - .,.5,
200
--- .,0
R"O
~-.
S~S
1(,...
'
30
20
10
Source: F. V. Lawrence, "The Predicted Influence or Weld Residual Stresses on Fatigue Crack Initiation," in Residual Stress for
Designers and Metallurgists, Larry J. Vander Walle, Ed., American Society for Metals, Metals Park OH, 1981, p 113
351
.fIN
AllOY
7075T6
7050T6
7075T7
7050-T7
W
0
:::l
f-
::J
Q..
::iE
UTS
kSi (MPa)
810(558)
887(612)
73.2(505)
74.1(511)
<>:
TYS
ks; (MPa)
769(530)
82 7(570)
658(454)
63.8(440)
./,
fl. 2"
95
12.0
11.0
14.0
z
<i:
a:
f-
en
...J
10- 2
/7050.T6
<>:
:-~-~ ----.
--"""""''''''''""""",""
f0
f-
-------
"""'''''''''''''1;;./11
-'-
-'-_ _............
10 7
REVERSALS TO INITIATION, 2N
Cyclic strain versus initiation life for laboratory-fabricated high-strength 7XXX aluminum alloys,
Plots of elastic-strain amplitude versus life have seen relatively little use for commercial alloys, but plots oflog total strain
amplitude versus life have been used more frequently to compare materials. This approach offers the advantage that both
high- and low-strain fatigue may be characterized with one plot.
As illustrated above, fatigue resistance at low total strain amplitude is governed by the elastic-strain amplitude. Fatigue lives
for total strain amplitudes less than about 5 X 10-3 generally
increase with increasing strength. On the other hand, fatigue
lives for total strain amplitudes greater than about 10-2 generally
increase with increasing ductility.
Source: T. H. Sanders, Jr., and J. T. Staley, "Review of Fatigue and Fracture Research on High-Strength Aluminum Alloys," in
Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, p 472
352
Zn
Mg
Cu
5.5
2.2
2.3
to
to
to
6%
2.4%
2.4%
6%Zn
2.2% Mg
0.4% Mn
1.2
Low humidity
Aged 3 h at 121C
(250F) + 9 h at
163C (325F)
Alloy 7050
1.0
0.8
High humidity
0.6
0.4
0.2
% Zr
%Mn
0.1
0.4
0.1
0.4
%Cu
1.0
2.3
1.0
2.3
Effect of dispersoid type (based on composition) on fatigue crack propagation life of 7050 alloy sheet.
Source: J. G. Kaufman and J. S. Santner, "Fracture Properties of Aluminum Alloys," in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, p 192
Q.
.400
UJ
::J
I...J
Q.
~ 3
(J)
(J)
UJ
7050 AR
7050 HR
II:
:n
200
...
105
CYCLES
106
TO FAILURE
(HR).
As indicated in the above graph, grain shape showed no perceptible difference in life over a range of stress amplitudes.
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure,
American Society for Metals, Metals Park OH, 1979, P 238
353
354
300
~ od~
\
m
a..
~
tl
c
200
0\0
&
6-
<i:
0
"-
\6,
'
0 0'0Q,
""4_
6
100
L..-
10
TS
632
627
573
567
'6
\
0
6'h
CJ)
.~
\~
\
If)
If)
PS
587
600
516
488
7075 TMP
... 7075 TMP
o 7075-T651
67075-T6
"""'---
'lIQ .Q..
--_~~
"""'---
.........
........
Cycles to failure
Effect ofthermomechanical processing (TMP).on the unnotched fatigue properties of the commerical AI-Zn-MgCu alloy 7075. PS = proof stress (MPa); TS = tensile
strength.
Detailed studies of the processes of fatigue in metals and alloys have shown that the initiation of cracks normally occurs at
the surface. It is here that strain becomes localized due to the
presence of pre-existing stress concentrations such as mechanical notches or corrosion pits, coarse (persistent) slip bands in
which minute extrusions and intrusions may form, or at relatively soft zones such as the precipitate-free regions adjacent to
grain boundaries. Density has also been found to improve the
fatigue performance of certain alloys, although this effect arises
in part from an increase in tensile properties caused by such a
treatment (see above diagram). It should be noted, however,
that the promising results mentioned above were obtained for
smooth specimens. The improved fatigue behavior has not been
sustained for severely notched conditions, and it seems that the
resultant stress concentrations override the more subtle microstructural effects that have been described.
Source: I. J. Polmear, Light Alloys. Edward Arnold Ltd. London, England, and American Society for Metals, Metals Park OH,
1981.P 41
1:1
400
Q.
~
ILl
;:)
I-
300
:J
Q.
~
c:(
I/)
I/)
200
ILl
II::
l-
I/)
100
104
105
106
CYCLES TO
107
108
FAILURE
Effect of inclusion density on the stress-life behavior oftwo 7XXX alloys: high-inclusion density, alloy 7075; low-inclusion density, alloy 7475,
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure,
American Society for Metals, Metals Park OH, 1979, P 233
355
356
400
'E
z
7075
oY;I~Nm-' ]
7075TMT 600
:gw
300
a:
~
til
Cl
Z
200
a:
w
100-'----,..---.-----,-----.----.--10'
10'
10'
107
10'
CYCLES TO FAILURE
Source: G. Lutjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys," in Aluminum Transformation Technology and
Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and
D. E. Embury, Eds., American Society for Metals, Metals Park OH, 1980, p 195
fi
. . . . .
?<
::'::)
2.0 X 106
Total life
Overload ratio
= 1.8
Applied every
4000 .cycles
Overload ratio
= 1.8
Applied every
8000 cycles
'0
1.0 X 106
E
::::J
0.5 X 106
Constant
amplitude
(Overload ratio
= 1.0)
Overload ratio
=1.4
Applied every
4000 cycles
I
Relative ranking of fatigue life of 7075 and 7050 aluminum alloys under constant
amplitude and periodic single overload conditions.
Source: J. G. Kaufman and J. S. Santner, "Fracture Properties of Aluminum Alloys," in Application of Fracture Mechanics for
Selection of Metallic Structural Materials, James E. Campbell, William W. Gerberich and John H. Underwood, Eds., American
Society for Metals, Metals Park OH, 1982, P 197
357
358
~
160
AIR
NaCl
'0
;f
-0
---
.0... - - 0 -
.; 120
'"
~
a:
I;;
u
::;
80
MODE I LDADING
'"
,I
7
10
AIR
Nael
;f
.;
'"
'"
....a:
'"
u
120
uJ
::;
--_~
80
u
>u
40
0 ....._
10~
MODE J LOADING
........-1.-1.........................._
o
0
--0
........- - ' - - '.....................' - - _........- - ' - - ' - ' "....., ...............
Tests performed on a commercial 7075 alloy in a mode 3 loading condition (torsion) indicated that
the reduction in fatigue resistance associated with cathodic charging was considerably less than it was
under mode I loading (note above charts). Although total immunity to corrosion fatigue was not
observed, the slight reduction in fatigue resistance can be associated with conditions that did produce a
true mode 3 loading condition both on a micro-scale and on a macro-scale.
To summarize the aluminum alloy results, it appears that corrosion reactions liberate hydrogen,
which effectively embrittles the region in the vicinity of a crack tip. The specific details of the embrittlement are not known, but it appears that dislocation transport of the hydrogen is involved. It has been
speculated that hydrogen may collect at the semicoherent precipitate-matrix interface, thus explaining
the reported fracture plane; however, a great deal more research will have to be performed before a
more definitive answer will be available.
Source: D. J. Duquette, "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys," in
Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 356
200
7075 AI T6
0.5 M NoCI
276 MN/m (40 ksil mean stress
RT
- 28
24
N-
E 150
20
<,
'iii
V>
V>
w
cr 100
l-
V>
u
u
u>-
:i
50 -
- 16
""
12
A AIR
B CORROSION
C PRE CORRODED/AIR FATIGUE
D PRE CORRODED/HEAT TREATED /AIR FATIGUE
L.LLL.J...U _ _L
I
I I l l d _ _ .I_-.l...0--4
5
10
106
10
0
7
10
The effects of corrosion and pre-corrosion on the fatigue lives of a 7075T6 alloy. Note that re-solutionizing and re-aging the alloy after precorrosion results in a significant increase in fatigue resistance.
Source: D. J. Duquette, "Fundamentals of Corrosion Fatigue Behavior of Metals and Alloys," in Hydrogen Embrittlement and
Stress Corrosion Cracking, R. Gibala and R. F. Hehernann, Eds., American Society for Metals, Metals Park OH. 1984, P 265
359
2U
.-... ,.
-,
~"
' ~~'
I~
L .. .' ,.
.....
10
13v NaCI
.---- .
N,
It had been previously observed that halide ions are particularly damaging to the fatigue behavior of Al alloys; however, if
the alloy is cathodically charged during stressing, sulfate ions
prove to be equally damaging, particularly at long NJ , At lower
NJ the slight decrease observed in cr solutions appears to be
associated with damage to the passive film, as shown in the
above S-Ndata, In SO~ solutions, a crack must initiate to break
the protective film to allow access to the bulk alloy, Cathodic
charging of the high-purity analog of the 7075 alloy also shows a
reduction in fatigue resistance, In many cases, fatigue crack
initiation in the equiaxed-grain high-purity alloy is intergranular, and at more active cathodic potentials there is a tendency toward a higher percentage of transgranular cracking.
Source: D. J. Duquelle, "Fundamentals of Corrosion Fatigue Behavior of Metals and Alloys," in Hydrogen Embrittlernent and
Stress Corrosion Cracking, R. Gibala and R, F. Hehemann, Eds., American Society for Metals, Metals Park OH, 1984, P 266
361
362
o PAlMlIIC ACIO
" ANODIZED AND WATER SEALED
50
41
40
"j
ij
!:1'"
}I
I;;;
......'------------~
}O
I~
4
10
4
10
NUMBER OF CYCLES
lOB
o PAlMIIIC ACID
" ANODIZED AND WATER SEALED
21
~20
"
ol-
----------
ij
~15
t;;
10
10
I~
104
NUMBER OF CYClES
Tension fatigue test of 7075-T6 aluminum alloy sheet, notch factor K T = 2.37.
363
45
BAREPQISHffi
o ANODIZm AND NOI WAHR SALD
o PALMIIIC ACtO
40
000
101
6
10
NUMBER OFCYCLES
The three charts show the effects of notch designs and surface treatments on fatigue properties
of aluminum alloy 7075-T6 sheet; the table shows the effects of 17 surface treatments.
Surface treatment
No. of cycles
to failure
Polished
.
125,000
Anodized and water-sealcd
125,000
Propionic acid 2,800,000
Vale ric acid 15,000,000
Caproic acid 9,200,000
Octanoic acid 12,300,000
Decanolc acid
.. 7,500,000
Lauric acid 8,600,000
Myristic acid 11,600,000
Surface treatment
No. of cycles
to failure
Source: Irvin R. Kramer. "Improvement of Metal Fatigue Lifebya ChemicalSurface Treatment, "in Fatigue-An Interdisciplinary
Approach, John J. Burke, Norman L. Reed and Volker Weiss, Eds., Syracuse University Press, Syracuse NY, 1964, pp 250,251
364
1 -
10
0.00
.20x
I;
.33e
.50l>
.700
10- 1
f--
f--
a...
~
.80-
10
R
0.00
.20
.33e
.50l>
.700
.80 -
f--
Q)
>.
Q)
U
>.
o
.....
E
E
.....u
10- 2
E
E
f--
~10
<I
I
z
.....
"
C
"
lie:
~
It
10- 3 f--
...
1-
~
~
~
.....
.g
~
o
o
101
10
20
30
10
30
AK(MPavm)
Fatigue crack propagation in aluminum alloy 7075-T6 showing the effect of R ratio
and the applicability of the Forman, Keraney, and Engle relation. The scatter in the
data is much less in the latter.
The above diagrams show that data scatter is much less when the rate daldN is
computed according to the equation due to Foreman et al. R< O. The proposed Foreman equation is:
da
C(t::.K)1I1
dN
(I -R) K c -
Source: Marc Andre Meyers and Krishan Kumar Chawla, "Mechanical Metallurgy: Principles and Applications," Prentice-Hall,
Inc., Englewood Cliffs NJ, 1984, p 716
VACUUM
-2
10
oJ
.9!
10
u
>.
E
E
z
Ci
"tl
"tl
,~
10
12
15
20
11K [ MNm-JI2 )
30
40
50
Cold deformation also increases the fatigue crack propagation rate as shown in the above graph, which compares an undeformed structure with 10% and 20% cold rolled structures.
Source: G. Liitjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys," in Aluminum Transformation Technology and
Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and
D. E. Embury, Eds.. American Society for Metals. Metals Park OH, 1980, P 207
365
366
c:--,---,--,r-r-,.,..-----,---r--____:::>
78910
20
1
(MPa m / 2 )
Comparative fatigue crack growth rates for aluminum alloys 2024-T3 and 7075-T6 in air ofvarying humidity.
Source: I. J. Polmear, Light Alloys. Edward Arnold Ltd, London, England, and American Society for Metals, Metals Park OH,
1981,p79
50
o
30*>
60*
1000
2000
I..LI
0
l:::Ji=? 4
-' - '
<t:a..
::?:::?:
""<t:
~~
2
00
LL.
I..LI::?:
=>""
-'<t:
<t:::c
>0
>2
L58
a.. I..LI
Vl
00
FATIGUE CYCLES
Source: O. Buck and G. A. Alers, "New Techniques for Detection and Monitoring of Fatigue Damage," in Fatigue and Microstructure, American Society for Metals, Metal Park OR, 1979, P 137
367
368
.25" OPEN
HOLE
0_
.75" 0.0.
2
(26 J/cm / SIDE)
LASER-SHOCKED
CONTROL SPECIMENS
SPECIMEN NO.
NO. OF CYCLES
TO FAILURE
51,500
n,loo
2
3
SPECIMEN NO.
NO. OF CYCLES
TO FAILURE
1
2
473,800
520,700
AVG.
SCATTER
497,250
1.1
385,000
AVG.
SCATTER
169,600
7.5
LASER-SHOC KED
CONTROL SPECIMENS
SPECIMEN NO.
1
2
3
AVG.
SCATTER
NO. OF CYCLES
TO FAILURE
41,500
74,300
109,300
75,033
2.63
SPECIMEN NO.
NO. OF CYCLES
TO FAILURE
1
2
171,800
266,200
AVG.
SCATTER
218,950
1.5
The fatigue test specimens were 0.25 inch thick by 1.5inches wide and approximately 9.5 inches long, as shown
in the above sketch. The specimen blanks were laser-shock processed, and then the 0.25-inch-diameter hole was
bored through the center of the laser-shock-processed area. The diameter of the laser-shock-processed area is
three times the fastener hole diameter. All of the specimens had machined surfaces of less than 125RMS. All of
these open-hole specimens were fatigue tested to failure at a maximum net section stress of20,000 psi, and an R=
0.1 under constant-amplitude load control. Three control specimens for each material were tested to establish
the typical fatigue life for the material. Two LSP specimens were tested for each material to establish the degree
of improvement due to the laser-shock processing.
The fatigue test results for the 7075-T6 material are summarized in the upper tabulation. The LSP specimens
showed three times better fatigue lives on the average and much less scatter than the unprocessed material. The
results for the 7475-T73 material are summarized in the lower tabulation; these show the same typically large
increases in fatigue life and reduced scatter. It should be noted that the 7075-T6 material shows better fatigue
resistance than the 7475-T73 material, whether or not it is laser-shock processed. This is largely due to the
differences in dislocation/precipitate interactions that result from the T6 and T73 heat treatments. The dislocations appear to shear through the precipitate particles in the T6 condition. The precipitate particles are apparently so strong in the T73 condition that the dislocations just loop around the particles.
Source: William F. Bates, Jr., "Laser Shock Processing of Aluminum Alloys," in Source Book on Applications of the Laser in
Metalworking, Dr. Edward A. Metzbower, Ed., American Society for Metals, Metals Park OH, 1981,pp 256-258
0-
-0
c:::::=:=========$$:Jr====
.072"
SYD HOLE
LOAD
CYCLES TO
FAILURE
Iff
The full load transfer joint shown in the above sketch was made from
7075-T6 clad aluminum alloy and fatigue tested. The purpose of this test was
to evaluate the fatigue life improvement of laser-shack-processed fastener
holes when the holes are loaded by the Hi-Lokfastener in bearing. A secondary purpose was to find out if the cheaper Hi-Lok fastener system in a
laser-shack-processed hole would show as good a fatigue life as the much
more expensive Taper-Lok fastener system. The above bar chart shows the
test results for three different stress levels. At each stress level, three specimens with standard holes and three specimens with laser-shack-processed
holes were tested. The specimens tested at the l4-ksi stress level showed severe fretting at the intersection of the hole wall with a badly galled area of the
fretted faying surface. All of the fatigue origins occurred at or near the hole
wall corners on the faying surface.
Source: William F. Bates, Jr., "Laser Shock Processing of Aluminum Alloys," in Source Book on Applications of the Laser in
Metalworking, Dr. Edward A. Metzbower, Ed., American Society for Metals, Metals Park OH, 1981, pp 262-263
369
370
400
n
I
:::E
u
If)
If)
300
UJ
a:
Iii
o
Z
200
~
Z
a:
UJ
<
100-'--,----r-----.,-----.-----,-10'
10
10'
10'
10'
CYCLES TO FAILURE
Source: G. LUtjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys," in Aluminum Transformation Technology and
Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and
D. E. Embury, Eds. American Society for Metals, Metals Park OH, 1980, p 193
::r
300
UJ
...::>
::::;
0..
::r
Vl
Vl
200
UJ
a::
Vi
100L~,----------~;----------'-=--------~;-
__
CYCLES TO FAILURE
Source: G. Lutjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys." in Aluminum Transformation Technology and
Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and
D. E. Embury, Eds., American Society for Metals. Metals Park OH. 1980, p 192
371
372
400
"
c,
::E
300
::::i
a,
::E
<I:
lI)
lI)
0:
200
l-
ll)
CYCLES TO FAILURE
The above chart shows the grain-size effect in a stresscontrolled test for a high-purity 7075 alloy (X-7075) aged to contain shearable precipitates. Since the flow stress is determined
by the interaction of dislocations with the coherent precipitates,
the yield stress is approximately the same for both alloys. Opti- .
cal examinations of the specimen surfaces show that cracks nucleate much earlier in specimens with the large grain size. Cracks
nucleated at intense slip bands for both grain sizes.
Source: Edgar A. Starke. Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure." in Fatigue and Microstructure.
American Society for Metals. Metals Park OH, 1979. P 225
-I
10
..
A: 24h 100'C
VACWM
C:48h 180'C
AIR
E
u
-)
10
_,
10
:E
a
'0
-i
10
I06..L..----,--r--..,.....--..--..,.....---.------r----r--..,......
Influence of environment (laboratory air) on fatigue
crack propagation rates for underaged (A) and overaged
(C) condition. X-7075, CT specimens,R = O.I,f= 30 Hz.
Source: G. LOtjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys," in Aluminum Transformation Technology and
Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and
D. E. Embury, Eds., American Society for Metals, Metals Park OH, 1980, P 204
373
374
-2
10
GS 200.lJrn
GS
-J
46.lJrn
10
>-
.!2
-4
E 10
.g
-5
,,/~
3.5NoCI
10
I06.L----.-----.---'-r---..-----r-----.---...----.7
10
15
20
30
40
50
t:. K [MNrn-3I2 )
Influence of environment (3.5% NaCI) on fatigue crack
propagation rates for two different grain sizes. X-7075,
24 h 100 DC, CT specimens,R = 0.1,1= 30 Hz.
Source: G. Lutjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys," in Aluminum Transformation Technology and
Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and
D. E. Embury, Eds., American Society for Metals, Metals Park OH, 1980, P 204
Q.
LLJ
300 r--~"-T'"""...-r'1""""'-r---'----r-"'T"""T"T"l"T~---'----"--"-"""""""'"
co..
'PO
\
~ 250 I---~ID-----':~_----_--t-----____l
:J
Q.
~
~
~o.
....
"
200t------O-~~--""'~-+__---____l
X-7075
150 t----::------:-~-:-----t-----_l
0---0 If = 0,
8h 160C
--If = 0.5,4h 160C
105
CYCLES
Effect of grain-boundary ledges on the stress-life
behavior of an alloy containing nonshearable
precipitates and PFZ.
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure,
American Society for Metals, Metals Park OH, 1979, P 230
375
376
24 h 100C
-1
VACUUM
10
n
~
u
'"
-)
10
E
E
u
-.
10
"0
-5
10
-6
10
40
50
Source: G. Lutjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys," in Aluminum Transformation Technology and
Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and
D. E. Embury, Eds., American Society for Metals, Metals Park OH, 1980, P 207
r------------------------------.
o
SPF Conditions:
= As pr......d
A = SPF - highsup.rpl.stic st"inl
o = SPF - low luporpl.stic st"inl
T= 516C
i=21104S-1
70
-+ = No,.ilu"
R =+0.1
60
AS PROCESSED SN CURVE
MAX
STRESS 50
(Ksi)
40
AOC>AO
-...
6.
...
(]I
_-------0-+0-+
3D
20 I-.....I._..I....I..I-I._..................I-.....I._...."""'"-"u.._"'-..........I."""".........._
1~
,~
10'
10'
10'
....."""'"-"U
10'
CYCLES TO FAILURE
Source: C. Bampton, F. McQuilkin and G. Stacher, "Superplastic Forming Applications to Bomber Aircraft ." in Superplastic
Forming. Suphal P. Agrawal. Ed.. American Society for Metals. Metals Park OH. 1985. p 77
377
378
,.....
0
Q..
-..J
250
- II
UJ
0
:J
-<9=0.5
...J
Q.
o <9=0
200
16h 160C
6h 160C
~
<{
I/l
I/l
UJ
0:
150
II)
7475
o
I
CYCLES
1.
TO FAILURE
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure,
American Society for Metals, Metals Park OH, 1979, P 232
7475 T-6
FINE GRAIN
~ 1 X 10-'
(.)
.....
7075 T-6 -
(J)
::I:
(.)
1 X 10- 5
<,
<
o
1 X 10- 6
1 X 10- 1
1.----I._.L.I.....L..J..J.._...l----JL...-..JL.I...l.----I._.....L....J...J..J
10
100
1000
This comparison of conventional, coarse-grain, nonsuperplastic aluminum alloy 7075 with superplastic alloy 7475 shows
almost an order-of-magnitude reduction in crack growth for the
superplastic material.
Source: C. Bampton, F. McQuilkin and G. Stacher, "Superplastic. Forming Applications to Bomber Aircraft," in Superplastic
Forming, Suphal P. Agrawal, Ed., American Society for Metals, Metal Park OH, 1985, p 77
379
380
400
:g
300
I-
:::;
Q.
<l:
200
~ J~----,--;,------------'--r:---""""'----""'-3
10
10'
10
la'
CYCLES TO FAILURE
Source: G. LUtjering and A. Gysler, "Fatigue and Fracture of Aluminum Alloys," in Aluminum Transformation Technology and
Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and
D. E. Embury, Eds., American Society for Metals, Metals Park OH, 1980, P 195
"t
IJ.
o
400
2.:
SF1-T6/T73
300~-T6
S
~
~200
.~~o
x
In
10 3
10 4
10s
10 6
REVERSALS TO FAILURE (2N,)
Source: G. Williams and K. M. Fisher, "Squeeze Forming of Aluminium-Alloy Components," in Production to Near Net Shape:
Source Book, C. J. Van Tyne and B. Avitzur, Eds., American Society for Metals, Metals Park OH, 1983, p 367
381
382
150
ST and quenched
85
200
310
175
270
lU
a..
(/] 120
(/]
Q)
(/]
Cl
'16
90
~
.~
oho
~ ~....
0_
~
<i:
&0
L60
10&
87
X 72
7
10
48
~
lOll
Number of cycles
Fatigue (SoN) curves for the alloy AI-SMg-O.SAg
in different conditions.
Source: I. J. Polrnear, Light Alloys. Edward Arnold Ltd. London, England, and American Society for Metals, Metals Park OH.
1981, P 42
r----------------...,
a,
" ",
GI
.':::.
.. _- "
........
24h at 150GC
...........
::---'.::: ....
'
...
(\J
-...
e--
...... ,
:.:~~
Q.
......
<.&>
<l
o Failure AI-Zn-Mg
2Nf
The above chart shows Coffin- Manson life plots of two averaged AI-Zn-Mg alloys. The small-grained AI-Zn-Mg-Zr alloy
has a much longer life than does the large-grained AI-Zn-Mg
alloy. The improvement in life is attributed to increasing the
cycles to crack initiation, as indicated in the chart. A convergence is noted for long lives (low plastic-strain amplitudes) for
this strain-controlled test. Since the fine-grained material
hardens more than the other at low strains, the stress to enforce
the applied strain is greater at long lives, and this affects the life
improvement due to the fine grains.
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure,
American Society for Metals, Metals Park OH, 1979. P 228
383
384
4h at 120C
.96h at 150C
10.
c..
wiN
..........
<l
I.
AI-Zn-Mg
0.1
10 1
Since the strain localization occurs in a region free of solute, overaging the matrix precipitates or adding dispersoids does not homogenize
the deformation. This is clearly illustrated by comparing the Coffin
Manson life curves of underaged and overaged specimens of largegrained AI-Zn-Mg alloy (see above chart). The tensile yield strength
and strain to fracture are approximately the same for both specimens.
The underaged alloy has shearable precipitates, which results in strain
localization, the formation of intense slip bands, and early crack nucleation under cyclic loading. Overaging was one method described for
homogenizing deformation; however, this method is not effective for
large-grained material. Preferential deformation in the PFZ also leads
to strain localization and results, for this particular case, in the same
fatigue life. Dispersoids distributed throughout the matrix would not
inhibit strain localization in the PFZ for the same reason.
Source: Edgar A. Starke, Jr., and Gerd Liltjering, "Cyclic Plastic Deformation and Microstructure,"in Fatigue and Microstructure,
American Society for Metals, Metals Park OH, 1979, P 227
70
g
Q.
60
~
-b
.......
0
<,
50
+1
c-
<,
~o
<,
o Al
D
<,
_-- -----
Al (Cu)
<,
0.....
40
10
10
10
10
Source: Marc Andre Meyers and Krishan Kumar Chawla. "Mechanical Metallurgy: Principles and Applications," Prentice-Hall,
Inc .. Englewood Cliffs NJ, 1984, P 707
385
386
400
l>
50
'in
"'"
40
30
..
20
E
OJ
E
'x
::;;;
7091T7E69 (2 lots)
7090T7E71
Open symbols-longitudinal
Solid symbols-long transverse
60
O~;"
300
10
o,
::;;;
~-
200 E
OJ
E
..
'x
::;;;
100
Source: Robert H. Graham, "Wrought Aluminum PIM Alloys," in Powder Metallurgy-Applications, Advantages and Limitations, Erhard Klar, Ed., American Society for Metals, Metals Park OH, 1983, P 240
12-68. P1M Alloys 7090 and 7091 vs 11M 7050 and 7075 Products
60
r-----.-----.----.-------.----.-------.---~-------,400
A 7091-T7E69 (2 lots)
7090-T7E71
Open symbols-longitudinal
Solid symbols-long transverse
o
50
40
30
60
O~"
300 ..
a.
0.013 in.
2:
~~
200
:>
E
:>
..
'x
20
10
2:
Stress ratio R = 0
Ambient air'
100
O'----:----'-,-------'-c-----'-:c-----L,---'--:---'----:----'-,------'
10 7
103
10'
10"
10
10 9
10 2
10 10
Cycles
Source: Robert H. Graham. "Wrought Aluminum PIM Alloys." in Powder Metallurgy-Applications. Advantages and Limitations. Erhard Klar, Ed. American Society for Metals, Metals Park OH, 1983. p 240
387
388
971B-in. R
o
o
Forged
'u;
40
Forged 601AB-T6
?:
-=fL-i-----3-"1!I
201 AB-T6
0.300-in. diam
c.
201AB-T6
202AB_T2~
.~ 20
x
2''"
60 1AB-T2'>.
601A~
_ _..::::===_=====a=_=_=_=_=_-1
OL------''------''-----'----'-----'-----'----'
10'
10'
10
Cycles to failure, N
10'
Source: John D. Generous and Wayne C. Montgomery, "Aluminum PI M-Properties and Applications." in Powder MetallurgyApplications. Advantages and Limitations, Erhard Klar, Ed.. American Society for Metals, Metals Park OH, 1983, p 214
~ 1.9;mm (0.075
E--t--]- E= f-- t
tf--+-1=
7.62 mm . _ _ 8.39 mm
(0.300 in.)
(0.3.30 ln.I
diarn.
diarn.
400
'"
::E
..
300
1;;
::>
-1.0
in.)
60
-,
0..
Ii
70
60 sharp' V
50
40
.'"
~
1;;
::>
'x
'x
'"
::E
'"
::E
Smooth
200
30
a---
100
..
0
10'
-- -
10'
10'
---
10'
20
10
-~
10'
0
10'
Cycles
Rotating-beam fatigue strength for die forgings of P 1M alloy X709I-T7E76 and ingot
metallurgy alloys 7075-T7352 and 2014-T6I. For P/M X709I-T7E76: 0 ,smooth,
transverse direction; ,notched, transverse direction; - denotes test specimen did
not fail in number of cycles indicated.
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society ForMetals, Metals Park OH, 1984,p 469
389
390
t."
L10."" J,~.L
mm
10.,)'
~t~
400
70
in.]
60
=3
K,
ul
50
Ul
e
1il
300
E
E
:>
'x
co
::E
-"
ul
Ul
e
1il
'iii
40
'x
co
m:a
0
200
-.we
7175-T73~ ~
It
30
o ~
10'
r r
103
10'
::E
20
Band for
100 _forgings (six lots)
Stress ratio: R = 0.0
:>
10
o
10'
10'
10'
10'
10'
Cycles
Comparison of axial-stress notch fatigue strength of P 1Malloy X7091-T7E69 die forgings and ingot metallurgy alloy 7175-T736 die forgings. 0 , longitudinal direction,
one lot; , short transverse direction, two lots; - denotes test specimen did not fail in
number of cycles indicated. Stress ratio: R = 0.1.
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals, Metals Park OH, 1984,P 469
'(f3_-------------------"'"
crack orientation L - T
specimens: SEN , GNP, DCB , CT
R 0.0 - 0.1 , 0.1 - 30 Hz
ambient
temperature
~
2cro - T651
2024 - T 3
2024-T~
2048 - T851
2219 - T851
2219 - T87
20
cyclic stress
2618 - T 6
5456 -H117
5456-H321
6061 - T651
7'005 - T63
7039 - T6X31
30
7005 - T63
7050 -T736
7fJ15 - T651
7175 - T736
7475 -T651
40
7050 - T 736
7075 - T 6
7079-T651
7106 - T 63
7175 - T 736
7475 - T 651
50
60
Source: Markus O. Speidel, "Aluminum as a Corrosion Resistant Material," in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and D. E.
Embury, Eds., American Society for Metals, Metals Park OH, 1980, P 615
391
392
7475 - T651
7005 -T63
2618 - T6
7175 - T 736
7075 - T 651
2219 - T 87
5456 - H 321
,----,
~
~
~
10- 5
'------'
~~
6061 -T651
2048 - T851
-6
10
7050 - T736
5456 - H 117
10- 7
Q)
::J
.21
....
....ro
crack orientation L - T
specimens: CNP I DCB I CT
environment: salt water
ambient temperature
R 0,0-0,1 I 0,1 - 30 Hz
-10
10 0,......--+---+----+----+----+---.....
10
20
30
40
50
60
cyclic stress intensity range,
~K , [MN. m-~]
Source: Markus O. Speidel, "Aluminum as a Corrosion Resistant Material, "in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn, Chubut, Argentina), C. A. Pampillo, H. Biloni and D. E.
Embury, Eds., American Society for Metals, Metals Park OH, 1980, p 614
500
400
300
250
'"
~
200
<;;
150
rf.
t:
en
~
~
201---"':
18t----
161----+~
141----+----,.:-'
~
iU
":;'l
1.
II)
100
121----'--+---'--~
75
8!---'---'-f--'-----if--"""'""-,..-t----'
71-----+--.-:.+------1------+--,,-:~
50
61-----+---+--...--:'":-...;,......1----'---+..,.....-..:.,......+---~
106
lOS
109
Source: Robert C. Juvinall, Fundamentals of Machine Component Design, John Wiley & Sons, New York NY, 1983, p 207
393
394
26
...
24
... ...
, , 1\
,~
"
4
16, [,>..
22
,,1,- 3,
... ...
5)
~ 1-1',
S-N CURVES
CASTING ALLOY AL - 195
DEGREE
FATIGUE* DEGREE
FATIGUE*
POROSITY STRENGTH POROSITY erRENGTH
- - 5 - - 9,000
;t
2
10,500
6
8,150
rr;ooo
1'>,
,""1\
i"
,~
I,
" ...
x 18
"
f' ~"
41- I - -
- .
' ... ,
8./
...
...
~I'
....
"of-.
f'
'"
I"'-\l:i:
I'
"o
.......
8
} DEGREE
:-:-- POROSITY
10 Q
,,5
10
. I-
",I'. ,
1'1
10
"I",
I""' l ......
"o
12
A~~~
-- f - -
.~
....
...
-- -
,~
r-,
_~
"'I'
1,000
.... REPRESENTATION OF
NO FAILURE
'I\.
<,
,'\..
~ .......
9,000
9,000
\,
,1"\
20
3
4
1)
IV
,,"'.
..
I-
...
r-,
r>-
.. .. .. r-...... .....-r--
10,,6
101
Source: N. E. Promisel, "Evaluation of Non-ferrous Materials," in Materials Evaluation in Relation to Component Behavior
(Proceedings of the Third Sagamore Ordnance Materials Research Conference), Syracuse University Research Institute, Syracuse
NY, 1956, P 65
~300
w
o
=>
l-
ii 200
2
(/)
~ 100
a::
In
O'-:--_--L
2
10
1~
--'4
'-::-_ _--'-
1~
--'---'
1~
REVERSALS TO FAILURE(2N,)
S-N curves for aluminum casting alloy LM25-T6; chill
cast versus squeeze formed.
Source: G. Williams and K. M. Fisher, "Squeeze Forming of Aluminium-Alloy Components," in Production to Near Net Shape:
Source Book, C. J. Van Tyne and B. Avitzur, Eds., American Society for Metals, Metals Park OH, 1983, p 367
395
396
<,
s:
u
c
...5.4
Vi
Q;
~5.2
'"
c
--
Ol
C
<0
a:
"E
4.6
OJ
(J)
~ -,
~
OJ
---
l'
'~
I--
::f 5.0
e
Vi
'0 4 .8
~"
<,
......
Q(~I'("
1110
~fiedAir ....
_______ T-D
r-...
I'
~
----
If) 0
.... PlJr i f iltr)fJ
I---
II
10
- --
~-- 1---
'ea"
--I" "~~.
InAir"_
4.4 6
10
.... iA
\.\~I
\
---
~
5.10
The effect of air and water vapor on the fatigue life of annealed copper.
Source: D. J. Duquette. "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys," in
Fatigue and Microstructure, American Society for Metals, Metals Park OH, 1979, P 336
Cu
prestrain 20"1.
annealed
"prestrain 40/.
_-
A-
Applied plastic-strain amplitude versus fatiguelife curves for copper at long life.
Helgeland was the first to observe and claim a fatigue limit for
copper (actually the plateau stress, although it was not recognized as such at the time). Unfortunately, his results were apparently contradicted by those of Kettunen, who observed failures
at stresses down to 17.7 MPa. This difficulty was resolved by
Laird, who showed that Lukas and Klesnil's long-life CoffinManson plots showed failures to occur only down to the plasticstrain fatigue limit; at lower strains, no failures were observed in
the testing time available (see above chart). However, Lukas et
al. also carried out stress-cycling tests, in which they monitored
the plastic strain. Specimens that had been stress-cycled yielded
a plot of saturation plastic-strain amplitude versus life, where
failures occurred at strains as low as 10-5 . The difference between these tests is that in strain cycling, the stress is low in the
initial cycles and increases to saturation, whereas in stress
cycling, full application of the load in the first cycle causes a
large strain in a soft material. This initial large strain creates the
PSB cell structures, which would not otherwise form in a
constant-strain test. Since Kettunen applied the full load to his
specimens, failures were observed at stresses below that of the
plateau. Helgeland, on the other hand, although he was stress
cycling, imposed a low stress at the start of his tests and increased it gradually to the chosen value.
Source: Campbell Laird. "Mechanisms and Theories of Fatigue," in Fatigue and Microstructure, American Society for Metals.
Metals Park OH, 1979, P 195
397
398
'"
Cl.
::E
22 ]1
150
s:."
...en
s:."
'C,
18
1;;
Q)
::J
::J
en
en
'+>
'"
u.
~
'+>
100
14
'"
u.
10
Stress cycles
Source: Metals Handbook, 9th Edition. Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society
for Metals, Metals Park OH, 1979, P 289
.
50
'"
<,
a.
0
0
0
40
I'--~
....0
::l
><
II!,,,)"
+~
20
.........
r-,
Q)
l J)
"0
<,
........
<fl<fl
.... 30
<,
I'
~~~
...e
~ a
t;.1--
~?"
I"'--.:!""...
1'; I-
Q)
--
f_
:: ..
r-rx_
IJ..
-i
-+
10
10
107
Cycles for Failure ,N
Legend
o Group G (60% Cold
I:>. Group D (40% Cold
o Group A (20% Cold
Anneal I Grain Size
... Anneal 3 Grain Size
x Anneal 4 Grain Size
+ Anneal 5 Grain Size
Drown)
Drawn)
Drawn)
0.0/2 mm.
0.026 mm.
0.051 mm.
0.131 mm.
Source: George M. Sinclair, "Some Metallurgical Aspects of Fatigue.vin Fatigue-An Interdisciplinary Approach, JohnJ. Burke,
Norman L. Reed and Volker Weiss, Eds., Syracuse University Press, Syracuse NY, 1964, P 69
399
400
25
160 f---~~~-+-------+------+-------+--------l
rf
140 f------~
20 ]
:::E
~
~
.J::.'
1;;
1201--------+------'
1;;
Ql
:::l
Ql
'"
:::l
.~
'"
.~
u.
100 f - - - - - - - + - - - - - - - + -
------+-------1
80 f-------+--------+-------I-----=
0-
15
10
Stress cycles
S-N curves (scatter band) for copper alloy C83600 (leaded red brass).
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society
for Metals, Metals Park OH, 1979, P 406
u.
If
300
:E
~c
e
t:
200
Cll
;:)
.~
'"
u.
100
- 50
<ttIOo. 0
- 40
~~
..................
--
.r."
",
30
~~
n
t:
Cll
- 20 .;::;c:
u.
- 10 '"
;:)
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, ArnericanSociety
for Metals, Metals Park OH, 1979, p 35
401
402
300
40
D-
.,;
Ii
1;;
Cl
..
c:
';;
...
Ql
<{
';;;
~
250
30
200
.....~
Cl
';;
E
Ql
150
100
104
20
105
106
107
108
109
Stress cycles
S-N curves (scatter band) for copper alloys C87500 and
C87800 (silicon brasses) tested at room temperature.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society
for Metals, Metals Park OH, 1979, P 416
'"
200
0-
25
::i:
Ii 150
20
'"
Ci:
...~..
Cl
Cl
...'"E
'iii
.>t
15
100
50
104
105
106
107
108
sc
.....
'"
10 Ci:'"
109
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society
for Metals, Metals Park OH, 1979, p 421
403
404
25
0~
't!'
160
140
20
Cl
.g
E
l!l
<i:
120
~_----=l15
100
80
104
105
Number of stress cycles
S-N curves (scatter band) for copper alloy C93700 (highleaded tin bronze) tested at room temperature.
Source: Metals Handbook. 9th Edition. Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society
for Metals, Metals Park OH. 1979, P 426
405
5/16-io.
5/16-io.
3/8-io.
I----+I---J----o
0
-r----
-~
I
:0
0
0
0
0
10
10
'Q,
a
0
0
..!
uc-,
.....
M
0
Results of fatigue tests on copper alloy tubes before and after salt spray exposure.
The tubes made from the copper alloy failed in the range of
105 to 106 cycles. After exposure for 180 days to salt spray, the
fatigue performance level was not lowered (see plot above).
Brazed steel tubes, prior to salt exposure, failed in the same
test in the range of 105 to 107 cycles. After 30-days exposure to
salt spray, the resistance to fatigue was 105 to 106 cycles. After
90-days exposure, the steel tubes showed no fatigue strength in
this particular test.
Source: Donald K. Miner, "An Effective Solution to the Problem of Hydraulic Brake Line Corrosion," in Source Book on Copper
and Copper Alloys, American Society for Metals. Metals Park OH, 1979, p 356
406
1----+----+----+-----t----2F------i
150
120
UTS: 101.5 ksi
~
ili
ili
90
iii
.....
""
60
30 I-:JIC---I
..,.<........,h~---fI+_-I+_---1t+_-H__-I::l=-_H-+++-_I_If--I
.3
G _----'
30 '--_
30
---'-
-'-
60
90
-L.
120
--'-_ _- - - '
150
180
Two variable-speed, flat-plate testing machines of the fixeddeflection type were used for the test work. These machines have
a speed range of 750-2000 cpm with a maximum deflection of I
in. The yield, ultimate tensile strength and 1.5times the ultimate
are plotted in the graph above. The fatigue limit at zero mean
stress was determined and was found to be 22.0 ksi.
Source: J. M. Cieslewicz. "A Modified Goodman Diagram to Predict the Fatigue Limits of Copper Alloy 955." in Source Book on
Copper and Copper Alloys. American Society for Metals. Metals Park OH, 1979, p 40
If.
:2
125
+1
100
t:
75
E
::J
E
50
..
'x
:2
-.
\.
20
<,
/Unnotched
~
+1
"",-...............
Unolched
/(K t ~ 2)
~.
E
E
::J
Vnotched
/IKt = 31
'x
~
25
Cycles of stress
150
If.
125
:2
+1
~.
100
75
E
::J
'x
~
50
- 20
<,
<;
<,......
/20 oC (68FI- 16 ~
.,
i!
200C 1392F)_ 12 :;;
E
~25~ c (482 of)
::J
E
'x
~
25
Cycles of stress
S-N curves for magnesium alloy sand castings, QE22AT6, showing effects of type of notches (upper graph) and
testing temperature (lower graph). Rotating-beam (Wohler) tests. Machine speed was 2960 Hz.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society
for Metals, Metals Park OH, 1979, P 589
407
408
175
150
rr.:::;;
125
...
100
1;;
E
::l
E
'K
:::;;
20 c (68 IF)
-, r-,
20
<,
75
I'-...
50
]
15
--
Unolched K t=2110
...
E
::l
E
'K
:::;;
25
106
Cycles of stress
175
25
Unnolch.d
150
rr.:::;;
+~ 100
~
E
20
125
75
'I\.
"'---
-,
I"--.
10
01
g
E
::l
E
'K
:::;;
50
25
Cycles of stress
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society
for Metals, Metals Park OH, 1979, P 590
409
Meanstress,ksi
20
20
20
.16
.16
~
Ii
:t12
-l---I----l
e
~
."
!i
4-----11-----l
t---+--+-----j---t---+--....=j4
100
!II
ell
f---t---+---+----'=-+--+--, '4
125
125
Mean stress,MPa
M&an stress,MPa
Meanstress,ksl
10
15
20
.150,------,,--,-----,,---y-,-----...---,---r-.
20
10
'150 r ---,----i---.---'T...---.:n--r--T--.
O
20
20
Cast plates
'1251---+---+-- 107 cycles. loaded in bending
.16
"6
Xi
~-
.12
tl2
---f----f----;
--+--1---;
I---t---+----f---""'---+-...:....-=J '4
!II
1-----J-=...."""-+-----1F"'-...;...f---+--..., 4
125
Mean stress,MPa
Mean stress,MPa
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society
for Metals, Metals Park OH, 1979, P 532
'"
.~
.
~
410
0.8
0.7
~r\
';:;
a::'"
0.6
0.5
0.4
100
200
300
400
500
600
Temperature,OC
Ratio of the fatigue limit of molybdenum at various
temperatures to its tensile strength at the same temperature.
Source: Metals Handbook, 9th Edition, Volume 2, Properties and Selection: Nonferrous Alloys and Pure Metals, American Society
for Metals, Metals Park OH, 1979, P 774
4000
I'~
......
~~
A
'iii
3000
--....r- ....
a.
ui
'"
~
in
2000
.......
:--.... ........
~ :--....
~
~
"'r-..... ~
-- r- --...........,
I~
<,
4 6 8
......
f4.. ......
. . 'N
.........
~~
1000
.....
102
Cycles to failure
..... ...........
4 6 8
10 3
':).04
0.2
The fatigue strength of soldered joints is a complex and difficult subject to examine. Because solder alloys are strain-rate
sensitive and have large elongation capabilities, the performance of fatigue tests under constant stress causes progressive
and rapid relaxation of the joint, and conversely, tests under
constant strain do not reflect a practical application situation.
The influence of the rate of stress cycling in terms of rate of
straining on the fatigue life of copper soldered joints with
60%Sn-40%Pb alloy is presented in the above graph.
Source: Metals Handbook, 9th Edition. Volume 6, Welding, Brazing, and Soldering, American Society for Metals, Metals Park
OH, 1983, P 1095
411
412
0.030
0.040
o Bimetal
I
Trimetal
~
OJ
:::l
......
..
200
Cl
',J
OJ
.~
Q;
100
a:
0.25
0.50
0.75
1.00
Babbitt thickness, mm
Source: Metals Handbook, 9th Edition, Volume 3. Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Metals, American Society [or Metals, Metals Park OH, 1980, P 806
300
en
.;;
200
300
r---,c-r.----~-----,
200 I------'\-t---+------t
.:?
'"
>
.;;
'" 100
Gi
I----~r_--t-----f
II:
OL-_ _L--_ _L - - _ - - - - l
200
100
150
50
Bearing temperature,
The alloy lining was 0.05 to 0.13 mm (0.002 to 0.005 in.) thick,
on steel backing. Bearing load: 14 MPa (2000 psi). As indicated,
operating temperature markedly influences fatigue life.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Metals, American Society for Metals, Metals Park OH, 1980, P 813
413
414
"'
400
a..
:2
'" 300
~'"
'"
E
6,
- 70
r-;
.~~
- 60
Cold rolled
.-. .... -
Annealed
. 200
)(
"'
:2
100
.;;;
50
.><
::i
40 ~
'"
E
J
30 E
'x
20 :2"'
10
SON curve for unalloyed grade 3 titanium. Data were obtained by rotating-beam testing at room temperature.
Source: Metals Handbook, 9th Edition. Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Metals, American Society for Metals, Metals Park OH, 1980, P 376
'"
Q.
400
'"
'"
E
::>
.Sx
'"
300
-------
- 80
_40C (-40 F)
- 60
20C (6SoF)
- 50
.;;;
..><
:i
E
::>
I
315C (600 of)
- 40 .Sx
30
200
100
70
'"
20
104
Number of stress cycles
S-N curves for unalloyed titanium, grade 4, at subzero, room, and elevated temperatures. Data were obtained by rotating-beam testing of unnotched, polished specimens
machined from annealed bar stock.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Metals, American Society for Metals, Metals Park OH. 1980.0378
415
416
STRESS
CI
a..
::lE
CONTROLLED
600
ILl
0
::>
I-
:J
a..
Ti-24%V, A.Q.
::lE
<l:
400
lI)
lI)
ILl
It:
l-
ll)
200
10
CYCLES
TO
FAILURE
Stress-life curves of two Ti-V alloys that undergo cyclic hardening (Ti- 24%V) and
cyclic softening (Ti-32%V).
Cyclic-response curves indicate that the Ti-24%V alloy undergoes extensive cyclic
hardening, whereas Ti-32%V undergoes cyclic softening, as indicated above. Hardening is caused by incomplete reversibility of twinning.
Source: Edgar A. Starke, Jr., and Gerd Lutjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure,
American Society for Metals, Metals Park OH. 1979, p 236
...
800
a::2: 600
~
~l>
~o
t: 400
rt:-
:I
e~
60 t:
Ground
Hand finished
.>I.
80
100 .;;;
Shot peened
40
.
200
)(
';(
...
20 ~
::2:
8
10
Lifetime, stress cycles
...
a::2:
l:/"
e
t:
800
Ultrasonic machined
100
Slab milled
600
80
Chem milled and annealed
Chem mille
400
E
E 200
';(
60
40
:I
::2:
0
104
...
105
..
106
107
800
100 ]
....... 600
t: 400
E
:I
...
)(
::2:
20 ::2:...
0
108
a::2:
E
:I
E
';(
...
,:;
...
......
~
.>I.
200
~are
- 80
notch (K r = 2.4)
~~
l>Al>A
V-notch (K r = 3.2)
'I
16
~
~l>l>
60
40
E
E
:I
'x
- 20 ::2:...
108
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Melals, American Society for Metals, Metals Park OH, 1980, p 382
417
418
10
10-4
Ti-6AI-4V
NI
ELI
24 to -269C
(75 to -452 of)
b 10-41-----+--~~--I~----4-----=1
Cl
OJ
.,
:J
.Cl
'"
u.
.,;m
Data on fatigue crack growth rates for Ti-5AI-2.5Sn and Ti6AI-4V alloys are plotted above. These data indicate that the
exposure temperature has no effect on the fatigue crack growth
rates for Ti-5AI-2.5Sn and Ti-6AI-4V(NI). However, over part
of the 11K range, the fatigue crack growth rates for Ti-6Al4V(ELI) are higher at cryogenic temperatures than at room
temperature at the same 11Kvalues.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Metals, American Society for Metals, Metals Park OH, 1980, P 765
SURFACE
GRIND + PEEN
Ah ...
b5
43
44
70
lO
20
END MILL
PERIPHERAL CUT
Gentle
Abusive
150
J67
67
14
14
73
28
39
12
II
145
I 45
I",
Off- Sta nda rd
147
'".
85
I
20
48
120
~5
iOff-S'.n""."
43
55
)83
,v.
Gentle
Abusive
ECM +
PEEN
Gentle
HAND
GRIND
ECM
Gentle
Cony.
Abusive
SURFACE
GRIND
40
60
80
Bar chart presentation showing effects of various machining and grinding operations
on fatigue characteristics of titanium alloy Ti-6AI-6V-2Sn.
Source: Norman Zlatin and Michael Field, "Procedures and Precautions in Machining Titanium Alloys," in Titanium and Titanium
Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982, P 355
419
420
iii 120
ANNEALED PLATE
(MINI
"
vi
en
w
a:
l-
ll)
10('
:2
:>
:2
X
<{
:2
80
60 -'--
---,
"T"
"T'"
--,
10
NUM8ER OF CYCLES
OJ
S-N curves showing endurance limits for titanium alloy powder consolidated by HIP at
900C (1650 OF).
Note that most data points obtained in this phase fell within the representative data
band for annealed forgings. In the specimen designated with an "M"subscript, low fatigue
endurance was apparently associated with failure initiation at an inclusion, This shows
that a clean powder is required for parts that are fatigue-critical and must operate with the
equivalent of fully forged properties.
Source: R. H. Witt and W. T. Highberger, "Experience With Net-Shape Processes for Titanium Alloys," in Production to Near Net
Shape: Source Book. C. J. Van Tyne and B. Avitzur, Eds., American Society for Metals. Metals Park OH. 1982. P 277
ANN PLATf'
120
tOO
tgTReeS,
KSI
80
JO
<,
22-5'0~-fOl<'Sl-I/-IR
'7------:
40
20
~4
~~
~~
~7
CVClP8 TO {"AILURf"
S-N curves showing that HIP Ti-6AI-6V-2Sn is equivalent to annealed plate of
the same composition.
Source: W. Theodore Highberger. "Manufacture of Titanium Components by Hot Isostatic Pressing," in Production to Near Net
Shape: Source Book. C. J. Van Tyne and B. Avitzur, Eds.. American Society for Metals. Metals Park OH, 1983, P 304
421
422
SU RFACE
GRIND
END MILL-
39
41
-, 68
I-l'G~e.:.:n~tl~e_....,.."....
Cony
I 17
~IO
lZO
I-l'G~en~ltI.:i-e
"T""
'"
8Z
36
77
...1 47
I
ZO
40
I
60
80
Bar chart presentation showing the effects of specific machining and grinding operations on fatigue characteristics of titanium alloy Ti-6AI-2Sn-4Zr-2Mo.
Source: Norman Zlatin and Michael Field, "Procedures and Precautions in Machining Titanium Alloys,"in Titanium and Titanium
Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982, P 355
-50
+50
100
150
1200
150
1000
.;;;
800
If.
.><
~
t:
E
:>
E
:;;
10D ~
t:
600
E
E
:>
'x
'x
~ 400
~
50
10 7 cvcles lifetime
200
0
-800
-600
--400
-200
+200
400
600
800
1000
0
1200
Minimumstress, MPa
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Metals, American Society for Metals, Metals Park OH, 1980, P 385
423
424
1600
'"
Q.
:2
220
---
1400
STA
200
~.
'"
E
::J
E
1200
'"
:2
1000
'x
180
160
140
800
1
103
'"
.><
~.
E
::J
E
'x
'"
:2
120
105
Number of cycles
Low-cycle axial fatigue curves for Ti-6AI-2Sn-4Zr-6Mo. STA
(solution treated and aged) condition: 1 hat 870 C (1600 OF),
water quench, age 8 h at 595C (1100 OF) and air cool. DA
(duplex annealed) condition: 15 min at 870C, air cool, then 8h
at 540 C (1000 F) and air cool. All fatigue tests conducted at a
stress ratio of R = 0.1. Open symbols indicate fatigue tests; solid
symbols, tension tests.
Source: Metals Handbook, 9th Edition, Volume 3. Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Metals. American Society for Metals, Metals Park OH, 1980, P 395
500
'"
e,
:::E
400
'"
'"
~
'" 300
E
:>
E
'x
:::E'"
200
80
Tension-tension tests
, R = 0,25; K r = 3.5
'.
'in
60
'"
'"
I'----
.~
- 40
E
:>
E
'x
,....
'"
:::E
- 20
100
Stress cycles
Source: Metals Handbook. 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Metals, American Society for Metals, Metals Park OH, 1980, P 403
425
426
1000 ~o
'" 900
Q.
:!:
~
...e
800
'"
E
:::>
700
)(
'" 600
:!:
.---- --
Unnotched
lili
140
r---
'u;
.:,t.
120
A
100
<,
",'
'"
~
'"
E
:::>
E
'x
~ to-..
500
'"
:!:
----
80
I--
Stress cycles
500
'"
Q.
400
\~
:!:
o RT
~2050C (400F)
'" 100
:!:
o 425 c (800 of)
40
~
'u;
.:,t.
r-,
E 200
:::>
E
60
Kr = 3.0
\0
'"
'" 300
'x
Notched
'"
'"
E
:::>
E
20 'x
'"
:!:
I
Stress cycles
Source: Metals Handbook, 9th Edition.Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Metals, American Society for Metals, Metals Park OH, 1980, P 399
10-4-r-----r---...."7T""-------,
(25)
MA Ti'-6A/-4V
iI
i
i
10-5
(025)
Ti"-IOV2Fe-3A/
70.6
(0025)/i'...
O--.....L.L---
20
(II)
LlK, ksi
(22)
vm.
J,
...J
O'---4..J.O - - - - - - - - '
(33) (44)
(MPoW;;)
Comparison of fatigue crack growth rates. Data for Ti10V-2Fe-3AI,R = 0.05,F= 1 to 30 Hz; for MA Ti-6AI4V,R= 0.08,F= 1 to 25 Hz; for RA Ti-6AI-4V,R= 0.08,
F= 6Hz.
Fatigue crack growth rates in air have been found to lie in the
scatter band for mill annealed (MA) Ti-6AI-4V, as shown
above. At high ~Kvalues, Ti-IOV-2Fe-3Al approaches the performance of Ti-6AI-4V in the recrystallized annealed (RA)
condition.
Source: Wayne A. Reinsch and Harry W. Rosenberg, "Three Recent Developments in Titanium Alloys,"in Titanium and Titanium
Alloys: Source Book, Matthew Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982, P 375
427
428
17-15. Ti-1 OV-2Fe-3AI: S-N Curve; Notched Bar Fatigue Life for a
Series of Forgings Compared With Ti-6AI-4V Plate
rrmox,IOJpsi(MPo)
70-r-----~----------____..,
(48J)
60
(414)
50
(345)
40
(216)
30
(201)/i.l-O-4----------'-----"':::OO"------.J
Cycles tofaIlure
Comparison of notched fatigue lives for Ti-lOV-2Fe-3AI forgings and Ti-6AI-4V plate. Data for Ti-lOV-2Fe-3AI,R = O.OS,F
=KT= 2.9; for STA Ti-6AI-4V plate,R = O.l,K T= 3.
Fatigue characteristics of Ti-lOV-2Fe-3Al are equal to or superior to those of Ti-6AI-4V. Notched fatigue results are shown
above. Data from a series of die forgings have shown that the
mean value for fracture toughness is 49.1 ksi ~. (54 MPa
Jill), with a standard deviation of 2.3 ksi~. (2.5 MPa Jill).
K/scc in 3.5% NaCl is typically about 90% of the x;
Source: Wayne A. Reinsch and Harry W. Rosenberg, "Three Recent Developments in Titanium Alloys,"in Titanium and Titanium
Alloys: Source Book, Matthew Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982, P 375
429
-100
-80
-60
-40
-20
+20
40
60
80
100
120
140
160
180
200
1400 r-_,_----,~-_,_-_,___,...---.-_r-_r""__,-__,-~""__.--t-er____,"'_,,,._-r____,, 200
180
160
140
120
100
"~" 600
~
80
~~
:;;
E
"K
At room
temperature
400
60
40
200
- - Unnotched
-600
-400
-200
+200
400
600
Edgenotched, Kr = 3.0
800
1000
20
1200
Minimum stress,ksi
Minimum stress,ksl
1400
-100
-80
-60
-40
-20
+20
40
60
80
100
120
140
160
180
160
140
~
0;;;
:;;
800
120 ""
~~
~E
~
"K
100
600
80
"E
o~
:;;
60
400
Al315
-c (600 OF)
40
200
- - Unnotched
- - Edgenotched, Kr=3.0
-400
-200
+200
400
600
800
1000
20
1200
Constant-life fatigue diagrams for Ti-13V-llCr-3AI, STA (solution treated and aged) condition, longitudinal orientation. Data arefor axial fatigue of edge-polished sheet specimens of material solution treated and
aged to room-temperature tensile strength of 1203 MPa (174.5 ksi), Corresponding yield strength was 1080
MPa (156.7 ksi); at 315C (600 OF), the tensile strength was 1078 MPa (156.3 ksi) and the yield strength was
876 MPa (127.0 ksi), Tests were conducted at a speed of 60 Hz.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels. Tool Materials and Special-Purpose
Metals, American Society for Metals, Metals Park OH, 1980, P 401
430
e,
::E 600
::i
~on 400
E
:::l
E
'j( 200
co
.a...a..
~ Smooth
--...:
--
100 ';;;
80
60
on
40
::E
106
Number of cycles
S-N curves for titanium alloy Ti-6AI-4V (rotating beam) showing effects
of STA (solution treated and aged) versus annealed conditions, and effect of notches.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Metals, American Society for Metals, Metals Park OH, 1980, p 389
700
650
m
a..
~
CIl
CIl
Q)
;:,
CIl
0)
c:
550
.~
'"c:
L.
2:! 500
~
450
106
Endurance (cycles)
Rotating-cantilever fatigue (S-N) curves for three testing directions in
57 mm thick, forged and annealed Ti-6AI-4V bar.
These curves show that fatigue properties are lowest in the long
transverse direction. This result has been attributed to the fact that
Poisson's ratios are also sensitive to crystal orientation, these ratios
being higher in the longitudinal and short transverse directions because stressing occurs parallel to the basal planes. Higher ratios imply
greater constraint, which means that the levels of strain will be reduced
and the fatigue strength enhanced in these two directions. The differences observed in fatigue strengths in the longitudinal and short transverse directions have been attributed to relative changes in grain
shapes that also occur during processing.
Source: I. J. Polmear, Light Alloys. Edward Arnold Ltd, London, England, and American Society for Metals. Metals Park OH,
1981, p 193
431
432
80
o!';.
iii
'" 60
.,;
w
'"
....a:
'"
::;
:::l
~
x
::;
40
20
6.
ISR TEE
EXTRUSION
OL-_ _'--_---JL...--l----JL.......I_ _---L_ _---L_-L.........-..L _ _......._ _......._ ...L.............L _ _-'-_ _...... _'--..L-J
103
105
CYCLES TO FAILURE
S-Nfatigue data for isothermally rolled tees versus extruded material. The notched fatigue behavior of the
ISR tees is as good as or slightly better than that of the extrusion.
Source: W. T. (Ted) Highberger, Govind R. Chanani and Gregory V. Scarich, "Advanced Titanium Metallic Materials and Processes for Application to Naval Aircraft Structures," in Production to Near Net Shape: Source Book, C. J. Van Tyne and B. Avitzur,
Eds., American Society for Metals, Metals Park OH, 1983, P 124
I 1111
co
o,
~
w'"".""~
grade Ti-6AI-4V
~
1000
I
~
"""
~
~~ ~ ~ ~ ~
'"
<I>
U;
~
125
800
~~
~
~ ~~
~ ~"
E
:J
"~
X
co
150
~~ ~
600
I I
400
10'
o Witness blocks
1lJ1I Radius failures
~~
~ ~i'-
~ ~ t:S ~"
100
10'
E
:J
E
"xco
:\I'ci
~ :\"
~~
~ ~t'- ~ ~ ~ ~"
105
10'
'"
<I>
1il
75
10'
Cycles to failure
900
co 700
n,
~
e'"
<I>
1il 500
E
I. I I I
I
o Engme mount supports
Wrought
standard
grade
Ti-6AI-4V
:J
"xco
~ 300
100
10'
~ ~~
~ ~"
~
C'\
125
~
~ ~ ~ -<: ~
100
e'"
<I>
75
~ r-;~ -:~" ~ ~ ~~
~ C'\~~ ~
~ ~ ~ ~\ ~ ~
~ ~ ~ ~~ ~ ~~
r~ ~ 88 ~ ~~
':':";lo,;
10'
"iii
-""
105
10'
1il
E
:J
50
~ r-,~ ~"
'x
co
~
25
10'
Cycles to failure
S-N curves for titanium alloy engine mount supports. Top: Data are for the standard wrought
grade;R = O.I,K,= 1.0, load controlled smooth specimens. Bottom: Data are for isostatically
pressed alloy powder, notched specimens;R = 0.1, K,= 3.
Source: Metals Handbook. 9th Edition, Volume 7. Powder Metallurgy, American Society for Metals. Metals Park OH. 1984.P 654
433
434
~
z
~
<Jl
<Jl
560
a:
In 420
,om,,,":,,,,
R.T
200C
o 340C
z
~280
a:
w
~ 140
<l:
1LO"':'~
---..L.::~-----~------.L:;------'
1(f
CYCLES
10~
TO
10~
FAILURE
Effect of simultaneous fretting and fatigue ofshot-peened Ti-6AI-4V at room temperature, 200, and 340 C at a mean tensile stress of 140 MNm2
Room temperature fretting was found to have little effect on the fatigue strength
at 107 cycles. Fretting at 200 C lowered the fatigue strength by approximately IS%;
furthermore, the fretting fatigue life in the overstress region (70 MNm- 2 above the
nonfretted run out stress) was lowered by two orders of magnitude compared with
results in the absence of fretting. At 340 C, fretting similarly reduced specimen life
7
at overstress levels; however, more importantly, the fretting fatigue strength at 10
cycles was reduced to approximately 40% of that found under room temperature
conditions.
The gross result of fretting normally is fatigue failure brought about by surface
damage in conjunction with normal or transient high stresses in a component. It
should be said at the outset that visual assessment of fretting mildness or severity is
inconclusive by itself, in that the presence of more or less fretting debris on a
microscopic examination is not necessarily relevant to the loss of surface integrity.
It may in fact be misleading and should not be relied upon for assessing the severity
offatigue life degradation. It is the stress state acting in concert with stress raisers
(e.g., pits, tears, cracks) which determines the actual fatigue propensity.
az R c
U;
80
:<:
en
lol
'"
0:
f.<
'"
60
~
~I
-..;
I I I III
f.<
...l
"-
..;
zo
10 5
III
""'-- I
\J
Z
E=:
..;
Z
0:
lol
II
ENDUR.
LIMIT
SURF.
FINISH
GENTLE MILL
GENTLE GRIND
66
6l
41
35
CHEM. MILL
51
zo
ABUSIVE MILL
az
59
ABUSIVE GRIND
13
65
CONDITION
r-, .....
----
10,6
I<SI -;v;:-
I
I
II
107
CYCLES TO FAILURE
S-N curves for beta-rolled titanium alloy Ti-6AI-4V. Curves show the effects of the various
finishing operations on fatigue.
Source: Norman Zlatin and Michael Field, "Procedures and Precautions in Machining Titanium Alloys," in Titanium and Titanium
Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982,p 354
435
436
-..
tJ
a..
:::E
650
0"0.2
(J"02
ILl
0
::>
I...J
600
Q..
:::E
550
UI
UI
ILl
a:
I-
500
UI
450 4
10
CYCLES
TO FAILURE
In order to establish microstructural effects on fatigue behavior, comparisons should be made on materials having the same
yield stress, especially for stress-controlled tests. This is illustrated above where it is shown that two titanium samples having
different yield strengths have different stress-life behavior when
tested at 500C (930 OF).
Source: Edgar A. Starke, Jr., and Gerd LOtjering, "Cyclic Plastic Deformation and Microstructure," in Fatigue and Microstructure,
American Society For Metals, Metals Park OH, 1979, P 237
II II
0-AS RECEIVED I
e- H T 400"r. 2 HRS.
.~
"'" 40
1'.,
iii
II:
~
I
\.~
~
30
Ul
-,
~
Iii
20
kr
- BlOKE IN GRIPS
CYCLES
Flexural tests of the sheet specimen were made at 1725 CPM. Results are indicated by the
open points in the above SoN diagram. The endurance limit was not reached at stresses as low as
20,000 psi.lfan estimated limit of 19,000 psi is chosen, the endurance ratio would be only 0.155,
a value considerably lower than for any other known metal or alloy. Most investigators have
obtained normal values around 0.3 in similar tests.
Several of the sheet fatigue specimens developed fatigue cracks away from the milled specimen
edges. The cracks did not appear to be associated with any visible surface imperfection. For these
reasons, it was assumed that the sample was abnormal, rather than the test procedure. Very
careful oil-powder and fluorescent powder tests, supplemented by metallographic examination,
failed to reveal any surface cracks, even when the sheet was flexed to open any incipient hairline
defects.
It was considered possible, though not probable, that residual stresses from cold rolling were
acting in a deleterious manner. If so, a moderate temperature stress relief might help. Brief
experiments soon disclosed that temperatures at least as high as 400 OF(205C) did not lower
the hardness; in fact, the hardness may have increased very slightly. Knowing this, a set of sheet
fatigue specimens was stress relieved for two hours at 400 OF(205C). The solid points in the
graph above represent the results obtained with these specimens. The endurance limit was not
altered significantly. A definite shift to the left in the upper portion of the curve was evident,
although the direction of shift was opposite to that, had the heat treatment released undesirable
stresses.
437
438
E 7EUI
~--
. e __
~
_
---==='~II
--
~~
e.
.Iee
.28e
--_. _--
..
C
~_==
--
--.~--
----
....
.60e
I.ee
2.e8
CYCLES TO FAILURE (IN MILLIONS)
le.ee
6.00
Sol. Cutting
Fluids
Neutral
75
62.5
12.5
Chlorinated
65
57.5
12.5
67.5
12.5
Sulfurized
Source: V. A. Tipnis and J. D. Christopher, "Machinability Testing for Industry," in Machinability Testing and Utilization of
Machining Data, American Society for Metals, Metals Park OH, 1979, p 26
SURFACE
GRIND
I 62
Gentle
~13
HAND
GRIND
r;~ntl~
Gentle
80
80
157
I 30
Ab .. etve
164
r;~ntl~
I 66
Ah ... lu~
~""n",
CHEMICAL
MILLING
nu_~
67
84
177
u ........
35
65
'II
59
I l2
...
20
165
51
14 5
I
20
40
60
80
Bar chart presentation showing relative effects of various machining and grinding operations on fatigue characteristics of titanium alloy Ti-6AI-4V.
Source: Norman Zlatin and Michael Field, "Procedures and Precautions in Machining Titanium Alloys,"in Titanium and Titanium
Alloys: Source Book, Matthew J. Donachie, Jr . Ed., American Society for Metals, Metals Park OH, 1982, P 354
439
440
Gentle
SURFACE GRIND
14
67
Il
160
ECM FRONTAL
161
I 40
ECM TREPAN
I
lO
J
40
60
80
Bar chart presentation showing effects of various metal removal operations on the fatigue
characteristics of titanium alloy Ti-6AI-4V.
Source: Norman Zlatin and Michael Field, "Procedures and Precautions in MachiningTitanium Alloys."in Titanium and Titanium
Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982, P 355
130
120
.
..
... 110
I
<,
b lOa
Load
axis
.....
II [loio la
........
........................
90
80 L--
........
.....L
--------J
Nt - cycles
SON curves showing the effect of texture on the fatigue strength of Ti6AI-4V. Fatigue strength is greater when the stress axis coincides with
the direction of a high density of basal poles.
Source:J. C. Williams and E. A. Starke.Jr., "The Role of Therrnomechanical Processing in Tailoring the Properties of Aluminum
and Titanium Alloys," in Deformation, Processing, and Structure, George Krauss, Ed., American Society for Metals, Metals Park
OH, 1984, P 334
441
442
900
-o~-
::;:
.,
-[]o-
\
\
800
-0
-,
<,
.-
a.
E
700
II>
II>
'"
Air
...... ....
-- - - - - - - -
600
<fl
10 4
10 5
106
Cycles to Failure
10 7
Source: J. C. Williams and E. A. Starke, Jr., "The Role of Thermomechanical Processing in Tailoring the Properties of Aluminum
and Titanium Alloys, "in Deformation, Processing, and Structure, George Krauss, Ed., American Society for Metals, Metals Park
OH, 1984, p 335
-,
AIr
-,
<,
............
_-----
lOS
Cycles to Failure
Q.
..
<l:
~
31/2 %
Ti-6AI-4V
900
IJ)
800
\
\
700
Nael Solution
\
\
-,
-,
-,
<,
600
~------
104
105
106
10 7
Cycles to failure
b
O! + ,8- processed Ti6Al4V, showing the
effects oftexture and environment on fatigue strength. (a)
Tested in air. (b) Tested in 3'12% NaC\, These data show
that testing in an aqueous 3'12% NaCI solution reduces
fatigue strength when the stress axis is along [0001].
Source:J. C. Williams and E. A. Starke, Jr.. "The Role of Thermomechanical Processing in Tailoring the Properties of Aluminum
and Titanium Alloys," in Deformation. Processing, and Structure, George Krauss, Ed .. American Society for Metals, Metals Park
OH, 1984, p 335
443
444
Kmin+lJ.KalK 1c
.::"
1
..,z
.....
..,e
o
'"
10- 10
L-_ _...L<...
log
lJ.K-
Schematic plot showing characteristic shape offatigue crack growth rate (doldN) versus cyclicstress-intensity (~) curves.
Source: J. C. Williams and E. A. Starke, Jr., "The Role of Thermomechanical Processing in Tailoring the Properties of Aluminum
and Titanium Alloys," in Deformation, Processing, and Structure, George Krauss, Ed., American Society for Metals, Metals Park
OB, 1984, P 338
17-32. Ti-6AI-4V: Fatigue Crack Growth Rates for ISR Tee, and
Extrusions
10-
rr========::::::;,-----,
Ti-6AI-4V
R = O.l,ROOM TEMPERATURE
TL AND LT ORIENTATION
ISR TEE
o EXTRUSION
W
...J
000
>-
~
:r:
o
z
~
"tl
w
I-
a:
:r:
a:
(9
~
a:
o
w
::J
(9
lL
o
o
-7 ':------:~--........-_:';:-.......-7:........~::_":~
10 10
20
40
60
80 100
Source: W. T. (Ted) Highberger, Govind R. Chanani and Gregory V. Scarich, "Advanced Titanium Metallic Materials and Processes for Application to Naval Aircraft Structures," in Production to Near Net Shape: Source Book, C. J. Van Tyne and B. Avitzur,
Eds., American Society for Metals, Metals Park OH, 1983, P 124
445
446
: : (INCHES/CYCLE)
I.IE
0
~
_11
~
-
,~
'
o
10
100
dK (KSI {INCH)
0- HEAT 3
b - HEAT 2
ORIENTATION - RW
0- HEAT 1
R FACTOR
-to. 30
Source: M. J. Harrigan, M. P. Kaplan and A. W. Sommer, "Effect of Chemistry and Heat Treatment on the Fracture Properties of
Ti-6AI-4V Alloy," in Titanium and Titanium Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals,
Metals Park OH, 1982,P 65
-.jm
100
10
I
WATER-QUENCHED
AIR-COOLED
FURNACE-COOLED
A'A
k~
'!:/lAC
10.3
'"
o
>..
u
':-
.s
100
:/
10- 4
:'/
:/
1I
~
"U
:/
:,:"
,.":'
10. 5
I
0.1
10
~K
ksi-
100
Jin":
From the data presented above, it can be seen that air cooling, per se, produced
little or no change in the cyclic crack growth compared to the mill-annealed base
(22
material. The slightly decreased crack growth rates above a IJ.K of 20 ksi
MPa VITi) are, more probably than not, the result of the higher fracture toughness
of the air-cooled material. However, both water quenching and furnace cooling
resulted in fatigue crack growth rates noticeably different from those measured for
the mill-annealed base material. As shown, furnace cooling had a consistently detrimental effect on the crack growth rate while water quenching produced $!eatly
increased crack growth rates above a stress-intensity range of 18 ksi yin. (20
MPa . ~). The accelerated growth rate above 18ksi \!Ill. (20 MPa .
may be
attributed to the proximity of the maximum stress intensity to the critical value.
The critical stress-intensity value for water quenching was an exceptionally low 38
ksi .jll1. (42 MPa . Jill).
Jill.
Jill)
Source: R. E. Lewis, J. G. Bjeletich, T. M. Morton and F. A. Crossley, "Effect of Cooling Rate on Fracture Behavior of MillAnnealed Ti-6AI-4V," in Titanium and Titanium Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for
Metals, Metals Park OH, 1982, P 90
447
448
u
>u
10- 5
....
Q>
.s
a.
~
.J::.
0,
...,
I
I
I
I
Q>
'i0
10- 6
-"
u
Sinusoidal
loading
--Dwell at
maximum load
10-7
100
10
Stress intensity
factor range (MPa m 1/ 2 )
A phenomenon which may be unique to certain titanium alloys is the effect of dwell periods at maximum load on rates of
growth of fatigue cracks. This effect is shown schematically
here, and increases in the rate of crack growth of as much as 50
times may occur compared with results obtained in tests on the
same material subjected only to sinusoidal stress cycles. Dwell
effects are maximized in alloys containing substantial amounts
of the a-phase which have a preferred texture such that stressing
is normal to the basal planes, whereas they appear to be insignificant if stressing occurs parallel to the basal planes of the aphase, or if the microstructure is homogeneous and fine grained.
Particular attention has been paid to a / {3 alloys, e.g., Ti-6AI4V, in which dwell effects have also been found to decrease with
increasing amounts of the {3-phase in the microstructure. In all
cases, dwell effects disappear when stressing occurs at temperatures above 75C (165 "Fj.and they are generally considered to
arise from the preferential diffusion of hydrogen, during the
dwell period, to regions of localized hydrostatic tension ahead
of an advancing crack. Such an accumulation of hydrogen
would tend to embrittle this region, and it has even been suggested that brittle plates of TiH 2 may be formed.
Source: I. J. Polmear, Light Alloys, Edward Arnold Ltd, London, England, and American Society for Metals, Metals Park OH,
1981, P 200
alP
I -. ", .'
" ..
Ti-6AI-4V
-SA
10
20
50
100
Work on Ti-6Al-4V rolled plate has indicated that the superior fatigue performance with the {3-annealed condition is associated with relatively slower rates of
crack propagation (above graph). This effect, in turn, is attributed to the slower
progress of cracks through the Widmanstatten microstructure, particularly at stress
intensities below a critical value at which desirable crack branching occurs within
packets ofthe a-laths.
Source: I. J. Polmear, Light Alloys, Edward Arnold Ltd, London, England, and American Society for Metals, Metals Park OH,
1981, P 179
449
450
70
~o
o
~OHIP
------00------
40
0.1.
Source: J. H. Moll, V. C. Petersen and E. J. Dulis, "Powder Metallurgy Parts for Aerospace Applications," in Powder MetallurgyApplications, Advantages and Limitations, Erhard Klar, Ed., American Society for Metals, Metals Park OH, 1983, P 286
120~
100
'GM"S'$,
K'S/
80
~ Q I/IPC'tCl.
o
15S0i:-I-5 1t:.'6/-II-IR
40t
''''50.~-1'51t:'8/- 3/-1J:?
22'50~-I0/t:!9I-II-IR
ANN P!AT~
/0'-
10-5
-'-:-_ _----',
1
/0
/0'
C~ESTO~(L~
SoN curve for HIP'd Ti-6AI-4V and annealed plate. According to the
above data, fatigue results for Ti-6AI-4V are within the required range
for plate properties from MIL-T-9046.
Source: W. Theodore Highberger, "Manufacture of Titanium Components by Hot Isostatic Pressing," in Production to Near Net
Shape: Source Book, C. J. Van Tyne and B. Avitzur, Eds., American Society for Metals, Metals Park OH, 1983, P 304
451
452
a..
42.2 (60)
:E
VI'
VI
Q.l
.......
en
28.1 (40)
104
10,
70.3 (100 )
56.2 (80 )
'u;
-:-- r----
42.2 (60 )
li.
--
~li.
Pli.
Ali.
o~l
:E
....~'
28.1 (40 )
O As received powder (SM 772)
li. -80 mesh screened powder (SM 768)
en
14.1 (20 )
io,
10,
High-cycle fatigue (HCF) data were developed on Ti-6AI-4V (Std) by Williams International in a program to apply near-net-shape HIP technology to
a compressor rotor part for the F-I07 cruise missile engine. In this study, two
size fractions of powder were used: -35 mesh (as-received) and -80 mesh.
There was no difference in HCF test results between the two sizes. Roomtemperature and 700 OF (370C) S/ N curves are shown above.
Source: J. H. Moll. V. C. Petersen and E. J. Dulis, "Powder Metallurgy Parts for Aerospace Applications,"in Powder MetallurgyApplications, Advantages and Limitations, Erhard Klar, Ed., American Society for Metals, Metals Park OH, 1983, P 286
1000
140
co
Q..
'iii
iii
III
.><
iii
800
III
e
li)
100 ti
E
:J
E
E
:J
E 600
'x
co
'x
co
60
400
200
10'
10'
10'
10'
Cycles to failure
10'
10'
SON curves showing comparison of smooth axial fatigue behavior of Ti-6AI-4V blended elemental and prealloyed P 1M compacts with wrought annealed material. Tested at room
temperature, R = O.
Source: Metals Handbook. 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals, Metals Park OH, 1984,P 753
453
454
Q..
1200
'"
1000
en
1ii
800
E
::J
E
600
CO
400
103
'(ij
.>t:
10
106
10
10
f/l
~f/l
100
E
::J
E
60
104
'"
140
'x
CO
~
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals, Metals Park OH, 1984, P 44
ro
a..
~
160
120
800
'ijj
..".
vi
u>
III
III
~
1;;
~
1;;
E
::J
E
80
E
::J
E
'x
ro
'x
ro 400
40
o"10
-'-
...L...-
--'
....L..
---I0
108
Cycles to failure
SON curves (bands) for titanium alloy Ti-6AI-4V processed by various fabrication processes.
The inconsistentfatigue life ofthe hot isostatically pressed product is usually casued by inclusions in the compact.
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals, Metals Park OH, 1984,P 439
455
456
10
10-4
10
11M and P/M Ti-6AI-4V
at room
temperature
(laboratory air),
R = 0.1, at 5 to
30 Hz
102
10-2
Q)
Q)
~
c
>
10- ~
E
E
......ai
i...
ai
III
.s::. 10- 5
Recrystallization
anneal
s:
...0
...001
10- 4 ~
.>0:
(J
(J
...
...
o
III
III
10- 6
10- 5
Source: Metals Handbook, 9th Edition, Volume 7, Powder Metallurgy, American Society for Metals, Metals Park OH, 1984, p 752
130
<,
<,
110
0---,
<,
l!l'
c:
<,
u;
>l
TEST CONDITIONS:
CONSTANT AMPLITUDE, R ~.1
K ~ 1.0
T
STRESS RELIEVED, FLUSH WELD BEAD
90
a:
t-
IJ)
:E
:::>
:E
"""-
CD
~- ~
70
:E
LEGEND
50
---/:::,.
BASE METAL
--0
SSEB WELDEO
10
S-N curve for titanium alloy plate-base metal versus SSEB-welded. Results show that the constantamplitude fatigue life of SSEB weldments in O.440-in.-thick plate equals that of the base metal.
Source: R. H. Will, J. G. Madora and H. P. Ellison, "Sliding-Seal Electron-Beam Welding of Titanium," in Source Book on
Electron Beam and Laser Welding, Melvin M. Schwartz, Ed., American SccietyIor Metals, Metals Park OH, 1981, P 87
457
458
120
TEST CONDITIONS:
CONSTANT AMPLITUDE, R
KT~
~.1
10
1lX'
80
o
~
LEGEND:
40
--- 0
SSEB WELDED
---- 6
BASE METAL
20
SON curves for titanium alloy plate-base metal versus SSEB-welded. Results show that the constantamplitude fatigue life of SSEB weldments in O.940-in.-thick plate equals that of the base metal.
Source: R. H. Witt, J. G. Madora and H. P. Ellison, "Sliding-Seal Electron-Beam Welding of Titanium," in Source Book on
Electron Beam and Laser Welding, Melvin M. Schwartz, Ed., American Society for Metals, Metals Park OH, 1981, p 87
459
O.OSQ-IN.THICK BASEMETAL
Ti-6AI-4V STOA CURVES
120
ci
s"
100
en
en'
en
w
a:
Ien
::E
SO
60
40
20
103
10 5
CYCLES
SoN curves for EB weldments that were flawless (lower two curves). Upper curve shows scatter band for base
metal (O.080in.-thick Ti-6AI-4V STOA.).
Source: R. Witt, A. Flescher and O. Paul, "Weldability and Quality of Titanium Alloy Weldments," in Titanium and Titanium
Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982,P 313
460
a
a
120
O.oaOIN.THICK
aASEMETAL
Ti6AI4V STOA CURVES
100
II
a:
in
~'
ao
a:
Ii;
X
<l:
:E
60
40
20
103
105
CYCLES
SON curves for ED-welded Ti-6AI-4V titanium alloy showing effects of porosity.
Above are shown experimental data obtained for porosity-containing ED welds which are
superimposed on a set of curves for the base material (0.080-in.-thick Ti-6AI-4V STOA sheet) at
various K, factors. For the points within the boundaries of the band, radiography indicated
scattered porosity (0.003 to 0.005 in. in diameter). For points below the lower boundary of the
band, radiography indicated either linear or heavily scattered porosity.
Source: R. Witt, A. Flescher and O. Paul. "Weldability and Quality of Titanium Alloy Weldments," in Titanium and Titanium
Alloys: Source Book Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982, P 312
461
0.250-IN.-THICK BASE-METAL
Ti-6AI-4V STOA CURVES
120
0
~ 100
Cii
~
iii
lI)
w
a:
BO
lll)
::E
60
D
40
o
105
CYCLES
SON curves for porosity-containing gas metal-arc welds. In the above graph the experimental fatigue
data for porosity-containing GMA W weldments are superimposed on SoN graphs for Ti-6AI-4V
STOA material (0.25 in. thick) for various K 1 factors.
Source: R. Witt, A. Flescher and O. Paul. "Weld ability and Quality of Titanium Alloy Weldments," in Titanium and Titanium
Alloys: Source Book, Matthew J. Donachie, Jr., Ed., American Society for Metals, Metals Park OH, 1982, p 313
462
III
0..
en
III
w
It:
t;
---
100
80
60
-'"
..
'r.I
.. Q=
(!)
5z
40
20
It:
W
C[
f 7500 CPM
Kr
HEAT TREATED
ELECTRON BEAM WELDED
a STRESS RELIEVED TI-6AI-4V
=1.0
I
6
10
CYCLES TO FAILURE
Room temperature rotating-beam fatigue life of unwelded and electron-beam-welded Ti-6AI-4V titanium
alloy in fully heat-treated condition. Decrease in fatigue strength ofthe weldment relative to the parent metal
did not exceed 12%.
Source: S. M. Silverstein, V. Strautman and W. R. Freeman, "Application of Electron Beam Welding to Rotating Gas Turbine
Components," in Source Book on Electron Beam and Laser Welding, Melvin M. Schwartz, Ed., American Society for Metals,
Metals Park OH, 1981, P 169
.....
....
ci
Legend
$1!!~iJfift:
100
II
a:
.....
'iii
.....
'@
80
Ul
Ul
E
E
:;,
oco
.0
-,
Gl
Ul
60
<,
')(
IV
::iE
40
20
<,
<,
-.....
E
---
----
4 runouts
.....
l3Gl
Ul
E
:;,
E
400
')(
----
10 7
Cycles to failure
IV
300
::iE
10 8
The fatigue properties of welds as shown above indicate that under proper
welding conditions, laser welds can be made in Ti-6AI-4V which exhibit base
metal fatigue characteristics. The best laser weld failures initiated at sites in
the base metal, whereas other weld failures originated at undetected small
pores. Where failures initiated in the base metal, it was concluded that no
porosity or weld defects of sufficient size to preferentially initiate fatigue
fractures was present.
Source: E. M. Breinan, C. M. Banas and M. A. Greenfield. "Laser Welding-The Present State-of-the-Art," in Source Book on
Electron Beam and Laser Welding. Melvin M. Schwartz. Ed., American Society for Metals, Metals Park OH, 1981, P 289
463
464
800 t- o
..
700
:E
600
Do
~.
...'"
~
E
::J
E
'j(
:E
500
400
300
200
120
oooeJO
- 90
0000y 00
0
0
00'17001:>.61:>.
- 60
~f:>6'17
00
+000#
00
'';;
.>t
.n
...e'"'"
E
::J
E
..
'j(
:E
0
0
+o~ooo
30
oc:P
100
Notch fatigue strength of as-cast Ti-6AI-4V. Each symbol represents fatigue data from a different source. Stress ratio,R, typically was +0.1; stress concentration factor, K/, was mostly 3.0,
but a few tests were run at K/ = 1.0.
Source: Metals Handbook, 9th Edition, Volume 3, Properties and Selection: Stainless Steels, Tool Materials and Special-Purpose
Metals, American Society for Metals, Metals Park OH, 1980,P 411
'"0.....
N
e
U
-,
Cl
-"
to
en
en
w
a::
l-
en
0.5
CYCLES
S-N curves for zirconium grade 702, showing effects of notches
and elevated temperature (400C, or 750 OF) on fatigue
characteristics.
Source: Donald R. Knittel, "Zirconium." in Corrosion and Corrosion Protection Handbook, Philip A. Schweitzer, Ed., Marcel
Dekker, Inc. New York NY, 1983. p 198
465
466
'W:....
::f ....0.1:-:!-::
o ...............
x~x
10 6
107
NUMBER OF CYCLES
50
(5.Q.~
~ I 40
Vl
llJ
[3 ~ 30
11::11-0
VI Cl:
olE
llJ
-I
::::I
(3
x,oX,
'o'x
VS
'-.""',
'"'ox
lbX~',
"
"'80x.. .
XX
x
-_
--o-J!l<u_x)(""
....x....
20
0",,--
-0_
0_
10
105
106
NUMBER OF CYCLES
The S-N curves shown above indicate that: (1) the welding
practice is of no great importance; and (2) the cast steel box
design is superior to a weldment design.
Source: Steel Castings Handbook. 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980,
P 7-6
0.9
0.6
0.7
0.6
0.5
:I:
l-
:I:
I-
lI)
lI)
W
.J
e> ~ 0.4
Z
It: It:
l- I-
e>
~
<l: W
u, I-
RANGE
TENSILE STRENGTH
0.3
0.2
0.1
0
0.5
MEAN
- 0.1
0.6
STRESS
TENSILE STRENGTH
- O. 2
- 0.3
-0.4
Surface condition has a significant effect on fatigue life and fatigue limit. A
highly polished smooth test specimen can exhibit twice the fatigue strength of
a rough machined sample. A good design approach is to use the notched
fatigue limit asa design value. For cast steels a O.OOIS-in. (O.0381-mm)root
radius circumferential notch in a rotating beam fatigue specimen reduces the
fatigue limit by about 0.7 of the unnotched value. This is sufficient to account
for variations in surface finish and minor surface discontinuities. The above
diagram shows that even severe surface discontinuities, not normally permitted by workmanship standards, do not reduce the fatigue limit by much more
than the 0.7 value.
The above emphasis on surface discontinuities is due to the fact that subsurface discontinuities which do not have a crack-like sharpness and which
do not significantly reduce the load-bearing area of a component generally
have little effect on fatigue performance.
Source: Steel Castings Handbook, 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH, 1980,
P 7-6
467
468
125
.......:::_II__-___7'''-------_+---------~
7oo~------~......,;~~-..-==-"""'-"""",------,,L----+----------d
100
6oo~--------___I--=~:::""....,--=_=-=""'l-"",""----------l
~ 4oot-----------1------------'''''f.,=---'''......, -----'''''-..---=c.j
Ul
Ii
l'
50 Ul
3oot-----------1----------+----"'-,--=---~c.j
2oot-----------1----------+---------''''''''',----~c.j
25
loot-----------1----------+------------''-l
0'-------------'------------'------------' 0
103
106
Number of cycles to failure
50
100
150
1000
I II
800
MaChine[5h
D..
::;;
600
-:V
]
'!l
.~ 400
200
125
~
100
75
/
--- >-_L --1,....--t.>
......
"....
400
600
800
1000
1200
50
----1---
1400
]
~
.~
As-forgedor
decarburized
u.
200
250
200
1600
-...
1800
.;;;
u.
25
2000
Application of small-scale laboratory fatigue testing to the analysis of components or assemblies introduces additional variables. One is the effect of surface condition. The curves in the top
curve above demonstrate that the fatigue strength of steel specimens varies markedly, depending
on whether the surface is polished, machined, hot rolled, or as-forged. The steel tested was an
unidentified wrought low-alloy steel heat treated to 269 to 285 HB, equivalent to a tensile
strength of 876 MPa (127 ksi) and a yield strength of 696 MPa (101 ksi). Sample preparation
required that the specimens be machined and polished after heat treatment and that rolling or
forging precede heat treatment. For a fatigue life of one million cycles, the fatigue limit was 393
MPa (57 ksi) for the ground specimens, 317 MPa (46 ksi) for the machined specimens, 207 MPa
(30 ksi) for the as-rolled specimens, and only 152 MPa (22 ksi) for the as-forged specimens.
The curves in the bottom graph apply to steels with tensile strength ranging from 345 to 2070
MPa (50 to 300 ksi) and are approximations from several independent investigations. Sample
preparation for "as-forged or decarburized" specimens at the 965 MPa (140 ksi) tensile-strength
level include 4140-type steels rough machined from bar stock, heated to approximately 900C
(1650 OF) in a gas-fired muffle for 20 to 30 min, very lightly swaged from an original 7,47-mm
(0.294-in.) diameter to a final diameter of 7.16 mm (0.282 in.), and air cooled. Heat treatment
consisted of austenitizing in a salt bath at approximately 830C (1525 OF)for 45 min, oil quenching, tempering in air for 1 h at approximately 620C (1150 OF), and water quenching. Forging
and heat treating produced a surface decarburized to a depth of about 0.06 mm (0.0025 in.).
These specimens exhibited a fatigue strength, at 106 cycles, of about 310 MPa (45 ksi), compared
with 470 MPa (68 ksi) for samples that were not forged but were machined or polished and free
of decarburization. Decarburization lowers the strength levels obtained by heat treatment.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 355
'" 50
~ (345)
tJl
40
0.. (276)
o
g 30
_ (207)
'E
:.:::i
20
~ (138)
Cl
10
(69)
.50
0
tx:
Q)
~
.45
.40
Cl
u..
.35
6.4
6.6
6.8
7.0
Density, gr/cucm
7.2
Source: Kurt H. Miska, "Powder Metal Parts," in Source Book on Powder Metallurgy, Samuel Bradbury, Ed., American Society
for Metals, Metals Park OH, 1979, P 3
469
470
150
00
100
o
o
200
400
Tensile strength)
500
600
N / mm
Source: Per Lindskog, "The Effect of Phosphorus Additions on the Tensile, Fatigue, and Impact Strength ofSintered Steels Based
on the Sponge Iron Powder and High-Purity Atomized Iron Powder," in Source Book on Powder Metallurgy, Samuel Bradbury,
Ed., American Society for Metals, Metals Park OH. 1979, P 46
40
As - Sintered, Tensile
Strength 67,OOOpsi
10
106
10
Cycles to Fuilure , N.
SON diagrams representing fatigue behavior of7.0 g/ cm 3 density,
4 Ni-0.48 C steels, and effect of quenching and tempering on tensile and fatigue strength.
Source: A. F. Kravic and D. L. Pasquine, "Fatigue Properties of Sintered Nickel Steels," in Source Book on Powder Metallurgy,
Samuel Bradbury, Ed., American Society for Metals, Metals Park OH, 1979, P 28
471
472
r---,...-----,---,----...------,---,---....,.--=__--,
LEGEND
50
Smooth As-Sinfered
Smooth Quenched
Tempered
Ui
40
(L
o
o
o
I
~
E 30
--'
.,
:::J
0>
" 20
LL
10
20
40
60
80
100
120
t40
160
Relation between fatigue limit and tensile strength (fatigue ratio) of sintered nickel
steels.
A plot ofthe fatigue ratio (above) indicates an average smooth value of 0.4 up to
150,000 psi tensile strength. Thus the averagefatigue ratio for sintered nickel steel is
0.4 which is apparently independent of density level, alloy content, and state of heat
treatment and therefore can be used to predict the fatigue behavior of other sintered
nickel steels.
Source: A. F. Kravic and D. L. Pasquine, "Fatigue Properties ofSintered Nickel Steels," in Source Book on Powder Metallurgy,
Samuel Bradbury, Ed., American Society for Metals, Metals Park OH, 1979, P 30
40
Smooth
2.2 Kt Notched
10
106
107
SoN curves for 0.48% carbon-4.0% nickel alloy steel in the as-sintered condition (7.0
g/ em! density). The two curves demonstrate the effect of a notch on fatigue
characteristics.
Source: A. F. Kravic and D. L. Pasquine, "Fatigue Properties of Sintered Nickel Steels," in Source Book on Powder Metallurgy,
Samuel Bradbury, Ed., American Society for Metals, Metals Park DB, 1979, P 33
473
474
0-0-'in
0--
a.
o 30
o
o
2.2 Kt Notched
If)
20
+
If)
10
106
107
Cycles to Failure, N.
SN curves for 0.48% carbon-4.0% nickel alloy steel in the quenched and tempered
condition (7.0 g/cm! density). The two curves demonstrate the effect of a notch on
fatigue characteristics.
Source: A. F. Kravic and D. L. Pasquine, "Fatigue Properties of Sintered Nickel Steels," in Source Book on Powder Metallurgy,
Samuel Bradbury, Ed., American Society for Metals, Metals Park OH, 1979,P 34
40
'"
a..
'"
'"
e
l""- I'--.
Nilrided
Nitriderl
30
=
=
=-
en
--.........
r-
to-.
Not nitrided
r--
20
r---..r--
10
lQ5
Smooth (K=1)
Notched (K=2)
I I III
I IIII
Not nitrided
10 8 lQ5
Cycles to Failure
S-N curves for sintered powders (low-carbon; 1 to 5% copper, 7.1 g/cm 3 density). As
shown above, notches greatly lower fatigue strength, particularly of those that were
not nitrided.
Source: "Nitriding Improves Fatigue Resistance of P 1M Parts," in Source Book on Nitriding, American Society for Metals, Metals
Park OH, 1977, P 292
475
476
<;
I I III I I II I III r
30
I II
Density - 7.0 g per eu em
Nitrided
20
r-..... r--...
0
0_
""" t--.....
<11<II
en 10
Not nilrided
Nilrided
Not nitrided
11
10 6
10 8 !OS
10 6
Source: "Nitriding Improves Fatigue Resistance of PI M Parts." in Source Book on Nitriding, American Society for Metals, Metals
Park OH, 1977, P 292
I"":iZ ~
~'''':': :::
'Vi
a.. 40
=
=. ~
=
Sintered Iron
::fit k:t
Nitrided
.:.:.:.:.:.: ;:::;:::::; :::;::;::
;:;:; ::: :: :.:.:.:.:.:.
::::; :::::'~
00
::::::;:;::: :.:.:.:.:.:
~:::::: ;~~~;~~~~~~ U~
Not nihided
~'" .....:....
_:
20
10 6
10 7
00
10 8 10 5
.....':':":
~ ::::;:
Nilrided
:;:;:;:;:::::
::: :~ .........
:.:~:~:~:~:~: ~:~:~ :~:~ ~:
m
it ~ : mIt ~~tt :~t~ ~~~~
"""" <' t
106
Not nilrided
::: ~~
~~~~~~~~~~~~~ ~~~~~~~~~ ~~~~;~ .':'
:::::::::
t~~ jtm~~ :::::;:;: ~ ~ ~: ~m ~
10 7
108
Cycles to Failure
Left: Effect of salt bath nitriding on fatigue strength of ductile iron. Right: Effect ofsalt
bath nitriding on P 1M parts. Specimens were made from a 3% copper sintered iron
ranging from 6.2 to 7.0 g/cm l in density. All specimens were unnotched and were
heated in a nitriding salt at 565C (1050 OF) for two hours.
Source: "Nitriding Improves Fatigue Resistance ofP I M Parts," in Source Book on Nitriding, American Society for Metals, Metals
Park OH, 1977, p 291
477
478
125
a Mild sleel
o Brass
~0~4;:---------:<----------,:-:':T------~IOV
5
6
10
Cycles 10 failure,
10
log scale
SoN curves for composite of mild steel clad (by the explosion technique) with
brass.
Initially, the composite has greater fatigue strength than either brass or
mild steel alone, but above about 106 cycles the values for the composite drop
to about that of mild steel but still remain substantially higher than for brass
alone.
Source: S. K. Banerjee and B. Crossland, "Mechanical Properties of Explosively-CJadded Plates," in Source Book on Innovative
Welding Processes, Melvin M. Schwartz, Ed., American Society for Metals, Metals Park OH, 1981, P 148
0____~
.
' Slainl e s s
2 /5
-e
12-5
'"
0
o~~-'====t
10 '1:
.e
0_
OJ'
"l::I
75 ~
~
5
'"
:::
~
/05
Cycles 10 failure,
/0 6
log scale.
SoN curves for composite of mild steel clad (by the explosion technique) with
austenitic stainless steel. Here it is seen that fatigue characteristics of the composite are nearly the same as for stainless steel, and substantially higher than the
fatigue strength of the unclad mild steel.
Source: S. K. Banerjee and B. Crossland, "Mechanical Properties of Explosively-Cladded Plates," in Source Book on Innovative
Welding Processes, Melvin M. Schwartz, Ed., American Society for Metals, Metals Park OH, 1981, P 148
479
480
c::J
600,------------------------,
~
500 1 - - - - - - -
:2
s: 400
a,
c
1-------
300
Cl>
.~ 200
'"
u,
100
SAE 1000
SAE 1015
SAE 1035
SAE 1045
SAE 1060
600
'"
500
0-
:2
-S' 400
'"
300
Cl>
::l
.g 200
'"
u,
100
0
SAE 1035
SAE 5134
Nitralloy
Bar charts showing increases in fatigue limit that may be obtained by nitrocarburizing (gas or liquid processes).
The amount of improvement in fatigue strength of nitrocarburized materials, as determined with unnotched Wohler test
specimens, depends on the hardness and depth of the diffusion
zone. The potential for improvement in fatigue strength lessens
with increasing carbon and alloy content.
Source: Metals Handbook, 9th Edition, Volume 4, Heat Treating, American Society for Metals, Metals Park OH, 1981, p 269
481
80
"is~
en
60 J----,---------t
.".
t-----t
+ I---l~-~----t
'"
c,
g 401-----1
=-
20 1-----11_
5134
SAE
Basic fatigue characteristics are directly related to carbon content, as indicated in the
above bar charts for carbon and alloy steels (unnotched test bars). Tufftriding these steels
shows results which prove that fatigue strength increases inversely with carbon content;
that is, the lower the carbon, the greater percentage increase in fatigue strength by
Tufftriding.
Source: Edward Taylor, "Tufftride: Only Skin Deep"," in Source Book on Nitriding, American Society for Metals, Metals Park
OH, 1977, P 280
482
60
After Hardening
111111/
I
I
55
....
III
.c
..10:
11111
NltraUoy
1015
1141
8.620
4620
4340
Il5
Atmosphere nitriding studies showing the interrelationships of steel composition and nitriding by the gas process, and the effect on fatigue strength from nitriding.
Source: J. A. Riopelle, "Short Cycle Atmosphere Nitriding," in Source Book on Nitriding, American Society for Metals, Metals
Park OH, 1977, P 286
0.4
340
320
:2
300
...en
c
~
'"
Q)
280
''::;
'"
0.8
1,2
1.6
45
'iii
.:.l
-,
J::."
-,
'"
u,
260
1
c
1;;
Q)
<,
::s
en
40
r-, .....
'fJ
u.
240
10
20
3D
40
35
50
Thickness, /.1m
Effect of coating thickness on the fatigue strength of a carbonmanganese steel.
Source: Metals Handbook, 9th Edition, Volume 5, Surface Cleaning, Finishing, and Coating, American Society for Metals, Metals
Park OH, 1982, P 232
483
484
by
40
200 1-----,,(---+----+----+----+--_=
20
o
Minimum stress,MPa
Wire
diam
Spring
on
Spring
No.
mm
in.
mm
1
2
3
4
5
6
0.81
0.81
1.22
2.59
3.07
4.50
0.032
0.032
0.048
0.102
0.121
0.177
9.52
6.35
15.88
22.22
22.22
22.22
Free
length
in.
Spring
index
mm
in.
Total
turns
Active
tums
Total
tested
0.375
0.250
0.625
0.875
0.875
0.875
10.7
6.8
12.0
7.6
6.2
4.9
22.10
26.97
44.45
60.20
57.15
57.15
0.87
1.062
1.75
2.37
2.25
2.25
6.0
7.0
7.0
7.0
7.5
7.5
4.2
5.2
5.2
5.2
5.7
5.7
16
28
38
43
35
25
Data are average fatigue Iimita from S-Ncurv.. for 185 unpeened springs of various wire diameters run to 10 million cycles oCetrese.Allstresees were
corrected for curvature using the Wahl correction factor. The springs were automatically coiled, with one tum squared on each end, then baked at 260'C
(500 'F) for 1 h, after which the ends were ground perpendicular to the spring axis. The test load was applied statically to each spring and a check made for
set three times before fatigue testing. The springs were all tested in groups of six on the same fatigue testing machine at ten cycles per second. After
testing. the unbroken springs were again checked for set and recorded. Number 4 springs, tested at 1070 MPa (166 ksi) max strese, had undergone about
2,",%set after 10 million strese cycl.., but the stresees were not recalculated to take this into account. None of the other springs showed appreciable set.
The tensile strengths of the wires were according to ASTM A228.
By means of the Goodman diagram many fatigue-limit test results can be shown on the same
diagram as indicated above. In this diagram, line OM represents the minimum stress for the
cycle; the plotted points represent fatigue limits for the respective minimum stresses used. The
vertical distances between these points and the minimum stress reference line represent the stress
ranges. Some scatter may be expected, at least partly attributed to normal changes of tensile
strength with wire diameter. Line UT is usually drawn to intersect line OM at the average ultimate shear strength of the various sizes of wire.
Source: Metals Handbook, 9th Edition. Volume I, Properties and Selection: Irons and Steels. American Society for Metals. Metals
Park OH. 1978, P 293
175
~
~
::i 1000f-------""....,,-=-::l-::--------+f
00
---l150
125
750t---------+----.::::!!oo""""'::-l----------l
500L ;--104
106
la)
Type of wire
Number
of
springs
Oil tempered
Music wire
Avg
~
104
105
(a) Springs were made of minimum quality music wire 0.59 mm (0.022 in.) in diameter. Spring
diameter was 5.21 mm (0.205 in.);D/dwas 8.32. Minimum stress was zero. Stresses corrected
by Wahl factor. (b) Life of springs used in a hydraulic transmission. They were made of
oil-tempered wire (ASTM A229) and music wire (A228). Wire diameter was 4.75 mm (0.187
in.), outside diameter of spring was 44.45 mm (1.750 ln.), with 15 active coils in each spring.
The springs were fatigue tested in a fixture at a stress of 605 MPa (88 ksi), corrected by the
Wahl factor.
The upper graph is a typical S- N diagram showing results of compression testing coil
springs, where the minimum stress is zero while maximum stress is shown by points on the
chart (see spring and testing details given in caption). The lower graph shows an alternate
method of presenting fatigue data for steel springs.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, p 292
485
486
90
80
<:
e
"
70
'in
<:
111
60
50 ~
40
~~ r-...
<,~tressed
-----
/peened
in bending
""-Not peened
I-
/Peened
~Stressed
in torsion
'\:"Not peened
Shot peening is often used to improve fatigue strength of springs by prestressing the surface in compression as indicated in the chart above. Shot
peening can be applied to wire 1.6 mm (1/16 in.) or more in diameter, and
slightly smaller wire using special techniques. The kind of shot used is important; better results are obtained with carefully graded shot having only a few
broken, angular particles. Shot size may be optimum at roughly 20% of the
wire diameter. However, for larger wire, it has been found that excessive
roughening during peening with coarse shot lessens the benefits of peening,
apparently by causing minute fissures. Also, peening too deeply leaves little
material in residual tension in the core; this negates the beneficial effect of
peening, which requires internal tensile stress to balance the surface
compression.
Shot peening is effective in largely overcoming the stress-raising effects of
shallow pits and seams. Proper peening intensity is an important factor, but
more important is the need for both the inside and outside surfaces of the
spring to be thoroughly covered. An Almen test strip necessarily receives the
same exposure as the outside of the spring, but to reach the inside, the shot
must pass between the coils and is thereby much restricted. Thus, for springs
with closely spaced coils, a coverage of 400% on the outside may be required
to achieve 90% coverage on the inside.
Cold wound steel springs normally are stress relieved after peening to restore the yield point. A temperature of 230C (450 OF) is common because
higher temperatures degrade or eliminate the improvement in fatigue
strength.
The extent of improvement in fatigue strength to be gained by shot peening, according to one prominent manufacturer of cold wound springs, is
shown in the above diagram.
Source: Metals Handbook. 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 297
100
<,
600
-.......
.
<,
o Not peened
500
r-.. l"-
Peened
110
100
/Peened
sn
BO
400
~-
10
I::
60
50
106
Number of cycles to failure
BO0
I 10
100
100
.....
",0
600
<,
500
r-,
400
90
r--
BO
10
I-
60
50
106
Number of cycles
(0
failure
BOO
600
500
400
0""
110
100
90
:ii
~~
80
.........
10
.
o'
il
60
60
106
Number01 cycles to failure
Design stresses. Springs were made from 15.9 to 27.0-mm (% to 1-1/16-in.) diam 8650 and
8660 hot rolled steel and heat treated to between 429 and 444 HB. Springs were shot peened to
an average arc height of 0.008 in. on the type C almen strip at 90% visual coverage.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH. 1978, p 304
487
488
~ 540
:<:
Mean
stress:
637 MPa
(92 ksi)
.; 490 -
-o
::s
~ 440
0&-
0.
(J)
- 64
..>:
:r:
C
'tl
0=, ,
:r:
48
Cycles to Failure
,
,
\
H
111
106
~ 50
.\.
p::
- 57
10 5
104
, \
52 -
.,-l
(/)
(/)
o SUP7
_ SUP7-Nb-V
390
(/)
(/)
Q)
H
+J
- 71
ocj-
- 78
e:tJ-
o SUP7
SUP7-Nb-V
I
I
I
5x104 10 5 2x10 5
2x10 4
Cycles to Failure
Fatigue life of coil springs: (left) not corroded and (right) corroded,
Compositions of HSLA Springs Tested
SUP7
SUP7-Nb-V "
Si
Mn
Cu
Cr
AI
0.58
0.56
2.09
1.94
0.83
0.79
0.014
0.014
0.008
0.008
0.09
0.09
0.14
0.09
0.025
0.021
Nb
0.15
0.18
Fatigue tests on coil springs at a hardness of 50 HRC were performed to examine the feasibility of S UP7-Nb-V to the actual suspension coil springs. When the coil springs were free of
corrosion, the result was as shown above (left), in which SUP7-Nb-V has comparable fatigue life
to that of SUP7 in any stress amplitude. When the coil springs were corroded, on the other hand,
the result was rather different. The corroding condition was as follows: an exposure in a chamber
filled with saltwater mist for 1.08 X 104 s (3 h) and a keeping in the atmosphere for 7.56 X 104 s
(21 h). After 10and 20 cycles ofthe corroding, the coil springs were loaded with the surface stress
of 490 340 MPa (71 50 ksi). The fatigue life of the coil springs subjected to 20 cycles of the
corroding are shown above (right). This time, different from the case in the graph at left, there
appears a remarkable difference between SUP7 and SUP7-Nb-V. Measurement ofthe surface
corrosion depth of the two steels showed no difference.
Source: Toshiro Yamamoto, Ryohei Kobayashi, Toshio Ozone and Mamoru Kurimoto, "Precipitation Strengthened Spring Steel
for Automotive Suspensions,"in HSLA Steels-Technology & Applications, American Society for Metals, Metals Park OH, 1984,
P 1022
U)
U)
ILl
lX:
E-<
U) '.-1
1380/200
III
Q::<:
ILl <,
>-<<1l
...:Il:lo
l:lo;:;:
c,
><
150
"
..... "
,.
'.
......d
,,
-.
690/100
~
~ u
'.
.,
f
"
104
-1070
-950
-565
-207
-69
,
345/5
-------
PEAK
RESIDUAL
STRESS, MPa
105
..
106
_ . -
+400
107
CYCLES
Residual stress and unidirectional bending fatigue data for strain-peened SAE
5160 steel. Applied strain during peening curve a, -0.60%; curve b, -0.30%;
curve c, 0% (conventional peening); curve d, preset only; curve e, +0.30%; and
curve f, +0.60%.
Source: V. K. Sharma and D. H. Breen, "Some Aspects of Incorporating Residual Stresses in Gear Design." in Residual Stress for
Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981, p 82
489
490
t:
30r--------~A_r/hr7'T_V.l+_-------------____l
'0
.8
20
E
:l
101------r;..."...,1'A -V/WA-f/.HV/l-V.l1-V'A-t/'J--Y./lH/.,-+-r.,1'A - - - - - - - - _ _ _ _ l
Here are results from simulated service fatigue tests offront suspension torsion bar springs of
5160H steel. Size of hexagonal bar section was 32 mm (1.25 in.), Mean service life, 134,000
cycles; standard deviation, 37,000 cycles; coefficient of variations, 0.28. It must always be
considered that results from actual or simulated service testing are likely to vary considerably
from results of laboratory testing as shown above.
Source: Metals Handbook, 9th Edition, Volume I, Properties and Selection: Irons and Steels, American Society for Metals, Metals
Park OH, 1978, P 677
II
3.0
u-' 2.0
....
.3
u
'"
u..
Q)
...J
1.0
0.5
10 4
106
10 8
The life factor depends on the required life in cycles. For a single mesh the number of revolutions and the number of cycles are equal. For a gear which has more than one mating member,
the life must be equal to the required number of revolutions multiplied by the number of mating
gears.
Source: "Bending and Contact Stresses in Hypoid Gear Teeth," in Source Book on Gear Design, Technology and Performance,
Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 127
491
492
(/)lI'l
(/)0
LLl ......
0:::
E-><
(/)
,......
~ '~
6 . 90/ 1. 0
...... 0..
0 .........
Zc;l
LLlo..
a:l::'<:
'-'
3.45/.5
CYCLES
Bending fatigue design curves for carburized gears having different amounts of
circumferential residual stress at the root-fillet surface,
The bending fatigue design curves for case-carburized gears with the circumferential root-fillet residual stress varying from +138 MPa (20 ksi) to
-690 MPa (100 ksi) are shown above. From these curves the residual stress
factors at various life cycles were calculated as the ratio of the allowable
bending stress for gears with -483 MPa (70 ksi) residual compression to the
allowable stress for gears with + 138 (20), 0, -276 (40), and -690 MPa (100
ksi) residual stresses.
Source: V. K. Sharma and D. H. Breen, "Some Aspects of Incorporating Residual Stresses in Gear Design," in Residual Stress for
Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981, P 86
L50
L50
L50
3. 45/ 5L....::--
'-;-
----'----;;-
----'
105
CYCLES
'"
....
0
1.2
....
Vl
Vl
1.1
....
'"
Vl
...J
:::>
1.0
SHOT CLEANED
Vl
'"
0.9
o. 8 '--.......4- -L..,,----'----;-----..L-:;--.....L.-:;---I
5
7
6
10
10
10
10
10
CYCLES
Top: Allowable (LSO) bending stress design curves for as-carburized,
shot cleaned, and shot peened carburized gears. Bottom: Residual stress
factor computed from the upper chart. K a for carburized shot cleaned
gears equals 1; that is, the allowable SoN curve for carburized shot
cleaned gears is used for design purposes.
Source: V. K. Sharma and D. H. Breen, "Some Aspects of Incorporating Residual Stresses in Gear Design," in Residual Stress for
Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981, pp 77, 78
493
494
U'l
U) ......
U)
13.80/2.0
L90
Jl.lX
c:<:
E- .......
10.35/1.5
U) 'M
L50
Ul
t:lp..
z .......
'"
6.90/1.0
......
00..
L10
z ......
Jl.l:'<:
!Xl '-'
3. 45/ . 5 L..5
10
...I...-
.1....-
--'
10
CYCLES
Probability-stress-life design curves for shot cleaned carburized steel gears having a root-fillet surface finish of 5Jl in.; l.e., KJl = 1.
Source: V. K. Sharma and D. H. Breen, "Some Aspects of Incorporating Residual Stresses in Gear Design," in Residual Stress for
Designers and Metallurgists, Larry J. Vande Walle, Ed., American Society for Metals, Metals Park OH, 1981, p 74
400
(St)
OR
CONTACT
GlO~
~}s,
300
(Se)
STRESS,
KSI
200
G5~
G90
~~
_-}st
10 6
10 7
CYCLES TO FRACTURE OR PITTING
S-N curves showing the wide difference in cycles to failure between bending and contact stress.
Source; D. H. Breen, "Fundamental Aspects of Gear Strength Requirements," in Source Book on Gear Design, Technology and
Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 66
495
496
5,600
.c
..J
c::
- 5,200
Q>
::I
eo
1-c:
4,800
c:
0...
4,400
10 5
10 6
Cycles
Source: G. H. Walter, "Computer Oriented Gear Steel Design Procedure," in Source Book on Gear Design, Technology and
Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 85
)(
G90
G50
GIO
'" 5
..0
c:
::::l
CI::
aI-
BENDING
FRACTURE
----
SPAlLlNG
: : : : - - - FATIGUE
~1~~----MIXED
..............
PITTING
2L--
-L..
'-........
'-........G90
1.. . . . .
....
'-.....: G50
GIO
----I
-.l
105
CYCLES
T-N curves for carburized six-pitch pinion, four-square gear tests.
The above fatigue data show torque versus cycles to breakage, pitting and
spalling for a six-pitch pinion test. Note that there is a mixed area where
failure can occur from anyone or a mixture of the three modes.
Source: D. H. Breen, "Fundamental Aspects of Gear Strength Requirements," in Source Book on Gear Design, Technology and
Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 66
497
498
C1
Ul 105
0:
....Q.
ai
...J
or
.!!!,.
"
..
.. .. . .
....... .--_ev
.. .. . ..
...
......
<,
<,
.....
.. .
<, ~......
.......
a
0:
fill...
1-..........:
.....
Iii
105
CYCLES
FOR
"
10 5
RUPTURE - (N)
<,
107
Fatigue life data for hypoid gears. Sloping line indicates minimum confidence level.
Source: "Gleason Method for Estimating the Fatigue Life of Bevel Gears and Hypoid Gears." in Source Book on Gear Design,
Technology and Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 386
-------- ----..........-..........
..........
---- 1;,0--..
060
--
---- -- ............
.a"" . ... . ..........
---
co
00
0..ll........
<,
<,
~<QI o
:--..
<, ~~~
<,
....'
I~
'l,
.............
"
n
no
lI)
lI)
a:::
l-
ll)
105
106
CYCLES FOR RUPTURE - (N)
Fatigue data-composite for results obtained by testing various gear designs,
Source: "Gleason Method for Estimating the Fatigue Life of Bevel Gears and Hypoid Gears," in Source Book on Gear Design,
Technology and Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 386
499
500
....,..
~
g.
i.
1'---.",-
III
a:
w
CL
ai
.J
~
I
III
:la:
10
.-
<,
<;
,..,..~"
a..
<,
... -.........:
IIII
.........
10 5
C.YCLE:S
10&
fOR
107
RUPTURE -(N)
Graph offatigue life for spiral bevel and Zerol bevel gears. Sloping line indicates the minimum
confidence level.
Source: "Gleason Method for Estimating the Fatigue Life of Bevel Gears and Hypoid Gears," in Source Book on Gear Design,
Technology and Performance, Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, P 385
501
--~
3
~
I
5
5%
n0
50%
...
Gl
....
:..:i
1'000.
.......
Gl
......
C
Gl
95%
0.9
0.8
-0
c;:
C
0.7
0.6
10J
<i
>
.....Gl
105
106
Both strength and durability are fatigue phenomena and therefore display a relationship between stress and life. The life factor for strength may be obtained from the above
data.
Source: "Bending and Contact Stresses in Hypoid Gear Teeth," in Source Book on Gear Design, Technology and Performance.
Maurice A. H. Howes, Ed., American Society for Metals, Metals Park OH, 1980, p 127
502
...
5%
I
l>
50%
0
u,
...::;
po".
I>
.......
I
I
.......
,
I
1
0.9
0.8
95%
I
I
)]
0.6
103
1O~
106
108
10 7
The life factor is obtained from the graph above. This depends upon the required life in cycles.
For a single mesh the number of revolutions and the number of cycles are the same. For a gear
which has more than one mating member the life must be equal to the required number of
revolutions multiplied by the number of mating gears. When the required life is less than
6,000,000 cycles on the pinion, the life factors will be different on gear and mating pinion.
In cases where the load varies, the designer may wish to determine the equivalent life at maximum torque. One suggested method is as follows:
L cp=60L H
T~
[k lnpl+ k n n ( T)
2
5.68
+k)npJ
(Tf; )
5.68
+ ..
+kl/npl/
T)
i;
5.68 ]
n,
L CG=L cp NG
Source: "Bending Stresses in Bevel Gear Teeth," in Source Book on Gear Design. Technology and Performance. Maurice A. H.
Howes, Ed., American Society for Metals, Metals Park OH, 1980, p 149
503
=
=
=-
30
.6.
",,-
,9
25
""
ll...
.8
"C
,~
20
CL
CL
<I:
"C
co
0
.....J
E
15
6 CuI gears
Forged gears
:::I
E
';(
co
::;;:
10
5
lQ3
104
10 5
10 6
1Q7
Cycles 10 Failure
Fatigue data shown in this chart proved that teeth on precision forged AMS 6265
helicopter pinions have a higher fatigue limit than cut teeth. Loads shown are applied
actuator loads. Tooth loads are approximately 33% greater.
Source: "How Gearmaking Methods Compare," in Source Book on Powder Metallurgy. Samuel Bradbury, Ed" American Society
for Metals, Metals Park OH, 1979, P 347
504
90
0
0
0_
80
"0
0
0:::
co 70
e
lJ)
lJ)
en 60
1% failure
m
c
'j
0
z:
50
40
30
- - - Cui gears
Forged gears
10 5
Cycles 10 Failure
S-N curves for spur gears forged and cut from 8620H steel.
As shown above, results of beam fatigue tests indicate that precision flow-forged
gear teeth are about 20% higher in fatigue strength than cut teeth.
Source: "How Gearmaking Methods Compare," in Source Book on Powder Metallurgy, Samuel Bradbury, Ed., American Society
for Metals, Metals Park OH, 1979, P 346
4
5.
6
7
8
3
99._
_5.0
SINTERING
TEMP TIME
ALLOY
2~F I MIN
4600V
4600Y
2100-F 3MIN
2350F 3MIN
2350F 6MIN
4600 V
4600V
4600V
4615
2100-F 6 MIN
(BAR STOCK)
80.0
::>
400
300
...J
u.
20.
z
w
a:
w
Q.
60.0
w
a:
'00
8
60
40
20
/
1.0
.1
.2
--(
.3
TIME (HR.)
.-
LEGEND
~.
,
2
3
!!::ill
2000
2000
4615
SINTERING
TEMP TIME
2350F "'6MiN
2350F 3 MIN
(BAR STOCK)
95.0
80.0
600
w
a:
40.0
300
u.
20.0
a:
w
Q.
10.0
80
60
40
2.0
1.0
.1
.2
.8
1.0
34~6810
TIME (HR.)
Top: Weibull distribution charts for fatigue testing of actual gears and
pinions made from 4600V alloy with various sintering times and
temperatures as shown, compared with cut pinions (4615 bar stock).
Bottom: Similar to graph at top exceptfor alloy 2000 and 4615bar stock.
Source: P. C. Eloff and L. E. Wilcox, "Fatigue Behavior of Hot Formed Powder Differential Pinions." in Source Book on Powder
Metallurgy, Samuel Bradbury, Ed., American Society for Metals, Metals Park OH, 1980, p 308
505
506
24-23. Gears and Pinions: P/M Grades 4600V and 2000 vs 4615;
Percent Failure vs Time
99.9
LEGEND
NO.
950
2
3
80.0
1-1
ALLOY
4600V
2000
4615 8AR STOCK
95'/. CONFIDENCE
LIMITS
40.0
w
0::
60.0
'i
30.0
--l
<l:
20.0
lJ...
IZ
0::
W
(L
10.0
8.0
6.0
4.0
2.0
1.0
.1
5 6
10
TIME (HR.)
Fatigue data for actual gears tested in specially designed machines. Presented
here are Weibull distributions for the three types of alloys tested.
Since the data from the two powder alloys fell into two groups, it was decided to
fit one Weibull curve to all of the data points from each alloy. This was done to
obtain more data points for each curve. The results are shown in the graph above,
which also graphically indicates the 95%confidence limits on the BID lives. It isplain
that the 4600V pinions have superior fatigue life at the stress level of 92,400 psi, and
the slope ofthe Weibull curve indicates uniform deoxidation of preforms and therefore less scatter (steeper Weibull slope) in the fatigue data. In the case ofthe 4600V
alloy, the sintering temperature should have little effect on deoxidation, since the
major alloying constituents, nickel and molybdenum, are readily reducible by CO
at temperatures even below 1150 C (2100 OF).
Source: P. C. Eloff and L. E. Wilcox, "Fatigue Behavior of Hot Formed Powder Differential Pinions." in Source Book on Powder
Metallurgy, Samuel Bradbury, Ed., American Society for Metals, Metals Park OH, 1980, P 313
24-24. Gear Steel AMS 6265: Parent Metal vs Electron Beam Welded
2
100
4 68l
4 681
4 681
4 681
I
900
10
10
i~
50'
Source: N. F. Bratkovich, W. L. Mcintire and Robert E. Purdy. "Electron Beam Welding-Applications and Design Considerations for Aircraft Turbine Engine Gears.t' in Source Book on Electron Beam and Laser Welding, Melvin M. Schwartz, Ed., American Society for Metals, Metals Park OH, 1981, P 199
507
508
z-,
~. ;.>.
>\0
40
.........~
.~
30
. .......
~
" ~
,....
I~
-- _..
3.10 4
ll.i
O,I..l,fl
(I ~ (; (I
.::--.
~~.
20
--- -- -
:~~.. :. ;. ........
i\:
.- .
. .-
.. --- ..
O.~o2
0, 'lG r\(W/m
~
1.)C G
-J-L
LW--+
SoN curves for various tooth profiles (50% survival probability in the short life
range).
The Woehler curves shown above are based (in the sloping section) on a
survival probability of 50% at the number of cycles indicated. The horizontal
sections of the curves are based on the highest load that can be carried for a
minimum life of five million cycles.
Source: H. Winter and M. Hirt, "The Measurement of Actual Strains at Gear Teeth, Influence of Fillet Radius on Stresses and
Tooth Strength,"in Source Book on Gear Design, Technology and Performance, Maurice A. H. Howes, Ed .. American Society for
Metals, Metals Park OH, 1980, P 102
I ,
I I
f/
.....
II ~
II'
;-- 70
Of
II
if
50
~ ...
il
71'
If
30
f
,I:
20
J~!'I
10
8
A -1-
l-
-- _..
----f I - - 1-I
6
4
3.,0'< 4
- --.-
8 1.'0:>
. .
t. 5
,-W
(;
. -
1.1
--~
Source: H. Winter and M. Hirt, "The Measurement of Actual Strains at Gear Teeth, Influence of Fillet Radius on Stresses and
Tooth Strength,"in Source Book on Gear Design, Technology and Performance, Maurice A. H. Howes, Ed.. American Society for
Metals, Metals Park OH, 1980, P 102
509
510
1040 steel
I
100
95
4037 steel
615
s:
:::iE
~g'
650
"v;
"'".
~
625
0.0
:0
~
)(
600
.n .
Cl
c
90 :0
c
..
.0
)(
:::iE
~o.
of>
~-
~~-
85
515
o
>--1-
The bolts (7'8 by 2 in., 16 threads to the inch) had a hardness of 35 HRC. Tensile properties of
the 1040 steel at three-thread exposure were: yield strength, 1060 MPa (154 ksi); tensile
strength (axial), 1200 MPa (175 ksi); tensile strength (wedge), 1190 MPa (173 ksi). For the
4037 steel: yield strength, 1110 MPa (161 ksi); tensile strength (axial), 1250 MPa (182 ksi);
tensile strength (wedge), 1250 MPa (182 ksi).
In general, if bolts made of two different steels have equivalent hardnesses throughout
identical sections, their fatigue strengths will be similar (see above S- N data), 'as long as
other factors such as mean stress, stress range, and surface condition are the same. If the
results of fatigue tests on standard test specimens were interpreted literally, high-carbon
steels would be selected for bolts. Actually, steels of high carbon content (more than
0.55% carbon) are unsuitable because they are notch sensitive.
The principal design feature of a bolt is the threaded section, which establishes a notch
pattern inherent in the part because of its design. The form of the threads, plus any mechanical or metallurgical condition that also creates a surface notch, is much more important than steel composition in determining the fatigue resistance of a particular lot of
bolts.
Source: ASM Committee on Carbon and Alloy Steels. "Threaded Steel Fasteners." in Quality Control Source Book. A. K. Hingwe,
Ed .. American Society for Metals, Metals Park OH, 1982. P 206
24-28. Bolts: S-N Data for Roll Threading Before and After Heat
Treatment
450 r - - - , - - - - , - - - - - , - - - ,
Roll threaded before
heat treatment
60
375 1 - - - t - - - - - 1 - - - - t - - - - - j
50
3001---t-----1----t-----j
40
225
30 U5
150 f-----11-
20
751----t------I---'
L-
! -_ _-'----_ _
10
~_~
104
450 , - - - , - - - - , - - - - - . - - - ,
60
375
-
50
300
40
'"
:;;:
0-
::i 225 I - - - t - -
1"
U5
150 f - - - - t - - - - t - - - + - - - - - j 20
75
1---+---+---+---::::1 10
Source: ASM Committee on Carbon and Alloy Steels, "Threaded Steel Fasteners," in Quality Control Source Book, A. K. Hingwe,
Ed., American Society for Metals, Metals Park OH, 1982, P 202
511
512
24-29. Power Shafts, AMS 6382 and AMS 6260: Electron Beam
Welded vs Silver Brazed Joints
.. ELECTRON BEAM WELDED JOINTS
60
o
o
50
III
S/N-2
Z
:l
:I:
..
40
z
+1
w
30
0:
I-
>0:
o
!;i
20
0:
>
--
r-- l I-
.. ----
~~
S/N-2
SIN-I
ROOM TEMP.
i----
r---r--.
..
S/N-3
o0..
....
SIN-I
:---
--. "-
-I---
10
CYCLES TO FAILURE
S-N curves for electron beam welded versus silver brazed power shafts made from AMS 6382 and
6260 alloy steels.
In the welded shafts, failures occurred apart from the weld, while in the brazed units all
failures occurred in the brazed joints.
Source: S. M. Silverstein, V. Strautman and V. R. Freeman, "Application of Electron Beam Welding to Rotating Gas Turbine
Components," in Source Book on Electron Beam and Laser Welding, Melvin M. Schwartz, Ed., American Society for Metals,
Metals Park OH, 1981, p 187
24-30. Axle Shafts, 1046, 1541 and 50854 Steels: S-N Data for
Induction Hardening vs Through Hardening
513
:\, B & C
D-
~on
if.
'""
if.
v: 1 SO
?i
..9."Il!':.? ....
,.....
-,
tr.
".
.... ..........
:~
...
"
.....
u:: ] on
>c:::
<
~
0;
., ............ --
\',
.......
-. -'-."
'-'-'-B
......... ::::.:::-:.:::-::.: ~
"
50
APPLIED
~
STRESS GRWIE~T
200
400
600
DISTA.'JCE FRO~l SURFACES (.001")
..... 140
IF.
~
u: 100
u:
..... .......~
.
"
..
.~
'
':-.."
l-
i/;
800
60
.,
-, . "
f:S
u:
20
10 2
]0 3
10 4
"
:"::..,,:,,'~ c
.'-._.-E
----B
lOS
10 6
Induction hardened 0.40% carbon steel axle shafts were developed to replace through hardened alloy steel
shafts for both product- and cost-improvement purposes. This was accomplished after a rather comprehensive
bench-test program, which examined variables such as surface hardness, core hardness, gradient strength, distortion, composition, and surface-condition effects. The more promising approach was then subjected to chassis, proving-ground, and in-service testing. Some interesting reflections can result from examining some of the
fatigue data that were generated. That the through hardened concept was vulnerable can be surmised by considering the stresses developed in a full-floating splined shaft loaded in torsion. The stresses are a maximum at the
surface and drop linearly to zero at the center. At the spline, the stresses drop more rapidly at the onset due to the
stress concentration caused by the spline. The upper graph shows the stress gradient in the body area when the
shaft is loaded to 110,000-psi shear stress. Also plotted on this graph are the shear yield-strength gradients
(converted from hardness) of the production alloy shaft, along with several experimental induction hardened
shafts. One would expect the through hardened shaft to have a surface-origin failure and to be lower in strength,
since its surface is the lowest hardness. Also, the high strength of the center ofthe shaft is essentially wasted, since
it is lowly stressed. Gradient strengthening by induction hardening provides a means of providing a better
strength match for the stress gradient. The lower graph gives the fatigue curves established for shafts having the
strength gradients shown in the upper graph.
Source: D. H. Breen and E. M. Wene, "Fatigue in Machines and Structures-Ground Vehicles," in Fatigue and Microstructure,
American Society for Metals. Metals Park OH. 1979, P 88
514
990
Carbo
of
95'0
0 Grou 0
II Group Q
900
Group
800
----
70'0
1750
1800
1900'
Quench
Sliding
..
~I
21
21
Slope
Cor r .
Coef.
.89
.97
.98
2.199
I. 463
2.961
vacuum
------
AJl
600
500
40'0 -
Data corrected to
= 400 ksi
Sc___
____
A= 0.5
30'0
200
~
a:
:>
:;;
.... 10'0
I-
Z
~
a:
~
e,
5'0
4'0
30
"0
Q
1'0
0'5
0'4
0'3
02
105
0"
106
4
6 7 8 9 ,
8
10
7
2
c v c i es
6 7 9 ,
6 7,19
S-N data, Weibull probability plot: Effect of three carburizing temperatures on surface fa-
tigue for carburized 8620" steel. All were slow cooled and reheated for quenching. This
technique improved fatigue characteristics compared with direct-quenched rollers.
Source: S. L Rice. "Pilling Resistance of Some High TemperatureCarburized Cases," in Source Book on Gear Design. Technology
and Performance. Maurice A, H, Howes, Ed" American Society for Metals, Metals Park OH, 1980, p 234
=...,
1-
.'0
OF
9"0
'0' 0 1-0=--Gc..r_o_u....
p_N
Sliding
\
Corr.
Coef.
r-
21
.90
1750
+f. ._ _r----1
.97
21
1800
.97
21
1900'
10.0 ~.u!!J"--"S'---'-"=--=----'-".:.-----=:..:..:.:...:..,~J_------+f
...- r _ - - - - _ j
70.0
Group P
~a"'cull
...umlll-.
_f_.=_"~-----_c,.-_r'--:_---__1
6&:~(--------------_:_
. .+-------__,f.;;;;:"M,./_------_I
'Ool---------------.'f--------r..-f---i~-------__I
4o01---------------;.t----------tl~-'":f_--------__I
Data Corrected to
3001--1>:c-~W9__1<!;i_-----.+------__::~+...,,'-----------_f
A = 0,5
200
. JOOj----------r--='---------if---------------_f
I-
U
0:
.. "0 - - - - - - - - f - : ; . ; o - - - - - - - - - - T - . H........---------------1
"0 t - - - - - r - - - - - - - - + - j ' - - - - - - - - - - - - - - - - - - - _ _ I
N
'-0
D"
0"
D"
03
0'\
,105
345678"
C 't'
c i es
4567."
45.71.
S-N data, Weibull probability factor: Effect ofthe three carburizing temperatures on
surface fatigue for carburized 8620H steel. All were direct quenched from the temperatures shown above.
Source: S. L. Rice, "Pitting Resistance of Some High Temperature Carburized Cases," in Source Book on Gear Design, Technology
and Performance, Maurice A. H. Howes, Ed.. American Society for Metals, Metals Park OH, 1980, P 233
515
516
III
OJ
s:
u
c .300
E
E
I
7
0
.250
en
:.::
<{
i=
a::
en
<{
.150
:.::
a::
u
;:!
iii
5
<.9
Z
(136MPa)
.200
a::
U
<.9
Z
i=
a::
.100
<{
I-
en
.050
1000
10,000
CYCLES TO
100,000
FAILURE. N f -
The variation of the fatigue life, NJ, with the starting flaw size a.; is shown in the
diagram above. The fatigue life increases dramatically at very small a i values. The
far curve shows that in the long life regime the final crack size has only a small effect
on Np This is because fatigue crack growth rates are very low at low ~Kvalues and
hence the greatest fraction offatigue life is spent at the smaller crack sizes. Since the
controlling parameter is ~K, low life for small crack sizes is possible at high cyclic
stresses. The second set of curves shows that doubling the cyclic stress range reduces
the fatigue life by about an order of magnitude. Also, if the starting ~K value is
high, the final crack size has a larger effect on the cyclic life. The above diagram
shows the importance of adjusting both the cyclic stress and starting flaw size to
optimize the fatigue life.
Source: Steel Castings Handbook. 5th Edition, Peter F. Weiser, Ed., Steel Founders' Society of America, Rocky River OH. 1980,
P 4-17
24-34. Notched Links, Hot Rolled Low-Carbon Steel: S-N Data for
Component Test Model
... - Component Tests
I. 0 r - - - - - - - - - - - . : . . . - - - - - - - - - - - ,
~:
~.., _
......
0.7
~~notched
~~~ ~/2Z7!?2Z
.~
0.6
_./Coml
---------_--=--~~~nent Tests
0.5
<, .l
~~
Local-Strain Model
--A
rr .............& ~1
......
04
.
0.3
S-N /
Model
R=-1
~@;/?/!?!Z2Z
0.2L..--------1------.........L~------J
7
loll
lOll
Fatigue Life. Nj cycles
10
COMPONENT TEST MODEL. The most straightforward life prediction model of a component is
developed from fatigue tests of the component itself. The component is cyclically loaded in a manner
that attempts to simulate service and the model is the plot of the test results. The cyclic load or nominal
net-section stress is plotted versus cycles to crack initiation N; or to failure Np
Example: Component fatigue tests were performed on the notched link ofthe previous examples; the
results are listed in the table and shown graphically above. The fatigue strength for crack initiation at
one million cycles is:
iJ.S
= 0.47
Sli
106 cycles
The three models-S-N, local strain, and component tests-are compared above, and the three
predictions for fatigue strength (iJ.S) at one million cycles are, respectively, 28, 42, and 47% of the
tensile strength Sli' The local strain prediction is closer to the behavior observed in the component test
than the S-N prediction.
Component Test Results, Hot-Rolled Carbon Steel Notched Link
Tensile
strength,
Yield
strength,
Cycles to
crack initiation,
Su,MPa
S}",MPa
N;
417
388
366
366
402
417
366
402
388
242
236
239
239
235
242
239
235
236
0.818
0.880
0.936
0.690
0.628
0.604
0.444
0.404
0.418
1.36 X
1.69 X
2.62 X
6.99 X
8.24 X
1.05 X
1.64 X
2.21 X
2.99 X
10'
10'
10'
10'
10'
105
10"
10"
106
Source: Harold S. Reemsnyder, "Constant Amplitude Fatigue Life Assessment Models," in Proceedings of the SAE Fatigue Conference P-I09, Society of Automotive Engineers. Inc .. Warrendale PA, 1982. P 127
517
518
.;;;
.:.!
tl' 120
~
:;;
E
::>
E 100
'x
'"
80
60
Annealed plate
(min)
-..;
10'
10'
m
10
10'
Number of cycles
Source: J. H. Moll. V. C. Petersen and E. J. Dulis, "Powder Metallurgy Parts for Aerospace Applications," in Powder MetallurgyApplications. Advantages and Limitations. Erhard Klar, Ed.. American Society for Metals. Metals Park OH, 1983, P 288