Atlas of 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 1
Introduction I Fatigue Crack Initiation 4 Fatigue Crack Propagation 12

SECTION 1: S-N Curves That Typify Effects of Major Variables
I-I. 1-2. 1-3. 1-4. 1-5. 1-6. 1-7. 1-8. 1-9. 1-10. I-II. 1-12. 1-13. 1-14. 1-15. 2-1. 2-2. 2-3. 2-4. 2-5. 2-6. 2-7. 2-8. 2-9. 2-10. 2-11. 2-12. 2-13. 2-14. 2-15. 2-16. 2-17. 2-18. 2-19. 2-20. 2-21. 2-22. 2-23. 2-24. 2-25.

27

S-NCurves Typical for Steel 27 S-NCurves Typical for Medium-Strength Steels 28 S-NDiagrams Comparing Endurance Limit for Seven Alloys 30 Steel: Effect of Microstructure 31 Steel: Influence of Derating Factors on Fatigue Characteristics 32 Steel: Correction Factors for Various Surface Conditions 33 Fatigue Behavior: Ferrous vs Nonferrous Metals 34 Comparison of Fatigue Characteristics: Mild Steel vs Aluminum Alloy 35 Carbon Steel: Effect of Lead as an Additive 36 Corrosion Fatigue: General Effect on Behavior 37 Effect of Corrosion on Fatigue Characteristics of Several Steels 38 Steel: Effect of Hydrogen on Fatigue Crack Propagation 39 Relationship of Stress Amplitude and Cycles to Failure 40 Strain-Life and Stress-Life Curves 41 Fatigue Plot for Steel: Ultrasonic Attenuation vs Number of Cycles 42

SECTION 2: Low-Carbon Steels: Flat-Rolled, Weldments and Tubes

43

Typical S-N Curve for Low-Carbon Steel Under Axial Tension 43 AISI 1006: Effects of Biaxial Stretching and Cold Rolling 44 AISI 1006: Weldment; FCAW, TIG Dressed 45 AISI 1006: Weldment; Shear Joints 46 AISI 1006: Weldment; Lap-Shear Joints 47 AISI 1015: Effect of Cold Working 48 A533 Steel Plate: Fatigue Crack Growth Rate 49 A514F Steel Plate: Fatigue Crack Growth Rates 50 A514F and A633C: Variation in Fatigue Crack Growth Rate With Orientation 51 A514F: Scatterbands of Fatigue Crack Growth Rate 52 A633C Steel Plate: Scatterbands of Fatigue Crack Growth Rates 53 Low-Carbon Steel Weldment: Effects of Various Weld Defects 54 Low-Carbon Steel Weldment: Effect of Weld Reinforcement and Lack of Inclusions 55 Low-Carbon Steel Weldment: Effect of Weld Reinforcement and Lack of Penetration 56 Low-Carbon Steel Weldment: Computed Fatigue Strength; Weldment Contained Lack of Fusion 57 Low-Carbon Steel Weldment: Effect of Reinforcement and Undercutting 58 Low-Carbon Steel: Transverse Butt Welds; Effect of Reinforcement 59 A36/E60S-3 Steel Plate: Butt Welds 60 A514F/EllO Steel: Bead on Plate Weldment 61 A36 and A514 Steel Plates: Butt Welded 62 A36 Plate Steel: Butt Welded 63 Low-Carbon Steel Tubes: Effect of Welding Technique 64 Low-Carbon Steel: Effect of Applied Anodic Currents in 3% NaCI 65 Low-Carbon Steel: Effect of pH in NaCI and NaOH 66 Low-Carbon Steel: Effect of Carburization and Decarburization 67

v

VI

Contents

2-26. A514B Steel: Effect of Various Gaseous Environments on Fatigue Crack Propagation 68 2-27. Cast 1522 and 1541 Steels: Effect of Various Surface Conditions 69 2-28. Cast A216 (Grade WCC) Steel: Fatigue Crack Growth Rate 70

SECTION 3: Medium-Carbon Steels, Wrought and Cast
3-1. 3-2. 3-3. 3-4. 3-5. 3-6. 3-7. 3-8. 3-9. 3-10. 3-11. 3-12. 4-1. 4-2. 4-3. 4-4. 4-5. 4-6. 4-7. 4-8. 4-9. 4-10. 4-11. 4-12. 4-13. 4-14. 4-15. 4-16. 4-17. 4-18. 4-19. 4-20. 4-21. 4-22. 4-23. 4-24. 4-25. 4-26. 4-27. 4-28. 4-29. 4-30. 4-31. 4-32. 4-33. 4-34. 4-35. 4-36. 4-37.

71

AISI 1030 (Cast) Compared With AISI 1020 (Wrought) 71 AISI 1035: Effect of Gas and Salt Bath Nitriding 72 AISI 1040: Cast vs Wrought 73 AISI 1045: Relationship of Hardness and Strain-Life Behavior 74 AISI 1141: Effect of Gas Nitriding 75 Medium-Carbon Steels: Interrelationship of Hardness, Strain Life and Fatigue Life 76 Medium-Carbon Steel: Effect of Fillet Radii 77 Medium-Carbon Steel: Effect of Keyway Design 78 Medium-Carbon Steel: Effect of Residual Stresses 79 Medium-Carbon Cast Steel: Effect of Changes in Residual Stress 80 Medium-Carbon Cast Steel: S-NProjection (Effect of Applied Stress) 81 Medium-Carbon Cast Steel: Effect of Applied Stress (Shot Blasting) 82

SECTION 4: Alloy Steels: Low- to High-Carbon, Inclusive

83

Medium-Carbon Alloy Steels, Five Grades: Effect of Martensite Content 83 Medium-Carbon Alloy Steels, Six Grades: Hardness vs Endurance Limit 84 Medium-Carbon Alloy Steels: Effect of Specimen Orientation 85 4027 Steel: Carburized vs Uncarburized 86 4120 Steel: Effect of Surface Treatment in Hydrogen Environment 87 4120 Steel: Effect of Surface Treatment in Hydrogen Environment 88 4120 Steel: Effect of Various Surface Treatments on Fatigue Characteristics in Air vs Hydrogen 89 4130 Steel: Fatigue Crack Growth Rate vs Temperature in Hydrogen 90 4135 and 4140 Steels: Cast vs Wrought 91 4135 and 4140 Steels: Cast vs Wrought 92 4140,4053 and 4063 Steels: Effect of Carbon Content and Hardness 93 4140 Steel: Effect of Direction on Fatigue Crack Propagation 94 4140 Steel: Effect of Cathodic Polarization 95 Cast 4330 Steel: Effects of Various Surface Conditions 96 4340 Steel: Scatter of Fatigue Limit Data 97 4340 Steel: Strength vs Fatigue Life 98 4340 Steel: Total Strain vs Fatigue Life 99 4340 Steel: Stress Amplitude vs Number of Reversals 100 4340 Steel: Effect of Periodic Overstrain 101 4340 Steel: Estimation of Constant Life 102 4340 Steel: Effect of Strength Level on Constant-Life Behavior 103 4340 Steel: Notched vs Unnotched Specimens 104 4340 Steel: Effect of Decarburization 105 4340H Steel: Effect of Inclusion Size 106 4340 Steel: Influence of Inclusion Size 107 4340 Steel: Effect of Hydrogenation; Static Fatigue 108 4340 Steel: Effect of Hydrogen 109 4340 Steel: Effect of Nitriding 110 4340 Steel: Effect of Nitriding and Shot Peening III 4340 Steel: Effect of Induction Hardening and Nitriding 112 4340 Steel: Effect of Surface Coatings 113 4340 Steel: Effect of Temperature on Constant-Lifetime Behavior 114 4520H Steel: Effect of Type of Quench 115 4520H Steel: Effect of Shot Peening 116 4620 Steel: Effect of Nitriding 117 4620 Steel: P/M-Forged 118 4620 Steel: P/M-Forged at Different Levels 119

Contents

VII

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.

4625 Steel: P/M vs Ingot Forms 120 4640 Steel: P/M-Forged 121 High-Carbon Steel (Eutectoid Carbon): Pearlite vs Spheroidite 122 52100 EF Steel: Surface Fatigue; Effect of Finish and Additives 123 124 52100 EF Steel: Surface Fatigue; Effect of Surface Finish and Speed 52100 EF Steel: Surface Fatigue; Effect of Lubricant Additives 125 52100 EF Steel: Surface Fatigue; Effect of Lubricant Viscosity, Slip Ratio and Speed 126 52100 EF Steel: Rolling Ball Fatigue; Effect of Oil Additives 127 52100 Steel: Carburized vs Uncarburized 128 8620H Steel: Carburized; Results From Case and Core 129 8620H Steel: Effect of Variation in Carburizing Treatments 130 8620 Steel: Effect of Nitriding 131 8622 Steel: Effect of Grinding 132 Cast 8630 Steel: Goodman Diagram for Bending Fatigue 133 Cast 8630 Steel: Effect of Shrinkage 134 Cast 8630 Steel: Effect of Shrinkage on Torsion Fatigue 135 Cast 8630 Steel: Effect of Shrinkage on Torsion Fatigue 136 Cast 8630 Steel: Effect of Shrinkage on Plate Bending 137 Cast 8630 vs Wrought 8640 138 8630 and 8640 Steels: Effect of Notches on Cast and Wrought Specimens 139 Nitralloy 135 Steel: Effect of Nitriding 140 AMS 6475: Effects of Welding 141 Medium-Carbon, ICr-Mo-V Steel Forging: Effect of Cycling Frequency 142 EM 12 Steel: Effect of Temperature on Low-Cycle Fatigue 143 Cast 0.5Cr-Mo-V Steel: Effects of Dwell Time in Elevated-Temperature Testing 144 Cast 0.5Cr-Mo-V Steel: Effect of Environment at 550°C (1022 OF) 145 Cast C-0.5Mo Steel: Effect of Temperature and Dwell Period on Cyclic Endurance at Various Strain Amplitudes 146

SECTION 5: HSLA Steels
5-1. 5-2. 5-3. 5-4. 5-5. 5-6. 5-7. 5-8. 5-9. 5-10. 5-11. 5-12. 5-13. 5-14.
5-15.

147

5-16. 5-17. 5-18.

HI-FORM 50 Steel vs 1006 147 HI-FORM 50 Steel vs 1006: Stress Response 148 HI-FORM 50 Steel Compared With 1006, DPI and DP2 149 HSLA vs Mild Steel: Torsional Fatigue 150 Proprietary HSLA Steel vs ASTM A440 151 Comparison of HSLA Steel Grades BE, JF and KF for Plastic Strain Amplitude vs Reversals to Failure 152 Comparison of HSLA Steel Grades BE, JF and KF for Total Strain Amplitude vs Reversals to Failure 153 Comparison of a Dual-Phase HSLA Steel Grade With HI-FORM 50: Total Strain Amplitude vs Reversals to Failure 154 AISI 50 XF Steel: Effects of Cold Deformation 155 AISI 80 DF Steel: Effects of Cold Deformation 156 Comparison of Three HSLA Steel Grades, Cb, Cb-V and Cb-V-Si: Strain Life From Constant Amplitude 157 Comparison of Stress Responses: DPI vs DP2 Dual-Phase HSLA Steels 158 Dual-Phase HSLA Steel Grade: Stress Response for As-Received vs Water-Quenched 159 Dual-Phase HSLA Steel Grade: Stress Response for As-Received vs Gas-JetCooled 160 S-N Comparison of Dual-Phase HSLA Steel Grades DPI and DP2 With 1006 161 Comparison of Dual-Phase HSLA Steel DP2 With HI-FORM 50 162 Comparison of Cyclic Strain Response Curves for Cb, Cb-V and Cb-V-Si Grades of HSLA Steel 163 Fatigue Crack Propagation Rate: Effect of Temperature for Two HSLA Steel Grades 164

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

SECTION 6: High-Strength Alloy Steels
6-1. 6-2. 6-3. 6-4. 6-5. 6-6. 6-7.

176

HY-130 Steel: Effect of Notch Radii 176 300 M Steel: Effect of Notch Severity on Constant-Lifetime Behavior 177 TRIP Steels Compared With Other High-Strength Grades 178 Corrosion Fatigue: Special High-Strength Sucker-Rod Material 179 Corrosion Fatigue Cracking of Sucker-Rod Material 180 181 Hydrogenated Steel: Effect of Baking Time on Hydrogen Concentration Hydrogenated Steel: Effect of Notch Sharpness 182

SECTION 7: Heat-Resisting Steels

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 540°C (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

SECTION 8: Stainless Steels

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 650°C (1200 OF) 204

Contents

ix

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

x
SECTION 9: Maraging Steels

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

SECTION 10: Cast Irons
10-1. 10-2. 10-3. 10-4. 10-5. 10-6. 10-7. 10-8. 10-9. 10-10. 10-11. 10-12. 10-13. 10-14. 10-15. 10-16. 10-17. 10-18. 10-19. 10-20. 10-21. 10-22. 10-23. 10-24. 10-25. 10-26. 10-27.

245

Fatigue of Cast Irons as a Function of Structure-Sensitive Parameters 245 Gray Iron: Fatigue Life, and Fatigue Limit as a Function of Temperature 246 Gray Iron: S-N Curves for Unalloyed vs Alloyed 247 Gray Iron: Effect of Environment 248 Class 30 Gray Iron: Modified Goodman Diagram for Class 30 249 Class 30 Gray Iron: Fatigue Crack Growth Rates for Class 30 250 Gray Irons: Torsional Fatigue for Various Tensile Strength Values 251 Gray Irons: Torsional Fatigue Data for Five Different Compositions 252 Gray Irons: Thermal Fatigue-Effect of Aluminum Additions 253 Gray Irons: Thermal Fatigue-Effect of Chromium and Molybdenum Additions 254 Gray Irons: Thermal Fatigue-Room Temperature and 540°C (1000 OF) 255 Gray Irons: Thermal Fatigue Properties-Comparisons With Ductile Cast Iron and Carbon Steel 256 Cast Irons: Thermal Fatigue Properties for Six Grades 257 Ductile Iron: Effect of Microstructure on Endurance Ratio-Tensile Strength Relationship 258 Ductile Iron: Effect of Microstructure on Endurance Ratio-Tensile Strength Relationship 259 Ductile Iron: S-N Curves for Ferritic and Pearlitic Grades, Using V-Notched Specimens 260 Ductile Iron: S-N Curves for Ferritic and Pearlitic Grades, Using Unnotched Specimens 261 Ductile Iron: Fatigue Diagrams for Bending Stresses and Tension-Compression Stresses 262 Ductile Iron: Effect of Surface Conditions-As-Cast vs Polished Surface 263 Ductile Iron: Fatigue Limit in Rotary Bending as Related to Hardness 264 Ductile Iron: Effect of Rolling on Fatigue Characteristics 265 Ductile Iron: Effect of Notches on a 65,800-psi-Tensile-Strength Grade 266 Ductile Iron: Fatigue Crack Growth Rate Compared With That of Steel 267 Malleable Iron: S- N Curve Comparisons of Four Grades 268 Pearlitic Malleable Iron: Effect of Surface Conditions on S-N Curves 269 Pearlitic Malleable Iron: Effect of Nitriding 270 Ferritic Malleable Iron: Effect of Notch Radius and Depth 271

SECTION 11: Heat-Resisting Alloys
II-I. 11-2. 11-3. 11-4. 11-5. 11-6. 11-7. 11-8. 11-9. 11-10. II-II. 11-12. 11-13. 11-14.

272

A286: Effect of Environment 272 A286: Effect of Frequency on Life at 593°C (1095 OF) 273 A286: Fatigue Crack Growth Rates at Room and Elevated Temperatures 274 Astroloy: S-N Curves for Powder vs Conventional Forgings 275 Astroloy: Powder vs Conventional Forgings Tested at 705°C (1300 OF) 276 FSX-430: Effect of Grain Size on Cycles to Cracking 277 FSX-430: Effect of Grain Size on Fatigue Crack Propagation Rate 278 HS-31: Effect of Testing Temperature 279 IN 738 LC Casting Alloy: Standard vs HIP'd Material 280 IN 738 LC: Effect of Grain Size on Cycles to Failure 281 IN 738 LC: Effect of Grain Size on Cycles to Cracking 282 IN 738 LC: Effect of Grain Size on Fatigue Crack Propagation Rate 283 IN 738 LC: Fatigue Crack Growth Rate at 850°C (1560 OF) 284 Inconel 550: Axial Tensile Fatigue Properties in Air and Vacuum at 1090 K 285

Contents

xi

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.

Inconel625: Effect of Temperature on Cycles to Failure 286 Inconel 706: Effect of Temperature on Fatigue Crack Growth Rate 287 Inconel "7I3C": Effect of Elevated Temperatures on Fatigue Characteristics 288 Inconel "7I3C" and As-Cast HS-31: Comparison of Two Alloys for Number of Cycles in Thermal Fatigue to Initiate Cracks 289 Inconel 718: Effect of Frequency on Fatigue Crack Propagation Rate 290 Inconel 718: Relationship of Fatigue Crack Propagation Rate With Stress Intensity 29I Inconel 718: Relationship of Fatigue Crack Growth Rate With Load/Time Waveforms 292 Inconel 718: Fatigue Crack Growth Rate in Air vs Helium 293 Inconel 718: Effect of Environment on Fatigue Crack Growth Rate 294 Inconel 718: Fatigue Crack Growth Rate in Air Plus 5% Sulfur Dioxide 295 lnconel 7I8: Fatigue Crack Growth Rate in Air at Room Temperature 296 Inconel 718: Fatigue Crack Growth Rate in Air at 316°C (600 OF) 297 Inconel 718: Fatigue Crack Growth Rate in Air at 427°C (800 OF) 298 Inconel 718: Fatigue Crack Growth Rate in Air at 538°C (1000 OF) 299 Inconel 718: Fatigue Crack Growth Rate in Air at 649°C (1200 OF) 300 Inconel 718: Fatigue Crack Growth Rates at Cryogenic Temperatures 301 Inconel 718 and X-750: Fatigue Crack Growth Rates at Cryogenic Temperatures 302 Inconel X-750: Effect of Temperature on Fatigue Crack Growth Rates 303 Jethete M I52: Interrelationship of Tempering Treatment, Alloy Class, and Testing Temperature With Fatigue Characteristics 304 Lapelloy: Interrelationship of Hardness and Strength With Fatigue Characteristics 305 MAR-M200: Effect of Atmosphere on Cycles to Failure 306 MAR-M509: Correlation of Initial Crack Propagation and Dendrite Arm Spacing 307 MAR-M509: Correlation Between Number of Cycles Required to Initiate a Crack and Dendrite Arm Spacing 308 MERL 76, P/M: Axial Low-Cycle Fatigue Life of As-HIP'd Alloy at 540°C (1000 OF) 309 Nickel-Base Alloys: Effect of Solidification Conditions on Cycles to Onset of Cracking 310 Rene 95 (As-HIP): Cyclic Crack Growth Behavior Under Continuous and HoldTime Conditions 3I I Rene 95: Effect of Temperature on Fatigue Crack Growth Rate 312 313 S-8 I6: Effect of Notches on Cycles to Failure at 900°C (1650 ° F) Udimet 700: Fatigue Crack Growth Rates at 850°C (1560 OF) 314 U-700 and MAR-M200: Comparison of Fatigue Properties 315 Waspaloy: Stress-Response Curves 316 X-40: Effect of Grain Size and Temperature on Fatigue Characteristics 317 Cast Heat-Resisting Alloys: Ranking for Resistance to Thermal Fatigue 318

SECTION 12: Aluminum Alloys

319

12-1. Corrosion-Fatigue Properties of Aluminum Alloys Compared With Those of Other Alloys 319 12-2. Comparisons of Aluminum Alloys With Magnesium and Steel: Tensile Strength vs Endurance Limit 320 12-3. Aluminum Alloys (General): Yield Strength vs Fatigue Strength 321 12-4. Comparison of Aluminum Alloy Grades for Crack Propagation Rate 322 12-5. Alloy 1100: Relationship of Fatigue Cycles and Hardness for HO and H 14 Tempers 323 12-6. Alloy 1100: Interrelationship of Fatigue Cycles, Acoustic Harmonic Generation and Hardness 324 12-7. Alloy 2014-T6: Notched vs Unnotched Specimens; Effect on Cycles to Failure 325 12-8. Alloy 2024-T3: Effect of Air vs Vacuum Environments on Cycles to Failure 326 12-9. Alloy 2024-T4 Alclad Sheet: Effect of Bending on Cycles to Failure 327

Alloy 7075-T6: Effect of R-Ratio on Fatigue Crack Propagation 12-46. Alloy 7075-T65I: Fatigue Life as Related to Harmonic Generation 12-49. Alloy 7075-T6: Effects of Corrosion and Pre-Corrosion 359 360 12-42. Alloys 7075-T6 and 7475-T73: Effect of Laser-Shock Treatment on Fatigue Properties 368 12-50. Alloy 2048-T851: Fatigue Crack Propagation Rates in LT and TL Orientations 339 340 12-22. Alloy 7075: Effect of TMT on Cycles to Failure 12-39. Alloy 2024-T4: High-Cycle vs Low-Cycle Fatigue 328 12-11. 5154-H34 and 6061-T6: Effect of Alloy on Fatigue Characteristics of Weldments 345 12-28. 7XXX Alloys: Cyclic Strain vs Crack Initiation Life 351 12-34.xii Contents 12-10.5% NaCl Environment on Cycles to Failure 12-43. Alloy 7075-T6: Effect of Surface Treatments and Notch Designs on Number of Cycles to Failure 362 364 12-45. 5086-H36. Alloys 7075 and 7050: Relative Ranking for Constant Amplitude and Periodic Overload 357 12-40. Alloy 2024-T4: Dependence of the Average Rocking Curve Halfwidth 7J on Distance From the Surface 330 12-13. Alloy 2219-T851: Probability of Fatigue Failure 344 12-27. Alloys 5083-0/5183: Fatigue Life Predictions and Experimental Data Results for Double V-Butt Welds 349 12-32. Alloys 2024 and X2024: Effect of Alloy Purity on Cycles to Failure 331 12-14. Alloys 2024-T4 and 7075-T6: Effect of Product Form and Notches 335 12-18. Alloy 2219-T851: Effect of Strain Amplitude on the Relaxation of Residual Surface Stress With Fatigue 342 12-25. Alloy 2219-T851: Dependence of Relaxation Behavior on the Cyclic Hardening Parameter 341 12-24. Alloy 7075-T73: Effect of a 3. Alloy 7075-T6: Effect of Laser-Shock Treatment on Hi-Lok Joints 369 . Alloys 5086-H34. Alloys 3003-0. Alloys 2024 and 2124: Relationship of Particle Size and Fatigue Characteristics 332 12-15. Alloy 7050: Influence of Alloy Composition and Dispersoid Effect on Mean Calculated Fatigue Life 352 353 12-35. Alloy 7075: Effect of Predeformation on Fatigue Crack Propagation Rates 365 12-47. Alloy 2048-T85I: Modified Goodman Diagram for Axial Fatigue 12-23. Alloys 7075 and 7475: Effect of Inclusion Density on Cycles to Failure 355 356 12-38. Alloy 5083-0 Plate: Effect of Temperature and Humidity on Fatigue Crack Growth Rates 347 12-30. Alloy 7050: Effect of Grain Shape on Cycles to Failure 12-36. Alloy 5083-0 Plate: Effect of Orientation on Fatigue Crack Growth Rates 346 12-29. Alloy 2048-T851: Longitudinal vs Transverse for Axial Fatigue 337 12-20. Alloy 2219-T851: Relationship of Fatigue Cycles to Different Depth Distributions of Surface Stress 343 12-26. Alloys 7075 and 2024-T3: Comparative Fatigue Crack Growth Rates for Two Alloys in Varying Humidity 366 367 12-48. Alloys 2024-TJ and 7075-T6: Summary of Fatigue Crack Growth Rates 334 12-17. Alloys 5083-0/5183: Predicted Effect of Stress Relief and Stress Ratio on Fatigue Life of Butt Welds 350 12-33. 6061-T6. Alloy 2024-T4: Relationship of Stress and Fatigue Cycles 329 12-12. 7075-T73 and 2024-T3: Comparative Resistance to Axial-Stress Fatigue 348 12-31. Alloys 2024-T4 and 2124-T4: Comparison of Resistance to Fatigue Crack Initiation 333 12-16. Alloy 7075: Effect of Environment and Mode of Loading 358 12-41. Alloys 2024-T351 and 7075-T73XXX: Comparison of P / M Extrusions and Rod 336 12-19. Alloy 7075: Effect of Cathodic Polarization on Fatigue Behavior 361 12-44. Alloy 2048-T851: Notched vs Unnotched Specimens at Room and Elevated Temperatures 338 12-21. T6 and T651): Effect of Thermomechanical Processing on Cycles to Failure 354 12-37. Alloy 7075 (TMP.

12-60. 12-67. Effects of Notches and Testing Temperature 408 14-3. 14-55. 12-58. 12-56. Effect on Reversals to Failure 395 SECTION 13: Copper Alloys 396 13-1. Copper Alloys C87500 and C87800 (Silicon Brasses): S-N Curves. 12-71.Contents xiii 12-51. Mg-AI-Zn Casting Alloys: Effects of Surface Conditions on Fatigue Properties 409 . Magnesium Casting Alloy QE22A-T6: Effects of Notches and Testing Temperature 407 14-2. 12-59. Copper Alloy C92200 (Navy "M" Bronze): S-N Curves. 12-73. Copper: Effect of Air and Water Vapor on Cycles to Failure 396 13-2. Scatter Band 400 13-6. 12-63. Scatter Band 403 13-9. Scatter Band 402 13-8. 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. 192: Effect of Salt Spray on Tubes 405 13-1 I. 12-57. 12-54. Copper: Applied Plastic-Strain Amplitude vs Fatigue Life 397 13-3.N Curves. Copper Alloy C83600 (Leaded Red Brass): S-N Curves. Copper Alloy No. 12-70. 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. 12-69. 12-65. Copper Alloy C86500 (Manganese Bronze): S-N Curves. Scatter Band 401 13-7. 12-53. 12-68. 12-61. 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. Copper Alloy CI 1000 (ETP Wire): Effect of Temperature on Fatigue Strength 398 13-4. 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. 12-52. Copper Alloy C93700 (High-Leaded Tin Bronze): S-NCurves. Copper Alloy C26000 (Cartridge Brass): Influence of Grain Size and Cold Work on Cycles to Failure 399 13-5. 12-76. 12-72. 12-62. 12-74. Copper Alloy 955: Goodman-Type Diagram 406 SECTION 14: Magnesium Alloys 407 14-1.5Ag: 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. 12-66. Scatter Band 404 13-10. 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. Magnesium Casting Alloy QH2 IA-T6: S. 12-64. 12-75.

Grade 4: S-N Curves for Three Testing Temperatures 415 17-3. Unalloyed Titanium. Grade 3: S-N Curves for Annealed vs Cold Rolled 414 17-2. Ti-5AI-2. Ti-6AI-4V: Fatigue Crack Growth Rates 446 17-34. Babbitt: Variation of Bearing Life With Babbitt Thickness 412 16-3. Ti-8Mo-2Fe-3AI: S-NCurves. Ti-6AI-4V: Effect of Texture and Environment on Cycles to Failure 443 17-31. Ti-6AI-6V-2Sn (HIP): S-N Curves for Annealed Plate vs HIP 421 17-9. Ti-6AI-4V: Relative Effects of Machining and Grinding Operations on Endurance Limit 439 17-27. Ti-6AI-4V: Effect of Texture on Fatigue Strength 441 17-29. Molybdenum: Fatigue Limit Ratio vs Temperature 411 16-1. Ti-6AI-4V: Effect of Dwell Time on Fatigue Crack Growth Rates 448 17-36. Ti-6AI-4V: Effect of Condition and Notches on Fatigue Characteristics 430 17-18. Ti-6AI-4V: Fatigue Crack Growth Rates for ISR Tee. and Extrusions 445 17-33. Ti-6AI-6V-2Sn (HIP): S-N Curves for Titanium Alloy Powder Consolidated by HIP 420 17-8. Ti-6AI-4V: Effect of Fretting and Temperature on Cycles to Failure 434 17-22. Tin-Lead Soldering Alloy: S-N Data for Soldered Joints 411 16-2. Ti-6AI-6V-2Sn: Effects of Machining and Grinding 419 17-7. 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-IOV-2Fe-3AI: S-N Curve. Ti-6AI-4V: Effect of Yield Strength on Stress-Life Behavior 436 17-24. Ti-6AI-2Sn-4Zr-2Mo: Bar Chart Presentation on Effects of Machining and Grinding 422 17-10. Ti-6AI-4V: Comparison of Wrought vs Isostatically Pressed Material for Cycles to Failure 433 17-21. Ti-6AI-4V: Fatigue Crack Growth Data 449 17-37. Ti-IOV-2Fe-3AI and Ti-6AI-4V: Comparison of Fatigue Crack Growth Rates 427 17-15.xiv SECTION 15: Molybdenum SECTION 16: Tin Alloys Contents 410 410 15-1. Ti-I3V-IICr-3AI: Constant-Life Fatigue Diagrams 429 17-17. Ti-24V and Ti-32V: Stress Amplitude vs Cycles to Failure 416 17-4. Ti-6AI-4V P / M: Comparison of HIP'd Material With Alpha-Beta Forgings for Cycles to Failure 450 . SAEI2 Bearing Alloy: Effect of Temperature on Fatigue Life 413 SECTION 17: Titanium and Titanium Alloys 414 17-1.5Sn: Effects of Notches and Types of Surface Finish 417 17-5. Notched Bar Fatigue Life for a Series of Forgings Compared With Ti-6AI-4V Plate 428 17-16. 'Solution Treated and Aged Condition 425 17-13. Ti-6AI-4V: Fatigue Crack Growth Rates 444 17-32.5Sn and Ti-6AI-4V: Fatigue Crack Growth Rates 418 17-6. Ti-6AI-4V: Effect of Stress Relief on Cycles to Failure 437 17-25. Ti-6AI-4V (Beta Rolled): Effect of Finishing Operations on Cycles to Failure 435 17-23. Ti-6AI-4V: Interrelationship of Machining Practice and Cutting Fluids on Cycles to Failure 438 17-26. Ti-6AI-4V: Effect of Final Cooling on Fatigue Crack Growth Rates 447 17-35. Notched vs Unnotched Specimens in Axial Fatigue 426 17-14. Ti-5AI-2. Unalloyed Titanium. Ti-6AI-4V: Effects of Various Metal Removal Operations on Endurance Limit 440 17-28. Ti-IOV-2Fe-3AI: S-N Curves. Ti-6AI-4V: Effect of Isothermally Rolled vs Extruded Material on Cycles to Failure 432 17-20. Ti-6AI-4V: Effect of Complex Texture on Cycles to Failure 422 17-30. Ti-6AI-4V: Effect of Direction on Endurance 431 17-19.

PI M (Nickel Steels): Relation Between Fatigue Limit and Tensile Strength for Sintered Steels 472 21-5.Contents xv 17-38.) 469 470 21-2. Steel Castings (General): Effects of Discontinuities on Fatigue Characteristics 467 SECTION 20: Closed-Die Forgings 468 468 (See also under specific grades of alloys.) 19-1. Ti-6AI-4V: Comparison of Fatigue Crack Growth Rate. on carbon and alloy steels. Closed-Die Steel Forgings: Effect of Surface Condition on Fatigue Limit SECTION 21: Powder Metallurgy Parts 469 (See also under specific alloys. No Copper): Effect of Density and Nitriding on Cycles to Failure 476 21-9. PI M (Sintered Iron. P/M (Low-Carbon. Ti-6AI-4V: Comparison of Specimens Processed by Various Fabrication Processes for Cycles to Failure 455 456 17-43. P/M: Effect of Nitriding on Ductile Iron and Sintered Iron (3%Cu) for Cycles to Failure 477 21-1. Ti-6AI-4V EB Weldments: Effects of Porosity on Cycles to Failure 460 17-48. 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 P/M: Effect of Powder Mesh Size on Fatigue Properties 17-40. Ti-6AI-4V PI M: Comparisons of HIP'd Material With Annealed Plate for Cycles to Failure 45 I 452 17-39. Low-Carbon. PI M: Relation of Fatigue Limit to Tensile Strength for Sintered Steels 21-3. PI M vs II M 17-44. 1-5%Cu): Effects of Notches and Nitriding on Cycles to Failure 475 2 I-8. Steel Castings (General): Effect of Design and Welding Practice on Fatigue Characteristics 466 19-2. Ti-6AI-4V: S-N Diagram for Laser-Welded Sheet 464 17-51. P/M: Relation of Density to Fatigue Limit and Fatigue Ratio SECTION 22: Composites 478 22-1. P/M (Nickel Steels): Effect of Notches on Cycles to Failure for the As-Sintered Condition 473 21-6. Zirconium 702: Effects of Notches and Testing Temperature on Cycles to Failure 465 SECTION 19: Steel Castings 466 (For other data on steel castings see Sections 3. PI M (Nickel Steels): Effect of Notches on Cycles to Failure for the Quenched and Tempered Condition 474 21-7.) 20-1. Brass/ Mild Steel Composite: Comparison of Brass-Clad Mild Steel With Brass and Mild Steel for Cycles to Failure 478 22-2. Ti-6AI-4V: Base Metal vs SSEB-Welded Material for Cycles to Failure 458 17-46.4 and 5. Ti-6AI-4V P/M: Comparison of Blended Elemental. Ti-6AI-4V: Unwelded vs Electron Beam Welded Material for Cycles to Failure 462 463 17-50. Ti-6AI-4V Gas Metal-Arc Weldments: Effects of Porosity on Cycles to Failure 461 17-49. Ti-6AI-4V EB Weldments: Base Metal Compared With Flawless Weldments 459 17-47. PI M (Nickel Steels): As-Sintered vs Quenched and Tempered for Cycles to Failure 471 2 I-4. Ti-6AI-4V (Cast): S-N Diagram for Notched Specimens SECTION 18: Zirconium 465 18-1. Ti-6AI-4V: Base Metal vs SSEB-Welded Material for Cycles to Failure 457 17-45. Stainless Steell Mild Steel Composite: Comparison of Stainless-Clad Mild Steel With Stainless Steel and Mild Steel for Cycles to Failure 479 .

Gears. Carbon-Manganese Steel: Effects of Nickel Coating on Fatigue Strength 483 SECTION 24: Test Results for Component Parts 484 24-1. Gears. Stress vs Cycles to Rupture 498 24-16. 8620H Carburized: T-N Curves for Six-Pinion. Carburized Low-Carbon Steel: Relation of Life Factor to Required Life 491 24-9. 42 CrMo4 (German Specification): Endurance Test Results in the Weibull Distribution Diagram 509 24-27. Carburized Low-Carbon Steel: Probability-Stress-Life Design Curves 494 24-12. Coil Springs. 8650 and 8660 Steels: Relation of Design Stresses and Probability of Failure 487 . Coil Springs: Effects of Shot Peening on Cycles to Failure 486 24-4. Four-Square Tests 497 24-15. Front Suspension Torsion Bar Springs. Gears. AMS 6265: S-N Data for Cut vs Forged 503 24-21. Gears. Carbon and Alloy Steels (Six Grades): Effects of Nitriding on Fatigue Strength 482 23-4. 8620H Carburized: Bending or Contact Stress vs Cycles to Fracture or Pitting 495 24-13. Gears and Pinions: PIM Grades 4600V and 2000 vs 4615. Coil Springs: S-N Data for Oil-Tempered and Music Wire Grades 485 24-3. Gears. Percent Failure vs Time 506 24-24. 5160H Steel: Distribution of Fatigue Results for Simulated Service Testing 490 24-8. 8620H: S-N Data for Cut vs Forged 504 24-22. Hypoid. 1040 and 4037 Steels: Maximum Bending Stress vs Number of Stress Cycles 510 24-28. Gears and Pinions: PIM 4600V vs 4615. Carbon and Alloy Steels (Seven Grades): Effects of Nitrocarburizing on Fatigue Strength 480 23-2. Zero I and Spiral Bevel Gears. Weibull Distributions 505 24-23. 8620H Case Hardened: Relation of Life Factor to Cycles to Rupture 501 24-19. 24-5. Coil Springs. Gears. 1541 and 50B54 Steels: S-N Data for Induction Hardening vs Through Hardening 513 24-31. Carbon and Alloy Steels (Seven Grades): Effects of Tufftriding on Fatigue Characteristics 481 23-3. Steel Rollers. Power Shafts. Bevel Gears. Gears. Hypoid Gears. 8620H Carburized: Minimum Confidence Level. Carburized Low-Carbon Steel: Effect of Shot Peening on Cycles to Failure 493 24-11. HSLA Steels: Effects of Corrosion on Cycles to Failure 488 24-6. Carburized Low-Carbon Steel: Bending Stress vs Cycles to Failure 492 24-10. 1046. 8620H Carburized: S-N Scatter Band and Minimum Confidence Level 500 24-18. Spur Gears. Coil Springs. Gears. Bolts. Spiral Bevel and Zero I Bevel Gears. 5160 Steel: Maximum Applied Stress vs Cycles to Failure 489 24-7. Axle Shafts. Music Wire (Six Sizes): Data Presented by Means of a Goodman Diagram 484 24-2. 42 CrMo4 (German Specification): S-N Curves for Various Profiles 508 24-26. AMS 6382 and AMS 6260: Electron Beam Welded vs Silver Brazed Joints 512 24-30. 8620H Carburized: S-NScatter Band and Minimum Confidence Level 499 24-17. Low-Carbon Steel Case Hardened: Relation of Life Factor to Cycles to Rupture for Various Confidence Levels 502 24-20. Leaf Springs. Gears.xvi Contents SECTION 23: Effects of Surface Treatments 480 23-1. 8620H Carburized: A Weibull Analysis of Bending Fatigue Data 496 24-14. Bolts: S-N Data for Roll Threading Before and After Heat Treatment 511 24-29. Gears. Gears. Gear Steel AMS 6265: Parent Metal vs Electron Beam Welded 507 24-25. 8620H Carburized: Effects of Carburizing Temperature and Quenching Practice on Surface Fatigue 514 .

8620H Carburized: Effects of Carburizing Temperature and Quenching Practice on Surface Fatigue 515 24-33. Steel Rollers.Contents xvii 24-32. Linkage Arm. Cast Low-Carbon Steel: Starting Crack Size vs Cycles to Failure 516 24-34. Notched Links. Ti-6AI-6V-2Sn: Fatigue Endurance of HIP-Consolidated Powder 518 . Hot Rolled Low-Carbon Steel: S-N Data for Component Test Model 517 24-35. Fuselage Brace.

the most important feature of the fatigue process is nucleation of one or more cracks under the influence of reversed stresses that exceed the flow stress. Prediction of Fatigue Life The fatigue life of any specimen or structure is the number of stress (strain) cycles required to cause failure. 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. If anyone of these three is not present. Fatigue crack propagation may be caused by cyclic stresses in a benign environment. The cyclic stress starts the crack. a plastic zone (or region of deformation) develops at the defect tip. however. making analytical prediction of fatigue life difficult. Fatigue fractures are caused by the simultaneous action of cyclic stress. This zone of high deformation becomes an initiation site for a fatigue crack. . compression load may do so. On the microscopic scale. Small changes in the specimen or test conditions can significantly affect fatigue behavior. and current studies with fracture mechanics test specimens are beginning to provide satisfactory design criteria. In crack propagation testing. the designer may rely on experience with similar components in service rather than on laboratory evaluation of mechanical test specimens. and the metallurgical condition of the material. Although compressive stress will not cause fatigue. Laboratory tests. or between a maximum tensile stress and a maximum compressive stress. fracture mechanics methods are used to determine the crack growth rates of preexisting cracks under cyclic loading. fatigue environment.) Fatigue cracks initiate and propagate in regions where the strain is most severe. or by the combined effects of cyclic stresses and an aggressive environment (corrosion fatigue). The stress usually is cycled either between a maximum and a minimum tensile stress. thus producing only tensile and compressive stresses. are essential in understanding fatigue behavior. including stress level. 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 Crack Initiation Most laboratory fatigue testing is done either with axial loading. tensile stress and plastic strain. Therefore. followed by development of cracks at persistent slip bands or at grain boundaries. or if the material has an appreciable work-hardening rate. In crack initiation testing. or in bending. however. stress state. the stresses also may be above the static yield strength. Because most engineering materials contain defects and thus regions of stress concentration that intensify strain. The crack propagates under the applied stress through the material until complete fracture results. Laboratory fatigue tests can be classified as crack initiation or crack propagation. 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. Under the action of cyclic loading. (In low-cycle fatigue. most fatigue cracks initiate and grow from structural defects. This number is a function of many variables. cyclic wave form. specimens or parts are subjected to the number of stress cycles required for a fatigue crack to initiate and to subsequently grow large enough to produce failure. the tensile stress produces crack growth (propagation).1 Fatigue Testing Introduction Fatigue is the progressive. fatigue cracking will not initiate and propagate.

Two commonly used stress ratios are the ratio. each table of fatigue strengths must specify the number of cycles for which the strengths are reported. 1 Nomenclature to describe test parameters involved in cyclic stress testing S-N Curves. 2340 steel-= 100 Q) 48 HRC _ u (notched) .--"". Smax I---~:------. Sa = S. Stress is plotted on either a linear or a logarithmic scale.----r---..' -_ _ Time Fig. their S-N curves continue to drop at a slow rate at high numbers of cycles. Three typical S-N curves are shown in Fig. For these types of metals.}075-T6 50 25 U5 ~.!./ 2 = (Smax .. The stress ratio is the algebraic ratio of two specified stress values in a stress cycle.1 Number of cycles to fracture... 2. If the stress is cycled between a maximum stress and no load. 2. the mean stress is zero. Instead. making the test a sustained-load creep test rather than a fatigue test. Because there is no standard number of cycles.. N Fig.. of the alternating stress amplitude to the mean stress (A = Sal S m) and the ratio.. Fatigue limit Sf <I)" .. to failure using a logarithmic scale for the number of cycles. the stress ratio R becomes -1.._ . j-a~Stress ratio (R) = ..---- Smin Ol-\-. E <ll <Il <Il _. known as the fatigue limit or endurance limit. The horizontal portion of an S-N curve represents the maximum stress that the metal can withstand for an infinitely large number of cycles with 50% probability of failure and is known as the fatigue (endurance) limit.. Nomenclature to describe test parameters involved in cyclic stress testing are shown in Fig.~ 1 --cr.. is given an algebraic minus sign. of the minimum stress to the maximum stress (R = Sminl Smax)' 1100 1000 :2: Q) a.. = Smax .. if the stresses are partially reversed. In the completely reversed cycle test.. S. 0. Fatigue limit. S m = (S max + Smin) / 2. The mean stress. the stress cycle usually is maintained constant so that the applied stress conditions can be written Sm± SO' where S mis the static or mean stress. <ll 900 800 700 600 500 400 300 200 f- I 150 2340 steel o~ '"' 48 HRC ~ (unnotchedl'r I = 1::-. S a' is one half the range of stress.~-----Ic-I-'--~r--f--+---L. the material can endure an infinite number of cycles without failure. the stress ratio R becomes a positive number less than 1. which is the stress to which the metal can be subjected for a specified number of cycles.. 2 Typical S-N curves for constant amplitude and sinusoidal loading During a fatigue test. ~ ... <I)" 75 0. 2) and titanium. Sf E <ll <Il <Il U5 ~ l Aluminum alloy """Q. The fatigue strength of nonferrous metals at 100million (108) or 500 million (5 X 108) cycles is erroneously called the fatigue limit. For the low-cycle fatigue region (N< 104 cycles) tests are conducted with controlled cycles of elastic plus plastic strain. the stress ratio R becomes zero...:. The results offatigue crack initiation tests usually are plotted as maximum stress. minimum stress. S m' is the algebraic average of the maximum and minimum stresses in one cycle. A stress ratio R of 1indicates no variation in stress. Low-Cycle Fatigue. R. Fatigue Limit and Fatigue Strength. The result- ing plot of the data is an S-N curve. If the stress is cycled between two tensile stresses. .Smin)/2. S" is the algebraic difference between the maximum and minimum stresses in one cycle.. u ~ ~ If the stresses are fully reversed. The number of cycles of stress that a metal can endure before failure increases with decreasing stress. 1.. and Sa is the alternating stress. Sf' Most nonferrous metals do not exhibit a fatigue limit. fatigue strength rather than fatigue limit is reported. which is equal to half the stress range.. N.2 Fatigue Testing The latter is considered a negative tensile stress. The range of stress... A.- 125 'iii . as shown by the curve for aluminum alloy 7075-T6 in Fig. Below this limiting stress. the S-N curve becomes horizontal at a certain limiting stress. For some engineering materials such as steel (see Fig.. '1. or stress amplitude to number of cycles. and therefore is known as the minimum stress.Smin' The stress amplitude. R becomes a negative number less than 1... Applied stresses are described by three parameters.

versus the number of cycles to failure (Fig. Stress is concentrated in a metal by structural discontinuities.K._-.. or scratches.-_ _. frequency. Crack length is measured as a function of elapsed cycles.I. dK. MPa\ m Fig.). 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. dE p' is the difference between the total strain range and the elastic strain range.-_ _--I. holes. --L. which act as stress raisers. Fatigue notch sensitivity. This value may be stated as percentage. o 5.3 10 -5 ~ >- ~ I-~=---+_--+----+---__l "tl 10-4 o~ 10. A common definition of fatigue notch sensitivity is q = (KJ ..' 10.. K J. in which q may vary between zero (where K J = 1) and 1 (where KJ = K. dE eo is equal to the stress range divided by the modulus of elasticity..-.2 U c:: 10. The fatigue notch factor.4 <:10. For determination of K" the greatest stress in the region of the notch is calculated from the theory of elasticity. The plastic strain range.1).K. Fatigue crack growth data are typically presented in a log-log plot of da/dNversus s« (Fig. and these data are subjected to numerical analysis to establish the rate of crack growth. 4 Fatigue crack propagation rate data in 7075-T6 aluminum alloy (R <0) LIVE GRAPH Click here to view .5 L. and the stress-concentration factor. including environment. Significant structural life may remain in the cyclic growth of the crack to a size at which a critical failure occurs.L-_ _. 3).3 Q) 10. q. .._ _-L.._ _. there are a number of additional factors that may exert a strong influence. such as notches..._. dE. temperature. K" is the ratio of the area test stress in the region of the notch (or other stress concentrators) to the corresponding nominal stress. K" for a specimen of a given size containing a stress concentrator of a given shape and size. the existence of a crack does not necessarily imply imminent failure of the part. However.J 10. 1. 3 10. 10 15 20 25 30 35 Fatigue Crack Propagation In large structural components.1 10 3 10 5 10 Cycles to failure Fig. ksivTn. -l.--. Stress-Concentration Factor. and grain direction. for a material is determined by comparing the fatigue notch factor. Crack growth rates are expressed as a function of the crack tip stress-intensity factor range. The total strain range is separated into elastic and plastic components.3 Typical plot of strain range versus cycles-to-failure for low-cycle fatigue cling of notched specimens that have been precracked in fatigue..6 o 10 20 30 40 J. and cracked specimen geometry. Fatigue crack propagation testing usually involves constant-load-amplitude cy- 10..3 L. The objective of fatigue crack propagation testing is to determine the rates at which subcritical cracks grow under cyclic loadings prior to reaching a size critical for fracture.1 I---+_--+----.-----r-"T"T"--r--r-~--. The growth or extension of a fatigue crack under cyclic loading is principally controlled by maximum load and stress ratio.:--et-'~-__l 10. For many metals and alloys. 4). da l d N. as in crack initiation. The stress-concentration factor. Under controlled strain testing. The stress-intensity factor is calculated from expressions based on linear elastic stress analysis and is a function of crack size. or equivalent values are derived experimentally. :J. fatigue life behavior is represented by a log-log plot of the total strain range. .Introduction rather than with controlled load or stress cycles. load range. the elastic strain range.l)f(K.

4 Fatigue Testing Fatigue Crack Initiation Crack initiation tests are procedures in which a specimen or part is subjected to cyclic loading to failure. features such as surface and metallurgical imperfections can act to concentrate stress locally. As a result. The foregoing processes also occur in alloys and heterogeneous materials. Because it is a synergistic effect of fatigue and corrosion. A variety of crystallographic features have been observed to nucleate fatigue cracks. An unlimited number of gaseous and liquid mediums may affect fatigue crack initiation in a given material. cuts. fatigue cracks initiate at the surface. second-phase particles. Although crack initiation tests conducted on small specimens do not precisely establish the fatigue life of a large part. The number of load cycles that a component under low overstress can endure is high. dents. alloying that (1) enhances cross slip. Surface imperfections such as scratches. the term highcycle fatigue is often applied. In addition to the geometric features of a part. Fretting corrosion. alloying usually raises the flow stress of a metal. initiation of multiple cracks is more likely. is increased. thus offsetting its potentially detrimental effect on fatigue crack nucleation. and other features that disturb the structure. and complete fracture results. and twin boundaries are common nucleation sites. In pure metals. inclusions. Relation to Environment. As the magnitude of the nominal stress increases. such tests do provide data on the intrinsic fatigue crack initiation behavior of a metal or alloy. burrs. Fatigue Testing Regimes The magnitude of the nominal stress on a cyclically loaded component frequently is measured by the amount of overstress-that is. Crack Nucleation. whereas nucleation at twin boundaries is associated with active slip on crystallographic planes immediately adjacent and parallel to the twin boundary. In low-cycle fatigue. are crack nucleation sites. especially in a material with an appreciable workhardening rate. which occurs from relative motion between joints. such data can be utilized to develop criteria to prevent fatigue failures in engineering design. spacing between fatigue striations. Nucleation at grain boundaries appears to be a geometrical effect. and other manufacturing flaws are the most obvious sites at which fatigue cracks initiate. may also accelerate fatigue crack initiation. a fatigue crack is initiated at a highly stressed region of a component subjected to cyclic loading of sufficient magnitude. however. the amount by which the nominal stress exceeds the fatigue limit or the long-life fatigue strength of the material used in the component. At high strain rates. Except for instances where internal defects or special surfacehardening treatments are involved. even in the absence of inherent grain boundary weakness. complex welded. . However. and automotive and aerospace components. (2) enhances twinning. alloying and commercial production practices introduce segregation. or (3) increases the rate of work hardening will stimulate crack nucleation. Examples of the use of small-specimen fatigue test data can be found in the basis of the fatigue design codes for boilers and pressure vessels. riveted. the cyclic stress may be above the static yield strength. Fatigue cracks initiate at points of maximum local stress and minimum local strength. and the region of final fast fracture is increased in size. On the other hand. Also. The local stress pattern is determined by the shape of the part and by the type and magnitude of the loading. The crack then propagates in progressive cyclic growth through the cross section of the part until the maximum load cannot be carried. Crack Initiation. slip band extrusion-intrusion pairs at free surfaces. Corrosion fatigue describes the degradation of the fatigue strength of a metal by the initiation and growth of cracks under the combined action of cyclic loading and a corrosive environment. tubular holes that develop in persistent slip bands. or bolted structures. In general. thus. which indicate the progressive growth of the crack front. corrosion fatigue can produce a far greater degradation in strength than either effect acting alone or by superposition of the singular effects. Grain boundaries in polycrystalline metals. Fatigue Cracking· Fatigue cracks normally result from cyclic stresses that are below the yield strength of the metal. Generally. All of these phenomena have a significant influence on the crack nucleation process. A large portion of the total number of cycles in these tests is spent initiating the crack. this appears to be the preferred site.

. gripping devices. this distinction is made by determining whether the dominant component of the strain imposed during cyclic loading is elastic (high cycle) or plastic (low cycle). The total strain range df consists of an elastic strain component df e = dalE and a plastic strain component df p • The width of the hysteresis loop depends on the level of cyclic strain. the stressstrain curve is O-A-B. High-cycle fatigue data are presented graphically as stress (S) versus cycles-to-failure (N) in S-N diagrams or S. 5 Classification of Fatigue Testing Machines Fatigue test specimens are primarily described by the mode of loading: • • • • • Direct (axial) stress Plane bending Rotating beam Alternating torsion Combined stress Testing machines. Because the stress in high-cycle fatigue tests is usually within the elastic range. df. stress calculated using the specimen dimensions and the controlled load or deflection applied axially. The arbitrary. and drive (loading) system. yielding begins in compression at a lower stress C due to the Bauschinger effect. depending on the fixturing used. or maximum stress on the S-axis is made using simple equations from mechanics of materials. however. Fatigue Testing Machine Components Whether simple or complex. In reloading in tension. in flexure. stress range. but commonly accepted. may be universaltype machines that are capable of conducting all of the above modes ofloading. The slope of this line in the region where plastic strain dominates has shown little variation for the large number of metals and alloys tested in low-cycle fatigue. 7. a hysteresis loop develops. i. When the level of cyclic strain is small. These are described in the Introduction to this Section along with the symbols and nomenclature commonly applied in fatigue testing. For tests conducted under constant df. The common method of presenting low-cycle fatigue data is to plot either the plastic strain range. During initial loading. The load frame is the structure of the machine that reacts to the forces applied to the specimen by the drive system. or in torsion. the stress range da usually changes with an increasing number of cycles. Typical load train components in an electrohydraulic axial fatigue machine are shown in Fig. versus N.N curves. the hysteresis loop becomes very narrow. Fig. When plotted using log-log coordinates. dividing line between high-cycle and low-cycle fatigue is considered to be about 104 to 105 cycles. controllers. which in turn depends on the properties of the metal as well as the magnitude of the nominal stress. Figure 5 illustrates a stress-strain loop under controlled constant-strain cycling in a low-cycle fatigue test.Fatigue Crack Initiation Low-cycle fatigue is the regime characterized by high overstress. test specimen. all fatigue testing machines consist of the same basic components: a load train.e. The load train consists of the load frame. Upon unloading. In practice. Figure 6 is an example of the typical presentation of low-cycle fatigue test results. the calculation of stress amplitude. The dimensions of this loop are described by its width df (the total strain range) and its height da (the stress range). Presentation of Fatigue Data. 5 Stress-strain loop for constantstrain cycling . the average value being Y2. df p' or the total strain range. a straight line can befit to the dfp-Nplot. This power-law relationship between dfpand Nis known as the CoffinManson relationship. and monitors.

and maintain the controlled test parameter(s). fixtures attached to an oscillating platen of a rotating-eccentric-masstype machine can facilitate axial... N Fig... Pressure transducers are used in hydraulic or pneumatic actuator devices..' load drop... or hydraulically through a piston limited by stops. or rotating grips. '" "1 <Ii C r-..1 ..~ ". Some devices provide an output signal to the controller. or to a readout device in the case of uncontrolled parameters... The control of timevarying deflection or displacement can be obtained in mechanical systems by cam-operated deflection levels. .. bending..... Loading fixtures to alter the mode of loading provide versatility. the magnitude of force and displacement initially set by the control system remains constant throughout the test. Fixtures can be designed to convert the axial force provided by a hydraulic actuator to perform four-point bending or torsion testing.. Q.. Sensors are required to measure the load. IU r--. . ~ ~ n ~ s C 1Il ~ .. 0 10 Cycles to failure.6 Fatigue Testing Ia-. Electromagnetic excitation can be used to excite a mass or inertia system to load a specimen. <.. Control in most simple machines and drive systems is obtained via the open-loop mode. In electrohydraulic machines. strain. Similarly. the motors drive hydraulic pumps to provide service pressure for control of the motion and force of a hydraulic piston actuator. a rotating eccentric mass. 01 10. and cycle count.. or deflection limit).. a::: 1Il <II 10- 2 ~ ~ ~ O.. 7 Schematic of the load train in an electrohydraulic axial fatigue machine The drive system is the most significant feature of a fatigue testing system and usually is electrically powered.. Common sensors are load cells (resistance strain gage bridges calibrated to load) inserted in the load train. displacement.. The controls and controllers manually or automatically initiate power and test. deflection...n ~ r-.. extension. 00 ... In such systems. Control Systems. --... 6 Low-cycle fatigue curve (t1€p versus N) for type 347 stainless steel " Fig... and torsion fatigue testing of specimens. levers. The simplest systems use electric motors to act on test specimens via cams. Controllers also terminate the test at a predefined status (failure. adjust.

Moore-type machines (Fig. but often have partial closed-loop features to continuously correct mean load. 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. and versatility in fatigue testing systems. The grips shown in Fig. thickness. Grip failure sometimes occurs prior to specimen failure. The R. these simple. 9a) can operate up to 10000 rpm. Fig. therefore.Fatigue Crack Initiation 7 {a} {bl {e} {dl lei Ifl (g) (a) Standard grip body for wedge-type grips. These can be obtained as component systems and can be upgraded as required. 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. The high cyclicfrequency of operation of these types of machines enables testing to long fatigue lives (> 108 cycles) within weeks. Proper gripping is not simply the attachment of the test specimen in the load train. The load is transmitted to the specimen through a flexure system. 8 are typical of those used for axial fatigue tests. Frequently. in a small-diameter specimen. satisfactory gripping evolves after specimen design development. A hydraulic actuator typically is used to apply the load in axial fatigue testing. or diameter. (g) Snubber-type wire grips for flexible wire or cable. these are open-loop systems. (d) Universal open-front holders. In all bending-type tests. (c) Flat grips for specimens for use in standard grip body. only a very small volume of material is under test. Rotating Beam Machines. Generally. bolts. 8 Grip designs used for axial fatigue testing Grips. an axial fatigue testing machine must be able to apply a greater force than a static bending machine to achieve the same stress. which provides straight-line motion to the specimen. Bending Fatigue Machines The most common types of fatigue machines are small bending fatigue machines. 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. Cantilever beam machines. In general. monitoring. (e) Adapters for special samples (screws. Servohydraulic closed-loop systems offer optimum control. R.) for use with universal open-front holders. etc. studs. Care must be taken in grip design and specimen installation in the grips to prevent misalignment. 9. in which the test specimen has a tapered width. 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. Typical rotating beam machine types are shown in Fig. Electromagnetic or magnetostrictive excitation is used for axial fatigue testing machine drive systems. (f) Holders for threaded samples. inexpensive systems allow laboratories to conduct extensive test programs with a low equipment investment. Electromechanical systems have been developed for axial fatigue studies. (b) V-grips for rounds for use in standard grip body. only the material near the surface is subjected to the maximum stress. 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. Specially . This system automatically and steplessly restores the preset load through the hydraulic piston. particularly when low-load amplitudes and high-cycle fatigue lives are desired in short test durations.

Rota ry actu ato r Fig. in which rotary actuators are incorporated in a closed-loop testing system (Fig. (b) Single-end rotating cantilever testing machine. Rolling contact fatigue of ball and roller bearings under controlled lubrication conditions is a specialized field of fatigue testing. special devices have been used. bearings.. in which linear motion is changed to rotational motion by the use of cranks. Rolling contact fatigue testers usually are constant-load machines in which a Hertzian contact stress between two rotating bearings is applied until occurrence of fatigue failure by pitting or spalling is indicated by a vibration or noise level in the system..:.~~....- ---I I I L---- Angular display feedback . 10). 9 Schematic of rotating beam fatigue testing machines Program . Special-Purpose Fatigue Testing Machines To perform fatigue testing of components that are prone to fatigue failure (gears..:. etc.....:l:::.).. Moore testing machine... sometimes as modifications to an existing fatigue machine.:.-----i----r--... Multiaxial Fatigue Testing Machines Many special fatigue testing machines have been designed to apply two or more modes of loading.. to specimens to de- .. Fig. Specimen I I I I I Hydraulic power supply Hydraulic service manifold Displacement transducer . 10 Schematic of a servohydraulic torsional fatigue testing machine designed torsional fatigue testing machines consist of electromechanical machines.R.Torque feedback . wire.. in which a length of the test wire is used as the beam and is deflected (buckled) a known amount and rotated. and servo hydraulic machines.8 Fatigue Testing A ~Load (a) (b) (a) Four-point loading R.C:::}:. in or out of phase. Wire testers are a modification of rotating beam machines..

Stress concentration can also arise from surface roughness and metallurgical stress raisers such as porosity. Because actual machine elements invariably contain stress raisers such as fillets. and decarburization. An optimum way of minimizing fatigue failure is the reduction of avoidable stress raisers through careful design and the prevention of accidental stress raisers by careful machining and fabrication. The ratio of the maximum stress in the region of the notch (or other stress concentration) to the corresponding nominal stress is the stress-con- [c] ~---~ ~3-~ ~ .) l Effect of Stress Concentration Fatigue strength is reduced significantly by the introduction of a stress raiser such as a notch or hole. 11. selected on basis of ultimate strength of material R. i. 5"43' . and holes. inclusions.25 mm (1.40 in. The effect of stress raisers on fatigue is generally studied by testing specimens containing a notch. smoothly blended radii to eliminate any stress concentrations in the transition. 4. The presence of a notch in a specimen under uniaxial load introduces three effects: (1) there is an increase or concentration of stress at the root of the notch.) selected on basis of ultimate strength of material R.) D. 75 to 250 mm (3 to 10 in. avoiding very low load amplitudes where sensitivity and response of the system are decreased. The test section in the specimen is reduced in cross section to prevent failure in the grip ends and should be proportioned to use the upper ranges of the load capacity ofthe fatigue machine. D 9 Fatigue Test Specimens A typical fatigue test specimen has three areas: the test section and the two grip ends. local overheating in grinding. (c) Rotating beam specimen. 12.) (d) D. screw threads. (d) Plate specimen for cantilever reverse bending. (bl Rotating cantilever beam specimen. press fits. Fig.)"] k 50 ~m_1 -$-O~2m.) ( ~=====-t-$ ~R D. 12.48 in. Several types of fatigue test specimens are illustrated in Fig. (2) a stress gradient is set up from the root of the notch toward the center of the specimen.50 in.) Ie) (a)Torsional specimen.7 mm (0. The grip ends are designed to transfer load from the test machine grips to the test section and may be identical..e.) (a) 30 mm (13/16 in.) :cD mm (% in. particularly for axial fatigue tests. selected on basis of ultimate strength of material R.Fatigue Crack Initiation termine the properties of metals under biaxial or triaxial stresses.50 in.) Tapered D.0 in.) D 38 mm (1V2 in.5 to 10 in.) Ib) I'" 90 mm (3:6 i ~ 19 ~~-®R~ ~ '\ I 12 mm (0. keyways. 5 to 10 mm (0. and (3) a triaxial state of stress is produced at the notch root. usually a V-notch or a U-notch. The transition from the grip ends to the test area is designed with large. Ie) Axial loading specimen. 11 Typical fatigue test specimens .7 mm (0. 90 to 250 mm (3. The design and type of specimen used depend on the fatigue testing machine used and the objective of the fatigue study.8 mm (3116 in.) 38 mm (1'12 in. fatigue cracks in structural parts usually initiate at such geometrical irregularities.20 to 0.

Experimental data on the size effect in fatigue typically show that the fatigue limit decreases with increasing specimen diameter. the surface may be subjected to oxidation and corrosion. Effect of Test Specimen Size It is not possible to predict directly the fatigue performance oflarge machine members from the results oflaboratory tests on small specimens. The data for notched specimens usually are plotted in terms of nominal stress based on the net cross section of the specimen. Horger's data for steel shafts tested in reversed bending (Table 1) show that the fatigue limit can be appreciably reduced in large section sizes. At a mean stress equal to the ultimate tensile strength of the material. Effect of Mean Stress A series of fatigue tests can be conducted at various mean stresses. the fatigue-notch factor is based on the fatigue strength at a specified number of cycles. The two straight lines and the curve shown in . Changing the size of a fatigue specimen usually results in variations oftwo factors. It is extremely difficult to prepare geometrically similar specimens of increasing diameter that have the same metallurgical structure and residual stress distribution throughout the cross section. K. Typically. In most cases. (3) material. The change in amount of surface is significant. an increase in diameter usually decreases the stress gradient across the diameter and increases the volume of material that is highly stressed. Values of KJhave been found to vary with (1) severity of the notch. The problems in fatigue testing of large specimens are considerable. the permissible amplitudes steadily decrease. As the mean stress increases. it is more useful to know how the mean stress affects the permissible alternating stress amplitude for a given life (number of cycles). free of machining grooves and grinding scratches. and (3) changes in the residual stress condition of the surface. Precise determination of this phenomenon is difficult..50 6. Additionally. Table 1 Effect of specimen size on the fatigue limit of normalized plain carbon steel in reversed bending Specimen diameter mm in. This usually is accomplished by plotting the allowable stress amplitude for a specific number of cycles as a function of the associated mean stress. i. fatigue life increases as the magnitude of surface roughness decreases. Therefore. because fatigue failures usually initiate at the surface. increasing the diameter increases the volume or surface area of the specimen.30 l. or can be measured using photoelastic plastic models. First. Factors that affect the surface of a fatigue specimen can be divided into three categories: (1) surface roughness or stress raisers at the surface. Decreasing surface roughness minimizes local stress raisers.(see the Introduction to this Section). all fatigue cracks initiate at the surface.e. In general. the permissible amplitude is zero. For design purposes. The effect of notches on fatigue strength is determined by comparing the S-N curves of notched and unnotched specimens. and the results can be plotted as a series of S-N curves.00 248 200 144 36 29 21 Surface Effects and Fatigue Generally. the fatigue strength of large members is lower than that of small specimens. Figure 12 illustrates the effects that various surface conditions have on the fatigue properties of steel. special attention must be given to the surface preparation of fatigue test specimens. values of K. For materials that do not exhibit a fatigue limit. The effectiveness of the notch in decreasing the fatigue limit is expressed by the fatigue-notch factor. a size effect exists. fatigue properties are very sensitive to surface conditions. A description of applied stresses and S-N curves can be found in the Introduction to this Section. and (5) stress level. (2) changes in the properties of the surface metal. Secondly.6 38 152 0. (4) type of loading. a metallographic finish.10 Fatigue Testing cent ration factor. Surface Roughness. and few fatigue machines can accommodate specimens with a wide range of cross sections.can be calculated using the theory of elasticity. At zero mean stress. K p This factor is the ratio of the fatigue limit of unnotched specimens to the fatigue limit of notched specimens. Fatigue limit MPa ksi 7. for plain or notched specimens loaded in bending or torsion. the allowable stress amplitude is the effective fatigue limit for a specified number of cycles. is necessary. In some situations. Except in special cases where internal defects or case hardening is involved. (2) type of notch.

The assumption is made that the application of n.-----..._ 1 _ .-------r-----r------.Fatigue Crack Initiation 1000 .~==~===~~=~===t===~==~ ~--_t_---~2. S CI)" / Fig./ N..___f-----~ 30 E 0 u 2 5 10 20 50 100 200 500 Life. it is necessary to predict fatigue life under various stress amplitudes.t. The test is then continued to fracture at a different amplitude.. and Soderberg line....::~~. for which the average number of cycles to failure is N.. 13 Effect of mean stress on the alternating stress amplitude 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._ + .. 12 Effect of surface conditions on the fatigue properties of steel (302 to 321 HB) Fatigue strength.----.. using the fatigue strength as determined experimentally..-----. These tests do not simulate service conditions. Goodman's original law included the assumption that the fatigue limit was equal to one third of the tensile strength. 11 r ~~. Fig. This is a hypothesis first suggested by Palmgren and restated by Miner./N. and that failure will occur when "'i..= _ ...) = 1._1_--~~=_---="""" .. this has since been generalized to the relation shown in Fig. The most widely used method of estimating fatigue under complex loading is provided by the linear damage law. the stress can be changed from one stress amplitude to another at regular intervals. Su OJ C Mean stress.. Stress Amplitude...... and numerous alternative theories of cumulative linear damage have been suggested. such tests are known as block._ _ l 100 'iii ~ ~ ~ ~ 600 500 W ~ 70 60 ~ 90 -"" vi U.. Some considerations of redistribution of stresses have been clarified.. ~ ~ ~ ~ > 400 50 40 ~ ~ > Qi Qj ~ Q) > ~ ~ ~ 300 t-----t-----t---+----+----+--"""""~--~""""""----+--___I u o E 200 I----t----+---+---+-----+---+---+--~. but may serve a useful purpose for assessing the linear damage law and indicating its limitations..----r---... and is sometimes known as Miner's rule. The straight line joining the alternating fatigue strength to the tensile strength is the modified Goodman law.---.. causes an amount of fatigue damage that is measured by the cumulative cycles ratio n. Because stress amplitude varies widely under actual loading conditions. 13. Gerber's parabola. 13 represent the three most widely used empirical relationships for describing the effect of mean stress on fatigue strength. Alternatively. Gerber's parabola vi ~ ~ / Modified Goodman line Tensile strength.cycles at a stress amplitude S. 800 900 ro 700 a.. Sm As shown by the modified Goodman line. or interval.. See text for discussion._ t _ ...(n.. . tests. This method is not applicable in all cases... but there is as yet no satisfactory approach for all situations. 1000 cycles Fig....

the majority has been concerned with the nominal stress required to cause failure in a given number of cyclesnamely.K"" which corresponds to the stressintensity factor range below which cracks do not propagate.K curves have three distinct regions. Fig.c 1.. 100 10 20 50 Corrosion fatigue is the combined action of repeated or fluctuating stress and a corrosive environment to produce progressive cracking. 14) exhibits a fatigue crack growth threshold.K plot was constructed of data on five specimens of ASTM A533 HI steel tested at 24 0 C (75 0 F).K)" where C and n are constants for a given material and stress ratio. Fracture mechanics methodology enhances the understanding of the initiation and propagation of fatigue cracks and assists in solving the problem of designing to prevent fatigue failures. 14 Fatigue crack growth behavior of ASTM A533 B1 steel Fatigue Crack Propagation Test Methods The general nature of fatigue crack propagation using fracture mechanics techniques is summarized in Fig. Once fatigue cracks have formed. corresponding to the load cycle applied to a specimen is illustrated... When failure occurs by corrosion fatigue.K versus da/ dN..~ 6~ I ~ I 1 ! I I Region 3: unstable crack t---- I I I : W '0 Ie? if I r. the corrosive aspect also may accelerate the rate of crack growth. fil.K.. da/ dN. the magnitude of cyclic stress and the number of times it is applied are not the only critical loading parameters. Usually. are dependent on the material. Timedependent environmental effects also are of prime importance. S-N curves. stress-cycle frequency.10. This is described by the power-law relationship: da dN = C(I:!. ambient room air. ksi\. Although considerable fatigue data are available.I growlh r-- 'I 1 Region 2: power-taw behavior Fatigue Crack Propagation Fatigue failure of structural and equipment components due to cyclic loading has long been a major design problem and the subject ofnumerous investigations. Usually. da dN ~ C(~Kln dO I~ ~~o rapid . Test conditions: R = 0. environmental effects are deleterious to fatigue life. and stress ratio all affect the cracking processes.1 I. a straight line usually is obtained on a log-log plot of I:!. it is difficult to distinguish between fatigue crack initiation life and fatigue crack propagation life. Preexisting flaws or crack-like defects within a material reduce or may eliminate the crack initiation portion of the fatigue life of the component. I:!. A plot of similar shape is anticipated with most structural alloys.K (Region II in Fig. The behavior in Region I (Fig.K. MPa \ m Yield strength of 470 MPa (70 ksi). At intermediate values of I:!. a ~ q I I ] i . however. however. the absolute values of da/dNand I:!. A logarithmic plot of the crack growth per cycle. 14. With this type of testing.. In corrosion fatigue. versus the stressintensity factor range. .. 24°C (75 OF).growth I I I i sx. 14). 5 Region 1: slow crack I. The da/ dN versus I:!.9' r.. 6 I. .12 Corrosion Fatigue Fatigue Testing Stress-intensity factor range UK). Results of fatigue crack growth rate tests for nearly all metallic structural materials have shown that the da/ dN versus I:!.j> I I 8 10 20 30 40 50 60 80 100 Suess-tntensnv teeter range UK). stresswave shape. producing cracks in fewer cycles than would be required in a more inert environment. such data are obtained by testing smooth or notched specimens. I:!.

the stress ratio and mean stress have negligible effects on the rate of crack growth in Region II.4 i ~ s: ~ 0 e Ol ~ e Ol -" ~ U -" 0 10.2 Ql "0 ~ > o o • • 12 Ni steel 10 Ni steel HY-80 steel HY-130 steel Ql E E 10.4 "0 > ~ c:: .: <: ~ ~ <: ~ where a is given in inches. or some other elastic-plastic fracture mechanics approach is more appropriate than ~K for correlating the data.. unstable behavior occurs.. Fatigue lives greater than about 106 cycles correspond to growth rates below 10-8 in] cycle.. For applications involving fatigue lives of up to about 106 load cycles.... crack-opening displacement..66 X 1O-8(~K)2. 14. low-toughness metals. crack length measurements on the sides of the specimens can be used as representations of through-thickness crack growth behavior. the increasing crack length during constant load testing causes the peak stress intensity to reach the fracture toughness... MPaVm Fig. R = P min/Pmax' is in the range 0 < R < 1. The second possibility. as shown in Fig: 15. Standardized testing procedures for measuring fatigue crack growth rates are described in ASTM Standard E 647. and the unstable behavior is related to the early stages of brittle fracture. and the load ratio.-_.25 13 Stress-intensity factor range (." 10 102 10. 14).5 value for K/c tests. The upper bound of scatter can be obtained from: da dN = 0. is common for ductile metals.3 10. and these require special testing procedures.:lKl. respectively. This method applies to medium to high crack growth rates-that is.:lK). plastic limit load behavior. co U 2 x 10. The specimen thickness-to-width ratio.Fatigue Crack Propagation Fatigue crack growth rate data for some steels show that the primary parameter affecting growth rate in Region II is the stress-intensity factor range and that the mechanical and metallurgical properties of these steels have negligible effects on the fatigue crack growth rate in a room-temperature air environment. in which specimen sizes normally used for fatigue crack growth rate testing behave in a linear elastic manner at K levels equal to K/c. When plastic limit load behavior causes unstable crack growth. K'n of the material.. ksiV'Ti1.25. because the limitations of linear elastic fracture mechanics have been exceeded.9 X 10-7 ui. B/W. Data for four martensitic steels fall within a single band. For some steels. 16 and 17). ASTM E 647 describes the use of centercracked specimens and compact specimens (Fig. 15 Summary of fatigue crack growth data for martensitic steels under consideration by ASTM. Here.5 10. loading fixtures frequently can be used.5 t--t-l~tf--+--+--+----I L. particularly if K/cis high. trapezoidal) do not affect the rate of crack propagation per cycle of load for some steels in benign environments. the frequency of cyclic loading and the waveform (sinusoidal. square. There are two possible causes of this behavior..I. triangular. Second. At high ~Kvalues (Region III in Fig. both the maximum and minimum loads are tensile. which are related to the threshold of fatigue crack growth illustrated in Fig. First.6 10. For tension-tension fatigue loading. The first possibility is operative for high-strength. the use of the J-integral concept.. Also.125 and 0. and ~K is given in ksiyTr. the maximum B/W values for centercracked and compact specimens are 0. the procedures recommended in ASTM E 647 can be used.] cycle). the growing crack reduces the uncracked area of the specimen sufficiently for the peak load to cause fully plastic limit load behavior. ~K values have no meaning. is smaller than the 0. the K.L-_.-_.I. above 10-8 tu] cycle (3. For this type of loading. With the thinner specimens. resulting in a rapid increase in the crack growth rate just prior to complete failure of the specimens.6 '---_. Procedures for growth rates below 10-8 in] cycle are ~ 2:! s: ~ 10. A ratio of .L_--L--' 2 5 10 20 50 100 200 Stress-intensity factor range (...

The incremental polynomial method eliminates some of the scatter in growth rate that is inherent in fatigue testing. R and the critical K value.sthe machined notch. This prevents the measurement of erroneous growth rates from a group of data points that are spaced too closely relative to the precision of data measurement and relative to the scatter of data.1 is commonly used for developing data for comparative purposes. Figure 18illustrates a typical a versus N plot. For constant-amplitude loading. such as the commonly used Forman equatl?n. The most widely accepted relationship is that proposed by Paris.K) W where Cand n are material constants of the same types as those in the Paris equation..01 in..K)" dN (1 . The most commonly used methods. This method. or other chemical reaction on cyclic loading. a is the crack length.K. between data points (see Fig. however. the Forman equation describes the frequently observed increase in da/ dN asso- .03 Wfor 2a W 6. K" at which rapid frac~ ture of the specimen occurs (Region III in Fig.02 W for 0. An advantage of the Forman equation is that it describes the type of accelerated da/dNbehavior that is often observed at high values of 6. Other relationships based on the Paris equation. The minimum crack-length interval.6. 18)should be 0. are the secant and incremental polynomial methods.. are used to represent the variation of da/ dN with other key variables. The secant method consists of the slope of the straight line connecting two adjacent data points.a ~ 0. Data Analysis..60 !I I 2a n i.25 ~!!. 14) of the crack growth regime.40 Fatigue crack growth rate data can be calculated by several methods.) or ten times the crack-length measurement precision..) R= 0.04 Wfor 0.a.~ 0.R)(K.25 mm (0. and the slope of this expression is the growth rate.a ~ 0.40 ~ !!. Numerous relationships have been generated to correlate crack growth rate and stress-intensity data. 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. Additionally. .75 mm (3 in. 6. however.. corrosion.a ~ 0..02 W for W ~----+-----J_l !I ! I specimen thickness. Testing often is performed in laboratory air at room temperature. Crack measurement intervals are recommended in ASTM E 647 according to specimen type. results in more scatter in measured crack growth rate. which is defined as the standard deviation on the mean value of crack length determined for a set of replicate measurements. 14). a set of crack-length versus elapsed-cycle data (a versus N) is generated.60 W 6. although simpler.60 W a For center-cracked tension specimens: 6.01 Wfor -. including load ratio. which is not described by the Paris equation. B is the Fig. < 0. 16 Standard center-cracked tension specimen for fatigue crack propagation testing when the width (WI of the specimen '. Pmax and P min' generally held constant.a~ -r W 0.~ 0. For compact-type specimens: 6. but of different values.a ~ 0.14 Fatigue Testing Two holes W/3 diam 6.Polynomial expression (parabola) to typically five to seven adjacent data points. The incremental polynomial method fits a secon~-order .::: 0. with the specimen loading. The Forman equation is: da = C(6.60 ~ > 0. any gaseous or liquid environment and temperature of interest may be used to determine the effect of temperature.

10 N Fig.Fatigue Crack Propagation Two holes 15 0. (J Oo c 1.. 14) complicates acquisition of valid and consistent data..0 in. and testing procedures under this regime. because the crack growth behavior becomes more sensitive to the material. the fatigue mechanisms of the material that slow the crack growth rates are more significant.0 i~' I ~ 2.) an = 0. 16 for explanation of symbols) ciated with an increase in R from 0 toward I.6W 0.20W ~------1. 18 Crack growth versus constantamplitude stress cycles for a Fe-10Ni8Co-1 Mo high-strength steel .. 17 Standard compact-type specimen for fatigue crack propagation testing (see Fig. environment. The precise definition of the cyclic crack growth rate threshold..2 50 E E 45 .. the less complex Paris equation may be used.w-------.a ----. ~ 35 .275W t t t t 0.6 C <iI s: 0.~I Allowable thickness: W/20 s B s W/4 Minimum dimensions: W = 25 mm (1.4 Q) . When only ~Kin Region II is involved..6W r 1 0.+ . When it is necessary to describe the effect of K approaching Kc> or the effect of R on da / dN. the Forman equation can be used to represent the da/dNbehavior.~ ( ) 1 . 55 ( 2.. ~K'h' varies significantly.25Wdiam 0..2 00 v u 30 25 20 125 n ro U .0 1.r= (J 40 Cyclic Crack Growth Rate Testing in the Threshold Regime Cyclic crack growth rate testing in the lowgrowth regime (Region I in Fig.J 1.0 175 200 225 250 3 150 275 300 Cycles...25W -------:. c 1.8 1.~I Fig.>< OJ ~ ".>< 'rP0a. Within this regime.275W r ~~_1_ 1 + .

It can also be small compared to the plastic zone size (10 to 1000 /. Consideration of the range of application of the stress-intensity solution of a specimen configuration is very important.K. m/cycle). linear elastic fracture mechanics might be replaced with elastic-plastic fracture mechanics. For example. the specimen must be predominantly elastic. It is extremely expensive to obtain a true definition of IJ. In this situation.2. unlike the requirements for plane-strain fracture toughness testing. environment. However./" / / \'. The method by which a notch is machined depends on the specimen . Stressintensity factor solutions for center-cracked tension and compact-type specimens are given in Table 2.8 to 10-9 to 10-10.Lm) when crack closure./ Short cracks that behave as long cracks \ \ \--- \ \ Short cracks that become non propagating cracks Stress-intensity factor range Fig. assuming that the appropriate equipment is available for controlling the test and measuring the crack dimensions. the expression for the center-cracked tension specimen is valid for 2a/W< 0. Because the duration of the tests increases greatly for each additional decade of near-threshold data (10. Short cracks that join long crack behavior Behavior of Short Cracks Recently. and growth mechanisms are of concern. it has been well documented that short cracks may behave differently from large cracks when plotted in the standard form of cyclic crack growth rate versus stress intensity.16 Fatigue Testing The most accurate definition would be the stressintensity value below which fatigue crack growth will not occur. Generally.Lm) when the concepts of continuum mechanics are of interest. A short crack is difficult to define. The stress state is considered to be a controlled test variable. The material characteristics.. The use of stress-intensity expressions outside their applicable crack-length region can produce significant errors in data.h. and in some materials a true threshold may be nonexistent. The crack may also be physically small (500 to 1000 /. Because the loading mode of different specimens varies significantly. each specimen geometry must be considered separately. / .95. To follow the rules of linear elastic fracture mechanics. and applied load will dictate whether the specimen is predominantly elastic. etc. such as the IJ.Kthat corres~onds to a fatigue crack growth rate of 10-8 to 10.0 tn] cycle (3. the expression given in Table 2 for the compact-type specimen is valid for a/ W > 0. crack length. Notch Preparation. the stresses at the crack tip do not have to be maintained in a plane-strain state. Figure 19 schematically illustrates the possible behavior of short cracks. The size of the specimen must also be appropriate. crack tip shape. 16 and 17). 19 Typical short crack behavior Selection of Test Specimens Selection of a fatigue crack growth test specimen is usually based on the availability of the material and the types of test systems and crackmonitoring devices to be used. Many stress-inten- sity expressions are valid only over a range of the ratio of crack length to specimen width (a/W). The two most widely used types of specimens are the centercracked tension specimen and the compact-type specimen (see Fig. the precise design requirements should be determined in advance of the test.j cycle). It may be small compared to the microstructure of the material to be studied (I to 50 /.9 X 10-7 to 10-9 in. any specimen configuration with a known stressintensity factor solution can be used in fatigue crack growth testing. However. specimen size.Lm). // / q. designers are more interested in the near-threshold regime.

Each of these requires a different stress-intensity solution. the step size of the load during precracking should be minimized. Grips designed for compact-type specimens are illustrated in Fig. where a = .64a . For example.Fatigue Crack Propagation Table 2 Stress-intensity factor solutions for standardized (ASTM E 647) fatigue crack growth specimen geometries 17 Center-cracked tension specimens (Fig. a where a = . and instability of cracks. such as the compact-type specimen grip.2 W W material and the desired notch root radius (p).Tl« 3 - 5. electrical (eddy current and resistance). 20.) in aluminum alloys.K= Ii V W ~ 2W sec 2 tra 2a . common practice is to initiate the pre cracking at a load above that which will be used during testing and to subsequently reduce the load. To decrease the amount of time needed for precracking to occur. It is also advisable to consider lateral movement above and below the grips. ultrasonic. For a center-cracked tension specimen less than 75 mm (3 in.886 + 4.. but is generally acceptable only for aluminum alloys.25 mm (0. Wider specimens generally require additional pins. and acoustic emission monitoring techniques.6a ) 4 a . A similar notch root radius in low.P(2 + a) 2 + 14. Load generally is reduced uniformly to avoid transient effects. a single pin grip is generally suitable. 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.).K - ru. Gripping arrangements for compact-type and center-cracked tension specimens are described in ASTM E 647.010 in. this percentage may be decreased. In grips that are permitted to rotate. 16) sr I:!. As the crack approaches the final desired size.) in width. Gripping of the specimen must be done in a manner that does not violate the stress-intensity solution requirements. the pin and the hole clearances must be designed to minimize friction. such as in the grips..95 id 10r-< W Compact-type specimens (Fig.010 in. milling or broaching is required. Crack-length measurements are made at intervals such that a nearly even distribution of da] dN versus ~K is achieved. friction gripping. nickelbase superalloys. Crack growth can be arrested above the threshold stress-intensity value due to formation of the increased plastic zone ahead of the tip of the advancing crack. 17) _ I:!. 3/2 (0. it is possible to produce a grip that permits rotation in the loading of the specimen. growth. Precracking of a specimen prior to testing is conducted at stress intensities sufficient to cause a crack to initiate from the starter notch and propagate to a length that will eliminate the effect of the notch.25 mm (0. Therefore.13. or it is possible to produce a rigid grip. The minimum amount of extension be- . Reduction in the maximum load should not be greater than 20% of the previous load condition. and titanium alloys. Crack-Length Measurement Techniques Several different techniques have been developed to monitor the initiation. electrical discharge machining may be necessary to produce a notch root radius of p ~ 0. compliance. or some other method to provide sufficient strength in the specimen and grip to prohibit failure at undesirable locations.and medium-strength steels can be produced by grinding. expression va liId f or-» 0.32a By W(l-a) I:!. expression va1'" 2a 0. in a singleedge notched specimen. Sawcutting is the easiest method. For high-strength steel alloys. including optical (visual and photographic). For a notch root radius of p ~ 0.

The optical technique is straightforward and.) tween readings is commonly about 0. However. the greater the amount of 0 measured for a specific value of tensile load. 21 Schematic of the relationship between compliance and crack length ure 21 illustrates that the more deeply a specimen is cracked.10 in. many fatigue crack growth rate tests are conducted in simulatedservice environments that obscure direct observation of the crack. and crack extension under these loading modes can be similarly determined. with the displacement measured along. Specimen load is simultaneously measured by an electronic load cell and conditioner / amplifier system. if the specimen is carefully polished and does not oxidize during the test. subjective.18 Fatigue Testing p t (a) ~--al~ p Fig. Compliance can also be defined for shear and torsional loads applied to cracked specimens.). the process is time consuming. Compliance Method of Crack Extension Measurement The compliance of an elastically strained specimen containing a crack of length a measured from the load line to the crack tip is usually expressed as the quotient of the displacement. In addition. the load line. (b) C(a1) = 01 /P Fig. 20 Grips designed for fatigue crack propagation testing of compact-type specimens (courtesy of MTS Systems Corp.25 mm (0. A generalized schematic of the circuits involved is shown in Fig. and the tensile load. P. Fig- (b) (a) C(aD) p = 0D/P. 22. 0. and can be automated only with complicated and expensive video-digitizing equipment. produces accurate results. and the output is directed to the same data-acquisition system. or parallel to. .

25CrIMo steel in air. can be determined through the use of the relevant calibration curve for the particular test piece geometry concerned. The fundamental data output requirement is the number of cycles to failure. rota- . 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.. Displacement gage -+----. gaseous hydrogen. the crack length to width ratio. in general. Similarly. a/ W.. its electrical resistance increases. for the load range applied to the specimen. noise-free. 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. the magnitude of the disturbance depending directly on the size and shape of the discontinuity. ) x-v recorder Load 'Lr a Gage condition t /P'~ I ± 10 V dc ± 10 V dc a. 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. or the number of cycles to failure under controlled cyclic loading conditions. Electric Potential Crack Monitoring Technique The electrical potential. Crack Growth Studies..Fatigue Crack Propagation 19 Specimen . and thus the potential difference between two points spanning the crack rises. as defined by the application. With increasing crack length. A variety of electromechanical fatigue testers have been developed for different applications. 22 Components of a compliance measurement system The required sensitivity of the systems depends on specimen geometry and size. and control capabilities. where it has been utilized to monitor almost all mechanisms of subcritical crack growth and most notably to follow fatigue crack growth. 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 uncracked cross-sectional area of the test piece decreases. a 1 a2 J=1 Load cell Load cell condition Fig. 23 for tests on a 2.. amplitude of loading (maximum loads and minimum loads). By monitoring this potential increase. an approximately I V de change in signal from the load cell is required for accurate calculation of the compliance. Vo . and hydrogen sulfide environments. By far the most useful application of the electrical potential method has been in measurements of crack length during crack propagation.) of deflection is satisfactory. Forced-displacement. amplified output on the order of I V dc per I mm (0. Typical crack propagation rates derived from direct current potential measurements are shown in Fig. For the application of crack growth monitoring. Variables associated with fatigue-life tests are frequency of loading and unloading. Va' and comparing it with some reference potential. or potential drop..04 in. forced-vibration.

iI t4. efficient. Typically < 450 N « 1001bf) .05 0.05 to 0. U . .4 mm (l.. 678910 E I I I I I I 'I II . Although load can be monitored in such sys- .10-.1 s: ~ . . Other specialized electromechanical systems are available to perform specific tasks. R r 20 I 30 40 ~I 50 I 60 10 80 I I • • '" • 0 0 l> Environment Moist air Dry hydrogen Air Dry hydrogen Hydrogen sulfide Molstair Dry hydrogen 0. Yes ..1 0. They effectively reproduce service environments that impart fixed.20 Fatigue Testing Table 3 Parameter Comparison of electromechanical fatigue systems Forced displacement Forced vibration Tension Compression Reverse stress Bending Frequency range Load range Type: Control Mode Maximum deflection Advantages Disadvantages . Simple. . very limited applications (soft samples) Open-loop Displacement Yes Yes Yes Yes Fixed. Iii ited control (open loop) Alternating stress intensity (:>K). Hz 50 50 5 5 5 I"'" '" / # !II ~# nO od 00 '" ~ • fill' ~ . > u 10-.75 Frequency. Fixed . .'l>~ ~ rP 'U .25Cr-1 Mo steel (SA542-C12) at R =0. resonance.1 0.75 0. vertical displacement and is used to fatigue the specimen. An electric motor-driven flywheel is used to carry a loading arm at a variable distance from the center of rotation. reciprocating displacements to a component or test piece. .f- oOd9 . 1800 rpm Up to 220 kN (500001bf) Open-loop Load 25. - 10" . Versatile. Yes Yes .. ksi vln.05 0..1 0.75 in air. . ~ rPJ pO .00 in. Fig. This rotational displacement is transformed into a guided. much in the same manner as a connecting rod in an automotive engine. One lattice -. hydrogen. spacing -. u ~ :: t ~ lJ I Threshold :>K". 0 0 7 8 9 10 10 • t 6 20 30 40 50 60 10 80 90 Alternating stress intensity (:>K).~ ~ ~ I~rt V~ ~. 23 Fatigue crack propagation data over a wide spectrum of growth rates tional bending. MPa\m Data derived from direct current potential measurements in martensitic 2. straightforward No load control. per cycle e rn ~ U ..10' ~ ~ 1) - 10. durable Fixed frequency. Forced-Displacement Systems Forced-displacement motor-driven systems are the simplest type of electromechanical fatigue testers. Yes . and hydrogen sulfide at ambient temperature. and servomechanical systems are discussed in this article and are compared in Table 3.

the loading frequency of the device is fixed at 1800 rpm. In operation. is achieved by preloading the inertia compensator spring. Because w.040 in. load-controlled system with the ability to accommodate up to 25 mm (1. This technique involves the use of frame-support compensator springs. high precision Low frequency only Open-loop Rotation/ bending fficient. Through special fixturing. The magnitude of the dynamic load is determined by placing the rotating mass at a known distance from the axis ofrotation (r). This technique has two requirements: the rotating frequency (w) must be kept constant and the mass of mounting frame (M) must be kept constant.) Fully closed-loop.) Fully closed-loop. is the acceleration of the frame in the z direction. inexpensive fatigue machines. limited applications tems. durable. and Kz is the spring-compensated displacement in the z direction. The mean or static load.0 mm (0. the forced-vibration rotating eccentric mass system is an open-loop. is tuned to equal Kz. Thus. the natural frequency of the spring (K)/mass (M) system is tuned to the revolutions per minute of the motor. (inertial) where a. forced-vibration devices are capable of testing in tension.Fatigue Crack Propagation 21 Rotational bending Resonance Servomechanical No No Yes Yes 0-10000 rpm Yes Yes Yes Yes 40-300 Hz Up to 180 kN (400001bf) Closed-loop Load l. and masses must be added or removed from the frame to compensate for fixturing to keep M constant. com- . Forced-Vibration Systems Forced-vibration motor-driven systems were the first production fatigue testers in commercial use. M. Thus. high frequencies. m. F. The centrifugal force produced by the rotating eccentric mass (m) is transmitted through the vertically guided frame to the test piece. Because Ma. displacement. simple .otational bending only. Because the centrifugal force usually is totally absorbed by the mounting frame (of mass M). The horizontal component of the centrifugal force is absorbed by the restraining flexure plates. the force on the specimen. Consequently. the load generally drops as failure progresses. K. and large loads are not required. and K are known.0 in. which is a function of specimen characteristics. onto which the dynamic load is superimposed. strain 100 mm (4 in. F = Mw 2r. These systems typically are custom-built. the inertial reaction is separated from the centrifugal force in such a way as to transmit only the centrifugal forces to the specimen. The centrifugal forces of an imbalanced rotor is used to impart a cyclic load to the test piece. neither the specimen nor the rotating eccentric mass (m) "sees" an inertial reaction from the frame. The rotating mass is mounted in a frame that is guided by flexure plates to restrict movement to vertical motion only. Therefore. the fixed displacement precludes the ability to control load. an electric motor is used to rotate an eccentric mass via flexible couplings.) of total sample deflection at loads up to 220 kN (50000 lbf) using special fixtures. because the inertial effects of the frame are totally compensated for by the frame support springs (not the specimen). used primarily for bend tests on soft samples in which load control. extremely efficient Operating frequency directly proportional to sample stiffness Yes Yes Yes Yes 0-1 Hz Up to 90 kN (200001bf) Closed-loop Load. is calibrated directly as a function of r as follows: F = Mw 2r (centrifugal) - + Kz (spring compensated) Ma.

Compression Fig. the dynamic load is achieved by varying the width of the pulse to the magnet beneath the lower mass. Although servo-controlled. bending. or reverse st~ess. senses the load and automatically triggers the electromagnet to achieve self-tuning capability. ) A Bending mom. The load applied to the specimen is assumed to be a function of r. which is large at the supported end of the specimen ~nd zero at the free end..". but incorporates solid-state technology to achieve fully closed-loop control of mean and dynamic loads. and reverse-stress fatigue tests. The dynamic load. achieved high loads (up to 90 kN. or 40000 lbf). high loads (up to 180 kN. or 20000 lbf). is electronically ma~n­ tained at a preset command level through solidstate closed-loop circuitry. compression. Gage area Drive motor Hotation 0~:f: I Tension '----. respectively. torsion. A strain gage load cell. in the form of a sine wave. like the mean load. the Amsler resonant fatigue testers were instrumental in obtaining the vast amount of fatigue data currently available. in series with the specimen.22 Fatigue Testing pression. Rotational Bending Systems Rotational bending systems effectively apply reversed loading to the outer surface of rods or shafts. The resonant system is based on a similar principle. 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. By applying a known static force at the end of the shaft. linked by the specimen to achieve vibration-free resonance. is subjected to a nonuniform bending moment. in which the specimen is used. The preload is maintained automatically during the test. a bending moment can be applied to the test section. the open-loop nature of the system prevents direct load measurement or control. The cantilevered specimen. the outer surface of which oscillates between tension and compression during each rotation. low power consumption (around 750 W maximum for closed-loop sys- . thereby maintaining a constant. Figure 24 illustrates the rotating-beam operating mechanism and the resulting stress distribution in the specimen. bending.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. To produce a more meanmgful uniform bending moment throughout the tes~ piece. in which closed-loop load control. During resonance. It is based on a resonant spring/mass system. The high efficiency of resonant systems makes them well suited to high-cycle fatigue tests. like a spring. mean-load-maintenance systems are available. Capable of tension. controllable dynamic load. 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. maintain preset loading conditions (S~1Ut­ ting down at preset load levels or frequencies). and power the electromagnet. a specially designed tapered specimen should be used or bending moments should be applied to each end of the specimen. which is characteristic of closed-loop systems. however. and a graduated scale is provided to permit reasonably accurate setup. torsional. the electromagnet restores any hysteresis energy lost during the previous cycle. This system uses dual opposing masses (unlike the single oscillating mass/ seismic base of earlier systems). The fatigue load. and consumed minimal energy. as an integral part of the oscillating mechanism. 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. The dynamic load is achieved by pulsing an electromagnet at the natural frequency of the spring/ mass system. The mean load is achieved by physically moving the upper mass up or down to achieve tension or compression.

In addition. primarily intended for low-cycle and creep-fatigue studies. The closed-loop servomechanical system is. or an open-loop condition) and when it is closed (providing feedback to the system). Displacements may have to be controlled (often for many days) to within a few microns. with the modular concept of servohydraulic systems. With any type of control system. See text for details and explanation of symbols.Fatigue Crack Propagation Electronic demand signal 23 1-----. the objective is to obtain an output that relates as closely as possible to the programmed input. K. S I' has been added to the diagram to permit analysis of the system when it is open (no feedback. at a frequency of 1 Hz over a force range of 0 to 100 kN (0 to 22000 lbf). and forces can range from 100 kN to just a few newtons. In a fatigue testing system. resolution. Extreme demands for sensitivity. stability. a power supply. it may be desired to vary the force on a specimen in a sinusoidal manner. 25(b).---. and high throughput are required. 25 Simplified block diagram for a negative-feedback closed-loop testing machine tems). the interaction between system components is illustrated. 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. and a solid-state closedloop electronic control console. by virtue of its design. and reliability are imposed by fatigue evaluations. Usually. rep- . The transfer functions of each of these blocks can be combined to produce the more simplified diagram shown in Fig. Fig. in particular. In this section. 25(a) represent a group oftypical components of a testing machine. Closed-Loop Servomechanical Systems The most recent development in electromechanical fatigue testers is based on an electric actuator/load frame assembly. a load cell. 25. Servohydraulic Fatigue Testing Systems Servohydraulic testing machines are particularly well suited for providing the control capabilities required for fatigue testing. The blocks shown in Fig. An overview of the basic principles of operation of negative-feedback systems is provided in Fig. a load frame. Closed-loop systems compare live feedback signals to an input command signal to maintain accurate control of preset conditions. The system closely resembles its servohydraulic counterpart in that it consists of an actuator. the principles underlying closed-loop servo systems are discussed briefly. Placement of the switch. and a brief description of their operating principles and characteristics is provided. This wide range of performance can be obtained with servomechanisms in general and. The only practical means to accomplish this with precision is through the use of a negative-feedback closed-loop system. (b) Transfer functions.c (bl (a) Typical components. These systems tolerate minimal hysteresis and produce optimum testing results when used with stiff metallic samples. The equation governing this simplified open-loop system is: C= KoD where Crepresents the controlled output.

In a fatigue machine. or actuators or cylinders. A servo-controller incorporates numerous other compensatory features. and system startup and shutdown The servo-valve controls the volume and direction of flow of hydraulic fluid between the hydraulic power supply and the hydrauI~c ram. Load Frames. resistance to extraneous loading. Unfortunately.b. overload rating. For example. load capacity. such as: • Auxiliary functions such as recorder signal conditioning. The analog is: 1 V represents 1000 N. deflection constant. the output pressure is approximately proportlOn~1 to the square of the input current when the flow IS constant. fatigue rating. An external applied force causes the elastic deformation of an internal member to which a strain gage bridge has been attached. Servohydraulic System Components Many commercially manufactured units are available for each component in a typical servohydraulic testing system. to the control device in the system (usually a flow-control servo-valve). Within the control loop. Some load cells are designed specifically for fatigue evaluations. The force available during dynamic operation depends on the pressure drop and flow characteristics of the servo-valve. Reference should be made to the load/flow/pressure characteristics supplied by the servo-valve manufacturer. or wear. because it can be influen"ced by several common system variations. K is seldom a constant. The programmer supplies the command signal to the system. An electronic sign~l that is proportional to the resistance change m the bridge and to the applied force can thus . calibration. The behavior of the hydraulic components may change with tempera~ure.24 Fatigue Testing resents the open-loop transfer function. Also. it compares the command signal with a signal produced by the controlled parameter (stress or strain. Hydraulic rams. The effective area of the piston is therefore equal to the cross-sectional area of the piston minus the cross-sectional area of the piston rod. . These rams usually are double ended to provide the greatest lateral rigidity and to produce the balanced flow and f~rce characteristics desirable for push-pull testmg. furnish the forces and displacements required by the testing system. the reaction forces to the specimen and to the housing of • Means to adjust the gain or proportional band of the system • Controls to modify the feedback or correction signals for improved stability • Controls to adjust the mean level and amplitude of the command signal(s) • Controls to enhance and adjust servo-valve response . Therefore. • Means to monitor the system error signal (a measure of how well the command and feedback signals agree) • Capability to select various command and feedback signals . the output is simply proportional to the system demand if K is a constant. or the~r gain may vary. 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. and the mechamcal components may vary because of thermal effects or friction. The servo-controller makes most of the adjustments necessary to optimize system performance. Load Cells. 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). Characte~ISt1CS of the device are such that the output flow IS approximately proportional to the input current when the output pressure is constant. which is generally an analog of the desired behavior of the controlled parameter. and D represents the electronic demand signal. hysteresis. contamination. The strain gage load cell is the most widely used force-measuring and feedback device in closed-loop fatigue machines. which can supply large force~ a~d displacements to the specimen. it is the intermediary between the low-power servo-controller and the hydraulic ram. Variable features include sensitivity.e produced. In this instance. linearity. for example) and relays a correction signal. The system can then be adjusted to produce the correct output. the maximum force capability of the ram will approach the hydraulic supply pressures multiplied by the effective area. natural resonant frequency. Most commercially available cells are very competitive with respect to these features. The electronic components may drift slightly. and compatibility with t?e testing machine and fixtures. For example. temperature stability. if needed. Under static conditions (very little flow). Any change in the programmer signal will result in a corresponding change in the controlled parameter.

. additional flow is required from the servo-valve. compressive as well as tensile forces). The extra rigidity can be obtained by increasing the diameter of the support columns or by utilizing three. Therefore. This is generally considered necessary in the design of fatigue machines. but for fatigue purposes the frames should be customized. When a deflection occurs in the load frame. The requirements of good high-frequency response demand that there be high axial stiffness in the load frame. because fatigue specimens must be subjected to fully reversed loading (i. some means is usually provided to refine the alignment with manual adjustments when necessary. In addition. this deflection should be minimal in comparison with the deflection imparted to the specimen. lateral rigidity must be increased to resist bending. In addition. A strain-gaged specimen can be used to make this evaluation.Fatigue Crack Propagation the ram are supplied by the load frame.e. Many styles ofload frames are available.or four-column configurations. 25 . Exceptional alignment is required of load frames used in fatigue evaluations to minimize undesirable bending forces.

.

%h Park OH.A stress ratio R oi I indicates no variation in stress. If the stresses are fully reversed. S-.. A family of S-. to fracture using a logarithmic scale for the number of cycles. If the stress is cycled between a maximum stressand no load.. the stress ratio R becomes a positive number less than I. lithestress isqcled bewssn two tensile stresses. l!TS and I’S indicate ultimate tensile strength and yield strength. Stress ratio is the algebraic ratio oftwo specified stress Lalues in a stress cycle. hlcr& . p 667 Edann.‘Sm) and the ratio. The results of fatigue tests are usually plotted as maximum stress or stress amplitude to number of cycles.Vcur\es typically halea fairI> straight slanting portion at low cycles changing into a straight.ofthealternatingstressamplitudeto the mean stress (A = Sa. with a sharp transition between the two. horizontal line at higher cycles. and the test \\ould becomea sustained-load creep test rather than a iatigue test. An S-.I’cur!e usually represents the median life for a given stress-the life that half the specimens attain.V curves for a material at various stress ratios. respectiveI). R becomes a negatke number less than I. R. A. in uniaxial tensile testing... S-N Curves Typical for Steel 27 Schematic S-. .ersed. the stress ratio R becomes zero./ S. The resulting tune of data points is called an S-. Volume I. 1978.Vcur\e. Proprrr~rrnnd Srlei~~on. For carbon and lo\{-allo! steels.l-1. Twocommonl~ used stress ratiosare the ratio. of the minimum stress to the maximum stress (R= S. the stress ratio R becomes -I: ii the stresses are partially re\. Source Metals Handbook. . Scatter of fatigue li\es can cover a \ery ibide range. lronsand Sreelr. . Stress is plotted on either a linear or a logarithmic scale.V.Z’curves for a material tested at various stress ratios is shown schematically in the above curves.Amcrwan So&t! ior hlctals.).

as does the scatter from longest to shortest life. for purposes of illustration this is the starting place. many nonferrous metals and some very high-hardness ferrous metals tend to continue their downward path at very large numbers of cycles.L_ _. for the specimens can sustain no higher load without fracture.g c~ g>:. while the other supposedly "identical" specimens or parts had lives of from about 1.28 1-2. Q. One such specimen is shown at the 60 percent level fracturing at about 150 cycles. as the S-N curve starts to flatten out.000) Number of cycles to Iracture S·N curves that typify fatigue test results for testing of medium-strength steels. When the load is dropped to 70 percent.. Dropping now to 50 percent of the single-load fracture strength.000 to 10. Again dropping the load. Intuitively it is known that if the maximum load (or stress) is lowered to 90 percent of the tensile strength. 20 10 O~ _ _-L-_ _-L-_ _. S-N Curves Typical for Medium-Strength Steels 100 Fracture region (all specimens fractured) s: 80 Fatigue .L 10' (10) 103 _ 10° (1) (1. it will require more than one load application to fracture the specimens. the lives get longer and the scatter in fatigue life increases to perhaps·5 to I.. this lone early fracture specimen is simply ignored. they will run longer with a fatigue life scatter of perhaps 3 to I. The test . they are not). might be found.Iraclure band rn ~ ViS: lii rz: CIl 0. although it is possible that some metallurgical reason.000) 10' 10 6 107 (100. Now. the problem is when to stop the tests. but in fact the scatter in life from longest to shortest is on the order of more than 2 to I. if the single-load fracture strength of the specimens is considered to be 100 percent. and normal fatigue fractures have no obvious plastic deformation. If the load is dropped to 80 percent of the single-load fracture strength and ten more specimens are tested. the results are placed as points at the top of the left axis at one load application. Invariably. The cause of such an early "anomaly" is often sought in vain.. 0_ .000.000. If ten specimens are fractured. such as a large inclusion on the surface. Frequently.000) (10.000) 10' (10.. Because the scale is logarithmic. of course. the fatigue lives increase dramatically. The ten points shown in the diagram at 90 percent represent the possible life to fracture of each of the ten specimens. there is at least one specimen that inexplicably fractures far earlier than any of the others in the same group. plastic deformation of the test specimen is likely to be great. the points appear to be relatively close. This flattening out is characteristic of ferrous metals oflow and moderate hardness. now to 60 percent of the single-load fracture strength. such as in bending a paper clip or wire coat hanger to make it fracture. As an explanation.000) (1. At this high stress. ~ i! CIl :: (. Actual parts are not intentionally designed to operate in this regime.L_ _. even for theoretically identical specimens (which._ . in actual fatigue testing. the fatigue lives again increase.000 cycles.L .. 40 30 Fatigue limit Infinile-lile region ~-' 0.'" 70 60 50 Finile-Iife region (no specimens fractured) II .' .L_ _. which is not unusual.

they should have infinite life. However. for it did not actually fracture. purpose. Also. characteristic of ferrous metals but not of most nonferrous metals. that is. For example. American Society for Metals. If ten million is selected as the end point. 29 Source: Donald J. five hundred million cycles is sometimes used in the aluminum industry. five million. or cycles per minute. S-N Curves Typical for Medium-Strength Steels (continued) machine will be needed for another test specimen after a very long test time. This is typical of certain structural parts in aircraft which have their histories carefully recorded so that they may be inspected and/ or replaced as their fatigue lives are used up in service. depending upon the metal. growing fatigue cracks must not be permitted to exceed the critical flaw size characteristic of the metal and the stress state. depending upon the rate ofloading. because specimens that are tested at stresses below the curve should run indefinitely. and the point shown with an arrow pointing to higher values. Frequently. The region below the lowest portion of the S-N curve is called the infinite-life region. or even one million. The leveling of the S-N curve is the fatigue limit.1-2. the region to the left of the sloping part ofthe S-N curve is called the finite-life region. the test must be stopped at that figure even if a specimen is unbroken. 1985. and urgency of the tests. for at the higher stress levels the test specimens or parts will eventually fracture in fatigue.P 135-137 . Wulpi. Understanding How Components Fail. cycles is selected as the end point. Metals Park OH.

• .". To determine the endurance limit of a metal. The results of fatigue tests are commonly plotted on diagrams in which values of stress are plotted as ordinates and values of number of cycles of stress for fracture are plotted as abscissas.L.·c ~ . N for number of cycles).. J II. Oil quenchedan 1 drawn' ~~ \ ~ K"'!> ~.. 1..000 cycles.fio. . £ ~ r-... Succeeding specimens are then tested. Specimens stressed below the endurance limit will not rupture.000 to 50. ~.000 90.000 10.000 '~6QOOO ~...S. as nearly as can be determined.... ~~e>/ l' 'wI. Such diagrams are called S-N diagrams (S for stress. Source: Fatigue and Creep Tests of Metals.30 1-3.I II.. P 220 . I II. . each one at a lower stress..C:P~ l'fo~"ea-lea I <:0.- .1 III a Not: 4 10 .N diagrams for duralumin and monel metal do not indicate well-defined endurance limits.N diagrams are drawn using semilogarithmic plotting as shown in the above diagram.l.sttel. The first specimen is tested at a relatively high stress so that failure will occur at a small number of applications of stress.. " ~ 30POO 20.Q' 10. I I.... J.. thus indicating a well-defined endurance limit. I r.1 ao'tLerMcarOon.. 0. the S.mfn ~/c..000.000 80.. which presents the results for various typical materials.7quenchedolnd - 0."lot rUf(furr I I 10 5 106 10' Number of c~c1e5 for rupture.t50.Ioq scale 108 I Typical S-N diagrams for determining endurance limit of metals under reversed flexural stress... rr (nd(cO/fis flf!f.!.000 0.. for values of N ranging from 1.I III .t>e: ~ '! ts« as rolled :l. ("o$e<tand drawn firn I. 40.000 C <Q-/'r. it is necessary to prepare a number of similar specimens that are metals tested. The number of repetitions required to produce failure increases as the stress decreases. the S-N diagrams become horizontal." ~ I· v. cen/-carbon 'steel O'i''7ve".."". c-J..000 <I) dJ. The S.~~n ~aJfn i?o.000 <J'~ 'x RC'-1. S-N Diagrams Comparing Endurance Limit for Seven Alloys 100.Jpe. and for most nonferrous metals. ! r.000.. In general.

40% carbon range. Above 40 HRC. Above 40 HRC. Below 40 HRC..00% carbon steel. the total ratio range is from 0.. whereas below 40 HRC the reverse was true.. including low-carbon martensites. much of the early work was with medium-carbon steels with intermediate hardnesses. Many believe that tempered martensite gives optimum fatigue properties.. Source: D... bainite had better fatigue properties at the same hardness than did martensite.P 77 ..40% carbon martensite is about 55 HRC. " 0..26 to 0.9 "i :--:. factors other than microstructure become more significant.. Steel: Effect of Microstructure I. but below 40 HRC the carbides had a "pearlitic mode.5 c 0.2 0.5 is a conservative number. The results of some of his work are summarized in the above bar chart. They used 5ll00. Breen and E.4 :> o e 0.7 'e 0. Borik and Chapman determined the endurance limit of bainite and martensite in the range 36 to 61 HRC.8 '<:: "i 0. H. 0 r-- 31 . especially nonmetallic content and residual stress.55 to 0. the carbide associated with the bainite was very fine and well-distributed. His data confirm that 0. Metals Park OH..1-4. "Fatigue in Machines and Structures-Ground Vehicles.----.62 for highly tempered martensites. Wene.62..1 -g w o Effect of steel microstructure on endurance ratio. When untempered martensite is included. a 1." which was less favorable in resisting fatigue. American Society for Metals.. Untempered 0.3 0.6 e 0. the carbides in the martensite are spheroidal. They explained the results in terms of carbide morphology and distribution. M. 0. Only limited data are available for other structures. However.. One of the more extensive investigations on influence of microstructure was conducted by Cazaud. 1979. They concluded that above 40 HRC..." in Fatigue and Microstructure. These data were also for steels in the 0. he found ratios varying from 0..

. a larger volume of metal is subject to high stress in a large part than in a small-diameter part. The 0. there is a higher probability of a critical-size nonmetallic inclusion to be in that volume. H. nonmetallic-inclusion content. Sinceaxial tests.. It must be borne in mind that this is a \er) rough estimateand that the cur\es shown in the above graph are thought to be touard the conser\ati\e side of scatter bands. .32 l-5. b’ene. C. and carbon content 31 higher hardnesses. C. Steel: Influence of Derating Factors on Fatigue Characteristics Derating factors for influence of surface condition on fatigue.J’inches and 0. as in bending. p 72 Vshlslcs. hlerals Park OH.1 and 3. is onl! reasonably accurate in the low and intermediate hardness ranges because of limitations related to microstructure. hf. ‘.0 inches.and S.. Earl) work b> Horger firmly established that large-diameter samples of the same metallurg) were not asgood in bending fati@e as weresmall samples. b> their nature. Source: 0. In the presence ofa stress gradient. in Far~gue and hlcroaucture. test fairly large volumes at maximum stress. Since a large volume is subject to maximum stress. 1979. The graph above gibes C.9 for diameters between 0. is significant. is usuall) taken at I. they also gibe lower-bound results. The fatigue properties established by testing large specimens are thought to represent the lower bound for a large number of small samples. the size factor. “Fatigue in hlashines and S~ruc~urcs~Ground American So&t) for Metals. since it is impossible to represent such variable conditions by a single cur\e. factors for various surface conditions. Brcen and E.5 relation for S.0 for diameters less than O. It should be obvious that these factors are approximate.

0 0. . .0 1.0 From chart above (aI A lower value 10. Correction factors for surface roughness (kJ.4 in. where: d c 0. .n. .0 0. conditions on fatigue limit of steels at Source.9 0.9rh Parh OH. .58 1. K. Comparative effects of various surface various levels of tensile strength. . Ironsand Srscls. .85) may be used to account for known or suspected undetermined bending because of load eccentricity. . type of loading (14). 0. . < d 5 2 in. 1. 19% p 6’1 Edlrlon. American SO~ISI) ior hlclalr.l-6.0 Kd. 0. Properr~erand Selcc[ia. hleralr . . Volume I.9(a) 1. . Metals Handbook. for fatigue life of steel ports. Steel: Correction Factors for Various Surface Conditions 33 Value for loading Factor Bending Torsion in Tension K. . .4 in. and poti diameter (KJ.9 1.06 to 0.

and consequently all the research in the field. p 689 KumarChauls. nonferrous metals. etc.E. ferrous metals. rockets. with considerable success. to fatigue problems. .34 l-7. It is this class of materials which has shown catastrophic failures in fatigue. pressure Lessels. airplane fuselages. Source: Marc Andrt Meyersand Krishan Engleuood Chfk NJ. etc. 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.. entire structures or very large structural elements (e.. This can beattributed to the use of materials such as high-strength alloys. the behavior of a material under conditions of fatigue has been studied by obtaining the S-. Such an endurance limit does not exist for nonferrous metals (curve B above). is the endurance limit. 198-l. gears. Starting in the late 1950s. resulting in monolithic structures meant to undergo high cyclic stresses in service.g. B. Inc. The relation between Sand . is not a single-ralue function but serves to indicate a statistical tendency. in general. For steels. S.Appllcaionr.. was confined to moving mechanical components (e.is the number of cycles to failure. Traditionally. axles. Up until the 196Os. it must be pointed out. hlechanlcal hle~allurg): Prmaplesand .V.‘l’cur\es (see above). Prcm~ce-Hall. and it is for this kind of material that fracture mechanics is being applied. together with the advances in the fabrication technolom. where S is the stress and . Fatigue Behavior: Ferrous vs Nonferrous Metals S (stress)--N (cycles to failure) curves. almost all fatigue failures.).) have been studied and tested for fatigue. 4.g.

Comparison of Fatigue Characteristics: Aluminum Alloy Mild Steel vs 35 Typical bending (R = -1) fatigue rous and nonferrous metals.l-8. the point on the cun’e where the number of cycles to failure becomes a straight line-essential11 infinity. . curves for fer- Here it is noted the lack of the “knee” for the aluminum allo) compared with steel: that is.

-----.. Carbon Steel: Effect of Lead as an Additive 120~-. Syracuse NY. V> Fatigue limit of leaded and nonleaded alloy steels as a function of ultimate tensile strength. Source: George M...---.---.-----....36 1-9.. John J... Eds. p 68 . Sinclair. 1964.--.. Reed and Volker Weiss. Burke.--r----r--r---. although usually at the cost of a minor (usually) loss in mechanical properties. Lead is often added to steels to improve machinability. The interrelationship of lead additions with tensile strength and fatigue limit is summarized in the above graph.. Syracuse University Press.---. Norman L. IOo/-----1I--+----1--+-----+~~ o BOI-------t---t----r-----: o o n."in Fatigue-An Interdisciplinary Approach. "Some Metallurgical Aspects of Fatigue..

however. In the presence of a corrosive environment the situation will be different. Whitwham. If a specimen is subjected to alternating stress (tension and compression in turn) over a range insufficient to cause immediate fracture. Edward Arnold (Publishers) Ltd and American Society for Metals. Metals Park OH. the stress range below which the material will endure some specified number of cycles (the number must be stated). should be indefinitely long as shown above. it might be hoped that the dislocations would merely move to and fro along the plane. in absence of corrosion. so that gliding will then start on another parallel plane. following glidingplanes inclined at such an angle as to provide high resolved shear stress. In practice a large number of cycles can be withstood without apparent damage. (2) if it exceeds a certain level (the fatigue limit) the gliding will become so irregular. although stress-corrosion cracking is often intergranular. convenient to determine an endurance limit-namely. An Introduction to Metallic Corrosion. they followed grain-boundaries for short distances. Corrosion Fatigue: General Effect on Behavior 37 t Fatigue without corrosion Fatigue limit Logarithm of number of cycles needed for fracture _ Effect of alternating stresses with and without corrosion. gliding may occur within some of the grains. preferential attack may reasonably be expected even below the fatigue limit. Disorganized atoms along a gliding-plane may require less activation energy to pass into a liquid than more perfectly arrayed atoms elsewhere. but in material as we know it. studying corrosion-fatigue cracks on steel. and roughening along the original glidingplane will make movement difficult. There are exceptions to both rules. In the end. bands of material will have become disorganized. It is. first locally. This means that there is no "safe stress range" within which the life should be infinite. producing gaps. 3d Edition. although mainly transgranular. P 160 . certainly. found that. which later will join up into cracks. as to cause separation between the moving surfaces. slight irregularities will prevent smooth gliding indefinitely. while the atoms are in motion along a gliding-plane. below the fatigue limit. and that no damage would result. where such boundaries chanced to run in a convenient direction. but when the disloca tions reach a grain-boundary they are halted. Source: Ulick R. corrosion-fatigue cracks are usually transgranular. gliding will cease altogether. If the material were ideal. the only changes still produced by the alternating stress being elastic. there will be failure. Thus above the fatigue limit (after a time which is shorter at high stress ranges). It should be noted that. the life. retracing their movement along the gliding-plane when the stress is reversed. and ultimately one of two things must happen: (1) if the stress range is low. Evans. 1982.1-10.

.ln. 'Copper' steel 40 80 60 . Metals Park OH.in. C Tensile: 81000 Ib. Cr.38 1-11. Effect of Corrosion on Fatigue Characteristics of Several Steels l00r------------. Cr. ~ e ~ 60 :E "" 'e l!! 25 104 105 . Two main procedures are available for corrosion-fatigue tests: One-stage tests. and the residual strength is estimated by measuring either (a) the endurance limit in the absence of corrosive influences (i.Zsq. Cr. 100 .5 12·9')(.5 .e. g- 0·14')(.lsQ. The logarithm of the number of cycles needed to produce breakage is generally plotted against the stress range. Here the corrosion fatigue is interrupted after a definite number of cycles. Here the corrosion fatigue is continued until breakage. C Steel (annealed) Tensile: 103500 Ib./sq.e Ii. 35 i:1 <: :> 0 ill 80 e ill :E " e '" e 'E (/) 0 103 104 105 106 107 lOB :E " <: '" 'E (/) ~ 40 " 20 0 103 "R". C Tensile: 89600 Ib. P 165 .in. Edward Arnold (Publishers) Ltd and American Society for Metals..0·11')(.0·14% Vo.1£ (/) ~ 50 " vanadium 20 steel (hardened and temperedl " E O'BB')(. 45 lOB .0'46')(.0'2')(. (b) the number of cycles needed to produce fracture in the absence of corrosive influences at some definite stress.lsq.lsq. C lOB 0 104 Tensile: 150600 Ib. ~ 66 . i: 106 107 lOB 104 105 106 107 105 106 107 lOB N = Cyclesto fracture 1I0g scale) ---Denotes testsmadein air 6-----" "stream of fresh water II saline riverwater Typical curves showing the number of cycles needed to produce fracture at different stress ranges in absence and presence of corrosion. " ~ 40 20 0 O 104 c. l' 09')(. An Introduction to Metallic Corrosion.in gIi."~ (hardened and tempered I Tensile: 65700 lb..~ 30 104 \4 105 chromium steel 27')(.e '0 i:1 <: ill :> 0 c. 105 106 ~ 107 (hardened and tempered) :c u . lOB 106 107 40 Chrome- . Source: Ulick R. Evans.e '0 w "0 :> 0 gc.O'9B')(. (c) the tensile strength. as in the above curves selected by Gough from McAdam's experimental data. or (d) the shock resistance (Izod number). 1982./sQ. Two-stage tests.in. the stress which can be withstood for some definite number of cycles. C.in. 3d Edition.5 :c u 20 ii . Cu Tensile:61500 Ib. .

---.2 Hz) have the highest practical importance. P 390 . Source: M.--. whereby low frequencies «10. 1 6 7 0 9 0 S-N type of fatigue curve... Under these circumstances a structural component can be subject to fatigue which is conventionally described by an S-N curve relating the cycle life. Metals Park OH...-.. "Design of Equipment to Resist Hydrogen Fatigue Service. Windgassen.--. the external load changing with time. 1982.---_. In non-aggressive environments an endurance limit can be defined below which no fatigue failure occurs. A disadvantage of this approach is that S-N curves do not differentiate between crack initiation and crack propagation." in Current Solutions to Hydrogen Problems in Steels.. The number of the cycles corresponding to the endurance limit presents initiation life primarily..-. Consequently S-Ntype data do not necessarily provide information regarding safe-life predictions in structural components.. Particularly...--.---.. Interrante and G. Kesten and K.. C...--.-. M. I i I 1+--.. G.---.--.+---. American Society for Metals... these are likely to reduce or even eliminate the crack initiation portion of the fatigue life.. whereas the number of cycles for crack initiation at a high value of applied stress is negligible.-. to applied stress. Steel: Effect of Hydrogen on Fatigue Crack Propagation 39 PR PAGAIION LIfE I -. In the majority of all cases.. S.. as in the above chart.....----.c--..... N. Pressouyre. if the structure contains surface irregularities different from those of the test specimens.-F. Eds.r+--.-.1-12.

R = -I equivalent stress for R = 0. but substitution of the fatigue strength at specific finite li\es can also be used (see chart a abole). Relationship of Stress Amplitude and Cycles to Failure (a) Finite life ASR diagram. D. Source.Vcur\e Designers have the ability to calculate the component’s stresses using classical formulas or the computer-based finiteelement-analysis (FEA) techniques.Y diagram.e. H Brscn and E hl Hens. rial’s R = -I S-.” in Fatigue and hlwrosrruc~ure. IYTY. Life estimations from the diagram can be done using such information as is shottn in charts a and b abo\. Here the hno\\n stress range at some Rvalue is conierted to an equivalent completely reversed (R= . (b) for R = 0.I) stress. .6 loading.6 loading using R = -I equivalent The ASR diagrams normally use theendurance-limit fatiguestrength value. or stress concentration. and thisequkalent stress is applied to the matefor the life estimate. showing life prediction stress. Both of these methods examine the elements for the maximum stresses that are normall> in the areas of a discontinuity. \ ehlcles.40 l-1 3. “Faugue m hlachlner and StrucrtmpGround Metals Park OH. showing S-. p 6’ Amencan So&l! ior hkrals.

and J c is the fatigue-ductility exponent.1-14. The summation curve is analogous to the stress-life. A schematic representation of these relationships and their superposition is shown in the above diagram. E is the modulus of elasticity. if the strain amplitudes are replaced by their respective stress amplitudes.Life Curves 41 2Nf Reversals to failure (lag scale) Strain-life and stress-life curves. Jr. The intersection of the Basquin and Coffin-Manson plots is normally defined as the transition between high. American Society for Metals. the fatigue life should be related to the plastic-strain amplitude. 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::. and Gerd Lutjering. "Cyclic Plastic Deformation and Microstructure. 2N is the number of reversals to failure. Source: Edgar A. Coffin and Manson independently proposed a relationship between the plastic-strain amplitude and the cycles to failure of the form: AEp/ 2 = EJ (2NJ).E. Strain. the ductility) of a particular material. Consequently.E/2 = aa = aJ (2NJ)b where I::. 1979./2 is the elastic-strain amplitude.E. p 211 . Metals Park OH. Fatigue damage is caused by cyclic plastic strain. Wohler diagram. where Ej is the fatigue-ductility coefficient. and is the fatigue-strength coefficient. the regime of low-cycle fatigue depends on the properties (for example. his the fatigue-strength exponent.and low-cycle fatigue.Life and Stress. a.. Starke. and consequently." in Fatigue and Microstructure.

respectively.152 l l 5 10. Fatigue Plot for Steel: Ultrasonic Attenuation vs Number of Cycles 0. Results are interpreted in terms of a series of microcracks being formed. The attenuation started to increase at about 6 X 105 fatigue cycles (65% of fatigue life). American Society for Metals. Alers. Source: O.5 X 105 cycles (85%). an additional pulse was observed.168 0. the attenuation is primarily determined by the transmission coefficient of this single crack.42 1-15. it will reflect part ofthe pulse. Metals Park OB.232 ADDITIONAL PULSE OBSERVED 0. As soon as a macrocrack has been formed (by coalescence of microcracks). with the acoustic pulse reflected at the back surface of the material. The measurements have been performed in a pulse-echo mode. Thus.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.200 « :::> Z « ~ 0.::: 0. probably at the surface.216 0 .249 ~ t:I E 0. As soon as the microcracks are sufficiently deep. p 135 . arriving earlier than the one reflected from the back surface. they will change the bulk attenuation.264 0. 1979. the attenuation curve (versus fatigue cycles) becomes discontinuous. The above chart shows their results obtained on cold rolled steel bars. as may be noticed in the above chart.4 dBATIENUATION CHANGE OBSERVED CD :5! z 0." in Fatigue and Microstructure. At roughly 7. Buck and G. A.184 0. "New Techniques for Detection and Monitoring of Fatigue Damage. After that.

18 \ <t: 16 ~ --fATIG-U-E-LlMi. failure will not take place until this has been done a certain number of times.2-1." in Failure Analysis: The British Engine Technical Reports. would be of the order of 16 t.i. Hutchings and Paul Unterweiser. a graph such as the one above will be obtained. For example. TYPICAL FATIGUE CURVE FOR M.p.. If testing is continued in this manner. since if the stress applications are continued for a sufficient number of times. Further.p. the concept of a fatigue limit does not apply. while at a lower stress still..i. If the same steel was tested under conditions of reversed bending stresses a value of the order of± 12 t. If. UNDER REPEATED AXIAL "TENSION (f MIN. may be found. This value is known as the fatigue limit of the material.p. Hutchings.p. P 344 .s. If the results from such a series of tests are plotted.TEtlSION) I~ 14 o I 2 I 4 I G -8 10 ~ \2. = 0) rc't- I ~ 22 « )( I I '-' III III uJ 0:: III i 20 \ t J- 0 W .\ONS.i. a stress value will ultimately be found at which fracture will not occur.s. in these circumstances.s. Metals Park OH.i. 1981. 0. a sample of mild steel may have a maximum stress of 27 t. "Fatigue Failure of Components of Lifting Machinery. It must also be pointed out that where corrosive conditions operate in addition to fluctuating stresses. a stress of say 25 t. R. the corresponding value ofthe stress being known as the fatigue limit.s. no matter how many stress repetitions are applied. Source: F. F. however. 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. Typical S-N Curve for Low-Carbon Steel Under Axial Tension 28 43 26 ~ I c ~ ~ 24 1. failure from "corrosion-fatigue"may occur and. the number ofload cycles required to cause failure will be even greater. ultimate failure will occur. is applied repeatedly to the same material. American Society for Metals.MILLIONS The term "fatigue" refers to the failure of metals from repetitions of stress rather than from a single application.:r-(REPEATEO. Eds.i. NUMBER OF STRESS APPLlC".p. R.s.J 0.S. most nonferrous metals and alloys do not possess a fatigue limit. the curve tending to run parallel to the abscissa after approximately IOmillion cycles (for steel). 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. when subjected to a single application ofload as in an ordinary tensile test.

J oct z 4 0 ff- oct 2 a 10 3 104 loS 10 6 REVERSALSTO FAILURE...2 00. the fatigue life remained approximately the same or decreased slightly when compared to that ofthe undeformed material. 0. unlike BBS.. Thus. C'l .J oct z 4 0 ff- oct 2 r 103 104 Runouts loS 10 6 REVERSALSTO FAILURE. 2N.4 a 0.. "Effect of Cold Formingon the Strain-Controlled Fatigue Properties ofHSLA Steel Sheets...4 a 0.2 00. AISI 1006: Effects of Biaxial Stretching and Cold Rolling 8 cP Eoff C'l . Strain-life plots for two modes of deformation for 1006 steel. b 8 Eeff )( ~ s c f- . After cold rolling. American Society for Metals. in contrast. Included is the data band for the undeformed material. Metals Park OH. P 217 ." in H'Sl.A'Stccls-c-Technology & Applications. Source: John M.r 6 . at small strain amplitudes the fatigue life increased as a result of the prior deformation. Another difference between the two deformation modes was that the scatter ofthe data was larger after BBS than after CR. 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). 2N.6 w· :::> :::i ::l!: Do « a: fVl .. The effect of balanced biaxial stretching on fatigue life was as follows: at large strain amplitudes (/:. CR appeared to cause no reduction in fatigue life at short lives.EI/2~ ~ 2. b )( .5 X 10-3 ) . those in the bottom chart are for cold rolling. Thus.. Plots in the top chart are for biaxial stretching.r 6 "j :::> :::i ::l!: Do « a: lii . Charpentier..44 2-2. 0. Holt and Philippe L. 1984. BBS was somewhat more detrimental to the fatigue life than CR.6 ~ s c f- .

3mml Smooth Specimen --<>-.-. Metals Park OH...0 0-. Frederick V. 10 5 10 6 NT' CYCLES TO FAILURE Fatigue strengths of FCA W/TlG.._ oro ~ '-._ '-''''lJ. American Society for Metals.dressed joints compared to those without TlG dressing for AISI I006steel (unwelded)... FCAW. "Weld Fatigue ofTlG-Dressed SAE-98QX HSLA Steel..--fr-._...!.. The improvement in fatigue provided by TlG dressing tbe welds is obvious.. Lawrence.0 0 roo 6_ --rrtr-lS" i:J_ _ tr 6 ~D.~_ 0 0 0 -o 0 .TIG-Dressed --~-- _..... Source: Kon-Mei Ewing.2-3...As-Welded _ .p 557 . o DO ~.."in HSLA Steels-Technology & Applications. TIG Dressed 45 SAE-1006 R' 0... t' 0... Houchens._. AISI 1006: Weldment..g 0 00 0 . and Albert F..13" 13._ '--0 . 1984. Pei-Chung Wang.1. Jr.

p 562 . Source: Kon-Mei Ewing. AISII0061ap-shear joints. AISI 1006: Weldment. Lawrence." in HSLA Steels-Technology & Applications.. Jr. Metals Park OH.1 • EXPERIMENT PREDICTION <J • • • . Pei-Chung Wang. Houchens. 1984. (a. Note that the results and predictions compare closely. Frederick V.46 2-4. and Albert F. ur= 32 KSI 10 5 106 I NT CYCLES Total fatigue life predictions and experimental results for FCA W. "Weld Fatigue ofTIG-Dressed SAE-980X HSLA Steel.77 I R = 0. American Society for Metals. Shear Joints CJ) ~.. or \0 1 SAE 1006 LAP-SHEAR WELDS Kfmax = 2.-32KSI \.

Pei-Chung Wang.p 562 . €~"-'2KSI a. Lap-Shear Joints 47 (f) <I ~ 10 1 SAE 1006 LAP-SHEAR WELDS Kfmax = 2. Frederick V. CYCLES Total fatigue life predictions and experimental results for FCAW.2-5. ~ 0 . the prediction and actual results are very close. 1984. "in HSLA Steels-Technology &Applicalions. 10 2 vi <l CTr =32 KSI • 101 NT. and Albert F. Metals Park OH. Lawrence. Here.. AISI 1006: Weldment. Houchens. R = 0. American Society for Metals.77. Source: Kon-Mei Ewing.1 EXPERIMENT PREDICTION • • • ... "Weld Fatigue ofTIG-Dressed SAE-980X HSLA Steel. Jr. AISI 1006 lap-shear joints.

05 1.48 2-6. American Society for Metals.15%C mild steel. P 215 . 1981. Reversed bending fatigue tests were carried out by using notched test pieces of the same material.0 I<l: '" u. which shows a definite correlation. Wear was determined in sliding between the end surfaces of cylinders at a speed of 0. M ." in Fundamentals of Friction and Wear of Materials.56 mjs under the loads . Source: Yoshitsugu Kimura. E ~ E E E 0. Metals Park OH. AISI 1015: Effect of Cold Working 5. Wear experiments were conducted in a rotating cylinder machine as described above with a machine oil as the lubricant. Attempts have been made to determine effects of cold-working on the resistance to wear and fatigue of a O.20 <. "The Role of Fatigue in Sliding Wear.0 0. Fully annealed material was then cold-worked to different degrees and the specimens were machined from it. Ed.0 I- '" uJ I<l: ui -O.I0:J :E ::J :: lLJ <l: a: a: 2. ():124 N and 0:147 N in machine oil. Rigney.15 0<II M I 0 3.0 .x: 4. Care had been taken to avoid the effects of work hardening during machining by electrolytically polishing the sliding surface.:82 N.15%C mild steel. % Comparison of effects of cold working on wear rate and fatigue limit of fully annealed 0. oL--!------~---_:::_---_::::__'O o 20 40 60 DEGREE OF COLD-WORKING. The wear rate and the fatigue limit are compared with the degree of cold-working in the above chart. David A. Fatigue limit (.5 mm in diameter.0 0..) was determined by reversed bending fatigue tests of notched plate specimens 25 mm wide and 4 mm thick having a central hole 1.

based on the work of Paris et al. 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::.KJ' dN Source: J... Gerberich. shows a logarithmic plot of the crack growth per cycle. W.0. m 1/2 Fatigue crack growth behavior of AS33 steel. daj dN.6 Stross-intensity factor rango. l! ti 10. James E. A plot of similar shape is expected with most structural alloys.. (b) an intermediate region of power-law behavior described by the Paris equation: ~=C(t::. William W. with a yield strength of 470 MPa (70 ksl).10..2 10.K in which fatigue cracks grow extremely slowly or not at all below a lower limit of t::. K. Metals Park OH.K. . ksi • in.. Campbell. P 18 . Underwood. corresponding to the load cycle applied to a sample. l!! 10-7 l! u '" :iE Region 2: power-law behavior '" :iE ~ 10. ambient room air.. e ~ e u 10.6 ~ e .K called the threshold of t::.6 I Region 3: I rapid I unstable I crack growth I I I I I I I 10. This figure.6 ~ co ~ . 24°C (75 OF). Eds.. '" u .4 10. .4 ie at ~ .K are dependent on the material.5 Z U '" . Test conditions: R= 0.K plot shown is from five specimens of ASTM A533 B-1 steel tested at 24°C (75 OF).' Region 1: slow crack growth 01 00 I 1 . Gerberich and John H..0.K. A533 Steel Plate: Fatigue Crack Growth Rate Stross-intensity factor rango. "Concepts of Fracture Mechanics. H. t::. Underwood and W. American Society for Metals. K.3 a <9':! ~ E E ~ co ~ Z ".. 1/ 2 49 50 10.. versus the stress-intensity factor range. The da] dN versus t::.. t::. 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.K. the absolute values of daj d N and t::. MPa .2-7." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. The material was ASTM AS33 B-1 steel. 1982. 1! 10...

-_ : ~I=-:-:=:':::. 1981. The increased isotropy in the CaT over the CON steels is evident with the through thickness (ST. ..K." in Wear and Fracture Prevention. ksl\ ii1.. 100 10 I /.---.. Wilson. A514F Steel Plate: Fatigue Crack Growth Rates MPavm 20 10' 100 20 100 A514F CON Quality '(J ' I ~ 10" < A514F CaT Quality .50 2-8. SL) orientation having the fastest growth rate in the CON steel and showing the greatest improvement by CaT.i l'. lS----1 u- -::'O·· 15-----·< TL-. Source: Alexander D. p 196 . /. Metals Park OH. ~ E E ~ z ! .--. 'The Effect of Inclusions on the Properties of Constructional Steels. American Society for Metals.. '00 Plots of fatigue crack growth rate versus range of stress intensity factor (best fit lines) for A514F plates.

It has previously been shown that higher strength level steels tend to be more adversely affected by inclusions associated in groups.2-9.10·' A633C A514F CON~ caTD 1110"' 3. P 197 . "The Effect ofInclusions on the Properties of Constructional Steels. "in Wear and Fracture Prevention. such as present in CON steels. Metals Park OH.frii) A633e CON~ CaT 0 ICaT I 4. Also there generally appears to be more anisotropy in the ASl4F steels of both quality levels. this figure indicates that there is a more substantial enhancement in FCP behavior for ASI4F. American Society for Metals.10" 20110-' daldN. Wilson. inches/cycle Comparison of fatigue crack growth rate variation with orientation for A633C and AS14F plates at two tJJ( levels. A514F and A633C: Variation in Fatigue Crack Growth Rate With Orientation 51 5. These data show that the CaT improvement in FCP growth rate takes place only at higher L1Klevels.10-' ~§~~ A514F 8x10" 12110" 16. Source: Alexander D. 1981. Additionally.10-' ~Kof 50 kst/iil (55MPa.

. .p 197 . ksi\/fil.52 2-10. 0 :-----. Wilson. : .. Metals Park OH.. I I 10" I I I ~ I I I I I I I I I ~ I A514F 2'/."in Wear and Fracture Prevention. .. American Society for Metals. I I 100 I I I I I I . ... A514F: Scatterbands of Fatigue Crack Growth Rate MPavm 20 10. 1981. "The Effect oflnclusions on the Properties of Constructiona ISteels. Source: Alexander D. .ln(57mm)Gage CjCON ClCaT 10" } E E 10·' L-_ _-----''--_-'------'_'-'-'--'---'-L-_ _-----' 10 100 6 K ... Summary scatterbands of fatigue crack growth rate versus range of stress intensity factor encompassing all data points in 6-orientation testing comparing CON and CaT quality A514F plates.

Wilson.5 GI 'fi 10"' A633C 4In(102mm)Gage z C=JCON DCaT 10-' I E ~ 10-' 10 L-_ _---'-_---''---'------'-----'----'--'--'---'--_ _------"-' . P 196 . K • kslv'ln: 100 Summary scatterbands of fatigue crack growth rate versus range of stress intensity factor encompassing all data points in 6-orientation resting comparing CON and CaT quality A633C plates. Metals Park OR. 1981.K levels are displayed. A633C Steel Plate: Scatterbands of Fatigue Crack Growth Rates 20 10-' 53 MPavm 100 10-' GI I . as well as the improved isotropy of the CaT steels.c. Source: Alexander D."The Effect of Inclusions of the Properties of Constructional Steels.2-11." in Wear and Fracture Prevention. In this presentation the generally faster FCP growth rates for the CON steels at higher 6.. American Society for Metals.

c '. Volume 6. Slag near surface b.. I poli:he~ Plai~ pl~te I I .-... and Soldering...--I-.mean bank f.. • ·0 Data . P 848 .18 10 • 2 4 I 10' 2 4 6 B 10' 2 4 6 6 8 10 . Brazing. Fatigue life.~ 0 0 40 -in . ... American Society for Metals.0/.. Slag at midthickness V Lack of penetration I " ~ ch on .. Source: Metals Handbook... Welding. IA - _C e» r----.r-. . 1983. Metals Park OH. 'flo 9- 10 8 f- V V 11 V 111 11 I~ 6 f- 1<:.' curve ~- ..r-9 . 1"-1- .. presented as a comparison with fatigue life of the plate.ll . ~ ~ I.54 2-12... 9th Edition... Low-Carbon Steel Weldment: Effects of Various Weld Defects 100 80 60 I I I r- -."C - ri if. o VV OJ" --uv .t: + 1 standard deviat ion data bank ~ 20 Low-carbon steel reinforcement off LC?cation of failure: o Plate or weld edge • Porosity in weld . cycles S-N curves showing effect of various weld defects on fatigue life of a low-carbon steel weldment..

l. P 850 ..... Fatigue strength of a weldment containing slag inclusions as a percentage of the mean fatigue strength of a sound low-carbon steel weld..6 0. . .8 1.: "0 C ::J g (. Low-Carbon Steel Weldment: Effect of Weld Reinforcement and. Source.. Volume 6. "0' Reinforcement intact Reinforcement removed 0...-- .. ---- . c-I_---~I""'-_~=__iI_---___l = ----- 1..'" u. -S ~ CI> '" ::J 40 _---"" ___-~.. Welding. . and Soldering.. Metals Park OH. Metals Handbook..4 0.j!.l- -J o 0.2-13. Brazing. 9th Edition..- . '" 201_----1_----1-- Ol. 1983.0 Slag inclusion length..2 0.. in.Lack of Inclusions 55 -e ..l- .J. American Society for Metals.

Fatigue strength ofa weldment containing lack of penetration as a percentage ofthe mean fatigue strength ofa sound low-carbon steel weld... I em orcement mtact I I 60 " Reinforcement remived <.....- 0.... ~ r-- 40 r----.. 1'-.. Welding.10 / -.. ..06 0.04 0.r--r-. r-... . A 100 oob cycles ~ \/ -: 0. 9th Edition.02 0. in. P 849 . 1983.12 0.i t--.. ----<.08 0.. Metals Park OH...... Brazing.... ~ . ... fl.. Volume 6.56 2-14.... -. - ~I R . ...16 2 000 000 cycles 20 o o 0. Source: Metals Handbook.. American Society for Metals. Low-Carbon Steel Weldment: Effect of Weld Reinforcement and Lack of Penetration 100 \\ 80 \\ -.. and Soldering..14 Lack of penetration half depth...~'" <..

. <.. Source: Metals Handbook. <.. -g 60 '0 -S '" 40 ~ Ii: Ql " .Reinforcement rejOved """- " . in.14 0.2-15. '" } 2 000 000 cycles 20 0... P 849 . American Society for Metals.10 0.. b..... Volume 6.02 0. 1983. • LL 5. Brazing. 9th Edition. Welding._ r-~ 1 100 000 cycles J ~...." ~ ~ ~ ""' -r-.08 0. Low-Carbon Steel Weldment: Computed Fatigue Strength. "... <.-::::.fi. Computed fatigue strength of a weldment containing lack offusion as a percentage ofthe mean fatigue strength oCa sound lowcarbon steel weld.06 0... ..04 0.. Metals Park QH. Qj 80 ... Reinforcement intact I I ' I .16 o o 0... <.12 Lack of fusion half depth.' s ~ \' \ 1\ '\ III r-...--. and Soldering..::-. Weldment Contained Lack of Fusion 100 57 \ ~ .. .t.

Source: Metals Handbook.00 000 cycles - 10-. :::::- t::-I-. r': t::'. 20 --.r--r-.58 2-16. in.i'--- r::: t:-.. Volume 6.12 0. a> :::l r-. ~ . Welding.t-- • 00 000 cycles t-- "k 0. o t: 5 g 60 \ [\ <. Metals Park OH... fl.10 0. r ""'--..r:. ~ Cl> '. to-.. ...ern orcement Intact ~ ... Brazing.08 0. 1983. American Society for Metals.. I . P 848 .06 0. Effect of depth of undercut in terms of percentage of fatigue strength of a sound low-carbon steel weld.. and Soldering. 9th Edition. fi a> ~ 40 I" '-....Q2 a a 0.Reinforcement removed . co r-. ' u..1-.. ~ t:. r.r--""'- I .'/ .16 Undercut depth. Low-Carbon Steel Weldment: Effect of Reinforcement and Undercutting 100 80 *' '0 '0' ~ ~~ ~ \~ ~ I """".14 0.04 0.

5 • •• 2... ----t--~ t." in Proceedings of the SAE Fatigue Conference P-109..'6~-------~ __ C ~.. ~ ......-3 ~ Q.... Low-Carbon Steel: Transverse Butt Welds. ~REINFORCEMENT OFF . . Society of Automotive Engineers.r .8mm 200 100 Tr==r?J r h 3'-.-. Effect of Reinforcement • • 'I 59 r . "Fatigue Considerations in Welded Structure.. Warrendale PA.... ~ • • -. Nelson... . ~ ........-. h =3... o __• ~_ 0 0 • 0 - -~- E (/) ro x ~ •• 300 Q.• REINFORCEMENT <.~.---r---"I---y---r-T"""T"" ..--""--"~J.2-17..------.-.r:~ CYCLES Influence of weld reinforcement on fatigue strength (R=O) of transverse butt welds of quenched and tempered carbonsteels.. . 600 • ~"""""O<UNWELDED . 1982.-. *• . From these data it is evident that removal ofthe reinforcement (weld dressing) improves fatigue strength and fatigue life... P 206 .: . ON 't:. Source: Drew V.. . '1-0--"- .. ~ O~ <. <...3 ..0 500 • • 1.... •• j~--... Inc.

5I.' +35 ksl 300 200 20 100~ 7O:e 10 ~ 8 6 /!l361 E60S-3 Double-V Bull Welds ~ma.." in Residual Stress for Designers and Metallurgists..60 80 40 2-18. =0 CT.'2. Lawrence. Larry J.--. Ed.T"T. (16-mm) A36/E60S-3 butt welds. Metals Park OH. = + 35 ksl 2 10 10 N1 ' Cycles 7 Fatigue crack initiation life predictions and experimental results for ¥a-in.... A36/E60S-3 Steel Plate: Butt Welds IOOr:---. American Society for Metals.""TT--. Vande Walle..P 114 . 1981. I = 5/8 in. 600 400 6Or---====:=--_ CT.--r-r-. "The Predicted Influence of Weld Residual Stresses on Fatigue Crack Initiation. Source: F.----.. vi 5O<l 4 R=0. S S 30 20 OCT.-r-. V.-rn"----r--r--r-OT.

.120kl' ---0 "" ·-120kli 200 200 vi .0 . "The Predicted Influence of Weld Residual Stresses on Fatigue Crack Initiation. V. Vande Walle. A514F/E1·10 Steel: Bead on Plate Weldment 61 ~14F/EII(j Begd On plgte Kfmal • 31~1 DJ • 0. so <I vi 00 0 . Larry J. 100 <I -:---_• --- 10 ~ l.01In. Lawrence." in Residual Stress for Designers and Metallurgists. Metals Park OH._ "'---- 30 20 20 10 Total fatigue life predictions and experimental results for A514F /EllO weldments with tensile and compressive residual stresses. Source: F.. Ed. American Society for Metals.• R -0.. . t -112 ln.. P 113 .2-19. "".. 1981.

... Lawrence. -60·. 1-I/Zin.-90·. The results for the high-strength. Metals Park OH.-90·.254mmJ 10 . quenched-and-tempered steels (upper chart). I -1/2 In. result in a predicted variation oftotal fatigue life with stress ratio R.) and stress ratio (R).... The observed dependence of N» on stress ratio does. 1~'r--~--r-r""""~"'--~-~~""""~"'--~-'-"""'~~"'6oo 60 400 300 R'O 40 20 A36/E60S-3 Bull Weld K.15.s. Source: F.D. 30 20 10 Predicted effect of stress relief and stress ratio on A514/EllO butt weld fatigue life. Mild steels can have appreciable residual stresses. (I2. 01... ' 30 20 10 10< NT' Cycle. The total fatigue life of such materials is strongly influenced by both residual stress (a. 50 ~ '" .-s)... . .' . indicate that such materials can sustain high residual stresses which do not relax..OIIn. but. Ot' 0.. ~ 8 6 .) is often very long (= 500. 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.. P 112 . 1981. Stress relief or mechanically induced compressive residuals should be highly effective.' -60·.3.. An intermediate case is mild steel as shown in the lower chart.01 In (O.35. 10' Predicted effect of stress relief and stress relief and stress ratio on A36/E60S-3 butt weld fatigue life. A36 and A514 Steel Plates: Butt Welded 514F / E 110 Bull Wold 10 K.7mml --.. there are large amounts of plasticity at the notch root even at long lives (106 cycles).."_ ·2. (0254mm) 100 ~ 70 ::f.000 cycles). Vande Walle.(I27mm) 8 6 --. American Society for Metals.. Larry J.62 2-20. -0 s~s K. Ed... since the transition fatigue life (N." in Residual Stress for Designers and Metallurgists. "The Predicted Influence of Weld Residual Stresses on Fatigue Crack Initiation. ~-.-O s~s K r_ ..w.. however. r NT' Cy(lu ~ . V.

. Metals Park OH. Vande Walle.The results obtained using the model agree with the experimentally observed behavior.. The above chart shows the qualitative behavior of N[ predictions. Source: F. A36 Plate Steel: Butt Welded 100 80 63 600 60 Zero Mean Slress 40 200 20 .2-21. R = 0. 1981.. Lawrence. .atian A36 Butt Weld (HAl) KImox =3 I R-O a r =+35 ksi 100 70 2: e ~~ 30 20 4 2 10 IIO~ 10' N I • Cycles Mean stress relaxation behavior influence on fatigue crack initiation life (A36 HAZ material. American Society for Metals. V. a os may be larger than a r. pIli ." in Residual Stress for Designers and Metallurgists. consequently. "The Predicted Influence of Weld Residual Stresses on Fatigue Crack Initiation. K f = 3. a r = +35 ksi (242 MPa) ). VI <I 10 8 6 --\---====::::::::d Mean Slress Effects With No Rela.. Ed. Larry J. Materials such as high-strength steels exhibit very little notchroot plasticity.

Low-Carbon Steel Tubes: Effect of Welding Technique . B .' .. V. 1971. The Welding Institute. "Low Cycle Fatigue of Welded Structural Steels: A Material Manufacture and Design Approach..f----..- .~ O·B 0·4 0·2 '-'0 "- -----_.~ ~~ .. .. '~r::-...-.'~ ' ~/ . P 193 ..longitudinal or helical welding (550 or 600 angle). Vol 2. . " ._ •. ~ ~ ~ %-.. ~/... '~ ~D o I I I I I I I I I I I I I I I I I I I I I I 1 10 Fatigue strength of welded tubes: A .-- i'. "" ~ 1-. '. C ._.~:. Source: R. .helical welding (500 angle).~ ~'/~ .ij'/W A_ .helical welding (700 angle).. Salkin. '/8 ~ ~..""' ~... Abington Cambridge.• _ f--------. . ~. " I\. D .--' 1·0 ----- 0·8 '--'. ~c ~..-:~~ /.~'-." in Proceedings of the Conference of Fatigue of Welded Structures. .. " " " '" . -~/ .~ b-...'.64 2-22.- I~'" V'~ ~ --~-.unwelded or welded without filler metal..

. Source: D.L--L. "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys.'_ _-'--------'-_'--.J'--_ _." in Fatigue and Microstructure.2 /-LA/cm-. 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... American Society for Metals. Additionally. A similar result has been reported for deaerated 3% NaCI solution in which corrosion rates were controlled by applied anodic currents (see above chart)... but that at concentrations below 1/40 molal. These observations indicate that a critical corrosion rate is a necessity to initiate corrosion-fatigue failures. increasing over-all corrosion rates over a long range of rates has little effect on corrosion-fatigue resistance.'_ _-'--------'-_'--L-JL.-'LIlk< 10' 10' 10' CYCLES TO FAILURE Effect of applied anodic currents on the fatigue lives of low-carbon steel in deaerated 3% NaCI solution. 1979.. P 344 . the effect approaches that of distilled water. Note the independence of fatigue life at currents greater than-: 40 /-LA/cm 2.2-23. although corrosion rates increase in an almost linear manner with solution ion concentration. Metals Park OH.l.. J. the absence of an applied stress effectand the reappearance of a fatigue limit at currents less than ~ 0. Duquette..l-L. Low Carbon Steel: Effect of Applied Anodic Currents in 3% NaCI 65 . 120 E ~ ~ 100 >-t-- w o ~ 80 040 ~ IX: IX: ~ 60 w 20 OL-_ _. The effects of salt concentration and temperature on the fatigue behavior of steels have been studied. The corrosion rate ofthe steel in this solution is virtually zero in the absence of applied currents.

and that a high pH provides diffusion barriers (ferrous hydroxide) to oxygen on the surface. but a given number of cycles was found to produce greater damage at low frequencies.. Low-Carbon Steel: Effect of pH in NaCI and NaOH 60 I I 55 / I NORMAL SOLUTION 50 45 ~ . showed that a fatigue limit is regained. 1979. Duquette. the effect of pH of aqueous solutions on corrosion-fatigue behavior has not received extensive study.1. For low-alloy steels in fresh water. which produce pits in the metal surface. Metals Park OH. A study of the effect of 0. Higher fatigue limits at high pH are explained in terms of a "better and more perfect film barrier. a given time was found to produce more damage at a higher frequency. but at a frequency of 5 cycles/min. a frequency of 1450 cycles/min produced failure in 106 cycles or II Y2 hours. For example. J. failure occurred in 0.1 II: UJ ~ 35 '" 30 25 pH :10 t- 20 . These investigators suggested that corrosion fatigue is a result of differential aeration cells. In general. P 346 . this limit improving at still higher pH values (above chart)." in Fatigue and Microstructure.1 N HCl on the fatigue life of steels showed greater damage in this medium than in neutral potassium chloride solutions." Source: D.H =12. 40 Q "'. at a pH above 12.66 2-24. To date. Tests conducted in alkaline media.5 5 8 10 The effect of pH on the fatigue behavior of low-carbon steel in NaCI+NaOH. or 400 hours. The effect of stress frequency on corrosion fatigue has been studied by a number of investigators but is still not completely understood. American Society for Metals.11 X 106 cycles. "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys. 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.

The two exceptions may be in combination rolling and sliding contact fatigue..... "Fatigue in Machines and Structures-Ground Vehicles.. Breen and E.. Wene. From the carburizing process an intergranular oxide network may develop... where the oxide network may enhance low-cycle bending fatigue-somewhat the same as does decarburization. M.. INTRINSIC' FOR DECARB MATERIAL .DECARBED • -COMPOSITE LIFE Influence on fatigue SoN curve of soft surface caused by decarburization. ." in Fatigue and Microstructure. This oxide may be an alloy oxide which causes alloy depletion in grain-boundary areas. Parts that were made from low-carbon steel.. American Society for Metals.. Low-Carbon Steel: Effect of Carburization and Decarburization 67 CJ)A CJ) • t) ~ ~ B ..P 80 . The effect on high-cycle bending fatigue is deleterious. Metals Park OH. Source: D. have special microstructural factors that must be considered.~ ~•••• NON-DECARBED MATERIAL . . 1979.2-25. this condition is thought to detract from fatigue properties.. but have high-carbon surfaces resulting from carburizing. H... These concepts are shown schematically in the above chart. as is decarburization. As a rule..

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. "Environmental Effects 11:Fatigue-Crack Growth in Metals and Alloys. Subsequent interaction with the crack allows the degradation mechanism to take place.. The main influence ofthe environment is to supply the active atoms to the vicinity of the crack tip. 1979. Metals Park OH. L. 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.4 2 AK = 29. It could also originate as an enriched sulfur layer associated with a propagating crack.7 MN/m 3/ 16 14 0.2 0.. The origin of the element (such as sulfur) Onthe surface could result from its presence in the gas phase (such as for hydrogen sulfide). 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. A514B Steel: Effect of Various Gaseous Environments on Fatigue Crack Propagation r------.68 2-26.1 4 2 The fatigue-crack propagation of ASTM A514B steel in various gaseous environments.::::::j""". Marcus.:. Source: H... American Society for Metals.3 0. 18 0. P 371 ." in Fatigue and Microstructure.

C.specimens were used to secure these data. Ed. Weiser. The nominally 1. P 15-29 . Source: Steel Castings Handbook. 1980.006 0. carburization. 5th Edition. This effect. Rocky River OH. Decarburization of the surface lowers fatigue resistance.4 I% C. The depth of decarburization (0. (2 mm). Peter F.006 0. that of carburization (1. respectively.2-27. respectively. and decarburization on the endurance ratio of normalized and tempered cast steel with cast surfaces. is shown in the above diagram.22% C and 0. along with the beneficial effects of carburization and shot peening on the endurance ratio of cast low alloy plate specimen in bending. Cast 1522 and 1541 Steels: Effect of Various Surface Conditions 69 NONE 0. Plate bending fatigue . ALMEN The effect of shot peening.. (1.15% C at the surface) was 0.2% Mn steels with 0.5 mm). were normalized and tempered to 78 and 95 ksi (538 and 65~ MPa) ultimate tensile strength.012 SHOT PEENING INTENSITY . Steel Founders' Society of America.08 in.06 in.05% C at the surface) was 0.012 NONE 0.

8 Z "0 <.--__. Source: Steel Castings Handbook...--'-_ _---' 10 20 40 60 100 200 STRESS INTENSITY FACTOR ~K.? 10-7{. RANGE.6 . Steel Founders' Society of America._~__.4 3r---. This is observed for a wide variety of materials and is illustrated in the above diagram for an ASTM-A2l6. Cast A216 (Grade WCC) Steel: Fatigue Crack Growth Rate 20 40 60 80 100 200 6 ._r_--" lTyS : 48 ksi 2 (331 MPo) TEST TEMP: 7soF ( 24°C) TEST FREQUENCY: 600 cpm WOL TYPE SPECI MENS UPPER SCATTER BAND ( SLOPE n : 3 ) """ w 2 -. E 10. Grade WCC cast steel. 8 6 4 u 0:: U <l: . 5th Edition. The equation do I dN= CoI1K" is sometimes referred to as the Paris law and predicts a linear plot of log dol dN versus log 11K with slope n. Weiser.---. 10.5 I W 0:: ~ I I<l: 0:: w 8 I0:: ~ 2 I o ~ 6 4 I~ 0 0:: ~ {. 0 Z "0 <...6 '--_ _-'-_-'----'_L...? U 0:: '<l 2 o._.I U u I r -... Rocky River OH. 6 "0 o "0 I 2 4 10. ksi ~ Fatigue crack growth rate as a function of ilK for A216 (grade weC) cast steel... Some materials show a significant influence of the mean load or Klevel on fatigue crack growth rates. 0 ~.. Ed. Peter F. The ratio ofKmin to K max is used to express the mean load conditions.. p4--16 .-I-L.70 2-28.. 1980...

although a straightforward approach is simply to compute the fatigue life from the expected cyclic plastic strain amplitude in service.....Ep produce a smaller change in the computed cyclic life than similar errors in the elastic strain range. AISI 1030 (Cast) Compared With AISI 1020 (Wrought) (\J 71 <.' <t ~ '00"::0.1 /WROUGHT SAE 1020 .l. /CAST SAE 1030 1 0....--r--.. Z /O..... 51hEdition.l. 0.. Rocky River OH.0001 ~_L--. IIII I.--r---r--. Ed. I ILl C ~ C .L... "...." 1)'0' ~O <i 0: 0....01 PLASTIC I- :J Q... Weiser.. 1980. Steel Founders' Society of America. Peter F..001 " t.::-~-. --r--. 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.....W O\' ' 0. Errors in computing or estimating 6..3-1. A number of techniques are available for computing the lowcycle fatigue life.2N f Low-cycle strain-control fatigue behavior of carbon steel..P 4-13 .:~..:=--~ 7 0 10 10 102 106 10 REVERSALS TO FAILURE .O. <I 'f "''0.. Source: Steel Castings Handbook..l..

P 286 . 1977. American Society for Metals. featuring the effects of gaseous atmosphere and salt bath nitriding on fatigue strength. A. AISI 1035: Effect of Gas and Salt Bath Nitriding 50 A1MQSPH£RE 40 NITRIDED SALT BATH PRQCT_'S 0 § lC QUENCIU:D AND TEl1PERID AT l050 F (565 C) '0 ~ HUKBER OF CYCLES Torsional fatigue strength of AISI 1035 steel-stress vs number of cycles for completely reversing torsional fatigue. "Short Cycle Atmosphere Nitriding." in Source Book on Nitriding. Metals Park OH. Source: J. Riopelle.72 3-2.

Source: Steel Castings Handbook.). R.. Rocky River OH. = 2. Moore rotating beam tests. 45 en WROUGHT CAST } NO NOTCH 300 .><: 94 (6481 90(62~ 56 (386) 56 (3B6) 25 27 187 170 350 ~ ~ J.3-3. R. both notched and unnotched. The same fatigue characteristics as those of cast steel. These effects are illustrated above. the cast steel shows much less notch sensitivity to variations in the values of the test parameters than wrought steel. Steel Founders' Society of America. or when standard lathe-turned surfaces are employed in the rotating beam bending fatigue test. 1980. are obtained when a notch is introduced. 5th Edition. however. AISI 1040: Cast vs Wrought 73 TENSILE YIELD STRENGTH STRENGTH ELONG. HARDNESS ksi MPo ksi MPo % BHN CAST WROUGHT en 50 .2. K. etc. Ed. Peter F. Cast steel suffers less degradation offatigue properties due to notches than equivalent wrought steel. Weiser. P 15-10 . Under the ideal laboratory test conditions and test preparation (uniform section size. f3 f3 g: ~ 0: 40 en ::::> ~ 35 250 ~ ~ ~ ~ ::::> 200 NOTCHED 150 5 10 6 10 7 10 5(30 ~ 25 <t <t X CYCLES TO FAILURE Fatigue characteristics (S-N curves) for cast and wrought 1040 steel in the normalized and tempered condition. polished and honed surfaces. the endurance limit of wrought steel is higher. When the ideallaboratory test conditions are replaced with more realistic service conditions.

"Control of Fatigue Resistance Through Microstructure-Ferrous Alloys.. __ .. AISI 1045: Relationship of Hardness and Strain-Life Behavior ..P 458 ... provides guidelines for optimizing material processing for specific situations.280 2211 I 10 2 10 3 10 4 10 lOll f 10 6 7 10 Reversals la Failure.. <l "0 C\l .01 >-. Strain-life curves at various hardnesses are presented in the diagram above to demonstrate the range of properties attainable by tempering.. 0.. 0....001 ." in Fatigue and Microstructure.3 3 0 . used in conjunction with life-prediction models. Landgraf. C3: c: (/) +0 E 1:1 .:::--=----_410 .. 1979.74 3-4. Source: R. Metals Park OH. 2N Strain-life behavior of medium-carbon steel as a function of hardness..~_ _ _ 1l911HB '. W. American Society for Metals. Such information.1 SAE 1045 or 0.~ Q.

" in Source Book on Nitriding.3-5. Riopelle. Source: J.\YEII ATHOSI'HERE NITRIDED niDI AT 1050 r (565 C) 40 NlHIEI OF C'l'CLlS S-N curves for 1141 steel-gaseous-atmosphere nitrided vs not nitrided (quenched and tempered only)-showing stress vs number of cycles for completely reversing torsional fatigue. A. "Short Cycle Atmosphere Nitriding. P 287 . AISI 1141: Effect of Gas Nitriding 75 ATHOSI'HERE NIUIDED QU!NCBED AND TDO'UlD GROUND to IlDlOVE COHPOUND I. 1977. American Society for Metals. Metals Park OH.

Metals Park OH. Volume I.001 0.76 3-6.30~ 200 . ~ 0.01 <.~ -." . Source: Metals Handbook. \. HB t: S '0 n. 9th Edition. 1978. P 673 . ~ ~ 400 <.0~ \. Strain Life and Fatigue Life 1. Medium-Carbon Steels: Interrelationship of Hardness. - 0. Properties and Selection: Irons and Steels. 300<.. -200- .'\. ~ t\ 600 s: c: 0 . ~ . 1\ \.. 0.1 1\\'\\ ~500~00 " " 600 \\ \ en c. . American Society for Metals.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. "'-. \ Hardness.

American Society for Metals. Hutchings and Paul Unterweiser.IEVER ~4 5EPARATE COLLAR The fillet radius at a change in diameter should be made as great as possible... A. Metals Park OH.g.000 I'lUMBlR OF C'{CLES . r 'Y-. Eds.ys 0+ ---- I II 300.·2·/3"~ (i·Q·If.000 1. In such cases the stress-raising effect can be moderated very considerably by adopting one of the expedients illustrated above. F. if the inner race of a rolling bearing must abut against a shoulder formed by the change in diameter.u ~S- D/ 2 d: d.--- 77 60 <-. "Some Common Stress Raisers in Engineering Parts. DIcJ.000." in Failure Analysis: The British Engine Technical Reports.000. ~ I~ 0: li. Source: G.000. 1981.. O/cl. e.. SCALE) ~A1 UNMODIFIED ~&j RE-EI1TRAN. Medium-Carbon Steel: Effect of Fillet Radii . t-- ~ on --~~p~~l f:o ~ ~ o o x I ~/lf: c 1ft'! d ~O'47" r--.REl.0 FAILURE (LOG.3-7.000 50. P 108 . ~40 ~ o z I""" Z 30 ~ o d I I - ~(l'O'I7 d. Cottell. This cannot always be done. I Q . ~ gj 50 loil oil III ~ . FILLET EXTERNAL 5TRE55.000 10.:2 {)r ~ r. R.

Effects of various keyway designs on fatigue life are shown above.. F.. Source: G.000 PROFILED .\LURE (LOG. American Society for Metals.000 1.YWAY 70.. R. Metals Park OH. Hutchings and Paul Unterweiser. Where bending stresses are predominant the cracks usually run transversely in the region of the keyway end. DF C'(CLES.. p 109 .000 10.78 3-8.. A.R KEYWAY PROFILED Kf. Medium-Carbon Steel: Effect of Keyway Design 5LEO-RUNNl:.000.O FI'. Cottell. Eds...EV." in Failure Analysis: The British Engine Technical Reports..000 NUMBER. "Some Common Stress Raisers in Engineering Parts.000.StALE) Keyways are severe stress raisers from which fatigue cracks are very liable to develop...( ' o MEDIUM-CARBON S. 1981...EEL (NORMALISE-D) I I I ~ I o 200. but where torsional stresses predominate they originate at the root at one side and may cause a portion of the shaft to peel off or they may lie diagonally across the bottom.

. "Prediction of Stresses Generated During the Heat Treating of Case Carburized Parts.... Fully reversed fatigue tests on smooth bar specimens in medium carbon steels fully hardened show. P 44 ." in Residual Stress for Designers and Metallurgists. Medium-Carbon Steel: Effect of Residual Stresses +600 79 +400 +200 NEAR-SURFACE RESIDUAL STRESS MPa O~-----------~------(= -200 -400 +3000 po IN/IN TESTS STOPPED AT 10 7 CYCLES MAX. Residual stress measurements are usually made in the direction of the applied stress. 1981.-----------. 1... Vander Walle.. that fatigue life increases directly with surface and near-surface residual compressive stress (see above chart).. Larry J. Metals Park OH. as expected. American Society for Metals.0 10 0.. Source: J. Ed. • • • -600L---------.1 UFE CYCLES Ie 10 6 Fatigue life relationship to near-surface residual 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.L.J.3-9. Alan Burnett.

--+- +--_ _--< --+ Ill" CYCLES --+- -+- +--' Change in residual stress with cycles at constant applied stress." in Residual Stress for Designers and Metallurgists. pp 127-128 . TEMP~ '" II -5 111 4 CYCLES ~ ti "a. the slope for each line could be determined. The upper chart shows residual stresses existing on the completion of individual tests.. Metals Park OH.1. Since the initial and final residual stress values were known for all bars. "A Quantitative Evaluation of Surface Residual Stress and Its Relation to Fatigue Performance. Neff. NORH"LIZE - -' Ul ILl ~ :§ 5 <. ~HOT BLASi' lJl III '4 A ' lil.. -.80 3-10. there was a question on the manner in which the residual stress changed during the progress oftesting. 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. /981. i- ~ .>: I . III -411 .. The lower chart also points to the fact that the rate of change in residual stress is dependent on the level of applied stress. Medium-Carbon Cast Steel: Effect of Changes in Residual Stress 25 211 .. Ed. Vander Walle. To explore this point. two shot blasted bars were tested with applied stress levels of 40 and 55 ksi.. J. The results are shown in the lower chart. Source: P. The test on each specimen was interrupted periodically to measure the residual stress at that time. -. Since the initial residual stress was known. Larry J. American Society for Metals.4'~ :a---~ <.. ~ Residual stress at completion of testing. NDR"""L. The similarity to S-N curves is apparent with the exception that the curve for normalized bars (R 1 =+22 ksi) is inverted.IZE ~N~ 15 e.

.

.

'" " ~ 6. particularly those with coarse pearlite. % 40 20 For specimens having comparable strength levels. ~ <. Volume I. :~ 500 All specimens 36 HRC 450 '"I'--. Pearlitic structures. resistance to fatigue depends somewhat on the microstructure.N curves for pearlitic and spheroidized structures in a eutectoid steel are provided in chart 4-40 (p 122).. Properties and Selection: Irons and Steels. ::. --80 6.4042 _4340 05140 6. I--. have poor resistance to fatigue. the fatigue limit will be lower (see graph above).. P 676 .4-1. . S. 9th Edition. American Society for Metals. n 70 I--. Metals Park OH. A tempered martensite structure provides the highest fatigue limit. if the structure as-quenched is not fully martensitic. Source: Metals Handbook.I'--6. Five Grades: Effect of Martensite Content 100 83 100 650 c. However. 1978. 80840 90 ·Vi ~ ~. 01340 . 400 60 100 60 Martensite.. Medium-Carbon Alloy Steels. 600 ] .~ 550 u.

1979. a variety of residual-stress conditions resulted. 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. Six Grades: Hardness vs Endurance Limit 160 150 140 Vi a. Carbon. "Fatigue in Machines and Structures-Ground Vehicles. American Society for Metals. Zurburg and Erickson show a very tight linear relation up to about 40 HRC. consequently. P 73 .SAE SAE . Above that hardness. governs the outcome. however.SAE G . the relation deviates from linearity. These data from Garwood. The tempering temperatures were necessarily sufficiently high to obtain 40 HRC. The sequence of transformation from surface to center. Breen and E. Because response to tempering is dependent on carbon and alloy levels. 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.SAE 6- E --' Q) u c ~ c w :J "0 50 20 30 Rockwell 4063 5150 4052 4140 4340 2340 40 50 Hordness 60 "c" Relation of hardness and fatigue strength for several steels. because it affects temperability.84 4-2._ SAE 0 .a. Source: D. it was necessary for samples of different grades to be tempered at different temperatures to achieve the same hardness. Medium-Carbon Alloy Steels. 0 0 0 H-ll 1----1I---t---I----1t-"'1 Aus t ernpered 130 120 110 100 90 • . seemingly depending on carbon content. is in an intermediate role here." in Fatigue and Microstructure. Metals Park OH.SAE Q . M. the residual stresses were reduced to a very low level. together with the temperature gradients. Wene. H.

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.~ . Metals Park OH. Medium-Carbon Alloy Steels: Effect of Specimen Orientation 85 BOO Steel 100 Avg tensile No.3 mm (0.: s: 600 . P 677 . Longitudinal 4027 4063 4032 Tests 11 1179 171 12 1682244 11 1627236 37 to 39 47 to 48 46 to 48 4140 Hardness.5 (a) Number of fatigue specimens. Properties and Selection: Irons and Steels.250 in. 50 longitudinal and 50 transverse specimens were tested. Source: Metals Handbook. Specimens for each grade were from the same heat of steel. for 4340 steel.5 1254 182 47. 5. For 4140 steel. but the tensile and fatigue specimens were heat trated separately. u. of strength Hardness. 1978. BO 400 ~ u. were taken from production billets. 9th Edition.5 1682 244 47 to 48. HRC 30 200 X4340 32 4027 44 4063 46 4032 4B Transverse Tests 4027 10 4063 9 4032 10 1130 164 34 to 39. 60 "g 40 . tests(a) MPa ksl HRC ::. Through hardened and tempered specimens.) in diameter.5 to 48. Fatigue limit is for 100 million cycles. 5. 10 longitudinal and 10 transverse.4-3.~ . 6. Volume I.>! . thus accounting for one discrepancy in hardness readings between the chart and the tabulation above. American Society for Metals.

the thinnest case offers some improvement but tends toward the baseline as a result of subsurface failure initiation. "0 .005 ji ~ ..~ E c Ul .e o 0.-_.03S"Ca. O... seems close to optimum.... Source: R.. ~ 0. W. American Society for Metals._--'-. At longer lives. 0. 1979... The thickest case.01 • 4027 Carburlzed b.002 0. Landgraf. As predicted." in Fatigue and Microstructure.. Bending-fatigue results supporting the validity of the effect of carburizing are presented in the above curves.-_'-. all carburized specimens show inferior low-cycle resistance.86 4-4. which shows the greatest life improvement and has been found to exhibit surface failure initiation._--' 102 103 104 lOS 10 1 Reversals 10 Follure.OOS"Ca.. Metals Park OR.001 L...... 4027 Steel: Carburized vs Uncarburized 0. An uncarburized baseline curve is compared with curves for three case thicknesses.... 0. "Control of Fatigue Resistance Through Microstructure-Ferrous Alloys.-_--'-_----'--.02 (\J Bending Fatigue '" <l .----'-.. P 463 ..OIS" Ca. D O....2N f Bending-fatigue results for uncarburized and carburized 4027 steel.

This is because the galvanizing produces a relatively high tensile residual stress and the deposits possess generally poor ductility. The base metal was mechanically polished before surface treatment. m bose metot .22Mo steel. pot galvanizing and Ni-plating have been performed after mechanical as well as after chemical polishing of the base metal. pol. The results after chemical polishing are given above in the form of dashed columns. 1982. G. Metals Park OH. The galvanizing such as Ni." in Current Solutions to Hydrogen Problems in Steels. the effect of surface treatment on fatigue life is summarized.23C-O. American Society for Metals. C. P 380 . H1 (10MPo) £ =aOO3 sec:' Cl bose metal .mech.98Cr-O.2%. Interrante and G.4-5. For comparison. prior to surface treatmMt surtoce treatments Effect of surface treatment on fatigue life in hydrogen environment for a O.annealed . In the above bar chart. 4120 Steel: Effect of Surface Treatment in Hydrogen Environment 87 200°6 1600 600 25 Cr '"'A 4 . Eds.4~. Presouyre. Source: Kyong-Tschong Rie and Werner Kohler.chem.= 1.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. "Improvement of the Resistance of Metals to Cyclic Plastic Loading in High Pressure Hydrogen Environment.. M. pot.

Source: Kyong-Tschong Rie and Werner Kohler. C.6% t =0. M. Interrante and G.la/ed I I Ia/hed - -1 '25 erMa. 1982. Metals Park OH. Eds. P 379 .-----~---~-----.-------..003 s-' H2 (10 MPaJ number af cycles N Effect of surface treatment on cyclic now curve in hydrogen environment. It can be seen that the fatigue behavior in hydrogen environment can be improved by some surface treatments. hammered I / alvanized I e- 300 ~ III III I 200 Ni. The above graph shows the cyclic strain hardening and softening curves for different surface treatments. American Society for Metals. Pressouyre. G. "Improvement of the Resistance of Metals to Cyclic Plastic Loading in High Pressure Hydrogen Environment. 4120 Steel: Effect of Surface Treatment in Hydrogen Environment 500'." in Current Solutions to Hydrogen Problems in Steels. -annealed <a=!o.88 4-6. .

g s ~ 25 erMa' -annealed -+~ i: =0..003 sec-' -H1. G.0 .. C. Eds. e 2.. 4120 Steel: Effect of Various Surface Treatments on Fatigue Characteristics in Air vs Hydrogen 89 ." chem..2 60 0. 0..' 0. 1.J0. M.. 1982. "Improvement of the Resistance of Metals to Cyclic Plastic Loading in High Pressure Hydrogen Environment.22 Mo (4120).-+-._ pol galvanized 0•• Ni-ptatea 0.....'. p 380 . . polished . Interrante and G.. 0. Steel contained 0..:~ .. . .0 "'l . Pressouyre.98 Cr and 0. '''. ".. polished 100 200 500 1000 2000 cntxa! nurrtJer of cycles Na 5000 MXJ Fatigue life curve for various surface treatments of steel in hydrogen environment and in air.IOMPo) S --air <II ... American Society for Metals.+ mech. Metals Park OH.4-7. Source: Kyong-Tschong Rie and Werner Kohler.23 C." in Current Solutions to Hydrogen Problems in Steels.

U <Il V> E I 3. oK-I Crack growth rate versus temperature in hydrogen gas." in Hydrogen Embrittlement and Stress Corrosion Cracking.2 I 4.8 4. Johnson. "Keynote Lecture: Overview on Hydrogen Degradation Phenomena. and ofthe unique strain rate and temperature dependence of hydrogen embrittlement.3 I 0 O"ty = 1330 MN m. Nelson and Williams reported the first complete investigation ofthe kinetics of crack growth in high strength steel exposed to hydrogen gas (see graph above). Eds. R. Hehemann.0 liT.90 4-8.2 O"ty = 1190 MN m. 4130 Steel: Fatigue Crack Growth Rate vs Temperature in Hydrogen 10. P 17 .6 5.3 / 2 o . it is nevertheless always several orders of magnitude larger than the diffusivities of other species. Although the diffusivity of hydrogen does vary significantly among metals and alloys. hydrogen transport is a prominent feature of discussions of hydrogen-induced crack growth kinetics. Source: Herbert H. F.4 3.. 1984. for 4130 steel with yield strengths of 1330 and o2 1190 MN m • The striking characteristic of hydrogen which sets it apart from other causes of embrittlement is its large diffusivity. American Society for Metals.. Consequently. Metals Park OH.2 I o O"tu =1310 MN m- 2 K =40 MN m.2 I O"tu = 1660 MN m. Gibala and R.

....-- ·-A-A_ 40 '. ~~ 106 . Weiser. cast steel: 112..4-9. transverse. Steel Founders' Society of America....107 105 CYCLES TO FAILURE SoN curves of a normalized and tempered AISI 4140 wrought steel in the longitudinal and transverse direction and cast 4135 steel normalized and tempered..7 ksi (763 MPa). ~"80--0~ 06.. . In general. 5th Edition.. ..6 .~-.\. Tensile strength for wrought steel: longitudinal. Ed. P 3-16 .7 ksi (770 MPa). .. ~ 0 I (f) (f) 55 ""8<.. 30 25 104 ~.0 ksi (758 MPa).. or fatigue property values of rolled steel are averaged. 110..>t: 60 "<. a 4140 rolled steel was tested in fatigue in the longitudinal and transverse position and compared with a similar Cr-Mo cast steel..__ . Rocky River OH. 110. One example of this is shown in the S-N curves presented above.. impact.. 1980. • • A 100 <Il .. UJ O-B~ (f) (f) I- a: 50 (f) ••• 45 40 \~\A UJ I60 a: (f) \\. Source: Steel Castings Handbook.6. they will be about the same as properties of cast steel. 4135 and 4140 Steels: Cast vs Wrought 85 80 4135 CAST STEEL 75 4140 (LONGITUDIAL) }NO. 35 . 80 o.~& . 0-- o 6. Peter F.....6 ....32 4140 (TRANSVERSE) 91 UNNOTCHED NOTCHED 6.. For these.. if the longitudinal and transverse ductility. 0 0 70 65 D~-& °o~ o ......

. en =* 200 106 CYCLES TO FAI LURE Fatigue characteristics (S-N curves) for cast and wrought 4100 series steels.. 1980. For ferrous alloys a true endurance or fatigue limit is reached below which fatigue failure is not observed. Secondly. 4135 and 4140 Steels: Cast vs Wrought STRENGTH YIELD TENSILE (MPo) ksi CAST 4135 WROUGHT 4140-L -T 80 75 WROUGHT CAST ELONG % 113 (779) B7 (560) 43 110 (758) 80 (552) 61 III (7651 81 (558) 30 ksi (MPol BHN 223 217 217 _ 500 ~ 70 65 60 55 . Ed. Peter F.~RANSVERSE ... ... ------~ o 400 11. <. Rocky River OH. . both notched and unnotched.1__ -. finally. the presence of a notch equalizes the fatigue properties of cast and wrought steels. b I W o ~ ::i <l UNNOTCHED' fh .. The above data also illustrate that the fatigue limit of notched test specimens is substantially below that of unnotched samples when the fatigue limit is computed on the basis of nominal stress.. Weiser. First. . 5th Edition... LONGITUDINAL .. AND TRANSVERSE .. ~ <.. while the fatigue properties of cast steel are lower than those obtained with the wrought steel. Steel Founders' Society of America... a fatigue limit is evident... The number of cycles to failure ofa structure subjected to the above stress history can be expressed in terms of the SoN curve shown above. P 4-8 . . it has less anisotropy.. below a certain cyclic stress amplitude. ' . -. Source: Steel Castings Handbook. That is. And. .. .92 4-10. quenched and tempered to the same hardness. .. 50 « ~ \ \ --- I en en lr 300 I- w I- lr en en w en 45 40 35 30 \ \ NOTCHED LONGITUDINAL ''-. The data presented in the above S-N curves illustrate several important points.. The fatigue life increases as the cyclic stress amplitude decreases. . fatigue failure will not occur for any arbitrarily large number of cycles.---NO FAILURE "" I '. . 11.

HAC 50 60 70 ·in ::. Source: Metals Handbook. ~. 9th Edition. 4140. 100 90 80 70 60 30 40 Hardness.--------. .~ ~ :J . when steels are hardened to 45 HRC or higher an increase in carbon content can increase fatigue limit.--------.~ u. 4053 and 4063 Steels: Effect of Carbon Content and Hardness 1000. 93 o to 2 micro-in. American Society for Metals.-------------. Volume I.. p 676 . As shown above. Properties and Selection: Irons and Steels. 6001-------: Effect of hardness and carbon level on fatigue limit of alloy steels.~ U. Although other alloying elements may be required in order to attain desired hardenability. 110 7001-------j-----. . they have little effect on fatigue behavior.finish 900. 8001--------/-------t--. sc ·E ~ :J .4-11. Metals Park OH. 1978.1--------/-------t--------t 140 130 120 ~ Q.--------.

. Mechanical Metallurgy: Principles and Applications._ ... which describes the crack propagation rate in stage II for a series of metals.. The Paris power law.' . 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. 'tJ'tJ E C121(j' cI2 Ill"' - 10· 1-1-_ _' .' .. is very useful because of its extreme simplicity.. Prentice-Hall.-lJ 10 20 304050 100 5 100 AK (MPaV'ffi) a AK(MPaV'ffi) b Fatigue crack propagation in an AISI 4140 steel: (a) longitudinal direction (parallel to rolling direction)... p 714 . >U E E U 10:' .. it can be considered that the parameter ~K describes uniquely the crack growth rates for many engineering applications. the structure of material can influence fatigue crack growth rates drastically... The above charts illustrate the directionality in the fatigue crack propagation rate in an AISI 4140 steel... 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. 10-' Q) ~ U ..' . Also. Thus. due to the presence of elongated inclusions.... However.. (b) transverse direction (perpendicular to rolling direction). 1984.... >U .._ .. and plate containing a partially through-thickness crack -all fall on the same line.. 4140 Steel: Effect of Direction on Fatigue Crack Propagation 10' 10-' L T .' 10· L. Inc._ .' .... Source: Marc Andre Meyers and Krishan Kumar Chawla. Q) ~ ... 'tJ'tJ E .' . Englewood Cliffs NJ.94 4-12.. the value of m can change a lot. The exponent m has a much higher value in the transverse direction than in the longitudinal (rolling) direction.L_ _-'--_---J'---'--'--'-_--. For example. through-thickness central crack in a plate...

_ " .. American Society for Metals. The use of cathodic protection to prevent corrosion fatigue of steels depends sensitively on the hardness of the steel. At higher hardness values. w ~ -1.. cathodic protection of a 4140 steel was shown to be feasible for hardness values of Rockwell C 40.0 -1. Source: D. "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys. . P 360 .- . 4140 Steel: Effect of Cathodic Polarization 95 Carras. 1979. presumably due to hydrogen embrittlement (note above chart)... (fj LLi-O.~ • o (/) -0.r <{ o I0. For example. Metals Park OR. At potentials large enough to inhibit corrosion fatigue for softer steels.8 • .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.. an improvement in fatigue resistance is observed for moderate cathodic potentials... but complete protection is not possible.4 I Potential ." in Fatigue and Microstructure. Duquette.4-13..6 ~ >-0. J. a decrease in fatigue resistance is observed.

MICROINCHES· Relationship between surface roughness and endurance ratio (endurance limit divided by tensile strength) of quenched and tempered cast 4330 steel rUTS = 165-185 ksi (1138-1276 MPa)]. 5th Edition.JS~IL~ICAGR. Fully reversed plate bending tests. Ed. .:::::====~~:. SILICA PRIME INV.RMS ..5 1. Steel Founders' Society of America. a DRAG MACHINED 200 400 600 800 1000 SURFACE ROUGHNESS .CHROM ITE COPE WASH GR. INV CRYOLITE HIGH L. WASH _ _ _ _==--========. or conventional castings produced with special mold washes.:. Source: Steel Castings Handbook..\ ~ w <l: 0:: """"'.10. 0.96 4-14.S. Weiser. fully machined and polished plate specimens. Rocky River OH.rnrn.. Peter F. 1980. or better than. performed equal to. PROPR.2 .RMS.2 o / f===:J~ . P 15-29 . Cast 4330 Steel: Effects of Various Surface Conditions SURFACE ROUGHNESS. ZIRCON INV. 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. SILICA ALUMINA C 6 C 16 WASH ZIRCON GR. S.5 3 FUL L Y MACHINED ...5 2 2. Plate bending tests for quenched and tempered low alloy cast 4330 steel indicate that investment cast surfaces.:=~:::..

. Rotating-beam fatigue specimens tested at 10 000 to II 000 rpm..".... 110 7001-------1---------11--------1--------:::J 100 e :... American Society for Metals. From these data it is evident that scatter increases as strength level is increased...+ .. These data show survival after I a million cycles of AISI-SAE 4340 steel with tensile strengths of 995.. 9th Edition....t . 191. and 267 ksi).~~~k:... ~ c '" ...t . Volume I.. 1 heat 50 300 '-L-----''----~~ _:_:_' 800 1100 2000 Interrelationships of alternating stress. tensile strength and expected percent survival for heat treated 4340 steel..~ ~ ~ 50% survival Ii 600 f-------t--------::.~ 70 ~ ~ E 400 f . and 1840 MPa (144...i 60 Approximately 1000 specimens. Coefficients of variation range from 0."" "' 1.rr-- - - - 240 rr-r -- - - 280 .4-15..' . 4340 Steel: Scatter of Fatigue Limit Data Tensile strength..---.. 1320.... P 678 .. Source: Metals Handbook.Io--"""'-----t------=90% 90 .: 500 ~ 1----..~l=____=.. Metals Park OH..... ksi 97 12 800 iir=0- - - - 160 ---'-T-'------ - - 200 ----..o>'-=::::::=:1f===-----l 99% 80 c '" .20.17 to 0. Properties and Selection: Irons and Steels.. 1978.

1978. . 9th Edition. American Society for Metals.ro.98 4-16. P 672 . 4340 Steel: Strength vs Fatigue Life '. 2N f Typical data for strength versus fatigue life for annealed 4340 steel. = 174 ksi / ' F. Metals Park OH.o9 / I Fatigue strength / = exponent = slope b = -0..17412N.)b .tigue strength eOr'ie. = 0. Source: Metals Handbook. Properties and Selection: Irons and Steels. Volume I.0'.09 - Reversals to failure. 0'.ent 0.{2N.

.4-17..-...57 + 0 0062(2N r O.------.... ..2 ~ = l1€p + 2 ~ E = 0 5Bl2N r O.. Properties and Selection: Irons and Steels.:----1----+-----+------1 10-4l::100 --'-:- ---lL::- -'-:: -'--:-- -'-::-- -..1 " . 99 10-11----+---"O:-"'. 9th Edition. 1978. ... Volume I.1 ..... Source: Metals Handbook.L:--~"___J Typical data for total strain versus fatigue life for annealed 4340 steel.------..--I-----r..+ .. .09 .-... American Society for Metals. 10-2 1 . . 4340 Steel: Total Strain vs Fatigue Life 100....-----r-------. ...... P 672 .." ".. Metals Park OH... .....

aj-the fatigue strength coefficient (equal to stress intercept at 2NJ = 1). p 697 . It is convenient to consider separately the elastic and the plastic components of strain. f ~ 3 10 b ~ C ~IN 0 10 2 10 1 Stress amplitude (!:J. 1984. It was observed that fatigue life increased with decreasing b. The above chart shows an application of this relation to SAE 4340 steel. Mechanical Metallurgy: Principles and Applications. based on energy considerations. Source: Marc Andre Meyers and Krishan Kumar Chawla.. Prentice-Hall.e. Morrow.. Inc. Of course. This relation is an empirical representation of the S.100 4-18. and b the fatigue strength exponent.or strain-controlled) increases with a reduction in n'. showed that the fatigue strength exponent is given by: n' b=--1 + 5n' where n' is the cyclic hardening coefficient.aj2) versus number of reversals (2 N J) for AISI 4340 steel. Ej is the ductility coefficient in fatigue and is equal to strain intercept at 2NJ = 1. a o the true stress amplitude. the better it is for fatigue. 4340 Steel: Stress Amplitude vs Number of Reversals 4 10 ( V = Uf =1200 MPa C Q. Thus. Englewood Cliffs NJ. and c is the ductility exponent in fatigue. the higher the material coefficient aj. twice the number of cycles): _e_=_o_= _ _ (2N)b ~E 2 E a (a'J) E f where ~Ee12 is the elastic strain amplitude. The plastic strain component is better described by the Manson-Coffin relation: ~E -p-= 2 Ej( 2NJ r where ~Ep12 is the plastic strain amplitude.N curve above the fatigue limit. 2NJ is the number of reversals to failure. NJthe number of cycles to failure. There is evidence that ajis approximately equal to aJ' the monotonic fracture strength. The elastic component can be readily described by means of a relation between the true stress amplitude and the number of reversals (i. the fatigue life under elastic cyclic conditions (whether stress.

American Society for Metals. Metals Park OH. c e ~ 150 0 0 ~ ~ strain at beginning of test • Periodicoverstrain Vi e 800 100 600 2 10 3 10 104 105 106 7 10 Number of cycles to failure Overstrain superimposed on constant strain may have a significant effect on fatigue life.>l 1J. 101 e " c '" ~ Vi 1200 1000 o No overstrain or single over0 200 . Source: Metals Handbook. 4340 Steel: Effect of Periodic Overstrain 2000 250 1500 ::.. 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). P 681 . Volume I.4-19. 1978. 9th Edition. Properties and Selection: Irons and Steels.

Estimating fatigue parameters from the Brinell hardness number provides more conservative estimates. and the methods may not apply for every material. MPa Potter has described a method for approximating a constant-lifetime fatigue diagram for unnotched specimens. Using this method. American Society for Metals.00 600 800 1000 1200 1'00 Minimum Sites'. The calculated lines correspond well with the experimental lines. Here is presented a comparison between a calculated constant-life fatigue diagram (solid lines) and experimentally determined data (dashed lines). Volume I.0) that corresponds to the ultimate tensile strength. Metals Park OH. a series of points corresponding to different lifetimes are calculated and plotted along the diagonal line on the left side (R = -1). ksi 600 --400 ~200 200 . These results are only approximations. 4340 Steel: Estimation of Constant Life Minimum SIren.102 4-20. Each of these points is connected by a straight line to the point of the other diagonal (R = 1. Properties and Selection: Irons and Steels. 9th Edition. Source: Metals Handbook. 1978. the predicted lines represent lower stresses than the actual data. Generally. P 681 . A comparison between the estimated constant-lifetime diagram and the experimentally determined diagram is given in the above illustration.

1035 MPa (150 ksi).0 200 c. Metals Park OH. .. 1000 150 ~ ~ "'" 800 ~ "1""""') O't. ksi 103 1. Properties and Selection: Jronsand Steels. It may be noted that lives of the specimens at the three higher strength levels are about the same. 1380 MPa (200 ksi) and 1790 MPa (260 ksi). :... Source: Metals Handbook.() .4-21. American Society for Metals. 9th Edition. 4340 Steel: Effect of Strength Level on Constant.1. hardened and tempered to tensile strength levels of 860 MPa (125 ksi).. :9~ 'x 600 " E E E 100 60 0 ~ 'x E 400 """~~'1. All lines represent fatigue lifetimes of one million cycles.Life Behavior Minimum stress. MPa Constant-lifetime fatigue diagram for AISI-SAE 4340 alloy steel (bar). Volume I. P 669 . the scatter in data is at least as great as any real differences in fatigue life among specimens. 1978... ~ ____<_ ____L_ ____'___ _ ___'____ _ _~ 10 6 cycles lifetime 50 200 o'----_---'-__---'--__--'--__-'---__ _ __''"____ -1200 -1000 "'1100 -tiOO -400 -200 L __ _ _ '_ _____' 0 200 400 600 800 1000 1200 1400 1600 Minimumstress.

Solid lines represent data obtained from unnotched specimens.' -_ _ -'0 1400 -een -il00 -400 -200 200 400 600 800 1000 1200 Minimum stress.. 1978.L-1000 ' -_ _---'- ..' ' .L- -'--_ _.. p 667 .-_ _. Metals Park OH. hardened and tempered to a tensile strength of 1035 MPa (150 ksi)...104 4-22. 100 600 ~ "w E ~ :!E 400 60 200 O'--_ _.L-_ _. 4340 Steel: Notched vs Unnotched Specimens Minimum stress. Source: Metals Handbook. dashed lines represent data from specimens having notches with K.. 9th Edition. Volume I.. "~ ~ E E ~ . American Society [or Metals..' - .' " ' " -_ _--'---_ _. Properties and Selection: Irons and Steels.3. = 3...MPa Constant-lifetime fatigue diagram for AISI-SAE 4340alloy steel (bar).ksl o 160 ~ ~ 800 1.

.. as indicated in the above S-N curve.08 to 0.4-23.. which reduce the fatigue limit of the material. 1200 :. it significantly reduces the fatigue limits of steel. 9th Edition. If.) on AISI-SAE 4340 notched specimens that were heat treated to a strength level of 1860 MPa (270 ksi) reduces the fatigue limit almost as much as a notch with K.=3.: . Hardening a part with a decarburized surface can also introduce residual tensile stresses. American Society for Metals. 4340 Steel: Effect of Decarburization 1800 250 1500 0 105 200 0 Ie. ~ .76mm (0. Results of research studies have indicated that fatigue properties lost through decarburization can be at least partially regained by recarburization (carbon restoration in the surfaces).030 in.. Number of cycles to failure Decarburization is the removal of carbon from the surface of a steel part. 1978. Decarburization offrom 0.. Metals Park OH. Volume I. the decarburized surface layer is weaker and therefore less resistant to fatigue than the core. 0> C . 100 ~ ~ ~ <t 50 >- 300 .~ ~ 600 ~ r.~ c 900 •• 0 0 0 o Not decarburized 0 0 • •• • Decarburized 0 0 - 150"" ::i e' :.. Properties and Selection: Irons and Steels. Source: Metals Handbook.. . P 674 .003 to 0. When subjected to the same heat treatment as the core of the part.

one lot (lower curve) contained abnormally large inclusions. o Small inclusions • Large inclusions s: :.. <. 1978. Properties and Selection: Irons and Steels. Points on the lower curve represent the cycles to failure for a few specimens from one bar selected from a lot consisting of several bars of 4340H steel.r--. steels can be selected on the basis of such inspection. Metals Park OH. Source: Metals Handbook. 4340H Steel: Effect of Inclusion Size 1000 --.) were detected in the other specimens. P 673 . ~ 900 1. which reduces the number and size of nonmetallic inclusions. 140 c en .13 mm (0..00075 in. Large spherical inclusions. No spherical inclusions larger than 0.005 in.: . Volume 1. 9th Edition. Vacuum melting.) in diameter. the other lot (upper curve) contained small inclusions.~ <. about 0. the inclusions were identified as corundum or silicate particles. Large nonmetallic inclusions can often be detected by nondestructive inspection.: 700 s E 800 ~ ~ - 1 100 Number of cycles to failure Fatigue life of two lots of AISI-SAE 4340H steel. ~ ~.02 mm (0..106 1100 4-24. increases the fatigue limit of 4340. American Society for Metals. were observed in the fracture surfaces of these specimens.

however. J. Tomkins. p 115 . 1981. As this is an order-of-magnitude greater than dislocation substructure sizes. Bressers. more complex alloys planar slip behavior dominates.. Applied Science Publishers Ltd. For stronger.. Ed. As well as debonding. London. "Fatigue: Mechanisms. the latter being particularly operative in hcp metals. Typical initiated crack sizes are l-lO Mm. 4340 Steel: Influence of Inclusion Size Xl0 200 190 180 STRESS RANGE. such an initiated crack will behave as in a continuum. Other microstructural sites for initiation are discontinuities such as grain and twin boundaries. producing a surface crack." in Creep and Fatigue in High Temperature Alloys. The interaction of slip bands with second-phase particles (inclusions. Variations in fatigue life relate to variations in inclusion size.: :I 0001- • SURFACE INCLUSIONS a SUBSURFACE INCLUSIONS I--~~ CYCLES SoN curve and dependence oflife on inclusion size for AISI 4340 steel. at ambient temperatures it is the dislocation substructures which dominate initiation. thus inhibiting initiation. The above SoN curves show the results of this process for a low alloy steel. making localized slip bands the initiation sites cfthe random notch-peak topography generated by shearing a pack of cards. Ib/on 2 170 160 150 140 130 i20 103 3 107 • •• tal • DIAMETER. oxide or carbide particles can crack under concentrated localized stresses. England. Usually. precipitates) can produce a local stress concentration which cracks the interface. For such materials crack initiation can occupy a significant fraction of life. For materials with lower stacking fault energy cross-slip and PSB formation is difficult. on " " .4-25. Source: B.

As hydrogen content is decreased by baking. 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.e.~ \ \ 0.5 hr . ductility as measured by total elongation to fracture or reduction of area may be decreased..~. Source: George Krauss. There are many embrittling effects of hydrogen on steels: the ultimate strength of a steel may be reduced. and sufficient baking eventually restores the strength of charged specimens to that of uncharged specimens. and crack growth may be significantly accelerated. "\\ \ + I'\. 1980. C« .01 \ \.. Hydrogen may promote a transition from a ductile to brittle fracture mode or it may reduce ductility without a change in fracture mode. Metals Park OH. the lower the static fatigue limits.. the static fatigue limit increases. an indication that a critical combination of hydrogen concentration and triaxial stress state is required for crack initiation. the sharper the notch. In general. The graph above shows the effects of baking at 150°C (300 OF) on the static fatigue (sustained loading) of the hydrogen-charged specimens. The horizontal portions of the curves in the graph above are designated as static fatigue or endurance limits. American Society for Metals. '" ~ + 0\ \ 1 <. Increasing baking time effectively lowers hydrogen content even in the plated specimens. -300 ksi + \ Uncharged ++- '--- Bake 24 hr - '00 "\ Bake 18 hr ul (/) (.x: Ai> '~ -. 275 250 225 . The specimens used to obtain the above data were notched and therefore the static fatigue limits hold for that particular notch geometry. 0. 4340 Steel: Effect of Hydrogenation. i.108 4-26.) ~ 200 175 150 125 100 75 50 e\ "0 a.. p 223 . Bake 7 hr Bake 3 hr " . 1 Bake 12 hr :\ ~ \ t-.. the stress level below which failure would not occur no matter what the duration of stress application.1 100 • 1 10 Fracture lime. 1000 Static fatigue curves for quenched and tempered 4340 notched specimens charged with hydrogen and baked at 150°C (300 OF) for the times shown. Static Fatigue Normal notch strength = . Principles of Heat Treatment of Steel. hours Bake 0.

EMBRIT.. Syracuse University.. 1956... P 508 .000 1000 10 100 CYCLES TO FAILURE <> • • • . Schematic representation ofthe effect of cycling rate on theS-N curve of hydrogen-containing 4340 steel. r-I Cf) Cf) co ~ RATE UNEMB..4-27.." in Materials Evaluation in Relation to Component Behavior (Proceedings of the Third Sagamore Ordnance Materials Research Conference).. Syracuse NY.... Source: George Sachs..000 psi.. .. heat treated to a strength level of 250. "Test Methods for Evaluating Hydrogen Ernbrittlement.000 PSI ~ co P-t 0 0 0 H • . ~ .33 RPM 0 10. PLATED IN LAB 1000 RPM COMMERCIALLY PLATED 50 250 RPM 0 PLATED IN LAB 200 RPM 'V PLATED IN LAB .. 4340 Steel: Effect of Hydrogen 109 400 350 300 250 200 150 100 4340 250.

--lAYER ATMOSPHERE GROUND TO RDlOVE COMPOUND ATMOSPHERE NITRJDED TJfEN ------------ HITfUDED 5'> QUENCHED AND TD1PERED AT 1050 f (565 C) HUHBER OF CYCLES 10 ~ 10 5 S-N curves for 4340 steel. American Society for Metals.-. 4340 Steel: Effect of Nitriding ----. gaseous atmosphere nitrided versus not nitrided (quenched and tempered only). A. Metals Park OH. "Short Cycle Atmosphere Nitriding." in Source Book on Nitriding. 1977. Riopelle. showing stress versus number of cycles for completely reversing torsional fatigue. P 287 . Source: J.110 4-28.

.---1 heat treated :Ii 350 L -_ _--'105 . ----'_. Shot peening is useful in recovering the fatigue resistance lost through decarburization of the surface. P 674 .>. 160 Nitridedcrankshafts 111 120 800 100 ~ 6001----+----' Normally 500 I----+-crankshafts --+--+--+----+----+-----.------._ .L. Shot peening and skin rolling are two methods for developing compressive residual stresses at the surface of the part. 4340 Steel: Effect of Nitriding and Shot Peening 1200 r-----...... Properties and Selection: Irons and Steels....---. Mechanical working of the surface of a steel part effectively increases the resistance to fatigue. Metals Park OH.------..' ' .' ..---. Source: Metals Handbook.-_ _---.4-29. raising the fatigue limit from 275 MPa (40 ksi) after decarburizing to 655 MPa (95 ksi) after shot peening..------.. 1978. 9th Edition. American Society for Metals.-----.---------'5. ----' Cvcles to failure Comparison between fatigue limits of crankshafts (S-Nbands) and fatigue limits for separate test bars.. decarburized specimens were shot peened. Volume I.' _ . which are indicated by plotted points at right.-----------. The improvement in fatigue life of a crankshaft that results from shot peening is illustrated in the above curves.' ...

Kenneth E.112 4-30. Barnes and Joseph F. "Combining Nitriding With Induction Heating Pays a Bonus. \ \._ 150 \ a. '" \ 'r-. 1977. Levy. Libsch. Source: Sander A. P 241 . \ e u. ~-130 \ "I' 120 "1\"'r--. Metals Park OH. 4340 Steel: Effect of Induction Hardening and Nitriding 160 .. American Society for Metals. As ni ide :5 140 C!. fatigue tests of AISI 4340 steel in various surface hardened conditions show that combined treatments produce endurance limits between those developed by separate treatments." in Source Book on Nitriding. ilrid ~ DU nducti n har en d Indu tio a dened 110 10 5 10 6 10 7 Cycles to Failure As demonstrated in the above S-N curves.

180 113 0--6 _ O~---------<>~ 170 ---- Cr + SFl <>~Chromate Cd+ ~ O .LUJ SoN curves (axial tension) of bare and coated 4340 steel in air environment. they will become discontinuous at a relatively low stress level owing to the development of fatigue cracks... Swindlehurst..5t so so so 90 110 165 90 60 60 Axial Tension R = 0. "Corrosion Fatigue Behavior of Coated 4340 Steel for Blade Retention Bolts of the AH-I Helicopter..-----'---'-.3 -14.uJ. contain the data obtained in air for 4340 steel.5% NaC1 Stress MN/m 2 13B 552 621 621 621 758 1138 621 414 4. it may be easier for the environment to reach the crack tip.u.3 -14.w.8 140 2000 cpm RT ~TUngsten Carbide SFl + Cr + we . But in NaCl solution.LJL. E.4-31.-. P 275 .• Plenum Press. improved the fatigue strength. (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. Since the Cr and WC hard (brittle) coatings have a relatively low intrinsic fatigue strength in comparison with the steel..u. 20 80 3. particularly the chromium.-----'---'-.-. % -81 -24 -14. Eds . Air Stress Test Rotating Bending R = -1 Condi tion Bare Cd + Chromate Cr + Ory Film' WC + Dry Film' WC + Cori cone + Dry Film' Bare Cd + Chromate Cr + SFL' WC + SFL' WC + Cori cone + SFL' 'Shot peened tCompared to bare alloy air value ~Compared to coated alloy air value ksi 105 105 95 90 MN/m 2 724 724 655 621 621 1103 1138 1207 965 965 Change.u-. ~ ~160 ~1f8are j ISO o R • 0. which remains to be elucidated.Shot Peened lJOL-----.---'----'-L..5 -14.3 ks.5 t -57f -57~ The above SoN curves.Lu. 2 0 -43. Jr.---.u.l. 1980. % 0 -9. Fatigue data at 107 cycles showed that the cadmium and chromium electroplates. Source.John J. significantly greater reductions in axial fatigue strength of the coated alloys were observed due to environmental effects.1 +9.---.5 -62.5 -12..4 -12.8t -4B. in conjunction with the table.8 160 165 175 140 140 +3. bare and coated. M. In the case of the axial tension test (high steady tensile load).-----.Yin Risk and Failure Analysis for Improved Performance and Reliability.w.. New York NY.3 -31.L.3 -14. Levy and C.4 Change. They were similar in both rotating bending and axial tension fatigue tests.L..LlllJJ.. 6~ -62. 4340 Steel: Effect of Surface Coatings 190r---------------------------.. Burke and Volker Weiss.

:2 '" 800 100 600 ". "00 E :J E Q)lj.. MPa '"-_ _--'---_ _---'--_ _----'0 800 1000 1200 1400 Constant-lifetime fatigue diagram for AISI-SAE 4340 alloy steel (bars) hardened and. dashed lines represent data from specimens having notches withK1 = 3. All lines represent lifetimes of ten million cycles. 9th Edition. American Society for Metals.9 s 400 . Source: Metals Handbook.. 4340 Steel: Effect of Temperature on Constant-Lifetime Behavior Minimumstress... tempered to a tensile strength of 1035 MPa (150 ksi) and tested at the indicated temperatures.. 107 cycles lifetime ~~ ~ ". ksi 150 c.. J'J' '1". 1978.. P 669 . Volume I. Properties and Selection: Irons and Steels.114 4-32. 6'a 1:> 50 200 0"--_ _-'----_ _--'--_ _---'--_ _---'_ _--'''"---_ _--'---_ _---'--_ _----' -800 600 -1000 -200 200 400 --1l00 -400 Minimum stress. Metals Park OH. ~ "x " E E ~ "x "1'/. ~~. '" e 1.. Solid lines represent data obtained from unnotched specimens...0 .3.

::-~ ••••••••••••• 5 4 3 2 10 -----. when the comparison was made with higher-hardenability steels.'::. which resulted in fracture origins below the surface at the case-core junctures..4-33.. with sufficient gradient strengths and thus all fracture origins at the surface... 1979...~. the difference was very slight.:.... H... where distortion could be a factor-the results might turn out differently for the marquenching... Wene...... This is consistent with what is known concerning the differences in residual stress...N Effect of quench type on fatigue of carburized differential cross. American Society for Metals. though still in favor of direct quenching. I 40 30 20 ~4".~~... Source: D. 4520H Steel: Effect of Type of Quench 115 z . Metals Park OH.:l ~ ~ l=l 10 8 6 ~1ARQUENCH /-. P 92 . which in this case would have been the only other contributing factor.. "Fatigue in Machines and Structures-Ground Vehicles. DIRECf (OIL) QUENCH (COLD OIL) / ..:. Since 4520H steel is relatively low in hardenability for the part involved... / -.." in Fatigue and Microstructure. Breen and E. However..:l 0 0 0 .. the depth-hardening characteristics of the two groups were significantly different... In other instances-such as for gears.:::•••••••••••• (400 0F OIL)' ~.... The marquenched group had shallower case depths.. M... --- ~ 10 3 10 4 CYCLES . p:\ . ~....

the shot must also be hard to be effective... tests were run to determine the amount of increase to be expected..116 40 ~ . "Fatigue in Machines and Structures-Ground Vehicles... The net gain was small. .. H. 4520H Steel: Effect of Shot Peening 30 20 ~':".": . P 93 . M. CQ 4-34.... although not recommended because of the high risk of having grinding cracks...: / .. Peening parts that had marginal strength gradients improved the strength at the surface but moved the failure origin to a subsurface location.. Shot peening was found to provide significant fatigue-strength improvements. The above chart shows some of the results.J . 0 0 0 ~. Peening surfaces that had suffered grinding damage was found to be very beneficial. It was also determined that to gain significant improvement the hardness of the shot used was very important.':'.. Shot peening is known to increase fatigue strength.:. .. Metals Park OH. if. -- ~ F l:>: 2 10 10 2 CYCLES . Since carburized surfaces are very hard. c: u.J 10 8 6 5 4 3 ---~ SHOT PEE~D lJNPEE\'ED~ __•••••••••••• :::..:. American Society for Metals... hence. Breen and E. ----... Source: D.. § . 1979.." in Fatigue and Microstructure... Wene...r\ Effect of shot peening on fatigue of carburized differential cross.

"Short Cycle Atmosphere Nitriding.4-35. A. Metals Park OH.---.-QUENCHED AND TEMPERED AT 1050 F (565 CJ --- ATMOSPHERE NITRIDED THEN GROUND TO RI1'IOV£ COMPOUND r. nitrided versus not nitrided (quenched and tempered only).-. 1977. showing stress versus number of cycles for completely reversing torsional fatigue. 4620 Steel: Effect of Nitriding 117 -.-. P 286 ." in Source Book on Nitriding. Source: J.~Y~p Al110SPHERE NITRIDED HUHBER OF CYCLES 10~ 1') 5 S-N curves for 4620steel. Riopelle. American Society for Metals.

The diagram above shows the effect for fatigue resistance.0 Axial fatigue of P 1M-forged 4620 steel as a function of height strain during forging.-------'-----Time LI-! ... Lynn Ferguson.. Metals Park OH. Source: B.. P 100 . For re-pressed parts. 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... American Society for Metals. Upsetting increases longitudinal toughness while toughness in the through-thickness direction remains at a relatively constant level. 4620 Steel: P1M-Forged 100 x 80 Height strain oE 60 r..I_ _.' - _ 4. In general. . Advantages and Limitations.. Ed.0 Log cycles 7.0 6." in Powder Metallurgy: Applications.' . throughthickness toughness is slightly lower than longitudinal toughness. Erhard Klar. 1983.0 5. American Society for Metals._ . sensitive properties improve as the level of upsetting is increased during the forging process. "Part II: Fully Dense Parts and Their Applications.-_ _' . although the cyclic stress state also influences fatigue behavior.118 ksi 4-36.

. :2: x co ..4-37.-----. fatigue limit increases as deformation (level of strain) increases.£ CI) CI) E 500 cil ~ 1'------+ 30% 60 400 o~ 0 7.-------..'+-----+ (JrnaK Height strain. 4620 Steel: P/M-Forged at Different Levels 700 .. Volume 7.---------. HIH o 71% 56 & 65% 42% 80 . Metals Park OH. Source: Metals Handbook. Powder Metallurgy. 100 119 600 a. 1984. American Society for Metals.r--------.------.. p416 .0 8. 9th Edition.0 Log cycles SoN curves for P 1M-forged 4620steel at various levels offorging deformation. As shown.

Shown is a 10% life (L IO ) of 563 hr for P/M material compared with L IO life of 192 hr for ingot material. American Society for Metals. 9th Edition. as shown by a typical Weibull plot. h Fatigue life characteristics of P 1M roller bearing cups. Metals Park OH. 1984. 4625 Steel: P1M vs Ingot Forms 100 50 Ingot material L. P 620 . Powder Metallurgy.o 5 500 2000 5000 Life.o 192 h 10 / \ V h I 163 1000 1/ 1/ o 100 200 P/M material L.120 4-38. Source: Metals Handbook. Volume 7.

Wrought steel. American Society for Metals. Source: B. Lynn Ferguson. R. The most impressive statistic is that the P / M-forged parts passed the Army ambient and low-temperature firing endurance tests. Fatigue data for P/M forged 4640 are shown above. .~. 8 c'i) 80 dir 3-1/4 60 r- '" '. . Metals Park OH.----. Water-atomized 4600 steel powder was blended with graphite and compacted in the split punch tooling. "Part II: Fully Dense Parts and Their Applications...------..4-39. . Erhard Klar. ...365-1 ':3. 0. 1983.-------.-------. American Society for Metals.000 psi.-----."in Powder Metallurgy: Applications. tested in the '/ " longitudinal direction -. p 103 . '0 " ~ 'Scatter band of SAE 4340 ~ u~ 32 Specimen configuration 21 f '0.r------.. Ed. '-----------'~ '------_Q: 9-7/8 R. and these data fall within the scatter band for 4340 steel. Moore fatigue curve for P/M-forged 4640 steel hardened and tempered to 33 HRC and a yield strength of 138. 40 '--------'---------''------'-------'---' 105 10· 10' 103 10' Cycles of stress R. Advantages and Limitations. 4640 Steel: P1M-Forged 100 121 .

...... 1978..78 C. .. (c) 0.. 0. % .. Source: Metals Handbook..... .. 0... 9th Edition. 28.....9 17. MPa (ksi) 490(71)(b) 248(36)(c) Elongation in 2 in.. 5 0 Spheroidite Pearlite 45 '" ~... Metals Park OH.. c: ~ 35 ~ « Tensile strength.1% offset yield strength.016 S and 0. i------j ~ 40 'f.. Properties and Selection: Irons and Steels. American Society for Metals. on p 83).. 0.. MPa (ksi) 641(93) 676(98) Yield strength.011 P. . (b) Lower yield point...8 Hardness.22 Si.122 4-40. 7 6 10 10 Number of cycles to failure Both pearlitic and spheroidized structures have notably lower fatigue strength than martensitic structures (see 4-1. Volume I. the fatigue properties of spheroidized structures are superior to those of pearlitic structures for eutectoid steels..8 Reduction in area.... High-Carbon Steel (Eutectoid Carbon): Pearlite vs Spheroidite Property 350.... % 57.. P 677 . 92 89 (a) Composition 0..27 Mn. HB .7 25. As is shown above.

4-41. 52100 EF Steel: Surface Fatigue; Effect of Finish and Additives

123

600

4

2

8

6

12

10

...
.:<

500

'"

1 2 3 4 5 6 7 8 9 10

11
12

til

. ... ..
E

.;

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

' .." 450

3.10
N

~ 400

2. 76

~

E

.~ x
~

'"

350

2.41

300

2.07

Mean Predicted Cycles to Failure

Effect of surface finish and additive on mean predicted surface fatigue life. 52100 EF steel, high slip, high speed.

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

124

4-42. 52100 EF Steel: Surface Fatigue; Effect of Surface Finish and Speed

600

II

4.14 3.79

... 500
Ul

3.45

""

. '" rl ..
'" e
~
x

::: 450

:;
400

/---e---i
1

3.10 GLl GL2 PLl PL2 GU GL2 PLl PL2 GLl GL2 PLl PL2 GH2 GH2 GH2 PH2 PH2 PH2 10 7 Mean Predicted Cycles to Failure
N

2
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

350

:2
300

5E 5E 5E 5E 5V 5V 5V 5V 8V 8V 8V 8V 5E 5V 8V 5E 5V 8V

2. 76 ~
to

e

9

5

2

2.07

Effect of steel, surface finish and speed on mean predicted fatigue life. Low-viscosity mineral oil, no additive.

Interaction effects of steel with speed and of surface finish with slip and speed on fatigue life are shown in the above graph. The direct steel effects are nonsignificant. The effect of surface finish is shown in the difference between the two line groups 13, 14, 15 (ground) and 16, 17, 18 (polished). The difference in the line groups 4, 8, 12 (low slip) and 16, 17, 18 (high slip) again brings out the very large detrimental effect of high slip on life. Higher speed decreases life with the maximum effect observable on 8620 CV steel (compare lines II and 12), on polished specimens at low slip ratio.

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

4-43. 52100 EF Steel: Surface Fatigue; Effect of Lubricant Additives

125

600

.... .:< '" '" '" "
~

HLN 1.2 HL\ 1.2 HLC L2 ~1.0 1.2 HI.N 112 HL\ 112 HLC 112 ~LO 112

4.14 J.79

J.45

90% Con fidence
Band
~

J.IO
N

::l
x

"
E

.

400

2. 76

~

E

.~ x J50
:E

.

"

2.41

JOO

2.07

250 L...;--'----'--'L..L__'_'..u..L..;-_ _'____'L....L..L..L.J..u.~--'----'L....IL..L--'-'..u.....--'----'L....IL..L--'-'L..1..L! I. 72 IO

J

10

6

10 7

:ican l'r-ed i c t ed Cyc l e s to Fa i lure

SoN diagram for 52100 EF steelshowing the effect oflubricant additives on surface fatigue. The specimens had a ground finish, and a lowviscosity oil was used. Additives were used for I, 2, 3 and 4; the favorable effect of the additives is obvious.

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

126

4-44. 52100 EF Steel: Surface Fatigue; Effect of Lubricant Viscositv, Slip Ratio and Speed

600 550 -;: 500
.Yo

1 2 3 4 5 6 7 8 9 10

.

11
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

'

l/}

:: 400

'"
E

. .
~

t

:: 450

f---<T---i

3.10
N

-!:

2.76

G

~ 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

Mean Predicted Cycles to failure

SoN diagram for 52100 EF steel.

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

4-45. 52100 EF Steel: Rolling Ball Fatigue; Effect of Oil Additives
SOO , - - - - - - - - - - - - - - - : : - - - - - - - - - - - - - - , 3 4 5

127

400

(OMe bl & 68'"

BASI 011 OA'"

2.76

300
~

2.07

vi
w
~

~
D<

N
<,

E

~
Co

\
200

z e

::c

W

" " "
:>
)(

"'4
.}Is-,f....
~o

1.38

o ()"''''
""'.. F

ot

-,
----'HASiIllOAU. 'I(DI(1I0 BY MUllIPll UQRfUION ANAlYSIS

10
l,. LIFE. 10 6 CYClES

100

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

128

4-46. 52100 Steel: Carburized vs Uncarburized
99. 95. 90. 80. 70. 60. 50. 40. 30.

0
ILl

...J
IL.

ex 20.
10.

...
0

Z ILl

a:::

ILl

a.. 5.0 4.0 3.0
2.0
I. 0
L..-_-'------'-----'----'''--'--'l..J...J...L--_-'------'-----'----'''--'--'LJ...LJ

I

10
MILLIONS OF STRESS CYCLES

100

Rolling Contact Fatigue Tests
Bar specimens, 0.973 em (0.383 in.) in diameter, about 8 em long, were machined from spheroidize-annealed 52100 steel. Two pieces were copper plated to prevent carburizing, then, along with two unplated pieces, were austenitized at 815°C (1500 OF) for two h in a carburizing atmosphere, oil-quenched and tempered for I.5 h at 175°C (350 OF). After finish grinding to 0.953 mm (0.375 in.), pieces were fatigue tested using a Polymet Model RCF-I testing machine with a computed maximum hertzian contact stress of 503 MPa (729 ksi). A Weibull plot, shown above, of the 16 tests on each type of specimen shows that pieces with a carburized surface had a fatigue life about 50% longer at all failure rate levels than pieces which were subjected to the same thermal cycle, but not carburized. The nonparametic Walsh test for statistical significance indicated at a 99.5% level of confidence, the two batches of fatigue test data came from different populations.

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

4-47. 8620H Steel: Carburized; Results From Case and Core

129

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

130

4-48. 8620H Steel: Effect of Variation in Carburizing Treatments

300

Single --......... reheat .........

• -------~rr"~
<lID - /""":
Direct quench ---./ 0

.0

o

o

o

100

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
Cycles to failure

107

The above S- N curves show results of a study of the effect of martensite morphology, including the effects of micro cracking on fatigue resistance of a carburized 8620 steel. These specimens, which were directly quenched from the carburizing temperature, had the coarsest structure and the highest density of microcracks, some of which were directly exposed on the specimen surfaces bychemical polishing. The single-reheat specimens had a finer austenite grain structure and therefore finer martensite plates and a lower density of microcracks. Since the retained austenite content and hardness profiles of the direct and single-reheat specimens were identical, the improved fatigue resistance of the single-reheat specimens is attributed to the smaller size of the microcracks and their lower density in the finer structure. The best fatigue resistance was shown by the double-reheat specimens.

Source: George Krauss, Principles of Heat Treatment of Steel, American Society for Metals, Metals Park OH. 1980, P 264

4-49. 8620 Steel: Effect of Nitriding

131

50

§
~

....

_-- .... _--

;
iO

~

-- .... --

--

.THO~l'\IrRr NITRIDw TNr.N GROUND TO RJ>1OVr COMPO''''" lAnR

:;;
QlJ[HCHED AHD TD4PER£D

AnlOSI'IlERE NITRIDED

AT 1010 F (\6\ C)

tHJKBER OF CYCUS

10

7

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

4-50. 8622 Steel: Effect of Grinding
EFFECT OF GRINDING BURN
C/)

40

~ 30 ...... 20

o o o

~
_

"':'.:.:-. - • _ 10 8
6

...

.....

............... . ...................

. ....

-.-.-.-.
....

8622 MATERIAL

..... ............

4

........

2

GRINDING WITHam BURl, - • - . - '" SEVERE DAHAGE BY GRINDING •••••••• •••••• (REFER FIG. (~4))
10 2 10 3 10 4 10 5 CYCLES TO FAILURE 106

10

Influence of grinding quality on fatigue properties of carburized differential cross.

C/)

z ......
I

~

40 30 20 10 8 6 4 2

~~

§
P-

"2'::.".::- •

....-::-.:.......

.....::: ........

8822 MATERIAL

o
0 ..... 0

...... ........::: . ........
~

'"

~

Z ......

t::l

~

GRIND AFfER H. T. GRIND BEFORE H.T.

- . _.-

ffi
~

10

10 2 10 3 10 4 lOS CYCLES TO FAILURE

10 6

Influence of grinding sequence on fatigue of carburized differential cross.

As shown in the above charts, grinding has an important influence on fatigue. Elimination of grinding damage resulted in drastic improvement in fatigue performance (upper chart). However, it was also determined that a high-quality ground part gave better fatigue performance than when the carburized surface was left unground (lower 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, pp 91-92

4 III I- 0:: 0:: I- :n III .6 0.0 CAViTIES CAST STEEL . it will be observed from the Goodman diagram above that the results of the notched [0.:.5 l: l: lZ III Ul e> ~ 0. 1980.a- X- 0. Weiser.:=.0381 mm) radius] R.J ..::.::. Moore fatigue data.4 Goodman diagram for bending fatigue for normalized and tempered cast 8630 steel.:.SOUND SLAG INCLUSION HOT TEARS 133 1. Steel Founders' Society of America.0. (0.:=--=-""-=~ RANGE 0. R. R I CONTAINED WITHIN THE BAND ARE THE FOLLOWING DISCONTINUiTIES WELD -INCOMPLETE PENETRATION WELD . R.2 III I- 0. MOORE FATIGUE SPECIMEN (NOTCHED 0.3 III e> lL => ~ ~O. Moore fatigue specimen fall below those of the other bending fatigue values.9 OoB 0. It must be remembered that the discontinuities were very severe and exceeded all ASTM classes of nondestructive inspection standards. .SLAG WELD . In many cases..1 .3 .. Source: Steel Castings Handbook.7 0. Peter F. Data here show that severe discontinuities lower the fatigue strength of cast steel. 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. However. Ed.:. MOORE FATIGUE SPECIMEN I UNNOTCHED I R.0. P 15-32 . therefore. R. Cast 8630 Steel: Goodman Diagram for Bending Fatigue LEGEND -.R.. design.1 0 o. 5th Edition.SOUND R.0015 in.UNDERCUT WELD . The allowable discontinuities described in the ASTM standards are therefore expected to exert a somewhat less damaging effect on fatigue behavior..4-51. Rocky River OH.s MEAN STRESS TENSILE STRENGTH . introduces a safety factor. based upon notched R.MACHINE .:.001!! in.SOUND AS WELDED .0.2 .0 ..0.

134 4-52.6 . Source: Steel Castings Handbook.) w 0. (SUB SURFACE) ~ ~ 0.4 (/)1WZ 00:: W 138 135 137 ~ 0. --. Ed. It 0. • W 0-. As shown in the chart here.26 0.17 0. 1980.!l 138 (~) (951) (951) (931) (945) 0.3 6. (EXTENg~ SURF. UTS END. Cast 8630 Steel: Effect of Shrinkage 0.2 Zl0:: I-.. Steel Founders' Society of America.1 ( EXTENDS TO ~URFACE )".13 0.5 (/) (/) g: ::c 0.1..28 o ::::>W i= Ii.r------.--_ _---' 104 10 5 106 10 7 ----' CYCLES TO FAILURE Effect of shrinkage on plate bending fatigue of quenched and tempered cast 8630 Ni-Cr-Mo steel. plate bending tests (completely reversed tension and compression with cast-to-size specimens) oflow-alloy cast 8630 steel indicate only minor effects of Class 2 internal shrinkage. 5th Edition. Rocky River OH. !. "" flLASS 2 SHRINK 6. Weiser. Peter F..-~ NO FAILURE - 0'--------'-----.--------.J <3: CLASS 6 SHRIN~~A~~ ~.--------. P 15-30 .J z_(/) I:::=0.----.K CLASS 2 SHRIN ~ t.

Peter F. 0-0 NO. SOUND NO FAILURE CLASS 6 SHRINK CLASS 2 SHRINK 2 . . . 0.5 135 . 15 NO.627 MPa) - OCr t.J w x I <.. .. Cast 8630 Steel: Effect of Shrinkage on Torsion Fatigue 0... .91 ksi (579 .... Weiser.2 ICLASS 6 W I. ~0 """"-- ° rSOUND Z(f) 0.--:. Rocky River OH./ -1-..4-53. Source: Steel Castings Handbook. Ed... . Steel Founders' Society of America.4 - ... P 15-31 . 13 - x-x CYCLES TO FAI LURE Effect of shrinkage on torsion fatigue properties of annealed cast 8630 steel.t. .1 r- ~~SHRINK SHRINK-...0O'::CLASS . 5th Edition.-------r-----r--------.. 1980. ~.. I I I TENSILE STRENGTH (f) (f) w 0...------.J <1 zlOC ~~ 0. . X ~ 9t..84 .. . 13 NO.3 • oc Z w <1 w OC (f) rt...

.. _ 0 o 0 o~--o_ X_X_ o~ 0 00 ... Rocky River OH. 12 NO.1 I- 0--<> - - x-x SHRINK - I I I 10 5 10 6 10 7 CYCLES TO FAILURE Effect of shrinkage on torsion fatigue properties of water quenched and tempered cast 8630 steel.~?... Steel Founders' Society of America. o Xo .2 I0-0 NO.------r----~----. XC o . Source: Steel Castings Handbook. 12 .3 ~ Z «<. Ed.. .5 . 0... Cast 8630 Steel: Effect of Shrinkage on Torsion Fatigue 0. Weiser. Peter F.. I I I T ENSILE STRENGTH (917-951 MPo) W lr en en 0. 1980. p 15-31 .136 4-54... 15 NO.4 I- 133-138 ksi I-I en I- o lr en en ~--1 W lr IIW <t W 0..0 _ ..__---___... 5th Edition. Z in <tW j::Z ZI- SOUND CLASS 6 CLASS 2 NO FAILURE SHRINK lr W ~ <t 0....

. WZ :::>W ~a:: l:i. RATIO . 1980. Cast 8630 Steel: Effect of Shrinkage on Plate Bending 0.. \ AXSOUND 6..5 CLASS 2 Lt ~tn ~..... Rocky River OH.32 ..3 W ~ . Steel Founders' Society of America.0.J 104 CYCLES TO FAILURE Effect of shrinkage on plate bending fatigue of normalized and tempered cast 8630 Ni-Cr-Mo steel.. Weiser.4-55. 0 ° '--6_ 0_ a:: I6 STEEL UTS ~ I MPo 1 END. P 15-31 .. Source: Steel Castings Handbook.4 W ( SUB SURFACE 1 Z_en I-Z <l:W Z I.-------.J SHRINK ----- 0..-------. 6 "'0'0 6''00..J....-----. Ed. ..- ° II 14 83 B4 (5721 (5791 ----' ..... 0. 5th Edition..-.2 L. Peter F..6 137 en en W a:: I-J: en t... o_"'\.35 .J <l: 0...

0. U5 300 40 200L----------'----------'-0. Volume I.N curves in the graphs above compare wrought 8640 and cast 8630 steel in two different conditions of heat treatment. c.------=""""'ct----------t---------.1 1 10 Millions of cycles to failure 600 Ouenched and tempered to 80 286 HB 500 :.01 0. but the two steels are practically identical in the notched fatigue test..0. but is reduced considerably by notches or a rough cast surface. 1978..-----------. :. the wrought 8640 is superior. 9th Edition.--------.----------. Metals Park OH. ~ .138 4-56. p 397 .. Cast 8630 vs Wrought 8640 600. Properties and Selection: Irons and Steels. This is significant because most articles fabricated from either wrought or cast steel contain more than one notch and more than one type of notch. 40 0. American Society for Metals. sc U5 e 400 60 ~. In both of these comparisons. Notched Unnotched Wrought 8640 0 • 80 Cast 8630 '" '" 5001--------.1 Millions of cycles to failure -----J 10 The fatigue limit for smooth-machined specimens is generally about one half the tensile strength. The S. Source: Metals Handbook. IJ) 0: 60 " ::i 400 1-----------+-----2-"""=-+-~---------l ~ ~ ~ 300 I-----------+"""-~""'_=------t------------j Normalized and tempered to 220 HB .

~!i'±i~_+--___l 80 5001---1-- "'----+---t----j 60 ] 300 u.~ " 40 20 100 L -_ _-'---_ _-'----_ _--'---_ _--'---_ _-'---_ _-'---_ _---l 500 600 700 800 900 1000 1100 1200 Tensile strength.4-57. P 397 .. 9th Edition. Metals Park OH. MPa 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. Source: Metals Handbook.----. ksi 139 700 r---..---. 80 100 120 140 160 100 . Volume I. 1978. Properties and Selection: Irons and Steels.-----. note curves above. American Society for Metals.-. • Cast steel 600 I---I-------Ir----t-----t-----.----. 8630 and 8640 Steels: Effect of Notches on Cast and Wrought Specimens Tensile strength.-. .------.------y-----.

p 287 . showing stress versus number of cycles for completely reversing torsional fatigue..." in Source Book on Nitriding.. A...... ... gaseous atmosphere nitrided versus not nitrided (quenched and tempered only)......... Nitralloy 135 Steel: Effect of Nitriding 60 .. 1977. . ... American Society for Metals.. ATHOSPHEAE NITAIDED THEN CADUND TO RDlDVI: CDHPOUND lAYER 50 AntDSPHERE NITRIDED --- 40 QUENCIlED AND mtPERED AT 1010 r (16' C) MlHIER OF CYCLES 10 7 S-N curves for aluminum-bearing nitriding steel (Nitralloy 135)...140 4-58.. ... ...... ... Riopelle. "Short Cycle Atmosphere Nitriding. Source: J.... ...

P 346 . Howes. American Society for Metals.- .'~ " " ----. Maurice A.---.B') WEWDE't>+ST~~~!l <..R Me OR 1:... "Review of Gear Testing Methods.. R.UE R:.VM)f ~TI r." in Source Book on Gear Design.. i ~eL..\:)ED ( E. 1980. IO~ '\..~~ \~EL. H.... Metals Park OH. _ ~-- 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. AMS 6475: Effects of Welding 141 120 _110 \ "". ':I... 100 If) V) (/) 90 lJJ ~ ~ " '.OR~ .4-59. I --+ -.. C e. 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.. Technology and Performance. Ed. " AMS b 4-75 (CE. Bowen... "-. Source: Charles W.CSMOO Ir-H i)Pec. Moore rotating specimen test. . i 80 i"o .

J.1 1.."....U.lllL----'--l-L. Peter F. 1Cr-Mo-V Forging: Effect of Cycling Frequency 600 500 400 300 'E... P 15-55 . Source: Steel Castings Handbook.. 5th Edition.JL..L1.""""" a:: ::> --l400 g 300 (f) ~ w 200 --l t> U 100 L-.L.---r-. Weiser... Rocky River OH..WJ ~:] ~': .--rTTTTrr-.L.-n'TTTT"--'--'-rTT'1~---...L..75 % w 500 . Medium-Carbon.142 4-60.2 10..-.L.Lll...0 100... no dwell period.0 FREQUENCY .I. Ed...'! 0...0 10.CYCLES PER MINUTE Influence of cycling frequency on the fatigue properties of forged lCr-Mo-V steel at 1049 OF (566°C).J..---'---'L..-..L.-.-. Steel Founders' Society of America..L..J...L.:. 1980...U.~:~ ::~ ::::' : : : :J 10.L--L..

American Society for Metals.. EM12 Steel: Effect of Temperature on Low-Cycle Fatigue 143 I II- 1= 1lef"rmal:ion ""'90li'T (%) III- -~ 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 I I I r I 10 10~ Low-cycle fatigue ofEM12 at 20 and 550°C (68 and 1020 OF). p 114 . Ed. "Properties of EM 12. the effect of hold time in compression is slightly detrimental to fatigue life. Ashok K." in Ferritic Steels for HighTemperature Applications.4-61. Jean-Roger Donati. Khare. Felix Pellicani and Michel Weisz. As holds true for other ferritic steels. Source: Philippe Berge. 1983.

z W 0:: ::J -6-----6_ ~ 0..J >U I 1.. ..0 I 10 DWELL PERIOD .5 % STRAIN -6----_6 _ -l ~ 100 f0- --..5-hour dwell was added to each cycle... Source: Steel Castings Handbook..5Cr-Mo-V steel was tested at 1022 ° F (550 0C).' « ~ o o "J -l 1..JL-_-'-_-'---.... 1980. up to 10 hours..15-57 .A-= g (f) ~ 1..I.. Peter F.l...l-..-r-r-r-rr-r-r-r-r-r-r-r-r--r 6 REVERSE BENDING MATERIAL A MATERIAL B MATERIAL C PUSH ...J.l. when a D. Steel Founders' Society of America. . Weiser.JL.. . 5th Edition. Ed. have relatively little additional effect beyond that induced by the D.Temperature Testing ~ 3... a 20% drop in fatigue life in reverse bending resulted when a D. The lower diagram shows that extended dwell periods..r ..144 4-62.LLJ.CONTI NUOUS CYCLE t:l Z <l 0:: LL -l <l I 0.. Cast O..PULL Z +......1 L-_----'-_.h Effect of dwell periods on fatigue characteristics of low-alloy cast steel......... .5-hour dwell.5 % STRAIN W U ...r-tr-'r-r-r-t-t-t-r-r-r-r-r-r-r-r-r-. Rocky River OH....5Cr-Mo-V Steel: Effects of Dwell Time in Elevated..LJ--:-_-' 10 2 10 3 CYCLES TO FAILURE Nf 'f-- ___ I .... .L-..-L-L.0 g LL W ~ ° MATERIAL A ....0 .. pp 15-56 . As the upper diagram shows...

... steam. p 15-55 .L-'--...L.• Steel Founders' Society of America. 5th Edition... Weiser...---'---'--'--'--'-'. and vacuum (no dwell period). Peter F. Rocky River OR 1980...:---'-----'-.J'--'--'c..L. Cast 0..5Cr-Mo-V Steel: Effect of Environment at 550°C (1022 OF) 145 . 10 CYCLES TO FAILURE .L.L'--.L.....N Fatigue endurance behavior of cast 0.. Source: Steel Castings Handbook... IN AIR + IN STEAM x IN VACUUM PUSH ..4-63.L. -l 1 0 ~ REVERSEO BEND TESTS 0.JL..L.5Cr-Mo-V steel at 1022 of (550°C) in air.. Ed .PULL TESTS IN AIR ° ~ o o Z « W a:: o I '-::2--'-----'-.' z "e o " I ~ ~ u.

--. Steel Founders' Society of America.5Mosteel at various strain amplitudes.-----r---.....' -_ _. 5th Edition.0 % CONTINUOUS CYCLE 30 MIN.L-_ _--'-_ _.OF Influence of temperature and dwell period on the cyclic endurance of C-O. 103 If) 0.146 4-64.----.. Ed..0 % 1. 1980 pIS-55 .' 400 600 800 1000 200 1200 TEMPERATURE .5 % W . Weiser.-----..6 % 0.5 % 0. L -_ _.7 % 1.5Mo Steel: Effect of Temperature and Dwell Period on Cyclic Endurance at Various Strain Amplitudes TEMPERATURE 10 4 100 200 300 0 c c:---.--.' . Rocky River OH. Source: Steel Castings Handbook..' -_ _. Cast C-O... DWELL TIME 10 '--_ _..J U U I ~ W U Z 0:: <t 0 => w 10 2 z 1. Peter F..

.. Source: N." in Wear and Fracture Prevention.. Lazaridis and S.-:- _ 105 REVERSALS TO FAILURE..5-1.L. HI-FORM 50 Steel vs 1006 o HI-FORM 50 147 ~ 0. P 214 . Metals Park OR...0.001 '-u 3 10 ... U ---'10 4 . Bhat.. P... t. "Fatigue Behavior of Cold Rolled Dual Phase Steels.1006 VI .. ::::.010 ::l I STRAINED lAND AGED .. 2Nf Strain-life data for AISI 1006 and HI-FORM 50 (a columbium-bearing HSLA steel) in the strained-and-aged condition. ~ <I: Z <I: u >. ::E ... American Society for Metals. 1981.

N 10~ Stress response of strained-and-aged AISI 1006 and HIFORM 50 steels..148 5-2. Source: N. which simply reflects the significantly lesser degree of strain hardening of the 1006 and HI-FORM 50 compared with the dual-phase steels.0065 "-"-~C)oO-OC>-_~0.FORM50 o-o-~ gJOO ~ 0.------' 2 3 1 10 10 CYCLE NUMBER..-----..L.-----. 1981. Lazaridis and S...:------L-.I. Bhat. "Fatigue Behavior of Cold Rolled Dual Phase Steels. P 209 . P. American Society for Metals. Metals Park OH. HI-FORM 50 Steel vs 1006: Stress Response ~ ~ ~ ~ 0500L ~400 ~1.::------'--. The imposed constant total strain amplitudes are indicated on the graph.002 1006 STRAINED AND AGED OL." in Wear and Fracture Prevention. The degree of softening of these two steels is less compared with that of dual-phase steels.

and consequently weight.. American Society for Metals. Here it can be seen that all three high-strength steels offer substantial increase in load carrying capacity at the same gauge when compared to the plain low-carbon steel. p 212 . U 0 3= 400 cyo-~ _000-0 DP 1 a~ 300 >Uv.. DP1 and DP2 500 149 o . Lazaridis and S..: ~ II> ~:.L-_--'-_----'L-_.. 1981. "Fatigue Behavior of Cold Rolled Dual Phase Steels.010 CYCLIC STRAIN AMPLITUDE Comparison offour steels: AISIl006... P. .. . HI-FORM 50 Steel Compared With 1006. Bhat..J.008 .. Metals Park OH..5-3.006 .-_.004 .. ~~ 200 100 II> ! o /0 . HI-FORM 50 (a columbium-bearing HSLA steel)..-_--'-_ _ . Source: N.." in Wear and Fracture Prevention. This confirms the potential for gauge. and a carbon-manganese dual-phase HSLA steel (DP2).002 ~HI-FORM50 '-1006 DP 2 « STRAINED-AND-AG ED OL. a lean phosphorus-bearing dual-phase HSLA steel (DPl). reduction that can be realized from the use of higher-strength steels..

.81) -'-0-. experiments were conducted.. To determine whether the foregoing basic test results apply to the frame models.811 IWIK A SPC(1. The three high-strength steel combinations showed virtually the same torsional fatigue strength values as those of the mild steel (1." in HSLA SteelsTechnology & Applications.. fatigue strength differs with the class of high-strength steel but virtually no differences of that nature are seen in the low-stress.01 SPC(O.:.81 --0-. "Criteria of High Strength Steels for Applying to Automobile Frame Components. Source: M. HSLA vs Mild Steel: Torsional Fatigue \.·06-.8 mm) and each ofthe three high-strength steels.IAI'I'C4011.I APFC45It.111 . American Society for Metals. low-cycle range. ~ I A D I spc(o.r.21) SPC : MLD STEEL APFC: IIGH STllEHOTH STEEL : THICKNESS(mm) I -..81) spc(O.. \\.... Melals Park OH. Takahashi...2 mm) combination.150 300 5-4.:. 10' 10' NUMBER OF CYCLES SoN curves showing torsional fatigue of automobile frame steels.. indicating the possibility of gauge reduction..'.~ ~..Ol ISPcIO.. highcycle range. ~"" ~~ ~ 50 ~ -. ~ ~. 1984. P 498 .APFC5O(1. In the high-stress.. 0: 1'-. The above chart presents the torsional fatigue behavior of the frame models fabricated with the mild steel (0. " .• _ --~ .

001 t---t------jt---t---t--~ 0.0004 L.. Source: Metals Handbook.. "tl . 'i! en E '" e Proprietary HSLA 690 MPa (100 ksi] min UTS 0. P 672 . American Society for Metals. Metals Park OH..- _ _L -_ _. Proprietary HSLA Steel vs ASTM A440 0.01 t---=~~f___t--_t--_+----l .5-5.-_~L. 9th Edition.----T""""--T""""----.----. Volume I.-_~L.-----. 1978.03 .e Q.L-_ _____I 102 103 104 105 Cycles to failure Total strain versus fatigue life for ASTM A440 having a yield strength of about 345 MPa (50 ksi) and for a proprietary quenched and tempered HSLA steel having a yield strength of about 750 MPa (110 ksi). 151 0.. Properties and Selection: Irons and Steels.

1 /. 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. JF and KF for Plastic Strain Amplitude vs Reversals to Failure 2. 1984. Comparison of HSLA Steel Grades BE.871(2Nf)-o·8396 R 2= 0. 2Nf 10 7 Plastic strain amplitude vs reversals to failure for Cb (BE). 0. Metals Park OH.968 ~ ::iE -e a: Ien 0.001 10 2 10 3 10 4 10 5 10 6 REVERSALS TO FAILURE. Bhat. Such a regression line is drawn as the solid line in this chart. "Influence of Composition Within a Grade on the Fatigue Properties of HSLA Steels. P 588 . For plastic strain-life relationship the statistical analysis indicates that there are no significant differences between the three steels (F-ratio is not significant). Cb. Source: Shrikant P. 0. American Society for Metals." in HSLA Steels-Technology & Applications.V (JF) and Cb-V-Si (KF) steels.Aj-pl= 1.152 5-6.0 t 'ill- ~N ui ::J e.0 •• 1.Q1 0 :( z· • • BE(Cb-) • JF(Cb-V) • KF(Cb-V-SI) • en i= :5 e. This is further illustrated in the above chart.

. Comparison of HSLA Steel Grades BE. 1984. Bhat. American Society for Metals. JF and KF for Total Strain Amplitude vs Reversals to Failure ' 153 . P 587 . :::l: < I- < a: en I- z . Strain-life behavior: The strain-life curves for the three steels are compared in this graph.. It is clear that when plotted as total strain versus reversals to failure.. Source: Shrikant P." in HSLA SteelsTechnology & Applications. ::::II- o w 2 ::::i e.1'------'----'----'----'-----' 10 2 10 3 10 4 10 5 10 6 REVERSALS TO FAILURE. Cb-V (JF) and Cb-V-Si (KF) steels. Metals Park OH.5-7. 2N f Total strain amplitude vs reversals to failure for Cb (BE).J -e 8 12 0.. "Influence of Composition Within a Grade on the Fatigue Properties of HSLA Steel. the three steels behave similarly and the differences between them are minor.

American Society for Metals....154 5-8.: .. Lazaridis and S.. ::::.... o w".' .0 .: : ..0 DUAL PHASE 1 0 =' .. P 213 .. Source: N. P.1--:10 5 __ REVERSALS TO FAILURE. 16 AS-RECEIVED STRAINED AND AGED HI-FORM SO . Data for HI-FORM 50 (a columbium-bearing HSLA steel) are included for comparison.." in Wear and Fracture Prevention. Metals Park OH. 1 ' .... Comparison of a Dual-Phase HSLA Steel Grade With HI-FORM 50: Total Strain Amplitude vs Reversals to Failure ~ 1. 2Nf Total strain amplitude versus life data for DPI (a lean-phosphorus HSLA steel) in the as-received and strained-and-aged conditions... '" '" u .4 u 103 10 . Bhat. "Fatigue Behavior of Cold Rolled Dual Phase Steels. u « >.- « z :e . 1981..

. Charpentier. P 218 .." in HSLA Steels-Technology & Applications.4 00. the effect of prior deformation was stronger for 50 XF than for 1006 in that both the decrease in life at large strain amplitudes and the increase in life at small strain amplitudes were greater in 50 XF than in 1006. 2Nf 8 '? 0 .r 00. some specific differences were apparent. 1984. 2Nf Total strain amplitude versus reversals to failure for AISI 50 XF HSLA steel.6 w· ~ I- ::::i e.5-9. I( Eeff A 0. AISI 50 XF Steel: Effects of Cold Deformation 8 155 '? 0 . Upper chart: after balanced biaxial stretching.J e.r 6 00 0.57 ~ s 0 . Source: John M. American Society for Metals.. Metals Park OH..20 0.. lower chart: after cold rolling. for example...2 ~ s 0 . z ~ 4 CIl Il- ct a: . "Effect of Cold Forming on the Strain-Controlled Fatigue Properties of HSLA Steel Sheets..30 6 D w· ~ !: . Holt and Philippe L. Although the effects of prior deformation by BBS or CR on the strain-life behavior of 50 XF were generally similar to those in 1006.J z I- <l: 2 0 103 104 105 106 o IREVERSALS TO FAILURE. <l: ~ 4 ct lii a: ... I( IJo tJ> A D Eeff 0.J <l: 2 e I103 104 105 106 REVERSALS TO FAILURE.

Holtand Philippe L.08 00." in HSLA Steels-Technology & Applications. Biaxial Stretching ~ s c -r 6 w· ::l I- :::i a- « In III: z ~ 4 Solid Symbols-D8ta for Specimens Tr8nsverse to Hot Rolling Direction b .06 a 0. Source: John M. 1984.. the fatigue life appeared to be unaffected by the mode of deformation and the specimen orientation. the fatigue life appeared to remain unchanged or to increase very slightly as a result of deformation.J ~ Runoutl 2 REVERSALS TO FAILURE.156 5-10. p 218 . Metals Park OH. In this steel. at least for the effective strain levels investigated (see graph). 2Nf Strain-life curves after deformation for AISI 80 DF HSLA steel. "Effect of Cold Forming on the Strain-Controlled Fatigue Properties ofHSLA Steel Sheets.0 '90. American Society for Metals.16 1 Uniaxi81 T8nsion 881. Also. Charpentier. AISI 80 OF Steel: Effects of Cold Deformation b 8 po • 'lI Effective Str8in 8nd Mode of Deformation )( 4 60.

1 ~ t 0." in HSLA Steels-Technology & Applications. ::l!i 4( 4( ~ ::::i 0. Metals Park OH.1 • TOTAL ..01 • 10 3 10 4 10 5 10 6 REVERSALS TO FAILURE..VERSALS TO FAILURE. ~ 0. Cb.001'----'----'----'----L..001 '-----''---------'-----'-----'-----' 10 4 10 5 10 2 10 3 10 6 REVERSALS TO FAILURE. Source: Shrikant P.1 E ~ 4( :.J l!f a: 0. Bhat. 2Nf Strain-life curves for the Cb. 2 N f Strain-life curves for the Cb steel.5-11.:. Cb-Vand Cb-V-Si: Strain Life From Constant Amplitude 2. 0.---' 10 2 10 3 10 4 105 10 6 RI.: Z . 10 7 0.01 :..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. American Society for Metals.. Comparison of Three HSLA Steel Grades.. "Influence of Composition Within a Grade on the Fatigue Properties of HSLA Steels.: en I- z a: 0..-----------------.-----------------. ::l!i 4( I- :::> c w en I- a: 0. P 583 . 2 N f -Strain-life curves for the Cb-V steel. 2.V-Si steel.PLASTIC AELASTIC :::> ::::i 0.01 lii .. 157 w c I- ~ t 0. 1984.

P.... 009--o_-.." in Wear and Fracture Prevention.. ..... Lazaridis and S. ...=... Metals Park OR.::J" 1 0' 2 : c I °0.... "Fatigue Behavior of Cold Rolled Dual Phase Steels.. p 209 .:-----0.. ~ v ~200 0--C ~_ "'. Source: N. Comparison of Stress Responses: DP1 vs DP2 Dual-Phase HSLA Steels 600 -----..... :E ::: . 1981..0065 ~boo ..10' '-::- '-::- ---'''--:- --1 Comparison of stress response of strained-and-aged DPl (a lean phosphorus HSLA steel) with that ofDP2 (a carbon-manganese HSLA steel) for the total strain amplitudes indicated....... ...:=:..0025 STRAINED AND AGED 0'-::10° '-... American Society for Metals. ~ .0065 -......003 --. ~ :-... Bhat..158 5-12. -~ . -_11 ~ : 5 0 1 _ '==. 0.>-""""_ _ ~-0.

L. Bhat." in Wear and Fracture Prevention. American Society for Metals. v 6 u ~ 400'r-o-OCo--o-----...002 0.- .. '" ~ o ::::.5-13...:-- ..-::- --'-:- --' 10 2 10 3 CYCLE NUMBER...0.o-O. Dual-Phase HSLA Steel Grade: Stress Response for As-Received vs Water-Quenched 159 ..L.. 1981. P. Lazaridis and S......005 ~300 v v200 >- Ol. P 208 ......005.-"d""--ov-.--<-->--e. Source: N. N Stress response of a water-quenched dual-phase steel in the as-received condition for total strain amplitudes of 0.I . "Fatigue Behavior of Cold Rolled Dual Phase Steels...002 and 0. Metals Park OH.

o . Lazaridis and S.... P 208 . _ _ 0.-:.160 5-14.006. "Fatigue Behavior of Cold Rolled Dual Phase Steels. > e C Cl00 OLI ----''-. Bhat.-:... N Stress response of the gas-jet-cooled dual-phase steel in the as-received condition for total strain amplitudes of 0.o-_ _. 1981.." in Wear and Fracture Prevention... American Society for Metals. Dual-Phase HSLA Steel Grade: Stress Response for As-Received vs Gas-Jet-Cooled .0025 and 0...... Source: N._0..- --l.500 :e Go E oo .. P.- --l. Metals Park OH.- ----'-:. .._-"''-oA. '" ~-----o....- I~ .0025 >- ~hoo ._..- .-o''-_..- ---' CYCLE NUMBER...006 = ~ ~ v v 3 0 0 r ._ -.--o_o(a o .

.

Metals Park OH.:---------"---=. Bhat.... 1981.J 10 5 REVERSALS TO FAILURE. Source: N. "Fatigue Behavior of Cold Rolled Dual Phase Steels.010 DUAL PHASE 2 0 1'" AS-RECEIVED STRAINED AND AGED HI-FORM SO .--- ....162 5-16.. Lazaridis and S.- 0.. American Society for Metals.L. Comparison of Dual-Phase HSLA Steel DP2 With HI-FORM 50 0.-10 4 10 3 ." in Wear and Fracture Prevention.001L. 2Nf Strain-life curves for DP2 (a carbon-manganese HSLA steel) in two conditions compared with HI-FORM 50 (a columbium-bearing HSLA steel).. P. p 214 ..

.

Metals Park OH. g' i I: 1/ / E '.5 10.5 at room temperature to 20.~ I.6 18 22 26 30 36 6K. (Pl b . =lOfLm R =0. The effect of temperature is seen to decrease the crack propagation rate with decreasing temperature at low values of 6. a crossover occurs wherein higher growth rates were observed. Gerberich and James P. / Q) HSLA. ! .164 5-18. William W.6at room temperature to 10. where it ranged from 3.2 G. n. = 10JLrn R=O.8 14 n 7. Lucas. as the stress intensity increases.6 8. I. I Prof6 . . 01: I/. ./ . 1/ t>A • '/: 10 .1 o 300K t> 233K 1//. "Near-Threshold Behavior of HSLA Steels. P. This behavior has also been seen in iron binary alloys where n increased from 3. 01 ' I / t>. 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). This crossover is further reflected in the increase in the Paris law exponent. Source: Khlefa A.f/:/ '/ .S.S. Esaklul. American Society for Metals.1 o 300K t> 233K Cl 173K o 123K Q) u o "E >.: t>. p 569 ." in HSLA Steels-Technology & Applications. MPa -m 1/2 The effect of test temperature on the fatigue crack propagation rates in the Paris law regime for two HSLA steels in the as-received condition.K.1 6. o 0123K~I1!I: . However.9 at l23K. as shown in the above charts..6 5. I I <. Fatigue Crack Propagation Rate: Effect of Temperature for Two HSLA Steel Grades HSLA-1 G.pI .8at l23K for HSLA-l.1 15. The large increase is a result ofthe change in the fracture mechanism from ductile transgranular fracture to cleavage. /?~ I 0. ' Pi T(K) 300 233 173 123 T(K) n 300 233 173 123 3.8 12. o~ . f o 173K o >. /. 1984.

06 0.014 0.35 • R-0.01 0.IO 10 • R-O.35 0.IO 10' 10 R- R-0. MPa-m l12 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. William W. Effect of R-Ratio and Test Temperature on Crack Propagation of H SLA Steel Grade 1 165 HSLA-1 G. 0.005 0. Compositions of HSLA-l and HSLA-2 Alloy C Mn Nb SI P S Al Ni Cr Fe Rem Rem HSLA-1 HSLA-2 . . Lucas. p 571 .005 0. 1984. MPa-m" 12 14 It III 20 2 24 4 II II 7 II 910 1214161920 24 6K.07 0.35 • R-0.51 0." in HSLA Steels-Technology & Applications.03 <0. Gerberich and James P. "Near-Threshold Behavior of HSLA Steels..01 0. 10pm T -123K • R-O.01 0.70 z ~ "0 c lO -. 10pm T-300K HSLA-1 G.01 0.01 Source: Khlefa A. Esaklul.03 0.70 0.03 0. Metals Park OH. e. American Society for Metals.S.S.5-19. 4 II II 7 II 9 10 6K.

10JLm R' 0. Furthermore. WilliarnW.173 K . Esaklul. where HSLA-2 showed lower crack propagation rates and higher threshold stress intensities than HSLA-I..m112 Fatigue crack propagation behavior of two HSLA steels tested at temperatures of 300. MPa .. Effect of Test Temperature on Fatigue Crack Propagation Behavior for Two HSLA Steel Grades HSLA-1 G." in HSLA Steels-Technology & Applications.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. Lucas.0 MPa-m 1/ 2) compared to HSLA-I (5. u II) ~ o o 300K 233K 173K 123K E 4 1I 6 7 8 9 10 12 14 16 18 20 24 4 5 6 7 8 9 10 12 14 16 1820 24 . The stress intensities amplitude. This difference is also reflected at low temperatures.233 K o .5 MPa-m 1/2). 1984.0-2. 11K. Source: Khlefa A. "Near-Threshold Behavior of HSLA Steels.. 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.233.5 MPa-m 1/2 higher in HSLA-2 than in HSLA-l. The 2. 173 and 123K in the as-received condition. IOlJm R-O.. for constant growth rates of IO.0-2. 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. Gerberich and James P.1 e300K ..166 5-20.6K.. Metals Park OH.s and 10-9 ta] cycle are 2.123 K u >. The threshold stress intensityl1K'h' is also higher for HSLA-2 (8.S. see p 165).h (D t 11K. 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. 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.S. MPa-m"2 6K. P 569 . (300K) are the same for both steels. American Society for Metals.I HSLA-2 G.

q ~ $~ .=2C~ on '<: 10: .....:. 1984...100 .."" F--". Richter..::. T~ '00 • )00 i~! "" E I~ .ES ... ..1--.... I~. 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. "Cold and Hot-Rolled Microalloyed Steel Sheets in Opel Cars-Experience and Applications...3 ." c-c 0= .:.. Source: Klaus E.~ I I i CYCLES " Stress-cycles curves of welded samples of different materials under tension load.E."in HSLA Steels-Technology & Applications.~E lotAlERrAl I~ . WATEAW. I~% .. <. OStEJOOn-l 5152.).wac... Metals Park OH. tiS '·0. ~ ~--.' .. by the heat influence during the welding operation.3 (according to DIN). Stress-Cycle Curves for Weldments of Different HSLA Steel Grades lHlER OYIW.. which is responsible for increased strength.00NXT10N9 .. . S"mESS . . ".~ ~~ ~ """ ".~ (00 ~~~ IoU5 I '.-"---dEl .~:~.LMTS OF WB. would be more favorable in this respect..'. the strength of which results from the chemical composition. I-NJIIB[A ~ .flI..0- --"'''''': ..B£R ~ I CYCLES Stress-cycles curves of welded samples of different materials under alternating bending load. such as St 52.--. " ~.DED SAt. .. .ES lHlER OYIW..... Fatigue data were derived from testing welded samples under tension and bending loads.". OSIE 3eON 5' 52. :. and it is thought that a higher-strength manganese-alloyed steel..E"'B '-PO -.DED SAt. ~ .-1 . An explanation for this is the loss of the thermal-mechanical effect.. .- sc. JO~ .LMTS OF WB. -.po'..~ ~ ......M"A.f=:-~ ~-.IC LOAD .::"" ". -'ed'" .J I "".cotI:XTJ:)N:S oOJtrElDl!1 167 S"mESS ._. USIW 22J.IC lCW> .. American Society for Metals..('.5-21.".. 00 . " ~ USIW 22->€. P 487 . ""'-'~"---dE]r ~ "" ~~ ~ =Jl 'II'EI. It was surprising that under both types of load the HSLA steel and the soft unalloyed steel hardly differed in fatigue strength.•.....flI.

Houchens. after welding.1---. 1984.. Lawrence.5 ksi) and 117 MPa (17 ksi).. toO.168 5-22..1." in HSLA Steels-Technology & Applications. and Albert F..13" 13...SAE·1006.-. FCAW .. Source: Kon-Mei W.. FCAW .. Jr.----. Smooth -'-0-'. However. The I06-cycle fatigue limit stresses for FCAW SAE 1006 and SAE 980 X steels were between 114 MPa (16..SAE-980X. .. Ewing. American Society for Metals.. SAE 1006 and SAE 980 X steels exhibited similar fatigue properties over the 104-1 06-cyclelife range studied. -. The fatigue strengths of the smooth HSLA steel were higher than that of the low-carbon steel.1.__ -----6- • -'"l:>"'&--C'IS"" 6-"tl.SAE-980X. Frederick V... Smooth . P 556 .3mml SAE-1006.. Pei-Chung Wang.. "Weld Fatigue of TIGDressed SAE-980X HSLA 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). Weldments (FCAW): SAE 980 X Steel vs 1006 --6-- R=0.. ~ _ 10 5 NT' CYCLES TO FAILURE 106 Fatigue properties of smooth and FCA W SAE 1006 and SAE 980 X steels.. ----. Metals Park OH.

grinding or TI G-treatment can be used to improve the weld geometry. For making full use of an increased static strength for a steel subjected to severe fatigue... . special attention must be paid to the configuration of the welds.5-23. The notch effect at the weld toe is decreased and the fatigue properties can be improved. Metals Park OH.. P 259 . "Formable Hot-Rolled Steel With Increased Strength. For welded joints the fatigue strength in the endurance range 105-2 X 106 is mainly dependent upon the weld geometry and is therefore roughly the same irrespective of the static strength of the steels. After welding. For unwelded parent metal the fatigue strength of a steel is improved with increasing static strength. Sheet thickness 5 mm and ultimate tensile strength 767 MPa..00 TlG·treUed fillet weld Untreated butt weld Untreated fillet 100 50 '-- . Standard-Wohlerdiagram (log-log scale) with pulsating load (R=min stress/max stress=O)..L- . 1984.-.. American Society for Metals. 105 CYClES TO FAILURE 106 Fatigue strength for DOMEX 640 XP.L- -'- .L.. Source: Tony Nilsson. Another solution is to place the welds in areas where the stresses are low." in HSLA Steels->Technology & Applications. Weldments (TIG): DOMEX 640 XP Steel Welded Joints vs Parent Metal 169 1000 000 "00 100 600 r---~ __ Parent IIletal TIC-treated bull weld sou 400 ! JOO ~ '..

American Society for Metals. P 563 .1 EXPERIMENT .52 I ~ . <I ~ 10 2 en • I 3 10 101 105 NT..PREDICTION . Houchens..170 5-24.•••. 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. Lawrence. Weldments (FCAW Dressed by TIG): Fatigue Life Estimates Compared With Experimental Data for SAE 980 X Steel 10 CJ) ~ 2 103 --.:::::::::::: 10 1 Kf = 2.e'!: • <I en TIG DRESSED SAE 980 LAP-SHEAR WELDS R = 0." in HSLA Steels-Technology & Applications. Jr. The other beneficial effects such as removal of slag intrusions and inclusions were not considered. (U r =-87KSI .. :E . "Weld Fatigue of T1GDressed SAE-980X HSLA Steel. and Albert F. 1984. Source: Kon-Mei W. Frederick V. Metals Park OH. 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 smaller flank angle contributed significantly to the increased fatigue strengths of TIG-dressed weldments.'=ur = 87 KSI J!. Ewing. 106 108 CYCLES Total fatigue life estimates compared to the experimental data for the FCA WITIG-dressed SAE 980 X steel. -. Pei-Chung Wang.

II __ ~ -----ZP-n---....It.-._ o '- ''''"00'_ . and Albert F. Lawrence..-""."'"b- 0 '[jCD-. Pei-Chung Wang..As-Welded --06--...-(}-.. P 558 .3mml ---0--- TIG-Dressed _.~ ~00 0 o o 0 -0.I\ 0 ~o 0 o '-'il." in HSLA Steels-Technology & Applications. t= 0. . Metals Park OH. American Society for Metals..5-25. a significant improvement in fatigue characteristics can be obtained by TlG-dressing the welds..13" 13. "Weld Fatigue of TIGDressed SAE-980X HSLA Steel. Ewing..._. SAE 980 X Steel Weldment (FCAW): Smooth Specimen vs TIG-Dressed vs As-Welded 171 SAE-980X R= 0. 00 0 -c. Houchens. o Jgt.1. Jr. 1984. From these data. '1I'o-.0 0 o'-'n- 10 5 10 6 NT' CYCLES TO FAILURE Fatigue properties of FCA WITI G-dressed SAE 980 X steel compared to the smooth specimen and as-welded data. Frederick V. Source: Kon-Mei W.Smooth Specimen ~-------4.

Metals Park OH. Frederick V. 1984.. "Weld Fatigue of TIGDressed SAE-980X HSLA Steel."in HSLA Steels-Technology and Applications. Jr. Houchens. SAE 980 X lap-shear joints. and Albert F.172 5-26. American Society for Metals. Source: Kon-Mei W. SAE 980 X Steel Weldment (FCAW): Lap-Shear Joints <] en 10 1 SAE980 LAP-SHEAR WELDS Kfmax =3. Lawrence. Ewing. CYCLES Total fatigue life predictions and experimental results for the FCAW.49 I R = 0. p 562 . Pei-Chung Wang.1 • EXPERIMENT PREDICTION 105 106 NT.

"= -II -~ ". it seemed necessary to examine the dynamic strength of fusion welded joints of HSLA steels in more detail. v."in HSLA Steels-s-Technology & Applications... OSIE380T'4 ...: I~: '~ I ~ .. - . '00 ~~: .. Tensile load and alternating bending load were the selected types for dynamic test.. n .-.5-27.~"".: ~ ."'£. Microalloyed HSLA Steels: Properties of Fusion Welds 173 STRESS . ~% '-~ USIW 22."E.f'lES )IQ. -..... '-.l1.LMTS . in comparison to a soft unalloyed. The above charts show the respective stress-cycle curves. "..' I I I lA.-. . I I~"" '0(>0'· 101" oi~hO£l .S lHlER OYNAlolC Lp"o - COIClIJ1DN8 ~~~-"£'S~~n IUTERlAl I~% '. "Cold and Hot-Rolled Microalloyed Steel Sheets in Opel Cars-Experience and Applications. ".. ~~ . Richter.-" ~~ . STRESS ... ~ "-'-dEl 10 a: WELDED SN.I. American Society for Metals........11' ' 10&1 0'" ".. .'.. '--I I 00 . P 487 . Due to the preferred crack location in the welded areas.5 lHlER OYNAlolC Lp"o - COIClIJ1DN8 w(lO~O'"" . .LMTS ".._':'0 .. "-"-dEi 10 a: WELDED SN... Source: Klaus E. - SI S23 '-=" '" f".co '00 '00 . -II c« q -..Orh(i· BfK>'Mj·Y«=ST 1. -..- l:SIW n.~ :.C..C'l>~ """" ~ ~--'= =---= ~~ I --~ r-''!!::='.. I~ " "00'.:. testing the steel used for the crossmember at a minimum yield strength of 380 N /mm 2.... I I III I fUo'8(A OF CVCl£S- Stress-cycles curves of fusion welded samples of different materials under alternating bending load.!~ 'l.IM8ER(JFC'fC~t'S_ Stress-cycles curves of fusion welded samples of different materials under tension load..~' C' lQl.:.. Metals Park OH... hot rolled steel sheet..........l[R'."~'" 051E 38CN >l "3 iec -= lC~ nn :.TC'SS'O\·T~ST ·....... 1984.. '00 ..~ ~~ ::...f'lES )10.• ':. o~· -.

174 5-28. kips AND NUGGET ROTATION 68 N • degree 5. I1P. LOAD AMPLITUDE 6P.0 0.0 0.0 1.0 2.5 GroupC Fatigue test results for the 2.0 0. Straight lines were fitted through the data.5 5. kN Group K 4 5 6 7 LOG CYCLES TO FAILURE. Fatigue test results for 1.2 1. .081 inch) thick sheet with various weld diameters.06 mm (0.0 1. Results of spot-weld fatigue tests are presented in the four plots (above and on the facing page) for the stated conditions.5 LOAD AMPLITUDE 6P.0 2.040 inch) thick sheet of different stiffnesses. 116 N' for each test as a function of cyclic life. Each curve shows the load amplitude. degree 5. and nugget rotation values. N. Microalloyed HSLA Steels: Properties of Spot Welds LOAD AMPLITUDE 6P. kips AND NUGGET ROTATION 68 N .02 mm (0.0 0.

0 2. l01611'WT1140n.040 inch) thick sheet with single and multiple welds.0 1.0 o ~::S'.." in HSLA Steels-Technology & Applications..0 LOAD AMPLITUDE dP. ..2 1.0 1.. Nf Fatigue test results for 1. 66 .- o - "W-IOI61M11~O"1 ' 02 _ " " .0 0.5 ::~' GroupN ~ ~ -.roOlI" I LOAD AMPLITUDE dP..1 0211Yn1004.0 0.0 de dP 1...0 1.0 0.5 5.0 0. 0 . 1984.1·~oe""". dP de 4 5 6 7 LOG CYCLES TO FAILURE.0 0.5 1. Davidson. American Society for Metals.0 2.0 2.0 2. Source: James A. degree 2.0 0.5 Group E 5..02 mm (0.5 5. Metals Park OH. ..0 0.0416 . kN 2.5 Group 0 5.. "Design-Related Methodology to Determine the Fatigue Life and Related Failure Mode of SpotWelded Sheet Steels. kips AND NUGGET ROTATION d(~N' degree 5.0 1...0 0. ' 0-66In"lf02fj1fl1 5. p 542 .. 5.. W .=I~~:" 175 LOAD AMPLITUDE dP.0 1.5 d0 dP Fatigue test results for variations in specimen width and thickness... kips AND NUGGET ROTATION :>SN.

NjXI03 Cycles to fatigue-crack initiation versus nominal stress amplitude. 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. The lower graph shows I:1K/ vp plotted versus N. As expected. A narrow-spread family of curves results. is plotted versus l:1a.5mm 6..!m side notch. Fine and R. NI xI0 3 Same data as in upper graph but plotted versus Ill( /. where N.40 0. E. gave the most rapid initiation at a given stress. the cycles to fatigue-crack initiation. in HY-130 steel for notches of constant depth but various radii of curvature.. "Fatigue-Crack Initiation and Near-Threshold Crack Growth. HY-130 Steel: Effect of Notch Radii ~ ::. is defined as the number of cycles to give a 50-J. Source: M. Barsom and McNicol used this parameter to compare N j . for notched specimens with various radii of curvature. ~ :::!: 4000 3000 2000 1000 600 500 400 300 200 2 3 4 5 6 8 10 2 3 4 56 8 I 2 3 4 56 8 I~ ~ 800 " ~ <l p=0. 1979. ~a.4 3.JP rather than ~a. Metals Park OR. O.176 6-1.20mm CYCLES TO FATIGUE-CRACK-INITIATION.. the sharpest notch. The results are shown in the above graphs. Ritchie. lowest p.2 0: ~ tii 1." in Fatigue and Microstructure.6 ~ z « z 0. There is a wide spread in the curves.80 0.20 2 3 4 5 6 8 10 3 CYCLES TOFATIGUE-CRACK-INITIATION. 800 <l 600 b IQ. In the upper graph. American Society for Metals. a. g 400 ~ 200 100 80 ~ 60 40 20 2 3 4 5 6 8 10 2 3 4 5 6 8 10 2 w· :J 9. N. pp 256-257 . Curvature ofthe notch and I:1Kis the stress-intensity amplitude computed for an imaginary crack whose length is the same as the notch depth.

300 M Steel: Effect of Notch Severity on Constant-Lifetime Behavior Minimum stress..~ ~ E :> ~..L..".. ::. 1978..------'25~0 -1500 -1000 --500 2000 1000 1500 Constant-lifetime fatigue diagram for 300 M alloy steel. Source: Metals Handbook..::-----. Metals Park OH. 9th Edition..: 1500 ~ 200 1.ksi 177 300 s: ::.. Solid lines represent lifetimes obtained from unnotched specimens..------._-_--L....6-2. Properties and Selection: Irons and Steels.L._-----. Volume I. hardened and tempered to a tensile strength of 1930 MPa (280 ksi). Dashed lines represent lifetime of ten million cycles for specimens having the indicated notch severity._------"L-_-_.. American Society for Metals. P 670 ..L-_-_--L. .: 100 Notch 500 severity (107 cycle IifeUmel _~00'::0:----...~ E E :> E ::.: 1000 .

1984.2' .. ~ i 800 . which may be particularly effective in the fatigue-threshold regime.1. and also exerts a beneficial influence in high-strength TRIP steels. In lower-strength austenites. This growth retardation may be due to crack-closure effects arising from the transformation volume change. transformation is detrimental to low-cycle fatigue life. Under stress control." in Deformation. as illustrated by comparison with other high-strength steels in the above graph. the fatigue life of lower-strength austenites is greatly enhanced by transformation. also revealed transfermation enhancement of fatigue life.178 6-3.. Ed. but a small amount of transformation may be beneficial at high cycles. p 419 . in which thermodynamic stability was varied by heat treatment.1 1200 1400 1800 2lXXl 2200 1400 TRIP ~175 u ~150 1200 ~ i . Metals Park OH. Olson. Source: G.1) vs ultimate tensile strength for TRIP steels compared with other high-strength steels. :J SR 4340 IlXXl"/. fatigue limits in excess of the yield strength are observed. "Transformation Plasticity and the Stability of Plastic Flow. Smooth-bar fatigue properties appear to be dominated by transformation hardening. 4l«l . which is desirable under stress-control conditions (reducing strain amplitude) but generally undesirable under strain-control conditions (increasing stress amplitude). George Krauss. the lowcycle fatigue properties of high-strength TRIP steels are found to be degraded by transformation under controlled total strain amplitude. Processing. B.. for a stress ratio (R= amin/ a m• x ) of 0. although to a much lesser extent. ~125 v. TRIP Steels Compared With Other High-Strength Grades MN/m2 1600 IlXXl 200 R • 0. particularly at low ~K. and Structure. 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. under controlled total strain amplitude. Investigation of the smooth-bar fatigue properties of high-strength TRIP steels at R= 0.l!loo 600 Fatigue strength at 10' cycles (R = 0. Such enhancement allows the achievement of exceptional fatigue strength at high ultimate strength levels. American Society for Metals. transformation reduces fatigue life under conditions of controlled plastic strain amplitude. Similarly.

IN ACID BRINE WITH INHIBITOR 10' FAILURE 0 10' to' Effect of corrosion and corrosion inhibitors on the SoN curve for highstrength steel (sucker-rod material)..000 179 . Source: Joseph F. American Society for Metals." in Corrosion: Source Book. B ce F AB. OH.000 lL lL ::i 0 . P 380 .000 102 10~ t04 NUMBER OF CYCLES UI UI 50. After the first brittle crack is initiated.000 10. No. Metals Park. lUI Q. Corrosion Fatigue: Special High-Strength Sucker-Rod Material 10. "Corrosion Fatigue Cracking of Oil Well Sucker Rods.. so the fatigue curve drops from AB to CD.000 80. 1984.. Ed.. Coburn. Seymour K. .. Chittum. a: 40.2 is the slow step in the process and electrochemical action is the slowest part of this step. Deceleration of the slow stage with a corrosion inhibitor will raise the S-Nfatigue curve from CD to EF..000 20.6-4..IN AIR C0 . as shown in the graph.IN ACID BRINE EF. Thus. the effect of corrosion can be illustrated with curves of stress vs logarithm-of-number-of-reversals-to-failure for sucker-rod steel.000 30. Corrosion accelerates cracks propagation.

Coburn.180 6-5. American Society of Metals... Source: Joseph F. Observations of sucker-rod crack penetration as a function of reversal accumulation are possible using a bending apparatus and a magnetic fluorescent powder technique.. no penetration is apparent in the first 40-60% of the specimen's fatigue life. Corrosion Fatigue Cracking of Sucker-Rod Material . During bending. PERCENT OF FAILURE Corrosion fatigue cracking of sucker rods. the cracking accelerates until catastrophic failure occurs. Seymour K. A crack eventually appears and progresses through the specimen. I0 :J: ~.6 ~ Ii ct: 0 . When the penetration reaches a certain percentage of the cross section.8 Ul 0 ~. 1984. Penetration vs reversal curves resemble the one shown above when the stress is well in excess ofthe endurance limit. Ed. showing stages in the fatigue process. even though intrusions and extrusions may form earlier. ISTAGEI 0 RELATIVE NUMBER OF CYCLES.4 ~." in Corrosion: Source Book.2 UJ oJ I I I I I I I 0 FIRST STAGE SECOND STAGE I I !THIRD. Metals Park OH. "Corrosion Fatigue Cracking of Oil Well Sucker Rods. P 378 . Chittum. This graph shows typical progress of a crack at high stress plotted against number of cycles.

"The Role of Hydrogen and Other Interstitials in the Mechanical BehaviorofMetals. Cedric D. static fatigue curves. ~. This behavior is sensitive to hydrogen concentration as shown above.1 \ • <- 1- 1000 1 10 Fracture Time. These are.01 -.000 psi. a stress below which failure will not occur for an indefinite period of time."in Hydrogen Damage Source Book.~ 0 0 0 "\\'\ -\ " 0\ \ 1 Bake --- 24 h r - Bake 18 hr 200 175 150 125 100 ( /) "0 Q) .6-6. . ~ i---:::::::::::: o~ -. Troiano.. (f) (f) Q) \ \ . in essence. In an unnotched specimen. and static fatigue limit increase with decreasing hydrogen concentration.5 hr 1 <t \ . Sharp-notch specimens. Hydrogenated Steel: Effect of Baking Time on Hydrogen Concentration Normal Notch Strength = 300. 181 300 275 250 a.000 psi strength level... ~ i'.. and the lower critical stress may be considered a static endurance limit-that is. 230.. yet delayed failure will occur after 24 hours or longer of baking time at 150 °C (300 OF). Beachem. + + ~~ :---:- + Uncharged +-+- \ <. even after 24 hours at 150°C (300 OF). there is still a substantial stress range.- Bake 12 hr Bake 7 hr Bake 3 hr Bake 0. rupture time. over which delayed failure will occur. 1977. Source: Alexander R.. a. 225 (f) o. a. Ed. Metals Park OH.1 l \ -\ ~ ~ 75 50 0. Hours 100 Static fatigue curves for various hydrogen concentrations obtained by baking different times at 150°C (300 OF).. where it may be seen that all delayed-failure parameters-notch strength. full recovery ofthe ductility as measured by the reduction of area can be attained in less than 20 hours at 150°C (300 OF)... P 154 .000 psi i'-. 0. American Society for Metals. of the order of 60. Also.

Hydrogenated Steel: Effect of Notch Sharpness T~. Baked 0..001 in .25 in. 1977. a 225 200 175 150 125 100 I~~ 1\ '\ 0 0 0 ~\ -a_" ""' -\. Source: Alexander R.- >-- \ 75 50 0.020 in. ... >-- >-- ~\\ Notch Radiu's = 0.182 300 275 250 en 0.5 hour at 150°C (300 oF).. \ Radiu~ = 0. 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). 6-7. a1lus = 2'In.. Notch Radius = 0. J..1 Static fatigue curves for specimens of different notch sharpness.01 .. Notch « 0. 0 01'\. Beachem. Metals Park OH. P 155 ." in Hydrogen Damage Source Book.. American Society for Metals. • I Notch Radius = 0. en en i'\ "0 \ \\ . "The Role of Hydrogen and Other Interstitials in the Mechanical Behavior of Metals.. Cedric D.. ~. Troiano.10 100 1000 Fracture Time.010 in.. Hours 0. The variation of lower critical stress with notch severity is shown in this diagram. 0. It-.. Notch ( J) Q) Q) L. Ed.

. H.. p 83 .. ' . O. . o < r l.... Ed. American Society for Metals... . Elihu F......5%Mo steel at 275 and 775 K..." in Source Book on Materials for Elevated.~ :....5%Mo Steel: Effect of Hold Time in Air and Vacuum at Different Temperatures + 10 I 183 - KEY CYCliNG CONTINUOUS llLL < e . :--... ..:--..: ......... __ . .... II AIR VAC .. VI w .. AIR 275 K VAC ....J ....7-1.. Bradley. =:::::..:.' < Z IX: IX: IVI w Z ~ 10 0 .. I::NV. Cook and R... ".. AIR 775 K VAC. loO .... P.... TEMP. . ::--- --....:~:-:-: ..:.. " ... Metals Park OH.Temperature Applications. .~""--. ' ~ 30 on HOLD .""'".. Source: R.. 1979..:-:. CYCLES TO FAILURE Effect of hold time in air and vacuum upon the fatigue endurances of a O.«:»r..'. "=':... "Environment-Dependence of the Mechanical Properties of Metal sal High Temperature.~ ... .. ._~.. Skelton. " ....

25 V): Effect of Liquid Nitriding 1000 '\ 900 . Curves A and C are for hardened and tempered (not nitrided) specimens.. . together with the increased strength of the nitrided layer.. Source: Metals Handbook. 9th Edition. American Society for Metals.=2.80 13 U . 0.184 7.~ ~ . 1978.90 Mo.5 Cr. these residual stresses. >- 500 400 \ ~ \ . Nitriding introduces residual compressive stresses at the surface of steel parts.40 105 106 107 Number of load cycles Effect of nit riding on fatigue behavior of DIN 14 CrMoV 69 steel (0.25 V). P 541 .120 . Band D are for liquid nitrided specimens. 0. The increase in fatigue strength that results from nitriding is illustrated in these S-N curves. Properties and Selection: Irons and Steels.2 DIN 14 Steel (1.14 C. 1.140 :'\. A and B are for smooth specimens.s IS I' \ '.60 C 300 . ""-8 800 700 .5 Cr. Metals Park OH.90 Mo. 0. C and D are for notched specimens K. -'" 0A .100 ~ 600 13 . Volume I. 0. increase the fatigue resistance of the part.

. (Cycle used was approximately up 5 s. hold 5 s..J -.25Cr-1. -~ . Elihu F.=.3 0. Metals Park OH..) Source: R. American Society for Metals.~~~.0 2. 0... Bradley.OMo steel in sodium.0 8. H.25Cr-l. 2...7-3.. • • .I 10~ CYCLES TO FAILURE Influence of cyclicstrain range upon fatigue endurance of 2. P...0Mo Steel: Influence of Cyclic Strain Range on Endurance Limit in Various Environments 10.: VI a: 1.0 6.. "Environment-Dependence of the Mechanical Properties of Metals at High Temperature. 1979..0 KEY • ENVIRONMENT Na (30 ppm OF °21 AIR HELIUM Na \300 ppm OF 02) 185 U 3. Ed... ...0 fu . Skelton..2 .. P 83 . Cook and R." in Source Book on Materials for Elevated-Temperature Applications. down 5 s.. hold 5 s.0 X ~XX z ... C.4 0.6 >u 0.. and helium at 865 K. air. ~ • u 0. .. ~ .8 x~ '~ 11K 'U .

OMo steel at 425. Metals Park OH.25Cr-l. P 659 .186 7-4. 1978.1 2 10 103 Cycles to failure The results of strain-controlled fatigue tests of 2. Volume I. 9th Edition. Within this range. 2. OJ' "Q °FI 2 . Properties and Selection: Irons and Steels.25Cr-I. ?f!. Source: Metals Handbook. test temperature had relatively little effect on number of cycles to failure. 1000 and 1100 OF) on specimens of annealed 2. E c 'iii '" J5 0.'= ~ l1 • 540 595 800. American Society for Metals. 1000 1100 a. 540 and 595°C (800.25Cr-1.0Mo Steel: Effect of Elevated Temperature Testing temperature ~C o 425 .OMo steel are presented in these S-N curves.2 0.

25Cr-l. Volume I. Furthermore.03% results in a reduction in fatigue strength.OMo steel. 2. American Society for Metals.OC I I I 600 700 800 900 1000 1100 Temperature. Metals Park OH.25Cr-1.Oc 187 104 :e B ~ e u :G > u 103 350 400 450 500 550 Temperature. p 659 .7-5. because of variations in strain aging effect. Strain-controlled fatigue tests have also shown (note above) that reducing carbon content to 0. 1978.0Mo Steel: Effect of Elevated Temperature and Strain Rate Temperature. 91h Edition. Source: Metals Handbook. Properties and Selection: Irons and Steels. of Effect of elevated temperature on strain-controlled fatigue behavior of annealed 2. specimens from one heat with a higher carbon content ran longer at 427°C (800 OF) than at 316 °C (600 OF).

the stress-intensity factor range increased as the crack length was increased..6 l.f + . 2. MPa vmm Variations in fatigue crack growth rate with test temperature for specimens of 2.OMo steel tested in air.. Properties and Selection: Irons and Steels. Source: Metals Handbook. Metals Park OH.._F____I1____l 1100 10-6 Q> U ~ Q> E 10-4 E :i s: ~ .. cyclic frequency was 400 per minute. e u 6 .. P 660 .4 10-31--+--oC• 205 o 370 6595 of ..Jiil. As shown.188 7-6.. 10 20 60 Specimens were subjected to cyclic loading at a constant maximum load...05..l 400 700 1---+-6-455-850---+--------+--------.25Cr-l..+ . Stress ratio was 0. 10-6 s«. ksi.£ s: U \! ~ ~ 0 l..>< u ~ .~____J<-----.l 1. e ...... Volume I.+ .. stress intensity factor.. stress intensity factor. :e' :!'! "D . 10 20 60 I--+-------+-------+--------+-----i Testing temperature 10... .>< :!'! "D :e' 10...... American Society for Metals. 9th Edition.25Cr-1.0Mo Steel: Effect of Temperature on Fatigue Crack Growth Rate 6K. 1978.

or that may develop cracks in service.7-7. Metals Park OH. Properties and Selection: Irons and Steels. Ibl Data shown above indicate that in elevated temperature tests at a given stress-intensity factor range. p 661 . 9th Edition.rm: 189 2 5 2 I 0-4 4C"7f'/ 5 5 40 CfKO / ! vr:~400C~ / ~ Frequency. Volume I. 2. crack growth rate increases as cyclic frequency is decreased. Source: Metals Handbook. These fracture mechanics data may be applied to the design of structural components that may contain undetected discontinuities. ksi 2 5 vrn: 2 I 2 3 ~ E . cycles/min o 2 I 5 400c:pm 2 I 4 • 40 5 2 2 5 4K..05. stressintensity laflOr. stress intensity factor.. MPaym (. t 1/ Frequency.25Cr-1. (a) Tested at 510 °C (950 OF). Stress ratio was 0. siren intemity laClor. 1978. E 5 4C~ 40c"m III 1-// 5 2 II.r:/ I 5 2 1 • 40 6 2 2 5 ilK.. cvcles/mln o 4 /. (b) tested at 595°C (1100 OF). ksi.1 IJJ. MPa V.0Mo Steel: Effect of Cyclic Frequency on Fatigue Crack Growth Rate b.K. slreu intensily 'actor. American Society ForMetals.K.

. 2 Hz 0.. z· -. Gerberich and A.V ~ . Vi>. P 333 . ~ ~ - . 2. 112 ) because ofthe hydrogen environment. In addition.Jl i>. For example..2SCr-1Mo steel tested in air and in 138-kPa hydrogen gas.190 7-8." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. Materials Selection and Failure Analysis. m 112 (4. i>.5 Hz I } - i>. '" s: i0 .l' °• oOe • .h ({) that there is a large increase in growth rate for the low test frequencies but not for the higher ones. "Design.f 5 I ~I Threshold I I • • • ° 50Hz 0 5 H. This is illustrated in the above graph for 2. Such mulitple effects are poorly understood and are clearly possible in a large number of material/environment systems. U Ol . there is a true threshold. Metals Park OH.1 Mo Steel R = 0. W. James E. Source: W.. J I J I j I I I 4 6 78910 20 40 60 80 100 Stress-intensity factor range.. '" 8 10. m 1/2 Fatigue crack growth rates in 2. Gerberich and John H. MPa .0Mo Steel: Fatigue Crack Growth Rates in Air and Hydrogen 10-5 I 10-6 1i Gl ~ -.h . in. t:J. Suresh et a/demonstrated that dry hydrogen may produce afrequency-sensitiveenvironmental effect analogous to SCC at intermediate t:J. the K""'j of about 22 MPa . Gl :::I 10-9 l- u. American Society for Metals.. Because the sustained-load threshold for this steel is on the order of 90 M Pa . 't:l 't:l E vi>.Kvalues greater than K . 50 Hz § 10. Ill. Therefore. m 112 (82 ksi . i>..h ({) . • v. .OMo steel in air and in hydrogen. D v i>. o .25Cr-l.. 112).10 f- f: • 0°.11 3 I . m 1/2 ( 20 k' .05 Frequency Environment Air 138 kPa H2 - • V 2 Hz. Underwood. Eds. 2% Cr ..K. W. • - 10-7 . which appears to be frequency-insensitive but which nevertheless decreased by about 30% to SA MPa . Gunderson. however.K values and a frequency-insensitive environmental effect near the threshold. 1982. .9 ksi in.? K th : SI ~ lt can be seen for t:J. Corrosion fatigue descriptions are further complicated by the fact that the environment may produce multiple effects. ~ •• ..::.25Cr-1.dt - ~ u Ol . al '" 0' oil' 10-8 o~ . this regime may be considered to be one where superposition might apply. Campbell. William W.. L1K. 1/2) gives K .

0 C. Properties and Selection: Irons and Steels. 2.25Cr-1.0Mo steel tested in strain-controlled cyclic loading at (a) 480°C (900 OF) and (b) 540 °C (1000 OF).0Mo Steel: Effect of Holding Time 191 120days E c Compressive or tensile hold Type of strain hold Strain range.25Cr-l. • • o X indicates zero hold lime Cycles to failure (b) Time-to-failure/cycles-to-failure diagrams for annealed 2. Source: Metals Handbook.7-9. American Society for Metals. which are approximately parallel.4 to 2. are for increasing periods of holding time at maximum strain levels in either tension or compression.} 0.0% with no holding period at maximum strain. 1978.0 Compressive Tensile Both loJ 4 10 Cycles to failure 1. 9th Edition. In these "time-to-failure/cycles-to-failure" diagrams. 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. failure occurred in less than I000 cycles regardless of holding time or whether the stress was tensile or compressive. Volume I.5 :::2 } . Hold time indicated on graph is length oftime that specimens were held (during each cycle) in the state of maximum tensile strain (open symbols) or compressive strain (filled symbols). 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. Metals Park OH. For all tests at 2% strain.% 2. The other curves. pp 662-663 . The vertical curves are drawn through the number of cycles to failure for each particular cyclic strain. Strain amplitude indicated by shape ofsymbols and figures along zero-hold-time line.

192 7-10._. 5.0 ~ OL..25Cr-1.0 z ••• I- cr 2.. Peter F..25Cr-1..------. Ed.------r----r---r---r--r---._ _-L-_ _---1.• Steel Founders' Society of America. A2l7.l._---1... Weiser.. Cast 2.L-_ _--'-_ _.0Mo Steel. Rocky River OH..0 • • • \ I 4. Grade WC9. at 540°C (1000 OF).-. J 100 200 400 600 1000 2000 CYCLES TO FAILURE 4000 Fatigue properties of 2.0 .0Mo centrifugally cast pipe.0 lLJ o Z cr <l 3. Centrifugally Cast: Fatigue Properties at 540°C (1000 OF) 60. P 15-55 . 1980.. Source: Steel Castings Handbook..J <r (f) g 1. 5th Edition.

.5 .000 cal...----. as determined by ductility measurements after rising load tests on hydrogen-charged materials. H11 Steel: Crack Growth Rate in Water and in Water Vapor 193 0....." in Hydrogen Embrittlement and Stress Corrosion Cracking. c o Activation Energy 9.3 "0 H -II Steel 230 ksi Y.. K ... F. o ~ • o Water o Relative Humidity 100% at Test Temperature • II II II II o u .Jl: U . Hehemann. P 18 .. R..1 .01 .. 1984./gm-atom o a::: Q) . Tested at Higher Temperature Crack growth rate versus temperature for an HII steel in water and water vapor.30 ksi IT'" c E 0. Johnson.-----. American Society for Metals.--"'"""'T--"'"""'T--~--. "Keynote Lecture: Overview on Hydrogen Degradation Phenomena. Gibala and R. 0. Eds..005 80° F..7-11. S. show a characteristic behavior that resembles closely that seen with crack growth rate measurements and external hydrogen environments... It is of considerable interest that the strain rate and temperature dependence of hydrogen embrittlement. Source: Herbert H..03 o ...05 s: . 0.. Metals Park OH.

Ed. Ashok K. J. although inferior at relatively high strain ranges.. 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.194 7-12. P 95 . Metals Park OH.0Mo Steel: Creep-Fatigue Characteristics 9% Cr 1% Mo AT 525°C WITH OR WITHOUT TENSILE HOLD TIMES TO hOWELL (TOTAL STRAIN RANGE. FROM (2311 cr « 10° z ~ cr Vl ~ 9% Cr 2% Mo NbV AT 550°C (PLASTIC STRAIN RANGE. also shown for direct comparison are the continuous cycling fatigue data for the low-C. suggests superior endurance may be attained in the high-cycle region. FROM (24)) ~--5 MIN. Source: S. 9. From the limited evidence. American Society for Metals." in Ferritic Steels for High-Temperature Applications. 1983. 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.0Cr-1. Khare. FROM (22)) t ~ UJ l-' Z LOW C 9% Cr 2% Mo AT 550°C CONTINUOUS CYCLING (TOTAL STRAIN RANGE. Sanderson. "Mechanical Properties and Metallurgy of 9%Cr I%Mo Steel. 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. 9%Cr-2%Mo variant which. TENSILE CYCLING DWELL 10' CYCLES TO FAILURE Illustrating the elevated temperature low-cycle fatigue and creep-fatigue properties of normalized and tempered 9% Cr Mo variants.

-__ 100 e a... Sikka and W.. Source: S. R.. C.0Crl.0Mo Modified Steel: Stress Amplitudes Developed in Cycling 195 -------------------------200 t .. v... Nb -----. 10 100 1000 10000 CYCLES This chart shows stress amplitudes (tensile and compressive) that developed in the course of cycling the modified Fe-9. Weertman.:"::-=-=_. Kim.. J. 9. S. 1984.... ::E Fe 9Cr IMo BV.OMo steel through a total strain range of 0." in Nondestructive Evaluation: Application to Materials Processing. J. Wolf.....5% at 649°C (1200 OF)...:_:'":_."'="::=-:-:..with 30 sec T H --- -200 r.. Otto Buck and Stanley M. Dotted curve indicates continuous cycling..._. Eds. "Microstructural Evaluation ofa Ferritic Stainless Steel by Small Angle Neutron Scattering..7-13. p 175 . Glinka..--.. Metals Park OH. solid curve indicates cycling with a 30-s hold at maximum tensile strain. American Society for Metals.. B.---.. Jones.:_::. Spooner.. Fatiguing was carried out in vacuum.. ----Continuous Cycle (/) l- en w a:: (/) o -100 .0Cr-1..

Spooner. R. B.04 0. Source: S.5 -/0. Metals Park OR. J. Glinka.ldO vs q for specimens of modified Fe-9. continuously •.Nb 0 0 Not deformed (N aT) Fotl(~ued 8 6 4 2 0 -2 -4 • • (649°C.06 0. A Et = 0. J. Jones. 6 Foti~ed ( 649°C.0Mo steel which have undergone various types of deformation.M~fb28 8 8 ~ ~ ~ ~ 0 ! 0 8 0 S 0 ~ 0 0 0 ~ S q 0. Eds. 1984. Wolf. Otto Buck and Stanley M." in Nondestructive Evaluation: Application to Materials Processing. S.48 nm. A = 0. 9.02 0. P 175 . V. cycles with 30 sec tension hold holn) • 'i 0 Crepl R~9~~.08 <I") Curves of dI.0Cr-1.0Mo Modified Steel: Effect of Deformation cb: dQ In 10 8 s Fe 9Cr IMo 8 V. "Microstructural Evaluation of a Ferritic Stainless Steel by Small Angle Neutron Scattering.196 7-14. A€t 79 = 0. Kim. American Society for Metals.0Cr-1.5 %. A magnetic field of ~28 kg was applied to the specimens during the SANS measurements. 10000 cycles. Sikka and W.. C. Weertman.

American Society for Metals. .K. These data are based on the "effective stress intensity factor. Campbell.. Gerberich and John H. 1982..." in Application of Fracture Mechanics for Selection of Metallic Structural Materials.. Kerr' Fatigue crack growth rate data reported by Walker for Y2-hard type 301 stainless steel sheet are summarized in the above graph..6 80 Effective stress-intensity factor. tested at 24 0 C (75 OF).. '" u. E...J L 10. If tests are made at several load ratios to determine m.. Eds. '" 10.. "t:l . andR ratios of 0... E E i3 10.~ / / / '" / 10-5 1/2 hard tvpe 301 24°C (75°F) 0.063 to 0. MPa • m 1/2 ax Scatter band offatigue crack growth rates of Y2·hardtype 301 stainless steel.. i3 ~ .4 GI I >u '<. Campbell... .. Metals Park OH. '" i0 '" GI' I / / / / / / 10......1/2 20 40 60 80 100 197 10... then the effects of other load ratios may be estimated.---J'---'---'---'-.807 at a frequency of 10 Hz.. '" u :::I GI '" u . to account for the effect of the range of stress ratios. Type 301 Stainless Steel: Scatter Band for Fatigue Crack Growth Rates K.c .. rather than on fj.. ksi • in.R)m where m is determined empirically and R is the load ratio (minimum load/maximum load) on cyclic loading. "Fracture Properties of Wrought Stainless Steels. Kerr is defined as follows: Kerr = K max (I .66 7 ..807 based on effective stress-intensity factor. 10 Hz. K... when compared at given values of fj..063 < R < 0.4 / / / / / / / / / / / L-.1. Underwood..If ..3 GI ~ .807 .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..K.063 to 0.. Source: J. 10.Rl o.2 / / / / z· "t:l .. "t:l . William W...3 / // / / / / z· "t:l .J'- --'-_ _. The crack growth rate law then becomes: da/dN= C[Kmax(l..8-1. P 114 .J." Kerr.. James E...1f = Km [1 ... The data were obtained in air at room temperature over a series ofload ratios (R) from 0.

.30 to +95 °C (-22 to +203 OF). summarized in this graph.198 8-2. 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. Gerberich and John H. ksi . "Fracture Properties of Wrought Stainless Steels.. Campbell. This effect of warm working has been observed for other austenitic stainless steels. tested in argon . Source: J. James E.. indicating that the humidity and/ or oxygen in the air influenced the growth rates. Fatigue crack growth rates in laboratory air at 20°C (68 OF) were higher than for corresponding conditions in argon. American Society for Metals.28 in. MPa .The results. fatigue crack growth rates at a given t::..5 Gl U I"'" 20°C (68°F) L-T "'" 10. William W.E --" ' "tl "tl > c. Type 301 Stainles Steel: Effects of Temperature and Environment on Fatigue Crack Growth Rate tlK. 1/ 2 20 10. This transformation effect would be most noticeable in type 301. P 113 .K value increased as the temperature increased over the testing temperature range. Campbell.0 1. Eds.. tested in argon Fatigue crack growth rates for type 301 stainless steel have been reviewed by Pineau and Pelloux in the temperature range from .. E. Metals Park OR.) thick at a cyclic frequency of 20 Hz with a sinusoidal waveform at a load ratio (R) of 0. All specimens were tested in dry argon except one series that was tested in laboratory air. were obtained on compact specimens 7 mm (0.6 -. in.7 30 40 50 60 Stress-intensity factor range. resulting in a substantial increase in strength.Annealed. These differences are attributed to the extent of the strain-induced transformation at the crack tip. tested in air . For the annealed specimens tested in argon. because it is less stable than the other alloys in the UNS S3xxxx series. The warm worked specimens were reduced 65% at 450 to 500°C (840 to 930 OF)...> 2' Type 301 2 X 10. 1982.Warm worked." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. Underwood. tl K.. m 1/2 -----Annealed.

.. Coffin.8-3. 430°C 650°C 8160C (f) a: ." in Fatigue and Microstructure. 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.. showing frequency-modified elastic and plastic strains at three temperatures in air. « z (f) Data of Berling and Slot for AISI 304stainless steel. Representation of the behavior here utilizes fatigue equations known as frequency-modified fatigue equations. interest in the fatigue problem in the power-generation industry generally involves elevated temperature. F... 1979. American Society for Metals. Source: L. The importance offrequency or strain-rate effects is shown in this chart. Metals Park OH.... Type 304 Stainless Steel: Effect of Temperature on Frequency-Modified Strains 199 AISI 304 STAINLESS STEEL o • o • LIJ ..) 6 . In contrast to most other segments of our technology. Laboratory testing on both smooth specimens and specimens designed for crack growth is performed with temperature and frequency or strain rate as parameters. They describe the elastic and plastic strains versus fatigue life and include the frequency or strain rate of the test. P 13 . These data are for solution-treated AISI 304 stainless steel subject to triangular wave shapes at equal-loading and reverse-loading strain rates. "Fatigue in Machines and Structures-Power Generation.J « <...

~K. Watson.. James E. American Society for Metals.. William w. and Smith. 1982..... "t:l "t:l s: . tested at 427°C (800°F)~ 80 100 2 X 10-4 / I Annealed. tested at 25°C (77°F) Cold worked 10-5 "t:l -. tested at 427°C (800°F) - I I I I ~// I . illustrated in this graph.. A comparison of fatigue crack growth rate data by Shahinian.. Cl ~ /it' 10.1: " / / / 10-4 . E. Type 304 Stainless Steel: Fatigue Crack Growth RateAnnealed and Cold Worked ~K.. "t:l z' '" u .. '':.3 > u E E z' '<. U '<. Gerberich and John H.. 1/ 2 10 20 30 40 60 Cold worked 25%. shows that the high-~K crack growth rates were lower for the cold worked specimens than for the annealed specimens. / " 30 Annealed... type 304 stainless steel components are fabricated in the cold worked condition to improve strength properties. in. U -<.. Underwood. U.----lU 10. Campbell.6 10 20 40 60 80 100 Stress-intensity factor range. Source: J." in Application of Fracture Mechanics for Selection of Metallic Structural Materials..~ -" u IJI. Eds. > u . Metals Park OH. tested at 25°C (77°F) Type 304 -'--_--'-_. "Fracture Properties of Wrought Stainless Steels. Campbell. In some applications. 10. MPa • m 1/2 Fatigue crack growth rates for annealed and cold worked type 304 stainless steel at 25 and 427°C (77 and 800°F).. Crack growth rates were higher for the specimens tested at 427°C (800 OF) than for corresponding specimens tested at room temperature._---'_.. ksi .' 25%.. s: ::> . p 120 .4 '" / ...17 Hz.. and an R ratio of O.L..200 8-4.L..' . 0.. ~ '" e Cl I' / y/~ / .

1982." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. William W. James E. 0. 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.. and an R ratio of O.1 (Bamford). Gerberich and John H. and Smith). CiK.K range. Watson. Metals Park OH.4 10. However. Underwood. 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. Campbell.1/2 201 10 20 40 60 80 100 10.17 Hz with an R ratio of zero (Shahinian. American Society for Metals. Source: J. However. p 122 . 2 and 0. At the lower end of the t:. Eds. fatigue crack growth rates in humid air are substantially greater than crack growth rates in dry air. MPa • m 1/2 Effect of humidity on fatigue crack growth rates for type 304 stainless steel tested at room temperature. "Fracture Properties of Wrought Stainless Steels. variations in R ratios influenced the fatigue crack growth rates in the pressurized water reactor environment. Campbell. E. Type 304 Stainless Steel: Effect of Humidity on Fatigue Crack Growth Rate CiK. in.5 ~Roomair 10.17 Hz.8-5.7 were no greater than the fatigue crack growth rates in air at the same temperature with an R ratio less than 0.4 / / / Type 304 25°C mOF) 20 30 40 Stress-intensity factor range. ksi .

there tends to be a slight increase in the fatigue crack growth rate at a given 11Klevel. 1982. l1K.._--'-_.. "Fracture Properties of Wrought Stainless Steels.0 minute is included in each loading cycle.I--'- ....~ ~ "t:J '" 10.0 min 10...4 Unaged Aged o • • A Hold time Zero 0.. precipitation of M 23C 6 carbides is essentially complete. m 1/2 Effect ofaging at 593 ° C (1100 OF)for 5000 h. Underwood..l--_ _. William W. and hold times of 0. When a holding time of 0. P 121 ..5 Z .1 and 1. Gerberich and John H. or 1100 OF) are shown in this graph. Campbell. Source: J. American Society for Metals..3 u 10..1 min 1. as reported by Michel and Smith. Campbell.. in.0 min for each cycle." in Application of Fracture Mechanics for Selection of Metallic Structural Materials.. Results offatigue crack growth rate tests on specimens that were tested in the unaged and aged conditions (5000 hours at 593 °C. Eds.17 Hz and an R ratio of O.1 or 1. After aging for 5000 hours at this temperature. E.----I_L-. ksi . > lJ -. 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. James E.... Metals Park OH...L. an evaluation of the effect oflong-time aging at service temperatures is important.6 Type 304 593°C (1100°F) 10-5 L-.I. Type 304 Stainless Steel: Effect of Aging on Fatigue Crack Growth Rate l1K.. on fatigue crack growth rates of L-T oriented specimens of type 304 stainless steel tested in air at 0.L-J.1/2 10-4 10.. Because austenitic stainless steels are expected to give long service life.202 8-6.JL-. MPa .J 10 20 40 60 80 100 Stress-intensity factor range.

. E. Type 304 Stainless Steel: Effect of Temperature on Fatigue Crack Growth Rate . >u E E 10-3 538"C (1000"-><-/ z· ~ '" " tJi .. Source: J. II> u .. As shown in this graph..8-7. 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.. and by others..066 Hz and anR ratio of 0 to 0. in. II> :::l 10.. American Society for Metals. Gerberich and John H....1K.. James E. . MPa .. m l/Z Effect of testing temperature on fatigue crack growth rates for annealed type 304stainless steel tested in air at 0.4 '" 10.. 1/ .66 Hz for the room-temperature tests and 0.1K. 10-4 . William W. Campbell.E z· . for several different maximum alternating loads.05. " '" " 10-6 316°C 1600°F) 20 40 60 100 Stress-intensity factor range. Metals Park DB. for load ratios of 0 to 0. Eds... 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." in Application of Fracture Mechanics for Selection of Metallic Structural Materials.05. These data. are for specimens of both the L-T and T-L orientations.067 Hz for the elevated-temperature tests. reported by James and Schwenk. "Fracture Properties of Wrought Stainless Steels.: ~~~ Type 304 10.5 t' I~ ~' 10 f! t<. II> y . Campbell. >- . Underwood.. P 115 ...5 c:. '" u. i0 u u E ~ . ksi .'" ... and for cyclic frequencies from 0.'" "" .033 to 6.1!z 203 10 20 40 60 100 649O C l 1 2 c:.. 1982..

: .... I3 v l3 ( k. A damage function is approximated by the quantity U.. since. ~ -tI Q o o STRAIN RATE.I -.. for the same inelastic strain.. 101 105 Ostergren's damage relation for AISI 304 at 650°C (1200 OF)..lIN... "Fatigue in Machines and Structures-Power Generation.. I ) = C = 1.' -.... P 23 ..!:iE p ...:-----''------''-----'---'--'-=----'----'----'---'---'-:-----'----'--... IN.204 8-8. is the maximum stress in the cycle and !:iE p is the inelastic strain range... where u.. .. Metals Park OH... a positive mean stress provides a greater hysteresis energy than does a compressive mean stress. The use of the tensile-stress quantity. American Society for Metals.. F.. Source: L. Type 304 Stainless Steel: Damage Relation at 650°C (1200 OF) o ()...895 k = 0.. The method is effective in accounting for hold-time effects. as indicated in the chart above.lSEC 4 x 10'3 HOLDPERIOD. The damage function was proposed by Ostergren and is based on the frequencymodified fatigue approach.1 0 0 3 30 3 0 0 0 C O"rtl€pN. L... -...756 101' . 1979.158 x 10 5 13 = 0... MINUTES TENSION COMPRESSION o 0 I 0 10 0 30 0 60 0 Ll 'i7 180 o I 0 30 I o ~ 4 x 10'3 4 x 10'4 4 X 10'5 o o (3 3 30 30 0. "in Fatigue and Microstructure.-.. 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... in conjunction with the plasticstrain range.. provides a means of accounting for loop unbalance. Coffin.

..+-.. • 24°C (75 of) o -196°C (-320 OF) 0-269 °c (-452 OF) 10-4 f--+---+-------4I::L-+------=l 5 10 50 Stress intensity factor range. Properties and Selection: Stainless Steels..... Type 304 Stainless Steel: Fatigue Crack Growth Rate at Room and Subzero Temperatures 205 5 10-4 10-3 1-. 11K. Tool Materials and Special-Purpose Metals... 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.. 1980. American Society for Metals. 9th Edition. p 756 . Volume 3.8-9. Source: Metals Handbook.. Metals Park OH..+-'''''10. crack growth rates are nearly the same at room and cryogenic temperatures.

E > u tJ) -"C "C z Cll .... Types 304 and 304L Stainless Steel: Effect of Cryogenic Temperatures on Fatigue Crack Growth Rate ~K. Eds.... Campbell. 1/ 2 5 20 40 60 80 ~ ..>t!. . . while for type 304L.... E E "C .. '" u Gl u .= i! 0 .206 8-10. MPa • m 1/2 Fatigue crack growth rates for annealed types 304 and 304L stainless steel at room and cryogenic temperatures.. William W...6 Stress-intensity factor range. "Fracture Properties of Wrought Stainless Steels. American Society for Metals. -452°FI '" 10. the calculations will be conservative for cryogenic exposure. 1982. Source: J.3 Gl Z U ~i ~ . -269°C (-320.. Metals Park OH. IL tJ) ~ 10. 20 to 28 Hz. Fatigue crack growth rate data obtained by Tobler and Reed on specimens of types 304 and 304L stainless steel (annealed) at temperatures in the range from room temperature to liquid helium temperature (-269°C..4 Type 304L -196..." in Application of Fracture Mechanics for Selection of Metallic Structural Materials..5 -. E Type 304L 22°C (72°FI 10. Campbell. Furthermore. or -452 OF)are shown in this graph. if design calculations for type 304L are based on room temperature fatigue crack growth rates. The data for type 304 were scattered over the range shown. Gerberich and John H. Underwood. Cll "C U Gl 10. James E.1. ksi • in. the data at room temperature described one curve and the data at the cryogenic temperatures described the other curve. P 123 . E. ~K. and an R ratio of 0. These results indicate that cryogenic fatigue crack growth rates for type 304 do not deviate significantly from room temperature fatigue crack growth rates over the /:!"K range studied.

. E E 10..... >- . co / / / / / / / / Type 304 538°C (1 0000 F) R = 0.05 with two different waveforms at 0. 1982.. Underwood..8-11... 2 10 20 '/ 30 40 50 10. 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. American Society For Metals.3 Z· '1::J .4 ( ( / / / / / / / / / I / / / II / / Waveforms . James E..4 CIl U ~ .05 10. "Fracture Properties of Wrought Stainless Steels.!! u 10.5 u . CIl :J u." 10.. Eds.E Z· '1::J rYYY\ /VVV\ } 0... m 1/2 Scatter band of fatigue crack growth rates for annealed type 304 stainless steel at 538 °C (IOOO°F)in air at anR ratio of 0.. Gerberich and John H..6 10 20 30 40 50 Stress-intensity factor range. MPa . Type 304 Stainless Steel: Fatigue Crack Growth Rate in Air With Variation in Waveforms ~K. in.." . ~J<. Source: J.. Campbell. Campbell.P 117 . 207 ksi .. ..067 Hz. Metals Park OH." in Application of Fracture Mechanics for Selection of Metallic Structural Materials.... The data presented in this graph were obtained in tests with a sawtooth waveform.067 Hz ~ co ... William W. E. co '1::J 2l' s: e u co u i0 ~ .

Overload effects are also important in retarding crack growth. Coffin.-_ _.. Source: L.L.....CYCLES TO FAILURE Effect of hold time on life for AISI 304 stainless steel." in Fatigue and Microstructure.. Metals Park OH.. American Society for Metals._ _.001 L. as can occur in long tensile-strain hold-time tests. Z ILl C> ~ 180 60 30 LABORATORY TESTS AISI304 STAINLESS STEEL 650°C z 0.. given a fixed period of cycling.. Substantial damage can result from these wave shapes.l.---L----L---l... particularly when the hysteresis loop is severely unbalanced.. "Fatigue in Machines and Structures-Power Generation..-_-. P 19 .01 « a::: a::: en -J ~ z « o DO 01 60 10 0 30 30 o I 0 I o NO HOLD TIME o TENSILE HOLD TIME IN MINUTESAS INDICATED g ~ 0.L. 1979.-.000 LIFE .L...._ _--'--~ 100 1000 10.L-.-_--L.-_----L_. He observed that the crack growth rates were greater as the loading rate increased and the unloading rate decreased.208 8-12. Wave-shape effects are also important in fatigue crack growth. Type 304 Stainless Steel: Effect of Hold Time on Cycles to Failure z .l.. as has been studied by Barsom. F.

L-_---'_---'_..1 '--_ _---'_ _. 1982.1 and 1. in. as shown here.f J I I " " ~ 10. Source: J. the fatigue crack growth rate is greater for specimens tested with no holding time (continuous cycling) than for specimens held at maximum load for 0. m 1/2 Fatigue crack growth rates per unit of time ida]dt) for annealed type 304 stainless steel for continuous cycling (0. cyclic loading has a more damaging effect than static loading on crack growth per unit of time.0minhold .L-_.:E.17 Hz). Type 304 593°C (1100°F) 10.L.1 .0 minute per cycle. Type 304 Stainless Steel: Effect of Hold Time and Continuous Cycling on Fatigue Crack Growth Rates 6K... American Society for Metals. Gerberich and John H.0-min hold times at maximum load for each cycle at 593 ° C (1100 OF). and for anR ratio of O. MPa . The same trend was observed for tests at 593°C (liOO OF).r~ 1." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. As shown above.1 or 1. "Fracture Properties of Wrought Stainless Steels.1 min hOld-. James E... ksi . based on dal dt. E E I I 0. Therefore. The lowest fatigue crack growth rates occurred for specimens with the longest holding time. 1/ 2 209 10 20 40 60 80 100 10. for 0. Campbell. Campbell..2 / . E. William W..8-13...I 20 30 40 60 80 100 Stress-intensity factor range... Metals Park OH. P 118 . Underwood.. 6K. Eds.

as shown here for tests at 538°C (1000 OF). en u.0014 Hz Q) 0.05 10. Source: J. Gerberich and John H.4 -. Campbell.. American Society for Metals. :: u ~ e en u . MPa .. 1982. P 116 .4 U -E E z' --" ' "tl "tl >o 10-3 l!l' ~ u 10. increasing the cyclic frequency will decrease the crack growth rate over part of the ~K range. Q) . Underwood. d K. Eds. '" Type 304 538°C (lOOO°F) R = 0. "Fracture Properties of Wrought Stainless Steels. E. Campbell.10: 10.:.. James E. ksi . m 1/2 Effect of variation in cyclic frequency on fatigue crack growth rates for annealed type 304stainless steel at 538 ° C (1000 0 F) for an R ratio of 0.6 10 20 30 40 50 Stress-intensity factor range.5 Q) s: . Metals Park OH. in. For fatigue crack growth rate tests on specimens of annealed type 304 stainless steel at elevated temperatures. Type 304 Stainless Steel: Effect of Cyclic Frequency on Fatigue Crack Growth Rate dK." in Application of Fracture Mechanics for Selection of Metallic Structural Materials.0067 Hz 10.5 z' --" ' "tl "tl >u '':.05 in air with a sawtooth waveform. William w.210 8-14..1/ 2 10 20 30 40 50 0..

Coffin. -c .05. Ib/lin)3/2 5x101 STRESS INTENSITY FACTOR RANGE......8-15.. ~K.. foIW/(m?/2 Effect of frequency on the fatigue crack growth behavior of type 304 tested in an air environment at 538°C (1000 OF). American Society for Metals.: co: 0 <.. -e :::I <.003 cpm 'il 0. Source: L.. "Fatigue in Machines and Structures-Power Generation... ." in Fatigue and Microstructure. kg/lmm?/2 5x 101 Il . Type 304 Stainless Steel: Effect of Frequency on Fatigue Crack Growth Behavior 211 SiRESS INTENSITY FACTOR RA:\iGE.:l ANNEALED TYPE 304 S S. . "' 10-5 :I: . P 14 . oM. TESTED IN AIR AT 53S"C 1l000°f) R· 0. 400 cpm STRESS INTENSITY FACTOR RANGE..l::! ~ c: u s: z· ::!2 't:l ~ co: . Metals Park OH... ~K. F... [45] + 10-6 0..:l u :><: e:( co: y ~ .4 cpm 00 4cpm o o 4 10 40cpm 4000cpm li. Ref. 1979.

S :: z· '" . MPa .. u..K.4 Shielded metal arc E E Z ~ -"t:l "t:l c:.17 Hz..:. .K.4 Submerged arc 10. E. Metals Park OH. Campbell.. m 1/2 Stress-intensity factor range.10. Type 304 Stainless Steel Welded With Type 308: Fatigue Crack Growth Rates t. . P 125 ...K. American Society for Metals. In general.S z· . '" . 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.5 -. Eds. Underwood. "Fracture Properties of Wrought Stainless Steels..212 8-16.. t.3 c:. > u Ql co -e s: "t:l oi . 1/ 2 20 40 60 80 20 40 60 80 10. James E. ksi . Campbell..4 10.. -"t:l "t:l c:.4 o SMAW 1 • SMAW 2 24°C (75°F) 20 40 10-6 60 80 100 20 593°C (l100°F) 40 10-6 60 80 100 Stress-intensity factor range.e s: 0 "t:l "t:l ~ . Gerberich and John H. 1I2 t. 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.K. it is important to have fracture information on weldments. > u E E Shielded metal arc Ql 10... m 1/2 Fatigue crack growth rates for annealed type 304 base metal and type 308 weld metal at 24 and 593°C (75 and 1100 OF)..... -. Ql 10.5 -.2 c:. ksi . 1982.. Source: J. t." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. 0. 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 593°C (llOO°F). in. 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.:. in. u. 10. j!l' u u Ql ~ z· 3 ~ . William W. -. u . MPa . ~ . > u Ql • 10. e i0 ~ ~ 10.

. <II 213 ~" 300 Type 310. Properties and Selection: Stainless Steels.." E ~ 100 10 5 10 ~ "' 106 No..~ x :::l ".. longitudinal ~~-~ 40 '" . Source: Metals Handbook. . American Society for Metals.~ ~ Type 304. 1980.. Metals Park OH. P 32 . '" 30 '" 20 . '" ~ .><: 'iii ~ 200 E :::l . Types 304 and 310 Stainless Steel: Effect of Direction on S-N 400 Q. transverse Type 310. longitudinal 50 ...8-17. 9th Edition..SoN curves for two grades of stainless steel. Volume 3. Tool Materials and Special-Purpose Metals. of stress cycles 0 108 .

E..however... "Fracture Properties of Wrought Stainless Steels... L-T orientation.17 Hz.. and 348 stainless steel in air at room temperature and 593°C (1100 OF). . Gerberich and John H.. 316.. Metals Park OH. fatigue crack growth rate tests were made on singleedge-notch cantilever specimens oftypes 321 and 348 stainless steel from L-T orientation at 0. AK. Eds. 1/ 2 20 10 60 100 40 10-2 .I. Source: J.-r----r----._--l. and 348 Stainless Steel: Effects of Temperature on Fatigue Crack Growth Rates AK. William W. 321.J.. The curves above show that. James E.4 u Q) -E E z· --" ' a> "0 "0 > Ll 10. 4 X 10-4 10..321. and 348 all fall within a narrow band. and an R ratio of O. and Watson.-.. As reported by Shahinian... whereas specimens of type 348 had the highest fatigue crack propagation resistance.. ksi ." in Application of Fracture Mechanics for Selection of Metallic Structural Materials.17 Hz with an R ratio of zero at room temperature and at elevated temperatures to 593°C (1100 OF). Smith. 316. fatigue crack growth rates in air increased with increasing testing temperature.214 8-18.3 .. in. the fatigue crack growth rates for types 304. 3: ~ Ll e Cl ~ l'Il Ll .~ Cl l'Il IL.L..---. American Society for Metals. at room temperature.._. Campbell..!!! . Campbell. Underwood. -z· -. s: . specimens of type 3I6 had the least fatigue crack propagation resistance. 0. 10-4 L -_ _. As for types 304 and 3I6.l. 316. Results of tests on specimens of types 304 and 321 were nearly the same at 593°C (1100 OF) in air. 1982. P 138 . Types 304... For tests at 593°C (1100 OF). 321.------r---.J 10 20 40 60 10-6 80 100 Stress-intensity factor range.~ "0 "0 l'Il > Ll Ll Q) 10-5 ::::I . over the 11Krange studied. MPa • m 1/2 Fatigue crack growth rates for annealed types 304.

.05.. "Fracture Properties of Wrought Stainless Steels.. does not affect the results of fatigue crack growth rate tests. ...1/2 r . These data provide further evidence that a wide variation in grain size. ~K.r .. ksi • 215 in. 10. 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. Types 309S and 3 lOS stainless steel are the low-carbon versions of types 309 and 310. Gerberich and John H.r . Type 309S Stainless Steel: Effect of Grain Size on Fatigue Crack Growth Rate ~K. The results are plotted above.. American Society for Metals.7 10 20 40 100 Stress-intensity factor range. Metals Park OH.6 10.... Underwood. Eds.. MPa • m 1/2 Fatigue crack growth rates for annealed type 309S stainless steel for two grain sizes. . E... They have higher chromium and nickel contents than type 304 and consequently have better corrosion resistance and more stable austenite than type 304..05 at room temperature in air. Specimens with the smaller grain size had substantially higher yield and ultimate tensile strengths than the specimens with the larger grain size. Hz 10 15 20 25 30 Grain size 45 fJm 480 fJm -0-0- 10. P 126 . Campbell..8-19.3 10 20 40 100 10." in Application of Fracture Mechanics for Selection of Metallic Structural Materials.. and the associated variation in strength level.. Campbell. 1982.. James E. Source: J.. at frequencies from 10 to 30 Hz and an R ratio of 0. William W.2 10-4 Type 309S Testing frequency. Fatigue crack growth rates were obtained on tension-tension loading at frequencies from 10 to 30 Hz and at an R ratio of 0..

co ... Type 3105 20 40 60 100 Stress-intensity factor range.269 0 C (72.216 8-20.320. James E. > ("l e en -" o co .. Campbell.. Because of its high nickel content. ksi • in.. 20 40 10-4 . MPa • m 1/2 Fatigue crack growth rates for annealed type 31OSstainless steel at 22..1..5 z· -"C "C co . "Fracture Properties of Wrought Stainless Steels. Eds.. P 127 .---. ~K..4 en '':. -452°F) u 10..3 E E -. and anR ratio of 0. Gerberich and John H. -269°C (-320.c . American Society for Metals.. :::l 10.---. William W.-------r---r-.. . Type 31 OS Stainless Steel: Effect of Temperature on Fatigue Crack Growth Rate ~K. Metals Park OH.. and .. .. ~ ("l Base metal -196.' co z· .5 .______.. -196. with corresponding data for SMAW welds with type 316 filler metal. Source: J. it is often considered for cryogenic applications that require a high degree of austenite stability on thermal cycling and strain cycling. 1982.. Underwood. 1/ 2 . co u. Fatigue crack growth rate at various temperatures is illustrated above. Campbell... and -452 0 F). E. Therefore." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. -."""T""'l~---.... type 3lOS stainless steel is completely stable at all cryogenic temperatures and with any amount of cold working.. U -"C "C > ("l 10. 10 to 28 Hz..

K level (Shahinian). >0..4 Type 316 593°C (1100°F) 10. Source: J. z' "C <0 Cl "C ~ 10. The weld was produced by the submerged arc method using type 316 welding rod. m 1/2 Fatigue crack growth rates in type 316 base metal and weld metal in the unirradiated and irradiated conditions at 593 ° C (BOO OF)in air [fluence 1. James E. 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. The curve shown above for unirradiated weld metal tested at 593°C (1100 OF) represents fatigue crack growth rates substantially lower than those for the unirradiated base metal at any given I:>. E.2 X 1022 n] cm-. Irradiation slightly reduced the fatigue crack growth resistance of the weld metal.17 Hz and at an R ratio of zero. 1/ 2 217 10-2 20 40 60 100 4 X 10-4 Ql 13 '<.3 Ql 13 '<. liK. MPa . s: 0 fl' E > u .. Weldments were stress-relief annealed at 482°C (900 OF). ~ u . Eds . 1982. but its fatigue crack growth resistance was better than that of the unirradiated base metal.. Specimens were single-edge-notch specimens for cantilever loading and were tested at 0. Smith." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. Metals Park OH.8-21.6 20 40 60 100 Stress-intensity factor range. Campbell. Type 316 Stainless Steel: Growth Rate of Fatigue Cracks in Weldments liK. Gerberich and John H. <0 10. Underwood.1 MeV at 410°C (770 OF)]. American Society for Metals. ksi • in. "Fracture Properties of Wrought Stainless Steels. p 134 .J Cl u.5 E u Ql :::J '. William W. > u E E 10-4 z' ~ "C <0 10.~ '<. and Hawthorne). Campbell.

1/2 10 60 80 20 40 10 2 . Campbell. Type 316 Stainless Steel: Fatigue Crack Growth RatesAged vs Unaged AK.2 III 10. cold worked." in Application of Fracture Mechanics for Selection of Metallic Structural Material. 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.S . MPa . E.-------r--. and hold times during cycling.. 10..1 U J!l' E e .L~ 10. III :J u.0055 to 6.L-_--l. ksi ... on fatigue crack growth rate of20% cold worked type 316 stainless steel at 593 ° C in air.intensity factor range. Campbell._.. in.. Eds. the fatigue crack growth rates were highest for the specimens subjected to the lowest cyclic frequency.L--l.05.. "Fracture Properties of Wrought Stainless Steels...3 10.. .... 1982.---r------.. m 1/2 Effect of exposure at 593 °C (1100 OF) for 5000 h.1 min 1.2 10-4 -..0 min • • • 10. u i 10.. Source: J..... at 538°C (lIDO OF) and at an R ratio of 0. William W.5 10. p 133 .c: e 10. American Society for Metals.0 min 60 10 20 40 10.66 Hz. --.. Over the 11Krange studied. 0 0 tested at 593 C (1100 F ) III U E E z ~ "lJ .218 8-22.. Metals Park OH. James E.".6 80 100 Stress.5 v L-_. Type 316. "lJ > u Z "lJ E u en . Underwood..4 0 Ii.. AK. .--. Aged Hold time Zero 0. Gerberich and John H.

..c . e i ~ Q) '" Q)' 10-3 U >u '<..1 min 1. Type 316 Stainless Steel: Fatigue Crack Growth RatesEffect of Aging . 112 219 20 40 60 80 10-1. .. p 130 . ksi • in. . LL.0 min 10-6 10-5 10 20 40 60 80 100 Stress-intensity factor range.~ "tl ~ "tl '" 10-5 Type 316 593°C (1100°F) Unaged Aged 0 en '.. American Society for Metals.... "Fracture Properties of Wrought Stainless Steels.1K. Underwood. :::l '" 10-4 G • • . m 1/2 Effect of exposure in air at 593°C (1100 OF) for 5000 h. and hold times. Campbell. A Hold time Zero 0. on fatigue crack growth rates for annealed type 316 stainless steel at 593°C in air.. e en u u z· '<.:. Metals Park OH. E E U 593°C (1100°F) z· ~ "tl 10. E.. Eds.." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. James E. Source: J. . --r-----. MPa .. Gerberich and John H.10 10-3 10-2 Q) ~ -.1K... 1982.-------.. William W.----.4 Q) .. Campbell.--.8-23..

Type 316 Stainless Steel: Effect of Temperature on Fatigue Crack Growth Rate dK. Source: J. dK. Similar results have been reported by Shahinian for tests on cold worked type 316 at 427°C (800 OF). R = 0. "Fracture Properties of Wrought Stainless Steels. E.3 Ql 1) 10. Eds. Underwood.4 10. Campbell. 0. ksi • in. American Society for Metals. Results of tests on compact specimens of 20% cold worked type 316 stainless steel at frequencies of 0. 1982. 3 Hz at 24°C. m 1/2 Fatigue crack growth rates of20% cold worked type 316stainless steel for various temperatures. William W.5 -"'C III "'C >u 2" 10-6 Type 316 Cold worked 10. Campbell.67 Hz at elevated temperatures.05 are summarized in the above graph. Metals Park OH." in Application of Fracture Mechanics for Selection of Metallic Structural Materials.0 Hz and at an R ratio of 0. James E. Gerberich and John H.67 and 3.5 10-4 -. MPa .05..5 10 20 40 60 100 Stress-intensity factor range. 1/ 2 10 20 40 60 100 10.220 8-24. P 132 . Curves are averages for L-T and T·L specimens at each temperature in air.

Campbell.0067 Hz----." in Application or Fracture Mechanics for Selection of Metallic Structural Materials.6 over the range of 11K values studied. James has shown that the spread from high to low values of fatigue crack growth rates is no greater than that represented by a factor of 2.• American Society for Metals.5 <. 1I2 221 10 20 0. the trend is for the crack growth rate to increase as the frequency is decreased. Eds . Type 316 Stainless Steel: Effect of Cyclic Frequency on Fatigue Crack Growth Rate dK. In studying heat-to-heat variations in fatigue crack growth rates for specimens from three heats of type 316stainless steel.0067 to 6.. ksi • in.. Campbell. e '" 0067 s: .5 U II> I 1." '" U II> Cl . 30 40 50 60 10.4 U II> 10.8-25. . P 131 ... E.6 10 Stress-intensity factor range. but there is more scatter than for type 304. Underwood.4 f -. "Fracture Properties or Wrought Stainless Steels. Source: J. As may be observed above in tests at frequencies in the range from 0.05. One heat was produced by air melting..1 u "'.67"' 10. dK. Gerberich and John H. 10. and a third by double vacuum melting.. James E. > u z· "C "C '" ::> u.: e Cl "'-y' //J '~ )7 ~0.. 1982. '" Type 316 538°C (lOOO°F) 10. Metals Park OH..67 Hz at 538°C (1000 OF).. another by vacuum arc remelting. William W.3 -~ "C > o E E z· ai . MPa • m 1/2 Effect of variation in cyclic frequency on fatigue crack growth rate of annealed type 316 stainless steel in air at 538°C (1000 OF) and an R ratio of 0.

. but the data also are applicable to design of equipment for fossil fuel power stations. Gerberich and John H. Underwood.222 8-26." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. 1I2 10 20 30 40 60 80 100 10.4 :i " 316 ' .J 10./ I . E..-_ _'-_J.J-J. Eds..5 / /'1 I 10.. Its improved yield strength compared with that of type 304 stainless steel is an advantage for these applications./ 593·C . MPa • m 112 Effect of testing temperature on fatigue crack growth rates for annealed type 316 stainless steel tested in air at 0. Source: J.. Campbell./ . 6K. Type 316 Stainless Steel: Fatigue Crack Growth Rate in the Annealed Condition 6K.4 y: /: "I l' / u . >- ~ .17 Hz and an R ratio of O. ---t! . Metals Park OH.. Campbell. ksi .----'_L-l. "Fracture Properties of Wrought Stainless Steels..I-J.' (1100·F) / p'" /1 10. American Society for Metals. and chemical plants. Effects of elevated temperature on crack growth rate are summarized in the graph above...F) 10.5 z· ~ . P 129 . petrochemical refineries. William W... James E. so it is advantageous to use type 316 rather than type 304 for critical cryogenic applications. 1982.1 /1 //' fi. The austenite stability in type 316 is greater than that in type 304.' I :. 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.3 370·C (700.'I "I I' 10.2 482·C {900· ""1/. in.6 10 20 30 40 60 80 100 Stress-intensity factor range..

8-27." in Source Book on Materials for Elevated-Temperature Applications. "Environment-Dependence of the Mechanical Properties of Metals at High Temperature.0 V> l- u 1. 1979. Cook and R.0 . Source: R. based on cyclic strain and cycles to failure. Helium.4 EXPOSED SPECIMENS 0.. and Air) on Cycles to Failure 10. Bradley.1 3 4 6 8 103 2 CYCLES TO FAILURE Effect of environment on fatigue characteristics of type 316 stainless steel at 92SK.0 4. P. Ed" American Society for Metals. H. P 84 . Elihu F. Type 316 Stainless Steel: Effect of Environment (Sodium.0 6.. Skelton.6 0.2 • 2 10 • • FATIGUED IN SODIUM FATIQJEDIN AIR FATIGUED IN HELIUM EXPOSURE MEDI~ Na EXPOSURE CONDITION 286 hrs AT 92S K ue ppm OF 02) 6 8 104 0.J u 0.0 8. Metals Park OH.0 t' 223 DATA POINTS o x a ex: ~ FATIGUE TEST MEDIA Na (10 ppm OF 02) AIR HELIUM z 2.8 >u 0.

... 40 60 100 ---.5 Types 316 and 321 649°C (1200°F) Dry nitrogen Dry argon 8 10 20 40 60 100 Stress.intensity factor range. Eds... however. Campbell.." <0 . Gerberich and John H." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. Metals Park OH.-..224 8-28. Types 316 and 321 10-6 2SoC (77°F) Room air Wet air 10. American Society for Metals. ksi . 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.:. In air. E. James E. 6K.=----. William W.. fatigue crack growth rates for types 316 and 321 also were the same.. U u " > Type 316 649°C (1200°F) Roomair 10-5 E z· ~ .. Fatigue crack growth rates in room air at room temperature were the same for types 316 and 321 stainless steel. p 135 . in. Source: J.. Fatigue crack growth rate data at 25°C (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..~ . Campbell. "Fracture Properties of Wrought Stainless Steels. 1982..~ :> " tn u..:.1I2 8 10 20 10-3 r----.:." f! . in tests at 649 °C (1200 OF)in dry nitrogen..c E Type 316 10-' 2SoC 177°F) Dry air Wet nitrogen Dry nitrogen fi > -!:! U " ~ u 12 tn f! z· ~ . MPa • m 1/2 Effect of gas environments on fatigue crack growth rates for types 316 and 321 stainless steel at 25 and 649°C (77 and 1200 OF). Furthermore. Types 316 and 321 Stainless Steel: Effects of Gaseous Environments on Fatigue Crack Growth Rates aK.-.. Underwood.>< u ..----..

.. "Fracture Properties of Wrought Stainless Steels. ksi .17 Hz). with 0. MPa . '" Unaged Aged 10.." in Application of Fracture Mechanics ForSelection of Metallic Structural Materials. in.. James E. Campbell....J 10 20 40 60 80 100 Stress-intensitv factor range. Source: J.6 • Hold time Zero 0.K. Results oftests by Michel and Smith on specimens of annealed type 321 stainless steel that had been aged at 593°C (1100 OF) for 5000 hours and then tested at 593°C have shown that long-time exposure at the service temperature does not reduce the fatigue crack propagation resistance in air. American Society for Metals. Campbell. 1982. E. Gerberich and John H. Fatigue cycling with holding times of 0. .8-29.l----JL-.1 min 1. William W. 10..D minute on each cycle increased the crack growth rates slightly.. . Type 321 Stainless Steel: Effect of Hold Time on Fatigue Crack Growth Rates .5 ..0min 10' '--_-'- .. > B ~ e '" u "" e u ~ ..6K. Aged specimens tested with zero holding time had lower crack growth rates than corresponding specimens that were not aged (see above graph)..1 and I. " ~ u 10.3 E E ~ z· .3 10.~ u.5 0 • .ll...J. as shown in the figure. P 139 .0-min hold time at anR ratio of 0 at 593°C (1100 OF). 1/ 2 225 10-' 10 20 40 60 80 10. Metals Park OH. 10-' 10..~ ~ z· .6 .L_ _. Underwood. ::I 10-' " 10.1 and 1..2 U > u Type 321 593°C (1100°F) . m'l2 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. Eds.

MPa . "Fracture Properties of Wrought Stainless Steels.0 M sodium phosphate solution at pH 10 and at 10 and 40 Hz and in 1. fatigue crack growth rates were the same as for water at the same temperature." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. 10 20 30 40 60 80 100 Stress-intensity factor range.5 L.pH2tol0. ~K.... 25°C pH 10. water. Type 403 Stainless Steel: Effect of Environment on Fatigue Crack Growth Rate ~K. Campbell. William W...5.0 M sodium chloride solutions were made with the solutions at pH levels of 2. P 147 . 100°C 10..--L---L----I---L. Source: J..-----.. Metals Park OH. Fatigue crack growth rates in 0. Gerberich and John H. At 100 °C (212 OF). 100°C pH 10. E.J. and 10 and with an open circuit. --' Exposure to water at 25°C (77 OF) resulted in intermediate crack growth rates between those in air and those in water at 100 DC.-_ _---L_ _L.L....K values above 20 MPa· m 1/2 (18 ksi· in.------. American Society for Metals.01 M sodium chloride at pH 10 and 100°C were the same as those in water at 100 "C..-. Tests in the 0.. ksi • in. 1/ 2 10 20 30 40 60 80 10-3 rr----.--..01 M (molar) and 1. For tests in the 1.pH 7. Underwood. m 1/2 Fatigue crack growth rates in type 403 stainless steelin air.6 In 1M NaCI solution "1:J .. the fatigue crack growth rates were higher than at 40 Hz at 6. 25°C -.J. At lower cyclic frequencies.226 8-30. 100°C 10. 7. and aIM NaCI solution at 10 Hz and an R ratio of 0.. as shown on a different scale in the above graph. I / I / II / / 10-4 / / Air 10-5 Type 403 In H20 ~ III z pH 7. Eds. fatigue crack growth rates in 1.. 1/2)..0 M sodium silicate at pH 10 and at 10 Hz were practically the same as those in air. Campbell.--.... 1982.0 M sodium chloride solution at 100°C (212 OF) (see graph)... James E.

Type 403 Modified Stainless Steel: Scatter of Fatigue Crack Growth Rates 227 '--_-'-_-'-. tested at 10 Hz and an R ratio of 0.3 MPa (1200 psi) Upper boundaries of fatigue crack growth rate scatter bands for three heats oftype 403 modified stainless steel in the heat treated condition... MPa ._ . K.. ll.. E...." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. American Society for Metals. "Fracture Properties of Wrought Stainless Steels.' ...L. P 145 . Eds. James E.. m 1/2 .' - ----I 10-6 20 40 60 100 200 300 Stress-intensity factor range.. 1982... Underwood.w. Campbell.083 or 0._ _. Metals Park OH..L. William W.L.067.8-31.... Campbell.. The curves representing the upper boundaries of the scatter bands of the fatigue crack growth rate data indicate that there is some heat-to-heat variation in fatigue crack growth rate properties for these heats... Gerberich and John H. exposure at 27 1°C (520 OF)in distilled water at a pressure of 8.. Furthermore.Heat 484 in room temperature air Heat 634 in room temperature air Heat 933 in room temperature air Heat 933 in 271°C (520°F) distilled water at 8....LJ. Source: J....3 MPa (1200 psi) increased the fatigue crack growth rate.

Results oftests with specimens in the 4..5% NaCl solution at room and elevated temperatues. The material for these tests apparently had been heat treated to a yield strength of approximately 827 MPa (120 ksi)." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. Each specimen was rotated at 600 cycles per minute (10 Hz) while at constant load with the salt water solution flowing over the notched section. The specimens were one inch in diameter in the test sections. Calculations for maximum stress-intensity factors were based on equations for solid round bars subjected to bending loads.228 8-32.5% solution.L. "Fracture Properties of Wrought Stainless Steels. Campbell. Type 422 stainless steel contains nickel. Campbell. Gerberich and John H. l.:. molybdenum.. 1/2 10 20 40 60 80 100 200 71°e (160°F) 57°e (135°F) 10- 4 I z a> ... 1982. P 150 . Source: J. and tungsten. 10Hz. Eds. and 71°C (160 OF) are shown above. Underwood.:::: ~ "tl c ~ Type 422 ---L_---l_. The effects of sodium chloride solutions and elevated-temperature exposure on fatigue crack growth rates were determined by Eisenstadt and Rajan in tests of notched round rotating beam specimens in which the numbers of test cycles were marked by minor stress interruptions that produced marking rings. William W..L... Metals Park OH. Type 422 Stainless Steel: Fatigue Crack Growth Rates in Precracked Specimens Koff' ksi • in. Increasing the temperature of the solution substantially increased the fatigue crack growth rates..J ~_---l 10-5 20 200 Fatigue crack growth rates in precracked round rotating beam specimens of type 422 stainless steel in 4.. E.. American Society for Metals.. James E.. Tests with several concentrations of salt solution indicated that the maximum corrosive effect was obtained with the 4. and an R ratio of-I. as well as 12%chromium to improve properties.. 57°C (135 OF).5%sodium chloride solution at room temperature...

..--------------------------... UNNOTCHED 229 x 80- o A • o CANTILEVER TESTS FOR 108 CYCLES 60r- 1-----------.030·inch root radius x o A o .x o x 0% delta ferrite • 5% delta ferrite 15/16% delta ferrite A 20% delta ferrite o 20 . Elihu F.. Source: J. Briggs and T. 12 to 30 inches in diameter.. ksi 1900 F (l 040 C)oil quench + 1200/1400 F(650/760 C) 100. American Society for Metals. Z. Ed. p 121 . Metals Park OH.1 60· notch 0.. Parker. VIBRATING NOTCHED 401Kt = 2.8-33. Bradley. including effects of varying amounts of delta ferrite. I 20 I I I 100 ksi 40 60 80 TRANSVERSE FATIGUE STRENGTH Transverse fatigue strength as related to longitudinal fatigue strength for type 422 stainless steel. 1979. D. "The Super 12%Cr Steels.. Type 422 Stainless Steel: Fatigue Strength-Longitudinal vs Transverse CLASS II (Crucible 422) turbine-wheel forgings." in Source Book on Materials for Elevated-Temperature Applications.

. P 121 .230 8-34." in Source Book on Materials for Elevated-Temperature Applications. D.... Briggs and T.. Type 422 Stainless Steel: Effect of Temperature on Fatigue Strength CLASS II (Crucible 422) (VacuumMelted) 15% delta ferrite 1800 F (980 C)oil quench + tempered to a tensile strength of 131/138 ksi ksi 140 130 f120 .. 1979. Bradley. Z. American Society for Metals.......I 101 700 F(370 C) . "The Super 12% Cr Steels... showing effect of temperature on fatigue strength.""'-ee_ .- IJJ a:: 100 "- J- (/) 80 70 90 '.. room temperature ROTATING CANTILEVER-BEAM TESTS I 105 • -. Elihu F. Source: J.. Ed.. • 108 601()~ I 106 CYCLES SON curves for vacuum-melted type 422 stainless steel with 15% delta ferrite. Parker..(/) (/) 110 . Metals Park OH.

. Bradley. D. .. Briggs and T.Tempered 1150 F (620 C) en 80 -~ "-. a lr-+ 0- 0- 70 'LONGITUDINAL ELECTROMAGNETIC CANTILEVER-BEAM TESTS 107 CYCLES S-N curves for type 422 stainless steel. "The Super 12%Cr Steels." in Source Book on Materials (or Elevated-Temperature Applications.Tempered 1200 F(650 C) --. . Type 422 Stainless Steel: Effects of Delta Ferrite on Fatigue Strength CLASS II (Crucible 422) 3/ -inch-diameter bar stock or 3/ -inch-thick plate 4 4 1900 F (1 040 C)oil quench ksi 110' 231 100 en en LLI 0: I- 90 \ ~. Source: J.-v.Tempered 1150 F ~ 15~20 '~~v_ % DeltaFerrite TensileStrength. V.. 140 5% Ferrite. which demonstrate the adverse effects of lOS delta ferrite on fatigue strength. 00% Ferrite.. p 121 . Parker..8-35.. ksi 155 160 .. American Society for Metals. Z. Elihu F.. 1979. Ed.. Metals Park OH.. .: 15/20% Ferrite. 0 (620 C) o 0. ..

ksi • in. P 156 . however. "Fracture Properties of Wrought Stainless Steels. Results of fatigue crack growth rate tests on specimens of 17-4PH stainless steel under comparable conditions are presented here. E.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:!. Source: J. 5 10.1 10-1 10 100 Stress-intensity lactor range. Underwood.5% NaCI solution.'/2 10 100 10-2 I . Metals Park OH.o.K.~ I~ I I 'I '---Hll00 . 17-4 PH Stainless Steel: Fatigue Crack Growth Rates in Air vs Salt Solution ." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. William W.5 17·4 PH '--'-'. had fatigue crack growth rates only slightly higher than those of comparable specimens tested in air with continuous cycling. Gerberich and John H.Kvalues...232 8-36. u.3 il! '" ~ u ".. Specimens in the H II 00 condition tested in a salt solution with a one-minute holding period.1.K. Campbell. . American Society for Metals." .05 1 min hold Salt soln 10-4 10. Campbell.. 1982.o. MPa • m1/2 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. Those specimens that were tested in the HI 050 condition at a stress ratio of 0. James E. Eds. R =0. 10-6 10.4 .

8-37. 15-5 PH Stainless Steel: Fatigue Crack Growth Rates in Air vs Salt Solution
LlK, ksi • in. 1/2 10 100

233

)
'I Hl~50 ~I // R - 0.67 I:/, 1 min hold I 1//
'li

"

.
~

RT.ir

E
E

10- 3


~
."

..

!l'
.t::

e
10-'

Hll00--.......,' R = 0.05 ~ 1 min hold " I I 'I Salt soln

I

1

,,1

I/Il

l,' l'
---Hl050 R = 0.05 10Hz Sine wave RT air

.. .§.
u

e '" u "" t!
TO

i

,

I'

~

VI 'I1 ,r--HllOO R = 0.05
1 min hold RT air

u,

I
10-5
H 1100 :----""'/' R = 0.05 10 Hz I Sine wave RTair

10- 6

I
15·5 PH
...1...

1-

--L.

.J1O-

7

10

100

Stress·intensity factor range, LlK,MPa • m 1/2

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

234

8-38. PH 13-8 Mo Stainless Steel: Fatigue Crack Growth Rates at Room Temperature
~K. ksi . in.1/2

10

20

40

60 80100

10-2

200 4 X 10-4

Gl

U

---e
Z
"tl

> u E E

10- 4
Gl

.s:

.....
u
Gl

l'O

U

Gl'

..

l'O

10- 3

e
Cl

;:

--.~
"tl

> U

Z'

l'O "tJ

~

E u
Cl

:::l

';; l'O u..

PH 13-8 Mo H1100

10-5

10- 4
20 40 60 80 100 200 Stress-intensity factor range, ~K. MPa • m 1/2 Fatigue crack growth rates in cantilever beam specimens of PH 13-8 Mo (HII00) stainless steel, at L-T orientation, 0.17 Hz, and an R ratio of 0, in room temperature air. Data are based on the stress-intensity-factor range as shown.

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

8-39. PH 13-8 Mo Stainless Steel: Fatigue Crack Growth Rates in Air and Sump Tank Water
toK, ksi • in. 1/2
6 8 10

235

20

40

60

10-2
R = 0.3 STW
L-T

10- 4 10- 3

10- 4

10- 6

PH 13-8 Mo H1000

6

8 10

20

Stress-intensity factor range, toK, MPa • m 1/2 Fatigue crack growth rates in compact specimens of PH 13-8 Mo stainless steel in the HI 000 condition for room temperature tests at I Hz, R ratios of 0.08 and 0.3, L-T and T-L orientations, in low-humidity air (LHA) or sump tank water (STW).

Effects of increasing the load ratio,R, on fatigue crack growth rates in low humidity air (LHA) in sump tank residue water (STW) for specimens of PH 13-8 Mo (H 1000)are shown above. The highest fatigue crack growth rates in this series were obtained on specimens tested at an R ratio of 0.3 in STW. Increasing the load ratio from 0.08 to 0.3 had a marked effect on the growth rates.

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

236

8-40. PH 13-8 Mo Stainless Steel: Fatigue Crack Growth Rates at Subzero Temperatures
~K,

ksi . in. 1/ 2

4

6

8 10

20

40

60

100
10- 4

10- 5

> u '-:
.!:
1:1 "nl 1:1

u

Q)

z

10-6

PH 13-8 Mo H1000
10- 7

4

6

8 10

20

40

60

100

Stress-intensity factor range, ~K. MPa . m1/2
Fatigue crack growth rate scatter band for compact specimens from rolled bar and extrusions of PH 13-8 Mo stainless steel in the HIOOO condition for room temperature tests in low-humidity air and in sump tank water at frequencies of 1 and 6 Hz and anR ratio of 0.08 for L-T and T-L orientations.

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 -54°C (-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

8-41. PH 13-8 Mo Stainless Steel: Constant-Life Fatigue Diagram
Minimum stress, ksi
-150 1600 1400 1200 -100

237

o

+50

100

150

200

200

'" a.. :2 1000
",'

150

Jl

'" ~ ... '"
:l

'" :2

'x

E E

800 600 400 200
S)<::l
(c)

'li~
.,-,; 'b~

S)<::l l'l'b

\e"""

100

~fl,

.S x '" :2

:l

E

Axial fatigue Unnotched specimens Longitudinal and
transverse or-

50

ientations

o '--_-'--_...l-__....J...._----L_--''------''L-_-'--_-L.._--'--_----L_--'_ _-'----_...l-_-'
-1200 -1000 -800 -600 -400 -200 0 +200 400 600 BOO 1000 1200 1400 1600

Minimum stress, MPa
Constant-life fatigue diagram for PH 13-8 Mo stainless steel, condition HI ODD.

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

238

8-42. Types 600 and 329 Stainless Steel: S-N Curves for Two Processing Methods
1000 800 600
o,

I
I I I I

::2:

'"

.~

600 Steel (STAMP)-+--+--I

400

.L

I

IV ••

Vi

Ul

'" e
200

I I I. AISI 329 (electroslag remelted)

I

.

0.57
'\1'

I

100 80 60 40
'iii
-'"
",'

0.51

Ul
20

'" e

100 10'

10' Cycles to failure Tensile strength, MPa(ksl) 760 (110) 630 (91) Yieldstrength (0.2% olTset), MPa (ksi) 600 (87) 500 (73)

10'

10'

Steel STAMP 600 Electroslag-remelted 329

Mechonlcal properties Elongationin Reduction 50 mm (2 ln.), % in area, % 26 29 54 65

Impact energy,

Fatigue
strength,

J (R·lb)
25 (18) 35 (25)

MPa (ksl) 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

8-43. Grade 21-6-9 Stainless Steel: Effect of Temperature on Fatigue Crack Growth Rates
AK, ksi • in.1/2

239

10 21-6-9
> u ...... E E
"t:l ...... III "t:l

20

40

60

I

10- 4

u

Q)

10- 3

2:

22 to -196°C / (72 to -320°F) / -269°C (-452° F)

or.
0

oj ... ..
III

/
10- 4

Q)

U

s ..
Cl

~

III

u u

.

/ /
/
10 20

10- 5

u ......

>

,E
2:
"t:l

"t:l ...... III

Q)

'+:i
III

Cl

:l

u.

I
10- 6
40 60 100

Stress-intensity factor range, AK, MPa • m 1/2
Fatigue crack growth rates in specimens of annealed 21-6-9 stainless steel at 22, -196 and - 269 0 C (72, - 320 and -452 0 F), 20 and 28 Hz, and an R ratio of 0,1.

Similar tests made with specimens of 22-13-5 stainless steel showed fatigue crack growth rates that were nearly the same as shown here for 21-6-9.

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

240

8-44. Kromarc 58 Stainless Steel: Effect of Cryogenic Temperatures on Weldments
dK, ksi • in. 1/ 2

20

40

60

100

200

10- 3

10- 4

I I
20 40

Kromarc 58
Base metal Weld metal

...... -::':::--_--:'-_....L.._----''--_ _.L------' _

10- 6

60

100

200

Stress-intensity factor range, dK, MPa . m 1/2
Fatigue crack growth rates for solution treated Kromarc 58 base metal in air at room temperature, and base metal and weld metal at-269°C (-452 OF) in liquid helium, at 10 Hz and anR ratio of 0.1.

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 -269°C (-452 OF).Fatigue crack growth rate data for the base metal at room temperature and at - 269°C and for the weld metal at -269°C 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

8-45. Pyromet 538 Stainless Steel: Effects of Welding Methods on Fatigue Crack Growth Rates
AK, ksi . in. 1/ 2

241

10

20

30 40

60 80 100

200

10- 4

U

Gl

u ......

> E E

10- 3

",-SMAW 24 and -269°C (75 and -452° F)


't:l ...... 't:l

I

I
> u ......

'" ..r:
.~

!l
~
0
~

'"

~

E u
Gl

u

'"
10- 4

I

I

/"GTAW 24°C (75°F)

u

Gl

10- 5 z·
~
't:l

.~

'"

:l

.;:;

u.. '"

'"
Pyromet 538 welds

20

30 40

60 80100

200

Stress-intensity factor range, AK, MPa . m 1/2
Fatigue crack growth rates in weld metal in Pyromet 538 stainless steel at room temperature and -269°C (-452 OF) and at

10Hz.

The base metal was solution annealed prior to welding. One set of welds was made by the gas tungsten arc welding (GT AW) process with 21-6-9 filler wire, and the other was made by the shielded metal arc welding (SMA W) process with IN 182 covered electrodes. Results of these tests are summarized in the graph above. Specimens with SMA W welds had the same fatigue crack growth rates at room temperature and at -269 "C (-452 ° F). Specimens welded by the GT AW process had higher crack growth rates at -269 °C than at room temperature. Examination of the microstructures near the fracture surfaces for the specimens tested at - 269 ° C showed tha t there was 6 to 7% delta ferrite (produced by welding) in the weld metals along with induced martensite. The SMAW weld metal was fully austenitic.

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

242

8-46. Duplex Stainless Steel KCR 171: Corrosion Fatigue

400

KCR 171 Whit. wat.r
pH ·4. IS

T • ISO·C
300
CI
Q.

~

_ 0 ....... 0 ...

(\/

200

l;
~

o
100

6 Hz

t:>

o

o

20 Hz 100 Hz
167Hz

Nf

eyel ..

Rotating bending S-Ntests were carried out in 50°C (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

=::j 'iii If) '" e ...- --'- ---L L. p 451 ..r .. and 300 Maraging Steel: S-N Curves for Smooth and Notched Specimens 1500r . Grades 200... Improved fatigue properties can be obtained by shot peening and by nitriding. and I8Ni(300) grades are summarized in the S-N curves shown above... Metals Park OH. Volume I. Fatigue crack growth rates in maraging steels obey the da] dN= (t1K)m relationship commonly observed in steels and are similar to those of conventional steels.- 104 105 106 Number of stress cycles 107 0 8 10 ~ Fatigue properties of maraging steels are comparable to those of other high-strength steels... 250. v. 500 o L... 243 css c.. Source: Metals Handbook..----~----_. 9th Edition...:.l '" ~ 1000 _ _-+- .9-1. CZZl Em 18Ni(300) 18Nj(250) 18Ni(200) 200 150 . 1978. Smooth-bar and notched-bar fatigue properties for I8Ni(200). I8Ni(250). American Society for Metals.-. Properties and Selection: Irons and Steels...

LJ..c \ ~ ELASTIC ...LUJ. Inc.....J. Source: Marc Andre Meyers and Krishan Kumar Chawla. Grade 300 Maraging Steel: Fatigue Life in Terms of Total Strain 1 18% Ni morolling (300) 10 -1 .ll.: TOTAL --... p 700 ..1." Prentice-Hall...244 9·2.uJJ... -)-Trrr~ ...U.--L.. "Mechanical Metallurgy: Principles and Applications.. An actual example for this method of determining fatigue life is presented above for a maraging steel..J.h \ PLASTIC~\ 10 -3 \ L-J.. 1984. A.lJ 1 10 Superposition of elastic and plastic curves gives fatigue life in terms of total strain.LJ. \ • .l.JLlJJ.--I.LWL--I..JWJJ. Englewood Cliffs NJ.LJ.LOAD CONTROL /(J'f/E 10 -2 -- \ '\ ~\ .-l-u....U-l...l.LI..JJJ.

Walter has shown that the fatigue properties of irons are highly dependent on volume of graphite and its morphology and distribution. the modulus at very small strains. D. Wene.1 I 10 MEAN FATIGUE LIMIT VERSUS (BHN) Fatigue of cast irons as a function of structure-sensitive parameters: Bhn. Sharp-edged flakes are greater stress raisers than rounded-edge flakes and spheroids. ofthe graphite volume. Metals Park OH. D." in Fatigue and Microstructure. . Source: D..: • . He was able to reduce these factors to some easily measurable parameters. Since the graphite present detracts from the matrix load-carrying area. is controlled mostly by the graphite morphology and to some degree by the graphite volume. as well as the matrix structure. Bhn is largely a measure of the matrix hardness and. American Society for Metals.J ~ ·S~ • "B" BAR BAR . and Bhn.10-1. which gave good correlation with fatigue properties over a rather wide range of irons (see graph). H. 1979. the more graphite. the better the fatigue performance. is controlled mostly by the volume of free graphite and to some degree by the graphite shape. thus the higher the D. thus the higher the Bhn. Fatigue of Cast Irons as a Function of Structure-Sensitive Parameters 245 50 20 iii a: 10 § I&. Eo. p 86 . These easily measured properties are put to good use in industry as specification means and process-control criteria. to some degree. since they are measures offatigue-related properties. "Fatigue in Machines and Structures-Ground Vehicles. the damping capacity. Eo. the higher the stress on the remaining matrix-thus lower fatigue performance.. It is reasonable that these parameters relate to fatigue performance. elastic modulus (Eo) and damping capacity (D). Breen and E. the poorer the fatigue performance. M.

1. Volume I.oC • Notched specimens. stress based on net area .84 C. of 400 200 400 600 Fatigue limit 800 1000 I Fatigue life I 350 - .31 Cr. (Ref 5) Typical fatigue life for as-cast gray iron of the above composition (left). _____0 17 - '- ~ 100 - -.07 P. Gray Iron: Fatigue Life. <.37 Cu.40 - 300 250 0IV :2 ::.12 S. Metals Park OH.~ - 20 50 100 Number of cycles to failure 400 500 " 1o ~ 600 o o Notched specimens • Un notched specimens Composition: 2. ". Source: Metals Handbook.-% . American Society for Metals. 0. Properties and Selection: Irons and Steels. 0. -.52 Si. .' 200 cil e 150 <.50 . . 0. and Fatigue Limit as a Function of Temperature Temperature.. 1. 0. 0.246 10-2." -~ 200 300 Temperature. P 21 .05 Mn. 9th Edition. Effect of temperature on fatigue limit for the same gray iron (right). 1978.20 Ni.

N curve and relates the maximum applied stress to the logarithm of the number of cycles for failure.-(f) en ~ 2 • en (f) L. The maximum stress that will allow this number of cycles is established as the endurance limit.10-3. and conversely a larger number of stress cycles can occur at a lower maximum stress level before a fatigue crack is initiated.. <J) (f) . 1% Ni.. Charles F. Gray Iron: S-N Curves for Unalloyed vs Alloyed 247 34 32 30 28 26 24 22 20 18 10' '\. 0 0 0 (f) ... Source: Iron Castings Handbook.. 1... As the crack progresses it increases the stress concentration. Inc. 0. It is always a brittle type of fracture even when occurring in ductile metals.. ________________. When the number of cycles without failure exceeds ten million.4% Cr.. Walton. Iron Castings Society. • ...0% Mn o 0 230 220 210 200 190 180 170 o. Ed. """. The number of stress applications that will induce a fatigue failure is less at higher maximum stress values.6% Mo. L. A plot of this relation for a metal is called an S. 1981. complete failure occurs as it would under an excessive steady stress. Two typical S-N curves for a plain and alloy high-strength gray iron are presented above.. the endurance life is considered infinite for bodycentered-cubic ferrous metals. 0. Plain Iron • Alloy Iron. p 246 .1. <J) (f) .. When the cr~ss section of the remaining metal becomes insufficient to support the maximum load. Endurance or ~ Fatigue Limit .. <U a. or the fatigue strength or fatigue limit.../-----------1----0-0 Knee 160 150 140 130 -L - 10 5 10 6 10 7 Cycles To Failure A fatigue crack starts in an area of high stress concentration after a large number of loading cycles. and the rate of propagation under the cyclic loading increases._--_.

.500 DSt (100 MPal 15 14 10· 10· 10 7 100 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. Walton..920 Water 14. Iron Castings Society. . . .. only borax was not completely effective for the pearlitic irons. n 0 0 0 19 18 17 Fatigue Strength 18. Environment Fatigue strength psi MPa 124 100 39 108 124 124 124 124 Fatigue strength reduction factor Air 17. . . 1981. . Ed.00 1. . .f) Q) . .00 1.. p 255 .248 21 20 - 10-4. . additional tests were made with a demineralized water spray and various known corrosion inhibitors. 17.. Annealed ferritic gray irons were similarly affected by the sodium chloride solution. and a spray of three-percent sodium chloride solution.20 1. . . Charles F. Source: Iron Castings Handbook.. 5.920 3% sodium carbonate 17.23 3..920 1. +-' Q) (. . . The corrosion fatigue program involved testing in air.. a spray of demineralized water. Gray Iron: Effect of Environment 140 'ea.25% potassium chromate 17.560 3% sodium chloride..920 0. Of the various alkaline inhibitors and soluble oils investigated.680 3% "Sobenite"* 17.f) en (JJ .00 * "Sobenite" is a mixture oj 10paris sodium benzoate to 1 part sodium nitrite. 16 +-' 110 _ 14. ..000 PSI 130 Q.600 1% borax 15. The S-N curves and table above indicate that both the demineralized water and three-percent sodium chloride sprays reduced the fatigue strength of a pearlitic gray iron. Inc.00 1. co 1124 MPa) ~ 120 en (JJ (.920 3% soluble oil.14 1.

ijit----~ a.- .P 251 .. Walton... 1000 psi A modified Goodman diagram relates the endurance limit to an allowable working stress when it is superimposed on a steady stress. Source: Iron Castings Handbook.l...... alternating stress is superimposed on a steady stress and requires special consideration... 1981. .... (f) Ul 50 Cl :il c S 0 0 -50 -101'.. 100 o o o CL eo ~ iii ~ . Charles F...J o 10 20 30 Mean Stress... Class 30 Gray Iron: Modified Goodman Diagram Mean Stress..10-5. of which a modified form is shown here......L.......<... Ed.. In many engineering applications. A method of relating the effect of the combined static and alternating stresses on the endurance limit has been developed into the Goodman diagram. Iron Castings Society. Inc.+.. MPa 30 ~---"""T---r--""---T""""'---"""'"2 249 50 100 150 200 200 Area of Finite Life 150 20 '(ij I--------+---~.

1981. p 250 . where dol dN is the crack growth per cycle and 6. If the initial flaw size can be determined from experience or by utilizing nondestructive inspection and the critical flaw size calculated using the fracture toughness value K. 3300 Ib (910 kg) Ib (1000 kg) Ib (1130 kg) Ib (1130 kg) Ib (1360 kg) Ib (1500 kg) (9 . Class 30 Gray Iron: Fatigue Crack Growth Rates Stress Intensity Factor Range. Charles F. For such specimens.. Ed. Iron Castings Society. The endurance limit approach to design utilizes fatigue data taken on smooth.4 ~ '" s: a: s: (9 e ~ o i3 '" U • 2000 o 2200 • 2500 o 2500 6 3000 .. Inc.:>L e ~ U • 10 20 30 40 60 80100 Stress Intensity Factor Range. fatigue crack initiation may take 80 to 90 percent of the total lifetime while crack growth is only 10 to 20 percent of the lifetime..~ c Z ~ "0 z Band for Wrought Ferrite-Pearlite Steels Load Range 10-8 a: 2i III ~ '" ~ "0 III 10. D.c' then crack growth rate data may be used to calculate the number of cycles required to grow a crack from an initial size to a critical size where final fracture occurs.K in MPavm 10 20 40 60 100 o >s: Q) ~ o c o >~ E E 10-' Q) .K is the stress intensity range. Source: Iron Castings Handbook. defect-free test specimens.K in ksi VIiicli Fatigue crack growth rate. These results are presented in the above chart for a class 30 gray iron. D.. Walton.250 10-6. Only limited fatigue crack growth rate data are available on cast irons. Such flaws allow fatigue cracks to initiate in a relatively small number of cycles so that the lifetime of the component depends principally on the crack growth rate.

000 psi (400 MPa) / 50 (j) "0 Q) Q) ~ en en 300 40 +-' <U li} 200 a:: '0 . Ed. 100 . p 253 .~ E ~ 24. Charles F. 1981. Walton. ::J ~ Mean Stress -100 Torsional fatigue strength for three levels oftensile strength with various mean stresses.000 psi (166 MPa) 14. Source: Iron Castings Handbook. Inc. Iron Castings Society.J o c ~ "0 <U Q..000 psi (97 MPa) 1000 psi 0 Q. Gray Irons: Torsional Fatigue for Various Tensile Strength Values MPa 1000 psi 60 400 251 58.10-7...

. OJ U lO L. 1981.63 0.0 .I: C Q) c . pp 288. 1.74 1. 0 0.59 0.75 -. OJ 20 U L...49 0. Ed. Inc.43 3.68 1.59 0.5 10 U .65 1.87 0..50 0.252 40 10-8. Gray Irons: Torsional Fatigue Data for Five Different Compositions r---------------------. 289 .I: C Q) ...._-'-----'-----'300 400 500 600 800 1000 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).25 Q) .45 3.. Iron compositions are as follows: Composition. Walton.J ~ 0..44 3..38 0.077 Source: Iron Castings Handbook. Charles F.21 0. Iron Castings Society. % Iron C Si Mn Cr Mo Ni Cu Sn A B C 0 E 3.60 0.45 3.I: U E 30 E .58 0.50 en 1...30 0.66 0..57 0.J ~ lO U I- o lO 0.30 0.69 1.39 0.58 0.25 I----l_ _----l --' lO O'--__ '---""'--_-'----_-'-----I.97 0. -.43 1. 1..30 0.

e.. Inc.!: ... Iron Castings Society.5% Mo D o c A 2% AI. Gray Irons: Thermal Fatigue-Effect of Aluminum Additions ...---...t----+-----. p 434 .4% carbon and 2% aluminum was highly resistant to thermal crack propagation. This graph shows results of a thermal fatigue test in which notched disc specimens were alternately heated to 800 OF (425°C) and cooled to 200 OF (95°C) in two fluid beds..>4F---t----+------l .. Source: Iron Castings Handbook.-----r----r----.10-9.----r----. 0. Ed.. Charles F.6% Si (65% Ferrite) <> 2% Si ..175 D --+---t"7""'==------j4 E E s: +-' g> . o . 2% Si (4% Ferrite) .050 I--------if------""--+---=-'f----+----+---'::=--'""=i 2 4 6 8 10 12 Number of Thermal Cycles (X 1000) Thermal fatigue resistance of different alloyed gray irons.----.100 t---+----+---+---..5% Cu..075 t---+----+---.J ~ 2 U ro . Walton.".150 o 0..5 253 Gray Iron 3% AI..-=-_ _--1 I 3 Q) . demonstrating that a peariitic gray iron containing 3...125 .. 0. 1981.

.: .._ _----l_ _----J 300 400 500 0 600 Number of Cycles The depth of cracks resulting from the thermal cycling of gray irons between 860 of (460°C) and room temperature.. Q) o . supplies a less than desirable influence on thermal fatigue cracking. Gray Irons: Thermal Fatigue-Effect of Chromium and Molybdenum Additions 4 3.15 '0 . by adding relatively large quantities of molybdenum and copper.. Ed.. In this case.10 . as shown above.-- .. it must be remembered that this improvement is related to and dependent on the temperature cycle and base iron composition. V /~ ... 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.L- J. en <Il U '0 s: o Q.05 o '--__-L..254 10-10. E ~ I . ~ .c o c en o ~ L. ..' o 100 200 ..._ _.. Charles F.10 CrNiMo / 2 .-:. Inc.. ' U .. CrNiMo0'" CrMo( 0 4 en Q) ... ~-.. L. .~~ ~. Iron Casings Society. . Walton...c a.-1--.2-3.. It has also been indicated that the development of an acicular matrix structure. Q) 3.t.15 Bridge Cracked I 2 ~ »-> Unalloyed ____~ CrNi _ '?"' . Alloying with molybdenum and chromium provided superior thermal fatigue resistance compared with irons that contained other alloying additions or no alloying at all.7% C 3 1.::.:-..-' . However.05 ! E . Source: Iron Castings Handbook. 1981. . p 288 .3% C 3 I Bridge Cracked I --.

the number and size of thermal fatigue cracks become larger. and for temperatures above 900 OF(500 "C). a compromise must be made in selecting the most appropriate metal for each type of service.and elevated-temperature strength on the thermal fatigue resistance of irons having similar carbon equivalents.• Iron Castings Society. resistance to oxidation and structural change. p 285 . Because some of these properties are in opposition. For good resistance to thermal fatigue. As the maximum temperature to which the gray iron is cycled and number of cycles increase. Walton. Ed .10-11. Gray Irons: Thermal Fatigue-Room Temperature and 540°C (1000 OF) Tensile Strength. Those irons with higher room and elevated-temperature tensile strengths (achieved by alloying) generally display higher thermal fatigue strength.• 1981.-_ _---' --'- -'- '---' 10 20 30 40 50 Tensile Strength. matrix structures. gray irons should have high thermal conductivity. and elastic moduli. Inc.D E :J Z 500 o Lo ' . high strength at both room and elevated temperatures. 1000 psi Curves showing relation between the number of thermal cycles for cracking and tensile strength at room temperature and 1000 ° F (540 o q . . The above curves illustrate the influence of room. a low modulus of elasticity. MPa 255 o 100 200 300 1500 <J) Q) o 0 >1000 '0 Q. Charles F. Source: Iron Castings Handbook. thermal conductivities.

38 Mn 0. Iron Castings Society.0--------=-.37 AI 1..--------.---.. Gray Irons: Thermal Fatigue Properties-Comparisons With Ductile Cast Iron and Carbon Steel Maximum Cycle Temperature. Ed.23 0.L------~-----~~ 700 800 900 Maximum Cycle Temperature.43 3.---.78 0.50 Si 2. Walton. 1981.20 0.51 Cu C D Source: Iron Castings Handbook.. Gray 1 6~0. Compositions of the four gray irons are as follows: Analysis. Charles F. % Iron A B C 3.22 0..88 0. Inc. and carbon steel.60 2.24 0.49 3.22 0.83 Cr 0.30 Mo 0.. 10' (J z d> c ~ o Steel Ductile Iron B <f) u <Il (J >- '0 . 287 .0 :J Q> 10' E Z AI 1m".21 Sn 2.256 10-12.r--------.48 3.. . F 1o'. 1200 1400 1600 :.37 2.77 Other 0... pp 286.84 0.37 0. ductile iron.32 0. C The above curves show the variation of the number of cycles to cracking with the maximum temperature of the cycles for gray iron.

030 0. Inc.05 0.015 0. Iron Castings Society. Ed.10-13.. Cast Irons: Thermal Fatigue Properties for Six Grades 257 Ferritic Compacted Graphite Pearlitic Compacted Graphite Ferritic Ductile Pearlitic Ductile Alloyed Ductile 6 8 10' 2 4 6 8 10' Number of Cycles The number of thermal cycles required to produce thermal fatigue cracking in cast irons.13 0.48 %5 2.96 3.54 Cu 1. 396 .47Cu %Mg Alloys 0.05 0.52 3. Analysis Class 35 Gray Iron Ferritic Compacted Graphite Pearlitic Compacted Graphite Ferritic Ductile Pearlitic Ductile Alloyed Ferritic Ductile %C 2.55 2.25 2.78 0.60 3.030 1.84 %Mn 0.06 0.61 2.50 0. Charles F.030 0.25 0.31 %P 0.015 0. 1981. pp 393.67 3.40 0. Compositions are tabulated below.12Cr 0.02Mo Source: Iron Castings Handbook.07 0..52 3.34 4.90 2. Walton.07 0.05 0.

Charles F. p 341 ..4 ..3 • j( . the increase is less than proportional. Source: Iron Castings Handbook.258 10-14. MPa 300 ~o o . Ductile Iron: Effect of Microstructure on Endurance Ratio-Tensile Strength Relationship Tensile Strength. Iron Castings Society.. 140 • • •• 160 60 100 Tensile Strength." II: Q) J. as illustrated above.~. 500 I 700 I 900 I 1100 I 1300 0.o~ "x~x: x o . 0.. Walton. 1981. but as with other ferrous metals. 0000 'U W c ~K \ · Tempered Martensite <. ferritic irons is different from that of the irons with a matrix of pearlite or tempered martensite.. ~. Ed. Inc.~ "\~t x Pearlitic '" g ~ ::J 0. 1000 psi In general the fatigue limit for ductile iron increases with tensile strength.5 -o\'ox> Ferritic... The relation between the tensile strength and the endurance ratio for the annealed.

ksi 259 50 75 100 125 150 175 0..I _ .10-15. This is indicated in the graph above by the shallower slope for martensite. MP.. Endurance ratio is defined as endurance limit divided by tensile strength. ~ 1i e "0 O.. Properties and Selection: Irons and Steels.. Volume I. For tempered martensite ductile iron._ _ + .. regardless of structure. P 45 .4I------_+_---'''--------'''r__+-----.~ . The influence of tensile strength and structure on the endurance ratio of ductile iron is indicated in this graph..._ + _ .. Source: Metals Handbook.. there may be little value in specifying a higher-strength ductile iron for a structure that is prone to fatigue failure._ _ 1 ::> w <: 0.. Because the endurance ratio of ductile iron decreases as tensile strength increases. 1978. "ki~---T___+-----__1I_-----+__-----__+--_t .31------t------+-----t--------1r---'''''''=----+----j • 200 Tensile strength.o. Metals Park OH..-+--'''o. the improvement in fatigue strength due to an increase in tensile strength is greater than for pearlitic or ferritic structures.... 9th Edition. Ductile Iron: Effect of Microstructure on Endurance Ratio-Tensile Strength Relationship Tensile strength.. American Society for Metals.....51-----e-.

. 9th Edition. ~ 40 ~ O -6 2S I -.---------.t) 46° V-notched 3001-----------1r--------l--------l---------l . : il. Volume I. P 43 . Bottom: Similar to above except for as-cast pearlitic ductile iron. 2001--------=zilJ!l ISOI--------+----== 20 loo'f--------f-------+-------1-----------l Fatiguelift. for annealed ductile iron.. 1978.. Metals Park OH.-+----: ~ o i!' ~ e :.260 10-16..r. Source: Metals Handbook. 3S01-------1-------+------+---------1 ~v..---------..-----------. cyeles 106 400. All test specimens were V-notched (45°).r--------. tchod o F_I6G-40-18 ...-------.. Properties and Selection: Irons and Steels. including scatter bands. Using V-Notched Specimens 400'..-------. Ductile Iron: S-N Curves for Ferritic and Pearlitic Grades.--------... 2001-----------1r----" 1 20 ISOI---------1I--------f--------f---------j 1001-----------1r-------f-------1--------l 106 Fatiguelife.. 3S01-----------lI----------l--------l'---------l Pllrlitlc (80-66-06 •• -e.odl 3ool----------+------+-------t--------I ~ g. cycles Top:S-N curves.. American Society for Metals.-------.

e :lE ~ '0 ] 2501--------t-200 S J = ~ 1501--------t------+------+--------I 1001--------+------+------1-------.-------:--'. Volume I.. Ductile Iron: S-N Curves for Ferritic and Pearlitic Grades. go ~ '0 ii 2501--------. F.OOr--------. 1978. Bottom: Similar to lower graph on the opposite page.------' :lE t:. p 44 . 9th Edition.-------. Data in table pertain to graphs on this and the opposite page.. but here the specimens were unnotched.--------. Metals Park OH.10-17.... 3501---------f~ 3001--------. American Society for Metals.26 0. cYcles .00r-------. but here the specimens were unnotched.rillc: (60....24 1. 106 20 ~ = ~ : 1 Faligue life.43 0..67 Top: Similar to upper graph on the opposite page.cycles Grade Tenolle alrenl!h MPa kal Unnolched Endurance UmII Endurance MPa kat rallo Notched Endurance Umil MPa kat Endul' ance ratio Slreaa coneenlrallon lacier 60·40-18 80-55-06 480 680 70 99 205 275 30 40 0.4o-18 ann••IId) unnotc'*l 261 3501--------t------+------+--------I 3001--------t----::.r--------.40 125 165 18 24 0.-------+------.1--------1 ~ a J200 1501-------+------+--------1------.--------' 106 10' Faltgue lIle.. Source: Metals Handbook. 20 5:0'"'·-------'<--------'.----------. Using Unnotched Specimens .-----------.. 1001-------+------+--------1------. Properties and Selection: Irons and Steels.67 1.------.

-:-_-=-'-. Tensile Strength 1000 psi 55 61 72 87 102 MPa Min. pp 344. Ductile Iron: Fatigue Diagrams for Bending Stresses and Tension-Compression Stresses MPa 1000 psi 100 600 MPa 1000 psi 600 80 400 5 400 60 JOO ·E ::. Minimum properties of the irons are given in the table below. Ed. Yield Strength 1000 psi 36 41 51 61 72 MPa Min.~~I:--- 1001 psi 200 400 600 MPa -40 L-_-:-'-:-_-=-~_--:-'-. Charles F. Inc. Iron No.P... Elongation Percent 17 12 7 2 2 38 42 50 60 70 25 28 35 42 50 Source: Iron Castings Handbook.----_-. 40 ~ c 200 -0 c: W 5 ~ 200 c: o c Q) 20 5 100 -0 W o -20 -200 -40 -200 -JO -100 ~--:-f\P'A"':ld'-?-:!:---!c:---f::---. 345 and 346 . Walton.262 10-18. M 100 200 JOO 400 600 o Mean Stress Mean Stress Fatigue diagrams indicating endurance limits for five grades of ductile iron under bending stresses (left) and tension-compression stresses (right). Iron Castings Society. 1 2 3 4 5 Min. 1981.

ksi 263 n8 90 100 120 350 n60r--_":.... Ductile Iron: Effect of Surface ConditionsAs-Cast vs Polished Surface Tensile strength......_-.'::'_... 9th Edition... Volume I..._r----r-... MPa Tests made on 10... Metals Park OH.--~.:.....---f---+-----i 40 . 1978.. The endurance limit is much higher for the polished specimens than it is for the as-cast specimens. 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._':"'.6-mm (O. Properties and Selection: Irons and Steels......_ ___'__ _ __ ' 800 900 1000 1100 Tensile strength.. 50 g 45 a.~-._--.<:=' 200 1----+---+--7"'~_+_---' 25 150 _ __'___ _-'-_ _ 400 600 500 700 ~ __L....... American Society for Metals.... which have relatively rough surfaces....417-in..-. Fully reversed stress (R = -1).10-19. Source: Metals Handbook....' " ' .. p 45 .r:.... ::2: '" 3001---+----+---i--7"'q..) diameter specimens. _ _.

Q) 60 ::J '.. I z- ... 260 "0 . I / I . 0) <U 0) 300 .. 'E . 40 <U u.. ~ <U 'E ::i Q) .. •• • ~ 300 .. Q) . <U U.. ::J 0) ::i 300 '.. Charles F.400 0..- ::i 360 ::J '.. Walton. Brinell lD 200 a: 160 s 0 ~ 'iii 00 0 0 60 . 60 ::J 0) .. / . Ductile Iron: Fatigue Limit in Rotary Bending as Related to Hardness iii 60 0- 0 0 0 'E ::i . p 347 ....264 10-20.s c Q) 0) • • • • Hardness.. . • • • • • . Bottom: Relation between rotary bending fatigue limit and matrix hardness for ductile iron.:. Ed" Iron Castings Society. Inc" 1981...• .:.. .... c c 0) 0) "0 Q) lD 30 e a: 0 <U 20 100 ----I200 ... c 0) 40 r- "0 c Q) lD ~ A. Source: Iron Castings Handbook. <U ~ 360 'E Q) ...s "0 c Q) 260 lD z- a: 0 <U 30 100 I I 200 300 400 600 600 700 a: 0 <U Micro-Vickers Hardness Number Top: Relation between Brinell hardness and fatigue limit in rotary bending for ductile iron. u.~ u. • •• • -----•• 400 400 0..~ .:..

Iron Castings Society.Unnotched.. ~ <0 en (J) u:. p 348 . Ductile Iron: Effect of Rolling on Fatigue Characteristics 265 75 Rolling Pressures (Pounds) 500 70 406 580 768 65 60 '(ii 400 55 o. o 0 0 . Ed. Charles F. More than a 60% increase in the endurance limit was obtained with a rolling pressure that was insufficient to depress the surface a measurable amount. Inc .• 1981.. . Source: Iron Castings Handbook..----------. especially when this method is applied to stressed radii or notches.. Walton. ~ 50 "".10-21. The improvement in fatigue properties obtained by various rolling pressures on ductile iron is indicated above. " " 83 45 . " n. Unrolled "-----------j 300 40 35 30 Unrolled 200 10' 10' 10' 10' Number of Cycles of Stress (Log Scale) Fatigue strength of ductile iron can be increased substantially by cold working.

417" ~ 175 W (J) Ul ~ V-Notched I'" ! 0 o -i 0.. Walton...4'72" I 225 I ~ ~()OOO -"'0- 'iij 0..• 1981. 1 I i t .800 psi (454 MPa). The endurance ratio is 0. Charles F. - 150 125 1--0-000-- 16 10' 10' 10' 10' 100 Number of Cycles The unnotched and notched fatigue properties of an annealed ductile iron with a tensile strength of 65. Ductile Iron: Effect of-Notches on a 65. '\.700" W "1 0. p 341 . 32 <.750" ~ 0"". ~ 0 \ . Unnotched 28 200 n.266 10-22.67. eo o o o 0 '\.0" ~ 20 t 0.25 mm Root Had.250 ~ 0}17" 10. . Source: Iron Castings Handbook. 0 0 t 0. Inc.. Iron Castings Society..41 and the notch sensitivity ratio is 1. Ed.800-psiTensile-Strength Grade I 36 2.75"R I <.

p 349 ... 1981. t.:....:. 2700 Ib (1225 kg) • 2200 Ib (1000 kg) <...9 ..L-_L-..! e ~ U 10-' '--_..K in ksi yinch Fatigue crack growth rate of annealed ferritic ductile iron. t.L-.....K in MPa 20 267 v'rri Q) Q) <3 o >- <3 <.I.10-23.L.. compared with that of ferritic-pearlitic steels...... o >- E E Band for Wrought Ferrite-Pearlite Steels c '" <Ii m a: s: 10-' z ~ eo "0 2 a: s: '" o o (910-' o ~ U ...... Source: Iron Castings Handbook.. :? 10-' o c c Z "0 "0 <. Ed.-I.J.-_L-.. Charles F. Ductile Iron: Fatigue Crack Growth Rate Compared With That of Steel Stress Intensity Factor Range. Inc.I 10 30 40 60 80100 Stress Intensity Factor Range...I. Walton.I...J._ _.. Iron Castings Society.! ~ ~ • 3300 Ib (1500 kg) o 1650 Ib (750 kg) .L-_ _..L.

0 for the higher-strength irons. Inc. Unmachined and notched surfaces do reduce the fatigue strength. coining. Walton. 350 <tI 0 0 0 c.2 for the lowerstrength irons to over 2. or shot peening can increase the fatigue life of a component significantly. Iron Castings Society.268 10-24... low-cycle fatigue. :2: 300 ~ --l -0 -0 0 --l <tI 40 260 30 L- -'- ~ 10' 10- 10' Number of Cycles The effect of cast surfaces on four grades of malleable iron was also studied in high-stress.. Malleable Iron: S-N Curve Comparisons of Four Grades 60 r--------.iii 0. 400 50 .p 311 .---------.. Source: Iron Castings Handbook. Design with adequate sections that are well blended to reduce stress concentrations is most effective in reducing the possibility of a fatigue failure.. The reduction factor is as low as 1. 1981.. Ed. The results with a 95% confidence limit are presented in this S-N diagram. Charles F.. Inducing compressive stresses into the surface by rolling.

25 mm) deep. Charles F. Source: Iron Castings Handbook. as shown above. smooth machined surfaces. Pearlitic Malleable Iron: Effect of Surface Conditions on S-N Curves 70 '00 60 C- 269 400 <0 o 0 0 a. and machined notches on the fatigue behavior of pearlitic malleable irons. Walton. 60 300 40 30 20 10' 200 ~ III en ~ en III Vi ~ Vi 10' 10' 10' Number of Cycles The influence of as-cast surfaces. Ed.. Iron 1 is grade 60003 and Iron 2 is grade 80002. The resulting data are shown in this diagram.10-25.. Inc.050 in. p 310 . Surface finish has an important influence on fatigue properties. Samples of malleable grades 60003 and 80002 were tested in fatigue with "as-cast" and machined surfaces. Iron Castings Society.. Samples of the 60003 grade were also included with a machined surface containing a sixty-degree notch that was 0. 1981. (1.

oil quenched. 1977. A. P 287 ." in Source Book on Nitriding. Metals Park OH.000 to 2. Samples were austenitized.0 1 PIECE 9 PIECP. Source: J. Pearlitic Malleable Iron: Effect of Nitriding t----50. "Short Cycle Atmosphere Nitriding.01.S ATHOSPIlr. Riopelle.270 10-26. tested by tension-compression. This chart indicates an increase in fatigue life of 750. and tempered to 241-269 HB prior to nitriding or testing without nitriding.160 POUND~ TENSION 7.700.RE NITRIDED COHPRESSIOH 11.000 cycles attained by nitriding.--------------------------------II) 5 HUMBER OF CYCLES Effect of gaseous atmosphere nitriding on fatigue characteristics of pearlitic malleable iron.600 POUNDS 1. American Society for Metals.

006 008 ~ tNotch 200 :..5 2. 100 ~ r-. . but fatigue strength decreases as notch depth increases.~ r--. I I -5 0> 150 <. Metals Park OH. Volume I. in. mm 1. I I 30 ~ t.13 mm or 0.r-s- u. . Source: Metals Handbook.. The graph above summarizes the effects of notches on fatigue strength... Properties and Selection: Irons and Steels.25 mm or 0. Ferritic Malleable Iron: Effect of Notch Radius and Depth Depth of notch.005 in. P 65 ..5 1..0 Effect of notch radius and notch depth on fatigue strength of ferritic malleable iron. 1978.75 mm or 0. or from 170 to 205 MPa (25 to 30 ksi). ~ 50 -• - E: 20 ~ '" 5.10-27. 9th Edition.. 10 o o o 0.. ~ 0. 271 250 o 002 004 rad.. notch radius has little effect on fatigue strength.... c.ius o 0.~ ~ '" 5. Fatigue strength of unnotched ferritic malleable iron is approximately 50% of the tensile strength. As a rule. u.0 Depth of notch.. • 0. American Society for Metals. .010 in..030 in..

J. Note absence of frequency effects in vacuum. virtually aU of the degradation in fatigue life at elevated temperatures can be attributed to environmental interactions. These results are not unambiguous. 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). P 343 . Source: D.001 c Q. since Koburger has shown a frequency effect in high-cycle fatigue for directionally solidified eutectic alloys when tested in air and in vacuum. Duquette.01 o AIR 593·C • VACUUM 593·C AVACUUM 20·C a: en en ~ ~ 0. The primary difference in these results may be related to the lack of intergranular cracking in eutectic alloys.272 11-1. particularly at elevated temperatures. Additionally. American Society for Metals. -J Plastic-strain range versus fatigue life for A286 ferrous alloy in air and in vacuum af 593°C (1095 OF). Coffin has suggested that for a number of materials. "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys. Metals Park OH. Numbers adjacent to test points indicate frequency in cycles per minute. it was noted that tests performed in vacuum showed transgranular crack nucleation and propagation versus intergranular nucleation and propagation in air at elevated temperatures." in Fatigue and Microstructure. 1979. A286: Effect of Environment A286 z : z c C) w VACUUM 0.

. with little or no effect on samples tested in a vacuum. As indicated.:::::::0 /' /' .0 ~---: -- l!>.. Source: L. P 13 . A286: Effect of Frequency on Life at 593°C (1095 OF) 273 /..-- ~ :0 ". "Fatigue in Machines and Structures-Power Generation." in Fatigue and Microstructure. =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 . decreasing frequency has a degrading effect on fatigue life ofsamples tested in air.11-2..------ --:.. F. Metals Park OH.:.. 1979.--------.--:. American Society for Metals.MI.AIR Effect of frequency on life of notched fatigue bars of A286 at 593°C (1095 OF) in air and vacuum../ / ) / A 286-593°C Kr=3. Coffin.::--:: /' --1>-.0 ..

I ill A·2S6 10-6 I 10 20 30 40 60 SO 100 Stress-intensity factor range. Gerberich and John H. Precipitation hardening occurs on aging in the range from 700 to 760°C (1300 to 1400 OF)for 16 hours. American Society for Metals. quench in oil or water. James E. / . AK. 1982. The austenitic precipitation-hardening stainless steel A286 (heat-resistant alloy) is the main representative in this category. It contains titanium and small amounts of vanadium and aluminum. This variation results in improved room temperature properties but less desirable stressrupture properties. and age at 700°C (1300 OF) for 16 hours. Ti) and Ni 4Mo(Fe.274 11-3. and at L-T.1I2 10 20 30 40 60 SO 100 10-3 53S·C (1000. (AI. E. Cr) Ti on aging. which precipitate as intermetallic compounds such as Ni. Underwood. / / / "'-24·C (75·F) ~ U .5 2' ~ "tl 316·C (600·F) 10-6 I . for one hour followed by quenching in oil or water). Source: J.67 Hz (elevated temperatures). ksi • In.OS. p 161 . Eds. Other combinations of heat treatments may be used depending on the application. MPa ' m1/2 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. anR ratio ofO.F)~' A' / / / t 10./ 10-4 / (i I" / I. Campbell. A286: Fatigue Crack Growth Rates at Room and Elevated Temperatures AK.. Metals Park OH. and R-C orientations. "Fracture Properties of Wrought Stainless Steels.4 427·C (SOO·F) . or 1800 OF. T-L. One variation is to re-solution treat at 900°C (1650 OF)for two hours. A / 10-6 01 ." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. Various mill forms of the alloy are usually supplied in the annealed conditionCondition A (980°C. R-L. Campbell. William W.

as shown above. <J<: w 60 s s <II 0 <J<: . B. American Society for Metals. Source: M. Athey and J. Cycles to first indication (crack) were comparable to conventional material. Allen. Astroloy: S-N Curves for Powder vs Conventional Forgings 100 Conventional Forgings I '" 275 o 80 I Powder Forging 2 x ' . \I) \I) ~ '" >- . 40 20 0 104 10 5 10 6 CYCLES S-N curves for conventional and powder forgings of Astroloy (notched versus smooth). Samuel Bradbury. 1979.11-4. R.. M.. P 97 . L.. "Application of Powder Metallurgy to Superalloy Forgings. Ed.. Crack propagation as judged by the number of additional cycles from first indication to failure was slower than conventional material. Moore. Metals Park OH." in Source Book on Powder Metallurgy. Testing was performed using standard methods at 705°C (1300 OF) and a combination of steady and vibratory stresses for which comparative data were available..

"Application of Powder Metallurgy to Superalloy Forgings. M. Source: M. 1979. Athey and J.000 psi o = Failure Powder Forging x = Crack Conventional Forging 10 4 CYCLES Astroloy tested in high-cycle fatigue at 705 °C (1300 OF).276 11-5. Ed. L." in Source Book on Powder Metallurgy. R. Samuel Bradbury. Metals Park OH. Allen. B. Astroloy: Powder vs Conventional Forgings Tested at 705°C (1300 OF) Steady Stress = Vibratory Stress = 40. Vibratory stress levels were selected to facilitate a direct comparison between conventional and powder forgings. American Society for Metals. Moore.p 97 ..

& .. FSX-430: Effect of Grain Size on Cycles to Cracking lSI 4)0 0.. P 336 ..) .1~"..0 IZO 160 CYCLES TO CRACK 240 zeo no SoN curves for alloy FSX-430.." in Superalloys: Source Book. Source: Eric Bachelet and Gerard Lesouit.~'I 0. Jr. c . \ t---O-I... Donachie.05 10. \ > 0.4 277 large grains ~ \ fine grain a if '" Z II: • \ . American Society for Metals..11-6..IG HI MIN I I I N ~-LG 0.. II: ~ c .0161 . . Metals Park OH. Ed." 0. ~ . "Quality of Castings of Superalloys... (O. 1984.. Matthew J. ~ ~cb \ \ II: . ~ 0 II: c ~.2 o 40 .01 10. showing effect of grain size on cycles to cracking..0"' 11. II: ~ !: II: ~ .01 .7621 0. L "ZOO 0.1 AVG MAl HZ E E C ~ '\ ( \ i .....

.01 GROWTH RATE IN llNlAl! flAIlGE ODZ . Donachie.4 CRACK AVG CflACK I ~-LG I I I 2 I t-{j----I I 0.0 1.03 II: (O.lCTCLEI 2.0151 ~ .. i .5 1. )~ large grlolinB I fine grains o 0... FSX-430: Effect of Grain Size on Fatigue Crack Propagation Rate o 0.TUI~ 0.OJ 0. American Society for Metals.O! .. Source: Eric Bachelet and Gerard Lesoult..04 (1.0 Fatigue crack propagation rate-effect of grain size on fatigue characteristics of FSX-430.2 . Jr.. Matthew J.~oeio !: '" C e.278 11-7. ~ :c a: z· . Ed. I ... . I ) (o.01 _ (O./CYCLE I O.5 GROWlH RAlE IN LINEAR flANGl (IO·"n.. II: c It' 0. Q. "Quality of Castings of SuperalIoys. P 337 .Z:>41 E E ~ ." in Superalloys: Source Book..". 1984.. % IE .04 ~ J--O-1-fG 0.. '" I I I 0....3 I a: O. Metals Park OH. < 0)1 C < < 0.

... Metals Park OH.. 1979. a. Source: ASM Committee on Heat-Resistant Castings. Ed. American Society for Metals..: -- :-1500 1400 1200 1300 Testing temperature..r---. ( Fatigue strength 100 million cycles Aged 50 hr at 1350 F 50 o o Q 40 30 20 1100 """"-< "'.. HS-31: Effect of Testing Temperature 70 279 60 .. Elihu F.. ! . after aging at 730°C (1350 OF).11-8. "Heat-Resistant Alloy Castings."in Source Book on Materials for ElevatedTemperature Applications... (["---. for 100 million cycles.....u. F Effect of testing temperature on fatigue strength of HS-31 casting alloy. Bradley. p 237 ..

. . ..... High-cycle fatigue properties ofnimocast alloy IN 738 LC tested at 850°C (1560 oF). 180 160 11M> 120 <. <.. . IN 738 lC Casting Alloy: Standard vs HIP'd Material 220 CII I y-"STANDARD CONDITlON" MATERIAL <. Metals Park OH.... 200 • :z... P 340 ... -"<... ~ Ul E-i - 107 CYCLES TO FAILURE SoN curves for casting alloy IN 738 LC... "Quality of Castings of Super alloys..280 11-9.. Donachie.. • <. .... Jr. American Society for Metals. MATERIAL <.. <... Ed. tI$...." in Superalloys: Source Book.... Matthew J...... Po. 1984. Source: Eric Bachelet and Gerard Lesoult.... +1 Po. Ul Ul " ... E E '-...

Jr... Ed.... + a. ~ . "Quality of Castings of Superalloys. I 200 180 160 140 120 -. . MATERIAL • .... 111 t- • ' • . • .... p 340 . • .. High-cycle fatigue properties of extra-fine-grain and conventional material tested at 850°C (1560 OF).... • ...... American Society for Metals.. Donachie. Source: Eric Bachelet and Gerard Lesoult.. <. IN 738 LC: Effect of Grain Size on Cycles to Failure 281 240 / N STANDARD CDNDITIDN I 2: 220 E E " a.........." in Superalloys: Source Book.. FINE GRAINS . Metals Park OH. 111 111 W 0: .... • • e· 8 10 6 10 7 10 CYCLES TO FAILURE SoN curves for alloy IN 738 LC.... 1984. . .......11-10...... ..... Matthew J..

..20 160 200 CYCLES TO CRACK 2.-o--i ~ fG -LG M'N...--<O)-----~ <..JUI 0.04 .... IN 738 LC: Effect of Grain Size on Cycles to Cracking 'N .. Ed. showing the effect of grain size on number of cycles to cracking.03 a: 10.40 :tIO 320 S-N curves for alloy IN 738 Le.. ..... a: ..2 ! IO.. HZ N ••.!i a: 0. American Society for Metals. ~~)-~.." in Superalloys: Source Book. c e:: c 0...282 0... - c 0.ZS41 'i ! .02 ~ w .J .. I I MAX. Metals Park OH..738 . ~ 11. ::l 0 _ O. Donachie.0161 . Jr..4 11-11. • .. . 1984... P 335 ... f ...0' o. 2: ~ L large grains fine grains eo ... : .!lOI1 c a: oJ ~ o . "Quality of Castings of Superalloys.."2 I 0.. Source: Eric Bachelet and Gerard Lesoult. Matthew J..

0 GIIOWTH RAT( IN LINEAR RANG[ llli'in.GZ & ! E . Donachie.." u n l l 283 O.01'1 IN ' ) I i . s (0.. . .~~I t ... SoN curves for alloy IN 738 Le..01 .!>08) O..0 1..' 1..5 1..04 . .. 4 0.OJ I 0.. ! ~ .. '" ...[ l . Jr...11-12. •. O. ." in Superalloys: Source Book. IN 738 LC: Effect of Grain Size on Crack Propagation Rate c.•. Metals Park OB. American Society for Metals. Ed... Source: Eric Bachelet and Gerard Lesoult. L • 0... • Ci QOJ (0.0..7611 f- .GI . P 336 .) f- ~ . 0.. I I CRACI( I /IIIG CU~K I I t-O-i :E-C»i I FG I LG o 0.ROWTH IlATE IN L1I1EAR 11&".lCYCLL. 1984. showing the effect of grain size on crack propagation rate. . % ..01 O.01.. . Matthew J.2 f- f 0.: . "Quality of Castings of Superalloys.

Jr. Metals Park OH.grained R t'<J 0.11 Waveform: Sinusoidal o 10 20 30 40 -3/2 ~K(MNm ) 50 60 Fatigue crack growth rate at 850°C (1560 OF) in various grain sizes of alloy IN 738 LC. American Society for Metals.." in Superalloys: Source Book. 1984.1 10.284 11-13. Source: Eric Bachelet and Gerard Lesoult.p 341 000- . Ed.0 fine-grained 10..5 do dN (m/cycle) 10. achie.10 Abo coarse . IN 738 LC: Fatigue Crack Growth Rate at 850°C (1560 OF) 10.6 100Hz • Alloy I IN 738 LC Temperature I 850°C . "Quality of Castings of Superalloys.' Matthew J.

as shown above. Skelton.11-14." in Source Book on Materials for Elevated-Temperature Applications.125..667. (a) Ratio of cyclicto mean stress= 0. P. Ed. relative to air. although the effect was much smaller than that seen in lead. In reversed bend tests on lead at 500 cycles/min.. cycles I I II lid I I "" II 100 1000 LIFE. Inconel 550: Axial Tensile Fatigue Properties in Air and Vacuum at 1090 K 50 40 30 285 2 (a) . At high temperature (l090 K) vacuum also improved endurance. indicating a possible strengthening effect of air. h 10000 Axial tensile fatigue properties oflnconel550 at 1090 K in air and vacuum. and pure oxygen. Bradley. Cook and R. C> 20 10 50 'iii 0- a: VI vi VI w t- ·VACUUM °AIR Z <Cl: w 40 30 20 10 105 ::E (b) 10 107 LIFE.. At all strain levels vacuum endurances exceeded those in air. American Society for Metals. (b) Ratio of cyclic to mean stress = 0. Elihu F. which exceeded those in oxygen. 1979. H. p 81 . of the Co-base alloy S-816 and the Ni-base alloy Inconel550. Endurances for the nickel-base alloy converged at low stresses. Snowden demonstrated a difference oftwo orders of magnitude in fatigue life between vacuum. "Environment-Dependence of the Mechanical Properties of Metals at High Temperature. air.. Intermittent stress-free exposure to air had no effect on the lifetime in vacuum. Metals Park OH. Source: R. Testing frequency = 33 Hz.

. Volume 3.9 mm (0. 1980. Metals Park OH. Inconel 625: Effect of Temperature on Cycles to Failure 600 500 427°C .).3) --- 760°C ('1400 of) 29 Oc (85 of) 871°C (1600 of) _ i ~ 20 100 Cycles 10 failure S-N curves for hot rolled solution treated Inconel625 bar 15. P 143 ..004 ln. Properties and Selection: Stainless Steels. Average grain size was 0.) in diameter at various temperatures. Source: Metals Handbook.10 mm (0. - 29°C (85 of) 538°C (1000 °F)_ 60 649°C (1200 of) I- ~ Ii 300 200 NOICh~~ specimens (Kt~3.625 in.286 11-15.. <0 '- :. <. 9th Edition.~ e:-- Joo 80 of) 400 a.. Tool Materials and Special-Purpose Metals. American Society for Metals.

Results of FCP tests at room temperature and at temperatures as low as -269°C (-452 OF) for Inconel 706 are shown above. m1/2 Fatigue crack growth rates of Inconel 706 forged billet (vacuum induction melted/vacuum are remelted) at an R ratio of 0. p 297 . Source: Stephen D. u 10. aK. AC. II. Eds.. E.4 i e Cl ~ I! u Ql ~ I! U Ql u ::J '':. III Cl ::J '':.1 and a frequency of 10 Hz. Metals Park OH."in Application of Fracture Mechanics for Selection of Metallic Structural Materials. "Fracture Properties ofSuperalloys. "C "C III III Ql' ~ i e Cl s: -269°C (--452°F) 10. At equivalent 11K values.. Campbell. aK. Antolovich and J. AC.11-16. ksi • in.. double aged 730°C (1350 OF) 8 h.1/2 10 20 30 40506080 100 200 300 10. .4 287 -196°C (-320°F) Ql 10-3 Ql U -E E z -"C III > u U "C l!l III s: .5 -.. MPa .E z· -. Gerberich and John H. Underwood. Campbell. James E. 10-6 20 30 40 50 60 80 100 200 300 Stress-intensity factor range. hold 8 h. FC to 620°C (1150 OF). the fatigue crack growth rates for this alloy are slightly lower at subzero temperatures than at room temperature. William W.. III Cl II. 1982. Heat treatment: 980°C (1800 OF) I h. Inconel 706: Effect of Temperature on Fatigue Crack Growth Rate Stress-intensity factor range. American Society for Metals..

LJ........-.--'--..LJ... P 235 ." Tests were performed at two different elevated temperatures as shown..... Metals Park OH. Source: ASM Committee on Heat-Resistant Castings..... 1979. Elihu F.288 11-17.J.1." in Source Book on Materials for ElevatedTemperature Applications.2? 0....L.I... SoN curves for Inconel "713C..J. "Heat-Resistant Alloy Castings. Bradley..~ . :i 3 a 1--+-+++t--~.---L....L.. I 10 100 Millions of cycles to failure ~ <i E 10 ....----L~ .---l.. Ed.... Inconel"713C": Effect of Elevated Temperatures on Fatigue Characteristics 50 "in g 4 a I-+-+-f-Ht--~H-Yt---+--+ o c... American Society for Metals...1 OL-.*h-r--+--P't"+l:>---1--+t-t-l Q) iii o "- Ol 2 a I--+--+--r-f-.---r---t-+-+--i---i--t-++t---+--t-+-+i NOTE: Higher fatigue strength at 1500 F than at 1200 F is consistent with tensile strength relations in graph shown above left...

crack. Inconel "713C" and As-Cast HS-31: Comparison of Two Alloys for Number of Cycles in Thermal Fatigue to Initiate Cracks . Metals Park OH. Elihu F..r I Thermal fatigue Cycles to develop I. erne k avg1 HS-31 Inconel 1713C" 289 Material Thermal fatigue Cycles to first crack Material HS-31 Inconel 1713C" ri a avg -1& I 2 3 Thousands of cycles IIIII a I 2 3 Thousands of cycles l-rnin cycles. P 235 . Bradley. 8 -m. 100 to 1700F 3 tests each motertol Thermal fatigue properties of HS-31 compared with those of Inconel "713C." Left: Number of cycles required to initiate cracks.11-18. Ed. Right: Number of cycles required to develop VB-in. "Heat-Resistant Alloy Castings." in Source Book on Materials for ElevatedTemperature Applications. American Society for Metals. Source: ASM Committee on Heat-Resistant Castings. 1979.

with the crack path following the boundaries of the largest grains. 61<.--1.290 11-19.---l 8 10 15 20 30 40 50 60 Stress-intensity factor range.. Antolovich and J.Hz } 05 X 10-' H. / 5 X 10-4 - 1> I & - o o o /.. Campbell._--'_. because different modes of deformation may have occurred at different strain rates.L.. 2 0 \ o o 00 o"... Metals Park OH... One may be inclined to attribute the increase in FCP to either creep or environment.~/ / • /S /ili. Gerberich and John H.. "Fracture Properties of Superalloys. 1982.. and the crack surfaces showed an increased amount of intergranular fracture with decreasing frequency. James E. Eds..5 Hz.----I.• a I / tL { -& /& 20 Hz 2Hz _ a &&/ & // ~/ 0. but this may not be the case.. o 5 X 10-5 Io a~a / I... MPa . American Society for Metals.---/ 5 X 10-6 / &1 1.. the results are shown above. Below 0.L...--'---'--'-._ _--1. E. the FCP rate was more rapid.. Source: Stephen D. William W..". P 294 .. Underwood.. Campbell. The effect of frequency at 550°C (1025 OF) was studied using a sine wave. m1/2 Variation ofFCP rate (da/dN) with stress-intensity factor range (LlK) and frequency at 550°C (1025 OF) (sinusoidal load) for specimens oflnconel 718.. "in Application of Fracture Mechanics for Selection of Metallic Structural Materials..5 Hz & . Inconel718: Effect of Frequency on Fatigue Crack Propagation Rate I I I I I I 5 X 10-3 f- o o 5 X 10.L.

where the effect of temperature is to increase the FCP rate. American Society for Metals.L.. • ••• • • • • ••• • •• •• •• .. Li K. E.. .._~ : Twins No twins 20 30 40 50 60 10. Source: Stephen D. Eds. James E..11-20..-. especially at LiK levels......_--'-_.. Underwood.-......... •• ....... An important effect is the hydrostatic state of stress in the tip region. 1982..... Campbell.. "in Application of Fracture Mechanics for Selection of Metallic Structural Materials. • • •• • • ... William W. Campbell... Metals Park OH. This idea has been considered for Inconel 718...6 . Gerberich and John H...... "Fracture Properties of Superalloys... 8 10 15 Stress-intensity factor range. m 1/2 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.L.--L . P 290 .L. Inconel718: Relationship of Fatigue Crack Propagation Rate With Stress Intensity 291 • 25°C (77°F) • 550°C (1025°F) 20 Hz 5 X 10-4 5 X 10-5 5 X 10-6 • :• •• ••••• ••• • • .. Antolovich and J. The FCP response is shown above for 20 Hz.L.J... MPa .

/ rI . Source: Stephen D. Inconel 718: Relationship of Fatigue Crack Growth Rate With Load/Time Waveforms a K./ (3) 'Jr¥ 4l I. Underwood. Gerberich and John H. '/2 10 20 30 40 60 • /'v (2) 5 X 10-2 Hz { _ rxr D (1) 2 Hz /'v (3) I I (2) 2 5 X 10. 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.6 8 10 15 20 30 40 50 60 aK. Eds.. Top: Various forms of cyclic stress fluctuations used at 550 °C (1025OF)at a frequency of5X 10. MPa • m. Antolovich and J. E. Bottom: FCP rates at 550°C under sinusoidal. James E..292 11-21. P 295 . 1982. "Fracture Properties of Superalloys. as shown in the lower graph.4 Sinusoidal ~ ~ rf· / I (1) ~ . as shown in the top graph.5 z· ~ 5 X 10. Metals Park OH. Campbell.0 2X 10-6 ! "tl " tl / I I I. . ksi • in./2 Load/time waveforms and FCP rates for specimens of Inconel 718.6 - t6 Sinusoidal 2 Hz .6 2 X 10. To separate out the possible effects of creep or environment from deformation mode. the authors used triangular and square wave shapes..2 Hz. _ -Q r:t • -.. triangular and square loads. William W.Hz 5 X 10. 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. American Society for Metals. "in Application of Fracture Mechanics for Selection of Metallic Structural Materials. Campbell.

.... Here it is evident that crack growth rate at constant temperature is lower in inert gas.. .. -"tl "tl Z' 10-3 ::.3 u . MPa ........ Campbell.. ksi • 293 iny2 10 20 3040 6080 5 X 10-1 1'"T'"--"""--T"-"""-T""T""1'""'I""l • = Air o = He 10. Gerberich and John H.I-l'-'-. 0.... -. William W...¥ 0> ... .. m1/2 Fatigue crack growth rate data for Inconel 718 in air and in helium.....4 . "Fracture Properties of Super alloys."in Application of Fracture Mechanics for Selection of Metallic Structural Materials._ _L .. Eds. Metals Park OH. 10-5 10-4 . 650°C (1200 OF). 0> . Inconel 718: Fatigue Crack Growth Rate in Air vs Helium ~K. 10 20 40 60 80100 ~K. Campbell.. P 287 . .J ' . . ~ -E E 10..2 Gl . . > u .. Antolovich and J.. Temperature..-. .. Source: Stephen D.. 1982.. Underwood. Frequency.::.' "tl "tl ... American Society for Metals.11-22..5 U i0 10. James E.1 Hz.. u.I: z' . E.... u u .

1 Hz."in Application of Fracture Mechanics for Selection of Metallic Structural Materials. "Fracture Properties of Super alloys. E. E CI 10.. Antolovich and J. From the data above it becomes obvious that fatigue crack growth rate increases greatly in aggressive environments compared with exposure to helium alone..L.4 e "'u" .. Source: Stephen D. 1/ 2 10 20 30 40 60 80 5 X 10-1 10. James E.' ..L.4 L-_----I_~---L___L. MPa • m 112 Fatigue crack growth rate data for Inconel 718 in helium + 0. Gerberich and John H. ~ E E III -~ .... Campbell.294 11-23.J..5 z· -"'C "'C III U ~ .t: 10-2 z· "'C . ksi • in.::...5% H 2S + 5% S02 40 60 80100 10.. .5% hydrogen sulfide and helium + 5% sulfur dioxide. American Society for Metals. i -.... Metals Park OH. 1982.. Campbell. III III U 10-3 :::l ''. William W. 0.5 • = 0= He He + 0. Temperature. P 288 . Eds.3 u .... 10. 650°C (1200 OF). Frequency.. Inconel 718: Effect of Environment on Fatigue Crack Growth Rate ~K..I. Underwood.J 10 20 ~K. CI u.

J. Campbell.. American Society for Metals.3 -"C "C ~ Q > u II> 10. P 289 . MPa • m' /2 Fatigue crack growth rate data for Inconel 718 in air + 5% sulfur dioxide. Underwood. Source: Stephen D. The attack on the surfaces of unstressed specimens in aggressive S02 environments was minimal..2 -'" i .... 1982.." ~ e e u u '" 10. (The effect of air plus 5% S02 was similar to the effect of air alone. which was used to establish a baseline." ._-'---'--'-. E. '" 10-4 '--_ _. In the air. Inconel 718: Fatigue Crack Growth Rate in Air Plus 5% Sulfur Dioxide AK.. Antolovich and J.J 10 20 40 60 80100 AK. '" z· II> E E Q 10.J. cracking was generally transgranular with well-defined striations... oxygen-bearing and sulfur-bearing environments. ksi • in.11-24. William W. Metals Park OH. "Fracture Properties of Superalloys.3 .. "in Application of Fracture Mechanics for Selection of Metallic Structural Materials.. indicating that an important effect of the environment was to degrade the boundary strength by mechanisms that were not clearly defined.S ~ "C > i z .1 Air + 5% S02 10..4 ~ .L.) It was observed that in the helium atmosphere.L. :s II> II.L. Eds.. 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. Campbell. Gerberich and John H. but the S02 environments caused substantial increases in FCP... It was suggested that oxygen diffusion along grain boundaries and localized oxidation may have occurred.. James E.. the crack path changed from transgranular to intergranular. ' / 2 10 20 30 40 60 80 295 5 X 10..

Inconel718 Tested in air at 24°C (75°F) 500 < f < 600 cpm. I .05 and at a frequency of 0. Campbell. Metals Park OH. ... Gerberich and John H. 803 0 Heat III 20 40 60 80 I:!> Spec. heat I 0 Spec.. heat II I I (b) 20 40 60 80 Stress-intensity factor range. 1290" / (CHT) 'f / ~ E E Z' ~ III : 10-3 Spec.CHT= conventional heat treatment.in Application of Fracture Mechanics for Selection of Metallic Structural Materials. 1982. 158 & 803'" (CHT) / / . Inconel 718: Fatigue Crack Growth Rate in Air at Room Temperature Spec. All testing was done at R = 0. James E.296 11-25. p 276 . 1283. 253. Antolovich and J. MPa • m 1/2 Fatigue crack growth rate behavior ofInconel718 tested in air at 24 ° C (75OF). "Fracture Properties of Super alloys. v' 0 '" " ~. Eds. 158 } heat II V Spec. R = 0.. Campbell. American Society for Metals. (!. Underwood. heat I 0 Spec. 1290.I:!> 0 Conventional heat treatment Modified heat treatment I:!> Spec.67 Hz. Source: Stephen D.05 / . .. William W. ~K. E.

Campbell.. MPa .--L.4' . E. R = 0..L.L--''--J.. Source: Stephen D. heat Spec..I IJ.... Ll.. (a) o Spec.I1J..05. P 277 . 1982." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. ..I 0 .. IJ.J-I 40 60 80 Stress-intensity factor range. .L--L./ Modified heat treatment IJ.0 . William W._--'_.67 Hz. 210. Antolovich and J... Campbell.-_ _--I. CHT = conventional heat treatment.... heat o Heat III Spec.... American Society for Metals.-L-'-..11-26.. Metals Park OH.6 L.J-'40 60 80 20 .~ Inconel718 Tested in air at 316°C (600°F) f = 40 cpm.. 1282.L. heat I II 10.-. "Fracture Properties of Superalloys..../ t:J. 254. 1294. All testing was done at R = 0.. Eds. and at a frequency of 0.-_. . . Gerberich and John H. m 1/2 Fatigue crack growth rate behavior of Inconel 718 tested in air at 316°C (600 OF). heat I II (b) IJ. Inconel 718: Fatigue Crack Growth Rate in Air at 316 °C (600 OF) 297 Spec. Underwood. James E.. K. 210 (CHT) Spec../ . -: / / Conventional heat treatment .05 / IJ. y "IJ. 20 o Spec. 1294 (CHTl / .

MPa • m1/2 Fatigue crack growth rate behavior of Inconel 718 tested in air at 427°C (800 OF).._ . 255... . Underwood.. "Fracture Properties ofSuperalloys._ .. 20 40 60 80 20 40 60 80 Stress-intensity factor range.II .: ~ (CHT) lID...... r::I:~--__r--.... heat II o Spec... Source: Stephen D. heat I Spec. p 278 ... L -..____. 1982. 1291. T. "in Application of Fracture Mechanics for Selection of Metallic Structural Materials.. heat II I :1 Spec.___. I II <9 .67 Hz...298 11-27...... CHT = conventional heat treatment. Gerberich and John H.. . Antolovich and J.. Inconel 718: Fatigue Crack Growth Rate in Air at 427 °C (800 OF) 10-1 ..___r.. American Society for Metals..-__r-......L---L----'----''---L. 162. R = 0.----r--.L . Metals Park OH. James E.05 and at a frequency of 0. 162 (CHT1..05 10.' . Campbell. William W.. 1286..___.2 D. E. AK.... Campbell. /. I II .... 1291". L . All testing was done at R = 0. Eds....:I Inconel718 Tested in air at 427°C (800°Fl f = 40 cpm.... heat I spec..___..~ /'1 0 0 0 10-4 // 10-5 L. o Spec..' . I 0 ----'_ _....____..../1 10-3 Spec.. Conventional heat treatment Modified heat treatment D.

E. Eds. 251} heat II Spec.3 10. 1982.-.l... heat I Spec.L:--JL. "in Application of Fracture Mechanics for Selection of Metallic Structural Materials.___r-. heat II heat III Modified heat treatment Spec._r_1r_T"T"']r:__--__r--. CHT = conventional heat treatment. R = 0.05 and at a frequency of 0.67 Hz..J'--_ _----'_ _-'------'_.1.. MPa ... James E.L~ 20 40 60 80 Stress-intensity factor range. P 279 .JL. LiK.-~-. William W.2 10. "Fracture Properties of'Superalloys.-. Campbell.... Underwood.1 1:""""----. Campbell. Source: Stephen D..--L_. 250 10.. American Society for Metals.5 '--_ _----'C--_.--.4 Conventional heat treatment Spec... m 1/2 Fatigue crack growth rate behavior of Inconel 718 tested in air at 538°C (1000 OF).05 10..... All testing was done at R = 0._r_1r_T"""1"":I 299 Inconel 718 Tested in air at 538°C (1000°F) f = 40 cpm. Inconel 718: Fatigue Crack Growth Rate in Air at 538 °C (1000 OF) 10.. Antolovich and J.-L.L. 1284.11-28.L. 1288. heat I Spec. Metals Park OH.. Gerberich and John H. 165.l.

Conventional heat treatment Modified heat treatment II o Spec. l1K. Campbell. Antolovich and J. William W.. All testing was done at R = 0. 252. Inconel 718: Fatigue Crack Growth Rate in Air at 649°C (1200 OF) 1 10. heat I Spec.300 11-29. "Fracture Properties ofSuperalloys. heat heat III 40 Spec. American Society for Metals.--r""'T"r-r""'] Inconel 718 Tested in air at 649°C (1200°F) f = 40 cprn. James E. Campbell. Metals Park OH. heat II o 20 Stress-intensity factor range.05 Spec. E. 1982. 156. m 1/2 Fatigue crack growth rate behavior of Inconel 718 tested in air at 649°C (1200 OF). Underwood.3 . P 280 .05 and at a frequency of 0.--."inApplication of Fracture Mechanics for Selection of Metallic Structural Materials.~---r---r-'---'-"'T1-r:r-----r----./ / !SJ / 10.4 b. 1289. MPa . heat I Spec.67 Hz. Source: Stephen D. CHT = conventional heat treatment. Gerberich and John H. 1289 (CHT) • V·' / 10./ . 1281. Eds. R = 0.

Underwood. Metals Park OH. American Society for Metals.3 22° C (72° F) 0. Heat treatment: 980°C (1800 OF) '% h.r: ~ e Cl .1 and a frequency of 20 Hz. :I u.! ~ CJ 10. James E.54 cm 'V -196° C (-320° F) thickness o -269° C (--452° F) o • Q) U E E ~ "0 e 10. double aged 720°C (1325 OF) 8 h. E. <tI 5 10 50 100 Stress-intensity factor range. AI<. Campbell. At the constant frequency the effect of higher temperature is to increase the FCP rate.:.:. Gerberich and John H. AC. FC to 620°C (1150 OF).51 cm thickness z <tI ~f ~ .4 Q) Cl . Antolovichand J. P 298 . "Fracture PropertiesofSuperalloys. Eds. MPa • m 1/2 Fatigue crack growth rates of Inconel 718 forged bar at an R ratio of 0. Source: Stephen D.. AC. Campbell... William W. hold 10 h."in Application of Fracture Mechanics for Selection of Metallic Structural Materials.11-30. Inconel 718: Fatigue Crack Growth Rates at Cryogenic Temperatures 301 l> 22° C (72° F) } -78° C (-108° F) 2. 1982.

302 11-31. William W. Inconel718 and X-750: Fatigue Crack Growth Rates at Cryogenic Temperatures Inconel 718 (Ref 8. American Society for Metals.48) ~ 27 to -269°C (80 to -452°F) 5 10 50 100 Stress-intensity factor range. results of FCP tests depend on both melting practice and thermomechanical processing. Metals Park OB. James E. Source: Stephen D. Underwood. Antolovich and J. along with room temperature FCP data for Inconel 718 from Shahinian et al. 1982.44) ~ 22 to -269°C (72 to -452°F) Inconel 718 (Ref 8. Gerberich and John H. However. A comparison of FCP values from room temperature to -269°C (-452 OF)for Incone1718 and Inconel X-750 is shown in the above chart.P 300 . E. The FCP data for these two alloys overlap in the t!K range shown. Eds. ~K.. MPa • m 1/2 Fatigue crack growth rates for Inconel718 and Inconel X-7S0in the subzero temperature range. Under some conditions. the FCP rate for Inconel 706 is slightly less than those for Inconel 718 and Inconel X-750 at corresponding temperatures and t!Klevels. Campbell. "Fracture Properties of Super alloys. Campbell.49) • • 22°C (72°F) Inconel X-750 (Ref 8."in Application of Fracture Mechanics for Selection of Metallic Structural Materials.

11-32. Campbell. 1982. Antolovich and J. Metals Park OH. Gerberich and John H. E. American Society for Metals. . m 1/2 Fatigue crack growth rates of Inconel X-750 at an R ratio of 0. Eds. Campbell. Underwood. "in Application of Fracture Mechanics for Selection of Metallic Structural Materials. the effect of higher temperature is to increase the FCP rate. Inconel X-750: Effect of Temperature on Fatigue Crack Growth Rates 303 !E E u CIl A • 22°C (72°F) -196°C (-320°F) o -269°C (-452°F) 10.~ CJ 10-4 CIl II. '" 100 Stress-intensity factor range. Heat treatment: solution treated and double aged. Source: Stephen D..3 ~ "tl :!l' E Z ie ~ en b 5. Within this frequency range.1 and at frequencies of 20 to 28 Hz. "Fracture Properties of Superalloys. William W. P 299 . llK. James E.. MPa .

. i!: 60 50 0" 800 1 Ii. '" ~ ~ . Bradley. 1979..-k"." in Source Book on Materials for Elevated-Temperature Applications. and Testing Temperature With Fatigue Characteristics ksl 100r------------------. Elihu F. "The Super 12%CrSteels. D. and with fatigue ratio for Jethete M152.6 0.od.. Ed. Jethete M 152: Interrelationship of Tempering Treatment. o ~ 40 20 LONGITUOINAL SPECIMENS ROTATING·8EAM TESTS FOR \0' CYCLES 0_ 50 o temperedal1200 Fl650 ClIorone hour toa tensilestreneth abouI150ksl 0' 200 400 I 600 1000 F °O!----.. Source: J. P 123 . Right: Influence of alloy class and testing temperature on fatigue strength for the same alloy. Metals Park OH.304 11-33. Alloy Class.. Z...---=----...4 =ksi 100 0-_0______ x x tempered at 510 F(300 Cllor one hour to alensile slren&lh 0labout205 ksl )(--------x 80 ~ . :::> o 0. 0: G .--__=! I 1200 I 450 200 300 400 600 C TEST TEMPERATURE Left: Interrelationship of prior tempering treatment and testing temperature with limiting fatigue stress....!:--. Parker.--""*'. Briggs and T.. American Society for Metals.

Elihu F. p 123 .----.. 90.." in Source Book on Materials for Elevaled-Temperature Applications..5/4. rough e:rinding andlough machining. marlemper 650 F(345 C) + lemper 1150/1600 F(620/B70 C) CANTILEVER ROTATING·BEAM TESTS mechanically polished. "The Super 12%Cr Steels. Lapelloy: Interrelationship of Hardness and Strength With Fatigue Characteristics ksi 305 9o.. CLASS II (Lapelloy) 2000 F C1 095 Cl sail quench. o Including surface lolline. Ed.----. D.----:-!-::-----:*"--""IBO ksi Left: Relation between surface hardness and mean fatigue limit for Lapelloy.. Metals Park OH. American Society for Metals. Source: J. 1979. Briggsand T.-!.. Right: Relation between tensile strength and mean fatigue limit for the same alloy.---------------.0 mlcrelnehes alhersurface Irealmenls. Bradley.------------. BO rms 2..5/40 microlnches I 70 95% conlidencelimits 50 50 4~'='00. rms5.". Parker. Z.11-34.

Elihu F. where it was suggested that oxide in cracks could prolong life in air at low stresses. in single crystals of the alloy Mar-M200. "Environment-Dependence of the Mechanical Properties of Metals at High Temperature.. Ed. Bradley. or sulfur dioxide. American Society for Metals. P 81 . A thin oxide film. but oxide formed during pre-exposure did not. P. Mar-M200: Effect of Atmosphere on Cycles to Failure 10 8 o 6 10 • o • AIR DRY AIR WET AIR VACUUM PREOXIDIZED SPECIMEN. suppressed surface crack initiation. air endurances were less than those in vacuum at room temperature whilst the reverse was true at high temperature (above)." in Source Book on Materials for Elevated-Temperature Applications. Metals Park OH. formed during testing. Skelton. Also. Cook and R. Source: R. where tests in purified argon gave shorter endurances than those in air. 1979. Crossovers have also been seen in a ferritic stainless steel and a Nil Cr alloy in the range 875-1025 K. impure argon. VACUUM TESTED 1 I 1 10 10 8 o z w w V> V> a: '" a: V> 1 l- 10 6 4 1 10 10 5 CYCLES TO FAILURE S-N curves showing fatigue life at 10 Hz of single-crystal low-carbon alloy Mar-M200 at 295-1200 K. H. Convergence of air and vacuum data was noted for AISI 3I6 steel at 1090 K.306 11-35. and a crossover of the air and vacuum curves occurred for nickel.

.0 2. E E N""- c 'fg :il~ 4.. p 338 .0 307 U U >. .5 1.:: . 20 .. . or: c 0 '1'0 u 2.. . or: c 0 :il~ iii 0\ . 4.. u 3.......L- ~ c 2.. Ed.0 ..5 L-.5 u N""- .. - 3.. 0 no ~ u .0 u >. . "Quality of Castings of Superalloys..... American Society for Metals... Jr. 0 :£ ... Mar-M509: Correlation of Initial Crack Propagation and Dendrite Arm Spacing 5. Matthew J...11-36. Metals Park OH.' in Superalloys: Source Book..r.... 0 3.. u u iii 0\ ~ ij :!.5 :!. Source: Eric Bachelet and Gerard Lesoult. 1984. c .. Donachie. 40 Dendrite Arm Spacing 60 1~1 Correlation between the initial crack propagation rate and the dendrite arm spacing for Mar-MS09.

-'- ---' -L. P 337 .. 1984.." in Superalloys: Source Book... Donachie.. a .... Ed. Mar-M509: Correlation Between Number of Cycles Required to Initiate a Crack and Dendrite Arm Spacing 300l r-----~---____r---- .. Matthew J...PresentStudy z ~ e :::J IlXXl OL.. American Society for Metals.. ---' o 20 40 60 100 Dendrite Arm Spacing (1111) Correlation between the number of cycles required to initiate a crack and the dendritic arm spacing for cast alloy Mar-MS09. ~ .---__r---_.-_. Source: Eric Bachelet and Gerard Lesoult.-_--L. u ~ ~ . "Quality or Castings or Superalloys.L.308 11-37. Metals Park OH.....§ c u '0 .. 2000 . . Jr.

MERL 76. American Society for Metals. H.. Petersen and E. Metals Park OH. Source: J. Erhard Klar. Ed.0 fA m' 50 40 30 Life. P1M: Axial Low-Cycle Fatigue Life of As-HIP'd Alloy at 540°C (1000 OF) 100 90 00 309 80 0 'in ~ 0 = 70 60 KT 1.11-38. J. C. V. 1983. p 275 ." in Powder MetallurgyApplications. Dulis. Moil. cycles Axiallow-cycIe fatigue life ofas-HIP'd P/M alloy MERL 76 at 540°C (1000 OF) at notch severities as indicated. "Powder Metallurgy Parts for Aerospace Applications. Advantages and Limitations.

-_-L.J ..L....J Creln 10 Ihl anul 01 crackIng Talal eycln 10 !raw crack to 2·5 "'''' Bar chart showing effect of solidification conditions on cycles to onset of cracking and total number of cycles required to grow cracks to 2. r .----1_..:s.. ....._"""'---1_..L-_---1L. 1984... ..J .... 0... Ed.. P 339 . Jr... Q ... Nickel-Base Alloys: Effect of Solidification Conditions on Cycles to Onset of Cracking 1000 F"t Solidification (Condition f) SlolI Solidlficotion (Condillon S) .. Source: Eric Bachelet and Gerard Lesoult..... ~ r U s ~ a: .5 mm in several nickel-base casting alloys.. Q ~ . "Quality of Castings of Superalloys. Q... 100 . o (.L_... Metals Park OH._.... ..L." in Superal1oys: Source Book. Matthew J.. '" .. American Society for Metals. Donachie... 10 .310 11-39....L_-L. .... . Q a: .... :> ... r :z :> CD M21 71) 1I~71a IN!U M21 713 IN738 '''939 lC LC lC lC "~_ _.

. L1K. IV '--.11-40. William W.1. the oxidation products in the crack tip region will not be cracked and.--'-_---1. If the stresses are sufficiently low. Rene 95 (As-HIP): Cyclic Crack Growth Behavior Under Continuous and Hold-Time Conditions Stress-intensity factor range... 10. f! -"C "C IV Z' Ql' .. MPa • m 1/2 Cyclic crack growth behavior for as-HIP Rene 95 under both continuous and hold-time conditions at 650 °C (1200 OF).6 8 10 20 40 60 80 100 Stress-intensity factor range. Melals Park OH. Antolovich and J. IV Cl U. ~ 2 Cl f! Ql Cl ~ ~ f! U Ql ::3 ''.1-_"'-----'-1 " 10.. 1982. u u IV Ql' ~ u 0 .~ U E E -"C "C IV z' . James E.4 U. Eds. Campbell.4 10. Campbell._ _. the rate of crack growth would be expected to increase due to the severely degraded region in the crack tip zone. P 284 .. >u 10-2 15-minute hold time at maximum tensile stress 10-3 >u '<.. Underwood.1 Stress-level dependence U Ql Ql -<. --_ _.c 10. Gerberich and John H. The effect of environment need not always lead to more rapid crack growth.. ksi • in... Such effects would be pronounced at high temperatures and long hold times and have actually been observed in Rene 95.' -_ _. L1K. as shown in the above chart. Source: Stephen D...... "Fracture Properties of Superalloys. in some systems.. E.:... Once the stress intensity is high enough to crack the oxides.. Cl ::3 ''.3 . 1/ 2 6 8 10 20 40 60 80 100 10-2 311 10.:. an elevation of the threshold might occur..5 10. "in Application of Fracture Mechanics for Selection of Metallic Structural Materials.c . American Society for Metals.. .. 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.

K levels except 22 MPa·M 1/2 (20 ksi-in. Antolovich and J..! 10.r :i ~ l! . • F 10_2 20 100 1000 2000 { ~ U ...312 11-41. Underwood. . "in Application of Fracture Mechanics for Selection of Metallic Structural Materials.4 '= .. ·C bJ( Effect of temperature on fatigue crack growth rate at constant for Rene 95. Eds.~ '" 10. American Society for Metals. This possibility is considered in more detail in an analysis of FCP properties of Rene 95.b"). Gerberich and John H. . :::J '" 11 e :::J . "Fracture Properties of Super alloys. 1982. where the data are at least suggestive of a minimum.. i' ~ . James E. Metals Park OH. P 282 . '" o Testing tempereture. . leading to the hypothesis of a strong environmental effect.K level and frequency may be written as: da dN = Aexp . The FCP rate was plotted as a function of temperature for a given !:J. It is noteworthy that there is a minimum in the FCP rate at all !:J.3 ~ ..Q(!:J... Source: Stephen D. Campbell. as shown in the above chart.. Because any environmental interaction is thermally activated.. E 10-4 ..K range. Campbell.. " ~ e '" 11 10-6 . the crack growth rate at a given !:J. l! z· .K) is the apparent activation energy. Rene 95: Effect of Temperature on Fatigue Crack Growth Rate Testing temperature. William W.5 z· -. the life of the subsurface crack was much greater than that of the surface crack. E.K)/ RT where A is a constant and Q(!:J.. 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.

.......4 0 ~ Reversed stress fatigue --00"-- r-v 140 'x ltJ ~ 'Iii E :::J E "- 106 Number of cycles SoN diagram for 8-816 heat-resisting alloy tested at 900°C (1650 OF).......11-42..... ..---r---..--r---r---r--. ----__ = 3. 20 A::. 8-816: Effect of Notches on Cycles to Failure at 900°C (1650 OF) 40 . 280 313 ~ vi ~ a. p 245 . ~ ltJ 30 210 'Iii E :J "- vi ~ 'n..oo . K.r--... Source: High-Temperature Fatigue....----...---..r----..~ E 0-_-0.-. notched (broken curve) versus unnotched (solid curve).

"in Application of Fracture Mechanics for Selection of Metallic Structural Materials. '" Stage I 0 . > t.'" 10. e 01 '" Q)' ~ .2 u Q) -E E z· "tl "tl Q)' '<.t: ~ ~ t. P 285 . E. Q) ::J 01 .. Udimet 700: Fatigue Crack Growth Rates at 850°C (1560 OF) AK. u.E 10. 1/ 2 10 20 40 60 80100 10. Metals Park OH. William W.l . Eds.l z' ~ s: . Campbell.4 KN Ib c 582 1310 v 711 1600 01067 2400 l> 1244 2800 160 3600 ° 10. ksi • in. Campbell. American Society for Metals. Crack growth rates for this alloy are greatly accelerated by increases in temperature.l ~ 10-3 e ::J -" t.314 11-43.4 "tl > t. Gerberich and John H. James E..6 40 60 80 100 20 AK.01 u..3 10.l '" ... Source: Stephen D. 1982. "Fracture Properties of Super alloys.. Underwood.. ~ e 01 t.l u Q) -.5 m = 16 10. Antolovich and J. MPa • m 1/2 Crack growth rates in terms of stress-intensity factor range for Udimet 700 at 850 0 C (1560 0 F).l Q) -" t.

.M200 I I I --·-WROUGHT POIYCRYSTAlliNE UDiMEl 700 I I I --_..CONVENTIONAllY CAST MAR. 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.. Metals Park OH.1---•.2 l- V) - '--... Bradley.I-.COLUMNAR GRAINE\i AND SINGlE CRYSTAl w C> Z a: Z a: 10-2 -.:-.~~ -c - r-- -. directionally solidified and mono crystal ..~ I':::-- . -.. 1979.. ~ o I1400 0 F . 1-.... 103 . Shank.. -. "The Development of Columnar Grain and Single Crystal High Temperature Materials Through Directional Solidification.. I- I- -. 1700 0 F M~R.--. .. p 358 .. E. U-700 and Mar-M200: Comparison of Fatigue Properties I 315 .. 1'-. American Society for Metals." in Source Book on Materials for Elevated-Temperature Applications...11-44. r!--. Elihu F. Source: Francis L. Ed.~OO I I . Mar-M200.--. Versnyder and M. I I -c -( w. r--. '.

This is accompanied by a change in precipitate-dislocation interaction from one of shearing to that of dislocation looping or bypassing the precipitates. During aging of precipitation-hardenable alloys. At some point the energy associated with the accommodation strains exceeds that necessary to create a precipitate-matrix interface. 1200 c 1100 STRAIN CONTROLLED :::l :J a. CYCLES 1000 Stress-response curves for Waspaloy having nonshearable precipitates. American Society for Metals. Jr. Source: Edgar A. and accommodation strains build up." in Fatigue and Microstructure. Metals Park OH. Starke. 1979. and the precipitates become partly incoherent. the coherent precipitates grow. P 217 . deformation becomes more homogeneous. Waspaloy: Stress-Response Curves 0 ::iE Ul a. "Cyclic Plastic Deformation and Microstructure. Local softening is thus prevented. and the cyclic-response curve shows hardening to saturation. ::iE t- « V) V) Ul 0:: t- WASPALOY V) 10 100 N.. or to failure.316 11-45. and Gerd Lutjering. Since the reasons for strain localization have been removed. as illustrated above.

2 -.6 1Il ~ Z . 0..5 0.. p 335 . "Quality of Castings of Superalloys..4 < z 0.. American Society for Metals.. Metals Park OH.. 2 5 10. < } -'---} ---- at 750 DC at 650 DC z L!J 0.11-46. . X-40: Effect of Grain Size and Temperature on Fatigue Characteristics 317 ~ % 2 ~ 0. Matthew J.." in Superalloys: Source Book.3 0::...J I- w Small -e 0. 1984. Donachie.000 Source: Eric Bachelet and Gerard Lesoult. Jr.1 2 100 5 1000 CYCLES TO FAILURE S -N curves for X-40 showing effects of grain size and temperature on fatigue characteristics of this alloy.. Ed.

Rapid heating and cooling may. Ed. Cast Heat-Resisting Alloys: Ranking for Resistance to Thermal Fatigue IlJ a:: :J it ='300 o F-200 . An example of results from thermal fatigue data is presented above. Source: Steel Castings Handbook. Peter F... Rocky River OH. 1980. This is particularly true ifthe temperature changes are frequent or rapid. 5th Edition. Fatigue is a condition in which failure results from alternating load applications in shorter times. P 19-7 . Good design helps minimize the external restraint to expansion and contraction. 40 ~ z m 30 20 15 10 ILl ILl ~ 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. such test results have been useful in considering alloy selection questions. and nonuniform within or between casting sections.J ~150 <l ::!: ILl ~IOO a:: ffi ~ a:: o ~ 50 u. impose temperature gradients within the part causing the cooler elements of the component to restrain the hotter elements. or at lower stresses.. "Thermal fatigue" denotes the condition when the stresses are primarily due to hindered expansion or contraction. 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. Steel Founders' Society of America. and in identifying the superior thermal fatigue resistance of nickel predominating grades and the good performance of some HH type compositions. than expected from constant-load properties. Nevertheless. however. for some industrial applications. This graph offers a ranking of many cast heat-resistant high-alloy grades relative to their resistance to thermal fatigue. these thermally induced stresses may exceed those resulting from the mechanicalloads. Weiser. Finite-element computer analysis has shown that.318 11-47.

n -6A1-4V) 300 40 .~ ferritic _ stainless martensilic_ \ steels 1) 8 e 0 As shown above the corrosion-fatigue strength of bare aluminum alloys is superior only to that of magnesium alloys. Chubut.>< L-J t 'iii 0 J:.g. z duplex staness steels. C. I bw alloy aluminum m nickel alloys (e. Biloni and D. A.g.carbon steels. aJloys 600 and fn» ~ 20 .. Argentina).:.. titanium alloys 50 (e.'"""' 400 60 ~ -€. c: iii e ~ -F ." in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn. Pampillo. copper ... 1980. 0 iii .. Speidel.·-/ . Embury...nickel alloys . N • 108 cycles ambient temperature t '. 70 R· -1 ) (sea water . Eds. R. . "Aluminum as a Corrosion Resistant Material. . Careful surface protection may bring the corrosion-fatigue strength up into the range of bare stainless steels or copper-nickel alloys. P 617 . American Society for Metals.~ 0 s ~ t 200 30 ~ 100 r. Metals Park OR.!j 0. Corrosion-Fatigue Properties of Aluminum Alloys Compared With Those of Other Alloys 319 500.. Source: Markus O.12-1..S!' .------typical corrosion fatigue strength I -. E..

/ ':>. as a general rule.-_----I._---l 100 200 300 Tensile strength (MPa) Fatigue ratios (endurance limit/tensile strength) for aluminum alloys compared with those of magnesium and steel.... in contrast to steels.o~'b ()~ / / 6. Edward Arnold Ltd.I.. the more an alloy is dependent upon precipitation hardening for its total strength.-_ _. r0- . It should also be noted that the fatigue ratios are lowest for age-hardened aluminum alloys and. the lower this ratio becomes. _ _---L_ _-'-_ _. J.. which shows relationships between fatigue endurance limit (S X 108 cycles) and tensile strength for different alloys. Metals Park OH.320 12-2. It is well known that.. P 39 . the increases that have been achieved in the tensile strength of most nonferrous alloys have not been accompanied by proportionate improvements in fatigue properties.. Source: I...'= u ~ E::::: 150 :. 1981. Light Alloys. • O~---JL. and American Society for Metals. London. This feature is illustrated in the graph above. England. Polmear..= (J) 0) 0)- / c i:) u "O~ moo 100 ~O C )( UJ~ x 50 Aged aluminium alloys Non-heat treatable aluminium alloys Magnesium alloys Steels 400 500 600 700 x 0 6. 200 ll.. Comparisons of Aluminum Alloys With Magnesium and Steel: Tensile Strength vs Endurance limit 250 6. "3 / 6.L..

Chubut.:: R R ~ ~ . N-5x1Q1 20 000 I I I ~§ ~ I ~ I II ~~ ~ I <0 ~ I ~ ~~ f2 ..5 X 101 1 latiguestrength in sea water I N _10 fatigue strength • 8 in river water I N . Eds. . Source: Markus O. Argentina).. 1980.. 30 s: c l!! OJ 000 -Orr-----l:I 100 o 0CF. In other words. so does the ultimate tensile strength. 60 Hertz I stnooth specimens I ambient temperature • 60 + '. Embury. The above graph shows that the same is true for the corrosion-fatigue strength: there is as yet not a single commercial aluminum alloy available with a high-cycle corrosion-fatigue strength significantly higher than all the other aluminum alloys.--. Here aluminum alloys are listed in order of increasing yield strength. l!! 200 OJ c e. I I ~ Ie ~ 'Ie . H.and high-strength aluminum alloys all have about the same fatigue strength. Metals Park OH. Speidel.10 70 4()() R . corrosion fatigue is still a limiting factor for the application of aluminum alloys.. Biloni and D. N . 300 + 0 s: • • • corrosion fatigue strength 01 aluminum alloys '--' "" 'iii 40 0 e.air ." in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn. .:. An analogous conclusion can be drawn from a review of corrosion fatigue tests with smooth aluminum alloy specimens as shown in 'the above graph. Aluminum Alloys (General): Yield Strength vs Fatigue Strength 321 500 o o • • • ultimate tensile strength yield strength fatigue strength in air . Thus.-1 . American Society for Metals. C. P 616 . Pampillo. Cl) iij ~ !2 <0 <0 <0 ~ ."'"' 50 6 z -E . E. "Aluminum as a Corrosion Resistant Material. A.12-3. As the yield strength goes up.. but the fatigue strength in air soon reaches a limit which is roughly the same for alloys of greatly different yield strength. medium..

"Aluminum asa Corrosion Resistant Material.-. C.T 736 2219 ...... Source: Markus O..3 r--.T 6 ~ Q) '----' ~ E Z ~ 10 5 10-6 ~ <I 10.1 0.9 typical scatter in experiments crack orientation L .. Pampillo... A. 1980..T specimens CNP.7 10 8 ~ 10.. Embury. Chubut. Eds. Biloniand D. Argentina). 7050 . American Society for Metals. Speidel.T 651 2618. Metals Park OH.K I [ MN· m.3f2] Many commercial aluminum alloys show similar fatigue crack propagation rates in air.. ° ---+--------------------4 10 20 40 I 50 60 cyclic stress intensity range .. DCB I CT ambient temperature environment air R =0.....t:...---.. as indicated in the above comparison.1 -30 Hz I I I 1)..T 851 7079 .. p 613 . "in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn. H. E.0-0. Comparison of Aluminum Alloy Grades for Crack Propagation Rate 1O.322 12-4.T 736 7175 .

in the case of well-annealed Al 1100 (RO). accompanied by pronounced slip-step formation on the surface. American Society for Metals. introduced by the cold work. which can be observed after about 25% of the fatigue life has been expended. 1979. A. This latter part of the fatigue life is called the saturation stage offatigue.. Alloy 1100: Relationship of Fatigue Cycles and Hardness for HO and H14 Tempers 50r----. On the other hand.---. because the initial dislocation-loop length is much shorter. the material will harden in the early stages of fatigue (see S-N curves above) as the dislocation density in the bulk of the material increases."New Techniques for Detection and Monitoring of Fatigue Damage. Source: O." in Fatigue and Microstructure.. P 128 . during this initial rearrangement.12-5. The micro crack density is about the same for both materials. Metals Park OH. For instance. the material will soften in the early stages offatigue (above curves) as the dislocation density.---. Slip-step formation in this situation is much less pronounced than it is during hardening. Buck and G.-------r--r-----r--r-----r--r--. In either case.------. 323 • LU ~ 30 Z ----------- o :I: cz:: <: o ~ ><:: 20 z 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. Microcrack initiation is often triggered by a dislocation rearrangement. the dislocations form a cell structure with individual dislocations of long loop length shuttling to and fro between the cell walls. in the case of the cold-worked material Al 1100 (R 14). decreases.---r----." and finally to microcracks. or "extrusions-intrusions.---r--. A1ers. during which the dislocation shuttling leads to local instabilities.

apparently because of an immediate rep inning of the long loops by interstitials in this alloy. Alers.. 1979. however. At the same time. increased. Alloy 1100: Interrelationship of Fatigue Cycles.324 12-6... Buck and G... the cell structure that developed (with its individual dislocations within the cells) became quite large. P 131 . Acoustic Harmonic Generation and Hardness 8xlf3 . Metals Park OH. 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. Apparently.. 0 4 AI 1100 ZO z ~ 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. 40 oe: « ::J: 0 0 ll.. 16xlf 3 A 80 ~ Cla -e 4 zlA I 60 Vl Vl .." in Fatigue and Microstructure. Using 3D-MHz longitudinal waves. since the initial amplitude of harmonic generation was small. so that a change of the dislocation-induced harmonic generation. A 2d . A. "New Techniques for Detection and Monitoring of Fatigue Damage.. During fatigue softening.. American Society for Metals. ~ .. Source: O. z 0 -c '" K.H. the dislocation-loop length prior to fatigue was quite short.. Application of this technique to highstrength aluminum alloys failed. the surface hardness (Knoop) decreased...

12-7.. As is true for most alloys. The Technology of Aluminium and Its Alloys. Ltd. notches greatly reduce the fatigue properties of aluminum alloys. England. London. Source: P. C.. Effect on Cycles to Failure 325 50 ~ 40 )( N.. p 43 . Varley.. Alloy 2014-T6: Notched vs Unnotched Specimens. Butterworth & Co. cr: :n VI c: 20 10 Effect of notch on fatigue of 2014-T6. ~ 30 VI . 1970.

with little or no effect at high crack-growth rates.. Alloy 2024-T3: Effect of Air vs Vacuum Environments on Cycles to Failure 40 ~ 35 Ii Q) . the increase in fatigue life caused by vacuum becoming greater at lower stresses. ~ .5 x 10 -3 J.326 12-8..5 x 10. Source: D.. E ::J . J. ... the majority of S... the effect of environment apparently becoming less important at decreasing stresses.. environment appears to be most effective early in the crack-growth process. . however.N curves diverge at decreasing stresses..3.5 x 10 -3 .. as shown in the above chart.. I '~ r"". "0 III .. 30 c Cl '" 2 . In contrast to this behavior.3 ~ 15 s 10 4 6 8 10' 2 En'durance Limit I I I I III"' 4 Number of Cycles-to-Failure 6 8 10 3 The effects of air versus vacuum on the fatigue life of a 2024-T3 aluminum alloy. "~ l/) . For most materials. American Society for Metals.. "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys.N curves at low stresses.Ultrahigh Vacuum 0.... P 337 ." in Fatigue and Microstructure... <. Metals Park OH.§ 20 )( ~~ J 7 2 1... Additionally. 25 .. aluminum and aluminum alloys have been shown to exhibit conflicting results. <.. r- • c - 2. For example.. 1979..Air . a 2017-T4 alloy tested in air and at 2 X 10-6 torr and a 2024-T3 alloy tested in air and at 10-10 torr in rotating bending exhibit convergence of S.. ~.. <.. Duquette......

18 mm (Ys in. Source: Metals Handbook.02 mm (0. 1979.4"-3 0. sheet 1. (5) Bend radius. flattened in annealed condition. 0.12-9..01 Millions of cycles to failure Effects of bending on fatigue characteristics of aluminum alloy sheet. Metals Park OH. (3) Bend radius. 3.1 -----2 1 20 10 0.). The sheet represented by curve I was not bent.) thick was annealed.::... 3. All other sheet was bent 90° in the annealed condition.. 3.). e . Volume 2. Flattening (unbending) was done in either the annealed condition (curve 2) or the solution heat treated and quenched condition (curves 3. Alloy 2024-T4 Alclad Sheet: Effect of Bending on Cycles to Failure 400 350 50 40 'w . American Society for Metals. and then fatigue tested.). flattened in quenched condition after 14 days of storage at -18 to -12°C (0 to 10 OF).1 ~. solution heat treated and quenched. 9th Edition.040 in.. VI' VI en e 100 .18 mm (Ys in.18 mm (Ys in.: 30 327 :E (/) rf. 300 250 200 150 Stress ratio.. (2) Bend radius.59 mm (1/ 16 in.18 to . Properties and Selection: Nonferrous Alloys and Pure Metals. Details of bending and flattening were as follows: (I) Not bent. flattened in quenched condition after 3 days of storage at-18 to-12°C (0 to IOOF). 1.I2 ° C (0 to IO° F)..). P 35 . flattened in quenched condition after 3 days of storage at . 4 and 5). For the data here. (4) Bend radius.

Source: Sigmund Weissmann and William E.1 Mo RADIATION I I PANGBORN e I ill I A /. Otto Buck and Stanley M.. p 195 . Wolf. American Society for Metals. "Determination of Strain Distributions and Failure Prediction by Novel X-ray Methods. / C::P 5 % ~ Q 20 / / / / ~ 10 / / ~ a: a: 8 w / / 0 0 0." in Nondestructive Evaluation: Application to Materials Processing. After correcting for the difference in initial /30 values. exhibit similar behavior throughout most of the fatigue life. HIGH CYCLE (R. although radically different in strain history.0 Dependence of 13 on n/nior low-cycle fatigue and bulk properties of high-cycle fatigue of AI 2024. . Alloy 2024-T4: High-Cyclevs Low-Cycle Fatigue 40 ~ III '5 LOW CYCLE t R -0 2 C. ] C 30 / / / / /.328 12-10. it can be seen in the above diagram that the two fatigue processes. Metals Park OB. Eds. 1984. RAOIAJlON I -_. Mayo.1 1.5 FRACTION OF LIFE In/n.

= 0. Otto Buck and Stanley M..000 12 : Number of fatigue cycles.. '" « > '" e 0' .000 10. Source: Sigmund Weissmann and William E. This was more pronounced for the surface grains (Cr radiation). '" E c:: I~ s c... Alloy 2024-T4: Relationship of Stress and Fatigue Cycles 329 32 30 28 u ~ 26 24 22 20 18 16 14 . the 13 value increased during the first several hundred cy- Here is shown that for the maximum stress of241 MPa with R KO'I cles. 5. Eds. Metals Park OB..z .000 15. "Determination of Strain Distributions and Failure Prediction by Novel X-ray Methods. Mayo." in Nondestructive Evaluation: Application to Materials Processing. American Society for Metals.. P 194 .12-11. 0 VI 0 -.c:: . . .1.000 20. 1984. N Dependence of 13 on number of cyclesN at various stress levels of AI 2024-T4.. Wolf.

' - . o b 14 13 _ _ _ _ _ ~o (inillal ~alf.. Metals Park OH. 19 18 ~-_:_------. Otto Buck and Stanley M.L--_ _.. ..l. American Society for Metals... N F' of Al 2024-T4..330 12-12.. corresponding to the static yield stress... 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.-_ _..----_1 o ICQ.. Mayo.. Source: Sigmund Weissmann and William E....i~_~_) _ 12 IIL.._ _---J o 50 100 I'm 150 200 250 Dependence of the average rocking curve halfwidth 11 on depth distance from surface for different fractions of corrosion fatigue lives.O----. are shown above. Eds. It may be seen that the minimum 13 values at the surface layers were larger than those in the interior. P 193 . "Determination of Strain Distributions and Failure Prediction by Novel X-ray Methods.. and typical results ofthe dependence of 13 on depth distance for an alloy cycled with a = 276 MPa." in Nondestructive Evaluation: Application to Materials Processing.. Alloy 2024-T4: Dependence of the Average Rocking Curve Halfwidth 13 on Distance From the Surface 20.. Wolf... X-ray rocking curve measurements were carried out as a function of depth distance from the surface...--------------------.". 1984..

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. Edward Arnold Ltd.~ <i: ~ 180 + 140 ~ 105 ---'106 Cycles to failure o ---'----' Effect of reducing the concentration of submicron particles in an AI-Cu-Mg alloy.12-13. England. Alloys 2024 and X2024: Effect of Alloy Purity on Cycles to Failure + 331 260 tV a. which is the metastable nature of the metallurgical structure under conditions of cyclic stressing. and American Society for Metals. The disappointing fatigue properties of age-hardened aluminum alloys are also attributed to an additional factor. Metals Park OH. P 40 . J. which causes softening there and leads to a further concentration of stress. so that the whole process of cracking is accelerated. 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.. London. Localization of strain is particularly harmful because the precipitate may be removed from certain slip bands. X2024 is a highpurity version of the commercial alloy 2024. Factors which prevent the formation of coarse slip bands should assist in this regard.. 0 :2: ~ C C CIl CIl 220 CIl Ol . 1981. 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. Light Alloys. The fatigue behavior of age-hardened aluminum alloys would therefore be improved if fatigue deformation could be dispersed more uniformly. Po 1mear.

: o u Q) e ..2 5 8 10 15 20 10. "Fracture Properties of Aluminum Alloys. Campbell.~ e.:::. UI 'u Q) '0.. LiK. '" .. G.332 12-14. . MPa • m 1/2 Comparison of typical particle sizes in aluminum alloys with crack advance per cycle on fatigue loading.: u . Alloys 2024 and 2124: Relationship of Particle Size and Fatigue Characteristics Ii K..:." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. Metals Park OK 1982. S.. Eds.:.. c.. William W..'" ~ 0 en 1.7 10-6 "---'~_ _"""""_ _"""""---''--_''''''''''_'''''''''---' 10. ..s: . Source: J... P 191 .3 u Q) --'" . . ksi • in. Santner. en U. . -~ '" o 20 Stress-intensity factor range. .3 10-5 10.4 10-6 10-5 10. '" 'C 'C > u 10-4 10. Underwood.... Gerberich and John H.. '" 1 . Q) t 'C Q) ... 1/ 2 3 10.5 > e Z -'" 'C 'C z· Q)" s: '" . American Society for Metals.8 u Q) E E -. The above graph represents Staley's work in summarizing the role of particle size on fatigue crack growth in aluminum alloys. Q) :l :l ''. Kaufman and J. Q)" . ~ 0 en . en '~ U. James E.

J AA z 6 10 "0 0 2.. in keeping with the larger grain size.. Source: M. Ritchie.0 102 3 10 5 10 7 10 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).... 1979." in Fatigue and Microstructure.r::: 0 . O. The 2124 alloy studied had 1/ 10 the inclusions of the 2024 alloy studied (0. Fine and R. P 251 .2 vol% compared to 2 vol%) but a larger grain size (45 /-lm compared to 20 /-lm in the transverse direction normal to the loading direction). Thus.8 ~ 2024 T4 oNjj 2124 T4 ANjj Nij =Cycles to ~ 15 fLm Crock ~ 700 600 500 ~ fJl fJl Ql z <.r::: 0 fJl .6 0 0 A A 0 400 300 in .. ~ Z ~ N co 2. "Fatigue-Crack Initiation and Near-Threshold Crack Growth. They also formed more easily in 2124-T4 than in 2024-T4 at high stresses. E.4 AA in CIl . but it is more resistant at low stresses.2 0 0 Ql fJl 2. fJl fJl Ql 200 z "0 0> 0 co A Ao .. E 2. at high stresses 2124-T4 is less resistant to fatigue crack initiation than is 2024-T4. Metals Park OH. slip-band cracks not associated with inclusions formed at the lowest stress studied.12-15. Alloys 2024-T4 and 2124-T4: Comparison of Resistance to Fatigue Crack Initiation 333 Mechanically Polished R=-I 2.. American Society for Metals. as shown in the above chart. With 2124-T4.

..6 ..~ z "C >u '<. American Society for Metals.. Alloys 2024-T3 and 7075-T6: Summary of Fatigue Crack Growth Rates t.K.-. Underwood.5 104 . with only small systematic effects of composition.. Santner. MPa • m 1/2 Summary of fatigue crack growth rate data for aluminum alloys 7075-T6 and 2024-T3. fatigue crack growth rates are found to fall within a relatively narrow scatter band... William W. Campbell..6 10 100 Stress-Intensity factor range.. Metals Park OH.~ Cl U.334 12-16.. Kaufman and J.. fabricating practice or strength level.... 1982.. 10-7 10. of Aluminum Alloys.... In general. U G> u E u ~ 10- 5 Note: Bounds defined by mean curves of separate investigations "C . . S.1/2 10 100 10... "Fracture Properties. as illustrated by the data in the above chart. Eds.. James E. 01 10.. Considerable use has been made of the fracture mechanics approach in measurement of fatigue crack growth rates in aluminum alloys." in Application of Fracture Mechanics for Selection of Metallic Structural Materials..4 2024-T3 8 investigations i Cl ~ 10... Source: J.1 r . ksi .. in.P 189 . G. 7075-T6 9 investigations 10.. AK.. . Gerherich and John H. These data have been generated by methods comparable to those of ASTM Method E647 for measuring fatigue crack growth rates.

> 17) rotating-beam specimens from various product forms of 2024-T4 and 7075-T6 aluminum alloys. the effects of composition and temper are even less pronounced and generally are of no practical significance. . 'q".!. Based on SoN data of smooth and sharply notched specimens and of similar tests of specimens designed to simulate joints in structures. Metals Park OH. .. \ \ " ~----I ::E '" 276 rn lUI a: w rJ) ' 207 1---+---''':-. 1979. T. "... ~---l·~-~~ ..::.~~. 1'" .'.. .:.-I ~ .-. f~.I. Jr. . ~. "Review of Fatigue and Fracture Research on High-Strength Aluminum Alloys.--'~-r-''"' . ~ • ~' r-'. ::.. .c---+----jf----+----j c. .:'.l. Notches have been employed to provide stress concentration. ...\ . Alloys 2024-T4 and 7075-T6: Effect of Product Form and Notches 335 414 345 ::E _ ... and environment. ''':"... airframe manufacturers determined that fatigue performance of alloy 7075-T6 was unquestionably inferior to that of alloy 2024-T3. I---L. American Society for Metals. I.-::'~---+---+---j .-:-l. 1''''.-----+---j--+_ 2024-T4 ---+----1 I '" rn UI e.- 69 : • . Numerous methods have been developed to evaluate response of materials to cyclic deformation. Source: T. . ."~'~r.~ ~-c •• ~."00. axially stressed and sheet flexure. the SoN response for aluminum alloys tends to level out as the number of applied cycles approaches 500 million.. . From fatigue results for aluminum alloys obtained with smooth specimens .. 108 '---->":" -.N response is strongly influenced by a number of conditions.. . For example.• and J.''-...1 ...#_." in Fatigue and Microstructure. As shown in the above S... ~...... When severely notched specimens are used.. H.-... stress ratio. rather wide variations can exist without causing appreciable differences in fatigue strengths. --}.-- .138 ~c-'-cc-:--''''. there is a lack of significant differences in the fatigue strength of the joints of the various alloys.-"'~_ _+-_--j I. "-. " 276 l- a: w 207 I 138 69 0 414 rJ) '"'-I' . and special specimens have been used to simulate a variety of other conditions." : ~gmg ~~~TE • FORGINGS • EXTRUSIONS 010 NOT FAil 1~ "e'. Typical examples are depicted above. :.--+~-t---+- I 7075-16 345 I----i---"~~.. I".. __ _ _ 1~ ! I ...N diagrams. ~. in the number of cycles where a "level out" condition is attained.12-17. -.~":''..'\ ~. .":-- --+ . P 470 . <. 1'"..j...~. 1~ 10' 109 CYCLES Fatigue performance of smooth and notched (K.:~.t-. Despite these laboratory data.J. _ _ .."' ~ • ---~.."-j--""'-f'--. ::""II'jl ':'. including surface condition.s -.' o 1~ - ----L 1~ . Staley. The earliest method was by use of S-Nplots." • t': . The S. the following conclusions have been drawn. Sanders. Basic specimens include rotating beam..§:--. As in the case of simple notch fatigue tests. users discovered that certain aluminum alloys performed decidedly better than others in service when fluctuating loads were encountered. The various alloys differ widely in their response to fatigue testing-specifically. ." ~ <..:.·t'i:::::~" J -.

053 in. * .013 in..I . American Society for Metals.) 400 1-_ _ <0 +-__-+-_ _ ~=1B ---l_+_f --F -+-_ _--+ t -_ _-1 60 50 0..35 mm (0...) 6...336 12-18. Source: Metals Handbook. 0 . Alloys 2024-T351 and 7075-T73XXX: Comparison of P1M Extrusions and Rod 500 r------. X7091-T7E69 in the longitudinal direction.:. X7091-T7E69 in the long transverse direction.l 20 -:::~~~~~"'F...33 mm 10.p 468 ..) K..:::::j::~-__jl--I 100 f -_ _ 2024-T351 rod and bar band - ~"t----I-~ 7075·T73XXX L -_ _---' 10 products band ... • . 6.. 70 1... bar and products...-----.+ .---~---_.X7090·T7E71 in the longitudinal direction.:.._---__r---__..L ..t .I f .....+ .43 mm (0. 10' 3 10 .1 .. Volume 7...r------..denotes test specimen did not fail. = 3 300 f----I------1I-------1I-----j----j-----t-----l ~ 'x <0 E :J E M 'x <0 ::E E :J E ::E 200 1 ..L_ _~ oL -_ _-.L. 1984.. 5 10 0 10' 10' 10' 10' 10' Cycles Comparison of room-temperature axial stress notch fatigue strength ofP 1Malloy extrusions and ingot metallurgy alloy rod.. .t ..253 in... Metals Park OH.l. 9th Edition... Powder Metallurgy..../ .. ::E vi ~ III III Stress ratio: R = 0 Ambient air Notch tip radius: e = 0.

1979. 337 70 I..12-19. 9th Edition. American Society for Metals. 50 ~ '" 40 E :::J 30 20 107 'x 200 '" :i: 100 103 E '" :i: 'x E 10 4 105 106 Numberof stress cycles SoN axial fatigue curves for unnotched specimens of aluminum alloy 2048·T851 plate. P 80 . Longitudinal I • • . Metals Park OH..1 I. :i: '" ::I" 400 '" 60 "" ::I" ~ E :::J '" 300 o 0 I. Volume 2. Properties and Selection: Nonferrous Alloys and Pure Metals. showing effects of R value and direction upon fatigue properties.... Alloy 2048-T851: Longitudinal vs Transverse for Axial Fatigue 500 Q. Source: Metals Handbook.. Longtransverse • • R = 0.

Metals Park OH. 24°C (76 of) 120 °c 1260 of) . Alloy 2048-T851: Notched vs Unnotched Specimens at Room and Elevated Temperatures 500 :::E s: Ii: E ..01 R ~ 0.. T 60 ] 50 24°C 176of) g E 176°C 13600F)~_ -=- /120oC 1260 of) 40 ~ 30 20 :::E ! 1 10 Number of stress cycles . Properties and Selection: Nonferrous Alloys and Pure Metals. P 82 ... .338 12-20. 1979.. 9th Edition.~ ~J 176°C 1360 OF) 30 _ 20 :::E 100 "i :::E E --.1 Number of stress cycles S-N curves for unnotched (upper graph) versus notched (lower graph) specimens of aluminum alloy 2048-T851 plate. :::E g ~ 300 E ~ 200 400 Notched (K t '" 3.1 - ~ s. Volume 2. American Society for Metals.40 g § 50 ] .~ ~ ~ 400 ~ 300 200 Unnotched R ~ 10 0.1 r-.. Source: Metals Handbook.

Source: Metals Handbook.4 1 10. Volume 2. p 81 . showing propagation data for both LT and TL (longitudinal and transverse) crack orientations.2 1 10 10 LT crack oriant!tion TL crack orianLtion " .12-21. Alloy 2048-T851 : Fatigue Crack Propagation Rates in LT and TL Orientations 1 339 AK. 1979.: 1 61 1 10 Fatigue crack propagation in aluminum alloy 2048-T851 plate.. 2 AK.:! 10-6 f M~7. Metals Park OH.: 1 ! 10. Properties and Selection: Nonferrous Alloys and Pure Metals. American Society for Metals. kSI'in.:! ~ . 9th Edition.1 61 10 M-a. ~ .5 G rot" -1i ~ E {.5 ~.. 10-8 / l~3.6 10-5 . kSi'in.5 "iI G .:! 0 ~ ..

ksi ::.. Metals Park OH. 1979. Properties and Selection: Nonferrous Alloys and Pure Metals.340 12-22.~ . American Society for Metals. P 81 . Source: Metals Handbook.: E ~ 200 . 9th Edition. ~ " E 'x ::.: s: 60 ~ 1. Volume 2..: Minimum stress. MPa Modified Goodman diagram for axial fatigue of un notched specimens of aluminum alloy 2048-T851 plate. Alloy 2048-T851: Modified Goodman Diagram for Axial Fatigue Minimum stress. E E " ::.

Larry J. Values used were: (J= 6 X 10. Ed.. which affects the near surface ductility in this alloy. "The Relaxation of Machining Stresses in Aluminum Alloys During Fatigue. . American Society for Metals. Metals Park OH. Morris.J o 10 20 30 40 50 CYCLES (x 10-3) The dependence ofrelaxation behavior on the cyclic hardening parameter..5 for 50% RH and (J= 2 X 10-5 for 0% RH.r-----.l.. 1981.J..------r-------. L. P 184 . Alloy 2219-T851: Dependence of Relaxation Behavior on the Cyclic Hardening Parameter O~-----r-------.88 a yleld for both samples... . R. The cyclic stress amplitude was 0. Source: M.. (J.." in Residual Stress for Designers and Metallurgists.. (J was varied by changing the relative humidity (RH).- ---I..!- ...12-23. 341 o o o 0% RH • 50% RH -300 L-- . Vander Walle.. James and W.

James and W.-------. Ed. Source: M.64 a YIELD -300 o~-----'--------l----..-------l---- 10 20 30 40 . R.------..342 12-24.J-.. Larry J.. L.0004 was used to fit the data for all specimens. Surface milling produced the shallowest stress gradient and resulted in the slowest rate of relaxation of the surface stresses.----. Vander Walle. "The Relaxation of Machining Stresses in Aluminum Alloys During Fatigue.. 0. Metals Park OH. Residual stress measurements were also made in a direction transverse to the applied stress axis.r------.3) The effect of strain amplitude on the relaxation of surface residual stress with fatigue... The solid curves are the predicted mean residual stress values during fatigue. The residual stress values were measured parallel to the external stress axis. A value of f3 = 0. Within experimental error. 1981. A comparison of measured to predicted values of residual stress during fatigue is made for four "as machined" specimens in the above chart. the cyclic relaxation rate was the same as in the longitudinal direction.. The symbols are the residual stress value measured by the x-ray diffraction peak shift technique.7" YIELD o 0. American Society for Metals.. Morris. P 182 .. Alloy 2219-T851: Effect of Strain Amplitude on the Relaxation of Residual Surface Stress With Fatigue Or-------." in Residual Stress for Designers and Metallurgists. 50 CYCLES (x 10.

------r-----r-----." in Residual Stress for Designers and Metallurgists. R. o 10 20 30 40 50 CYCLES (x 10-3) The relaxation behavior oftwo samples having different depth distributions of residual stress. James and W.. Metals Park OH..Surface Stress O.. this involves slip at an acute angle to the surface. 343 o o • • -300 l-- --L -L- .71 ayield' Relaxation of a compressive surface stress requires an expansion of the material normal to the surface. 1981. "The Relaxation of Machining Stresses in Aluminum Alloys During Fatigue.- ----I .. = sand blasting. Of necessity. The effect of humidity on relaxation is therefore simply to make it more difficult for dislocations to penetrate to the surface. Source: M. If the slip does not penetrate the surface.. It is known that humidity increases the rate of cyclic hardening of a thin (less than I Jlm) layer at the surface.. P 183 .• American Society for Metals. a. Supporting this picture are our observations that the relaxation rate in Al 2219-T851 is more rapid in dry air...003.. f3 = 0. Note the difference in the peak cyclic stress.-----r-----. a= 0. the residual stress cannot relax. f3 = 0. Morris.L..12-25. Larry J. L.91 ayield'. Vander Walle.. 0= rolling (10%reduction). Ed. Alloy 2219-T851 : Relationship of Fatigue Cycles to Different Depth Distributions of .012. a = 0.

t' in Fatigue and Microstructure. the dashed lines indicate the percentage offatigue life expended. 1979. Source: O. "New Techniques [or Detection and Monitoring of Fatigue Damage.344 12-26. Alloy 2219-1851: Probability of Fatigue Failure LOG (NUMBER OF FATIGUE CYCLES) Schematic curves of constant probability for failure (actual failure = 100%). Metals Park OH. A. American Society for Metals. p 104 . Buck and G. The solid line in the graph represents failure. The exact location of these lines is highly sensitive to the material and its microstructure as well as the influences of environment. Alers.

weld reinforcement removed.. " 'x 100 10 50 ::.12-27.. Volume 2. 5154-H34 and 6061-T6: Effect of Alloy on Fatigue Characteristics of Weldments 300 40 260 345 If. " .. Source: Metals Handbook. Specimens were from 9. American Society for Metals.. 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. differences disappear until. Properties and Selection: Nonferrous Alloys and Pure Metals. R = O. As the load is decreased. . 30 g E E :t! 20 160 g E E 'x ::. 200 ::.) plate. Number of cvcles The fatigue life of welded joints at high loads varies with the alloy. at about one to ten million cycles of axial loading (R = 0). P 195 . axial loading. 9th Edition. 1979. Metals Park OH.. Typical data are given in the above graph for three aluminum alloys.5-mm (Ys-in. Alloys 3003-0.

dry air . . Source: J.1K." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. . From the data shown above there is obviously no great effect of specimen orientation on fatigue crack growth rates. Cll 'C 10. William W. S.0. u Z 10..) R = 1/3. American Society for Metals. > E E t. Metals Park OH.and 196-mm (7. P 193 .8 in.4 ~ -.l -. Eds. 112 5 10. m 1/2 Effect of orientation on fatigue crack growth rates in 180. Kaufman and J.7-in. "Fracture Properties of Aluminum Alloys. ksi • in.and 7.4 Z . James E. 1982.346 12-28. Santner.6 T-S 10-5 Compact specimen thickness = 46 mm (1.3 u 10-5 10. Cll 'C 'C -. Gerberich and John H.E 'C -. Campbell. Alloy 5083-0 Plate: Effect of Orientation on Fatigue Crack Growth Rates . G.1K.) 5083·0 plate. MPa .. f = 13 Hz Room temperature. Underwood.

7 5 ~K. Alloy 5083-0 Plate: Effect of Temperature and Humidity on Fatigue Crack Growth Rates ~K. Gerberich and John H. 10 MPa • m1/2 50 Effect of temperature and humidity on fatigue crack growth in 180-mm (7.4 10.12-29.3 Ql U E E z· ~ "C ~ . Santner..E III z· ~ "C 10-6 Compact specimen thickness = 46 mm (1.2 RT.L-_--L. Growth rates in water solutions of sodium chloride are similar to those in moist air. G. Source: J.) 5083-0 plate. T-L orientation '-. American Society for Metals..0-in. P 195 . Underwood.I . . James E..) R = 1/3. Campbell. S./ I .8 in. f=13&18Hz. ksi • in... Metals Park OH. 1/2 347 10. As shown in the above graph. Ql 10. growth rates for alloy 5083-0 are appreciably higher in moist air than in dry air. '--~ 10-5 10. 1982..5 ~ U . Kaufman and J. 10-4 . "Fracture Properties of Aluminum Alloys. William w. Eds. • 10." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. moist air~i .

Fatigue strengths of aluminum alloys are lower in corrosive environments such as seawater and other salt waters than they are in air.. such as 5xxx and 6xxx alloys. 7075-T73 and 2024-T3: Comparative Resistance to Axial-Stress Fatigue 0. 1979. Metals Park OH. Volume 2. ~ ~ 0. than in less resistant alloys. . long-period tests. Alloys 5086-H34. 9th Edition..1 5086H34 5086H36 6061T6 7075T73 2024T3 Ratio of axial-stress fatigue strength of aluminum alloy sheet in 3% NaCI solution to that in air.4 ~ . American Society for Metals. Like stress-corrosion cracking of aluminum alloys..5 U .3 a: 0. As shown in the above bar chart. 5086-H36. Properties and Selection: Nonferrous Alloys and Pure Metals.1 kHz 0. 6061-T6. corrosion fatigue requires the presence of water. . corrosion fatigue is not appreciably affected by test direction. such as those of the 2xxx and 7xxx series. In contrast to stress-corrosion cracking.~ 0. because fracture resulting from this type of attack is predominantly trans granular. P 220 . Source: Metals Handbook.2 0. especially when evaluated by low-stress.348 12-30. such corrosive environments produce smaller reductions in fatigue strength in alloys of the more corrosion-resistant types.6 0.7 R~O Frequencv = 1.

V Butt Welds 30 20 2 s~s K ' mol =2. (10-mm) butt welds. American Society for Metals...12-31.: <I 1II 10 8 CT.. Larry J.. • + 18 ksi 10~ 10 I 4 10 NT' Cycles Total fatigue life predictions and experimental results for 5083-0/5183 3/8-in..--I'""T'"TT"n ~ ~ 349 60 400 40 300 200 .n.. V.----.-r.--.--. R =0..0Iln." in Residual Stress for Designers and Metallurgists. Ed. Lawrence. CT. Vander Walle. Alloys 5083-0/5183: Fatigue Life Predictions and Experimental Data Results for Double V-Butt Welds IOOr---r-..=+18kSi 6 4 KI~ 5083 -0/5183 Double . 1981. p 114 . "The Predicted Influence of Weld Residual Stresses on Fatigue Crack Initiation.--I'""T'"T'TTT"--. Source: F.. Metals Park OH. 1= 3/8 .----r-.60. 0ls0.T"T"TT.

p 113 . Source: F. '" 112ir\ 112 7mm) Kt. the aluminum weld considered here (5083/5183) exhibits little dependence upon either residual stress or stress ratio..5. .. Because of the high notch-root plasticity during the first few cycles. Ed." 3. before the material cyclically hardens.0111\ (0.. Larry J. V.."90". "The Predicted Influence or Weld Residual Stresses on Fatigue Crack Initiation. 1981.. Lawrence. American Society for Metals. --..' "SO"... 30 20 10 Predicted effect of stress relief and stress ratio on 5083-0/5183 butt weld fatigue life. even though the relaxation of the stabilized mean stress (uos) is very slow-as indicated in the above chart.350 12-32.2~41Ml) . ' ~-.·0 400 300 200 R"O j 2 S~S 1(.HI.. Metals Park OH." in Residual Stress for Designers and Metallurgists. Alloys 5083-0/5183: Predicted Effect of Stress Relief and Stress Ratio on Fatigue Life of Butt Welds 1OO~"""""""""""'--''''''''''''''''''''''~-~''''''''''''''''''''''''''''''"T''T"r--''''''---'---'''''''''''''''''''''''''600 80 ~3-o1!l183 Bull Weld eo 40 °1" 0. Vander Walle....

.. Source: T.8(440) .and low-strain fatigue may be characterized with one plot.T6 en ... 2N Cyclic strain versus initiation life for laboratory-fabricated high-strength 7XXX aluminum alloys.J f- f0 f- <>: :-~------. 1979...1(511) <>: TYS ks. 10 7 REVERSALS TO INITIATION. but plots oflog total strain amplitude versus life have been used more frequently to compare materials. ::iE 0 :::l f- AllOY 7075·T6 7050·T6 7075·T7 7050-T7 UTS kSi (MPa) 810(558) 887(612) 73./11 -------'-_ _. Sanders. T. "Review of Fatigue and Fracture Research on High-Strength Aluminum Alloys." in Fatigue and Microstructure. and J.2(505) 74.."" 1O-3 ':'":-_ _---'100 -'-':-_ _----J'-_ _--'- """'''''''''''''1.2 /7050.0 a: z <i: 10. American Society for Metals. fl.. This approach offers the advantage that both high. ~ --"""""''''''''""""". fatigue resistance at low total strain amplitude is governed by the elastic-strain amplitude.. On the other hand. As illustrated above. (MPa) 769(530) 82 7(570) 658(454) 63. fatigue lives for total strain amplitudes greater than about 10-2 generally increase with increasing ductility. p 472 .fIN W ::J Q... 2" 95 12./.. 7XXX Alloys: Cyclic Strain vs Crack Initiation Life 100 351 .. Plots of elastic-strain amplitude versus life have seen relatively little use for commercial alloys. Jr..12-33.. Metals Park OH.0 14. Fatigue lives for total strain amplitudes less than about 5 X 10-3 generally increase with increasing strength..0 11. Staley.. H.

2 Low humidity Aged 3 h at 121°C (250°F) + 9 h at 163°C (325°F) Alloy 7050 Alloy 7050 sheet 1.3 1.3 to 6% to 2.4% Mn Aged 3 h at 121°C (250°F) + 9 h at 163°C (325°F) 1.4 %Cu 1.1 0.3 Effect of dispersoid type (based on composition) on fatigue crack propagation life of 7050 alloy sheet. There is some evidence that new processing practices may provide the fine microstructures needed to enhance fatigue resistance. The influence of alloy composition on dispersoid effect is shown in the above bar chart.0 2. "Fracture Properties of Aluminum Alloys.352 12-34. James E.5 Mg 2.1 0. Whereas dispersoid type appears to have a relatively small effect on mean calculated life." in Application of Fracture Mechanics for Selection of Metallic Structural Materials. Gerberich and John H. Kaufman and J. Campbell. Alloy 7050: Influence of Alloy Composition and Dispersoid Effect on Mean Calculated Fatigue Life Zn 5.2 % Zr %Mn 0. The general trend in this chart is that for more finely dispersed particles the fatigue crack propagation life is increased.2% Mg 0.. American Society for Metals. S. Metals Park OH.0 1. Underwood. p 192 .6 0.4 0. 1982.4% to 2.2 Cu 2. The potential of intermediate working (commonly referred to as ITMT treatments) remains attractive but has not been proven for notched specimens. Eds.8 High humidity 0. Santner. the smaller precipitates provided by aging produce a much larger effect.0 2. Source: J.0 0.4% 6%Zn 2.4 0. William W. G.

105 106 CYCLES TO FAILURE Stress-life curves for two 7050 alloys having fine.J Q. Metals Park OH. Jr. grain shape showed no perceptible difference in life over a range of stress amplitudes.400 UJ ::J I.. equiaxed grains (AR) and pancake-shaped grains (HR). o ~ . Alloy 7050: Effect of Grain Shape on Cycles to Failure 353 Q. P 238 . "Cyclic Plastic Deformation and Microstructure. and Gerd Lutjering.12-35.. Starke.. Source: Edgar A. American Society for Metals. o ~ « 3 (J) (J) UJ • 200 • • 7050 AR • 7050 HR II: :n ." in Fatigue and Microstructure. 1979... As indicated in the above graph.

that the promising results mentioned above were obtained for smooth specimens. England. or at relatively soft zones such as the precipitate-free regions adjacent to grain boundaries. TS = tensile strength. and it seems that the resultant stress concentrations override the more subtle microstructural effects that have been described..P 41 ..~ e c CJ) If) If) 0\0 & 200 \ 0 6'h 6"- 0 0 0'0Q. It should be noted. and American Society for Metals.. London.. T6 andT651): Effect of Thermomechanical Processing on Cycles to Failure 300 ~ od~ \ m a.."""'--- ' ""4_ 'lIQ .354 12-36.. 7075 TMP o 7075-T651 67075-T6 PS 587 600 516 488 TS 632 627 573 567 \ \ tl . although this effect arises in part from an increase in tensile properties caused by such a treatment (see above diagram).. Edward Arnold Ltd. Source: I.. The improved fatigue behavior has not been sustained for severely notched conditions...... J.. Polmear. Metals Park OH. Light Alloys.. . PS = proof stress (MPa). <i: ~ E \6. Density has also been found to improve the fatigue performance of certain alloys. It is here that strain becomes localized due to the presence of pre-existing stress concentrations such as mechanical notches or corrosion pits. 1981.... 10 4 Cycles to failure Effect ofthermomechanical processing (TMP).Q.on the unnotched fatigue properties of the commerical AI-Zn-MgCu alloy 7075. ~ \~ \ '6 \ • 7075 TMP .. Alloy 7075 (TMP. --_~~ """'--. Detailed studies of the processes of fatigue in metals and alloys have shown that the initiation of cracks normally occurs at the surface. however.. 6 100 L. coarse (persistent) slip bands in which minute extrusions and intrusions may form.

American Society for Metals. alloy 7475.:) IQ. Metals Park OH. low-inclusion density. 1979. Source: Edgar A.. "Cyclic Plastic Deformation and Microstructure. alloy 7075. Alloys 7075 and 7475: Effect of Inclusion Density on Cycles to Failure 355 Q. ~ c:( 300 :J I/) I/) 200 ILl II:: I/) l- 100 104 105 106 107 FAILURE 108 CYCLES TO Effect of inclusion density on the stress-life behavior oftwo 7XXX alloys: high-inclusion density. ~ 1:1 400 0 ILl . P 233 . Starke.12-37. and Gerd Lutjering. Jr." in Fatigue and Microstructure.

Chubut. There is evidence in the literature that a uniform dislocation density introduced by cold working improves the fatigue life also in connection with FTMT. Eds. Argentina). "Fatigue and Fracture of Aluminum Alloys. The above graph shows an example taken from the work of Ostermann. Metals Park OH.. Embury. American Society for Metals. Pampillo.-----.356 12-38. Gysler.---. p 195 . Biloni and D.I~Nm-' ] u ~ 'E z 7075TMT 600 :g w a: til ~ 300 Cl Z ~ a: w z 200 « ~ 100-'----.--10' 10' 10' 107 10' CYCLES TO FAILURE Influence of TMT on S·N curves (R = -1).. Source: G. 1980. H.-----.. Most of these improvements are due to an increased yield stress.----. C. E. Lutjering and A. Alloy 7075: Effect of TMT on Cycles to Failure n 400 7075 oY." in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn. A.

• . Thus. G. "Fracture Properties of Aluminum Alloys. Alloys 7075 and 7050: Relative Ranking for Constant Amplitude and Periodic Overload 357 2. '0 j Z 1. S. information on the variation in load level during fatigue cycling is required for correct characterization of the fatigue behavior of aluminum alloys.• . and it clearly provides more useful information for alloy-development investigations. Gerberich and John H.tant amplitude life Life Increase due to retardation ::'::) Overload ratio = 1. Overload ratio = 1. P 197 .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.5 X 106 Constant amplitude (Overload ratio = 1. For example.cycles . as illustrated by the data for alloys 7075 and 7050 in the above graph.8 Applied every 4000 . Campbell. Crack-growth retardation is caused by tension overloading during fatigue testing. Underwood. quite different results are obtained in constant-amplitude tests than in tests with single overloads every 4000 or 8000 cycles.. William W.8 Applied every 8000 cycles .• .• . Metals Park OH. Kaufman and J. James E.0 X 106 E ::::J 0. 1982.• • Total life ?< c~ns. The variable-amplitude test is believed to be more sensitive to alloy difference." in Application of Fracture Mechanics for Selection of Metallic Structural Materials.12-39.0 X 106 fi . American Society for Metals. Eds. Source: J.0) Overload ratio =1. Santner.

' .. 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)......' .. (below) under mode 3 loading.. Although total immunity to corrosion fatigue was not observed... thus explaining the reported fracture plane.. Alloy 7075: Effect of Environment and Mode of Loading o '0 -0 ~ 160 AIR ........... The specific details of the embrittlement are not known..358 12-40..'." in Fatigue and Microstructure.... it appears that corrosion reactions liberate hydrogen... but it appears that dislocation transport of the hydrogen is involved........ J.. "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys.. a great deal more research will have to be performed before a more definitive answer will be available.. u u '" ::.. Duquette....' ..I 7 10 o • AIR Nael .. 80 ~ MODE I LDADING '" ..f .._.....0 - ~ a: I.. 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... --_~ 80 o 0 • --0 40 MODE J LOADING • • .. It has been speculated that hydrogen may collect at the semicoherent precipitate-matrix interface.-1. To summarize the aluminum alloy results.... . '" u u >u ::. Metals Park OH.. uJ 120 '" '" a: ...-1._ Fatigue behavior of 7075 aluminum alloy in air and aerated sodium chloride solution: (above) under mode 1 loading... 1979. American Society for Metals..' .' ".... which effectively embrittles the region in the vicinity of a crack tip.... Source: D... .....0...._ 10~ .......... however.....f ..... 0 ....... 120 --- • NaCl ... P 356 ...

Fatigue resistance of high-strength aluminum alloys is severely affected by corrosive solutions. Alloy 7075-T6: Effects of Corrosion and Pre-Corrosion 200 359 ------------------7075 AI T6 0. For example. Metals Park OH.0Zn-2.. indicating at least partial reversibility of the damaging phenomenon and strongly suggesting a solid-solution effect arising from environmental interaction.12-41.. The reduction in life at low Nfis associated with pits which form at nonmetallic inclusions. 1984.5Mg-1..U _ _L I I I l l d _ _ . R.16 "" 12 A AIR B CORROSION C PRE CORRODED/AIR FATIGUE D PRE CORRODED/HEAT TREATED /AIR FATIGUE L.28 A N<.5 M NoCI 276 MN/m (40 ksil mean stress RT . Duquette. Note that re-solutionizing and re-aging the alloy after precorrosion results in a significant increase in fatigue resistance. F. If specimens are pre-corroded and tested in laboratory air. Gibala and R.J. P 265 . equivalent to a low-temperature bake. American Society for Metals. Eds. Hehernann..I_-. Source: D. J. the above diagram shows the results of fatigue tests performed on the 7075 alloy under simultaneous exposure to cyclic stresses and a corrosive environment (curve B) compared to tests performed in laboratory air (curve A).5Cu) indicate that localized hydrogen embrittlement may be responsible for the poor corrosion fatigue resistance of these alloys. there is also a significant reduction in fatigue resistance (curve C).LLL. If the alloy is re-solutionized and aged. especially chloride solutions. "Fundamentals of Corrosion Fatigue Behavior of Metals and Alloys. and this behavior has been attributed either to preferential dissolution at the tips of the growing cracks or to preferential adsorption of damaging ionic species. 24 20 'iii E 150 z ~ l- w cr 100 V> V> V> u u >u :i 50 - ~ .0--4 5 10 106 10 ~ 8 4 0 10 7 The effects of corrosion and pre-corrosion on the fatigue lives of a 7075T6 alloy. Experiments on a 7075-T6 commercial alloy and on a highpurity analog of the alloy (AI-5.l." in Hydrogen Embrittlement and Stress Corrosion Cracking.. a significant amount offatigue resistance is regained.

.

. a crack must initiate to break the protective film to allow access to the bulk alloy. 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. Hehemann.. Source: D. ~ N. -." in Hydrogen Embrittlernent and Stress Corrosion Cracking.12-43. Duquelle.. Eds. 1984. P 266 . "Fundamentals of Corrosion Fatigue Behavior of Metals and Alloys. Gibala and R.. American Society for Metals.. In many cases.. ~" ' ~~' . Cathodic charging of the high-purity analog of the 7075 alloy also shows a reduction in fatigue resistance. Effect of cathodic polarization on the fatigue behavior of 7075 AI alloy in NaCI and Na Z S0 4' It had been previously observed that halide ions are particularly damaging to the fatigue behavior of Al alloys.. In SO~ solutions. sulfate ions prove to be equally damaging. if the alloy is cathodically charged during stressing. . fatigue crack initiation in the equiaxed-grain high-purity alloy is intergranular.. F. 13v NaCI .. J. I~ 10 ~ . particularly at long NJ . / " " 13v Na2S04 . . . Metals Park OH. R. and at more active cathodic potentials there is a tendency toward a higher percentage of transgranular cracking. -.. however.. Alloy 7075: Effect of Cathodic Polarization on Fatigue Behavior 361 2U L .' .

10 10 4 I~ 4 10 NUMBER OF CYClES Tension fatigue test of 7075-T6 aluminum alloy sheet. notch factor K T = 2.'------------~ 4 10 I~ 4 10 NUMBER OF CYCLES lOB Tension fatigue test of 7075-T6 aluminum alloy sheet. }O '" !:1 }I ..37..362 12-44. Alloy 7075-T6: Effect of Surface Treatments and Notch Designs on Number of Cycles to Failure 50 o PAlMlIIC ACIO " ANODIZED AND WATER SEALED 41 ~ ~ 40 ij "j I... .. notch factor K T = 1. ·}D o PAlMIIIC ACID " ANODIZED AND WATER SEALED 21 ~ ~20 ol- " o 0 0 ij ~15 ---------- t....

Syracuse University Press. John J.000. 40 minutes.000 Sheet was anodized: 15% sulfuric acid. of cycles to failure Surface treatment No.800.. 125.251 .000 Anodized and water-sealcd ••• 125.600.600.000 Decanolc acid .000 Palmitic acid ••••••••••••••• 30.000 psi.500. 1964. Eds. 7.000.000 Myristic acid ••••••••••••••• 11.000 Propionic acid •••••••••••••• 2.. Reed and Volker Weiss. 23 °C. Stress amplitude: 26. The three charts show the effects of notch designs and surface treatments on fatigue properties of aluminum alloy 7075-T6 sheet.000 Docosanoic acid ••••••••••••• 6. Burke.000 Lauric acid ••••••••••••••••• 8.000. Kramer.700. the table shows the effects of 17 surface treatments.000 Hexandeiamine ••••••••••••••• 3. Norman L. Surface treatment No. Source: Irvin R.000 Vale ric acid •••••••••••••••• 15.200.000 Dodecyl alcohol ••••••••••••• 7. of cycles to failure Polished .000 Octanoic acid ••••••••••••••• 12. Syracuse NY.000 Dodecylamine •••••••••••••••• 18.000 Caproic acid •••••••••••••••• 9.700.300. "Improvement of Metal Fatigue Lifebya ChemicalSurface Treatment.000 Sebacic acid •••••••••••••••• 13. pp 250.500.000.000 Stearic acid •••••••••••••••• 8.000.363 45 • BARE·PQISHffi o ANODIZm AND NOI WAHR S£AL£D o PALMIIIC ACtO 40 000 o o ANOD Izm AND HOI WATER S£AUD 20 101 6 10 NUMBER OFCYCLES Flexure fatigue test of 7075-T6 aluminum alloy sheet.000 Octyl alcohol ••••••••••••••• 6. 15 amp/sq ft. "in Fatigue-An Interdisciplinary Approach.000.

00 .50l> .. 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....K)1I1 dN (I -R) K c - s« Source: Marc Andre Meyers and Krishan Kumar Chawla. Englewood Cliffs NJ. Keraney. and Engle relation.00 f-- . >. The scatter in the data is much less in the latter..20 • .2 f-- E E I ~10­ <I " C z . The proposed Foreman equation is: da C(t::.g ~ o o 101 _ o o 2 5 10 20 30 I 3 I 10 I 30 AK(MPavm) tJ. a.33e .700 . R< O.. o ..80- 10 2 f-- U U >.K (MPa Ifffi ) Fatigue crack propagation in aluminum alloy 7075-T6 showing the effect of R ratio and the applicability of the Forman.." Prentice-Hall. 10. .. Inc. ~ ~ 1- ~ .50l> ..364 12-45..20x ..33e .. Alloy 7075-T6: Effect of R-Ratio on Fatigue Crack Propagation o 3 1 - R I I I I 10 0.. 10... E E u . 1984..3 f-- lie: ~ It I " ..700 I.... p 716 .80 - o 10. ~ Q) c R 0.1 Q) f-- . "Mechanical Metallurgy: Principles and Applications.

R = O. Argentina).. which compares an undeformed structure with 10% and 20% cold rolled structures.~ 7 10 12 15 20 11K [ MNm-JI2 ) 30 40 50 Influence of predeformation by cold rolling on fatigue crack propagation rates for 7075. 1980.j= 30 Hz. H. Alloy 7075: Effect of Predeformation on Fatigue Crack Propagation Rates 365 VACUUM 10 -2 . Ih 100°C. Liitjering and A. E E z Ci "tl "tl . Chubut. Biloni and D. Source: G. Embury. vacuum. P 207 . American Society for Metals. Gysler. A.I. "Fatigue and Fracture of Aluminum Alloys. Metals Park OH.12-46. SEN specimens. Pampillo. Cold deformation also increases the fatigue crack propagation rate as shown in the above graph. E." in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn.9! ~ 10 oJ u >. C. Eds.

p79 .. Relationships between rate of growth of fatigue cracks and stress intensity for the alloys 2024-T3 and 7075-T6 are shown above. Source: I.-----.---r--____:::> 4 78910 20 1 a« (MPa m / 2 ) Comparative fatigue crack growth rates for aluminum alloys 2024-T3 and 7075-T6 in air ofvarying humidity. 1981. Metals Park OH. Edward Arnold Ltd.. Light Alloys.. Polmear.r-r-. Other 2xxx series alloys show rates of crack propagation similar to tha t of 2024-T3 over most of the range of test conditions.--. since this type oftest avoids uncertainties associated with crack initiation.. In general. Alloys 7075 and 2024-T3: Comparative Fatigue Crack Growth Rates for Two Alloys in Varying Humidity 10" c:--.366 12-47. J.---. and American Society for Metals. It is now common to use precracked specimens to assess comparative resistance of alloys to stress-corrosion cracking. these alloys have rates of crack growth that are close to one-third those observed in the 7xxx series alloys. England. London.

I." in Fatigue and Microstructure. ::?:::?: ""<t: 2 00 LL.' ~~ 3 I. Recent experiments on flexural-fatigue specimens (aluminum alloy 7075-T651) clearly show the potential of harmonic generation for fatigue monitoring. American Society for Metals. the harmonic had increased by about a factor of four.. Metal Park OR. A.12-48. "New Techniques for Detection and Monitoring of Fatigue Damage. -' . Alloy 7075-T651: Fatigue Life as Related to Harmonic Generation FATIGUE LIFE EXPENDED 367 50 o 30*> 60* I.LI::?: =>"" -'<t: <t:::c 2 L58 a. The above chart shows the peak value of the harmonic generated as a function of fatigue life. Buck and G... Source: O. Alers. At 60% ofthe fatigue life expended. 1979.. P 137 .LI 0 l:::Ji=? 4 <t:a.LI Vl >0 >£2 00 1000 FATIGUE CYCLES 2000 Peak value of normalized second harmonic generation as a function of fatigue life..

The LSP specimens showed three times better fatigue lives on the average and much less scatter than the unprocessed material.000 psi.1 1 2 3 n.700 497." in Source Book on Applications of the Laser in Metalworking. Source: William F.63 NO.t CONTROL SPECIMENS SPECIMEN NO.250 1. whether or not it is laser-shock processed. 1981.300 75. The specimen blanks were laser-shock processed.5inches wide and approximately 9. OF CYCLES TO FAILURE 41. and an R= 0. SCATTER 7. and then the 0. SCATTER AVG. The fatigue test results for the 7075-T6 material are summarized in the upper tabulation.1 under constant-amplitude load control.25 inch thick by 1. Alloys 7075-T6 and 7475-T73: Effect of Laser-Shock Treatment on Fatigue Properties 0_ . The diameter of the laser-shock-processed area is three times the fastener hole diameter. The results for the 7475-T73 material are summarized in the lower tabulation.950 1. Jr. 1 2 3 LASER-SHOC KED SPECIMEN NO.800 520. OF CYCLES TO FAILURE LASER-SHOCKED SPECIMEN NO. SCATTER FATIGUE TEST RESULTS FOR 7475-T73 ALUMINUM "MAX = 20 KSI NET R= • I CONTROL SPECIMENS SPECIMEN NO. OF CYCLES TO FAILURE 473. SCATTER The fatigue test specimens were 0.033 2. This is largely due to the differences in dislocation/precipitate interactions that result from the T6 and T73 heat treatments. NO.25-inch-diameter hole was bored through the center of the laser-shock-processed area.75" 0.500 AVG. Metzbower.500 74. 1 2 NO. Edward A.25" OPEN HOLE . these show the same typically large increases in fatigue life and reduced scatter. Ed.600 51. Bates. Three control specimens for each material were tested to establish the typical fatigue life for the material.0. LASER SHOCK PROCESSED AREA 2 (26 J/cm / SIDE) Fatigue Test Specimen Geometry FATIGUE TEST RESULTS FOR 7075-T6 ALUMINUM "MAX = 20 KSI NET R = . OF CYCLES TO FAILURE 171. All of these open-hole specimens were fatigue tested to failure at a maximum net section stress of20.pp 256-258 . All of the specimens had machined surfaces of less than 125RMS. Dr.loo 385.5 AVG.5 inches long.368 12-49.800 266.000 169.. The precipitate particles are apparently so strong in the T73 condition that the dislocations just loop around the particles. 1 2 NO.300 109..5 AVG. The dislocations appear to shear through the precipitate particles in the T6 condition.200 218. Two LSP specimens were tested for each material to establish the degree of improvement due to the laser-shock processing. as shown in the above sketch. It should be noted that the 7075-T6 material shows better fatigue resistance than the 7475-T73 material. "Laser Shock Processing of Aluminum Alloys. Metals Park OH. American Society for Metals.

Jr.... All of the fatigue origins occurred at or near the hole wall corners on the faying surface." in Source Book on Applications of the Laser in Metalworking..072" . Bates. Dr.. 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. Ed.. 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.. Source: William F. Metzbower.. Metals Park OH..... "Laser Shock Processing of Aluminum Alloys.. Alloy 7075-T6: Effect of Laser-Shock Treatment on Hi-Lok Joints 3/16~ HI-LOKS & TAPER-LOKS (4 PLACES) 369 0- -0 c:::::=:=========$$:Jr==== Joint Fatigue Test Specimen Geometry 6 10 .. Edward A.-. At each stress level.. American Society for Metals. 1981... 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..12-50.. IZZI SYD HOLE ~ LASER SHOCKED HOLES 5 LOAD CYCLES TO FAILURE Iff 5 "MAX = 25 KSI NEJ "MAX = 17 KSI NET "MAX· 14 KSI NET Fatigue Test Results for Laser-Shack-Processed 7075-T6 (Clod) Hi-lok Joints The full load transfer joint shown in the above sketch was made from 7075-T6 clad aluminum alloy and fatigue tested.--== 0=----. pp 262-263 . three specimens with standard holes and three specimens with laser-shack-processed holes were tested..

" in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn. 1980. Pampillo. Argentina). p 193 . Embury. A. Source: G. E. Biloni and D. Gysler. Chubut. This is important because the removal of these small inclusions would have the opposite effect on fatigue life. Eds. The large Fe.----r-----.and Si-containing inclusions are detrimental to the fatigue life of smooth specimens.-10' 10· 10' 10' 10' CYCLES TO FAILURE Influence of Fe and Si content on SoN curves (R = -1).370 12-51. Comparing two alloys. one containing these inclusions (Commercial Purity 7075) and the other one not (High Purity 7075). Metals Park OH. shows the improvement in fatigue life due to the removal of these inclusions (see the above SoN curves).-----..• American Society for Metals. because these inclusions serve as easy nucleation sites for cracks. C. The alloy termed High Purity 7075 in this figure still contains Cr and therefore the small Cr-containing inclusions. LUtjering and A. "Fatigue and Fracture of Aluminum Alloys. Alloy 7075 (High Purity): Effect of Iron and Silicon on Cycles to Failure n I 400 :::E u If) If) z E 300 UJ Iii o a: ~ a: Z ~ Z 200 UJ < 100-'--.-----. H.

. 1980. 371 ::r 300 UJ ::> . The resulting improvement in fatigue life due to the grain size reduction for this crack nucleation mechanism is shown in the above S-N curve.. Pampillo. Gysler. American Society for Metals. Embury. Biloni and D. "Fatigue and Fracture of Aluminum Alloys.----------'-=--------~. C." in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn. H. A.. Also for low-cycle fatigue it was found that red ueing the grain size of 7XXX series alloys results in increased fatigue life of smooth specimens in the averaged condition. ::::.- o __ CYCLES TO FAILURE Influence of grain size on S-N curves (R = -1. Argentina). or 320 OF). f= 100 Hz) for X-7075 with PFZ (20 hat 160°C. a ::r « Vl Vl 0. E.. Metals Park OH. Again. Lutjering and A. Chubut. the tensile yield stress was equal for both grain sizes.12-52.. Source: G. Eds.----------~. p 192 . Alloy X-7075: Effect of Grain Size on Cycles to Failure o 0. 200 UJ a:: Vi 100L~.

Since the flow stress is determined by the interaction of dislocations with the coherent precipitates. 1979. ::E 0 w a. P 225 . and Gerd Lutjering. 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.. Alloy X-7075: Effect of Grain Size on Stress-Life Behavior 400 " c. Opti. Jr. Source: Edgar A. Metals Park OH. 24 h at 100 °C (212 OF). cal examinations of the specimen surfaces show that cracks nucleate much earlier in specimens with the large grain size. "Cyclic Plastic Deformation and Microstructure." in Fatigue and Microstructure. Starke. the yield stress is approximately the same for both alloys.. Cracks nucleated at intense slip bands for both grain sizes. ::E <I: lI) lI) 300 ::::i ~ ll) l- 0: w 200 CYCLES TO FAILURE Aluminum alloy X-7075.372 12-53. American Society for Metals.

Source: G... X-7075.....I.. Argentina). The aggressive environment has a much more pronounced effect on the underaged condition. Pampillo.--.. H. American Society for Metals. A.---.------r----r--. CT specimens. Gysler.. Chubut..L. LOtjering and A.. To illustrate this point. P 204 .. E u Z E 10 :E a '0 -i 10 I06. C. the above graph shows the comparison between underaged and overaged microstructure also for tests performed in laboratory air.f= 30 Hz.. In laboratory air the cracks propagate still along slip bands at low dajdN rates..--r--..----... Air vs Vacuum 373 A: 24h 100'C -I VACWM 10 n C:48h 180'C AIR .... Embury. Alloy X-7075: Effect of Environment..R = O. u -) ~ 10 _. Influence of environment (laboratory air) on fatigue crack propagation rates for underaged (A) and overaged (C) condition. 1980. Metals Park OH..12-54." in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn. leading even to an opposite ranking of the alloy conditions.--..... Biloni and D. A basic correlation between microstructural parameters and fatigue crack propagation rate can only be determined so clearly if the tests are performed with the exclusion of any aggressive environment. E.. Eds. "Fatigue and Fracture of Aluminum Alloys.

!2 >- E -4 E 10 z ~ . The influence of environment is larger for the large grain size.5% NaCI) on fatigue crack propagation rates for two different grain sizes. 24 h 100 DC.5% NaCI solution (note above curves)./~ n 3..5NoCI 10 I06." in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn.lJrn GS 46. CT specimens. 1980.L----. For this highly aggressive environment the cracks propagate at low dal dN rates along grain boundaries in a complete brittle fashion. Chubut.1. A. K [MNrn-3I2 ) Influence of environment (3. The same tendency is observed for the grain size dependence of crack propagation ifthe tests are carried out in a 3. Gysler.1= 30 Hz.7 10 15 20 30 40 50 t:. Pampillo.-----r-----.---'-r---. Biloni and D.---..R = 0. Argentina). "Fatigue and Fracture of Aluminum Alloys. Metals Park OH. Source: G. Lutjering and A. Embury..----. C.374 12-55. X-7075. Alloy X-7075: Effect of Environment on Two Different Grain Sizes -2 10 GS 200.-----..g -5 . American Society for Metals. E. H.lJrn ~ u 10 -J . P 204 .. Eds.

Metals Park OH. American Society for Metals. ~ ~ 200t------O-~~--""'~-+__---____l " X-7075 150 t----::-----±-:-~-:-----t-----_l 0---0 If = 0.. c 300 r--~"-T'""". one cold-worked 50% to produce grain-boundary steps. the effective slip length within the PFZ is drastically reduced (similar to a small grain size) with corresponding improvement in resistance to fatigue-crack nucleation.. and Gerd Lutjering. crack nucleation occurred at inclusions for both alloys. Alloy X-7075: Effect of Grain-Boundary Ledges on Cycles to Failure . . :J Q. One method that may be employed to reduce the slip length in the PFZ is thermomechanical processing.-r'1""""'-r---'----r-"'T"""T"T"l"T~---'----"--"-"""""""'" co. 8h 160°C --If = 0.. Source: Edgar A.4h 160°C 105 CYCLES Effect of grain-boundary ledges on the stress-life behavior of an alloy containing nonshearable precipitates and PFZ. At low stress amplitudes and long fatigue lives. "Cyclic Plastic Deformation and Microstructure. Starke.5. The cold work drastically reduced the incidence of grain-boundary cracking and improved the fatigue life at high stress amplitudes... •• \ o LLJ \ 'PO \ ~ 250 I---~ID-----':~_----_--t-----____l b ~o.- 375 ! Q." in Fatigue and Microstructure. If enough cold deformation is employed to introduce steps (or "ledges') into the grain boundaries. P 230 . The above chart shows the results of a stress-controlled test for two high-purity 7075 alloys.12-56. 1979. Jr... This effect is most likely due to stress concentration at inclusions..

Eds. CTspecimens. Metals Park OH.1. "Fatigue and Fracture of Aluminum Alloys.. they normally increase fatigue crack propagation rates also indirectly by their effect on grain size and shape.! = 30Hz. Biloni and D.R = 0. Source: G.376 12-57. E." in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn. Chubut. 1980. Furthermore. C. Alloys X-7075 and 7075: Effects of Chromium Inclusions on Fatigue Crack Propagation -1 24 h 100·C VACUUM 10 n ~ u -) E E u ~ '" 10 z 10 -. H. Embury. Lutjering and A.24 h 100 °C. Gysler. Pampillo. -5 ~ "0 a 10 -6 10 7 40 50 Influence of Cr-containing inclusions on fatigue crack propagation rates by comparing aluminum alloys X-7075 and 7075. the small inclusions have a much stronger influence on fatigue crack propagation because they lower the reversibility of slip and they crack within the plastic zone ahead of the crack tip. vacuum. P 207 . A. Argentina). American Society for Metals. As shown above.

"""'"-"U 10' CYCLES TO FAILURE S·N curves for a superplastic aluminum alloy: fine-grain 7475..... Alloy 7475-T6: S-N Diagram for a Superplastic Fine-Grain Alloy 80 377 r------------------------------.low luporpl.. F... o SPF Conditions: T= 516·C i=21104S-1 70 = As pr.. 1985. American Society for Metals.I... "Superplastic Forming Applications to Bomber Aircraft .......I-._ 1~ ....stic st"inl o = SPF ..~ 10' 10' 10' . Suphal P.... All testing was done with smooth specimens._"'-.I..I.. Metals Park OH..ilu" 60 R =+0.... p 77 .._. An even more dramatic improvement is obtained in damage tolerance.. Agrawal. Source: C....highsup..._. (]I o _ -....rpl..I-I..I. Ed.....stic st"inl -+ = No.. Tests on fine-grain 7475 alloy have shown improved fatigue life as superplastic strain is increased.."""".I. _-------0-+0-+ 3D 20 I-........."""'"-"u..d A = SPF ..12-58.. 0 .... McQuilkin and G. Bampton..1 AS PROCESSED S·N CURVE MAX STRESS 50 (Ksi) 40 AOC>AO o 6. as shown in the above SoN diagram....." in Superplastic Forming.._. Stacher.

Metals Park OH.. there will be no shear stress parallel to the grain boundary. If the stress axis is parallel or perpendicular to the long grain dimension.. and Gerd Lutjering.J Q." in Fatigue and Microstructure. Alloy 7475: Effect of Alignment of Grain Boundaries on Cycles to Failure 300 0 .. "Cyclic Plastic Deformation and Microstructure... Starke.. -. American Society for Metals. P 232 .. I I CYCLES TO FAILURE Effect of alignment of grain boundaries-and alignment plus steps in grain boundaries-on the stress-life behavior of a 7475 aluminum alloy containing nonshearable precipitates and PFZ.378 12-59.. 0 Q. Grain-boundary alignment is then as effective in restricting deformation in the PFZ as are steps produced by thermomechanical treatment: this is shown by the stress-life curves in the above graph. and preferential deformation within the PFZ will be restricted.. Source: Edgar A. Jr. 200 ! ~ <{ I/l I/l UJ 0 0: II) ~ 150 7475 0 o I 1..J ~ 250 I I -I UJ 0 :J ~ o <9=0 -<9=0.5 16h 160°C 6h 160°C . 1979.

_.J.1 1..L.L..-.... "Superplastic. F.....5 ~ ::I: (. 1 X 10... Forming Applications to Bomber Aircraft.l.L. Ed._. 1985.. . McQuilkin and G.) (J) . p 77 ... as Related to Fatigue Crack Growth 1 X10-] r . coarse-grain.I...6 1 X 10.J...J.. Bampton. Source: C.JL... nonsuperplastic aluminum alloy 7075 with superplastic alloy 7475 shows almost an order-of-magnitude reduction in crack growth for the superplastic material. o Z < o 1 X 10... Agrawal..I.r . ...J..._.) w 7075 T-6 - Z <..----I. American Society for Metals..... Stacher.12-60. 379 7475 T-6 FINE GRAIN ~ 1 X 10-' (..J 1 10 100 1000 K (I-R)M-l (KSI "'lINCH) This comparison of conventional.l----JL." in Superplastic Forming.... Alloy 7475-T6: Superplastic vs Nonsuperplastic.----I.. Metal Park OH..... Suphal P......

thus retarding crack nucleation as indicated by arrows on the graph. ~ ~ 300 :::. These small inclusions.. A. Source: G. LUtjering and A. inhibit the formation of intense slip bands. Metals Park OH. 24 h 100 0 C. P 195 . Gysler. which also may have contributed to the observed improved fatigue behavior. E. American Society for Metals.------------'--r:---""""'----""'-I I ~ 200 10 3 10' 10 5 la' I • CYCLES TO FAILURE Influence of Cr-containing inclusions on SoN curves (R = -1.. as in the tensile test. Chubut. Argentina). the grain size of the 7075 alloy was somewhat smaller as compared to that of X-7075. H. 1980.1= 100 Hz) comparing aluminum alloys x-7075 and 7075. Eds." in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn. Embury. Pampillo. j The above S.--. <l: ~ J~----. Biloni and D. C.N curves compare results obtained from testing commercial 7075 alloy with the alloy X-7075 which does not contain Cr. Due to these small Cr-containing inclusions. "Fatigue and Fracture of Aluminum Alloys. (Arrows indicate crack nucleation visible by LM at u a= ± 200 MNm -2. Alloys X-7075 and 7075: Effect of Chromium-Containing Inclusions on Cycles to Failure 400 :g IQ.380 12-61.

Eds. Williams and K. The results show good fatigue properties for squeeze-formed material." in Production to Near Net Shape: Source Book. it not being possible to obtain samples of sufficient size from the transverse direction.12-62.~~o x In 10 3 . Fisher. not from separately made testpieces. 1983. M. Aluminum Forging Alloys: Stress Amplitude vs Reversals to Failure 381 "t z E ~ 400 SF1-T6/T73 2. o • x 7075-T73 squeeze formed SF1-T61T73 squeeze formed 6082-T6 squeeze formed 6082-T6 extruded bar §300~-T6 S ~ ~ ~200 ~ . Van Tyne and B. C. which in one case compare favorably with conventionally extruded material. The above chart presents results from push-pull. This further substantiates the claim that squeeze formings in general are comparable with forgings with respect to mechanical performance. Source: G. J. Avitzur.. p 367 . about mean zero. The tests have been carried out on samples cut from actual components. the data are in the longitudinal direction. 10 4 10s 10 6 REVERSALS TO FAILURE (2N. The results from conventionaJly extruded AA 6082 (H30) are included for reference: in this case. "Squeeze Forming of Aluminium-Alloy Components. Metals Park OH.: IJ. fatigue tests which have been carried out on a servohydraulically controJled machine.) SON fatigue data for several squeeze-formed forging-type aluminum alloys compared with extruded AA 6082-T6. American Society for Metals.

5%Ag stress IMPa) strength IMPa) 0 150 ST and quenched Aged 1 day 175°C • 85 200 175 L60 310 270 lU a. 1981.5Ag: Effect of Condition on Fatigue Characteristics 180 ~------'--------r------.2% proof stress may be raised to 200 MPa after aging for one day at 175°C (350 OF). AI-5Mg-O. 0. (/] Q) X Aged 70 days 175°C ~ ~ (/] (/] 120 • Cl '16 c c 90 ~ <i: &0 ~ oho . which approximately equals the 0. Light Alloys. It is well known that binary AI-Mg alloys such as AI-5Mg.SAg in different conditions. Polrnear.. P 42 . and American Society for Metals..382 12-63..~ ~~ . which minimizes formation of coarse slip bands during fatigue. England. The same applies for an AI-5Mg-0. Source: I. and the 0. 0_ ~ 10& 7 10 87 X 72 48 ~ lOll Number of cycles Fatigue (SoN) curves for the alloy AI-SMg-O. The silver-containing alloy responds to age hardening at elevated temperatures due to the formation of a finely dispersed precipitate. The fact that microstructure can have a greater influence upon the fatigue properties of aluminum alloys than the level of tensile properties has been demonstrated for an AI-Mg alloy containing a small addition of silver. London. display a relatively high level of fatigue strength. Edward Arnold Ltd.2% proof Tensile AI-5%Mg-0.----. in which the magnesium is present in solid solution. This result is attributed to the interaction of magnesium atoms with dislocations. J.5Ag alloy in the as-quenched condition.. and the above diagram shows that the endurance limit after lOS cycles is ±87 MPa.2% proof stress.. Metals Park OH.

. and this affects the life improvement due to the fine grains..... 24h at 150GC . the AIZn-Mg-Zr. Jr. The improvement in life is attributed to increasing the cycles to crack initiation. P 228 .. The above chart shows Coffin.. 1979.... The small-grained AI-Zn-Mg-Zr alloy has a much longer life than does the large-grained AI-Zn-Mg alloy.... (\J -....Manson life plots of two averaged AI-Zn-Mg alloys. as indicated in the chart... the stress to enforce the applied strain is greater at long lives... _.. Metals Park OH.&> . .:~~ Q. ' .-'. GI .... small grain size. " ". ::--. Crack Initiation AI-Zn-Mg-Zr 10 2Nf Effect of grain size on the strain-life behavior of an alloy having nonshearable precipitates plus PFZ...." in Fatigue and Microstructure.. Starke.. . A convergence is noted for long lives (low plastic-strain amplitudes) for this strain-controlled test.. AI-Zn-Mg and AI-Zn-Mg-Zr: Effect of Grain Size on Strain-Life Behavior 100 383 r----------------. <l o Failure AI-Zn-Mg • Failure AI-Zn-Mg -Zr .. American Society for Metals. The AI-Zn-Mg alloy had large grain size." . :.... and Gerd Lutjering. Source: Edgar A. "Cyclic Plastic Deformation and Microstructure..::: .12-64.... Since the fine-grained material hardens more than the other at low strains.':::.. a. <. e-- .

AI-Zn-Mg: Strain-Life Curves of a Large-Grained Alloy 10. this method is not effective for large-grained material. for this particular case. overaging the matrix precipitates or adding dispersoids does not homogenize the deformation.. Since the strain localization occurs in a region free of solute. or 250 OF) and nonshearable precipitates plus PFZ when overaged (96 h at 150 "C.... Source: Edgar A. . however."in Fatigue and Microstructure. Jr.. Starke. AI-Zn-Mg 0. in the same fatigue life. 1979. and Gerd Liltjering. wiN c. or 300 OF). and early crack nucleation under cyclic loading.1 10 1 Strain-life curves of large-grained AI-Zn-Mg alloy having shearable precipitates when underaged (4 h at 120 ° e.384 12-65. American Society for Metals... Metals Park OH. Overaging was one method described for homogenizing deformation.. Dispersoids distributed throughout the matrix would not inhibit strain localization in the PFZ for the same reason. The tensile yield strength and strain to fracture are approximately the same for both specimens. which results in strain localization. the formation of intense slip bands. "Cyclic Plastic Deformation and Microstructure. • 4h at 120°C . P 227 ...96h at 150°C <l I. The underaged alloy has shearable precipitates. Preferential deformation in the PFZ also leads to strain localization and results.. 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).

under conditions where initiation of fatigue crack is more important than its propagation). at high N's). Source: Marc Andre Meyers and Krishan Kumar Chawla.12-66. <.. 0 _-.. 0. It would be expected that the treatment suggested above will produce a great effect in large fatigue life regimes (i. 1984. N...e. Aluminum With a Copper Overlay: Stress Amplitude vs Cycles to Failure 385 - 70 b g Q.... increases with decreasing load amplitude (i.. The above graph shows this phenomenon in the case of pure aluminum and aluminum with a copper surface layer. 40 10 4 10 7 Stress amplitude U a versus number of cycles to failureNj for AI and AI with a Cu layer. / NJ . P 707 ... 10 5 10 6 Al (Cu) <..----<.. Englewood Cliffs NJ. Note the pronounced improvement in the latter at large N j • The fraction of fatigue life spent in crack nucleation. <. 60 ~ D ." Prentice-Hall. ~o 0 c- 50 +1 o Al D <.e. "Mechanical Metallurgy: Principles and Applications.. Inc .

. 100 10 Ol-_ _-'-_ _--'--_ _--'-_ _-'-_ _--'-_ _--'-_ _--'_ _--. . "Wrought Aluminum PIM Alloys. Advantages and Limitations..... Ed... 7075 and 2024 (an 11M alloy often selected for its resistance to fatigue) at one million or more cycles." 300 Notch tip radius = 0.013 in. The notched axial fatigue strengths of alloys 7090 and 7091 are 35 to 40% higher than those of alloys 7050." in Powder Metallurgy-Applications. E OJ E 30 200 E OJ E ~ 20 Band for 2024·T351 rod and bar Stress ratio R " 0 Ambient air 'x ::. Metals Park OH.386 12-67.I ndicates did not fail /to:D ::. American Society for Metals. Graham.0 4 "'" 40 . o.. ~- ~ 'x ::. 1983. Source: Robert H. KT = 3... P 240 .J 10· 10'0 10· 10· 10' 10' 107 10' Cycles S-N diagram that provides a comparison of notched axial fatigue strength for P 1M alloy 7090 and 7091 extrusions vs 11M alloy 2024-T351 rod and bar. P/M Alloys 7090 and 7091 vs Extruded 2024 60 400 50 'in o l> 7091·T7E69 (2 lots) 7090·T7E71 Open symbols-longitudinal Solid symbols-long transverse 60° O~.. Erhard Klar. .

Source: Robert H.-------.---~-------. "Wrought Aluminum PIM Alloys..-----. Graham. a.• American Society for Metals. Notch tip radius KT = 3. Ed.-------'-c-----'-:c-----L.013 in.----. Advantages and Limitations. E :> 100 10 O'----:----'-.------' 10 7 103 10' 10" 10· 10 9 10 2 10 10 Cycles S-N diagram that compares notched axial fatigue strength for P 1M alloy 7090 and 7091 extrusions vs 11M 7050 and 7075 products.12-68. 1983.-------. The notched axial fatigue strengths of alloys 7090 and 7091 are 35 to 40% higher than those of alloys 7050..400 387 o 50 A 7091-T7E69 (2 lots) 7090-T7E71 Open symbols-longitudinal Solid symbols-long transverse Indicates did not fail O~" 300 . P1M Alloys 7090 and 7091 vs 11M 7050 and 7075 Products 60 r-----.---'--:---'----:----'-. as shown above.----." in Powder Metallurgy-Applications. Metals Park OH. 2: ~~ 200 30 ~ E :> E 'x 2: 20 Stress ratio R = 0 Ambient air' Band for 7050 and 7075 products .7075 and 2024 (an 11M alloy often selected for its resistance to fatigue) at one million or more cycles. Erhard Klar.0 = 60° ] g E 40 0. p 240 .

P1M Aluminum Alloys: Typical Fatigue Behavior 9·71B-in. American Society for Metals.388 12-69. Ed. Montgomery.. The above S-N diagram shows typical fatigue behavior of specimens of alloys 6OIAB. "Aluminum PI M-Properties and Applications..~ 20 E 60 1AB-T2'>." in Powder MetallurgyApplications. p 214 . Metals Park OH. Fatigue is an important design consideration for P / M parts subject to dynamic stresses. 20lAB and 202AB in the T2 (as-cold-formed after sintering) and/ or T6 tempers. Source: John D. N OL------''------''-----'----'-----'-----'----' 10' 10' Typical fatigue behavior of alloys 601AB. 1983.::::===_=====a=_=_=_=_=_-1 601A~ 10' 10· Cycles to failure. x '" 2' . Erhard Klar. R o o o Forged £-=fL-i-----3-"1!I 0. diam ?: 'u. Generous and Wayne C. 202AB_T2~ _ _. x 40 Forged 601AB-T6 201AB-T6 201 AB-T6 c. 201AB and 202AB.300-in. Advantages and Limitations.

::E '" Ii ~ 300 1..12-70. E E ::> . _ _ 8. diarn.) : 60· sharp' V 70 in.3. ~ 3 ] 0 40 0 ~ 1. longitudinal direction Dashed lines represent bands for 2014-T61 (nine lots smooth. Volume 7.30 ln.075 -1. . Powder Metallurgy..t tf--+-1= I -. For P/M X709I-T7E76: 0 . • Notch tip radius: e ~ 1. Metals Park OH. longitudinal direction . • . five lots notch I..) (0.E= f-. transverse direction.I diarn.01 in.denotes test specimen did not fail in number of cycles indicated.39 mm (0. 1984. Source: Metals Handbook. ~ 1.9.. transverse direction. 10' Cycles -• a--20 --10' -~ 10 0 10' 10' 10' 0 10' Rotating-beam fatigue strength for die forgings of P 1M alloy X709I-T7E76 and ingot metallurgy alloys 7075-T7352 and 2014-T6I.19 mm (0.480 in.'" E ::> 'x '" ::E 200 Smooth 0 '" ::E 30 'x E 100 Solid lines represent bands for 7075·17352 !three lots).. E--t--].) I 50 0.) 389 400 7.notched. . P1M Aluminum Alloys: Comparison With Specimens Made by Ingot Metallurgy 500 251 mm (9'10 in. American Society ForMetals.p 469 .62 mm .300 in.smooth.254 mm (0.) K.mm (0. 9th Edition.0 60 12.

two lots. 1984. 9th Edition. longitudinal direction.we Band for 100 _forgings (six lots) Stress ratio: R = 0.1. Stress ratio: R = 0. .] 70 60 ul =3 - 50 'iii -" ul e 1il :> Ul 300 e 1il 40 :> Ul 'x co ::E 200 E E E • m:a 0 -. • .L Notch tip radius: K. ~t~ e = 0.P 469 .500 t.390 12-71. Source: Metals Handbook.."" J. Volume 7. one lot." 400 L10.)' in. Metals Park OH.013 in.33 mm (0.) mm 10. 0 . P 1M Aluminum Alloys: Comparison With Forged 7175 for Cycles to Failure . Powder Metallurgy.denotes test specimen did not fail in number of cycles indicated.0 o 10' r r 103 10' 7175-T73~ ~ 10' • • •It o • • •~ • 'x co 30 E ::E • ~ 20 10 o 10' Cycles 10' 10' 10' Comparison of axial-stress notch fatigue strength of P 1Malloy X7091-T7E69 die forgings and ingot metallurgy alloy 7175-T736 die forgings. short transverse direction. American Society for Metals.~.

A.0 .T 6 5456 -H117 5456-H321 6061 . H.30 Hz ambient temperature 11 AI . Chubut. 1980.T736 7475 -T651 ~ 2cro . Various Aluminum Alloys: Comparison of Grades for Corrosion-Fatigue Crack Growth Rates.T specimens: SEN . Speidel. E. P 615 . 0. DCB . 6K.T63 7050 -T736 7fJ15 ." in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn.T 651 50 60 iltensity range. Metals Park OH.T851 2219 . American Society for Metals.alloys in salt water: 2048-T851 2219 .1 .T 87 2618 -T6 5456-H117 5456-H321 6061 -T651 7005 . [MN 'm-¥2] Comparison of scatterbands of corrosion-fatigue crack growth rates and fatigue crack growth rates of many commercial aluminum alloys.T 63 7175 .0. Embury. Pampillo.. Air vs Salt Water 391 '(f3_-------------------"'" crack orientation L . GNP.alloys in air: 2048 . CT R • 0. Biloni and D.T651 2024 . C.T651 7'005 .T6X31 30 40 7050 . "Aluminum as a Corrosion Resistant Material.T651 7175 .T 3 2024-T~ 18 AI.T87 20 cyclic stress 2618 .T 6 7079-T651 7106 .12-72. Source: Markus O.T 736 7475 .T 736 7075 .T851 2219 .T63 7039 . Argentina). Eds.1 .

p 614 .1 I 0. E..5 E '------' 6061 -T651 ~~ 10 -6 2048 . curves for growth rate are somewhat higher than the air-test scatterband.. Speidel. C.T736 5456 .30 Hz -10 10 0.----.T6 7175 .T851 7050 .. . ::J ~K .T specimens: CNP I DCB I CT environment: salt water ambient temperature R ·0.... but at very low and very high stress-intensity ranges. Pampillo.7 Q) ~ typical experimental scatter crack orientation L . Chubut. As shown in the above graph. Source: Markus O... Various Aluminum Alloys: Comparison of Grades for Corrosion-Fatigue Crack Growth Rates in Salt Water 163~------------------.--+---+----+----+----+---. Embury.. 7475 . ro .H 117 10. no significant difference between fatigue and corrosion fatigue crack growth rates is observed.. 10 20 30 40 50 60 cyclic stress intensity range.0-0.T651 7005 -T63 2618 . Eds..T 87 5456 . 1980. "Aluminum as a Corrosion Resistant Material.H 321 . Argentina).. ~ ~ ~ 10. m-~] Corrosion-fatigue crack growth rates in salt water for aluminum alloys exceed the scatterband. [MN. American Society for Metals. Biloni and D. H. Metals Park OH.. A.T 736 7075 . "in Aluminum Transformation Technology and Applications (Proceedings of the International Symposium at Puerto Madryn..1 ..T 651 2219 .392 12-73..21 ..

..... (To give a "feel" for the time required to accumulate this many cycles...' -.. Fundamentals of Machine Component Design.. Juvinall....+ . an automobile would typically travel nearly 400. and Influence of Casting Method on Fatigue Life 393 500 400 300 '" ~ 250 200 201---"': 18t---161----+~ t: e en ~ <. John Wiley & Sons. 1983.1----'---+..:.-:~ 61-----+---+--.+-. Source: Robert C..000 miles before anyone of its cylinders fired 5 X 108 times.......-:...... ~ ~ 150 100 75 50 ~ rf.-t----' 71-----+--.. New York NY.: " :.+------1------+--... Note the absence ofa sharply defined "knee" and true endurance limit..'l 8!---'---'-f--'-----if--"""'""-.-.. It will also be noted in the graph that there is an overlapping of fatigue strength for the sand and permanent mold casting methods (same alloy).) As is true for most metals and alloys.... 1 0 1 ... p 207 .. In the absence of an endurance limit. ~ II) § E 141----+----. the fatigue strength at 108 or 5 X 108 cycles is often used...--:'":-.... This is typical of nonferrous metals...12-74.+---~ 106 Life N (cycles (log)) lOS 109 Representative S-N curves for various aluminum alloys are shown in the above graph.. Various Aluminum Alloys: Wrought vs Cast....:-' 121----'--+---'--~ ~ iU 1... the wrought versions of aluminum alloys have greater fatigue strength than the cast (see graph)...

t 2 10..."'.... I- r-.. .. 8. . 'I\.~ -.195 DEGREE FATIGUE* DEGREE FATIGUE* POROSITY STRENGTH POROSITY erRENGTH . <...POROSITY ."1\ 1'>. ' ....6 10 - 101 Interrelationship offatigue properties with degree of porosity for AL-195 casting alloy.. 22 " .150 rr. ... ...500 6 8.... Source: N. . . [.. P 65 ..~ ....12 . 1\ 2 " 4 5) ~ 1-1'. .000 .I'.~ "I".1... ~ ~ .. A~~~ I 20 ...9. '" I"'-\l:i: I' ".ooo 9.. . 24 .....1"\ x 18 ~ " . ~ . .... .. REPRESENTATION OF NO FAILURE " . I- ..000 1) IV i" 3 4 1 1....3.. .. ~ . Aluminum Casting Alloy AL-195: Interrelationship of Fatigue Properties With Degree of Porosity 26 ... .. 6 f' ~" -. ~ - f' I""' l ./ "'I' . J'... 16. . . "~ r-.. r-- 10 Q .. "Evaluation of Non-ferrous Materials. Syracuse NY..000 9......" in Materials Evaluation in Relation to Component Behavior (Proceedings of the Third Sagamore Ordnance Materials Research Conference). 1956..I .. .~ . ..394 12-75. ~I' 1'1· "o .>.."o "of-. r-...~ \. _~ r-... r>- . .5 10 .. Promisel.....5 .. .. ..000 8 * AT 10 000 000 CYCLES .. Syracuse University Research Institute. E. ~ . I. <.....f ...'\.. I- 10 8 } DEGREE :-:-. ~ .. S-N CURVES CASTING ALLOY AL .... . 41.

M. Avitzur.. "Squeeze Forming of Aluminium-Alloy Components. The results are presented in the above chart." in Production to Near Net Shape: Source Book. which includes for reference the results of similar tests carried out on conventionally cast LM25. A servohydraulically controlled fatigue machine was used to execute push-pull tests about mean zero. which were cut from a bracket component. 1983. Williams and K. Eds. American Society for Metals. It can be seen that a significant improvement in the fatigue performance has been achieved by squeeze forming this type of alloy. Source: G. Aluminum Casting Alloy LM25-T6: Squeeze Formed vs Chill Cast.) S-N curves for aluminum casting alloy LM25-T6. J. chill cast versus squeeze formed. Van Tyne and B. p 367 . Fatigue tests have been carried out with LM25 samples. Fisher. Metals Park OH. C.12-76. Effect on Reversals to Failure N 395 I 400 ii 200 « (/) l- => ~300 w o 2 z E a:: ~ 100 LM 25-T6 chill cast In O'-:--_--L 10 2 1~ --'4 '-::-_ _--'- --'---' m 1~ 1~ REVERSALS TO FAILURE(2N.

2 -- '~ <. I n A i r " _ II I 'ea" ---- ~ ~ - 4.396 13-1. American Society for Metals.... J. ~ OJ (J) ~~--I" " .. ~ Q(~I'(" ~-. 3 Oxygen must be a gas (preoxidation or intermittent exposure is not effective).. Based on these experiments.. iA 1110 o ::f 5. ~" I Vi Q. --- l' ~ \ .\~I \ - _______ T-D If) 0 . because the majority of life is concerned with crack propagation (environment has little or no effect on nucleation and initial growth). c . Scale] .. - I--- ..5.. P 336 .6 s: u <. 1979. Cycles to Fracture (Log.+ Indicates Specimen Unbroken The effect of air and water vapor on the fatigue life of annealed copper.. Source: D.6 I' r-.4 ~ -... Duquette.. the investigators concluded that: Fatigue cracks form early. Copper: Effect of Air and Water Vapor on Cycles to Failure 5... Metals Park OH. Alternate static exposure to air and dynamic exposure to vacuum do not affect fatigue life. and SoN curves diverge as applied stresses are reduced (see graph). ~ I '" c I-- ~5.---- "E 4.. The effect of atmospheric oxygen on fatigue life of copper has been investigated..0 e Vi '0 4 . 2 Oxygen and water vapor are the primary damaging constituents in air (water vapor alone is effective).8 a: Ol C <0 OJ \. PlJr i f · iltr)fJ ~fiedAir .. oxygen and water vapor reduce fatigue life in copper.10 7 Endurance. "Environmental Effects I: General Fatigue Resistance and Crack Nucleation in Metals and Alloys." in Fatigue and Microstructure.1--- .4 6 10 10 7 ~ 5..

full application of the load in the first cycle causes a large strain in a soft material. although it was not recognized as such at the time). Helgeland. "Mechanisms and Theories of Fatigue. Copper: Applied Plastic-Strain Amplitude vs Fatigue Life 397 Cu • prestrain 20"1. imposed a low stress at the start of his tests and increased it gradually to the chosen value. Helgeland was the first to observe and claim a fatigue limit for copper (actually the plateau stress. his results were apparently contradicted by those of Kettunen.13-2. the stress is low in the initial cycles and increases to saturation. _- A- Applied plastic-strain amplitude versus fatiguelife curves for copper at long life. who observed failures at stresses down to 17." in Fatigue and Microstructure. whereas in stress cycling. on the other hand. in which they monitored the plastic strain.7 MPa. However. where failures occurred at strains as low as 10-5 . Metals Park OH. Lukas et al. failures were observed at stresses below that of the plateau. who showed that Lukas and Klesnil's long-life CoffinManson plots showed failures to occur only down to the plasticstrain fatigue limit. The difference between these tests is that in strain cycling. Unfortunately. although he was stress cycling. This initial large strain creates the PSB cell structures. also carried out stress-cycling tests. Specimens that had been stress-cycled yielded a plot of saturation plastic-strain amplitude versus life. Source: Campbell Laird. which would not otherwise form in a constant-strain test. at lower strains. This difficulty was resolved by Laird. no failures were observed in the testing time available (see above chart). American Society for Metals. 1979. Since Kettunen applied the full load to his specimens. • annealed "prestrain 40/. P 195 .

26 ::E e Cl.398 13-3....... ::J .. '" 10 Stress cycles Rotating-beam fatigue strength of electrolytic tough pitch copper... 9th Edition. Copper Alloy C11 000 (ETP Wire): Effect of Temperature on Fatigue Strength 200 r . Properties and Selection: Nonferrous Alloys and Pure Metals. American Society for Metals. CllOOO wire..... .... 1979." .. en 1. s:.. Source: Metals Handbook... '" e 150 22 ]1 s:. 18 100 ~ Q) e '+> u. en '+> en ::J '" 14 u.. Metals Park OH..." 'C. ..) diam. Volume 2. P 289 ....... H80 temper when tested at various temperatures.... 2 mm (0.08 in.

Q) ...N Legend o Group G (60% Cold I:>. 6 Influence of grain size and cold work on fatigue strength of copper alloy C26000 (cartridge brass).026 mm.. P 69 ... I' <... Group D (40% Cold o Group A (20% Cold • Anneal I Grain Size . 40 . Source: George M. x + - -+ 10 10 107 Cycles for Failure .. ~~~ .131 mm.. II!. JohnJ.. 0.. I- t. 0..vin Fatigue-An Interdisciplinary Approach...... therefore.1-- :: .. Reed and Volker Weiss.. Norman L. <. Syracuse NY.13-4. Burke. All specimens were prepared from the same heat and..x _ -i IJ. I'--~ . Copper Alloy C26000 (Cartridge Brass): Influence of Grain Size and Cold Work on Cycles to Failure 60 399 50 a.051 mm... Changes in grain size and in degree of cold work which result in increased tensile strength or hardness usually result in improved fatigue strength. 30 0 . Sinclair.)" +~ l J) ~ >< 20 ~?" I"'--. The above S-N curves illustrate this and indicate the influence of grain size and cold work on the fatigue strength of alpha brass. r-r...... ... 1'0. Anneal 3 Grain Size x Anneal 4 Grain Size + Anneal 5 Grain Size Drown) Drawn) Drawn) 0. had the same nominal composition.. "Some Metallurgical Aspects of Fatigue....... 0. Syracuse University Press.0/2 mm.. 0 0 0 '" "0 <. ::l Q) <fl<fl r-. 1964..e f_ ~ a 1'.:!""... Eds.

Metals Park OH.~ '" u.' :::E ~ ~ Ql t» 1201--------+------' 1.. Scatter Band 180 ~------. P 406 .. Ql s :::l 1.---------r--------''-------~_-----~ 25 160 f---~~~-+-------+------+-------+--------l rf 140 f------~ 20 ] .. Source: Metals Handbook.+ .. Volume 2. American Society for Metals.J::.400 13-5.. 1979. Copper Alloy C83600 (Leaded Red Brass): S-N Curves. 80 f-------+--------+-------I-----= 0- 10 Stress cycles S-N curves (scatter band) for copper alloy C83600 (leaded red brass)....~ '" 100 f .. ------+-------1 15 u.... 9th Edition.. Properties and Selection: Nonferrous Alloys and Pure Metals. :::l ...+ - .

. '" .10 Number of stress cycles S-N curves (scatter band) for copper alloy C86500 (manganese bronze).::. All testing was performed at room temperature.. 200 Cll '" u........ ArnericanSociety for Metals.... p 35 .13-6. Metals Park OH. Properties and Selection: Nonferrous Alloys and Pure Metals. Scatter Band 400 401 If 300 :E <ttIOo.... 9th Edition. 30 t: . 1979. Volume 2.r.. u.40 ~~ n ~ ." ".~ 100 -- . 0 . Copper Alloy C86500 (Manganese Bronze): S-N Curves.20 ..:) e Cll c c: . Source: Metals Handbook.50 ~ c t: . ..:) e ~~ .

Ql E . . P 416 . 9th Edition. ~ ~ 1. 300 40 250 200 150 100 4 10 30 '. . 20 105 106 Stress cycles S-N curves (scatter band) for copper alloys C87500 and C87800 (silicon brasses) tested at room temperature. Properties and Selection: Nonferrous Alloys and Pure Metals..402 13-7. American Society for Metals. 1979.... Ii e Cl c: '. Scatter Band D- .. Copper Alloys C87500 and C87800 (Silicon Brasses): S-N Curves.. '. Volume 2... Metals Park OH.. Ql E .. Cl <{ . ~ ~ . ~ 107 108 109 Source: Metals Handbook....

.. .. Cl 105 106 107 108 '" 10 Ci: 109 . Scatter Band 403 0- ::i: '" 200 25 20 15 'iii . American Society for Metals. '" '" 100 50 104 s c ~ . Metals Park OH....>t s Ci: ~ Ii 150 Cl E . Copper Alloy C92200 (Navy "M" Bronze): S-N Curves.. Properties and Selection: Nonferrous Alloys and Pure Metals. 1979.13-8. . Volume 2. '" Number of stress cycles S-N curves (scatter band) for copper alloy C92200 (Navy "M" Bronze. Source: Metals Handbook. p 421 .. 9th Edition. or steam bronze) tested at room temperature.

1979.404 13-9. 9th Edition.g <i: e E l!l 't!' 20 105 Number of stress cycles S-N curves (scatter band) for copper alloy C93700 (highleaded tin bronze) tested at room temperature. . Source: Metals Handbook•. Cl 160 140 120 100 80 104 ~_----=l15 . Properties and Selection: Nonferrous Alloys and Pure Metals. Copper Alloy C93700 (High-Leaded Tin Bronze): S-N Curves. P 426 . Volume 2. Metals Park OH. Scatter Band 180 25 0~ . American Society for Metals.

Tube Oia.." in Source Book on Copper and Copper Alloys. Source: Donald K. 1/4-io.. After 90-days exposure. 1979. the fatigue performance level was not lowered (see plot above).. The tubes made from the copper alloy failed in the range of 105 to 106 cycles. Brazed steel tubes.. Copper Alloy No. 0 b 0 0 a - 0 .! U u c-. prior to salt exposure. I----+I---J----o I -r---- 0 i 0 0 0 0 0 0 0 -~ I :0 10 ~ 10 'Q. American Society for Metals. 192: Effect of Salt Spray on Tubes liAs Received" Cycles to Failure 405 Cycles to Failure After 180-0ays Exposure to Salt Sorcv 3/8-io. After exposure for 180 days to salt spray. Tube Die. Metals Park OH. 3/8-io. ~ M 0 ~ N Results of fatigue tests on copper alloy tubes before and after salt spray exposure. 3/16-io.13-10. Miner. After 30-days exposure to salt spray. 5/16-io. . 1/4-io. the steel tubes showed no fatigue strength in this particular test. 5/16-io.. 3/16-io. "An Effective Solution to the Problem of Hydraulic Brake Line Corrosion. p 356 . the resistance to fatigue was 105 to 106 cycles. failed in the same test in the range of 105 to 107 cycles.

...." in Source Book on Copper and Copper Alloys....5 UTS: 152025 ksi 120 UTS: 101.. Cieslewicz. Source: J. Two variable-speed.406 13-11..<....5times the ultimate are plotted in the graph above. Copper Alloy 955: Goodman-Type Diagram 150 1----+----+----+-----t----2F------i 1. flat-plate testing machines of the fixeddeflection type were used for the test work. 1979.' o 30 120 150 180 Goodman-type diagram (after Creech) for annealed copper alloy 955.. The yield. These machines have a speed range of 750-2000 cpm with a maximum deflection of I in.h~---fI+_-I+_---1t+_-H__-I::l=-_H-+++-_I_If--I . --'-_ _.. "A Modified Goodman Diagram to Predict the Fatigue Limits of Copper Alloy 955.5 ksi ~ °ili °ili 0 90 . ultimate tensile strength and 1. p 40 ...3 °i o ~ G 30 '--_ _----' ---'60 -'90 -L. American Society for Metals.0 ksi. iii "" ~ 60 30 I-:JIC---I . The fatigue limit at zero mean stress was determined and was found to be 22.. M. Metals Park OH.

1979. ::J 'x :2 . E ~25~ °c (482 of) ::J E 'x ~ - /20 oC (68°FI.. 'x ~ E i! 200°C 1392°F)_ 12 :.... E ::J <. Machine speed was 2960 Hz.16 ~ I I . Cycles of stress S-N curves for magnesium alloy sand castings.20 <. 20 <.... QE22AT6.... :2 +1 125 100 75 50 25 <. Metals Park OH.14-1.. ~. E ::J E "".. American Society for Metals.. /Unnotched U·nolched /(K t ~ 2) V·notched /IKt = 31 1 1 ~ +1 t: ~ ~ ~. Rotating-beam (Wohler) tests.. 9th Edition. showing effects of type of notches (upper graph) and testing temperature (lower graph).. 'x ~ E E Cycles of stress 150 If..-. :2 +1 -... P 589 . \. .... Source: Metals Handbook.. Volume 2. Magnesium Casting Alloy QE22A-T6: Effects of Notches and Testing Temperature 150 125 100 75 50 25 407 If. Properties and Selection: Nonferrous Alloys and Pure Metals..

. 10 :::. ~ <. Volume 2.. :::.d I 25 \ 'I\. '-. Cycles of stress S-N curves for magnesium alloy castings. 125 100 75 50 25 -. 9th Edition... -106 g . Properties and Selection: Nonferrous Alloys and Pure Metals. .. QH21A-T6.. Rotating-beam (Wohler) tests... U·nolched K t=2110 E ::l E 'K :::.. r-. 1. . ] 15 . :::.. 20 rr.... American Society for Metals.. Metals Park OH. showing effects of notches (upper graph) and testing temperature (lower graph). 1979.. Effects of Notches and Testing Temperature 175 150 25 \ "- 20 °c (68 IF) rr. E ::l E I'-.408 14-2. +~ 100 ~ ~ ~ E 75 50 25 -.Unnolched 20 .. machine speed 2960 Hz. "'--- 20°C (68 OF) 15 ] 01 g E ::l E 'K ~50 °c 1480 OF) I"--. P 590 .. ..... Source: Metals Handbook.. 'K :::... Magnesium Casting Alloy QH21A-T6: S-N Curves. Cycles of stress 175 150 125 Unnolch...

.. Small radii.------.MPa 100 M&an stress..=j±4 f---t---+---+----'=-+--+--.....:n--r--T--..---..ksi 20 ±20 ±20 .MPa Effect of type of surface on fatigue properties of cast Mg-AI-Zn alloys.---r-.~ --+--1---.§ ~ '" . ksi 409 Meanstress. Source: Metals Handbook..16 . notches or fretting corrosion are more likely to reduce fatigue life than are minor variations in composition or heat treatment.-=J '4 1-----J-=.. 9th Edition."""-+-----1F"'-. ±4 125 Mean stress.. '4 125 Mean stress. Mg-AI-Zn Casting Alloys: Effects of Surface Conditions on Fatigue Properties Meanstress...150.-----.. loaded in bending . Properties and Selection: Nonferrous Alloys and Pure Metals. Metals Park OH... 1979.12 tl2 ---f----f----. ksi 10 15 20 ..---y-.ksl 125 Mean stress.-----. as shown in the above curves.16 "6 Xi ~- ." Ii -l---I----l !II .MPa Meanstress..MPa Mean stress. O ±20 Cast plates '1251---+---+-. ~ !i ~ 4-----11-----l ell t---+--+-----j---t---+--.--..----i---. Volume 2. P 532 .:.---'T.. ±20 o 20 10 '150 r ---.... American Society for Metals. !II . Machining improves fatigue properties of castings.14-3.16 ~ :t12 e ~ .---..f---+--.107 cycles.. I---t---+----f---""'---+-.

Molybdenum: Fatigue Limit Ratio vs Temperature 0.5 ~r\ \ o 100 200 300 Temperature.7 '. American Society for Metals. 0.8 0. Metals Park OH. 9th Edition. P 774 .4 400 500 600 Source: Metals Handbook. 0 ~ a:: '" ! 0. Properties and Selection: Nonferrous Alloys and Pure Metals.410 15-1.OC Ratio of the fatigue limit of molybdenum at various temperatures to its tensile strength at the same temperature.9 0. 1979.:.6 0. Volume 2.

f4.... ~ -. 1983..r. ~ ~ .--. P 1095 ................ . Metals Park OH.. Because solder alloys are strain-rate sensitive and have large elongation capabilities.. a... 9th Edition..... ~~ 1000 ... . I'~ 4000 ~ ~~ A .. I~ . Tin-Lead Soldering Alloy: S-N Data for Soldered Joints 5000 411 I ..04 0.... Welding. Volume 6. 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.. and conversely........ ~/min I I 2 .........16-1.. .. Source: Metals Handbook.. ':).r.. ... 'Speed of testing I .. <.. Brazing.. in 2000 "'r-. American Society for Metals.... tests under constant strain do not reflect a practical application situation.. '" ~ ui --..... :--.2 - a 1 2 2 4 6 8 102 Cycles to failure 2 4 6 8 10 3 2 The fatigue strength of soldered joints is a complex and difficult subject to examine. 'iii 3000 ~ :--... 'N ..... ........ the performance of fatigue tests under constant stress causes progressive and rapid relaxation of the joint... and Soldering.....

Use is made of this principle (see graph above).50 0. mm Variation of bearing life with babbitt thickness for lead or tin babbitt bearings...005 in. Properties and Selection: Stainless Steels. Source: Metals Handbook.040 ~ • Trimetal . 9th Edition. 0.030 o Bimetal I 0.) thick. American Society [or Metals.75 1. usually no more than 0. . ..13 mm (0. Volume 3.00 Babbitt thickness. Babbitt: Variation of Bearing Life With Babbitt Thickness Babbitt thickness.412 16-2. 100 0. in. '. Tool Materials and Special-Purpose Metals. Metals Park OH.J :::l OJ 200 Cl . One of the most useful concepts in bearing-material design came with the recognition that the effective load capacities and fatigue strengths of lead and tin alloys were sharply increased when these alloys were used as thin layers intimately bonded to strong bearing backs of bronze or steel. 1980.. in which the surface layer is composed of a lead or tin alloy. Bearing load was 14 MPa (2000 psi) for all tests.. in both two-layer and three-layer constructions.25 0. P 806 .~ OJ a: Q.

05 to 0.. Properties and Selection: Stainless Steels. 200 300 en . operating temperature markedly influences fatigue life. ~ 200 I------'\-t---+------t ...l 200 100 150 50 Bearing temperature.. °c Varlanon of bearing life with temperature for SAE12 bimetal bearings.005 in.. SAE 12 Bearing Alloy: Effect of Temperature on Fatigue Life Bearing temperature. P 813 .13 mm (0. Bearing load: 14 MPa (2000 psi).----~-----. on steel backing. The alloy lining was 0.. Volume 3.. Source: Metals Handbook.) thick. Tool Materials and Special-Purpose Metals.002 to 0.. ° F 413 300 r---.c-r. As indicated. American Society for Metals. '" Gi II: '" 100 I----~r_--t-----f OL-_ _L--_ _L .16-3. Metals Park OH. 1980.:? > . 9th Edition._ .

Grade 3: S-N Curves for Annealed vs Cold Rolled 500 400 6. 9th Edition.>< 40 ~ ::i J 30 E '" E "' 20 :2 'x - 10 Number of stress cycles SON curve for unalloyed grade 3 titanium. Properties and Selection: Stainless Steels... a.70 r-. Data were obtained by rotating-beam testing at room temperature. Metals Park OH...~~ 0 .-.§ 200 )( .. . 50 .. . Tool Materials and Special-Purpose Metals.. Source: Metals Handbook. ~ .414 17-1. P 376 . Volume 3. Unalloyed Titanium.60 Cold rolled Annealed J '" E "' :2 100 - . American Society for Metals. :2 "' '" '" 300 ~ . 1980.

60 . Metals Park OH.S x 30 '" '" 100 104 20 S-N curves for unalloyed titanium. 9th Edition. Unalloyed Titanium. and elevated temperatures. Data were obtained by rotating-beam testing of unnotched. at subzero. Grade 4: S-N Curves for Three Testing Ternperatu res 600 500 Q. polished specimens machined from annealed bar stock. .>< . room.. American Society for Metals. Tool Materials and Special-Purpose Metals. Volume 3.0378 ... Source: Metals Handbook. Properties and Selection: Stainless Steels.40 .S x ~ ------Number of stress cycles . grade 4.17-2. 415 ~ '" '" 400 '" 300 200 ~ ::> E .50 ~ ::> ~ :i E I 315°C (600 of) ..80 _40°C (-40 ° F) 20°C (6SoF) - 70 . 1980.

.Q.. American Society for Metals. 400 ::lE lI) lI) ILl It: l- ll) 200 10 3 CYCLES TO FAILURE Stress-life curves of two Ti-V alloys that undergo cyclic hardening (Ti. Source: Edgar A. 1979. p 236 . A. CI STRESS CONTROLLED ::lE ILl 0 600 I- ::> :J a. as indicated above..24%V) and cyclic softening (Ti-32%V). Ti-24V and Ti-32V: Stress Amplitude vs Cycles to Failure a. "Cyclic Plastic Deformation and Microstructure." in Fatigue and Microstructure.416 17-3. Jr. and Gerd Lutjering. <l: Ti-24%V. Cyclic-response curves indicate that the Ti-24%V alloy undergoes extensive cyclic hardening. whereas Ti-32%V undergoes cyclic softening. Starke. Hardening is caused by incomplete reversibility of twinning. Metals Park OH.

. stress cycles SON curves for annealed titanium alloy Ti-5AI-2. stress cycles o a::2: ... 600 notch (K r = 2. .:. .17-4.>I.( ::2: .. p 382 . Tool Materials and Special-Purpose Melals. .. .. e t: :I . 800 100 ] ~are t: 400 :I e . Properties and Selection: Stainless Steels. 1980.4) ~ .80 a E ..§ ::2: . 800 ~l> ~o ~ Shot peened 100 . Bottom graph shows fatigue strength as affected by type of notch. 20 ::2: 0 8 10 '. Source: Metals Handbook. 0 104 5 10 106 Lifetime.5Sn (rotating-beam tests). t: 400 E :I e rt:Ground ~ 80 Hand finished - 60 t: 40 '. Metals Park OH.2) ~l>l> 'x . Top and center graphs show fatigue strength for different types of surface finish..20 ::2: 8 10 - 40 :I E E o Lifetime... Ti-5AI-2.......... )( 200 ~~ l>Al>A 'I 16 ~ 60 A V-notch (K r = 3.§ 200 )( ::2: E E ...>I.( ~ e . 20 ~ 8 10 Lifetime. Volume 3.. ..( E :I E a::2: ..5Sn: Effects of Notches and Types of Surface Finish 417 a::2: 600 ~ . . l:/" 800 600 400 Ultrasonic machined Slab milled Chem milled and annealed Chem mille 100 80 60 40 ... 9th Edition.. American Society for Metals. ... stress cycles 107 . ~ E E 200 '...

1980. 9th Edition.SSn and Ti-6AI-4V: Fatigue Crack Growth Rates Stress intensity factor range.j~t. over part of the 11K range. s«.---f c:l 10-31---.Cl '" u. However. the fatigue crack growth rates for Ti-6Al4V(ELI) are higher at cryogenic temperatures than at room temperature at the same 11Kvalues. Data on fatigue crack growth rates for Ti-5AI-2. MPa . ksi Viii. :J Stress intensity factor range..f--+_--__l 0C(75 0F) ~ {24 E ELI -196°C (-32°F) E -269°C (-452 of) ~' Ti-6AI-4V ~ i ~ ~ ~ 24 to -269°C (75 to -452 of) b 10-41-----+--~~--I~----4-----=1 Cl e u OJ NI ELI .5Sn and Ti6AI-4V alloys are plotted above.418 17-5...5Sn and Ti-6AI-4V. Properties and Selection: Stainless Steels.. Ti-SAI-2.m Fatigue crack growth rates for Ti-5AI-2.452 of) ----=".. Volume 3. Tool Materials and Special-Purpose Metals.. American Society for Metals. Metals Park OH. These data indicate that the exposure temperature has no effect on the fatigue crack growth rates for Ti-5AI-2. 5 10 10-4 NI24 to -269 0C . P 765 .5Sn and Ti-6AI-4V(NI). . Source: Metals Handbook.(75 to .. /iK.

" I I 85 ~5 48 120 I I I o 20 40 60 80 ENDURANCE LIMIT. 42 Rcl Surface R aug h ne s s . )83 150 43 55 HAND GRIND Gentle Abusive J67 67 14 14 END MILL· PERIPHERAL CUT Gentle Abusive 73 I 45 28 39 ECM I". American Society for Metals.17-6. .n"". Donachie.. Jr. AA 419 SURFACE GRIND Gentle Cony. Off. Ti-6AI-6V-2Sn: Effects of Machining and Grinding Ti-bAI-bV-lSn (STA.Sta nda rd 147 12 II 145 ECM + PEEN '". Ed.v. KSI Bar chart presentation showing effects of various machining and grinding operations on fatigue characteristics of titanium alloy Ti-6AI-6V-2Sn.. Matthew J." in Titanium and Titanium Alloys: Source Book. P 355 . iOff-S'.. Source: Norman Zlatin and Michael Field. Metals Park OH. 1982. "Procedures and Precautions in Machining Titanium Alloys.. Abusive I b5 I 20 lO 43 44 70 SURFACE GRIND + PEEN Gentle Ah .

Ti-6AI-6V-2Sn (HIP): S-N Curves for Titanium Alloy Powder Consolidated by HIP iii 120 " vi en w l- ANNEALED PLATE (MINI a: ll) 10(' :2 :> :2 X <{ :2 80 60 -'-- ---..420 17-7. This shows that a clean powder is required for parts that are fatigue-critical and must operate with the equivalent of fully forged properties. T. AS MACHINED (SPEC W "M" SU8SCRIPT WAS EVALUATED 8Y METALLOGRAPHY) o HIP RUN 2. Metals Park OH. "T" "T'" --. Highberger. low fatigue endurance was apparently associated with failure initiation at an inclusion. VAC ANN AT 1300"FI24 HR iii S-N curves showing endurance limits for titanium alloy powder consolidated by HIP at 900°C (1650 OF). "Experience With Net-Shape Processes for Titanium Alloys. Avitzur. Witt and W. Source: R. VAC ANN AT 1300·FI16 HR OJ HIP RUN 1. J. C. In the specimen designated with an "M"subscript. 1982. H. P 277 . VAC ANN AT 1300·FI2 HR rn HIP RUN 2. American Society for Metals. VAC ANN AT 1300·FI24 HR HIP RUN 4. Van Tyne and B. Note that most data points obtained in this phase fell within the representative data band for annealed forgings. Eds." in Production to Near Net Shape: Source Book. 10 NUM8ER OF CYCLES 5 mHIP RUN 2.

17-8. RAfIIG[ O~ PROGRAM l?ATA '7------: ~~ ~7 40 20 ~4 ~~ CVClP8 TO {"AILURf" S-N curves showing that HIP Ti-6AI-6V-2Sn is equivalent to annealed plate of the same composition. American Society for Metals. C." in Production to Near Net Shape: Source Book. 1983. Theodore Highberger. Van Tyne and B. / eeo F-16KeI-2 /lR \ / 22-5'0~-fOl<'Sl-I/-IR ANN PLATf' tOO \ tgTReeS. Ti-6AI-6V-2Sn (HIP): S-N Curves for Annealed Plate vs HIP 421 120 . Metals Park OH. J. Eds. Avitzur. Source: W. P 304 .. "Manufacture of Titanium Components by Hot Isostatic Pressing. KSI 80 €JO \ <.

.. Ed....422 17-9..:n~tl~e_... KSI Bar chart presentation showing the effects of specific machining and grinding operations on fatigue characteristics of titanium alloy Ti-6AI-2Sn-4Zr-2Mo. 68 I Cony I 17 ~IO 39 41 lZO END MILL- I-l'G~en~ltI. Source: Norman Zlatin and Michael Field.. Jr.:i-e "T"" '" 8Z PERIPHERAL CUT .... Matthew J. American Society for Metals.:·v~e .". AA SU RFACE GRIND I-l'G~e. Ti-6AI-2Sn-4Zr-2Mo: Bar Chart Presentation on Effects of Machining and Grinding Ti-bAl-lSn-4/'... Donachie. "Procedures and Precautions in Machining Titanium Alloys. 47 1 36 77 o I ZO I I 60 I 80 I 40 ENDURANCE LIMIT.. -. A"'lb~u~.:.r-lMo (STA. P 355 . Metals Park OH. 1982.:.."in Titanium and Titanium Alloys: Source Book. 36 Re! Surface Roughness.

9th Edition. Ti-6AI-2Sn-4Zr-2Mo: Constant-Life Fatigue Diagram Minimumstress. ksi -100 1200 -50 0 +50 100 150 423 1000 150 . 10D ~ 600 If. Volume 3. MPa Constant-life fatigue diagram for duplex annealed Ti-6AI-2Sn-4Zr-2Mo sheet. Metals Park OH.04 ln.) thick.. 1 mm (0. .17-10. Source: Metals Handbook.. P 385 .>< 800 ~ t: 'x E :> E :. Properties and Selection: Stainless Steels.. American Society for Metals.. 1980. Tool Materials and Special-Purpose Metals. t: :> ~ 400 50 10 7 cvcles lifetime 200 ~ 'x E E 0 -800 -600 --400 -200 +200 400 600 800 1000 0 1200 Minimumstress.

All fatigue tests conducted at a stress ratio of R = 0. Source: Metals Handbook. then 8h at 540 ° C (1000 ° F) and air cool. STA (solution treated and aged) condition: 1 hat 870 ° C (1600 OF). DA (duplex annealed) condition: 15 min at 870°C. Volume 3. '" ~ '" :2 'x '" E ::J E ~ :2 1000 800 1 'x E ::J E 140 '" 103 Number of cycles 120 105 Low-cycle axial fatigue curves for Ti-6AI-2Sn-4Zr-6Mo. :2 '" ~. 9th Edition. tension tests. 1400 1200 --- 220 STA 200 180 160 . air cool. Open symbols indicate fatigue tests.424 17-11. P 395 .1. Properties and Selection: Stainless Steels. solid symbols. Metals Park OH. age 8 h at 595°C (1100 OF) and air cool. water quench. Ti-6AI-2Sn-4Zr-6Mo: Low-Cycle Axial Fatigue Curves 1600 Q.>< ~. American Society for Metals. 1980. Tool Materials and Special-Purpose Metals.

Tool Materials and Special-Purpose Metals. Solution Treated and Aged Condition 600 500 e.. American Society for Metals. Metals Park OH.5 'in I - '" '" ~ '" 300 '" :::E 'x E :> E ~ • 60 ~ ~ '" '" 200 100 I'---- . air cooled. P 403 . Data are for 1.25.) thick sheet solution treated 10 min at 790 0 C (14S0 0 F). 9th Edition. R = 0.20 '" :::E 'x E :> E Stress cycles SoN curve for Ti-8Mo-2Fe-3AI titanium alloy in the solution treated and aged condition.17-12.40 . 425 :::E '" 400 '. 1980.S-mm (0. Ti-8Mo-2Fe-3AI: S-N Curves. Properties and Selection: Stainless Steels. ... Source: Metals Handbook.~ . 80 Tension-tension tests . and aged 8 h at 480 0 C (900 OF). Volume 3.060-in.. K r = 3.

~ Unnotched I :!: ~ '" 900 800 700 E :::> e . .. '" :!: 'x E :::> E ---- 80 I-- Stress cycles 500 Q. Top: results of unnotched bars tested at room temperature.. P 399 . Ti-10V-2Fe-3AI: S-N Curves. ". Notched vs Unnotched Specimens in Axial Fatigue 1100 1000 ~o Q. 9th Edition. ~ to-.:. Metals Park OH.1 and a frequency of 20 Hz.§ )( '" .---..t. Bottom: fatigue characteristics of notched specimens tested at elevated temperature.t. Tests were conducted at a stress ratio ofR= 0. Source: Metals Handbook. ~ 40 ~ ~ '" '" 0 20 'x E :::> E '" :!: I Stress cycles Axial fatigue of Ti-l0V-2Fe-3AI bar stock in the STOA (solution treated and overaged) condition. Tool Materials and Special-Purpose Metals. ~ lili r--A 'u.0 60 'u.Volume 3. Specimens were taken from round bars 7Smm (3 in.426 17-13. furnace cooled. ~ '" 300 '" o RT ~2050C (400°F) '" :!: 100 o 425 °c (800 of) 'x E 200 :::> E r-. . American Society for Metals.-- ~ ~ 140 .:. :!: '" 400 \~ \0 Notched I Kr = 3. overaged 8 h at 565°C (1050 OF) and air cooled.) in diameter that had been solution treated 1 h at 760° C (1400 OF). 1980.' 120 100 '" ~ '" '" :!: 600 500 <. Properties and Selection: Stainless Steels.

.J O'---4.6 20 J."7T""-------."in Titanium and Titanium Alloys: Source Book..08. ksi vm. American Society for Metals..08...17-14.L..) Comparison of fatigue crack growth rates...R= 0..... Source: Wayne A.F= 1 to 30 Hz. Rosenberg. as shown above.L--... Matthew Donachie. for RA Ti-6AI-4V. P 375 . Reinsch and Harry W.J.. Ti-10V-2Fe-3AI and Ti-6AI-4V: Comparison of Fatigue Crack Growth Rates do/d~ in/cycle (pm/cycle) 427 10-4-r-----r---. .R = 0. Jr. Ti-IOV-2Fe-3Al approaches the performance of Ti-6AI-4V in the recrystallized annealed (RA) condition. . (22) (33) (44) (MPoW.' O (II) LlK. for MA Ti-6AI4V. (25) MA Ti'-6A/-4V 10-5 (025) • i i i i I Ti"-IOV2Fe-3A/ (0025)/i'O--.. Fatigue crack growth rates in air have been found to lie in the scatter band for mill annealed (MA) Ti-6AI-4V. "Three Recent Developments in Titanium Alloys.... Metals Park OH.. 1982. Data for Ti10V-2Fe-3AI. F= 6Hz. 70.. At high ~Kvalues..R= 0.F= 1 to 25 Hz.05. Ed.

Notched fatigue results are shown above.l-O-4----------'-----"':::OO"------. Fatigue characteristics of Ti-lOV-2Fe-3Al are equal to or superior to those of Ti-6AI-4V..OS. Metals Park OH.J 30 Cycles tofaIlure Comparison of notched fatigue lives for Ti-lOV-2Fe-3AI forgings and Ti-6AI-4V plate. American Society for Metals. Rosenberg.3 ksi~.428 17-15.1 ksi ~.5 MPa Jill). Notched Bar Fatigue Life for a Series of Forgings Compared With Ti-6AI-4V Plate rrmox.l. Reinsch and Harry W.IOJpsi(MPo) 70-r-----~----------____. Source: Wayne A. Jr..5% NaCl is typically about 90% of the x.R = O.9. with a standard deviation of 2.R = O. Data from a series of die forgings have shown that the mean value for fracture toughness is 49. "Three Recent Developments in Titanium Alloys. Matthew Donachie.. (54 MPa Jill). 1982. for STA Ti-6AI-4V plate. (48J) 60 (414) 50 (345) 40 (216) STA Ti·6AI-4V plate (201)/i.. P 375 . Ti-1 OV-2Fe-3AI: S-N Curve. Ed.F =KT= 2. (2. K/scc in 3.K T= 3. Data for Ti-lOV-2Fe-3AI."in Titanium and Titanium Alloys: Source Book.

5 ksi).17-16. MPa 429 40 60 80 100 120 140 160 180 200 1400 r-_.. 400 Al315 60 -c (600 OF) 40 20 200 . P 401 .._-r____.0 800 1000 1200 20 -600 -400 -200 +200 400 600 Minimum stress.-~""__. Metals Park OH... Tool Materials and Special-Purpose Metals. E ~ "K ~ 60 40 200 . Source: Metals Handbook. MPa Constant-life fatigue diagrams for Ti-13V-llCr-3AI.___. at 315°C (600 OF)..7 ksi). Data arefor axial fatigue of edge-polished sheet specimens of material solution treated and aged to room-temperature tensile strength of 1203 MPa (174.Unnotched - - Edgenotched. 1980. Ti-13V-11 Cr-3AI: Constant-Life Fatigue Diagrams Minimum stress. STA (solution treated and aged) condition. Tests were conducted at a speed of 60 Hz.-_r-_r""__."'_... Kr = 3.3 ksi) and the yield strength was 876 MPa (127.0 -400 -200 +200 400 600 800 1000 1200 Minimum stress. 9th Edition._----.. longitudinal orientation.. the tensile strength was 1078 MPa (156.. Kr=3.._-_. American Society for Metals.Edgenotched..Unnotched . Volume 3.ksi Minimum stress.--t-er____. Properties and Selection: Stainless Steels.-__.0 ksi)..---...ksl 1400 -100 -80 -60 -40 -20 +20 40 60 80 100 120 140 160 180 160 140 ~ :. 120 "" 100 E o~ ~ E ~ ~~ E " ~ "K 600 80 :. Corresponding yield strength was 1080 MPa (156. 200 180 160 140 120 -100 -80 -60 -40 -20 +20 ~ ~~ " "~ 600 ~ At room 400 temperature E 100 80 :.~-_. • 800 0.

1980. Volume 3. STA stock -r:"---~ Smooth bar. Ti-6AI-4V: Effect of Condition and Notches on Fatigue Characteristics 800 e.a. STA stock 100 '..a.430 17-17.. American Society for Metals.: bar...5). annealedrock Notched bar (K t = 3. ~ 80 60 g on "" Notched bar. Properties and Selection: Stainless Steels.. p 389 .. and effect of notches. annealed stock 40 106 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. Tool Materials and Special-Purpose Metals. Metals Park OH. 9th Edition. co . Source: Metals Handbook. ~ Smooth ::E 600 ~ 400 on E :::l E 'j( 200 ::E co ::i - ~ -- --..

England. Metals Park OH.:. 106 Endurance (cycles) Rotating-cantilever fatigue (S-N) curves for three testing directions in 57 mm thick. ~ . 1981. and American Society for Metals. these ratios being higher in the longitudinal and short transverse directions because stressing occurs parallel to the basal planes.17-18. which means that the levels of strain will be reduced and the fatigue strength enhanced in these two directions. Edward Arnold Ltd. forged and annealed Ti-6AI-4V bar. Light Alloys. 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. Polmear. These curves show that fatigue properties are lowest in the long transverse direction. CIl Q) CIl CIl 0) .~ c: 550 c: '" 2:! 500 ~ 450 L. p 193 .. London. This result has been attributed to the fact that Poisson's ratios are also sensitive to crystal orientation. J. Ti-6AI-4V: Effect of Direction on Endurance 431 700 650 m a. Source: I. Higher ratios imply greater constraint.

. RoO............ _'--.L _ _.. "Advanced Titanium Metallic Materials and Processes for Application to Naval Aircraft Structures._ ...' 2. 1983... T. :::l 40 AXIAL FATIGUE.......... Scarich.- 80 w '" a: .--l----JL... Chanani and Gregory V." in Production to Near Net Shape: Source Book.L _ _-'-_ _. P 124 ... HOLE ISR TEE EXTRUSION 20 6.L-J 103 105 CYCLES TO FAILURE S-Nfatigue data for isothermally rolled tees versus extruded material..L. Metals Park OH._ _........ Source: W.. o OL-_ _'--_---JL.... Avitzur... Govind R..... The notched fatigue behavior of the ISR tees is as good as or slightly better than that of the extrusion...432 17-19. (Ted) Highberger... I ROOM TEMPERATURE K.......... J. '" ::..... ..8 FLAT SPECIMEN WITH 0. 60 iii o!'. ~ x « ::...... Van Tyne and B. Eds..I_ _---L_ _---L_-L..... American Society for Metals... '" ..-.050·IN. C. Ti-6AI-4V: Effect of Isothermally Rolled vs Extruded Material on Cycles to Failure 100 ..

Ti-6AI-4V: Comparison of Wrought vs Isostatically Pressed Material for Cycles to Failure 1200 433 o. notched specimens. 9th Edition. load controlled smooth specimens.R = 0. ~ ~~ . I I I I o Engme mount supports 125 n.= 1. Top: Data are for the standard wrought grade.= 3.17-20. Bottom: Data are for isostatically pressed alloy powder.1. Powder Metallurgy.~ ~ <: ~ 100 "iii -"" ~ 75 1il E :J '" e E <I> 50 'x co ~ ~ ~ 88 ~ ~~ 10' ~ ~ r-. Metals Park OH.. 1984.""~~ grade Ti-6AI-4V I 1111 150 <I> ~ '" U.I.R = O. -: ~ ~ ~ ~ ~~ ~ ~ ~ ~\ ~ ~ ~ ~ ~ ~~ ~ ~~ r':':". Source: Metals Handbook. ~" 25 100 10' 10' 105 10' Cycles to failure S-N curves for titanium alloy engine mount supports.K.P 654 .0. K. ~ co 1000 w'"". Volume 7. ~ co 700 1il 500 E :J '" e E <I> Wrought standard grade Ti-6AI-4V ~ ~~ ~ ~ ~" ~ C'\ "x co ~ 300 ~ ~" ~ C'\ ~ ~ r-. 800 E :J I ~ ~ """ ~ ~~ ~ ~ ~ ~ ~ 125 <I> 1il ~ '" "~ X ~ co ~ ~~ ~ ~" 600 • Engine mount supports o Witness blocks 1lJ1I Radius failures I I ~~ ~ 5 10 ~~ ~ ~ ~~ ~ ~ ~i'- 400 10' ~ ~ :\I'ci ~ ~ ~ :\" ~ ~t'~ ~" 10' ~ ~ t:S ~" 100 ~ "x co E :J E 75 10' 10' Cycles to failure 900 I.lo. American Society for Metals.

.::~-----~------. fretting similarly reduced specimen life 7 at overstress levels..om. 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. however. and 340 °C at a mean tensile stress of 140 MNm·2• Room temperature fretting was found to have little effect on the fatigue strength at 107 cycles. Source: Practical Observations of Fretting Fatigue Cracks. Ti-6AI-4V: Effect of Fretting and Temperature on Cycles to Failure ~ z ~ <Jl <Jl • R.------' 1(f CYCLES 10~ 10~ TO FAILURE Effect of simultaneous fretting and fatigue ofshot-peened Ti-6AI-4V at room temperature. the fretting fatigue strength at 10 cycles was reduced to approximately 40% of that found under room temperature conditions. tears.. the fretting fatigue life in the overstress region (70 MNm. p 180 . furthermore. It is the stress state acting in concert with stress raisers (e. At 340 °C.L. o 340°C z ~280 a: z w ~ 140 <l: °1LO"':'~ ---. more importantly. Fretting at 200 °C lowered the fatigue strength by approximately IS%.":.g. cracks) which determines the actual fatigue propensity.T • 560 Fatigue (all test 200°C In 420 o a: w ~ . It should be said at the outset that visual assessment of fretting mildness or severity is inconclusive by itself.L:.. It may in fact be misleading and should not be relied upon for assessing the severity offatigue life degradation. 200. in that the presence of more or less fretting debris on a microscopic examination is not necessarily relevant to the loss of surface integrity... pits.434 17-21..2 above the nonfretted run out stress) was lowered by two orders of magnitude compared with results in the absence of fretting.

17-22. ZERO MEAN STRESS TEMPERATURE: 75' F ENDUR. 1982.< 60 -..l .6 10 ---- I I ABUSIVE MILL az 59 ABUSIVE GRIND 13 65 10 5 II 107 CYCLES TO FAILURE S-N curves for beta-rolled titanium alloy Ti-6AI-4V... FINISH U..v. Z 0: lol f. Donachie. MILL 51 zo 40 f. LIMIT SURF... American Society for Metals. Source: Norman Zlatin and Michael Field. PERIPHERAL END MILLING. CHEMICAL MILLING MODE: CANTILEVER BENDING. Ti-6AI-4V (Beta Rolled): Effect of Finishing Operations on Cycles to Failure FATIGUE CHARACTERISTICS OF BETA ROLLED TITANIUM 6AI-4V... ~ ~I I ""'-.. Jr. Matthew J..." in Titanium and Titanium Alloys: Source Book.. Metals Park OH....p 354 . Ed. "zo r-.. 435 az R c METAL REMOVAL PROCESSES: SURFACE GRINDING... "Procedures and Precautions in Machining Titanium Alloys.. :0- III CHEM.. ..< .. . 80 :<: lol '" 0: \J en '" Z E=: .. Curves show the effects of the various finishing operations on fatigue..... .I ~ II CONDITION I<SI -...:- I I I I III GENTLE MILL GENTLE GRIND 66 6l 41 35 I .

Starke. Ti-6AI-4V: Effect of Yield Strength on Stress-Life Behavior 700 -. In order to establish microstructural effects on fatigue behavior. 24h 500°C 650 • 0"0. P 237 .J Q. Source: Edgar A.2 (J"02 910 MPo o eooppm 02 990 MPo • 1900ppm 02 I. 1979. ::> :::E 600 • « 550 UI UI ILl IUI a: 500 450 4 10 CYCLES TO FAILURE Effect of yield strength on the stress-life behavior of two Ti-6AI4Valloys. :::E ILl 0 Ti . "Cyclic Plastic Deformation and Microstructure. especially for stress-controlled tests." in Fatigue and Microstructure. tJ a. Jr.. This is illustrated above where it is shown that two titanium samples having different yield strengths have different stress-life behavior when tested at 500°C (930 OF). American Society For Metals..... comparisons should be made on materials having the same yield stress.. and Gerd LOtjering. Metals Park OH.6AI-4V.436 17-23.

~ ~ ~ 30 -. The solid points in the graph above represent the results obtained with these specimens. failed to reveal any surface cracks. rather than the test procedure. Office of Naval Research. it was assumed that the sample was abnormal. If so. Brief experiments soon disclosed that temperatures at least as high as 400 OF(205°C) did not lower the hardness. a value considerably lower than for any other known metal or alloy.000 psi. Results are indicated by the open points in the above SoN diagram.3 in similar tests. Most investigators have obtained normal values around 0.000 psi is chosen.155. o I \. The endurance limit was not altered significantly.S..lfan estimated limit of 19.H T 400"r.). the hardness may have increased very slightly. Very careful oil-powder and fluorescent powder tests. The cracks did not appear to be associated with any visible surface imperfection. a set of sheet fatigue specimens was stress relieved for two hours at 400 OF(205°C). 2 HRS. Flexural tests of the sheet specimen were made at 1725 CPM. supplemented by metallographic examination. A definite shift to the left in the upper portion of the curve was evident. Ti-6AI-4V: Effect of Stress Relief on Cycles to Failure 60 437 \ 0-AS RECEIVED I e. II II . although the direction of shift was opposite to that. For these reasons.17-24. It was considered possible. The endurance limit was not reached at stresses as low as 20. a moderate temperature stress relief might help. that residual stresses from cold rolling were acting in a deleterious manner. even when the sheet was flexed to open any incipient hairline defects.000 T. though not probable. Several of the sheet fatigue specimens developed fatigue cracks away from the milled specimen edges.BlOKE IN GRIPS CYCLES Flexural fatigue tests of titanium sheet (123.~ II: iii "'" 40 1'. Knowing this. in fact. ~ Iii 20 ~ Ul ~ ~ - T 0 kr . p 97 . had the heat treatment released undesirable stresses. the endurance ratio would be only 0. Source: Titanium Symposium.

p 26 . B = low stress grinding.ee 2. chlorinated. E 7EUI .5 Sol. Christopher. 1979. Influence of Chlorinated and Sulfurized Cutting Fluids On High Cycle Fatigue Properties of Ti-6Al-4V Machined Surfaces at 75°F 10 7 Cycle Fatigue Strength (ksi) Low Stress Abusive End Milling Grinding Grinding 75 65 62. D.----.. C = end milling.5 57." in Machinability Testing and Utilization of Machining Data. e __ ~-- ---==='~II -- ~ _ C ~_== --_.. A. Tipnis and J. American Society for Metals. -. .. - ~~ e.--= ----= .28e ---- --. A = abusive grinding.e8 CYCLES TO FAILURE (IN MILLIONS) Alternating stress vs cycles to failure in high cycle fatigue of machined titanium surfaces using neutral. "Machinability Testing for Industry. 6.438 17-25.••••• --. _-- .5 12. Cutting Fluids Neutral Chlorinated Sulfurized 12. Metals Park OH.5 67..00 le..60e I..5 12.5 Source: V. Ti-6AI-4V: Interrelationship of Machining Practice and Cutting Fluids on Cycles to Failure lee.. and sulfurized soluble cutting fluids.~-- - .Iee .ee .

.~ntl~ I 66 I l2 Ah . 1982.... Ti-6AI-4V: Relative Effects of Machining and Grinding Operations on Endurance Limit Tio 6AI-4V B ETA ROLLED.. etve I 30 164 177 80 80 END MILLEND CUT Gentle u . Metals Park OH. Source: Norman Zlatin and Michael Field.. Matthew J.. l2 R c Surfac e Roughness..17-26."in Titanium and Titanium Alloys: Source Book.. American Society for Metals. Donachie.• Ed. P 354 . lu~ 'II 59 CHEMICAL MILLING ~""n". "Procedures and Precautions in Machining Titanium Alloys.~ntl~ I 62 35 65 HAND GRIND 157 Ab . I 51 14 5 I I 20 165 I I o 20 40 60 80 ENDURANCE LIMIT. nu_~· . KSI Bar chart presentation showing relative effects of various machining and grinding operations on fatigue characteristics of titanium alloy Ti-6AI-4V... 67 84 END MILLPERIPHERAL CUT r.. AA 439 SURFACE GRIND Gentle ~13 r. Jr ..

"Procedures and Precautions in MachiningTitanium Alloys. Jr.. P 355 . Source: Norman Zlatin and Michael Field. KSI Bar chart presentation showing effects of various metal removal operations on the fatigue characteristics of titanium alloy Ti-6AI-4V.. 35 R c Surface Rouglme9s. Ti-6AI-4V: Effects of Various Metal Removal Operations on Endurance Limit Ti. Metals Park OH. AA SURFACE GRIND ECM FRONTAL ECM TREPAN Gentle I 67 14 Il 160 I 40 I 161 I o lO J 40 I 80 I 60 ENDURANCE LIMIT. American Society for Metals. Ed. Donachie.440 17-27.6Al-4V ANNEALED. Matthew J. 1982."in Titanium and Titanium Alloys: Source Book.

. C.. George Krauss.... Source:J.. .. "The Role of Therrnomechanical Processing in Tailoring the Properties of Aluminum and Titanium Alloys. American Society for Metals." in Deformation........ Processing... Ti-6AI-4V: Effect of Texture on Fatigue Strength EFFECTS OF TEXTURE ON SMOOTH FATIGUE LIFE 441 130 120 ..1 . 1984. Metals Park OH...... A.Jr... Starke. and Structure..cycles SON curves showing the effect of texture on the fatigue strength of Ti6AI-4V.. 90 Ti-6AI-4V Re-X Anneal R =0.....L. E <.... 80 L-- --------J Nt .. b lOa Load axis II [loio la ~ . Williams and E... Ed..17-28....... P 334 .. .... . Fatigue strength is greater when the stress axis coincides with the direction of a high density of basal poles.... 110 I .L . ..

... 1984. Williams and E.... "in Deformation.. and Structure. Starke. These data show that a mixed texture lowers the high-stress end oftheS-N curve preferentially.. Metals Park OH.. p 335 . Ti-6AI-4V: Effect of Complex Texture on Cycles to Failure Ti-6AI-4V ~ Air 900 E \ \ \ \ -[]o-o~\ ::.. Jr.. <. . A. -0 800 -. E II> II> 700 « <fl -. George Krauss. Processing.442 17-29..: z ...- ~ a.. C. "The Role of Thermomechanical Processing in Tailoring the Properties of Aluminum and Titanium Alloys. Ed..10 7 '" 600 10 4 10 5 106 Cycles to Failure S-N curves showing the effect of more complex texture on fatigue strength of Ti-6AI-4V. . Source: J. American Society for Metals. ..

(b) Tested in 3'12% NaC\. Metals Park OH. Processing.... "The Role of Thermomechanical Processing in Tailoring the Properties of Aluminum and Titanium Alloys.. a Ti-6AI-4V 31/2 % <l: ~ . American Society for Metals.. _----- lOS Cycles to Failure Q.processed Ti·6Al·4V. Ed . -.. -. . Jr. showing the effects oftexture and environment on fatigue strength.17-30.. C.. p 335 . -. George Krauss. 1984. E Nael Solution 900 \ IJ) \ 800 \ \ \ 700 \ \ -... Starke.8.." in Deformation.. Ti-6AI-4V: Effect of Texture and Environment on Cycles to Failure TI-6AI-4V AIr 443 -. A. <. 600 104 ~------ 105 Cycles to failure 106 10 7 b SoN curves for O! + . Source:J. Williams and E. These data show that testing in an aqueous 3'12% NaCI solution reduces fatigue strength when the stress axis is along [0001]... (a) Tested in air. and Structure. <.

. Most structural materials show variations in near-threshold FCP rate and in tiK.444 17-31. American Society for Metals.K- Schematic plot showing characteristic shape offatigue crack growth rate (doldN) versus cyclicstress-intensity (~) curves. P 338 . Jr.. Williams and E. In contrast. Moreover. Ed. At the highest crack growth rates shown above. Metals Park OB. These factors include microstructure and texture.. Source: J.. However. and Structure. Ti alloys show significant variations in FCP rate over the entire range.. since the majority of the lifetime of a crack component is spent in the low-FCP-rate regime. Processing... it is of little interest and will not be discussed further here. e .. but fewer show significant variations in FCP rate in the Paris law regime. 1984.KalK 1c . It can be seen that at higher growth rates there is a linear portion of the curve.1 0 L-_ _..L<. the FCP rate curve bends upward. factors which control FCP at rates less than ~ 10-6 tu] cycle are probably most important. Ti-6AI-4V: Fatigue Crack Growth Rates Kmin+lJ.. since crack growth rates are uncontrollably rapid in this latter regime. 10... This linear portion was represented as tiK'" by Paris and Erdogan and is now frequently referred to as the Paris law regime of fatigue crack growth. This is controlled by fracture toughness. "The Role of Thermomechanical Processing in Tailoring the Properties of Aluminum and Titanium Alloys. Starke." in Deformation.. George Krauss.. A.. _ log lJ.. C. o '" .:»:" 1 z .

" in Production to Near Net Shape: Source Book... and extrusions.. T. J.ROOM TEMPERATURE TL AND LT ORIENTATION • ISR TEE o EXTRUSION 000 o W ... along with data for the extrusion. it can be seen that particularly at lower stress intensities the fatigue crack rate for the (lSR) isothermally rolled tee is faster than that for the extrusions.. Ti-6AI-4V R = O.~K. Source: W. This is probably due to the extrusions being beta formed while the ISR tees are alpha-beta formed.J U ~ :r: o >z z ~ "tl w· « a: :r: I- a ~ ~ a: (9 o « a: o w ::J (9 ~ lL « o o 10 10 0 -7 ':------:~--..-7:.... In the chart. Fatigue crack growth results for ISR tee. Scarich.-----. Metals Park OH.~::_":~ 20 40 60 80 100 STRESS INTENSITY FACTOR RANGE. Data for both TL and LT orientations are shown above. Chanani and Gregory V.. Van Tyne and B. Ti-6AI-4V: Fatigue Crack Growth Rates for ISR Tee.... Govind R.:-.-_:'. Eds.. Avitzur....... KSI viN. 1983. In comparing individual results. (Ted) Highberger.. C. and Extrusions 102 445 rr========::::::.l.17-32. P 124 . American Society for Metals.. no differences were seen between TL and LT. "Advanced Titanium Metallic Materials and Processes for Application to Naval Aircraft Structures..

HEAT 3 b . ~ 0 : : (INCHES/CYCLE) 0 I. "Effect of Chemistry and Heat Treatment on the Fracture Properties of Ti-6AI-4V Alloy." in Titanium and Titanium Alloys: Source Book. Ed.HEAT 2 ENVIRONHENT .HEAT 1 -to.RW R FACTOR 0. Harrigan. 1982. Donachie. P..IE ~ _°11 . Kaplan and A.30. Sommer. Jr.RECRYSTALLI ZATI ON ANNEAL Fatigue crack growth rates for three different heats of Ti-6AI-4V titanium alloy.446 17-33. Matthew J. when recrystallization-annealed.~ ~ - o o '§ o 10 dK (KSI {INCH) 100 0. J. American Society for Metals. Metals Park OH.K tested at an R factor of +0.P 65 . The crack growth rate is shown as a function of 6. W.. behaved similarly with decreasing oxygen and aluminum. 30 HEAT TREATHENT . M. Ti-6AI-4V: Fatigue Crack Growth Rates . The fatigue crack growth rate in the RW orientation for this alloy. Source: M.LOW-HUHIOITY AIR ORIENTATION .

From the data presented above.1. (20 MPa .. Lewis. A. G. The slightly decreased crack growth rates above a IJ. Crossley. M.s 10. and 1450 OF (787. may be attributed to the proximity of the maximum stress intensity to the critical value. J. air cooled. per se. Metals Park OH.~ CLUDED FOR COMPARISON) FC WQ 100 10. Matthew J. Jill) Source: R. both water quenching and furnace cooling resulted in fatigue crack growth rates noticeably different from those measured for the mill-annealed base material. American Society for Metals. ~). Jill). However.3 0q. 1982. As shown. Ed. it can be seen that air cooling.. Donachie.K of 20 ksi MPa· VITi) are." in Titanium and Titanium Alloys: Source Book. The accelerated growth rate above 18ksi \!Ill.1 10 100 ~K I ksi- Jin": Effects of final cooling rate on fatigue crack growth rate in duplex-annealed Ti-6AI-4V. plate. Jill. P 90 . MPa • 447 -. 1/2 h. Morton and F. "Effect of Cooling Rate on Fracture Behavior of MillAnnealed Ti-6AI-4V. 5 Z .jm 100 10 I CONSTANT AMPLITUDE FATIGUE: R = 0. 1775 OF (968. I h. more probably than not. u '" ':- . T. E. (20 MPa .4 "C). I-in. the result of the higher fracture toughness of the air-cooled material.jll1. The critical stress-intensity value for water quenching was an exceptionally low 38 ksi . Bjeletich... cooled as noted.17-34.3 o >. 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.4 I Y :'/ k~ '!lAC :/ :/ I ~ "U 10. 10 TO 20 HZ MAXIMUM SCATTER OF ACTUAL DATA POINTS FROM ANY CURVE IS LESS THAN 40% WQ AC FC A'A WATER-QUENCHED AIR-COOLED FURNACE-COOLED MILL-ANNEALED (DATA IN. (42 MPa .":' I :" :. produced little or no change in the cyclic crack growth compared to the mill-annealed base (22 material. :/ 1I :/ 0. Jr. Ti-6AI-4V: Effect of Final Cooling on Fatigue Crack Growth Rates ~K.

. 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. Q> .. Metals Park OH. London.. In all cases. England. Particular attention has been paid to a / {3 alloys. Ti-6AI4V. J. Ti-6AI-4V: Effect of Dwell Time on Fatigue Crack Growth Rates 10.J::. e. P 200 .g.. Such an accumulation of hydrogen would tend to embrittle this region.s ~ .-... Q> u >u 10. ~ a. 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. This effect is shown schematically here. dwell effects disappear when stressing occurs at temperatures above 75°C (165 "Fj.and they are generally considered to arise from the preferential diffusion of hydrogen. during the dwell period.4 . and it has even been suggested that brittle plates of TiH 2 may be formed. and American Society for Metals.6 -" u I I I I Sinusoidal loading U 10-7 10 --Dwell at maximum load 100 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.. to regions of localized hydrostatic tension ahead of an advancing crack. Edward Arnold Ltd. or if the microstructure is homogeneous and fine grained.448 17-35. Source: I. in which dwell effects have also been found to decrease with increasing amounts of the {3-phase in the microstructure. whereas they appear to be insignificant if stressing occurs parallel to the basal planes of the aphase.. Light Alloys. 1981. m 10. I 'i 0 0.5 ( I . Polmear.

and American Society for Metals. K = 1.1. England. Ti-6AI-4V -SA 10 20 50 100 Left: Fatigue crack growth rates for Ti-6AI-4V rolled plates in the .' " . 1981. . is attributed to the slower progress of cracks through the Widmanstatten microstructure. London. ".17-36. Light Alloys. Above: Branching of fatigue cracks within the Widmanstiitten packets of the a-laths. Tests conducted at 5 Hz using compact tension specimens. P 179 . Edward Arnold Ltd. in turn. Stress-intensity factor range LlK(MPa m1J2 ) 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).5 h 1038 0 C. Metals Park OH. Polmear. Source: I. air-cool to room temperature. BA = 0. J. Ti-6AI-4V: Fatigue Crack Growth Data 0 449 • Annealed 2 hours at 705 C. This effect..0 • alP 0 t r ~ I -. air-cooled after forging transus 1005 C " Axial loading: smooth specimens. Ratio of minimum to maximum load = 0.a-annealed (HA) and mill annealed (MA) conditions. particularly at stress intensities below a critical value at which desirable crack branching occurs within packets ofthe a-laths.

. J.cr::ti±==I==:±=~:::l. "Powder Metallurgy Parts for Aerospace Applications.lll_--. Advantages and Limitations. Source: J.-'~ 106 10 Cycles to failure Notched fatigue strength of HIP'd P 1M Ti-6AI-4V compared with fatigue strength ofalpha-beta processed forgings.l. H. V.. 1983. American Society for Metals... Ed.K. Petersen and E.. P 286 .1.J_L-L. alpha-beta-processed forgings Upper and lower limits of as listed in AFML TR-73-301 30 LJ.Ll.R= 0.. Moll. Ti-6AI-4V P/M: Comparison of HIP'd Material With Alpha-Beta Forgings for Cycles to Failure r: 70 ~o o ~OHIP ------00------ o 40 ~-_-:--. = 3. Dulis. Hz = 30.. Erhard Klar..450 17-37." in Powder MetallurgyApplications. C.. Metals Park OH.

Theodore Highberger. Ti-6AI-4V P/M: Comparisons of HIP'd Material With Annealed Plate for Cycles to Failure 140 451 120~ 100 r o 'GM"S'$. According to the above data.3/-1J:? ANN P!AT~ 20 1----. Source: W. 1983. Metals Park OH. Avitzur." in Production to Near Net Shape: Source Book. 1 /0'- 10-5 /0· /0' C~ESTO~(L~ SoN curve for HIP'd Ti-6AI-4V and annealed plate.'6/-II-IR 40t I • a 22'50·~-I0/t:!9I-II-IR ''''50. P 304 . C.17-38.. fatigue results for Ti-6AI-4V are within the required range for plate properties from MIL-T-9046. K'S/ 80 ~ Q I/IPC'tCl. Van Tyne and B. American Society for Metals.-_ _--'--_ _---"I 10~ -'-:-_ _----'.~-1'51t:'8/. "Manufacture of Titanium Components by Hot Isostatic Pressing. J. Eds.£ o 15S0i:-I-5 1t:.

There was no difference in HCF test results between the two sizes. Ali. two size fractions of powder were used: -35 mesh (as-received) and -80 mesh. Roomtemperature and 700 OF (370°C) S/ N curves are shown above."in Powder MetallurgyApplications... ~ 42. Petersen and E.1 (20 ) io. H.. Ti-6AI-4V P/M: Effect of Powder Mesh Size on Fatigue Properties 70. Bottom: Properties at 700 OF (370°C)... VI' VI Q.452 17-39. en 28. Metals Park OH.2 (80) 'Vi ~ m :E a. o~l :E en ~' . Dulis. 1983. Advantages and Limitations.. O As received powder (SM 772) li. Moll. Erhard Klar. In this study.. Ed.1 (40 ) -- ~li. ~ -:--r---- li.. -80 mesh screened powder (SM 768) 14.. 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.2 (60) . Pli. 10. . American Society for Metals.l 42. Top: Room-temperature properties. J. P 286 . Source: J..2 (60 ) 28.3 (100) 56.1 (20) 0 104 ·10. 70. "Powder Metallurgy Parts for Aerospace Applications.2 (80 ) 'u. C. V.1 (40) o As received powder (SM 772) 0-80 mesh screened powder (SM 768) 14.3 (100 ) 56.

9th Edition. Source: Metals Handbook. Powder Metallurgy.>< iii e li) 'x co ~ III 800 ~ 100 ti E :J E 600 'x co ~ E :J E 400 60 200 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. Ti-6AI-4V P/M: Comparison of Blended Elemental. also. microstructure determines the ability of the compact to accommodate foreign particles and resist crack initiation. This situation is achieved by careful control of cleanliness (powder handling) and microstructure. Blended elemental compacts.P 753 . conditions of subsequent handling. Cleanliness depends on the environment in which the powder is produced. 140 'iii iii III co ~ . R = O. chemcial heterogeneity may lead to areas of similarly aligned alpha plates. Prealloyed powder compacts exhibit fatigue behavior equivalent to that of Ij M materials.17-40. The fatigue behavior of titanium PIM compacts is compared to wrought products in the graph above. and microstruture developed by compaction. American Society for Metals. Metals Park OH. compete well with many titanium alloy castings in fatigue strength. 1984. The blended elemental material is inferior to prealloyed compacts and II M materials. Cleanliness dictates the amount of contamination contained in the final product.. however. Volume 7. This is caused by residual chlorides and consequent porosity. Tested at room temperature. Prealloyed and Wrought Material for Effect on Cycles to Failure 1200 453 1000 Q.

Powder Metallurgy. Source: Metals Handbook. Metals Park OH. American Society for Metals. P 44 .454 17-41. 9th Edition. Volume 7. co ~ 1200 1000 800 600 400 103 104 10 5 '(ij ..>t: '" 1ii ~ CO en 140 100 60 106 10 7 ~ f/l '" f/l E ::J E x ~ 10 8 'x CO ~ E ::J E Number of cycles to failure Fatigue chart presentation showing a comparison of fatigue behavior of Ti-6AI-4V compacts with ingot metallurgy material. 1984. Ti-6AI-4V: P/M Compacts vs 11M Specimens: Cycles to Failure Q.

. Metals Park OH. 1984. ~ III ro 800 120 'ijj . Powder Metallurgy.P 439 . 9th Edition. Volume 7..". 'x 400 ro ~ E ::J E 'x ro ~ E ::J E 40 o"10 3 -'- ... American Society for Metals. The inconsistentfatigue life ofthe hot isostatically pressed product is usually casued by inclusions in the compact. ---I0 108 Cycles to failure SON curves (bands) for titanium alloy Ti-6AI-4V processed by various fabrication processes.L. Source: Metals Handbook.1 Annealed 455 160 a.- --' . ~ 1.... Ti-6AI-4V: Comparison of Specimens Processed by Various Fabrication Processes for Cycles to Failure 1200 Axial fatigue smooth Room temperature R = 0.L. III u> 80 vi ~ 1......17-42.

. Volume 7.. 01 III .. p 752 .5 ai . Powder Metallurgy.. 10 3 1 10 11M and P/M Ti-6AI-4V at room temperature (laboratory air).~ E E 3 u Q) .. R = 0. 10. MPaViTl Comparison of fatigue crack growth rate of Ti·6AI·4V P 1M compact with 11M material heat treated to various conditions. 9th Edition.>0: 10.6 10. o 10. ..... ~ 0 .s::. The fatigue crack growth rate of blended elemental and pre alloyed compacts is equivalent to 11M material with the same microstructure. ksiVTrl.456 17-43.. P/M vs 11M Stress intensity.. ~ (J Recrystallization anneal s: i. Ti-6AI-4V: Comparison of Fatigue Crack Growth Rate.5 . Source: Metals Handbook.. III (J Stress intensity. at 5 to 30 Hz 2 10 10-2 10-4 Q) U ~ c III > 10. Metals Park OH. 1984. ai 0 ..1. American Society for Metals.4 ~ U ..

American SccietyIor Metals. P.17-44. Schwartz. 0 >l u. FLUSH WELD BEAD 0---. J. Ed. 1981.0 T STRESS RELIEVED.440-in. Melvin M. G. 50 BASE METAL SSEB WELDEO --0 10 6 NUMBER OF CYCLES (LOG SCALE) TO FAILURE S-N curve for titanium alloy plate-base metal versus SSEB-welded. <. Will. P 87 .~ 70 :E LEGEND ---/:::. Results show that the constantamplitude fatigue life of SSEB weldments in O. Metals Park OH. Madora and H.. Source: R." in Source Book on Electron Beam and Laser Welding. Ti-6AI-4V: Base Metal vs SSEB-Welded Material for Cycles to Failure 457 130 ~ 110 <. 0 TEST CONDITIONS: CONSTANT AMPLITUDE. R ~.1 K ~ 1. c: <. Ellison. 90 l!l' w a: ~ IJ) t- """- CD :E :::> :E « x ~. <.. "Sliding-Seal Electron-Beam Welding of Titanium. H.-thick plate equals that of the base metal.

P. Schwartz.1 10 1lX' STRESS RELIEVED. Ed. "Sliding-Seal Electron-Beam Welding of Titanium.458 120 17-45.0 SSEB WELDED ---. R KT~ ~. American Society for Metals. J. FLUSH WELD BEAD 80 o o ~ LEGEND: 40 --. Ti-6AI-4V: Base Metal vs SSEB-Welded Material for Cycles to Failure TEST CONDITIONS: CONSTANT AMPLITUDE. Metals Park OH. Results show that the constantamplitude fatigue life of SSEB weldments in O. Madora and H. Witt. 1981. H. G.6 20 BASE METAL SON curves for titanium alloy plate-base metal versus SSEB-welded. p 87 .940-in." in Source Book on Electron Beam and Laser Welding. Ellison. Melvin M.-thick plate equals that of the base metal.. Source: R.

-thick Ti-6AI-4V STOA.17-46. Matthew J." in Titanium and Titanium Alloys: Source Book.OSQ-IN.080·in. A. Witt.. 1982.. "Weldability and Quality of Titanium Alloy Weldments. Upper curve shows scatter band for base metal (O. Donachie. Source: R. Ed. American Society for Metals. Flescher and O. Paul. Ti-6AI-4V EB Weldments: Base Metal Compared With Flawless Weldments 140 0 0 459 o 120 0 O.P 313 . Jr. Metals Park OH.).·THICK BASE·METAL Ti-6AI-4V STOA CURVES ci " s en ~ 100 o 0 o o o o en' en w a: Ien ::E X « SO 60 40 ~ 103 10 5 CYCLES 20 SoN curves for EB weldments that were flawless (lower two curves).

Flescher and O. Witt.·THICK aASE·METAL Ti·6AI·4V STOA CURVES ~ d II 100 a: ~ in ao w X ~' Ii. For the points within the boundaries of the band.oaO·IN. :E <l: a: 60 40 20 103 105 CYCLES SON curves for ED-welded Ti-6AI-4V titanium alloy showing effects of porosity. Metals Park OH.005 in. Above are shown experimental data obtained for porosity-containing ED welds which are superimposed on a set of curves for the base material (0. radiography indicated either linear or heavily scattered porosity.. radiography indicated scattered porosity (0. Source: R. A.003 to 0. 1982. Donachie. American Society for Metals. For points below the lower boundary of the band. P 312 ." in Titanium and Titanium Alloys: Source Book Matthew J. Ti-6AI-4V EB Weldments: Effects of Porosity on Cycles to Failure a a 120 a a a O. Paul.460 140 17-47. factors. Jr. in diameter). Ed.080-in.-thick Ti-6AI-4V STOA sheet) at various K.. "Weldability and Quality of Titanium Alloy Weldments.

"Weld ability and Quality of Titanium Alloy Weldments. thick) for various K 1 factors." in Titanium and Titanium Alloys: Source Book. 1982..25 in. Jr. 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. Flescher and O.. p 313 . Metals Park OH. Paul. Ti-6AI-4V Gas Metal-Arc Weldments: Effects of Porosity on Cycles to Failure 140 0. American Society for Metals.17-48.-THICK BASE-METAL Ti-6AI-4V STOA CURVES 461 D D 120 0 ~ 100 D Cii ~ iii lI) lll) a: w BO ::E ~ 60 D D 40 D o o A 105 CYCLES SON curves for porosity-containing gas metal-arc welds. Source: R. Matthew J. Witt. Ed. Donachie.250-IN. A.

0 • HEAT TREATED ELECTRON BEAM WELDED a STRESS RELIEVED TI-6AI-4V Room temperature rotating-beam fatigue life of unwelded and electron-beam-welded Ti-6AI-4V titanium alloy in fully heat-treated condition.. Schwartz. R. American Society for Metals. M. Strautman and W. 1981. Q= 5 z S C[ z It: W HEAT TREATED TI-6AI-4V 20 f· 7500 CPM =1. V. "Application of Electron Beam Welding to Rotating Gas Turbine Components..I . en III It: w 100 80 t." in Source Book on Electron Beam and Laser Welding. Ed. P 169 . o I 6 10 CYCLES TO FAILURE ~ . (!) 60 40 --Kr . Source: S. Decrease in fatigue strength ofthe weldment relative to the parent metal did not exceed 12%. Silverstein. Melvin M. -'" 'r.. Freeman..462 17-49. Metals Park OH. Ti-6AI-4V: Unwelded vs Electron Beam Welded Material for Cycles to Failure III 0.

it was concluded that no porosity or weld defects of sufficient size to preferentially initiate fatigue fractures was present... Ed. Greenfield. of all un-welded control )( 600 N specimens 80 ~ • . ::iE IV <. "Laser Welding-The Present State-of-the-Art.. (0.. :.. American Society for Metals. E IV 400 ')( ---- 300 ::iE 20 10 8 S-N diagram for laser-welded titanium alloy sheet.17-50. Where failures initiated in the base metal. M. C.. A.0 0 'iii Ul Ul o 0.. <... Source: E.... M. 60 40 ')( E E <. P 289 .. E c Gl Ul -... Metals Park OH..584 cm) sheet • 0. ." in Source Book on Electron Beam and Laser Welding.. . Schwartz.230 in. Melvin M. Ul l3 Gl E :.¥ II $1!!~iJfift: '@ oco .. whereas other weld failures originated at undetected small pores..140 in.. Ti-6AI-4V: S-N Diagram for Laser-Welded Sheet 120 463 Legend ..... The fatigue properties of welds as shown above indicate that under proper welding conditions..356 cm) sheet .. Banas and M.. . The best laser weld failures initiated at sites in the base metal. Mean curve for plasma-arc welds 700 co I 0 of mean fatigue strength .. 1981. ci 100 a: ... laser welds can be made in Ti-6AI-4V which exhibit base metal fatigue characteristics. Lower bound data for pIa welds J on flat sheet with filler 10 6 10 7 Cycles to failure --- ---- 4 runouts .. (0. Breinan...

but a few tests were run at K/ = 1.. +000# :E ~. Volume 3. Stress ratio. 1980. typically was +0. K/.61:>. '" ~ 'j( E ::J E ~f:>6'17 00 '" e . of stress cycles Notch fatigue strength of as-cast Ti-6AI-4V.+°oOOx O +0 0 o-tCo 00 00 :E . o 00+-tp.0..n :E .~66 '17 'Ii' '17 o 000 01:>. Metals Park OH..R..1. Ti-6AI-4V (Cast): S-N Diagram for Notched Specimens 1000 900 800 t.. 700 600 500 400 300 200 100 oooeJO 0000y 00 0 0 00'17001:>. 9th Edition. '" 'j( ..0.60 + 0 0 0 ''. Properties and Selection: Stainless Steels. American Society for Metals.>t .P 411 .Q) +0 x ooo. . Each symbol represents fatigue data from a different source. Tool Materials and Special-Purpose Metals. Source: Metals Handbook.90 .o Do o 0 0 120 . stress concentration factor..464 17-51. . E ::J E +o~ooo oc:P 30 No. was mostly 3.

. Inc. showing effects of notches and elevated temperature (400°C. U -" Cl to en en l- w a:: en 0." in Corrosion and Corrosion Protection Handbook...• New York NY.18-1... zirconium and its alloys exhibit a fatigue limit behavior similar to most ferrous alloys. Ed. Schweitzer. Philip A. "Zirconium. Marcel Dekker. or 750 OF) on fatigue characteristics. Knittel. N e -. Zirconium 702: Effects of Notches and Testing Temperature on Cycles to Failure 465 '"0 . p 198 . Source: Donald R. 1983.5 CYCLES S-N curves for zirconium grade 702. As indicated above.

. Steel Founders' Society of America. o (5.. '"'ox S-N CURVE -CYCLES OFSTRESS COMPUTED STRESS [3 ~ 30 11::11-0 VI Cl: Vl llJ lbX~'. 5th Edition. P 7-6 . comparing hand weldments with machine weldments. u.-- X XX x -_ --o-J!l< x)("" u_ .2-BW HAND WELDED.. ~ X "" STRESS VS COMPUTED S-N CURVE -CYCLES OF "" . 1980......... Steel Castings (General): Effect of Design and Welding Practice on Fatigue Characteristics 'W:..1:-:!-:: x~x x 10 6 NUMBER OF CYCLES 7 10 S-N curve for cast box designs....2 ..x....466 19-1... ... Rocky River OH.. Ed. 'o'x VS '-. and (2) the cast steel box design is superior to a weldment design.BW MACH WELDED.""'. Q The S-N curves shown above indicate that: (1) the welding practice is of no great importance..~ ~ I 40 ~ 50 x. Source: Steel Castings Handbook. ::f . STRESS RELIEVED-o .Q....... " ~ olE llJ 20 10 "'80x.....0.. STRESS RELIEVED- 105 106 NUMBER OF CYCLES SoN curves of box weldments. 0"..... x STRESS .x ~~~ o .oX. ~ x Xx 00 . Weiser. Peter F.. All weldments were stress relieved at 1100 F (593 Qq. 0_ -0_ 0 (3 3 -I ::::I X . ..

Weiser. Source: Steel Castings Handbook.4 Goodman diagram for bending fatigue for normalized and tempered 8630 cast steel.4 Z AVERAGE YIELD STRESS RANGE lI) It: It: l.0 467 0. This is sufficient to account for variations in surface finish and minor surface discontinuities. (0.6 0.7 of the unnotched value. The above diagram shows that even severe surface discontinuities.1 STRESS TENSILE STRENGTH .OOIS-in. (Machined notch of R.0381-mm)root radius circumferential notch in a rotating beam fatigue specimen reduces the fatigue limit by about 0.7 0. 1980.1 0 0. Steel Castings (General): Effects of Discontinuities on Fatigue Characteristics 1.5 0. do not reduce the fatigue limit by much more than the 0.7 value.2 0.6 0. Steel Founders' Society of America.5 :I: lW :I: IW e> ~ 0.19-2.0015-in. A good design approach is to use the notched fatigue limit asa design value.. (O.0. 2 . Moore specimen: 60° included angle. R.9 0.O.0381mm) root radius.3 TENSILE STRENGTH W e> ~ W . 0. 5th Edition. For cast steels a O.3 -0.6 MEAN . A highly polished smooth test specimen can exhibit twice the fatigue strength of a rough machined sample. Ed. Peter F. I- ~ 0.IlI) 0. Rocky River OH. not normally permitted by workmanship standards. 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. P 7-6 .0.] Surface condition has a significant effect on fatigue life and fatigue limit.J ~ <l: W u.

9th Edition.--.---=c. 1978.. ksi 50 1000 100 150 200 250 I II 800 D....j 2oot-----------1----------+---------'''''''''. 600 v 125 ~ 100 ~ ] u.. l' 50 Ul ~ 4oot-----------1------------'''''f..>-_L --. MPa -:V 75 As-forgedor ] .::>..--=---~c......06 mm (0... 207 MPa (30 ksi) for the as-rolled specimens.. American Society for Metals.--=_=-=""'l-"".. heated to approximately 900°C (1650 OF) in a gas-fired muffle for 20 to 30 min. ::_t--"'-....L_-=... 25 o 200 400 1400 2000 Application of small-scale laboratory fatigue testing to the analysis of components or assemblies introduces additional variables.--. Forging and heat treating produced a surface decarburized to a depth of about 0.468 20-1..).----------~ Ground and polished 8oot-----...~~-.16 mm (0.. Volume I... tempering in air for 1 h at approximately 620°C (1150 OF). depending on whether the surface is polished.. of about 310 MPa (45 ksi).. The curves in the top curve above demonstrate that the fatigue strength of steel specimens varies markedly..... . and air cooled..~ ~ decarburized 50 u.j Ii 25 loot-----------1----------+------------''-l 0'-------------'------------'-----·-------' 0 103 106 Number of cycles to failure Tensile strength. and only 152 MPa (22 ksi) for the as-forged specimens.. the fatigue limit was 393 MPa (57 ksi) for the ground specimens. Sample preparation required that the specimens be machined and polished after heat treatment and that rolling or forging precede heat treatment. Closed-Die Steel Forgings: Effect of Surface Condition on Fatigue Limit 9oor--.------. .""----------l ~ Ul 500 ~--------___I---___.""""=__--=l 75 ~ .....) diameter to a final diameter of 7... oil quenching. Properties and Selection: Irons and Steels. hot rolled.-=..j . Decarburization lowers the strength levels obtained by heat treatment. very lightly swaged from an original 7.:::_II__-___7'''-------_+---------~ 100 7oo~------~..... compared with 470 MPa (68 ksi) for samples that were not forged but were machined or polished and free of decarburization._~---------. Metals Park OH. 3oot-----------1----------+----"'-.. 317 MPa (46 ksi) for the machined specimens. . 600 800 1000 1200 Tensile strength..294-in.0025 in. Source: Metals Handbook. machined.).> 1.. These specimens exhibited a fatigue strength..~~<?' 125 .=_-:=. equivalent to a tensile strength of 876 MPa (127 ksi) and a yield strength of 696 MPa (101 ksi).. One is the effect of surface condition. For a fatigue life of one million cycles. The steel tested was an unidentified wrought low-alloy steel heat treated to 269 to 285 HB. MaChine[5h ::.r------------.... or as-forged.47-mm (0.. / t. Heat treatment consisted of austenitizing in a salt bath at approximately 830°C (1525 OF)for 45 min..L----+----------d 6oo~--------___I--=~:::""..282 in. at 106 cycles.----~c.. P 355 .. 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.=---'''. -----'''''-.-==-"""'-"""".. and water quenching.. ----1--1600 1800 -... 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...~ 400 '!l 200 " ..

For similar P / M and wrought parts. 1979.21-1. P 3 .:::i ~ (138) 20 ~ Cl (69) 10 . Parts containing nickel show improved fatigue resistance compared to iron-carbon steels.4 6.45 .. However.50 ~ 0 ." in Source Book on Powder Metallurgy. fatigue strengths of P/M parts generally are more stable and uniform than for wrought parts.6 6. (276) o g 30 _ (207) 'E :. the ultimate tensile strength to fatigue strength ratios are the same. Ed. and high-density nickel steel parts can be case hardened to improve wear and fatigue properties.40 . ~ Cl Q) ~ 6. Metals Park OH. Samuel Bradbury.. Fatigue strength is best at high densities. Miska..8 7. American Society for Metals.0 Density. gr/cucm 7.35 tx: u. "Powder Metal Parts.2 The relationship of fatigue strength to density is shown above. P/M: Relation of Density to Fatigue Limit and Fatigue Ratio 469 '" 50 ~ (345) tJl 40 0. Source: Kurt H.

Triangles are values for materials without phosphorus. 1979. P 46 .. Fatigue." in Source Book on Powder Metallurgy. Ed. closed circles to PASC materials. Source: Per Lindskog. open circles correspond to PNC materials. Samuel Bradbury. "The Effect of Phosphorus Additions on the Tensile. P/M: Relation of Fatigue Limit to Tensile Strength for Sintered Steels N/mm 2 150 • 0 00 • 100 o o •o 400 Tensile strength) 200 500 N / mm 2 600 Fatigue limit of different sintered steels as a function of tensile strength. American Society for Metals. Metals Park OH. and Impact Strength ofSintered Steels Based on the Sponge Iron Powder and High-Purity Atomized Iron Powder.470 21-2.

0 g/ cm 3 density. "Fatigue Properties of Sintered Nickel Steels. Metals Park OH. Source: A.48 C steels. American Society for Metals. N.1 Quenched and Tempered _ Tensile Strength 105. 4 Ni-0. Ed.Sintered.OOOpsi 10 106 10 7 Cycles to Fuilure . Kravic and D." in Source Book on Powder Metallurgy. Tensile Strength 67. Pasquine.N curve usually shows a distinct fatigue limit. P/M (Nickel Steels): As-Sintered vs Quenched and Tempered for Cycles to Failure 471 40 / r 0.21-3. One of the characteristics of the fatigue behavior of wrought steels is that the S. As-sintered nickel steels possess distinct fatigue limits occurring between 106 and 108 cycles. F. 1979.000 psi As . A typical SoN curve for an as-sintered nickel steel is shown above. L. Samuel Bradbury. and effect of quenching and tempering on tensile and fatigue strength.. SON diagrams representing fatigue behavior of7. P 28 . This is most marked in wrought plain carbon steels and usually occurs between 105 and 107 cycles.

. Pasquine. American Society for Metals. Samuel Bradbury.... Ed. "Fatigue Properties ofSintered Nickel Steels... 2.------. Metals Park OH. 0> " 20 LL 10 - o o 20 40 60 80 100 120 t40 160 Tensile Strength-IOOO PSI Relation between fatigue limit and tensile strength (fatigue ratio) of sintered nickel steels.2 Kt Notch Ouenched-B Tempered Ui 40 (L o o o I ~ E 30 :::J --' . A plot ofthe fatigue ratio (above) indicates an average smooth value of 0. L. alloy content.---.000 psi tensile strength..---.4 which is apparently independent of density level. P 30 ..4 up to 150.----. Thus the averagefatigue ratio for sintered nickel steel is 0. F. Kravic and D.472 21-4. and state of heat treatment and therefore can be used to predict the fatigue behavior of other sintered nickel steels...2 Kt Notch As-Sinlered t:J.---. 2. Source: A. P/M (Nickel Steels): Relation Between Fatigue Limit and Tensile Strength for Sintered Steels 60 r---...." in Source Book on Powder Metallurgy.. LEGEND • Smooth As-Sinfered Smooth Quenched 0 50 a Tempered . 1979.--=__--.-----..

21-5. P1M (Nickel Steels): Effect of Notches on Cycles to Failure for the As-Sintered Condition
50 r--"-'-"'-''"T'"l"'T'T''r---''-'-'''-''"T'''l''T''Mr---r--''T''""T''"T'''''''''TT''I---'''''

473

40

Smooth

2.2 Kt Notched

10

106

7 10

Cycles to Fa.ilure IN.

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

474

21-6. P1M (Nickel Steels): Effect of Notches on Cycles to Failure for the Quenched and Tempered Condition

0-0-'in

o o

o 30
2.2 Kt Notched

a.

0--

If)

+ If)

e

~

20

10

106
Cycles to Failure, N.

7 10

S·N 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

21-7. P1M (Low-Carbon, 1-5%Cu): Effects of Notches and Nitriding on Cycles to Failure

475

40

--.........

Nilrided

l""- I'--.
Nitriderl

'" a..

30

= = ='" '" e
20

r-

to-.

Not nitrided
r--

en

Smooth (K=1)

Notched (K=2)

-

r---..r--

Not nitrided

10 lQ5

I I III

I IIII

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

476

21-8. P1M (Sintered Iron, Low-Carbon, No Copper): Effect of Density and Nitriding on Cycles to Failure

30

Density - 6.4 g per eu em
"iii
D.0 0_
<11<II

<; I I III I I II I III r
r-..... r--...

I II Density - 7.0 g per eu em
""" t--.....

Nitrided

20
Not nilrided

o

en 10

~

Not nitrided

Nilrided

11
10 6

10 8 !OS
eye les to Fa ilure

10 6

SoN curves for P/M parts. As shown above, the greater the density, the higher the fatigue strength of sintered iron powder (low carbon; no copper; notched; K = 2). Nitriding in a salt bath is especially beneficial, it will be noted. Bath temperature was 565°C (1050 OF); nitriding time was two hours.

Source: "Nitriding Improves Fatigue Resistance of PI M Parts." in Source Book on Nitriding, American Society for Metals, Metals Park OH, 1977, P 292

21-9. P/M: Effect of Nitriding on Ductile Iron and Sintered Iron (3%Cu) for Cycles to Failure
60
Ductile Iron
I"":iZ ~
'Vi

477

Sintered Iron

I

~'''':': :::

= = ~ =.
20

a.. 40

::fit k:t

I

Nitrided .:.:.:.:.:.: ;:;:; ::: :: :.:.:.:.:.:. ;:::;:::::; :::;::;:: Not nihided

~'" .....:....

00

.....':':":

~

~ ::::;:

Nilrided :;:;:;:;::::: ::: :~ .........•:.:~:~:~:~:~: ~:~:~ :~:~ ~:

~

::::; :::::'~
00

::::::;:;::: :.:.:.:.:.: ~:::::: ;~~~;~~~~~~ U~

~ ~ ~ mmmm ~~t~r ~~r m= ~

_:

-

m it ~ : mIt ~~tt :~t~ ~~~~
"""" <' t
106

Not nilrided

::: ~~~~~~~~~~~~~ ~~~~~~~~~ ~~~~;~ .':' ~~ ::::::::: t~~ jtm~~ :::::;:;: ~ ~ ~: ~m ~
10 7 108

10 6

10 7

10 8 10 5
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 565°C (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

478

22-1. Brass/Mild Steel Composite: Comparison of Brass-Clad Mild Steel With Brass and Mild Steel for Cycles to Failure
r----------r-------,.---------,15
20

12·5

• Brass clad mild sleel

5

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

22-2. Stainless Steel/Mild Steel Composite: Comparison of Stainless-Clad Mild Steel With Stainless Steel and Mild Steel for Cycles to Failure
...-----------r----------,---------,15
20

479

2 /5
-e

0____ . ~

' Slainl e s s

• staintess clad mild steel steet 0 Mild sleel

12-5

'"

0 o~~-'====t

=
_

10 '1:

~

.e
"l::I
OJ'

0_

7·5 ~

~
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

480

23-1. Carbon and Alloy Steels (Seven Grades): Effects of Nitrocarburizing on Fatigue Strength
~ Normalized

c::J

Normalized and treated in cyanide-base salt bath (treatment 1). 90 mm (3.5 in.), 570°C (1060 of), water quenched

600,------------------------,
~

500 1 - - - - - - -

:2
s: 400

a,

1-------

~

c

300

.~ 200
u,

Cl>

'"

100

o
600
0-

SAE 1000

SAE 1015

SAE 1035

SAE 1045

SAE 1060

:2
c

'"

500

-S' 400

~
u,

'"
Cl>

300

-

::l .g» 200

'"

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

23-2. Carbon and Alloy Steels (Seven Grades): Effects of Tufftriding on Fatigue Characteristics
Tufftrided, 90 min, 1050 F, water-quenched Normalized and tufflrided,

481

80

90 min, 1050 F, water-quenched "" 1---, ~ I------'r---.....:..;'--T''''''-=..:~..:...;.;.'''''''':....:l.::..:..;,;.:..;;;.;.'__, ~ ~
en
"is~

60 J----,---------t
c,

.".

t-----t

+ I---l~-~----t

'"

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

23-3. Carbon and Alloy Steels (Six Grades): Effects of Nitriding on Fatigue Strength

60

After Atmoaphere NitridJng

After Hardening

111111/ I I

55

....
..10:
III

.c

I
I

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

23-4. Carbon-Manganese Steel: Effects of Nickel Coating on Fatigue Strength
Thickness, mils

483

340

o

0.4

0.8

1,2

1.6

320
:2 £ 300
~

\

\

... en
c
Q)

~ '"
~

280

''::;

'"
10

45
.:.l

'iii

-,

J::."

-,

1» c
1;;
Q)

e
::s

u,

'"
260

<,

r-, .....
~
40

40

u.

'fJ

en

240

o

20

3D

35 50

Thickness, /.1m Effect of coating thickness on the fatigue strength of a carbonmanganese steel.

The reduction in fatigue strength produced by electro less nickel deposits is affected by the thickness of the coating. Thicker deposits have the greatest effect on fatigue strength, This is illustrated in the above graph, which shows the reduction in strength of a carbon-manganese steel (Werkstoff St52) produced by different thicknesses of a 5% boron-nickel.

Source: Metals Handbook, 9th Edition, Volume 5, Surface Cleaning, Finishing, and Coating, American Society for Metals, Metals Park OH, 1982, P 232

the unbroken springs were again checked for set and recorded. with one tum squared on each end.2 4.0 7.0 7.15 0..875 0. the plotted points represent fatigue limits for the respective minimum stresses used. had undergone about 2. Source: Metals Handbook.07 4.22 22.22 2.5 7.875 10.81 1. The test load was applied statically to each spring and a check made for set three times before fatigue testing.625 0.MPa Spring No. after which the ends were ground perpendicular to the spring axis. The tensile strengths of the wires were according to ASTM A228.032 0.".20 57. P 293 . Music Wire (Six Sizes): Data Presented Means of a Goodman Diagram by 40 200 1-----. Wire diam mm in. Total turns Active tums Total tested 1 2 3 4 5 6 0.22 0.• By means of the Goodman diagram many fatigue-limit test results can be shown on the same diagram as indicated above.%set after 10 million strese cycl. then baked at 260'C (500 'F) for 1 h.52 6.0 7.7 16 28 38 43 35 25 Data are average fatigue Iimita from S-Ncurv. Metals Park OH. American Society for Metals. line OM represents the minimum stress for the cycle. After testing.Allstresees were corrected for curvature using the Wahl correction factor.5 4.8 12.9 22.7 5. Spring on mm in.121 0.. Some scatter may be expected.22 22.10 26.250 0.37 2.2 5.875 0.6 6.50 0. Line UT is usually drawn to intersect line OM at the average ultimate shear strength of the various sizes of wire.88 22.45 60. The vertical distances between these points and the minimum stress reference line represent the stress ranges.25 6.062 1.032 0. The springs were automatically coiled.375 0. Properties and Selection: Irons and Steels. tested at 1070 MPa (166 ksi) max strese.2 5.87 1.2 5.7 6. Coil Springs.59 3. but the stresees were not recalculated to take this into account. Number 4 springs.35 15.25 2. 1978. 9th Edition. Spring index Free length mm in.97 44.(---+----+----+----+--_= 20 o Minimum stress.0 7.048 0. In this diagram.484 24-1.15 57.75 2..2 5.81 0. at least partly attributed to normal changes of tensile strength with wire diameter. Volume I.102 0..177 9.0 7. for 185 unpeened springs of various wire diameters run to 10 million cycles oCetrese. None of the other springs showed appreciable set. The springs were all tested in groups of six on the same fatigue testing machine at ten cycles per second.

.. Metals Park OH. p 292 . corrected by the Wahl factor. Minimum stress was zero..-=-::l-::--------+f 00 ---l150 125 ~ * 750t---------+----.205 in. Properties and Selection: Irons and Steels...59 mm (0.. The upper graph is a typical S...... .. (b) Life of springs used in a hydraulic transmission. The lower graph shows an alternate method of presenting fatigue data for steel springs....___==! 105 106 Number of cycles to failure la) Type of wire Number of springs Oil tempered 6 ~ r Avg Music wire 8 ~ 104 105 Number of cycles to failure (b) (a) Springs were made of minimum quality music wire 0.. ...::::!!oo""""'::-l----------l 500L ....--104 .. Source: Metals Handbook..) in diameter. Wire diameter was 4.21 mm (0... American Society for Metals....32....'-.75 mm (0...187 in. 200 1250t---------+--------+------------j ~ ~ 485 175 ~ ::i 1000f-------"".. Coil Springs: S-N Data for Oil-Tempered and Music Wire Grades 1 5 0 0 ......750 ln... outside diameter of spring was 44.. 9th Edition.--'-::-.022 in.24-2. They were made of oil-tempered wire (ASTM A229) and music wire (A228).. Spring diameter was 5. Volume I.). Stresses corrected by Wahl factor.45 mm (1.D/dwas 8. 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 springs were fatigue tested in a fixture at a stress of 605 MPa (88 ksi). 1978..)..--r-r-.-.N diagram showing results of compression testing coil springs.. with 15 active coils in each spring.).

The kind of shot used is important. Proper peening intensity is an important factor. peening too deeply leaves little material in residual tension in the core. However. angular particles. An Almen test strip necessarily receives the same exposure as the outside of the spring. for springs with closely spaced coils. Volume I. A temperature of 230°C (450 OF) is common because higher temperatures degrade or eliminate the improvement in fatigue strength. it has been found that excessive roughening during peening with coarse shot lessens the benefits of peening. P 297 . Thus. which requires internal tensile stress to balance the surface compression. Coil Springs: Effects of Shot Peening on Cycles to Failure 100 90 80 t e " <: 70 s 111 'in <: ~ ° 60 50 ~ ~~ r-. is shown in the above diagram. a coverage of 400% on the outside may be required to achieve 90% coverage on the inside. and slightly smaller wire using special techniques. Shot peening is effective in largely overcoming the stress-raising effects of shallow pits and seams.. but more important is the need for both the inside and outside surfaces of the spring to be thoroughly covered. Shot peening can be applied to wire 1. Properties and Selection: Irons and Steels. 1978. 9th Edition. better results are obtained with carefully graded shot having only a few broken. <. this negates the beneficial effect of peening. but to reach the inside.~tressed I- /peened in bending 40 ~ ""-Not peened /Peened - ----- ~Stressed '\:"Not peened in torsion Number of cycles to failure Shot peening is often used to improve fatigue strength of springs by prestressing the surface in compression as indicated in the chart above.486 24-3. Metals Park OH. Also. for larger wire. Cold wound steel springs normally are stress relieved after peening to restore the yield point.. the shot must pass between the coils and is thereby much restricted. apparently by causing minute fissures.6 mm (1/16 in. American Society for Metals. Source: Metals Handbook.) or more in diameter. Shot size may be optimum at roughly 20% of the wire diameter. according to one prominent manufacturer of cold wound springs. The extent of improvement in fatigue strength to be gained by shot peening.

tr--o r--.---.i1u.. Springs were made from 15.24-4. is made quite apparent. 1978. Such data on springs hot wound from bars with as-rolled surfaces are limited.. l"- . 400 "" f'--. American Society for Metals..0-mm (% to 1-1/16-in.. I'-. 8650 and 8660 Steels: Relation of Design Stresses and Probability of Failure 800 487 9J". 90 500 r-. .b!lIty 011. The value of peening.fl--Not peened /Peened sn BO ~- • Peened 10 60 400 I:: 50 106 Number of cycles to failure BO0 50% p."u. <. p 304 . and interpretation is therefore difficult. Springs were shot peened to an average arc height of 0.. o' il 0 106 Number01 cycles to failure Design stresses.. p'~b'blllty of f. The desirability of conservative design in cyclical service is illustrated in the three charts above.i1~..l 100 600 0"" 110 500 .. Coil Springs.r----106 (0 r-- BO ~ 10 60 50 I- Number of cycles failure BOO J". Pertinent test data are given above.0 <..) diam 8650 and 8660 hot rolled steel and heat treated to between 429 and 444 HB.e I 10 100 100 600 ". 400 r---. Properties and Selection: Irons and Steels. 0 10 60 60 ..bility 01 f.. Source: Metals Handbook.~ <.. 500 o Not peened r-. Metals Park OH. 100 90 :ii 80 0 .. p. however..9 to 27.r-rr---..oti. on the type C almen strip at 90% visual coverage.lb... ~~ . 9th Edition. 110 100 100 600 -..008 in... in which the minimum stress used was low... Volume I. .

\.09 1.79 0.58 0. "Precipitation Strengthened Spring Steel for Automotive Suspensions. there appears a remarkable difference between SUP7 and SUP7-Nb-V.15 o 0.>: :r: (/) p:: 0 •.56 X 104 s (21 h)."in HSLA Steels-Technology & Applications. the result was rather different. Toshio Ozone and Mamoru Kurimoto.14 0. After 10and 20 cycles ofthe corroding.71 .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. 490 - 0&ocj- .56 2.025 0.57 r stress: 490 ± 340 t~Pa (71 ± 50 ksi) 52 U -o ::s ~ 440 ~ . the coil springs were loaded with the surface stress of 490 ± 340 MPa (71 ± 50 ksi). Source: Toshiro Yamamoto. in which SUP7-Nb-V has comparable fatigue life to that of SUP7 in any stress amplitude. The corroding condition was as follows: an exposure in a chamber filled with saltwater mist for 1.94 0.. the result was as shown above (left).08 X 104 s (3 h) and a keeping in the atmosphere for 7.64 . •.014 0. Metals Park OH.09 0. The fatigue life of the coil springs subjected to 20 cycles of the corroding are shown above (right). P 1022 . American Society for Metals.-l (/) . Ryohei Kobayashi. 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. on the other hand. . Measurement ofthe surface corrosion depth of the two steels showed no difference. \ . This time.008 0. 390 o SUP7 _ SUP7-Nb-V I C I r \ \ I C 'tl ~ 50 H +J (J) 104 10 5 Cycles to Failure 106 :r: H 111 48 0=.488 24-5. When the coil springs were corroded.78 . HSLA Steels: Effects of Corrosion on Cycles to Failure ~ 540 :<: Mean stress: 637 MPa (92 ksi) e:tJ- . • \ • \ 0 \ 0 ~ (/) (/) Q) e 0.021 o 0.008 0. different from the case in the graph at left.. Compositions of HSLA Springs Tested C Si Mn P S Cu Cr AI Nb v SUP7 SUP7-Nb-V " 0. \ .014 0.83 0.09 0.09 0.. Coil Springs. When the coil springs were free of corrosion. 1984.

..... ------. " -. " -69 . . Leaf Springs. b----_ a '. This stress-free ASR diagram will be used to predict the endurance limit for the other specimens containing peak stress values of . 106 345/5 • _ .10)... and curve f.1070 (.•• +400 104 105 107 CYCLES Residual stress and unidirectional bending fatigue data for strain-peened SAE 5160 steel.138). .. curve b.I55). Ed.. >-<<1l .. Sharma and D.-. Larry J. ~ ~ u PEAK RESIDUAL STRESS. +0. Source: V.. • « .. . Vande Walle. -950 (..30%. curve c. . . . -69 (. then the S-N curves (see above) were obtained from the same samples by testing in unidirectional bending.. 5160 Steel: Maximum Applied Stress vs Cycles to Failure 2070/ 300 250 1380/200 150 III 489 U) U) ILl E-< lX: U) '.24-6."'.. Breen.. Metals Park OH. ..60%. In this illustration. preset only.d '.. -0.-1 Q::<: ILl <. MPa -1070 -950 -565 -207 " . the endurance limit corresponding to the specimen strain-peened to produce a residual stress of -565 MPa (-82 ksi) will be used to develop a stress-free ASR diagram for 5160 spring steel (48 HRC). 1981. Leaf spring specimens of SAE 5I60 steel quenched and tempered to 48 HRC were shot peened under various conditions of applied strain to introduce a wide range of residual stress.... . K. ..:Il:lo l:lo. curve e..60%.: c. H.. p 82 . 0% (conventional peening). The predicted endurance limit will be compared with the values determined experimentally. -207 (-30). and +400 MPa (+58 ksi). -----. +0. -0. curve d. ...30%. "Some Aspects of Incorporating Residual Stresses in Gear Design. ." in Residual Stress for Designers and Metallurgists.. American Society for Metals.. Applied strain during peening curve a..-.:. f ••c . ~ >< 690/100 .

Mean service life.25 in. Source: Metals Handbook.490 24-7. Size of hexagonal bar section was 32 mm (1. Properties and Selection: Irons and Steels._ _ _ _ l E Z Service life... 1000 cycles Here are results from simulated service fatigue tests offront suspension torsion bar springs of 5160H steel.HV A /l-V.1' -V/WA-f/. 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.. coefficient of variations.-+-r. P 677 . American Society for Metals. 134.. 25 lots...l1-V'A-t/'J--Y... 300 parts '" ~ t: 30r--------~A_r/hr7'T_V. Front Suspension Torsion Bar Springs.8 :l '0 20 101------r..... 9th Edition.. 1978.. Metals Park OH.. Volume I.".. 5160H Steel: Distribution of Fatigue Results for Simulated Service Testing 40.000 cycles.. 0./lH/.1'A .).-------------------------------. 37.000 cycles..28... standard deviation.l+_-------------____l .

0 4.0 3. H....J 1.. P 127 . Gears. Q) '" u . Ed. Carburized Low-Carbon Steel: Relation of Life Factor to Required Life 5. American Society for Metals." in Source Book on Gear Design.3 u. Technology and Performance. Howes. Maurice A..24-8.0 0. For a single mesh the number of revolutions and the number of cycles are equal.0 491 II ~ u-' 2. For a gear which has more than one mating member.5 10 4 106 Required Life in Cycles 10 8 The life factor depends on the required life in cycles.. Metals Park OH. Source: "Bending and Contact Stresses in Hypoid Gear Teeth.0 .. . the life must be equal to the required number of revolutions multiplied by the number of mating gears. 1980.

0.45/. Source: V.. 0. Ed.. 6 . ~ '~ . "Some Aspects of Incorporating Residual Stresses in Gear Design.. P 86 . -276 (40)....... 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).. 1981.. '-' 3. H. 0::: (/) E->< ." in Residual Stress for Designers and Metallurgists. 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. American Society for Metals. and -690 MPa (100 ksi) residual stresses. Metals Park OH.492 24-9.... Breen. Vande Walle..... Larry J. K... Sharma and D. Gears. 0 0 . Carburized Low-Carbon Steel: Bending Stress vs Cycles to Failure (/)lI'l (/)0 LLl .. 90/ 1...l a:l::'<: LLlo.5 CYCLES Bending fatigue design curves for carburized gears having different amounts of circumferential residual stress at the root-fillet surface... Zc...

45/ • 5L..1 Q « :::> .-:. The L50 design curves for as-carburized.. Carburized Low-Carbon Steel: Effect of Shot Peening on Cycles to Failure 493 L50 L50 L50 3.- ----' 105 CYCLES '" . Ed... indicating a more significant effect of residual stress at higher life cycles.. The lower chart shows the residual stress factors calculated from the S-N curves in the upper chart.- ----'----.2 « .-L...L-:....... American Society for Metals. Ka Based on Dynamometer Tests The dynamometer test data obtained on testing sets of gears with different magnitudes of residual stresses can be used to develop the S.----'----...--. u . Gears. The S-N curve for shot peened gears is derived from the results published by Alman and Black... Vande Walle. Ka for shot cleaned gears equals one. 1981.-----. shot cleaned. and shot peened gears are shown in the upper chart." in Residual Stress for Designers and Metallurgists. Bottom: Residual stress factor computed from the upper chart.... H. K a for carburized shot cleaned gears equals 1. The data for as-carburized and shot cleaned gears were obtained on testing six-pitch test pinions on a Four Square Dynamometer... '" Vl w Vl Vl 1. Sharma and D. that is. According to these results the effective bending stress for shot cleaned gears at 108 cycles is approximately 20% higher than as-carburized gears and approximately 15% lower than shot peened gears.. 78 .. which means the residual stress has almost no influence on the fatigue properties of a material at high loads. the allowable SoN curve for carburized shot cleaned gears is used for design purposes.24-10.N curves necessary to calculate the Ka factor.. that is.::-- '-... pp 77. shot cleaned. At low cycles. Breen.. 0 1.0 e e SHOT CLEANED e e '" w Vl 0. The value of Ka deviates from unity with increasing cycles.. It is assumed that the S-N curve for shot cleaned gears is used for design purposes. K. This is obviously because ofthe stress relaxation caused by the cyclic plastic deformation accompanying low-cycle fatigue.J 1.L.9 ~ 10 10 o. and shot peened carburized gears... Metals Park OH. 8 '--. Larry J. the residual stress factor seems to approach one. "Some Aspects of Incorporating Residual Stresses in Gear Design. Source: V.---I 4 5 7 6 10 10 10 8 CYCLES Top: Allowable (LSO) bending stress design curves for as-carburized.

The allowable stress-life diagram characteristic for each material..I. t:lp.80/2.l:'<: !Xl '-' z . Carburized Low-Carbon Steel: Probability-Stress-Life Design Curves U'l o U) . "Some Aspects of Incorporating Residual Stresses in Gear Design. Larry J.0 L10 3... Metals Park OH.5 10 . Ed.. 1981. The applied root-fillet bending stress is then calculated to predict the gear life from the stress-life design curves such as those shown above. l. Ul Jl..- --' 10 8 CYCLES Probability-stress-life design curves for shot cleaned carburized steel gears having a root-fillet surface finish of 5Jl in. U) 'M c:<: Jl.... a designer usually begins with a preliminary selection of the tooth widths and other design parameters based on past practices and empirical approaches recommended by AGMA...... KJl = 1...35/1. U) 13...90/1.lX L90 L50 z . H. 5 L.. p 74 .. American Society for Metals. Breen..... The procedure is reiterated to optimize the design so that the calculated life is just equal to the required life with an appropriate level of safety...." in Residual Stress for Designers and Metallurgists... '" 6.. Vande Walle. heat treatment. 45/ ..5 E...e. Gears.1.494 24-11. Bending Stresses in Gears In designing gears for a new application.0 10.- ... K.. 00... and surface treatment is usually obtained on testing acceptable commercial quality gears on a dynamometer.... Sharma and D. .. Source: V.

H. 8620H Carburized: Bending or Contact Stress vs Cycles to Fracture or Pitting 500 BENDING (St) 495 400 300 200 G5~ OR CONTACT (Se) GlO~ G90 ~~ ~}s." in Source Book on Gear Design. Metals Park OH. Breen. P 66 . Source. D. Gears. STRESS. Technology and Performance.. Ed. Howes. KSI _-}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. H.24-12. "Fundamental Aspects of Gear Strength Requirements. Maurice A. 1980. American Society for Metals.

Maurice A. Obviously. H.. Howes. Although the experimental steel had a significantly lower case hardenability. "Computer Oriented Gear Steel Design Procedure.400 10 5 Cycles Weibull analysis of bending fatigue data from gear tests indicates that gears made from either the experimental CH steel or 8620H have equivalent durability. American Society for Metals.c c:: . case hardenability.J . 10 6 Metallurgical data gathered on these gears established the adequacy of the experimental steel (a CRAT steel-computer harmonizing by application tailoring). thus representing an efficient use of alloy hardenability in CRAT steels.. Walter.5.800 4. P 85 .600 . 1980. Gears. though not excessive. Technology and Performance. 8620H Carburized: A Weibull Analysis of Bending Fatigue Data 5.. it quenched out to a 100% martensite plus austenitic structure at the root-fillet surface.200 Q> 1-0 eo c: ::I 4. Source: G. Metals Park OH.496 24-13. it had adequate.. Ed. c: 0. H." in Source Book on Gear Design.

0 '" 5 . H.. H..... .....FATIGUE C a I- CI:: 3 1. 1980. pitting and spalling for a six-pitch pinion test." in Source Book on Gear Design...G90 '-.. 4 c: ::::l BENDING FRACTURE t w ~1~~----MIXED .... ... Gears. Metals Park OH..... Breen..24-14... ...... P 66 . Ed. Technology and Performance. American Society for Metals.. The above fatigue data show torque versus cycles to breakage...: G50 GIO -... Four-Square Tests 8 7 M 497 0 )( 6 G90 G50 GIO ---- . PITTING SPAlLlNG : : : : .... '-. four-square gear tests. Maurice A. Source: D. '-. "Fundamental Aspects of Gear Strength Requirements. 2L-105 -L. Howes. 8620H Carburized: T-N Curves for Six-Pinion..l .. Note that there is a mixed area where failure can occur from anyone or a mixture of the three modes. .. ----I CYCLES T-N curves for carburized six-pitch pinion.

or ....: . fill.. " ....498 24-15.. Maurice A. .. ~... .!!!.. .... <...... .. Howes... 1980.. " <.J ai . . .. American Society for Metals. Sloping line indicates minimum confidence level.. . 1-.. M ~ <. .... Ed." in Source Book on Gear Design.. .. 105 CYCLES FOR 10 5 RUPTURE . .• : a Iii 0: . 8620H Carburized: Minimum Confidence Level.. Hypoid Gears.. ..........·--_ev .. . Technology and Performance....(N) 107 Fatigue life data for hypoid gears.. Stress vs Cycles to Rupture i C1 Ul 105 0: ... Q.. Source: "Gleason Method for Estimating the Fatigue Life of Bevel Gears and Hypoid Gears.. ... P 386 .. <. H... Metals Park OH. .

.... 8620H Carburized: S-N Scatter Band and Minimum Confidence Level 499 -------.... American Society for Metals.... 1980..... .. . 060 o 0 co <." in Source Book on Gear Design. . ~ ....0--.' . ........ ....... " n n no a::: w ll) l- 5 10 106 CYCLES FOR RUPTURE .. .....-..(N) Fatigue data-composite for results obtained by testing various gear designs. ~<QI o <... Technology and Performance.. . ----... Source: "Gleason Method for Estimating the Fatigue Life of Bevel Gears and Hypoid Gears. -0 ~ :--. P 386 ....... .. • .-....ll. <... . ·8·~ 'l...a"" .. --0 00 0. H... Hypoid... ... Howes.... PROPOSED DESIGN LIMIT(NOT OVER 5 PER CENT FAILURES) 1 lI) lI) I~ . Maurice A.. : 0 0 <. ... .. -..----.24-16.. ---..1. ~~~ <. ... Zero I and Spiral Bevel Gears..... ... Metals Park OH... . Ed...... .... ~"'-"'. ---.... ...

Technology and Performance.... .... ... . .YCLE:S fOR RUPTURE -(N) Graph offatigue life for spiral bevel and Zerol bevel gears... .... 1980.. 1'--. . -..: . III d 10 5 1'---. Ed.. .' <. Sloping line indicates the minimum confidence level.. .....~" ~ a.. . American Society for Metals... . Metals Park OH. . .. .. . 8620H Carburized: S-N Scatter Band and Minimum Confidence Level ... -----.. .".. ~ ........." in Source Book on Gear Design...... .. ... P 385 . Maurice A. ... .... . Howes. . 10 5 10& 107 C..... . .. .J ~ I :l a: IIII III .. <. w CL <... . .. ..500 24-17. .. Spiral Bevel and Zerol Bevel Gears.. .. « ~ i.. <.... g.. .-.. Source: "Gleason Method for Estimating the Fatigue Life of Bevel Gears and Hypoid Gears.. ai . .- a: <.. H..

. Gl ..:i . Maurice A. American Society for Metals.. :. 1980.24-18... Howes.. Source: "Bending and Contact Stresses in Hypoid Gear Teeth.. The life factor for strength may be obtained from the above data. 95% 1 0. Metals Park OH.... H. p 127 ..: -0 U 0 0. Gl . 8620H Case Hardened: Relation of Life Factor to Cycles to Rupture 5 4 501 ~ ~ --~ ~ 3 ~ • 5% <i > u ." in Source Book on Gear Design..... Ed.. Gears..9 0. n 0 I 5 2 50% 1'000.6 10J 105 106 Cycles for Rupture Both strength and durability are fatigue phenomena and therefore display a relationship between stress and life..8 0.. Technology and Performance.7 c.... Gl C Gl C .

k 2.. n po". Ed. 0 u. k). .. 95% I I 1 0. Low-Carbon Steel Case Hardened: Relation of Life Factor to Cycles to Rupture for Various Confidence Levels - 3 ~ . Howes.68 + . .k) k.68 +k)npJ (T ) f. . For a single mesh the number of revolutions and the number of cycles are the same..8 )] 0. Maurice A. n P) n PI/= pinion rpm corresponding to torque loads T" T 2... L H = required total life in hours. . p 149 . Metals Park OH.. One suggested method is as follows: L cp=60L H T~ [k lnpl+ k n n ( T) 2 5..000. T"respectively.. ::. I l> 2 5% .6 103 . 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.. 50% I> . Source: "Bending Stresses in Bevel Gear Teeth. 1980. Bevel Gears. +kl/npl/ ( T) i... T) . 5.=proportion of time at torque loads T" Tb T) . the designer may wish to determine the equivalent life at maximum torque. T 2.9 0. T 1.. .. . T) ." in Source Book on Gear Design. n PI' n n. T" = torque loads where T I is maximum torque and T" is minimum torque which will produce a stress above the endurance limit. This depends upon the required life in cycles. Technology and Performance.. . When the required life is less than 6. the life factors will be different on gear and mating pinion.502 5 4 24-19. I ~ I I I l 1O~ 106 Cycles for Rupture 10 7 108 The life factor is obtained from the graph above. 5.68 ] where L cp = required equivalent life in pinion cycles at maximum torque.. In cases where the load varies. American Society for Metals..000 cycles on the pinion.. TI/ respectively. The required equivalent life in gear cycles at maximum torque may be obtained by multiplying the life in pinion cycles by the gear ratio: L CG=L cp NG n. H...

AMS 6265: S-N Data for Cut vs Forged 35 .J E :::I ::..: E '.- 30 25 20 15 10 5 lQ3 . 1979. Loads shown are applied actuator loads. Source: "How Gearmaking Methods Compare.8 "C .9 ll... P 347 . Samuel Bradbury..." in Source Book on Powder Metallurgy. 6 CuI gears • Forged gears "" CL CL ..J 503 = = =".( co 104 10 5 Cycles 10 Failure 10 6 1Q7 Fatigue data shown in this chart proved that teeth on precision forged AMS 6265 helicopter pinions have a higher fatigue limit than cut teeth. Metals Park OH.24-20. .. Gears...6. Tooth loads are approximately 33% greater..c .~ <I: "C co 0 .. Ed" American Society for Metals..

.. 0 0 0_ 90 80 "0 0 0::: lJ) lJ) en 60 'j§ 0 co 70 e 1% failure m c z: 50 40 30 .504 24-21. P 346 . Metals Park OH. 1979. Spur Gears. Source: "How Gearmaking Methods Compare." in Source Book on Powder Metallurgy. As shown above. a.Cui gears Forged gears 10 5 Cycles 10 Failure S-N curves for spur gears forged and cut from 8620H steel.. 8620H: S-N Data for Cut vs Forged 120 110 100 "u. Ed. results of beam fatigue tests indicate that precision flow-forged gear teeth are about 20% higher in fatigue strength than cut teeth. Samuel Bradbury.. American Society for Metals.

0 . American Society for Metals.2 .0 80.0 a: w Q.2 .. "Fatigue Behavior of Hot Formed Powder Differential Pinions." in Source Book on Powder Metallurgy. a: w Q. Bottom: Similar to graph at top exceptfor alloy 2000 and 4615bar stock.0 ¢ Z 3 W 300 20. . C. 1980. ::> . ~ 40.. Gears and Pinions: P/M 4600Vvs4615.0 1. Eloff and L.) .95. 6 7 8 3 4600Y 4600 V 4600V 4600V 4615 2~F I MIN SINTERING TEMP TIME 2100-F 3MIN 2350·F 3MIN 2350·F 6MIN 2100-F 6 MIN (BAR STOCK) / / / / / / / / / / w a: u.0 400 300 4 5.) Top: Weibull distribution charts for fatigue testing of actual gears and pinions made from 4600V alloy with various sintering times and temperatures as shown.J / 3 --( / ¢ / 20._ _5.24-22.1 .3 / / / / / / I / / / / TIME (HR. Source: P.0 60. 0 10.0 80 60 40 2.8 1. p 308 . Wilcox. 0 z w ~ '00 8 60 40 20 / 1.0 80.0 . 505 ALLOY 4600V 99.0 34~6810 TIME (HR. Weibull Distributions LEGEND ~. E. Samuel Bradbury. Ed. 2 3 2000 2000 4615 !!::ill TEMP TIME 2350·F "'6MiN 2350·F 3 MIN (BAR STOCK) SINTERING w a: u. compared with cut pinions (4615 bar stock)..0 600 LEGEND ~.1 . Metals Park OH.

1 2 3 4 5 6 8 10 TIME (HR. Gears and Pinions: P/M Grades 4600V and 2000 vs 4615.0 <l: IZ 0 w 0:: W (L 10.0 60. which also graphically indicates the 95%confidence limits on the BID lives. The results are shown in the graph above.. 950 80. it was decided to fit one Weibull curve to all of the data points from each alloy.. nickel and molybdenum. P 313 .) Fatigue data for actual gears tested in specially designed machines.0 ALLOY 4600V 2000 4615 8AR STOCK 95'/." in Source Book on Powder Metallurgy. American Society for Metals. Ed. Samuel Bradbury. It isplain that the 4600V pinions have superior fatigue life at the stress level of 92.0 . Since the data from the two powder alloys fell into two groups.0 4.0 'i w 0:: --l lJ. This was done to obtain more data points for each curve.506 24-23.9 LEGEND NO. are readily reducible by CO at temperatures even below 1150 °C (2100 OF). Wilcox.400 psi.0 1. 1980.0 20.0 6.. since the major alloying constituents. C.0 2. Percent Failure vs Time 99. "Fatigue Behavior of Hot Formed Powder Differential Pinions. Eloff and L.0 8. Presented here are Weibull distributions for the three types of alloys tested. In the case ofthe 4600V alloy. CONFIDENCE LIMITS / / / t / 2 3 1-1 / / 3 40. ~ 30. E. and the slope ofthe Weibull curve indicates uniform deoxidation of preforms and therefore less scatter (steeper Weibull slope) in the fatigue data. the sintering temperature should have little effect on deoxidation. Metals Park OH. Source: P.

Source: N. In the weld evaluations made. Ed. F. Gear Steel AMS 6265: Parent Metal vs Electron Beam Welded 2 4 68l 507 100 2 4 681 2 4 681 2 4 681 I 900 10 10 i~ 50' SON diagram for AMS 6265-parent metal versus electron beam welded. The welded specimens failed in the weld zone at 86% joint efficiency. Bratkovich. Other gear materials tested resulted in compara ble weld joint efficiencies. Schwartz. American Society for Metals..24-24. it was demonstrated that electron beam mechanical properties were comparable or better than welds made with other fusion welding processes such as gas tungsten arc and metallic arc welding. 1981. Metals Park OH. Mcintire and Robert E. W. excellent mechanical properties were found. "Electron Beam Welding-Applications and Design Considerations for Aircraft Turbine Engine Gears. Melvin M. L. Purdy. P 199 . In general.t' in Source Book on Electron Beam and Laser Welding.

)C G -J-L LW--+ SoN curves for various tooth profiles (50% survival probability in the short life range).. ~ . .i i\: .>..... --. H. .I. Howes. 'lG r\(W/m 1....10 4 J 0. .. .~ . Gears. I • ~ O. P 102 ..-.. Technology and Performance."!...."in Source Book on Gear Design...fl (I ~ (.O.. (I ....508 24-25.:.. . :~~. "The Measurement of Actual Strains at Gear Teeth.- ll. Hirt..-. .~o2 20 -.... --. .. . 3. Metals Park OH.. Maurice A. 1980. 42 CrMo4 (German Specification): S-N Curves for Various Profiles 80 z-...:' ~~~ ~~. The horizontal sections of the curves are based on the highest load that can be carried for a minimum life of five million cycles... I~ ... Winter and M.::--.... i 30 ~:~ ! .. :.. Influence of Fillet Radius on Stresses and Tooth Strength._..:::~ ".~ ~~ " . Ed . The Woehler curves shown above are based (in the sloping section) on a survival probability of 50% at the number of cycles indicated. American Society for Metals. Source: H. >\0 z 40 ~.l.

Ed. Also.I: I I I II' II ~ I . L so. American Society for Metals. life values L IO . Influence of Fillet Radius on Stresses and Tooth Strength. 6 4 ----f I . . A more adequate approximation by the theoretical distribution is achieved by a three-parameter Weibull distribution. Winter and M. H.1 --~ Endurance test results in the Weibull distribution diagram.."in Source Book on Gear Design. P 102 . 5 (. This formulation produces a minimum endurance Lo.which represents in such a probability grid the test results for gears of one tooth form. A -1·- l[ -. - 3. From this curve we are able to read. 50 and 10 percent survival probability.._.. or 90.. and LJo for 10.70 Of 50 il 71' J 30 If 20 10 8 J~!'I L . An example is shown in the graph above..--.'0:> 2 3 t.W= if (n + I) for test i out of n test results sorted to the number ofload cycles at which fracture occurred. Scale of ordinates is the failure probability A = 1. the above chart shows the compensating curve which results from the formulation of the three-parameter Weibull distribution.24-26.0'< 4 6 8 1. The test points are approximated with a straight line.. Gears. Metals Park OH. Technology and Performance. 1980..A are adjusted to the test points. Source: H. The parameters ')I. ~ . .-. .. 42 CrMo4 (German Specification): Endurance Test Results in the Wei bull Distribution Diagram 95 509 . "The Measurement of Actual Strains at Gear Teeth.-W 8 1. Howes. which is reached by all test pieces. .. f/ I II if f . Hirt. . respectively.1-I ~ . and 90 percent failure probability..50. Maurice A.

'as long as other factors such as mean stress. Hingwe. Bolts. which establishes a notch pattern inherent in the part because of its design. K. For the 4037 steel: yield strength. :::iE . The principal design feature of a bolt is the threaded section.55% carbon) are unsuitable because they are notch sensitive.. and surface condition are the same. "'". The form of the threads. If the results of fatigue tests on standard test specimens were interpreted literally.. tensile strength (axial). )( 650 • - 95 "v. if bolts made of two different steels have equivalent hardnesses throughout identical sections. ~ 625 0.0 )( ~ 600 . 1250 MPa (182 ksi). Source: ASM Committee on Carbon and Alloy Steels. P 206 . 1250 MPa (182 ksi). 16 threads to the inch) had a hardness of 35 HRC. 1982. 1040 and 4037 Steels: Maximum Bending Stress vs Number of Stress Cycles 100 0 • 0 1040 steel I - 100 • 4037 steel 615 :::iE s: 0 ~ g' :0 .0 ~ - ~ . "Threaded Steel Fasteners. tensile strength (wedge). tensile strength (axial). steels of high carbon content (more than 0. stress range. Metals Park OH. high-carbon steels would be selected for bolts. 1060 MPa (154 ksi). Ed .510 24-27. Tensile properties of the 1040 steel at three-thread exposure were: yield strength.N data). • • 0 ~- ~~- - 85 515 o 0 >--1- Number of stress cycles The bolts (7'8 by 2 in. 1110 MPa (161 ksi).n . c 90 :0 c Cl . of> .. their fatigue strengths will be similar (see above S. In general. tensile strength (wedge). is much more important than steel composition in determining the fatigue resistance of a particular lot of bolts. A." in Quality Control Source Book. plus any mechanical or metallurgical condition that also creates a surface notch. Actually. American Society for Metals. 1200 MPa (175 ksi). ~o. 1190 MPa (173 ksi).

..6.. K...t .t ... Lower graph represents five different lots that were heat treated to average hardnesses of 23.. Other factors being equal.. 1982. .j 50 3001---t-----1----t-----j 40 225 30 U5 150 f-----11- ~ 20 751----t------I---' 10 o L- ! -_ _-'---." in Quality Control Source Book.j 20 Roll threaded after heat treatment ~ 75 1---+---+---+---::::1 10 Number of stress cycles S-N curves showing fatigue limits for roll-threaded bolts..t ... Ed.. Source: ASM Committee on Carbon and Alloy Steels. "Threaded Steel Fasteners. 31. Hingwe. then roll threaded..7.. This is true for any strength category..... Bolts having higher hardnesses in each category had higher fatigue strengths...7.. The cold work of rolling increases the strength at the weakest section (the thread root) and imparts residual compressive stresses...6. 26. .. Upper graph represents four different lots of bolts that were roll threaded. similar to those imparted by shot peening..... a bolt with threads properly rolled after heat treatment-that is. . 27.. 27. ... P 202 .. and 33.t ..0 HRC.. Bolts: S-N Data for Roll Threading Before and After Heat Treatment 450 r . then heat treated to average hardness of 22.... . and 32......- 1" U5 150 f .6 HRC. A..4..+ ..... Metals Park OH.6. American Society for Metals._ _ ~_~ 104 450 . 29... ....: ::i 225 I ...1 .. .24-28.. free from mechanical imperfections-has a higher fatigue limit than one with cut threads..3.t . Roll threaded before heat treatment 511 60 375 1 . 60 375 300 40 0- 50 '" :.

~~ S/N-2 :--- --. Source: S. Silverstein. while in the brazed units all failures occurred in the brazed joints.l I- i---- > 10 r---r--.• .i m 0: >0: o 20 - -- r-. 1981..512 24-29. Strautman and V.. V. Melvin M. S/N-3 40 z +1 w 5 0: 30 I- o !. "- . ELECTRON BEAM WELDED JOINTS • 60 SILVER BRAZED JOINTS TEST TEMPERATURE. p 187 . Ed. Freeman... "Application of Electron Beam Welding to Rotating Gas Turbine Components. Power Shafts. 500-600 oF Q o o x III 50 o o 0.. M. . AMS 6382 and AMS 6260: Electron Beam Welded vs Silver Brazed Joints .. R. SIN-I . Metals Park OH. ---- SIN-I ROOM TEMP. u :I: Z :l S/N-2 . failures occurred apart from the weld. American Society for Metals..." in Source Book on Electron Beam and Laser Welding. -I--- 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.. Schwartz.

1979. composition.TI-JROllGH HARDE\ED 50B46 n 2 f:S i/._.... .40% carbon steel axle shafts were developed to replace through hardened alloy steel shafts for both product." .. .:..'~ .. B & C ~on if.... 513 '"" if. along with several experimental induction hardened shafts. ~ u:: ] on 50 \'.. Metals Park OH.: ~ APPLIED ~ STRESS GRWIE~T 200 -.. The upper graph shows the stress gradient in the body area when the shaft is loaded to 110.and cost-improvement purposes...... At the spline.. ~ 60 . ".'·-... . Wene. The stresses are a maximum at the surface and drop linearly to zero at the center..000-psi shear stress.."Il!':.-E ----B c lOS 10 6 " 10 4 NO. distortion. H.. .:::-::. M.. . . " . Source: D..:::-:... .... proving-ground." - ·'-'-'-B 400 600 DISTA.. American Society for Metals..9....001") 800 IF.. 140 u: 100 u: l- A.. the high strength of the center ofthe shaft is essentially wasted... Induction hardened 0." in Fatigue and Microstructure. and in-service testing. :~ -... -- . ~ . since its surface is the lowest hardness.. 1046. gradient strength. which examined variables such as surface hardness.. Breen and E.. the stresses drop more rapidly at the onset due to the stress concentration caused by the spline... 1541 and 50854 Steels: S-N Data for Induction Hardening vs Through Hardening :\. One would expect the through hardened shaft to have a surface-origin failure and to be lower in strength. ::::. . Axle Shafts.... OF CYCLES TO FAILURE Top: Axle-shaft strength gradients in terms of shear yield strength. B & C . " ~ . Also plotted on this graph are the shear yield-strength gradients (converted from hardness) of the production alloy shaft.. core hardness.THROUQI HARDENED 50B54 D- ?i tr. Some interesting reflections can result from examining some of the fatigue data that were generated.. P 88 . "Fatigue in Machines and Structures-Ground Vehicles. since it is lowly stressed...... .. \ ~ " >c::: < 0. . Bottom: Fatigue performances of axle shafts as a function of strength gradient... The more promising approach was then subjected to chassis. Also. Gradient strengthening by induction hardening provides a means of providing a better strength match for the stress gradient.... u: 20 10 2 ]0 3 -. " :"::....' ':-.. .rxructrox HARDE\'ED 1541 D 1046 E . .. and surface-condition effects. I1'I1lUCf IO~ HARDENED 1541 1046 E . This was accomplished after a rather comprehensive bench-test program.~ . 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 lower graph gives the fatigue curves established for shafts having the strength gradients shown in the upper graph.. v: 1 SO .24-30...'JCE FRO~l SURFACES (....

. This technique improved fatigue characteristics compared with direct-quenched rollers. Ed" American Society for Metals. a: :> . Metals Park OH.Reheat Carbo of Quench Sliding 95'0 . Steel Rollers. 1980. 463 2.961 ------ ·vacuum 60·0 50·0 40'0 30'0 Data corrected to Sc___ ____ = 400 ksi A= 0. Maurice A..89 . 8 10 2 4 5 6 7. Howes.97 . 5'0 4'0 3·0 "0 Q 1'0 0 R 0'5 0'4 0'3 0·2 105 0" I 106 4 5 6 7 8 9 . Source: S." in Source Book on Gear Design. All were slow cooled and reheated for quenching. ..98 90·0 80·0 70'0 AJ·l • Group ---- 0 Grou 0 II Group Q R 1750 1800 1900' ~I 21 21 2..514 24-31. Z . 8620H Carburized: Effects of Carburizing Temperature and Quenching Practice on Surface Fatigue 99·9 99·0 SAE 8b2011 . L Rice.5 20·0 ~ :. Slope B Coef. H. Weibull probability plot: Effect of three carburizing temperatures on surface fa- tigue for carburized 8620" steel.19 S-N data. 2 7 c v c i es 3 4 5 6 7 • 9 . 10'0 I~ a: ~ U e.199 I. Technology and Performance. p 234 . . "Pilling Resistance of Some High TemperatureCarburized Cases. Cor r ..

...T ...." 45. American Society for Metals.t----------tl~-'":f_--------__I Data Corrected to 30·01--1>:c-~W9__1<!.. Source: S.105 345678" 2 C 't' c i es 3 4567..... 1980.. .---------------1 "0 t . Technology and Performance.5 20·0 ~ . "0 ..-----=:. o ...j ' .. OF .. H.97 21 1900' 10...:.... S-N data.....~J_------+f r _ ...._ _r----1 . Rice..24-32.........- r- o Group P 70..-_r'--:_---__1 6&:~(--------------_:_ +-------__. 21 . Weibull probability factor: Effect ofthe three carburizing temperatures on surface fatigue for carburized 8620H steel. Steel Rollers..u!!J"--"S'---'-"=--=----'-"../_------_I . ..:"M.. Ed.-f---i~-------__I 4o·01---------------.....:....: . .:.:... JO·Oj----------r--='---------if---------------_f I- z U 0: .+ .. Howes... 8620H Carburized: Effects of Carburizing Temperature and Quenching Practice on Surface Fatigue 99 ..:.r . Coef... P 233 .71.. _f_...• 515 1SAE 8620H ...97 21 1800 ....0 ~a"'cull . L. Maurice A.+------__::~+. "Pitting Resistance of Some High Temperature Carburized Cases..'-----------_f c A = 0....i_-----. Metals Park OH.90 1750 +f.._ j . All were direct quenched from the temperatures shown above. p_N Sliding \ Corr.f .. H....r_o_u..0 ~...f._ _ I S N P '-0 D" 0" D" 0·3 0'\ . 'O·ol---------------......Direct Quenched Carbo =.'·0 9"0 '0' 0 1-0=--Gc..'f--------r. umlll-." in Source Book on Gear Design...=_"~-----_c...

. The second set of curves shows that doubling the cyclic stress range reduces the fatigue life by about an order of magnitude. Also.:! en a:: 3 . Since the controlling parameter is ~K. 1980.516 24-33. Weiser.9 Z i= . with the starting flaw size a. Ed. Cast Low-Carbon Steel: Starting Crack Size vs Cycles to Failure s: u c . The fatigue life increases dramatically at very small a i values. the final crack size has a larger effect on the cyclic life. if the starting ~K value is high. 0. P 4-17 . Steel Founders' Society of America. is shown in the diagram above.300 7 0 III OJ E E I .000 CYCLES TO FAILURE. Source: Steel Castings Handbook. The variation of the fatigue life. Rocky River OH.63 in.:: <{ N W iii 5 U :. N f - Fatigue life of a linkage arm as a function of starting crack size.150 4 0 N W U en :.000 100.:: <{ a:: u a:: U <. low life for small crack sizes is possible at high cyclic stresses. 5th Edition. Linkage Arm.. NJ.9 Z i= a:: <{ I- en 1000 10. 116 m rn ) .100 2 . 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.050 <.200 af.250 6 (136MPa) . The above diagram shows the importance of adjusting both the cyclic stress and starting flaw size to optimize the fatigue life. Peter F.

~ _. or to failure Np Example: Component fatigue tests were performed on the notched link ofthe previous examples.. . Reemsnyder. ..MPa N. respectively...MPa S}".24 X 1.880 0.936 0. ... _ ~~~ ~/2Z7!?2Z ...62 X 6. and component tests-are compared above..628 0..Component Tests 517 ~: 0. Su. Cycles to crack initiation. . Society of Automotive Engineers.....69 X 2.. Yield strength. 28.S) at one million cycles are.3 R=-1 rr . The fatigue strength for crack initiation at one million cycles is: iJ... 417 388 366 366 402 417 366 402 388 242 236 239 239 235 242 239 235 236 0./Coml ~~notched <." in Proceedings of the SAE Fatigue Conference P-I09. & .L~------J 7 10· loll lOll Fatigue Life.2L.47 Sli 106 cycles The three models-S-N..S = 0..24-34.l ~~ Local-Strain Model 0. ~1 .... Hot Rolled Low-Carbon Steel: S-N Data for Component Test Model .. 0 r .7 0.... S-N / Model --A ~@.21 X 2.36 X 1.444 0...6 I. ~. .604 0.... 1982. 42.404 0. ---------_--=--~~~nent Tests 0..... The most straightforward life prediction model of a component is developed from fatigue tests of the component itself.. Inc . local strain.5 04 . The cyclic load or nominal net-section stress is plotted versus cycles to crack initiation N. Hot-Rolled Carbon Steel Notched Link Tensile strength.690 0..../?/!?!Z2Z I 0..... "Constant Amplitude Fatigue Life Assessment Models.. and the three predictions for fatigue strength (iJ.--------1------.418 1.99 X 8. The component is cyclically loaded in a manner that attempts to simulate service and the model is the plot of the test results. P 127 ... the results are listed in the table and shown graphically above.64 X 2. Notched Links.. Warrendale PA. : ..05 X 1. Component Test Results. .. Nj • cycles ~ COMPONENT TEST MODEL..... 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....818 0..99 X 10' 10' 10' 10' 10' 105 10" 10" 106 Source: Harold S.

P 288 . Moll. Ti-6AI-6V-2Sn: Fatigue Endurance of HIP-Consolidated Powder .. J. E ::> E 100 ~ Annealed plate (min) '" 80 -. 3: HIP run #2.518 24-35. 2: HIP run #2. The fatigue endurance limit of H lP powder developed in this program is given in the SoN diagram above.. vacuum annealed for 2 h at 1300 of. Advantages and Limitations. A fuselage brace made from HlP'd Ti-6AI-6V-2Sn powder was used to establish the flight worthiness of a HlP'd P/M airframe component. vacuum annealed for 24 hat 1300 OF. Significance of boxed numbers is as follows: 1: HIP run #1. 5: HIP run #4.. H. Dulis. . Petersen and E. Ed. The tensile and toughness properties developed compare well with the average values for forgings. Metals Park OH. Erhard Klar. vacuum annealed for 16 h at 1300 ° F. Source: J... American Society for Metals.. C. 4: HIP run #2.. 10' 10' Number of cycles m 10· 60 10' Fatigue endurance of Ti-6AI-6V-2Sn powder consolidated by HIP at 1650 of (900°C). V. 1983. "Powder Metallurgy Parts for Aerospace Applications. vacuum annealed for 24 h at 1300 OF. as machined. Here the HIP data points lie within the band for annealed forgings and plate.! tl' 120 'x ~ :. Fuselage Brace." in Powder MetallurgyApplications.:.

Sign up to vote on this title
UsefulNot useful